Understanding Trust in Organizations: A Multilevel Perspective 9781138327580, 9781138327597, 9780429449185

Citation preview

UNDERSTANDING TRUST IN ORGANIZATIONS

Understanding Trust in Organizations: A Multilevel Perspective examines trust within organizations from a multilevel perspective, bringing together internationally renowned trust scholars to advance our understanding of how trust is affected by both macro and micro forces, such as those operating at the societal, institutional, network, organizational, team, and individual levels. Understanding Trust in Organizations synthesizes and promotes new scholarly work examining the emergence and embeddedness of multilevel trust within organizations. It provides a much-needed integration and novel conceptual advances regarding the dynamic interplay between micro and macro levels that influence trust. This volume brings new insights into how trust in groups, networks, and organizations forms, and why employees can differ in their trust in leaders and teams. Providing rich and nuanced insights into how to develop, maintain, and restore trust in the workplace, Understanding Trust in Organizations is a critical resource for scholars, graduate students, and researchers of industrial and organizational psychology, as well as practitioners in fields such as human resource management and strategic management. Nicole Gillespie is the KPMG Chair in Organizational Trust and a Professor of Management at the University of Queensland, Australia, and an International Research Fellow at the Centre for Corporate Reputation, Oxford University, UK. Her research focuses on the development and repair of trust in organizations and in the context of technological disruption.

C. Ashley Fulmer is an Assistant Professor of Management at the J. Mack Robinson College of Business, Georgia State University, USA. Her research centers on trust dynamics in organizations, affect and emotions in management, and levels of analysis theory and research. Roy J. Lewicki is the Irving Abramowitz Professor of Management and Human Resources Emeritus at the Ohio State University, USA. He is a leading scholar in the fields of negotiation, conflict management, and trust.

SIOP Organizational Frontiers Series Series Editors Angelo DeNisi

Tulane University, USA

Kevin Murphy

University of Limerick, Ireland

Editorial Board Derek R. Avery

Franco Fraccaroli

Paul Sparrow

Jill Ellingson

Susan Jackson

Hannes Zacher

Wake Forest University, USA University of Kansas, USA

University of Trento, Italy Rutgers University, USA

Lancaster University, UK

Jing Zhou

Rice University, USA

Leipzig University, Germany

The Organizational Frontiers Series is sponsored by the Society for Industrial and Organizational Psychology (SIOP). Launched in 1983 to make scientific contributions accessible to the field, the series publishes books addressing emerging theoretical developments, fundamental and translational research, and theory-driven practice in the field of Industrial-Organizational Psychology and related organizational science disciplines including organizational behavior, human resource management, and labor and industrial relations. Books in this series aim to inform readers of significant advances in research; challenge the research and practice community to develop and adapt new ideas; and promote the use of scientific knowledge in the solution of public policy issues and increased organizational effectiveness. The Series originated in the hope that it would facilitate continuous learning and spur research curiosity about organizational phenomena on the part of both scientists and practitioners. The Society for Industrial and Organizational Psychology is an international professional association with an annual membership of more than 8,000 industrial-organizational (I-O) psychologists who study and apply scientific principles to the workplace. I-O psychologists serve as trusted partners to business, offering strategically focused and scientifically rigorous solutions for a number of workplace issues. SIOP’s mission is to enhance human well-being and performance in organizational and work settings by promoting the science, practice, and teaching of I-O psychology. For more information about SIOP, please visit www​.siop​.org.

The Self at Work Fundamental Theory and Research Edited by D. Lance Ferris, Russell E. Johnson, and Constantine Sedikides Workforce Readiness and the Future of Work Edited by Frederick L. Oswald,Tara S. Behrend, and Lori L. Foster Vocational Interests in the Workplace Rethinking Behavior at Work Edited by Christopher D. Nye and James Rounds Creativity and Innovation in Organizations Edited by Michael D. Mumford and E. Michelle Todd Social Networks at Work Edited by Daniel J. Brass and Stephen P. Borgatti The Psychology of Entrepreneurship New Perspectives Edited by Michael M. Gielnik, Melissa S. Cardon, and Michael Frese For more information about this series, please visit www​.r​​outle​​dge​.c​​om​/SI​​OP​-Or​​ganiz​​ation​​al​-Fr​​ontie​​rs​-Se​​r ies/​​book-​​serie​​s​/SIO​P

UNDERSTANDING TRUST IN ORGANIZATIONS A Multilevel Perspective

Edited by Nicole Gillespie, C. Ashley Fulmer, and Roy J. Lewicki

First published 2021 by Routledge 605 Third Avenue, New York, NY 10158 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 Taylor & Francis The right of Nicole Gillespie, C. Ashley Fulmer, and Roy J. Lewicki to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. With the exception of Chapter 8, no part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Chapter 8 of this book is available for free in PDF format as Open Access from the individual product page at www​.routledge​.com. It has been made available under a Creative Commons Attribution-Non Commercial-No Derivatives 4.0 license. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Gillespie, Nicole, editor. | Lewicki, Roy J., editor. | Fulmer, C. Ashley, editor. Title: Understanding trust in organizations: a multilevel perspective / edited by Nicole Gillespie, Roy J. Lewicki and C. Ashley Fulmer. Description: New York: Routledge, 2021. | Includes bibliographical references and index. Identifiers: LCCN 2020057304 (print) | LCCN 2020057305 (ebook) | ISBN 9781138327580 (hardback) | ISBN 9781138327597 (paperback) | ISBN 9780429449185 (ebook) Subjects: LCSH: Trust. | Business ethics. | Organizational behavior. | Leadership. Classification: LCC BF575.T7 U53 2021 (print) | LCC BF575.T7 (ebook) | DDC 158.2–dc23 LC record available at https://lccn​.loc​.gov​/2020057304 LC ebook record available at https://lccn​.loc​.gov​/2020057305 ISBN: 978-1-138-32758-0 (hbk) ISBN: 978-1-138-32759-7 (pbk) ISBN: 978-0-429-44918-5 (ebk) DOI: 10.4324/9780429449185 Typeset in Bembo by Deanta Global Publishing Services, Chennai, India

This volume is dedicated to Dr. Graham Dietz, a wonderful trust scholar whose research and contributions to the community continue to inform and enrich our understanding of trust.

CONTENTS

Series Foreword xii Acknowledgmentsxiv List of Contributors xv PART I

Introduction1 1 A Multilevel Perspective on Organizational Trust Nicole Gillespie, C. Ashley Fulmer, and Roy Lewicki

3

2 Trust Conceptualizations Across Levels of Analysis C. Ashley Fulmer and Cheri Ostroff

14

PART II

Multilevel Trust Processes and Dynamics 3 Divergence in Collective Trust Audrey Korsgaard and Paul Bliese 4 The Relationship Between Trust and Attributions: A Levels-of-Analysis Perspective Edward Tomlinson and Luke Langlinais

43 45

66

x Contents

5 Cascading Influences and Contextualized Effects: A Model of Multilevel Control–Trust Dynamics Chris Long 6 Trust Me or Us?: A Multilevel Model of Individual and Team Felt Trust by Supervisors Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam 7 Trust Repair: A Multilevel Framework Nicole Gillespie, Steve Lockey, Matthew Hornsey, and Tyler Okimoto

87

121 143

PART III

Embedding Trust in Organizations 8 Network Trust Bill McEvily, Akbar (Aks) Zaheer, and Giuseppe Soda 9 The Tangled Ties of Trust: A Social Network Perspective on Interpersonal Trust Stephen Jones and Priti Shah

177 179

205

10 Multilevel Theorizing of How Gender Influences Trust and Creativity233 Hye Jung Eun, Roy Chua, and Mengzi Jin 11 Multilevel Trust in Uncertain Contexts: ODID You Hear We Have a New VP? Roger C. Mayer and Michele Williams 12 Multilevel Trust and Human Resource Management Rosalind Searle and Rami Al-Sharif

256 277

13 Trust Cues in Artificial Intelligence: A Multilevel Case Study in a Service Organization Lisa van der Werff, Kirsimarja Blomqvist, and Sirpa Koskinen

307

14 Employee Trust in Organizations Across Cultures: A Multilevel Model S. Arzu Wasti and Çetin Önder

334

Contents  xi

PART IV

Conclusion and Way Forward 

359

15 Multilevel Trust: Reflections, Insights, and a Future Research Agenda Guido Möllering, Nicole Gillespie, and Roy Lewicki

361

Index373

SERIES FOREWORD

This work is all about trust. If there ever was a book responsive to the needs of my country – the United States of America – this one surely fits the bill. At the time of this writing, the manifestations of lack of trust are all around. At the national level there are millions of Americans who appear not to have much trust in our institutions of government (voting procedures or the formal electoral process) nor in the tools of civil society (the courts or police and safety services). Millions of Americans appear to no longer have trust in the media, whether in old forms (e.g., newspapers) or new ones (e.g., blogs or social media). Aligning with the focus of this SIOP Frontiers Series volume, reports of the lack of trust regarding how employees currently see the way that their organization is being managed surface regularly in the news. While this volume cannot be expected to cover in detail the effects of all ‘levels’ of analysis when it comes to the nature and potential role of trust, it does a very fine job describing their potential impact on the people in work organizations. It explores how trust, as a core aspect of the human condition, operates.That is, it informs how it might come about, might be sustained, or can be destroyed. It addresses how trust shapes relations among individuals, members of a work group or team, or how it will affect the experience of working in a business unit. It also provides insights regarding trust repair. What I find particularly compelling about this volume is the way that Gillespie, Fulmer, and Lewicki as editors have been able to line up such a fine set of authors. Collectively, they do an excellent job of capturing the nexus of forces that serve to create the ‘ecology’ in which trust exists and will have its effects. In doing so they closely examine the dynamic nature of trust, which, if poorly managed, often leads to undesirable outcomes. However, and more optimistically, throughout

Series Foreword  xiii

the volume, they point out that, if better understood and directed, trust has the potential to promote valued ends for the people involved, their organization, and by extension, civil society. Thanks to this excellent work we now have a SIOP Frontiers book available for those who have such goals in mind. Richard Klimoski

ACKNOWLEDGMENTS

We sincerely thank Resham Kay for her valuable assistance and excellent attention to detail in compiling and formatting the chapters of this volume. We also thank Christina Chronister and Danielle Dyal at Taylor & Francis for their guidance through the publication process, Rich Klimoski for his guidance and feedback on the original proposal, and Neal Ashkanasy for his encouragement to edit a book on multilevel trust. Finally, we would like to thank each of the chapter authors for their support and valuable contributions to advancing a multilevel understanding of trust: without your insights and willingness to venture into unknown territory, this volume would not have come to fruition. We also thank the contributors who participated in and provided feedback at two workshops on multilevel trust at the Academy of Management conference. Nicole Gillespie C. Ashley Fulmer Roy J. Lewicki

CONTRIBUTORS

Editors Nicole Gillespie is the KPMG Chair in Organizational Trust and a Professor

of Management at the University of Queensland, Australia, and an International Research Fellow at the Centre for Corporate Reputation, Oxford University, UK. Her research focuses on the development and repair of trust in organizations, with current work focused on understanding trust in challenging contexts such as trust in artificial intelligence, after trust failures, during organizational change and disruption, in complex stakeholder environments, and in cross-cultural relations. Her research appears in leading journals such as Academy of Management Review, Journal of Applied Psychology, Journal of Management, Organization Studies, Journal of Business Ethics, Business Ethics Quarterly, Sloan Management Review, and Human Resource Management. Nicole has written commissioned reports on building and repairing trust for the Institute of Business Ethics and designing trustworthy organizations and trust in AI for KPMG, as well as a policy note for the UK Parliament. She has conducted commissioned research and executive education on trust for a range of private and public sector organizations including the World Economic Forum. She is the Deputy Editor of the Journal of Trust Research and on the editorial board of Leadership Quarterly. C. Ashley Fulmer is an Assistant Professor of Management at the J. Mack Robinson

College of Business, Georgia State University, USA. She received her Ph.D. in Organizational Psychology from the University of Maryland, USA, and previously served on the faculty of the National University of Singapore, Singapore, and the University of Iowa, USA. Her research centers on trust dynamics in organizations, affect and emotions in management, and levels of analysis theory and research. Ashley’s work has been published in outlets such as Harvard Business Review, Academy

xvi Contributors

of Management Review, Journal of Applied Psychology, Journal of Management, Journal of Organizational Behavior, Personnel Psychology, and Psychological Science. Of particular relevance to this volume is her comprehensive review article on “Trust across multiple organizational levels” published in the Journal of Management in 2012 (over 800 citations). She is currently an Associate Editor of the Journal of Trust Research and on the editorial boards of Academy of Management Review and Personnel Psychology. Roy Lewicki is the Irving Abramowitz Professor of Business Ethics and the Professor

of Management and Human Resources Emeritus at the Max M. Fisher College of Business, Ohio State University, USA. He has a B.A. degree from Dartmouth College, USA, and a Ph.D. in Social Psychology from Teachers College, Columbia University, USA. Professor Lewicki maintains research and teaching interests in the fields of trust development, negotiation, conflict management and dispute resolution, managerial leadership, organizational justice, and ethical decision making, and has published many research articles and book chapters on these topics. He is a Fellow of the Academy of Management and the Organizational Behavior Teaching Society. He is the author/editor of 36 books, including Negotiation (Lewicki, Saunders, & Barry, 2014) and Essentials of Negotiation, (Lewicki, Barry, & Saunders, 2015) – the leading academic textbooks on negotiation – and Mastering Business Negotiations (Lewicki & Hiam, 2007), a book for managers. He has extensive management consulting and training experience worldwide.

Contributors Rami Al-Sharif is a Lecturer in Human Resource Management at Adam Smith

Business School, University of Glasgow, UK, where he also earned his Ph.D. in Management. His research interests include organizational trust, perceived fairness, identity threat, stereotype threat, young people’s work and crime, and the role of human resource management policies and practices in providing a trustworthy, fair, and safe-identity work environment for stigmatized groups to work and produce, and for their talent to flourish. Prior to becoming an academic, Dr. Al-Sharif worked for over four years in the banking sector. Paul Bliese is the Jeff B. Bates Chaired Professor in the Department of Management

at the Darla Moore School of Business, University of South Carolina, USA. He received a Ph.D. from Texas Tech University, USA, and a B.A. from Texas Lutheran University, USA. After graduate school, he worked for 22 years at the Walter Reed Army Institute of Research. Throughout his career, Dr. Bliese has led efforts to use statistical methods to answer complex organizational problems and advance theory and practice. He developed and maintains the multilevel package for R and has been influential in supporting the R community. He was an Associate Editor for the Journal of Applied Psychology from 2010 to 2017 and is currently the Editor in Chief for Organizational Research Methods.

Contributors  xvii

Kirsimarja Blomqvist is a Professor of Knowledge Management at the School

of Business and Management at LUT University, Finland. Her research focuses on trust, knowledge, innovation, and new digital forms of organizing. Currently, she is leading a national research consortium studying the future of work. She has published her research in Research Policy, Technovation, Scandinavian Journal of Management, and California Management Review. Kirsimarja serves as an Assistant Editor for the Journal of Trust Research and a board member for FINT, an international network of trust researchers. She engages actively in societal interaction and is a member of the Academy of Finland Research Council for Culture and Society. Roy Chua is an Associate Professor of Organizational Behavior and Human

Resources at the Lee Kong Chian School of Business, Singapore Management University, Singapore. He received his Ph.D. from Columbia Business School, Columbia University, USA. His current research focuses on the impact of culture and gender on creativity and innovation. Hye Jung Eun is a Ph.D. candidate in Organizational Behavior and Human

Resources at Lee Kong Chian School of Business, Singapore Management University, Singapore. Her current research interests include creativity, gender, and inequalities within organizations. Matthew Hornsey is a Professor of Management at the University of Queensland,

Australia. He has published over 150 papers, many of which examine the psychology of communication within and between groups. He is a Fellow of the Academy of the Social Sciences in Australia, and is on the editorial board of numerous journals, including Journal of Business Research, British Journal of Social Psychology, Group Processes and Intergroup Relations, Social Psychological and Personality Science, and Social Influence. Mengzi Jin is an Assistant Professor of Organization and Strategy at the Guanghua

School of Management, Peking University, China. She received her Ph.D. from Singapore Management University, Singapore. Her current research focuses on gender diversity and organizational creativity and innovation. Stephen Jones is an Assistant Professor in the School of Business at the University

of Washington Bothell, USA. He received his Ph.D. from the Carlson School of Management, University of Minnesota, USA, and his MBA from Brigham Young University, USA. His research focuses on trust, conflict, and collaboration within and between organizations and emphasizes the role of social networks in interpersonal and inter-organizational relations. His research has appeared in Administrative Science Quarterly, Journal of Applied Psychology, Strategic Management Journal, and Organization Science.

xviii Contributors

Audrey Korsgaard is a Professor in the Management Department at the Darla

Moore School of Business, University of South Carolina, USA. She earned a Ph.D. from New York University, USA, and a B.A. from Rutgers University, USA. She has studied trust building in a variety of leadership roles, including middle management, top management teams, virtual teams, and boards of venture-capital-backed firms. She has published over 70 articles, book chapters, and proceedings on these topics. Formerly the Associate Editor of the Journal of Management, she currently serves on the board of numerous management journals. Sirpa Koskinen holds an MSc (Economics and Business Administration) in

Knowledge Management and Leadership from LUT University, Finland. Sirpa’s master’s thesis explored the relationship between trust and AI. In the near future, she will apply for a doctoral program and continue research in the area of AI and trust. Long W. Lam holds a Ph.D. from the University of Oregon, USA, and is a

Professor of Management at the University of Macau, China. Dr. Lam is currently doing research on felt trust, customer mistreatment, proactive behaviors, and dirty work. His research has appeared or been accepted for publication in the Journal of Applied Psychology, Journal of Management, Human Relations, Journal of Organizational Behavior, Journal of Vocational Behavior, Journal of Occupational and Organizational Psychology, Asia Pacific Journal of Management, and Journal of Business Ethics, etc. He is currently the Senior Editor of the Asia Pacific Journal of Management, the Advisory Editor of Journal of Human Resource Management in Taiwan, and a member of the Editorial Review Board of the Journal of Trust Research. Luke Langlinais is a Ph.D. candidate in Management at the John Chambers

College of Business and Economics at West Virginia University, USA. He holds an MBA in Healthcare Administration and a Bachelor of Science in Marketing from the University of Louisiana at Lafayette, USA. His primary research interests involve organizational behavior topics related to social influence in organizations, with an emphasis on how individuals rebuild trust in workplace relationships. Dora C. Lau holds a Ph.D. from the University of British Columbia, Canada, and

is an Associate Professor at the Chinese University of Hong Kong, Hong Kong. Her research interests include demographic diversity and fault lines, relational trust, team dynamics, upper-echelon composition and organizational impact, family business challenges and management, and Chinese management. She has published extensively in top-tiered journals such as Academy of Management Review, Academy of Management Journal, and Journal of Applied Psychology. She has served as the Associate Editor of the Journal of Trust Research and also guest-edited a special issue for the Asia Pacific Journal of Management, namely “Leadership in Asia.”

Contributors  xix

Steve Lockey is a Postdoctoral Research Fellow in Organizational Trust at the

University of Queensland, Australia. He received his Ph.D. in Management at Durham University, UK, and received the Business School’s Outstanding Thesis award. Steve’s primary research areas relate to trust development, repair, and measurement. The influence of emotions on trust processes is an area of particular interest. He is currently examining trust in artificial intelligence (AI), with a focus on understanding public attitudes and understanding of AI systems, antecedents of trust in AI, and the vulnerabilities stakeholders face when interacting with or being impacted by AI. Other work includes the examination of how leadership impacts employee well-being, behavior, and performance in the context of policing. Steve has worked closely with several police forces in the United Kingdom, and his research has informed policy in this context. His work is published in Business Ethics Quarterly and the International Journal of Police Science and Management. Chris Long is the Paul Naughton Associate Professor of Management at the

Tobin College of Business, St. John’s University, USA. In his award-winning research, he examines how leaders foster the achievement of a variety of performance objectives in complex and dynamic business environments. His current work focuses both on how leaders balance their efforts to apply controls, demonstrate trustworthiness, and promote fairness and how the combined actions that leaders take influence the cognitions, emotions, and behaviors of other stakeholders (e.g., subordinates, authorities). His research has appeared in leading journals including Academy of Management Annals, Academy of Management Journal, Accounting, Organizations, and Society, Annual Review of Organizational Psychology and Organizational Behavior, and Organization Science. He holds a Ph.D. in Management from Duke University, USA, and a Master in Public Policy from the John F. Kennedy School of Government at Harvard University, USA. Chris also currently serves as an officer in the United States Army Reserve where he works on issues related to psychological health (e.g., PTSD), human performance, and human–machine teaming. Roger C. Mayer is a Professor of Management, Innovation, and Entrepreneurship

at the Poole College of Management, North Carolina State University, USA. He received a Ph.D. in Organizational Behavior and Human Resource Management from the Krannert School of Management at Purdue University, USA. Mayer’s interdisciplinary research is focused on trust, employee decision making, attitudes, effectiveness, police–public trust, and human–technology interactions. A leading scholar on trust in organizations, his research has been published in many premier scholarly journals. He has been a PI or a contributing researcher on numerous grants. Mayer has worked in a wide variety of organizations and industries, including firms in finance, research, construction, steel, and offshore oil drilling. He serves on the editorial boards of Journal of Management, Journal of Managerial

xx Contributors

Psychology, and Journal of Trust Research. He speaks frequently to business, government, legal, medical, and civic groups on such topics as trust, leadership, negotiation, diversity, and influencing others. Bill McEvily is the Jim Fisher Professor of Leadership Development and a

Professor of Strategic Management at the Rotman School of Management, University of Toronto, Canada. Professor McEvily teaches courses on social networks and strategic change and implementation in MBA and Executive programs, and courses on organizational theory in the Ph.D. program. His research explores social networks as an organizational and strategic resource. Professor McEvily has published research articles in leading academic journals in the fields of management, psychology, sociology, and economics. Thomson Reuters named Professor McEvily to its list of “Highly Cited Researchers” in 2014 and again in 2015. He is currently an Associate Editor at Annals of Academy of Management and has previously served as a Senior Editor at Organization Science and as a guest editor for special issues of Management Science and Organization Science. Prior to joining Rotman, Professor McEvily was on the faculty at Carnegie Mellon University, USA, and he earned his Ph.D. in Strategic Management and Organization from the University of Minnesota, USA. Guido Möllering is the Director of the Reinhard Mohn Institute of Management

at Witten/Herdecke University, Germany, where he also holds the Reinhard Mohn Endowed Chair of Management. He earned his Ph.D. in Management Studies at the University of Cambridge, UK, and his habilitation (postdoctoral degree, venia legendi) in Business Administration at Freie Universität Berlin, Germany. His main areas of research are inter-organizational relationships, organizational fields, and trust. Professor Möllering has published several books, notably Trust: Reason, Routine, Reflexivity (2006), and articles in leading journals such as Organization Science and Journal of International Business Studies. He is a Senior Editor of Organization Studies and the Editor-in-Chief of the Journal of Trust Research. Tyler Okimoto is a Professor in Management and the Deputy Head of the

Business School at the University of Queensland, Australia. He received his Ph.D. in Organizational Psychology from New York University, USA. Tyler’s research aims to better facilitate collaboration and consensus between diverse points of view and to understand the role of leadership in overcoming those challenges. He often examines consensus/collaboration as a conduit for social justice in organizations and society, both how a lack of consensus contributes to injustice and inequality, and how people can effectively collaborate to move past conflict and repair harmonious relationships. His work has been published in leading journals such as Journal of Personality and Social Psychology, Journal of Applied Psychology, Journal of International Business Studies, Leadership Quarterly, Business Ethics Quarterly, and Science Advances.

Contributors  xxi

Çetin Önder is a Professor of Management and Organizations at the Business

Administration Department, Social Sciences University of Ankara, Turkey. He received his Ph.D. in Management and Organizations from Sabancı University, Istanbul, Turkey. He has been lately interested in changes in the higher education sector, evolution of labor union populations, diffusion of management practices across countries, and political ties of organizations. His research has been published in journals like Leadership Quarterly, Industry and Innovation, Journal of Management and Organization, Higher Education, Scientometrics, and Research Evaluation. Cheri Ostroff is a Research Professor in the Centre for Workplace Excellence at the

University of South Australia, Australia. Her research focuses on how the practices and features of an organization influence the behavior and attitudes of individuals, and, conversely, how individuals’ attributes collectively influence team and organizational functioning and effectiveness. Professor Ostroff received the Distinguished Scientific Award for Early Career Contributions in Applied Research/Statistics from the American Psychological Association (APA), the Distinguished Career Contributions Award from the Society for Industrial-Organizational Psychology (SIOP), the Scholarly Achievement Award from the Human Resources Division of the Academy of Management (AoM), an Outstanding Teacher Award from Teachers College Columbia University, and the Leadership in Teaching and Learning Award from UniSA. She received the Decade Award from the Academy of Management Review. She has been listed as a most-published author in the organizational psychology field and is ranked in the top 250 most influential management scholars worldwide (top 1% of 25,000) and in the world top 2% in 2019. She has served on multiple journal editorial boards in both psychology and management. She also served in the five-year (elected) leadership track for the Organizational Behavior Division of the Academy of Management. Rosalind Searle holds the chair in HRM and Organizational Psychology at the

Adam Smith Business School at the University of Glasgow, UK, and is Director for the European Association of Work and Organizational Psychology (EAWOP) Impact Incubator. She is a Chartered Occupational Psychologist and a Fellow of the British Psychological Society (BPS) and of the ‘Chartered Institute for Personnel and Development (CIPD). She has a Ph.D. from Aston University, UK. Her research focuses on organizational trust and HRM, trust and controls, change, and counterproductive work behaviors. She is co-editor for the Routledge Companion to Trust (2018) and the Edward Elgar Frontiers in Trust Research book series. She is an Associate Editor for Group and Organization Management and the Journal of Trust Research. She serves on editorial boards of Human Relations, Journal of Management, and International Perspectives in Psychology: Research, Practice, Consultation (IPP). Her research appears in leading international journals (e.g., Human Resource Management, Journal of Organizational Behavior, International Journal of HRM, Organization Studies, and Long Range Planning) and in commissioned

xxii Contributors

research for regulators (e.g., Professional Standards Authority), government (e.g., Welsh Audit, Scottish and English Governments), and private organizations (e.g., energy sector). Priti Shah is an Associate Professor in the Department of Work and Organization at

the Carlson School of Management, University of Minnesota, USA. She received her Ph.D. in Organizational Behavior from the Kellogg School of Management at Northwestern University, USA. Her research focuses on interpersonal relationships at work, teams, and decision making. In particular, she uses a social network approach to investigate dynamics within teams, focusing on trust and conflict. Her work appears in the Academy of Management Journal, Administrative Science Quarterly, Journal of Applied Psychology, Journal of Organizational Behavior, Organization Behavior and Human Decision Processes and Organization Science, and she is currently on the editorial boards of the Journal of Applied Psychology, Journal of Organizational Behavior and Small Group Research. Giuseppe Soda is a Professor of Organization Theory and Social Networks at

the Department of Management and Technology and the SDA Bocconi School of Management, Bocconi University, Italy. He earned his Ph.D. in Management and Organization from Bocconi University, Italy. His research has looked at both when and how organizational actors should leverage greater connectivity within and across their boundaries to enhance performances. More precisely, he investigates how organizational architectures, organizational networks, inter-firm collaborations, and their interplay influence organizational-level outcomes. His work has been published in the leading management and organization theory journals, receiving the Bocconi University’s Research Excellence Award for several years. Edward Tomlinson is a Professor of Management in the John Chambers College

of Business and Economics at West Virginia University, USA. His primary research interest is interpersonal trust, and his work in this area has appeared in several high-impact management journals such as Academy of Management Review, Journal of Management, Journal of Applied Psychology, Journal of Organizational Behavior, and Journal of Trust Research. Lisa van der Werff is an Associate Professor of Organizational Psychology at

Dublin City University Business School, Ireland, a Research Director of the Irish Institute on Digital Business, and the incoming President of FINT, an international network of trust researchers. Lisa’s research focuses on trust development across a range of organizational contexts and includes interdisciplinary research on the antecedents of trust in disruptive technologies. Lisa serves on the editorial boards of Human Relations and the Journal of Trust Research.

Contributors  xxiii

S. Arzu Wasti is a Professor of Management and Organization Studies at Sabancı

Business School, Sabancı University, Turkey. She received her Ph.D. in Industrial Relations and Human Resource Management from the University of Illinois at Urbana-Champaign, USA. Her cross-cultural research on organizational commitment, sexual harassment, workplace incivility, and organizational trust has appeared in such journals as Journal of Applied Psychology, Journal of Cross-Cultural Psychology, Journal of International Business Studies, and Leadership Quarterly. She is a recipient of several research awards such as the Science Award in the social sciences by the Scientific and Technological Research Council of Turkey and the Turkish Academy of Sciences Encouragement Award. Michele Williams holds a Ph.D. from the University of Michigan, USA, is an

Assistant Professor of Management and Entrepreneurship, the John L. Miclot Fellow in Entrepreneurship, and the Diversity, Equity, and Inclusion Faculty Fellow at the University of Iowa, USA. She conducts interdisciplinary research on the micro-foundations of collaboration, innovation, and equity among team members and boundary spanners from diverse groups and from different organizations. At the interpersonal level, she examines the influences of psychological processes, such as perspective taking, interpersonal sensitivity, and emotion regulation, on how interpersonal trust and cooperation evolve in peer, leader–member, transgressor–victim, and strategic relationships. At the group and organization levels, she examines how social categorization processes and the social construction of gender influence women as team members, leaders, boundary spanners, consultants, and entrepreneurs. She has authored papers and case studies, published in notable academic journals, and speaks at academic and professional conferences on her topics of research. Michele is on the editorial boards of Organization Science and the Journal of Business Venturing and an Associate Editor at the Journal of Trust Research. Akbar (Aks) Zaheer is a Professor and the Curtis L. Carlson Chair in Strategic

Management at the Carlson School of Management, University of Minnesota, USA. He received his Ph.D. in strategic management from the Massachusetts Institute of Technology, USA, and his Master’s in Business from the Indian Institute of Management in Ahmedabad, India. His current research examines the antecedents and consequences of interfirm and organizational networks, and the antecedents and consequences of trust in organizations, interfirm exchange, and the context of phenomena such as innovation and strategic alliances. He has published extensively in many journals, receiving the School’s Outstanding Research Award in 2014. He serves as Dean of the Fellows of the Strategic Management Society and is an elected Fellow since 2014. He was a guest editor of a special issue of Organization Science on “Trust in an Organization Context” and of Academy of Management Review on “Repairing Relationships,” and currently serves as a guest editor of Academy of Management Review on contemporary perspectives on trust.

xxiv Contributors

Julie N.Y. Zhu holds a Ph.D. from the University of Macau, China, and is cur-

rently an Assistant Professor in the School of Economics and Management at Fuzhou University, China. Her research interests include felt trust, workplace incivility, creativity, and workplace status. Her research has appeared or been accepted for publication in the Asia Pacific Journal of Management, International Journal of Hospitality Management, Journal of Managerial Psychology, Journal of Business Psychology, and Personnel Review.

PART I

Introduction

1 A MULTILEVEL PERSPECTIVE ON ORGANIZATIONAL TRUST Nicole Gillespie, C. Ashley Fulmer, and Roy Lewicki

Welcome Welcome to Understanding Trust in Organizations: A Multilevel Perspective. Approximately three years ago, we examined the thriving research literature on organizational trust and trust repair, viewed through multiple lenses: different conceptualizations of the core constructs, different disciplinary perspectives, different contexts in which trust was being observed, and the role of trust as an antecedent, mediating, moderating, or outcome variable. An area we noted that was ripe for concerted effort and systematic inquiry was how trust was being addressed at various organizational levels. While much of the early trust literature was developed around interpersonal constructs, a parallel set of literatures has developed around team and group trust and organizational trust.We believed that the trust literature had reached enough of a level of maturity that these multilevel perspectives – at the individual, team, and organizational levels and beyond – needed to be brought together and conceptualized; hence the purpose of this volume. We invited leading scholars from around the world whose expertise and previous work could offer insight into a multilevel perspective on organizational trust. The challenge we gave to our contributors was to break new conceptual ground to advance our understanding of how trust functions and operates across multiple levels of analysis. We committed to providing our authors with the space and freedom to take risks in their theorizing and exploration of multilevel trust. Reminiscent of the process used to develop other influential volumes on trust (e.g., Searle, Nienaber, & Sitkin, 2018), we invited our contributors to exchange ideas and give feedback to each other on their developing papers at workshops at the 2018 and 2019 meetings of the Academy of Management, as well as to present their papers at the 2020 Academy of Management. DOI: 10.4324/9780429449185-1

4  N. Gillespie, C. A. Fulmer, and R. Lewicki

We the editors have enjoyed the opportunity to assemble this collection of invited chapters and work with the contributors as they developed and strengthened their chapters to form the excellent contributions within this volume. We are also grateful that the Society for Industrial and Organizational Psychology (SIOP) recognized the value of this collection of work and was willing to publish it as a volume in their Organizational Frontiers Series.

Why This Book? Trust is an issue of great organizational, national, and societal importance. Trust underpins and supports effective leadership, teamwork, employee and stakeholder relationships, knowledge sharing, and innovation, as well as broader organizational effectiveness and market participation (Fulmer & Gelfand, 2012; Searle et al., 2018). Yet recent public surveys, such as those by the Edelman Trust Barometer, the Gallup organization, and the Pew Research Centre, reveal a disconcerting trend of declining trust in business, government, and some societal institutions and organizations. This has resulted in unprecedented interest in and focus on trust in organizations, including business, government, and nongovernmental organizations, as well as CEOs, line managers, organizational specialists, and academics. Interest in trust research has increased accordingly over the last quarter-century. The field has now reached a stage where it is time for it to focus attention away from the fundamental nature of trust, to explore its nuances and complexities, including its dyadic nature, temporal fluctuations, multiparty dynamics, contextual influences, and trust as context. Importantly, it is becoming more apparent that a full understanding of trust in the workplace requires a multilevel assessment of its processes and dynamics, embedded in multiple networks and systems and enacted by multiple organizational actors and entities (Fulmer & Gelfand, 2012; Gillespie & Dietz, 2009). The seminal definition of trust in the organizational literature – the willingness of a party to be vulnerable to the actions of another party based upon positive expectations of the intentions or behavior of another (Mayer, Davis, & Schoorman, 1995; Rousseau, Sitkin, Burt, & Camerer, 1998) – is multilevel in nature: both the trusting party (trustor) and the referent of trust (trustee) can be individuals, groups, organizations, and institutions, and trust can take place and be examined at these different levels of analysis. As Schoorman, Mayer, and Davis (2007) note, their seminal model of organizational trust was developed to support the investigation of trust across multiple levels. Research has since studied trust across different levels of analysis in a great variety of organizational contexts (Fulmer & Gelfand, 2012). Despite the inherently multilevel nature of the trust concept, for many years, measurement tools and research evidence on interpersonal trust at the individual level have been used to understand trust at other higher levels of analysis: within and between groups and organizations. This approach assumes that the nature of

A Multilevel Trust Perspective  5

trust and its nomological network do not differ across levels of analysis, but are rather isomorphic. However, this approach is questionable. Trust in the workplace – whether between individuals or within and between teams, networks, and organizations – is embedded in a multilevel system where top-down influences from the system and bottom-up effects of member interactions jointly affect trust. Yet research has largely focused on a single level of analysis in studying trust and its antecedents and consequences. Furthermore, research at different levels of analysis, such as interpersonal trust, team trust, and trust within organizations, has developed independently with little cross-fertilization, creating silos in our knowledge of trust. Recent work has identified both similarities and differences in trust across levels (Fulmer & Gelfand, 2012) and highlighted the opportunities afforded by considering trust at different levels (e.g., Currall & Inkpen, 2002; De Jong & Dirks, 2012; Gillespie & Dietz, 2009; Schilke & Cook, 2013;Vanneste, 2016). The increased interest and importance in understanding trust from a multilevel perspective is highlighted by a recent special issue of the Journal of Trust Research (Fulmer & Dirks, 2018), and calls in prominent reviews for future trust research to adopt a multilevel perspective (e.g., Bachmann, Gillespie, & Priem, 2015; Costa et al., 2018; Fulmer & Gelfand, 2012; Kramer & Lewicki, 2010). As these reviews point out, by taking level effects into account, researchers can gain a more nuanced and realistic understanding of organizational trust and provide more precise and relevant insights for practitioners.

Aims and Scope of the Volume This book aims to both synthesize and promote new scholarly work examining the emergence and embeddedness of multilevel trust within organizations. It advances understanding of how trust within organizations is affected by both macro and micro forces, such as those operating at the societal, institutional, community, network, organizational, group/team, and individual (e.g., leaders, managers, and employees) levels. To our knowledge, it is the first volume that takes a focused multilevel perspective to understand organizational trust. This approach differs radically from prior scholarship that separately considers trust at different levels of analysis and hence is less equipped to uncover the embedded nature of trust and the dynamic interplay between micro and macro levels that influence trust in the workplace. The book provides much-needed integration as well as novel conceptual and empirical insights on the various multilevel trust dynamics and processes that play out in organizations. We see this as a timely and necessary springboard for further development of work on multilevel trust. We have brought together this volume with multiple audiences in mind, including academics working in the fields of industrial and organizational (I/O) psychology, management, organization studies, sociology, political science, and

6  N. Gillespie, C. A. Fulmer, and R. Lewicki

education, as well as practitioners in the fields of I/O psychology, human resources, organizational development, change management, and management in general. Together, these leading international scholars in trust research summarize what we know and what we need to know about multilevel trust. It provides readers – both scholars and practitioners – with a more comprehensive and nuanced view of the nature and dynamics of trust, and rich insights on how to develop, maintain, and restore trust.

What Is a Multilevel Analysis and Why Is It Critical to the Field at This Time? A multilevel understanding of trust in organizations is required because trust in organizational settings can occur between a wide range of parties at different levels of analysis, including but not limited to between co-workers and between a leader and a follower at the individual level, within a team and between departments at the team or unit level, and among a firm’s employees and across different organizations at the organizational level (Fulmer & Gelfand, 2012; Mayer & Gavin, 2005; Rousseau et al., 1998). The levels-of-analysis approach enables the analysis of a phenomenon or relationship of interest at the level appropriate to a research question where the phenomenon or relationship exists. For example, a research question that focuses on trust between different teams should be examined at the team level (rather than at the individual level). However, specifying the level of analysis concerns more than just identifying the level where the trust construct of interest resides; it also serves to ensure and strengthen the alignment of the level of the theory, measurement, analysis, interpretation, and implications to ensure it is appropriate to the level of the research question. The field of trust has made significant progress in important directions across levels of analysis, with several recent studies explicitly considering trust dynamics in organizations across different levels of analysis. For instance, at the organizational level, research now recognizes the impact of trust as a context that shapes employee perceptions, attitudes, and interactions (Lumineau & Schilke, 2018), and that reciprocity between boundary spanners influences inter-organizational trust (Vanneste, 2016). At the team level, research now shows dynamics such as changes and spirals in trust over time (Drescher, Korsgaard, Welpe, Picot, & Wigand, 2014; Ferguson & Peterson, 2015). At the dyadic level, the concurrent and reciprocal perspective where trust from both parties is considered has gained increased attention (Fulmer & Gelfand, 2012; Jones & Shah, 2016; Korsgaard, Bliese, Kautz, Samson, & Kostyszyn, 2018). At the individual level, the relationship between one’s trust in different targets and between trust desired and received have been considered (Baer, Frank, Matta, Luciano, & Wellman, 2020; Fulmer & Ostroff, 2017). Importantly, the levels-of-analysis approach has widened our conceptualizations of trust in organizations, resulting in exciting new areas for improving our

A Multilevel Trust Perspective  7

understanding of trust in organizations. Some examples include trust meta-accuracy, which refers to the extent to which one’s felt trust is accurate (Campagna, Dirks, Knight, Crossley, & Robinson, 2020), and trust asymmetry and consensus, which may be examined at the dyadic, team, and unit levels (De Jong & Dirks, 2012; De Jong, Gillespie, Williamson, & Gill, 2020; Korsgaard, Brower, & Lester, 2015). Many of these themes require a more systematic multilevel treatment, and the chapters in this volume seek to address this. In contrast to this existing research which has developed independently, our volume represents a rare and concerted effort to bring together prominent trust scholars to theorize trust concepts and relationships explicitly through a multilevel lens. Despite the strong growth of trust research, theoretical work has been notably less prevalent in recent years. For example, the Academy of Management Review has published no papers with ‘trust’ in the title in the decade from 2010 to 2019. The trust field is rich with earlier theoretical contributions that advance understanding of trust at single or multiple levels (e.g., Gillespie & Dietz, 2009; Kim, Dirks, & Cooper, 2009; Lewicki, McAllister, & Bies, 1998; Mayer et al., 1995; Wicks, Berman, & Jones, 1999), which has supported a plethora of empirical work and firmly established the role of trust in organizational sciences. We hope that the multilevel trust conceptual work put forth in the chapters of this volume can serve as foundations supporting and advancing research in future decades on trust in organizations.

An Overview of the Volume and Contributions We organize the book into four parts. This introductory part first gives an overview of the book and the contributions of each chapter. Then in Chapter 2, entitled Trust conceptualizations across levels of analysis, C. Ashley Fulmer and Cheri Ostroff set out to clarify the conceptualization, operationalization, and theorization of trust across levels of analysis. They differentiate levels of trustors (e.g., ‘I’ versus ‘we’) and trustees (e.g., another individual or unit as a whole), and consider composition (shared trust across individuals or entities), dispersion (extent of difference in trust across individuals or entities), and compilation (pattern or configuration of trust across individuals or entities) models. In doing so, they highlight how the conceptualization and meaning of trust changes across levels based on the trustor focus, trustee target, and higher-level trust form.They demonstrate that the levels-of-analysis approach is not merely a methodological tool but a theoretical means that can enrich and deepen how we view trust in organizational settings. In Part II, the focus is on examining critical trust processes and dynamics from a multilevel perspective. In Chapter 3, entitled Divergence in collective trust, Audrey Korsgaard and Paul Bliese focus on collective trust by members of a unit. Specifically, they consider mutual trust within dyads and trust within a group and an organization. Following recommendations from the levels-of-analysis literature, research on collective trust typically demonstrates a level of agreement in

8  N. Gillespie, C. A. Fulmer, and R. Lewicki

trust among members. Nevertheless, considerable within-unit variability often remains. Korsgaard and Bliese propose that both trust emergence and trust divergence are predictable and consequential states for groups and organizations. They present a theoretical framework that examines trust divergence as a result of a developmental phase or destabilizing events, and as a chronic state.They also offer analytic approaches and best practices for examining the impact of trust convergence and divergence over time. Chapter 4 by Edward Tomlinson and Luke Langlinais is titled The relationship between trust and attributions: A levels-of-analysis perspective. This chapter adopts a levels-of-analysis perspective to consider the relationship between trust and the attribution dynamics that affect trust judgments. The authors begin by reviewing and synthesizing the extant literature on trust and attributions, and then specifically examining multi- and cross-level research on attributions. The authors conducted an integrative analysis to provide insights and observations on how to incorporate this emerging levels-of-analysis approach to better understand the attribution–trust relationship. Special consideration is given to how social influence processes may affect an individual’s attributions. Implications and recommendations are provided for ways to incorporate different levels of analysis in researching trust–attribution dynamics. Chapter 5, Cascading influences and contextualized effects: A model of multilevel control–trust dynamics by Chris Long offers a multilevel framework of control–trust dynamics that illustrates how the decisions and actions of senior leaders, mid-level managers, and front-line employees interconnect in organizational environments. After providing a map of the conceptual landscape encompassing multilevel control–trust dynamics, the chapter highlights key factors that influence actors’ trust-building and control activities at various organizational levels. It explains how, at each level (senior leaders, managers, employees), actors’ control- and trustbuilding actions, as well as their attitudes and exhibited levels of performance, are influenced by the decisions and actions of other actors above and below them in the organizational hierarchy. Using this conceptual map to outline key relationships, the chapter then proposes an agenda for future multilevel research, providing scholars tools to develop more complete and accurate theoretical models of control–trust dynamics in organizations. In Chapter 6, Trust me or us? A multilevel model of individual and team felt trust by supervisors, Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam focus on the concept of employee felt trust (i.e., feelings and perceptions of being trusted) by the supervisor. Adopting a multilevel approach, they present a theoretical model where they identify supervisory trusting behaviors that engender felt trust at the individual and team levels and the impact of felt trust by supervisor at these two levels on individual and team performance. Furthermore, they discuss the top-down process of team prototypicality and the bottom-down process of team identification that connect between individual and team felt trust by supervisor.The framework thus offers a nomological network of felt trust by supervisor between individuals and within teams.

A Multilevel Trust Perspective  9

In the final chapter in this section, Chapter 7, Trust repair: A multilevel framework, Nicole Gillespie, Steve Lockey, Matthew Hornsey, and Tyler Okimoto integrate insights from the management and social psychological literatures on trust repair to develop an integrated multilevel framework of trust repair. They offer two propositions. First, trust repair is not isomorphic across levels but rather can take on different forms and dynamics. Second, the complexity and challenge of trust repair increases as one moves from the individual (single transgressor and trustor) to the collective level (group transgressor and trustor). Their framework identifies four factors that together explain why trust repair is different and more complex at the collective level: (1) increased third-party involvement, (2) (inter) group dynamics and processes, (3) a wider repertoire of strategies available for trust repair, and (4) the indirect and direct experience of the violation by trustors. They conclude with a research agenda for advancing a multilevel understanding of trust repair. Part III features seven chapters examining how trust is embedded within and influenced by organizational systems, structures, processes, and practices. In Chapter 8 on Network trust by Bill McEvily, Akbar (Aks) Zaheer, and Giuseppe Soda, the authors move beyond the extensive focus on relational trust in the organizational literature by broadening the conceptualization to include its inherent generalizability across a network. Specifically, they introduce the concept of network trust. Central to this conceptualization is the idea that apart from forming trust as a result of direct interaction, trust also flows through the indirect connections linking individuals to one another, and emerges from the inherent design features of a network itself. They further conceptualize network trust as comprising two forms: secondhand trust and prototrust. Secondhand trust refers to the partial spillover of relational trust to socially proximate, indirectly connected actors. Prototrust refers to the latent potential for confident positive expectations to emerge between two actors who are neither directly nor indirectly connected. Drawing on network theory, they articulate the logics (in terms of mechanisms, indicators, and contingencies) by which secondhand trust and prototrust operate. They conclude with a call to treat network trust as a novel form and with an agenda for considering the unique understandings that network trust permits. In Chapter 9, Stephen Jones and Priti Shah continue the focus on social networks in their chapter on The tangled ties of trust: A social network perspective on interpersonal trust. Drawing on the fact that trust is embedded in a complex web of relationships and influenced by the social context at multiple levels, they argue that a social network approach helps understand the determinants and outcomes of interpersonal trust. They explain how social networks can be used to model individual, team, division, and organizational factors that influence trust, and to understand how trust influences organizational relationships. The chapter provides a high-level overview of the relational and structural components of social networks, reviews the existing literature on how social networks influence

10  N. Gillespie, C. A. Fulmer, and R. Lewicki

interpersonal trust, and discusses how trust may influence the formation of social ties. They conclude with new avenues for future trust research using social networks. In Chapter 10, Hye Jung Eun, Roy Chua, and Mengzi Jin examine Multilevel theorizing of how gender influences trust and creativity. Integrating the literatures on trust, gender, and creativity, this chapter takes a multilevel approach to analyze how gender influences trust at the individual, dyadic, social network, and group levels – which in turn leads to different creativity-related outcomes for men and women, such as gender gaps in creativity and innovation achievement. These authors draw on prior work indicating women are primarily perceived as warm and men are largely seen as competent, together with gender differences in social network properties and the effects of gender composition in groups, to analyze gender differences in trust. They propose a set of propositions to help explain the effect of gender on affect-based trust and cognition-based trust, as well as potential barriers and opportunities for women in creativity-related processes. In Chapter 11, Multilevel trust in uncertain contexts: ODID you hear we have a new VP?, Roger C. Mayer and Michele Williams introduce the construct of Organization Dissociative Identity Disorder (ODID) – the extent to which an organization member perceives that the organization, groups, or individuals representing the organization hold independent or conflicting logics, identities, values, and goals that carry implications for employees’ perceptions of their leaders and their trust. They outline a model where contextual factors such as external crises, internal crises, and changes in strategic direction can lead to ODID. As a result, a supervisor’s ability to fulfill promises is restricted and employees have lower perceptions of the supervisor’s trustworthiness. The chapter concludes with directions for future trust research that incorporates the employee perceptions of ODID at multiple levels of analysis. Chapter 12 by Rosalind Searle and Rami Al-Sharif focuses on Multilevel trust and human resource management. They review how HRM policies and practices influence trust in either the employing organization, management, or supervisors and co-workers, as well as the mediating role that trust plays between HRM policies and delivery of organizational outcomes. They then focus on the question of how HRM processes and practices influence trust pathways to produce different outcomes. They conclude with a future research agenda focused on the need to study distrust, and the importance of understanding how HRM practices and trust processes are experienced by minority groups and women, who may receive less favorable employment outcomes. In Chapter 13, entitled Trust cues in artificial intelligence: A multilevel case study in a service organization, Lisa van der Werff, Kirsimarja Blomqvist, and Sirpa Koskinen examine how various stakeholders draw cues that inform their decisions to trust in the complex and multi-level system of artificial intelligence (AI) development and adoption. This chapter first provides a review of the literature on micro and macro trust cues in the context of AI. The chapter then draws on a case study of

A Multilevel Trust Perspective  11

an AI pilot program in the service industry to examine the within-level and crosslevel antecedents of trust in AI for stakeholders in technology management, developer, and user roles. The authors highlight the need for further research on the role of emotion and heuristic cues for trust, particularly in terms of their potential for cross-level influence, as well as the trustworthiness of complementary suppliers and institutions, such as regulation, legal frameworks, and culture. In Chapter 14, Employee trust in organizations across cultures: A multilevel model, S. Arzu Wasti and Çetin Önder present a multilevel model on the effects of societal context on organizations and individuals in influencing employee trust in the organization. They differentiate between modern versus neotraditional institutional systems at the societal level and their different impacts on organizational situationscapes and individual members within the organization. Organizations thus differ in their cultures, strategies, structures, and practices. Both societal and organizational variables in turn shape employees’ norms and schemas that shape how they evaluate their particular trust-related organizational experiences. Through the model, the chapter provides a more contextualized view of the development of employee trust in organizations that is subject to influences from multiple levels of analysis. In the final part, we join forces with Guido Möllering, Chief Editor of the Journal of Trust Research, to reflect on key insights about multilevel trust from across the volume and set out a future research direction. We reflect on the multiple ways that trust is examined across levels within the volume, including in relation to top-down and bottom-up processes, vertical and horizontal analyses, stable and dynamic relationships, and convergent and divergent trust. We conclude that trust should not be assumed to be isomorphic across levels and that the multilevel approach enables a novel and nuanced understanding of trust. We discuss ongoing gaps in our knowledge and identify four promising future research directions for multilevel trust: taking a social network approach, understanding contextual influences, focusing on emerging societal and organizational trust challenges, and taking a processual approach. In conclusion, we wish to thank all of our authors for their excellent contributions. We hope that the innovative theory and applications provided by these essays further stimulate research on multilevel trust dynamics within and beyond organizations in the future.

References Bachmann, R., Gillespie, N., & Priem, R. (2015). Repairing trust in organizations and institutions: Towards a conceptual framework. Organization Studies, 36(9), 1123–1142. https://doi​.org​/10​.1177​/0170840615599334 Baer, D. M., Frank, M. E. L., Matta, D. F. K., Luciano, D. M. M., & Wellman, D. N. (2020). Under trusted, over trusted, or just right? The fairness of (in)congruence between trust wanted and trust received. Academy of Management Journal. https://doi​.org​/10​.5465​/ amj​.2018​.0334

12  N. Gillespie, C. A. Fulmer, and R. Lewicki

Campagna, R. L., Dirks, K. T., Knight, A. P., Crossley, C., & Robinson, S. L. (2020). On the relation between felt trust and actual trust: Examining pathways to and implications of leader trust meta-accuracy. Journal ofApplied Psychology,105(9),994–1012.https://doi​.org​/ 10​.1037​/apl0000474 Costa, A. C., Fulmer, A., & Anderson, N. R. (2018). Trust in work teams: An integrative review, multilevel model, and future directions. Journal of Organizational Behavior, 39(2), 169–184. https://doi​.org​/10​.1002​/job​.2213 Currall, S. C., & Inkpen,A. C. (2002).A multilevel approach to trust in joint ventures. Journal of International Business Studies, 33(3), 479–495. https​:/​/do​​i​.org​​/10​.1​​057​/p​​algra​​ve​. ji​​​bs​.84​​91027​ De Jong, B. A., & Dirks, K. T. (2012). Beyond shared perceptions of trust and monitoring in teams: Implications of asymmetry and dissensus. Journal of Applied Psychology, 97(2), 391–406. https://doi​.org​/10​.1037​/a0026483 De Jong, B. A., Gillespie, N., Williamson, I., & Gill, C. (2020). Trust consensus within culturally diverse teams: A multistudy investigation. Journal of Management. http:​/​/dx.​​ doi​.o​​rg​/10​​.1177​​/0149​​20632​​​09436​​58 Drescher, M. A., Korsgaard, M. A., Welpe, I. M., Picot, A., & Wigand, R. T. (2014). The dynamics of shared leadership: Building trust and enhancing performance. Journal of Applied Psychology, 99(5), 771–783. http://dx​.doi​.org​/10​.1037​/a0036474 Ferguson, A. J., & Peterson, R. S. (2015). Sinking slowly: Diversity in propensity to trust predicts downward trust spirals in small groups. Journal of Applied Psychology, 100(4), 1012–1024. http://dx​.doi​.org​/10​.1037​/apl0000007 Fulmer, A., & Dirks, K. (2018). Multilevel trust: A theoretical and practical imperative. Journal of Trust Research, 8(2), 137–141. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​.15​​31657​ Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​.org​/ 10​.1177​/0149206312439327 Fulmer, C.A., & Ostroff, C. (2017).Trust in direct leaders and top leaders:A trickle-up model. Journal of Applied Psychology, 102(4), 648–657. https://doi​.org​/10​.1037​/apl0000189 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34(1), 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 Jones, S. L., & Shah, P. P. (2016). Diagnosing the locus of trust: A temporal perspective for trustor, trustee, and dyadic influences on perceived trustworthiness. Journal of Applied Psychology, 101(3), 392–414. https://doi​.org​/10​.1037​/apl0000041 Kim, P. H., Dirks, K. T., & Cooper, C. D. (2009). The repair of trust: A dynamic bilateral perspective and multilevel conceptualization. Academy of Management Review, 34, 410– 422. https​:/​/do​​i​.org​​/10​.5​​465​/A​​MR​.20​​09​​.40​​63188​7 Korsgaard, A., Bliese, P., Kautz, J., Samson, K., & Kostyszyn, P. (2018). Conceptualizing time as a level of analysis: New directions in the analysis of trust dynamics. Journal of Trust Research, 8(2), 142–165. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​.15​​16557​ Korsgaard, M. A., Brower, H. H., & Lester, S. W. (2015). It isn’t always mutual: A critical review of dyadic trust. Journal of Management, 41(1), 47–70. https://doi​.org​/10​.1177​/ 0149206314547521 Kramer, R. M., & Lewicki, R. J. (2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annals, 4(1), 245–277. http:​/​/dx.​​doi​.o​​rg​/10​​.5465​​/1941​​6520.​​201​0.​​48740​3

A Multilevel Trust Perspective  13

Lewicki, R. J., McAllister, D. J., & Bies, R. J. (1998). Trust and distrust: New relationships and realities. Academy of Management Review, 23(3), 438–458. https://doi​.org​/10​.2307​/ 259288 Lumineau, F., & Schilke, O. (2018). Trust development across levels of analysis: An embedded-agency perspective. Journal ofTrust Research, 8(2), 238–248. http:​/​/dx.​​doi​.o​​rg​/ 10​​.1080​​/2151​​5581.​​2018​.​​15317​​66 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi​.org​/10​.2307​/258792 Mayer, R. C., & Gavin, M. B. (2005). Trust for management and performance: Who minds the shop while the employees watch the boss? Academy of Management Journal, 48(5), 874–888. https​:/​/do​​i​.org​​/10​.5​​465​/A​​MJ​.20​​05​​.18​​80392​8 Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Schilke, O., & Cook, K. S. (2013). A cross-level process theory of trust development in interorganizational relationships. Strategic Organization, 11(3), 281–303. https://doi​.org​/ 10​.1177​/1476127012472096 Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007).An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32(2), 344–354. https​:/​/ do​​i​.org​​/10​.5​​465​/a​​mr​.20​​07​​.24​​34841​0 Searle, R. H., Nienaber, A. M., &. Sitkin, S. B. (2018). The Routledge companion to trust. New York: Routledge. Vanneste, B. S. (2016). From interpersonal to interorganisational trust: The role of indirect reciprocity. Journal of Trust Research, 6, 7–36. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​15​. 11​​08849​ Wicks, A. C., Berman, S. L., & Jones,T. M. (1999).The structure of optimal trust: Moral and strategic implications. Academy of Management Review, 24(1), 99–116. https://doi​.org​/ 10​.2307​/259039

2 TRUST CONCEPTUALIZATIONS ACROSS LEVELS OF ANALYSIS C. Ashley Fulmer and Cheri Ostroff

Introduction Trust in organizational settings is a complex and inherently multilevel phenomenon (Fulmer & Gelfand, 2012). Although the importance of incorporating levels of analysis issues into trust research has been recognized for at least 25 years (e.g., Mayer, Davis, & Schoorman, 1995; Rousseau, Sitkin, Burt, & Camerer, 1998) and the field of trust in management and organizational psychology has increasingly integrated trust and levels of analysis theories (Fulmer & Dirks, 2018), there continues to be inconsistency, and often confusion, in the meaning and application of levels issues to the examination of trust across different levels of analysis. Different conceptualizations and approaches to trust will necessarily depend on whether the research focus is an individual, a dyadic, a higher unit level (e.g., team and organization), or a multilevel perspective. To date, advances in clarifying trust at different levels of analysis have been made, such as distinguishing between the level of analysis and the target of trust as an individual, team, or organization (e.g., Fulmer & Gelfand, 2012), clarifying different forms of trust between two parties in a dyad (Korsgaard, Brower, & Lester, 2015), examining trust consensus at the team level (De Jong, Gillespie, Williamson, & Gill, 2020; Fulmer, 2012), the recognition that trust in organizations is both top-down and constrained by the organizational context as well as bottom-up and emergent as individuals interact and form collective and shared perceptions of trust (e.g., Lumineau & Schilke, 2018), and the notion that concepts of trust shift as levels of analysis shift from individuals to teams or units (e.g., Costa, Fulmer, & Anderson, 2018). Our knowledge about the embedded nature of trust (e.g., individuals embedded in units that are in turn embedded in organizations) and its contextual DOI: 10.4324/9780429449185-2

Trust Conceptualizations  15

influences has significantly improved. However, much of the work on trust and levels of analysis has developed in a fragmented manner, with researchers typically focusing on issues involving one level or limited attention to the role of levels, such as the relationship between interpersonal trust and team trust (e.g., Costa et al., 2018) or delineating different targets of trust (e.g., leaders, teams, organization) at different levels of analysis and their differential antecedents and consequences (e.g., Fulmer & Gelfand, 2012). Further, the preponderance of theory in research across levels has assumed compositional models (Chan, 1998; Kozlowski & Klein, 2000), assuming functionally similar constructs across individual and higher unit (e.g., teams and organizations) levels of analysis in trust constructs. Constructs at unit levels of analysis can assume both composition models, where the higherlevel construct is represented by the average of the lower-level construct if sufficient agreement exists, and compilation models, where the higher-level construct is based on the pattern of the lower-level construct. Both composition and compilation models are relevant in considering trust. Thus, a broader overview and theoretical integration of the different conceptualizations of trust across levels of analysis theories can help provide greater clarity in understanding how trust operates in organizations. Without such an understanding, researchers risk adopting trust conceptualizations and operationalizations that are ill-suited to their research question, making valid comparisons among research impossible. In what follows, we delineate different conceptualizations of trust as one moves from the individual to higher levels, across different trustor foci (e.g., ‘I’ versus ‘we’) and different trustee targets (e.g., another individual or unit as a whole). Compositional (similarity in trust across individuals or entities), dispersional (extent of variance or consensus across individuals or entities), and compilational (pattern or configuration of trust across individuals or entities) forms of trust are considered. In doing so, we highlight how the conceptualization and nature of trust change across levels based on trustor foci, trustee targets, and trust forms at higher levels. Greater attention to levels of analysis will allow researchers to avoid common pitfalls by ensuring that the construct is defined precisely for the appropriate levels of analysis, to consider how and why constructs of trust can emerge at higher levels of analysis, to ensure that theoretical rationales are based on the appropriate levels of analysis, and to avoid misspecification of levels such as applying a group construct to individuals or vice versa. In this way, the levels of analysis approach is not merely a methodological tool for aggregation from a lower level to a higher level of analysis but a theoretical means that can expand our current perspective on trust in organizational contexts.

Defining Trust Across Levels Trust at the individual level of analysis has frequently been defined as an individual’s psychological state of being willing to be vulnerable based on positive

16  C. Ashley Fulmer and Cheri Ostroff

expectations of another individual or entity (Mayer et al., 1995; Rousseau et al., 1998). While the definition of trust has largely converged among scholars at the individual level of analysis, ambiguities remain regarding definitions of trust at higher levels of analysis. This reflects the increasing complexity of trust as researchers move up from the individual level to higher levels of analysis. Definitions of trust at higher levels have largely borrowed from the definition at the individual level.Trust at higher unit levels of analysis (e.g., teams and organizations) has been defined as a shared psychological state among unit members of being willing to be vulnerable based on positive expectations (Fulmer & Gelfand, 2012). However, there is little clarity about the form of ‘sharedness’ of this psychological state. In general, there has been little discussion about the degree and type of sharedness needed beyond statistical aggregation requirements (Bliese, 2000), which limits our understanding of trust in organizations. For example, if considering trust in a unit leader, should sharedness be based on whether each individual indicates they trust the leader and the extent of similar degrees of trust across unit members? Or should sharedness be based on whether members perceive that unit members overall trust the leader? Neither is necessarily right or wrong, or better than the other, but drawing such distinctions has important implications for the meaning of constructs at different levels of analysis. Alternate conceptualizations of trust at higher levels are also indicated from a levels perspective. For example, the extent to which unit members agree or share the same level of trust can also have meaning in and of itself about unit trust but is rarely considered (De Jong et al., 2020; Fulmer, 2012). Further, the configuration or pattern of trust can be a meaningful construct, for example, if there is a bifurcated group with two subgroups whereby members in a subgroup trust one another but not the members of the other subgroup or if core members of the team strongly trust their leader but others do not.These patterns may be meaningful for understanding trust dynamics beyond just the overall level of trust in a group. By incorporating these considerations from the levels of analysis perspective in trust research, our understanding of trust concepts can be refined as well as expanded.

Levels of Analysis Fundamentals Organizations are nearly decomposable systems, meaning that they can be parsed into distinct levels, but the levels are related to one another (Simon, 1973). Being nearly decomposable means that while relationships and mechanisms can be studied at any particular level as relatively independent. Interactions among levels are also important considerations and it is impossible to fully understand any organizational phenomenon without understanding the context and nature of the nesting within organizations (Roberts, Hulin, & Rousseau, 1978). For simplicity, we decompose into the individual, dyadic, and unit levels. A unit can be a group, team, or organization that encompasses multiple individuals and dyadic relationships.

Trust Conceptualizations  17

The levels perspective requires that constructs be specified at each relevant level, with consideration given to how constructs are related within and across levels. Importantly, in conceptualizing any organizational phenomenon, the level of theory, level of measurement, and level of interpretation must be very clearly specified and be consistent (Klein, Danserau, & Hall, 1994; Roberts et al., 1978; Kozlowski & Klein, 2000). The level of theory is the level at which the focal construct exists. For example, in trust, individual-level trust could be defined as the feeling of trust existing in each individual in the organization, while team trust could be defined as existing at the team level based on a shared feeling of trust across members. The level of measurement is the level at which variables are measured. However, the level of measurement does not necessarily originate at the same level as the level of theory. For example, for team-level trust, the level of measurement could be the team as a whole by asking team members to agree on a single score of trust, or it could originate from responses of individual members that are then aggregated to represent the team level. The level of interpretation is the level at which interpretations of results are drawn. In levels of analysis theory, interpretations can only be drawn about the level from which the theory and measurement are derived. The fallacy of the wrong level occurs when the level of theory, measurement, and interpretation are inconsistent, such as interpreting results from a team-level study as if it applies to individuals.While seemingly simplistic on the surface, misalignment between the levels of theory, measurement, and interpretation are relatively common in research studies (see, for example, Gooty, Serban, Thomas, Gavin, & Yammarino, 2012). Although levels of analysis issues have been extensively specified and examined for a number of organizational constructs, particularly for organizational climate (e.g., Chan, 1998; Kozlowski & Klein, 2000; Schneider, González-Romá, Ostroff, & West, 2017), there are some unique challenges in developing multilevel perspectives of trust in organizations that require extra attention on the part of trust researchers. First, trust is, at its core, a construct residing in individuals. Although the definition of trust has evolved over time and multiple definitions continue to exist, in general, trust is recognized as being a psychological state of an individual who is willing to be vulnerable to the trustee(s) based on perceptions, expectations, and beliefs about the trustee(s) (Fulmer & Gelfand, 2012). Thus, by definition, trust resides within an individual as a psychological phenomenon.This foundational definition and the specification of where the construct resides have important implications for the meaning of trust at higher levels of analysis. For example, and as discussed below in more detail, if a researcher is interested in the shared extent to which team members perceive they trust one another and the trustor focus is referred to in measurement as ‘we’ (rather than ‘I’), the definition of trust shifts from a psychological state residing in individuals to a perceptual judgment of the degree of trust across team members. The shift from the individual to other levels, therefore, influences how scholars can study and interpret trust across levels.

18  C. Ashley Fulmer and Cheri Ostroff

A second unique feature of trust is that it can be viewed from an egocentric or bi-directional perspective. From an egocentric view, trust is based on trust residing in the individual, dyad, or unit. Yet, trust also entails bidirectional and even reciprocal notions where one party’s trust influences the other’s and vice versa (Korsgaard et al., 2015) when considered at dyadic or higher levels, such as the degree to which two or more members each trust one another. The bidirectional nature of trust can affect how scholars conceptualize trust, particularly at the dyadic and higher levels of analysis, and how they theorize and measure trust within and across levels. A third distinguishing feature of trust is that trust is directed toward something – the target of trust can be another person, a group of persons (e.g., team or management), or an entity or context (e.g., the unit or organization as a whole). Thus, trust can be directed toward a leader or colleague.Yet, trust can also be viewed as directed toward an entity or context, such as trusting a team or trusting an organization (Den Hartog, 2003; Wang, Mather, & Seifert, 2018), trusting the HR function (e.g., Becker, Huselid, & Ulrich, 2001), or trusting the organizational system of policies, procedures, and norms (e.g., Long & Sitkin, 2018). In this latter view, a conceptualization can be that of a psychological state of trusting an entity like the organization without reference to any specific individuals within that entity. Here, trusting the collective group can involve the willingness to be vulnerable to those in the group as well as the embedded unit systems, processes, and culture (e.g., Gillespie & Dietz, 2009). A slightly different conceptualization takes this further – when the trustor focus becomes an entity, the conceptualization becomes more akin to the meanings that people impute to their work context, a concept that people use to describe their environment in relation to themselves in terms of the ‘goodness’ or ‘badness’ of the context (e.g., James, James, & Ashe, 1990; James et al., 2008). In this case, the meaning attributed to the context is in terms of the degree of trustworthiness or lack thereof. Note that these definitions are all relevant views of trust but differ slightly in their underlying meaning and may have different implications for developing relevant theories. Therefore, researchers should attend very specifically to the levels of theory, measurement, and interpretation, and specify where the construct originates theoretically, the degree to which the theoretical basis is egocentric or reciprocal, and the target of trust (e.g., an individual or entity or context). From this foundation, more nuanced conceptualizations of trust in organizations can be delineated as illustrated in Tables 2.1 and 2.2.

Composition, Dispersion, and Compilation Models Building on the importance of theory, measurement, and interpretation is the approach to conceptualizing and computing trust as a higher-level construct from the individual level. Distinctions have been drawn between composition and compilation models (e.g., Chan, 1998), with some prominent levels of analysis authors considering it a continuum between the two (e.g., Kozlowski & Klein, 2000).

Psychological state resides within individual

Psychological state resides within individual

Individual– bi-directional Psychological state resides within individual; emergent property of dyad when shared

Dyadic– decomposed Perceptual state resides in dyad; emergent property of dyad when shared

Dyadic–joint

Trustor focus

‘I’ I trust X

‘I’ I trust X; X trusts me

‘I’ I trust X

‘We’ We trust each other

Level of measure Individuals Individuals Each individual Each individual within a dyad within a dyad in dyad in dyad or a unit or a unit Example trustee Single individual Single individual Single individual Dyad target

Level of theory

Individual– egocentric

Unit–joint

Unit–joint

‘We’ We trust one another (Continued)

Unit as whole

Each individual in Unit as whole unit ‘I’ ‘I’ I trust member X; I I trust my unit trust member Y members

Individuals in unit

Individuals in unit Individuals in unit

Psychological state Psychological Perceptual state resides within state resides resides in unit; individual; within emergent emergent individuals property of unit property of unit and is shared; when shared when shared emergent property of unit when shared

Unit–decomposed

TABLE 2.1 Examples of compositional models of trust and the changing nature of the construct

Trust Conceptualizations  19

Conceptual considerations

Level of data analysis

Individual

Individual– egocentric

TABLE 2.1 Continued

Dyad

Dyadic– decomposed Dyad

Dyadic–joint Unit

Unit–decomposed Unit

Unit–joint

Requires Trust can be Requires Average of Requires individuals to low to high individuals multiple dyadic members consider their on average to consider interpersonal to average own trust and trust can and weigh trusts in unit; or weigh level toward be shared trust levels of reflects extent trust across a target other person of serial dyadic different individual trust in group members and consider the amount of trust reciprocity

Individual

Individual– bi-directional

Requires members to average or weigh trust across different members and to consider trust levels of other team members Construct becomes ‘perception-like’

Unit

Unit–joint

20  C. Ashley Fulmer and Cheri Ostroff

Unit–decomposed dispersion

Unit–joint dispersion

Unit–decomposed compilation

Unit–joint compilation

Dyad

Dyad

Level of data analysis

Team

Team

Team

‘I’ I trust unit members

‘I’ I trust X

Trustor focus

‘We’ We trust one another

‘We’ We trust each other

Single individual

Example trustee target ‘I’ I trust member X; I trust member Y

Each individual Individuals in unit Individuals in Individuals in unit in dyad unit Dyad Each individual in unit Unit as whole Unit as whole

Team (Continued)

‘We’ We trust our team

Unit as whole

Individuals in unit

Psychological state resides Perceptual state; Psychological state Perceptual Psychological state Perceptual state; in individual; emergent emergent resides in individual; state; resides in individual; pattern of property of dyad property of emergent property emergent pattern of individual dyad of unit property of individual responses responses is unit is property of unit property of unit

Dyadic–joint dispersion

Level of measure Each individual in dyad

Level of theory

Dyadic–decomposed dispersion

TABLE 2.2 Examples of dispersion and compilation models of trust

Trust Conceptualizations  21

Conceptual considerations

Degree of dispersion or difference represents extent of trust similarity in dyad

Dyadic–decomposed dispersion

TABLE 2.2 Continued

Unit–decomposed dispersion

Requires Degree of dispersion individuals or variance to consider represents average trust levels of extent of trust other person; similarity degree of dispersion or difference represents extent of similarity in perceptions about trusting one another

Dyadic–joint dispersion

Unit–decomposed compilation

Requires Pattern represents members configuration to average of trust in the or weigh unit; can take trust across multiple forms different such as bifurcated, members subgroups, min/ and to max, or social consider network patterns trust levels of other team members; degree of dispersion represents extent of similarity in perceptions of trusting team

Unit–joint dispersion

Requires members to consider perceptions of trust of other team members; pattern represents configuration of perceptions of trust in the unit; can take multiple forms such as bifurcated, subgroups, min/max, or social network patterns

Unit–joint compilation

22  C. Ashley Fulmer and Cheri Ostroff

Trust Conceptualizations  23

Composition models are based on the notion of isomorphism and functional similarity of constructs across levels of analysis (see Table 2.1). In simple terms, this posits that the two constructs at the two different levels are generally similar and related to one another. A few examples include psychological climate and unit climate (James, 1982), individual and team self-efficacy (DeRue, Hollenbeck, Ilgen, & Feltz, 2010), individual and unit goal orientation (Dragoni, 2005), individual and team empowerment (e.g., Chen, Kirkman, Kanfer, Allen, & Rosen, 2007). In trust, compositional models have been explicated between interpersonal trust and team trust (e.g., Costa et al., 2018). Composition models typically rely on the average or aggregated score across individuals to represent the higher-level construct. With the exception of additive models (Chan, 1998), composition models require that some degree of agreement or similarity in responses be demonstrated across members of the focal unit in order to justify using the aggregated score to represent a higher-level construct. The level of theory, measurement, and interpretation must be considered in composition models. Composition models pertain to how a construct at one level of analysis relates to another form of that construct at another level of analysis (e.g., Chan, 1998; Rousseau, 1985). In direct consensus composition models, unit or higher-level trust derives its conceptual meaning from consensus among the lower-level units or individuals in the unit.This is distinguished from referent-shift models, whereby consensus formed across lower-level units or individuals is theoretically distinct from the original lower-level units (Chan, 1998). The nuances of considering direct consensus versus referent-shift models are often glossed over in trust research but need attention. In organizational climate, for example, it has been proposed that the unit of theory is fundamentally the individual; however, aggregation of individuals’ climate perceptions can result in an emergent property of unit climate provided there are shared perceptions among members (e.g., James, 1982). Climate is the overall meaning derived from aggregating individuals’ perceptions of their work context, i.e., an aggregation of individuals’ psychological climate perceptions given consensus in these individual perceptions (James et al., 2008). In climate, a healthy debate continues over direct consensus and referent-shift models (Schneider et al., 2017). Applying this idea to trust, referent-shift models whereby a change in the trustor focus (e.g., from individualized perceptions of ‘I’ to generalized unit perceptions of ‘we’) can represent a shift in the unit of theory assume that the construct resides in the unit itself and is a property of the unit, not individuals. We urge trust researchers to consider applications of these nuanced levels concepts, as the examples in Table 2.1 illustrate different meanings and forms of trust by utilizing these concepts. In addition to composition models, dispersion and compilation models are also relevant for trust research (see Table 2.2). Dispersion models (Chan, 1998) represent a categorically different form of composition models and focus on variance among the lower-level units or individuals. The standard deviation or average deviation among responses from unit members is typically used to represent the

24  C. Ashley Fulmer and Cheri Ostroff

extent of dispersion, or alternately, the degree of consensus (e.g., Burke, Finkelstein, & Dusig, 1999; González-Romá & Hernández, 2014). The extent of variance is purported to represent a meaningful theoretical construct that reflects some type of group dynamic pertaining to the construct of interest (e.g., González-Romá & Peiró, 2014; Lindell & Brandt, 2000; Schneider, Salvaggio, & Subirats, 2002). Notably, the extent of consensus or variability among members on the construct of interest is independent of the level of the construct (how high or low the trust). In trust research, the extent of consensus on trust (low dispersion) was found to be related to team performance, independent of the level of trust in the team (De Jong et al., 2020). Compilation models depart from composition models in that the lower-level individuals or units form a particular pattern or configuration that represents the unit construct (Kozlowski & Klein, 2000). In this case, the amount of the construct varies (e.g., the levels of trust vary in the group), but it is not random variance or a simple extent of variance (high versus low). The variance takes a particular shape or form that represents something meaningful about the trust pattern in the group. A multitude of patterns is possible, for example, bimodal (two distinct subgroups, one that trusts and one that does not), minority group (one smaller subgroup that differs from a majority), central node (one person that everyone trusts but other members do not trust each other), or entire social network patterns of trust.Very few researchers have considered trust from a compilation perspective (cf. Choi, Özer, & Zheng, 2020), so there is much room for considering patterns of trust (see Table 2.2). In sum, a multitude of levels concepts should be considered in trust research, particularly the level of theory, measurement, and interpretation and how the definition of trust subtly or substantially changes when different types of levels concepts are integrated with trust. Table 2.1 provides some illustrative examples of trust considered at the individual, dyadic, and unit levels from composition, dispersion, and compilation perspectives. Of note is that the level of theory and ultimately the conceptual definition shifts when viewing trust as residing in an individual or as being a property of the unit. These nuanced considerations can add clarity to future trust research.

An Illustration of Trust Concepts and Levels Using the foundational levels of analysis concepts provided above, we show in more detail how the application of such concepts to trust can enhance research. We develop a small heuristic model to illustrate how consideration of a few core trust concepts coupled with a few core levels concepts reveal the complexity of trust research across levels of analysis and can expand thinking on the topic (see Table 2.3). In particular, as highlighted above, one unique aspect of trust is that the target can be an individual or a collective or unit as a whole, with different bases for

Trust Conceptualizations  25

TABLE 2.3 Illustrative framework crossing trustee target, trustor focus, and levels of analysis

Trustee target An individual Trustor focus

An individual Quadrant 1 Trust example: ‘I trust my leader.’ Individual level Individuals’ own trust in an individual.

A unit

A unit Quadrant 2 Trust example: ‘I trust my team.’ Individual level Individuals’ own trust in a group of individuals as a whole.

Unit level Unit level People in a unit share similar People in a unit share similar levels of trust in an individual, levels of trust in a unit, vary in these levels of trust, or vary in these levels, or exhibit a meaningful pattern exhibit a meaningful in their levels of trust in an pattern in their levels of individual. trust in a unit. Quadrant 3 Quadrant 4 Trust example: ‘We trust our Trust example: ‘We trust the leader.’ team.’ Individual level Individuals’ perceptions of their unit’s trust in an individual.

Individual level Individuals’ perceptions of their unit’s trust in a unit as a whole.

Unit level Unit level People in a unit share similar People in a unit share similar levels of how much the levels of how much the unit as a whole trusts unit as a whole trusts a an individual, vary in unit, vary in these levels, these levels, or exhibit a or exhibit a meaningful meaningful pattern in how pattern in how much the much the unit as a whole unit as a whole trusts a trusts an individual. unit.

trusting an individual at an interpersonal level compared to trusting a collective or unit as a whole. For example, researchers examining inter-organizational relationships rarely draw conceptual distinctions between the trust employees have in organizational representatives and the trust they have in the organization overall in which these representatives work (Long & Sitkin, 2018). This oversight is unfortunate as trust can shift meanings and theoretical bases between the two. We integrate the target of trust as an individual or a unit with the trustor focus as an individual or a unit (i.e., direct consensus or referent shift). Consideration of the

26  C. Ashley Fulmer and Cheri Ostroff

trustee target (who or what is being trusted) and the trustor focus (‘I’ trust or ‘we’ trust) results in four quadrants with distinct, but related, conceptualizations of trust. These forms of trust can also be viewed from composition, dispersion, and compilation models across different levels of analysis.

Trustee Target Versus Trustor Focus Within a trust relationship, distinctions are drawn between the trustor and trustee, where the trustor has a level of trust in the trustee. In our framework, the trustor is the focal unit of interest (e.g., the individual or unit in whom trust resides in the study) and the trustee is the target of trust (e.g., who or what is to be afforded the trustee’s trust). A wide range of trustees has been examined in management and organizational psychology. The trustees most commonly studied in research have been coworkers, a supervisor, team, management, and one’s own organization (Fulmer & Gelfand, 2012). Additional examples of trust targets include customers, clients, negotiation counterparts, another team or organization, the government, and the medical system (e.g., Inkpen & Curral, 2004; Kerler & Killough, 2009; Kong, Dirks, & Ferrin, 2014; Sutton, He, Edmonds, Sheppard, & Sheppard, 2019; Van der Meer, 2010; Weibel et al., 2016).

Trustee Target Affording trust to an individual is not necessarily conceptually equivalent to affording trust to an entity such as a team, unit, organization, or management in general in the organization. Placing one’s trust in an individual such as a leader, a coworker, or a negotiation partner has been the foundation of research on interpersonal trust (Fulmer & Gelfand, 2012; Six & Sorge, 2008). The theoretical basis assumes that trust is based on dynamic interpersonal interactions between the parties and that the focal individual (trustor) makes cognitive, evaluative, and affective judgments about the degree to which they should trust their focal person or trustee (e.g., Lewicki, Tomlinson, & Gillespie, 2006; McAllister, 1995; Mayer et al., 1995). Deciding to place trust in a unit or entity as a whole, such as one’s team, another team, the organization, or leaders as a whole in the organization, is based on more complex dynamics. Although the unit or entity is comprised of people, the cognitive, evaluative, and affective judgments of the trustor are not necessarily based on personal interactions with each individual in the unit or entity. Trust within a team is also subject, in part, to team dynamics such as norms, goals, task arrangements, and leadership, and trust within an organization is partly influenced by organization cultures, strategies, policies, and processes (Costa et al., 2018; Gillespie & Dietz, 2009). Further, deciding to trust the team as a whole may be based on deep interpersonal interactions with only one or two members, superficial interactions with all team members as a group, a mix of interpersonal

Trust Conceptualizations  27

and team interactions, assumptions based on observations, or other mechanisms. In this case, the focal individual subjectively weighs and calculates some overall feeling of trust for the unit, but the weighing process may be idiosyncratic and unknown. Moreover, some research suggests that employees can project their trust level onto higher-level units, such as leaders as a whole in the organization, based on their trust experience with a single individual, such as their direct supervisor (Fulmer & Ostroff, 2017). The trust process of projecting trust to a collective higher-level unit is based on notions of trust transfer (e.g., Stewart, 2003) and has a different theoretical basis than the trust process of giving one’s trust to a particular individual based on interpersonal interactions, again highlighting how different conceptualizations take on different meanings and different theoretical underpinnings. To further illustrate this point, in our framework, we differentiate between the trustee target as an individual (e.g., a leader or a negotiation counterpart) and the trustee target as a whole entity or unit that encompasses a group of individuals (e.g., a team or an organization), consistent with theoretical notions about contextual analysis (e.g., Ostroff, 2019).

Trustor Focus Consistent with the levels of analysis literature (Chan, 1998; Kozlowski & Klein, 2000) noted above, we draw a distinction between direct consensus and referentshift measurement for the trustor.1 Research has assessed the trustor as the individual’s own psychological state of trust, typically from the perspective of ‘I,’ such as “I trust….” In more recent years, some researchers have shifted the trustor focus to ‘we,’ such as “We trust…” or “In this organization, employees trust…” when measuring unit-level trust. This shift fundamentally changes the nature of the construct for trust. Using ‘I’ as the trustor focus in a direct consensus model retains the conceptual notion that trust resides as a psychological state that exists within the individual. When these psychological states are similar or shared across individuals in a unit, then the average, aggregate, or some combination across individual responses is calculated and taken to represent the degree of unit-level trust. In contrast, changing the trustor focus to ‘we’ assumes that unit-level trust is a property of the unit or system that exists independently of any individual within the unit. Here, individual trustors are asked to describe their general evaluation of the extent to which they and others in the unit feel trust, as if it exists as an independent

1 The term ‘referent’ has been used in the trust literature to refer to the target of trust (e.g., Fulmer & Gelfand, 2012) and in the levels of analysis literature to denote the focal entity or context (e.g., ‘I,’ ‘we,’ ‘this organization’) in measurement (e.g., Chan, 1998). To avoid confusion, we use the term ‘trustor focus’ to denote the focal person/entity in measuring trust and the term ‘trustee target’ for the person/entity whom the trustor trusts.

28  C. Ashley Fulmer and Cheri Ostroff

construct of the unit. As such, it becomes more of a perception of trust, almost akin to perceptions of climate, as opposed to a collective psychological state of trust that exists in each individual in similar ways. In the following section, we highlight the implications of examining two trustee targets and two trustor foci to produce four forms of trust with related but distinct trust conceptualizations (see Table 2.3). The four forms of trust can be examined across different levels of analysis, including but not limited to the individual, dyad, unit, and organizational levels. The level of analysis concerns the level of the outcomes of interest. For example, an individual-level study focuses on an individual’s trust and individual antecedents or outcomes of trust. Here, an individual’s or employee’s trust is viewed as independent of the context or unit in which it is embedded. At the unit level, the focus is on relationships between some measure of collective, aggregated, or unit trust and unit-level antecedents or outcomes. At the unit level, the focus is on relationships among contextual or collective responses.We also note that cross-level studies are possible, such as examining features of the unit and the impact on an individual’s trust or the influence of unit trust levels on an individual’s own trust.

Trustee Targets and Trustor Focus at Different Levels The majority of the trust literature is dominated by the individual level of analysis. Yet, in the past 10 years, there has been growing interest in understanding unit or collective trust, such as trust in teams, between teams, or between organizations (Baer et al., 2018; Serva, Fuller, & Mayer, 2005; Weibel et al., 2016). In doing so, some researchers have begun to explicate the processes by which individuals’ trust can converge or coalesce to become unit-level trust (e.g., Fulmer & Dirks, 2018) by developing a theoretical compositional model between individual interpersonal trust and trust among team members (e.g., Costa et al., 2018). Other researchers have noted that the antecedents and consequences of trust differ at the individual level compared to the unit level (e.g., Fulmer & Gelfand, 2012), while still others have begun tentative explorations into the meaning of dispersion or variance in trust across unit members (e.g., De Jong et al., 2020; Fulmer, 2012) or in network patterns between organizational trust actors (e.g., Choi et al., 2020). In our quadrant descriptions in Table 2.3, in a composition model, we view trust at a higher level as only existing when members share sufficiently similar levels of trust. In other words, trust must be shared among members in a unit to justify using an aggregate score to represent unit trust. If there is insufficient agreement among unit members’ trust, unit trust is deemed not to exist. Dispersion in trust is represented as the degree of variability (or consensus) in trust. Withinindividual dispersion can be examined as the degree to which an individual’s trust varies over time. At the unit level, dispersion reflects the degree of variance among members in trust. Compilation models are also relevant and increasingly complex as they denote an entire pattern of trust across different members. Some examples

Trust Conceptualizations  29

include minimum member domination, maximum member domination, subgroups or bifurcation, skewness, centralized trust in one or a few members, or an entire configural network pattern (cf. Barrick, Stewart, Neubert, & Mount, 1998; DeRue et al., 2010; González-Romá & Hernández, 2014; Kozlowski & Klein, 2000; Sinha, Janardhanan, Greer, Conlon, & Edwards, 2016).

The Four Trust Forms For illustrative purposes, we delineate four trust forms in a two-by-two framework of trustor focus and trustee target (see Table 2.3). We provide additional details about each of these four forms to explore how levels of analysis perspectives result in alternative views of trust conceptualizations.

Quadrant 1: Individual Trustor Focus and Individual Trustee Target The individual trustee target and trustor focus are reflected in Quadrant 1 in Table 2.3, where the target of trust is an individual (e.g., a leader, a colleague, a customer, or each individual member in a unit) and the trustor focus is the self (e.g., I trust my leader; I trust this coworker). This is the most commonly examined form of trust in the trust literature (Fulmer & Gelfand, 2012). It is used in research on individual-level and unit-level interpersonal trust for trust in a single target. At the individual level of analysis, it is one part of the dyadic trust between individuals. It is one-directional in the sense that it reflects interpersonal trust toward a trustee but does not examine reciprocity. At the unit level of analysis, a compositional model can be evoked whereby trust is deemed to reside within the individual, each individual indicates their degree of trust in an individual, and if the levels of trust are similar to one another, can be aggregated to represent the overall level of trust contained in the unit. For example, team trust in the leader can be represented by the aggregated or average level of trust each individual member holds in their leader if sufficient agreement in these levels of trust is demonstrated (Costa et al., 2018). Here, the theoretical basis of trust is retained within the individual; only if the amount or level of trust is similar across individuals does unit-level trust in the leader become an emergent property of the group. An extension of this notion is a focus on the degree of dispersion or consensus in trust – for example, how much similarity or difference there is in trust in the leader across members in a unit. In this definition, the absolute level of trust (high or low) is not the issue; it is the extent to which members feel a similar (or dissimilar) amount of trust in the focal individual (Fulmer, 2012). The degree of similarity or difference in trust can have a direct impact on outcomes or it can interact with the level of trust whereby both the mean amount of trust and the extent of variability around this mean matter for outcomes.

30  C. Ashley Fulmer and Cheri Ostroff

In contrast, a compilation model can be adopted, with trust residing within each individual in the unit. While there are many potential combinations of trust across the unit members, a compilation model could compare, for example, the consequences of a pattern of high trust in a leader among half the unit and low trust in the leader among the other half with a unit in which there is moderate trust in the leader across all members. Theoretically, different patterns of trust in their leader across group members could reflect leader–member exchange differentiation (e.g., Buengeler, Piccolo, & Locklear, 2020; Henderson, Liden, Glibkowski, & Chaudhry, 2009), different interaction patterns among team members that create subgrouping or faultlines (e.g., Lau & Murnighan, 1998), or some other group process such as work interdependence or physical location differences. Rather than explicate the wide variety of possible configurations and theoretical bases for different configurations, our point is to highlight that the theoretical underpinnings shift as one moves from composition to dispersion to compilation models, even when adopting the same individual trustee target and the same individual trustor focus and with trust theoretically residing within individuals.

Quadrant 2: Individual Trustor Focus and Unit Trustee Target In the second form of trust (Quadrant 2 in Table 2.3), the trustor focus is an individual while the trustee target is a unit or collective (e.g., I trust my team; I trust the organization; I trust management). As with Quadrant 1, this conceptualization also places trust as residing within the individual. However, it differs in that it requires individuals to consider an entity as a whole in determining their level of trust. This latter aspect moves trust from an individualized interpersonal process to some combinatorial form of trust that the focal person must calculate in their head to form their overall level of trust in the entity. This form of trust has been examined at the individual level and the team level (Baer et al., 2018; Chou,Wang, Wang, Huang, & Cheng, 2008). The difference between this form of trust and that of the first quadrant is that the trust target (the trustee) changes from an individual to a unit or a group of individuals. A potential challenge with this form of trust concerns how individuals’ idiosyncratic (individual-level) or collective (aggregated at the unit level) trust is decided without distinguishing among particular members. When a whole unit is the trust target, there is no way to ascertain how the trustor determines his or her level of trust in the unit. For example, one team member may base their team trust primarily on their relationship with their leader, while another may base their team trust on some averaging of their relationships with everyone in the team. The two trust scores could be equal but conceptually they convey very different things about the meaning of trust in the team. In this form of trust, these potential differences in forming one’s overall trust for a unit are treated as an error not of substantive interest, and researchers need to be aware of this in drawing interpretations about trust.

Trust Conceptualizations  31

At the unit level, the aggregate of individuals’ trust in their unit changes the nature of the construct. In a composition model, the level of trust in the unit must exhibit some degree of sharedness in order to justify using the aggregate score to represent the unit. The fact that individuals come to form their overall judgment of trust in the team in different ways, even if the level of trust is similar across unit members, makes it difficult to interpret as a compositional construct. The aggregated construct represents the notion that team members all feel a similar level of trust in their team and that level can be high or low, but what team trust itself means and how people share that meaning become blurred. In a sense, team members are being asked to examine their context (e.g., team, unit, organization) overall and make an evaluative judgment about how much they trust (or do not trust) that context in general, making it more akin to an evaluation of the context rather than reflecting interpersonal dynamics inherent in trust. The construct has merit, but researchers need to be clear about its underlying conceptualization and the theoretical basis for how there might be some coalescing of trust across unit members. Similar issues are relevant for dispersion and compilation models in this quadrant. For example, high dispersion across team members could reflect true dissensus in the team or it could reflect different ways that different individuals calculated trust across unit members to derive an overall feeling of trust in the team. Compilation models could be derived to examine different patterns or networks of trust in the team across members, but again, care must be taken in the interpretation, given that the mental averaging to derive the feeling of trust in the team can vary across individuals. Again, the construct is valid provided that care is taken in interpreting the meaning of trust.

Quadrant 3: Unit Trustor Focus and Individual Trustee Target Quadrant 3 in Table 2.3 represents the situation whereby the trustor focus is a unit and the trustee target is an individual (e.g., members in this team trust the team leader). This form of trust has been used to examine trust in an individual (e.g., unit leader) at the unit level of analysis with unit-level antecedents and outcomes (e.g., Carter & Mossholder, 2015; Schaubroeck, Lam, & Peng, 2011). Less research has examined how an individual’s view of how much the team trusts a particular individual relates to individual antecedents or outcomes. Conceptually, this form of trust can be interpreted as perceptual and evaluative rather than a psychological state of interpersonal trust. That is, individuals are asked to assess their context or set of peers, appraise how much members trust an individual, and then combine this across all team members to form a perception of how much the team trusts a particular individual. It assumes that trust is a property of the unit or system, not necessarily residing in any one individual in the unit. To the extent that trust theoretically is a property of a unit, theory development must originate from that perspective to explicate how and why a

32  C. Ashley Fulmer and Cheri Ostroff

level of trust belongs to the unit as a whole, across all members. This is akin to studies of organizational cultural values or referent-shift models of organizational climate. The construct can be measured from individuals but reflects what the unit members believe are the normative expectations in the unit as a whole rather than what is expected of them personally (e.g., James et al., 2008). Here, members respond about the level of trust they perceive to be held in the unit toward a particular individual, rather than the level of trust they personally hold toward that individual. Dispersion reflects differences in members’ views about the extent to which their unit as a whole trusts a particular individual. Dispersion, in this case, may be a function of poor unit dynamics that do not allow for the formation of consistent perceptions across members or could reflect differences in the way different members amalgamate their beliefs about how others in the unit (or even how many in their unit) trust the individual of interest. As a compilation model, this form conceptualizes trust as a pattern in members’ perceptions of trust in a particular individual across members of the unit. For example, consistent with organization climate research (Schneider et al., 2017), subgroups can exist that are in frequent communication and hence more likely to share similar trust perceptions than unit members that are not in communication.This pattern in one team may be distinct from the social processes of another team where a single member, for example, influences the views of the majority of other members about how much trust there is in a particular individual.

Quadrant 4: Unit Trustor Focus and Unit Trustee Target With the fourth form of trust (Quadrant 4 in Table 2.3), both the trustor focus and the trustee target are a unit (e.g., we trust our organization; we trust each other in this team).This form of trust has been frequently used to examine trust in a unit at the unit level of analysis (e.g., Cogliser, Gardner, Gavin, & Broberg, 2012; De Jong, Dirks, & Gillespie, 2016; Joshi, Lazarova, & Liao, 2009). At the individual level, this form of trust concerns an individual’s own perceptions of a unit’s trust in a unit (e.g.,Verburg et al., 2018) but has rarely been examined. Conceptually, this model of trust is the furthest removed from trust residing in the individual as a psychological state. That is, trust is deemed to be a property of the unit that exists outside of any individual and the target of trust is a collection of individuals that individuals must view as a single entity to trust rather than as individuals comprising a unit. With both the trustee target and trustor focus as the unit, individuals need to form perceptions about how others in a unit trust a group of individuals, regardless of their personal level of trust. From a theoretical standpoint, the underlying mechanisms that lead to such trust perceptions may be more based on interactions and interdependencies among members than based on interpersonal trust dynamics per se. In this case, the notion of generalized exchange may be relevant as a theoretical foundation whereby each participant

Trust Conceptualizations  33

provides something to others without requiring direct reciprocation from that person, and as others do the same, each member eventually receives reciprocation indirectly (Yoshikawa, Wu, & Lee, 2018). At the individual level, this form of trust becomes an idiosyncratic perception about the degree that a unit trusts a collective entity, i.e., an individual perception about the context. At the unit level, the conceptual underpinning in a compositional model is that trust is shared perceptions that unit members have about a unit as a whole. It reflects an overall evaluation of the context in terms of the extent to which the unit trusts a whole entity, and when shared, forms a higherlevel emergent property of the unit. It is more akin to trust climate in a unit (e.g., Menges, Walter,Vogel, & Bruch, 2011) than trust as an interpersonal psychological state. A high degree of dispersion or dissensus, in this case, indicates that the construct does not exist as a property of the unit, which negates the theoretical basis of trust perceptions as being a shared property of the unit. In a compilation model, trust would be reflected as a pattern in the perceptions among team members about the unit’s trust in a unit, with different patterns reflecting different team dynamics or different consequences. For example, in a unit with one or two members who have been isolated from the rest of the unit, perceptions of unit trust might be low, but the majority might perceive trust to be higher, whereas another unit could have the same aggregate score because everyone has a moderate level of trust. These two patterns reflect different mechanisms of trust formation at the unit level.

Extensions The quadrants in Table 2.3 illustrate how meaning and conceptualizations of trust require differentiation based on different levels of analysis concepts. For the purpose of honing in on meaning shifts, we focused primarily on the target of trust and trustor focus. However, trust is complex and multifaceted (Rousseau et al., 1998), and each slight shift in level or approach requires careful conceptualization. As mentioned earlier, one aspect of trust in organizations is that it can be viewed from an interpersonal or from an entity or contextual perspective. While we have focused largely on trust within units, such as within a team, trust can extend between units or between organizations. And the target of trust between units can be another focal individual or can be the other unit as a whole. When considering trust between units, different theoretical bases for understanding trust between teams are needed when liaisons or individual actors represent their unit versus when units overall give valuations of how much another unit as a whole can be trusted (Long & Sitkin, 2018). An example of this is evidenced in trust in supply chain networks about executives’ beliefs of the extent to which another partner in their supply chain can be trusted (e.g., Choi et al., 2020). Here, the notion of an individual trustee, ‘I trust,’ is extended to a collective partner as a whole that exists outside of the focal organization.

34  C. Ashley Fulmer and Cheri Ostroff

Another extension pertains to understanding reciprocity or bidirectionality in trust. Theories of interpersonal trust emphasize that trust can be considered unilaterally (how much a focal person trusts) but that trust is dynamic between people and entails some degree of reciprocity. This unique aspect of trust adds another layer of complexity in level considerations. To illustrate, with an individual as the trustee target and the trustor focus the individual, as in Quadrant 1 in Table 2.3 (e.g., I trust person X), both parties can indicate their level of trust as a psychological state toward the other person. At the dyadic level then, the degree of reciprocity (or lack thereof) can be calculated or derived statistically. Alternately, when the trust target is a unit or entity as a whole and measurement is from the individual, as in Quadrant 2 (e.g., I trust my team), reciprocity is implicit in the focal individual’s response; by asking individuals to focus on the amount of trust in the team as a whole, evaluations will likely include consideration of reciprocal degrees of trust among team members. Only when each individual is asked how much they trust each other member would it be possible to more objectively calculate reciprocity. This notion is akin to subjective and objective fit (Edwards, Cable, Williamson, Lambert, & Shipp, 2006) where subjective fit reflects an implicit weighting of how much a person fits their context across multiple aspects of the context, and objective or actual fit measures the individual and context separately and then calculates the amount of overlap. In trust, subjective trust can occur when individuals are asked to mentally combine the degree to which each member trusts each other member, for example, to form a judgment of overall reciprocal trust. In contrast, asking each member the degree to which they trust every other member allows the researcher to calculate the degree of reciprocity independently and objectively.

Implications and Conclusions Our primary purpose in this chapter has been to emphasize the importance of incorporating levels of analysis concepts into trust theory and research. By focusing on the trustee target (e.g., individual or unit) and trustor focus (e.g., ‘I trust’ or ‘we trust’), we illustrate how the inclusion of a levels perspective allows researchers to conceptualize trust in a variety of ways. Importantly, definitions of trust shift across levels, trustee targets, and trustor foci, both expanding the domain of trust research and highlighting the critical need for researchers to better specify trust within a levels of analysis framework. Many of the different ways trust can be conceptualized noted above have been applied in research, but the nuances of what they mean are not always provided or clarified. Without clarity and direct attention to levels issues, the alignment between the trust conceptualization, theory, measures, and interpretations of the findings is likely to be suboptimal. The level of theory, measurement, and interpretation must be specified for, match, and be consistent with the level(s) of interest in the study. Too often, this level of precision is neglected in theory and empirical research. For example, using individual-level theories such as delineating an affective process through which an

Trust Conceptualizations  35

individual comes to trust a colleague does not automatically apply to unit-level or collective trust in peers. Assuming so is a fallacy of the wrong level. Rather, a researcher must specify how the unit level construct emerges from individual trust such as through ongoing interactions and event cycles with one another (Morgeson & Hofmann, 1999), vicarious learning or observation of the group, leader communication, similar cognitive structures, or other mechanisms (Fulmer & Ostroff, 2016). Similar issues apply when considering dispersion, which requires researchers to explicate why and how trust may vary between individuals as well as what it means and whether it is likely to be a positive or negative attribute of the unit. The most complex are compilation models, where a researcher must consider the array of possible patterns of trust across members (e.g., all high, core members high with others low, bifurcated, fragmented, a minimum number with high trust, etc.). In this case, theoretical development is challenging and requires not just hypothesizing about the patterns, but also specifying how, why, and which different patterns may be more or less important for key outcomes. Taken together, trust theory and research from a multilevel view are still largely in their infancy, making them exciting avenues for researchers. Likewise, moving from the trustor focus of ‘I’ to one of ‘we’ requires rethinking of theoretical bases to match a new conceptualization of trust from the perspective of the self to that of perceiving the trust level of the unit as a whole as ‘we.’ This has been demonstrated in a meta-analysis in the organizational climate literature in which referent-shift (‘we’) measures of perceptions of unit climate were a stronger predictor of unit performance while direct consensus (‘I’) perceptions of unit climate were more strongly related to unit attitudes (Wallace et al., 2016). The authors suggest that referent-shift measures of climate are more cognitively based whereas direct consensus ‘I’ perceptions are based more on affective feelings. The same change in theoretical underpinnings, from affective to cognitive, may be true in the domain of trust when the focal trustor changes from ‘I’ to ‘we.’ Again, this highlights the need for trust researchers to reconsider some of the theoretical bases when moving across trustee targets, trustor foci, levels, and forms. Once the constructs of interest have been determined and consideration given to the development of theories that specify how the constructs are defined and related within and across levels, attention can be devoted to the function of a construct. That is, does the construct at one level have more similar relationships to outcomes than that construct at another level (Morgeson & Hofmann, 1999)? For example, does trust in a leader specified with an individual trustor focus – ‘I trust my leader’ – relate to the same outcomes and in similar strengths as trust in a leader specified with a unit trustor focus – ‘we trust our leader’? While some distinctions have been drawn such as meta-analytically examining differences between trust in different targets (e.g., Colquitt, Scott, LePine, 2007; De Jong et al., 2016; Dirks & Ferrin, 2002), there is still room for improvement in explicating how and why different forms of trust with different trustee targets and different trustor foci at different levels differentially relate to outcomes.

36  C. Ashley Fulmer and Cheri Ostroff

Another difference between trust and other constructs is that individuals can experience a feeling of being trusted or felt trust (Salamon & Robinson, 2008; Zhu, Lau, & Lam, 2021). For example, an employee can feel a level of trust their team places in them and the team as a whole may feel a level of trust the employee places on the team. That is, feelings of being trusted, when more or less than desired, have implications for responses at work (Baer, Frank, Matta, Luciano, & Wellman, 2020) and may differ for different targets and different levels of analysis. Felt trust can be added as another layer to our illustrative framework so that the trustor and trustee of felt trust are differentiated as well as the level of analysis. While the strengths and weaknesses of the different forms of felt trust are beyond the scope of this chapter, we would encourage researchers to carefully choose the trustor focus, trustee target, and level of a felt trust form, clearly indicate them in the research, and make sure they are aligned with the rest of the research components just as we recommend for studying trust. Finally, we note that trust is dynamic, develops over time, and may change over time (Fulmer & Gelfand, 2012; Korsgaard, Bliese, Kautz, Samson, & Kostyszyn, 2018; Lewicki et al., 2006).While theory and research have focused on how interpersonal trust develops and changes over time, a dynamic framework of trust at higher levels of analysis is lacking. For example, how feelings of trust in one another among team members develop, emerge, and maintain over time may evoke a different process related to the dynamics of multiple team members and the team as a whole than trust between a leader and subordinate over time where power dynamics are inherent in the relationship.

Conclusion Research on trust has been adopting the levels of analysis approach to understand the complexities of trust within organizations, to expand the nomological network of trust’s antecedents and consequences, and to identify the strengths and weaknesses of the literature as a whole (Fulmer & Gelfand, 2012).We demonstrate in this chapter that trust research can further benefit from the levels of analysis approach to expand and refine current perspectives on trust. The levels of analysis approach provides not only methodological tools to study trust across different levels but is a theoretical lens to enrich and deepen our knowledge of trust, with the potential to increase the precision of research findings and their implications for practice in organizations.

References Baer, M. D., Frank, E. L. Matta, F. K., Luciano, M. M., & Wellman, N. (2020). Under trusted, over trusted, or just right? The fairness of (in)congruence between trust wanted and trust received. Academy of Management Journal, 64(1), 180–206. https://doi​.org​/10​.5465​/ amj​.2018​.0334

Trust Conceptualizations  37

Baer, M. D., van der Werff, L., Colquitt, J. A, Rodell, J. B., Zipay, K. P., & Buckley, F. (2018). Trusting the “look and feel”: Situational normality, situational aesthetics, and the perceived trustworthiness of organizations. Academy of Management Journal, 61(5), 1718–1740. https://doi​.org​/10​.5465​/amj​.2016​.0248 Barrick, M. R., Stewart, G. L., Neubert, M. J., & Mount, M. K. (1998). Relating member ability and personality to work-team processes and team effectiveness. Journal of Applied Psychology, 83(3), 377–391. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​3​.3​.3​​77 Becker, B. E., Huselid, M. A., & Ulrich, D. (2001). The HR scorecard: Linking people, strategy and performance. Harvard Business School Press. Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations: Foundations, extensions, and new directions (pp. 349–381). Jossey-Bass. Buengeler, C., Piccolo, R. F., & Locklear, L. R. (2021). LMX differentiation and group outcomes: A framework and review drawing on group diversity insights. Journal of Management, 47(1), 260–287. https://doi​.org​/10​.1177​/0149206320930813 Burke, M. J., Finkelstein, L. M., & Dusig, M. S. (1999). On average deviation indices for estimating interrater agreement. Organizational Research Methods, 2(1), 49–68. http:​/​/dx.​​ doi​.o​​rg​/10​​.1177​​/1094​​4281​9​​92100​4 Carter, M., & Mossholder, K. (2015). Are we on the same page? The performance effects of congruence between supervisor and group trust. Journal of Applied Psychology, 100(5), 1349–1363. https://doi​.org​/10​.1037​/a0038798 Chan, D. (1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83(2), 234–246. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​3​.2​.2​​34 Chen, G., Kirkman, B. L., Kanfer, R., Allen, D., & Rosen, B. (2007). A multilevel study of leadership, empowerment, and performance in teams. Journal of Applied Psychology, 92(2), 331–346. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​.2​.3​​31 Choi, E. W, Özer, Ö., & Zheng, Y. (2020). Network trust and trust behaviors among executives in supply chain interactions. Management Science, 66(12), 5823–5849. https:// doi​.org​/10​.1287​/mnsc​.2019​.3499 Chou, L. F., Wang, A. C., Wang, T. Y., Huang, M. P., & Cheng, B. S. (2008). Shared work values and team member effectiveness:The mediation of trustfulness and trustworthiness. Human Relations, 61(12), 1713–1742. https://doi​.org​/10​.1177​/0018726708098083 Cogliser, C. C., Gardner, W. L., Gavin, M. B., & Broberg, J. C. (2012). Big Five personality factors and leader emergence in virtual teams: Relationships with team trustworthiness, member performance contributions, and team performance. Group & Organization Management, 37(6), 752–784. https://doi​.org​/10​.1177​/1059601112464266 Colquitt, J.A., Scott, B.A., & LePine, J.A. (2007).Trust, trustworthiness, and trust propensity:A meta-analytic test of their unique relationships with risk taking and job performance.Journal of Applied Psychology, 92(4), 909–927. https://doi.org/10.1037/0021-9010.92.4.909 Costa, A. C., Fulmer, C. A., & Anderson, N. R. (2018). Trust in work teams: An integrative review, multilevel model, and future directions. Journal of Organizational Behavior, 39(2), 169–184. https://doi​.org​/10​.1002​/job​.2213 De Jong, B. A., Dirks, K. T., & Gillespie, N. (2016). Trust and team performance: A metaanalysis of main effects, moderators, and covariates. Journal of Applied Psychology, 101(8), 1134–1150. https://doi​.org​/10​.1037​/apl0000110

38  C. Ashley Fulmer and Cheri Ostroff

De Jong, B., Gillespie, N.,Williamson, I., & Gill, C. (2020).Trust consensus within culturally diverse teams:A multistudy investigation.Journal of Management.https://doi​.org​/10​.1177​/ 0149206320943658 Den Hartog, D. (2003). Trusting others in organizations: Leaders, management and co-workers. In B. Nooteboom & F. Six (Eds), The trust process in organizations: Empirical studies of the determinants and the process of trust development (pp. 125–46). Edward Elgar Publishing. DeRue, D. S., Hollenbeck, J., Ilgen, D., & Feltz, D. (2010). Efficacy dispersion in teams: Moving beyond agreement and aggregation. Personnel Psychology, 63(1), 1–40. https​:/​/ do​​i​.org​​/10​.1​​111​/j​​.1744​​-6570​​.200​9​​.0116​​1.x Dirks, K.T., & Ferrin, D. (2002).Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology, 87(4), 611–628. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​021​-9​​010​​.8​​7​.4​.6​​11 Dragoni, L. (2005). Understanding the emergence of state goal orientation in organizational work groups. Journal of Applied Psychology, 90(6), 1084–1095. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​021​-9​​010​​.9​​0​.6​.1​​084 Edwards, J. R., Cable, D. M., Williamson, I. O., Lambert, L. S., & Shipp, A. J. (2006). The phenomenology of fit. Journal of Applied Psychology, 91(4), 802–827. https​:/​/do​​i​.org​​/10​. 1​​037​/0​​021​-9​​010​​.9​​1​.4​.8​​02 Fulmer, C. A. (2012). Getting on the same page: How leader build trust consensus in teams and its consequences [Unpublished dissertation.] University of Maryland. Fulmer, C. A., & Dirks, K. (2018). Multilevel trust: A theoretical and practical imperative. Journal of Trust Research, 8(2), 137–141. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​.15​​31657​ Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​.org​/ 10​.1177​/0149206312439327 Fulmer, C. A., & Ostroff, C. (2016). Convergence and emergence in organizations: An integrative framework and review. Journal of Organizational Behavior, 37(S1), S122–S145. https://doi​.org​/10​.1002​/job​.1987 Fulmer, C. A., & Ostroff, C. (2017). Trust in direct leaders and top leaders: A trickle-up model. Journal of Applied Psychology, 102(4), 648–657. https://doi​.org​/10​.1037​/ apl0000189 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34(1), 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 González-Romá, V., & Hernández, A. (2014). Climate uniformity: Its influence on team communication quality, task conflict, and team performance. Journal of Applied Psychology, 99(6), 1042–1058. https://doi​.org​/10​.1037​/a0037868 González-Romá,V., & Peiró, J. M. (2014). Climate and culture strength. In B. Schneider & K. Barbera (Eds.), The Oxford handbook of organizational climate and culture (pp. 496–531). Oxford University Press. Gooty, J., Serban, A., Thomas, J. S., Gavin, M. B., & Yammarino, F. J. (2012). Use and misuse of levels of analysis in leadership research: An illustrative review of leader–member exchange. The Leadership Quarterly, 23(6), 1080–1103. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​. 20​​​12​.10​​.002 Henderson, D. J., Liden, R. C., Glibkowski, B. C., & Chaudhry,A. (2009). LMX differentiation: A multilevel review and examination of its antecedents and outcomes. The Leadership Quarterly, 20(4), 517–534. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​09​.04​​.003 Inkpen, A. C., & Currall, S. C. (2004).The coevolution of trust, control, and learning in joint ventures. Organization Science, 15(5), 586–599. https://doi​.org​/10​.1287​/orsc​.1040​.0079

Trust Conceptualizations  39

James, L. R. (1982). Aggregation bias in estimates of perceptual agreement. Journal of Applied Psychology, 67(2), 219. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.6​​7​.2​.2​​19 James, L. R., Choi, C. C., Ko, C. E., McNeil, P. K., Minton, M. K., Wright, M. A., & Kim, K. (2008). Organizational and psychological climate: A review of theory and research. European Journal of Work and Organizational Psychology, 17(1), 5–32. https​:/​/do​​i​.org​​/10​. 1​​080​/1​​35943​​207​01​​66255​0 James, L. R., James, L. A., & Ashe, D. K. (1990). The meaning of organizations: The role of cognition and values. In B. Schneider (Ed.), Organizational climate and culture (pp. 40–84). Jossey-Bass. Joshi, A., Lazarova, M. B., & Liao, H. (2009). Getting everyone on board: The role of inspirational leadership in geographically dispersed teams. Organization Science, 20(1), 240–252. https://doi​.org​/10​.1287​/Orsc​.1080​.0383 Kerler, W. A. III, & Killough, L. N. (2009). The effects of satisfaction with a client’s management during a prior audit engagement, trust, and moral reasoning on auditors’ perceived risk of management fraud. Journal of Business Ethics, 85(2), 109–136. https​:/​/ do​​i​.org​​/10​.1​​007​/s​​10551​​-00​8-​​9752-​x Klein, K. J., Dansereau, F., & Hall, R. J. (1994). Levels issues in theory development, data collection, and analysis. Academy of Management Review, 19(2), 195–229. https​:/​/do​​i​.org​​/ 10​.5​​465​/a​​mr​.19​​94​.​94​​10210​​745 Kong, D.T., Dirks, K.T., & Ferrin, D. L. (2014). Interpersonal trust within negotiations: Metaanalytic evidence, critical contingencies, and directions for future research. Academy of Management Journal, 57(5), 1235–1255. https://doi​.org​/10​.5465​/amj​.2012​.0461 Korsgaard, A., Bliese, P., Kautz, J., Samson, K., & Kostyszyn, P. (2018). Conceptualizing time as a level of analysis: New directions in the analysis of trust dynamics. Journal of Trust Research, 8(2), 142–165. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​.15​​16557​ Korsgaard, M. A., Brower, H. H., & Lester, S. W. (2015). It isn’t always mutual: A critical review of dyadic trust. Journal of Management, 41(1), 47–70. https://doi​.org​/10​.1177​/ 0149206314547521 Kozlowski, S. W. J., & Klein, K. J. (2000). A multilevel approach to theory and research in organizations: Contextual, temporal, and emergent processes. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research and methods in organizations: Foundations, extensions, and new directions (pp. 3–90). Jossey-Bass. Lau, D., & Murnighan, J. (1998). Demographic diversity and faultlines: The compositional dynamics of organizational groups. Academy of Management Review, 23(2), 325–340. https://doi​.org​/10​.2307​/259377 Lewicki, R. J., Tomlinson, E. C., & Gillespie, N. (2006). Models of interpersonal trust development: Theoretical approaches, empirical evidence, and future directions. Journal of Management, 32(6), 991–1022. https://doi​.org​/10​.1177​/0149206306294405 Lindell, M. K, & Brandt, C. J. (2000). Climate quality and climate consensus as mediators of the relationship between organizational antecedents and outcomes. Journal of Applied Psychology, 85(3), 331–348. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​5​.3​.3​​31 Long, C. P., & Sitkin, S. B. (2018). Control–trust dynamics in organizations: Identifying shared perspectives and charting conceptual fault lines. Academy of Management Annals, 12(2), 725–751. https://doi​.org​/10​.5465​/annals​.2016​.0055 Lumineau, F., & Schilke, O. (2018). Trust development across levels of analysis: An embedded-agency perspective. Journal of Trust Research, 8(2), 238–248. https​:/​/do​​i​.org​​/ 10​.1​​080​/2​​15155​​81​.20​​​18​.15​​31766​

40  C. Ashley Fulmer and Cheri Ostroff

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​95​. ​95​​08080​​335 McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations.Academy of Management Journal,38(1),24–59.https://doi​.org​/ 10​.2307​/256727 Menges, J. I., Walter, F., Vogel, B., & Bruch, H. (2011). Transformational leadership climate: Performance linkages, mechanisms, and boundary conditions at the organizational level. The Leadership Quarterly, 22(5), 893–909. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​11​.07​​.010 Morgeson, F. P., & Hofmann, D. A. (1999). The structure and function of collective constructs: Implications for multilevel research and theory development. Academy of Management Review, 24(2), 249–265. https://doi​.org​/10​.2307​/259081 Ostroff, C. (2019). Contextualizing context in organizational research. In S. E. Humphrey & J. M. LeBreton (Eds.), The handbook of multilevel theory, measurement, and analysis (pp. 39–65). American Psychological Association. Roberts, K. H., Hulin, C. L., & Rousseau, D. M. (1978). Developing an interdisciplinary science of organizations. San Francisco: Jossey-Bass. Rousseau, D. M. (1985). Issues of level in organizational research: Multi-level and crosslevel perspectives. Research in Organizational Behavior, 7, 1–37. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Salamon, S., & Robinson, S. (2008). Trust that binds: The impact of collective felt trust on organizational performance. Journal of Applied Psychology, 93(3), 593–601. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​021​-9​​010​​.9​​3​.3​.5​​93 Schaubroeck, J., Lam, S. S. K., & Peng, A. C. (2011). Cognition-based and affect-based trust as mediators of leader behavior influences on team performance. Journal of Applied Psychology, 96(4), 863–871. https://doi​.org​/10​.1037​/a0022625 Schneider, B., González-Romá,V., Ostroff, C., & West, M. A. (2017). Organizational climate and culture: Reflections on the history of the constructs in the Journal of Applied Psychology. Journal of Applied Psychology, 102(3), 468–482. https://doi​.org​/10​.1037​/ apl0000090 Schneider, B., Salvaggio, A. N., & Subirats, M. (2002). Climate strength: A new direction for climate research. Journal of Applied Psychology, 87(2), 220–229. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​021​-9​​010​​.8​​7​.2​.2​​20 Serva, M. A., Fuller, M. A., & Mayer, R. C. (2005). The reciprocal nature of trust: A longitudinal study of interacting teams. Journal of Organizational Behavior, 26(6), 625– 648. https://doi​.org​/10​.1002​/job​.331 Simon, H. A. (1973). The organization of complex systems. In H. H. Pattee (Ed.), Hierarchy theory. Braziller. Sinha, R., Janardhanan, N. S., Greer, L. L., Conlon, D. E., & Edwards, J. R. (2016). Skewed task conflicts in teams: What happens when a few members see more conflict than the rest? Journal of Applied Psychology, 101(7), 1045–1055. https://doi​.org​/10​.1037​/ apl0000059 Six, F. E, & Sorge, A. (2008). Creating a high-trust organization: An exploration into organizational policies that stimulate interpersonal trust building. Journal of Management Studies, 45(5), 857–884. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-6486​​.200​7​​.0076​​3.x

Trust Conceptualizations  41

Stewart, K. J. (2003).Trust transfer on the world wide web. Organization Science, 14(1), 5–17. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​​.1​.5​​.1281​0 Sutton, A. L., He, J., Edmonds, M. C., Sheppard, V. B., & Sheppard, V. B. (2019). Medical mistrust in black breast cancer patients: Acknowledging the roles of the trustor and the trustee.Journal of Cancer Education,34(3),600–607.https​:/​/do​​i​.org​​/10​.1​​007​/s​​13187​​-01​8​​1347-​3 Van der Meer, T. (2010). In what we trust? A multi-level study into trust in parliament as an evaluation of state characteristics. International Review of Administrative Sciences, 76(3), 517–536. https://doi​.org​/10​.1177​/0020852310372450 Verburg, R. M., Nienaber, A.-M., Searle, R. H., Weibel, A., Den Hartog, D. N., & Rupp, D. E. (2018). The role of organizational control systems in employees’ organizational trust and performance outcomes. Group & Organization Management, 43(2), 179–206. https://doi​.org​/10​.1177​/1059601117725191 Wallace, J. C., Edwards, B. D., Paul, J., Burke, M., Christian, M., & Eissa, G. (2016). Change the referent? A meta-analytic investigation of direct and referent-shift consensus models for organizational climate. Journal of Management, 42(4), 838–861. https://doi​.org​/10​.1177​/ 0149206313484520 Wang, W., Mather, K., & Seifert, R. (2018). Job insecurity, employee anxiety, and commitment: The moderating role of collective trust in management. Journal of Trust Research, 8(2), 220–237. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​.14​​63229​ Weibel, A., Den Hartog, D., Gillespie, N., Searle, R., Six, F., & Skinner, D. (2016). How do controls impact employee trust in the employer? Human Resource Management, 55(3), 437–462. https://doi​.org​/10​.1002​/hrm​.21733 Yoshikawa, K.,Wu, C-H, & Lee, H-J. (2018). Generalized social exchange and its relevance to new era workplace relationships. Industrial and Organizational Psychology, 11(3), 486– 492. https://doi​.org​/10​.1017​/iop​.2018​.100 Zhu, J., Lau, D., & Lam, L. (2021).Trust me or us? A multilevel model of individual and team felt trust by supervisors. In N. Gillespie, C. A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. Routledge.

PART II

Multilevel Trust Processes and Dynamics

3 DIVERGENCE IN COLLECTIVE TRUST Audrey Korsgaard and Paul Bliese

Introduction Trust can be specified at a variety of levels of analysis – the individual, the dyad, the group, and larger collectives. For the purposes of this chapter, we use the term collective trust to refer to the trust of social entities involving two or more individuals. Collective forms of trust can involve multiple levels such as dyads nested within groups nested within larger units or organizations. Collective trust originates in the behaviors and attitudes of individuals and is shaped by interactions between individuals and the context in which those interactions occur. In many situations, collective trust emerges over time, by which we mean that individuals in the same collective become more similar with respect to their trust-related attitudes and behaviors as they interact. It is entirely possible, however, that under certain circumstances collective trust can diverge such that members become less similar over time. Group characteristics, such as interdependence, and attributes of the broader context, such as compensation policy, likely play a role in whether patterns of collective trust are convergent or divergent over time. In this chapter, we explore the theoretical foundations of collective trust with a specific theoretical focus on understanding when collective trust does or does not converge over time. Theory and research on multilevel trust have largely assumed quasi-isomorphism for the impact of higher-level trust (Fulmer & Gelfand, 2012). For example, dyadic, group, and organizational trust are all posited to predict average levels of cooperation in the same way that an individual’s trust in a partner predicts the individual’s cooperation. While theory and research propose quasi-isomorphism across levels, relationships between collective trust and collective outcomes such as average cooperation may be non-isomorphic more often than not. That is, relationships at the collective level may differ in meaningful ways from their DOI: 10.4324/9780429449185-3

46  Audrey Korsgaard and Paul Bliese

lower-level counterparts because the processes leading to collective trust are arguably distinct from the processes leading to individual trust. Collectives – dyads, groups, and organizations – represent distinct social entities. For this reason, attributes of the social entity may influence trust in the entity as a whole. That is, group members may develop trust in the group because it possesses certain attributes such as norms and routines that promote collective competence and character (Breuer, Hüffmeier, Hibben, & Hertel, 2020; Costa, Fulmer, & Anderson, 2018). Similarly, organization trust is, in part, a product of attributes of the organization such as policies, roles, and coordinating mechanisms that assure the effective functioning of the collective and the mutual benefit of its members. Research on higher-level trust has assumed that trust is a consensus-based attribute wherein the defining feature is the mean (i.e., shared) level of trust. For collective means to provide unique insights into how relationships vary across levels, it is important for the means to differ in meaningful ways across collective entities (Bliese, Maltarich, & Hendricks, 2018).The existence of meaningfully differing consensus-based means is often demonstrated by statistical indices such as the ICC (Bliese, 2000). The ICC quantifies the degree to which members within a collective coalesce relative to differences across collectives. Most published research on collective trust finds evidence that collectives differ in meaningful ways with respect to the ICC values. However, in our experience ICC values for collective units are rarely above 0.25 for trust and other variables of importance to groups such as cohesion, leadership, or safety climate (see Bliese, Maltarich, Hendricks, Hoffman, & Adler, 2019).Values of this magnitude indicate that a sizable amount of the variance (e.g., 75% or more) in individual ratings of trust are not explained by group membership. Assuming a consensus-based approach treats residual variance in trust as random error. As such, the approach effectively ignores the idea that patterns of trust divergence might be unique influences on collective and individual outcomes (Cole, Bedeian, Hirschfield, & Vogel, 2011). In this chapter, we consider the implications for moving beyond simply assuming that the residual variance in trust is nothing more than random error; rather, we consider ways in which patterns of residual variance may help make inferences about trust processes (e.g., Cole et al., 2011) and the ways in which collective trust may converge or diverge over time. We begin by making a case for trust divergence as a meaningful collective attribute and then develop a series of propositions regarding the dynamics of trust divergence and convergence. We then address approaches to modeling trust divergence and convergence.

Divergence in Dyadic and Collective Trust Trust divergence refers to meaningful differences in trust levels within a social entity. In the case of dyads, divergent trust involves asymmetry in trust levels where one party necessarily trusts more or less than the other. Thus, in dyads, trust divergence reflects two main features: the degree of difference between trust

Divergence in Collective Trust  47

levels and the direction of those differences (i.e., which party has a higher degree of trust). In collectives involving more than two parties, divergent trust may reflect a variety of different configurations or patterns. For example, trust divergence in a newly formed group is likely to reflect varying individual propensities toward trust and risk taking, which are likely to strongly influence trusting attitudes and behaviors in the absence of shared experience. In such cases, each party may have a unique level of trust. On the other hand, a group experiencing ongoing intragroup conflict may exhibit a coherent pattern of divergence in which trust within coalitions is high and trust between coalitions is low. Viewing divergence in trust as a meaningful property implies that the lack of trust congruence is a predictable and meaningful state for a social entity, suggesting that the degree of congruence or divergence can be treated as a variable of substantive interest. The causes and consequences of dispersion are likely to be theoretically distinct from individual-level trust. Increasingly, scholars are recognizing that social entities can simultaneously possess composition and compilation features that both have unique effects and potentially interact (Cole et al., 2011). The simultaneous and joint effects of level and dispersion have been documented for a variety of collective phenomena including climate, leadership, conflict, and satisfaction (Bliese & Britt, 2001; Cole, Bedeian, & Bruch, 2011; Dineen, Noe, Shaw, Duffy, & Wiethoff, 2007; Jehn, Rispens, & Thatcher, 2010; Schneider, Ehrhart, & Macey, 2013).

Why Trust Divergence Matters There are compelling theoretical reasons to expect that divergence will have social psychological importance to the members of the collective, and the collective as a whole. When trust divergence exists, some individuals trust more than they are trusted, which bodes poorly for those who trust more. Individuals are likely to presume reciprocity in relational attitudes (Kenny & DePaulo, 1993): if Jack trusts Jill, Jack believes he is trusted by Jill. For an applied example, a leader’s trust in followers predicts how much they feel trusted by their followers, regardless of how much their followers actually trust them (Campagna, Dirks, Knight, Crossley, & Robinson, 2019). Individuals who trust more than they are trusted are therefore likely to overestimate the trust from their counterparts. Being trusted offers psychological benefits to the trustee. Trust is associated with greater risk taking in the relationship, which is manifested in a number of ways, including delegating, refraining from monitoring and controls, relying on the trustee, deferring to the trustee, and sharing sensitive information (Colquitt, Scott, & Lepine, 2007).These behaviors are generally valued by trustees. Delegation and refraining from monitoring grant trustees greater autonomy and control over their work. Reliance and deference convey to trustees that their contributions are impactful and valued. Sharing sensitive information signals inclusion to trustees. If trustees have inflated expectations for treatment, they are likely to have less

48  Audrey Korsgaard and Paul Bliese

favorable perceptions of their treatment along these dimensions. That is, they are likely to feel they lack the autonomy they deserve, their contributions are not valued, and/or they are being excluded. In sum, individuals who overestimate how much they are trusted by a partner are likely to have inflated expectations for the quality of treatment they receive from the partner, which may adversely affect their assessment of subsequent interactions with the partner. Further, individuals who overestimate how much they are trusted are likely to believe they are not receiving what they deserve from the relationship.Trust is foundational to cooperative exchanges (Blau, 1964); an imbalance of trust can lead to a perceived imbalance of voluntary exchanges between parties. Individuals who trust less are likely to contribute less to joint endeavors than individuals who trust more. As a result, individuals who overestimate how much they are trusted are likely to have inflated expectations of the reciprocation of benefits with the partner. As a result, they are apt to feel that they are not treated fairly (Rupp & Cropanzano, 2002). In summary, for individuals who trust more than they are trusted, trust divergence can lead to feelings of disempowerment, exclusion, and injustice, which are likely to erode trust, organizational attitudes, and performance. Consistent with this logic, research shows that overestimation of being trusted by a partner negatively impacts subsequent trust in the partner (Brion, Lount, & Doyle, 2015). Trust asymmetry may also undermine the attitudes and behavior of individuals who trust less than they are trusted. The same phenomenon of presumed reciprocity is likely to color their perceptions of interactions with their partner. That is, individuals who trust less than they are trusted underestimate the level of trust their partners have in them, leading them to infer less favorable motivation for their partner’s well-intentioned behavior. A trusting partner’s willingness to delegate or rely on an individual may be interpreted by the individual as an attempt to offload extra work without recompense. In fact, felt trust is associated with higher perceived overload (Wang & Huang, 2019) and, as a result, lower performance (Baer et al., 2015). Thus, trust asymmetry may be deleterious for both partners, regardless of their own levels of trust. The limited research on trust asymmetry supports this notion, finding that asymmetry in trust has unique consequences beyond the mean levels of trust. For example, Carter and Mossholder (2015) found that incongruence between leaders’ trust in their groups and groups’ trust in their leaders was negatively associated with the groups’ task performance and citizenship behavior. Similarly, Kim, Wang, and Chen (2018) found that incongruence between leaders’ and followers’ trust was negatively associated with task performance and citizenship behavior. While both studies found similar effects of absolute differences in trust, Carter and Mossholder (2015) also found that groups that underestimated their leaders’ trust performed at lower levels, whereas Kim et al. (2018) found that individuals who overestimated trust performed at lower levels. The units of analysis (groups versus individuals) may explain the differences in these findings but further research is needed to resolve these inconsistencies.

Divergence in Collective Trust  49

Trust divergence in collectives of more than two people is a more complex issue, as it can be manifested in a variety of distinct patterns. Divergence in groups may reflect a relatively random pattern of trust levels but it may also organize into various configurations of high- and low-trust relationships. In some cases, trust in one group member may be much higher or lower than trust in the remaining members of the group. Examples of this would be an informal group leader who is trusted more than the remaining group and an abusive worker who is not trusted by most of the group even though the remaining members trust each other. Another configuration could be an inner circle in which trust is high whereas trust among members of the outer circle is low. Yet another possible configuration involves coalitions or subgroups in which trust within subgroups is high but the trust between subgroups is low. For example, fault lines activate ingroup–outgroup distinctions, strengthening trust between members of the same subgroup while also undermining trust between members of different subgroups (Thatcher & Patel, 2012). Within collectives with divergent trust, the quality of dyadic interactions is likely to be undermined in the same way as described above for dyadic trust: individuals who trust more than they are trusted are likely to perceive imbalances in exchanges, and those who trust less than they are trusted are likely to misinterpret interactions in ways that erode trust. In addition, the pattern of collective divergence is likely to have distinct consequences for individuals, dyads, and the collective as a whole. For example, consider a group with an uneven distribution of trust in which one member of the group is trusted more. Given that trust in leadership is positively associated with cooperation and performance (Colquitt et al., 2007), it is likely that the group will fare better if the trusted member is in an informal leader role as opposed to a less critical role. A different group dynamic would be observed in groups with fault lines. Low trust between fault line subgroups may undermine team coordination and effectiveness. Overall, divergence in trust is likely to undermine the potential benefits of high trust within collectives. For example, De Jong and Dirks (2012) found that the average divergence in trust between team members weakened the positive relationship between group-level trust and group performance. Theory and findings on trust divergence suggest that it is valid to relax the assumption that collective trust is necessarily mutual. Instead, it is appropriate to assume that collective trust has, to varying degrees, two features: mutual level and divergence, and that these features can co-occur in various patterns.What remains unexplored are the processes and factors that lead to the emergence of different degrees and patterns of divergence.We explore these processes in the section that follows.

Toward a Theory of Trust Divergence and Convergence As an emergent process, the development of collective trust is a dynamic phenomenon. At different points in time, collective trust can manifest varying degrees of convergence or divergence, depending on the conditions at that point in time.

50  Audrey Korsgaard and Paul Bliese

Drawing on the dynamics of trust development, we advance two assumptions regarding the emergent processes in collective trust. The first assumption is that individuals within collectives are likely to possess divergent trust levels in the initial phases of relationships. The second assumption is that as relationships develop, trust between individuals will tend toward convergence. These assumptions are rooted in the dynamics of trust development. Trust is built on individual disposition and context (Kramer, 1999). Combined, these broad classes of factors help to shape initial beliefs regarding a partner’s trustworthiness (Mayer, Davis, & Schoorman, 1995). Individuals may, for example, approach a new relationship with a trusting stance because they are predisposed to trust (Rotter, 1967) and because institutional forces or structural assurances are in place to compel trustworthy behavior (Kramer, 1999; McKnight, Cummings, & Chervany, 1998). These contextual factors are emphasized in stage models of trust development (e.g., Lewicki & Bunker, 1995; Shapiro, Sheppard, & Cheraskin, 1992) which posit that relationships often begin with deterrence-based (or calculus-based) trust wherein structural assurances deter violating trust. This phase is followed by knowledge-based trust in which direct experience with others informs on their trustworthiness. Trust at this stage develops as a function of the quality of interactions between members and how individuals make sense of the outcomes of these interactions. The third level of trust, identification-based trust, develops as the trustor comes to understand and value the needs, choices, and preferences of the trustee. Transition and growth in this stage require not only high-quality interactions but identity-building factors, such as the development of a collective identity or the perception of shared values. Shared context and shared interests early in the relationship can lead members of the collective to have similar levels of trust in each other from the outset (e.g., McKnight et al., 1998). Even so, work groups are not always formed with a clear purpose or explicit goals and priorities. Thus, individual differences can produce idiosyncratic and divergent beliefs and expectations that lead to divergent initial levels of trust. Idiosyncratic beliefs and expectations may not reflect the reality of the situation but nevertheless guide behavior. For example, individual differences in trust propensity and risk propensity are likely to influence risk taking in the relationship. Similarly, differences in values and norms can produce divergent expectations and evaluation of initial interactions, leading to divergent trust levels. For example, De Jong, Gillespie, Williamson, and Gill (2020) found that cultural diversity in newly formed groups was associated with dispersion in trust in the group. Thus, we propose: Proposition 1: Early in relationships, collective trust is likely to be divergent, reflecting individual differences that have not been shaped by shared experiences. Drawing on theory of the self-reinforcing dynamics of trust (Ferrin, Bligh, & Kohles, 2008; Korsgaard, 2018), the second assumption is that trust between

Divergence in Collective Trust  51

individuals is likely to converge over time. The convergence of trust between individuals suggests a mutually reinforcing effect whereby trust is reciprocated with trust. Reciprocal trust is supported in theory and research that indicates that trusting behaviors such as cooperation and delegation are both motivated by trust and a sign of trustworthiness (Ferrin, Bligh, & Kohles, 2007). Jack’s cooperative behavior – actions that benefit Jill – is a sign of Jack’s trustworthiness. Jill accordingly trusts Jack and is therefore willing to cooperate in return. Jill’s cooperation thereby reinforces Jack’s trust. In short, one party’s trust, by motivating cooperative behavior, begets another party’s trust. Empirical evidence supports the principle of mutually reinforcing trust. Using a multi-round prisoner’s dilemma game, Ferrin et al. (2008) found that actor trust predicted actor cooperation and, similarly, actor cooperation predicted partner trust such that trust and cooperation of both parties increased over time. Further, while not formally tested, the correlations between parties’ trust increased over time, offering informal evidence of convergence. In a study of teams of project managers and developers, Serva, Fuller, and Mayer (2005) examined the cycles of trusting and trust over four observation periods. They found that the managers’ trust in developers predicted risk-taking behavior positively, which impacted developers’ trust in managers. Developers’ trust, in turn, led them to engage in risk-taking behavior toward the managers, which positively impacted managers’ trust in developers. Overall, these ideas can be formalized in a second proposition: Proposition 2: All else being equal, trust between parties is likely to converge over time.

Three Circumstances of Divergent Collective Trust A critical implication of Proposition 2 is that trust between and among parties will eventually converge. However, theory and research also suggest that collectives predictably experience divergent states of trust.Viewed through a developmental and dynamic lens, we propose that trust divergence can occur in three distinct circumstances. First, divergent trust may be a relatively temporary state reflecting the developmental stage of the group. For example, as suggested in Proposition 1, when groups initially form, they may have divergent expectations and perceptions, leading to misalignments between trusting and trustworthy behaviors between members. In such cases, convergence of trust may evolve through subsequently shared experience (Brattström, Faems, & Mähring, 2019). Second, divergent trust may be the result of destabilizing events (Vanneste, Puranam, & Kretschmer, 2014; Ballinger & Rockmann, 2010) that significantly alter the balance of the relationship. Whether this form of divergent trust subsequently converges hinges on the nature of the disruption and the quality of experiences going forward. Third, divergent trust may reflect a chronic state driven by stable contextual influences. In some cases, this may lead to dissolution of relationships. The implications of these phenomena are discussed below.

52  Audrey Korsgaard and Paul Bliese

Trust Divergence as a Developmental Phase As noted above, in the early stages of relationships, divergent levels of trust are likely to exist. In the absence of direct knowledge of trustees, idiosyncratic beliefs and expectations that are not grounded in the current reality of the relationship are liable to influence trust. However, with experience, individuals go through a process of bias correction (Vanneste et al., 2014). That is, as individuals interact and experience outcomes that are counter to their initial expectations, they adjust their trust levels accordingly. Bias correction paves the way for the influence of shared context and self-reinforcing processes that are likely to encourage the convergence of trust. We propose that the way in which bias is corrected depends on whether an individual’s initial trust was higher or lower than that of their counterpart (Vanneste et al., 2014). When individuals’ initial trust is higher than others’, they are likely to experience negative outcomes in interactions with those who are less trusting toward them. As noted above, individuals are likely to presume that they are trusted as much as they trust (Brion et al., 2015; Campagna et al., 2019). Individuals who trust more than they are trusted are apt to have relatively higher expectations for the behavior of their exchange partners. These unrealistically high expectations can lead them to perceive exchanges as unfair and to question the trustworthiness of their partners. In such cases, bias correction should lead to lower trust and trusting behavior in the future, bringing the individual more in line with others within the collective. Conversely, when individuals’ trust is lower than that of others, their expectations of their partners’ behavior will be relatively lower. As a result, their interactions are likely to exceed their expectations, leading to growth in trust and trusting behavior. While in both cases of over- and underestimation, individuals are likely to adjust their trust levels, the adjustment for individuals with initially more optimistic levels of trust is likely to be greater. A pervasive phenomenon in psychology is that individuals react more strongly to negative events than to positive events (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001). Trust is no exception: the impact of deception on perceptions of untrustworthiness is stronger than the impact of honesty on perceptions of trustworthiness (Wang, Galinsky, & Murnighan, 2009) when the outcomes are of the same magnitude. Thus, compared to upward adjustments to trust, downward adjustments to trust are likely to be greater. Over time, trust divergences should resolve toward lower average trust levels. For example, Ferguson and Peterson (2015) found that diversity in the propensity to trust subsequently led to lower levels of group trust. Research simulating the emergence of generalized trust supports this notion, finding that, depending on the initial density of trusting behavior, societies are likely to go through a “valley of death” as trust initially converges (Fang, Kimbrough, Pace, Valluri, & Zheng, 2002). Specifically, over time, trust levels degenerate, converging at low levels before rebuilding to higher levels of generalized trust.

Divergence in Collective Trust  53

Proposition 3: When divergent trust converges, the collective’s average level of trust is likely to have declined. Proposition 4: As divergent trust approaches convergence, it is likely to do so at a level near the lower trustor’s initial level, as opposed to the higher trustor’s initial level. This phenomenon is illustrated in Figure 3.1a, which displays the trust ratings of a five-member group over six observational periods. This hypothetical pattern depicts a group that formed at Time 1. Lacking strong contextual cues or shared experiences, members’ trust in the group is driven by predisposition and idiosyncratic beliefs, resulting in an initial divergence of trust. In early stages of the group’s interactions, at Times 2 and 3, team members adjust their trust levels to reflect the behavior of others in the group, with high-trusting members adjusting more than low-trusting members. By Time 4, members’ trust levels have converged. Note that the trend line for average trust is downward until the group achieves consensus on trust, at which point, the shared understandings of trust are likely to reinforce trust going forward. Once trust converges, exchanges between

Divergence as a developmental phase 5

Trust

4

3

2

1

0

1

2

3 Time

4

5

Person 1

Person 2

Person 3

Person 4

Person 5

Average

FIGURE 3.1A Divergence

6

as a developmental phase.

54  Audrey Korsgaard and Paul Bliese

parties are likely to be more balanced, laying the groundwork for a positive selfreinforcing cycle that enables growth in trust.

Trust Divergence as a Result of Destabilizing Events Events are observable actions or circumstances that occur at a discrete moment and involve the interaction among persons or between persons and the context (Morgeson, Mitchell, & Liu, 2015). Any exchange between trustor and trustee is an event that may incrementally and/or temporarily impact trust levels. Destabilizing events are ones that produce a lasting imbalance in the exchange relationship. That is, destabilizing events differentially impact the benefits, risk, or vulnerability of one party more than others. Destabilizing events can result from changes in the context (i.e., exogenous changes; Vanneste et al., 2014) or from the choices and actions of the parties involved (i.e., anchoring events; Ballinger & Rockmann, 2010). Changes to the context alter the value of the relationship such that the potential benefit of fulfilled trust, or the downside of unfulfilled trust, dramatically changes, resulting in concomitant changes in trust. Context changes that destabilize the relationship involve altering relative power or dependence, or the potential impact one party has on the outcomes of others. For example, suppose a company has downsized and restructured the roles of the remaining employees to assume greater responsibilities and discretion. With the newly defined roles, employees have the potential to do greater damage if they do not act in a trustworthy manner. Therefore, managers may trust employees less because they are required to perform at a higher level and with greater discretion than they formerly warranted. Another form of destabilizing event is an anchoring event (Ballinger & Rockmann, 2010), in which the actions that one party produces have powerful consequences for the other party. Anchoring events can be either positive – wherein trustworthy behavior exceeds expectations, and the trustee is viewed as altruistic – or negative – wherein trust is violated and the trustee is viewed as competitive. Anchoring events are so profound as to lead to lasting changes in the terms of the exchange relationship in ways that deviate from reciprocal exchange. Instead of being based on reciprocity, exchanges are based on altruism (following a positive anchoring event) or competition (following a negative anchoring event). In a relationship with a balance of exchanges, trust is liable to be mutual.When an anchoring event occurs, it will likely lead one party’s trust to markedly increase (in the case of a positive anchoring event) or decrease (in the case of a negative anchoring event) relative to the other party, leading to an imbalance of trust. In sum, both exogenous changes and anchoring events can destabilize a relationship characterized by mutual trust, which is likely to result in divergence in trust. Proposition 5: Destabilizing events in the form of exogenous changes and anchoring events are likely to result in divergence trust.

Divergence in Collective Trust  55

The path to convergence will differ depending on whether the event is positive or negative. For example, research on trust violations offers substantial insight into the impact of negative events. Depending on the sense-making, a trust violation (Kim, Dirks, & Cooper, 2009) can result in an enduring loss of trust. Recovery from lost trust requires considerable effort on the part of the trustee to rebuild the relationship and restore perceptions of trustworthiness (Kramer & Lewicki, 2010). If the trustor is firm in the belief that the trustee intentionally violated trust, the restoration of trust will require consistent behavioral demonstrations of trustworthiness over time (Kim et al., 2009). Restoration may also be stalled or scuttled by the trustee’s negative reaction to being accused because losing someone’s trust undermines self-efficacy and leads to a withdrawal of effort (Lau, Lam, & Wen, 2014). Less is known about positive destabilizing events. In one investigation involving multiple rounds of the trust game, participants were presented with an exogenous positive change: the opportunity to make substantially more money by trusting. This change led to a sharp increase in trust that persisted through the course of the experiment (Kautz, Korsgaard, & Bliese, 2019). Given that negative events have a stronger impact on distrust than do positive events on trust (Wang et al., 2009), it is likely that convergence of trust takes longer after a negative event than a positive event. Figure 3.1b illustrates a hypothetical example of a destabilizing event. In this case, Time 1 reflects a point in the life of the group when trust has converged and starts Destabilizing events 5

Trust

4

3

2

1

0

1

2

3 Time

4

5

Person 1

Person 2

Person 3

Person 4

Person 5

Average

FIGURE 3.1B Destabilizing

events.

6

56  Audrey Korsgaard and Paul Bliese

to grow through subsequent observational periods until the point of the event. At Time 4, a negative event occurs that causes two members of the group to be more vulnerable to the remaining members of the group (e.g., a change in technology or workflow design).The change leads to a divergence of trust within the group overall, reflecting the differences in vulnerability of the subgroups. Specifically, there is a dramatic drop in trust for the minority subgroup. Note that the trend line for average trust reflects an overall drop in trust. After the event, the group moves toward convergence, but it is likely to be at a lower level of average trust. With respect to both positive and negative destabilizing events, we propose that: Proposition 6a:Trust convergence after a negative destabilizing event is likely to take longer than trust convergence after a positive destabilizing event. Proposition 6b: Trust after a negative destabilizing event is likely to converge at a level that is lower than the level of mutual trust prior to the event, whereas trust after a positive destabilizing event is likely to converge at a level that is higher than the level of mutual trust prior to the event.

Trust Divergence as a Chronic State Relatively stable characteristics of the collective and the context can lead to a chronic state of divergence in collective trust. The composition of the collective contributes to trust divergence (Korsgaard, Brower, & Lester, 2015). When individuals differ in characteristics that make them more or less prone to trust, such as diversity in trust propensity (Ferguson & Peterson, 2015) or gender (Buchan, Croson, & Solnick, 2008), trust divergence is likely. Further, as noted earlier, divergence in trust may be driven by fault lines (Thatcher & Patel, 2012). Structural differentiation within the collective, including differences in power, status, dependence, and access to information, which create varying degrees of vulnerability within the collective, may also contribute to divergence in trust (Korsgaard et al., 2015). Noise and ambiguity in the social context can lead to trust divergence, as information regarding the intentions and actions of parties is obscured, making it difficult to make trust judgments grounded in reality. In cases where social context is ambiguous, idiosyncratic beliefs weigh more heavily, which, as noted above, are likely to contribute to divergence in trust. Finally, some social contexts may actively produce divergence. For instance, a toxic manager who creates and maintains strong in-groups versus out-groups represents a social context that would likely produce a chronic state of divergence Figure 3.1c illustrates one scenario involving the process leading to chronic divergence. This figure depicts a scenario involving the activation of a fault line within a group. Early in the formation of the group, trust is dispersed due to

Divergence in Collective Trust  57

Divergence as a chronic phase 5

Trust

4

3

2

1

0

1

2

3 Time

4

5

Person 1

Person 2

Person 3

Person 4

Person 5

Average

FIGURE 3.1C Divergence

6

as a chronic phase.

idiosyncratic beliefs and dispositions. Over time, group members become aware of differences that create ingroup–outgroup effects.This leads to a pattern of high trust with fellow in-group members and low trust with out-group members. The majority group is likely to have more trust in the group overall due to the proportion of positive relationships within the group. In contrast, the minority group is liable to have more negative interactions, resulting in lower trust in the group overall, despite their trust within their own subgroup. Generally, trust convergence is preferable to divergence in collectives. For individuals who trust more than others, exchanges are apt to fall below expectations, which over time is likely to lower their trust to levels that match those of the others. Given that people benefit less from low trust relationships, they are likely to exit them. However, barriers to exit may prevent dissolution of the relationship (Vanneste et al., 2014). For example, factors that limit mobility, like non-compete agreements, high unemployment rates, and job embeddedness, create barriers to exit for employees. Therefore: Proposition 7: Chronic trust divergence is more likely in relationships with high barriers to exit.

58  Audrey Korsgaard and Paul Bliese

Approaches to Modeling Divergence and Convergence The phenomena described above illustrate how the degree of convergence in trust can increase or decrease over time in predictable ways, underscoring the dynamic nature of convergence. Increasingly, scholars have called for the direct examination of the dynamics of emergence (Kozlowski, Chao, Grand, Braun, & Kuljanin, 2016). There are four main approaches to directly investigating emergent processes: qualitative, computational modeling, dynamic social network analysis, and the consensus emergence model (Lang, Bliese, & de Voogt, 2018). The qualitative approach is perhaps the most frequently used method to study emergence (Kozlowski et al., 2016). Ideally, this approach should involve methods such as ethnography and participant observation that enable the researcher to observe processes as they unfold. Direct qualitative methodology “situates the observer in the midst of the people and system undergoing change” (Kozlowski et al., 2016, p. 587). For example, Bijlsma-Frankema, Sitkin, and Weibel (2015) employed structured interviews and participant observation to model the emergence of distrust between judges and administrative staff following a reorganization of a court system. This approach enabled the authors to identify distinct stages in the process of distrust emergence. Such qualitative approaches offer rich insight for theory development. However, direct qualitative methods are better suited to theory development than to theory validation and are labor-intensive. A second approach to the study of emergent processes is computational modeling, wherein data are simulated given a set of specified theoretical parameters (Kozlowski et al., 2016). Computational modeling requires the researcher to develop formal theory, specify the mechanisms and processes of emergence, and identify the metrics of these processes and their outcomes. These elements are then designed into an agent-based simulation to test the theorized emergence process. For example, Fang et al. (2002) used computational modeling to simulate the impact of various trusting strategies on the emergence of collective trust in large-scale collectives. While computational modeling fosters theoretical refinement and validation, it involves the imposition of constraining assumptions that limit the ability to test contextualized theory (Kozlowski et al., 2016). Thus, computational models require empirical validation in natural settings. Another promising quantitative approach to modeling emergence is dynamic social network analysis (Fowler & Christakis, 2008), which examines the changes in the patterns of ties among individuals over time. Social network analysis offers numerous metrics, such as density and centralization, for assessing the configurations of trust ties, which can then be dynamically modeled using various growth modeling techniques. For example, Drescher, Korsgaard,Welpe, Picot, and Wigand (2014) demonstrated that shared leadership was associated with the growth in the density of trust ties within groups. A unique advantage of this approach is

Divergence in Collective Trust  59

that it allows for the examination of mechanisms for different configurations of divergent trust, such as fault lines and social isolation. However, network analyses require access to all individuals within the collective over multiple periods, which poses logistical and methodological challenges. The fourth approach is the consensus emergence model (CEM; Lang et al., 2018). The CEM employs mixed effect models (i.e., hierarchical linear modeling or random coefficients modeling) to examine systematic changes over time in residual variances when accounting for higher-level effects.The residual variance represents the lack of convergence in a higher-level phenomenon, and changes in residual variance represent movement toward or away from convergence. This approach has several advantages. First, it allows for the examination of convergence (changes in residual variance), independent of growth in trust level (i.e., changes in mean levels). Second, using advances in mixed model analyses, the CEM can estimate the effects of group- and individual-level predictors on changes in residual variance. Thus, it allows for the quantitative test of structural, social, and individual difference predictors of convergence. For instance, to demonstrate divergent trust as a developmental phase as depicted in Figure 3.1a, the CEM would propose that the residual variance would be negative over time. Formal tests can be conducted to determine whether the magnitude of the reduction in residual variance over time is likely to have occurred by chance (i.e., test the statistical significance of the decline). Similarly, to test the impact of destabilizing events such as depicted in Figure 3.1b, a variant of the model could be estimated that included a dummy coded variable for pre-event versus post-event. The CEM could then formally test whether this event-related dummy code was associated with a significant change (increase or decrease) in the residual variance. With respect to Figure 3.1c, the CEM could test whether ratings of leadership in the collective were related to levels of residual variance and/or to changes in the residual variance over time (see Lang, Bliese, & Adler, 2019). Finding that groups with poor levels of leadership had significantly more variance would provide support for the role of leadership in terms of maintaining chronic levels of divergence. The CEM is a flexible approach to modeling trust convergence that is amenable to both dyads and larger groups. However, unlike dynamic network modeling, it does not differentiate between different configurations of divergent trust (e.g., fault lines versus a workplace bully). Another limitation of the CEM is that the models return an omnibus estimate of the residual variance, from which inferences are drawn. In contrast, other approaches to modeling variance are often based on variables such as the standard deviation estimated for each group (e.g., Bliese & Britt, 2001) or a fault line estimate for each group (Thatcher & Patel, 2012). In short, the CEM allows one to make inferences and to model patterns of change in the residual variances over time, but the CEM results by themselves cannot be used to specifically identify collective entities that fit into one pattern versus another.

60  Audrey Korsgaard and Paul Bliese

Recommended Best Practices Each of the four approaches to the direct examination of emergence has its advantages and limitations, suggesting that a combination of methodologies will be needed to substantively advance scientific knowledge on the emergence of collective trust. Across these approaches, there are three key best practices. First, the study of emergence requires repeated observations – a longitudinal design ideally containing three or more observational periods to model change. If the goal is to understand destabilizing events, we recommend at least six periods to model events, with three measurement occasions before the event and three after (Bliese & Lang, 2016). The duration of longitudinal studies can vary considerably; for example, a brief experiment with 15 trials can be informative of emergent processes (e.g., Ferrin et al., 2008). But the overall duration and the interval of time between observations should be meaningfully linked to the context and, in particular, information about how long it takes parties to form personalized relationships based on direct experience. For example, van der Werff and Buckley (2017) examined the development of trust among newly hired accountants in four observation periods over three months. Their findings showed an initial growth in trust followed by a leveling off. This curvilinear form, consistent with theory on socialization, would not have been observed if they had limited their observations to the first two periods. A second recommended practice is to emphasize context in the study of emergence. Social and organizational context factors are powerful forces in the emergence of trust (Lumineau & Schilke, 2018).The study of context, however, should not be confused with the study of a phenomenon within a rich context (e.g., a case study) because, in such cases, context is a design constant. Rather, context must be a focal variable, meaning the emergence process should be examined across meaningfully different contexts. For example, Williams (2016) demonstrated the moderating role of group diversity on the relationship between dyadic diversity and felt trust. In this study, group diversity represented the social context of dyadic relations and necessitated observing individuals from different groups that varied in this contextual variable. Similarly, when examining the role of contextual factors in the emergence of dyadic or group trust, it is necessary to examine dyads and groups across different higher-level units. Note also that contextualizing trust does not imply that computational modeling or experiments, which tend to strip away the richness of context, are inappropriate.These approaches enable the isolation of contextual factors to ascertain a clearer causal impact. The third recommended practice is that emergence itself – not the level of trust – can be the focal outcome of the study of emergence processes. As illustrated in the preceding discussion of the interpretations of divergence, average levels of collective trust and the degree of divergence, while often co-occurring, are distinct features of collective trust. In qualitative studies, focusing attention on the process of convergence is imperative to disentangling the trajectory of

Divergence in Collective Trust  61

levels of trust from convergence. For quantitative research, that means developing unique metrics of convergence that can be tracked and modeled over time. As noted above, residual variance estimated in the CEM is a promising metric for convergence that is independent of mean levels.

Conclusion Theory on trust development rightly focuses on the level of trust, with an emphasis on the factors that contribute to higher or lower levels of trust. But, in extending theory of trust to collective levels – dyadic, group, and organizational – the exclusive focus on trust levels offers an incomplete picture of how collective trust develops. Individuals within collectives can vary substantially in the degree to which they share the same level of trust. This chapter has explored the meanings of and reasons for divergence in collective trust, which underscore the importance of directly examining the emergence of collective trust. We advance a framework and array of analytic approaches that should foster a more complete understanding of multilevel trust. Primarily, this chapter offers insights into the dynamics of the emergence of collective trust. It also offers important implications for the interplay between the degree of divergence and the overall levels of trust. Further, by highlighting the ability to differentiate between configurations of trust, we offer a bridge to understanding other phenomena that shape the social structures of the group, such as fault lines and shared leadership. Finally, the analytic approaches and best practices laid out in this chapter offer a basis for examining the impact of individual differences, context, and events on the ebb and flow of trust convergence over time. Doing so will provide a more complete picture of the nature and consequences of collective trust.

References Baer, M. D., Dhensa-Kahlon, R. K., Colquitt, J. A., Rodell, J. B., Outlaw, R., & Long, D. M. (2015). Uneasy lies the head that bears the trust:The effects of feeling trusted on emotional exhaustion. Academy of Management Journal, 58(6), 1637–1657. http://dx​.doi​.org​/ 10​.5465​/amj​.2014​.0246 Ballinger, G. A., & Rockmann, K. W. (2010). Chutes versus ladders: Anchoring events and a punctuated-equilibrium perspective on social exchange relationships. Academy of Management Review, 35(3), 373–391. https://doi​.org​/10​.5465​/amr​.35​.3​.zok373 Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than good. Review of General Psychology, 5, 323–370. http:​/​/dx.​​doi​.o​​rg​/10​​.1037​​/1089​​-2680​​​. 5​.4.​​323 Bijlsma-Frankema, K., Sitkin, S. B., & Weibel, A. (2015). Distrust in the balance: The emergence and development of intergroup distrust in a court of law. Organization Science, 26(4), 1018–1039. https://doi​.org​/10​.1287​/orsc​.2015​.0977 Blau, P. M. (1964). Exchange and power in social life. New York: Wiley. http://dx​.doi​.org​/10​. 4324​/9780203792643

62  Audrey Korsgaard and Paul Bliese

Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations: Foundations, extensions, and new directions (pp. 349–381). San Francisco, CA: Jossey-Bass. Bliese, P. D., & Britt, T. W. (2001). Social support, group consensus and stressor-strain relationships: Social context matters. Journal of Organizational Behavior, 22, 425–436. http://dx​.doi​.org​/10​.1002​/job​.95 Bliese, P. D., & Lang, J. W. (2016). Understanding relative and absolute change in discontinuous growth models: Coding alternatives and implications for hypothesis testing. Organizational Research Methods, 19(4), 562–592. http:​/​/dx.​​doi​.o​​rg​/10​​.1177​​/ 1094​​42811​​​66335​​02 Bliese, P. D., Maltarich, M. A., & Hendricks, J. L. (2018). Back to basics with mixed-effects models: Nine take-away points. Journal of Business and Psychology, 33, 1–23. http:​/​/dx.​​ doi​.o​​rg​/10​​.1007​​/s108​​69​-01​​​7​-949​​1-z Bliese, P. D., Maltarich, M. A., Hendricks, J. L., Hofmann, D. A., & Adler, A. B. (2019). Improving the measurement of group-level constructs by optimizing between-group differentiation. Journal of Applied Psychology, 104, 293–302. http://dx​.doi​.org​/10​.1037​/ apl0000349 Brattström,A., Faems, D., & Mähring, M. (2019). From trust convergence to trust divergence: Trust development in conflictual interorganizational relationships. Organization Studies, 40(11), 1685–1711. http:​/​/dx.​​doi​.o​​rg​/10​​.1177​​/0170​​84061​​​87891​​95 Breuer, C., Hüffmeier, J., Hibben, F., & Hertel, G. (2020). Trust in teams: A taxonomy of perceived trustworthiness factors and risk-taking behaviors in face-to-face and virtual teams. Human Relations, 73(1), 3–34. http:​/​/dx.​​doi​.o​​rg​/10​​.1177​​/0018​​72671​​​88187​​21 Brion, S., Lount Jr, R. B., & Doyle, S. P. (2015). Knowing if you are trusted: Does metaaccuracy promote trust development? Social Psychological and Personality Science, 6(7), 823–830. http:​/​/dx.​​doi​.o​​rg​/10​​.1177​​/1948​​55061​​​55902​​00 Buchan, N. R., Croson, R. T. A., & Solnick, S. (2008). Trust and gender: An examination of behavior and beliefs in the investment game. Journal of Economic Behavior & Organization, 68(3/4), 466–476. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jebo​​.200​7​​.10​.0​​06 Campagna, R. L., Dirks, K. T., Knight, A. P., Crossley, C., & Robinson, S. L. (2019). On the relation between felt trust and actual trust: Examining pathways to and implications of leader trust meta-accuracy. Journal of Applied Psychology. https://doi​.org​/10​.1037​/ apl0000474 Carter, M. Z., & Mossholder, K. W. (2015). Are we on the same page? The performance effects of congruence between supervisor and group trust. Journal of Applied Psychology, 100(5), 1349–1363. http://dx​.doi​.org​/10​.1037​/a0038798 Cole, M. S., Bedeian, A. G., & Bruch, H. (2011). Linking leader behavior and leadership consensus to team performance: Integrating direct consensus and dispersion models of group composition. Leadership Quarterly, 22(2), 383–398. https​:/​/do​​i​.org​​/10​.1​​016​/j​​. leaq​​ua​.20​​​11​.02​​.012 Cole, M. S., Bedeian, A. G., Hirschfeld, R. R., & Vogel, B. (2011). Dispersion-composition models in multilevel research a data-analytic framework. Organizational Research Methods, 14(4), 718–734. https://doi​.org​/10​.1177​/1094428110389078 Colquitt, J. A., Scott, B. A., & LePine, J. A. (2007).Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 92(4), 909. http:​/​/dx.​​doi​.o​​rg​/10​​.1037​​/0021​​-9010​​​.92​. 4​​.909

Divergence in Collective Trust  63

Costa, A. C., Fulmer, C. A., & Anderson, N. R. (2018). Trust in work teams: An integrative review, multilevel model, and future directions. Journal of Organizational Behavior, 39(2), 169–184. http://dx​.doi​.org​/10​.1002​/job​.2213 De Jong, B. A., & Dirks, K.T. (2012). Beyond shared perceptions of trust and monitoring in teams: Implications of asymmetry and dissensus. Journal of Applied Psychology, 97(2), 391. http://dx​.doi​.org​/10​.1037​/a0026483 De Jong, B., Gillespie, N.,Williamson, I., & Gill, C. (2020).Trust consensus within culturally diverse teams: A multistudy investigation. Journal of Management, http:​/​/dx.​​doi​.o​​rg​/10​​. 1177​​/0149​​20632​​​09436​​58 Dineen, B. R., Noe, R. A., Shaw, J. D., Duffy, M. K., & Wiethoff, C. (2007). Level and dispersion of satisfaction in teams: Using foci and social context to explain the satisfaction-absenteeism relationship. Academy of Management Journal, 50(3), 623–643. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​07​​.25​​52598​7 Drescher, M. A., Korsgaard, M. A., Welpe, I. M., Picot, A., & Wigand, R. T. (2014). The dynamics of shared leadership: Building trust and enhancing performance. Journal of Applied Psychology, 99(5), 771. http://dx​.doi​.org​/10​.1037​/a0036474 Fang, C., Kimbrough, S. O., Pace, S.,Valluri, A., & Zheng, Z. (2002). On adaptive emergence of trust behavior in the game of stag hunt. Group Decision and Negotiation, 11(6), 449– 467. https://doi​.org​/10​.1023​/A​:1020639132471 Ferguson, A. J., & Peterson, R. S. (2015). Sinking slowly: Diversity in propensity to trust predicts downward trust spirals in small groups. Journal of Applied Psychology, 100(4), 1012. http://dx​.doi​.org​/10​.1037​/apl0000007 Ferrin, D. L., Bligh, M. C., & Kohles, J. C. (2007). Can I trust you to trust me? A theory of trust, monitoring, and cooperation in interpersonal and intergroup relationships. Group & Organization Management, 32(4), 465–499. https://doi​.org​/10​.1177​/ 1059601106293960 Ferrin, D. L., Bligh, M. C., & Kohles, J. C. (2008). It takes two to tango: An interdependence analysis of the spiraling of perceived trustworthiness and cooperation in interpersonal and intergroup relationships. Organizational Behavior and Human Decision Processes, 107(2), 161–178. http:​/​/dx.​​doi​.o​​rg​/10​​.1016​​/j​.ob​​hdp​.2​​​008​.0​​2​.012​ Fowler, J. H., & Christakis, N.A. (2008). Estimating peer effects on health in social networks: A response to Cohen-Cole and Fletcher; and Trogdon, Nonnemaker, and Pais. Journal of Health Economics, 27(5), 1400–1405. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jhea​​leco.​​20​08.​​07​.00​1 Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​.org​/ 10​.1177​/0149206312439327 Jehn, K. A., Rispens, S., & Thatcher, S. M. (2010). The effects of conflict asymmetry on work group and individual outcomes. Academy of Management Journal, 53(3), 596–616. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​10​​.51​​46897​8 Kautz, J., Korsgaard, M. A., Bliese, P. (August, 2019). The Timing of Accelerating Events in Trust Growth, Academy of Management. Boston, MA, US. Kenny, D. A., & DePaulo, B. (1993). Do people know how others view them? An empirical and theoretical account. Psychological Bulletin, 114, 145–161. http:​/​/dx.​​doi​.o​​rg​/10​​.1037​​/ 0033​​-2909​​.​114.​​1​.145​ Kim, P. H., Dirks, K. T., & Cooper, C. D. (2009). The repair of trust: A dynamic bilateral perspective and multilevel conceptualization. Academy of Management Review, 34(3), 401–422. http:​/​/dx.​​doi​.o​​rg​/10​​.5465​​/amr.​​2009.​​​40631​​887 Kim, T. Y., Wang, J., & Chen, J. (2018). Mutual trust between leader and subordinate and employee outcomes. Journal of Business Ethics, 149(4), 945–958. http:​/​/dx.​​doi​.o​​rg​/10​​. 1007​​/s105​​51​-01​​​6​-309​​3-y

64  Audrey Korsgaard and Paul Bliese

Korsgaard, M. A. (2018). Reciprocal trust: A self-reinforcing dynamic process. In R. H. Searle & A-M. I. Nienaber & S. B. Sitkin (Eds.), Routledge companion to trust research. New York: Routledge. Korsgaard, M. A., Brower, H. H., & Lester, S. W. (2015). It isn’t always mutual: A critical review of dyadic trust. Journal of Management, 41(1), 47–70. https://doi​.org​/10​.1177​/ 0149206314547521 Korsgaard, M. A., Kautz, J., Bliese, P., Samson, K., & Kostyszyn, P. (2018). Conceptualising time as a level of analysis: New directions in the analysis of trust dynamics. Journal of Trust Research, 8(2), 142–165. http:​/​/dx.​​doi​.o​​rg​/10​​.1080​​/2151​​5581.​​2018​.​​15165​​57 Kozlowski, S. W., Chao, G. T., Grand, J. A., Braun, M. T., & Kuljanin, G. (2016). Capturing the multilevel dynamics of emergence: Computational modeling, simulation, and virtual experimentation. Organizational Psychology Review, 6(1), 3–33. http:​/​/dx.​​doi​.o​​rg​/ 10​​.1177​​/2041​​38661​​​45479​​55 Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50(1), 569. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​. psy​​c​h​.50​​.1​.56​9 Kramer, R. M., & Lewicki, R. J. (2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annals, 4(1), 245–277. http:​/​/dx.​​doi​.o​​rg​/10​​.5465​​/1941​​6520.​​201​0.​​48740​3 Lang, J. W. B., Bliese, P. D., & Adler, A. B. (2019). Opening the black box: A multilevel framework for studying group processes. Advances in Methods and Practices in Psychological Science, 2, 271–287. http:​/​/dx.​​doi​.o​​rg​/10​​.1177​​/2515​​24591​​​88237​​22 Lang, J.W. B., Bliese, P. D., & de Voogt, A. (2018). Modeling consensus emergence in groups using longitudinal multilevel methods. Personnel Psychology, 71, 255–281. http://dx​.doi​. org​/10​.1111​/peps​.12260 Lau, D. C., Lam, L. W., & Wen, S. S. (2014). Examining the effects of feeling trusted by supervisors in the workplace: A self-evaluative perspective. Journal of Organizational Behavior, 35(1), 112–127. https://doi​.org​/10​.1002​/job​.1861 Lewicki, R. J., & Bunker, B. B. (1995). Trust in relationships: A model of development and decline. In B. B. Bunker & J. Z. Rubin (Eds.), Essays inspired by the work of Morton Deutsch (pp. 133–173). San Francisco, CA: Jossey-Bass. Lumineau, F., & Schilke, O. (2018). Trust development across levels of analysis: An embedded-agency perspective. Journal of Trust Research, 8(2), 238–248. http:​/​/dx.​​doi​. o​​rg​/10​​.1080​​/2151​​5581.​​2018​.​​15317​​66 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​. 19​​95​.​95​​08080​​335 McKnight, D. H., Cummings, L. L., & Chervany, N. L. (1998). Initial trust formation in new organizational relationships. Academy of Management Review, 23(3), 473–490. http:​/​/ dx.​​doi​.o​​rg​/10​​.5465​​/amr.​​1998​.​​92662​2 Morgeson, F. P., Mitchell, T. R., & Liu, D. (2015). Event system theory: An event-oriented approach to the organizational sciences. Academy of Management Review, 40(4), 515–537. http://dx​.doi​.org​/10​.5465​/amr​.2012​.0099 Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35(4), 651–665. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-6494​​.1967​​​.tb01​​454.x​ Rupp, D. E., & Cropanzano, R. (2002).The mediating effects of social exchange relationships in predicting workplace outcomes from multifoci organizational justice. Organizational Behavior and Human Decision Processes, 89(1), 925–946. https​:/​/do​​i​.org​​/10​.1​​016​/s​​0​749-​​ 5978(​02)00​036-5​

Divergence in Collective Trust  65

Schneider, B., Ehrhart, M. G., & Macey, W. H. (2013). Organizational climate and culture. Annual Review of Psychology, 64(1), 361–388. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​-psy​​ch​11​​​3011-​​14380​9 Serva, M. A., Fuller, M. A., & Mayer, R. C. (2005). The reciprocal nature of trust: A longitudinal study of interacting teams. Journal of Organizational Behavior, 26(6), 625– 648. https://doi​.org​/10​.1002​/job​.331 Shapiro, D. L., Sheppard, B. H., & Cheraskin, L. (1992). Business on a handshake. Negotiation Journal, 8(4), 365–377. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1571​​-9979​​.1992​​​.tb00​​679.x​ Thatcher, S. M., & Patel, P. C. (2012). Group faultlines: A review, integration, and guide to future research. Journal of Management, 38(4), 969–1009. https://doi​.org​/10​.1177​/ 0149206311426187 van der Werff, L., & Buckley, F. (2017). Getting to know you: A longitudinal examination of trust cues and trust development during socialization. Journal of Management, 43(3), 742–770. https://doi​.org​/10​.1177​/0149206314543475 Vanneste, B. S., Puranam, P., & Kretschmer, T. (2014). Trust over time in exchange relationships: Meta-analysis and theory. Strategic Management Journal, 35(12), 1891–1902. https://doi​.org​/10​.1002​/smj​.2198 Wang, C. S., Galinsky, A. D., & Murnighan, J. K. (2009). Bad drives psychological reactions, but good propels behavior: Responses to honesty and deception. Psychological Science, 20(5), 634–644. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-9280​​.200​9​​.0234​​4.x Wang, H., & Huang, Q. (2019). Feeling trusted and employee outcomes:The double-edged sword of political behaviour. Personnel Review, 48, 1653–1668. https://doi​.org​/10​.1108​/ pr​-11​-2017​-0368 Williams, M. (2016). Being trusted: How team generational age diversity promotes and undermines trust in cross-boundary relationships. Journal of Organizational Behavior, 37(3), 346–373. https://doi​.org​/10​.1002​/job​.2045

4 THE RELATIONSHIP BETWEEN TRUST AND ATTRIBUTIONS A Levels-of-Analysis Perspective Edward Tomlinson and Luke Langlinais

Introduction Trust is a psychological state reflecting the willingness to be vulnerable to another based on confident positive expectations (Rousseau, Sitkin, Burt, & Camerer, 1998). In contrast, an attribution is a psychological judgment regarding the cause of a behavior or event (Fiske & Taylor, 2013). Researchers have begun to explore how trust (i.e., a psychological state) is related to attributions (as psychological judgments) (e.g., Tomlinson & Mayer, 2009), and this chapter is intended to continue developing an understanding of this connection. Because both trust and attributions are psychological phenomena, they ultimately originate at the individual level. However, it has been recognized that both trust and attributions can be situated in contexts that influence one’s psychological experience (Fulmer, 2018). One of the most salient contexts is an interpersonal relationship, and since relationship partners both influence and are influenced by one another (Ferrin, Bligh, & Kohles, 2015; Gilbert, Pelham, & Krull, 1988), this context has been fertile ground for research on how trust and attributions are related (for a recent review, see Tomlinson, 2018). However, scholars have recently called for more research to examine how this relationship is affected when considering phenomena at different levels of analysis (Fulmer & Gelfand, 2012).Therefore, this chapter will approach the trust-attribution relationship from a levels-of-analysis perspective. The term “levels-of-analysis” should be understood to encompass constructs that cross levels (e.g., individual team member attributions shaping team-level trust) or incorporate constructs at one level relating to other constructs at multiple levels (Fulmer, 2018). Toward that end, this chapter is organized as follows. First, we provide a brief summary of the extant research on the relationship between trust and attributions DOI: 10.4324/9780429449185-4

Trust and Attributions  67

to establish the basic foundation for our analysis. As already stated, both trust and attributions are ultimately “individual” in nature. Thus, it is no surprise that the bulk of research on this relation has not involved a levels-of-analysis perspective. However, pursuing the possibility that psychological phenomena can be influenced by various social influences (Deutsch & Krauss, 1965) entails a careful consideration of how levels of analysis might inform and enrich our basic foundation. Therefore, we will proceed to review relevant research incorporating a levels-ofanalysis perspective. In particular, we will examine multi- and cross-level research on attributions. Finally, we will close with observations and recommendations for subsequent research on how the trust–attribution nexus can be explored beyond the virtually exclusive focus on the individual level that has characterized prior work.

The Relationship Between Trust and Attributions Most trust researchers regard the formal study of trust to have begun with a publication by Morton Deutsch (1958). Coincidentally, the origin of the formal study of attributions is generally credited to Fritz Heider in the same year (1958). Over the next decade, a variety of specific theories regarding the attribution process were developed. In reviewing some of this work, Kelley (1967) asserted that “Attribution theory has important statements to make about the conditions and dilemmas [of establishing trust in interpersonal relationships]” (p. 235). In a much later review of trust research, Kramer (1999) observed that interaction histories [between actor and observer] give decision makers information that is useful in assessing others’ dispositions, intentions, and motives. This information, in turn, provides a basis for drawing inferences regarding their trustworthiness and for making predictions about their future behavior. (p. 575) Fortunately, we have a body of theoretical and empirical research that has explored the relationship between trust and attributions. To frame our analysis, we draw upon the Mayer, Davis, and Schoorman (1995) model of trust.This model specifies that trust (as the willingness to be vulnerable to a trustee) is a joint function of both trustee and trustor characteristics. Specifically, trust is predicted by the trustor’s perception of the trustee’s trustworthiness, as well as by the trustor’s propensity to trust. Trustworthiness is composed of three dimensions: ability, benevolence, and integrity. Ability refers to situation-specific or task-based competence. Integrity considers the consistency of the trustee’s actions and the match of these actions with the trustor’s values. Benevolence accounts for caring and good intentions. The outcome of trust is risk taking in the relationship

68  Edward Tomlinson and Luke Langlinais

(where the trustor moves from being willing to be vulnerable to actually engaging in vulnerability); this relationship is moderated by perceived risk. Finally, the outcome of risk taking in the relationship is assessed by the trustor in comparison to what was anticipated, and a feedback loop in the model is designated to represent the updated perceptions of trustworthiness for consideration in subsequent trustrelevant interactions. Notably, this model was specifically designed to accommodate a levels-of-analysis perspective (Schoorman, Mayer, & Davis, 2007). In terms of the attribution process, we will draw primarily from Weiner’s (1986) attribution theory. Interestingly, this was originally proposed as an intrapersonal theory (e.g., how the attributions individuals make for their own performance affect their subsequent motivation, emotions, and behavior), and later expanded to the case where observers make attributions for the causes of an actor’s behavior. Regardless, the theory indicates that an outcome (a behavior, an event) prompts individuals to search for its cause. Weiner’s theory posits that the resulting causal ascriptions (whatever they may be) will be meaningfully analyzed along three distinct dimensions. First, locus of causality indicates the attributor’s conclusion on the degree to which the cause is deemed internal to the actor or external (e.g., another person, or some situational factor). The second dimension is controllability, which indicates the attributor’s conclusion regarding how much control an actor had over the cause of an outcome; even though an outcome might be internal, it might not be controllable (e.g., a heart attack). Finally, stability indicates the attributor’s conclusion regarding how temporary or permanent the cause of an outcome is. Weiner’s theory predicts that this dimensional analysis leads to specific emotional reactions and subsequent behavioral reactions (with the latter based on conclusions regarding stability). Returning to the quotes above from researchers contemplating a relationship between trust and attributions, we can now begin to formulate a more detailed connection. Because an attribution is a suspected or inferred cause of a behavior or event, our initial analysis of the trust–attribution nexus indicates attributions are only relevant after an outcome in a trust-relevant episode.That is, the outcome (positive or negative) will be analyzed to determine its cause (i.e., the trustee’s ability, benevolence, integrity, or some cause unrelated to the trustee; Tomlinson & Mayer, 2009).The trustee’s perceived ability, benevolence, and/or integrity may be updated as a result of this attribution process, with a corresponding effect on subsequent trust. This process is depicted in Figure 4.1. This specific theoretical reasoning is generally consistent with an early study by Strickland (1958) finding that when a subordinate’s trustworthy behavior was regarded by a supervisor as internal (i.e., an indication of the subordinate’s actual motives), this led to more trust. However, if the same behavior was attributed to an external cause (i.e., the supervisor’s close monitoring), then this led to lower trust. In a fascinating twist, the results in this experiment were produced not by any actual difference among subordinates’ work performance; they were solely due to attributions the supervisors made for subordinate behavior driven by differences in

Trust and Attributions  69

Trustworthiness T1 -Ability -Benevolence -Integrity

Causal Ascription -Other -Ability -Benevolence -Integrity

Trust T1

Causal Attribution Analysis -Locus of causality -Controllability -Stability

Risk Taking

Trustworthiness T2 -Ability -Benevolence -Integrity

Outcome -Negative -Positive

Trust T2

The attribution process

FIGURE 4.1 The

(indirect) relationship between attributions and trust.

Source: Adapted from Mayer et al. (1995) and Tomlinson and Mayer (2009).

how much the supervisors were able to monitor their subordinates. Many subsequent studies at the individual level are consistent with the posited linkage shown in Figure 4.1, whereby trust eventually leads to an outcome that is subjected to attributional analysis; consequently, trustworthiness perceptions are updated for a subsequent trust-relevant episode (Tomlinson & Mayer, 2009). The attribution process is key to understanding how and why perceptions of trustworthiness are revised. The accuracy and characteristics of the attributions made regarding why a certain outcome occurred will affect future decisions regarding trust. Variations on this initial analysis have also appeared in the extant literature. For example, a stream of research by Kim and his colleagues (Ferrin, Kim, Cooper, & Dirks, 2007; Kim, Dirks, Cooper, & Ferrin, 2006; Kim, Ferrin, Cooper, & Dirks, 2004) indicates that (1) outcomes attributed to different forms of trustworthiness (ability versus integrity) appear to have important implications for subsequent attributional analysis and, hence, updated trustworthiness and trust, and (2) one does not need to have a personal trust-relevant interaction with the trustee prior to drawing attributions regarding an outcome to determine trustworthiness and trust (i.e., one can draw from a vicarious instead of first-hand experience).

Literature Review Whereas the prior section outlined the basic theory and empirical research on the relationship between trust and attributions at the individual level, this section delves into research that has expanded the frontier to include a levels-of-analysis perspective. Tables 4.1 and 4.2 provide a summary of the reviewed literature and ideas for future research.

70  Edward Tomlinson and Luke Langlinais

TABLE 4.1 Attributions as a predictor – multilevel research involving attribution theories

and suggested opportunities Article

Levels-of-analysis operationalization and key findings

Potential attribution–trust research questions

Sanders et al. (2008)

Distinctiveness and consistency Investigate whether higher-level attributions of the human resource attribution measures lead management (HRM) system to commitment due to the at the individual level (but not mediating effect of trust. consensus at the department level) predicted individual-level affective commitment. Dithurbide Team-level stability attributions Team-level stability attributions et al. (2009) predict team perceptions of team performance in of collective efficacy, while conjunction with unequivocal individual-level differences do not evidence of team success have a significant effect. (objective performance) may enhance trust in the team. Riolli and Group-level attribution style had a Group attributional style might Sommer significant effect on individual impact individual-level trust (2010) turnover. regardless of actual objective indicators. Vlachos et al. Managers’ unit-level genuine and self- Investigate higher-level social (2017) serving CSR causal attributions influence processes on an affect individual-level genuine employee’s attributions, such and self-serving CSR causal as the influence of important attributions; in turn, genuine others that an employee trusts. attributions affect employee advocacy on behalf of the organization.

While trust research has been conducted at higher levels, studies including attributions at higher levels have not been typical in this literature (Fulmer & Gelfand, 2012). More directly to the point of this chapter, we believe that attribution theories may serve as a useful lens to apply in multilevel trust research. Other theorists have posited the relevance of attributions in higher-level trust research (e.g., Gillespie & Dietz, 2009; Janowicz-Panjaitan & Krishnan, 2009), and the extant trust literature has already discussed how trust itself manifests at higher levels of analysis (e.g., De Jong & Dirks, 2012). However, we are unaware of any extant empirical trust research that has incorporated attributions in multi- or cross-level studies. Nonetheless, the broader literature on attribution theory (i.e., non-trust research) has incorporated a levels-of-analysis perspective (Hewstone, 1989). We chose to carefully review this work for insights it may have for further illuminating

Trust and Attributions  71

TABLE 4.2 Attributions as an outcome – multilevel research involving attribution theories

and suggested opportunities Article

Levels-of-analysis operationalization and key findings

Sherman Team-level performance affects and Kim individual-level internal (2005) attributions, but this effect is qualified by individual selfaffirmation (i.e., sense of self-worth independent of team membership). Cherpitel et Aggregated individual and socioal. (2006) cultural context factors influence the relationship between individual perceptions/behaviors and that individual’s causal attributions for his/her outcome. Chow and Individual perceptions of team success Feltz or failure as well as individual and (2008) aggregated collective efficacy beliefs were significant predictors of team attribution dimensions. Vlachos et Unit-level perceptions of charismatic al. (2013) leadership affect individuallevel (internal versus external) attributions, which impact job satisfaction. Van De The unit-level effect of highVoorde performance work systems and influences individual-level human Beijer resource attributions. (2015)

Sanders and Unit-level understanding of highYang commitment human resource (2016) management leads to individuallevel attributions and outcomes.

Potential attribution–trust research questions Investigate whether individuals on a failing team experience a selfserving bias and place greater blame on their team, hence negatively affecting team trust. Investigate macro-contextual variables that may influence the attribution–trust relationship.

Investigate the effect of attribution re-training on the process of making accurate attributions of and trust judgments in teams. Examine the effect of unitlevel perceptions of manager trustworthiness on employees’ attributions regarding organizational strategic activities. Investigate whether trust mediates the relationship between attributions regarding HR programs and employee outcomes (such as organizational commitment and job strain). Investigate how trust in relation to the communication of organizational systems may influence attributions made about those systems and management’s motivations.

levels-of-analysis research on the relationship between trust and attributions. In the review that follows, we organize the literature by specifying the empirical relationships among constructs, careful to note the level of each. We begin with research that has examined attributions as a predictor and conclude with research that has examined attributions as an outcome.

72  Edward Tomlinson and Luke Langlinais

Attributions as a Predictor We now turn to the empirical research that has used a levels-of-analysis approach with attributions as the predictor variable. A brief summary of each article will be provided followed by insights into what these cross-level studies mean in terms of attribution theory and recommendations on how they may inspire future work on the attribution–trust relationship. Distinctiveness and consistency attributions of the human resource management (HRM) system at the individual level (but not consensus at the department level) predicted individual-level affective commitment. Sanders, Dorenbosch, and de Reuver (2008) developed and tested a model based on earlier conceptual work by Bowen and Ostroff (2004). Bowen and Ostroff proposed a model to conceptualize the “strength” of a human resource management (HRM) system regarding how effectively information is communicated within an organization (i.e., creating a strong situation, a situation where there is a consistent understanding and response by individuals). Bowen and Ostroff referred to Kelley’s (1967) covariation model to advance propositions on how employees use distinctiveness, consistency, and consensus information to determine HRM system strength. (Kelley’s covariation model posits that individuals make decisions regarding locus of causality based on others’ behavior and seek information, based on multiple observations, to confirm or refute their judgments. We discuss Kelley’s model more completely in the final section of the chapter.) In this manner, Bowen and Ostroff put forth a model to describe how individual-level psychological processes (i.e., attributions) emerge to form a strong (i.e., shared) organizational climate at the aggregate level. For example, these authors claimed that employees determine the distinctiveness of an HRM system by considering cues like relevance and legitimacy of authority. Sanders et al. (2008) relied on this framework and created a model where distinctiveness was operationalized by relevance (i.e., employees’ approval of HRM practices) and legitimacy of authority (i.e., employees’ evaluation of managers’ performance), consistency was operationalized as employees’ within-respondent agreement on a measure of high-commitment HRM practices, and consensus was operationalized by the level of agreement on the same measure of highcommitment HRM practices between the line manager and HR manager of each department. We have strong reservations about the theoretical and empirical application of Kelley’s (1967) theory here, and we discuss these matters more fully in the final section of the chapter. For the moment, however, we point out that this is the earliest study on attributions as a predictor that attempted to conceptualize and measure an attribution construct at a higher level. Sanders et al. measured distinctiveness and consistency at the individual level, but consensus was measured at the department level. The argument that attributions can be meaningfully studied at higher (than the individual) levels was an important step forward in this stream of

Trust and Attributions  73

research and has been subsequently adopted by several other researchers. We also believe it would be useful to investigate whether higher-level attribution measures lead to commitment due to the mediating effect of trust. Team-level stability attributions predict team perceptions of collective efficacy, while individual-level differences do not have a significant effect. Invoking Weiner’s (1986) attribution theory, Dithurbide, Sullivan, and Chow (2009) studied the relationship between team-referent attributions, team performance, and collective efficacy beliefs in competitive coed recreational volleyball teams. Their findings indicated that both objective and subjective measures of team performance significantly predict collective efficacy. Team attributions of stability were measured with the Causal Dimension Scale for Teams (CDS-T) (Greenlees, Lane, Thelwell, Holder, & Hobson, 2005) and were also a significant predictor of collective efficacy (supporting the reverse causal order suggested by Chow and Feltz, 2008, which found that both individual and aggregated collective efficacy predicted team control attributions). This relationship was qualified such that when objective performance was high, higher stability attributions were related to higher subsequent collective efficacy. In other words, “for teams with better performance, more stable team-referent attributions resulted in higher subsequent collective efficacy beliefs” (p. 502). Individual-level variables were not hypothesized, but age, sex, and experience were analyzed post hoc. None of the individual-level variables predicted collective efficacy, and level-two tests provided further evidence that variance was explained at the team level. How might this apply to the attribution–trust relationship? While trust is not used in the definition of collective efficacy, the idea of multiple individuals believing in their team’s ability to achieve goals definitely implies a certain level of trustworthiness and, hence, trust. Stability attributions were significantly related to team collective efficacy when objective performance was high. Stable attributions of team performance in conjunction with unequivocal evidence of team success (objective performance) likely enhance trust in the team, which would in turn contribute to collective efficacy beliefs. Group-level attribution style had a significant effect on individual turnover. Riolli and Sommer (2010) argued and presented evidence indicating that attributional styles can be a group-level phenomenon. Drawing from the theoretical and scale development work of Kent and Martinko (1995), Riolli and Sommer define group attributional style as, “the group’s habitual and collective manner of explaining the causes of bad and good events happening to them” (2010, p. 55). They note that attributional style is a cognitive process rather than an affective disposition, such as positive and negative affectivity. The major implication is that as members of a group interact, they collectively develop schema regarding the causes of their experiences, which results in a group-level phenomenon.The study also indicated that group attributional style influences individual turnover behavior, even after controlling for related variables such as group potency and social identity. Further, groups with optimistic attribution styles experienced significantly less turnover

74  Edward Tomlinson and Luke Langlinais

at the individual level than groups with pessimistic attribution styles (despite the fact that there were no between-group differences in actual performance). This study used the Group Attributional Style Questionnaire (Kent & Martinko, 1995). For trust research, these findings may imply that group attributional style might also impact individual-level trust (e.g., toward managers, the organization) regardless of actual objective indicators. To the extent that groups develop a chronically optimistic or pessimistic attribution style, individuals’ trust may reflect shared sense-making and have downstream effects on turnover (or other variables such as performance, etc.). Managers’ unit-level genuine and self-serving CSR causal attributions affect individuallevel genuine and self-serving CSR causal attributions; in turn, genuine attributions affect employee advocacy on behalf of the organization.Vlachos, Panagopoulos, Bachrach, and Morgeson (2017) develop a multilevel social influence theory of how corporate social responsibility (CSR) affects employees. Corporate social responsibility is “a company’s discretionary involvement in business practices that appear to further economic, societal, and environmental well-being” (Vlachos, Panagopoulos, & Rapp, 2013, p. 577). This study investigates the effect of managers’ attributions for CSR on employee attributions. Employee attributions were hypothesized to predict employee advocacy on behalf of the organization. This line of reasoning was supported for genuine attributions (i.e., CSR initiatives were regarded as sincere efforts by the organization); while managers’ self-serving attributions (i.e., that the organization was engaging in CSR for self-serving purposes) affected employees’ self-serving attributions, this effect was not moderated by manager tenure, and employee self-serving attributions did not relate to employee advocacy. Notably, this was the only study we found in this section that controlled for trust in the manager. The results offer practical implications for managers to convey their genuine motives. The authors also encourage an “inside-out” approach, where employees are seen and treated as key stakeholders in the organization. These findings may influence attribution–trust research by encouraging deeper investigation of social influence processes on an employee’s attributions, such as the influence of important others that an employee trusts. Organizations should seek out leaders that employees can trust to have genuine motives for their prosocial behavior.

Attributions as an Outcome Team-level performance affects individual-level internal attributions, but this effect is qualified by individual self-affirmation (i.e., sense of self-worth independent of team membership). This article by Sherman and Kim (2005) centers on the hedonic bias (i.e., the tendency for winners to make more internal attributions than losers) noted in a litany of prior attribution research. The self-serving bias is a specific case of the general hedonic bias and refers to the tendency of individuals to attribute their successes to internal factors and their failures to external factors (Miller &

Trust and Attributions  75

Ross, 1975). Notably, when attached to a group, a similar phenomenon has been observed, whereby individuals attribute group success to themselves personally, and group failure to more external (to the individual member) causes. However, it is also possible that individuals in groups might exhibit a group-serving bias, where the group’s successes are attributed more to the group’s internal factors than their failures. Sherman and Kim (2005) conducted two studies to examine the tendency for individuals in groups to make self- and group-serving causal attributions for success and failure. The results indicated that participants that received individually based self-affirmation exhibited significantly reduced self-serving and group-serving biases. This suggests that when individuals felt more secure about their self-image, they were less motivated to protect themselves or their team by making biased attributions for the cause of success or failure. In terms of what this study means for a levels-of-analysis perspective in attribution theory, Sherman and Kim (2005) found that team-level performance (i.e., victory versus defeat) invoked a hedonic bias unless there was a self-affirmation that ostensibly reduced the psychological need for individuals to derive a sense of self-worth from team membership.These findings reveal that judgments about the group may be anchored to one’s self-image.This cognitive process where self- and group-serving judgments are invoked can be altered through self-affirmation by removing the psychological need to protect one’s self. They conclude, “Collective events (such as the victories and defeats of one’s group) affect feelings of selfworth, and individual events (such as self-affirmation) affect judgments about one’s group” (p. 118). How might this apply to the attribution–trust relationship? The aforementioned article by Sherman and Kim (2005) can be used to inform future research questions in the attribution–trust relationship in a team setting. Sherman and Kim (2005) posited and found that self-concept anchored judgments about their group, leading to hedonic bias. That is, unless participants were affirmed by reflecting on their individually based values, both self-serving and group-serving biases moved in tandem (higher internal attributions for winning, lower internal attributions for losing). No effects were evident with respect to external attributions. However, it is possible that in some contexts, individuals on a failing team may experience a self-serving bias and place greater blame on their team (self-serving and hedonic bias might not always be in tandem); this would be expected to negatively affect team trust. One can also envision a scenario where a group-serving bias after a failure event may contribute to declining trust in an out-group.These possibilities should be explored. Aggregated individual and socio-cultural context factors influence the relationship between individual perceptions/behaviors and that individual’s causal attributions for his/ her outcome. Cherpitel et al. (2006) used data from two large-scale studies that spanned 15 countries and a 19-year period to investigate how aggregated (withinstudy) individual factors (e.g., average log volume of alcohol consumed prior to injury) and socio-cultural factors (e.g., per capita consumption of ethanol) affect

76  Edward Tomlinson and Luke Langlinais

the causal attributions of emergency room patients’ beliefs that drinking alcohol is associated with their injuries. Their results suggest that individuals’ causal attribution of injury to drinking varies by aggregated individual drinking habits as well as typical drinking patterns within their particular society. Individuals that drink alcohol at least weekly are less likely than those that drink less frequently to attribute their injuries to drinking at low volumes. Interestingly, the frequent drinkers were more likely to attribute their injuries to drinking when they consumed alcohol at higher volumes. In addition, societies that had higher integration of alcohol consumption were less likely than those with low integration to attribute injuries to drinking when drinking at low levels, but more likely to make that attribution when drinking at high levels.There was surprisingly no predictive relationship between the time between individuals consuming their last drink and the occurrence of injury in whether or not alcohol was attributed as the cause. In terms of what this study means for a levels-of-analysis perspective in attribution theory, Cherpitel et al. (2006) suggest that aggregated contextual variables within a major social category (e.g., ethnicity, region, culture, country) are associated with causal attributions. This adds further support to the notion that context can shape attributions in important ways (Hewstone, 1989). How might this apply to the attribution–trust relationship? Future research exploring the attribution–trust relationship may consider investigating different macro-contextual variables that could have an influence. For example, organizational behavior research has found that Black employees are more likely to rate their managers lower in behavioral integrity (a dimension of trustworthiness) than non-Black employees (Simons, Friedman, Liu, & McLean Parks, 2007).While this study did not incorporate a levels-of-analysis perspective or include attributions, it is possible that Blacks (relative to other racial categories) tend to experience more cynicism and suspicion (Simons et al., 2007), and this may generate a more pessimistic attribution tendency (cf. Pettigrew, 1979) that reduces trustworthiness perceptions and trust in a specific referent. In this way, a major social category in which a team or individual finds themselves may serve as a strong force, unbeknownst to them, that affects their causal attributions and subsequent trust. Individual perceptions of team success or failure as well as individual and aggregated collective efficacy beliefs were significant predictors of team attribution dimensions. Chow and Feltz (2008) invoked Bandura’s (1997) construct of collective efficacy, which captures a team’s shared belief in their ability to attain goals through their combined capabilities. This study utilized a cross-level framework based on Weiner’s (1986) attribution theory to explain the relationship between individual perceptions of success/failure and collective efficacy (both individually and at the team level) on team causal attribution dimensions. From a levels-of-analysis standpoint, the cross-level framework accounts for both individual and team effects of collective efficacy. Individual subjective perceptions of team controllability attributions and team-level perceptions of collective efficacy were related to stability attributions. This suggests that the beliefs

Trust and Attributions  77

held by individuals and the team regarding the team’s collective efficacy affect the attributions the team makes regarding their performance. Their findings also indicated that gender plays a moderating role in the collective efficacy and team controllability attribution relationship, such that teams of females demonstrated a stronger relationship between these variables compared to males. This indicates that females may have a unique perspective on their teams’ collective efficacy and attributions regarding team control. The authors surmise that this gender effect could be due to females having higher levels of emotional communication and feedback, which may strengthen their perceptions of team control. They draw from their results to recommend attribution retraining as a potentially promising intervention: Attribution retraining focuses primarily on altering failure attributions that produce maladaptive behaviors, thoughts, and emotions. Following a failure experience, individuals are taught to attribute the cause of performance to controllable factors rather than internal ability-oriented factors…. Thus, applied techniques directed at enhancing the collective efficacy beliefs of teams (e.g. modelling, verbal persuasion, team goal setting) may indirectly influence team control attributions. (Chow & Feltz, 2008, p. 1187) How might this apply to the attribution–trust relationship? The concept of attribution retraining may have significant application in trust research, specifically in the trust repair literature. For example, Hatzakis (2009) suggests that some individuals may have an attributional style reflecting a tendency to attribute the causes of negative outcomes to the shortcomings of others. This may lead to errors in judgment due to inaccurate perceptions of the extent of influence and control others have over specific outcomes. If this process generalizes to teams, and teams can be retrained in the process of making attributions, this may affect the level of trust that team members are willing to extend to their team. For instance, if a trust violation is regarded as having an unstable cause (i.e., unlikely to recur), versus stable, a team member may be more willing to place trust in the team again (Tomlinson & Mayer, 2009). If organizations were to train team members in the process of making accurate attributions (as opposed to unduly pessimistic attributions), this may enhance their trust judgments. We note, however, that attribution retraining research has focused on how dysfunctional attribution patterns at the intrapersonal level can become more constructive.We are unaware of any research in this area at other levels of analysis. Unit-level perceptions of charismatic leadership affect individual-level (internal versus external) attributions, which impact job satisfaction. In their study on corporate social responsibility (CSR), Vlachos et al. (2013) examine how aggregated perceptions of charismatic leadership affect attributions that employees make about their organization’s motives for CSR activities, which in turn affect employee

78  Edward Tomlinson and Luke Langlinais

job satisfaction. They essentially argue that middle managers are the “face” of the company to employees, such that how managers behave reflects the characteristics and motives of the organization itself. Intrinsic attributions capture the belief that CSR activity is values-driven (i.e., out of genuine concern), while extrinsic attributions capture the belief that engagement in CSR is simply for promotional purposes. Vlachos et al. (2013) used a sample of manufacturing employees and collected questionnaire data. They ultimately found that charismatic leadership in middle managers positively affects intrinsic attributions employees make regarding the organization’s CSR activities, and in turn, these attributions are positively related to their job satisfaction. While this study focuses on charismatic leadership as the interpretive lens through which employees view their organization’s CSR activities, future research should consider trust in management. Charismatic leaders “stress self-sacrifice for the long-term good of the organization and/or the larger community” (Vlachos et al., 2013, p. 579). This bears a striking resemblance to perceptions of benevolence that individuals form when choosing to trust another individual. Future research on the attribution–trust nexus might examine the effect of unit-level perceptions of manager trustworthiness on employees’ attributions regarding organizational strategic activities. Further, parallel to Vlachos et al.’s recommendation to train managers to behave in a more charismatic fashion, training managers to exhibit more trustworthy behaviors may reap similar dividends. While we are not the first to suggest this implication, we are unaware of empirical research that actually tests this supposition. The unit-level effect of high-performance work systems influences individual-level human resource attributions. High-performance work systems (HPWS) refers to a collection of human resource management practices specifically geared toward maximizing employee and organizational performance.Van De Voorde and Beijer (2015) used multilevel data to examine the effect of HPWS (at the work unit level) on individual-level human resource attributions (i.e., that HR practices are intended to enhance employee well-being or performance, respectively), ultimately leading to organizational commitment and job strain. The sample for this study consisted of line managers and their employees. Line managers reported on the extent to which HPWS practices were applied to their work unit employees, while employees responded to questions measuring their attributions, commitment, and job strain. The results indicated that work units with more employees covered by HPWS practices were positively associated with individual employees attributing the organization’s motives to increasing both employee well-being and performance. Well-being attributions, in turn, were related to higher organizational commitment and lower job strain; performance attributions were related to higher job strain. Future attribution–trust research might investigate whether trust mediates the relationship between attributions regarding HR programs and employee outcomes (such as organizational commitment and job strain).

Trust and Attributions  79

Unit-level understanding of high-commitment human resource management leads to individual-level attributions and outcomes. Topically, this study bears a striking resemblance to the Van De Voorde and Beijer (2015) study reviewed above. Sanders and Yang (2016) investigated whether unit-level high-commitment human resource management (HC-HRM; the European terminology for HPWS) is more effective when employees attribute it to the organization’s management (i.e., that HC-HRM practices are a strong signal of their management’s intent). In terms of Kelley’s (1967) covariation principle of attribution theory, attributing HC-HRM practices to managers should occur when distinctiveness, consistency, and consensus are high (cf. Sanders et al., 2008). Their results revealed that when unitlevel HC-HRM practices are attributed as high in distinctiveness, consistency, and consensus, they are associated with higher employee affective commitment and innovative behavior. As we mentioned earlier, we have strong reservations about this application of covariation theory and will elaborate on this matter in the final section. Nonetheless, future attribution–trust research might investigate how trust in relation to the communication of organizational systems may influence attributions made about those systems and management’s motivations. HC-HRM systems seemingly function on a reciprocal trust-based relationship between the organization’s management and its employees. Indeed, Sanders and Yang (2016) describe such systems in terms of a social exchange perspective and assert that “employees perceive HC-HRM as benevolence on the part of their employer” (p. 205). If an organization focuses on the development and maintenance of trust, it may serve as a conduit to better communication and more beneficial attributions.

Potential Contributions of Attribution Theories in Cross- and Multilevel Research on Trust In this final section, we offer some integrative analysis and describe several important observations that we believe will be central to examining the attribution– trust relationship in cross- and multilevel research.

Theoretical Considerations Theoretical Perspectives Sherman and Kim (2005) did not invoke any particular attribution theory, while Riolli and Sommer (2010) cited several major attribution theories (Heider, 1958; Green & Mitchell, 1979; Kelley, 1967; Weiner,1986). Group attributional style (Kent & Martinko, 1995) was incorporated by Riolli and Sommer (2010). The primary attribution theories invoked in the cross- and multilevel studies on attributions were Weiner’s attribution theory (Chow & Feltz, 2008; Dithurbide et al., 2009;Vlachos et al., 2017) and Kelley’s covariation theory of attributions (Sanders

80  Edward Tomlinson and Luke Langlinais

et al., 2008; Sanders & Yang, 2016; Van De Voorde & Beijer, 2015; Vlachos et al., 2013, 2017).

Theoretical Concerns As we mentioned earlier, we have serious concerns regarding how Kelley’s (1967) covariation theory has sometimes been invoked and applied in this stream of research. Kelley’s covariation theory was developed to explain how individuals rely on social information collected over time to test and confirm their causal attributions with respect to the locus of causality.Taking the example,“John laughs at the comedian,” Hewstone (1989) explains “this outcome could be caused by something in the person (John), the circumstances (e.g., the occasion in which the outcome occurred), the entity or stimulus (the comedian), or some combination of these factors” (p. 22). Kelley posited that observers would assess accumulated information regarding consensus, consistency, and distinctiveness to determine which source of the cause is most likely. In Kelley’s theory, consensus refers to how the effect varies across persons; consistency refers to how consistent the effect is over time and modality; and distinctiveness refers to how the stimulus is distinct from other stimuli. For ease of exposition, each of these factors is described as being high or low, and various profiles point to specific loci of causality. For example, an entity locus is indicated by a high degree of all three factors. Bowen and Ostroff (2004) referred to this attribution theory when they developed propositions specifying that an HRM system high in distinctiveness, consistency, and consensus should enhance clarity of interpretation in the setting, thereby allowing for similar ‘cognitive maps’ or ‘causal maps’ to develop among people, as well as to create an ‘influence situation’ whereby individuals yield to the message and understand the appropriate ways of behaving. (p. 214, emphasis original) Their paper proposed various characteristics of HRM that they viewed as illustrating each of Kelley’s factors. For example, they argued that distinctiveness is captured by information such as legitimacy of authority (presence of organizational cues indicating that the HRM function has high status and credibility) and relevance (the HRM function aligns individual employee goals with the organization’s strategic goals). They argued that consistency is captured in this context by information such as the instrumentality of the HRM system (e.g., desired behaviors are consistently rewarded) and the validity of the HRM system (i.e., there is consistency between what the HR system purports to do and what it actually accomplishes). They argue that consensus is indicated by information regarding the level of agreement among principal HRM decision-makers and employees’ fairness perceptions.

Trust and Attributions  81

While we do not necessarily object to the specific cues Bowen and Ostroff outlined in their effort to articulate what is meant by and contributes to a “strong” HRM system, our reading of Kelley’s covariation theory indicates that it is potentially being misapplied here. First, covariation theory is about determining the locus of causality for an effect. In this context, they argue that the locus is the entity (HRM system), but what is the effect? From our reading, we presume they are modeling how employees attribute the locus of causality for employee work-related behaviors within the organization, but this question is never explicitly addressed. Stated differently, we believe it is imperative to specify whose behavior and what effect is in view to apply this attribution theory. Similarly, we are not convinced that some of their ways of operationalizing Kelley’s factors are faithful representations of the respective conceptualizations. For example, Bowen and Ostroff never explained their distinctiveness indicators in terms of how one organization’s HRM system compares to other organizations’ HRM systems. Distinctiveness is “the impression … attributed to the thing if it uniquely occurs when the thing is present and does not occur in its absence” (Kelley, 1967, p. 197). In short, while we can accept that their model points to information that social perceivers would attend to in drawing a conclusion regarding HRM system strength, we do not see their conceptual work (or the subsequent empirical work by Sanders and Yang [2016] and Sanders et al. [2008]) as definitively illustrating a contribution grounded in attributions. Attributions are a very specific and narrow type of social perception. It has been noted elsewhere that covariation theory is occasionally invoked incorrectly (Tomlinson, 2018), and we reiterate that concern here.

Theoretical Contributions On a more positive note, we would like to highlight what we saw as a very noteworthy theoretical contribution by Vlachos et al. (2017). Their multilevel model of how CSR affects employees investigated the effect of managers’ attributions for CSR on employee attributions. We regard this as a very significant theoretical contribution. Prior to this study, extant research had not explored how the attributions of other observers create social information that is used by focal observers evaluating actors. Due to causal ambiguity inherent in CSR motives, managers should be deemed an informative source when employees arrive at their attributional conclusions, and this effect should be more pronounced for managers with longer tenure (because they have more experience with the organization that informs their own attributions). Moreover, employee attributions were hypothesized to predict employee advocacy on behalf of the organization (another significant theoretical contribution, speaking to the extent to which employees would engage in important boundaryspanning behavior, as opposed to the normal outcome variables of satisfaction, commitment, etc.).

82  Edward Tomlinson and Luke Langlinais

Empirical and Methodological Considerations Chow and Feltz (2008) raised an important methodological issue when it comes to generating aggregate measures of attribution dimensions in light of their low consensus estimates for these variables. In their study, team attribution dimensions were measured with the Revised Causal Dimension Scale (McAuley, Duncan, & Russell, 1992), a commonly used scale. The respondent is asked to provide an attribution for the outcome under study and then respond to subsequent semantic differential items indicating the extent to which this attributed cause is internal, controllable, and stable. (A similar approach is used in the Causal Dimension Scale for Teams, validated using interactive sports teams, such as soccer.) The practical challenge here is that many outcomes have multiple causes, and researchers typically do not analyze the ascribed causes directly to ensure respondents are listing the same cause before providing ratings on attribution dimensions. Even on the same team, where the outcome is the same, team members may reach different attributions for the cause of that outcome, and as a result, they have different attributions in mind when responding to measures of the attribution dimensions. This measurement approach poses a challenge insofar as it is “unable to accurately detect whether team attributions represent an emergent group perception or merely an individual’s perception about the team” (p. 1187). As it has been cogently noted elsewhere, once we move beyond the individual level with a given construct, we must demonstrate a requisite degree of consensus, or shared perception of the construct, at that level (Fulmer, 2018). In other words, assessing aggregation statistics on causal attribution dimensions alone might not go far enough to establish shared perceptions on attributed causes. Even if we just consider the locus of causality, Hewstone (1989) points out that “the categories of internal and external causality are very broad, containing a heterogeneous collection of attributions” (p. 31). Measurement strategies like the Revised Causal Dimension Scale are designed to avoid the “fundamental attribution research error” (Russell, McAuley, & Tarico, 1987); researchers assume that certain causal ascriptions will be universally classified into certain causal attribution dimensions even though the individuals they study sometimes reach different conclusions (e.g., certain forms of ability might be interpreted as unstable instead of stable, etc.). Therefore, the Revised Causal Dimension Scale and similar measures allow researchers to “fast-forward” to the causal dimension profile to see how respondents interpret ascribed causes, regardless of what those causes are. But in a social context where we are interested in examining how attributions emerge at higher levels, it becomes more imperative for researchers to gain assurance that the ascribed cause itself, as well as the dimensional profile, is shared among perceivers at a given level. Certain content analytical tools can be used to assess shared perceptions of ascribed causes among members of a social unit (Hewstone, 1989).

Trust and Attributions  83

Finally, we note that several researchers included in our review created contextspecific attribution measures (e.g., Vlachos et al., 2017) and were not as interested in causal dimensions per se. Researchers that are interested in aggregating attribution dimensions should take great care to ensure the validity of their methods and measures. Perhaps offering and subsequently evaluating an open-text response on the perceived cause of the outcome of interest would allow for subsequent content analysis and tests of inter-rater agreement. Another suggestion would be to specifically prime the participants to the cause of interest in the study to ensure that the responses to the causal dimensions are made regarding the same cause. However, the identification of the mechanisms leading to these variations in the perceptions of the cause of the outcome may itself be a worthy direction for future research.

Conclusion While individuals engage in psychological activities that lead to judgments (like attributions) and states (like trust), these processes are better understood as social psychological. Stated differently, individuals interacting with others do not arrive at their judgments or states in a vacuum. Shared social contexts influence these judgments and states in important ways, and given the increasing reliance on various forms of interdependence in society (teamwork, cross-cultural interaction), we believe cross- and multilevel research can shed new insight on these important phenomena. We offer this review of how attribution theories – outside of the trust literature – have recently expanded their focus to include multilevel research, and we hope this work inspires other researchers exploring the nexus between attributions and trust.

References Bandura, A. (1997). Self-efficacy:The exercise of control. New York: Freeman. Bowen, D. E., & Ostroff, C. (2004). Understanding HRM-firm performance linkages: The role of the “strength” of the HRM system. Academy of Management Review, 29, 203–221. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​04​​.12​​73607​6 Cherpitel, C. J., Bond, J.,Ye,Y., Borges, G., Room, R., Poznyak,V., & Hao,W. (2006). Multilevel analysis of causal attribution of injury to alcohol and modifying effects: Data from two international emergency room projects. Drug and Alcohol Dependence, 82, 258–268. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.drug​​alcde​​p​.2​00​​5​.10.​​002 Chow, G. M., & Feltz, D. L. (2008). Exploring the relationships between collective efficacy, perceptions of success, and team attributions. Journal of Sports Sciences, 26, 1179–1189. https​:/​/do​​i​.org​​/10​.1​​080​/0​​26404​​108​02​​10182​7 De Jong, B. A., & Dirks, K. T. (2012). Beyond shared perceptions of trust and monitoring in teams: Implications of asymmetry and dissensus. Journal of Applied Psychology, 97, 391–406. https://doi​.org​/10​.1037​/a0026483 Deutsch, M. (1958). Trust and suspicion. Journal of Conflict Resolution, 2, 265–279. https​:/​/ do​​i​.org​​/10​.1​​177​/0​​02200​​2758​0​​02004​​01 Deutsch, M., & Krauss, R. M. (1965). Theories in social psychology. New York: Basic Books, Inc.

84  Edward Tomlinson and Luke Langlinais

Dithurbide, L., Sullivan, P., & Chow, G. (2009). Examining the influence of team-referent causal attributions and team performance on collective efficacy: A multilevel analysis. Small Group Research, 40, 491–507. https://doi​.org​/10​.1177​/1046496409340328 Ferrin, D. L., Bligh, M. C., & Kohles, J. C. (2015). The actor-partner interdependence model: A model of studying trust in dyadic relationships. In F. Lyon, G. Mollering, & M. N. K. Saunders (Eds.), Handbook of research methods on trust (2nd ed., pp. 240–250). Cheltenham: Edward Elgar. https​:/​/do​​i​.org​​/10​.4​​337​/9​​78178​​25474​​​19​.00​​032 Ferrin, D. L., Kim, P. H., Cooper, C. D., & Dirks, K. T. (2007). Silence speaks volumes: The effectiveness of reticence in comparison to apology and denial for repairing integrityand competence-based trust violations. Journal of Applied Psychology, 92, 893–908. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​.4​.8​​93 Fiske, S.T., & Taylor, S. E. (2013). Social cognition: From brains to culture (2nd ed.). Los Angeles: SAGE. https://doi​.org​/10​.4135​/9781446286395 Fulmer, C. A. (2018). Multilevel trust: Antecedents and outcomes of trust at different levels. In R. H. Searle, A. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 143–160). London: Routledge. https://doi​.org​/10​.4324​/9781315745572​-10 Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38, 1167–1230. https://doi​.org​/ 10​.1177​/0149206312439327 Gilbert, D. T., Pelham, B. W., & Krull, D. S. (1988). On cognitive busyness: When person perceivers meet persons perceived. Journal of Personality and Social Psychology, 54, 733– 740. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.5​​4​.5​.7​​33 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34, 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 Green, S. G., & Mitchell, T. R. (1979). Attributional processes of leaders in leader-member interactions. Organizational Behavior and Human Performance, 23, 429–458. https​:/​/do​​i​. org​​/10​.1​​016​/0​​​030​-5​​073(7​9)900​08-4 Greenlees, I. A., Lane, A., Thelwell, R., Holder, T., & Hobson, G. (2005). Team-referent attributions among sport performers. Research Quarterly for Exercise and Sport, 76, 477– 487. https​:/​/do​​i​.org​​/10​.1​​080​/0​​27013​​67​.20​​0​5​.10​​59932​1 Hatzakis, T. (2009). Towards a framework of trust attribution styles. British Journal of Management, 20, 448–460. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-8551​​.200​8​​.0059​​6.x Heider, F. (1958). The psychology of interpersonal relations. New York: John Wiley. https://doi​. org​/10​.1037​/10628​-000 Hewstone, M. (1989). Causal attribution: From cognitive processes to collective beliefs. Malden, MA: Blackwell Publishing, Inc. Janowicz-Panjaitan, M., & Krishnan, R. (2009). Measures for dealing with competence and integrity violations of interorganizational trust at the corporate and operating levels of organizational hierarchy. Journal of Management Studies, 46, 245–268. https​:/​/do​​i​.org​​/10​. 1​​111​/j​​.1467​​-6486​​.200​8​​.0079​​8.x Kelley, H. H. (1967). Attribution theory in social psychology. In D. Levine (Ed.), Nebraska Symposium on Motivation 1967 (pp. 192–238). Lincoln: University of Nebraska Press. Kent, R., & Martinko, M. J. (1995). The development and evaluation of a scale to measure organizational attributional style. In M. J. Martinko (Ed.), Attribution theory: An organizational perspective (pp. 53–75). Delray Beach, FL: St. Lucie Press. https://doi​.org​/ 10​.4324​/9781315137926-5 Kim, P. H., Dirks, K. T., Cooper, C. D., & Ferrin, D. L. (2006). When more blame is better than less:The implications of internal vs. external attributions for the repair of trust after

Trust and Attributions  85

a competence- v. integrity-based trust violation. Organizational Behavior and Human Decision Processes, 99, 49–65. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​0​​5​.07.​​002 Kim, P. H., Ferrin, D. L., Cooper, C. D., & Dirks, K. T. (2004). Removing the shadow of suspicion: The effects of apology versus denial for repairing competence- versus integrity-based trust violations. Journal of Applied Psychology, 89, 104–118. https​:/​/do​​i​. org​​/10​.1​​037​/0​​021​-9​​010​​.8​​9​.1​.1​​04 Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions Annual Review of Psychology, 50, 569–598. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​. psy​​c​h​.50​​.1​.56​9 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20, 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​95​.​ 95​​08080​​335 McAuley, E., Duncan, T. E., & Russell, D. W. (1992). Measuring causal attributions: The Revised Causal Dimension Scale (CDSII). Personality and Social Psychology Bulletin, 18, 566–573. https://doi​.org​/10​.1177​/0146167292185006 Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality: Fact or fiction? Psychological Bulletin, 82, 213–225. https://doi​.org​/10​.1037​/h0076486 Pettigrew,T. F. (1979).The ultimate attribution error: Extending Allport’s cognitive analysis of prejudice. Personality and Social Psychology Bulletin, 5, 461–476. https​:/​/do​​i​.org​​/10​.1​​177​/ 0​​14616​​7279​0​​05004​​07 Riolli, L., & Sommer, S. M. (2010). Group attributional style: A predictor of individual turnover behavior in a manufacturing setting. Journal of Business and Management, 16, 51–73. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23, 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Russell, D. W., McAuley, E., & Tarico, V. (1987). Measuring causal attributions for success and failure: A comparison of methodologies for assessing causal dimensions. Journal of Personality and Social Psychology, 52, 1248–1257. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​. 5​​2​.6​.1​​248 Sanders, K., Dorenbosch, L., & de Reuver, R. (2008). The impact of individual and shared employee perceptions of HRM on affective commitment: Considering climate strength. Personnel Review, 37, 412–425. https​:/​/do​​i​.org​​/10​.1​​108​/0​​04834​​808​10​​87758​9 Sanders, K., & Yang, H. (2016). The HRM process approach: The influence of employees' attribution to explain the HRM-performance relationship. Human Resource Management, 55, 201–217. https://doi​.org​/10​.1002​/hrm​.21661 Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007).An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32, 344–354. doi:10.5465/ AMR.2007.24348410 Sherman, D. K., & Kim, H. S. (2005). Is there an “I” in “Team”? The role of the self in groupserving judgments. Journal of Personality and Social Psychology, 88, 108–120. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.8​​8​.1​.1​​08 Simons, T., Friedman, R., Liu, L. A., & McLean Parks, J. (2007). Racial differences in sensitivity to behavioral integrity: Attitudinal consequences, in-group effects, and “trickle down” among black and non-black employees Journal of Applied Psychology, 92, 650–665. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​.3​.6​​50 Strickland, L. H. (1958). Surveillance and trust. Journal of Personality, 26, 200–215. https​:/​/ do​​i​.org​​/10​.1​​111​/j​​.1467​​-6494​​.1958​​​.tb01​​580.x​

86  Edward Tomlinson and Luke Langlinais

Tomlinson, E. C. (2018). The contributions of attribution theories to trust research. In R. H. Searle, A. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 245–266). London: Routledge. https://doi​.org​/10​.4324​/9781315745572​-17 Tomlinson, E. C., & Mayer, R. C. (2009). The role of causal attribution dimensions in trust repair. Academy of Management Review, 34, 85–104. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​. 35​​71329​1 Van De Voorde, K., & Beijer, S. (2015). The role of employee HR attributions in the relationship between high-performance work systems and employee outcomes. Human Resource Management Journal, 25, 62–78. https://doi​.org​/10​.1111​/1748​-8583​.12062 Vlachos, P. A., Panagopoulos, N. G., Bachrach, D. G., & Morgeson, F. P. (2017). The effects of managerial and employee attributions for corporate social responsibility initiatives. Journal of Organizational Behavior, 38, 1111–1129. https://doi​.org​/10​.1002​/job​.2189 Vlachos, P. A., Panagopoulos, N. G., & Rapp, A. A. (2013). Feeling good by doing good: Employee CSR-induced attributions, job satisfaction, and the role of charismatic leadership. Journal of Business Ethics, 118, 577–588. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10551​​-01​ 2-​​1590-​1 Weiner, B. (1986). An attributional model of motivation and emotion. New York: SpringerVerlag. https​:/​/do​​i​.org​​/10​.1​​007​/9​​78​-1-​​46​12-​​4948-​1

5 CASCADING INFLUENCES AND CONTEXTUALIZED EFFECTS A Model of Multilevel Control–Trust Dynamics Chris Long

Introduction Control–trust dynamics are important, fundamental facets of organizations (Bachmann, 2001; Cao & Lumineau, 2015; Long & Sitkin, 2018). Controls comprise a variety of communication, monitoring, and evaluative mechanisms that are used to direct individuals to achieve various organizational objectives (Dekker, 2004; Fayol, 1949; Ouchi, 1979). Trust describes one’s willingness to be vulnerable to another based on positive expectations of their intentions and behaviors (Rousseau, Sitkin, Burt, & Camerer, 1998). Control–trust dynamics have attracted significant attention from scholars for two reasons. First, because when appropriate levels of control and trust exist in organizations, actors tend to exhibit increased cooperation, commitment, performance, and motivation (Bijlsma-Frankema & Costa, 2005; Mayer, Davis, & Schoorman, 1995; Sitkin, 1995). Second, because control and trust build upon distinct psychological mechanisms that are often opposing or conflicting, it can be difficult to effectively manage control–trust dynamics in ways that facilitate the achievement of these outcomes (Anderson, Christ, Dekker, & Sedatole, 2013; Coletti, Sedatole, & Towry, 2005; Long, 2018). This leads control and trust to sometimes exist as mutually reinforcing, synergistic mechanisms, sometimes as opposing, antithetical mechanisms, and at other times, as orthogonal or seemingly unrelated mechanisms (Emsley & Kidon, 2007; Long & Sitkin, 2006; Poppo & Zenger, 2002). These opportunities and complexities have increased interest in control–trust dynamics to grow and foster some of the most active debates in organizational research (Long & Sitkin, 2018; Searle, Nienhaber, & Sitkin, 2018). One area where DOI: 10.4324/9780429449185-5

88 

Chris Long

scholars are beginning to focus more attention is on multilevel theoretical modeling and empirical testing (Currall & Inkpen, 2002;Weibel et al., 2015). Multilevel research contrasts with the vast majority of existing control–trust research that has focused on examining these dynamics at only one analytical level and previous multilevel work that has primarily focused on examining only organizational-level outcomes (Das & Teng, 2001; Schilke & Cook, 2013; Zaheer, McEvily, & Perrone, 1998). Concurrently evaluating multiple analytical levels increases our capacity to understand how employee trust, cooperation, and performance are impacted by relationships between individuals, organizations, and other institutional factors. This chapter aims to advance our understanding of these relationships by charting the complex conceptual landscape surrounding intra-organizational control–trust dynamics. It specifically examines how the trust that individual employees have in their managers, their organization, and the authority systems that surround these relationships are developed through dynamics that occur between actors at multiple levels of the organization. In detailing both general and specific aspects of these relationships, this chapter seeks to achieve four major objectives. First, it organizes and examines existing multilevel research on control–trust dynamics that informs employee trust in multiple referents. Second, it identifies conceptual opportunities to solve theoretical dilemmas and increase our understanding of how employee trust, cooperation, and performance are impacted by the interactions between managers, organizations, and their incumbent authority systems. Third, it presents a conceptual map of the overall theoretical landscape encompassing multilevel control–trust dynamics. Fourth, it presents an agenda for future research and identifies how examining these issues can increase our understanding of overall control–trust dynamics as well as employee trust development and performance. To provide a comprehensive picture and detail deficits in our knowledge of multilevel influences within organizations, the chapter first reviews the relevant literature on control–trust dynamics. After specifying key parameters, it introduces a theoretical framework that outlines the conceptual landscape of these dynamics between employees, managers, senior organizational leaders, and the systems that evolve from relationships between these entities. This chapter then discusses how the framework presented in this chapter provides scholars with opportunities to deepen and enrich their understanding of employee trust-development, cooperation, and related factors that influence performance.

Theory In an effort to establish a common baseline of understanding, the text below presents some key control and trust concepts in advance of evaluating primary elements of control–trust dynamics.

Multilevel Control–Trust Dynamics  89

Control Concepts Organizational control comprises a fundamental factor in all organizations (Merchant, 1985; Ouchi, 1979, 1980). Organizational controls are generally exercised within management control systems comprised of mechanisms that authorities use to guide their subordinates and other exchange partners (e.g., peers) to bring their goals, actions, and capabilities into line with those of the organizations that they represent (Long, Burton, & Cardinal, 2002; Merchant, 1985). Managers employ control mechanisms to inform their subordinates about desired performance standards, to provide those subordinates with resources that they can use in pursuing those standards, to conduct evaluations of their performances, and to reinforce (i.e., reward or punish) subordinates according to the levels of performance they achieve (Bradach & Eccles, 1989; Costa & Bijlsma-Frankema, 2007; Inkpen & Currall, 2004). Forms of control are differentiated according to their attributes or functions. While construct-labeling varies (Cardinal, Sitkin, & Long, 2010; Cardinal, Sitkin, Long, & Miller, 2018), output, process, and input mechanisms describe forms of control that are commonly used in control–trust research. These controls are generally applied directly by principals to agents and are differentiated according to the phase of the production process to which they are targeted (Cardinal, Sitkin, & Long, 2004; Cardinal et al., 2010). For example, output controls (i.e., outcome or results controls) are designed and deployed to impact product outputs in ways that help ensure that actors achieve end-state results and required outcome standards (Mintzberg, 1979; Snell, 1992). Process controls (i.e., behavior or activity controls) are used by authorities to motivate subordinates to perform their work tasks using particular behaviors, processes, and procedures (Ouchi, 1977). Input controls are generally applied at the beginning of the production process through selection mechanisms or approaches to training and socialization that prepare human and material resources for production efforts. As a result, input control mechanisms help subordinates gain important competencies or encourage employees to adopt particular goals, values, or identities (Van Maanen & Schein, 1979; Wanous, 1980).

Trust Concepts Trust is a dynamic concept that describes how individuals experience a psychological state where they “accept vulnerability based upon positive expectations of the intentions or behavior of another” (Rousseau et al., 1998, p. 395).1 Trust is

1 The terms ‘trustee’ and ‘trustor’ are used in this chapter to denote two individuals who enact control– trust dynamics. A trustee (usually a ‘controller’ or ‘superior’) identifies the principal in the relationship who is seeking cooperation from another individual to achieve a particular performance goal. The trustor (usually a ‘controlee’ or ‘subordinate’) is the agent in the relationship (Eisenhardt, 1989). This individual is deciding whether, how, and how much they will cooperate with the trustee in

90 

Chris Long

formed through attributional processes that individuals use to assess trustees’ words and deeds. Trustors’ attributions can be developed from instrumental, relational, or values-based assessments of trustees’ attributes, beliefs, and behaviors (Fulmer & Gelfand, 2012). Because authorities are often doing most of the control-based efforts in organizations, control–trust research tends to focus on assessments by subordinates of their managers, leaders, or other key decision makers (Connell, Ferres, & Travaglione, 2003; Long & Sitkin, 2018). However, an important note for multilevel work is that researchers have conducted initial examinations into the extent to which authorities trust their subordinates (Knoll & Gill, 2011; Ladegard & Gjerde, 2014; Werbel & Lopes Henriquez, 2009). The character of trust differs somewhat according to the referent at which that trust is directed and the type of information trustors use in their evaluations (Bhattacharya, Devinney, & Pillutla, 1998; Searle et al., 2018). The term “trust” generally refers to a holistic assessment of a trustee that is developed through more granular evaluations of trustees on various dimensions of trustworthiness. In control–trust research, scholars often distinguish between competence-based trust and goodwill-based trust (Hernandez, Long, & Sitkin, 2014; Sako, 1992). Competence-based trust describes a determination that trustees possess the capacity or ability to protect and promote a trustor’s values and interests (Das & Teng, 1998, 2001). Alternatively, goodwill-based trust develops as individuals observe a trustee’s willingness to actively protect and promote their values and interests through demonstrations of benevolence and integrity (Das & Teng, 2001; Long & Sitkin, 2018). Positive assessments of a trustee’s goodwill generally rely on evaluations of their benevolence and integrity. Benevolence describes a trustee’s willingness to accommodate a trustor’s specific needs and interests in their decisions and actions. Integrity describes beliefs that trustees uphold important values and reliably fulfill promises and obligations (Mayer et al., 1995). Goodwill and competence-based trust evaluations are relevant at various levels of analysis, depending on how trustors target their assessments. At the interpersonal level, trust is developed through evaluations of how the trustee treats them and other referents they observe. At the organizational or institutional level, subordinates examine senior organizational leaders’ communications, decisions, and actions to assess whether they will be able to promote and protect their personal and professional interests within the cultures and systems those leaders

working to achieve those performance objectives. It is important to note, however, that while the label ‘trustee’ generally connotes someone in a position of authority, this may or may not be the case. Observing that the basic parameters of control–trust relationships have been applied within various contexts and levels of analyses, ‘trustee–trustor’ control–trust dynamics that are described here are also present in peer relationships where one peer seeks to influence the activities of another peer (Korsgaard, Schweiger, & Sapienza, 1995). Similarly, this can characterize relationships between organizations where one firm is attempting to motivate the decisions and actions of another firm (Gulati & Nickerson, 2008; Mellewigt, Madhok, & Weibel, 2007; Sydow, 1998).

Multilevel Control–Trust Dynamics  91

foster (Gillespie & Dietz, 2009; Rousseau et al., 1998; Zucker, 1986). In general, employees tend to trust organizations where leaders foster enabling cultures and supportive employment practices that promote collaboration, transparency, and job security (Gillespie & Dietz, 2009; Malhotra & Lumineau, 2011).

Control–Trust Dynamics A fundamental and omnipresent dilemma faced by any individual in a position of authority is how to direct their subordinates in ways that engender their trust while effectively motivating them to achieve key performance objectives (Long, 2010; Strickland, 1958). As research has shown, when managers implement controls in ways that highlight coordination over subjugation, foster employee development, and clearly link the achievement of important outcomes with the provision of valued rewards, employee trust can increase even as control applications increase (Long & Sitkin, 2006; Stahl, Larsson, Kremershof, & Sitkin, 2011). Managers who are able to gain their employees’ trust while applying controls empower those subordinates to achieve relevant performance standards as they increase the overall quality of their superior–subordinate relationships. When applying controls, employees’ trust in their managers increases as they gain confidence that their superiors are working actively to protect and promote their values and interests (Weber, Malhotra, & Murnighan, 2004). It is also important to acknowledge that increasing amounts of research now report how managers (i.e., trustees) are conscious of the need to develop trusting relationships with their subordinates. For example, Whitener et al.’s work on trustworthiness-promotion (1998) suggests that managers are aware of the benefits of enhancing subordinates’ perceptions of their trustworthiness. Khodyakov’s (2007) study of Orpheus, a conductorless orchestra, shows how this ensemble’s musicians consciously work to balance musical control and performance concerns with their desires to engender trust and confidence in their motivations and abilities. A recent study by Long (2018, see also 2002, 2010) integrates qualitative and quantitative analyses to detail how managers systematically integrate the efforts they make to apply controls and demonstrate their trustworthiness in their ongoing attempts to elicit specific types of employee cooperation. It is important to note that within any superior–subordinate relationship, the development of trust in a work environment is often a dynamic and tenuous process.This is because, while subordinates look to controls for direction and motivation in doing work, complying with controls generally requires them to concede aspects of their individual discretion and personal control (Deci & Ryan, 1985; Ouchi, 1979). As individuals follow the directives of authorities, their strong drive for self-determination motivates them to engage in attributional processes to assess whether complying with a particular set of controls is a ‘good bet’ for them. Generally, compliance will enable subordinates to achieve their desired goals if a particular superior (who is applying those controls) can be relied upon to protect

92 

Chris Long

and promote that subordinate’s personal needs and interests (Deci & Ryan, 1985; Ferrin, Bligh, & Kohles, 2007). Subordinates’ positive assessments of superiors on this dimension increase subordinate confidence that their superior can be trusted. This, in turn, increases a subordinate’s willingness to cooperate with that superior in doing work. Conversely, a negative assessment of a particular superior’s intentions decreases subordinate trust and confidence in that actor, which lessens that particular subordinate’s willingness to fully comply and cooperate with that superior in doing work. Because more stringent organizational controls require greater concessions in personal agency and self-determination from subordinates, managers face an ongoing paradoxical dilemma. If managers exert excessive amounts of control, they may be viewed as acting inappropriately. In particular, they may compromise their employees’ sense of self-determination and, thus, fail to assure their subordinates they can be trusted to protect their interests (Blau, 1964; Fox, 1974; Shapiro, 1987). Conversely, if managers focus too much on accommodating their subordinates’ desires for selfdetermination and act to minimize their subordinates’ concessions of personal control, they may fail to sufficiently direct their subordinates’ work activities and guide them toward accomplishing desired performance objectives (Dekker, 2004; Spreitzer & Mishra, 1999).This latter situation can erode trust in an authority figure by making subordinates feel like their sense of self-efficacy, competence, and individual agency is diminished under that manager’s leadership (Long & Sitkin, 2018). Researchers agree that for superiors to achieve the level of cooperation they seek from subordinates, they must apply controls in ways that subordinates will view as trustworthy. Scholars have outlined several ways that this can be accomplished. First, managers need to choose enabling forms of control (Adler & Borys, 1996). On this point, Weibel (2007) outlines that managers tend to foster higher levels of trust when they signal to their subordinates that they appreciate them, will not overly restrict their autonomy, and will actively work to protect their interests (Sitkin & George, 2005). Scholars suggest that managers who subject their subordinates to lower levels of explicit monitoring make them feel less scrutinized and accommodate their desires for self-determination. These attributes are often fixtures of control systems that codify participatively set performance targets, supply subordinates with sufficient training and resources, and provide subordinates with choices about how they will attempt to accomplish those objectives. Research suggests that managers who apply controls with these characteristics are able to forge more positive, trusting relationships with their subordinates, who see their managers as protecting their sense of self-determination as well as promoting laudable goals, norms, and values (Inkpen & Currall, 2004; Malhotra & Murnighan, 2002). Conversely, managers can compromise trust when they specifically dictate and closely monitor their subordinates’ task efforts in ways that signal their interests in restricting their subordinates’ autonomy (Hartmann & Slapničar, 2009). In addition to communicating that they want to maintain a highly formal relationship with their subordinates, managers who apply controls in coercive ways tend to

Multilevel Control–Trust Dynamics  93

challenge their subordinates’ desires for self-determination. Because subordinates seek to maintain a level of discretion over their decisions and actions, they often interpret these activities as evidence that their manager is unwilling to protect and promote their interests. Subordinates in these situations tend to exhibit a decreased interest in cooperating with their managers based on concerns that those managers cannot be trusted to protect and promote their interests (Falk & Kosfeld, 2006; Ghoshal & Moran, 1996; Sitkin & Roth, 1993). These observations highlight how subordinates’ trust, as well as their motivations to perform and cooperate with authorities, are dependent both on ‘what’ controls managers apply as well as ‘how’ managers apply those controls in directing their subordinates’ work. Long and Sitkin (2018) recently highlighted this issue by suggesting that, while some controls may be better suited for trust-building initiatives, managers may apply any type of control in trustworthy or untrustworthy ways. Research on ‘how’ managers apply controls suggests that the strength with which managers administer controls, how consistently and accurately they implement control parameters, and whether managers use controls to exchange information or monitor subordinates comprise key factors that subordinates use in evaluating the trustworthiness of their managers (Vélez, Sánchez, & ÁlvarezDardet, 2008; Weibel, 2007).

Control–Trust Dynamics at Multiple Organizational Levels A complicating but very real aspect of these dynamics is that the choices managers make about control and trust are not formulated in a vacuum but are directly and indirectly influenced by key contextual elements that lie above, below, and around them in their organizations (Mishra & Mishra, 2013). To provide a picture of interactions between managers, their subordinates, their immediate supervisors, and organizational governance practices, Figure 5.1 details a multilevel map of the conceptual landscape surrounding control–trust dynamics. This perspective provides a comprehensive view of the milieu of strategic, organizational, and interpersonal factors that influence managers’ decisions and actions. In addition, it details the multilevel factors that impact the trust, cooperation, and performance that leaders and employees across organizational levels exhibit. Figure 5.1 outlines dynamics at three levels: organizational, managerial, and employee. The components of this conceptual map are developed using a broad review of both theoretical and empirical work in the control and trust literature (e.g., Cao & Lumineau, 2015; Long & Sitkin, 2018). At each level, the model displays key assessments that influence actors’2 control- and trust-building behaviors.

2 The term ‘actors’ is used in this description of the figure to encompass how, while assessments and actions are often taken by individuals, sometimes (especially at the organizational level) groups of individuals engage in these activities.

94 

Chris Long

1. Strategic Assessments a. Performance-Focused

b.Unit-Focused

c. System-Focused

Financials Operations Customers/Clients Products/Services

Managerial Goodwill Managerial Competence Performance Status Relational Status

Diagnostic Systems Belief Systems Boundary Systems Interactive Systems

Organizaon-Level 2. Strategic Activities a. Controls

b. Trust-Building

Requirements Restrictions Resources Rewards

Competence Consideration Consistency Coherence

7. Feedback

3. Managerial Assessments a. Subordinate-Focused

b. Organization-Focused

c. Self-Focused

Subordinate Goodwill Subordinate Competence Performance Status Relational Status

Organization Goodwill Organization Competence Performance Status Relational Status

Managerial Autonomy Managerial Competence Managerial Empowerment Managerial Propriety

Manager-Level 4. Managerial Activities a. Controls

b. Trust-Building

Input Controls Process Controls Output Controls

Ability-Promotion Benevolence-Promotion Integrity-Promotion

5. Subordinate Assessments c. Self-Focused

b. Organization-Focused

a. Manager-Focused

Subordinate Autonomy Subordinate Competence Subordinate Empowerment Subordinate Propriety

Organization Goodwill Organization Competence Performance Status Relational Status

Managerial Goodwill Managerial Competence Performance Status Relational Status

6. Subordinate Activities a.Instrumental Performance b. Relational Performance Goal Achievement Behavioral Appropriateness Overall Capability Efficiency/Effectiveness

Employee-Level

Cooperation Motivation/Commitment Value Congruence Resistance

FIGURE 5.1 Pictorial

depiction of the theoretical landscape encompassing multilevel control–trust dynamics

Multiple categories of assessments are presented to capture the range of considerations that influence actors’ behaviors at each level. Figure 5.1 depicts the multiple considerations that individuals address in making these assessments and taking these actions.The description of the model that is presented below details how superior–subordinate relationships and control–trust

Multilevel Control–Trust Dynamics  95

dynamics exist at all levels of organizations where individuals undertake controland trust-based decisions and actions. In the model, arrows are used to show how actions taken by a particular actor at one level influence the assessments of actors at other levels. A core proposition of this model is that actors at all levels of the organization (i.e., leaders, managers, employees) take actions in response to the various interpersonal and performance-based assessments they make. Within this text, these categories of assessments are labeled according to the number and the letter depicted in the model. The model also presumes that control–trust dynamics are enacted by actors who assess the extent to which other actors in the organization understand their interests and can be counted on to help them achieve their objectives. At each level, actors assess their performance and relational ‘status’ with other actors. Through these distinct but related status assessments, actors use direct evaluations and attributional processes to gauge the state and quality of their cooperative relationships with other actors (Long, 2018). They specifically examine whether appropriate levels of cooperation currently exist and what factors may be influencing those cooperation levels (Long & Sitkin, 2018). They also evaluate the performance and relational risks inherent in working with the particular targets of those evaluations (Das & Teng, 2001). Through ‘performance status’ assessments, actors assess whether other actors are performing in ways that will enable them to achieve their instrumental objectives. For example, actors assessing the performance status of their subordinates evaluate whether the managerial climate they have created is sufficiently motivating subordinates to accomplish their respective objectives. Performance status assessments of authorities lead actors to examine whether those who are directing them are enabling their success and generally view them as performing successfully in their respective roles. Using ‘relational status’ assessments, actors gauge the state of their interpersonal relationships with other actors. For example, in conducting relational status assessments of authorities, actors will examine how well they get along and ‘are liked’ by their managers. Actors conduct relational status assessments of their subordinates to, for example, gauge whether those whom they are directing view them as legitimate authorities and are happy working under their direction. Actors focus the actions they take as a result of these assessments on directly and indirectly influencing the thoughts and actions of others. They do this in order to enhance their capacities to achieve their personal and professional goals.The specific outcomes that actors are seeking to achieve will differ according to several factors including their position in the organization, the relationships with those whom they work, and other personal or situational factors. Actors achieve this by spending their time and energy evaluating factors that will enable them to most efficiently and effectively understand how to achieve their personal and professional objectives. Actors often balance multiple assessments when deciding actions to initiate. This model focuses on three categories of assessments at each level. The model

96 

Chris Long

specifies two categories of factors in their task environment that actors examine (e.g., subordinate- and organization-focused assessments for managers). In addition, the model outlines how actors at each level conduct an individually directed assessment (e.g., a self-focused assessment for managers).3 The model assumes that actors will process information from their task environment in developing their self-assessments. As a result, the model assumes that actors’ self-assessments exert the strongest influences on their action choices. Managing this multiplex of relationships requires actors to rely heavily on their information-processing capabilities and to focus on particular assessments at particular points in time. How actors focus their assessments will be determined by their particular priorities as well as other contingencies that highlight the importance of specific concerns in their task, relational, and organizational spans of accountability. Examining these dynamics in a multilevel context displays how the actions of others often directly influence the particular concerns on which an actor focuses at any given time. In addition to conducting multiple assessments, actors often concurrently play multiple, distinct but related roles. Across the range of other actors with whom they interact, focal actors can be concurrently serving as authorities to some and subordinates to others. As authorities, actors work to motivate their subordinates to cooperate with them in doing work. As subordinates, actors endeavor to achieve objectives within the parameters set by their authorities. One goal of these activities is to create an environment that is professionally, relationally, and instrumentally beneficial. The model assumes that actors continuously endeavor to achieve these objectives while striking an appropriate balance between varying demands. Across the relationships they maintain, actors process information and exert energies formulating attributions about the interests and intentions of others. Sometimes these attributions stimulate actors to initiate new trust-building and control actions with their exchange partners (authorities and/or subordinates). At other times, actors may choose to maintain more of a static posture or even reduce their control- and trust-building efforts in particular domains. Specific propositions regarding how individuals engage and react to these influences are outside the scope of this chapter. It is hoped that future researchers develop specific propositions around these observations. To assist readers in following the flow of the relationships described in the figure, key conceptual areas are numbered both in the figure and in the accompanying text. The multiple assessments individual actors conduct are described in boxes. Arrows in the figure are placed to show how specific types of actions impact subsequent assessments made by actors at other levels. The model shows

3  Note that the model outlines how the self-assessment that senior leaders engage is an evaluation of the efficacy of an organization’s internal performance management systems.

Multilevel Control–Trust Dynamics  97

that, at any one time, actors process influences from multiple organizational levels in deciding which actions they should take.4 The relationships outlined in Figure 5.1 are disaggregated and presented in detail below.

1. Strategic Assessments The discussion of the figure’s components begins at the strategic level, where topline organizational decisions are made, and broader-based policies and programs are developed and implemented. The pictorial representation of these relationships presented in Figure 5.1 begins by outlining why senior leaders (top management teams, senior executives) deploy particular control- and trust-building initiatives. While acknowledging that senior leaders must also balance a range of concerns, this section focuses on the proximate considerations impacting the directive actions that senior leaders take.This section specifically details how institutional-level considerations are processed by senior leaders as they make decisions about organization-level systems.

a. Performance-Focused Assessments Senior leaders are focused on implementing strategies and directing their organizations to achieve a range of performance objectives. As they evaluate their organization’s progress towards those objectives, they process information contained on explicit and implicit “scorecards” of key performance indicators (Kaplan & Norton, 2007). While scorecards may encompass a range of key metrics, data that senior leaders tend to evaluate include financial measures of performance (e.g., profits, gross margins, EBITA, etc.), qualitative and quantitative information on the status of an organization’s operations, customers, and clients, and their organization’s capacity to produce and deliver particular products or services. The quality and quantity of information that senior leaders use in these evaluations are dependent on existing organizational norms, performance management initiatives, system design considerations, and types of information senior leaders solicit.

b. Unit-Focused Assessments The control- and trust-building actions that senior leaders take are, in large part, determined by comparing their performance expectations against performance

4  It is important to note that the final category of actor assessments that are discussed at each level of the model (e.g., Self-Focused Managerial Assessments) are assumed to exert the strongest influences on the behavior choices of the actor at that level. As a result, the order with which particular categories of assessments are discussed does not follow a consistent sequence (e.g., categories listed left to right).

98 

Chris Long

information that they receive from key actors throughout the organization. Managers within the organization generate reports that provide higher authorities with a range of data points they use to assess how efficiently and effectively subordinate units are cooperating with them in fulfilling the organization’s mission and strategic objectives. In conducting these assessments, senior leaders use these records to gauge the overall health of the relationships between themselves and managers who are implementing the organization’s front-line strategic initiatives (i.e., relational status). Senior leaders also use these assessments to gauge whether the directives and resources they are providing to unit managers are enabling those individuals to assist in the efficient and effective attainment of the organization’s strategic objectives (i.e., performance status). The evaluations they make on these points are important because they influence both how senior leaders control and seek to build trust with managers within the organizations who are executing organizational strategies.

c. System-Focused Assessments Their desires to accomplish particular strategic objectives lead senior leaders to translate the data they receive from ongoing reviews of their managers and organization into assessments of various strategic control mechanisms. Simons’ (1995) “levers of control” framework describes four categories of strategic assessments used by senior leaders to direct decision-making and action across units and entire organizations: belief systems, boundary systems, diagnostic systems, and interactive systems. Authorities use belief systems to establish and maintain an organization’s explicit purpose and direction through the enactment of mission and ethos statements that promote its culture and values. Authorities enforce compliance using boundary systems that specify the relative desirability and appropriateness of individuals engaging in particular decisions and actions. Diagnostic control systems are deployed to require individuals within the organization to report operational successes and failures. Lastly, interactive control systems encompass directive and feedback mechanisms that authorities use to gather and evaluate information to evaluate organizational performance while directing the implementation of strategic initiatives. A distinctive feature of interactive systems is that subordinates participate directly in how these systems are developed and implemented.

2. Strategic Activities Senior leaders use diagnostic, belief, boundary, and interactive indicators to review the firm’s performance and evaluate how efficiently and effectively managers are achieving strategic objectives (Merchant, 1985; Miller, 1992). Based on these evaluations, senior leaders deploy controls in ongoing attempts to motivate subordinate managers’ decisions and actions.

Multilevel Control–Trust Dynamics  99

The overall goal of these actions is to clearly scope out each manager’s span of accountability, which is the “the range of performance measures used to evaluate a manager’s achievements” (Simons, 1994). “Achievements” in this context can mean many things. In some organizations it may mean revenues, levels of service provision, cost savings, or returns on assets. However, because managers may be rated on a range of metrics, the term managerial achievements can describe a varying array of formal and informal evaluative indicators. For example, managers can be rated on the level of job satisfaction their employees exhibit; the extent to which members of their units comply with institutional norms and requirements; their ability to acquire and manage human and physical capital; as well as more informal process-based assessments of how well their team members function over periods of time.

a. Control Activities A review of the control literature suggests that within these systems higher authorities deploy strategically focused controls through four control channels: requirements, restrictions, resources, and rewards (Langfield-Smith, 1997). Control channels specifically describe the particular ways through which control messages are relayed to subordinates. Each channel can be utilized for informative and evaluative purposes. When a channel is used for informative purposes, superiors use it to transmit information about what subordinates are being asked to do and how they are being asked to do it. When a channel is used for evaluative purposes, superiors are deploying controls to assess the level of performance that subordinates have achieved (Lumineau, 2017). Authorities leverage informative and evaluative elements of requirements, restrictions, resources, and rewards to motivate their subordinate managers to achieve strategic goals by directing their attention towards the objectives they want them to accomplish. For example, senior leaders will often specify requirements or standards that managers need to satisfy in doing their work. These include results managers are expected to achieve, processes that they need to follow in pursuing strategic objectives, and ways they should organize human and physical assets (Sitkin, Long, & Cardinal, 2020). In addition to requirements, senior leaders will often issue formal and informal restrictions in order to ensure that managerial conduct complies with existing laws, regulations, or norms (Sitkin & Bies, 1994). Strategic leaders direct the attention and actions of subordinate managers by aligning resource allocations in concert with strategic objectives that they are seeking to achieve (Ouchi, 1978; Simons, 1995). Lastly, senior leaders help to motivate subordinate managers to align their decisions and actions with their firm’s strategic objectives by providing them rewards (and reprimands) for achieving (or failing to achieve) particular performance indicators (Ferrin & Dirks, 2003).

100 

Chris Long

b. Trust-Building Activities As is discussed above, ‘what’ controls senior leaders apply provides managers with strategic direction and outlines key performance indicators. In addition, ‘how’ senior leaders actually implement controls more directly influences their capacities to build organizational trust and motivate managers to willingly comply with their directives. Strategic leaders in organizations build trust when they utilize control channels in ways that are viewed by their subordinates as competent, considerate, consistent, and coherent. Senior leaders are viewed as competent when they build policies that leverage resources, requirements, restrictions, and rewards in ways that establish a cogent context that motivates those within the organization to work towards compelling strategic objectives. Demonstrated mission-focus and strategic effectiveness can provide employees with evidence of organizational coherence (Gillespie & Dietz, 2009). For example, the perceived fairness of organizational rules, processes, and procedures, the transparency of communications, and a demonstrated solicitation of employee voice on important organizational decisions can become referential indicators of organizational consideration (Perrone, Zaheer, & McEvily, 2003). When these principles are consistently applied to policies, practices, and procedures across time and across individuals they provide subordinate managers with a sense of coherence and clear sets of directives to accomplish mission objectives (Hernandez et al., 2014).

3. Managerial Assessments While desirable, effectively balancing control- and trust-building initiatives is not an easy end-state for senior leaders to achieve. Managers who are charged with executing strategic imperatives process the actions of senior leaders through a prism of other influences that exist in and around them. They use the directives of higher authorities as a way of guiding their own control- and trust-based decisions. When senior leaders (i.e., and the overall organization) act in trustworthy ways by implementing organizational policies that their subordinate managers see as beneficial, those individuals will be more empowered to pursue organizational objectives. It is important to note, however, that the control- and trust-building actions of higher authorities exert strong, yet incomplete influences on managers’ own control- and trust-building decisions.To make these decisions, managers also conduct critical evaluations of both their interactions with subordinates as well as intrapersonal assessments of their personal and managerial capabilities. Encouraging leaders and cooperative subordinates will often increase managers’ positive perceptions of their legitimacy and authority (Long, 2010).

a. Subordinate-Focused Assessments Managers obtain information about their subordinates’ instrumental and relational performances from direct observation, information provided to them from

Multilevel Control–Trust Dynamics  101

various organizational stakeholders, as well as data transmitted to them through performance management systems. Managers use these evaluations to determine whether their subordinates are acting in trustworthy ways (i.e., competently and with goodwill) and what the risks are of working with these subordinates. The conclusions that managers make on these counts are influenced by managers’ relational status assessments that they use to gauge whether subordinates like working with them and view them as legitimate authorities. Managers also apply these assessments into holistic evaluations of their subordinates’ performances to determine how efficiently and effectively those subordinates are achieving their (i.e., and their organizational unit’s) performance objectives (i.e., performance status). From this, managers and organizations assess performance risks and the likelihood that they will be able to achieve their performance requirements (Das & Teng, 2001; Ouchi, 1979; Williamson, 1975). Doing this helps managers understand what they must endeavor to do to accomplish their personal and professional (e.g., organizational) objectives (Bijlsma-Frankema & van de Bunt, 2003; Long, 2018).

b. Organization-Focused Assessments How senior leaders utilize control channels and the qualities of relationships they produce through those interactions are important because they influence the subsequent control- and trust-building actions that their subordinate managers take (Righetti & Finkenauer, 2011). Managers in organizations routinely evaluate their relationships with higher authorities to determine whether higher authorities believe they are performing effectively in their roles and are providing them the resources necessary to facilitate their professional success (i.e., performance status). They also use these evaluations to gauge the overall status of their relationships with those authorities and whether authorities can be counted on to develop and maintain encouraging, positive relationships with them (i.e., relational status) (Ross, 1994). Managers who will tend to perceive their authorities as more trustworthy and worthy of higher levels of transparency and cooperation when those authorities share information and other resources, coordinate performance management activities with them, and accurately assess their achieved performances on strategically relevant dimensions (Gillespie & Dietz, 2009; Pillutla, Malhotra, & Murnighan, 2003).

c. Self-Focused Assessments Managers use interactions with both their superiors and subordinates to generate pictures of how well they are fulfilling their diagnostic, belief, boundary, and interactive responsibilities within their spans of accountability. As they do this, they assess the efficacy of their specific control choices and the overall approach they are using to direct subordinate actors within their own spans of control (Merchant & Van der Stede, 2007).

102 

Chris Long

Managers also use these evaluations to assess their personal capabilities. Through interactions with both their higher authorities and their own subordinates, managers gauge their autonomy of action as well as the extent to which others view them as competently and appropriately enacting their managerial role obligations. Managers compile these evaluations into comprehensive assessments of whether their subordinates and superiors see them as acting with propriety and view them as trusted, legitimate authorities. Managers receive these affirmations and are empowered by perceptions of themselves as capable of motivating the levels and types of cooperation necessary to accomplish important strategic objectives (Long & Sitkin, 2006; Sitkin & Bies, 1994; Sitkin & George, 2005; Spreitzer, 1995; Williamson, 1975).

4. Managerial Activities The assessments that managers make are important because these evaluations directly impact their ongoing control- and trust-building actions. The challenge that managers face in undertaking these activities is that they must concurrently attempt to gain the trust and confidence of at least two distinct audiences. At any one time, senior leaders are evaluating how confident they are that managers within their organization can and are achieving organizational objectives. At the same time, managers’ subordinates are also evaluating whether their superiors can be relied upon to cogently direct their work activities while protecting and promoting their individual needs and interests. During some periods, the values, needs, and interests of all involved parties (higher authorities, managers, and subordinates) are congruent and closely aligned. At these times, managers can more easily engender subordinate trust because higher authorities are deploying resources, requirements, restrictions, and rewards in ways consistent with their subordinates’ expectations and desires. At other times, the values, needs, and interests of these parties are incongruent and divergent. During these periods, managers will find it more difficult to foster trust among their subordinates because higher authorities may be deploying resources, requirements, restrictions, and rewards in ways that constrain or compromise managers’ capacities to competently, considerately, consistently, and coherently address their subordinates’ needs and interests.

a. Control Activities To engage these complex sets of issues, managers attempt to develop appropriate configurations of control- and trust-building activities by triangulating their desires to accomplish their personal objectives with the demands placed on them by both higher authorities and their subordinates (Jagd, 2010; Whitener, Brodt, Korsgaard, & Werner, 1998). Managers generally attempt to align the ways they direct their subordinates with the standards that their higher authorities expect

Multilevel Control–Trust Dynamics  103

them to accomplish. Managers combine these assessments with evaluations of their subordinates’ past and current performances to determine how they will deploy controls and build trust with those whom they are directing. In taking these decisions, managers are seeking to deploy formal and informal input, process, and output controls in ongoing attempts to both satisfy their strategic performance requirements while they attempt to impart the quality of task direction that will be motivating for their subordinates (Cardinal et al., 2004, 2010).

b. Trust-Building Activities Managers use their assessments to determine the efficacy of their current managerial approach and implement sets of trust-building activities that motivate the types and levels of cooperation they seek to engender with their subordinates (Long, 2018; Long & Sitkin, 2006). As managers implement controls, they leverage assessments of their capabilities and current requirements to help ensure that they make related efforts to demonstrate their ability, benevolence, and integrity (Spreitzer & Mishra, 1999; Whitener et al., 1998). As Long (2018) recently showed, managers seek to demonstrate their trustworthiness in ways that align closely with both the types of subordinate cooperation managers seek to generate and the controls they implement. How competently, consistently, and considerately managers implement the components of performance management systems (e.g., evaluations, rewards, feedback) provides important signals about their willingness and ability to protect and promote their subordinates’ interests (Dickson, Smith, Grojean, & Ehrhart, 2001; Gillespie & Dietz, 2009).

5. Subordinate Assessments Through interactions with their managers, subordinates become familiar with their responsibilities, gauge their capabilities, and assess their exposure to performance and relational risks (Mayer et al., 1995).

a. Manager-Focused Assessments Managerial control actions influence subordinates’ perceptions of their own selfdetermination, which is a key issue impacting their cooperation and motivation (Weibel, 2007). For example, managers’ applications of controls directly influence subordinates’ perceptions of their autonomy and task competence (Bandura, 1977, 1982; Deci & Ryan, 1985). The controls managers employ also impact how empowered subordinates feel by communicating to them both how much autonomy they have over the work they do and whether they have been acting appropriately in completing their tasks (Lee, Ashford, & Bobko, 1990; Skinner, 1996). Research demonstrates how formal controls tend to denigrate subordinates’ autonomy, competence, and sense of empowerment while more informal

104 

Chris Long

controls produce still strong but more positive effects on these attitudes (Das & Teng, 2001). The efforts managers make to apply controls and build trust directly impact subordinates’ evaluations of managerial trustworthiness and their assessments of whether their managers are seeking to take advantage of them. Concurrently, subordinates assess how managers view them as performing and whether they are providing them resources that are facilitating their professional success (professional status). Subordinates conduct relational assessments to also gauge whether their managers ‘like them’ and are interested in developing and maintaining positive interpersonal relationships with them (i.e., relational risk) (Ferrin et al., 2007; Skinner, 1996; Weibel, 2007). Concurrent with these assessments, subordinates critically examine the communications, decisions, and actions of their immediate supervisors to evaluate how willing and able managers are to accommodate their subordinates’ needs and interests (goodwill and competence). For example, assessments of how their managers implement organizational policies, processes, and procedures often serve as the basis for understanding how willing and able (i.e., competent) managers are to protect their subordinates’ values, needs, and interests (Tyler & Lind, 1992; McEvily, Perrone, & Zaheer, 2003). Subordinates will tend to view their managers as exhibiting more goodwill the more that they enact decisions, strategies, and policies that promote fairness, encourage participation, celebrate their contributions, and foster strategic consistency and coherence (Das & Teng, 1998; Stahl et al., 2011; Weibel, 2007).

b. Organization-Focused Assessments As they engage in these assessments, subordinates also gather information that they use in evaluating the overall trustworthiness as well as the general legitimacy of senior leaders’ organizational-focused decisions and whether those leaders will attempt to take advantage of them (Gillespie & Dietz, 2009; Mayer et al., 1995). Using this information, subordinates assess whether their senior leaders are creating a context where they are able to fulfill their professional responsibilities and achieve their required objectives (i.e., performance status). At the same time, they use assessments of their relational status to gauge the quality of their interpersonal relationships with organizational decision-makers and whether higher authorities value them as people (i.e., relational status) (Das & Teng, 2001; Ferrin et al., 2007).The extent to which managers enact performance-management systems in ways that are supportive, participative, transparent, and fair provides subordinates with additional but important signals about their senior leaders’ values and overall trustworthiness (Long, 2010; Malhotra & Lumineau, 2011). While distinguishable in the mind of the subordinate, how their senior leaders and managers act influence subordinates’ trust assessments of both entities (Fulmer & Gelfand, 2012; Klein, Dansereau, & Hall, 1994). If the manager acts strictly in implementing an organization’s policies and procedures, subordinates will tend to

Multilevel Control–Trust Dynamics  105

closely align their trust assessments and attribute their managers’ actions to their organization. Alternatively, managers who enact policies and procedures that are less strictly aligned with the dictates of senior leaders will be assessed by their subordinates as more independent. In these instances, subordinates will tend to evaluate their manager and senior leaders as separate entities (Ostroff, Kinicki, & Tamkins, 2003; Schulte, Ostroff, & Kinicki, 2006).

c. Self-Focused Assessments Subordinates evaluate the demands placed on them by higher authorities against the resources and rewards they possess to accomplish those objectives (Lee et al., 1990; Skinner, 1996). For example, subordinates develop their sense of selfdetermination through evaluations of the levels of autonomy that authorities are providing them.They combine these assessments with examinations of their overall competence in completing tasks and accomplishing key objectives (Bandura, 1977; Deci & Ryan, 1985). When subordinates feel confident, see authorities affirming that they are acting appropriately (propriety), and conclude that they maintain autonomy in their cognitions and actions, they tend to be more empowered to cooperate with those authorities in doing work (Weibel, 2007). If, however, authorities constrain their subordinates’ sense of autonomy, compromise their sense of self-efficacy, or otherwise make them feel incompetent or inappropriate, they will compromise their employees’ desires to collaborate with those authorities in accomplishing performance objectives.

6. Subordinate Activities a. Instrumental Performance The assessments that subordinates perform both of themselves and higher authorities within their influence spheres directly impact the efforts they make to achieve performance objectives (Long, Bendersky, & Morrill, 2011; Roth, Sitkin, & House, 1994; Şengün & Wasti, 2007). Subordinates’ assessments of their own capabilities and their interests influence the amounts of attention and effort they apply to their task- and goal-based responsibilities (De Cremer & Tyler, 2007; Weibel, 2007). In general, subordinates who feel more empowered and self-determined will demonstrate higher levels of competence and effectiveness as they more willingly pursue strategic objectives and work in ways that their superiors will view as appropriate (i.e., in alignment with organizational norms) (Spreitzer, 1995; Whitener, 2001).

b. Relational Performance The levels of self-determination that subordinates feel also influence how outwardly motivated and committed subordinates are to achieving organizational

106 

Chris Long

objectives. Subordinates who are empowered to work towards organizational objectives will express stronger desires to cooperate with their authorities in doing work (Spreitzer, 1995). The lack of resistance that these subordinates express toward organizational directives communicates that they currently maintain goals and values that are congruent with those expressed by their superiors (McAllister, 1995; Weick, Sutcliffe, & Obstfeld, 2008). Alternatively, subordinates who express low levels of motivation and commitment, who resist their superiors’ influence attempts, or who enact values that are incongruent with the organization provide signals to higher authorities that they may need to adjust their management approach in order to elicit the forms and types of cooperation they seek to engender (Knoll & Gill, 2011; Long, 2018; Nooteboom, 1996).

7. Feedback The information that higher authorities acquire from the performances that subordinates exhibit is a critical element in multilevel control–trust dynamics. Multilevel control and trust dynamics are enacted through cybernetic systems that generate performance information (Merchant & Otley, 2006). This performance information is assessed by actors who compare it against expected results. When expected levels of performance are not achieved, superiors adjust the elements of their system’s (control- and trust-building) initiatives in ongoing attempts to more effectively direct and empower their subordinates. For example, managers will closely examine the instrumental and relational performances of their direct reports. From these evaluations, managers develop assessments of their subordinates’ trustworthiness. They are able to examine how competent their subordinates are at performing their work responsibilities. In addition, they concurrently evaluate whether employees are acting with goodwill by fulfilling promises of performance expectations by enacting organizational goals, norms, and values through their decisions and actions (i.e., acting benevolently toward these entities) (Long & Sitkin, 2018). Leaders at the top of organizational hierarchies incorporate the data that they receive from performance management systems directly into their evaluations of how efficiently and effectively their organization is accomplishing its strategic objectives (financials, operations, customers/clients, products/services). They combine this data with the information that they receive from managers within the organization to determine both how individual units are performing and whether the managers who lead those units can be trusted. Both of these assessments are important for organizational leaders as they influence their determinations of how well an organization’s performance management system is working and whether adjustments need to be made to the ways leaders direct and motivate actors within the organization (Simons, 1995).

Multilevel Control–Trust Dynamics  107

Discussion By mapping out the theoretical landscape surrounding multilevel control–trust dynamics, this chapter moves control–trust research from examining just manager–subordinate interactions to conceptualizing the range of antecedent actions and resulting perceptions and attitudes that encompass control–trust dynamics within organizations (Long & Sitkin, 2018; Malhotra & Lumineau, 2011). Specifically, the framework presented in this chapter broadens perspectives often presented in control–trust research by highlighting how managers’ control- and trust-building actions, along with organization-level policies implemented by senior leaders, combine with subordinates’ past performances to impact their control and trust perceptions as well as a number of related job attitudes and behaviors. On this point, one potentially important contribution for control–trust scholars is that while past research has examined authorities’ implementation of controls, this work highlights the importance of understanding how control- and trust-building activities enacted by authorities throughout an organization impact the experiences of individuals they seek to direct (Long & Sitkin, 2018; Sitkin et al., 2020). Because control–trust dynamics often hinge on controlees endorsing particular controls as legitimate, how managers’ directives and related actions are actually perceived and experienced by controlees should become more integral components of control–trust theories and empirical research. Doing so would lead scholars to develop fully specified and relevant pictures of how subordinate trust, cooperation, motivation, and performance are generated through control–trust dynamics. Building from these ideas, future research can examine whether various authorities’ control- and trust-building choices combine to positively influence their subordinates’ self-determination perceptions of autonomy, competence, and relatedness (Deci & Ryan, 1985). Additional work developed on these issues can evaluate whether these conditions lead subordinates to generate a robust sense of agency sufficient to motivate them to cooperate with their authorities to perform well in their work tasks (Burger, 1987, 1989; Thompson, 1981). Another important contribution of this work is the picture it presents of how control- and trust-related decisions and actions taken by senior leaders cascade through the organization to influence the perceptions and behaviors of front-line employees. The framework specifically displays how key senior leader behaviors directly and indirectly impact the development of employee trust, cooperation, motivation, and performance (Kay, Whilson, Gaucher, & Galinsky, 2005; Kay, Gaucher, Napier, Callan, & Laurin, 2008). In addition, it details how the attitudinal and behavioral reactions of subordinates to multiple, higher authorities influence the ongoing decisions and actions that leaders take. Future research will be needed to build from these general propositions to fully assess how the dynamics described here influence leaders’ and subordinates’ control- and trust-related cognitions, emotions, and behaviors.

108 

Chris Long

Future Research Over the past 20 years, scholars have examined control–trust issues because these dynamics are integral components of organizational responses to increasing environmental uncertainty and complexity, globalization pressures, the changing nature of work, evolving market demands, and shifting contractual arrangements between organizations (Adler, 2001).These factors highlight the need to continue to evaluate the basic dynamics under which control and trust independently and jointly affect important outcomes. However, much work is left to be done as scholars acknowledge sentiments similar to those previously expressed by Emsley and Kidon (2007) that we remain “a long way from understanding how trust and control coexist, and many basic issues remain unresolved” (p. 829) (for similar assessments, see Bijlsma-Frankema & Costa, 2005; Langfield-Smith & Smith, 2003; Long & Sitkin, 2018). The multilevel perspective presented in this chapter seeks to clarify some of these issues by providing a framework for understanding how key factors within organizations jointly impact the enactment and experience of control–trust dynamics (Cao & Lumineau, 2015; Sitkin & Bies, 1994). The overall objective of this effort is to assist scholars in fostering conceptual consensuses around what elements in organizations influence managers’ choices regarding the control- and trust-building actions they undertake, how those actions directly affect subordinates’ perceptions and behaviors, and how the performances that are achieved through those activities impact the subsequent decisions and actions that senior leaders take. Scholars can use the figure and related discussion in this chapter to begin developing testable hypotheses of multilevel control–trust dynamics. For example, using the components of this figure, scholars can begin to chart the ways that subordinates’ perceptions of control and trust are influenced by ‘what’ controls managers apply and ‘how’ they apply those controls. Notably, this figure can also be used to help researchers conceptualize how factors above and below managers influence their control- and trust-building decisions and actions. This perspective is important because it can allow us to more accurately assess the institutional and interpersonal factors that concurrently impact both how managers direct their subordinates and communicate with higher authorities about those initiatives. Research conducted on these topics holds the potential to advance our understanding of the multilevel nature of trust in several important ways. For example, it highlights three fundamental ways how multilevel factors influence actors’ control- and trust-building decisions and actions up and down and organizational hierarchy. First, it outlines how the trust that subordinates have in their managers is partially determined by the institutional context within which that relationship resides. Second, it outlines how actions taken by frontline employees can influence institutional trust development through the performance information that managers communicate to senior leaders. Third, it

Multilevel Control–Trust Dynamics  109

outlines how managers’ control- and trust-building efforts are concurrently influenced by their relationships with senior leaders and the subordinates they are charged with directing.

Influences of Power and Authority The research that may emerge from this work promises to significantly increase our comprehension of power and authority relationships in organizations (Long, 2010). Surprisingly, and with few exceptions (e.g., Horak & Long, 2018;Yeung, Selen, Zhang, & Huo, 2009), the relationship between power and trust has still not generally been an explicit focus of control–trust researchers. However, the relationships between influence, dependence, and vulnerability that are endemic to power dynamics are a potentially important focus for future scholarship in this area. The limited research on this topic suggests that symmetric trust between exchange partners can lead to positive instrumental and relational outcomes (Bouty, 2017). For example, scholars have found that managers and subordinates who trust each other exhibit higher levels of organizational citizenship behaviors (Brower, Lester, Korsgaard, & Dineen, 2009; Dineen, Lewicki, & Tomlinson, 2006). Research also suggests that, over time, power dynamics can lead to persistent or even institutionalized control that fosters trust asymmetries (Schoorman, Mayer, & Davis, 2007). Trust asymmetries tend to breed lower levels of overall performance and increased goal and value incongruities because partners who trust too much may be too lenient on their exchange partners’ exploitation and risk-taking activities (Zaheer & Zaheer, 2006). Over time, those who trust too little may become overly sensitive and reactionary to actions they view as inappropriate (Goel, Bell, & Pierce, 2005). Future research should also investigate how perceptions of high and low power can significantly impact how individuals enact control- and trust-building activities (Greenwood & Buren, 2010). For example, work by Schoorman et al. (2007) suggests that authorities who perceive themselves as holding high status and power positions in organizations will often take risks and act in ways that subordinates might view as inappropriate. Because they perceive themselves to be powerful, these individuals may tend to less actively regulate their directive behaviors and be less willing to ensure that they align their control- and trust-building actions with their subordinates’ expectations (Van Dijke & Verboon, 2010). Future research will be needed to examine the complex effects that these decisions produce for actors within organizations.While work by Murrell, Blake-Beard, Porter, and Perkins-Williamson (2008) suggests that authorities who are perceived to have more power may more easily gain the trust of their subordinates, work by Chua, Wang, Wang, Huang, and Cheng (2008) suggests that actions subordinates perceive as inappropriate may actually increase perceptions of managerial integrity and ability but compromise perceptions of benevolence.

110 

Chris Long

Contextual Influences Changes in the strategic landscapes surrounding organizations provide a strong impetus and rich fertile ground for future research on multilevel control–trust dynamics. Consistent with this assertion, Biljsma-Frankema and Costa (2005) observe that changes that have occurred in organizational environments over the past 30 years have stimulated increasing scholarly interest in control–trust dynamics. Managers in these environments have encountered greater levels of task uncertainty that have highlighted both the limitations of traditional commandand-control mechanisms and the potential advantages of trust-based mechanisms in motivating subordinates’ cooperation and control (Mizrachi, Drori, & Anspach, 2007; Rus & Iglič, 2005; Woolthuis, Hillebrand, & Nooteboom, 2005). While work in this area is still in a nascent stage, what becomes clear from this research is that contextual factors outside of organizations can significantly impact control–trust dynamics within organizational boundaries (Mishra & Mishra, 2013). The framework presented in this chapter is important for this work because it provides scholars with a range of factors that may be directly and indirectly impacted by these contextual influences. For example, researchers could examine how the relationships that individuals have with external governing institutions influence both the control choices they make and the types of trust that develop between them, their authorities, their peers, their subordinates, and other organizational stakeholders (Grey & Garsten, 2001; Pearce, Branyiczki, & Bigley, 2000). Initial work by scholars in this area has shown that weaker and external formal institutions lead individuals to rely more on pervasive cultural norms inside organizational boundaries (Puffer, McCarthy, & Boisot, 2010; Tan, Yang, & Veliyath, 2009). Future scholarship in this area could employ co-evolutionary perspectives to evaluate how the decisions of senior leaders both shape and are shaped by various institutional and geopolitical influences (Lewin, Long, & Carroll, 1999). Further insights developed from this work can evaluate how cultural and geopolitical factors influence the ways that individuals conceptualize, enact, and respond to control–trust dynamics (Mizrachi et al., 2007; Reed, 2001; Roberts, 2001). Importantly, whether senior leaders adopt and how they implement technologically advanced monitoring systems represents a contextual factor that can fundamentally and significantly impact control–trust dynamics at multiple organizational levels. As organizations increasingly use electronic surveillance and other technologies to augment their capacity to monitor their employees’ activities, scholars are well advised to examine how these mechanisms impact control–trust dynamics at multiple organizational levels (Hongo, 2015; Long & Weibel, 2018; Son, 2015). For example, the implications of decisions that senior organizational leaders make to either adopt or avoid using these technologies will cascade through various levels of organizations in ways that can impact employee perceptions of privacy, autonomy, self-determination, and trust in

Multilevel Control–Trust Dynamics  111

organizations and their authorities. In addition to influencing individual perceptions, these more impersonal forms of control also hold the potential to significantly influence how cooperative, efficient, and effective employees are in performing their tasks (Ball, 2010).

Conclusion The theoretical framework of control–trust dynamics outlined in this chapter identifies and clarifies key relationships between individuals and their authorities, peers, and subordinates. Toward this end, the conceptual roadmap presented here details several important, multilevel factors that impact control–trust dynamics. It demonstrates how factors at one analytical level can produce direct and indirect cascading effects that permeate multiple, hierarchical levels. As scholars begin to tackle the challenges of doing more multilevel work on control–trust dynamics it will be increasingly important for them to examine the full complexity of control- and trust-based relationships in organizations. Specifically, by considering the relationships described here, scholars conducting multilevel control–trust research may gain insights that they can use to refine their theoretical ideas, craft empirical methods to test the relationships they posit, and assist the continued development of the overall knowledge base of the control–trust dynamics that comprise fundamental elements of organizational life.

References Adler, P. S. (2001). Market, hierarchy, and trust: The knowledge economy and the future of capitalism. Organization Science, 12(2), 215–234. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​2​.2​​. 2​​15​.10​​117 Adler, P. S., & Borys, B. (1996). Two types of bureaucracy: Enabling and coercive, Administrative Science Quarterly, 41, 61–89. https://doi​.org​/10​.2307​/2393986 Anderson, S.W., Christ, M. H., Dekker, H. C., & Sedatole, K. L. (2013).The use of management controls to mitigate risk in strategic alliances: Field and survey evidence. Journal of Management Accounting Research, 26(1), 1–32. https://doi​.org​/10​.2308​/jmar​-50621 Bachmann, R. (2001).Trust, power and control in trans-organizational relations. Organization Studies, 22(2), 337–365. https://doi​.org​/10​.1177​/0170840601222007 Ball, K. (2010). Workplace surveillance: An overview. Labor History, 51(1), 87–106. https​:/​/ do​​i​.org​​/10​.1​​080​/0​​02365​​610​03​​65477​6 Bandura,A. (1977). Self-efficacy:Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191. https​:/​/do​​i​.org​​/10​.1​​037​/0​​033​-2​​95X​​.8​​4​.2​.1​​91 Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37(2), 122–147. https​:/​/do​​i​.org​​/10​.1​​037​/0​​003​-0​​66X​​.3​​7​.2​.1​​22 Bhattacharya, R., Devinney, T. M., & Pillutla, M. M. (1998). A formal model of trust based on outcomes. Academy of Management Review, 23(3), 459–472. https://doi​.org​/10​.5465​/ amr​.1998​.926621 Bijlsma-Frankema, K. M., & Costa, A. C. (2005). Understanding the trust-control nexus. International Sociology, 20(3), 259–282. https://doi​.org​/10​.1177​/0268580905055477

112 

Chris Long

Bijlsma-Frankema, K. M., & van de Bunt, G. G. (2003). Antecedents of trust in managers: A “bottom up” approach. Personnel Review, 32(5), 638–664. https​:/​/do​​i​.org​​/10​.1​​108​/ 0​​04834​​803​10​​48838​8 Blau, P. M. (1964). Exchange and power in social life. New York, NY: Wiley. Bouty, I. (2017). Interpersonal and interaction influences on informal resource exchanges between R&D researchers across organizational boundaries. Academy of Management Journal, 43, 50–65. https://doi​.org​/10​.5465​/1556385 Bradach, J. L., & Eccles, R. G. (1989). Price, authority, and trust: From ideal types to plural forms. Annual Review of Sociology, 15, 97–118. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​.so.​​15​. 08​​​0189.​​00052​5 Brower, H., Lester, S., Korsgaard, M., & Dineen, B. (2009). A closer look at trust between managers and subordinates: Understanding the effects of both trusting and being trusted on subordinate outcomes. Journal of Management, 35, 327–347. https://doi​.org​/10​.1177​/ 0149206307312511 Burger, J. M. (1987). Increased performance with increased personal control: A selfpresentation interpretation. Journal of Experimental Social Psychology, 23(4), 350–360. https​:/​/do​​i​.org​​/10​.1​​016​/0​​​022​-1​​031(8​7)900​46-1 Burger, J. M. (1989). Negative reactions to increases in perceived personal control. Journal of Personality and Social Psychology, 56(2), 246. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.5​​6​.2​.2​​46 Cao, Z., & Lumineau, F. (2015). Revisiting the interplay between contractual and relational governance: A qualitative and meta-analytic investigation. Journal of Operations Management, 33, 15–42. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jom.​​201​4.​​09​.00​9 Cardinal, L. B., Sitkin, S. B., & Long, C. P. (2004). Balancing and rebalancing in the creation and evolution of organizational control. Organization Science, 15(4), 411–431. https:// doi​.org​/10​.1287​/orsc​.1040​.0084 Cardinal, L. B., Sitkin, S., & Long, C. (2010). A configurational theory of control. In S. B. Sitkin, L. B. Cardinal, & K. Bijlsma-Frankema (Eds.), Organizational control. Cambridge: Cambridge University Press. Cardinal, L. B., Sitkin, S. B. Long, C. P., & Miller, C. C (2018). The genesis of control configurations during organizational founding. In J. Joseph, O. Baumann, R. M. Burton, & K. Srikanth (Eds.), Organization design: Advances in strategic management (pp. 83–114). Bingley, UK: Emerald Publishing Limited. Chua, L., Wang, A., Wang, T., Huang, M., & Cheng, B. (2008). Shared work values and team member effectiveness: The mediation of trustfulness and trustworthiness. Human Relations, 61, 1713–1742. https://doi​.org​/10​.1177​/0018726708098083 Coletti, A. L., Sedatole, K. L., & Towry, K. L. (2005). The effect of control systems on trust and cooperation in collaborative environments. The Accounting Review, 80(2), 477–500. https​:/​/do​​i​.org​​/10​.2​​308​/a​​ccr​.2​​005​​.8​​0​.2​.4​​77 Connell, J., Ferres, N., & Travaglione, T. (2003). Engendering trust in manager-subordinate relationships: Predictors and outcomes. Personnel Review, 32(5), 569–587. https​:/​/do​​i​. org​​/10​.1​​108​/0​​04834​​803​10​​48834​2 Costa, A. C., & Bijlsma-Frankema, K. (2007). Trust and control interrelations new perspectives on the trust—control nexus. Group & Organization Management, 32(4), 392–406. https://doi​.org​/10​.1177​/1059601106293871 Currall, S. C & Inkpen, A. C. (2002). A multilevel approach to trust in joint ventures. Journal of International Business Studies, 33(3), 479–495. https​:/​/do​​i​.org​​/10​.1​​057​/p​​algra​​ve​.ji​​​bs​. 84​​91027​

Multilevel Control–Trust Dynamics  113

Das,T. K., & Teng, B. S. (1998). Between trust and control: Developing confidence in partner cooperation in alliances. Academy of Management Review, 23(3), 491–512. https://doi​.org​/ 10​.5465​/amr​.1998​.926623 Das, T. K., & Teng, B. S. (2001). Trust, control, and risk in strategic alliances: An integrated framework. Organization Studies, 22(2), 251–283. https://doi​.org​/10​.1177​/ 0170840601222004 Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum Press. De Cremer, D., & Tyler, T. R. (2007). The effects of trust in authority and procedural fairness on cooperation. Journal of Applied Psychology, 92(3), 639–649. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​021​-9​​010​​.9​​2​.3​.6​​39 Dekker, H. C. (2004). Control of inter-organizational relationships: Evidence on appropriation concerns and coordination requirements. Accounting, Organizations and Society, 29(1), 27–49. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​361-​​3682(​02)00​056-9​ Dickson, M. W., Smith, D. B., Grojean, M. W., & Ehrhart, M. (2001). An organizational climate regarding ethics: The outcome of leader values and the practices that reflect them. The Leadership Quarterly, 12(2), 197–217. https​:/​/do​​i​.org​​/10​.1​​016​/S​​1​048-​​9843(​ 01)00​069-8​ Dineen, B. R., Lewicki, R. J., & Tomlinson, E. C. (2006). Supervisory guidance and behavioral integrity: Relationships with employee citizenship and deviant behavior. Journal of Applied Psychology, 91(3), 622–635. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​1​.3​.6​​22 Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of Management Review, 14(1), 57–74. https://doi​.org​/10​.5465​/amr​.1989​.4279003 Emsley, D., & Kidon, F. (2007). The relationship between trust and control in international joint ventures: Evidence from the airline industry. Contemporary Accounting Research, 24(3), 829–858. https://doi​.org​/10​.1506​/car​.24​.3.7 Falk, A., & Kosfeld, M. (2006). The hidden costs of control. The American Economic Review, 96(5), 1611–1630. https://10.1257/aer.96.5.1611 Fayol, H. (1949). General and industrial administration. New York: Pitman. Ferrin, D. L., Bligh, M. C., & Kohles, J. C. (2007). Can I trust you to trust me? A theory of trust, monitoring, and cooperation in interpersonal and intergroup relationships. Group & Organization Management, 32(4), 465–499. https://doi​.org​/10​.1177​/ 1059601106293960 Ferrin, D. L., & Dirks, K. T. (2003). The use of rewards to increase and decrease trust: Mediating processes and differential effects. Organization Science, 14(1), 18–31. https​:/​/ do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.1​​8​.128​​09 Fox, A. (1974). Beyond contract:Work power and trust relations. London: Faber, Ltd. Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust. Journal of Management, 38, 1167–1230. https://doi​.org​/10​.1177​/0149206312439327 Ghoshal, S., & Moran, P. (1996). Bad for practice: A critique of the transaction cost theory. Academy of Management Review, 21(1), 13–47. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​96​.​ 96​​02161​​563 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34, 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 Goel, S., Bell, G. G., & Pierce, J. L. (2005). The perils of Pollyanna: Development of the over-trust construct. Journal of Business Ethics, 58, 203–218. https​:/​/do​​i​.org​​/10​.1​​007​/ s​​10551​​-00​5-​​1415-​6

114 

Chris Long

Greenwood, M., & Buren, H., III. (2010). Trust and stakeholder theory: Trustworthiness in the organisation–stakeholder relationship. Journal of Business Ethics, 95, 425–438. https​:/​/ do​​i​.org​​/10​.1​​007​/s​​10551​​-01​0-​​0414-​4 Grey, C., & Garsten, C. (2001). Trust, control and post-bureaucracy. Organization Studies, 22(2), 229–250. https://doi​.org​/10​.1177​/0170840601222003 Gulati, R., & Nickerson, J. A. (2008). Interorganizational trust, governance choice, and exchange performance. Organization Science, 19(5), 688–708. https://doi​.org​/10​.1287​/ orsc​.1070​.0345 Hartmann, F., & Slapničar, S. (2009). How formal performance evaluation affects trust between superior and subordinate managers. Accounting, Organizations and Society, 34(6), 722–737. Hernandez, M., Long, C. P., & Sitkin, S. B. (2014). Cultivating follower trust: Are all behaviors equally influential? Organization Studies, 35(12), 1867–1892. https://doi​.org​/ 10​.1177​/0170840614546152 Hongo, J. (2015, September 8). Who’s the boss? Hitachi looks to promote artificial intelligence. The Wall Street Journal. Retrieved from http:​/​/blo​​gs​.ws​​j​.com​​/japa​​nreal​​ time/​​2015/​​09​/08​​/whos​​-the-​​boss-​​hitac​​hi​-lo​​oks​-t​​o​-pro​​mote-​​arti​f​​i cial​​-inte​​llige​​nce/ Horak, S., & Long, C. P. (2018). Dissolving the paradox: Toward a Yin–Yang perspective on the power and trust antagonism in collaborative business relationships. Supply Chain Management: An International Journal, 23(6), 573–590. https://doi​.org​/10​.1108​/SCM​-01​ -2018​-0013 Inkpen, A. C., & Currall, S. C. (2004). The coevolution of trust, control, and learning in joint ventures. Organization Science, 15(5), 586–599. https://doi​.org​/10​.1287​/orsc​. 1040​.0079 Jagd, S. (2010). Balancing trust and control in organizations: Towards a process perspective. Society and Business Review, 5(3), 259–269. https​:/​/do​​i​.org​​/10​.1​​108​/1​​74656​​810​11​​07948​2 Kaplan, R. S., & Norton, D. P. (2007). Using the balanced scorecard as a strategic management system. Harvard Business Review, 85, 7–8. Kay, A., Gaucher, D. Napier, J., Callan, M., & Laurin, K. (2008). God and the government: Testing a compensatory control mechanism for the support of external systems, Journal of Personality and Social Psychology, 95(1), 18–35. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​51​4​. 9​​5​.1​.1​8 Kay, A. Whilson, J. Gaucher, D., & Galinsky, A. (2005). Compensatory control: Achieving order through the mind, our institutions, and the heavens, Current Directions in Psychological Science, 18(50), 264–268. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-8721​​.200​9​​. 0164​​9.x Khodyakov, D. M. (2007). The complexity of trust-control relationships in creative organizations: Insights from a qualitative analysis of a conductorless orchestra. Social Forces, 86, 1–22. https://doi​.org​/10​.1353​/sof​.2007​.0100 Klein, K., Dansereau, F., & Hall, R. (1994). Levels issues in theory development, data collection, and analysis. Academy of Management Review, 19, 195–229. https​:/​/do​​i​.org​​/ 10​.5​​465​/a​​mr​.19​​94​.​94​​10210​​745 Knoll, D. L., & Gill, H. (2011). Antecedents of trust in supervisors, subordinates, and peers. Journal of Managerial Psychology, 26(4), 313–330. https​:/​/do​​i​.org​​/10​.1​​108​/ 0​​26839​​411​11​​12484​5

Multilevel Control–Trust Dynamics  115

Korsgaard, M. A., Schweiger, D. M., & Sapienza, H. J. (1995). Building commitment, attachment, and trust in strategic decision-making teams:The role of procedural justice. Academy of Management Journal, 38(1), 60–84. https://doi​.org​/10​.5465​/256728 Ladegard, G., & Gjerde, S. (2014). Leadership coaching, leader role-efficacy, and trust in subordinates. A mixed methods study assessing leadership coaching as a leadership development tool. The Leadership Quarterly, 25(4), 631–646. https​:/​/do​​i​.org​​/10​.1​​016​/ j​​.leaq​​ua​.20​​​14​.02​​.002 Langfield-Smith, K. (1997). Management control systems and strategy: a critical review. Accounting, Organizations and Society, 22(2), 207–232. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​361-​​ 3682(​95)00​040-2​ Langfield-Smith, K., & Smith, D. (2003). Management control systems and trust in outsourcing relationships. Management Accounting Research, 14(3), 281–307. https​:/​/do​​i​. org​​/10​.1​​016​/S​​1​044-​​5005(​03)00​046-5​ Lee, C., Ashford, S. J., & Bobko, P. (1990). Interactive effects of “Type A” behavior and perceived control on worker performance, job satisfaction, and somatic complaints. Academy of Management Journal, 33(4), 870–881. https://doi​.org​/10​.5465​/256296 Lewin, A. Y., Long, C. P., & Carroll, T. N. (1999). The coevolution of new organizational forms. Organization Science, 10(5), 535–550. https://doi​.org​/10​.1287​/orsc​.10​.5​.535 Long, C. P. (2002). Balancing organizational controls with trust-building and fairness-building initiatives (Doctoral dissertation). Duke University. Long, C. P. (2010). Control to cooperation: Examining the role of managerial authority in portfolios of managerial action. In S. B. Sitkin, L. B. Cardinal, & K. Bijlsma-Frankema (Eds.), Organizational control. Cambridge: Cambridge University Press. Long, C. P. (2018). To control and build trust: How managers use organizational controls and trust-building to motivate subordinate cooperation. Accounting, Organizations and Society. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.aos.​​201​8.​​05​.00​6 Long, C. P., Bendersky, C., & Morrill, C. (2011). Fairness monitoring: Linking managerial controls and fairness judgments in organizations. Academy of Management Journal, 54(5), 1045–1068. https://doi​.org​/10​.5465​/amj​.2011​.0008 Long, C. P., Burton, R. M., & Cardinal, L. B. (2002). Three controls are better than one: A computational model of complex control systems. Computational & Mathematical Organization Theory, 8(3), 197–220. https://doi​.org​/10​.1023​/A​:1020767513183 Long, C. P., & Sitkin, S. B. (2006). Trust in the balance: How managers integrate trustbuilding and task control. In R. Bachmann & A. Zaheer (Eds.), Handbook of trust research, Cheltenham: Edward Elgar. Long, C. P., & Sitkin, S. B. (2018). Control–trust dynamics in organizations: Identifying shared perspectives and charting conceptual fault lines. Academy of Management Annals, 12(2), 725–751. https://doi​.org​/10​.5465​/annals​.2016​.0055 Long, C. P., & Weibel, A. (2018). Two sides of an important coin: Outlining the general parameters of control-trust research. In R. Searle, A-M Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust. New York: Routledge Taylor & Francis Group. Lumineau, F. (2017). How contracts influence trust and distrust. Journal of Management, 45(3), 1553–1577. https://doi​.org​/10​.1177​/0149206314556656 Malhotra, D., & Lumineau, F. (2011). Trust and collaboration in the aftermath of conflict: The effects of contract structure. Academy of Management Journal, 54(5), 981–998. https://doi​.org​/10​.5465​/amj​.2009​.0683

116 

Chris Long

Malhotra, D., & Murnighan, J. K. (2002). The effects of contracts on interpersonal trust. Administrative Science Quarterly, 47(3), 534–559. https://doi​.org​/10​.2307​/3094850 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​. 19​​95​.​95​​08080​​335 McAllister, D. J. (1995). Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https://doi​. org​/10​.5465​/256727 McEvily, B., Perrone,V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, 14(1), 91–103. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.9​​1​.128​​14 Mellewigt, T., Madhok, A., & Weibel, A. (2007). Trust and formal contracts in interorganizational relationships—substitutes and complements. Managerial and Decision Economics, 28(8), 833–847. https://doi​.org​/10​.1002​/mde​.1321 Merchant, K. A. (1985). Control in business organization. New York: Prentice Hall. Merchant, K. A., & Otley, D. T. (2006). A review of the literature on control and accountability. Handbook of Management Accounting Research, 2, 785–802. https​:/​/do​​i​.org​​/ 10​.1​​016​/S​​1​751-​​3243(​06)02​013-X​ Merchant, K. A., & Van der Stede, W. A. (2007). Management control systems: Performance measurement, evaluation and incentives. New York: Prentice Hall-Financial Times. Miller, G. (1992). Managerial dilemmas: The political economy of hierarchy. Cambridge: Cambridge University Press. Mintzberg, H. (1979). The structuring of organization: A synthesis of the research. Englewood Cliffs, NJ: Prentice Hall. Mishra, A. K., & Mishra, K. E. (2013). The research on trust in leadership: The need for context. Journal of Trust Research, 3(1), 59–69. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​13​. 77​​1507 Mizrachi, N., Drori, I., & Anspach, R. R. (2007). Repertoires of trust: The practice of trust in a multinational organization amid political conflict. American Sociological Review, 72(1), 143–165. https​:/​/do​​i​.org​​/10​.1​​177​/0​​00312​​2407​0​​72001​​07 Murrell, A. J., Blake-Beard, S., Porter, D. M., & Perkins-Williamson, A. (2008). Interorganizational formal mentoring: Breaking the concrete ceiling sometimes requires support from the outside. Human Resource Management, 47, 275–294. https:// doi​.org​/10​.1002​/hrm​.20212 Nooteboom, B. (1996). Trust, opportunism and governance: A process and control model. Organization Studies, 17(6), 985–1010. https​:/​/do​​i​.org​​/10​.1​​177​/0​​17084​​0696​0​​17006​​05 Ostroff, C., Kinicki, A. J., & Tamkins, M. M. (2003). Organizational culture and climate. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Handbook of psychology: Industrial and organizational psychology. Hoboken, NJ: John Wiley. Ouchi, W. G. (1977). The relationship between organizational structure and organizational control. Administrative Science Quarterly, 22, 95–113. http://dx​.doi​.org​/10​.2307​/ 2391748 Ouchi,W. G. (1978).The transmission of control through organizational hierarchy. Academy of Management Journal, 21(2), 173–192. https://doi​.org​/10​.5465​/255753 Ouchi, W. G. (1979). A conceptual framework for the design of organizational control mechanisms. Management Science, 25(9), 833–848. https://doi​.org​/10​.1287​/mnsc​.25​. 9​.833 Ouchi, W. G. (1980). Markets, bureaucracies, and clans. Administrative Science Quarterly, 25, 129–141. https://doi​.org​/10​.2307​/2392231

Multilevel Control–Trust Dynamics  117

Pearce, J. L., Branyiczki, I., & Bigley, G. A. (2000). Insufficient bureaucracy: Trust and commitment in particularistic organizations. Organization Science, 11(2), 148–162. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​1​.2​​.1​​48​.12​​508 Perrone,V., Zaheer, A., & McEvily, B. (2003). Free to be trusted? Organizational constraints on trust in boundary spanners. Organization Science, 14, 422–439. https​:/​/do​​i​.org​​/ 10​.1​​287​/o​​rsc​.1​​4​.4​​.4​​22​.17​​487 Pillutla, M. M., Malhotra, D., & Murnighan, J. K. (2003). Attributions of trust and the calculus of reciprocity. Journal of Experimental Social Psychology, 39(5), 448–455. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​0​022-​​1031(​03)00​015-5​ Poppo, L., & Zenger, T. (2002). Do formal contracts and relational governance function as substitutes or complements? Strategic Management Journal, 23(8), 707–725. https://doi​. org​/10​.1002​/smj​.249 Puffer, S. M., McCarthy, D. J., & Boisot, M. (2010). Entrepreneurship in Russia and China: The impact of formal institutional voids. Entrepreneurship: Theory and Practice, 34, 441– 467. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1540​​-6520​​.200​9​​.0035​​3.x Reed, M. I. (2001). Organization, trust and control: A realist analysis. Organization Studies, 22(2), 201–228. https://doi​.org​/10​.1177​/0170840601222002 Righetti, F., & Finkenauer, C. (2011). If you are able to control yourself, I will trust you: the role of perceived self-control in interpersonal trust. Journal of Personality and Social Psychology, 100(5), 874. https://doi​.org​/10​.1037​/a0021827 Roberts, J. (2001). Trust and control in Anglo-American systems of corporate governance: The individualizing and socializing effects of processes of accountability. Human Relations, 54(12), 1547–1572. https​:/​/do​​i​.org​​/10​.1​​177​/0​​01872​​670​15​​41200​1 Ross, A. (1994). Trust as a moderator of the effect of performance evaluation style on jobrelated tension: A research note. Accounting, Organizations and Society, 19(7), 629–635. https​:/​/do​​i​.org​​/10​.1​​016​/0​​​361​-3​​682(9​4)900​27-2 Roth, N. L., Sitkin, S. B., & House, A. (1994). Stigma as a determinant of legalization. In S. B. Sitkin & R. J. Bies (Eds.), The legalistic organization. Thousand Oaks: SAGE. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Rus, A., & Iglič, H. (2005).Trust, governance and performance the role of institutional and interpersonal trust in SME development. International Sociology, 20(3), 371–391. https:// doi​.org​/10​.1177​/0268580905055481 Sako, M. (1992). Price, Quality and Trust: Inter-firm Relations in Britain and Japan (No. 18). Cambridge: Cambridge University Press. Schilke, O., & Cook, K. S. (2013). A cross-level process theory of trust development in interorganizational relationships. Strategic Organization, 11(3), 281–303. https://doi​.org​/ 10​.1177​/1476127012472096 Schoorman, F., Mayer, R. C., & Davis, J. (2007). An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32, 344–354. https​:/​/ do​​i​.org​​/10​.5​​465​/a​​mr​.20​​07​​.24​​34841​0 Schulte, M., Ostroff, C., & Kinicki, A. J. 2006. Organizational climate systems and psychological climate perceptions: A cross-level study of climate-satisfaction relations. Journal of Occupational and Organizational Psychology, 79, 654–671. https://doi​.org​/ 10​.1348​/096317905X72119 Searle, R. Nienaber, A, & Sitkin, S. B. (2018). Routledge Companion to Trust. New York: Routledge.

118 

Chris Long

Şengün,A. E., & Wasti, S. N. (2007).Trust, control, and risk:A test of Das and Teng’s conceptual framework for pharmaceutical buyer-supplier relationships. Group & Organization Management, 32(4), 430–464. https://doi​.org​/10​.1177​/1059601106293959. Shapiro, S. P. (1987). The social control of impersonal trust. American Journal of Sociology, 623–658. https://doi​.org​/10​.1086​/228791. Simons, R. (1994). Levers of control. Boston: Harvard Business School Press. Simons, R. (1995). Control in an age of empowerment. Harvard Business Review, 73(2), 80–88. Sitkin, S. B. (1995). On the positive effect of legalization on trust. In R. J. Bies, R. J. Lewicki, & B. H. Sheppard (Eds.), Research on negotiations in organizations. Greenwich, CT: JAI Press. Sitkin, S. B., & Bies, R. J. (1994). The legalistic organization. Thousand Oaks, CA: Sage. Sitkin, S. B., & George, E. (2005). Managerial trust-building through the use of legitimating formal and informal control mechanisms. International Sociology, 20(3), 307–338. https:// doi​.org​/10​.1177​/0268580905055479. Sitkin, S. B., & Roth, N. L. (1993). Explaining the limited effectiveness of legalistic “remedies” for trust/distrust. Organization Science, 4, 367–392. https://doi​.org​/10​.1287​/ orsc​.4​.3​.367 Sitkin, S. B., Long, C. P., & Cardinal, L. B. (2020). Assessing the control literature: Looking back and looking forward. Annual Review of Organizational Psychology and Organizational Behavior. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​-org​​psych​​-0121​​​19​-04​​5321. Skinner, E. A. (1996). A guide to constructs of control. Journal of Personality and Social Psychology, 71(3), 549. https://doi: 10.1037/0022-3514.71.3.549 Snell, S. A. (1992). Control theory in strategic human resource management:The mediating effect of administrative information. Academy of Management Journal, 35(2), 292–327. https://doi​.org​/10​.5465​/256375 Son, H. (2015, April 8). JPMorgan algorithm knows you’re a rogue employee before you do. Bloomberg. Retrieved from http:​/​/www​​.bloo​​mberg​​.com/​​news/​​artic​​les​/2​​015​-0​​4​08/​​jpmor​​gan​-a​​lgori​​thm​-k​​nows-​​you​-r​​e​-a​-r​​ogue-​​​emplo​​yee​-b​​efore​​-you-​​do Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions, measurement, and validation. Academy of Management Journal, 38(5), 1442–1465. https:// doi​.org​/10​.2307​/256865 Spreitzer, G. M., & Mishra, A. K. (1999). Giving up control without losing control trust and its substitutes’ effects on managers’ involving employees in decision-making. Group & Organization Management, 24(2), 155–187. https://doi​.org​/10​.1177​/ 1059601199242003 Stahl, G. K., Larsson, R., Kremershof, I., & Sitkin, S. B. (2011).Trust dynamics in acquisitions: A case survey. Human Resource Management, 50(5), 575–603. https://doi​.org​/10​.1002​/ hrm​.20448 Strickland, L. H. (1958). Surveillance and trust. Journal of Personality, 26(2), 200–215. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-6494​​.1958​​​.tb01​​580.x​ Sydow, J. (1998). Understanding the constitution of interorganizational trust. In C. Lane & R. Bachmann (Eds.), Trust within and between organizations: Conceptual issues and empirical applications. Oxford: Oxford University Press. Tan, J., Yang, J., & Veliyath, R. (2009). Particularistic and system trust among small and medium enterprises: A comparative study in China’s transition economy. Journal of Business Venturing, 24, 544–557. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jbus​​vent.​​20​08.​​05​.00​7

Multilevel Control–Trust Dynamics  119

Thompson, S. C. (1981). Will it hurt less if I can control it? A complex answer to a simple question. Psychological Bulletin, 90(1), 89. https​:/​/do​​i​.org​​/10​.1​​037​/0​​033​-2​​90​9​.9​​0​. 1​.8​9 Tyler, T. R., & Lind, E. A. (1992). A relational model of authority in groups. Advances in Experimental Social Psychology, 25, 115–191. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​065-​​2601(​ 08)60​283-X​ Van Dijke, M., & Verboon, P. (2010). Trust in authorities as a boundary condition to procedural fairness effects on tax compliance. Journal of Economic Psychology, 31(1), 80– 91. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.joep​​.200​9​​.10​.0​​05 Van Maanen, J., & Schein, E. H. (1979). Toward a theory of organizational socialization. In B. W. Staw & L. L. Cummings (Eds.), Research in organizational behavior. Greenwich, CT: JAI Press. Vélez, M. L., Sánchez, J. M., & Álvarez-Dardet, C. (2008). Management control systems as inter-organizational trust builders in evolving relationships: Evidence from a longitudinal case study. Accounting, Organizations and Society, 33(7), 968–994. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.aos.​​200​8.​​02​.00​6 Wanous, J. P. (1980). Organizational entry: Recruitment, selection and socialization of newcomers. Reading, MA: Addison-Wesley. Weber, J. M., Malhotra, D., & Murnighan, J. K. (2004). Normal acts of irrational trust: Motivated attributions and the trust development process. Research in Organizational Behavior, 26, 75–101. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​191-​​3085(​04)26​003-8​ Weibel, A. (2007). Formal control and trustworthiness shall the twain never meet? Group & Organization Management, 32(4), 500–517. https://doi​.org​/10​.1177​/ 1059601106293961 Weibel, A., Den Hartog, D. N., Gillespie, N., Searle, R., Six, F., & Skinner, D. (2015). How do controls impact employee trust in the employer? Human Resource Management, 22(5), 1069–1092. https://doi​.org​/10​.1002​/hrm​.21733 Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. (2008). Organizing for high reliability: Processes of collective mindfulness. Crisis Management, 3(1), 81–123. http://dx​.doi​.org​/ 10​.4135​/9781446261262 Werbel, J. D., & Lopes Henriques, P. (2009). Different views of trust and relational leadership: Supervisor and subordinate perspectives. Journal of Managerial Psychology, 24(8), 780–796. https​:/​/do​​i​.org​​/10​.1​​108​/0​​26839​​409​10​​99679​8 Whitener, E. M. (2001). Do “high commitment” human resource practices affect employee commitment? A cross-level analysis using hierarchical linear modeling. Journal of Management, 27(5), 515–535. https​:/​/do​​i​.org​​/10​.1​​177​/0​​14920​​6301​0​​27005​​02 Whitener, E. M., Brodt, S. E., Korsgaard, M. A., & Werner, J. M. (1998). Managers as initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23(3), 513–530. https://doi​.org​/ 10​.5465​/amr​.1998​.926624 Williamson, O. E. (1975). Markets and Hierarchies. New York: Free Press. Woolthuis, R. K., Hillebrand, B., & Nooteboom, B. (2005). Trust, contract and relationship development. Organization Studies, 26(6), 813–840. https://doi​.org​/10​.1177​/ 0170840605054594 Yeung, J. H. Y., Selen, W., Zhang, M., & Huo, B. (2009). The effects of trust and coercive power on supplier integration. International Journal of Production Economics, 120(1), 66– 78. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.ijpe​​.200​8​​.07​.0​​14

120 

Chris Long

Zaheer, A., McEvily, B., & Perrone, V. (1998). Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9(2), 141–159. https://doi​.org​/10​.1287​/orsc​.9​.2​.141 Zaheer, S., & Zaheer, A. (2006). Trust across borders. Journal of International Business Studies, 37(1), 21–29. https​:/​/do​​i​.org​​/10​.1​​057​/p​​algra​​ve​.ji​​​bs​.84​​00180​ Zucker, L. (1986). The production of trust: Institutional sources of economic structure, 1840–1920. Research in Organizational Behavior, 8, 53–11.

6 TRUST ME OR US? A Multilevel Model of Individual and Team Felt Trust by Supervisors Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Introduction Interpersonal trust is central to effective management in organizations (Yang & Mossholder, 2010). Often, decisions to trust do not take place in a social vacuum but are constructed through the interactions between trustors and trustees (Konecki, 2019). As a result, to develop trust relationships, understanding the rationale for both “trusting” and “feeling trusted” is critical (Lau, Liu, & Fu, 2007). In organizations, supervisor trust has an important role in shaping subordinates’ attitudes and behaviors, and managers who initiate trusting relationships with their subordinates are thus often more effective than those who do not (Fehr, Fulmer, & Keng-Highberger, 2020; Lau & Liden, 2008; Nerstad et al., 2018;Whitener, Brodt, Korsgaard, & Werner, 1998). In this chapter, our discussion focuses on employees’ perception of being trusted (or ‘felt trust’) by their supervisor. Research on trust argues that interpersonal trust can be distinguished by cognition-based trust and affect-based trust, which derive from different bases of trust: cognitive judgments versus emotional ties (e.g., McAllister, 1995). To trust, trustors need to have a reason. By contrast, to feel trusted, trustees just need to perceive the counterpart’s willingness to take risks, regardless of the bases of trust. Consistent with this notion, extant literature on felt trust does not distinguish between cognition- and affect-based felt trust (e.g., Baer et al., 2015; Lau, Lam, & Wen, 2014; Lau et al., 2007). We aim to fill this gap by discussing ‘felt trust by supervisor’ as a unidimensional construct involving perception. Trust in supervisors is an important determinant of desired outcomes, such as enhanced work performance, organizational citizenship behavior, and reduced counterproductive behaviors (Colquitt, Scott, & LePine, 2007; Dirks & Ferrin, 2001). While many studies have focused on trust in supervisors, relatively few DOI: 10.4324/9780429449185-6

122  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

have examined felt trust by supervisor or how it develops and functions (Lau et al., 2007; Salamon & Robinson, 2008). This dearth is unfortunate because trust and felt trust are related but not necessarily equivalent in hierarchical relationships (Brower, Schoorman, & Tan, 2000; Lau et al., 2007, 2014). If trust cannot be perceived, it will not be influential. Moreover, previous research has demonstrated that being trusted by a supervisor has an independent and significant influence on employee performance through different mechanisms over and above trust in the supervisor (Brower, Lester, Korsgaard, & Dineen, 2009). Organizations are multilevel systems, and the complexity of multilevel issues is more pronounced in organizations with tall hierarchies. However, within the topic of felt trust, discussion on the multilevel nature of felt trust is lacking. While there are many ways to examine the multilevel effects of felt trust, we focus on two forms at the trustee level: individual and team felt trust by supervisor. Individual felt trust by supervisor refers to an employee’s perception that his or her supervisor is willing to assume risk given positive expectations of the employee (Lau et al., 2014). Team felt trust by supervisor refers to an emergent and dynamic shared psychological perception among team members of the extent to which their supervisor is willing to accept vulnerability given positive expectations of the whole team. It takes time for individuals’ perceptions to converge as a shared team property (Feitosa, Grossman, Kramer, & Salas, 2020). Although felt trust shares similar meaning at both levels, individual and team felt trust by supervisor emphasize different referents (i.e., individual and team). Given such difference, team and individual felt trust by supervisor are likely to be distinct in terms of antecedents and outcomes. At the individual level, felt trust by supervisor derives from a series of interpersonal exchanges between the supervisor and the focal subordinate and is likely to influence individual behaviors (Lau et al., 2014).The perception of being trusted by the supervisor as a team, by contrast, may be influenced by supervisors’ practices in managing team processes and performance (Salamon & Robinson, 2008). With different referent and supervisor practices, team felt trust occurs above and beyond the mere aggregation of individual felt trust. Management needs to understand the functions of felt trust at different levels so it can design effective interventions and practices. In addition, within-team dynamics and potential variance of sharedness in terms of leader trust perceptions add another layer of complexity to the construct of team felt trust. To determine the antecedents, outcomes, and cross-level effects of multilevel felt trust, we adopt social information processing and social identity perspectives. Supervisors’ sense-giving activities deliver important cues to subordinates to develop expected work patterns that promote adaptation in the work environment (Zaccaro, Rittman, & Marks, 2001). Trust acts as a social cue in supervisor–subordinate dyads, providing information about expectations of interaction patterns. As such, social information processing theory (Salancik & Pfeffer, 1978), which assumes that employees’ work attitudes and behaviors are cognitive products of their immediate social environment, is an important theoretical perspective in trust-related management research (Lau & Liden, 2008).

Multilevel Felt Trust  123

Supervisory trusting behavior at the team level -Delegation (e.g., devolution, decentralization)

Team felt trust by supervisor

-Monitoring removal (e.g., low formalization) Supervisory trusting behavior at the individual level -Delegation (e.g., authority delegation) -Monitoring removal (e.g., less micromanaging behaviors, use of less surveillance)

FIGURE 6.1 Conceptual

Team identification

Team performance

Team prototypicality

Individual felt trust by supervisor

Individual performance

framework.

In examining the relationship between individual and team felt trust by supervisor, we adopt the social identity perspective because identities of and identification with individuals and teams serve as potential connections between felt trust at different levels (Sluss & Ashforth, 2007). Prior studies on social influence, identity, and group decision-making have demonstrated the importance of social group membership and categorization in shaping attitudes and beliefs (e.g., Coats, Smith, Claypool, & Banner, 2000; Turner, 1991). We extend this line of research by exploring the interplays of multilevel felt trust by taking two team factors into consideration: team identification and team prototypicality (Coats et al., 2000; Hains, Hogg, & Duck, 1997). When individuals’ concepts of self and team membership overlap (i.e., “I identify with my team as a whole” or “This member is an embodiment of our team”), the social cues used to evaluate one referent are significant for the judgment of the other referent (Coats et al., 2000). In this chapter, we review the literature on felt trust and develop a conceptual framework that delineates how supervisors manage individual subordinates and their teams by making them feel trusted. First, we examine the antecedents and outcomes of felt trust by supervisor at the individual and team level separately. In doing so, we aim to enrich the understanding of the antecedents and consequences of felt trust within the same level and to test whether their nomological networks (i.e., at the individual and team levels) are similar and parallel. Second, we explore the cross-level interactions of individual and team felt trust (i.e., whether individual and team felt trust are related or under what conditions these constructs, at different levels, are related). Drawing on the social identity perspective, we propose two conditions through which felt trust by supervisor can be transferred between individuals and teams. Figure 6.1 depicts the proposed model.

Felt Trust Literature Review Prior trust research indicates that unlike leader–member exchange, trust is not necessarily mutual and reciprocal (Brower et al., 2000; Schoorman, Mayer, &

124  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Davis, 2007). Although mutual trust is important, most studies have focused on only one side of the dyad – trustors’ perceptions (Brower et al., 2009); the perspective of trustees is relatively less studied (for exceptions, see Baer et al., 2015; Lau et al., 2014). As trust itself is internal and cannot be observed directly, trustees may not know that they are trusted. Indeed, as felt trust reflects a person’s attempt to understand another person’s trust in him or her, an individual’s felt trust may not correspond to the counterpart’s actual trust in him or her (Campagna, Dirks, Knight, Crossley, & Robinson, 2019). In addition, even when trust is transferred into trusting behavior, it is subject to interpretation.Trust misinterpretation sometimes occurs from the difficulty in recognizing the motives behind the seemingly trusting behaviors (Lau & Lam, 2008). Most studies on interpersonal trust in vertical dyads have focused on trust in the supervisor or leader. Specifically, previous research (e.g., Aryee, Budhwar, & Chen, 2002; Chughtai, Byrne, & Flood, 2015; Holtz & Harold, 2008; Whitener et al., 1998) indicates that leadership styles and supervisory practices, such as transformational leadership, procedural and distributive justice, participative decisionmaking, and perceived organizational support, increase employees’ trust in leaders; in addition, research has shown that supervisory trust is associated with a wide range of attitudinal (e.g., job satisfaction, commitment), behavioral, and performance (e.g., organizational citizenship behavior, job performance) outcomes (for a review, see Dirks & Ferrin, 2001). Despite the abundance of research on trust, understanding of felt trust is limited. Only a few empirical studies, based on a social-exchange explanation, self-evaluative explanation, and normative appropriateness explanation, have demonstrated that felt trust positively influences organization-based selfesteem, organizational citizenship behavior, and job performance (Lau & Lam, 2008; Lau et al., 2014; Lester & Brower, 2003; Salamon & Robinson, 2008). According to conservation of resources theory, felt trust can be a double-edged sword, filling employees with pride while also bringing burdens (Baer et al., 2015). On the one hand, feeling trusted can lead to perceptions of increased workload; on the other hand, as feeling trusted is a signal of a positive reputation, trusted employees also experience heightened concerns about keeping the positive views of others intact (i.e., not to disappoint the trustor). Both perceived workload and the constant need to maintain one’s reputation are taxing and drain personal resources, such that felt trust can have a negative effect on job performance through emotional exhaustion. For example, Baer, Frank, Matta, Luciano, and Wellman (2020) show that employees who feel undertrusted or over-trusted by supervisors perceive the supervisors as less fair and therefore engage in less in-role and extra-role performance. As a dependent variable, felt trust by supervisor is likely to be influenced by supervisors’ moral leadership and autocratic leadership behaviors, supervisor–subordinate value congruence, demographic differences, and mastery climate (Lau et al., 2007; Nerstad et al., 2018).

Multilevel Felt Trust  125

Multilevel Felt Trust: Same-Level Examination At the individual level, trust and felt trust occur in a dyadic relationship between a supervisor and a subordinate. The focus of individual felt trust by supervisor is on specific employees’ perceptions of whether they are trusted by the supervisor. Supervisors decide to trust when vulnerability is offset by positive expectations of subordinates’ benevolence, integrity, or/and ability (Mayer, Davis, & Schoorman, 1995). Conversely, the team as an independent social entity is the primary focus of team felt trust. Thus, “we are trusted” and “I am trusted” are different. When team members perceive their supervisor as trusting the whole team, they all share the belief that the supervisor takes risks on the team relationship out of positive expectations that team members will cooperate efficiently and act responsibly and ethically to advance organizational goals. Team felt trust by supervisor as a collective belief is conceptualized and analyzed in referent-shift consensus compositional models (i.e., representing the aggregated degree of “we are trusted by the supervisor”) as having sufficient consensus among team members (Chan, 1998). By shifting the referent from individuals to teams, team felt trust is no longer the mere simple addition or average of individual felt trust. That is, rather than an individual’s own felt trust perception, team felt trust by supervisor is the extent to which every team member believes that the team as a whole has garnered the supervisor’s trust and whether there is within-team consensus in such beliefs. We contend that the development and impact of felt trust at the individual and team levels may be similar but not completely isomorphic. In the following section, we draw on social information processing theory to develop propositions of the antecedents and outcomes of felt trust at the individual and team level separately.

Social Information Processing and Felt Trust To adapt to the social context and maintain acceptance by others, people process and internalize social information from significant interactions and then delimit the proper types of behaviors accordingly (Baldwin, 1992; Salancik & Pfeffer, 1978). In the workplace, supervisors are important sources of social information due in part to their higher formal status (Lau & Liden, 2008; Salancik & Pfeffer, 1978). A critical way that supervisors communicate information is through their management behaviors. In daily interactions, subordinates observe their supervisors’ behaviors and draw inferences about what behaviors or attitudes are appropriate in the work context (Zaccaro et al., 2001). Previous research has recognized that information processors can be individuals or teams (De Dreu, Nijstad, & van Knippenberg, 2008; Hinsz, Tindale, & Vollrath, 1997). For example, Fehr et al. (2020) show that individual employees may evaluate whether their supervisors are trustworthy by interpreting their behaviors as ethical or unethical. The appropriateness of collective behavior, including whether voice is invited in teams, can also be delivered and clarified through supervisor behaviors, such as self-disclosure,

126  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

noting awareness of their own fallibility, and developing a nonthreatening atmosphere (Edmondson, 2003). Trust research suggests that supervisor trust creates a context in which employees interpret their roles in the supervisory relationship (Davis, Schoorman, & Donaldson, 1997; Hernandez, 2008). A supervisor’s expectations of and attitudes toward his or her subordinates influence the interpretation of the trust relationship. Employees who feel trusted believe that their supervisors assume risks and have positive expectations of their relationship (Lau et al., 2007). Supervisors deliver trust information through their observable supervisory behaviors. In their daily interactions, subordinates keenly attend to their supervisors’ trusting behaviors, interpret the behavioral cues to construct the perception of being trusted, and then behave appropriately in response to the trust (Brass & Burkhardt, 1993).

Supervisory Trusting Behavior Trust and trusting behaviors are distinct but related. Trust refers to the willingness to assume vulnerability while trusting behavior involves a trustor’s act of risk taking (Mayer et al., 1995). In this chapter, we focus on ‘supervisory trusting behavior’ and define it as behavior that communicates a supervisor’s acceptance of vulnerability and the positive expectations of subordinates’ intentions and behaviors. We propose two categories of supervisory trusting behaviors: delegation and monitoring removal. The former involves transferring decision-making authority, while the latter entails removing constraints. Both types of supervisory behaviors inherently involve relinquishing control, which requires supervisors to assume risks. These two behavior types also relate to reliance, because supervisors rely on subordinates’ skills, knowledge, and judgments to complete tasks satisfactorily (Gillespie, 2003). As the target for felt trust can be an individual employee or a group (Campagna et al., 2019), we describe supervisory behaviors at these two levels accordingly.

Delegation Delegation is an important aspect of trust (Baer et al., 2015; Brower et al., 2009). Delegation to individuals differs from delegation to teams, as the former is person-focused while the latter is both member- and team-focused. When supervisors delegate responsibility to individuals, they rely on individuals’ knowledge, skills, and abilities. When supervisors delegate responsibility to teams, they also take into account team effectiveness (e.g., conflict resolution, collaborative problem-solving, communication, coordination), as well as team members’ knowledge, skills, and abilities (Janssen, Van de Vliert, & West, 2004). Team- and individuallevel delegation are two distinct decisions, with the former involving shared decision-making and responsibility of teams and the latter dealing with decisions made individually (Kirkman & Rosen, 1999).

Multilevel Felt Trust  127

At the individual level, delegation entails assigning important responsibilities to individuals, giving them additional authority, and involving them in decisionmaking (Bass, 1990; Yukl, 1998). For example, a supervisor will permit a trusted assistant to handle administrative tasks by giving access to his or her private calendar and empowering the assistant to manage the schedule without consultation. When delegating authority, the supervisor assumes risks, as the subordinate may (1) behave opportunistically by taking advantage of reduced control and (2) handle tasks poorly due to incompetence. At the team level, delegation refers to a structure of devolution and decentralization in which teams are assigned important tasks and team members participate in decision-making and have collective authority and responsibility to execute and manage teamwork processes and performance (Richardson,Vandenberg, Blum, & Roman, 2002; Wageman, 2001). For example, a basketball coach may provide autonomy to team members to make training plans themselves and determine problem-solving approaches collaboratively and collectively.The risks of delegation are also inherent in team management, possibly resulting in (1) ineffective cooperation and coordination and (2) poor decision-making about teams’ marketing, development, and resource allocation. As team decisions and performance are highly visible, poor teamwork may have serious negative implications for both the supervisor and the team.

Monitoring We also theorize that less supervisory monitoring may further develop employees’ perception of felt trust. Monitoring refers to supervisory activities, including “the observation, examination, or recording of employee work related behaviors (or all of these), with and without technological assistance” (Stanton, 2000, p. 87). For example, supervisors may specify rules and standardize procedures to regulate employee behavior (Hirst, Van Knippenberg, Chen, & Sacramento, 2011). With the assistance of technological tools, such as electronic surveillance, they may also detect employees’ misconducts and record employees’ working hours and Internet usage (Spitzmüller & Stanton, 2006). In this chapter, we emphasize control-based monitoring, which is used to reduce deviance and lower costs, rather than needbased monitoring, which is used to gain tacit awareness of coworkers’ needs (see McAllister, 1995).While monitoring is an effective tool in improving productivity and containing costs (Holman, Chissick, & Totterdell, 2002; Martin & Freeman, 2003), it can have a negative impact on trust (De Jong & Dirks, 2012; Malhotra & Murnighan, 2002; Webber, 2008). As such, reduced monitoring may signify to employees that they are trusted by supervisors. Reducing monitoring of individuals also differs from that of teams. The former focuses on whether and how individuals complete their individual work tasks, while the latter focuses on whether and how team members coordinate and synchronize to achieve collective goals. At the individual level, supervisors may offer subordinates more freedom by engaging in less micromanaging behaviors and using fewer surveillance

128  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

technologies.They may also provide fewer specific directions for actions and procedures and thus encourage employees to engage in more discretionary behaviors without constraints (Spreitzer, De Janasz, & Quinn, 1999). For example, operators at Michelin can set their own work schedules and vacations, and design and monitor their own performance indicators (Carney & Getz, 2018). However, reducing monitoring can also lead to risks. Individual employees with less monitoring may engage in deviant and counterproductive behaviors, such as fraud, theft, and corruption (Pierce, Snow, & McAfee, 2015). At the team level, supervisors may reduce structural formalization by using less detailed policies and procedures and removing some bureaucratic constraints (Hirst et al., 2011). For example, after the purchase of laboratory equipment, a supervisor may simplify overly complicated reporting procedures and abandon inflexible reimbursement rules for teams to follow. In addition, supervisors can reduce the team monitoring climate, in which team members closely watch for teammates’ errors or performance discrepancies (Marks & Panzer, 2004), and thus encourage a more flexible environment for team members to execute their interdependent roles to reach team goals. However, reducing monitoring is also risky at the team level. Low team formalization and monitoring may result in social loafing, ineffective team cooperation, and unsatisfactory levels of team performance (Aiello & Kolb, 1995; George, 1992). Whether supervisors delegate authority and reduce monitoring depends on their confidence in individual subordinates’ or teams’ “capability, trustworthiness, and motivation to assume greater responsibility” (Leana, 1986, p. 762). Given the risks, delegating authority and reducing monitoring are more likely to occur when supervisors have positive expectations of the individuals or teams. For both individuals and teams, supervisory trusting behaviors provide the opportunity to gain more control over their work, which signals that their supervisors consider them trustworthy (Chen & Aryee, 2007; Gardner,Van Dyne, & Pierce, 2004; Pierce & Gardner, 2004). As such, we expect a high level of delegation and a low level of monitoring, as the manifestations of trust, to deliver social information about trust, thereby leading to enhanced felt trust. Proposition 1: Delegation is positively related to (1) individual felt trust by supervisor and (2) team felt trust by supervisor. Proposition 2: Monitoring is negatively related to (1) individual felt trust by supervisor and (2) team felt trust by supervisor.

Felt Trust and Performance Individual performance refers to employees’ efforts in tasks, duties, and responsibilities as formalized in job descriptions (Williams & Anderson, 1991). Previous

Multilevel Felt Trust  129

research suggests that supervisors play an important role in observing, appraising, and rewarding employees’ formal work activities (Becker & Kernan, 2003). We argue that felt trust by supervisor is positively related to individual employees’ job performance. People are social actors who learn to behave appropriately in social environments (Messick, 1999). In organizations, employees usually draw inferences of their roles and obligations by identifying their supervisors’ expectations and then engaging in proper responses (Baldwin, 1992). In trusting relationships, trustors expect their trustees to be competent and responsible, to not exploit their exposed vulnerability, and to live up to their expectations (Lau et al., 2014). As such, when subordinates recognize that they are trusted by their supervisors, they tend not to violate the expectations and act to complete the assigned work tasks conscientiously and responsibly (Deutsch, 1958; Salamon & Robinson, 2008). In addition, supervisor trust conveys a compliment about employees’ competence, integrity, and benevolence (Mayer et al., 1995). Trusted subordinates are likely to report positive self-evaluations of their competence and moral worth (Lau et al., 2014; Pfeffer, 1998). To maintain their reputation and the feeling that they are valued in the workplace, employees will exert significant efforts to enhance their performance (Baer et al., 2015). Proposition 3: Individual felt trust by supervisor is positively related to individual performance. Team performance refers to the extent to which the team effectively executes the required, intended, and expected tasks or duties (Chatman & Flynn, 2001; Walumbwa, Morrison, & Christensen, 2012). Compared with individual performance, team performance requires effective coalitions among team members, above and beyond the sum of individual performance (Walumbwa et al., 2012). While individual and team performance are two distinct concepts, previous studies have shown that they are positively correlated (Chen, 2005; Chen, Kirkman, Kanfer, Allen, & Rosen, 2007). High performance norms in teams can lead to better individual performance, and individual performance of team members may sometimes aggregate to influence the collective performance (Chen et al., 2007). Some studies, however, have cautioned that without effective coordination, individuals’ performance may not be able to converge into team performance collectively within the team (Cohen & Bailey, 1997; Li & Liao, 2014). When employees believe that their supervisor trusts their teams, they will likely develop responsibility norms and assign greater importance to the teams’ interests (Salamon & Robinson, 2008). In other words, trusted team members will collectively adopt an “ethical” cognitive frame and have a sense of obligation to act appropriately and monitor one another to ensure no one damages the supervisor’s trust (Salamon & Robinson, 2008). Specifically, they will make an effort

130  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

to advance collective goals by working to restrain detrimental acts, such as social loafing, and providing assistance to teammates when needed. Proposition 4:Team felt trust by supervisor is positively related to team performance.

Multilevel Felt Trust: Cross-Level Examination Team felt trust is not a simple addition of individual felt trust in a team, and individual felt trust by supervisor and team felt trust by supervisor are not automatically related. It is plausible that a supervisor may trust only a few team members but not the whole team. For example, most sales teams consist of only one or two star salespeople; the rest are average performers, and a few may even be laggards. In extreme cases, the few star salespeople may carry the whole team to meet set targets. When this happens, supervisors are likely to delegate more to those individuals than to the team as a whole. Even if all members have good individual performance or are trustworthy, this may not necessarily lead to a perceived trustworthy team (Klein & Kozlowski, 2000), because team success relies on team members’ collective efforts and synergy (Zaccaro et al., 2001) and effective interactions among individuals (Levine, Resnick, & Higgins, 1993). As such, supervisors’ positive expectations of teams are less about individual members’ competence, integrity, and benevolence and more about teams’ cohesion and collective ability. For example, successful delivery of a wedding banquet service in a hotel usually relies heavily on seamless integration of individual employees’ efforts. Individual members of a banquet team assume different roles and positions. For example, several staff members are likely to work in the kitchen preparing food, while pantry helpers and waiters take care of services.The work is intensely interdependent, and everyone must work closely together for the banquet to run smoothly and meet customers’ expectations. In such teams, fostering effective teamwork rather than one best performer is therefore important. Wedding banquets can also confront unexpected events that can easily disrupt the planned schedule (e.g., uninvited guests coming to surprise the bride and groom). While individual employees may still feel trusted, it is more important for team members to collectively feel trusted by the banquet supervisor to handle contingencies timely and effectively. In summary, as felt trust is a personal perception of a specific trust referent, it does not transfer easily between different referents or across levels (i.e., individual and team). Research on social identity suggests that team membership is not simply a context for individuals’ activities but is a part of individual psychology that critically shapes the way members perceive and interact with each other (Turner & Haslam, 2001).Thus, it is important to consider how the perception of felt trust shifts for these different referents. In particular, we adopt a social identity perspective to explain the possible situations in which individual felt trust and team felt trust influence each other.

Multilevel Felt Trust  131

Social Identity Theory and Felt Trust Social identity is “that part of an individual’s self-concept which derives from his knowledge of his membership of a social group (or groups) together with the value and emotional significance attached to that membership” (Tajfel, 1978, p. 63), and identification is about “the perception of oneness” between the individual and the social group (Ashforth & Mael, 1989, p. 21). When someone highly identifies with a particular social entity, the “person’s self-concept contains the same attributes as those in the perceived organizational/team identity” (Dutton, Dukerich, & Harquail, 1994, p. 239). Social identity theorists argue that people develop identification because doing so addresses various selfrelated needs such as self-enhancement and uncertainty reduction (Ashforth, Harrison, & Corley, 2008). Tajfel (1972) in particular suggests that people are motivated to seek positive social identity so that they can internalize team positive attributes as part of their selves (Sluss & Ashforth, 2007). When team attributes are distinct and desirable, members are likely to identify with their teams (Turner & Haslam, 2001). Identities in organizational contexts comprise values, goals, beliefs, stereotypic traits, knowledge, skills, and abilities that are central and distinctive (Ashforth et al., 2008). According to social identity theory, “identification implies an acceptance of those attributes as one’s own” and “the more an individual actually embodies those attributes, the more prototypical he or she is said to be” (Ashforth et al., 2008, p. 330). Specifically, prototypicality is the extent to which an individual is perceived as being representative of a group identity, so highly prototypical members embody the team more (Sluss & Ashforth, 2007; Sluss, Ployhart, Cobb, & Ashforth, 2012).We argue that both identification and prototypicality can influence the cross-level spillover effects of individual and group felt trust by supervisor. Specifically, we propose a top-down process, in which the team’s felt trust by supervisor (a collective concept) influences a member’s felt trust by supervisor (an individual concept), and also a bottom-up process, in which a member’s felt trust by supervisor affects the team’s felt trust by supervisor.

Top-Down Process Team identification is a specific type of social identification in which team members define themselves in terms of their membership in a particular team (Mael & Ashforth, 1992; Somech, Desivilya, & Lidogoster, 2009). If a member identifies strongly with his or her team, team identity will become part of his or her self-concept. As mentioned previously, as positively evaluated team membership is particularly appealing, employees are likely to draw on it to represent their selfconcept (Turner & Tajfel, 1986). Identification can spill over across levels of analyses (Sluss & Ashforth, 2007; Sluss et al., 2012). When individuals identify with a particular referent, sometimes

132  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

such identification may extend to another referent. The transfer between individual felt trust by supervisor and team felt trust by supervisor is likely to occur when the cognitive representations of the two referents (i.e., self and in-group) overlap or are closely related. We contend that in the case of high team identification, team felt trust by supervisor may engender individual felt trust by supervisor. Employees who strongly identify with their teams tend to evaluate themselves on the basis of ‘we’ rather than ‘I’ (Van Knippenberg, 2011) and internalize team characteristics and experience as their own (Mael & Ashforth, 1992; Mael & Tetrick, 1992). For example, individuals with high team identification working in teams of high collective efficacy may consider themselves very capable (i.e., high self-efficacy); however, when teams are not an important source of self-concept, individuals are less likely to perceive themselves as competent even if their teams perform well. Trust from supervisors represents positive evaluation and connotations, as it indicates capability, benevolence, and integrity (Colquitt et al., 2007). When the concept of the self is closely intertwined with that of the team, individuals necessarily consider the holistic appraisal of the teams in developing their personal evaluation (Fernández-Ballesteros, Díez-Nicolás, Caprara, Barbaranelli, & Bandura, 2002). In this case, as team membership is critical for self-concept, the experience of being trusted as a team will lead employees to believe that they are also individually trusted by their supervisors. By contrast, if employees dissociate themselves from their teams, they are less likely to pay attention to team-related cues such as team felt trust when evaluating themselves. Proposition 5: The positive relationship between team felt trust by supervisor and individual felt trust by supervisor is stronger when team identification is high versus low.

Bottom-Up Process Again, prototypicality refers to the extent to which an employee is representative of the team characteristics that define the team in comparison with other teams (Barreto & Hogg, 2017; Epitropaki, Kark, Mainemelis, & Lord, 2017; Giessner, van Knippenberg, van Ginkel, & Sleebos, 2013; Van Knippenberg, 2011). Team identity, as an amalgam of perceived team characteristics (e.g., values, beliefs, abilities), is abstract but can be observed by how the prototypical member acts and what he or she thinks and perceives (Ashforth et al., 2008). Some team members can be more prototypical than others by embodying and personifying the collective concept of the team (Sluss et al., 2012). The more prototypical the member, the more he or she exemplifies the standards, values, and norms the team members have in common. Research, for instance, has described a person as prototypical of Google if he or she, “as a member of Google, is bold, innovative, and proactive” (Ashforth et al., 2008, p. 330).

Multilevel Felt Trust  133

Because prototypical team members are more central than others as a representative of team identity, they may also emerge as team leaders (van Knippenberg, van Knippenberg, & van Dijk, 2000), and prototypicality can further make these leaders socially attractive and enhance their leadership effectiveness and influence (Hogg & van Knippenberg, 2003; van Knippenberg & Hogg, 2003; van Knippenberg & van Knippenberg, 2005). Previous research indicates that followers are more likely to endorse team prototypical individuals than others and comply with their suggestions and requests (e.g., Giessner, van Knippenberg, & Sleebos, 2009; Giessner et al., 2013; van Knippenberg & van Knippenberg, 2005). According to social identity theory, individuals tend to judge themselves and other members by how prototypical they are in the team (Ashforth et al., 2008). Prototypical members may serve as important carriers of trust information who may later influence the development of collective felt trust by supervisor among team members. When a prototypical member feels trusted by the supervisor, he or she believes that the supervisor also trusts the whole team. As the trusted prototypical member is a representative of the team’s standards, values, and norms, the supervisor’s trustworthiness assessments of the prototypical member and the whole team are likely to be similar. Prototypical members are trusted as a reliable source of identity-related information, and other members will pay close attention to, be influenced by, and endorse and support them (Barreto & Hogg, 2017). When prototypical members feel trusted by their supervisors and believe that their supervisors trust the group, other members will likely endorse and share similar thoughts. In addition, prototypical members will likely be motivated to unify other members’ perceptions because doing so might be a good way to maintain their prototypical status and advantage in the team. As mentioned previously, prototypicality is built on cognitive similarity within a team (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). If team members differ in their perceptions of the team characteristics, prototypicality will not exist. Besides, being highly prototypical allows team members to be “entrepreneurs of identity” who can manage perceptions of the team prototypes and, thus, the team identity (Barreto & Hogg, 2017). With such motivation and ability, prototypical members may proactively instill a sense of collective felt trust by transferring a positive affect, such as pride (Baer et al., 2015), among team members, and casting the supervisor’s attitudes in a positive light. Taken together, as prototypical members are deemed to embody the team (Barreto & Hogg, 2017), other members are likely to attribute their supervisor’s trust in prototypical members to the team. From this, individual felt trust by supervisor will generalize to team felt trust by supervisor when the focal individual is exemplary in the team. Proposition 6: The positive relationship between individual felt trust by supervisor and team felt trust by supervisor is stronger when the trusted individual has high versus low team prototypicality.

134  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Team Felt Trust by Supervisor and Individual Performance Trust researchers indicate that trustees are likely to take collective goals as a priority and engage in more cooperative behaviors to protect the trustor from threats (Davis et al., 1997; Hernandez, 2008). Collective felt trust leads team members to develop shared responsibility norms (Salamon & Robinson, 2008), which not only motivate them to fulfill their own duties but also encourage them to assist other team members to complete their work appropriately. In addition, supervisor trust is an important social cue for employees to shape their perceptions of trust and judge whether their teammates are dependable and reliable (Lau & Liden, 2008). Employees working in trustworthy teams are likely to believe strongly that they will receive back-up when they are in need of help and that their efforts will be converted into group success (Dirks, 1999; Palanski, Kahai, & Yammarino, 2011). In addition, as trust develops, they will spend less time ‘covering their backs’ and thus focus their cognitive and attentional resources more on individual task accomplishment (Colquitt, LePine, Zapata, & Wild, 2011; Mayer & Gavin, 2005; Palanski et al., 2011). Proposition 7: Team felt trust by supervisor is positively related to individual performance.

Conclusion As the development of interpersonal trust is embedded in the social structure of organizations (Fulmer & Gelfand, 2012), the social interaction regarding trust at the dyadic and team levels should be understood as mutually influencing. Crosslevel models (e.g., the antecedents and consequences related to trust at various levels, top-down, and bottom-up effects) are an exciting area for future trust research (Fulmer & Gelfand, 2012). Our study enriches the literature by focusing on trustees’ perceptions and exploring the nomological networks of each felt trust construct and their possible interrelations across levels. Recent world events have given the issue of trust more significance in management. Our discussion suggests that a trust-oriented management philosophy that guides employees through trust can offer real benefits to an organization. Specifically, this chapter highlights the importance of felt trust in management and demonstrates that the determinants of felt trust by supervisor and its impact on performance vary across different levels. Propositions in this chapter may help supervisors more accurately shape this important perception individually and collectively. Previous research has shown the importance of third parties in the formation of trust perceptions (Ferrin, Dirks, & Shah, 2006; Lau & Liden, 2008). We also extend the exploration of trust transfer and third-party influence by demonstrating how and under what conditions the perception of being trusted can shift

Multilevel Felt Trust  135

between different referents across levels. In doing so, we hope to stimulate interest in multilevel felt trust so as to entice researchers to devote increased attention to this topic. Future research might investigate the indirect effects of supervisory behaviors on individual felt trust through team felt trust or on team felt trust through individual felt trust. Doing so could help supervisors better understand how to manage the transmission of social information about trust beliefs among team members. The propositions proposed in this chapter are testable. We call for research to shed more light on the issues by examining the model empirically. As the current understanding of bottom-up processes is limited, it would be prudent to examine how higher-level phenomena or constructs emerge from lower-level ones (Gong, Kim, Lee, & Zhu, 2013). Two possible ways to identify member prototypicality seem viable given its definition. In an experiment, this variable would be manipulated by the similarity of goals and values between a participant and his or her team. In a field survey, team members would be asked to evaluate others’ prototypicality with a round-robin method. Member prototypicality could then be measured as the average of teammates’ scores. The context of our propositions is set in work organizations and vertical dyads. Future research might extend the context to nonbusiness, nonwork, and nonhierarchical relationships. In Gillespie’s (2003) behavioral trust model, our proposed felt trust antecedents are related to reliance but not disclosure. Prior felt trust research at work has also confirmed that reliance has a greater effect on employee work performance than disclosure (Lau et al., 2014). However, the lack of disclosure effects may be a contextual constraint. Thus, it would be worthwhile to examine how trustor disclosure in nonbusiness, nonwork, and nonhierarchical relationships affects trustee attitudes and behaviors. In conclusion, we argue that individual felt trust and team felt trust are two distinct variables with different antecedents and outcomes. By using appropriate supervisory behaviors to cultivate the feeling of being trusted, management can establish a foundation for trusting organizations with high performance.

References Aiello, J. R., & Kolb, K. J. (1995). Electronic performance monitoring and social context: Impact on productivity and stress. Journal of Applied Psychology, 80(3), 339–353. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​0​.3​.3​​39 Aryee, S., Budhwar, P. S., & Chen, Z. X. (2002).Trust as a mediator of the relationship between organizational justice and work outcomes: Test of a social exchange model. Journal of Organizational Behavior, 23(3), 267–285. https://doi​.org​/10​.1002​/job​.138 Ashforth, B. E., Harrison, S. H., & Corley, K. G. (2008). Identification in organizations: An examination of four fundamental questions. Journal of Management, 34(3), 325–374. https://doi​.org​/10​.1177​/0149206308316059 Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organization. Academy of Management Review, 14(1), 20–39. https://doi​.org​/10​.2307​/258189

136  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Baer, M. D., Dhensa-Kahlon, R. K., Colquitt, J. A., Rodell, J. B., Outlaw, R., & Long, D. M. (2015). Uneasy lies the head that bears the trust: The effects of feeling trusted on emotional exhaustion. Academy of Management Journal, 58(6), 1637–1657. https://doi​.org/ 10.5465/amj.2014.0246 Baer, M. D., Frank, E. L., Matta, F. K., Luciano, M. M., & Wellman, N. (2020). Under trusted, over trusted, or just right? The fairness of (in) congruence between trust wanted and trust received. Academy of Management Journal. Advance online publication. https:// doi​.org​/10​.5465​/amj​.2018​.0334 Baldwin, M. W. (1992). Relational schemas and the processing of social information. Psychological Bulletin, 112(3), 461–484. https​:/​/do​​i​.org​​/10​.1​​037​/0​​033​-2​​909​​.1​​12​.3.​​461 Barreto, N. B., & Hogg, M. A. (2017). Evaluation of and support for group prototypical leaders: A meta-analysis of twenty years of empirical research. Social Influence, 12(1), 41–55. https​:/​/do​​i​.org​​/10​.1​​080​/1​​55345​​10​.20​​​17​.13​​16771​ Bass, B. M. (1990). From transactional to transformational leadership: Learning to share the vision. Organizational Dynamics, 18(3), 19–31. https​:/​/do​​i​.org​​/10​.1​​016​/0​​​090​-2​​616 (9​0)900​61-S Becker, T. E., & Kernan, M. C. (2003). Matching commitment to supervisors and organizations to in-role and extra-role performance. Human Performance, 16(4), 327– 348. https​:/​/do​​i​.org​​/10​.1​​207​/S​​15327​​043H​U​​P1604​_1 Brass, D. J., & Burkhardt, M. E. (1993). Potential power and power use: An investigation of structure and behavior. Academy of Management Journal, 36(3), 441–470. https://doi​.org​/ 10​.2307​/256588 Brower, H. H., Lester, S.W., Korsgaard, M. A., & Dineen, B. R. (2009). A closer look at trust between managers and subordinates: Understanding the effects of both trusting and being trusted on subordinate outcomes. Journal of Management, 35(2), 327–347. https:// doi​.org​/10​.1177​/0149206307312511 Brower, H. H., Schoorman, F. D., & Tan, H. H. (2000). A model of relational leadership:The integration of trust and leader–member exchange. Leadership Quarterly, 11(2), 227–250. https​:/​/do​​i​.org​​/10​.1​​016​/S​​1​048-​​9843(​00)00​040-0​ Campagna, R. L., Dirks, K. T., Knight, A. P., Crossley, C., & Robinson, S. L. (2019). On the relation between felt trust and actual trust: Examining pathways to and implications of leader trust meta-accuracy. Journal of Applied Psychology. Advance online publication. http://dx​.doi​.org​/10​.1037​/apl0000474 Carney B., & Getz, I. (2018, September 10). Give your team the freedom to do the work they think matters most. Harvard Business Review. https​:/​/hb​​r​.org​​/2018​​/09​/g​​ive​-y​​our​t​​eam​-t​​he​-fr​​eedom​​-to​-d​​o​-the​​-work​​-they​​-t​hin​​k​-mat​​ters-​​most. Chan, D. (1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83(2), 234–246. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​3​.2​.2​​34 Chatman, J. A., & Flynn, F. J. (2001). The influence of demographic heterogeneity on the emergence and consequences of cooperative norms in work teams. Academy of Management Journal, 44(5), 956–974. https://doi​.org​/10​.2307​/3069440 Chen, G. (2005). Newcomer adaptation in teams: Multilevel antecedents and outcomes. Academy of Management Journal, 48, 101–116. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​05​​.15​​99314​7 Chen, G., Kirkman, B. L., Kanfer, R., Allen, D., & Rosen, B. (2007). A multilevel study of leadership, empowerment, and performance in teams. Journal of Applied Psychology, 92(2), 331–346. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​.2​.3​​31

Multilevel Felt Trust  137

Chen, Z., & Aryee, S. (2007). Delegation and employee work outcomes: An examination of the cultural context of mediating processes in China. Academy of Management Journal, 50(1), 226–238. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​07​​.24​​16238​9 Chughtai, A., Byrne, M., & Flood, B. (2015). Linking ethical leadership to employee wellbeing:The role of trust in supervisor. Journal of Business Ethics, 128(3), 653–663. https​:/​/ do​​i​.org​​/10​.1​​007​/s​​10551​​-01​4-​​2126-​7 Coats, S., Smith, E. R., Claypool, H. M., & Banner, M. J. (2000). Overlapping mental representations of self and in-group: Reaction time evidence and its relationship with explicit measures of group identification. Journal of Experimental Social Psychology, 36(3), 304–315. https://doi​.org​/10​.1006​/jesp​.1999​.1416 Cohen, S. G., & Bailey, D. E. (1997). What makes teams work: Group effectiveness research from the shop floor to the executive suite. Journal of Management, 23(3), 239–290. https​:/​/do​​i​.org​​/10​.1​​177​/0​​14920​​6397​0​​23003​​03 Colquitt, J. A., LePine, J. A., Zapata, C. P., & Wild, R. E. (2011). Trust in typical and highreliability contexts: Building and reacting to trust among firefighters. Academy of Management Journal, 54(5), 999–1015. https://doi​.org​/10​.5465​/amj​.2006​.0241 Colquitt, J. A., Scott, B. A., & LePine, J. A. (2007). Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance.  Journal of Applied Psychology, 92(4), 909–927. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​021​-9​​010​​.9​​2​.4​.9​​09 Davis, J. H., Schoorman, F. D., & Donaldson, L. (1997). Toward a stewardship theory of management. Academy of Management Review, 22(1), 20–47. https://doi​.org​/10​.2307​/ 259223 De Dreu, C. K., Nijstad, B. A., & van Knippenberg, D. (2008). Motivated information processing in group judgment and decision making. Personality and Social Psychology Review, 12(1), 22–49. https://doi​.org​/10​.1177​/1088868307304092 De Jong, B. A., & Dirks, K. T. (2012). Beyond shared perceptions of trust and monitoring in teams: Implications of asymmetry and dissensus. Journal of Applied Psychology, 97(2), 391–406. https://doi​.org​/10​.1037​/a0026483 Deutsch, M. (1958).Trust and suspicion. Journal of Conflict Resolution, 2(4), 265–279. https​:/​/ do​​i​.org​​/10​.1​​177​/0​​02200​​2758​0​​02004​​01 Dirks, K.T. (1999).The effects of interpersonal trust on work group performance. Journal of Applied Psychology, 84(3), 445–455. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​4​.3​.4​​45 Dirks, K. T., & Ferrin, D. L. (2001). The role of trust in organizational settings. Organization Science, 12(4), 450–467. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​2​.4​​.4​​50​.10​​640 Dutton, J. E., Dukerich, J. M., & Harquail, C.V. (1994). Organizational images and member identification. Administrative Science Quarterly, 39(2), 239–263. https​:/​/do​​i​.org​​/10​.1​​287​/ o​​rsc​.1​​2​.4​​.4​​50​.10​​640 Edmondson, A. C. (2003). Speaking up in the operating room: How team leaders promote learning in interdisciplinary action teams. Journal of Management Studies, 40(6), 1419– 1452. https://doi​.org​/10​.1111​/1467​-6486​.00386 Epitropaki, O., Kark, R., Mainemelis, C., & Lord, R. G. (2017). Leadership and followership identity processes: A multilevel review. Leadership Quarterly, 28(1), 104–129. https​:/​/ do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​16​.10​​.003 Fehr, R., Fulmer, A., & Keng-Highberger, F. T. (2020). How do employees react to leaders’ unethical behavior? The role of moral disengagement. Personnel Psychology, 73(1), 73– 93. https://doi​.org​/10​.1111​/peps​.12366

138  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Feitosa, J., Grossman, R., Kramer, W. S., & Salas, E. (2020). Measuring team trust: A critical and meta-analytical review. Journal of Organizational Behavior, 41(5), 479–501. https:// doi​.org​/10​.1002​/job​.2436 Fernández-Ballesteros, R., Díez-Nicolás, J., Caprara, G. V., Barbaranelli, C., & Bandura, A. (2002). Determinants and structural relation of personal efficacy to collective efficacy. Applied Psychology, 51(1), 107–125. https://doi​.org​/10​.1111​/1464​-0597​.00081 Ferrin, D. L., Dirks, K. T., & Shah, P. P. (2006). Direct and indirect effects of third-party relationships on interpersonal trust. Journal of Applied Psychology, 91(4), 870–883. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​1​.4​.8​​70 Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​.org​/ 10​.2139​/ssrn​.1873149 Gardner, D. G.,Van Dyne, L., & Pierce, J. L. (2004).The effects of pay level on organizationbased self-esteem and performance:A field study. Journal of Occupational and Organizational Psychology, 77(3), 307–322. https://doi​.org​/10​.1348​/0963179041752646 George, J. M. (1992). Extrinsic and intrinsic origins of perceived social loafing in organizations. Academy of Management Journal, 35(1), 191–202. https://doi​.org​/10​.2307​/ 256478 Giessner, S. R., van Knippenberg, D., & Sleebos, E. (2009). License to fail? How leader group prototypicality moderates the effects of leader performance on perceptions of leadership effectiveness. Leadership Quarterly, 20, 434–451. https​:/​/do​​i​.org​​/10​.1​​016​/j​​. leaq​​ua​.20​​​09​.03​​.012 Giessner, S. R., van Knippenberg, D., van Ginkel, W., & Sleebos, E. (2013). Team-oriented leadership: The interactive effects of leader group prototypicality, accountability, and team identification. Journal of Applied Psychology, 98(4), 658–667. https://doi​.org​/10​. 1037​/a0032445 Gillespie, N. (2003, August). Measuring trust in working relationships: The behavioral trust inventory [Paper presentation]. Academy of Management Conference, Seattle, WA. Gong,Y., Kim,T.Y., Lee, D. R., & Zhu, J. (2013).A multilevel model of team goal orientation, information exchange, and creativity. Academy of Management Journal, 56(3), 827–851. https://doi​.org​/10​.5465​/amj​.2011​.0177 Hains, S. C., Hogg, M. A., & Duck, J. M. (1997). Self-categorization and leadership: Effects of group prototypicality and leader stereotypicality. Personality and Social Psychology Bulletin, 23(10), 1087–1100. https​:/​/do​​i​.org​​/10​.1​​177​/0​​14616​​729​72​​31000​9 Hernandez, M. (2008). Promoting stewardship behavior in organizations: A leadership model. Journal of Business Ethics, 80(1), 121–128. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10551​​-00​7​​9440-​2 Hinsz, V. B., Tindale, R. S., & Vollrath, D. A. (1997). The emerging conceptualization of groups as information processors. Psychological Bulletin, 121(1), 43–64. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​033​-2​​909​​.1​​21​.1.​​43 Hirst, G., Van Knippenberg, D., Chen, C. H., & Sacramento, C. A. (2011). How does bureaucracy impact individual creativity? A cross-level investigation of team contextual influences on goal orientation–creativity relationships. Academy of Management Journal, 54(3), 624–641. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​11​​.61​​96812​4 Hogg, M. A., & van Knippenberg, D. (2003). Social identity and leadership processes in groups. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 35, pp. 1–52). San Diego, CA: Academic Press. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​065-​​2601(​03) 01​001-3​

Multilevel Felt Trust  139

Holman, D., Chissick, C., & Totterdell, P. (2002). The effects of performance monitoring on emotional labor and well-being in call centers. Motivation and Emotion, 26(1), 57–81. https://doi​.org​/10​.1023​/A​:1015194108376 Holtz, B. C., & Harold, C. M. (2008). When your boss says no! The effects of leadership style and trust on employee reactions to managerial explanations. Journal of Occupational and Organizational Psychology, 81(4), 777–802. https://doi​.org​/10​.1348​/ 096317907X251569 Janssen, O.,Van de Vliert, E., & West, M. (2004).The bright and dark sides of individual and group innovation: A special issue introduction. Journal of Organizational Behavior, 25(2), 129–145. https://doi​.org​/10​.1002​/job​.242 Kirkman, B. L., & Rosen, B. (1999). Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal, 42(1), 58–74. https://doi​.org​/10​.2307​/256874 Klein, K. J., & Kozlowski, S.W. (2000). From micro to meso: Critical steps in conceptualizing and conducting multilevel research. Organizational Research Methods, 3(3), 211–236. https://doi​.org​/10​.1177​/109442810033001 Konecki, K. T. (2019). Trust in symbolic interactionist research and in phenomenological investigation.  Polish Sociological Review,  207(3), 271–288. https​:/​/do​​i​.org​​/10​.1​​111​/j​​. 1475​​-682X​​.1972​​​.tb00​​229.x​ Lau, D. C., & Lam, L. W. (2008). Effects of trusting and being trusted on team citizenship behaviours in chain stores. Asian Journal of Social Psychology, 11(2), 141–149. https​:/​/do​​i​. org​​/10​.1​​111​/j​​.1467​​-839X​​.200​8​​.0025​​1.x Lau, D. C., Lam, L. W., & Wen, S. S. (2014). Examining the effects of feeling trusted by supervisors in the workplace: A self-evaluative perspective. Journal of Organizational Behavior, 35(1), 112–127. https://doi​.org​/10​.1002​/job​.1861 Lau, D. C., & Liden, R. C. (2008). Antecedents of coworker trust: Leaders' blessings. Journal of Applied Psychology, 93(5), 1130–1138. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​3​.5​.1​​130 Lau, D. C., Liu, J., & Fu, P. P. (2007). Feeling trusted by business leaders in China:Antecedents and the mediating role of value congruence. Asia Pacific Journal of Management, 24(3), 321–340. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10490​​-00​6-​​9026-​z Leana, C. R. (1986). Predictors and consequences of delegation. Academy of Management Journal, 29(4), 754–774. https://doi​.org​/10​.2307​/255943 Lester, S.W., & Brower, H. H. (2003). In the eyes of the beholder:The relationship between subordinates’ felt trustworthiness and their work attitudes and behaviors. Journal of Leadership & Organizational Studies,  10(2), 17–33. https​:/​/do​​i​.org​​/10​.1​​177​/ 1​​07179​​1903​0​​10002​​03 Levine, J. M., Resnick, L. B., & Higgins, E.T. (1993). Social foundations of cognition. Annual Review of Psychology, 44(1), 585–612. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​.ps.​​44​.02​​​0193.​​ 00310​1 Li, A. N., & Liao, H. (2014). How do leader–member exchange quality and differentiation affect performance in teams? An integrated multilevel dual process model. Journal of Applied Psychology, 99(5), 847–866. https://doi​.org​/10​.1037​/a0037233 Mael, F. A., & Ashforth, B. E. (1992). Alumni and their alma mater: A partial test of the reformulated model of organizational identification. Journal of Organizational Behavior, 13(2), 103–123. https://doi​.org​/10​.1002​/job​.4030130202 Mael, F. A., & Tetrick, L. E. (1992). Identifying organizational identification. Educational and Psychological Measurement, 52(4), 813–824. https​:/​/do​​i​.org​​/10​.1​​177​/0​​01316​​44920​​​ 52004​​002

140  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Malhotra, D., & Murnighan, J. K. (2002). The effects of contracts on interpersonal trust.  Administrative Science Quarterly,  47(3), 534–559. https://doi​.org​/10​.2307​/ 3094850 Marks, M. A., & Panzer, F. J. (2004). The influence of team monitoring on team processes and performance. Human Performance, 17(1), 25–41. https​:/​/do​​i​.org​​/10​.1​​207​/S​​15327​​ 043H​U​​P1701​_2 Martin, K., & Freeman, R. E. (2003). Some problems with employee monitoring. Journal of Business Ethics, 43(4), 353–361. https://doi​.org​/10​.2139​/ssrn​.348040 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi​.org​/ 10​.2307​/258792 Mayer, R. C., & Gavin, M. B. (2005). Trust in management and performance: Who minds the shop while the employees watch the boss? Academy of Management Journal, 48(5), 874–888. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​05​​.18​​80392​8 McAllister, D. J. (1995). Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https:// doi​.org​/10​.2307​/256727 Messick, D. M. (1999). Alternative logics for decision making in social settings. Journal of Economic Behavior & Organization, 39(1), 11–28. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​167-​​2681 (​99)00​023-2​ Nerstad, C. G., Searle, R., Černe, M., Dysvik, A., Škerlavaj, M., & Scherer, R. (2018). Perceived mastery climate, felt trust, and knowledge sharing. Journal of Organizational Behavior, 39(4), 429–447. https://doi​.org​/10​.1002​/job​.2241 Palanski, M. E., Kahai, S. S., & Yammarino, F. J. (2011). Team virtues and performance: An examination of transparency, behavioral integrity, and trust. Journal of Business Ethics, 99(2), 201–216. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10551​​-01​0-​​0650-​7 Pfeffer, J. (1998). The human equation: Building profits by putting people first. Boston, MA: Harvard Business School Press. https://doi​.org​/10​.1002​/npr​.4040170313 Pierce, J. L., & Gardner, D. G. (2004). Self-esteem within the work and organizational context: A review of the organization-based self-esteem literature. Journal of Management, 30(5), 591–622. https://doi​.org​/10​.1016​/j​.jm​.2003​.10​.001 Pierce, L., Snow, D. C., & McAfee, A. (2015). Cleaning house: The impact of information technology monitoring on employee theft and productivity. Management Science, 61(10), 2299–2319. https://doi​.org​/10​.1287​/mnsc​.2014​.2103 Richardson, H. A., Vandenberg, R. J., Blum, T. C., & Roman, P. M. (2002). Does decentralization make a difference for the organization? An examination of the boundary conditions circumscribing decentralized decision-making and organizational financial performance. Journal of Management, 28(2), 217–244. https​:/​/do​​i​.org​​/10​.1​​177​/ 0​​14920​​6302​0​​28002​​05 Salamon, S. D., & Robinson, S. L. (2008).Trust that binds:The impact of collective felt trust on organizational performance. Journal of Applied Psychology, 93(3), 593–601. https​:/​/do​​i​. org​​/10​.1​​037​/0​​021​-9​​010​​.9​​3​.3​.5​​93 Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 23(2), 224–253. https:// doi​.org​/10​.2307​/2392563 Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007).An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32(2), 344–354. https​:/​/ do​​i​.org​​/10​.5​​465​/a​​mr​.20​​07​​.24​​34841​0

Multilevel Felt Trust  141

Sluss, D. M., & Ashforth, B. E. (2007). Relational identity and identification: Defining ourselves through work relationships. Academy of Management Review, 32(1), 9–32. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​07​​.23​​46367​2 Sluss, D. M., Ployhart, R. E., Cobb, M. G., & Ashforth, B. E. (2012). Generalizing newcomers' relational and organizational identifications: Processes and prototypicality. Academy of Management Journal, 55(4), 949–975. https://doi​.org​/10​.5465​/amj​.2010​.0420 Somech, A., Desivilya, H. S., & Lidogoster, H. (2009). Team conflict management and team effectiveness: The effects of task interdependence and team identification. Journal of Organizational Behavior, 30(3), 359–378. https://doi​.org​/10​.1002​/job​.537 Spitzmüller, C., & Stanton, J. M. (2006). Examining employee compliance with organizational surveillance and monitoring. Journal of Occupational and Organizational Psychology, 79(2), 245–272. https://doi​.org​/10​.1348​/096317905X52607 Spreitzer, G. M., De Janasz, S. C., & Quinn, R. E. (1999). Empowered to lead: The role of psychological empowerment in leadership. Journal of Organizational Behavior,  20(4), 511–526. https​:/​/do​​i​.org​​/10​.​1​​002/(​SICI)​1099-​1379(​19990​7)20:​ 43.0.CO;2-L Stanton, J. M. (2000). Reactions to employee performance monitoring: Framework, review, and research directions. Human Performance, 13(1), 85–113. https​:/​/do​​i​.org​​/10​.1​​207​/ S​​15327​​043H​U​​P1301​_4 Tajfel, H. (1972). Social categorization. English manuscript of ‘La catégorisation sociale. In S. Moscovici (Ed.), Introduction à la psychologie sociale (Vol. 1, pp. 272–302). Paris: Larousse. https://doi​.org​/10​.2307​/2605125 Tajfel, H. (1978). Differentiation between Social Groups: Studies in the Social Psychology of Intergroup Relations. London: Academic Press. https://doi​.org​/10​.1086​/227378 Turner, J. C. (1991). Social influence. Pacific Grove, CA: Brooks/Cole. https​:/​/do​​i​.org​​/10​. 1​​177​/0​​02087​​2891​0​​34004​​01 Turner, J. C., & Haslam, S. A. (2001). Social identity, organizations, and leadership. In M. E. Turner (Ed.), Applied social research. Groups at work: Theory and research (pp. 25–65). Mahwah, NJ: Erlbaum. https://doi​.org​/10​.2307​/2578690 Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford: Blackwell. https​:/​/do​​i​.org​​/10​.1​​002​/ 9​​78111​​84308​​​73​.es​​t0539​ Turner, J. C., & Tajfel, H. (1986). The social identity theory of intergroup behavior. Psychology of Intergroup Relations,  5, 7–24. https://doi​.org​/10​.4324​/9780203505 984​-16 van Knippenberg, D. (2011). Embodying who we are: Leader group prototypicality and leadership effectiveness. Leadership Quarterly, 22(6),1078–1091.https​:/​/do​​i​.org​​/10​.1​​016​/ j​​.leaq​​ua​.20​​​11​.09​​.004 van Knippenberg, D., & Hogg, M. A. (2003). A social identity model of leadership effectiveness in organizations. Research in Organizational Behavior, 25, 243–295. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​0​191-​​3085(​03)25​006-1​ van Knippenberg, B., & van Knippenberg, D. (2005). Leader self-sacrifice and leadership effectiveness: The moderating role of leader prototypicality. Journal of Applied Psychology, 90(1), 25–37. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​01​0​.9​​0​.1​.2​5 van Knippenberg, D., van Knippenberg, B., & van Dijk, E. (2000). Who takes the lead in risky decision making? Effects of group members' risk preferences and prototypicality.  Organizational Behavior and Human Decision Processes, 83(2), 213–234. https://doi​.org​/10​.1006​/obhd​.2000​.2907

142  Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam

Wageman, R. (2001). How leaders foster self-managing team effectiveness: Design choices versus hands-on coaching. Organization Science,  12(5), 559–577. https​:/​/do​​i​.org​​/10​. 1​​287​/o​​rsc​.1​​2​.5​​.5​​59​.10​​094 Walumbwa, F. O., Morrison, E. W., & Christensen, A. L. (2012). Ethical leadership and group in-role performance: The mediating roles of group conscientiousness and group voice.  Leadership Quarterly,  23(5), 953–964. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​12​. 06​​.004 Webber, S. S. (2008). Development of cognitive and affective trust in teams: A longitudinal study. Small Group Research,  39(6), 746–769. https://doi​.org​/10​.1177​/ 1046496408323569 Whitener, E. M., Brodt, S. E., Korsgaard, M. A., & Werner, J. M. (1998). Managers as initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23(3), 513–530. https://doi​.org​/ 10​.2307​/259292 Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of Management, 17(3), 601–617. https​:/​/do​​i​.org​​/10​.1​​177​/0​​14920​​6391​0​​17003​​05 Yang, J., & Mossholder, K. W. (2010). Examining the effects of trust in leaders: A basesand-foci approach. Leadership Quarterly, 21(1), 50–63. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​. 20​​​09​.10​​.004 Yukl, G. A. (1998). Leadership in organizations. Upper Saddle River, NJ: Prentice-Hall, Inc. https://doi​.org​/10​.2307​/1244516 Zaccaro, S. J., Rittman, A. L., & Marks, M. A. (2001). Team leadership. Leadership Quarterly, 12(4), 451–483. https​:/​/do​​i​.org​​/10​.1​​016​/S​​1​048-​​9843(​01)00​093-5​

7 TRUST REPAIR A Multilevel Framework Nicole Gillespie, Steve Lockey, Matthew Hornsey, and Tyler Okimoto

Introduction Trust is vital to successful organizational functioning (Fulmer & Gelfand, 2012; McEvily, Perrone & Zaheer, 2003), yet it is also fragile and easily lost (Bachmann, Gillespie, & Priem, 2015). Trust failures in organizations are unfortunately all too common and happen at multiple levels, including the interpersonal, group, and organizational levels. While it is widely recognized that trust repair after a breach also occurs at multiple levels, most scholarly examination of trust repair to date has focused solely on one level of analysis, typically the individual level (Kramer & Lewicki, 2010; Schoorman, Mayer, & Davis, 2007; Lewicki & Brinsfield, 2017). The small body of extant work that examines trust repair at multiple levels remains isolated and fragmented, limiting both practical and scholarly understanding of whether and how the dynamics, mechanisms, and processes of trust repair may operate differently across levels. This is despite recent reviews arguing that trust processes and dynamics are not isomorphic across levels and referents (Fulmer & Gelfand, 2012), and conceptual and empirical work suggesting that the dynamics and processes of trust repair differ at the individual versus collective and organizational levels (e.g., Gillespie & Dietz, 2009; Gillespie, Dietz, & Lockey, 2014; Kim, Cooper, Dirks, & Ferrin, 2013). The aim of this chapter is to advance understanding of the multilevel nature of trust repair. We respond to calls to examine trust repair across levels in an integrated manner (Fulmer & Dirks, 2018; Fulmer & Gelfand, 2012; Gillespie & Dietz, 2009; Kim, Dirks, & Cooper, 2009) by integrating insights from the distinct scholarly literatures on trust repair in social psychology on the one hand, and organizational behavior, management, and organizational studies on the other. We integrate these literatures to answer two fundamental questions. First, in what DOI: 10.4324/9780429449185-7

144  Nicole Gillespie et al.

ways do the mechanisms, dynamics, and processes of trust repair differ across individual and collective levels of analysis, and in what ways are they similar? Second, given interpersonal relationships are embedded within groups and the broader organizational context, how does this embeddedness facilitate and/or constrain trust repair processes and dynamics? In answering these questions, we develop an integrated framework of the key factors influencing approaches to trust repair across individual and collective levels. While our focus is on trust repair in work contexts, we draw on research conducted in non-work settings that help bring insight into multilevel trust repair. Our chapter contributes to the trust repair literature in three key ways. First, we integrate insights from social psychological theories and frameworks of trust and moral repair focused at the interpersonal and group level (e.g., Philpot & Hornsey, 2008; Wenzel, Okimoto, Hornsey, Lawrence-Wood, & Coughlin, 2017; Wohl, Hornsey, & Bennett, 2012), with organizational behavior and management theories focused on trust repair at the interpersonal, organizational, and institutional levels (e.g., Bachmann et al., 2015; Gillespie & Dietz, 2009; Kim et al., 2009; Kramer & Lewicki, 2010). Second, we identify novel scholarly and practical insight into the similarities and differences in repairing trust at the individual and group levels of analysis. Third, we outline a research agenda for advancing a multilevel perspective on trust repair within organizations. In the next section, we begin with a brief review outlining the current conceptualization of trust violations and repair, and then provide an overview of commonly studied trust repair mechanisms and strategies. From this foundation, we then elaborate on how multilevel dynamics further complicate this framework, building toward our multilevel framework on trust repair.

Conceptualizing Trust Violations and Repair In the management literature, it is commonly accepted that trust comprises two key elements: (1) a willingness to be vulnerable to the actions or behaviors of another party based on (2) positive expectations of that party’s conduct (Mayer, Davis, & Schoorman, 1995; Rousseau, Sitkin, Burt, & Camerer, 1998). The other ‘party’ may be an individual, a group, an organization, or some other kind of entity (Schoorman et al., 2007).We view a trust violation as an event that causes a trustor’s positive expectations of and willingness to be vulnerable to another party to substantively diminish (Kim et al., 2009).When one party violates another’s trust, one or both parties may engage in the process of trust repair by undertaking action(s) to improve the victim’s positive expectations of the transgressor and willingness to be vulnerable to them (Kramer & Lewicki, 2010). As we argue in our chapter, third parties can also play a role in trust repair. While simple to describe, the process of trust violation and repair is dynamic and complex (Lewicki & Brinsfield, 2017). In their integrative review, Dirks and

Multilevel Trust Repair  145

colleagues (2009) identified three theoretical bases underpinning trust repair research: attributional, social equilibrium, and structural. More recently, Bachmann et al. (2015) drew on and extended these three theoretical bases to develop an integrated framework of five1 theoretical mechanisms for trust repair at the organizational and institutional levels: sense-making, relational, structural and (in)formal control, transparency, and transference. The sense-making mechanism integrates Dirks and colleagues’ attributional process and assumes that trustors require an understanding of what went wrong and why for effective trust repair to take place. The focus is on cognitive processes – making sense of the violation and gaining an accepted account of it (e.g., What happened? Who or what caused this to happen? Who was affected?). Repair activities focus on providing the wronged party with information that shifts and overcomes negative attributions and inferences, through strategies such as offering explanations, justifications, denials, or conducting investigations into the violation. The relational mechanism assumes that rituals and symbolic acts are required to resolve the negative emotions caused by the violation. This mechanism integrates Dirks and colleagues’ social equilibrium process, which argues that trust violations disrupt the relative standing of the parties involved and cause disequilibrium in the relationship and social context. Re-establishing equilibrium can be achieved by restoring the relative standings of the parties and reaffirming the norms that govern them through social rituals such as apologies, penance, compensation, and punishment. These repair tactics help answer the question of whether the transgressor is remorseful and whether they have learned a lesson and appropriately made amends with victims. The structural and (in)formal control mechanism facilitates trust repair by putting in place rules or (in)formal controls that constrain the likelihood of future untrustworthy behavior or transgressions occurring and encourage positive exchange. This includes strategies such as implementing new policies, codes of conduct, cultural reforms, incentives, sanctions, regulations, laws, or contracts that regulate distrust and prevent future transgressions. This mechanism integrates and extends the structural mechanism proposed by Dirks et al. (2009) by highlighting that cultural changes in values, norms, and assumptions are strong drivers of behavior, and are often required in addition to formal structural and regulatory processes to regulate untrustworthy behavior by groups. Transparency assumes that open and transparent information sharing and reporting facilitates trust repair by demonstrating to trustors that the group or organization is behaving in a trustworthy manner. It answers the question of whether there is evidence of renewed trustworthiness. This includes strategies such as

1 In the original article, formal and informal controls were separated; however, for the purposes of this chapter they are combined for parsimony.

146  Nicole Gillespie et al.

conducting independent audits and openly reporting the results and allowing ongoing monitoring. Transference assumes that trust repair can be facilitated by transferring trust from one (credible) party to another (discredited party), through processes of endorsement, affiliation, and reputational spillover, as well as certifications and awards. Research shows that people are more likely to trust another party or entity when they are endorsed by or otherwise trusted by a credible third party (Ferrin, Dirks, & Shah, 2006; Mueller, Carter, & Whittle, 2015), share a common affiliation or membership (Brewer, 1979), or have certifications or awards that signal trustworthiness (Gillespie et al., 2014). Table 7.1 provides an outline of these mechanisms and their underlying assumptions, as well as examples of selected scholarly papers relevant to each mechanism, at the individual and collective levels. As the last two rows show, research has not yet examined the full spectrum of trust repair mechanisms, with interpersonal trust repair research predominately focusing on only the first two mechanisms.

A Multilevel Perspective on Trust Repair Building on this foundational theoretical work, we review how each of the trust repair mechanisms features in the extant literature at various levels of analysis. To understand the multilevel nature of trust repair, we consider both the level of the transgressor (individual or collective) and the level of trust (i.e., whether the trust of an individual or a collective has been violated). From this, we identify and examine three distinct levels of trust repair (see Table 7.2). We use the term individual-to-individual to refer to interpersonal trust repair, where an individual transgressor seeks to repair the trust of the individual victim (e.g., a leader or employee repairing the trust of a co-worker).The term individualto-collective trust repair refers to situations where an individual seeks to repair the trust of a group, collective, or multiple others (e.g., a leader repairing trust with her team or with her organization). Finally, collective-to-collective trust repair refers to a collective group that seeks to repair the trust of another group or groups (e.g., an organization repairing trust with its stakeholders, or a supplier repairing trust with a procuring organization). As we explain later in the chapter (see the section on collective-to-collective repair), when a collective violates the trust of an individual, it is typically interpreted as an affront and violation toward the collective group to which the individual belongs. For example, the police using undue force in arresting a black American is likely to be viewed as an inter-group violation of police towards black citizens. Similarly, an organization paying a woman substantially less than her male colleagues performing the same role is seen as a violation towards female employees. Hence, collective-to-individual–type violations require the transgressing collective to engage in trust repair processes with the collective that the

Underlying mechanism

Assumption

Relational (social equilibrium)

Structural and (in) formal controls

Transparency

Transference

A shared Trust repair requires Trust repair requires Transparently sharing Trust repair can be facilitated understanding social rituals and rules or formal relevant information by transferring trust from or accepted symbolic acts to and informal about the a credible party to the account of the resolve negative controls to transgressing party discredited party trust violation is emotions caused constrain future and their ongoing required for trust by the violation untrustworthy functioning helps repair, including and re-establish the behavior and restore trust clarification social order in the hence prevent on whether relationship future trust a violation violations occurred and if so, what caused it and who is responsible Collective learning Remorse and Formal and informal Information sharing Trust transfer from third parties through redemption through control through reporting and reputation spillover cognition, social rituals and or monitoring attributions, and emotional expression social influence

Sense-making (attributional)

Trust repair mechanism

TABLE 7.1 Trust repair mechanisms used for restoring trust across multiple levels

Multilevel Trust Repair  147

What happened and Are those responsible why? remorseful? Who is responsible/ Have they been guilty? punished/learned Who was affected by their lesson? Have the transgression? appropriate amends been made to victims? Repair strategies Explanations, Apologies, punishment, justifications, penance, accounts, denials, compensation, apologies, ‘grassroots’ remorse, use reticence, of ingroup messenger, investigations, and redistribution of power, public inquiries resetting expectations, and removal of offenders from group Selected Benoit, 2006, 2017; Brown et al., 2008; examples – Gillespie & Dietz, Gillespie & Dietz, collective trust 2009; Gillespie et 2009; Gillespie et repair al., 2014; Kim et al., 2014; Hornsey & al., 2006; Pfarrer Wohl, 2013; Kim et et al., 2008; al., 2009; Okimoto et Tomlinson & al., 2019; Wenzel et Mayer, 2009 al., 2017; Philpot & Hornsey, 2011; Wohl et al., 2011, 2012; Stevens et al., 2015; Tomlinson et al., 2004

Fundamental questions

Is there evidence of renewed trustworthiness?

Do others trust this party?

Regulation, laws, rules, Open reporting, Third-party or group policies, controls, monitoring, endorsement/acceptance, contracts, codes of external audits, certifications, memberships, conduct, sanctions, public inquiries, affiliations, awards, and incentives, ‘hostage and whistleblower endorsements posting,’ promises, protection cultural reforms, training, leadership, and role modeling Bachmann & Inkpen, Augustine, 2012; Child Ferrin et al., 2006; Mueller et al., 2011; Eberl et al., & Rodrigues, 2004; 2015; McEvily et al., 2003; 2015; Gillespie Grimmelikhuijsen et Spicer & Okhmatovskiy, 2015 & Dietz, 2009; al., 2013; Pirson & Gillespie et al., Malhotra, 2011 2014; Poppo & Schepker, 2010; Sitkin & Roth, 1993; Stevens et al., 2015; Nakayachi & Watabe, 2005 (Continued)

Has the party reformed itself? Have changes been made to prevent future transgressions?

148  Nicole Gillespie et al.

Relational (social equilibrium)

Note: Adapted from Bachmann et al. (2015).

Selected Bagdasarov et al., Bagdasarov et al., examples – 2019; Bottom 2019; Bottom et al., interpersonal et al., 2002; 2002; De Cremer trust repair Elangovan et al., & Schouten, 2008; 2015; Kim et Desmet et al., 2011a, al., 2004, 2013; 2011b; Dirks et al., Ferrin et al., 2007; 2011; Ferrin et al., Shapiro, 1991 2007; Haesvoets et al., 2013, 2018; Kim et al., 2004, 2006; Ma et al., 2019; Schniter et al., 2013; Schweitzer, Hershey, & Bradlow, 2006

Sense-making (attributional)

Trust repair mechanism

TABLE 7.1 Continued

Dirks et al., 2011

Structural and (in) formal controls

Transparency Ferrin et al., 2006

Transference

Multilevel Trust Repair  149

150  Nicole Gillespie et al.

TABLE 7.2 Three levels of trust repair

Transgressor (or trustee)

Repairing trust of…(trustor)

After a violation involving…

Individual Individual

Individual Collective

Collective

Collective

Individual transgressor–individual victim Individual transgressor–collective victim Individual transgressor–individual victim* Collective transgressor–collective victim Collective transgressor–individual victim** Individual transgressor–individual victim**

* in situations where the transgression negatively affects the trust of the broader group to which the victim belongs, requiring collective level repair. In this situation, the violation is indirectly experienced by the collective. ** in situations where the violation is interpreted as an inter-group trust breach and hence the violation toward an individual victim is interpreted as a violation toward the broader collective group to which the victim belongs, and the transgression by an individual is interpreted as a transgression by the broader collective group to which the transgressor belongs.

individual victim belongs to (e.g., black American community; female employees), rather than just with the sole individual who was the direct victim of the transgression. Similarly, collective trust repair is typically required if a transgression by one member of a group towards a member of another group (i.e., individual-toindividual violation), is interpreted as an inter-group violation. For this reason, we discuss collective-to-individual trust repair within the section on collective-tocollective repair. A general proposition guiding our chapter is that trust repair mechanisms and dynamics are not isomorphic across levels – rather they can vary significantly depending on the level at which the trust repair process takes place. Our starting point is the observation that individual-to-individual trust violation and repair between two colleagues is likely to play out very differently from collective-to-collective organizational breaches where the concerns of multiple stakeholders need to be considered. Attributions regarding responsibility for the trust violation are typically easier to evidence at the individual than the collective level. In scandals involving an organization, who or what is responsible for the violation is often not clear and often requires investigation. Attributions of responsibility and causes may be contested, shaped by group dynamics, processes, and external influences, and further complicated by the fact that not all stakeholders are likely to perceive and make sense of the violation in the same way (Bachmann et al., 2015). Indeed, repair strategies that appease one group of stakeholders may further alienate another. Organizations also generally have a suite of different strategies to choose from in responding to a violation (Dirks et al., 2009; Gillespie & Dietz, 2009); for instance, an organization can fire a culpable transgressor, an option not available

Multilevel Trust Repair  151

to individual transgressors. Relatedly, some repair strategies may be inappropriate depending on the context; in the wake of an organizational trust violation, companies often compensate affected stakeholders in an attempt to make amends for their transgressions. However, in an interpersonal setting within the workplace, attempting to ‘buy back’ a colleague’s trust in such a manner is likely to be considered crass and inappropriate (Elangovan & Shapiro, 1998). These examples evidence the need for a more systematic review of the similarities and differences of trust repair across levels. The three levels of trust repair have received varying degrees of attention, and much work remains to be done to understand their various dynamics. In the next sections, we examine prior work on trust repair and closely related concepts at the individual-to-individual, individual-to-collective, and collective-to-collective levels, respectively. Our review is necessarily selective and aims to draw out insights to inform a forward-looking multilevel framework and research agenda for deepening understanding of trust repair across levels.

Individual-to-Individual (Interpersonal) Trust Repair Individual-to-individual trust repair involves dyadic, interpersonal relationships, where there are only one transgressor and one victim. Both parties typically have direct, first-hand experience of the violation; therefore, compared to collective contexts, it is relatively more clear what happened, why it happened, who caused it (facilitating sense-making), if that transgressor is remorseful, and what they need to do to make amends (facilitating relational repair). It is not surprising, then, that most research on individual-to-individual trust repair concerns tactics underpinned by sense-making and relational processes (see Table 7.1 bottom row for examples). In particular, scholarly research on individual-individual level trust repair has focused on verbal strategies for repairing trust, particularly apologies, denials, and explanations.This research is primarily experimental in nature and has shown that apologies are more than mere “cheap talk” and can positively influence attributions and sense-making about the violation and transgressor (Bottom, Gibson, Daniels, & Murnighan, 2002, p. 500), as well as future cooperation (Bottom et al., 2002; De Cremer & Schouten, 2008; Schniter, Sheremeta, & Sznycer, 2013). The more components an apology has, the more effective it is perceived to be (Lewicki, Polin, & Lount, 2016). However, underpinned by attribution theory, a series of studies (Dirks, Kim, Ferrin, & Cooper, 2011; Ferrin, Kim, Cooper, & Dirks, 2007; Kim, Ferrin, Cooper, & Dirks, 2004; Kim, Dirks, Cooper, & Ferrin, 2006) found the effectiveness of verbal trust repair strategies depends on the type of violation; apologies are more effective at repairing trust after a competence-based failure, and denials are more effective after integrity failures (Bagdasarov, Connelly, & Johnson, 2019; Ferrin et al., 2007; Kim et al., 2004, 2006). Apologies are also more effective than reticence

152  Nicole Gillespie et al.

– simply not responding to an allegation (Ferrin et al., 2007), both after abilityand integrity-based transgressions. This is because failure to respond at all to a violation indicates to the violated party that the transgressor does not care about the status of their relationship or any consequences that arise from the violation. In addition to facilitating sense-making and attributions about the violation, apologies can help repair the relational damage caused by a violation. Apologies can positively enhance the emotional states of both the violated party and the transgressor (Byrne, Barling, & Dupré, 2014). Affective and relational factors can also motivate the transgressing party to offer an apology (Schniter & Sheremeta, 2014). However, when violated parties experience high levels of negative emotion in the aftermath of a trust violation, apologies alone may not be sufficient to repair trust (Ma et al., 2019). In particular, an apology after an integritybased violation may result in more negative emotions if it is devoid of empathy (Bagdasarov et al., 2019), and insincere apologies from leaders can result in more negative reactions from subordinates than no apology at all (Basford, Offermann, & Behrend, 2014). At the individual–individual level, the most commonly studied substantive trust repair strategy is the offer of compensation (Lewicki & Brinsfield, 2017; see also Mullen & Okimoto, 2015). Experimental studies show that the offer of financial compensation after a trust transgression can incentivize violated parties to trust deceitful partners in future rounds of the trust game, particularly when coupled with an apology (Bottom et al., 2002; Desmet, De Cremer, & van Dijk, 2011a, 2011b; Haesevoets, Folmer, De Cremer, & Van Hiel, 2013). Recently, Haesevoets, De Cremer,Van Hiel, and Van Overwalle (2018) found that receiving compensation (compared to receiving no compensation) after a violation had a positive direct effect on trust, and was associated with increased activation in areas of the brain linked to forgiveness. Dirks et al. (2011) provide one of the few studies of structural and (in)formal controls in trust repair at the interpersonal level. Their experiments found that both compensation and regulation could be effective in repairing trust; however, regulation was only effective in repairing trust when participants perceived the transgressor to have repented for their actions. In sum, research on individual– individual trust repair has focused on and is underpinned by the sense-making (attributional) and relational mechanisms, with structural and control mechanisms far less apparent, and transparency and transference mechanisms rarely studied. Notably, this interpersonal trust repair research, drawn predominately from organizational behavior and management literatures, has rarely considered the potential influence of parties beyond the individual transgressor and trustor; but this does not mean the broader collective has no relevance. Indeed, social psychological research and theory have argued for the role of third parties in interpersonal trust repair processes as agents of influence that may compel or persuade the transgressor (and victim) to engage in the repair process (Okimoto, 2008; Yu, Yang, & Jing, 2017). In this case, the target of trust repair is still the

Multilevel Trust Repair  153

victim as a trustor (i.e., one-to-one, interpersonal trust), but the broader group acts as an adjudicator, looking after the victim’s interests. For example, following escalation of conflict where one co-worker has insulted another co-worker, an organizational representative (e.g., department leader or HR) may intervene, encouraging the two parties to repair the relationship so they can continue to cooperate productively. In such instances, the intervening group representative may seek a reparative response that moves toward reconciliation on behalf of the victim (Okimoto & Wenzel, 2014). As we explore in the next section, the line between the victim and third parties within the shared collective can be a thin one, and an individual-to-individual violation can be perceived as a violation against the broader collective.

Individual-to-Collective Trust Repair With ‘individual-to-collective’ trust violations, it is a single individual who has breached the trust of a collective, suffers the cost of the resulting trust deficit, and is held accountable for repairing it. This particular situation is quite common in organizational contexts, when an individual employee or leader violates the trust of his/her team or co-workers.The violation itself is construed as an affront to the broader group, and any subsequent trust repair strategies must target group-level concerns, and will be influenced by the group’s dynamics and processes. Collective-level trust can be lost directly (i.e., via a violation towards the collective group) or indirectly (i.e., via a violation towards a single group member). Often direct violations of collective trust by individual transgressors occur from a position of power or leadership; a leader might be caught in a lie (integrity violation), fail to deliver promised resources (ability violation), or make a decision favoring their own interests rather than the team’s interests (benevolence violation; Mayer et al., 1995). The complexity of maintaining multiple relationships is likely to be an exacerbating factor that contributes to leader-to-collective violations. The nature of their role may encourage leaders to engage in behaviors that enhance the trust of one party yet intentionally or unintentionally violate the trust of another. For example, leaders may feel the need to give special treatment to ‘star’ employees to retain them in a competitive employment market, yet have to balance that against the expectation to treat all employees equally, lest others perceive them to act unfairly. The embedded nature of leadership within hierarchies also opens up cross-level effects in trust violation and repair, where higher-level dynamics can influence lower-level trust violations. For example, a leader may promise a bonus to employees only later to be informed by their superiors that they need to cut costs. Dirks and Skarlicki (2004) term these tensions trust dilemmas. In such contexts, employees are likely to be more willing to forgive leader transgressions when attributed to external forces (e.g., directives of those above), rather than within the leader’s control (Tomlinson & Mayer, 2009), particularly after a competence failure (Kim et al., 2006).

154  Nicole Gillespie et al.

Research from the image-repair and face-saving literatures suggests that the strategies leaders use to respond to violations are often more defensive or qualified when targeting a collective than when targeting an individual (e.g., Brown, 1968). This is likely due to the influence of third-party advisors (e.g., lawyers) and the need to appeal to multiple audiences and stakeholders with varying values, goals, and interests (e.g., party members and supporters, as well as critics). For instance, George W. Bush frequently employed denial and defensive tactics, refusing to apologize for anything relating to the Iraq War in televised broadcasts preceding the 2004 US presidential election (Benoit, 2006). More recently, Benoit (2017) analyzed Donald Trump’s response to a video showing him making disparaging sexual remarks about women, which surfaced days before a presidential debate in the 2016 US election campaign. Trump offered a short conditional apology: “I apologize if anyone was offended” (Benoit, 2017, p. 249), and failed to acknowledge that any inappropriate behavior had taken place. In one of the only studies to directly examine the differences that arise when individual transgressors attempt to regain the trust of a group as compared to individuals, Kim, Cooper, Dirks, and Ferrin (2013) found that trust repair was generally more difficult with groups than individuals. They explain this through group polarization literature (Hinsz, Tindale, & Vollrath, 1997) arguing that after a trust violation, the involvement of multiple trustors (rather than just one), strengthens the negativity toward the transgressor, making trust repair more difficult. They find that individuals anchor their assessments to the harsher assessments of the group, which persist over time. They further find that this harshness by the group is further exacerbated when the transgressor provides an ‘ineffective response,’ meaning that the repair response (apology versus denial) was mismatched with the violation type (apology for integrity failure, denial for competence failure). The study supports the view that repairing the trust of a group differs from, and is more difficult than, repairing the trust of individuals. Repairing the trust of the collective is often a necessary response following individual-to-individual violations too. Even if the violation itself directly targeted or harmed one specific individual victim, trust violations rarely occur in a social vacuum (Brodt & Neville, 2013; Kim et al., 2013; McEvily, Zaheer, & Soda, this volume), and the broader group may nonetheless be indirectly impacted by the violation and feel diminished trust in the transgressor. Hence, within organizational contexts, individual-to-individual violations may nonetheless require individual-to-collective trust repair action. Examples of this dynamic abound. If a supervisor treats one of her direct reports poorly, that violation likely undermines the victim’s trust in the supervisor – but when the broader group observes or hears of that trust breach, it also undermines the broader group’s trust in that supervisor. In another example, a leader’s sexist attitudes or sexual harassment toward an employee destroys trust not only with the direct victim but also indirectly with other employees (as evidenced by the #MeToo movement).

Multilevel Trust Repair  155

Although trust research has tended to focus on the victim’s perspective, a growing body of research suggests that rebuilding trust with the broader group is critical following an interpersonal trust violation. The literature on third-party reactions to justice violations provides evidence that violations affect the views of others in the organization, not just the victim (see Skarlicki, O’Reilly, & Kulik, 2015). According to Skarlicki and Kulik (2004), employees care about violations against other organizational members for two reasons: self-interest and moral imperative.2 What happens to others (particularly in-group members; Tyler & Lind, 1992) arouses self-protective concerns in observers that something similar might happen to them in the future, an instrumental fear of being personally vulnerable (i.e., lacking trust). However, the plight of others also arouses a moralistic response to the violation of social norms about how trusted co-workers should behave (Cropanzano, Goldman, & Folger, 2005; Folger, 2001), a symbolic threat to the moral fabric (i.e., integrity) of the group, organization, and/or broader society (Durkheim, 1964; Tyler & Boeckmann, 1997; Vidmar, 2000). This is particularly salient when the transgressor has a leadership position and hence is a symbol of the organization’s values and cultures. Such leader violations trigger the need for trust repair not only with internal employees but also with external stakeholders, and depending on the nature and severity of the transgression, may trigger the need for trust repair in the organization itself (i.e. cross-level effects). Using the language from the trust literature, individual trust violations affect the broader group(s) by undermining their trust in the individual transgressor, as well as their trust in the integrity of the organization that the transgressor belongs to (Ferrin et al., 2007). As a consequence of the broader group’s concern over the integrity of the organization, trust repair in individual-to-collective contexts require relational mechanisms that reassert the norms and values of the broader group. Indeed, groups often respond to individual transgressors (e.g., CEOs) by removing the transgressor from the group (e.g., ‘heads will roll’) or requiring them to apologize and offer penance, which can be effective in restoring trust (Dirks et al., 2011; Ferrin et al., 2007; Ferrin, Cooper, Dirks, & Kim, 2018). Research has shown repeatedly that indirectly affected parties respond to injustice (and in particular the implied threat to morality) with punishment of the violator (i.e., ‘third-party punishment’; e.g., Fehr & Gächter, 2002; Turillo, Folger, Lavelle, Umphress, & Gee, 2002). Punishment of the transgressor allows third parties to alleviate their moral-emotional reaction (Gromet, 2012; Haidt, 2001) and re-establish the moral order by reasserting shared rules and norms (Okimoto & Wenzel, 2009; Dirks et al., 2009; Tyler & Boeckmann, 1997). The transgressor also has agency to restore

2 These two underlying concerns over self-interest and morality mirror the distinction in moral philosophy between consequentialism and deontology (see Brooks, 2012), and the distinction in psychology between utilitarianism and retributivism (see Carlsmith, 2008).

156  Nicole Gillespie et al.

the social order and help restore group trust in them through relational mechanisms such as demonstrating sincere commitment to the values and norms that were breached (Okimoto & Wenzel, 2009) and voluntarily offering apologies, penance, and self-regulation to the broader group. However, one of the challenges with restoring the trust of the collective is that members can diverge in their sense-making about the violation, and expectations for relational trust repair, as we explore further in the next section.

Collective-to-Collective Trust Repair Collective-to-collective trust repair is qualitatively distinct from trust repair efforts in which the transgressor is an individual, for a number of reasons. First, when the transgressor is a collective, it can be difficult to know who is responsible and what caused the violation. How culpable various individuals are within the group is often not immediately evident. Is this a ‘bad apple’ situation – where a few rogue individuals did the wrong thing within the context of an otherwise high-integrity organization – or a ‘bad barrel’ situation, in which the transgression was the result of a corrupt or dysfunctional group and culture (Kish-Gephart, Harrison, & Trevino, 2010)? Second, when the transgressor is a collective, the people engaging in the trust repair efforts may not be directly responsible for violating trust in the first place. Organizations and groups have parties other than the transgressor who can respond on the group’s behalf (Ferrin et al., 2018; Gillespie et al., 2014).This highlights a third fundamental difference: collectives and organizations have options and capacities for repairing trust that are unavailable to individuals, such as replacing ‘guilty’ agents with new agents who symbolize different values and imposing structural constraints (Ferrin et al., 2018; Gillespie & Dietz, 2009; Gillespie et al., 2014). In this context, those affected by a collective trust violation are faced with an unusually complex and ambiguous algorithm in terms of understanding basic sense-making and relational questions, such as “Who violated my trust?” “Who is sorry?” “Have those responsible been punished and made amends?” and “Have changes been made to prevent future violations?” One clear consequence is that collectives need to engage more elements in the repertoire of trust repair efforts. When the transgressor is an individual, the onus is on them as the transgressor to communicate genuine remorse and commitment to change, and the receiver decides whether or not to open themselves up to future re-engagement on the basis of these largely symbolic communication strategies. From this perspective, the best cue for whether the transgressor has changed is the extent to which they apologize and communicate regret. But at the collective level, the best cues for whether the collective transgressor has changed is the extent to which they understand and acknowledge what went wrong and enact reforms to prevent a reoccurrence of the transgression. That is, whereas judgments of trustworthiness after a transgression by an individual focus on the individual’s character

Multilevel Trust Repair  157

and actions, when a collective violates trust, perceivers focus primarily on the trustworthiness of the organization, including the organization’s purpose, leadership, systems, processes, culture, and control mechanisms, hence demanding repair strategies that directly reform these elements (Dirks et al., 2009; Gillespie & Dietz, 2009). As such, sense-making and relational strategies need to be supplemented with structural changes, both formal and informal (Gillespie & Dietz, 2009). Research on organizational trust repair suggests a range of strategies beyond sense-making and relational mechanisms are often used, including regulatory and structural changes, informal efforts to build an ethical culture and ethical leadership, a commitment to transparency when investigating, understanding, and atoning for the trust violation, and efforts to gain the endorsement or certification of trusted entities (Bachmann et al., 2015; Dietz & Gillespie, 2012; Gillespie et al., 2014; Hurley, Gillespie, Ferrin, & Dietz, 2013). In short, scholarly work in the organizational literature has demonstrated both the complexity of repairing trust after collective-to-collective violations and the necessity of integrating a broad range of trust repair mechanisms in order to effectively restore trust. Another complexity inherent in repairing trust with a diverse collective (e.g., stakeholders of an organization) is that the strategies taken to repair trust with one set of stakeholders (e.g., customers or investors) may further undermine the trust of another set of stakeholders (e.g., employees; Eberl, Geiger, & Aßländer, 2015; Gillespie et al., 2014; Lamin & Zaheer, 2012). For example, in response to their bribery scandal, Siemens introduced tight new rules and compliance regimes, which had the desired effect of improving external stakeholders’ trust in the company, but ‘threatened’ internal employee trust (Eberl et al., 2015, p. 1217). Similarly, in an inter-organizational context, Stevens, MacDuffie, and Helper (2015) found Nissan’s attempts to recalibrate trust with their suppliers via structural reforms worked well with newer suppliers, but past suppliers were less receptive because they did not take into account previous sacrifices those suppliers had made and thus did not restore a sense of equilibrium in the relationship. Finally, in a study of organizational responses to allegations of using international sweatshops, denial and defiance hindered reputational recovery with the public (‘main street’) but not investors (‘Wall Street’), while decoupling from the transgression aided recovery with investors (Lamin & Zaheer, 2012). This challenge extends not only across stakeholder groups but also within stakeholder groups – for example, repairing the trust of diverse employees. A good example comes from a case study of a British water company’s reintegration after an integrity-based scandal. Gillespie and colleagues (2014) found within-group variance in employee responses to trust repair attempts, with some employees feeling shame for the employer’s actions and accepting the need to pay a substantial fine to demonstrate penance, while others felt angry and wanted the board to defend the organization. Recognizing this diversity, the company engaged in a broad range of trust repair approaches (e.g., supporting employees in sense-making

158  Nicole Gillespie et al.

processes, working with the industry regulator and making voluntary policy amendments, apologizing, etc.), which ultimately culminated in winning the British ‘Utility of the Year’ award and restoring trust. Findings also suggest that the company’s early obfuscation and denials were not successful. Despite their utility in response to integrity violations within individual-to-individual experimental settings (Bagdasarov et al., 2019; Kim et al., 2004), denials are not typically considered an appropriate response for repairing organizational trust, and are ineffective – even trust-damaging – in the face of evidence of guilt (Arendt, LaFlenche, & Limperopulos, 2017; Gillespie et al., 2014).

Group Dynamics in Collective-to-Collective Trust Repair Part of the complexity of collective-to-collective trust repair is that it often occurs in an inter-group context – that is, where the transgressor group and the victim group share distinct and coherent social identities triggering a range of group dynamics, which often spiral into an ‘us versus them’ mentality. Examples include where one organization violates the trust of another organization, an organization causes harm to a definable community (e.g., environmental damage within a town), senior management violates the trust of employees, or one team violates the trust of another work team. To understand how intergroup dynamics influence collective-to-collective trust repair, we draw on the social psychology literature. Two long traditions of research have (independently) made the point that trust issues are particularly fraught in the intergroup context. One line of research involves observations of how people respond to mixed-motive situations such as the prisoner’s dilemma game. Famously, these games demonstrate that rational individuals often prioritize short-term self-interest over cooperation, even if it damages everybody’s long-term interests. However, in variants of the game in which people interact in groups, the outcomes are poorer again: people in groups are more competitive and less cooperative than when people operate as individuals (Insko et al., 2001; Wildschut, Pinter, Vevea, Insko, & Schopler, 2003). This tendency – labeled the interindividual–intergroup discontinuity effect – has been attributed to various psychological mechanisms, such as increases in both greed and fear in group contexts, which leads to a less cooperative and less trusting atmosphere. Another explanation is a sense of duty to favor the in-group, a competitive norm that is partly enabled by diffusion of responsibility: whereas individuals feel solely accountable for their decisions, group members’ moral accountability can be diffused across individuals, making them feel less responsible for their (hostile) actions. Another line of research that speaks about intergroup barriers to trust is the social identity literature (Tajfel & Turner, 1979). This approach draws a qualitative distinction between people’s personal identities – the idiosyncratic feelings, beliefs, behaviors, and memories that separate oneself from other individuals – and

Multilevel Trust Repair  159

people’s social identities – the feelings, beliefs, behaviors, and memories that flow from their group memberships. If people belong to a group that is an important part of their self-definition – and if that group membership becomes salient – then their sense of self will be suffused by the normative expectations of that group (Turner, 1991). The social identity approach assumes people are motivated to see their groups as positively distinct from others’ groups, a process of intergroup differentiation that helps nourish a secure and coherent sense of self. One consequence is people implicitly trust in-group members assuming they will look out for their interests, even in the absence of a history of positive reciprocal exchanges. In contrast, group members view outgroup members through a lens of suspicion and mistrust, a lens that might exist even in the absence of a history of negative exchanges (Tanis & Postmes, 2005; Worchel, 1979). Together, these lines of research make the case that there is a default ‘trust deficit’ when people see a violation through an intergroup as opposed to an interpersonal lens. As such, people are typically less willing to give the benefit of the doubt to transgressors and are less responsive to trust repair efforts when they construe the events as part of an intergroup (as opposed to an interpersonal) dynamic. Accepting trust repair efforts – such as apologies, compensations, and promises of reform – is in itself a trust-laden act and one that should theoretically be less plausible in the intergroup context. Another complicating aspect of intergroup trust repair is that transgressor and victim groups often fundamentally differ in how they appraise and make sense of the violation. The transgressor group often gravitates toward seeing the violation through an interpersonal lens: this is something a few ‘bad apples’ have done, and the onus is on those individuals to work it out (i.e., they see the transgressor as an individual). From the victims’ point of view, however, the same issue tends to be perceived as something ‘they’ have done to ‘us.’ This tendency for victims to see transgressions through a collective lens more so than perpetrators has been labeled the “appraisal gap” (Hornsey, Okimoto, & Wenzel, 2017). The appraisal gap stems from two intergroup dynamics. First, the tendency for people to perceive a sameness or equivalence among members of outgroups (i.e., the “outgroup homogeneity effect”; Judd & Park, 1988; Linville, Fischer, & Salovey, 1989; Quattrone & Jones, 1980). This tendency is exaggerated when the exemplars are negative ones because negative events tend to be more salient and memorable – and perceived to be more diagnostic of reality – than positive events (Barlow et al., 2012; Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001). In contrast, in-group members are predisposed to ‘bad apple’ accounts of transgressions because they are more cognitively equipped to see the individual variation within their group, and their perception of their own group is not hostage to a small handful of salient exemplars. Second, members of perpetrator groups have a basic psychological motivation to ward off guilt and shame by distancing themselves from responsibility for a trust violation. As such, ‘bad apple’ accounts can be seen as part of a broader repertoire of moral disengagement efforts group members

160  Nicole Gillespie et al.

can use to avoid the painful process of accepting responsibility for a transgression (Bandura, 1999; Castano, 2008). In sum, transgressor and victim groups often have different implicit understandings of what the violation represents, who the victims and transgressors are, and thus what is required to repair trust. Another line of research that underscores the difficulty of overcoming trust violations in intergroup contexts is work on “competitive victimhood” (Noor, Shnabel, Halabi, & Nadler, 2012; Sullivan, Landau, Branscombe, & Rothschild, 2012). This research identifies that a sense of victimhood can be a desirable commodity in an intergroup context as it gives the group moral leverage and invites sympathy from third parties. Collective victimhood can provide a rallying point around which group members experience solidarity, cohesion, and identity, and is one way group members can distract and distance themselves from their own (collective) moral transgressions toward others (Wohl & Branscombe, 2008). This research raises an interesting proposition: in intergroup contexts, there can be psychological and emotional payoffs to staying in a state of mistrust and unforgiveness. This is in contrast to the interpersonal literature, where the state of unforgiveness is negatively associated with well-being (e.g., Wade, Hoyt, Kidwell, & Worthington, 2014), to the point that scholars have construed interpersonal reconciliation and forgiveness as an act of self-compassion designed to preserve mental and physical health. The added complexity associated with collective trust repair means that transgressors might need to be more creative in terms of communicating remorse and penance. Often, the responsibility falls on an individual spokesperson to communicate these sentiments, often the leader of the organization. It is questionable, however, whether this is always sufficient in the context of collective-to-collective trust violations because victims might reasonably question the extent to which the spokesperson genuinely reflects the sentiment of the wider group. In the social psychological literature, there is some evidence that trust repair efforts are more effective when they incorporate the voices of the broader group constituency (e.g., ‘grassroots remorse’), and not just the leader (Okimoto, Hornsey, & Wenzel, 2019; Wenzel et al., 2017). The notion that intergroup trust repair faces unique challenges finds support in the parallel literature on apologies and forgiveness (much like trust repair, forgiveness is operationalized as a shift from negative behaviors, cognitions, and emotions toward more positive behaviors, cognitions, and emotions). In the interpersonal literature, the effect is both large and robust: when one person apologizes to another person, they are much more likely to be forgiven. Indeed, a meta-analysis showed that an apology is a bigger predictor of interpersonal forgiveness than any individual differences variable, any relationship-oriented construct (including the closeness of the relationship), and even the size of the transgression (Fehr, Gelfand, & Nag, 2010). However, when applied to the intergroup domain, the relationship between an apology and forgiveness is small and unreliable. Although there are a couple of instances in which the presence of an intergroup apology

Multilevel Trust Repair  161

has been shown to increase forgiveness (e.g., Brown, Wohl, & Exline, 2008), there are over a dozen other experimental studies that suggest no relationship between an intergroup apology and intergroup forgiveness (see Hornsey & Wohl, 2013; Hornsey, Wohl, & Philpot, 2015 for reviews). In these studies, collective apologies typically do have significant effects on perceptions of remorse and satisfaction with the response, but in the context of what seems to be high levels of skepticism: on average, perceptions of remorse and satisfaction with the response hover at or below the mid-point, whereas the perception that the apology was a result of ulterior motives nears the ceiling of the scale. Revealingly, this skepticism is reflected also in non-experimental approaches such as longitudinal survey research (e.g., Philpot & Hornsey, 2011; Wohl, Matheson, Branscombe, & Anisman, 2013) and interview studies following real-world apologies (e.g., Chapman, 2007). To date, the literature has been more successful at highlighting the challenges associated with restoring trust in intergroup contexts than in identifying workable solutions. However, two approaches show promise. One approach – most often attributed to the common in-group identity model but sometimes referred to as the dual identity model – is to make salient what the two groups share at the superordinate level (Gaertner & Dovidio, 2000). For example, when engaging in trust repair between two departments, it would be advised to use language, values, and arguments that focus on what the departments share at the organizational level. Theoretically, such an approach has the power to reconfigure an intergroup dynamic into an intragroup dynamic, with all that implies in terms of promoting trust and cooperation. Another approach is to have outgroup communications directed through an in-group messenger. In one study, Wohl, Hornsey, and Bennett (2012) examined Canadians’ responses to an apology from an Afghan defense minister after a friendly fire incident that cost Canadian lives. When issued directly from the Afghan defense minister, the gestures of remorse were met with high levels of cynicism and were unsuccessful at promoting forgiveness. However, in some conditions, the participants were told that the Afghan defense minister has delivered his gesture of remorse to a senior Canadian general, after which the Canadian general had relayed the message to the Canadian people. In these conditions, levels of forgiveness and trust in the sincerity of the message increased by nearly two points on a seven-point scale. Although the mechanism for this effect is unclear, it is consistent with a long history of research showing that messages are absorbed more readily when they come from in-group messengers than when the same message comes from an outgroup member (Esposo, Hornsey, & Spoor, 2013; Turner, 1991), and may reflect a process of trust transference from the in-group messenger to the outgroup. In sum, for the various reasons outlined above, trust repair at the collectiveto-collective level is qualitatively different from, and considerably more complex and ambiguous than, trust repair at the levels of individual-to-individual and individual-to-collective.

162  Nicole Gillespie et al.

Discussion Developing a more integrated understanding of trust repair across levels is vital for both scholarly and practical purposes. As we have evidenced and argued in this chapter, the dynamics and processes of trust repair can differ markedly across the individual and collective levels of analysis.We have shown how the embedded nature of trust violations within broader groups and collectives can both facilitate and constrain repair.Yet existing empirical and conceptual work has paid limited attention to the multilevel embedded nature of trust repair in work contexts. Our selective review supports the observation that extant research on trust violation and repair continues to focus on trust repair at one level of analysis, often the individual level (Fulmer & Gelfand, 2012; Kim et al., 2013; Bachmann et al., 2015), rather than tackling the many complex and interesting dynamics that imbue trust repair at collective levels and across levels. To facilitate and guide future research, we develop an integrative framework of trust repair across levels of analysis (see Table 7.3). This framework draws on and synthesizes the insights on trust repair across levels reviewed in this chapter, including insights from the management and organizational literatures as well as social psychological research on trust repair, forgiveness, and reconciliation. From these literatures, we identify six key factors that influence and change the nature of trust repair across levels (see top half of Table 7.3).We propose that the five theoretical mechanisms to trust repair and their underlying questions (Bachmann et al., 2015) apply across levels: however, the cues, dynamics, and processes involved vary across levels due to the six factors (see bottom half of Table 7.3). We find theoretical and/or empirical evidence in our review that suggests the trust repair mechanisms are seriously complicated as the number of people or groups involved in committing or contributing to the transgression (number of transgressors), or affected by the transgression (number of victims), increases from one to many. That is, as trust repair moves from the individual-to-individual to the collective-to-collective level (left to right in Table 7.3). The shift to the collective level changes the experience of the violation from one directly experienced by a single victim to a mix of directly and indirectly experienced by a collective of stakeholders and groups and strengthens the influence of third parties beyond the immediate transgressor and trustor on the trust repair process (shifting from minimal and private to stronger and public). The shift to the collective level also strengthens the influence of (inter)group dynamics and processes by opening up a broad range of group and intergroup dynamics that can influence the effectiveness of trust repair, such as the appraisal gap, in-group/outgroup biases, and competitive victimhood. Finally, the strategies and resources available for trust repair broaden at the collective level. As part of our multilevel trust repair framework, we make two general propositions. First, trust repair mechanisms, dynamics, and strategies are not isomorphic across levels – rather they can vary depending on the level at which the trust repair process takes place.

Trust repair mechanisms and related questions Sense-making mechanism What happened and why? (explanations) Who is responsible/guilty? Who was affected? Relational mechanism Are those responsible genuinely remorseful? Have those responsible endured appropriate punishment or penance? Have appropriate amends been made to victims? Structural and (in)formal control mechanism Has the party reformed itself? Have changes been made to prevent future transgressions? Will this happen again? Transparency mechanism Is there transparent evidence of renewed trustworthiness? Trust transference approach Do credible parties trust the transgressor? Can certifications or audits transfer trust?

Number of transgressors Number of victims Experience of violation (by trustor) (Inter)group dynamics and processes Third parties beyond transgressor and trustor Strategies and resources available

Factors influencing trust repair

Many Direct or indirect Moderate Moderate

Individual-to-collective

Weaker influence

Victim bases perceptions largely on the words and actions of individual transgressor

Often unclear requiring investigation, with divergence across groups due to group dynamics (e.g., ‘appraisal gap’)

Strong Strong (public) Typically broader

Many

Collective-to-collective

Collective bases on Difficult to ascertain if whole collective feels words and actions remorse. Often divergent views within and across of individual groups on what is required and appropriate for transgressor, as well trust repair as views of other group members Collective bases on words and actions of collective (noting there may transgressor, views of other group members, be divergence in view of external assessors, and reforms by the views) transgressing group Collective bases on words, actions, and reporting of collective transgressor, views of other group members, and views of external assessors Stronger influence

Known to transgressor Typically known to Known to victim transgressor Known to victims

One One Direct Weak Minimal (private) Typically limited

Individual-to-individual

Level of trust repair

TABLE 7.3 An integrative framework of trust repair across levels of analysis

Multilevel Trust Repair  163

164  Nicole Gillespie et al.

Second, due to the culmination of the six factors at the top of Table 7.3, the complexity and challenges involved in repairing trust increase as one moves from the individual to the collective level. As we evidence in our review, the collective level introduces challenges and complexities to trust repair. This does not mean that trust repair at the interpersonal level will necessarily be easier or require less effort, per se. Rather, there will be more complexity and challenges involved due to the increased number of transgressors and/or victims, trustors’ various direct and indirect experiences of the violation, the greater influence of group dynamics and processes, and the stronger involvement of third parties. The broader set of strategies and resources typically available at the collective level compared to the interpersonal level can help address some of these additional complexities and challenges. By identifying and theorizing the factors that influence, challenge, and complicate trust repair across levels within organizations, our framework provides guidance for future research on multilevel trust repair.

Future Research Agenda Third-Party and Group Influences on Trust Repair Our review and framework raise pertinent questions around the role of group dynamics and third parties in influencing trust repair at the collective level. In line with the relational approach to trust repair, the disequilibrium caused by the violation enables collective victims to not only decry the violation but also to ‘dictate’ to some extent the terms of the repair strategy (i.e., what is required to restore trust). This represents a new line of research on how collective level trust violations lead to ‘negotiations’ over the terms and conditions of the necessary or appropriate repair strategies, and more broadly, how the balance and enactment of power in the relationship between the transgressor and the collective victim influences trust repair dynamics and mechanisms. To date, there has been little exploration of the role of power and negotiations in the process of repair. We call for a multilevel approach to this line of enquiry, given that negotiation and power dynamics are likely to operate differently when the victim is a diverse collective, compared to an individual. A related line of future research is to examine how the diversity versus homogeneity of the collective set of victims influences repair processes, as well as unity in their approach and ‘demands’ for repair. For example, does victims’ use of collective power and unified voice (e.g., victims binding together to form a class action against a transgressing organization) facilitate trust repair (e.g., by clarifying sense-making of the violation from the victims’ perspective, and demands for relational and structural reforms) or potentially undermine it by triggering stronger ‘us versus them’ inter-group dynamics? Further investigation of hitherto underexplored intra-group processes may be another fruitful avenue of future research. For instance, while emotional contagion has been proposed to influence trust (Fulmer & Gelfand, 2012), its influence

Multilevel Trust Repair  165

on trust repair has yet to receive attention and warrants examination. Emotional contagion may change trust repair dynamics, and this may have cross-level effects. As an example, an individual may experience a trust violation at work and feel negative emotions. These emotions may spread to other members of the individual’s workgroup, impacting the group’s perceptions of the violation and changing the trust repair dynamics at play. Understanding the role of third parties and group dynamics on individual-tocollective trust repair is also a promising avenue for future research. We observed that trust repair responses by leaders toward collectives are often defensive and qualified. We speculated that it may be due to the desire to save face in public forums, as well as the influence of third party advisors, who act to protect and reduce the negative impact (even responsibility) of the transgression on the transgressor, as opposed to focusing on relational repair. An alternative explanation is that more defensive reactions may be due to the agency constraints on the leader (i.e., to represent and protect a particular constituent viewpoint, rather than adopting a more conciliatory approach to appeal to a diverse audience). Future research is required to test out these various potential influencers and rigorously examine how leaders attempt to repair trust with collectives, and whether, how, and why this differs from their approach to repairing trust with individuals. We proposed that third parties play a stronger role at the collective level than the individual level. However, this does not mean that third parties play no role at the individual level, with social psychological research suggesting third-party intervention in interpersonal violations (e.g., by management) can potentially facilitate or hinder trust repair (Okimoto, 2008; Okimoto & Wenzel, 2014). It is important to recognize the potential role the broader group might also play in influencing interpersonal trust repair. For example, often the organizational climate (e.g., forgiveness climate; Fehr & Gelfand, 2012) defines the norms dictating what constitutes an appropriate action following a violation, and even how important it is to repair trust. Thus, even without actively intervening, the trust repair expectations of the broader organizational group can shape the form of interpersonal trust repair. This is a ripe area for future investigation, given most interpersonal work relationships are embedded within teams, groups, and departments.

Deepening Insight on Trust Repair Mechanisms and Dynamics Across Levels As discussed in our framework, a distinguishing aspect of a trust violation at the collective level is that it may be directly or indirectly experienced by members of the collective. To date, there has been little empirical examination of how the experience of the violation (direct or indirect) may change the mechanisms or dynamics of trust repair. How does the experience of the violation influence attribution and sense-making processes, and expectations for relational repair? Are the emotional processes involved in trust repair different for those who have experienced the violation directly versus indirectly? Can the trust of those indirectly

166  Nicole Gillespie et al.

affected by the violation be repaired predominately by rebuilding the trust of those directly affected? In what contexts might trust repair between those directly and indirectly affected be decoupled and require distinct processes? As these questions highlight, this is a ripe area for future examination. As demonstrated in Table 7.1, empirical research on the repair of trust in individual-to-individual relationships (Dirks et al., 2011; Lewicki & Brinsfield, 2017) has largely focused on sense-making and relational mechanisms, such as apologies and denials. These two mechanisms are also widely studied in collective-level trust repair. In our review, we focus on how these mechanisms may operate differently across levels. However, it would be remiss of us not to acknowledge that research suggests some similarities in how these strategies function across levels, particularly the repair strategies employed by organizations and individual leaders (collective-to-collective and individual-to-collective). Potential reasons for this are that individual leader responses and organizational responses are perceived similarly, given leaders are often viewed as the embodiment of the organization (Ranft, Zinko, Ferris, & Buckley, 2006), and leaders (and their responses) are constrained within organizational, institutional, and constitutional structures. In contrast to the interpersonal level, considerable research on trust repair at the collective level has studied the role of structural and informal controls, and to a lesser degree the transparency and transference mechanisms. Based on our review and theorizing, and in line with proposition 2, we suggest that sense-making and relational mechanisms are harder to achieve at the collective level than at the individual level. In contrast, we speculate that structural and cultural controls and transparency mechanisms to repair trust may be easier to demonstrate at the collective level (e.g., through group and organizational change programs). We encourage future research to engage with and test these ideas.

Examining the Simultaneous Embedded Enactment of Trust Repair Across Levels We have discussed the three forms of trust repair (individual–individual; individual–collective; collective–collective) as though they operate separately. However, in practice, they can co-occur in an embedded manner. To illustrate, consider the context of two elite athletes of a team found to have cheated. The recovery of internal and external stakeholder trust is likely to involve each of the athletes individually engaging in trust repair with key individuals directly or indirectly harmed by the scandal (e.g., the captain, coach, family members), as well as with key groups (e.g., their team, the club, their fans). It is also likely to involve the captain and coach (as representatives of the team) repairing trust with the broader team and club. Understanding how such simultaneous, multilevel trust repair processes play out and what influences their combined effectiveness is a rich area for future research. For example, how important is alignment between the trust repair strategies across transgressors and representatives? Signaling theory (e.g.,

Multilevel Trust Repair  167

Przepiorka & Diekmann, 2013) would suggest that aligned messages from multiple sources might be more effective at changing perceptions, including trust perceptions, than single voices. We call for future research to examine how multiple leaders and actors at different levels within a transgressing organization, and at different levels of proximity to the violation, may simultaneously play a role in trust repair. One of the challenges of collective level trust repair is that the source of the trust violation might stem from one particular subgroup within the organization. Although the public figure charged with trust repair is often the organizational leader (e.g., the CEO) or spokesperson, the repair effort might be more appropriate (and ostensibly more effective) when coming from the offending subgroup. Not only does this target the specific ‘source of the problem,’ but it also leverages the general tendency for people to trust smaller groups relative to larger groups: in the absence of other information, numerical smallness seems to be an implicit cue for the benevolence of an entity (La Macchia, Louis, Hornsey, & Leonardelli, 2016). Okimoto and colleagues (2019) found that when those at the ‘grassroots’ of the transgression expressed remorse, it was more effective for forgiveness. Future research is required to examine if this effect translates to the repair of trust.

Examining the Effects of Levels Beyond the Organizational Boundary In this chapter, we restricted our focus to what occurs within or between groups and organizations, in line with the vast majority of trust repair literature. However, this ignores the fact that groups and organizations are embedded in and conditioned by higher-level institutions, as well as external forces and macro trends (e.g., global crises, pandemics, recessions, globalization, technological disruption, and automation) that are likely to affect trust breakdown and repair (Möllering, 2006; Siebert, Martin, Bozic, & Docherty, 2015). As an example, in their analysis of trust repair in banks in the aftermath of the global financial crisis, Gillespie, Hurley, Dietz, and Bachmann (2012) concluded that fundamental changes in the governance and regulation of the financial sector were required, in addition to organizational-level trust repair strategies. Hence, we recommend that future research on multilevel trust repair consider the role of the institutional, legal, and regulatory structures and trends within which groups and organizations are embedded (Gillespie & Siebert, 2018). Such analysis may help to explain why some groups and organizations fail to repair trust despite genuine efforts to do so (Child & Rodrigues, 2004).

A Contextualized and Processual Approach to Examining Multilevel Trust Repair An observation of the literature is that there is still an overreliance on laboratorybased work, particularly at the interpersonal level. This can lead to inappropriate

168  Nicole Gillespie et al.

assertions of what may work and be appropriate to repair trust within the workplace. For example, we might conclude from lab-based experimental work that compensation is effective in repairing interpersonal trust at work. However, a caveat is that these studies typically use a variant of an economic game (e.g., trust game, prisoner’s dilemma, dictator game), which has no lasting consequences for participants outside of the experiment. The external validity of these games has been questioned (Kee & Knox, 1970). Indeed, a recent lab-field experiment, systematic review, and meta-analysis by Galizzi and Navarro-Martinez (2019) found that economic games are mediocre at predicting social behavior in the field and social preferences from the past. Further, economic games do not incorporate all facets of trust (Ben-Ner & Halldorsson, 2010) and may not reflect a state of trust at all (Kee & Knox, 1970), but rather financial (non)cooperation. We encourage future research on multilevel trust to embrace a diversity of methods. In particular, we join other trust scholars in calling for more empirical work that examines the process of trust repair – that is, that employs methodologies that focus on the processual nature of trust repair over time (e.g., Bachmann et al., 2015; Lewicki, Tomlinson, & Gillespie, 2006; Möllering, 2013) – in the field. This is particularly important for multilevel research, given trust repair efforts and dynamics may occur at different rates across levels (e.g., interpersonal trust repair may occur more quickly than collective-level trust repair) and be heavily influenced by the context. Petriglieri’s (2015) examination of relationship repair between BP executives and the organization in the aftermath of the Gulf of Mexico oil rig explosion is an example of a processual field study. It found that repair was co-created by the organization and its members, over time. Another example is Bijlsma-Frankema, Sitkin, and Weibel’s (2015) case study approach to developing a dynamic process model of distrust development. As highlighted in this study, organizational trust failures not only diminish trust, but also typically trigger distrust, and overcoming perceptions of value incongruence helps overcome distrust, enabling trust to be rebuilt. Examining the processes that enable distrust to be overcome in the context of multilevel trust repair is another fertile ground for future research. In conclusion, we see the multilevel investigation of trust repair processes as a field of inquiry that is in its infancy. Our multilevel trust repair framework and selected review identify many more interesting, theoretically rich, and practically relevant questions than they answer. Our hope is that our paper will serve as an inspiration and useful guide for future research that aims to systematically investigate trust repair from a multilevel perspective.

References Arendt, C., LaFleche, M., & Limperopulos, M. A. (2017). A qualitative meta-analysis of apologia, image repair, and crisis communication: Implications for theory and practice. Public Relations Review, 43(3), 517–526. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.pubr​​ev​.20​​​17​.03​​.005

Multilevel Trust Repair  169

Augustine, D. (2012). Good practice in corporate governance: Transparency, trust, and performance in the microfinance industry. Business & Society, 51(4), 659–676. https:// doi​.org​/10​.1177​/0007650312448623 Bachmann, R., Gillespie, N., & Priem, R. (2015). Repairing trust in organizations and institutions: Towards a conceptual framework. Organization Studies, 36(9), 1123–1142. https://doi​.org​/10​.1177​/0170840615599334 Bachmann, R., & Inkpen, A. C. (2011). Understanding institutional-based trust building processes in inter-organizational relationships. Organization Studies, 32(2), 281–301. https://doi​.org​/10​.1177​/0170840610397477 Bagdasarov, Z., Connelly, S., & Johnson, J. F. (2019). Denial and empathy: Partners in employee trust repair? Frontiers in Psychology, 10, 19. https://doi​.org​/10​.3389​/fpsyg​. 2019​.00019 Bandura, A. (1999). Moral disengagement in the perpetration of inhumanities. Personality and Social Psychology Review, 3(3), 193–209. https​:/​/do​​i​.org​​/10​.1​​207​/s​​15327​​957ps​​​pr030​​3_3 Barlow, F. K., Paolini, S., Pedersen, A., Hornsey, M. J., Radke, H. R. M., Harwood, J., Rubin, M., & Sibley, C. G. (2012). The contact caveat: Negative contact predicts increased prejudice more than positive contact predicts reduced prejudice. Personality and Social Psychology Bulletin, 38, 1629–1643. https://doi​.org​/10​.1177​/0146167212457953 Basford, T. E., Offermann, L. R., & Behrend, T. S. (2014). Please accept my sincerest apologies: Examining follower reactions to leader apology. Journal of Business Ethics, 119(1), 99–117. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10551​​-01​2-​​1613-​y Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than good. Review of General Psychology, 5(4), 323–370. https​:/​/do​​i​.org​​/10​.1​​037​/1​​089​-2​​68​ 0​.5​​.4​.32​3 Ben-Ner, A., & Halldorsson, F. (2010). Trusting and trustworthiness: What are they, how to measure them, and what affects them. Journal of Economic Psychology, 31(1), 64–79. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.joep​​.200​9​​.10​.0​​01 Benoit, W. L. (2006). Retrospective versus prospective statements and outcome of presidential elections. Journal of Communication, 56(2), 331–345. https​:/​/do​​i​.org​​/ 10​.1​​111​/j​​.1460​​-2466​​.200​6​​.0002​​2.x Benoit,W. L. (2017). Image repair on the Donald Trump “Access Hollywood” video: “Grab them by the p*ssy”. Communication Studies, 68(3), 243–259. https​:/​/do​​i​.org​​/10​.1​​080​/ 1​​05109​​74​.20​​​17​.13​​31250​ Bijlsma-Frankema, K., Sitkin, S. B., & Weibel, A. (2015). Distrust in the balance: The emergence and development of intergroup distrust in a court of law. Organization Science, 26(4), 1018–1039. https://doi​.org​/10​.1287​/orsc​.2015​.0977 Bottom,W. P., Gibson, K., Daniels, S. E., & Murnighan, J. K. (2002).When talk is not cheap: Substantive penance and expressions of intent in rebuilding cooperation. Organization Science, 13, 497–513. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​3​.​5​.4​​97​.78​​16 Brewer, M. B. (1979). In-group bias in the minimal intergroup situation: A cognitivemotivational analysis. Psychological Bulletin, 86(2): 307–324. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​033​-2​​909​​.8​​6​.2​.3​​07 Brodt, S. E., & Neville, L. (2013). Repairing trust to preserve balance: A balance-theoretic approach to trust breach and repair in groups. Negotiation and Conflict Management Research, 6(1), 49–65. https://doi​.org​/10​.1111​/ncmr​.12003 Brooks, T. (Ed.). (2012). Hegel's philosophy of right. Hoboken, NJ: John Wiley & Sons. Brown, B. R. (1968). The effects of need to maintain face on interpersonal bargaining. Journal of Experimental Social Psychology, 4(1), 107–122. https​:/​/do​​i​.org​​/10​.1​​016​/ 0​​​022​-1​​031(6​8)900​53-X

170  Nicole Gillespie et al.

Brown, R. P., Wohl, M. J. A., & Exline, J. J. (2008). Taking up offenses: Secondhand forgiveness and group identification. Personality and Social Psychology Bulletin, 34, 1406– 1419. https://doi​.org​/10​.1177​/0146167208321538 Byrne, A., Barling, J., & Dupré, K. E. (2014). Leader apologies and employee and leader well-being. Journal of Business Ethics, 121(1), 91–106. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10551​​01​3-​​1685-​3 Carlsmith, K. M. (2008). On justifying punishment: The discrepancy between words and actions. Social Justice Research, 21(2), 119–137. https​:/​/do​​i​.org​​/10​.1​​007​/s​​11211​​-00​8​​0068-​x Castano, E. (2008). On the perils of glorifying the in-group: Intergroup violence, in-group glorification, and moral disengagement. Social and Personality Psychology Compass, 2(1), 154–170. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1751​​-9004​​.200​7​​.0004​​0.x Chapman, A. R. (2007). Truth commissions and intergroup forgiveness: The case of the South African truth and reconciliation commission. Peace and Conflict: Journal of Peace Psychology, 13, 51–69. https://doi​.org​/10​.1037​/h0094024 Child, J., & Rodrigues, S. B. (2004). Repairing the breach of trust in corporate governance. Corporate Governance: An International Review, 12(2), 143–152. https​:/​/do​​i​.org​​/10​.1​​111​/ j​​.1467​​-8683​​.200​4​​.0035​​3.x Cropanzano, R., Goldman, B., & Folger, R. (2005). Self-interest: Defining and understanding a human motive. Journal of Organizational Behavior, 26(8), 985–991. https://doi​.org​/10​. 1002​/job​.353 De Cremer, D., & Schouten, B. C. (2008). When apologies for injustice matter: The role of respect. European Psychologist, 13(4), 239–247. https​:/​/do​​i​.org​​/10​.1​​027​/1​​016​-9​​040​​. 1​​3​.4​.2​​39 Desmet, P. T. M., De Cremer, D., & Van Dijk, E. (2011a). In money we trust? Financial compensations as a means to repair trust in the aftermath of distributive harm. Organizational Behavior and Human Decision Processes,114,75–86.https​:/​/do​​i​.org​​/10​.1​​016​/ j​​.obhd​​p​.20​1​​0​.10.​​006 Desmet, P.T. M., De Cremer, D., & Van Dijk, E. (2011b).Trust recovery following voluntary or forced financial compensations in the trust game: The role of trait forgiveness. Personality and Individual Differences, 51, 267–273. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.paid​​.201​0​​. 05​.0​​27 Dietz, G., & Gillespie, N. (2012). The recovery of trust: Case studies of organizational failures and trust repair. London: Institute of Business Ethics. Dirks, K. T., Kim, P. H., Ferrin, D. L., & Cooper, C. D. (2011). Understanding the effects of substantive responses on trust following a transgression. Organizational Behavior and Human Decision Processes, 114(2), 87–103. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​1​​0​.10.​​003 Dirks, K. T., Lewicki, R., & Zaheer, A. (2009). Repairing relationships within and between organizations: Building a conceptual foundation. Academy Of Management Review, 34(1), 68–84. https​:/​/do​​i​.org​​/10​.5​​465​/A​​MR​.20​​09​​.35​​71328​5 Dirks, K.T., & Skarlicki, D. P. (2004).Trust in leaders: Existing research and emerging issues. In R. M. Kramer & K. S. Cook (Eds.), The Russell Sage Foundation series on trust: Trust and distrust in organizations: Dilemmas and approaches (pp. 21–40), New York: Russel Sage Foundation. Durkheim, E. (1964). Essays on sociology and philosophy (Vol. 1151). New York: Harper & Row. Eberl, P., Geiger, D., & Aßländer, M. (2015). Repairing trust in an organization after integrity violations: The ambivalence of organizational rule adjustments. Organization Studies, 36, 1205–1235. https://doi​.org​/10​.1177​/0170840615585335

Multilevel Trust Repair  171

Elangovan, A. R., Auer-Rizzi, W., & Szabo, E. (2015). It’s the act that counts: Minimizing post-violation erosion of trust. Leadership & Organization Development Journal, 36, 81–96. https​:/​/do​​i​.org​​/10​.1​​108​/L​​ODJ​-0​​7​​-201​​2​-009​0 Elangovan, A. R., & Shapiro, D. L. (1998). Betrayal of trust in organizations. Academy of Management Review, 23(3), 547–566. https://doi​.org​/10​.2307​/259294 Esposo, S. R., Hornsey, M. J., & Spoor, J. R. (2013). Shooting the messenger: Outsiders critical of your group are rejected regardless of argument quality. British Journal of Social Psychology, 52, 386–395. https://doi​.org​/10​.1111​/bjso​.12024 Fehr, E., & Gächter, S. (2002). Altruistic punishment in humans. Nature, 415(6868), 137– 140. https://doi​.org​/10​.1038​/415137a Fehr, R., & Gelfand, M. J. (2012). The forgiving organization: A multilevel model of forgiveness at work. Academy of Management Review, 37(4), 664–688. https://doi​.org​/10​. 5465​/amr​.2010​.0497 Fehr, R., Gelfand, M. J., & Nag, M. (2010). The road to forgiveness: A meta-analytic synthesis of its situational and dispositional correlates. Psychological Bulletin, 136, 894. https://doi​.org​/10​.1037​/a0019993 Ferrin, D. L., Cooper, C., Dirks, K., & Kim, P. (2018). Heads will roll! Routes to effective trust repair in the aftermath of a CEO transgression. Journal of Trust Research. 8, 7–30. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​17​.14​​19877​ Ferrin, D. L., Dirks, K. T., & Shah, P. P. (2006). Direct and indirect effects of third-party relationships on interpersonal trust. Journal of Applied Psychology, 91(4), 870–883. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​1​.4​.8​​70 Ferrin, D. L., Kim, P. H., Cooper, C. D., & Dirks, K. T. (2007). Silence speaks volumes: The effectiveness of reticence in comparison to apology and denial for responding to integrity- and competence-based trust violations. Journal of Applied Psychology, 92, 893–908. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​.4​.8​​93 Folger, R. (2001). Fairness as deonance. In S. Gilliland, D. Steiner, & D. Skarlicki (Eds.), Theoretical and cultural perspectives on organizational justice (pp. 3–33). Greenwich, CT: Information Age Publishing. Fulmer, A., & Dirks, K. (2018). Multilevel trust: A theoretical and practical imperative. Journal of Trust Research, 8(2), 137–141. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​.15​​31657​ Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​.org​/ 10​.1177​/0149206312439327 Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: The common ingroup identity model. Philadelphia, PA: Psychology Press. Galizzi, M. M., & Navarro-Martinez, D. (2019). On the external validity of social preference games: A systematic lab-field study. Management Science, 65(3), 976–1002. https://doi​. org​/10​.1287​/mnsc​.2017​.2908 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34(1), 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 Gillespie, N., Dietz, G., & Lockey, S. (2014). Organizational reintegration and trust repair after an integrity violation: A case study. Business Ethics Quarterly, 24(3), 371–410. https://doi​.org​/10​.5840​/beq2014437 Gillespie, N., Hurley, R., Dietz, G., & Bachmann, R. (2012). Restoring institutional trust after the global financial crisis. In R. M. Kramer & T. L. Pittinsky (Eds.), Restoring trust in organizations and leaders: Enduring challenges and emerging answers (pp. 186–216). New York: Oxford University Press. https​:/​/do​​i​.org​​/10​.1​​093​/a​​cprof​​:oso/​​97801​​99756​​0​87​. 0​​03​.00​​10

172  Nicole Gillespie et al.

Gillespie, N., & Siebert, S. (2018). Organizational trust repair. In R. Searle, A. Nienenbar, & S. Sitkin (Eds.), The Routledge companion to trust (pp. 284–301). London: Routledge. https://doi​.org​/10​.4324​/9781315745572​-20 Grimmelikhuijsen, S. G., Porumbescu, G., Hong, B., & Im, T. (2013). The effect of transparency on trust in government: A cross-national comparative experiment. Public Administration Review, 73(4), 575–586. https://doi​.org​/10​.1111​/puar​.12047 Gromet, D. M. (2012). Restoring the victim: Emotional reactions, justice beliefs, and support for reparation and punishment. Critical Criminology, 20(1), 9–23. https​:/​/do​​i​. org​​/10​.1​​007​/s​​10612​​-01​1-​​9146-​8 Haesevoets, T., De Cremer, D.,Van Hiel, A., & Van Overwalle, F. (2018). Understanding the positive effect of financial compensation on trust after norm violations: Evidence from fMRI in favor of forgiveness. Journal of Applied Psychology, 103(5), 578–590. https://doi​. org​/10​.1037​/apl0000271 Haesevoets, T., Folmer, C. R., De Cremer, D., & Van Hiel, A. (2013). Money isn’t all that matters: The use of financial compensation and apologies to preserve relationships in the aftermath of distributive harm. Journal of Economic Psychology, 35, 95–107. https​:/​/ do​​i​.org​​/10​.1​​016​/j​​.joep​​.201​3​​.02​.0​​03 Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108(4), 814–834. https​:/​/do​​i​.org​​/10​.1​​037​/0​​033​2​​95X​​.1​​08​.4.​​814 Hinsz, V. B., Tindale, R. S., & Vollrath, D. A. (1997). The emerging conceptualization of groups as information processors. Psychological Bulletin, 121(1), 43–64. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​033​-2​​909​​.1​​21​.1.​​43 Hornsey, M. J., Okimoto, T. G., & Wenzel, M. (2017). The appraisal gap: Why victim and transgressor groups disagree on the need for a collective apology. European Journal of Social Psychology, 47, 135–147. https://doi​.org​/10​.1002​/ejsp​.2279 Hornsey, M. J., & Wohl, M. J. A. (2013). We are sorry: Intergroup apologies and their tenuous link with intergroup forgiveness. European Review of Social Psychology, 24, 1–31. https​:/​/do​​i​.org​​/10​.1​​080​/1​​04632​​83​.20​​​13​.82​​2206 Hornsey, M. J., Wohl, M. J. A., & Philpot, C. R. (2015). Collective apologies and their effects on forgiveness: Pessimistic evidence but constructive implications. Australian Psychologist, 50, 106–114. https://doi​.org​/10​.1111​/ap​.12087 Hurley, R., Gillespie, N., Ferrin, D., & Dietz, G. (2013). Beyond rogue employees and bad apples: Engineering trustworthy organizations. MIT Sloan Management Review, 54(4), 75–82. Insko, C. A., Schopler, J., Gaertner, L., Wildschut, T., Kozar, R., Pinter, B. … Montoya, M. R. (2001). Interindividual-intergroup discontinuity reduction through anticipation of future interaction. Journal of Personality and Social Psychology, 80, 95–111. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​022​-3​​51​4​.8​​0​.1​.9​5 Judd, C. M., & Park, B. (1988). Out-group homogeneity: Judgments of variability at the individual and group levels. Journal of Personality and Social Psychology, 54(5), 778–788. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.5​​4​.5​.7​​78 Kee, H. W., & Knox, R. E. (1970). Conceptual and methodological considerations in the study of trust and suspicion. Journal of Conflict Resolution, 14(3), 357–366. https​:/​/do​​i​. org​​/10​.1​​177​/0​​02200​​2770​0​​14003​​07 Kim, P. H., Cooper, C. D. Dirks, K.T., & Ferrin, D. L. (2013). Repairing trust with individuals and groups. Organizational Behavior and Human Decision Processes, 120,1–14. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.obhd​​p​.20​1​​2​.08.​​004

Multilevel Trust Repair  173

Kim, P. H., Dirks, K. T., & Cooper, C. D. (2009). The repair of trust: A dynamic bilateral perspective and multilevel conceptualization. Academy of Management Review, 34, 410– 422. https​:/​/do​​i​.org​​/10​.5​​465​/A​​MR​.20​​09​​.40​​63188​7 Kim, P. H., Dirks, K. T., Cooper, C. D., & Ferrin, D. L. (2006). When more blame is better than less:The implications of internal vs. external attributions for the repair of trust after a competence- vs. integrity-based trust violation. Organizational Behavior and Human Decision Processes, 99, 49–65. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​0​​5​.07.​​002 Kim, P. H., Ferrin, D. L., Cooper, C. D., & Dirks, K. T. (2004). Removing the shadow of suspicion: The effects of apology versus denial for repairing competence- versus integrity-based trust violations. Journal of Applied Psychology, 89, 104–118. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​9​.1​.1​​04 Kish-Gephart, J. J., Harrison, D. A., & Trevino, L. K. (2010). Bad apples, bad cases, and bad barrels: Meta-analytic evidence about sources of unethical decisions at work. Journal of Applied Psychology, 95(1), 1–31. https://doi​.org​/10​.1037​/a0017103 Kramer, R. M., & Lewicki, R. J. (2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annals, 4(1), 245–277. https​:/​/do​​i​.org​​/10​.1​​080​/1​​94165​​20​.20​​​10​.48​​7403 La Macchia, S. T., Louis, W. R., Hornsey, M. J., & Leonardelli, G. J. (2016). In small we trust: Lay theories about small and large groups. Personality and Social Psychology Bulletin, 42, 1321–1334. https://doi​.org​/10​.1177​/0146167216657360 Lamin, A., & Zaheer, S. (2012).Wall Street versus Main Street: Firm strategies for defending legitimacy and their impact on different stakeholders. Organization Science, 23(1): 47–66. https://doi​.org​/10​.1287​/orsc​.1100​.0631 Lewicki, R. J., & Brinsfield, C. (2017). Trust repair. Annual Review of Organizational Psychology & Organizational Behavior, 4, 287–313. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​org​​psych​​-03​25​​16​-11​​3147 Lewicki, R. J., Polin, B., & Lount Jr, R. B. (2016). An exploration of the structure of effective apologies. Negotiation and Conflict Management Research, 9(2), 177–196. https:// doi​.org​/10​.1111​/ncmr​.12073 Lewicki, R. J., Tomlinson, E. C., & Gillespie, N. (2006). Models of interpersonal trust development: Theoretical approaches, empirical evidence, and future directions. Journal of Management, 32(6), 991–1022. https://doi​.org​/10​.1177​/0149206306294405 Linville, P.W., Fischer, G.W., & Salovey, P. (1989). Perceived distributions of the characteristics of in-group and out-group members: Empirical evidence and a computer simulation. Journal of Personality and Social Psychology, 57(2), 165–188. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​022​-3​​514​​.5​​7​.2​.1​​65 Ma, F., Wylie, B. E., Luo, X., He, Z., Jiang, R., Zhang, Y., & Evans, A. D. (2019). Apologies repair trust via perceived trustworthiness and negative emotions. Frontiers in Psychology, 10, 758. https://doi​.org​/10​.3389​/fpsyg​.2019​.00758 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi​ . org​/10​.2307​/258792 McEvily, B., Perrone,V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, 14(1), 91–103. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.9​​1​.128​​14 McEvily, B., Zaheer, A., & Soda, G. (2021). Network trust. In N. Gillespie, A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. New York, NY: Routledge. Möllering, G. (2006). Trust: Reason, routine, reflexivity. Oxford: Elsevier.

174  Nicole Gillespie et al.

Möllering, G. (2013). Process view of trusting and crisis. In R. Bachmann & A. Zaheer, (Eds.), Handbook of advances in trust research (pp. 258–306), Cheltenham: Edward Elgar Publishing. https://doi​.org​/10​.2139​/ssrn​.2109376 Mueller, F., Carter, C., & Whittle, A. (2015). Can audit (still) be trusted? Organization Studies, 36(9), 1171–1203. https://doi​.org​/10​.1177​/0170840615585336 Mullen, E., & Okimoto,T. (2015). Compensatory justice. In M. Ambrose & R. Cropanzano (Eds.), Oxford handbook of justice in work organizations (pp. 477–496). New York: Oxford University Press. Nakayachi, K, & Watabe, M. (2005). Restoring trustworthiness after adverse events: The signaling effects of voluntary “Hostage Posting” on trust. Organizational Behavior and Human Decision Processes, 97(1), 1–17. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​0​​5​.02.​​001 Noor, M., Shnabel, N., Halabi, S., & Nadler, A. (2012). When suffering begets suffering: The psychology of competitive victimhood between adversarial groups in violent conflicts. Personality and Social Psychology Review, 16, 351–374. https://doi​.org​/10​.1177​/ 1088868312440048 Okimoto, T. G. (2008). Outcomes as affirmation of membership value: Material compensation as an administrative response to procedural injustice. Journal of Experimental Social Psychology, 44(5), 1270–1282. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jesp​​.200​8​​.04​.0​​09 Okimoto,T. G., Hornsey, M. J., & Wenzel, M. (2019).The power of grassroots expressions of remorse for promoting intergroup forgiveness. Journal of Experimental Social Psychology, 80, 39–51. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jesp​​.201​8​​.10​.0​​03 Okimoto, T. G., & Wenzel, M. (2009). Punishment as restoration of group and offender values following a transgression: Value consensus through symbolic labelling and offender reform. European Journal of Social Psychology, 39(3), 346–367. https://doi​.org​/ 10​.1002​/ejsp​.537 Okimoto, T. G., & Wenzel, M. (2014). Bridging diverging perspectives and repairing damaged relationships in the aftermath of workplace transgressions. Business Ethics Quarterly, 24(3), 443–473. https://doi​.org​/10​.5840​/beq201471515 Petriglieri, J. L. (2015). Co-creating relationship repair: Pathways to reconstructing destabilized organizational identification. Administrative Science Quarterly, 60(3), 518– 557. https://doi​.org​/10​.1177​/0001839215579234 Pfarrer M., DeCelles, K., Smith, K., & Taylor, M. (2008). After the fall: Reintegrating the corrupt organization. Academy of Management Review, 33(3), 730–749. https​:/​/do​​i​.org​​/ 10​.5​​465​/a​​mr​.20​​08​​.32​​46575​7 Philpot, C. R., & Hornsey, M. J. (2008). What happens when groups say sorry: The effect of intergroup apologies on their recipients. Personality and Social Psychology Bulletin, 34, 474–487. https://doi​.org​/10​.1177​/0146167207311283 Philpot, C. R., & Hornsey, M. J. (2011). Memory for intergroup apologies and its relationship with forgiveness. European Journal of Social Psychology, 41, 96–106. https:// doi​.org​/10​.1002​/ejsp​.741 Pirson, M., & Malholtra, D. (2011). Foundations of organizational trust: What matters to different stake-holders? Organization Science, 22(4), 1087–1104. https://doi​.org​/10​. 1287​/orsc​.1100​.0581 Poppo, L., & Schepker, D. J. (2010). Repairing public trust in organizations. Corporate Reputation Review, 13(2), 124–141. https://doi​.org​/10​.1057​/crr​.2010​.12 Przepiorka, W., & Diekmann. A. (2013). Temporal embeddedness and signals of trustworthiness: Experimental tests of a game theoretic model in the United Kingdom, Russia, and Switzerland. European Sociological Review, 29(5), 1010–1023. https://doi​.org​/ 10​.1093​/esr​/jcs078

Multilevel Trust Repair  175

Quattrone, G. A., & Jones, E. E. (1980). The perception of variability within in-groups and out-groups: Implications for the law of small numbers. Journal of Personality and Social Psychology, 38(1), 141–152. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.3​​8​.1​.1​​41 Ranft, A. L., Zinko, R., Ferris, G. R., & Buckley, M. R. (2006). Marketing the image of management:The costs and benefits of CEO reputation. Organizational Dynamics, 35(3), 279–290. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.orgd​​yn​.20​​​06​.05​​.003 Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Schniter, E., & Sheremeta, R. M. (2014). Predictable and predictive emotions: Explaining cheap signals and trust re-extension. Frontiers in Behavioural Neuroscience, 8, 401. https:// doi​.org​/10​.3389​/fnbeh​.2014​.00401 Schniter, E., Sheremeta, R. M., & Sznycer, D. (2013). Building and rebuilding trust with promises and apologies. Journal of Economic Behavior & Organization, 94, 242–256. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jebo​​.201​2​​.09​.0​​11 Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007).An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32(2), 344–354. https​:/​/ do​​i​.org​​/10​.5​​465​/a​​mr​.20​​07​​.24​​34841​0 Schweitzer, M. E., Hershey, J. C., & Bradlow, E. T. (2006). Promises and lies: Restoring violated trust. Organizational Behavior and Human Decision Processes, 101(1), 1–19. https​:/​/ do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​0​​6​.05.​​005 Shapiro, D. L. (1991). The effects of explanations on negative reactions to deceit. Administrative Science Quarterly, 36(4), 614–630. https://doi​.org​/10​.2307​/2393276 Siebert, S., Martin, G., Bozic, B., & Docherty, I. (2015). Looking ‘beyond the factory gates’: Towards more pluralist and radical approaches to intraorganizational trust research. Organization Studies, 36, 1033–1062. https://doi​.org​/10​.1177​/0170840615580010 Sitkin, S. B., & Roth, N. (1993). Explaining the limited effectiveness of legalistic “remedies” for trust/distrust. Organization Science, 4(3), 67–392. https://doi​.org​/10​.1287​/orsc​.4​.3​.367 Skarlicki, D. P., & Kulik, C. T. (2004). Third-party reactions to employee (mis)treatment: A justice perspective. Research in Organizational Behavior, 26, 183–229. https​:/​/do​​i​.org​​/ 10​.1​​016​/S​​0​191-​​3085(​04)26​005-1​ Skarlicki, D. P., O’Reilly, J., & Kulik, C. T. (2015). The third-party perspective of (in)justice. In M. Ambrose & R. Cropanzano (Eds.), Oxford handbook of justice in work organizations (pp. 235–255). New York: Oxford University Press. Spicer, A., & Okhmatovskiy, I. (2015). Multiple paths to institutional-based trust production and repair: Lessons from the Russian bank deposit market. Organization Studies, 36, 1143–1170. https://doi​.org​/10​.1177​/0170840615585334 Stevens, M., MacDuffie, J., & Helper, S. (2015). Reorienting and recalibrating interorganizational relationships: Strategies for achieving optimal trust. Organization Studies, 36, 1237–1264. https://doi​.org​/10​.1177​/0170840615585337 Sullivan, D., Landau, M. J., Branscombe, N. R., & Rothschild, Z. K. (2012). Competitive victimhood as a response to accusations of ingroup harmdoing. Journal of Personality and Social Psychology, 102, 778–795. https://doi​.org​/10​.1037​/a0026573 Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S.Worchel (Eds.), The social psychology of intergroup relations. Monterey: Brooks/ Cole. Tanis, M., & Postmes,T. (2005). A social identity approach to trust: Interpersonal perception, group membership and trusting behaviour. European Journal of Social Psychology, 35, 413–424. https://doi​.org​/10​.1002​/ejsp​.256

176  Nicole Gillespie et al.

Tomlinson, E. C., Dineen, B. R., & Lewicki, R. J. (2004). The road to reconciliation: Antecedents of victim willingness to reconcile following a broken promise. Journal of Management, 30, 165–187. https://doi​.org​/10​.1016​/j​.jm​.2003​.01​.003 Tomlinson, E. C., & Mayer, R. C. (2009). The role of causal attribution dimensions in trust repair. Academy of Management Review, 34(1), 85–104. https​:/​/do​​i​.org​​/10​.5​​465​/A​​MR​. 20​​09​​.35​​71329​1 Turillo, C. J., Folger, R., Lavelle, J. J., Umphress, E. E., & Gee, J. O. (2002). Is virtue its own reward? Self-sacrificial decisions for the sake of fairness. Organizational Behavior and Human Decision Processes, 89(1), 839–865. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​749-​​5978(​02) 00​032-8​ Turner, J. C. (1991). Social influence. Thomson: Brooks/Cole Publishing Co. Tyler, T. R., & Boeckmann, R. J. (1997). Three strikes and you are out, but why? The psychology of public support for punishing rule breakers. Law and Society Review, 31(2), 237–265. https://doi​.org​/10​.2307​/3053926 Tyler, T. R., & Lind, E. A. (1992). A relational model of authority in groups. In P. Zanna (Ed.), Advances in experimental social psychology (pp. 115–191). Academic Press. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​0​065-​​2601(​08)60​283-X​ Vidmar, N. (2000). Retribution and revenge. In J. Sanders & V. L. Hamilton (Eds.), Handbook of justice research in law (pp. 31–63). New York: Kluwer Academic Publishers. Wade, N. G., Hoyt, W. T., Kidwell, J. E, & Worthington, E. L. (2014). Efficacy of psychotherapeutic interventions to promote forgiveness: A meta-analysis. Journal of Consulting and Clinical Psychology, 82, 154–170. https://doi​.org​/10​.1037​/a0035268 Wenzel, M., Okimoto, T. G., Hornsey, M. J., Lawrence-Wood, E., & Coughlin A.-M. (2017).The mandate of the collective: Apology representativeness determines perceived sincerity and forgiveness in intergroup contexts. Personality and Social Psychology Bulletin, 43, 758–771. https://doi​.org​/10​.1177​/0146167217697093 Wildschut, T., Pinter, B., Vevea, J. L., Insko, C. A., & Schopler, J. (2003). Beyond the group mind: A quantitative review of the interindividual-intergroup discontinuity effect. Psychological Bulletin, 129(5), 698–722. https​:/​/do​​i​.org​​/10​.1​​037​/0​​033​-2​​909​​.1​​29​.5.​​698 Wohl, M. J. A., & Branscombe, N. R. (2008). Remembering historical victimization: Collective guilt for current ingroup transgressions. Journal of Personality and Social Psychology, 94, 988–1006. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.9​​4​.6​.9​​88 Wohl, M. J. A., Hornsey, M. J., & Bennett, S. H. (2012). Why group apologies succeed and fail: Intergroup forgiveness and the role of primary and secondary emotions. Journal of Personality and Social Psychology, 102, 306–322. https://doi​.org​/10​.1037​/a0024838 Wohl, M. J. A., & Hornsey, M. J., & Philpot, C. R. (2011). A critical review of official public apologies: Aims, pitfalls, and a staircase model of effectiveness. Social Issues and Policy Review, 5, 70–100. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1751​​-2409​​.201​1​​.0102​​6.x Wohl, M. J. A., Matheson, K., Branscombe, N. R., & Anisman, H. (2013). Victim and perpetrator groups’ responses to the Canadian government’s apology for the head tax on Chinese immigrants and the moderating influence of collective guilt. Political Psychology, 34, 713–729. https://doi​.org​/10​.1111​/pops​.12017 Worchel, S. (1979). Trust and distrust. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 174–187). Monterey, CA: Brooks/Cole. Yu,Y.,Yang,Y., & Jing, F. (2017). The role of the third party in trust repair process. Journal of Business Research, 78, 233–241. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jbus​​res​.2​​​017​.0​​1​.015​

PART III

Embedding Trust in Organizations

8 NETWORK TRUST Bill McEvily, Akbar (Aks) Zaheer, and Giuseppe Soda

Introduction To what extent can the phenomenon of trust be accurately understood in the context of an isolated interpersonal relationship? Certainly, trust as a social judgment about the willingness to be vulnerable to the decisions and actions of another is shaped by our direct experiences with a focal individual. Yet, trust is also based on what we learn about a person indirectly through our interactions with others who have also had experiences with the same person and on the surrounding conditions that affect the social dynamics of trust. In the context of organizations, which consist of a web of formal and informal interactions for coordinating efforts, exchanging information, and making decisions (McEvily, Perrone, & Zaheer, 2003; McEvily, Soda, & Tortoriello, 2014; Puranam, 2018; Soda & Zaheer, 2012), an isolated interpersonal relationship would appear to be more of an anomaly than the norm. If correct, we see this as quite striking given that the bulk of organizational research on trust, and dominant models of trust (e.g., Mayer, Davis, & Schoorman, 1995; McAllister, 1995; Rousseau, Sitkin, Burt, & Camerer, 1998), focus exclusively on trust between a specific trustor and trustee who interact directly, while overlooking the influence of the broader network of interactions surrounding a trust dyad (De Jong, Kroon & Schilke, 2017). Indeed, we know of only a handful of empirical studies examining networks and trust

DOI: 10.4324/9780429449185-8

180  B. McEvily, A. Zaheer, and G. Soda

(e.g., Burt & Knez, 1995; Buskens, 1998; Chua, Ingram, & Morris, 2008; Ferrin, Dirks, & Shah, 2006; Gulati, 1995a and 1995b; Gulati & Sytch, 2007; Lau & Liden, 2008; Shazi, Gillespie, & Steen, 2015). Similarly, the chapter by Jones and Shah in this volume is one of the few theoretical treatments of interpersonal trust from a network perspective of which we are aware.While these initial contributions have both confirmed the enhanced explanatory power and clarified the underlying conceptual mechanisms of networks in models of relational trust (e.g., McAllister, 1995; Rousseau, Sitkin, Burt, & Camerer, 1998), the organizational literature has yet to consider the extent to which trust extends beyond dyads. We argue that a pervasive form of trust occurs among individuals who are not necessarily directly connected. Specifically, we introduce the concept of network trust, which we maintain is distinct in terms of its locus of operation, antecedents, and outcomes relative to established forms of trust (e.g., relational, presumptive, swift, institutional, generalized). Central to our notion of network trust is the idea that apart from forming as a result of direct interaction, trust also flows through the indirect connections linking individuals to one another and emerges from the inherent design features of the network itself. In this way, network trust is a multilevel phenomenon involving system-level (i.e., network-level) features that condition individual-level actions, which in turn aggregate to produce system-level outcomes (Coleman, 1990). To better illuminate our notion of network trust, we begin with some examples. The case of the tenure letter: take the situation of a letter of recommendation for a tenure candidate, Beth (see Figure 8.1). Lisa is Beth’s senior colleague chairing the tenure review committee. Bob is a letter-writer for Beth. Some of the letters for Beth’s case, including Bob’s, were controversial. Lisa decided to gather some additional information about the significance of Beth’s scholarly impact to help the committee better interpret Bob’s letter, so she reached out to her colleague Don for help, who was a co-author of Bob’s. In turn, Don asked Bob for some insights and Bob duly obliged. Don relayed the information to Lisa who then passed it on to the committee to inform their deliberations. Even though Lisa and Bob never interacted directly and do not know each other, Lisa trusted the information provided by Bob because of her trust in Don and Don’s trust in

Bob

Lisa Don

Legend Arrows indicate direction of trust Solid line arrow = relational trust (direct tie) Dashed line arrow = secondhand trust (indirect tie) FIGURE 8.1 Secondhand

trust with two degrees of separation.

Network Trust 

181

Bob. While it is possible that as a result of the transmission of information, Lisa and Bob initiate a direct relationship, it is not required for trust to function. The case of the alumni association: Avi is an alumnus of a leading MBA program. At a professional development event organized by the school, Avi met Barb. Even though Barb graduated four years earlier than Avi and did not have any acquaintances in common, when Avi described his start-up venture aimed at placing artisanal products from India at upscale retailers, Barb offered to introduce Avi to her classmate Claire, who was a senior manager at a luxury department store. When Barb contacted Claire she immediately agreed to the introduction and met with Avi the following week. Despite Barb and Avi being strangers, by virtue of their common affiliation to the school, Barb was comfortable referring Avi to Claire. The case of social trading: eToro is an open platform for online trading of currencies, stocks, and commodities. Joining the platform is free and requires a nominal deposit of funds to invest. All the traders on the site are visible to everyone, and all traders’ investments, transactions, and portfolios are fully transparent. More importantly, for each trader, their daily, weekly, monthly, and yearly financial performance, as well as their portfolio’s risk and volatility, is accessible to all. In addition, the platform allows traders to communicate with one another, but only through public posts that everyone can view. Traders can initiate trades on their own, but also choose to ‘copy trade’ the actions of another trader. To do so, a trader decides the percentage of their funds that they want to allocate to each ‘copied’ trader.The site then automatically executes all subsequent transactions by the copied trader on the copying trader’s account. Copy-trading activity is public and the top 100 highly copied traders are prominently displayed on the platform. Interestingly, in making copy choices traders place more emphasis on the status and social visibility than the financial performance of other traders. Stepping back from these examples, we make two observations about forms of trust that do not clearly conform to the concept of relational trust. First, we see trust occurring in instances between a trustor and trustee who do not have a direct relationship. In the case of the tenure letter, Lisa’s reliance on Bob’s private information is based on Don serving as a proxy for Lisa’s trust in Bob as opposed to her trust in Bob directly. Likewise, in the case of the alumni association, Claire’s willingness to spend time hearing a pitch for new products is based on Barb serving as a proxy for Claire’s trust in Avi. The case of the alumni association and the case of the social trading platform also illustrate the propensity for individuals to make themselves vulnerable to the actions and decisions of others based on a premise of trust rather than any personal knowledge of, or experience with, those strangers (e.g., Barb’s referral of Avi and traders copying other traders). Second, these two forms of trust occur within the bounds of a network. Thus, in each of the examples trust is not simply a dyadic element, but instead is situated in the larger social space connecting individuals as well as in the shared affiliation to a collective entity (see also Gunia, 2019 for related treatments in negotiations). We maintain that existing conceptualizations do not adequately capture these forms

182  B. McEvily, A. Zaheer, and G. Soda

of trust and we propose that the examples above, as well as other similar instances, fall into a class of trust that we refer to as network trust. In this chapter, we aim to move beyond the extensive focus on relational trust in the organizational literature by broadening the conceptualization of trust to include its inherent generalizability across a network. In doing so, we aim to broaden the scope of organizational scholarship on trust in order to more fully realize the potential of the intuition that social resources extend beyond dyads and to advance the view that it is not exclusively through direct relationships that the benefits of trust accrue and are realized.That is, trust exists and matters at the level of not only direct relationships, but also indirect connections across, and even lack of connection among members of, a network. In the remainder of this chapter, we define network trust and identify two separate forms that it takes: secondhand trust and prototrust.We then ground our definition in core concepts from network theory (reputation, status, and social control) and subsequently proceed to identify the logics (mechanisms, indicators, and contingencies) of the two forms of network trust. Next, we detail the effects of network trust, followed by a discussion of how network trust is distinct from and related to other trust constructs (e.g., relational, presumptive, swift, institutional, generalized). We conclude by exploring how the two forms of network trust can enrich the organizational literature and pave the way for fresh lines of inquiry.

Network Trust Defined We define network trust as generalized positive expectations about the motives, intentions, and behavior between actors who are not directly connected to each other but are part of a bounded social structure (i.e., the set of formal or informal relations among actors). As opposed to particularized forms of trust (e.g., relational) that are directed at a specific target, network trust is less focused on a single actor and at times extends to multiple members of a bounded social structure. At the same time, our conceptualization of network trust does not encompass the entire network as its point of reference for categorizing whether the members of the network trust one another overall (Gausdal, Svare, & Möllering, 2016). Rather, our notion of network trust resides between the dyadic and network levels as a feature of the social structure within which members are embedded. For the purposes of network trust, it is critical that members of the bounded social system generally agree upon and recognize themselves as part of that system. Network research points to two approaches to defining the boundary of a social system. The boundary can be defined from the vantage point of the actors themselves, or from the perspective of researchers imposing a boundary constructed to serve a particular analytical or conceptual objective (Laumann, Marsden, & Prensky, 1989). For our purposes, the critical issue is that the members of the social system widely agree on the boundary, such as when they recognize themselves as

Network Trust 

183

members, identify with each other on the basis of shared characteristics, or accept the categorization applied to themselves as meaningful. In this sense, we hew closer to the actor-defined view of defining network boundaries. We conceptualize network trust as comprised of two forms: second-hand trust and prototrust. Secondhand trust refers to the partial spillover of relational trust to socially proximate, indirectly connected actors (e.g., the case of the tenure letter and the case of alumni association), to the nth degree of separation, albeit with decay. The notion of Simmelian (1950) ties – a strong, reciprocal relationship that is supported by a common third party – is apropos in that trust in a common third party serves as a proxy for the disconnected actors’ trust in each other. Trust in the third party substitutes for relational trust between the disconnected actors as with, for example, referrals. As Granovetter (1985, p. 490) explained, “Better than the statement that someone is known to be reliable is information from a trusted informant that he has dealt with that individual and found him so.” By prototrust we mean the latent potential for confident positive expectations to emerge between two actors who are neither directly nor indirectly connected (e.g., in the case of the alumni association, Barb’s referral of Avi after meeting him for the first time, and in the case of eToro traders, copying other traders who are strangers). Prototrust enables the members of a bounded social system to activate trust. Prototrust is not trust per se, but rather refers to the conditions giving rise to the emergence of confident positive expectations between any two actors in a network, although it may or may not evolve into relational trust. Even if prototrust does not evolve into relational trust, it still allows two actors to make themselves vulnerable to one another (see Table 8.1).

TABLE 8.1 Relational trust and network trust definitions

Form of trust

Definition

Relational trust

Trustor’s positive expectations about the The case of the tenure trustee’s intentions based on information letter: Lisa trusts Don from within their direct relationship (Figure 8.1) Generalized positive expectations about The case of the tenure the motives, intentions, and behaviors letter: Lisa trusts Bob between actors who are not directly (Figure 8.1) connected to each other, but are indirectly The case of the alumni connected in a bounded social structure association: Claire trusts Avi Generalized positive expectations about The case of the alumni the motives, intentions, and behaviors association: Barb between actors who are neither directly nor trusts Avi indirectly connected to each other in a The case of the social bounded social structure trading platform: eToro

Network trust: secondhand

Network trust: prototrust

Example

184  B. McEvily, A. Zaheer, and G. Soda

Network Theory and Trust Whereas trust in the organizational literature is most commonly conceptualized in the context of an isolated individual dyad, network theory considers the relationships among interconnected sets of dyads, with triads being the most basic unit of analysis, and extending to larger and more complex configurations, commonly referred to as social structure. A distinguishing feature of network theory relative to other theories of organization is its focus on discretionary relationships, as opposed to those that are formally mandated or assigned by the organization. More specifically, network theory differs from other theories of organization in that the system of discretionary relationships describes and defines social space as a way of differentiating actors both horizontally, in terms of proximity and the flow of valued resources, and vertically, in terms of status and prestige. By horizontal network differentiation we mean the heterogeneity in locations, or positions, occupied by individual actors that defines their access to valued resources flowing through the network. Thus, networks serve as critical channels. Chief among network resources is information, particularly private information, that is not equally accessible to all. Private information flowing through networks includes, but is not limited to: factual knowledge, gossip, second-hand stories, half-truths, distorted facts, and outright lies (Burt & Knez, 1995). Since networks “penetrate irregularly and in differing degrees” (Granovetter, 1985, p. 491), different people hear about, learn about, understand, and believe different things, even polar opposite things, about the same individual. In this respect, what people ‘know’ about a person, i.e., the reputation of the person, can and does vary from complete ignorance to deep insight and, critically, informs the strength and types of social judgments they form, and therefore the very meaning, degree, and valence of trust (or distrust). From a network perspective, therefore, one can see the value of conceptualizing trust in terms of impressions shaped based on private information acquired through indirect channels. Vertical network differentiation, on the other hand, implies heterogeneity in the respect, or status, ascribed to individual actors. When actors are sorted into social positions that carry unequal rewards, obligations, and expectations, a status hierarchy is established. Status refers to the prestige, esteem, and admiration actors enjoy from others (Anderson, Srivastava, Beer, Spataro, & Chatman, 2006). Status is based on both innate attributes, reflecting underlying variations in actors’ qualities, and on social judgments that confer privileged positions to actors in a way that is largely independent of their innate qualities. Such judgments are particularly salient under conditions of uncertainty (Podolny, 1993). For our purposes, status is a combination of both innate quality and social judgments. As Gould (2002, p. 1146) argues, the reason positions with greater and lesser advantage exist is that judgments about relative quality are socially influenced. Socially influenced judgments amplify underlying differences, so that actors who objectively rank above

Network Trust 

185

the mean on some abstract quality dimension are over-valued while those ranking below the mean are undervalued – relative to the baseline scenario, in which social influence does not operate. Amplification occurs because observable interactions expressing judgments of quality are also cues to other actors seeking guidance for their own judgments. From a network perspective, status is related to trust in two ways. First, highstatus actors are trusted when their innate qualities or the social judgments about those actors are reflective of their ability, benevolence, and integrity. Second, those judgments are further reinforced as members of a network model their own judgments on those of other network members. Thus, status serves as a proxy for trust when social judgments about an actor’s intentions and motives ripple through a network. Network theory differs from other theories of organization not only with respect to how it differentiates actors horizontally and vertically in a bounded system of discretionary relationships but also in terms of how it defines and describes the governance of such social systems. In networks, governance (i.e., the framework of agreed-upon rules of organization) is emergent, collective, and based on social control as opposed to being mandated and based on formal authority. For instance, actors self-select into joining and opting out of networks, and by the same token, members are accepted into, and can be expelled by the members and or organizers of, a network. Likewise, members of a network often internalize the norms, expectations, and codes of conduct to the extent they share a social identity (Mehra, Kilduff, & Brass, 1998) with other members. From a network perspective, there are no legally binding contracts detailing performance duties and obligations, nor is there hierarchical fiat that serves as the ultimate arbiter of divergent preferences and priorities. Instead, order in the context of networks is a matter of socially defined, constructed, and maintained understandings. Taken together, network theory offers a distinctive lens through which trust can be understood. Most important is the idea that trust is able to operate in the absence of a direct relationship between a trustor and trustee by virtue of the bounded system of discretionary relationships that differentiates actors both horizontally (in terms of reputation) and vertically (in terms of status), as well as the framework of governance (in terms of social control). Using the network mechanisms of reputation, status, and social control, we now explain the logic of secondhand trust and prototrust.

Logics of Network Trust Secondhand Trust As noted previously, we define secondhand trust as trust between two actors who are not directly connected but are socially proximate to each other. Secondhand

186  B. McEvily, A. Zaheer, and G. Soda

trust is based on (one or more) intermediate third parties acting as proxies for trust between two disconnected actors. Third parties who broker trust in this way occupy the role of a trust “advisor” (Coleman, 1990; McEvily et al., 2003). More specifically, the two disconnected actors both have a relationship of mutual trust with the advisor. Returning to the case of the tenure letter, secondhand trust exists between Lisa and Bob, who are not directly connected to each other. Lisa trusts the information provided by Bob because she has a relational trust tie with Don and, in turn, Don has one with Bob. Thus, Don is not only a direct connection to Lisa, but also an indirect channel to Bob through which private information flows. The private information includes both the veracity, or reputation, of Bob and the details about Beth that Bob divulges. Critically, it is relationships of mutual, as opposed to unidirectional, relational trust with the advisor that undergird secondhand trust. Clearly, Lisa is vulnerable to misinformation from Don, as is Don from Bob.Yet, Bob is also vulnerable to Don mishandling sensitive information, as is Don to Lisa.Thus, for secondhand trust to function the advisor needs to be trusted by, and trust, both the trustor and trustee. Drawing on and extending the network bases of trust (McEvily et al., 2003), we now articulate the mechanisms, indicators, and contingencies of secondhand trust (see Table 8.2). As we explain in detail below, secondhand trust is based on the mechanism of transitivity. The primary network indicator for secondhand trust is the open triad. Key contingencies of secondhand trust include tie strength, social distance, and network closure.

Mechanism Relational trust gives rise to the potential for secondhand trust to emerge through the network process of transitivity. Formally, transitivity refers to a system of relationships among all three actors in a triad (Simmel, 1950; Granovetter, 1973; Krackhardt, 1999). When a focal actor (Don), who is strongly connected to two other actors (Lisa and Bob), facilitates a connection between those two actors,

TABLE 8.2 Logics of network trust

Form of network trust

Mechanisms

Indicators

Contingencies

Secondhand trust

Transitivity

Open triad

Tie strength Social distance Network closure

Prototrust

Social prospecting

(Dis)assortativity - Interest-based - Status-based

Governance veracity Network closure Identity authenticity

Network Trust 

187

transitivity occurs (Aven, 2015). In the context of secondhand trust, transitivity occurs when the relational trust between Don and Lisa, and between Don and Bob, is generative of a secondhand trust tie between Lisa and Bob. Note that the secondhand trust tie between Lisa and Bob is of a different kind than those between Lisa and Don and between Don and Bob. Rather than a relationship of direct mutual trust with each other, Lisa and Bob have an indirect, secondhand tie to one another through Don. The secondhand tie has the latent potential to evolve into a direct relationship between Lisa and Bob,1 although that is not necessary for secondhand trust to occur.

Indicators Secondhand trust is most directly observable in a system of triadic relationships; specifically, an ‘open’ triad (Granovetter, 1973; Burt, 1992) in which two of the actors are not directly connected to each other, but are connected to the same advisor with reciprocal trust ties.2 For instance, in Figure 8.1, the Lisa–Don–Bob triad is open in the sense that Lisa and Bob are only indirectly connected through Don. The structural configuration of an open triad by itself is necessary, but not sufficient to capture secondhand trust. In addition, the conditions giving rise to the need for trust – i.e., risk and interdependence (Rousseau et al., 1998) – are also required. Risk is inherent in the structural configuration. Interdependence, however, is likely to vary across open triads and needs to be activated by one or both of the disconnected parties. Secondhand trust may also be observable in open systems of relationships beyond triads, such as quads and larger.

Contingencies The incidence and intensity of secondhand trust are amplified (or diminished) by features of the first-order ties (e.g., between Lisa and Don, and Don and Bob) and the configuration of the network surrounding the secondhand trust triad (i.e., trustor, trustee, and advisor). Not all first-order ties and network configurations are equally potent in enabling secondhand trust. Tie strength. First-order tie strength – comprised of the frequency and duration of interaction, expressiveness, and reciprocation (Casciaro & Lobo, 2008; Granovetter, 1973; Krackhardt, 1990) – will act as a catalyst (Tortoriello, McEvily,

1 We note that the formation of such a direct relational trust tie is consistent with the core prediction of structural balance theory, whereby actors are motivated to eliminate strain or tension resulting from a triadic system of relationships of inconsistent valence (Cartwright & Harary, 1956; Heider, 1946, 1958; Hummon & Doreian, 2003). 2  For a triad to serve as an indicator of secondhand trust, both the secondhand trustor (Lisa in Figure 8.1) and the secondhand trustee (Bob) need to have positive and reciprocal relations of trust with the advisor (Don).

188  B. McEvily, A. Zaheer, and G. Soda

& Krackhardt, 2015) for secondhand trust.The stronger the trust in the first-order ties, the greater the confidence in and willingness to rely on the judgment of the trust advisor (Don). When both first-order ties are strong, the potential for secondhand trust is the greatest. However, if one first-order tie is weak, the stronger tie may compensate up to a point, but only to a limited degree and secondhand trust is less likely.Thus, secondhand trust is not simply a multiplicative function of first-order tie strength. Social distance. The logic of secondhand trust extends beyond two degrees of separation, e.g., beyond a friend of a friend (Watts & Strogatz, 1998). We believe, however, that secondhand trust will decay rapidly with increasing social distance in terms of the number of intermediaries on the shortest path between a potential trustor and trustee. As the number of intermediaries increases, the trustor and trustee increasingly rely on actors to whom one or both are not directly connected. For instance, as shown in Figure 8.2, if only Lisa is directly connected to Don and Bob is only directly connected to Deb, who in turn is directly connected to Don, Lisa and Bob are now three degrees of separation from each other as opposed to the two degrees separating Lisa and Bob in Figure 8.1. As a result, Don is able to vouch for Lisa and Deb, but not Bob, while Deb is able to vouch for Bob and Don, but not Lisa. Thus, neither trust advisor (Don and Deb) is able to vouch for both the secondhand trustor (Lisa) and secondhand trustee (Bob). Even so, both the trustor and trustee have direct relationships with one of the two trust brokers, which is why there continues to be the potential for secondhand trust. Further extending secondhand trust to four degrees of separation, involving three trust advisors (e.g., Don to Dan to Deb in Figure 8.3), one of whom (Dan) neither the secondhand trustor nor secondhand trustee is directly connected to, further diminishes the prospects for secondhand trust due to the limited veracity of information accessed and the heightened risks of the trustor and trustee relying on the referral of a stranger.That is, Dan is able to vouch for neither Lisa nor Bob since he does not have a direct relationship with either.

Bob

Lisa Don

Deb

Legend Arrows indicate direction of trust Solid line arrow = relational trust (direct tie) Dashed line arrow = secondhand trust (indirect tie) FIGURE 8.2 Secondhand

trust with three degrees of separation.

Network Trust 

189

Bob

Lisa Don

Dan

Deb

Legend Arrows indicate direction of trust Solid line arrow = relational trust (direct tie) Dashed line arrow = secondhand trust (indirect tie) FIGURE 8.3 Secondhand

trust with four degrees of separation.

Bob

Lisa Don

Tim

Legend Arrows indicate direction of trust Solid line arrow = relational trust (direct tie) Dashed line arrow = secondhand trust (indirect tie) FIGURE 8.4 Secondhand

trust with network closure.

Network closure.The occurrence of secondhand trust also depends on the extent to which there is network closure around a trust triad. Network closure exists when the members of a secondhand trust triad have mutual connections to common third parties outside the triad (Coleman, 1988; Burt, 2005). For instance, if Lisa, Don, and Bob are all connected to Tim as shown in Figure 8.4, Tim is a common third party to all three individuals and there is complete closure around the secondhand trust triad. In networks characterized by closure, information circulates rapidly and is relatively easy to calibrate and confirm. As a result, individuals are more likely to have common knowledge and shared understandings in closed relative to open networks (Reagans & McEvily, 2003). More critically, closed networks permit a more robust form of social control than open networks by sanctioning anti-social behavior and rewarding pro-social behavior (Coleman, 1990). Moreover, in closed networks, news of actors’ pro- and anti-social behavior (e.g., sharing versus withholding requested information, clarifying versus distorting sensitive details, etc.) also circulates rapidly and as a result, magnifies the reputational consequences of one’s behavior (Burt & Knez, 1995). Whereas in an isolated dyad, reputational consequences are limited to the counterparty in

190  B. McEvily, A. Zaheer, and G. Soda

the relationship, in a closed network, one’s reputation in the eyes of mutual third parties is also altered. Given this, actors tend more toward pro- rather than antisocial behavior in closed networks. Analogously, norms are easier to create and enforce in closed, relative to open, networks since actors can more readily coordinate expectations and sanction norm violation.Taken together, network closure around a secondhand trust triad will heighten the potential for secondhand trust between a secondhand trustor and trustee. Additionally, the potential for secondhand trust to emerge is likely to vary with the extent of network closure around a secondhand trust triad. Specifically, the degree of network closure around a secondhand trust triad can be partial, rather than complete. For instance, there would be partial network closure if Tim is connected to Don and Bob, but not Lisa (Figure 8.4). Even so, the force of social control in the form of reputation and norms would still exist and, therefore, heighten the prospects for secondhand trust, albeit less intensely than in the case of complete closure. The effect of partial closure is particularly interesting given that secondhand trust is amplified even though the third party (Tim) is not directly connected to the trustor (Lisa). Likewise, in the situation where partial network closure exists around the trustor rather than the trustee, if for instance Tim is connected to Lisa and Don, but not Bob, the potential for secondhand trust to emerge is heightened. Lastly, as the number of mutual third parties to whom the members of a secondhand trust triad are connected increases, the prospects for secondhand trust are further amplified. To summarize, secondhand trust is based on the transitivity of trust flowing through third-party intermediaries who connect two actors indirectly.A key proxy for trust transitivity is the open triad. The propensity for secondhand trust also increases with the strength of ties connecting the trustor and the trustee to the intermediary, increases with network closure, and decreases with social distance.

Prototrust Like secondhand trust, we see prototrust as a property of social structure (i.e., beyond the dyad). However, unlike secondhand trust, we see prototrust as a social-structural property that may systematically vary across dyads within the same network. Analogous to secondhand trust, prototrust is a form of trust that occurs among actors in a network who are not directly connected to each other. Prototrust differs from secondhand trust, however, in that an indirect connection (i.e., through an advisor) is not a defining feature of this form of trust. Rather, prototrust refers to the latent potential for confident positive expectations to emerge between two actors who are neither directly nor indirectly connected. Prototrust also differs from secondhand trust in that prototrust primarily occurs in affiliation networks, which involve joint participation or membership in collectivities, such as the case of the alumni association above, as well as other examples like social groups, clubs, and professional associations.Within affiliation networks, subgroups – such as activities, events, committees, organizations, and the like – exist, where

Network Trust 

191

members interact more intensively (Faust, 1997). Additionally, members may have multiple, overlapping subgroup memberships (e.g., two members participating in the same social activities, and events, and committees), in which interaction intensity increases even more. Thus, affiliation networks are often nested structures of primary membership in the bounded social system and secondary memberships on committees, events, and other subgroups. Prototrust is based on (1) taken-for-granted, background assumptions about what constitutes trustworthy behavior in the context of an affiliation network and (2) the capacity of the network to curate and match members with compatible interests. Basic assumptions about the expected behavior of other members of the affiliation network in pursuit of shared goals are the genesis of prototrust. In the absence of such assumptions, individuals may still affiliate within a network, but the potential for prototrust is limited due to uncertainty about expected behaviors. For instance, in the case of the alumni association, there is a strong belief in giving back and helping other alumni whenever possible. New members are tacitly socialized by both the alumni association and existing members. Through events and other activities organized by the association, new members have a chance to see other alumni engaging in expected behaviors. Similarly, becoming a member of an organized crime syndicate entails clear understanding and acceptance of the behavioral rules of involvement in crime, solidarity, and omertà, i.e., code of silence (Gambetta, 1993). The clearer the rules and the more the rules circumscribe behaviors, even if they are informal or tacit, the less the uncertainty and the greater the potential for prototrust. While necessary, background assumptions by themselves are not sufficient to initiate prototrust. In addition, members of affiliation networks are more likely to realize prototrust to the extent that the network facilitates the discovery of and connection with other members with whom their goals are aligned. The curating and matching of members can occur in a number of different ways but is often enabled by a network architect (McEvily & Zaheer, 2004).The primary activities performed by a network architect include the initial design of the network and recruitment of members, as well as the ongoing evolution of the network. In particular, the rules of affiliation and rules of engagement (e.g., participation, contributions, and value creation), determine the extent to which prototrust may arise. Rules of affiliation encompass both the principles, conventions, and expectations that govern attracting members to join the network and govern the inclusion of members into the network. Rules of engagement circumscribe the manner in which members may, and may not, interact with one other and the mechanisms of social control (e.g., sanctioning, ostracism, expulsion, etc.) that the members may exercise. Drawing on network theory, we now articulate the mechanisms, indicators, and contingencies of prototrust (see Table 8.2). As we explain in detail below, prototrust is based on the mechanism of social prospecting.The primary indicator for prototrust is assortativity. Key contingencies of prototrust include governance veracity, network closure, and identity authenticity.

192  B. McEvily, A. Zaheer, and G. Soda

Mechanism The inclination to connect with a stranger in an affiliation network is based on heuristic processes (Lewicki & Brinsfield, 2011; McEvily, 2011; Uzzi, 1997) in which an actor’s background assumptions about, and the perceived quality of, the prospective match are fitted together in order draw an inference about the value of connecting with the prospective match. We refer to this class of social judgments as “social prospecting.” Returning to the example of the alumni association, prototrust exists between alumni by virtue of the school attracting and selecting students who share common interests (career advancement, professional development, helping others, etc.) and shared social experiences (e.g., work, education, extra-curricular, etc.). Members of the alumni association are open to connecting with each other to the extent that the association has attracted like-minded individuals who accept, internalize, and reinforce a shared set of norms, expectations, and codes of conduct for appropriate behavior. These normative expectations are driven in part by shared social experience and in part by the alumni association’s network governance.

Indicators Prototrust at the network level is indicated by assortativity, which is defined as the tendency for actors in a network to preferentially connect with similar others (Newman, 2002). For instance, in the case of the alumni association, two individuals may connect based on their common interests in promoting gender equity and diversity in their respective organizations. Likewise, in the case eToro, two traders may connect on the basis of their shared interest in socially responsible investing. The specific form of similarity upon which assortativity is based varies depending on the nature of the context (social, professional, organizational). At the same time, matching may occur preferentially such that dissimilar actors connect, which is known as disassortativity (Uzzi & Spiro, 2005; Watts, 2004). For instance, in the case of the alumni association, two individuals with different years of work experience may connect to form a mentorship relationship. In the case of prototrust, both assortativity and disassortativity are operative. Assortativity underlying prototrust is based on factors such as common interests, while disassortativity could be based on factors such as status asymmetry. The potential for (dis)assortativity in a network is a function of the extent to which the network has tightly defined and enforced rules of affiliation and rules of engagement. Put another way, to the extent that the network is better able to curate and match members with compatible interests, the greater the assortative matching success of social prospecting among members. Likewise, to the extent that the network is better able to reveal underlying differences in quality that are relevant to the formation of a status hierarchy (Ertug & Castellucci, 2013; Podolny, 1993), the greater the disassortative matching success of social prospecting among members.

Network Trust 

193

Prototrust at the dyad level is indicated by the capacity of the network to differentiate members both horizontally (i.e., similar interests) and vertically (i.e., status). Differentiation, both horizontally and vertically, in a network refers to the distribution of attributes among members such that differentiation is lower when attributes are highly concentrated, and differentiation is higher when attributes are highly dispersed across members. Horizontal network differentiation clarifies the strength and overlap of interests shared by some, but not all, members. One common instantiation of horizontal differentiation in affiliation networks is via subgroups that enable, concentrate, and accelerate the flow of valued resources among members who share similar interests. Subgroups form organically by member initiation and may be enabled by structures put in place by the network architect. For instance, an organic member-initiated group indicative of horizontal differentiation might include industry-based, topic-based, or regional activities initiated by alumni. Similarly, horizontal differentiation in the alumni association may form cohort-based or interest-based (e.g., finance, consulting, marketing) groups at events such as reunions to help alumni meet and interact with others who share some commonality. Critically, both of these are examples of informal groups in the sense that members freely choose to join (or not) the group regardless of whether the group is initiated by the members or by the network architect. Another way that horizontal differentiation occurs in affiliation networks is by referral and recommendation algorithms that are intentionally designed by the network architect. Such algorithmic processes are pervasive in online networking platforms (e.g., LinkedIn, ResearchGate, Match​.c​om) of many forms. As these examples suggest, formal structures, systems, and rules create the context within which individuals choose whether or not to affiliate with other members. While the structures, systems, and rules are formally designed and maintained by the network architect, the choice to affiliate is informal in the sense that rather than being assigned to interact, individuals choose to do so. Vertical network differentiation in terms of status clarifies the perceived differences in quality among members and in social judgments about members independent of their innate quality. Status in affiliation networks often takes the form of rankings, recognition, and reviews. For instance, vertical differentiation is manifested in eToro (the case of social trading discussed earlier) through compilation and display of information on highly copied traders for all to see, which proxy for not only innate quality but also the aggregate social judgments of other members. Such rankings are intentionally devised and highlighted by the network architect in an effort to reduce uncertainty and promote the potential for relationship initiation.

Contingencies At both the network and dyadic levels, prototrust is amplified (or dampened) by the perceived reliability or veracity of network governance, the visibility of

194  B. McEvily, A. Zaheer, and G. Soda

the network configuration, and the extent to which members can discriminate between each other on the basis of authentic identities. Governance veracity. By veracity, we mean the extent to which members accept that the network applies the rules of affiliation and rules of engagement consistently and rigorously, such that members who share common interests and background assumptions with the existing members are selected into the network, while prospective members who do not share interests and background assumptions are screened out. Governance veracity is also relevant for allowing members to make better matches with other members who share the same interests through the creation of subgroups and algorithms. To be clear, we are not claiming that there is a change in the formality of the network structure due to the rules. The rules are formal in the sense that they are originated by the network architect, but the rules are more accurately understood as a framework for interaction, within which members decide for themselves whether or not to informally interact with certain other members. Network closure. Prototrust is also enhanced to the extent that the architect of the network provides information enabling members to view the network and differentiate each other in terms of status. In the context of affiliation networks, the social structure in which members are embedded is a further signal that can enhance prototrust. For instance, consider a new member of eToro. The actual structure of copy trading ties that she observes, which the architect makes transparent to all the members, influences prototrust in other members. The level of network closure she observers around others, particularly other high-status members, amplifies the potential for trust. Identity authenticity. Finally, the greater the extent to which members perceive others as authentic in their projected personas, the more the prototrust. Networks “confer social identity through the segmentation of social space into clusters and positions populated by actors who share common social characteristics and who are, therefore, social referents for each other” (Ibarra, Kilduff, & Tsai, 2005, p. 362). The degree of congruence, or lack thereof, between a member’s social identity and self-projected identity determines the authenticity of identity and, respectively, amplifies or attenuates the potential for prototrust. For instance, in eToro members are allowed to choose nicknames and avatars to represent themselves or to use their actual names and photographs. The latter are more likely to receive copy-trading ties since they are seen as more authentic.

Effects of Network Trust Taken together, the logics of secondhand trust and prototrust provide a wide variety of promising avenues for further research. A key priority for advancing the research agenda on network trust is exploring the extent to which, and ways in which, secondhand trust and prototrust matter for valued outcomes in and between organizations.

Network Trust 

195

Risk-Taking Outside of Relationships While in traditional models of relational trust, risks are concentrated at the level of the dyad, in network trust, risks are distributed and shared across larger systems of relationships. In this sense, network trust is ‘in the air’ and becomes a resource that is shared beyond just the two members of a dyad to other members of the network in close proximity, and in the case of prototrust, extending throughout the network to members who are disconnected from one another. As a consequence, the assessment of the risks associated with placing trust is based not on the properties of the dyad, but rather on the features of network structure and governance. A key implication of theorizing trust from a network perspective is that the concepts of secondhand trust and prototrust advance our understanding of the micro–macro links as posited by Coleman (1990). In his ‘bathtub’ model Coleman displays the links between the micro and macro levels of social systems (see Figure 8.5). Arrow A represents the effect of systemlevel features, in our case structural features of the network such as open relational trust triads and (dis)assortativity, on a system-level outcome, which is network trust. Arrow B shows how the system-level conditions the individual-level by means of mechanisms such as transitivity of relational trust for secondhand trust and social prospecting for prototrust.These mechanisms, in turn, influence generalized positive expectations about the motives, intentions, and behaviors between individuals at the micro level. Arrow C conveys the individual-level actions that occur as shaped by the system or macro level, which in our case constitutes risk

Antecedent Macro Factors - Social Structure

A

B

Constraints on Actors - Transitive Trust (secondhand trust) - Social Prospecting (prototrust)

FIGURE 8.5 Macro–micro

Consequent Macro Factors - Trust Contagion (secondhand trust) - Collective Engagement (prototrust)

D

C

Rational Actions by Actors - Risk-taking outside direct relationships Legend Bold text = Coleman’s original model Italicized text = network trust model

links of network trust.

Adapted from Source: Coleman, J. S. (1990). Foundations of Social Theory. Cambridge, MA: Belknap Press of Harvard University Press.

196  B. McEvily, A. Zaheer, and G. Soda

taking outside of direct relationships. Lastly, arrow D indicates the extent to which the individual-level actions aggregate to produce macro-level outcomes. In our case, while the aggregation generated by secondhand trust is the contagion of trust among network members, the aggregation created by prototrust is the engagement of individuals with the collective community in the form of citizenship, participation, cohesion, and solidarity. Taken together, our theory also extends to the percolation of trust between macro and micro levels of social systems in a way that identifies and details the mechanisms, actions, and links that underlie the notion of trust being ‘in the air.’

Substitute for Relational Trust From a collective perspective, then, networks that are able to enhance secondhand trust and prototrust are capable of yielding trust-like advantages on a large, distributed scale. For example, the costs associated with developing relational trust at the micro level are not only time-consuming and high but also concentrated in socially proximate relationships. In contrast, the production of trust in networks occurs at a relatively larger scale. Imagine a team of 20 people who, in order to engage in joint activities, have to develop relational trust with every other member of the team. The investment in terms of the number of relational trust ties to be activated is n(n–1)/2, or 190, assuming trust is reciprocal. By comparison, suppose the same group of 20 people are at a maximum distance from each other of two ties. By virtue of secondhand trust, the number of relational trust ties needed is reduced to as little as 19 (with a hub and spoke structure). The efficiency gains are achieved by substituting direct relational trust ties (171 in the example above) with secondhand trust ties (19 ties), which involve considerably lower investment than relational trust. Clearly, the efficiency gains are considerable; an order of magnitude lower for secondhand trust. At the same time, a question arises as to whether such efficiency of secondhand trust also translates into comparable effectiveness relative to relational trust. Further, for a team of 20 people that embodies the conditions for prototrust, the emergence of trust is potentially automatic, or swift, by virtue of the co-affiliation network ties. While secondhand trust is a substitute, prototrust is an enabler, precursor, or “lubricant” (Arrow, 1974) for relational trust. Both secondhand trust and prototrust establish the notion that risk taking in network settings is not solely based on direct relational ties.

Complement to Relational Trust In addition to acting as a substitute for relational trust, network trust may also serve as a complement. Returning to the tenure letter case (Figure 8.1), the result of Lisa’s secondhand trust in Bob may spill over to Lisa’s direct relational trust with Don and Don’s direct relational trust with Bob. For instance, if Lisa’s trust in Bob’s private information is well-placed, Lisa’s relational trust in Don is further enhanced. In this way, secondhand trust begets relational trust. At the same time,

Network Trust 

197

when Lisa’s secondhand trust is misplaced, her relational trust in Don is compromised. Here, misplaced secondhand trust corrodes relational trust, and the same sorts of spillovers of secondhand trust onto relational trust apply to the relational trust between Bob and Don. Note that such spillover effects need not be symmetric. For instance, if Lisa lacked discretion in how she handled the private information from Bob via Don, both Don’s trust in Lisa and Bob’s trust in Don would be compromised.

Positioning Network Trust As we argue in the preceding pages, network trust is not relational trust, which requires firsthand knowledge or experience. At the same time, network trust is related to, although distinct from, other trust constructs including presumptive, swift, institutional, and generalized trust, which we discuss below. Network trust is akin to presumptive trust (Kramer, 2010) in the sense that it involves generalized positive expectations in the context of a collective. Importantly, however, network trust differs from presumptive trust in terms of the unit of analysis. For presumptive trust, the unit of analysis is the average, or stereotypical, member of the collective as perceived by the trustor, which then provides the content for presumptive trust in “the collective as a whole” (Kramer & Lewicki, 2010, p. 259). In contrast, the unit of analysis for network trust is the social structural position of members in the network. Thus, while presumptive trust implies a set of diffuse expectations in an entire collectivity, network trust is enabled and shaped by features of networks and is directed toward specific members of the network. Although presumptive trust may extend to “individuals who are considered ingroup members,” it is based on the “generic features of all the members of that collective” (p. 259). Unlike such a diffuse conceptualization, network trust differentiates among the members of a collective and is best understood as being an embedded form of trust that percolates among members to differing degrees. That is, our concept of network trust identifies the process and flow of trust based on the features of a network and the relative positioning of members in the network. Network trust is also similar to, but different from, swift trust. Swift trust refers to the trust that forms in the context of temporary systems, characterized by high interdependence, high risk, and complex tasks among individuals who typically have never worked together in the past and have no expectation of working together again in the future. As Meyerson, Weick, and Kramer (1996) explain, Trust (or distrust) in temporary systems can develop swiftly because the expectations that are invoked most quickly tend to be general, task-based, plausible, easy to confirm, and stable, all of which implies the care of valuable things can be entrusted to those who seem to fit these institutiondriven categories. (1996, p. 175)

198  B. McEvily, A. Zaheer, and G. Soda

Whereas swift trust is grounded in institution-driven categories that allow roles to be invoked instantly, network trust is based on social structures that facilitate the flow of trust and provide the conditions for relational trust to potentially emerge. Further, network trust is not institutional trust – i.e., it is not “the safety one feels about a situation because of guarantees, safety nets, or structures” (McKnight et al., 1998). With respect to secondhand trust, it is not the effect of safeguards in shaping context that engenders trust, but rather the effect of actors and the configuration of actors that account for trust. More precisely, secondhand trust is based on the informal norms, expectations, values, and reputations that are widely held among a bounded set of actors. In terms of prototrust, it is the signals of assortativity that differentiate it from institutional trust. At the institutional level, affiliation is highly diffuse and extends to broad categories of membership (e.g., the nation-state, religion, etc.). For prototrust, network affiliation is crucial because it is one of the key bases upon which assortativity occurs. Moreover, relative to institutional trust, the signals of assortativity underlying prototrust are clearer and more informative for the creation of ties and the potential to realize relational trust. Thus, while institutional regulations, guarantees, and laws facilitate, for example, banking transactions by mitigating downside risks, those safeguards are not informative for differentiating among prospective transactors (e.g., banks). In contrast, prototrust is precisely the latent potential for confident positive expectations to emerge due to the assortativity of a network. Lastly, network trust is not generalized trust, which is defined as a belief in the benevolence in human nature in general.Yamagishi and Yamagishi (1994, p. 139) call this type of trust “general trust,” as it reflects “a belief in the benevolence of human nature in general.” Generalized trust is most frequently assessed at the societal level, using survey items such as “Generally speaking would you say that most people can be trusted or that you can’t be too careful in dealing with people?” Network trust applies within the boundary of the network and is based on expectations about members of the network.

Discussion and Conclusion Scholarly understanding of trust is concentrated at two extremes. On the one hand, trust is considered inherently personal in terms of the relational features of direct interactions. On the other hand, trust is treated as impersonal in terms of the institutional properties safeguarding exchange. The gulf between these poles remains conceptually bereft. Into this void, we propose a class of trust that is situated in the enduring pattern of social connections among actors – network trust. Given the widespread prevalence of social networks in and between organizations, it behooves us to understand the distinctive forms of trust to which networks give rise. Network trust is especially relevant in the organizational context where getting things done routinely requires relying on others with whom there is no direct

Network Trust 

199

connection (Krackhardt & Hansen, 1993) and where formal roles and structures do not explicitly specify how all decisions and actions are to be organized and coordinated (McEvily et al., 2014). As a result, informal arrangements for achieving organizational goals and outcomes emerge and are put in place based on socially devised understandings. At the same time, it is important to recognize that the informal side of organizations need not necessarily enable trust and at times may even undermine it or engender distrust due to, for instance, inter-departmental skullduggery, organizational politics, opportunistic behavior, and the like. Thus, organizations are a prime arena for examining network trust and distrust given the inherent interdependencies that exist and the discretion that individuals have, to varying degrees, to support the activities and role-responsibilities of their co-workers. We see a number of exciting implications for organizational scholars from examining trust through the lens of the network forms that we have conceptualized. The first-order implication of embracing network trust is to revisit the basic premise of the genesis and realization of trust. Thus far, scholarly understanding of trust has been heavily based on the psychological view of trust as personal and the sociological view of trust as impersonal. We maintain that there is also a distinct network view of trust that is multi-level and recognizes both structure (in terms of patterns of connections) and behavior (in terms of the actions taken by individuals under constraints).Viewed this way, network trust bridges a multi-level space between the micro and the macro, between the personal and impersonal, between the psychological and sociological. Network trust is a phenomenon in and of itself. Thus, while network trust could be considered in relation to other forms of trust (e.g., as a substitute or complement), the prime implication is to treat network trust as a novel form and consider the unique understandings that it permits. Indeed, we see a wide range of promising avenues for network trust to enrich scholarly understanding. Three areas in particular are ripe for discovery.

From Stocks to Flows Most organizational research on trust is principally concerned with explaining the level of trust within a relationship. In addition to informing our understanding of such ‘stocks,’ network trust introduces the potential to consider how trust ebbs and flows through a network. By virtue of the structural features of a network, trust and distrust have the potential to spread, as do trends, fads, gossip, and good ideas. Importantly, the ‘contagion’ of trust does not just happen on its own, but rather is agentic in the sense that it is intentionally passed along and accepted when individuals pursue interests that require them to rely on strangers for valued resources. At the same, network trust can be latent to the extent that it resides in network structures that can lie dormant for an extended period until triggered by a critical event.When network trust is activated at a large scale, it has the potential to fuel social movements for collective action. For instance, when a manager is

200  B. McEvily, A. Zaheer, and G. Soda

promoting a new initiative, program, or product that requires the buy-in and support of colleagues from across the organization who are not directly connected to the manager, those colleagues typically draw on the reservoir of network trust to create a well-spring of support, apathy, or resistance toward the initiative.

From Emergent to Designed Part of the allure of trust is its potential to enable actions that would be exceedingly costly or difficult to achieve in its absence. Analogous to conventional forms of capital in economic models (e.g., human, financial, physical), trust has been characterized as a type of social capital with similar value-generating properties (Coleman, 1988). As a factor of production, scholars have also considered the modes of production by which trust is constructed and reconstituted (Zucker, 1986a). Like the broader organizational literature, trust production modes are conceptualized in terms of personal (i.e., character-based and exchange processes–based) and impersonal (institutional-based) mechanisms. Alongside these modes, we maintain that informal networks of connections also create trust, albeit via a distinct production function. In some cases, the production of trust is emergent and automatic as a consequence of common shared experiences. In other instances, the production of trust in networks is more intentional and by design (Hurley, Gillespie, Ferrin, & Dietz, 2013). And in still other situations, both the emergent and intentional combine. Consider again the example of alumni networks. By virtue of graduating from the same educational institution, two alumni are members of a common affiliation network and to the extent that they are indirectly connected by other alumni, they may experience secondhand trust. At the same time, even if they are not indirectly connected, the fact that they belong to a community with shared values and identity creates the potential for prototrust. Further, a number of educational institutions organize reunions, events, and other activities with the express intent of creating opportunities for alumni to meet, reconnect, and interact. In this way, the alumni network strengthens the potential for trust by reinforcing the sense of shared identity and social norms and enhances the potential for network closure. Here, governance veracity is less salient given the alumni’s prior socialization into the network by virtue of being selected into and matriculating from the educational institution. The production of trust by networks is of course not limited to the alumni of educational institutions, but extends to shared prior organizational affiliations (e.g., McKinsey, GE, State Department). It is important to note that for each of these examples, the presence of an affiliation network is the minimum required necessary condition for prototrust. In addition, the production of prototrust for a given type of affiliation network (e.g., MBA alumni networks) varies depending on the intrinsic prestige of the institution as well as the design of the affiliation network in terms of creating opportunities for effective social prospecting through assortativity (i.e., status-based and interest-based matches).

Network Trust 

201

From Dyads to Networks The past three decades of organizational scholarship on trust has laid a critical foundation for understanding the nature of trust in and between organizations – how the willingness to be vulnerable has been investigated as a relational property between a pair of directly connected actors. The bulk of trust theory has been predicated on the dyadic level. Our understanding of the antecedents, formation, duration, dissolution, repair, concomitants, (a)symmetry, intensity, and outcomes, among others facets, have as their locus the dyad. How these dyadic elements link to macrolevel organizational dynamics remains a critical but relatively less studied aspect of scholarship. We argue that network forms of trust provide a bridge to discovering the contextual underpinnings of trust.The network perspective presents the opportunity to consider the ways in which system-level features influence trust beyond the micro-dyadic level to also encompass more network-level elements such as the governance and design of social systems to generate and deploy trust. Taken together, the network forms of trust we have proposed lay the foundation for moving from stocks to flows, from emergent to designed, and from dyadic to network. In so doing, we aim to promote a richer, deeper, and enhanced understanding of the nature of trust in organizational settings.

References Anderson, C., Srivastava, S., Beer, J., Spataro, S. E., & Chatman, J. A. (2006). Knowing your place: Self-perceptions of status in social groups. Journal of Personality and Social Psychology, 91(6), 1094–1110. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.9​​1​.6​.1​​094 Arrow, K. (1974). The Limits of Organisation. New York: W. W. Norton. Aven, B. L. (2015).The paradox of corrupt networks: An analysis of organizational crime at Enron. Organization Science, 26(4), 980–996. https://doi​.org​/10​.1287​/orsc​.2015​.0983 Burt, R. S. (1992). Structural holes:The social structure of competition. Cambridge, MA: Harvard University Press. Burt, R. S. (2005). Brokerage and closure: An introduction to social capital. Oxford, NY: Oxford University Press. Burt, R. S., & Knez, M. (1995). Kinds of third-party effects on trust. Rationality and Society, 7(3), 255–292. https​:/​/do​​i​.org​​/10​.1​​177​/1​​04346​​31950​​​07003​​003 Buskens, V. (1998). The social structure of trust. Social Networks, 20(3), 265–289. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​0​378-​​8733(​98)00​005-7​ Cartwright, D., & Harary, F., (1956). A generalization of Heider’s theory. Psychological Review, 63, 277–292. https://doi​.org​/10​.1037​/h0046049 Casciaro,T., & Lobo, M. S. (2008).When competence is irrelevant:The role of interpersonal affect in task-related ties. Administrative Science Quarterly, 53(4), 655–684. https://doi​. org​/10​.2189​/asqu​.53​.4​.655 Chua, R.Y. J., Ingram, P., & Morris, M. W. (2008). From the head and the heart: Locating cognition-and affect-based trust in managers' professional networks. Academy of Management Journal, 51(3), 436–452. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​08​​.32​​62595​6 Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120. https://doi​.org​/10​.1086​/228943

202  B. McEvily, A. Zaheer, and G. Soda

Coleman, J. S. (1990). Foundations of Social Theory. Cambridge, MA: Belknap Press of Harvard University Press. De Jong, B., Kroon, D. & Schilke, O. (2017). The Future of Organizational Trust Research: A Content-Analytic Synthesis of Scholarly Recommendations and Review of Recent Developments. P. van Lange, B. Rockenbach & T.Yamagishi, eds. Trust in Social Dilemmas. NY: Oxford University Press, 173–194. https://doi.org/10.1093/oso/ 9780190630782.003.0010 Ertug, G., & Castellucci, F. (2013). Getting what you need: How reputation and status affect team performance, hiring, and salaries in the NBA. Academy of Management Journal, 56(2), 407–431. https://doi​.org​/10​.5465​/amj​.2010​.1084 Faust, K. (1997). Centrality in affiliation networks. Social Networks, 19(2), 157–191. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​0​378-​​8733(​96)00​300-0​ Ferrin, D. L., Dirks, K. T., & Shah, P. P. (2006). Direct and indirect effects of third-party relationships on interpersonal trust. Journal of Applied Psychology, 91(4), 870. https​:/​/do​​i​. org​​/10​.1​​037​/0​​021​-9​​010​​.9​​1​.4​.8​​70 Gambetta, D. (1993). The Sicilian Mafia. The Business of Private Protection. Cambridge, MA: Harvard University Press. Gausdal, A. H. Svare, H., & Möllering, G. (2016) Why don't all high-trust networks achieve strong network benefits? A case-based exploration of cooperation in Norwegian SME networks, Journal of Trust Research, 6(2), 194–212. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​. 20​​​16​.12​​13173​ Gould, R. V. (2002). The origins of status hierarchies: A formal theory and empirical test. American Journal of Sociology, 107(5), 1143–1178. https://doi​.org​/10​.1086​/341744 Granovetter, M. (1973).The strength of weak ties. American Journal of Sociology, 78(6), 1360– 1313. https://doi​.org​/10​.1086​/225469 Granovetter, M. (1985). Economic action and social structure:The problem of embeddedness. American Journal of Sociology, 91(3), 481–510. https://doi​.org​/10​.1086​/228311 Gulati, R. (1995a). Does familiarity breed trust? The implications of repeated ties for contractual choice in alliances. Academy of Management Journal, 38(1), 85–112. https:// doi​.org​/10​.5465​/256729 Gulati, R. (1995b). Social structure and alliance formation patterns: A longitudinal analysis. Administrative Science Quarterly, 619–652. https://doi.org/10.2307/2393756 Gulati, R., & Sytch, M. (2007). Dependence asymmetry and joint dependence in interorganizational relationships: Effects of embeddedness on a manufacturer's performance in procurement relationships. Administrative Science Quarterly, 52(1), 32–69. https://doi​.org​/10​.2189​/asqu​.52​.1​.32 Gunia, B. C. (2019). The Bartering Mindset: A Mostly-Forgotten Framework for Mastering Your Next Negotiation. University of Toronto Press. https://doi​.org​/10​.3138​/9781487512163 Heider, F. (1946). Attitudes and cognitive organization. Journal of Psychology, 21, 107–112. https​:/​/do​​i​.org​​/10​.1​​080​/0​​02239​​80​.19​​​46​.99​​17275​ Heider, F. (1958). The Psychology of Interpersonal Relations. New York:Wiley. https://doi​.org​/ 10​.1037​/10628​-000 Hummon, N. P., & Doreian, P. (2003). Some dynamics of social balance processes: Bringing Heider back into balance theory. Social Networks, 25, 17–49. https​:/​/do​​i​.org​​/10​.1​​016​/ S​​0​378-​​8733(​02)00​019-9​ Hurley, R. F., Gillespie, N., Ferrin, D. L., & Dietz, G. (2013). Designing trustworthy organizations. MIT Sloan Management Review, 54(4), 75.

Network Trust 

203

Ibarra, H., Kilduff, M., & Tsai, W. (2005). Zooming in and out: Connecting individuals and collectivities at the frontiers of organizational network research. Organization Science, 16(4), 359–371. https://doi​.org​/10​.1287​/orsc​.1050​.0129 Jones, S., & Shah, P. P. (2021). The tangled ties of trust: A social network perspective on interpersonal trust. In N. Gillespie, A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. New York, NY: Routledge. Krackhardt, D. (1990). Assessing the political landscape: Structure, cognition, and power in organizations. Administrative Science Quarterly, 342–369. https://doi​.org​/10​.2307​/ 2393394 Krackhardt, D. (1999). The ties that torture: Simmelian tie analysis in organizations. In S. B. Andrews & D. Knoke (Eds.), Research in the sociology of organizations: Networks in and around organizations (vol. 16, pp. 183–210). Bingley, UK: Emerald Group Publishing Limited. Krackhardt, D., & Hanson, J. R. (1993). Informal networks. Harvard Business Review, 71(4), 104–111. Kramer, R. M. (2010). Collective trust within organizations: Conceptual foundations and empirical insights. Corporate Reputation Review, 13, 82–97. https://doi​.org​/10​.1057​/crr​. 2010.9 Kramer, R. M., & Lewicki, R. (2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annals, 4(1), 245–277. https​:/​/do​​i​.org​​/10​.5​​465​/1​​94165​​20​.20​​​10​.48​​7403 Lau, D. C., & Liden, R. C. (2008). Antecedents of coworker trust: Leaders’ blessings. Journal of Applied Psychology, 93(5), 1130–1138. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​3​.5​.1​​130 Laumann, E. O., Marsden, P.V., & Prensky, D. (1989). The boundary specification problem in network analysis. Research Methods in Social Network Analysis, 61, 87. Lewicki, R. J., & Brinsfield, C. (2011). Framing trust:Trust as a heuristic. In W. A. Donohue, R. G. Rogan, & S. Kaufman (Eds.), Framing matters: Perspectives on negotiation research and practice in communication (pp. 110–135). New York, NY: Peter Lang Publishing. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​. 19​​95​.​95​​08080​​335 McAllister, D. J. (1995). Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https:// doi​.org​/10​.2307​/256727 McEvily, B. (2011). Reorganizing the boundaries of trust: From discrete alternatives to hybrid forms. Organization Science, 22(5), 1266–1276. https://doi​.org​/10​.1287​/orsc​. 1110​.0649 McEvily, B., Perrone,V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, 14(1), 91–103. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.9​​1​.128​​14 McEvily, B., Soda, G., & Tortoriello, M. (2014). More formally: Rediscovering the missing link between formal organization and informal social structure. Academy of Management Annals, 8(1), 299–345. https​:/​/do​​i​.org​​/10​.5​​465​/1​​94165​​20​.20​​​14​.88​​5252 McEvily, B., & Zaheer, A. (2004). Architects of trust: The role of network facilitators in geographical clusters. In R. Kramer & K. Cook (Eds.), Trust and distrust in organizations: Dilemmas and approaches (pp. 189–213). New York, NY: Russell Sage Foundation. McKnight, D. H., Cummings, L. L., & Chervany, N. L. (1998). Initial trust formation in new organizational relationships. Academy of Management Review, 23(3), 473–490. https://doi​. org​/10​.5465​/amr​.1998​.926622

204  B. McEvily, A. Zaheer, and G. Soda

Mehra, A., Kilduff, M., & Brass, D. J. (1998). At the margins: A distinctiveness approach to the social identity and social networks of underrepresented groups. Academy of Management Journal, 41(4), 441–452. https://doi​.org​/10​.5465​/257083 Meyerson, D.,Weick, K. E., & Kramer, R. M. (1996). Swift trust and temporary groups. Trust in Organizations: Frontiers of Theory and Research, 166, 195. https://doi​.org​/10​.4135​/ 9781452243610​.n9 Newman, M. E. (2002). Assortative mixing in networks. Physical Review Letters, 89(20), 208701. https​:/​/do​​i​.org​​/10​.1​​103​/P​​hysRe​​vLett​​​.89​.2​​08701​ Podolny, J. M. (1993). A status-based model of market competition. American Journal of Sociology, 98(4), 829–872. https://doi​.org​/10​.1086​/230091 Puranam, P. (2018). The microstructure of organizations. Oxford University Press. https​:/​/do​​i​. org​​/10​.1​​093​/o​​so​/97​​80199​​67236​​​3​.001​​.0001​ Reagans, R., & McEvily, B. (2003). Network structure and knowledge transfer: The effects of cohesion and range. Administrative Science Quarterly, 48, 240–267. https://doi​.org​/10​. 2307​/3556658 Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Shazi, R., Gillespie, N., & Steen, J. (2015).Trust as a Predictor of innovation network ties in project teams. International Journal of Project Management, 33(1), 81–91. https​:/​/do​​i​.org​​/ 10​.1​​016​/j​​.ijpr​​oman.​​20​14.​​06​.00​1 Simmel, G. (1950). The triad. In K. H. Wolff (Ed.), The Sociology of Georg Simmel (pp.145– 169). New York, NY: Free Press. Soda, G., & Zaheer, A. (2012). A network perspective on organizational architecture: Performance effects of the interplay of formal and informal organization. Strategic Management Journal, 33(6), 751–771. https://doi​.org​/10​.1002​/smj​.1966 Tortoriello, M., McEvily, B., & Krackhardt, D. (2015). Being a catalyst of innovation: The role of knowledge diversity and network closure. Organization Science, 26(2), 423–438. https://doi​.org​/10​.1287​/orsc​.2014​.0942 Uzzi, B. (1997). Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly, 35–67. https://doi​.org​/10​.2307​/ 2393808 Uzzi, B., & Spiro, J. (2005). Collaboration and creativity:The small world problem. American Journal of Sociology, 111(2), 447–504. https://doi​.org​/10​.1086​/432782 Watts, D. J. (2004).The "new" science of networks. Annual Review of Sociology, 30, 243–270. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​.soc​​.30​.0​​​20404​​.1043​​42 Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of ‘small-world’ networks. Nature, 393(6684), 440. https://doi​.org​/10​.1038​/30918 Yamagishi, T., & Yamagishi, M. (1994). Trust and commitment in the United States and Japan. Motivation and Emotion, 18(2), 129–166. https://doi​.org​/10​.1007​/BF02249397 Zucker, L. G. (1986a). Production of trust: Institutional sources of economic structure: 1840 to 1920. In L.L. Cummings & B. Staw (Eds.), Research in Organizational Behavior, Vol. 8 (pp. 55–111). Greenwich, CT: JAI.

9 THE TANGLED TIES OF TRUST A Social Network Perspective on Interpersonal Trust Stephen Jones and Priti Shah

Introduction Interpersonal trust is complex because it inheres in relationships instead of individuals. It requires both a trustor who renders a trust judgment and a trustee who is the target of the trust. This basic notion points to three loci or centers of action (Jones & Shah, 2016) within a trust relationship: the trustor, the trustee, and the dyad – which is the combination of the two (Figure 9.1). Trust scholars have generated novel and important insights by examining these three trust loci in lab experiments and field observations (see Fulmer & Gelfand, 2012, for a review). Yet, much of the existing literature isolates the trust relationship from its surrounding context, attempts to control for the context, or simply ignores it. This is problematic because interpersonal trust develops and decays in a multilevel social field (Ferrin, Dirks, & Shah, 2006; Fulmer & Gelfand, 2012; McKnight, Cummings, & Chervany, 1998), which further amplifies its complexity.The social field both shapes trust relationships and is shaped by them. To cut through the complexity, we advocate for a social network perspective on interpersonal trust. A social network perspective allows one to examine the three loci within a trust relationship and to examine the broader structure of trust relationships at multiple levels in a social field. It embraces the complexity of relational concepts, multiple levels, and structures. Thus, social networks provide a common framework to examine trust within and between relationships. In effect, social networks provide us with a language and methodology to describe the context’s influence on the trustor, the trustee, and their dyadic trust relationship and, reciprocally, how these trust ties shape their surrounding context at multiple levels. There are two key components to a network perspective – one relational, the other structural (Barden & Mitchell, 2007; Granovetter, 1992) – which are DOI: 10.4324/9780429449185-9

206  Stephen Jones and Priti Shah

Structural component Relational component Trustee Trustor Dyad

Trustor. Individual rendering trust judgment Trustee. Referent for trust. Dyad. Trustor-trustee pair. Arrows indicate the direction of trust, from trustor to trustee. A double arrow indicates reciprocated trust.

FIGURE 9.1 Relational

and structural components of a trust network.

depicted in Figure 9.1. The relational component is concerned with the quality and nature of a social tie. It helps us understand how the trustor’s and trustee’s personality, attributes, and interactions inform trust and how trust within the dyad shapes other aspects of the trustor–trustee relationship. The structural component is concerned with the larger organization of social ties and affiliations surrounding a dyad. It helps us understand how the network’s structure and a dyad’s network position influence trust development as well as how trust relationships might change the structure of the network. Thus, the relational component examines the trustor, trustee, and dyadic attributes within a trust relationship and the structural component examines connections among trust relationships. The combination of both perspectives is important; the relational component provides insights into the psychological and behavioral mechanisms that influence trust development, and the structural component clarifies the positional and contextual mechanisms that generate and moderate interpersonal trust. Without the relational component, we may fail to recognize the trustor’s and trustee’s psychological and behavioral inputs that shape trust; without the structural component, we may fail to recognize the direct or moderating influence of the social context affecting trust development or, even worse, misattribute the contextual influences to the more commonly investigated trustor, trustee, or dyadic factors. A social network perspective also gives trust scholars a common language and methodology for studying multilevel antecedents and outcomes of interpersonal trust. Trust dyads are embedded in teams, functions, divisions, and organizations, and each level introduces a potential boundary for network analysis. For example, if a research question involves trust in teams, then network theory and methods may be applied to interpersonal trust networks within teams. Or if a research question involves trust in a division, then the same network approach may be applied to interpersonal trust networks within a division as well as the teams within the

The Tangled Ties of Trust  207

division. Ultimately, the macro level in multilevel research sets the boundary for network analysis, but the ability to examine trust within and between dyadic relationships remains consistent regardless of the level chosen. In this chapter, we first briefly review the concept of interpersonal trust and explain the basic elements of social network theory and analysis that can be used to examine it. Next, we use social network theory to explain the formation of interpersonal trust – both reviewing the literature and exploring future directions. Specifically, we explore how social network determinants can influence the trustor, trustee, and trustor–trustee dyad. Then we explore the nascent topic of how interpersonal trust may shape social networks. Finally, we conclude with the benefits of adopting a social network perspective to understand interpersonal trust and multiple levels.

Theoretical Foundation Trust Conceptualizations Scholars describe and conceptualize trust in different ways (Bhattacharya, Devinney, & Pillutla, 1998; Fulmer & Gelfand, 2012), which has led to a family of trust constructs in the literature. Some conceptualize interpersonal trust as positive expectations (Ferrin & Dirks, 2003; Ferrin et al., 2006) or trusting beliefs (McKnight et al., 1998) that one person has toward another. Specifically, they focus on perceived trustworthiness or its three dimensions – perceived ability, integrity, and benevolence (Mayer, Davis, & Schoorman, 1995) – to indicate positive expectations that a trustor has toward a trustee. Others have conceptualized interpersonal trust as willingness to be vulnerable (Colquitt, Scott, & LePine, 2007; Mayer et al., 1995) or trusting intent (McKnight et al, 1998). Willingness to be vulnerable indicates motivation, whereas perceived trustworthiness indicates a belief. At times willingness to be vulnerable is divided into trust based on affect or trust based on cognition (McAllister, 1995; cf. Lewicki & Bunker, 1995). In work contexts, it is often divided into a willingness to rely on another or a willingness to disclose sensitive information to another (Gillespie, 2003). Still others define trust as trusting actions (Kramer, Shah, & Woerner, 1995; Malhotra, 2004). The trusting actions definition is frequently used when experimental trust games are played in which subject pairs pass or return money to each other (Pillutla, Malhotra, & Murnighan, 2003; Sniezek & Van Swol, 2001). All three trust conceptualizations fit well within a social network perspective, but there are distinctions between perceived trustworthiness and willingness to be vulnerable on the one hand and trusting actions on the other. The first two cannot be explicitly observed by others but the latter can. Because trusting actions capture actual trustor behavior, the trustor’s and trustee’s recollections of the trusting behavior are apt to be similar. This means trusting actions can be assessed accurately by the trustor, trustee, or others who witness it – though perspectives

208  Stephen Jones and Priti Shah

regarding the motives underlying the trusting actions may differ. In contrast, only the trustor can provide an accurate assessment of her own trustworthiness perceptions or willingness to be vulnerable. As such, the trustee’s perceptions of how much the trustor trusts her is prone to be inaccurate (Campagna, Dirks, Knight, Crossley, & Robinson, 2019). This has implications for social networks because, as we will articulate below, trust can diffuse through a network. When trust is conceptualized as a belief or motivation, it must be communicated by the trustor directly to others to whom she is connected in order to diffuse through the network (Burt & Knez, 1995). In comparison, trusting actions can diffuse more readily as they are observable by others in the network without explicit effort by the trustor to transfer trust. Nevertheless, those who observe trusting actions must attribute motives to the trustor to make sense of them (Ferrin & Dirks, 2003), which may lead to discrepant interpretations. In comparison, direct communication from the trustor regarding her beliefs and motives is apt to transfer trust more precisely.Thus, using a conceptualization based on beliefs or motives, one may expect trust to diffuse quite powerfully but only within proximate network relations. In contrast, one may expect trust to diffuse more broadly but less potently with a trusting action conceptualization.

Social Network Background This section provides a lexicon for the different constructs that can be investigated from a social network perspective.We will focus on both the tie that connects the trustor and trustee as well as the different structural configurations that surround them. From a network perspective, an individual actor can be simultaneously a trustor and a trustee because actors both render trust judgments and are judged by other actors.

Relational Component The trustor and trustee together form a dyad with a trust tie between them. Trust is just one of many types of ties that connect one actor to another. Ties can be expressive or instrumental (Ibarra, 1993). Expressive ties are usually strong ties among similar others.They are affective in nature, providing a high degree of intimacy associated with friendship and social support (Shah, 2000). Instrumental ties are normally weaker connections often with dissimilar others that provide access to resources, expertise, information, political access, and advice (Shah, 2000). The delineation of expressive and instrumental ties aligns closely with trust’s affective and cognitive dimensions. Affect-based trust reflects positive beliefs about a trustee due to an emotional connection based on mutual caring and concern. It is forged through a strong interpersonal connection and most closely aligns with an expressive tie. Cognitive-based trust reflects positive beliefs about a trustee based

The Tangled Ties of Trust  209

on his or her reliability, dependability, professionalism, or expertise and closely aligns with an instrumental tie (Chua, Ingram, & Morris, 2008; Gibbons, 2004). Connections may also be affiliative, such as when two actors share friendship or group membership. Or they may capture resource flows such as information, help, or resources (Borgatti & Halgin, 2011). A key feature of these ties is whether they are symmetric or asymmetric. Affiliative and expressive ties are normally symmetric because they reflect a mutual connection and a strong degree of intimacy, respectively. In contrast, instrumental ties often represent directional flows of information or help and are usually asymmetric. Trust ties may be symmetric or asymmetric. Symmetric trust ties exist when both the trustor and trustee share the same level of trust, albeit mutually high or low (Korsgaard, Brower, & Lester, 2015). Asymmetric trust ties indicate a disparity in the level of trust between two actors with one trusting the other to a greater degree than the other way around. Affective trust ties that reflect an emotional connection are likely to be symmetric, whereas cognitive trust ties that reflect skills, expertise, or reliability may be asymmetric. Note, a perceptual or motivational measure of trust (i.e., perceptions of trustworthiness or willingness to be vulnerable) may also be asymmetric given variations in individuals’ perceptions, while a behavior-based tie is more likely to be symmetric if individuals are reporting on the same actions. Ties may be binary, weighted, or signed. In a binary trust network, a tie (value 1) represents the presence of trust and the lack of a tie (value 0) signifies the absence of trust (Wong & Boh, 2010). In a weighted trust network, a tie represents perhaps weak, moderate, or strong trust (value 1–3), and the lack of a tie (value 0) represents no trust (Burt, Bian, & Opper, 2018; Chua et al., 2008; Ferrin et al., 2006; Jones & Shah, 2016; Lau & Liden, 2008;Tasselli & Kilduff, 2018). In a signed trust network, a positive tie (value +1) represents the presence of trust, the lack of a tie (value 0) represents no trust, and a negative tie (value −1) signifies distrust – if trust and distrust are conceptualized on the same continuum, though scholars have argued against such a conceptualization (Lewicki, McAllister, & Bies, 1998; see also Marineau, 2017). Dyadic relations may also be multiplex, which occurs when actors share more than one type of tie. For example, two actors may share reciprocal trust ties, an affiliative workgroup tie, and reciprocal social support ties. In contrast, two actors who trust one another but are in separate workgroups and do not share social support would not have a multiplex relationship. Multiplex relations are stronger and are more likely to endure because of the multiple layers of connection, albeit they may require more resources to maintain than simplex ties (Ingram & Zou, 2008; Methot, LePine, Podsakoff, & Christian, 2016; Shah, Parker, & Waldstrøm, 2017).

Structural Component The network structure surrounding a focal dyad influences the dyad’s relationship, and it influences the behavior and judgments of individuals within the dyad.

210  Stephen Jones and Priti Shah

We review a few of the most common structures that influence actors and dyads. Third parties, triads, density, and structural equivalence affect the nature of dyadic relationships. Centrality, brokerage, and cognitive social structure affect trustors and trustees within dyads. We provide a graphical illustration of each of these structures in Figure 9.2. Third parties. Third parties are actors outside a focal dyad who have relationships with both actors within the dyad.These third parties can fundamentally alter the nature of the dyadic relationship. They act as indirect paths for information to flow between one dyad member to another (Burt & Knez, 1995; Ferrin et al., 2006; Lau & Liden, 2008). They can introduce actors to one another or act as arbiters when relations within a dyad are strained (Granovetter, 1983; Grosser, Obstfeld, Labianca, & Borgatti, 2019; Halevy, Halali, & Cohen, 2019). Thus, third parties can be the impetus for tie formation within a dyad. Triads. All dyadic and third-party ties become triads if all three actors are connected to each other, which are also known as Simmelian ties (Simmel, 1950). These Simmelian ties are particularly strong because they provide an opportunity to monitor and sanction behavior deemed unacceptable. An actor is constrained from exploiting another member of a triad because the third party acts as a witness, can remove support from the aggressor, or can warn others of the

Trustor. Trustor and trustee connected to the same third other. Trustor

Trustee

Centrality. The number of trustor or trustee ties. Trustor or trustee

Brokerage. A trustor or trustee who connects two or more otherwise disconnected groups. Trustor or trustee

Simmelian triad. Trustor and trustee connected to each other and a third other. Trustor

Trustee

Cognitive social structure (CSS). A trustor’s perception of network ties. Trustor 1

Trustor 2

Structural equivalence. A trustor and trustee who share a similar pattern of ties.

Trustor

Trustee (Actual 4-person network)

Trustor 3

Trustor 4 (Trustor’s perception)

FIGURE 9.2 Network

structure illustrations.

The Tangled Ties of Trust  211

malfeasance. The greater the number of Simmelian ties connecting an actor, the greater the constraint on behavior (Krackhardt, 1999). Density. Density is another variable with properties similar to Simmelian ties. Density is a measure indicating the ubiquity or sparseness of connections within a social network (Reagans, Zuckerman, & McEvily, 2004; Shah, Dirks, & Chervany, 2006). A density of 1.0 suggests that all individuals within a binary network are connected to each other, while a density of 0.5 indicates that only half the possible ties within a network actually exist. Dense networks create cohesion among the network members and allow for rich information flow and close monitoring (Balkundi & Harrison, 2006; Brass, 1984; Shah et al., 2006). The greater the density of a network, the greater the likelihood of Simmelian ties such that all actors will be aware of any potential transgressions. Structural equivalence. Structurally equivalent ties describe individuals who share a similar pattern of relationships with each other. That is, structurally equivalent actors occupy equivalent positions in the network (Burt, 1987; Galaskiewicz & Burt, 1991; Kilduff, 1990; Shah, 1998, 2000) such as when two managers are connected to bosses at similar levels and have ties to subordinates. Research shows that structurally equivalent actors influence one another. Competition is often cited as the reason, but socialization and ecological factors also play a role (Burt, 1987; Cartwright, 1965; Marsden & Friedkin, 1993). As such, influence among structural equivalent actors is due to common shared experiences, socialization practices, similar role demands and expectations, or an inherent rivalry when occupying the same position in the network. Centrality. Network position may also influence trust perceptions. Network centrality, one of the most popular measures of network position, refers to the number of connections an individual possesses in a network (i.e., the degree to which they are the hub of a network). There are different ways to calculate network centrality: degree centrality, which is a simple count of an individual’s connections; betweenness centrality, which is the number of times that an individual lies on the shortest path between two other actors in the network and provides an opportunity to control, distort, or omit information; or eigenvector (i.e., Bonacich) centrality, which accounts for the centrality of one’s connections and is often associated with power or access to information (Bonacich, 1987;Wasserman & Faust, 1994). Centrality reflects the relational activity of an individual (Marsden, 2002). It provides a measure of informal status or prestige in a network as one is considered an attractive interaction partner as centrality increases. All forms of centrality are also prone to what network scholars have called the Matthew effect: because central individuals have a strong reputation and many connections, they are apt to gain even more connections (Barabási & Albert, 1999; Merton, 1968; Rivera, Soderstrom, & Uzzi, 2010). However, as centrality increases, the need to manage and maintain existing relationships also increases, suggesting costs associated with this network position (Rivera et al., 2010). Friendship, for instance, can be costly to maintain, which limits the number of friendships an actor can garner.

212  Stephen Jones and Priti Shah

However, being trusted and trusting others have fewer limiting costs, so trust centrality can likely grow much larger than friendship centrality. Brokerage. Network brokers are individuals who connect two groups that would be otherwise disconnected. The space between these two unconnected groups is referred to as a structural hole. The greater the number of holes one spans the greater the brokerage potential. An individual connecting these structural holes is also in a position of power, for they have access to information, resources, or arbitrage opportunities long before others in the network (Burt, 2002a, 2004). This gives brokers a unique advantage to accrue benefits (Burt, 2007; Soda, Usai, & Zaheer, 2004; Zaheer & Soda, 2009). Brokers may also translate tacit knowledge across knowledge boundaries (Burt, 2007). They often span between functions or units and are tasked with ensuring communication flows between otherwise unconnected groups. Similar to the hazards of centrality, there are also costs associated with maintaining the ties needed to satisfy a brokerage role, particularly when the two groups one is spanning are each densely connected (Krackhardt, 1999; Tasselli & Kilduff, 2018). Cognitive social structure. The final network variable we focus on is perceptual. The cognitive social structure (CSS) captures an actor’s perception of the social network in which the actor is embedded (Krackhardt, 1987, 1990). While the network elements above are based on actual ties between actors, a CSS is a cognitive map an actor holds of her surrounding network. This is illustrated in the CSS explanation in Figure 9.2. While there are actual ties between the four actors in the illustrative network, each actor has a different subjective perception of the network, as shown in the four boxes. The actor in the top-left box believes all actors are tied in a dense network, whereas the actor in the bottom-right box believes no actors are tied. Only the actor’s perception in the bottom-left box matches the actual network. Because the perceptual network can be compared to the actual network, examining the CSS leads to measures of network accuracy.

Determinants of Interpersonal Trust from a Social Network Perspective In this section, we investigate how the different features of social networks discussed above influence interpersonal trust.We organize this section based on three key loci: the trustor, the trustee, and the trustor–trustee dyad (see Figure 9.1). For the trustor, we focus on how social network characteristics affect trustors’ experiences, access to information, or expectations, which subsequently influence their trust in others. The trustee section explores how facets of the surrounding social network systematically influence the behaviors and actions of trustees. In the dyad section, we investigate how the relational and structural social network features influence the mutual trust that is forged within the dyad. In so doing, we review existing literature on the influence of network factors on the trustors’ perceptions,

The Tangled Ties of Trust  213

the trustees’ behaviors and actions, and the dyadic trustor–trustee relationships and identify areas for future research.

Social Network Factors Influencing Trustors Early work on trust examined trustors’ individual or experiential differences that generated generalized trust expectations toward others (Gurtman, 1992; Jang, Livesley & Vernon, 1996; Rotter, 1971). This line of research views trust as an individual difference unique to the trustor that influences all of a trustor’s relationships. Similarly, in this section, we detail how social networks in which the trustor is embedded systematically may affect a trustor’s perceptions and motivations to trust others. We focus on the network effects of trustor centrality and a trustor’s perceptions of their surrounding social network (i.e., CSS). We note, however, that the effect of social networks on trust judgments and motivations is largely unexplored. We review the existing evidence and provide some avenues for further research. Central individuals possess higher status and power (Rossman, Esparza, & Bonacich, 2010; Wasserman & Faust, 1994), so research that connects status and power to trust easily applies to individuals in such positions. Lount and Pettit (2012) showed that higher-status individuals are more likely to trust than those in lower-status positions. Those with high status believe that trustees will want to act favorably toward them and perceive trustees as more benevolent. Thus, it is likely that more central individuals will increase their willingness to be vulnerable. However, there are counterarguments that may limit the degree to which central individuals are more likely to trust.Trustors gather information about trustees to determine insights into the trustees’ dispositions.The act of using information to make a trustworthiness determination is increasingly viewed by scholars as a type of attributional process (Kelley, 1973; Kelley & Michela, 1980; Reeder & Brewer, 1979) in which the trustor ascribes a level of ability, integrity, and benevolence to the trustee (Kim, Ferrin, Cooper, & Dirks, 2004; Weber, Malhotra, & Murnighan, 2004). This is important because social networks can influence attributional processes as well. One attributional argument in particular is important when using a social network perspective: behaviors are more diagnostic of disposition when those behaviors are distinct from others in the same social group (Kassin, 1979; Kelley & Michela, 1980). Group members set the normative expectations for how one in that group and situation should act; when one acts contrary to that norm, it indicates that one is expressing their disposition instead of conforming to a given situation.This idea easily transfers into social networks and the nature of a trustor’s relations with others in the social network. A central trustor is apt to have many positive relationships surrounding them, which is apt to raise the central actor’s normative expectancies of behaviors indicative of trust. To the extent that a trustee’s relationship with a trustor is positive, the trustee’s relationship to the trustor will be similar to other relationships and will not convey diagnostic information

214  Stephen Jones and Priti Shah

about the trustee’s disposition. But when the trustee’s relationship with the trustor is less positive, it will be diagnostic and provide useful information about a lack of trustworthiness. Thus, those who diverge from the trustor’s high expectations are apt to be judged more severely and to be viewed as less trustworthy. Contrariwise, when a trustor is on the periphery and has few close relationships, a positive relationship with a trustee may be particularly diagnostic, as the trustee will be viewed more favorably and will be trusted more. Moreover, motivated attribution theory suggests a greater tendency towards trusting others when one is less central because one’s dependence on any existing connections is amplified (Weber et al., 2004). A more peripheral trustor is likely to feel greater vulnerability as few others are apt to provide her with resources or information. This increased vulnerability causes distress, which can be resolved through cognitive means: trustors can self-servingly evaluate a trustee as more trustworthy to balance out the dependence (Emerson, 1962; Kruglanski, 1996; Kunda, 1990). Thus, as dependency increases through limited ties to others, so too do judgments of trustworthiness of those to whom the trustor is connected. The tension between these attributional arguments, which suggests that centrality may lessen trust, and Lount and Pettit’s (2012) status argument, which suggests central individuals are apt to trust more, has yet to be disentangled. It provides an opportunity for future research. Interestingly, scholars also argue that dependence, the obverse of power, can make trust judgments more extreme. Sniezek and Van Swol (2001) showed that dependent individuals’ trust was both higher and lower, whereas less dependent individuals’ trust was more moderated. They argue that dependent individuals rely more on trust because they are more vulnerable in their dependent position. This makes ascertaining trust more salient and perhaps leads to more extreme ratings because of their additional effort to make a correct judgment. Peripheral individuals in a social network are normally more dependent, and, thus, may make more extreme judgments. While our discussion above provides insight into how or why a trustor’s network position may influence trust perceptions, we posit that network position may also influence when trust perceptions are formed. While trust researchers have theorized about the evolution of trust over time (Lewicki & Bunker, 1995; Mayer et al., 1995; McKnight et al., 1998; Rousseau, Sitkin, Burt, & Camerer, 1998), scholars have yet to investigate at what point the initial decision to trust is made. In one exception to this, Jones and Shah (2016) found in a longitudinal study of all three dimensions of perceived trustworthiness that individuals formed perceptions of ability more quickly than they formed perceptions of integrity and benevolence. In the initial stages of new team formation, 70% of trustors were comfortable forming judgments of ability; in contrast, only 62% and 50% were comfortable forming judgments of integrity and benevolence respectfully. Moreover, once formed, perceptions of ability and integrity did not change much over time, while perceptions of benevolence increased steadily, indicating stronger

The Tangled Ties of Trust  215

information-gathering or dyadic influences at play. Incorporating a social network perspective into these findings suggests an interesting temporal perspective to forming trust judgments. Since a trustor’s network position determines access to information, it may also influence when the trustor feels confident making trust perceptions or when they are willing to be vulnerable to a trustee. Those in more information-privy central positions may be able to formulate trust judgments more readily than those positioned in the network periphery. Access to information may also enable a central trustor to more accurately assess the surrounding trust network. Specifically, a central trustor’s assessment of the CSS may be more consistent with the actual trust network than that of a trustor who is peripheral in the network (see Figure 9.2). These, too, provide avenues for future research. Last, social network research may provide insight into the emerging literature on felt trust. This body of work focuses primarily on the positive consequences of feeling trusted by one’s supervisor or teammates such as enhanced task performance, contribution to a team, and increased creativity (Lau, De Jong, & Lam, 2018). Albeit, recent work has begun to investigate negative consequences as well such as increased demands or stress (Baer, Frank, Matta, Luciano, & Wellman, 2020). Investigating a trustor’s perceptual network of who trusts whom within their surrounding network may provide insights into her own behavior. As such, felt trust can extend beyond feeling trusted by one’s supervisor or a generalized feeling of trust within one’s team. It can extend to a cognitive social structure of ties with surrounding others. The CSS is important because actors may base their trust perceptions or decisions to trust on the information derived from their perceptions of the network instead of the objective network itself. If the pattern of ties perceived is inaccurate, the individual may act mistakenly or ineffectively, failing to trust those who are worthy of trust or erroneously bestowing trust on those undeserving (cf. Campagna et al., 2019). Moreover, a trustor’s perception of the surrounding communication, advice, or friendship network may play a role in her tendency to forge cognitive or affective trust ties.

Social Network Factors Influencing Trustees In this section, we explore the broader social network in which the trustee is embedded and how elements of this network influence a trustee’s behavior or perceptions of trustee behavior. In their seminal work, Mayer and colleagues (1995) shifted the focus of trust research from understanding one’s inherent propensity to trust to uncovering information about a prospective trustee to determine whether they are worthy of trust. While the initial intent was to directly observe trustee behaviors, a social network perspective expands the reach of these observations to others to whom a trustee is connected. Greater trustee centrality suggests that many others are aware of a trustee’s behaviors and hold the trustee as an esteemed member of the focal network. Indeed, researchers have found that a trustee’s personal network size is positively

216  Stephen Jones and Priti Shah

related to perceptions of trustworthiness (Wong & Boh, 2010).While past research has investigated integrity-based trust, centrality in different types of networks may provide information pertaining to benevolence or ability too. For example, centrality in friendship networks may affirm the affective qualities of the trustee, leading to higher perceptions of benevolence. Centrality in instrumental or advice networks indicates the trustee has skills or expertise sought out by many others, leading to higher perceptions of ability. In contrast, centrality in networks based on negative ties may severely inhibit trust. For example, centrality in a relationship conflict network may be detrimental to perceptions of benevolence. Indeed, negative information is apt to diffuse more rapidly and forcefully than positive information (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001; Labianca, Brass, & Gray, 1998). As such, negative ties may cause more damage to perceptions of a trustee’s trustworthiness than positive ties offer benefits (Barsade, 2002; Bartel & Saavedra, 2000). In addition to the mere quantity of trustee contacts, structural attributes of these contacts are also important (Bonacich, 1987; Brass, 1984). A trustee’s connections act as third-party conduits through which information regarding the trustee travels (Burt & Knez, 1995). For instance, others trust individuals more when those individuals are trusted by their team leaders (Lau & Liden, 2008). Trustees also benefit from the network configuration of their contacts, which affects their contacts’ ability to disseminate information about their character. Wong and Boh (2010) investigated the role of trustee advocates (i.e., individuals who received social support and advice from the trustee in the past) and found that an advocate’s network heterogeneity, non-overlapping contacts, and network density are all positively related to perceptions of a trustee’s trustworthiness. All three of these features of an advocate’s network help expand the pool of people who can be made aware of a trustee’s positive character or attributes. Moreover, a trustee’s network size interacted with her advocates’ network attributes such that a trustee with a larger network benefitted from heterogeneity in her advocates’ network, but a trustee with a smaller network did not. These more nuanced findings suggest a complementarity between the trustee’s network and that of her advocates is needed to promote trust. Greater heterogeneity of an advocate’s network engenders greater credibility with regards to trust, but greater scope of a trustee’s network is needed to ensure a trustee’s visibility in the network (Wong & Boh, 2010). Overall, affective or instrumental networks facilitate trust perceptions when a trustee is more central, but occupying a brokerage position may adversely affect the trustee.As network brokers span the boundary across disconnected groups (i.e., structural holes), they control the flow of information, expertise, and resources across these groups. While this position is often cited as one of privilege given its associated arbitrage opportunities, people tend not to trust those who occupy this position (Coleman, 1988; Burt, 2001). As brokers control what flows from one group to another, information may be shared, distorted, or omitted for personal advantage. Additionally, broker trustees face difficulties managing conflicting

The Tangled Ties of Trust  217

demands of the groups they span, such that their actions may not always be in the best interest of those who occupy the two groups. These disadvantages are amplified when trustees are Simmelian brokers, or the sole connection between two groups who have densely connected ties (Krackhardt, 1999). Trustee loyalties are now torn between two groups that have strong normative expectations. However, recent research finds that some trustees possess personality traits that enable them to navigate the delicate balance of adhering to the normative expectations across these two densely connected yet separate groups. Simmelian broker trustees that exhibit diplomatic personalities and discreet communication styles are perceived as more trustworthy than those who are less diplomatic and more loquacious (Tasselli & Kilduff, 2018). We have more knowledge about how social networks affect trustees than how they affect trustors, as indicated by the greater breadth of literature in this section. However, there are more avenues worth pursuing to understand social network influences on trustees. For instance,Tasselli and Kilduff (2018) demonstrated that trustee attributes may interact with network position to affect individuals’ trustworthiness. They examined self-monitoring and the tendency to talk a lot, but other combinations of individual differences and network positions may be worth pursuing as well. Additionally, the mechanism that makes a central actor trustworthy could be further elucidated. Specifically, Podolny (2001) argues that there are two ways in which the effects of centrality are manifest. The first way is that ties act as pipes through which communication and information about the central actor can flow. This is the dominant view and is expressed by Wong and Boh (2010), Burt and Knez (1995), and others. The pipes allow for the diffusion of information. But there is a second way as well. In many situations, the ties that a central actor possesses are observable in a broader social field. Thus, those not tied to central actors nor their advocates may develop trust simply by observing central actors’ ties.The extent to which this second mechanism exists and the contexts in which it exists have not yet been explored.

Social Network Factors Influencing Dyads The sections above focused on network elements independently acting upon the trustor or the trustee. In this section, we embrace the relational nature of trust to explore how network elements simultaneously influence perceptions of dyadic trust. Direct mutual connections have long been cited as the basis of trust (Alfano, 2016; Blau, 1964; Nehamas, 2010; Thomas, 1987). Below, we examine how trust perceptions are influenced by both properties of the dyadic trustor–trustee tie and the third-party relationships surrounding this dyadic tie. Dyadic ties typically form due to homophily or proximity (Rivera et al., 2010). Demographic similarity, common background or group membership, or simply shared proximal space all provide the basis for forging dyadic ties. Common gender, race, nationality, or any form of social categorization commonality reduces

218  Stephen Jones and Priti Shah

the expected cost of forming relationships and lowers the perceived chances of misunderstanding or conflict (Ibarra, 1995; Kramer et al., 1995; Rivera et al., 2010). Thus, not only are actors more likely to form homophilous ties, but there is a greater tendency towards mutual trust across these ties. Indeed, evidence in the network literature finds that dyadic interpersonal ties provide the basis for affective trust, while ties that provide resources provide the basis for cognitive trust (Chen, Chang, & Hung, 2008; Chua et al., 2008). Proximity affects tie formation by increasing the opportunities for interaction and observing the actions of those in close proximity, making it easier to judge the trustworthiness of near actors (Rivera et al., 2010). Thus, homophily and proximity facilitate tie formation by building an initial degree of mutual trust. Dyadic ties can have several different relational properties. Trust ties can be symmetric or asymmetric. Much of the research foundation for trust in social exchange theory and existing models of interpersonal trust assumes some level of symmetry (Blau, 1964; Mayer et al., 1995). However, Korsgaard, Brower, and Lester (2015) have argued that may not be the case. They argue for a more nuanced perspective of dyadic trust: it may be reciprocated such that one party’s trust influences that of another; it may be mutual where both parties have similar levels of trust in each other; or it may remain asymmetric where the trustor and trustee differ in their trust perceptions or behaviors of each other. Asymmetric trust is seen to be quite disruptive both within the dyad and in the context of teams (Brower, Lester, Korsgaard, & Dineen, 2009; De Jong & Dirks, 2012). Interestingly, asymmetric trust may provide a launching point for the intersection of social network research with the growing literature on felt trust. This body of work suggests the imbalance between trust desired and felt trust has consequences for task performance, creativity, citizenship behavior, and stress, though researchers have found evidence that these consequences are both positive and negative (Baer et al., 2015; Baer et al., 2020; Lau et al., 2018; Lau, Lam, & Wen, 2014). Typically, research is conducted with a specific referent, be it a supervisor or teammate; a social network analysis would provide a more comprehensive mode of assessing trust discrepancies. Indeed, findings may reveal that a trustee is over-trusted by one facet of the network (supervisor to teammates) but under-trusted by another (teammates to supervisor). Exchanges within the dyad give the trustor access to direct information about the trustee. Not only do repeated dyadic exchanges foster relational cohesion often leading to mutual trust (Lawler & Yoon, 1993, 1996, 1998), but they also serve as an uncertainty reduction mechanism (Molm, Takahashi, & Peterson, 2000). The decision to trust leaves a trustor vulnerable to the trustee. As such, gathering first-hand knowledge of a trustee’s behavior, capabilities, or intent is critical to determining whether to trust (Mayer et al., 1995). Existing dyadic exchanges provide ample opportunity to make these observations. Repeated or future dyadic exchanges also provide the trustor an opportunity to sanction unscrupulous trustee behavior (Buskens & Raub, 2002), particularly if the trustee’s reputation can

The Tangled Ties of Trust  219

be damaged by gossip spreading through the surrounding network (Wu, Balliet, & Van Lange, 2016). Taken together, dyadic exchanges enable the trustor to gain greater confidence in whether to trust a prospective trustee, as trustors have access to both information of and controls for opportunistic behavior. The structural configuration of ties surrounding a dyad also plays an important role in the formation of trust ties. Actors near the focal dyad serve as important information sources that help with the decision of whether to trust. Specifically, actors can be influenced by their direct ties (i.e., cohesive referents), third-party ties (i.e., Simmelian ties), or structurally similar ties (i.e., structure equivalents). Cohesive actors are friends or other direct ties with whom we have close interpersonal relations. These are actors we rely upon to access information to manage surrounding uncertainty (Shah, 1998, 2000).The mechanism of influence is direct and cooperative as actors work together to form attitudes or perceptions or share information to facilitate this process for others (Ibarra & Andrews, 1993; Rice & Aydin, 1991). As trusting others may leave us vulnerable, we often look towards those we know best (i.e., cohesive ties) to garner relevant information. Indeed, our direct ties may provide valuable insight into the skills, expertise, reliability, and benevolence of prospective trustees. Alternatively, the behavioral or attitudinal mimicry often observed among structurally equivalent actors may result in trustors forging trust ties similar to those of their structural equivalents. Trustors may indirectly gather information by observing trust exchanges among structurally equivalent actors, assess the risks of opportunistic behavior, and determine whether to make themselves vulnerable to prospective trustees. In contrast, the competition among structurally equivalent trustors as they jockey for power and prestige within the social network may lead them to forge similar ties to maintain their network position. In the trust context, a trustor–trustee tie devoid of any connecting third parties may render the trustor vulnerable to a breach of trust, despite the trustor’s ability to sanction. As discussed above, the existence of Simmelian ties can serve as an informal policing mechanism such that any possible breach will be witnessed by others.The greater the number of Simmelian ties, the greater the awareness of any possible maleficence, rendering this type of behavior unlikely. Thus, as the density of ties in a network grows, so too does the likelihood of greater trust. Indeed, much of the network research on guanxi in China (i.e., strong ties indicative of enduring bonds based on common family, group, or geographic membership) relies heavily on the role of third-party intermediaries to vouch for the behavior of a trustor and trustee to help forge an initial trust tie (Burt, Bian, & Opper, 2018; Chou, Cheng, Huang, & Cheng, 2006; Chua, Morris, & Ingram, 2009).

Summary of Social Network Factors Influencing Trust The discussion above highlights different ways in which a trustor, trustee, or trustor–trustee dyad may be influenced by the social network in which they are

220  Stephen Jones and Priti Shah

embedded. Network position, albeit central or peripheral, can influence both trustor judgments and interpretations of trustee behavior. A trustor’s network position drives power dynamics, normative expectations, and information access as a means for influencing the decision to trust. Similarly, a trustee’s network position may influence how they are perceived by others. Those connected to a trustee serve as implicit or explicit advocates for the trustworthiness of the trustee. But trustees in brokerage positions occupy difficult positions and must satisfy differing expectations to be trusted, something that only those with a knack for diplomacy seem to do well. In moving the research focus from actual trust ties to perceived trust ties (i.e., CSS), there is also an opportunity to connect social network research with the emerging field of felt trust. Last, characteristics of the dyad and the structure surrounding the dyad influence trust. Homophily and proximity allow for greater interaction and reduce the hazards of trust. Third-party ties and structurally equivalent actors convey information that helps make the decision to trust.Third parties also possess sanctioning power. Thus, their mere presence fosters the trustworthy behavior of both parties in the dyad. While there exists empirical support for some of the social network effects we describe, there is still much that can be done. Many of our descriptions rely on only one or two studies that have yet to be replicated or extended. And while we highlight a few promising areas for research, we have just scratched the surface of the potential ways in which trust and social network research can be integrated to more accurately depict how trust forms.

Interpersonal Trust Influences on Social Networks Knowledge of how interpersonal trust influences social networks within organizations is very limited. McEvily, Perrone, and Zaheer (2003) proposed some novel ways by which trust structures organizational networks and mobilizes flows of information and commitment, but no scholars have empirically examined the topic. Simulation-based research on the related topic of trust diffusion in online reputation networks is developing (e.g., Hauke, Pyka, Borschbach, & Heider, 2010; Wang, Liu, Zhang, Xiong, & Lu, 2017), but it has not expanded into nonvirtual organizational settings. Thus, there is ample opportunity for scholars to explore the topic. McEvily, Zaheer, and Soda’s chapter in this volume is an excellent example of building theory in this area. In this section, we offer some arguments and conjectures as well to further this effort.

Influence of Interpersonal Trust Ties on the Diffusion of Trust in a Network Interpersonal trust may influence the diffusion of trust within a team, unit, or organizational network, and it may build other types of network ties such as friendship or advice. McEvily and colleagues (2003) used a microdynamics

The Tangled Ties of Trust  221

perspective to suggest how such changes may happen. They suggest that trust transferred through third-party ties is apt to increase trust density in a network. As discussed earlier, a trustor often uses a trusted third-party’s opinion of another person to help them formulate their own view of the person (Burt & Knez, 1995; Ferrin et al., 2006); thus, closing the trust triad increases the density in the trust network. One question that arises is whether trust can transfer through a chain of third parties. If person D is trusted by person C, C is trusted by person B, and B is trusted by person A, can trust transfer such that A will trust D? It is possible that trust could transfer through such chains as a form of endorsement much like gossip (Burt & Knez, 1995), but the credibility of the endorsement is apt to dissipate as the chain gets longer (Hauke et al., 2010; McEvily et al., this volume). Trust transferability is based on the strength of the endorsement and the strength of the trustor’s trust in the third party. If either is weak, then trust is less likely to transfer. In a longer third-party chain, any weak link may limit transfer. As such, a series of dyadic trust ties may over time impact the density of the overall trust network in a firm.

Influence of Interpersonal Trust on Structural Holes in the Network Another question is the extent to which trust transferability closes structural holes. Structural holes normally exist along team, unit, and organizational boundaries. If trust transferability is not able to leap across formal boundaries, then it is likely to only increase density within teams or within units – but it will not break down trust barriers between formal structures. It is likely the trust transferability helps initiate trust, but interaction between a trustor and trustee is needed to strengthen and cement it. Because formal boundaries tend to limit interaction, trust transferability may have limited effectiveness to build a tie between a trustor in one team or unit and a trustee in another. Furthermore, McEvily and colleagues (2003) argue that trust is likely to strengthen brokerage structures instead of building dense ties across team or unit boundaries. They propose that trust reduces redundancy in a network because team or unit members are willing to be vulnerable to brokers or liaisons who act on their behalf. Brokers are tasked with representing their team’s or unit’s interests to others in an organization. They must coordinate interunit work, resolve interunit conflict, and translate and transmit knowledge to other teams and units. If team or unit members do not believe the broker is effective or acting in their best interests, they may develop informal ties across boundaries to circumvent the broker. These trust network microdynamics raise a novel proposition: an organization’s social network will more closely mirror its formal structure and processes when there are many trustworthy individuals than when there are few trustworthy

222  Stephen Jones and Priti Shah

individuals in the organization. If we assume that an organization has designed its teams, units, and workflows to efficiently accomplish its goals and if we assume that trust ties will form between trustworthy co-workers when they interact, then the trust network should closely adhere to the team and unit structures and cross-unit workflows.Trust transitivity is apt to create dense clusters of trust triads within teams and units. And the willingness to rely on brokers for cross-team and cross-unit workflows is apt to sustain structural holes between teams and units. However, if trust is lacking, then informal networks may look quite different from the formal structure to compensate for the lack of trust. Teams are less likely to form dense clusters of trust ties, and brokerage ties are apt to be countered by additional boundary-spanning ties. This proposition offers an opportunity for new empirical tests as the transfer of trust across individuals has novel implications for the structural configuration of a firm’s network.

Interpersonal Trust and the Formation of Enduring Multiplex Ties McEvily and colleagues (2003) also argued that trust is likely to build multiplex ties. That is, when a dyad possesses a trust tie, it is likely that other types of ties such as friendship, advice, or interpersonal organizational citizenship behaviors (OCBIs) are apt to form and flow. Additionally, one dimension of trust may help another dimension form. For example, members of a newly founded team may become willing to rely on one another as they attempt to be efficient through role specialization. (If they do not develop a willingness to rely, then they would specialize less to check each other’s work.) This leads members to seek advice from one another based on their unique knowledge and skills. Willingness to rely may also lead to greater personal affinity toward team members and create expressive bonds or affective attachments.Team members may then become willing to disclose sensitive, personal information. Thus, a willingness to rely may lead to advice, friendship, and willingness to disclose ties. In another situation, two friends in separate teams may be willing to disclose sensitive, personal information because of their expressive tie. When they need work-related knowledge that does not exist in their respective teams, they may seek advice because there is a low cost to doing so (Casciaro & Lobo, 2008, 2015; Rivera et al., 2010). Indeed, Shazi, Gillespie, and Steen (2015) found that advice seeking in innovation contexts requires the seeker to first perceive the giver to be high in benevolence and ability.When multiplex ties form, they tend to be stronger than simplex ties as they provide greater returns from a single relationship, greater monitoring potential, and a greater degree of influence (Balkundi & Harrison, 2006; Rank, Robins, & Pattison, 2010). Thus, multiplex ties will tend to persist over time (Dahlander & McFarland, 2013) and might be more enduring through periods of interpersonal change. In times of organizational change – when organizations restructure or when jobs and individuals are removed – trust ties may be particularly relevant. Often

The Tangled Ties of Trust  223

change is enacted without regard to how the remaining employees are able to re-establish the social connections needed to continue their work (Shah, 2000). When key advice contacts are no longer available, employees may rely heavily on their existing trust network to help reestablish connections needed to complete their tasks. Indeed, past network research suggests that survivors suffering great losses in their friendship network are more likely to reconfigure their advice network post-layoff than those with a minimal loss of friends (Shah, 2000). And periods of organizational change may have the ability to spread a denser web of trust ties throughout an organization. Generally, ties decay following a power curve (Burt, 2000, 2002b), but they may also express vitality – that is, a tendency to persist – if they are forged in dense, multiplex relations (Maloney, Shah, Zellmer-Bruhn, & Jones, 2019). Moreover, trust ties may fall dormant but remain potential sources of network access (Levin, Walter, & Murnighan, 2011; Walter, Levin, & Murnighan, 2015). These ideas lead to another novel proposition: organizational change builds a network of active and dormant trust ties that do not mirror the organizational structure. Dense, multiplex clusters of ties may form within teams and units. But when teams disband or when individuals leave units, ties with former cluster members are apt to lose their multiplex nature.Yet, trust may be the key piece of a relationship that persists.Thus, as teams disband and new teams form or as employees move to different positions in an organization, they may create a network of trust ties that connect to distant or unconnected components of an organization. Such an informal trust network could be a unique avenue for accessing social capital (Levin et al., 2011; Maloney, et al., 2019; Walter et al., 2015).

Summary of Trust-Influencing Social Networks We have argued that trust transitivity (Ferrin et al., 2006; McEvily et al., this volume) and multiplexity (Casciaro & Lobo, 2008; Shazi et al., 2015) can generate dense trust networks – dependent on the trustworthiness of the individuals in the network. However, organizational change – both the leaving and entering of organization members and the reorganization of formal structures – can disrupt the emergence of greater trust density. Further, a lack of interaction between teams and divisions may limit the development of trust across intra-organizational boundaries. Thus, formal organizational structures have a substantial influence on trust networks. Yet, we do not expect trust networks to necessarily mirror the formal structures. Trust networks may adapt around untrustworthy individuals. For instance, if a boundary spanner such as a manager is not trustworthy, subordinates or superiors may build trust ties with others outside of the formal hierarchy that they would not have built if the manager was trustworthy. Moreover, trust ties across intra-organizational boundaries may develop when team members separate from one another, turning intra-team ties into boundary-spanning ties (Maloney et

224  Stephen Jones and Priti Shah

al., 2019). Thus, to understand the emergence of trust networks, it is important to consider prior trust relations, the trustworthiness of individuals, and formal organizational structures.

Conclusion Given its complexity and versatility, a social network perspective provides a useful way to examine and understand the multilevel nature of interpersonal trust. A social network perspective embraces trust’s relational nature; trust is easily modeled as a network tie between two actors. Social networks also easily model the social field in which a trust tie is embedded, whether the social field be a team, unit, or organization. Doing so enables scholars to investigate the contribution of trust to social network research, and vice versa, across different levels of analysis. Within a single network, for example, scholars can study trust at the unit level by examining unit density or other unit-level network structures as well as trust at the interpersonal level by observing individual ties. An individual’s position in a social network and the configuration of surrounding ties simultaneously influences one’s trust perceptions of others and whether one is viewed as trustworthy. One’s perception of the surrounding trust ties may also influence how subsequent ties are forged. At the dyadic level, trustor–trustee trust ties may be simultaneously influenced by their surrounding network structure and, in turn, have the potential to influence the pattern of ties as they create opportunities for multiplex ties. A social network lens investigating the trust across different levels also has important managerial implications. Trust ties provide the foundation upon which other communication, advice, and information-sharing ties are built. As such, knowing how the information structure of a firm influences these ties is imperative to fostering these ties.The research findings reviewed provide evidence of the importance of third parties in facilitating trust. When introducing new employees to teams or divisions, it may be more important to highlight where and with whom they may have worked before highlighting their skills and expertise. While the latter will be apparent through interaction, the former may be helpful to facilitate the initial trust tie. Moreover, cross-functional teams or rotations across units may be a way to ensure that connections are made across the firm that may later serve as a reference later in an employee’s career. In our research, we have seen the practical value of studying trust networks up close. We mapped the trust network of one municipal government, and the network map gave insight into how and where trust problems existed. When we debriefed the anonymized results with the government leader, it became clear that one division was experiencing a particular dearth of trust. Awareness of the network structure provided insight into the dysfunctional behavior observed among employees. Equipped with this information, the leader was able to intervene to rebuild the division’s trust.

The Tangled Ties of Trust  225

Scholars have increasingly turned to social networks to study trust, but much more can be done. Our review raises new avenues for future research regarding how networks affect a trustor’s decision to trust, a trustee’s trustworthy behavior, and the trustor–trustee relationship. It also suggests new research paths to develop knowledge about interpersonal trust’s impact on social networks. Trust is by its very nature a highly socialized phenomenon, embedded in a complex web of relationships. Adopting a social network perspective enables us to fully comprehend this complexity.

References Alfano, M. (2016). Friendship and the structure of trust. In A. Masala & J. Webber (Eds.), From personality to virtue: Essays on the philosophy of character (pp. 186–206). Oxford, UK: Oxford University Press. Baer, M. D., Dhensa-Kahlon, R. K., Colquitt, J. A., Rodell, J. B., Outlaw, R., & Long, D. M. (2015). Uneasy lies the head that bears the trust: The effects of feeling trusted on emotional exhaustion. Academy of Management Journal, 58(6), 1637–1657. https:// doi​.org​/10​.5465​/amj​.2014​.0246 Baer, M. D., Frank, E. L., Matta, F. K., Luciano, M. M., & Wellman, N. (2020) Under trusted, over trusted, or just right? The fairness of (In)congruence between trust wanted and trust received. Academy of Management Journal. https://doi​.org​/10​.5465​/ amj​.2018​.0334 Balkundi, P., & Harrison, D. A. (2006). Ties, leaders, and time in teams: Strong inference about network structure’s effects on team viability and performance. Academy of Management Journal, 49(1), 49–68. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​06​​.20​​78550​0 Barabási, A. L., & Albert, R. (1999). Emergence of scaling in random networks. Science, 286, 509–512. https​:/​/do​​i​.org​​/10​.1​​126​/s​​cienc​​e​.286​​​.5439​​.509 Barden, J. Q., & Mitchell, W. (2007). Disentangling the influences of leaders' relational embeddedness on interorganizational exchange. Academy of Management Journal, 50(6), 1440–1461. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​07​​.28​​22598​3 Barsade, S. G. (2002). The ripple effect: Emotional contagion and its influence on group behavior. Administrative Science Quarterly, 47(4), 644–675. https://doi​.org​/10​.2307​/ 3094912 Bartel, C. A., & R. Saavedra (2000). The collective construction of work group moods. Administrative Science Quarterly, 45(2), 197–231. https://doi​.org​/10​.2307​/2667070 Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than good. Review of General Psychology, 5(4), 323–370. https​:/​/do​​i​.org​​/10​.1​​037​/1​​089​-2​​68​0​. 5​​.4​.32​3 Bhattacharya, R., Devinney, T. M., & Pillutla, M. M. (1998). A formal model of trust based on outcomes. Academy of Management Review, 23(3), 459–472. https://doi​.org​/10​.5465​/ amr​.1998​.926621 Blau, P. M. (1964). Exchange and power in social life. New York: Wiley. Bonacich, P. (1987). Power and centrality: A family of measures. American Journal of Sociology, 92(5), 1170–1182. https://doi​.org​/10​.1086​/228631 Borgatti, S. P., & Halgin, D. S. (2011). On network theory. Organization Science, 22(5), 1168– 1181. https:// doi​.org​/10​.1287​/orsc​.110​0​.0641

226  Stephen Jones and Priti Shah

Brass, D. J. (1984). Being in the right place: A structural analysis of individual influence in an organization. Administrative Science Quarterly, 29(4), 518–539. https://doi​.org​/10​. 2307​/2392937 Brower, H., Lester, S., Korsgaard, A., & Dineen, B. (2009). A closer look at trust between managers and subordinates: Understanding the effects of both trusting and being trusted on subordinate outcomes. Journal of Management, 35, 327–347. https://doi​.org​/10​.1177​/ 0149206307312511 Burt, R. S. (1987). Social contagion and innovation: Cohesion versus structural equivalence. American Journal of Sociology, 92(6), 1287–1335. https://doi​.org​/10​.1086​/228667 Burt, R. S. (2000). Decay functions. Social Networks, 22(1), 1–28. https​:/​/do​​i​.org​​/10​.1​​016​/ S​​0​378-​​8733(​99)00​015-5​ Burt, R. S. (2001). Bandwidth and echo: Trust, information, and gossip in social networks. In A. Casella, & J. Rauch (Eds.), Networks and markets: Contributions from economics and sociology (pp. 30–74). New York: Russell Sage Foundation. Burt, R. S. (2002a). The social capital of structural holes. In M. F. Guillen, R. Collins, P. England, & M. Meyer (Eds.), The new economic sociology: Developments in an emerging field (pp. 148–190). New York: Russell Sage Foundation. Burt, R. S. (2002b). Bridge decay. Social Networks, 24(4), 333–363. https​:/​/do​​i​.org​​/10​.1​​016​/ S​​0​378-​​8733(​02)00​017-5​ Burt, R. S. (2004). Structural holes and good ideas. American Journal of Sociology, 110(2), 349–399. https://doi​.org​/10​.1086​/421787 Burt, R. S. (2007). Secondhand brokerage: Evidence on the importance of local structure for managers, bankers, and analysts. Academy of Management Journal, 50(1), 119–148. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​07​​.24​​16208​2 Burt, R. S., Bian,Y., & Opper, S. (2018). More or less guanxi:Trust is 60% network context, 10% individual difference. Social Networks, 54, 12–25. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.socn​​et​. 20​​​17​.12​​.001 Burt, R. S., & Knez, M. (1995). Kinds of third-party effects on trust. Rationality and Society, 7(3), 255–292. https​:/​/do​​i​.org​​/10​.1​​177​/1​​04346​​31950​​​07003​​003 Buskens, V., & Raub, W. (2002). Embedded trust: Control and learning. Advances in Group Processes, 19, 167–202. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​882-​​6145(​02)19​007-2​ Campagna, R. L., Dirks, K. T., Knight, A. P., Crossley, C., & Robinson, S. L. (2019). On the relation between felt trust and actual trust: Examining pathways to and implications of leader trust meta-accuracy. Journal of Applied Psychology, https://doi​.org​/10​.1037​/ apl0000474 Cartwright, D. (1965). Influence, leadership, control. In J. G. March (Ed.), Handbook of organizations (pp. 1–47). London: Routledge. https://doi​.org​/10​.4324​/9780203629130 Casciaro,T., & Lobo, M. S. (2008).When competence is irrelevant:The role of interpersonal affect in task-related ties. Administrative Science Quarterly, 53(4), 655–684. https://doi​.org​/ 10​.2189​/asqu​.53​.4​.655 Casciaro, T., & Lobo, M. S. (2015). Affective primacy in intraorganizational task networks. Organization Science, 26(2), 373–389. https://doi​.org​/10​.1287​/orsc​.2014​.0939 Chen, M. H., Chang,Y. C., & Hung, S. C. (2008). Social capital and creativity in R&D project teams. R&D Management, 38, 21–34. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-9310​​.200​7​​. 0049​​4.x Chou, L. F., Cheng, B. S., Huang, M. P., & Cheng, H. Y. (2006). Guanxi networks and members' effectiveness in Chinese work teams: Mediating effects of trust networks. Asian Journal of Social Psychology, 9, 79–95. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-839X​​. 200​6​​.0018​​5.x

The Tangled Ties of Trust  227

Chua, R. Y., Ingram, P., & Morris, M. W. (2008). From the head and the heart: Locating cognition- and affect-based trust in managers’ professional networks. Academy of Management Journal, 51(3), 436–452. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​08​​.32​​62595​6 Chua, R. Y., Morris, M. W., & Ingram, P. (2009). Guanxi vs networking: Distinctive configurations of affect- and cognition-based trust in the networks of Chinese vs American managers. Journal of International Business Studies, 40(3) 490–509. https​:/​/do​​i​. org​​/10​.1​​057​/p​​algra​​ve​.ji​​​bs​.84​​00422​ Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120. https://doi​.org​/10​.1086​/228943 Colquitt, J. A., Scott, B. A., & LePine, J. A. (2007).Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 92(4), 909–927. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​01​0​.9​​2​. 4​.9​0 Dahlander, L., & McFarland, D. A. (2013). Ties that last: Tie formation and persistence in research collaborations over time. Administrative Science Quarterly, 58(1), 69–110. https:// doi​.org​/10​.1177​/0001839212474272 De Jong, S. B., & Dirks, K. (2012). Beyond shared perceptions of trust and monitoring in teams: Implications of asymmetry and dissensus. Journal of Applied Psychology, 97, 391–406. https://doi​.org​/10​.1037​/a0026483 Emerson, R. M. (1962). Power-dependence relations. American Sociological Review, 27, 31– 41. https://doi​.org​/10​.2307​/2089716 Ferrin, D. L., & Dirks, K. T. (2003). The use of rewards to increase and decrease trust: Mediating processes and differential effects. Organization Science, 14(1), 18–31. https​:/​/ do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.1​​8​.128​​09 Ferrin, D. L., Dirks, K. T., & Shah, P. P. (2006). Direct and indirect effects of third-party relationships on interpersonal trust. Journal of Applied Psychology, 91(4), 870–883. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​1​.4​.8​​70 Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​.org​/ 10​.1177​/0149206312439327 Galaskiewicz, J., & Burt, R. S. (1991). Interorganization contagion in corporate philanthropy. Administrative Science Quarterly, 36(1), 88–105. https://doi​.org​/10​.2307​/2393431 Gibbons, D. E. (2004). Friendship and advice networks in the context of changing professional values. Administrative Science Quarterly, 49(2), 238–262. https://doi​.org​/10​. 2307​/4131473 Gillespie, N. (2003). Measuring trust in work relationships:The behavioural trust inventory, paper presented at the Academy of Management Annual Meeting, Seattle, USA, 1–6 August. Granovetter, M. (1983). The strength of weak ties: A network theory revisited. Sociological Theory, 1, 201–233. https://doi​.org​/10​.2307​/202051 Granovetter, M. S. (1992). Problems of explanation in economic sociology. In N. Nohria & R. G. Eccles (Eds.), Networks and organizations: Structure, form and action (pp. 25–56). Boston: Harvard Business School Press. Grosser, T. J., Obstfeld, D., Labianca, G., & Borgatti, S. P. (2019). Measuring mediation and separation brokerage orientations: A further step toward studying the social network brokerage process. Academy of Management Discoveries, 5(2), 114–136. https://doi​.org​/10​. 5465​/amd​.2017​.0110 Gurtman, M. B. (1992). Trust, distrust, and interpersonal problems: A circumplex analysis. Journal of Personality and Social Psychology, 62(6), 989–1002. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​022​-3​​514​​.6​​2​.6​.9​​89

228  Stephen Jones and Priti Shah

Halevy, N., Halali, E., & Cohen, T. R. (2019). Brokering orientations and social capital: Influencing others’ relationships shapes status and trust. Journal of Personality and Social Psychology. https://doi​.org​/10​.1037​/pspi0000213 Hauke, S., Pyka, M., Borschbach, M., & Heider, D. (2010). Augmenting reputationbased trust metrics with rumor-like dissemination of reputation information. In: K. Rannenberg,V.Varadharajan, & C. Weber (Eds.), Security and privacy – Silver linings in the cloud. IFIP Advances in information and communication technology (vol. 330, pp. 136–147). Berlin: Springer. https​:/​/do​​i​.org​​/10​.1​​007​/9​​78​-3-​​642​-​1​​5257-​​3​_13 Ibarra, H. (1993). Personal networks of women and minorities in management: A conceptual framework. Academy of Management Review, 18(1), 56–87. https://doi​.org​/ 10​.5465​/amr​.1993​.3997507 Ibarra, H. (1995). Race, opportunity, and diversity of social circles in managerial networks. Academy of Management Journal, 38(3), 673–703. https://doi​.org​/10​.5465​/256742 Ibarra, H., & Andrews, S. B. (1993). Power, social influence, and sense making: Effects of network centrality and proximity on employee perceptions. Administrative Science Quarterly, 38(2), 277–303. https://doi​.org​/10​.2307​/2393414 Ingram, P., & Zou, X. (2008). Business friendships. Research in Organizational Behavior, 28, 167–184. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.riob​​.200​8​​.04​.0​​06 Jang, K. L., Livesley, W. J., & Vernon, P. A. (1996). Heritability of the big five personality dimensions and their facets: A twin study. Journal of Personality, 64(3), 577–591. https​:/​/ do​​i​.org​​/10​.1​​111​/1​​467​-6​​494​.e​​​p9703​​13289​5 Jones, S. L., & Shah, P. P. (2016). Diagnosing the locus of trust: A temporal perspective for trustor, trustee, and dyadic influences on perceived trustworthiness. Journal of Applied Psychology, 101(3), 392–414. https://doi​.org​/10​.1037​/apl0000041 Kassin, S. M. (1979). Consensus information, prediction, and causal attribution: A review of the literature and issues. Journal of Personality and Social Psychology, 37(11), 1966–1981. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​.​3​​7​.11.​​1966 Kelley, H. H. (1973). The processes of causal attribution. American Psychologist, 28(2), 107– 128. https://doi​.org​/10​.1037​/h0034225 Kelley, H. H., & Michela, J. L. (1980). Attribution theory and research. Annual Review of Psychology, 31(1), 457–501. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​.ps.​​31​.02​​​0180.​​00232​5 Kilduff, M. (1990). The interpersonal structure of decision making: A social comparison approach to organizational choice. Organizational Behavior and Human Decision Processes, 47(2), 270–288. https​:/​/do​​i​.org​​/10​.1​​016​/0​​​749​-5​​978(9​0)900​39-C Kim, P. H., Ferrin, D. L., Cooper, C. D., & Dirks, K. T. (2004). Removing the shadow of suspicion: The effects of apology versus denial for repairing competence-versus integrity-based trust violations. Journal of Applied Psychology, 89(1), 104–118. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​9​.1​.1​​04 Korsgaard, M. A., Brower, H. H., & Lester, S. W. (2015). It isn’t always mutual: A critical review of dyadic trust. Journal of Management, 41(1), 47–70. https://doi​.org​/10​.1177​/ 0149206314547521 Krackhardt, D. (1987). Cognitive social structures. Social Networks, 9(2), 109–134. https​:/​/ do​​i​.org​​/10​.1​​016​/0​​​378​-8​​733(8​7)900​09-8 Krackhardt, D. (1990). Assessing the political landscape: Structure, cognition, and power in organizations. Administrative Science Quarterly, 35(2), 342–369. https://doi​.org​/10​.2307​/ 2393394 Krackhardt, D. (1999).The ties that torture: Simmelian tie analysis in organizations. Research in the Sociology of Organizations, 16, 183–210.

The Tangled Ties of Trust  229

Kramer, R. M., Shah, P. P., & Woerner, S. L. (1995). Why ultimatums fail. In R. M. Kramer & D. M. Messick (Eds.), Negotiation as a social process (pp. 285–308).Thousand Oaks, CA: SAGE. Kruglanski, A. W. (1996). Motivated social cognition: Principles of the interface. In: E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 493–520). New York: Guilford. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108(3), 480. https​:/​/do​​i​.org​​/10​.1​​037​/0​​033​-2​​909​​.1​​08​.3.​​480 Labianca, G., Brass, D. J., & Gray, B. (1998). Social networks and perceptions of intergroup conflict: The role of negative relationships and third parties. Academy of Management Journal, 41(1), 55–67. https://doi​.org​/10​.5465​/256897 Lau, D. C., De Jong, B., & Lam, L. W. (2018). Felt trust at work: A multi-theoretical and multi-referent examination. In G. Atinc (Ed.), Academy of Management Proceedings (Vol. 2018, No. 1, p. 13065). Briarcliff Manor, NY: Academy of Management. Lau, D. C., Lam, L. W., & Wen, S. S. (2014). Examining the effects of feeling trusted by supervisors in the workplace: A self-evaluative perspective. Journal of Organizational Behavior, 35(1), 112–127. https://doi​.org​/10​.1002​/job​.1861 Lau, D. C., & Liden, R. C. (2008). Antecedents of coworker trust: Leaders’ blessings. Journal of Applied Psychology, 93(5), 1130–1138. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​. 9​​3​.5​.1​​130 Lawler, E. J., & Yoon, J. (1993). Power and the emergence of commitment behavior in negotiated exchange. American Sociological Review, 58(4), 465–481. https://doi​.org​/ 10​.2307​/2096071 Lawler, E. J., & Yoon, J. (1996). Commitment in exchange relations: Test of a theory of relational cohesion. American Sociological Review, 61(1), 89–108. https://doi​.org​/ 10​.2307​/2096408 Lawler, E. J., & Yoon, J. (1998). Network structure and emotion in exchange relations. American Sociological Review, 63(6), 871–894. https://doi​.org​/10​.2307​/2657506 Levin, D. Z.,Walter, J., & Murnighan, J. K. (2011). Dormant ties:The value of reconnecting. Organization Science, 22(4), 923–939. https://doi​.org​/10​.1287​/orsc​.1100​.0576 Lewicki, R. J., & Bunker, B. (1995). Trust in relationships: A model of development and decline. In B. Bunker & J. Rubin (Eds.), Conflict, cooperation and justice (pp. 133–174). San Francisco, CA: Jossey-Bass. Lewicki, R. J., McAllister, D. J., & Bies, R. J. (1998). Trust and distrust: New relationships and realities. Academy of Management Review, 23(3), 438–458. https://doi​.org​/10​.5465​/ amr​.1998​.926620 Lount Jr, R. B., & Pettit, N. C. (2012). The social context of trust: The role of status. Organizational Behavior and Human Decision Processes, 117(1), 15–23. https​:/​/do​​i​.org​​/10​. 1​​016​/j​​.obhd​​p​.20​1​​1​.07.​​005 Malhotra, D. (2004). Trust and reciprocity decisions: The differing perspectives of trustors and trusted parties. Organizational Behavior and Human Decision Processes, 94(2), 61–73. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​0​​4​.03.​​001 Maloney, M. M., Shah, P. P., Zellmer-Bruhn, M., & Jones, S. L. (2019). The lasting benefits of teams: Tie vitality after teams disband. Organization Science, 30(2), 260–279. https:// doi​.org​/10​.1287​/orsc​.2018​.1254 Marineau, J. E. (2017). Trust and distrust network accuracy and career advancement in an organization. Group & Organization Management, 42(4), 487–520. https://doi​.org​/10​. 1177​/1059601115627529

230  Stephen Jones and Priti Shah

Marsden, P. V. (2002). Egocentric and sociocentric measures of network centrality. Social Networks, 24(4), 407–422. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​378-​​8733(​02)00​016-3​ Marsden, P. V., & Friedkin, N. E. (1993). Network studies of social influence. Sociological Methods & Research, 22(1), 127–151. https​:/​/do​​i​.org​​/10​.1​​177​/0​​04912​​41930​​​22001​​006 Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi​.org​/10​.2307​/258792 McAllister, D. J. (1995). Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https://doi​. org​/10​.2307​/256727 McEvily, B., Perrone,V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, 14(1), 91–103. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.9​​1​.128​​14 McEvily, B., Zaheer, A., & Soda, G. (2021). Network trust. In N. Gillespie, A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. New York, NY: Routledge. McKnight, D., Cummings, L., & Chervany, N. (1998). Initial trust formation in new organizational relationships. Academy of Management Review, 23(3), 473–490. https://doi​. org​/10​.5465​/amr​.1998​.926622 Merton, R. K. (1968). The Matthew effect in science: The reward and communication systems of science are considered. Science, 158, 56–63. https​:/​/do​​i​.org​​/10​.1​​126​/s​​cienc​​e​. 15​9​​.3810​​.56 Methot, J. R., Lepine, J. A., Podsakoff, N. P., & Christian, J. S. (2016). Are workplace friendships a mixed blessing? Exploring tradeoffs of multiplex relationships and their associations with job performance. Personnel Psychology, 69(2), 311–355. https://doi​.org​/ 10​.1111​/peps​.12109 Molm, L. D., Takahashi, N., & Peterson, G. (2000). Risk and trust in social exchange: An experimental test of a classical proposition. American Journal of Sociology, 105(5), 1396– 1427. https://doi​.org​/10​.1086​/210434 Nehamas, A. (2010). Aristotelian philia, modern friendship? In B. Inwood (Ed.), Oxford studies in ancient philosophy (pp. 213–247). Oxford: Oxford University Press. Pillutla, M. M., Malhotra, D., & Murnighan, J. K. (2003). Attributions of trust and the calculus of reciprocity. Journal of Experimental Social Psychology, 39(5), 448–455. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​0​022-​​1031(​03)00​015-5​ Podolny, J. M. (2001). Networks as the pipes and prisms of the market. American Journal of Sociology, 107(1), 33–60. https://doi​.org​/10​.1086​/323038 Rank, O. N., Robins, G. L., & Pattison, P. E. (2010). Structural logic of intraorganizational networks. Organization Science, 21(3), 745–764. https://doi​.org​/10​.1287​/orsc​.1090​.0450 Reagans, R., Zuckerman, E., & McEvily, B. (2004). How to make the team: Social networks vs. demography as criteria for designing effective teams. Administrative Science Quarterly, 49(1), 101–133. https://doi​.org​/10​.2307​/4131457 Reeder, G. D., & Brewer, M. B. (1979). A schematic model of dispositional attribution in interpersonal perception. Psychological Review, 86(1), 61–79. https​:/​/do​​i​.org​​/10​.1​​037​/0​​ 033​-2​​95​X​.8​​6​.1​.6​1 Rice, R. E., & Aydin, C. (1991). Attitudes toward new organizational technology: Network proximity as a mechanism for social information processing. Administrative Science Quarterly, 36(2), 219–244. https://doi​.org​/10​.2307​/2393354 Rivera, M. T., Soderstrom, S. B., & Uzzi, B. (2010). Dynamics of dyads in social networks: Assortative, relational, and proximity mechanisms. Annual Review of Sociology, 36, 91– 115. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​.soc​​.34​.0​​​40507​​.1347​​43

The Tangled Ties of Trust  231

Rossman, G., Esparza, N., & Bonacich, P. (2010). I’d like to thank the academy, team spillovers, and network centrality. American Sociological Review, 75(1), 31–51. https://doi​. org​/10​.1177​/0003122409359164 Rotter, J. B. (1971). Generalized expectancies for interpersonal trust. American Psychologist, 26(5), 443–452. https://doi​.org​/10​.1037​/h0031464 Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/AMR​.1998​.926617 Shah, N. P., Parker, A., & Waldstrøm, C. (2017). Examining the overlap: Individual performance benefits of multiplex relationships. Management Communication Quarterly, 31(1), 5–38. https://doi​.org​/10​.1177​/0893318916647528 Shah, P. P. (1998). Who are employees' social referents? Using a network perspective to determine referent others. Academy of Management Journal, 41(3), 249–268. https:// doi​.org​/10​.5465​/256906 Shah, P. P. (2000). Network destruction:The structural implications of downsizing. Academy of Management Journal, 43(1), 101–112. https://doi​.org​/10​.5465​/1556389 Shah, P. P., Dirks, K.T., & Chervany, N. L. (2006).The multiple pathways of high performing groups:The interaction of social networks and group processes. Journal of Organizational Behavior, 27(3), 299–317. https://doi​.org​/10​.1002​/job​.368 Shazi, R., Gillespie, N., & Steen, J. (2015). Trust as a predictor of innovation network ties in project teams. International Journal of Project Management, 33(1), 81–91. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.ijpr​​oman.​​20​14.​​06​.00​1 Simmel, G. (1950). The sociology of Georg Simmel. New York, NY: Free Press. Sniezek, J. A., & Van Swol, L. M. (2001). Trust, confidence, and expertise in a judge-advisor system. Organizational Behavior and Human Decision Processes, 84(2), 288–307. https:// doi​.org​/10​.1006​/obhd​.2000​.2926 Soda, G., Usai, A., & Zaheer, A. (2004). Network memory:The influence of past and current networks on performance. Academy of Management Journal, 47(6), 893–906. https:// doi​.org​/10​.5465​/20159629 Tasselli, S., & Kilduff, M. (2018). When brokerage between friendship cliques endangers trust: A personality–network fit perspective. Academy of Management Journal, 61(3), 802– 825. https://doi​.org​/10​.5465​/amj​.2015​.0856 Thomas, L. 1987. Friendship. Synthese, 72(2), 217–236. https://www​.jstor​.org​/ stable​/20116424 Walter, J., Levin, D. Z., & Murnighan, J. K. (2015). Reconnection choices: Selecting the most valuable (vs. most preferred) dormant ties. Organization Science, 26(5), 1447–1465. https://doi​.org​/10​.1287​/orsc​.2015​.0996 Wang, X., Liu, Y., Zhang, G., Xiong, F., & Lu, J. (2017). Diffusion-based recommendation with trust relations on tripartite graphs. Journal of Statistical Mechanics: Theory and Experiment, 2017(8), 083405. Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications (Vol. 8). Cambridge: Cambridge University Press. Weber, J. M., Malhotra, D., & Murnighan, J. K. (2004). Normal acts of irrational trust: Motivated attributions and the trust development process. Research in Organizational Behavior, 26, 75–101. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​191-​​3085(​04)26​003-8​ Wong, S., & Boh, W. F. (2010). Leveraging the ties of others to build a reputation for trustworthiness among peers. Academy of Management Journal, 53(1), 129–148. https​:/​/ do​​i​.org​​/10​.5​​465​/a​​mj​.20​​10​​.48​​03726​5

232  Stephen Jones and Priti Shah

Wu, J., Balliet, D., & Van Lange, P. A. (2016). Reputation, gossip, and human cooperation. Social and Personality Psychology Compass, 10(6), 350–364. https://doi​.org​/10​.1111​/spc3​. 12255 Zaheer, A., & Soda, G. (2009). Network evolution: The origins of structural holes. Administrative Science Quarterly, 54(1), 1–31. https://doi​.org​/10​.2189​/asqu​.2009​. 54​.1.1

10 MULTILEVEL THEORIZING OF HOW GENDER INFLUENCES TRUST AND CREATIVITY1 Hye Jung Eun, Roy Chua, and Mengzi Jin

Introduction Creativity is the generation of ideas and products that are both novel and useful (Amabile, 1983). Creativity is prized in a business setting – yet there is abundant evidence showing that the gender gap in career advancement is more severe in fields where creativity and innovation are particularly important (Joshi, Son, & Roh, 2015). In other words, in fields where creativity and innovation are salient, women appear to be less valued (Adams, Kräussl, Navone, & Verwijmeren, 2017; Jensen, Kovács, & Sorenson, 2018). For example, research (Adams et al., 2017) found that artworks by women were sold for less because of the creator’s gender. Compared to their male counterparts, female inventors are also more likely to get their patent applications rejected (Jensen et al., 2018) and receive less funding (Kanze, Huang, Conley, & Higgins, 2018). In addition, while 40% of beginning architects are women, women receive only 18% of architecture’s highest honors (Chang, 2014). All of these studies reinforce the pervasive concern that women lag behind in career advancement in fields where success depends on creative performance. We believe that investigating gender differences in trust may help us better understand these stark gender differences in creativity-related outcomes and that this research can also provide insights into understanding the biases that women face in creative endeavors at all levels of work (e.g., gender difference in patent acceptance, gender effect on selection for creative work). Given the increasing importance of trust and creativity in today’s businesses, researchers have investigated the relationship between trust and creativity. Prior

1 Acknowledgment: this research was conducted with the support of the Singapore Ministry of Education’s Social Science Research Thematic Grant (MOE2017-SSRTG-042).

DOI: 10.4324/9780429449185-10

234  Hye Jung Eun, Roy Chua, and Mengzi Jin

studies have uncovered the integral main effect of trust on creativity (e.g., Shazi, Gillespie, & Steen, 2015) as well as the mediating (e.g., Boies, Fiset, & Gill, 2015) and the moderating effects (e.g., Zhang & Zhou, 2014) of trust on creativity. For example, Shazi and colleagues (2015) found that perceived trustworthiness predicts the formation of ties for creative collaboration. Similarly, trust in teammates explains the positive effect of the inspirational motivation of leaders on team creative performance (Boies et al., 2015). Zhang and Zhou (2014) found that the relationship between empowering leadership and employee creativity is the strongest when both trust in leadership and uncertainty avoidance are high. More relevant to our discussion is a body of studies that has more specifically examined the differing effects of affect-based trust and cognition-based trust. Although earlier research also identified affective and cognitive elements of trust (Butler, 1991; Lewis & Wiegert, 1985), McAllister’s (1995) framework established a more nuanced understanding of the two elements of trust, allowing further research to explore the elements more independently. Affect-based trust is known as “trust from heart, a bond that arises from one’s own emotions and sense of the other’s feelings and motives,” whereas cognition-based trust refers to trust “from the head,” which is largely informed by evidence of the other’s competence and reliability (Chua, Ingram, & Morris, 2008, p. 2; Cook & Wall, 1980). Researchers have revealed that these two types of trust have differing effects. For example, Chowdhury (2005) found that affect-based trust promotes social ties whereas cognition-based trust fosters professional collaboration, both of which are important contributors to complex knowledge-sharing. Another line of research, albeit limited, has been devoted to understanding the effects of affect-based trust and cognition-based trust on creativity (e.g., Barczak, Lassk, & Mulki, 2010; Chua et al., 2008; Chua, Morris, & Mor, 2012). Empirical evidence has demonstrated distinctive effects of the two types of trust on creativity. Barczak and colleagues (2010) found that cognition-based trust in teams, but not affect-based trust, is positively associated with collaborative creativity. Chua, Morris, and Ingram (2010) found that affect-based trust is particularly associated with new idea-sharing. Another study (Cheung, Gong, Wang, Zhou, & Shi, 2016) similarly revealed that affect-based trust, but not cognition-based trust, buffers team members’ concerns about new idea-sharing. This study also found that affect-based trust moderates the relationship between functional diversity of teams and knowledge sharing in teams, which provides a foundation for our exploration of how affect-based trust influences new idea-sharing. Building on prior research on status and perceived creativity (Lau & Li, 1996) and creativity as a form of competence (e.g., Slabbert, 1994), we argue that cognition-based trust would have a main, but not sole, effect on perceived creativity. That is, the more one garners cognition-based trust from others, the more likely one is to be perceived as creative. These arguments provide an explanation as to why men are more readily associated with creative attributes (i.e., perceived creativity) (Elmore & Luna-Lucero, 2017; Proudfoot, Kay, & Kovel, 2015). This

Trust

Individual level

P1-P4

Creativity-related Outcomes

Affect-based trust

Gender differences in trust propensity and trustworthiness

Cognition-based trust

Individual level outcome Perceived creativity

Gender of dyad Dyadic level Affect-based trust

Gender similarity

P5-P7

Cognition-based trust

Gender of trustee

Dyadic level outcome Perceived creativity New idea sharing Collaborative creativity

Social Network level

P8-P12

Gender differences in Social network density Social network gender diversity

Affect-based trust

Gender difference in Social network centrality

Individual level outcome Perceived creativity Individual creativity

Cognition-based trust

Psychological safety Group level Gender diversity in group

P13-P14 Gender diversity in group

FIGURE 10.1 Overview

Affect-based trust

New idea sharing in group

Cognition-based trust

Group creativity

of hypothesized relationships across multiple levels.

236  Hye Jung Eun, Roy Chua, and Mengzi Jin

approach to understanding perceived creativity is insightful because the gender bias in perceived creativity may explain the gender gap in creative achievements, given that studies consistently showed no gender difference in actual creative performance (Amabile, 1983; Harris, 2004; Saeki, Fan, & Van Dusen, 2001). Overall, it appears that affect-based trust is associated with new idea-sharing whereas cognition-based trust is linked to whether or not one might be perceived as creative. Using the affect-based trust versus cognition-based trust framework, we integrate prior studies on trust, gender, and creativity to better understand how men and women are different in terms of trust. This allows us to better understand various creativity-related outcomes (e.g., perceived creativity, ideasharing). In doing our analysis, we develop propositions on gender, trust, and creativity. The contribution of the chapter is largely threefold. First, we provide a systematic understanding of how men and women differ in trust at individual, dyadic, social network, and group levels. Second, our discussion develops a set of testable propositions on how gender influences affect-based trust and cognition-based trust, which in turn leads to different creativity-related outcomes for men versus women. Lastly, we provide a future research agenda to further advance the scientific knowledge of trust, gender, and creativity at multiple levels. Contributing to the growing line of research taking a multilevel approach to trust (e.g., Costa, Fulmer, & Anderson, 2018; Fulmer & Dirks, 2018; Fulmer & Gelfand, 2012), we discuss gender differences in trust at different levels to explore how men and women differ in aspects of trust. We employ McAllister’s (1995) affect-based trust and cognition-based trust distinction because prior studies on gender provide extensive discussion on how men are perceived to be competent while women are seen as warm, which largely aligns with McAllister’s (1995) theoretical framework. Our discussion of these relationships is organized by levels of analysis – individual, dyadic, social network, and group levels. We also consider the effects of gender similarity at the dyadic level and gender diversity at the social network and group levels. Please refer to Figure 10.1 for the overview of the propositions across multiple levels.

Individual Level Prior research across various disciplines has shown gender differences in general trust propensity. This finding has been especially well documented in economics literature (Alesina & La Ferrara, 2002; Buchan, Croson, & Solnick, 2008; Chaudhuri & Gangadharan, 2003; Shaub, 1996; Spector & Jones, 2004). The consistent finding is that men are more likely to trust others. Researchers have paid attention to the gender differences in trust propensity because it has important implications for economic behaviors (Buchan et al., 2008); moreover, researchers have identified specific reasons underlying these gender differences. Using the General Social Survey (GSS), which consists of 1,500 responses from 1974 to 1994, Alesina and La Ferrara (2002) found that men are more likely to think

Gender Influences on Trust and Creativity  237

that “most people can be trusted”; the authors argue that those who have been historically discriminated against (i.e., women or black people) tend to be less likely to trust others. Using an investment game with 100 participants, Chaudhuri and Gangadharan (2003) showed that men exhibited greater trust in others than women, a finding that the researchers said could be attributed to the fact that women are more risk-averse than men. Other scholars explained that women trust others less because women are more sensitive to exploitative behaviors (Wang, Huang, Yin, & Ke, 2018). Although these studies do not specify types of trust, these findings provide reasons for us to argue that men tend to have higher levels of both affect-based trust and cognition-based trust in others compared to women because men occupy less vulnerable positions in society. Thus, we propose the following: Proposition 1: Compared to women, men in general have higher affect-based trust and higher cognition-based trust in others. To better understand gender differences in affect-based trust as well as cognitionbased trust, we direct our attention to the two well-established theories – the stereotype content model and the social role theory. The stereotype content model (SCM) proposes that stereotypes have two dimensions – warmth and competence (Fiske, Cuddy, Glick, & Xu, 2002; Fiske, Xu, Cuddy, & Glick, 1999) and that individuals use these two dimensions to “sort out” social worlds (Cuddy, Fiske, & Glick, 2008; Fiske, Cuddy, & Glick, 2007). Scholars found that when people interact with others, they are mostly interested in finding out how warm and how competent others tend to be (Fiske et al., 2002; Wojciszke, 2005). Social cognition researchers found that compared to men, women are more likely to be seen as warm and sociable as well as less competent (Fiske & Cuddy, 2006), which aligns with the social role theory of gender delineated below. A study also showed that judgments of a person’s trustworthiness correspond with judgments of that person’s warmth, especially for female subjects (Sutherland, Oldmeadow, & Young, 2016). The social role theory (Eagly, 1987), which stems from the observation that men and women have historically occupied different roles in society, also merits discussion. This theory helps to explain the gender stereotype that men are agentic (e.g., assertive, independent, competent) while women are communal (e.g., interpersonally concerned, emotionally expressive) (Wood & Eagly, 2012). In other words, while men have been more often assigned to tasks that require strength and competence, women have been more often seen as responsible for home- and family-related tasks (Harrison & Lynch, 2005). Perceivers’ beliefs and expectations about social groups are shaped by the roles those groups typically occupy (Eagly, 1987; Wood & Eagly, 2012). People expect stereotypical behaviors from each gender and therefore expect individuals to fill social roles that comport with that stereotype (Wood & Eagly, 2012).

238  Hye Jung Eun, Roy Chua, and Mengzi Jin

Eagly and Steffen (1984) explained that this social role theory is almost exclusively important in explaining gender stereotypes. Men are stereotypically seen as more competent as well as less warm, but ready to fulfill agentic social roles. Meanwhile, women are expected to be more interpersonally approachable and ready to take on communal roles. Integrating these perspectives with McAllister’s (1995) theory of affect-based trust and cognition-based trust, we argue that individuals in general have higher affect-based trust in women as well as higher cognition-based trust in men.The notion that individuals may have higher affect-based trust in women is also supported by the fact that women are stereotypically seen as more willing to help others (Ryan, Haslam, & Postmes, 2007). Given that a major antecedent of affect-based trust is the level of citizenship behaviors toward others (McAllister, 1995), it is reasonable to expect that individuals have higher affect-based trust in women. Thus, we propose that: Proposition 2: Individuals in general have higher affect-based trust in women than in men. Proposition 3: Individuals in general have higher cognition-based trust in men than in women. The gender differences outlined above prepare us for the subsequent discussion of gender differences in perceived creativity. Prior research has emphasized the importance of affect-based trust on promoting the exchange of new ideas (Chua et al., 2012) because sharing new ideas entails the risk of being ridiculed or receiving negative feedback (Nemeth, Personnaz, Personnaz, & Goncalo, 2004). Thus, since men tend to have more affect-based trust in others than women do, men are more likely to share their novel ideas with others.The more one shares novel ideas, the more likely one is to be perceived as creative. In addition, given that men are seen as more competent, their novel ideas would be more favorably received.Thus, the differential effects of gender on affect-based trust and cognition-based trust may explain why men are seen as more creative (Elmore & Luna-Lucero, 2017; Proudfoot, Kay, & Kovel, 2015). The arguments presented above regarding gender difference in perceived creativity is valuable because creativity evaluation usually results in greater workplace recognition and opportunities (Gupta & Singhal, 1993; Wachter & Estlund, 2012).To the extent that compared to women, men are seen as more creative because men have higher affect-based trust in others and are more willing to share their creative ideas, the discussion above advances a better understanding of why people stereotypically associate creativity with men more readily. Proposition 4: Men are typically perceived as more creative than women because (1) men are more likely to share their creative ideas due to their tendency to have higher affect-based trust in others and (2) individuals generally have higher cognition-based trust in men.

Gender Influences on Trust and Creativity  239

Dyadic Level As a majority of research has focused on trust as one’s willingness to be vulnerable to the other (Krasikova & LeBreton, 2012), scholars have acknowledged that research has “tended to examine trust as a static phenomenon and from one party’s (i.e., the trustor’s) perspective” (Fulmer & Gelfand, 2012, p. 1211). Trust at the dyadic level is bidirectional, in that it recognizes the role of both trustor and trustee (Korsgaard, Brower, & Lester, 2015). Trust at the dyadic-level perspective posits that trust is a function of properties and characteristic of the self (I), the specific partner with whom one is interacting (you), and the unique features, requirements, or constraints of the current situation (to do X) (Simpson, 2007). Simpson (2007) further noted that according to this perspective, change of one component would result in further changes in a dyad. Extending the previous discussion on gender difference in perceived creativity, this section examines whether and how women may be, in general, considered to be less effective partners in creative collaboration, as well as whether gender similarity could better facilitate idea-sharing. Previous research points to this sort of analysis. Mayer, Davis, and Schoorman (1995) proposed an integrative model that identifies trustworthiness (i.e., the ability, benevolence, and integrity of a trustee) and trust propensity (i.e., a trustor’s dispositional willingness to be vulnerable to others) as antecedents of trust. Other researchers (Jiang, Chua, Kotabe, & Murray, 2011; Mayer et al., 1995; Whitener, Brodt, Korsgaard, & Werner, 1998) emphasized that characteristics of trustees matter because they shape trustworthiness, and that interaction between trustor and trustee is important (Simpson, 2007).Taking this research one step further, the current section focuses on whether and how the gender of the trustee as well as gender similarity influence trust in dyads. Prior research has also specifically looked at the gender of the trustee as an important aspect (Spector & Jones, 2004). Bevelander and Page (2011) found that although women tend to socialize with other women in general, when it comes to risky tasks, women prefer to work with men, as men are perceived to be more competent and better able to provide needed resources. Similarly, studies on gender differences in social networks revealed that women prefer male contacts for instrumental purposes (Ibarra, 1992). Such patterns align with studies showing that compared with women, men are perceived to be more task-oriented and capable (Bales & Slater, 1955; Meeker & Weitzel-O’Neill, 1977; Strodtbeck & Mann, 1956). Keeping this research in mind, we expect that when the trustee is a man, cognition-based trust increases in the dyadic relationship, rendering the trustee to be perceived as an effective creative collaboration partner. Conversely, when the trustee is a woman, a higher threshold level of cognition-based trust in the trustee might be needed to facilitate creative idea-sharing in a dyad. Proposition 5: Regardless of the trustor’s gender, when the trustee is a man, cognitionbased trust increases in a dyadic relationship, which renders the trustee to be perceived as a better collaborative creativity partner.

240  Hye Jung Eun, Roy Chua, and Mengzi Jin

Proposition 6: Regardless of the trustor’s gender, when the trustee is a woman, a higher threshold level of cognition-based trust in the trustee from the trustor is needed to facilitate idea-sharing from the trustor with the trustee in a dyad for better collaborative creativity. Research we conducted provides preliminary evidence to support these propositions. We administered a network survey where participants shared information about their professional network. A total of 26 working professionals at an educational institution in Singapore shared information about their interactions. Among them, 23% were female and the average age was 40.34 (S.D. = 10.10). Among the sample, 65% identified themselves as local and 72% reported a master’s degree as the highest level of education they obtained.We were particularly interested in their gender and the gender information of alters2 in the dyadic relationship, as well as how effective group members thought each alter was as a creative collaborator. Each participant shared information on his or her interaction with each employee on a list of 37 employees. The data was restructured so that each observation is at the dyadic level. The data consists of 201 dyadic observations. We ran regression analyses and the results showed marginally significant results to support that regardless of the gender of a trustor, individuals tend to think that men alters are better creative collaboration partners (b = 0.22, p = 0.09). Also, the data showed that regardless of the gender of a trustor, a higher level of cognitionbased trust is needed for women trustees to be perceived as effective creative collaborators (b = –0.36, p = 0.04). Figure 10.2 depicts the interactive effect. Thus, our data provides initial evidence supporting our argument that men trustees are in general seen as better creative collaborators, as well as that women trustees require a higher threshold level of cognition-based trust from a trustor to be seen as effective creative collaborators. Our analysis above suggests that compared to men, women appear to need to prove themselves more to be considered effective creative collaboration partners. In other words, having men partners instead of women partners in dyadic relationships generally facilitates new idea discussion and hence greater collaborative creativity. Although we argue that having men partners can be beneficial for facilitating collaborative creativity, we also believe that gender similarity can promote idea-sharing through higher affect-based trust and that women can benefit more from having the same gender in dyads. Evidence indicates that affect-based trust is higher in same-gender dyads (e.g., Chua, Morris, & Ingram, 2009; McAllister, 1995), promoting psychological safety and facilitating collaborative creativity. This is because interacting with someone similar involves predictable behaviors and easier communication between actors (Zhou, Shin, Brass, Choi, & Zhang, 2009), prompting higher affect-based trust.

2 Alters refer to actors to whom a focal actor is directly tied in the social network.

Gender Influences on Trust and Creativity  241

FIGURE 10.2  Interactive

effect of gender of alter and cognition-based trust on perceived effectiveness as creative collaborator.

Spector and Jones (2004) also found that men’s initial trust is higher for a male newcomer compared to a female newcomer. Thus, while the gender of the trustee has a direct effect on cognition-based trust as illustrated above, gender similarity influences affect-based trust in a dyad. Although gender similarity overall can influence dyadic relationships, we expect that it has a stronger impact on women as they are more interpersonally oriented (Eagly & Johannesen-Schmidt, 2001). For instance, Chua and Jin (2020) found that relationship conflict hurts collaborative creativity in women–women intercultural collaboration more than in men–men intercultural collaboration. Although these authors did not explicitly discuss the role of trust, it is likely that women–women collaborative creativity suffers because relationship conflict reduces their affect-based trust in each other. Therefore, women–women dyads can benefit more from higher affect-based trust compared to men–men dyads, which can facilitate sharing of new ideas and collaborative creativity. Proposition 7: Gender of dyadic collaboration moderates the relationship between gender similarity and affect-based trust, such that the positive effect of gender similarity on affect-based trust is more salient in women–women collaboration compared with men–men collaboration, leading to greater idea-sharing.

242  Hye Jung Eun, Roy Chua, and Mengzi Jin

Social Network Level We next explore how gender differences within social networks lead to differences in affect-based trust and cognition-based trust, which in turn result in systematic differences in perceived creativity and individual creativity. Social capital is defined as “the benefits individuals derive from their social relationships and interactions” such as “emotional support, exposure to diverse ideas, and access to non-redundant information” (Ellison, Lampe, & Steinfield, 2010, p. 1). Social capital is differentially distributed across different social groups – for example, across gender groups (Lin, 2001). In this section, we examine three main social network properties that shape social capital – network density, network centrality, and network diversity (Borgatti, Jones, & Everett, 1998) – to explain how systematic differences in social network properties translate into differences in affect-based trust and cognition-based trust. This, in turn, influences creativity-related outcomes.

Network Centrality Social network centrality reflects how central one is in his or her network. A central actor has ties to many people, who are also tied to many others (Felmlee & Faris, 2013). Men typically hold more central positions than women because men tend to have more homophilous ties with other men, who also tend to have higher-status positions and connections to high-status others. This further increases network centrality (Ibarra, 1992). In parallel with our earlier discussion, men in social networks tend to have higher perceived status than women because they are more likely to be seen as competent (Eagly, 1987). A study (Wasko & Faraj, 2005) also showed that those with a high level of network centrality tend to contribute more knowledge, which can allow the actors to be seen as competent and higher-status. Building on status and creativity research (Lau & Li, 1996), we expect that alters in social networks have higher cognition-based trust in men than women because men tend to hold more central positions, which in turn renders them to be perceived as more creative. In sum, we theorize that men tend to have higher network centrality than women, which prompts alters to have higher cognition-based trust in men, making men be seen as more creative. It is worth noting some caveats when discussing gender, social network centrality, and creativity. Prior studies have found that women who engaged in roles that are not stereotypically associated with feminine attributes could be penalized (Rudman & Glick, 2001).Thus, if women were to take central positions in a social network, it may not make them be seen as creative in the same manner as it does men. Proposition 8: Alters in social networks have higher cognition-based trust in men than women because men tend to hold more central positions in those networks. Proposition 9: Alters perceive men to be more creative than women because men are connected to more high-status others; men are therefore perceived to be more competent (cognition-based trust).

Gender Influences on Trust and Creativity  243

Network Density Social network density refers to how extensively ties are interconnected (Dubini & Aldrich, 1991).The more connected individuals are in one’s network, the denser one’s network is. Research shows that women tend to have more intimate, smaller circles of contacts than men because women’s social networks are composed of close others, whereas contacts in men’s social networks tend to come from a wider variety of sources (Smith, 2000). Past research has indicated that dense networks promote individual and firm performance because they foster trust as well as the exchange of information (e.g., Ingram & Roberts, 2000; Uzzi, 1996). However, researchers more specifically found that social network density has a main effect on affect-based trust (Chua et al., 2008; Chua & Morris, 2006; Chua et al., 2009). Social network density would have implications for affect-based trust but not necessarily for cognition-based trust because a denser network would have a positive effect on socio-emotional support (House, Umberson, & Landis, 1988). In addition, a denser social network increases positive affect (Totterdell, Wall, Holman, Diamond, & Epitropaki, 2004), facilitating affect-based trust among actors. Our earlier discussion explained that affect-based trust can facilitate the exchange of ideas. In sum, women tend to have denser social networks, which have higher affect-based trust to facilitate idea-sharing. Proposition 10: Compared with men, women tend to have social networks that are higher on affect-based trust because their social networks are more likely to be high in network density. Proposition 11: Compared with men, women’s creativity can benefit more from dense networks because women’s denser networks and corresponding high affect-based trust allow for more idea-sharing.

Network Diversity A stream of research has shown that social network diversity is overall beneficial for creativity (e.g., Rodan & Galunic, 2004; Sosa, 2011). There are both direct and indirect evidence that men tend to have more diverse social networks than women. For example, Mehra, Kilduff, and Brass (1998) found that compared to men, women are more likely to build friendships within groups with other women. Smith (2000) also found that men are more likely to benefit from mobilizing their personal contacts for career advancement; she argued that this is because men’s personal contacts span a larger occupational range, thereby making them more diverse in terms of occupational characteristics. Finally, Chua (2018) found a significant gender difference in cultural diversity in social networks, in that men reported more culturally dissimilar ties (r = 0.05, p < 0.05). Having diverse contacts in terms of occupational characteristics would signal one’s competence within the network (Chua, 2018).

244  Hye Jung Eun, Roy Chua, and Mengzi Jin

Thus, men’s diverse social networks would promote alters’ cognition-based trust for men, making them be seen as creative. Proposition 12: Compared to women’s social networks, men’s social networks are more diverse, which fosters cognition-based trust, making men be seen as more creative than women by alters of both genders.

Group Level We have witnessed an exponential growth of diversity research (Giambatista & Bhappu, 2010), yet prior studies have not fully investigated the relationship between gender diversity in groups, trust, and creativity-related outcomes at the group level. As women increasingly collaborate in creative activities (Sonnenburg, 2004), the effect of gender diversity in groups on creativity is worth discussing. Research on the effect of gender diversity on relational conflict and task conflict, which might inform the discussion on the effect of gender diversity on affect-based trust and cognition-based trust, has shown inconsistent results. While some studies have shown that gender diversity increases group conflict such as relational conflict or task conflict (e.g., Jehn, Northcraft, & Neale, 1999), others found no effect (e.g., Pelled, Eisenhardt, & Xin, 1999). To reconcile these findings, one must take into account the complex nature of diversity. It is important to note that diversity researchers describe having diversity in a group context as “a double-edged sword” (Milliken & Martins, 1996), one that results in both benefits and difficulties. A line of diversity research with “the pessimistic view” relies on the similarity-attraction hypothesis (Byrne, 1971), social identity (Tajfel, 1978), and selfcategorization theory (Turner, 1982, 1985) to discuss challenges that result from diversity in groups. According to these theories, gender diversity in groups would make one’s gender identity salient and create an “us-versus-them” distinction. These processes would foster interaction within one’s own gender group, which inevitably decreases affect-based trust at the group level. Gender diversity could also create more task conflict because, in gender-diverse groups, members are less likely to develop effective communication (Curşeu & Schruijer, 2010), which reduces cognition-based trust at the group level. Thus, one might expect that group-level gender diversity would prevent the development of affect-based trust and cognition-based trust among group members. The other, more optimistic, perspective is also worth noting. It draws on the information-processing approach, which highlights that diversity in groups allows members to have access to others with various backgrounds and resources. This perspective largely explains why diversity in groups may lead to greater creative outcomes (Mannix & Neale, 2005). This approach also builds on the “value-indiversity” hypothesis (Cox, Lobel, & McLeod, 1991); specifically, varied inputs such as perspectives and knowledge from gender-diverse group members might create positive environments for constructive conflict where diverse opinions are valued and complex problems are solved (Mannix & Neale, 2005).These processes

Gender Influences on Trust and Creativity  245

would increase affect-based trust and cognition-based trust among group members. Thus, according to this perspective, gender diversity within groups may foster affect-based trust and cognition-based trust. Given the complex nature of diversity (Giambatista & Bhappu, 2010), instead of theorizing its main effects, we explore the moderating effect of psychological safety on the relationship between gender diversity and the two types of trust. Psychological safety is an important moderator when disentangling the effects of diversity on group-level outcomes (Edmondson & Lei, 2014). We believe that psychological safety, which is defined as a shared belief about the extent to which the group is safe to take interpersonal risks (Edmondson, 1999), would moderate the effect of gender diversity on affect-based trust and cognition-based trust; this, in turn, would lead to idea-sharing and group creative outcomes respectively. Specifically, a psychologically safe environment would help foster engaged interactions and interdependence among group members (Caruso & Woolley, 2008) in gender diverse groups, which increases affect-based trust. On the contrary, in gender-diverse groups where psychological safety is low, gender diversity would lead to greater ‘us–them’ distinctions and reduce affect-based trust at the group level. Building on research by Bradley and colleagues (2012) that argued that psychological safety encourages constructive disagreements among group members, we also argue that psychological safety can moderate the effect of gender diversity on cognition-based trust. Specifically, gender-diverse groups with high psychological safety might increase cognition-based trust among members because psychological safety promotes exchanging resources while encouraging constructive debates. However, gender-diverse groups with low psychological safety would likely reduce cognition-based trust among group members. As proposed above, we argue that affect-based trust has a main effect on idea-sharing.We also argue that increased cognition-based trust at the group level can facilitate and therefore promote group creative activities.Thus, we propose the following: Proposition 13: Psychological safety moderates the indirect effect of gender diversity in groups on idea-sharing via affect-based trust such that gender diversity in groups increases affect-based trust, which leads to greater idea-sharing when psychological safety is high. Proposition 14: Psychological safety moderates the indirect effect of gender diversity on group creativity via cognition-based trust, such that gender diversity in groups increases cognition-based trust, which leads to greater group creativity when psychological safety is high.

Discussion The current chapter has identified two key research gaps – the gender perspective on trust and the lack of a multilevel approach to trust analyses – and aimed to build a greater understanding of gender, trust, and creativity at the individual,

246  Hye Jung Eun, Roy Chua, and Mengzi Jin

dyadic, social network, and group levels. Our approach particularly advances the understanding of the multilevel nature of trust processes in organizations by demonstrating that knowledge at one level can be utilized to develop more contextrelevant knowledge at high levels. For example, the propositions at the individual level regarding gender differences in trust propensity informed more contextual exploration of trust and interaction at the dyadic level. As we theorize from the individual level to higher levels, our discussion of the gender effect on trust and creativity provides meaningful insights to better understand the social problem of the gender gaps in creativity-related outcomes. Thus, our work demonstrates the importance of taking cross-level and multilevel approaches in order to appreciate the complex nature of trust in organizations.

Future Research Directions Despite the voluminous research done to date regarding both trust and gender, our knowledge of how gender influences trust is still at a nascent stage. The current chapter has developed a set of propositions to explore how gender influences trust at multiple levels of analysis, which in turn may help explain the glass ceiling effect as well as other gender biases, specifically in creativity fields (Adams et al., 2017; Jensen et al., 2018; Joshi et al., 2015). Research to date does not fully explain why women make slower progress in career advancements in creative fields. Although Proudfoot and colleagues (2015) show initial evidence to argue that gender bias in perceived creativity may contribute to inequality, there is much more to be explored to better understand the phenomenon. Extending the research on gender stereotypes and creativity, our discussion has shown the possibility that the two types of trust might be important factors to examine to understand gender inequality in creativity-related achievements. We believe that our discussion of gender effect on trust and creativity can also provide insights on the broad discussion of bias, discrimination, and equality – as well as personnel selection and promotion and organizational performance – not just at the higher level but at all levels. Several new research directions emerge from our discussion on gender, trust, and creativity. First, future studies can test the proposed ideas empirically. The chapter has provided a set of interesting propositions that need more thorough investigation. For example, the discussion at the individual level provides a novel way of understanding gender differences in perceived creativity.We discussed the possibility that men may share their novel ideas with others more frequently due to their tendency to have higher affect-based trust for others and receive higher cognition-based trust from others, which makes them be seen as more creative. Future studies can examine this idea empirically to provide concrete evidence. It is also worth employing a neurobiological approach to examine the gender perspective on trust. Researchers have taken a neurobiological approach to understand trust (e.g., Aimone, Houser, & Weber, 2014; Baumgartner,

Gender Influences on Trust and Creativity  247

Heinrichs, Vonlanthen, Fischbacher, & Fehr, 2008) and processes involved in creative tasks (e.g., Fink et al., 2010). Future research can take a neurobiological approach to better investigate our suggested propositions, such as whether people develop affect-based trust toward women differently than they do toward men (proposition 2). Our discussion has not included organizational-level or cultural-level analyses. More studies are warranted to understand gender, trust, and creativity in broader contexts. For example, if an organization is largely composed of male employees, how would this lead to differences in trust and creativity dynamics? Such extended investigation may uncover boundary conditions for the proposed relationships. In a similar vein, future studies can extend our discussion to innovation. Creativity and innovation are highly interconnected, but they are also distinctive (Hammond & Farr, 2011). While creativity specifically refers to the idea-generation process, innovation encompasses broader processes including idea generation, idea selection, and implementation (Unsworth, 2001). Along these lines, one may investigate whether and how high affect-based trust bestowed on women and cognition-based trust bestowed on men influence other innovation processes such as idea selection and implementation. A study (Foss, Woll, & Moilanen, 2013) finds that women’s ideas are less likely to be implemented due to differences in the perceived value of ideas. A discussion of trust may generate alternative explanations to understand such a phenomenon. Our propositions focus on just one form of trust or the other (i.e., affect-based trust or cognition-based trust). However, we acknowledge the interplay between the two types of trust, as they will be present in some combinations. Relatedly, Shazi and colleagues (2015) found that the absence of affect-based trust makes cognition-based trust irrelevant when choosing a creative collaborator. Future research may take this perspective to extend our proposed ideas. For example, one may explore whether and how affect-based trust moderates the gender effect on idea-sharing via cognition-based trust in a dyad (proposition 6). Future work can also investigate the gender perspective on trust in cross-cultural or intercultural contexts. Past research has documented various cross-cultural differences in trust (e.g., Buchan, Croson, & Dawes, 2002) and investigated trust in intercultural relationships (Chua & Morris, 2009). Investigating how the discussion above may differ cross-culturally or in intercultural relationships can also generate interesting research questions. Chua and Jin (2020) recently found that the gender composition of dyads moderates the effect of intercultural conflicts on creativity; they posit that the level of information and ideas exchanged and integrated mediates the relationships.Yet the authors also acknowledge that these relationships may be explained by trust. Our study provides insights to address this potential alternative explanation. Also, one may investigate whether and how men and women differ in developing trust in intercultural relationships. Future studies can generate meaningful research regarding gender, trust, and culture as well as implications for creativity, innovation, and gender biases.

248  Hye Jung Eun, Roy Chua, and Mengzi Jin

Last but not least, future studies can also explore whether gender perspectives on trust have implications for other outcomes more broadly. The current chapter has focused on creativity-related outcomes, but future research endeavors can broaden this perspective to better understand whether and how the gender perspective on trust can contribute to explaining the more general gender biases in other performance contexts. For example, a recent body of studies found converging evidence to show that when in crisis, female leaders are preferred to men; this phenomenon is identified as “think crisis–think female” (e.g., Gartzia, Ryan, Balluerka, & Aritzeta, 2012; Ryan, Haslam, Hersby, & Bongiorno, 2011). Taking a gender perspective on trust, one could argue that during crises, female leaders are preferred because affect-based trust bestowed on leaders grows more important during difficult times when leaders are expected to be not only competent in solving problems but also able to rally employees’ emotions and motivation.

Conclusion In this chapter, we theorized how gender influences affect-based trust and cognition-based trust at the individual, dyadic, social network, and group levels.We also extended the gender perspective on trust to discuss creativity-related outcomes. Our discussion not only presents a systematic understanding of gender and trust but also presents novel accounts of what barriers and opportunities women face in creativity-related processes, as well as why they face those barriers and opportunities. We hope that our discussion demonstrates the importance of taking a multilevel approach to understanding trust and motivates further studies on it.

References Adams, R. B., Kräussl, R., Navone, M.A., &Verwijmeren, P. (2017, December 6). Is gender in the eye of the beholder? Identifying cultural attitudes with art auction prices. Identifying Cultural Attitudes with Art Auction Prices. http://dx​.doi​.org​/10​.2139​/ssrn​.3083500 Aimone, J. A., Houser, D., & Weber, B. (2014). Neural signatures of betrayal aversion: An fMRI study of trust. Proceedings of the Royal Society of London. Series B, Biological Sciences, 281(1782), 2013–2127. https://doi​.org​/10​.1098​/rspb​.2013​.2127 Alesina, A., & La Ferrara, E. (2002). Who trusts others? Journal of Public Economics, 85(2), 207–234. https​:/​/do​​i​.org​​/10​.1​​016​/s​​0​047-​​2727(​01)00​084-6​ Amabile,T. M. (1983).The social psychology of creativity:A componential conceptualization. Journal of Personality and Social Psychology, 45(2), 357–376. https​:/​/do​​i​.org​​/10​.1​​037​/ 0​​022​-3​​514​​.4​​5​.2​.3​​57 Bales, R. F., & Slater, P. E. (1955). Role differentiation in small decision-making groups. In T. Parsons & P. E. Slater (Eds.), The family socialization and interaction processes (pp. 259–306). Glencoe, IL: Free Press. https://doi​.org​/10​.2307​/2785640 Barczak, G., Lassk, F., & Mulki, J. (2010). Antecedents of team creativity: An examination of team emotional intelligence, team trust and collaborative culture. Creativity and

Gender Influences on Trust and Creativity  249

Innovation Management, 19(4), 332–345. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-8691​​.201​0​​. 0057​​4.x Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., & Fehr, E. (2008). Oxytocin shapes the neural circuitry of trust and trust adaptation in humans. Neuron, 58(4), 639–650. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.neur​​on​.20​​​08​.04​​.009 Bevelander, D., & Page, M. J. (2011). Ms. Trust: Gender, networks and trust – implications for management and education. Academy of Management Learning & Education, 10(4), 623–642. https://doi​.org​/10​.5465​/amle​.2009​.0138 Boies, K., Fiset, J., & Gill, H. (2015). Communication and trust are key: Unlocking the relationship between leadership and team performance and creativity. The Leadership Quarterly, 26(6), 1080–1094. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​15​.07​​.007 Borgatti, S. P., Jones, C., & Everett, M. G. (1998). Network measures of social capital. Connections, 21(2), 27–36. Bradley, B. H., Postlethwaite, B. E., Klotz, A. C., Hamdani, M. R., & Brown, K. G. (2012). Reaping the benefits of task conflict in teams: The critical role of team psychological safety climate. Journal of Applied Psychology, 97(1), 151–158. https://doi​.org​/10​.1037​/ a0024200 Buchan, N. R., Croson, R. T., & Dawes, R. M. (2002). Swift neighbors and persistent strangers: A cross-cultural investigation of trust and reciprocity in social exchange. American Journal of Sociology, 108(1), 168–206. https://doi​.org​/10​.1086​/344546 Buchan, N. R., Croson, R. T., & Solnick, S. (2008). Trust and gender: An examination of behavior and beliefs in the Investment Game. Journal of Economic Behavior & Organization, 68(3–4), 466–476. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jebo​​.200​7​​.10​.0​​06 Butler, J. K. (1991). Toward understanding and measuring conditions of trust: Evolution of a conditions of trust inventory. Journal of Management, 17(3), 643–663. https​:/​/do​​i​.org​​/ 10​.1​​177​/0​​14920​​6391​0​​17003​​07 Byrne, D. E. (1971). The attraction paradigm. New York: Academic Press. Caruso, H. M., & Woolley,A.W. (2008). Harnessing the power of emergent interdependence to promote diverse team collaboration. Diversity and Groups, 11, 245–266. https​:/​/ do​​i​.org​​/10​.1​​016​/S​​1​534-​​0856(​08)11​011-8​ Chang, L. C. (2014). Where are the women? Measuring progress on gender in architecture. Association of Collegiate Schools of Architecture (US). Retrieved from https​:/​/ww​​w​. acs​​a​-arc​​h​.org​​/reso​​urces​​/data​​-reso​​urces​​/wher​​e​-are​​-the-​​women​​-meas​​uring​​-prog​​ress-​​ on​-ge​​​nder-​​in​-ar​​chite​​cture​/ Chaudhuri, A., & Gangadharan, L. (2003). Gender differences in trust and reciprocity (Unpublished manuscript). University of Auckland, New Zealand. Cheung, S.Y., Gong,Y.,Wang, M., Zhou, L., & Shi, J. (2016).When and how does functional diversity influence team innovation? The mediating role of knowledge sharing and the moderation role of affect-based trust in a team. Human Relations, 69(7), 1507–1531. https://doi​.org​/10​.1177​/0018726715615684 Chowdhury, S. (2005). The role of affect-and cognition-based trust in complex knowledge sharing. Journal of Managerial Issues, 17(3), 310–326. Chua, R.Y. J. (2018). Innovating at cultural crossroads: How multicultural social networks promote idea flow and creativity. Journal of Management, 44(3), 1119–1146. https:// doi​.org​/10​.1177​/0149206315601183 Chua, R.Y. J., Ingram, P., & Morris, M. W. (2008). From the head and the heart: Locating cognition-and affect-based trust in managers’ professional networks. Academy of Management Journal, 51(3), 436–452. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​08​​.32​​62595​6

250  Hye Jung Eun, Roy Chua, and Mengzi Jin

Chua, R.Y. J., & Jin, M. (2020). Across the great divides: Gender dynamics influence how intercultural conflict helps or hurts creative collaboration. Academy of Management Journal, 63(3), 903–934. https://doi​.org​/10​.5465​/amj​.2016​.1319 Chua, R.Y. J., & Morris, M.W. (2006). Dynamics of trust in guanxi networks. In Y. R. Chen (Ed.), National culture and groups (pp. 95–113). Amsterdam: Elsevier. https​:/​/do​​i​.org​​/10​. 1​​016​/S​​1​534-​​0856(​06)09​005-0​ Chua, R. Y. J., & Morris, M.W (2009). Innovation communication in multicultural networks: Deficits in inter-cultural capability and affect-based trust as barriers to new idea sharing in intercultural relationships. Harvard Business School Organizational Behavior Unit Working Paper, (09–130). http://dx​.doi​.org​/10​.2139​/ssrn​.1403944 Chua, R. Y. J., Morris, M. W., & Ingram, P. (2009). Guanxi vs networking: Distinctive configurations of affect-and cognition-based trust in the networks of Chinese vs American managers. Journal of International Business Studies, 40(3), 490–508. https​:/​/do​​i​. org​​/10​.1​​057​/p​​algra​​ve​.ji​​​bs​.84​​00422​ Chua, R.Y. J., Morris, M. W., & Ingram, P. (2010). Embeddedness and new idea discussion in professional networks:The mediating role of affect-based trust. The Journal of Creative Behavior, 44(2), 85–104. https​:/​/do​​i​.org​​/10​.1​​002​/j​​.2162​​-6057​​.2010​​​.tb01​​327.x​ Chua, R. Y. J., Morris, M. W., & Mor, S. (2012). Collaborating across cultures: Cultural metacognition and affect-based trust in creative collaboration. Organizational Behavior and Human Decision Processes, 118(2), 116–131. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​1​​2​.03.​​009 Cook, J., & Wall,T. (1980). New work attitude measures of trust, organizational commitment and personal need non-fulfilment. Journal of Occupational Psychology, 53(1), 39–52. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.2044​​-8325​​.1980​​​.tb00​​005.x​ Costa, A. C., Fulmer, C. A., & Anderson, N. R. (2018). Trust in work teams: An integrative review, multilevel model, and future directions. Journal of Organizational Behavior, 39(2), 169–184. https://doi​.org​/10​.1002​/job​.2213 Cox,T. H., Lobel, S. A., & McLeod, P. L. (1991). Effects of ethnic group cultural differences on cooperative and competitive behavior on a group task. Academy of Management Journal, 34(4), 827–847. https://doi​.org​/10​.2307​/256391 Cuddy, A. J., Fiske, S.T., & Glick, P. (2008).Warmth and competence as universal dimensions of social perception: The stereotype content model and the BIAS map. Advances in Experimental Social Psychology, 40, 61–149. https​:/​/do​​i​.org​​/10​.1​​016​/s​​0​065-​​2601(​07) 00​002-0​ Curşeu, P. L., & Schruijer, S. G. (2010). Does conflict shatter trust or does trust obliterate conflict? Revisiting the relationships between team diversity, conflict, and trust. Group Dynamics:Theory, Research, and Practice, 14(1), 66–79. https://doi​.org​/10​.1037​/a0017104 Dubini, P., & Aldrich, H. (1991). Personal and extended networks are central to the entrepreneurial process. Journal of Business Venturing, 6(5), 305–313. https​:/​/do​​i​.org​​/10​ .1​​016​/0​​​883​-9​​026(9​1)900​21-5 Eagly, A. H. (1987). Reporting sex differences. American Psychologist, 42(7), 755–757. https​:/​/do​​i​.org​​/10​.1​​037​/0​​003​-0​​66X​​.4​​2​.7​.7​​55 Eagly, A. H., & Johannesen-Schmidt, M. C. (2001). The leadership styles of women and men. Journal of Social Issues, 57(4), 781–797. https://doi​.org​/10​.1111​/0022​-4537​.00241 Eagly, A. H., & Steffen,V. J. (1984). Gender stereotypes stem from the distribution of women and men into social roles. Journal of Personality and Social Psychology, 46(4), 735–754. https​:/​/do​​i​.org​​/10​.1​​037​/0​​022​-3​​514​​.4​​6​.4​.7​​35 Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. https://doi​.org​/10​.2307​/2666999

Gender Influences on Trust and Creativity  251

Edmondson, A. C., & Lei, Z. (2014). Psychological safety: The history, renaissance, and future of an interpersonal construct. Annual Review of Organizational Psychology and Organizational Behavior, 1(1), 23–43. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​-org​​psych​​03​14​​13​-09​​1305 Ellison, N. B., Lampe, C., & Steinfield, C. (2010). With a little help from my friends: How social network sites affect social capital processes. In Z. Papacharissi (Ed.), A networked self (pp. 132–153). Routledge. https://doi​.org​/10​.4324​/9780203876527​-13 Elmore, K. C., & Luna-Lucero, M. (2017). Light bulbs or seeds? How metaphors for ideas influence judgments about genius. Social Psychological and Personality Science, 8(2), 200– 208. https://doi​.org​/10​.1177​/1948550616667611 Felmlee, D., & Faris, R. (2013). Interaction in social networks. In J. DeLamater & A. Ward (Eds.), Handbook of social psychology (pp. 439–464). Dordrecht: Springer. https​:/​/do​​i​.org​​/ 10​.1​​007​/9​​78​-94​​-007-​​​6772-​​0​_15 Fink, A., Grabner, R. H., Gebauer, D., Reishofer, G., Koschutnig, K., & Ebner, F. (2010). Enhancing creativity by means of cognitive stimulation: Evidence from an fMRI study. NeuroImage, 52, 1687–1695. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.neur​​oimag​​e​.2​01​​0​.05.​​072 Fiske, S. T., & Cuddy, A. J. (2006). Stereotype content across cultures as a function of group status. In S. Guimond (Ed.), Social comparison processes and levels of analysis (pp. 249–263). New York: Cambridge University Press. https​:/​/do​​i​.org​​/10​.1​​017​/C​​BO978​​05115​​​84329​​.013 Fiske, S. T., Cuddy, A. J., & Glick, P. (2007). Universal dimensions of social cognition: Warmth and competence. Trends in Cognitive Sciences, 11(2), 77–83. https​:/​/do​​i​.org​​/ 10​.1​​016​/j​​.tics​​.200​6​​.11​.0​​05 Fiske, S. T., Cuddy, A. J., Glick, P., & Xu, J. (2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82(6), 878–902. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​022​-3​​514​​.8​​2​.6​.8​​78 Fiske, S. T., Xu, J., Cuddy, A. C., & Glick, P. (1999). (Dis)respecting versus (dis)liking: Status and interdependence predict ambivalent stereotypes of competence and warmth. Journal of Social Issues, 55(3), 473–489. https://doi​.org​/10​.1111​/0022​-4537​.00128 Foss, L., Woll, K., & Moilanen, M. (2013). Creativity and implementations of new ideas. International Journal of Gender and Entrepreneurship, 5(3), 298–322. https​:/​/do​​i​.org​​/10​. 1​​108​/I​​JGE​-0​​9​​-201​​2​-004​9 Fulmer, A., & Dirks, K. (2018). Multilevel trust: A theoretical and practical imperative. Journal of Trust Research, 8(2), 137–141. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​.20​​​18​. 15​​31657​ Fulmer, A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​. org​/10​.1177​/0149206312439327 Gartzia, L., Ryan, M., Balluerka, N., & Aritzeta, A. (2012). Think crisis-think female: Further evidence. European Journal of Work and Organizational Psychology, 21(4), 603–628. https​:/​/do​​i​.org​​/10​.1​​080​/1​​35943​​2x​.20​​​11​.59​​1572 Giambatista, R. C., & Bhappu, A. D. (2010). Diversity’s harvest: Interactions of diversity sources and communication technology on creative group performance. Organizational Behavior and Human Decision Processes, 111(2), 116–126. https​:/​/do​​i​.org​​/10​.1​​016​/j​​. obhd​​p​.20​0​​9​.11.​​003 Gupta, A. K., & Singhal, A. (1993). Managing human resources for innovation and creativity. Research-Technology Management, 36(3), 41–48. https​:/​/do​​i​.org​​/10​.1​​080​/0​​89563​​08​.19​​9​3​. 11​​67090​2

252  Hye Jung Eun, Roy Chua, and Mengzi Jin

Hammond, M. M., & Farr, J. L. (2011). The role of errors in the creative and innovative process. In D. A. Hofmann & M. Frese (Eds.), Errors in Organizations (pp. 67–96). Routledge. https://doi​.org​/10​.4324​/9780203817827​-10 Harris, J. A. (2004). Measured intelligence, achievement, openness to experience, and creativity. Personality and Individual Differences, 36(4), 913–929. https​:/​/do​​i​.org​​/10​.1​​016​ /s​​0​191-​​8869(​03)00​161-2​ Harrison, L. A., & Lynch, A. B. (2005). Social role theory and the perceived gender role orientation of athletes. Sex Roles, 52(3), 227–236. https​:/​/do​​i​.org​​/10​.1​​007​/s​​11199​​-00​ 5-​​1297-​1 House, J. S., Umberson, D., & Landis, K. R. (1988). Structures and processes of social support. Annual Review of Sociology, 14(1), 293–318. https​:/​/do​​i​.org​​/10​.1​​146​/a​​nnure​​v​ .so.​​14​.08​​​0188.​​00145​3 Ibarra, H. (1992). Homophily and differential returns: Sex differences in network structure and access in an advertising firm. Administrative Science Quarterly, 37(3), 422–447. https://doi​.org​/10​.2307​/2393451 Ingram, P., & Roberts, P. W. (2000). Friendships among competitors in the Sydney hotel industry. American Journal of Sociology, 106(2), 387–423. https://doi​.org​/10​.1086​/316965 Jehn, K. A., Northcraft, G. B., & Neale, M. A. (1999). Why differences make a difference: A field study of diversity, conflict and performance in workgroups. Administrative Science Quarterly, 44(4), 741–763. https://doi​.org​/10​.2307​/2667054 Jensen, K., Kovács, B., & Sorenson, O. (2018). Gender differences in obtaining and maintaining patent rights. Nature Biotechnology, 36(4), 307–309. https://doi​.org​/10​ .1038​/nbt​.4120 Jiang, C. X., Chua, R. Y., Kotabe, M., & Murray, J. Y. (2011). Effects of cultural ethnicity, firm size, and firm age on senior executives’ trust in their overseas business partners: Evidence from China. Journal of International Business Studies, 42(9), 1150–1173. https:// doi​.org​/10​.1057​/jibs​.2011​.35 Joshi, A., Son, J., & Roh, H. (2015). When can women close the gap? A meta-analytic test of sex differences in performance and rewards. Academy of Management Journal, 58(5), 1516–1545. https://doi​.org​/10​.5465​/amj​.2013​.0721 Kanze, D., Huang, L., Conley, M. A., & Higgins, E. T. (2018). We ask men to win and women not to lose: Closing the gender gap in startup funding. Academy of Management Journal, 61(2), 586–614. https://doi​.org​/10​.5465​/amj​.2016​.1215 Korsgaard, M. A., Brower, H. H., & Lester, S. W. (2015). It isn’t always mutual: A critical review of dyadic trust. Journal of Management, 41(1), 47–70. https://doi​.org​/10​.1177​ /0149206314547521 Krasikova, D.V., & LeBreton, J. M. (2012). Just the two of us: Misalignment of theory and methods in examining dyadic phenomena. Journal of Applied Psychology, 97(4), 739–757. https://doi​.org​/10​.1037​/a0027962 Lau, S., & Li,W. L. (1996). Peer status and perceived creativity: Are popular children viewed by peers and teachers as creative. Creativity Research Journal, 9(4), 347–352. https​:/​/do​​i​ .org​​/10​.1​​207​/s​​15326​​934c​r​​j0904​_6 Lewis, J. D., & Weigert, A. (1985). Trust as a social reality. Social Forces, 63(4), 967–985. https://doi​.org​/10​.1093​/sf​/63​.4​.967 Lin, N. (2001). Social capital: A theory of social structure and action. Cambridge: Cambridge University Press.

Gender Influences on Trust and Creativity  253

Mannix, E., & Neale, M. A. (2005). What differences make a difference? The promise and reality of diverse teams in organizations. Psychological Science in the Public Interest, 6(2), 31–55. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1529​​-1006​​.200​5​​.0002​​2.x Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​ 95​.​95​​08080​​335 McAllister, D. J. (1995). Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https://doi​ .org​/10​.2307​/256727 Meeker, B. F., & Weitzel-O'Neill, P. A. (1977). Sex roles and interpersonal behavior in taskoriented groups. American Sociological Review, 42(1), 91–105. https://doi​.org​/10​.2307​ /2117733 Mehra, A., Kilduff, M., & Brass, D. (1998). At the margins: A Distinctiveness approach to the social identity and social networks of underrepresented groups. Academy of Management Journal, 41(4), 441–452. https://doi​.org​/10​.2307​/257083 Milliken, F. J., & Martins, L. L. (1996). Searching for common threads: Understanding the multiple effects of diversity in organizational groups. Academy of Management Review, 21(2), 402–433. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​96​.​96​​05060​​217 Nemeth, C., Personnaz, B., Personnaz, M., & Goncalo, J. (2004). The liberating role of conflict in group creativity:A study in two countries. European Journal of Social Psychology, 34(4), 365–374. https://doi​.org​/10​.1002​/ejsp​.210 Pelled, L. H., Eisenhardt, K. M., & Xin, K. R. (1999). Exploring the black box: An analysis of work group diversity, conflict and performance. Administrative Science Quarterly, 44(1), 1–28. https://doi​.org​/10​.2307​/2667029 Proudfoot, D., Kay,A. C., & Koval, C. Z. (2015).A gender bias in the attribution of creativity: Archival and experimental evidence for the perceived association between masculinity and creative thinking. Psychology Science, 26(11), 1751–1761. https://doi​.org​/10​.1177​ /0956797615598739 Rodan, S., & Galunic, C. (2004). More than network structure: How knowledge heterogeneity influences managerial performance and innovativeness. Strategic Management Journal, 25(6), 541–562. https://doi​.org​/10​.1002​/smj​.398 Rudman, L. A., & Glick, P. (2001). Prescriptive gender stereotypes and backlash toward agentic women. Journal of Social Issues, 57(4), 743–762. https://doi​.org​/10​.1111​/0022​ -4537​.00239 Ryan, M. K., Haslam, S., Hersby, M., & Bongiorno, R. (2011). Think crisis–think female: The glass cliff and contextual variation in the think manager–think male stereotype. Journal of Applied Psychology, 96(3), 470–484. https://doi​.org​/10​.1037​/a0022133 Ryan, M. K., Haslam, S., & Postmes, T. (2007). Reactions to the glass cliff: Gender differences in the explanations for the precariousness of women's leadership positions. Journal of Organizational Change Management, 20(2), 182–197. https​:/​/do​​i​.org​​/10​.1​​108​ /0​​95348​​107​10​​72474​8 Saeki, N., Fan, X., & Van Dusen, L. (2001). A comparative study of creative thinking of American and Japanese college students. The Journal of Creative Behavior, 35(1), 24–36. https​:/​/do​​i​.org​​/10​.1​​002​/j​​.2162​​-6057​​.2001​​​.tb01​​219.x​ Shaub, M. K. (1996). Trust and suspicion; the effects of situational and dispositional factors on auditors' trust of clients. Behavioral Research in Accounting, 8, 154–174.

254  Hye Jung Eun, Roy Chua, and Mengzi Jin

Shazi, R., Gillespie, N., & Steen, J. (2015). Trust as a predictor of innovation network ties in project teams. International Journal of Project Management, 33(1), 81–91. https​:/​/do​​i​.org​​ /10​.1​​016​/j​​.ijpr​​oman.​​20​14.​​06​.00​1 Simpson, J. A. (2007). Psychological foundations of trust. Current Directions in Psychological Science, 16(5), 264–268. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-8721​​.200​7​​.0051​​7.x Slabbert, J. A. (1994). Creativity in education revisited: Reflection in aid of progression. The Journal of Creative Behavior. 28(1), 60–69. https​:/​/do​​i​.org​​/10​.1​​002​/j​​.2162​​-6057​​.1994​​​ .tb00​​720.x​ Smith, S. S. (2000). Mobilizing social resources: Race, ethnic, and gender differences in social capital and persisting wage inequalities. The Sociological Quarterly, 41(4), 509–537. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1533​​-8525​​.2000​​​.tb00​​071.x​ Sonnenburg, S. (2004). Creativity in communication: A theoretical framework for collaborative product creation. Creativity and Innovation Management, 13(4), 254–262. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.0963​​-1690​​.200​4​​.0031​​4.x Sosa, M. E. (2011). Where do creative interactions come from? The role of tie content and social networks. Organization Science, 22(1), 1–21. https://doi​.org​/10​.1287​/orsc​.1090​ .0519 Spector, M. D., & Jones, G. E. (2004). Trust in the workplace: Factors affecting trust formation between team members. Journal of Social Psychology, 144(3), 311–321. https​:/​ /do​​i​.org​​/10​.3​​200​/S​​OCP​.1​​44​​.3.​​311​-3​​21 Strodtbeck, F. L., & Mann, R. D. (1956). Sex role differentiation in jury deliberations. Sociometry, 19(1), 3–11. https://doi​.org​/10​.2307​/2786099 Sutherland, C. A., Oldmeadow, J. A., & Young, A. W. (2016). Integrating social and facial models of person perception: Converging and diverging dimensions. Cognition, 157, 257–267. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.cogn​​ition​​.20​16​​.09​.0​​06 Tajfel, H. E. (1978). Social categorization, social identity, and social comparison. In H. Tajfel (Ed.), Differentiation between social groups: Studies in the social psychology of intergroup relations (pp. 61–76). London: Academic Press. Totterdell, P., Wall, T., Holman, D., Diamond, H., & Epitropaki, O. (2004). Affect networks: A structural analysis of the relationship between work ties and job-related affect. Journal of Applied Psychology, 89(5), 854–867. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​9​.5​.8​​54 Turner, J. C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.), Social identity and intergroup relations (pp. 15–40). Cambridge: Cambridge University Press. Turner, J. C. (1985). Social categorization and the self-concept: A social-cognitive theory of group behavior. In E. J. Lawler (Ed.), Advances in group processes: Theory and research 2 (pp. 77–122). Greenwich, CT: JAI Press. Unsworth, K. (2001). Unpacking creativity. Academy of Management Review, 26(2), 289–297. https://doi​.org​/10​.2307​/259123 Uzzi, B. (1996). The sources and consequences of embeddedness for the economic performance of organizations: The network effect. American Sociological Review, 61(4), 674–698. https://doi​.org​/10​.2307​/2096399 Wachter, M. L., & Estlund, C. L. (2012). Research handbook on the economics of labor and employment law. Edward Elgar Publishing. https://doi​.org​/10​.4337​/9781781006115 Wang, W., Huang, S.,Yin, H., & Ke, Z. (2018). Employees' emotional labor and emotional exhaustion: Trust and gender as moderators. Social Behavior and Personality, 46(5), 733– 748. https://doi​.org​/10​.2224​/sbp​.6745

Gender Influences on Trust and Creativity  255

Wasko, M. M., & Faraj, S. (2005). Why should I share? Examining social capital and knowledge contribution in electronic networks of practice. MIS Quarterly, 29(1), 35– 57. https://doi​.org​/10​.2307​/25148667 Whitener, E. M., Brodt, S. E., Korsgaard, M. A., & Werner, J. M. (1998). Managers as initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23(3), 513–530. https://doi​.org​ /10​.2307​/25929 Wojciszke, B. (2005). Morality and competence in person-and self-perception. European Review of Social Psychology, 16(1), 155–188. https​:/​/do​​i​.org​​/10​.1​​080​/1​​04632​​805​00​​ 22961​9 Wood, W., & Eagly, A. H. (2012). Biosocial construction of sex differences and similarities in behavior. In M. P. Zanna & J. M. Olsen (Eds.), Advances in experimental social psychology (Vol. 46, pp. 55–123). San Diego, CA: Elsevier Inc. Zhang, X., & Zhou, J. (2014). Empowering leadership, uncertainty avoidance, trust, and employee creativity: Interaction effects and a mediating mechanism. Organizational Behavior and Human Decision Processes, 124(2), 150–164. https​:/​/do​​i​.org​​/10​.1​​016​/j​​ .obhd​​p​.20​1​​4​.02.​​002 Zhou, J., Shin, S. J., Brass, D. J., Choi, J., & Zhang, Z. X. (2009). Social networks, personal values, and creativity: Evidence for curvilinear and interaction effects. Journal of Applied Psychology, 94(6), 1544–1552. https://doi​.org​/10​.1037​/a0016285

11 MULTILEVEL TRUST IN UNCERTAIN CONTEXTS ODID You Hear We Have a New VP?1 Roger C. Mayer and Michele Williams2

Introduction Bill was hired as a supervisor in his department. Bill understood that his job was to improve the department’s performance. Shortly after he started, Bill noted that employees were breaking the company’s rules and formal procedures. In consultation with his manager, Bill began formal procedures to initiate performance management. Bill’s actions were supported by four levels of management above him who he trusted to support him in the conflict and who explicitly said they supported him. After one year, and after Bill had thoroughly alienated a number of misbehaving employees, the CEO fired Bill’s second-level manager, who removed Bill’s direct manager as well. In fact, the CEO replaced the three managers above Bill with managers who sought to take the organization in a very different direction. Bill was caught in between having several levels of new managers above him who did not support his formal actions with the employees, and the employees in his department – who were not only unionized but also empowered by the new direction and lack of higher-level support for Bill. The story above is true and is one example of many of which the authors are aware that have occurred in work organizations. Such situations are unfortunately all too common and put managers in a very uncomfortable situation, one which can be both career-damaging and undermining of relationships with subordinates.

1 The authors would like to thank C. Ashley Fulmer, Roy Lewicki, and Philip Bobko for their helpful comments on an earlier version of this chapter. 2 The authors contributed equally to this chapter.

DOI: 10.4324/9780429449185-11

Multilevel Trust in Uncertain Contexts  257

The purpose of this chapter is to consider how trust in managers is affected by how unified or dissociative an organization is in the eyes of employees who are evaluating their managers’ – and secondarily their employer’s – trustworthiness. We integrate literature on multiple organizational identities with trust literature to better understand how complex multilevel organizational influences impact the trustworthiness of managers by influencing their ability and incentives to follow through on promises to employees. We introduce the concept of Organization Dissociative Identity Disorder (ODID) that reflects the extreme case in which organizational forces (cultural changes, crises, conflicting organizational logics, and inconsistent routines) act to pull a manager in different directions, and inadvertently undermine his or her ability to act with ability, benevolence, and integrity. We first discuss the effect of multiple influences that individuals or situations can have on a supervisor and then turn to how they affect an employee’s trust in that supervisor.

Organizational Dissociative Identity Disorder Change and Vectors Kurt Lewin’s field theory (1951) spoke of similar forces as driving and restraining forces regarding a change (p. 218). In his seminal work on change, his purpose for this terminology was to present force field analysis, or forces acting to push forward or inhibit a specific change. Lewin wrote this groundbreaking work in an era of great stability, describing a given change at a point in time in a given direction (e.g., p. 205). He recognized that the magnitude of driving and restraining forces can change over time, but his focus was on how the balance of these “opposing forces” (p. 201) affected the focal change at a point in time.The driving and restraining forces he depicted in his figures were 180 degrees to one another. In more than a half-century since Lewin’s work, work organizations and the environments in which they operate have both become considerably more complex. Following his logic, but in consideration of the complexity of many concurrent forces in various directions (i.e., not simply one directed toward a single goal and another resisting that force), we instead refer to these forces as vectors, based on both Lewin’s reasoning and on the meaning of that word in physics. A vector has a direction and a magnitude, and importantly, the vectors need not be opposing. Lewin’s examples become too simplistic when (1) those who hold power within different parts of the organization either hold different objectives, change them repeatedly, or have self-serving agendas, or (2) other influences act either to change the magnitudes or directions of the vectors or to add new vectors. We consider some of these forces below but will look first at the result of a combination of these factors. The combination of these simultaneous vectors (examples of which are below) can pull a manager to respond in ways that are inconsistent and lack benevolence, or as we will explore, other factors of trustworthiness.

258  Roger C. Mayer and Michele Williams

The Concept of ODID Several decades ago, a popular novel was published entitled Sybil (Schreiber, 1973). The book depicted a woman afflicted with what was then called multiple personality disorder, currently referred to as dissociative identity disorder (DID) in the APA’s DSM-5 (Diagnostic and Statistical Manual, American Psychological Association, 2013). The novel depicted Sybil as having 16 distinct identities, all housed within a single person. These identities each had names and different, developed personalities. It is easy to imagine how difficult it would be for a family member or coworker to deal with Sybil, or with any person afflicted with DID. Conversations and agreements made with one personality may be irrelevant within hours when a different personality is in control of the person’s behaviors and reactions. Thankfully, DID is a relatively rare affliction for a person to have. According to WebMD (2020), studies indicate it afflicts about 1% of the population. Unfortunately, something akin to this syndrome seems to be much more common among organizations. Despite attempts to align various parts of an organization via such mechanisms as mission statements and speeches by top managers, individual managers in an organization don’t necessarily align in their agendas. A given employee may hear several inconsistent messages in a short period of time from different managers who control resources and opportunities that the employee wants. Rather than the organization being a coherent whole in which the employee places a greater or lesser amount of trust, it can come across to the employee as a combination of different non-aligned, opposing, or even random vectors pushing the employee toward different goals and actions. This lack of consistency of the vectors can come across as the organization having multiple ‘personalities.’ Each of the personalities may be more or less trustworthy to the focal employee for different reasons. Having provided a sense of the problem this chapter addresses, we offer the following definition: Organizational Dissociative Identity Disorder (ODID) is the strength of an organization member’s perception that her/his employing organization, groups within that organization, and/or individuals who represent the organization (e.g., leaders, founders) hold independent or conflicting logics, identities, values, incentives, and/or goals. ODID reflects the perceived extent of non-alignment that exists among the individuals, groups, and organizations that provide direction to the organization. As vectors become less aligned in their directions, or a shift in the relative strengths of misaligned vectors occurs, perceptions of ODID increase.While it can be argued that some extent of misalignment is inevitable and to be expected in any organization in today’s complex organizations, we propose the following boundary condition of ODID. We submit that the perception of ODID exists when the effects of the various conflicting vectors in an organization are such that the manager cannot find a path of action that adequately satisfies the demands in ways that are simultaneously effective, ethical,

Multilevel Trust in Uncertain Contexts  259

and consistent with past actions. While we define ODID as a continuum, the point at which ODID begins and the organization seems irrational to a particular manager or employee is idiosyncratic, because people experience the organization differently and people also have different capacities for coping and for rationalization of the problematic vectors. As the manager or employee judges the vectors to be clearly and significantly in conflict with one another so that it becomes unclear about how to proceed, then we propose that perceptions of ODID are likely to cause strain for the person, potentially leading to such responses as withdrawal, acting out, or leaving the situation. The research of Ashford, Lee, & Bobko (1989) found that factors including anticipated organizational change could lead to job insecurity, which then led to intention to quit, and lower commitment and satisfaction. In related research, Mayer and Gavin (2005) measured employees’ ability to focus attention. They found that employees who felt that the political situation in their company pulled them in different directions performed less organizational citizenship behaviors (OCB) directed at both other individuals and the organization. While various authors have written about trusting an organization (Fulmer & Gelfand, 2012 for a review), implying that an organization acts as a cohesive unit, we confront this supposition and argue that it may not be warranted. The Organizational Identity (OI) literature, for example, has recognized that it is not unusual for an organization to have multiple identities that may be somewhat inconsistent with each other (e.g., Glynn, 2000; Pratt & Rafaeli, 1997). When the employing organization is perceived as having multiple personalities, the incongruencies can easily reach a point wherein middle managers, lower-level managers, and employees see the result of the various vectors as resulting in a dysfunctional whole. To this point, we have discussed how conflicting vectors can contribute to a belief that the organization functions like it has multiple personalities (ODID). We turn next to the effect these vectors can have on trust relationships among the organization’s members, as well as how individuals trust the organization.

Multilevel Trust in Organizations: Background In the mid-1990s, several groups of authors broached the idea that trust was an important topic that deserved more attention in the organizational sciences (e.g., Hosmer, 1995; Kramer & Tyler, 1996; Mayer, Davis, & Schoorman, 1995; McAllister, 1995). These included papers focused on individuals’ trust in other individuals (Lewicki & Bunker, 1996), trust with an organization as a referent (Hosmer, 1995), and multilevel trust (Schoorman, Mayer, & Davis, 1996). Shortly afterward, interest in the topic bloomed, and it has become among the most popular topics of study in the organizational sciences (Fulmer & Gelfand, 2012).Thus, from what might be deemed the beginning of the modern era of organizational trust research, the topic has been considered at multiple levels.

260  Roger C. Mayer and Michele Williams

One prominent area of trust research has been trust in a supervisor or manager. Numerous studies have examined trust in supervisors, facility managers, and higher-level managers, either as individual referents or in combination (e.g., Colquitt, Scott, & LePine, 2007; Davis, Schoorman, Mayer, & Tan, 2000; Mayer & Gavin, 2005). These managers are at various levels of the organization, from first-line people to the CEO. Trust in a manager affects important organizational processes such as employees’ discretionary behaviors, willingness to follow, performance, and turnover (Burke, Sims, Lazzara, & Salas, 2007; Davis et al., 2000; Dirks & Ferrin, 2002; Fulmer & Gelfand, 2012). Although there is a positive relationship between trust in managers and these outcomes, there is also an important caveat that too much trust in a manager that is ill-placed can be dangerous, as an employee can be abused by that manager or left unsupported when the manager leaves suddenly.

Manager Trustworthiness in Context Trustworthiness Significant empirical evidence suggests that trust in managers is based largely on three fundamental bases of perceived trustworthiness delineated by Mayer et al. (1995): ability, benevolence, and integrity (e.g., Colquitt et al., 2007; Davis et al., 2000; Mayer & Gavin, 2005). Burke et al. (2007) assert that these dimensions of trustworthiness are helpful in understanding effective leadership. They assert that leaders can demonstrate trustworthiness showing (1) ability by setting a compelling direction, creating an enabling structure, or by their task knowledge, (2) benevolence by coaching, showing empathy, or creating a supportive context, and (3) integrity with reliability, fairness, accountability, and valuecongruent actions. Because managers are embedded within an organizational context, they may not always have the autonomy and support necessary to enact behaviors that demonstrate ability, benevolence, and integrity to their subordinates. For example, consider a situation wherein an employee and supervisor have been working together for a year. The supervisor has done nothing in that time that makes the employee question the supervisor’s word or commitments. Then, the supervisor’s manager is replaced with another manager from a unit with different values and priorities. Thus, the new manager has different ideas about what should be expected in terms of performance, scheduling, employee development, etc., and communicates these to the supervisor. The supervisor’s ability to honor prior commitments to the employee can suffer greatly and immediately due to the supervisor’s new reporting relationship. The vectors influencing the supervisor have changed, and the supervisor acting with integrity both toward the manager and toward employees suddenly comes into conflict. Depending upon the employees’ understanding of the demands on their direct manager, employees

Multilevel Trust in Uncertain Contexts  261

may question the manager’s ability to lead, the integrity with which they made their promises, and/or their concern for protecting the employees’ well-being (benevolence). As the effect of the varying vectors on the supervisor plays out in the relationship with the subordinate, previously agreed-upon aspects of the employment relationship are no longer honored. Even if the supervisor provides an explanation about the reason for the change, from the perspective of her or his employees, it still may be the supervisor’s responsibility to fulfill promises and support employees. Consequently, employees may view the supervisor as having less integrity and benevolence. And for reasons that have nothing to do with the supervisor’s intentions when making commitments, the context of their relationship with the employee is altered.The employee may be less willing to be vulnerable to the supervisor, and less willing to believe the supervisor’s promises and offers. Furthermore, because supervisors are an important representative of the organization (Kristof-Brown, Zimmerman, & Johnson, 2005; Levinson, 1965), employees may also feel their psychological contract has been violated (see, for example, Morrison & Robinson, 1997; Robinson & Morrison, 2000; Robinson & Rousseau, 1994), thereby decreasing their trust in the organization such that trust on multiple levels is affected.

Organizational Context We argue that middle managers, who are situated a level above front-line supervisors and often implement changes in an organization’s strategic direction or enforce organizational priorities from other units such as human resources or legal, may provide a conduit for important multilevel contexts influencing employees’ trust on multiple levels – i.e., employee trust in their direct supervisors and the organizations they represent. In this chapter, we focus on a wide range of contextual influences on ODID – conflicting logics, identities, values, incentives, and/or goals stemming from internal and external circumstances (see Figure 11.1). However, this is not to say that concerns about the negative experiences that managers and their subordinates may have with conflicting vectors are new to the organizational sciences. In fact, in a well-known paper now half a century old, Rizzo, House, and Lirtzman (1970) developed measures of role conflict and role ambiguity. Of the two, the former is directly related to the notion of multiple vectors described in this chapter. Rizzo et al. (1970) also refer to the concept of unity of command that was well accepted at the time: The principle of unity of command states that for any action an employee should receive orders from one superior only, and that there should be only one leader and one plan for a group of activities having the same objective is the idea that an employee should report to one and only one manager…the

262  Roger C. Mayer and Michele Williams

structure of the organization should keep a member from incompatible orders or incompatible expectations from more than one superior. (p. 150) Role conflict from incompatible orders or incompatible expectations from more than one superior increases as unity of command decreases. With matrix organizations, intra-organizational network structures, and organizations that are more team-based and less hierarchical in general, unity of command is less and less common, increasing the likelihood of higher levels of role conflict and ODID. Examining ODID enables us to place trust in a manager in the broader organizational context that may influence employees’ trust in their managers and the relative wisdom of that trust. Below, we explore how the context in which each level of management acts can have a significant effect on those managers’ behavior and subsequently on their direct reports’ interpretation of their integrity, benevolence, and overall trustworthiness. While the importance of the context for interpersonal trust relationships has gained widespread acceptance (e.g., Burt & Knez, 1996; Ferrin, Dirks, & Shah, 2006; Jones & Shah, 2016; Mayer et al., 1995; Williams, 2001, 2016), less attention has focused on organizational characteristics and dynamics that influence how employees perceive the trustworthiness of their managers. ODID vector sources Internal crises

External crises

Change in strategic direction

Organizational change/stability

Leader ABI-relevant behaviors Past experience with former leaders

Follower trust propensity Perceived leader trustworthiness (subordinate’s view of leaders’ behavior)

FIGURE 11.1 Organization-level

influences on perceived leader trustworthiness.31

3 This model focuses on influences on perceived leader (manager) trustworthiness. Supervisors and managers are considered as leaders in their own part of the organization. Concomitant influences on organizational trustworthiness are described in the text but not included on this figure.

Multilevel Trust in Uncertain Contexts  263

Organization-Level Influences on Trust in Managers There is a dearth of research that applies attribution theory to trust in a multilevel context (see Tomlinson and Langlinais, 2021).We use attribution theory to examine contextual influences on the vectors influencing employees and their trust in the direct supervisors. These contextual factors include middle managers above the level of the focal supervisor and three organization-level factors that affect the supervisors’ ability to fulfill their promises: (1) external crises, (2) internal crises, and (3) changes in strategic direction.

Attributions for ODID and Multiple Vectors Attribution theory provides a framework for understanding when employees’ trust in their supervisors, their organization, or both are likely to be influenced (Tomlinson & Mayer, 2009; Williams, Belkin, & Chen, 2020) by conflicting vectors originating from crises and changes in strategic directions. According to Weiner (1986), when something negative happens the individual experiences a general emotional reaction of displeasure. Subsequently, the person looks to identify the cause of the negative outcome. If, on the one hand, the causal ascription is within the focal manager, then trust in that manager will likely suffer. When a focal manager perceives ODID and experiences his or her own behavior as externally constrained, but his or her subordinates do not notice or understand the ramifications of the organization-level factors, subordinates may make internal attributions, finding the manager responsible for broken promises and failures to protect their well-being and, therefore, less trustworthy in terms of integrity and benevolence. On the other hand, if the source of the displeasure is deemed to be outside the manager’s control, this may or may not damage trust in the manager. If the causal ascription is that the organization caused the pain, the organization itself is likely to be seen as less trustworthy, but the perceived trustworthiness of the manager may not be affected. The reasoning for this is similar to that just described for an individual manager as a trustee; however, the causal ascription is internal to the organization, not the manager. In contrast, if the cause of the emotional displeasure is deemed to be the organization’s overriding of the focal manager’s judgment or wishes, then this may be seen as limiting the focal manager’s ability. Mayer et al. (1995) differentiated ability from competence, the former being situation-specific and the latter being relatively constant in the near term.Thus, if the organization changes directions or provides conflicting vectors for the manager, the subordinate may see that while the manager’s competence remained consistent, her/his ability declined because of the degraded organizational context and conflicting vectors to which the manager was no longer able to respond effectively. Thus, an ODID context, when perceived by a manager’s subordinates, not only makes the organization itself less

264  Roger C. Mayer and Michele Williams

predictable and less trustworthy, it may also damage the trustworthiness of the manager. Even when the subordinate sees that the manager has good intentions, if s/he is not able to fulfill promises or protect the subordinate’s interests, then the decay in the manager’s situational ability will also damage the subordinate’s trust in the manager. In sum, an ODID context may affect not only an employee’s trust in the organization but also the employee’s trust in the manager, even if the employee does not deem a given manager to be the cause of the displeasure. ODID can thus affect an employee’s trust on multiple levels: both the organization and the managers attempting to work within an ODID context.

Middle Managers as Conduits of Organizational Processes Organizations inevitably experience turnover. When middle managers above the level of the focal supervisor enter the organization, their behavior and alignment with a specific organizational or departmental identity can either support or damage the trust an employee has in her/his supervisor. Because middle managers can have important effects on the context of the focal supervisor’s relationship with the subordinate, the particular middle manager may not even be in the reporting line above the employee, but may be influential within a support function like HR or the organization’s legal department (Burke et al., 2007).Within the middle manager’s scope of authority may be the power to make decisions about interpretation of the benefits to which the employee is allowed, or the way the organization will treat its employees for legal reasons. Thus, a change in a middle manager, although a small-scale organizational change, may alter the directions or magnitudes of various vectors or bring new ones to the situation, leaving the employee questioning how valued s/he is by their direct supervisor and the organization, thereby affecting trust in the organization and in the supervisor.

Contextual Influence: External Crises External crises such as recessions, natural disasters, pandemics, and changes in regulatory environments can impact all aspects of organizational functioning such as the available workforce, firms’ physical locations, and their financial stability. For example, during the COVID-19 pandemic in 2020, travel restrictions affected hotels, airlines, restaurants, and their suppliers. The vectors affecting managers changed tremendously in a matter of weeks. Firms had to decide which workers and functions were essential and how to manage a workforce working from home. Frontline supervisors and managers were suddenly influenced by new or altered vectors from finance, human resources, and legal departments. Firm revenues placed many supervisors in a position where only a fraction of their direct reports could keep their jobs. For example, GE Aviation went through a round of voluntary and involuntary layoffs of roughly 25% of their workforce (Gryta, 2020).

Multilevel Trust in Uncertain Contexts  265

At one point, Disney furloughed 43,000 union workers while 200 remained at their jobs (Witten & Kimball, 2020). Other firms temporarily cut the salaries of workers and/or re-assigned them. For example, Scandinavian Airlines trained workers to assist in nursing homes and besieged hospitals during the COVID19 pandemic (Associated Press, 2020). In the United States, CVS Health hired furloughed workers from its business partners in the travel industry, Marriott and Hilton (Nathan-Kazis, 2020). In colleges and universities across the globe, budgets were frozen and HR departments were tasked with coming up with policies to oversee moving nearly all of their employees to working from home. Supervisors likely received new policies and imperatives from different levels within their organization and were then tasked with implementing those policies. In such a flurry of novel expectations, inconsistencies amongst vectors and sudden shifts in their magnitudes are apt to abound. More widespread financial disaster may prompt employees to place blame externally on uncontrollable circumstances (Tomlinson & Mayer, 2009) and not blame organizations or supervisors. However, crises that highlight flawed or unsafe organizational choices such as poor financial investment or lackadaisical safety standards may prompt an internal attribution. For example, while the COVID-19 crisis was worldwide, it also brought the poor working conditions in meatpacking plants and the previous lack of attention to employee health to the forefront of workers’ minds (Sallinger, 2020). For example, workers in a JBS meatpacking plant reported being afraid to come to work and that health precautions espoused to be in place by the company were not in fact being carried out (Sallinger, 2020). As new and changing vectors for safety were influencing organizations in the meatpacking industry, their middle managers, and their frontline supervisors, this situation could give rise to ODID. Moreover, when supervisors do not protect their employees, especially when other supervisors in the same or similar companies do protect their employees, it could negatively affect the employee’s psychological contract with the firm and the assessment of the supervisor’s ability to use their upward influence. It could damage the supervisor’s perceived benevolence if the employee interpreted the situation as reflecting that the supervisor did not care enough to use his/her influence on behalf of the employee. Finally, the situation could cause a reduction in the perception of integrity if the employee thought that rewards were no longer being distributed fairly. Such re-evaluations of trustworthiness are, of course, more likely if the employee makes causal attributions for the situation that are internal to the organization and the supervisor (Tomlinson & Mayer, 2009). In such situations, internal attributions may be likely because employees are under stress, and their stress levels may prevent them from observing or comprehending the new and changing vectors that are influencing their supervisors (Williams, 2007). For example, new vectors arising from governmental mandates, human resources policies, and new legal constraints may all influence supervisors and their managers. In very few situations are the attributions clear-cut. Henry Ward, CEO of Carta,

266  Roger C. Mayer and Michele Williams

foreseeing the possibility that workers would personally blame their managers for layoffs, took full blame for layoffs by officially absolving all managers. In a LinkedIn post, he provided a copy of the speech he made to his employees (Ward, April 15, 2020): It is important that all of you know I personally reviewed every list and every person. If you are one of those affected it is because I decided it.Your manager did not. For the majority of you it was quite the contrary. Your manager fought to keep you and I overrode them. They are blameless. If today is your last day, there is only one person to blame and it is me. In sum, external crises contribute to ODID by generating new vectors and changing directions and magnitudes of existing vectors, thus putting new pressures on frontline supervisors and middle managers.

Contextual Influence: Internal Crises Internal crises such as sexual harassment allegations create a similar situation in that new forces and vectors come into play. In this example, human resource policies and legal requirements restrict what managers can say both internally and publicly.Thus, managers who have always been transparent and supportive of gender equality may find themselves for months allowed to send only very general messages about ‘zero tolerance,’ as investigations occur and the legal rights of the accused are protected. Even when the legal vectors are known, doubts and gossip that ‘more could be done’ may arise. Consequently, employees may question managers’ concern for protecting them (benevolence) and the veracity of their support for gender equality (integrity).

Contextual Influence: Changes in Strategic Direction A consistent and uniform strategic direction can instill trust in managers who are likely to have role clarity and should be better able to fulfill promises without conflicting demands from their superiors. Such organizations are likely to have a clear and unified organizational identity. Organizational identity refers to the characteristics of an organization that its members believe are central, distinctive, and enduring (Albert & Whetten, 1985; Whetten, 2006). A shared organizational identity provides a clear guide for managers wishing to make decisions that keep their trustworthiness intact. In contrast, changes in the strategic direction of an organization such as those associated with mergers and acquisitions often create organizations with multiple identities, and consequently new vectors influencing the actions of lowerlevel managers. Organizations with multiple identities contain groups within the organization that have differing views on what is central, distinctive, and enduring

Multilevel Trust in Uncertain Contexts  267

about the organization (Pratt & Foreman, 2000). Although these identities are not necessarily in conflict, conflicting identities are the ones that may undermine trustworthy actions, perceptions of other people’s trustworthiness, and perceptions of the organization’s trustworthiness. In addition to the creation of multiple identities, strategic choices at the organization level can cause shifts in the status and dominance of pre-existing identities. Thus, although uniformity and consistency of the strategic direction and a uniform organizational identity can have a significant positive effect on lower-level managers’ ability to behave in a consistent manner, periods of strategic planning and change may have the opposite effect. These changes are designed to create vectors pushing managers in new directions, but the impact of potential misalignment of these vectors may have unintended influences on employees’ trust in their supervisors and in their organization. The final impact of creating new vectors and new organizational identities depends on the level of conflict versus synergy among the multiple new identities. Furthermore, these vectors may become less conflictual over time as the organization moves to a more stable relationship among the new vectors and new identities. We provide examples of the potentially cascading impact of strategic changes under two types of strategic changes: (1) mergers and acquisitions and (2) divestitures.

Mergers and Acquisitions When mergers and acquisitions bring different organizational identities together, organizations may attempt to keep those identities physically isolated from one another in different locations, integrate them into a single identity, or arrange them hierarchically. Each strategic choice has different implications for trust during the transition process. Although keeping identities physically isolated from one another in different locations may seem straightforward and allow lower-level managers to maintain the same level of trustworthy behavior, this would require adequate resources. The strategic benefits from mergers and acquisitions are often realized by creating economies of scale, which requires integration. Maintaining separate identities may drain resources, reduce the ability to achieve economies of scale, and result in political battles over resources. Thus, conflicting vectors may arise from top management targets, finance, and other competing departments, such that leaders may be less able to fulfill promises and support employees, and thereby be perceived as less trustworthy. In contrast to separation, aggregation refers to the “hierarchical arrangement of identities may facilitate organizational action by determining which identities are most important” as well as strengthening ties among identities (Pratt & Foreman, 2000, p. 33). This often happens when senior leaders decide to merge groups within an organization into a larger unit. For example, in 2016, Cornell University announced the merger of three schools: the School of Hotel

268  Roger C. Mayer and Michele Williams

Administration, the Johnson Graduate School of Management, and the Dyson School of Applied Economics and Management, under the overarching College of Business (Foderaro, 2016). Cornell announced that each school would retain the core features of its own identity (Foderaro, 2016), but the joint identity of the College of Business would be primary in a hierarchical arrangement and some functions such as admissions would have closer ties across schools. Ideally, aggregation leads to higher trust in all leaders and in the organization by aligning leaders’ actions with a primary identity. However, when aggregation is met with resistance, loss of trust and suspicion in the top leaders may result, which can weaken cooperation (Opperman, 2017), and political infighting can generate new vectors in addition to those from the initial organizational directives to merge and cooperate. As a result, non-primary unit managers may be left with fewer resources to fulfill promises and less ability to protect lower-level managers and employees from less advantageous policies. These non-primary unit managers may then be perceived as less trustworthy, having less integrity (willingness to follow through on promises), less benevolence (care for protecting well-being), and/or less ability (effectiveness). Alternatively, when the affected employees view the causes for the disruptions as organizational, their trust in their direct managers may not be impacted, while senior managers and the organization may be perceived as less trustworthy for not protecting the well-being of affected employees. Finally, integration occurs when top managers work to merge multiple identities into a distinct new whole (Pratt & Foreman, 2000). For example, in 1996, Beth Israel Hospital and neighboring New England Deaconess combined to form Beth Israel Deaconess in Boston. The goal was to integrate the identities of the nurses across the two hospitals, integrating the “primary nursing” model and professional identity of Beth Israel nurses, which granted nurses more autonomy and professional expertise, with the efficient, cost-cutting identity that Deaconess nurses held (Weinberg, 2003). Jointly the new identity would reflect high-quality, cost-effective nursing. When such integration is successful, trust in the organization and in frontline managers is increased. In our example, competing demands on nurse managers would be reduced and vectors would be aligned, allowing them to more easily fulfill their promises and support their employees. However, in practice integration can and, in our example, did turn into a political battle with one identity dominating and generating new vectors aimed at subordinating the other identity such that one organizational vector espoused an integrated identity, but in reality, a battle to arrange identities hierarchically was occurring (Weinberg, 2003) with conflicting vectors arising from new directions. When integration does not succeed, managers of the groups with subordinated identities may end up with fewer resources to fulfill promises and less ability to protect employees from less advantageous policies.These managers’ actions may then be perceived as less trustworthy, having less integrity (willingness to follow through on promises) and less benevolence (care for protecting well-being). Even in the limited cases when employees understand the organization-level constraints on their managers and

Multilevel Trust in Uncertain Contexts  269

thus maintain their beliefs in his or her benevolence and integrity, perceptions of managers’ ability to lead effectively may be impugned, also lowering perceptions of their trustworthiness.

Divestitures Economic pressure, a new strategic direction, or strategic opportunities may lead an organization to sell one or more of its units. As a result, the organization may lose a part of its uniform organizational identity or one of its subordinate identities (Pratt & Foreman, 2000). In some cases, a new strategic direction will intersect with regulatory pressure or requirements for divestiture. For example, to secure a merger and become Collins Aerospace, Rockwell Collins was required to sell business units related to two key aircraft safety features: (1) “[p]neumatic ice protection systems [that] remove ice from the wing of an aircraft by means of an inflatable rubber de-icing boot,” and (2) the “trimmable horizontal stabilizer actuators (THSAs) business. THSAs ensure that an aircraft maintains altitude during flight by adjusting the angle of the horizontal tail surface” (Friestad, 2019; United States Justice Department, 2018). Although this decision may lead to a more uniform organizational identity in the future, during the transition, trust in lower-level leaders of deleted units or those aligned with deleted identities may have been affected. If leaders from eliminated units are reassigned but retain old values, they may be perceived as less trustworthy because they may experience role conflict and make conflicting demands on subordinates or execute decisions that are inconsistent with the company’s remaining values and strategic direction. The leaders’ internalized values from their old units may partially or completely survive the transition, bringing forth inconsistent vectors from them. Thus, these leaders may both experience conflicting vectors, and also be the source of conflicting vectors influencing the supervisors reporting to them. Moreover, as these leaders are absorbed into other units as intact teams, the supervisors they lead may be perceived as less trustworthy because those supervisors may be unable to fulfill promises to subordinates or protect them from career harm in this new context. Employees may also experience psychological contract violations, thus lowering their trust in the organization.

Shadow of Past ODID Contexts We turn now to consider some individual factors that influence whether or not someone experiences a complex organization as being ODID. Within the same organization, different lower-level employees may be more or less equipped to cope with the strain that can arise from attempting to respond in appropriate ways to the multiplex vectors of expectations and demands. Likewise, an individual employee may pay greater or lesser attention to the organizational circumstances that give rise to these inconsistent vectors.

270  Roger C. Mayer and Michele Williams

It is well established that a person’s propensity to trust others affects the level of trust the person has in a specific other party (Fulmer & Gelfand, 2012; Jones & Shah, 2016; Mayer et al., 1995; Patent & Searle, 2019; Rotter, 1967). This has been described as akin to a personality trait, which presumably develops early in life and then continues to be relatively stable over time. Propensity is particularly important to trust judgments when the relationship is relatively new, as the trustor lacks an experience base with the particular trustee (Mayer et al., 1995). An employee has typically had experience with former managers, either within the same employer or at prior places of employment. These experiences with other managers have an impact on how the employee sees others who are in similar leadership roles. Unlike propensity (which is trust of others in general), they represent a category in the employee’s mind (see, for example, Foti, Fraser, & Lord, 1982). Experiences with others in the same category can be reasonably expected to act as a conditioned stimulus to the employee (Mayer, 2007), possibly evoking cognitive and emotional responses from the employee simply by virtue of being in the category of ‘organizational leader’ or ‘manager’ or ‘supervisor.’ These experiences may, of course, be positive or negative. Both past experiences and propensity to trust contribute to employees’ trust in managers by influencing how employees interpret the actions of their managers. This influence is seen most early in a relationship when, as Jones and Shah (2016) found, a large portion of the variance in trust comes from the trustor rather than the trustee or their relationship. When past experiences with managers involve an organization that was perceived to have ODID, the experiences are likely to promote angst because of the difficulties experienced with responding effectively. Such past experiences are likely to promote unwarranted responses to a new manager, including suspicion (Bobko, Barelka, Hirschfield, & Lyons, 2014), withdrawal (e.g., Eder & Eisenberger, 2007), or defensiveness (e.g., Ashforth & Lee, 1990). From the perspective of a new manager who seeks to build a relationship with the subordinate, and who lacks knowledge of the spectrum of the employee’s prior experiences, these reactions may appear irrational and counterproductive. While not an exhaustive list, propensity to trust and experience with other organizational managers are both examples of factors within the individual that will affect the perceived trustworthiness of the focal manager. Our intent here is not to provide a complete list of individual factors for this model, but to be mindful of the fact that a number of factors are already known to influence how much a follower trusts a manager outside of the top-down organizational issues considered earlier in this paper.

Conclusion and Future Research Directions We have discussed in this chapter a number of factors that make an employee’s work life more complex. Over the last half-century, much has been learned about

Multilevel Trust in Uncertain Contexts  271

the effects of factors such as crisis and change. Pratt and Corley, for example, consider “psychological outcomes (benefit or harm)” as the outcome of their Organizational Identities (OI) model (2007, p. 105). Less is known, however, about how these challenges accumulate or interact to affect how the employee trusts a manager or how the employee trusts the organization. We have argued that managers not only experience vectors pulling in various directions, but also must try to establish relationships with subordinates who are evaluating that manager’s trustworthiness. Enough various vectors in inconsistent directions, particularly if they are strong vectors, can lead to a manager’s perception that the organization has ODID, and moreover, to his or her subordinates’ perceptions that the supervisor is not behaving in a trustworthy fashion. The conflicting vectors associated with perceptions of ODID can affect trust either in the supervisor, the organization, or both, as posited above. Examination of ODID opens new areas of research. By examining the potential influences of the context in which managers and supervisors work, our framework suggests that research on trust in managers needs to capture organizational dynamics such as a large organizational change initiatives, crises, and turnover above the focal manager, especially among strategic decision-makers. Because organizational scholars often collect data at multiple time points, including measures of these events and the focal managers’ perceptions of ODID may provide a richer multilevel understanding of the relationship between trust in one’s direct leader, trust in the organization, and relevant work outcomes. Successfully carrying out this type of multilevel research will require the development of new validated measures. Measures that capture the perception of ODID, as well as organization change and crises, will be needed. During the COVID-19 pandemic, organizational researchers came up with new and innovative methods of capturing variance in turmoil and disruption across organizations, but validated measures will be important for identifying conditions associated with how strongly employees perceive that the organization has ODID, and for pinpointing the impact of managers’ and employees’ perceptions of ODID. Our framework suggests the importance of both looking at how organizational and broader socio-political phenomena influence trust in managers and also how employees’ historic interactions with people in the managerial role influence their perceptions of and responses to ODID. Researchers can ask new questions, for instance, whether employees with previous ODID experiences are more forgiving of the managers and less of the organization, or vice versa. When things go wrong, these employees may hold expectations for trustworthy behavior by supervisors that are lower, or alternatively, their ascriptions of supervisors’ control may be lower. We argue that understanding the potentially dynamic nature of trust in one’s immediate supervisor requires that researchers examine the context surrounding the supervisor and the historic contexts surrounding the employee. Hackman referred to this process as bracketing (Hackman, 2003). We contend that bracketing trust both ‘up a level’ with the broader context above

272  Roger C. Mayer and Michele Williams

the focal manager that the employee is evaluating as well as ‘down a level’ with the idiosyncratic context in the employee’s past is an underutilized method for understanding trust in leaders and managers. Moreover, longitudinal research examining trust in leaders and its change over time will be important to better understand how the impact of vectors shifts over time, and how the two levels of trust are impacted relative to one another as these shifts occur. For example, researchers might ask,“As an organization stabilizes after a change or crisis, does trust return to its previous level or does a trust differentiation occur between the supervisor and new employees versus the supervisor and those with longer tenure?” In addition to investigating trust’s downstream effects on in-role and extra-role performance, examining supervisor stress and well-being will help us disentangle the direct and indirect effects of perceptions of ODID over time. Because ODID can also be expected to cause other negative outcomes for the supervisor and potentially maladaptive responses to the strain, mediators such as resultant stress and beneficial or maladaptive coping responses may be uncovered as indirect paths between ODID and trust in the direct supervisor and/or the organization. This is not meant to be an inclusive list, but such mediators of ODID may include counterproductive work behavior (Kelloway, Francis, Prosser, & Cameron, 2010) like avoidance behaviors, attempts at revenge (Bies & Tripp, 1996), sabotage, and even violence. Organizations that wish to avoid such maladaptive supervisor behavior would do well to carefully study their internal environments both for evidence of the causes described in this chapter that can give rise to the perception of inconsistent vectors experienced by employees and managers and for reactions that may signal that they perceive the organization as having ODID. Questions to ask might include how clearly the mission is stated and made salient, and how clearly convinced employees are that the organization adheres to it in its decision-making. In this chapter, we advance understanding of the multilevel nature of trust processes in organizations by multiple levels of contextual factors and multiple layers of hierarchy that influence trust at two levels – trust in managers and trust in the organization. We have defined ODID as a continuum of perception by an employee or manager. We have considered a multilevel range of potential causes of this perception, including factors rarely examined as contexts for interpersonal trust such as broad external environmental shifts like a pandemic, broad organizational strategic changes in direction or mergers and divestitures, internal pandemonium caused by changes in middle managers and exertion of power by managers with self-serving agendas, and factors internal to the perceiver of ODID like experiences in prior organizations. Beyond the scope of this chapter, and we believe worthy of investigation, is how causal factors arising from these multiple levels of analysis interact to produce a greater perception of ODID. During the period of time when this chapter was written, a global pandemic spread, changing the landscape of the working world. At the time of writing, it

Multilevel Trust in Uncertain Contexts  273

was not clear how the changes would continue to spread, or to what extent they would become permanent. How will these changes, spawned from the environment outside of organizations, interact with factors within an organization and the within-employee factors as suggested above to increase employee perceptions of ODID? At least two very different possibilities exist: (1) employees may rationalize that there was nothing the organization or given managers could do and just accept the changes and whatever negative outcomes; (2) the organizational reactions to the change in the external environment could exacerbate any perceptions of ODID that may have existed previously.This is one example of multilevel interactions that hold promise to expand our understanding of organizational trust.

References Albert, S., & Whetten, D. A. (1985). Organizational identity. Research in Organizational Behavior, 7, 263–295. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington,VA: Author. Ashford, S. J., Lee, C., & Bobko, P. (1989). Content, cause, and consequences of job insecurity: A theory-based measure and substantive test. Academy of Management Journal, 32(4), 803–829. https://doi​.org​/10​.5465​/256569 Ashforth, B. E., & Lee, R. T. (1990). Defensive behavior in organizations: A preliminary model. Human Relations, 43, 621–648. https​:/​/do​​i​.org​​/10​.1​​177​/0​​01872​​6790​0​​43007​​02 Associated Press (2020, April 2). SAS airline employees train to assist in overwhelmed nursing homes, hospitals amid Coronavirus. USA Today. Retrieved from https​:/​/ ww​​w​.usa​​today​​.com/​​story​​/trav​​el​/ai​​rline​​-news​​/2020​​/04​/0​​2​/cor​​onavi​​r us​-s​​as​-ai​​rline​​empl​​oyees​​-trai​​n​-ass​​ist​-n​​ursin​​g​​-hom​​es​-ho​​spita​​ls​/51​​10477​​002/ Bies, R. J., & Tripp, T. M. (1996). Beyond distrust: “Getting even” and the need for revenge. In R. Kramer & T. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 246–260). Thousand Oaks, CA: SAGE Publications. https​:/​/do​​i​.org​​/10​.4​​135​/9​​78145​​ 224​36​​10​.n1​2 Bobko, P., Barelka, A. J., Hirshfield, L. M., & Lyons, J. B. (2014). The construct of suspicion and how it can benefit theories and models in organizational science. Journal of Business and Psychology, 26, 335–342. https​:/​/do​​i​.org​​/10​.1​​007​/s​​10869​​-01​4-​​9360-​y Burke, C. S., Sims, D. E., Lazzara, E. H., & Salas, E. (2007). Trust in leadership: A multilevel review and integration. The Leadership Quarterly, 18(6), 606–632. https​:/​/do​​i​.org​​/ 10​.1​​016​/j​​.leaq​​ua​.20​​​07​.09​​.006 Burt, R. S., & Knez, M. (1996). Trust and third-party gossip. In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 68–89). Thousand Oaks, CA: SAGE Publications. Colquitt, J. A., Scott, B. A., & LePine, J. A. (2007).Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 92(4), 909–927. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​. 4​.9​​09 Davis, J. H., Schoorman, F. D., Mayer, R. C., & Tan, H. H. (2000). The trusted general manager and business unit performance: Empirical evidence of a competitive advantage. Strategic Management Journal, 21(5), 563–576. https​:/​/do​​i​.org​​/10​.​1​​002/(​SICI)​1097-​ 0266(​20000​5)21:​53.0.CO;2-0

274  Roger C. Mayer and Michele Williams

Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology, 87(4), 611–628. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​7​.4​.6​​11 Eder, P., & Eisenberger, R. (2007). Perceived organizational support: Reducing the negative influence of coworker withdrawal behavior. Journal of Management, 34, 55–68. https:// doi​.org​/10​.1177​/0149206307309259 Ferrin, D. L., Dirks, K. T., & Shah, P. P. (2006). Direct and indirect effects of third-party relationships on interpersonal trust. Journal of Applied Psychology, 91(4), 870–883. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​1​.4​.8​​70 Foderaro, L. W. (2016, February 16). Cornell’s plan to merge hotel school gets an icy reception. New York Times. Retrieved from https​:/​/ww​​w​.nyt​​imes.​​com​/2​​016​/0​​2​/16/​​ nyreg​​ion​/a​​t​-cor​​nells​​-hote​​l​-sch​​ool​-a​​n​-icy​​-rece​​ption​​-for​-​​a​-pla​​nned-​​merge​​r​.htm​l Foti, R. J., Fraser, S. L., & Lord, R. G. (1982). Effects of leadership labels and prototypes on perceptions of political leaders. Journal of Applied Psychology, 67(3), 326–333. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.6​​7​.3​.3​​26 Friestad, T. (2019, October 29). Regulators ask Collins Aerospace to divest military GPS business in cedar rapids. Gazette. Retrieved from https​:/​/ww​​w​.the​​gazet​​te​.co​​m​/sub​​ject/​​ news/​​busin​​ess​/r​​egula​​tors-​​ask​-c​​ollin​​s​-aer​​ospac​​e​-to-​​dives​​t​-nav​​igati​​on​-bu​​sines​​​s​-in-​ ​cedar​​-rapi​​ds​-20​​19102​8 Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi​. org​/10​.1177​/0149206312439327 Glynn, M. A. (2000).When cymbals become symbols: Conflict over organizational identity within a symphony orchestra. Organization Science, 11(3), 285–298. https​:/​/do​​i​.org​​/ 10​.1​​287​/o​​rsc​.1​​1​.3​​.2​​85​.12​​496 Gryta, T. (2020, May 4). GE moves cut roughly 13,000 aviation workers. Wall Street Journal. Retrieved from https​:/​/ww​​w​.wsj​​.com/​​artic​​les​/g​​e​-mov​​es​-to​​-cut-​​rough​​ly​-13​​-000-​​ aviat​​ion​-w​​​orker​​s​-115​​88601​​106 Hackman, J. R. (2003). Learning more by crossing levels: Evidence from airplanes, hospitals, and orchestras. Journal of Organizational Behavior: The International Journal of Industrial, Occupational and Organizational Psychology and Behavior, 24(8), 905–922. https://doi​. org​/10​.1002​/job​.226 Hosmer, L. T. (1995). Trust: The connecting link between organizational theory and philosophical ethics. Academy of Management Review, 20(2), 379–403. https://doi​.org​/ 10​.2307​/258851 Jones, S. L., & Shah, P. P. (2016). Diagnosing the locus of trust: A temporal perspective for trustor, trustee, and dyadic influences on perceived trustworthiness. Journal of Applied Psychology, 101(3), 392. https://doi​.org​/10​.1037​/apl0000041 Kelloway, E. K., Francis, L., Prosser, M., & Cameron, J. E. (2010). Counterproductive work behavior as protest. Human Resource Management Review, 20(1), 18–25. https​:/​/do​​i​.org​​/ 10​.1​​016​/j​​.hrmr​​.200​9​​.03​.0​​14 Kramer, R. M., & Tyler, T. R. (1996). Trust in organizations: Frontiers of theory and research. Thousand Oaks, CA: SAGE Publications. Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C. (2005). Consequences of individual’s fit at work: A meta-analysis of person-job, person-organization, persongroup, and person-supervisor fit. Personnel Psychology, 58, 281–342. https​:/​/do​​i​.org​​/ 10​.1​​111​/j​​.1744​​-6570​​.200​5​​.0067​​2.x Levinson, H. (1965). Reciprocation: The relationship between man and organization. Administrative Science Quarterly, 9, 370–390. https://doi​.org​/10​.2307​/2391032

Multilevel Trust in Uncertain Contexts  275

Lewicki, R. J., & Bunker, B. B. (1996). Developing and maintaining trust in work relationships. In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 114–139). Thousand Oaks, CA: SAGE Publications. Lewin, K. (1951). Frontiers in group dynamics. In D. Cartwright (Ed.), Field theory in social science: Selected theoretical papers by Kurt Lewin, Dorwin Cartwright (pp. 188–237). New York, NY: Harper Torchbooks. Mayer, R. C. (2007). Employee loss of trust in management: Surviving in a new era. In J. Langan-Fox, C. Cooper, & R. Klimoski (Eds.), Research companion to the dysfunctional workplace: Management challenges and symptoms (pp. 125–135). Cheltenham: Edward Elgar Publishing Ltd. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi​.org​/10​.2307​/258792 Mayer, R. C., & Gavin, M. B. (2005). Trust in management and performance: Who minds the shop while the employees watch the boss? Academy of Management Journal, 48(5), 874–888. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​05​​.18​​80392​8 McAllister, D. J. (1995). Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https:// doi​.org​/10​.2307​/256727 Morrison, E. W., & Robinson, S. L. (1997). When employees feel betrayed: A model of how psychological contract violation develops. Academy of Management Review, 22(1), 226–256. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​97​.​97​​07180​​265 Nathan-Kazis, J. (2020, March 23). CVS Will Hire 50,000, Suggesting the Coronavirus pandemic hasn’t hurt sales. Barron’s. Retrieved from https​:/​/ww​​w​.bar​​rons.​​com​/a​​rticl​​es​/ cv​​s​-hir​​ing​-c​​orona​​virus​​-pand​​emic-​​sales​​-furl​​oughe​​d​-wor​​kers-​​aetna​​-hilt​​​on​-ma​​r riot​​t​515​​84973​​696 Opperman, M. P. (2017). Change in a loosely coupled organization –Two case studies (Unpublished Masters’ Thesis). Cornell University. https://doi​.org​/10​.7298​/X48K775D Patent, V., & Searle, R. H. (2019). Qualitative meta-analysis of propensity to trust measurement. Journal of Trust Research, 9(2), 136–163. https​:/​/do​​i​.org​​/10​.1​​080​/ 2​​15155​​81​.20​​​19​.16​​75074​ Pratt, M. G., & Corley, K. (2007). Managing multiple organizational identities: On identity ambiguity, identity conflict, and members’ reactions. In C. A. Bartel, S. Blader, & A. Wrzesniewski (Eds.), Identity and the Modern Organization (pp. 99–118). Mahwah, NJ: Lawrence Erlbaum. Pratt, M. G., & Foreman, P. O. (2000). Classifying managerial responses to multiple organizational identities. Academy of Management Review, 25(1), 18–42. https://doi​ . org​/10​.2307​/259261 Pratt, M. G., & Rafaeli, A. (1997). Organizational dress as a symbol of multilayered social identities. Academy of Management Journal, 40(4), 862–898. https://doi​.org​/10​.2307​/ 256951 Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970). Role conflict and ambiguity in complex organizations. Administrative Science Quarterly, 150–163. https://doi​.org​/10​.2307​/ 2391486 Robinson, S. L., & Morrison, E. W. (2000). The development of psychological contract breach and violation: A longitudinal study. Journal of Organizational Behavior, 21(5), 525–546. https​:/​/do​​i​.org​​/10​.1​​002​/1​​​099​-1​​379(2​00008​)21:5​ 3.0.CO;2-T Robinson, S. L., & Rousseau, D. M. (1994). Violating the psychological contract: Not the exception but the norm. Journal of Organizational Behavior, 15(3), 245–259.

276  Roger C. Mayer and Michele Williams

Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35(4), 651–665. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-6494​​.1967​​​.tb01​​454.x​ Sallinger, J. (2020, April 2). Coronavirus outbreak: Some JBS meat packing employees in Greeley afraid to go to work. Retrieved from https​:/​/de​​nver.​​cbslo​​cal​.c​​om​/20​​20​/04​​/03​/ c​​orona​​virus​​-jbs-​​meat-​​packi​​ng​-​em​​ploye​​es​-af​​raid Schoorman, F. D., Mayer, R. C., & Davis, J. H. (1996). Including versus excluding ability from the definition of trust. Academy of Management Review, 21(2), 339. Schreiber, F. R. (1973). Sybil. Chicago: Regnery Publishing. Tomlinson, E. C., & Langlinais, L. A. (2021).The relationship between trust and attributions: A levels-of-analysis perspective. In N. Gillespie, A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. New York, NY: Routledge. Tomlinson, E. C., & Mayer, R. C. (2009). The role of causal attribution dimensions in trust repair. Academy of Management Review, 34(1), 85–104. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​ 09​​.35​​71329​1 United States Justice Department (2018). Justice department requires UTC to divest two aerospace businesses to proceed with acquisition of Rockwell Collins (No.s 18–1279). Retrieved from https​:/​/ww​​w​.jus​​tice.​​gov​/o​​pa​/pr​​/just​​ice​-d​​epart​​ment-​​requi​​res​-u​​tc​-di​​vest-​​two​a​​erosp​​ace​-b​​usine​​sses-​​proce​​​ed​-ac​​quisi​​tion-​​rockw​​ell Ward, H. (2020, April 15). Carta’s covid-19 layoff. Medium. Retrieved from https:// link​.medium​.com​/tEdxxaBgI5 Web, M. D. (2020). Dissociative identity disorder (multiple personality disorder). Retri­ eved from https​:/​/ww​​w​.web​​md​.co​​m​/men​​tal​-h​​ealth​​/diss​​ociat​​ive​-i​​denti​​ty​-di​​sorde​​r​mul​​tiple​​-pers​​​onali​​ty​-di​​sorde​​r​#2. Weinberg, D. B. (2003). Code green: Money-driven hospitals and the dismantling of nursing. Ithaca, NY: Cornell University Press. Weiner, B. 1986. An attributional model of motivation and emotion. New York: Springer-Verlag. Whetten, D. A. (2006). Albert and whetten revisited: Strengthening the concept of organizational identity. Journal of Management Inquiry, 15(3), 219–234. https://doi​.org​/ 10​.1177​/1056492606291200 Williams, M. (2001). In whom we trust: Group membership as an affective context for trust development. Academy of Management Review, 26(3), 377–396. https://doi​.org​/10​. 2307​/259183 Williams, M. (2007). Building genuine trust through interpersonal emotion management: A threat regulation model of trust and collaboration across boundaries. Academy of Management Review, 32(2), 595–621. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​07​​.24​​35186​7 Williams, M. (2016). Being trusted: How team generational age diversity promotes and undermines trust in cross-boundary relationships. Journal of Organizational Behavior, 37(3), 346–373. https://doi​.org​/10​.1002​/job​.2045 Williams, M., Belkin, L. Y., & Chen, C. C. (2020). Cognitive flexibility matters: The role of multilevel positive affect and cognitive flexibility in shaping victims’ cooperative and uncooperative behavioral responses to trust violations. Group & Organization Management, 45(2), 181–218. https://doi​.org​/10​.1177​/1059601120911224 Witten, S., & Kimball, K. (2020, April 12). Disney World will furlough 43,000 union workers during coronavirus shutdown. CNBC. Retrieved from https​:/​/ww​​w​.cnb​​c​. com​​ / 2020​​ / 04 ​ / 1​​ 2 ​ / dis​​ n ey ​ - w​​ o rld-​​ w ill-​​ f urlo​​ u gh ​ - 4​​ 3 000-​​ u nion​​ - work​​ e rs ​ - d​​ u ring​​ coro​​​navir​​us​-sh​​utdow​​n​.htm​l

12 MULTILEVEL TRUST AND HUMAN RESOURCE MANAGEMENT Rosalind Searle and Rami Al-Sharif

Introduction In 2017, the BBC – along with other UK organizations with over 250 staff members – was required to show its gender pay gap. As part of this, it published staff salaries revealing large pay discrepancies in a number of departments. The differences were particularly marked within news and current affairs, causing significant grievances from two long-standing employees. The first, Carrie Gracie, was one of the organization’s most senior journalists and the China Editor for BBC News. She decided to resign, accusing her employer of having a “secretive and illegal” pay culture. She later went on to win her court case against the corporation over gender pay inequality, receiving both an apology and a payout that she donated to charity. A second, journalist Sarah Montague, was a long-standing presenter on Radio 4’s flagship news and current affairs program, Today. She described the level of trust breach, stating that she was “incandescent with rage” to learn that her previously regarded “very good salary [£133,000] for the job she loved” was in fact only a fifth of that of her co-presenter John Humphrys. She described how she “felt a sap” for “subsidizing other people’s lifestyles,” and as a consequence, her professional confidence was undermined. For her, the transparency of the pay and contracts revealed that not only was she the lowest paid of all the presenters on the Today program, but that she was on a different employment contract to the others, creating life-long financial consequences. She decided to remain with the organization, but she negotiated a much better contract and moved to a different program. These contrasting examples give insights into the huge affective and insidious cognitive impacts of human resource management (HRM) on trust. These impacts often extend to the organization to alter intrapersonal trust relationships with supervisors and colleagues. DOI:10.4324/9780429449185-12

278  Rosalind Searle and Rami Al-Sharif

HRM is an area of scholarship that focuses on the policies and practices adopted by organizations to regulate their employees, and therefore enable the employing organization to achieve its goals (Storey, 2007). Synthesis of HRM research indicates the dominance of three distinct kinds of study: that focused on the content of HRM policies and practices (what), exploration of processes (how), and finally, context (why) (Farndale, McDonnell, Scholarios, & Wilkinson, 2020). ‘Content’ studies consider either separately, or in combination, various policies and practices that extend over the whole employee–employment relationship cycle (Searle & Skinner, 2011). This cycle commences with HRM policies concerned with the attraction, selection, and entry of new employees by including the recruitment process as well as induction and onboarding. It then proceeds with policies focused on development and training, with access to voice and industrial relations, and those concerned with performance and reward management, and finally, those concerned with exit. Over recent decades there has been growing consensus about the interdependency of these areas, the value of examining collective HRM systems, and their relative importance (Boon, Den Hartog, & Lepak, 2019). Indeed, Searle (2018) argues that initial policies regarding entry are likely to be far more significant than others, as they set employees’ expectations and initial levels of organizational trust (Robinson, 1996). Of more critical interest to trust scholars is the second type of HRM study, particularly that which advances understanding of how these policies and practices influence the attitudes and behaviors of employees (Sanders, Shipton, & Gomes, 2014). Trusting an employing organization is a component of psychological attachment, and critical in underpinning employees’ commitment and cooperation in delivering the organization’s objectives (Ng, 2015). Therefore, it is unsurprising that there has been a focus on understanding how HRM policies influence the trust of employees in either managers or the organization itself. We shall see that this is an important area for multilevel study. The final type of HRM research is that which examines HRM in different contexts, with growing recognition of the significant but under-explored role of context (Johns, 2018). For example, the study by Gustafsson et al. (2020) on preserving organizational trust during a crisis contrasted different approaches and their impacts.The study showed important similarities among those organizations that were able to maintain trust. The relationships between trust and HRM are multi-faceted and multilevel. However, to date, research has focused predominantly on a single level of analysis to examine the impact of HRM on employee trust in either an employing organization (Searle et al., 2011) or their supervisors (Whitener, Brodt, Korsgaard, & Werner, 1998). Significant progress has been made in identifying the critical factors and their outcomes at these separate levels. However, research that more comprehensively explores HRM as a dynamic influencer of different levels of trust is in its infancy: for example, how trust at the management level may be influenced by HRM policies at the organizational level, or what policies affect

Multilevel Trust and HRM  279

team dynamics at the team level, and the distinct reactions of employees at the individual level.Yet it is also important for HRM researchers and practitioners to understand at which level their policies can have the greatest impact on trust. For example, is it through the team level that trust in the more abstract organization is built, or from the organizational level down to the team? There are growing calls for the study of these more complex processes that include the examination of multilevel and contextual effects between trust and HRM (Fulmer, 2018). This chapter reviews prior multilevel research of HRM and trust to consider the current and novel developments that emerge from adopting a multilevel or multi-referent approach.1 We build on previous empirical and conceptual perspectives of trust to outline how a multilevel perspective allows a deeper examination of the dynamic and reciprocal influences, as well as the cross-level effects, of HRM on employees’ trust in their leaders, co-workers, and teams, and in the organization. Trust in these various referents subsequently affects performance and other organizational outcomes. To do this, we undertake a systematic literature review of multilevel research on HRM and trust. First, we outline the review process and screening criteria. Then, we synthesize the main findings. In our discussion, we extend beyond these multilevel papers to critique the current literature and identify important missed or new agendas for future research.We conclude with a brief take-home summary of what can be learned from a review of multilevel research on HRM and trust.

Review Process In undertaking this review of multifaceted trust and HRM, we found a clear dearth of attention on these phenomena. We searched for papers using various search terms including ‘trust*,’ ‘Human Resources,’ and ‘HRM,’ and then separate policies including ‘recruitment,’ ‘hiring,’ ‘selection,’ ‘performance appraisal,’ ‘training,’ ‘performance management,’ ‘promotion,’ and ‘career progression.’ We searched five different databases, including EBSCO, PsycARTICLES, Psychology and Behavioral Sciences Collection, PsycINFO, and Scopus. Our inclusion criteria pertained to articles that (1) incorporated both trust and HRM practices, and (2) examined different levels of analysis. We identified 15 conceptual and empirical papers published between 1995 and 2019 (see Tables 12.1 and 12.2), following the seminal work of Mayer, Davis, and Schoorman (1995). The resultant papers are organized into three themes: the conceptual foundation of trust and trustworthiness in HRM research; types and levels of trust; and components of HR policies.

1 We use the term multilevel research to refer to studies that examine trust in different referents (e.g., trust in a supervisor, a team, the organization) and/or at different levels of analysis (e.g., when the trustor is at the individual, group, or organizational level).

Outcome and mediator Trust in organization Trust in supervisor

Capell et al. (2018)

Cooper et al. Outcome and Mediator (2019) Social climate (trust as a dimension)

Antecedent HRM inclusion policies and practices

Outcome, mediator, and moderator Trust in organization Trust in supervisor

Capell et al. (2016)

Social information Antecedent processing Well-being HRM

Antecedent HRM inclusion policies and practices

Antecedent Commitment HRM Control HRM

Outcome Trust in employees Trust in leaders Trust in organizations (rules, systems)

Ahteela and Vanhala (2018)

Psychological contract

HR conceptualization

Trust conceptualization and Trust development referent

Author

TABLE 12.1 Summary of research examining trust and HRM

•• Trust in employees’ competence is positively related to commitment HRM and control HRM. •• Trust in employees’ benevolence is positively related to commitment HRM and control HRM. •• Trust in employees’ competence and benevolence is higher in organizations with commitment HRM than those with control HRM. •• Trust in organization mediates the relationship between HRM inclusion policies and practices and stigma disclosure. •• Trust in supervisor mediates the relationship between HRM inclusion policies and practices and stigma disclosure. •• Trust in the organization mediates the relationship between HRM inclusion policies and practices and stigma disclosure. •• Trust in the manager mediates the relationship between HRM inclusion policies and stigma disclosure. •• Well-being HRM is positively related to social climate. •• Social climate mediates the relationship between wellbeing HRM and employee resistance.

Key results

280  Rosalind Searle and Rami Al-Sharif

Social exchange

Social exchange

Antecedent Leadership HRM systems Outcome Engagement

Outcome Organizational trustworthiness Moderator Trust in co-workers Trust in supervisor Trust in organization

Gillespie and Dietz (2009) Ho and Astakhova (2018)

Eva et al. (2019)

Social exchange Antecedent Social information HRM as a contextual processing factor Social categorization Swift trust Trust transformation Outcome Social exchange Antecedent Trust in top management Leadership team members Trust in leaders (CEO)

Outcome and moderator Trust in teams Trust in team members

Costa et al. (2018)

(Continued)

•• CEO participative leadership is positively related to trust in top management team members. •• CEO participative leadership is positively related to affective and cognitive trust in CEO. •• Affective trust in CEO mediates the relationship between the CEO participative leadership and team performance. •• Trust in top team members mediates the relationship between CEO participative leadership and team performance. •• An organization’s system shapes employees’ perceptions of the organization’s trustworthiness and can contribute to failures and effective trust repair. •• Trust in organization moderates the relationship between obsessive passion and perceived P–O fit.

•• Trust in teams resides at multiple levels of analysis. •• Trust in teams is subject to factors across levels in organizations. •• Trust in teams impacts performance and other relevant outcomes at both the individual and team levels.

Multilevel Trust and HRM  281

Antecedent Emotional intelligence Antecedent Socialization

Social exchange

Social exchange

Nerstad et al. Mediator (2018) Feeling trusted by supervisor Rezvani et al. Outcome and mediator (2018) Trust in team van der Outcome and mediator Werff and Trust in co-worker Buckley Trust in supervisor (2014) Trust in organization

Social exchange Transformation

Antecedent Mastery climate

Social exchange

Nam and Lee Outcome and mediator (2018) Trust in organization (aggregated to the organizational level)

Antecedent High-commitment HRM

Antecedent Change-oriented HRM

Outcome Trust in management

Lee et al. (2019)

Social exchange

HR conceptualization

Trust conceptualization and Trust development referent

Author

TABLE 12.1 Continued

•• Change-oriented HRM is positively related to trust in management. •• Trust in management is positively related to proactive behavior. •• Trust in management mediates the relationship between change-oriented HRM and proactive behavior. •• High-commitment HRM is positively related to trust in organization. •• Trust in organization mediates the relationship between high-commitment HRM and affective commitment. •• Trust in organization mediates the relationship between high-commitment HRM and turnover intention. •• Felt trust from a supervisor at the individual level mediates the relationship between perceived mastery climate and knowledge sharing. •• Trust in team mediates the relationship between team emotional intelligence and individual performance. •• Socialization increases trust behaviors (reliance and disclosure to co-workers) over time. •• Propensity to trust is positively related to initial intentions to engage in trust behaviors.

Key results

282  Rosalind Searle and Rami Al-Sharif

Mediator Trust in management (organization)

Wong (2018) Outcome Trust in supervisor Trust in organization

Whitener (2001)

Psychological contract

Social exchange

Antecedent Leadership Job security

Moderator High-commitment HRM

•• Perceived organizational support is positively related to trust in management. •• Trust in management is positively related to employee commitment. •• Trust in management mediates the relationship between perceived organizational support and employee commitment. •• Perceived job security is positively related to trust in organization. •• Trust in organization is negatively related to turnover intention. •• Subordinate–supervisor guanxi is positively related to trust in supervisor. •• Trust in supervisor is positively related to trust in organization. •• Trust in supervisor is positively related to OCB.

Multilevel Trust and HRM  283

# of studies found Ahteela and Vanhala (2018) Capell et al. (2016) Capell et al. (2018) Cooper et al. (2019) Costa et al. (2018) Eva et al. (2019) Gillespie and Dietz (2009) Ho and Astakhova (2018) Lee et al. (2019) Nam and Lee (2018) Nerstad et al. (2018) Rezvani et al. (2018) van der Werff and Buckley (2014) Whitener (2001) Wong (2018)

Employment and job security

X

X

2

X

X X

4 X

Selection

HRM components

X

1

Socialization

Author

Communication X

X

2

Pay and reward/compensation X

X

X

X

4

Decision-making participation X X

X

X

5 X

X

X X

3

Performance appraisal

TABLE 12.2 Summary of HRM components examined in research on trust and HRM

Training and development X

X X X

X

6 X

Diversity and inclusion X X

2

Rotation X

2 X

Teamwork X X

2

Engagement X

1

Leadership X X

2

Emotional intelligence X

X

2

284  Rosalind Searle and Rami Al-Sharif

Multilevel Trust and HRM  285

Findings Foundations of Trust in HRM Research Defining Trust There have been significant conceptual strides across the entire field of trust to achieve broad areas of consensus. Researchers of trust and HRM have become more unified in their consensus of trust as a psychological state (Rousseau, Sitkin, Burt, & Camerer, 1998). Critically, this perspective on attitudinal trust is likely to be influenced by dispositional factors (Patent & Searle, 2019), and more significantly by expectations and perceptions, most notably from the psychological contract (Robinson, 1996). Trust definitions have largely converged on a willingness to accept vulnerability based on confident positive expectations of another (Lewicki & Bunker, 1996; McAllister, 1995). Attention, however, has mainly been confined to the confident positive expectations that arise from suspension or diminution of perceived vulnerability. This confidence is derived from beliefs that the other’s actions (either an organization or individual) will at best be beneficial, or at worst not harmful (Robinson, 1996); such beliefs can change following perceived breaches or violations (Gillespie & Dietz, 2009; Searle & Ball, 2004). There is ongoing debate about vulnerability and its importance to the salience of trust. For some, risk is a prerequisite for trust (Mayer et al., 1995; Rousseau et al., 1998), while others contend that the holding of positive expectations negates matters of vulnerability and risk (Lewicki, McAllister, & Bies, 1998). In an HRM context, both risk and vulnerability are germane concerns that remain ripe for further investigation.To that end, it may be important to distinguish between trust as a belief that involves an intention (i.e., the willingness to become vulnerable) as opposed to trust as a behavior in which such vulnerability and the inherent risk can be more concrete (Colquitt, Scott, & LePine, 2007; Nienaber, Hofeditz, & Romeike, 2015). An important distinction in most trust–HRM studies is the inclusion of trust as a discrete variable in its own right, typically an outcome or a mediator and/or moderator, examined at different levels (see next section). However, occasionally trust is included within a bundle of items measuring a ‘social climate’ in which it is defined as a collective set of values, norms, and beliefs (Cooper, Wang, Bartam, & Cooke, 2019), yet it is measured alongside other related concepts including cooperation and shared codes and language.

Dimensions of Trust and Trustworthiness Prior empirical research has found HRM policies and practices are the foundations of trust, providing clues about the trustworthiness of another party (Searle et al., 2011).There has been some debate about the dimensions of trustworthiness

286  Rosalind Searle and Rami Al-Sharif

(McEvily & Tortoriello, 2011), with the ability, benevolence, and integrity (ABI) dichotomy dominating the field (Mayer et al., 1995). There is support for these three categories from single-level studies, revealing relationships between them and distinct justice elements (Colquitt & Rodell, 2011). Our review shows that multilevel studies can include different trustworthiness dimensions, such as exchanging reliability for integrity, and with greater granularity in the measurement of the interpersonal level, distinguishing between trust in other employees or in leaders, compared to that of a collective including the team and the more distal employing organization (e.g., Ahteela & Vanhala, 2018). At the organizational level, some scholars have suggested a quasi-isomorphism of trustworthiness dimensions operationalized at different levels by simply changing the trust referent in items (Ho & Astakhova, 2018). The resultant measures, perhaps unsurprisingly, show high inter-correlations. In contrast, others have argued that trust in an organization is different as it is impersonal, derived instead from more abstract elements including roles, systems, and reputation (Ahteela & Vanhala, 2018). For example, Gillespie and Dietz (2009) have conceptually argued for the relevance of ABI at the organizational level but noted that the nature and content of these dimensions, and the cues that influence them, differ markedly from the interpersonal level. There is support for the differentiation of trustworthiness in organizational referents compared to individual referents, with the more distal referent comprising two dimensions – ability, along with a combined benevolence and integrity (Searle et al., 2011). Other studies suggest more significant differences in the form of distinct antecedents between the two referents, with trustworthiness important for trust in a supervisor, while justice is more significant for trust in an organization (Tan & Tan, 2000). There is some need to explore these conceptual and measurement differentials, and multilevel research of HRM provides an important context to explore trust in these different referents. An important alternative conceptualization of the foundations of trust that is of particular value to the study of HRM is the differentiation of cognitive- from affect-based trust (McAllister, 1995). Cognitive-based trust concerns confidence derived from capability, professionalism, dependability, and reliability, and affectbased trust focuses on the emotional investment and perceived expressions of genuine care and concern (Luhmann, 1979). Empirical HRM individual-level studies using these dimensions reflect two distinct processes in operation, with cognitive-based trust reducing uncertainty, while affect-based trust deepens employee–supervisor relationships (Colquitt, LePine, Piccolo, Zapata, & Rich, 2012).This approach has further application for dynamic multilevel research, with McAllister (1995) arguing that cognitive-based trust precedes affect-based trust. This dichotomy might be useful in discerning whether their relative weight fluctuates over the employee–employment cycle. For example, cognitive-based trust might be more salient to safety matters, while affect-based trust has more value to downsizing policy.

Multilevel Trust and HRM  287

Trust and Distrust A further conceptualization that is particularly germane for HRM research concerns the more negative form of trust. Considerable attention has been directed towards whether trust is a bipolar dimension comprising either low trust or distrust, or whether distrust is actually something quite distinct (Lewicki et al., 1998). Earlier scholarship has often muddied any distinction by treating low trust and distrust as synonymous. However, although both trust and distrust refer to expectations of future action, distrust is materially different not only in the associated beliefs about the other party, but in the future decisions and actions that arise in order to reduce further vulnerability and mitigate exposure to the anticipated harm (BijlsmaFrankema, Sitkin, & Weibel, 2015; Saunders, Dietz, & Thornhill, 2014). Evidence confirms the distinction between trust and distrust; however, a more interesting question for HRM is whether trust and distrust co-exist (Benamati, Serva, & Fuller, 2010; Saunders et al., 2014; Vlaar, Van den Bosch, & Volberda, 2007). None of the multilevel papers we found includes distrust, an omission we will discuss further.

Felt Trust Prior research has examined trust in individual, group, and organizational referents. This perspective makes sense, as HRM practices are designed to increase employees’ trust in an organization or its agents. However, it has overlooked a potentially more significant trust concept for HRM study – the impact of being trusted. This is relevant given HRM policies are designed to increase self-efficacy, autonomy, performance, and skill, or to improve engagement and empowerment, and therefore they should lead to an increased feeling that the employee is trusted. Felt trust is distinct from other forms of trust as it incorporates a sense of obligation (Salamon & Robinson, 2008). Comparative multilevel research has found that in mastery climates, although both individuals and teams felt trusted by their supervisors, it was at the individual level where significant enhancement of knowledge sharing was found (Nerstad et al., 2018; for further discussion, see Zhu, Lau, and Lam’s 2021 chapter 6, this volume). An individual-level moderated mediation study showed that trust in the supervisor and feeling trusted interacted to reduce employees’ turnover intentions through reducing their workplace uncertainty, while feeling trusted by the supervisor enhanced their engagement levels (Skiba & Wildman, 2019). These distinct but complementary measures can be important to include in multilevel HRM research, enabling the testing of the critical process pathways through which HRM policies and practices might alter specific employee perceptions and behaviors.

Types of Trust and Levels As noted earlier, a multilevel examination of trust allows a distinction between different trust referents: trust in specific individuals (e.g., colleagues, supervisors)

288  Rosalind Searle and Rami Al-Sharif

or a collective such as team members, and trust in an employing organization (see Table 12.1). It enables simultaneous assessment of HRM regarding distinct interpersonal relationships, including co-workers and team members (Ho & Astakhova, 2018), and top management team members and Chief Executive Officers (CEOs) (Eva, Newman, Miao, Cooper, & Herbert, 2019). Multilevel research has revealed positive associations of trust at the team level with performance, innovation, creativity, and information-processing (Costa, Fulmer, & Anderson, 2018; Eva et al., 2019; Lee, Pak, Kim, & Li, 2019; Rezvani, Khosravi, & Ashkanasy, 2018), while for individuals it can enhance performance, proactivity, job engagement, and knowledge sharing (Cooper et al., 2019; Ho & Astakhova, 2018; Lee et al., 2019; Nerstad et al., 2018). The inclusion of trust in the employing organization offers the possibility of more complex research designs that enable cross-level exploration and the emergence of a more precise understanding regarding which HRM processes are important to which referents of trust (the content form of HRM study) and how they can create different outcomes (the process form of study) at distinct levels (Fulmer, 2018). For example, examination of the relationships and consequences of trust in supervisors and the organization identified how trust in the supervisors was critical for trust in an employing organization. It also revealed distinct outcomes, with trust in the organization reducing turnover intentions, while trust in the supervisor increased citizenship behaviors (Wong, 2018). Further, these types of studies have given an insight into the dynamics of trust by showing how cues change over time (van der Werff & Buckley, 2014). Other studies have explored the distinct mediating role of trust in these different referents; for example, identifying generic positive benefits of trust between team members, while trust in the team improved both individuals’ and the team’s performance (Rezvani et al., 2018). Other studies have shown different engagement pathways with high trust in the organization improving organizational engagement, while trust in coworkers and the supervisor increased job engagement (Ho & Astakhova, 2018). In contrast, Nam and Lee’s (2018) research showed high-commitment HRM practices (HCHRM) – systems designed to increase employee commitment – are important antecedents to organizational trust (aggregated to the organizational level), which in turn mediated the relationship between HCHRM and two individual-level outcomes (affective commitment and turnover intentions).

HRM Components HRM scholarship has identified different stakeholders involved in HRM policy, distinguishing conception and development from implementation and delivery (Guest & Conway, 2011). Important disparities can be found between leaders’ HRM visions, their operationalization by human resources employees, and their eventual delivery by supervisors (Searle & Dietz, 2012). In recognition of this multilevel complexity, Gillespie and Dietz’s (2009) conceptual framework

Multilevel Trust and HRM  289

of employee trust in their employing organization differentiates between HRM strategy, policies, and delivery. Cross-level studies of HRM and trust have dominated, with trust positioned as a consequence of HRM. This top-down exploration concerns consequences from the organization’s HRM for their employees. This research can be categorized into those focused on bundles of HRM practices or discrete single HRM practices (following Searle, 2018) (see Table 12.2).

HRM Policy Bundles Boon et al. (2019) argued the merit of recognizing HRM policies as a collective system. Some multilevel research has explored the impact of HCHRM; Whitener (2001) advanced insight into policy content choices in her HCHRM study that included selection, training, reward and compensation, and performance appraisal. At the individual level, she showed the partial mediation of trust in management and revealed more about the process (how) by which HCHRM operates. Her cross-level modeling showed that HCHRM positively influenced the relationship between perceived organizational support (POS) and trust in both management and organizational commitment. Such results confirmed the significance of interpersonal trust relationships with managers in implementing HRM, adding knowledge of beneficial processes (how). By contrast, Nam and Lee’s (2018) HCHMR study added further policies (content – what) to their bundle with enriched job design, teamwork, and decision-making participation. Their results again revealed a significant mediating process of organizational-level trust between HCHRM and two individual outcomes: affective commitment and turnover intention. However, the study did not show cooperative labor-management relations as a mediator in the relationship between HCHRM and trust in the organization. Another HRM bundle study compared distinct aspects of HRM content (what) by dichotomizing two distinct types of policies – control and commitment approaches (Ahteela & Vanhala, 2018). Control-focused HRM included outcome-based performance appraisal and structured job classification, whereas commitment-focused approaches comprised participative decision-making, job rotation and flexibility, behavior-focused performance appraisal, and training. Through their analysis they found that different levels of trust emerged, giving greater insight into the process pathways that are in operation. Specifically, commitment-focused HRM systems were found to have a stronger impact on employing organization trust and the trustworthiness of leaders, while controlfocused HRM systems positively influenced trust in the employing organization and the trustworthiness of co-workers. However, in the process of showing these distinct trust pathways, they argued there was further value to be gained in combining these HRM-focused practices. Lee et al. (2019) examined a different type of content, change-orientated HRM systems, showing the impact for two distinct levels of outcome: employees’ proactive behaviors and team-level innovation. Their two studies examined

290  Rosalind Searle and Rami Al-Sharif

and confirmed how such HRM systems could enhance individuals’ proactivity through changing who was selected, their training, and how they are motivated, and by providing specific opportunities for proactivity (through enhancing employee participation and job rotation provision). Critically, their work considered the central routes to increasing innovation, showing that by targeting efforts at individual employees their proactivity could be raised, and through this, team innovation improved. The role of trust in this study centered on a trust-in-management pathway and revealed it as an important mediator for individuals’ proactive behaviors. This study design has value not just in showing the critical content of HRM systems (what), but also in identifying the process pathways (how) through which such change can be achieved. HR professionals and researchers have gained fresh insight into the mechanism through which these HRM practices alter individuals’ attitudes. By attending to individuals rather than focusing on the group, they can deliver more creative groups. However, the study does have some limitations that can be common to such designs, most notably the simple aggregation of different HRM practices into an index, rather than by considering their distinctive consequences for trust in management, or for individuals’ proactive behavior. Using indexes can also hide conflicting processes; for example, motivation studies have revealed that extrinsic HR rewards can diminish intrinsic motivation (Gardner, Wright, & Moynihan, 2011) that other bundle elements of participation and job rotation were designed to promote. This raises the question about whether these policies might simply cancel each other out, or whether one effect is stronger than another. Boon et al. (2019) have underscored the value of more fine-grained designs that provide the means to assess the relative weights of these HR practices. Nor was Lee et al.’s (2019) work fully longitudinal, and so precise causality has had to be inferred through the combined results of their two distinct studies. It does suggest, however, that trust in management was an important factor through which to advance innovation in these organizations. Research has considered HRM policy bundles designed to promote well-being (Cooper et al., 2019) involving training and development, teamwork, information sharing, quality of work, quality of relationship with supervisor, and job security. The results showed that these HRM practices fostered a social climate of which trust was an element. However, closer scrutiny shows the measurement of trust was limited, with one question concerning interpersonal relations – “Employees in this organization have relationships based on trust and reciprocal faith” (Prieto & Pilar Pérez Santana, 2012). In this study, trust was a mediator in the relationship between these HR practices and employee resilience, which in turn enhanced employees’ job performance. This study identified both significant content for trust but also illuminated contexts (why) in which more positive individual and organizational outcomes occurred. However, its value to trust scholars is reduced by the study’s limited operationalization of trust and the lack of differentiation from other social climate factors.

Multilevel Trust and HRM  291

Single HRM Policies Distinct from aggregated HRM practices research, more defined contributions have been made to both HRM and trust scholarship through the examination of single HRM practices. These studies have contributed to our understanding of how distinct HRM practices can influence both trust in the organization and interpersonal-level trust, including trust in a supervisor or colleagues. Consideration of specific HRM content has advanced understanding of vulnerability, an area where further trust scholarship is required. Additionally, these studies have contributed knowledge of such trust both as a consequence and a mediator for other organizational outcomes. Further, they have been important to our insight into how trust processes operate, revealing the distinct pathways between, as well as within, distinct levels of trust. We start by reviewing studies examining trust in both the employing organization and interpersonal referents before considering those that focus on trust between different types of employees. There are three multilevel studies that provide insight into different and integrated roles of trust for different referents, notably in the organization and in supervisors. Collectively, these studies have increased our knowledge of the significance of interpersonal trust as a mediator for increasing trust in the organization by showing how such trust can have different outcomes. Wong’s (2018) study added insight into context and trust pathways by comparing the influence in China of the management practices and employment relations of state-owned and joint-venture enterprises (Wong, 2018). In these two distinct contexts, he dichotomized distinct inter-relationships and outcomes for trust in the organization and those for trust in a supervisor. In both contexts, an HR policy of job security significantly reduced intentions to quit and increased trust in the employing organization, and this trust, in turn, mediated the relationship between job security and turnover intentions. A second interpersonal trust pathway was found regarding supervisor–subordinate relations, termed ‘guanxi’ in this Chinese context, which was positively associated with increased trust in the supervisor.This interpersonal trust was a mediator between guanxi and trust in the organization. However, Wong found that this interpersonal trust was only associated with increased organizational citizenship behaviors in the joint ventures. In contrast, within state-owned enterprises, there was less citizenship behavior and it was linked to trust in the organization. Wong (2018) explained the reason for this difference in terms of ‘danwei’ which is a work unit with the powers to recruit, fire, and transfer employees, which has a more significant role for state enterprises. His study provides evidence of the distinct interactions between these different trust referents and of the distinct outcomes they produced, but more critically for both HR and trust scholars, it revealed the further complicated influences of context for these relationships (Johns, 2018). Employee engagement is an important HRM concern and has been examined in a study that considered whether interpersonal trust (in a colleague or supervisor)

292  Rosalind Searle and Rami Al-Sharif

or trust in the organization was a moderator in the relationship between two forms of job passion (obsessive and harmonious) and engagement (job or organizational) (Ho & Astakhova, 2018). The research considered the mediating role of fit by comparing perceived demands–abilities (D–A) and person–organization (P–O) fit. They found that only when trust in the organization was high did it become a significant positive moderator between obsessive passion and P–O fit, with posthoc analysis indicating an indirect impact where the two forms of interpersonal trust were high and formed a three-way interaction between obsessive passion and D–A fit. For both HRM and trust researchers, these results are significant in identifying the necessity of having interpersonal trust in both supervisors and colleagues in order to produce trust in the workplace. They consider how such trust can be built, for example, through distinct rewards including formal incentives for a supervisor’s or colleagues’ social support, and through the use of penalties such as formal warnings or ostracization. The results raise the significance of psychological safety for the flourishing of obsessively passionate workers, which requires multi-foci referents of interpersonal and organizational trust. Further, this study confirms the significance of high-trust workplaces as contexts that enable these important engagement mechanisms to operate. Capell et al. (2016, 2018) have expanded understanding of the significance of both trust in the organization and trust in supervisors for the successful operation of another HRM area, inclusion policies and practices. This work has conceptually and empirically advanced awareness of these two trust pathways, and revealed the complex moderation and mediation of trust in the organization, but also in the supervisor, to be central to employees’ decisions to disclose their concealable stigmas, such as religion, HIV status, and sexual orientation. They show how HRM inclusion policies and practices require these twin referents of organizational and interpersonal trust to be in place before employees are willing to make themselves potentially vulnerable by sharing such sensitive personal information. The four-wave longitudinal study by van der Werff and Buckley (2014) has advanced knowledge of interpersonal trust during a critical time for employees in terms of induction and socialization. Their work provides a more nuanced understanding of the dynamics of trust behaviors, showing shifts in the cues newcomers use as they get to know their work colleagues. Their latent growth model charts the dynamics to two behaviors that demonstrate interpersonal trust, intention to rely on a colleague, and the disclosure of information. The study shows the shifts in the trust cues from initial presumptive trust cues – which included three elements, rule-based trust, role-based trust, and group identification – to more personal trust cues, concerning their colleagues’ trustworthiness (ABI). Such work is valuable to HR professionals as it is concerned with a critical point in any employee cycle, the start of an employment relationship.They empirically showed when and how the basis for these trust behaviors changes with initial individual difference factors (trust propensity) and presumptive cues being replaced by more direct personal signals. The detailed longitudinal study revealed some non-linear

Multilevel Trust and HRM  293

dynamics to this interpersonal trust in their new colleagues, rapidly building on the newcomers’ trust predispositions to provide a period of stability in which information about two personal trust cues, competence and benevolence, are gathered and then replaced by presumptive cues.They showed how the third personal element, integrity, only became significant from newcomers’ third month and with regard to newcomers’ reliance behaviors. They found the influence of presumptive cues, critically role-based perceptions, was mediated by the personal trust characteristics, whereas rule-based perceptions only became significant to trust behavior reliance in the latter stages. Further, group identification was found to be important for early reliance behavior, but not for information disclosure, and had a subsequent significant and positive relationship with both trust behaviors. They offer important insights for trust scholars and HR professionals into how trust is built in new work relationships, with distinct cues linked to different aspects of vulnerability for newcomers. They confirmed the use of the sources of multiple information to help inform and shape newcomers’ interactions. Two other studies have explored trust between individuals. The first focused on the style of leaders using a four-wave multilevel and source study to examine the role of participative leadership style and top management team’s performance within the context of a new venture (Eva et al., 2019). This study dichotomized trust in the CEOs using McAllister’s (1995) cognitive- and affect-based approach and revealed affect-based trust alone was a critical mediator for the significant relationship between participative leaders and top management team’s performance. Further, they found intra-team trust was a significant mediator between this leadership style and the top team’s performance. Such insights can be of value to HR in the selection and development of new start-up leaders, but also in demonstrating the synergies between leaders’ behaviors and style and the building of trust within top teams. The last study in this area concerned emotional intelligence (EI) – argued to be an important foundation for trust, and which can be utilized by HR for selection and training. Rezvani et al. (2018) examined real project teams and found individual-level EI was associated with individuals’ job performance, while team-level EI had a positive relationship on the project team’s performance, but not the individual. While trust was a small component in this work, the findings confirmed trust in the team as a mediator for team-level EI regarding both individuals’ and team performance.These results empirically demonstrate the value of EI training and trust in a team for organizationally significant outcomes.

Discussion and New Research Agendas In addition to indicating important findings for HRM concerning distinct content, the significant trust processes and pathways, and particular context considerations, our review of current multilevel HRM and trust research reveals five omissions and shortcomings for these fields in relation to trust conceptualization,

294  Rosalind Searle and Rami Al-Sharif

breadth of HRM practices considered, diversity of social group examined, the external context of the organization, and the neglect of temporality and disruptive events.

Conceptualization of Trust Research into HRM and trust has been largely dominated by social exchange theory (Blau, 1964), as a progressive approach to trust. Significantly, this approach conceives trust as having a linear development, and so focuses on trust processes in terms of creation and building through reciprocal behaviors and experiences of the different parties. However, inherent to progressive approaches is their starting point, which has ranged from trust commencing from zero to a positive (i.e., trusting) or negative skew (i.e., distrusting) (Lewicki, Tomlinson, & Gillespie, 2006). Understanding the significance of this starting point for HRM is evident from research that has focused on breaches and shows different trajectories dependent on the initial trust level (Robinson & Rousseau, 1994). Social exchange theory can offer a nuanced perspective on trust reciprocity across various starting points and levels; however, it is crucial for scholarly research to consider whether other approaches might be of more value in this context. Social information processing theory (Salancik & Pfeffer, 1978) offers another less common approach to the progressive perspective, under which trust is regarded as developing over time in a more linear way through a sensemaking process of social networks and behaviors. This is an under-utilized conceptual approach with considerable merit for both HRM and trust studies. Limited multilevel research has adopted social information processing theory to explore the distinct forms of social information used by individuals to understand and adapt their attitudes, behaviors, and values to the various contexts or social environment in which they work. For example, the approach was useful in explaining how well-being–oriented HRM practices can generate positive social climates, and through such climates improve resilience and employee performance (Cooper et al., 2019). A different conceptualization of trust can also be found in discontinuous approaches, which do not view the development of trust as linear. Arguably, in a working context, individuals are likely to build ‘swift trust,’ a temporary form which HRM policies and practices are designed to promote and which encourages positive expectations during the critical early stages of an employment relationship (Meyerson, Weick, & Kramer, 1996). Understanding how swiftly trust can be built and stabilized has value for HRM policies, such as team-building interventions, or to support project or temporary teams, allowing talent to be retained and high performance to be achieved quickly. Social categorization theory is a further discontinuous approach that is significant in the study of identity, which shapes social-group formation and conflict (Tajfel & Turner, 1986; Turner, 1985). This approach suggests that individuals tend

Multilevel Trust and HRM  295

to classify themselves and others based on social categories, such as gender, ethnicity, religion, occupation, and organizational membership (Ashforth & Mael, 1989). These categories can often include assumptions about the congruence of values to these social groups, and this can be an important foundation for trust by reducing uncertainty and positive expectations (Kramer & Lewicki, 2010). In contrast, value incongruence has been shown to promote distrust by fostering suspicion or perceived conflict (Bijlsma-Frankema et al., 2015; Chambers & Melnyk, 2006; Sitkin & Roth, 1993). Despite its importance, especially to organizations, this perspective has received little attention from trust researchers. The theory has considerable merit for the exploration of the differential impact of HRM policies on different groups of employees, specifically those who feel they are negatively stereotyped or not from the majority – for instance, women and ethnic minority groups (Ghumman & Jackson, 2010;Van Laar, Derks, & Ellemers, 2013). A final approach to conceptualizing trust with an important opportunity for multilevel research pertains to transformational theories. For example, Lewicki and Bunker’s (1996) three-stage model of trust argues that individuals develop and transform their basis of trust as the relationships develop. Van der Werff and Buckley (2014) showed how the basis of trust was transformed in HRM socialization processes for new employees, leading to more identification-based trust. HRM can play a central role in a transformational context for many employees, with downsizing, labor unrest, or a pay dispute all being significant triggers for change.Yet, studies focusing on transformations and the role of HRM in processes that would result in shifts of trust into distrust, and vice versa, are scarce. Little attempt has been made to chart trust as processes or outcomes through these events. There are obvious challenges for researchers to design such studies, but it is important that we investigate transformations, particularly into distrust, which is argued to have a more enduring impact on employees.

Range of HRM Practices There are some clear omissions in HRM policies and practices that multilevel research examined in relation to trust. More critically, these gaps include regular and significant HRM practices, most notably, areas where trust is likely to be a notable concern, such as annual performance management reviews, reward and recognition, and change management. There is also little examination of HRM in trust breach and the role of HRM in trust repair. Gillespie and Dietz’s (2009) model has been tested at the organizational level (Gillespie, Dietz, & Lockey, 2014), but there is still work required to explore the consequences for different referents, such as whether trust is retained for some referents, say colleagues, but not the supervisor or the organization, and regarding different trust repair outcomes (see Gillespie and colleagues, 2021 chapter 7, this volume). As a result, important insights about whether trust can be restored at both supervisory and organizational levels are only partly addressed.

296  Rosalind Searle and Rami Al-Sharif

Multilevel study could extend the work of Capell et al. (2016, 2018) to illuminate the distinction between HRM policy content and its delivery through comparing trust in the organization with trust in the supervisor. For example, Searle and Skinner’s (2011) conceptual work on performance management has contended that, over time, trust in a good supervisor will become undermined by poor HRM systems. Scholars could add further measures to assess the consequences of poor HRM systems for workplace attitudes and behaviors of both the supervisor and the employee. Similarly, Gillespie and Dietz (2009) conceptually outlined the shifts towards distrust in breach, and research could empirically consider the impacts of HRM-led inquiry for trust repair and the consequences of more fine-grained differentiation between low trust and distrust. The omission of specific consideration of distrust, as opposed to trust, is a surprise given that HRM policies and practices have the potential to create and perpetuate distrust, rather than trust. Indeed, many innovation-led, financial, and security businesses include policies that are deliberately designed to create and maintain distrust (Siebert & Czarniawska, 2018). In addition, many HRM policies involve the formation of internal competition, which Sapegina and Weibel (2017) argue can have unintended consequences if it is based on unfair or opaque processes. For example, performance management systems often use forced distribution rating processes to annually remove the bottom-ranked 10% (Giumetti, Schroeder, & Switzer, 2015), which do not actually improve sustainable performance – instead causing harm even from short-term use (Mulligan & Bull Schaefer, 2011). These systems further undermine trust by tasking individuals to position outcomes as their own, rather than as the result of collective endeavors (Searle, Nienaber, Price, & Holtgrave, 2018). Through undermining the individual’s team identity, they reduce their identity with their employing organization.

Diversity Matters The design of many current studies aggregates employees’ perceptions and therefore fails to capture and consider perceptual divergence between employees. HRM and psychological studies have revealed that although employees might work in a team, individual difference factors can create important disparities in their personal experiences, and through this a likely variance in their levels of trust. The studies of Capell et al. (2016, 2018) considered concealable stigmas, but what about those more discernible differences between social groups? Studies have long shown inequity for HRM practices, particularly in recruitment and selection (Truxillo & Bauer, 1999), career progression (Wyatt & Silvester, 2015), and performance review even when delivered by the same supervisor (Byrne, Pitts, Wilson, & Steiner, 2012; Chory & Hubbell, 2008). More nuanced scrutiny confirms that although the volume of performance feedback can be similar for employees, the content given to ethnic minority-group members is less precise and actionable (Wyatt & Silvester, 2015). Meta-analytic study has confirmed significant gender

Multilevel Trust and HRM  297

and racial disparities in workplace treatment, but also that changes designed to make HRM systems fairer can often result in backlash from less impacted white male employees (McCord, Joseph, Dhanani, & Beus, 2017). Similarly, Apfelbaum, Grunberg, Halevy, and Kang (2017) found differences between two HRM approaches towards racial discrimination – colorblindness versus multiculturalism. Extant research on HRM policies that are designed to favor individuals from stigmatized groups, such as affirmative action, shows they can have unintended consequences; some individuals regard them as a mechanism to end historical discrimination (Moscoso, García-Izquierdo, & Bastida, 2010), while others feel that these tools result in the tainting of individuals’ real achievements and create doubt about their competence (Heilman, Block, & Stathatos, 1997). It is clear that very different experiences are likely to arise for employees based on differences in their gender, ethnicity, religion, age, sexual orientation, status, and disability, yet little attention has considered these variations of experience (for an exception, see Al-Sharif, 2019). What are the consequences in terms of different trust referents of such experiential differences? Multilevel study is an important means to further explore the nested relations and identities that exist in organizations, offering ways of distinguishing individuals’ career identity from workgroup identity and organizational identification (Searle et al., 2018). Similarly, exploration of propensity to trust and the impact of such individual difference factors is in its infancy, while there is recognition that it has a more important role at the onset of the HR journey (van der Werff & Buckley, 2014). Are there significant differences in the initial level of trust for newcomers in ethnic minority or marginalized groups? Insight into the starting points and dynamics of multi-referent trust perceptions are especially important for those with stigmatized identities and could assist in designing safe-identity work environments (Kalokerinos, von Hippel, & Zacher, 2014; Wehrle, Klehe, Kira, & Zikic, 2018).

Context Matters There is re-emerging recognition of the significance of context in the study of HRM (Johns, 2018). Our review of current work did reveal some important insights regarding trust processes from the study of different internal organizational contexts (e.g., Chinese organizations in Wong, 2018). However, the external context of an employing organization can result in constraints to HRM policy choices, combining influences of cultural, political, or societal values with industry norms and national legal and regulatory frameworks in determining the type and flavor of the organization’s HRM and, through this, different potential consequences for trust. However, context has been neglected (for an exception, see Al-Sharif, 2019). A better understanding of contextual dimensions can be critical to the takeup of more sensitive HRM policies, such as diversity and inclusion. While they

298  Rosalind Searle and Rami Al-Sharif

may vary for legal and religious reasons within the same organization across national borders, there can be other powerful influencing factors. Awareness of these contextual variables can be crucial to understanding what organizations need to do to influence individuals’ attitudes and behaviors at work, with prior study of stigmatized staff underscoring the significance of both trust in the organization and in those delivering the policy to successful implementation (Capell et al., 2016, 2018). Social and cultural capital can influence the operation of personal connections at work to create distinct obligations between individuals. While study has included such phenomena in a Chinese context – guanxi (Wong, 2018) – what about similar cultural artifacts in the Middle East, such as wasta (Harbi,Thursfield, & Bright, 2017), or elsewhere? Understanding how these nuances can alter the delivery, and therefore, the perceived fairness of HRM processes would be important to multinationals, potentially affecting employees’ trust in organizations and in their different agents.

Temporality and Disruptive Events The study of HRM and trust has largely failed to examine either episodic or event levels (Morgeson, Mitchell, & Liu, 2015). The study of triggers or disruptive events is in its infancy in organization studies, yet multilevel research can offer an exciting means to examine diverse antecedents, dynamics, processes, and outcomes. As argued earlier, in many ways, HRM policies can be conceived of as particularly disruptive for employees, with episodes including recruitment, career progression, job insecurity and loss, or retirement. These can cause short-term, and more often, enduring consequences not just for the individual, but also for their work groups and employing organizations. There is, therefore, merit in the inclusion of episodic and temporality aspects (Cojuharenco, Fortin, & German, 2014) for future studies of HRM, such as through the inclusion of turning-point analysis (Bullis & Bach, 1989a, 1989b) to assist in the exploration of the key events and experiences that can shape and alter trust perceptions over time. Attention to such areas would enable a more detailed examination of contexts that could illuminate the important conditions and mechanisms for change to employees’ attitudes and behaviors, and therefore into “spirals of trust” (Korsgaard, 2018; Korsgaard, Brower, & Lester, 2014). By means of such research, evidence could be provided that would improve HRM policies’ capacity to maintain a positive psychological contract throughout the employment relationship for all their employees. We need to discern the seminal HRM experiences for employees that either build or demolish their trust – whether interpersonally or in their employing organization. More complex study designs, including multi-method and multilevel, would offer the potential to increase understanding of when, how, and why these negative experiences occur; to provide evidence to change organizations, but also to support the development of HR policies to enhance employee

Multilevel Trust and HRM  299

well-being, rather than to deplete human beings as they seek to navigate a dramatic and dynamic employment landscape. An important example of the value of multilevel work in revealing dynamics is that of van der Werff and Buckley (2014), who revealed the transitions of trust clues and behaviors for newcomers. Dynamic study of this kind has merit for identifying the distinct elements of HRM policy content and its delivery that can be made more effective to build, maintain, or restore trust for organizations, or to ameliorate or reduce its decline. Organizations are facing increasing demands to improve their inclusion and diversity, to be able to identify, develop, and retain talent, and to be more productive. Inherent to these matters are issues of trust that require a more nuanced understanding of knowledge about shared and distinct antecedents, of the mediating and moderating influences of the different levels, and of the dynamic of processes and different pathways.

Conclusion This review into specific multilevel trust and HRM studies has revealed their value to organizations and research and has provided three important contributions for HRM. First, it shows that HRM policies and practices have critical contents and are an important context in which to build trust (Costa et al., 2018; Gillespie & Dietz, 2009). In all of the current studies – various bundles or single practice studies – these policies and practices are examined as antecedents that increase trust in some referents and levels, whether in the organization or, most often, in a supervisor or management. Sadly, despite prior work demonstrating its significance for HRM leaders’ decisions to include promotion and training for their staff (Tzafrir, 2005), the role of trust as an influencer of HRM strategic content has not been considered. Further, none of the studies has considered HRM policy content in terms of distrust. Second, and more significantly, this review reveals a value to HRM scholarship in discerning and advancing knowledge of the critical HRM policy content to the building of trust, whether in management (e.g.,Whitener, 2001) or the organization (e.g., Wong, 2018). However, in addition to the content, multilevel studies often reveal the process (the how) by identifying the complex and significant roles of trust as a mediator and moderator in HRM policies delivering organizationally significant outcomes such as innovation (Rezvani et al., 2018), well-being and resilience (Cooper et al., 2019), or job and organizational engagement (Ho & Astakhova, 2018). The studies have illuminated the distinct trust mechanisms and their dynamics within newcomers as they are socialized into a new workplace (van der Werff & Buckley, 2014), or the critical leadership style for successful new ventures (Eva et al., 2019). They provide nuance to HR professionals in revealing the critical trust referent pathways, such as a focus on individual change to offer team-level outcomes (Rezvani et al., 2018), or for supervisors focusing on building their individual relationships within a team (Nerstad et al., 2018). The studies

300  Rosalind Searle and Rami Al-Sharif

also distinguish the distinct impacts of different trust referents (Ho & Astakhova, 2018), and the criticality of complementary referents of trust for more effective diversity and inclusion policies to be successful (Capell et al., 2016, 2018). Finally, through our critical reflection on the limitations of the current multilevel literature on HRM and trust, an agenda for future research is outlined. We began this paper by contrasting the experiences of female employees with those of their male and white counterparts. In the light of #MeToo and Black Lives Matter, more attention needs to be directed toward understanding the divergent experiences of distinct social groups within a workforce, and the role of, and implications for, trust as well as distrust at different levels and referents. For example, can trust in an organization remain if the supervisor is distrusted? What are the consequences of this distrust for the supervisor, but also for the organization? Examining such experiences is important for trust scholarship to develop and re-balance study of the other central element of trust – vulnerability. Do different employees have lower initial levels of trust and greater distrust in their new employer, and anticipate unfair workplace outcomes (e.g., performance appraisal, career progression, and reward)? What happens to individual performance and engagement if their concerns are fulfilled? Or where the opposite occurs? Are there different antecedents and dynamics of trust and distrust for those in disadvantaged and stigmatized groups? What can such studies show us about the consequences of these inconsistencies for organizational reputation, and also to create more effective HRM policies (Al-Sharif, 2019)? Multilevel attention is central to understanding how different social networks and processes operate within the same working context (Moscoso et al., 2010; Truxillo & Bauer, 1999).

References Ahteela, R., & Vanhala, M. (2018). HRM bundles and organizational trust. Knowledge and Process Management, 25(1), 3–11. https://doi​.org​/10​.1002​/kpm​.1561 Al-Sharif, R. (2019). The impact of stereotype threat on talent management: A multi-level study on (Dis)Trust/(In)Justice perspectives of Muslims in the UK banking sector [Unpublished doctoral dissertation]. University of Glasgow. Apfelbaum, E. P., Grunberg, R., Halevy, N., & Kang, S. (2017). From ignorance to intolerance: Perceived intentionality of racial discrimination shapes preferences for colorblindness versus multiculturalism. Journal of Experimental Social Psychology, 69(Supplement C), 86–101. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jesp​​.201​6​​.08​.0​​02 Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organisation. Academy of Management Review, 14(1), 20–39. https://doi​.org​/10​.2307​/258189 Benamati, J., Serva, M. A., & Fuller, M. A. (2010). The productive tension of trust and distrust:The coexistence and relative role of trust and distrust in online banking. Journal of Organizational Computing and Electronic Commerce, 20(4), 328–346. https​:/​/do​​i​.org​​/ 10​.1​​080​/1​​09193​​92​.20​​​10​.51​​6632 Bijlsma-Frankema, K., Sitkin, S. B., & Weibel, A. (2015). Distrust in the balance: The emergence and development of intergroup distrust in a court of law. Organization Science, 26(4), 1018–1039. http://dx​.doi​.org​/10​.1287​/orsc​.2015​.0977

Multilevel Trust and HRM  301

Blau, P. (1964). Exchange and power in social life. Hoboken, NJ: Wiley. Boon, C., Den Hartog, D. N., & Lepak, D. P. (2019). A systematic review of human resource management systems and their measurement. Journal of Management, 45(6), 2498–2537. https://doi​.org​/10​.1177​/0149206318818718 Bullis, C., & Bach, B. W. (1989a). Are mentor relationships helping organizations? An exploration of developing mentee-mentor-organizational identifications using turning point analysis. Communication Quarterly, 37(3), 199–213. https​:/​/do​​i​.org​​/10​.1​​080​/ 0​​14633​​789​09​​38554​0 Bullis, C., & Bach, B. W. (1989b). Socialization turning points: An examination of change in organizational identification. Western Journal of Communication (includes Communication Reports), 53(3), 273–293. https​:/​/do​​i​.org​​/10​.1​​080​/1​​05703​​189​09​​37430​7 Byrne, Z. S., Pitts,V. E., Wilson, C. M., & Steiner, Z. J. (2012). Trusting the fair supervisor: The role of supervisory support in performance appraisals. Human Resource Management Journal, 22(2), 129–147. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1748​​-8583​​.201​2​​.0019​​3.x Capell, B.,Tzafrir, S. S., & Dolan, S. L. (2016).The disclosure of concealable stigmas: Analysis anchored in trust. Cogent Psychology, 3(1), 1121066. https​:/​/do​​i​.org​​/10​.1​​080​/2​​33119​​08​. 20​​​15​.11​​21066​ Capell, B., Tzafrir, S. S., Enosh, G., & Dolan, S. L. (2018). Explaining sexual minorities’ disclosure: The role of trust embedded in organisational practices. Organization Studies, 39(7), 947–973. https://doi​.org​/10​.1177​/0170840617708000 Chambers, J. R., & Melnyk, D. (2006). Why do I hate thee? Conflict misperceptions and intergroup mistrust. Personality and Social Psychology Bulletin, 32(10), 1295–1311. https:// doi​.org​/10​.1177​/0146167206289979 Chory, R. M., & Hubbell, A. P. (2008). Organizational justice and managerial trust as predictors of antisocial employee responses. Communication Quarterly, 56(4), 357–375. https​:/​/do​​i​.org​​/10​.1​​080​/0​​14633​​708​02​​44812​1 Cojuharenco, I., Fortin, M., & German, H. (2014). Organizational justice and time: A review of the literature on justice reactions over time and directions for future research. In A. J. Shipp & Y. Fried (Eds.), Time and work: How time impacts individuals (Vol. 1, pp. 163–190). New York, NY: Routledge. Colquitt, J.A., LePine, J.A., Piccolo, R. F., Zapata, C. P., & Rich, B. L. (2012). Explaining the justice-performance relationship: Trust as exchange deepener or trust as uncertainty reducer? Journal of Applied Psychology, 97(1), 1–15. https://doi​.org​/10​.1037​/a0025208 Colquitt, J. A., & Rodell, J. (2011). Justice, trust, and trustworthiness: A longitudinal analysis integrating three theoretical perspectives. Academy of Management Journal, 54, 1183– 1206. https://doi​.org​/10​.5465​/amj​.2007​.0572 Colquitt, J. A., Scott, B., & LePine, J. (2007).Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 92(4), 909–927. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​. 9​​2​.4​.9​​09 Cooper, B., Wang, J., Bartram, T., & Cooke, F. L. (2019). Well-being-oriented human resource management practices and employee performance in the Chinese banking sector: The role of social climate and resilience. Human Resource Management, 58(1), 85–97. https://doi​.org​/10​.1002​/hrm​.21934 Costa, A. C., Fulmer, C. A., & Anderson, N. R. (2018). Trust in work teams: An integrative review, multilevel model, and future directions. Journal of Organizational Behavior, 39(2), 169–184. https://doi​.org​/10​.1002​/job​.2213 Eva, N., Newman, A., Miao, Q., Cooper, B., & Herbert, K. (2019). Chief executive officer participative leadership and the performance of new venture teams. International Small

302  Rosalind Searle and Rami Al-Sharif

Business Journal: Researching Entrepreneurship, 37(1), 69–88. https://doi​.org​/10​.1177​ /0266242618808558 Farndale, E., McDonnell, A., Scholarios, D., & Wilkinson, A. (2020). Human Resource Management Journal: A look to the past, present, and future of the journal and HRM scholarship. Human Resource Management Journal, 30(1), 1–12. https://doi​.org​/10​.1111​/ 1748​-8583​.12275 Fulmer, C. A. (2018). Multilevel trust: Antecedents and outcomes of trust at different levels. In R. H. Searle, A.-M. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 143–160). New York, NY: Routledge. Gardner, T. M., Wright, P. M., & Moynihan, L. M. (2011). The impact of motivation, empowerment, and skill-enhancing practices on aggregate voluntary turnover: The mediating effect of collective affective commitment. Personnel Psychology, 64(2), 315– 350. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1744​​-6570​​.201​1​​.0121​​2.x Ghumman, S., & Jackson, L. (2010).The downside of religious attire:The Muslim headscarf and expectations of obtaining employment. Journal of Organizational Behavior, 31(1), 4–23. https://doi​.org​/10​.1002​/job​.601 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34(1), 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 Gillespie, N., Dietz, G., & Lockey, S. (2014). Organizational reintegration and trust repair after an integrity violation: A case study. Business Ethics Quarterly, 24(3), 371–410. https://doi​.org​/10​.5840​/beq2014437 Gillespie, N., Lockey, S., Hornsey, M., & Okimoto, T. (2021). Trust repair: A multilevel framework. In N. Gillespie, A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. New York, NY: Routledge. Giumetti, G. W., Schroeder, A. N., & Switzer Iii, F. S. (2015). Forced distribution rating systems:When does “rank and yank” lead to adverse impact? Journal of Applied Psychology, 100(1), 180–193. https://doi​.org​/10​.1037​/a0037191 Guest, D., & Conway, N. (2011).The impact of HR practices, HR effectiveness and a ‘strong HR system’ on organizational outcomes: A stakeholder perspective. The International Journal of Human Resource Management, 22(8), 1686–1702. https​:/​/do​​i​.org​​/10​.1​​080​/ 0​​95851​​92​.20​​​11​.56​​5657 Gustafsson, S., Gillespie, N. A., Searle, R., Hope Hailey, V., & Dietz, G. (2020). Preserving organisational trust during disruption. Organization Studies, 1–25. https://doi​.org​/10​. 1177​/0170840620912705 Harbi, S. A., Thursfield, D., & Bright, D. (2017). Culture, wasta and perceptions of performance appraisal in Saudi Arabia. The International Journal of Human Resource Management, 28(19), 2792–2810. https​:/​/do​​i​.org​​/10​.1​​080​/0​​95851​​92​.20​​​16​.11​​38987​ Heilman, M. E., Block, C. J., & Stathatos, P. (1997). The affirmative action stigma of incompetence: Effects of performance information ambiguity. Academy of Management Journal, 40(3), 603–625. https://doi​.org​/10​.5465​/257055 Ho,V.T., & Astakhova, M. N. (2018). Disentangling passion and engagement:An examination of how and when passionate employees become engaged ones. Human Relations, 71(7), 973–1000. https://doi​.org​/10​.1177​/0018726717731505 Johns, G. (2018). Advances in the treatment of context in organizational research. Annual Review of Organizational Psychology and Organizational Behavior, 5, 21–46. https​:/​/do​​i​.org​​/ 10​.1​​146​/a​​nnure​​v​-org​​psych​​-03​21​​17​-10​​4406 Kalokerinos, E. K., von Hippel, C., & Zacher, H. (2014). Is stereotype threat a useful construct for organizational psychology research and practice? Industrial and Organizational Psychology, 7(3), 381–402. https://doi​.org​/10​.1111​/iops​.12167

Multilevel Trust and HRM  303

Korsgaard, M. A. (2018). Reciprocal trust: A self-reinforcing dynamic process. In R. H. Searle, A. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 14–28). Routledge. Korsgaard, M. A., Brower, H. H., & Lester, S. W. (2014). It isn’t always mutual: A critical review of dyadic trust. Journal of Management. 41(1), 47–70. https://doi​.org​/10​.1177​/ 0149206314547521 Kramer, R. M., & Lewicki, R. J. (2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annuals, 4(1), 245–277. https​:/​/do​​i​.org​​/10​.5​​465​/1​​94165​​20​.20​​​10​.48​​7403 Lee, H. W., Pak, J., Kim, S., & Li, L.-Z. (2019). Effects of human resource management systems on employee proactivity and group innovation. Journal of Management, 45(2), 819–846. https://doi​.org​/10​.1177​/0149206316680029 Lewicki, R. J., & Bunker, B. (1996). Developing and maintaining trust in working relationships. In R. Kramer & T. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 114–139). Thousand Oaks, CA: SAGE. Lewicki, R. J., McAllister, D. J., & Bies, R. J. (1998). Trust and distrust: New relationships and realities. Academy of Management Review, 23(3), 438–458. https://doi​.org​/10​.5465​/ amr​.1998​.926620 Lewicki, R. J., Tomlinson, E. C., & Gillespie, N. (2006). Models of interpersonal trust development: Theoretical approaches, empirical evidence, and future directions. Journal of Management, 32(6), 991–1022. https://doi​.org​/10​.1177​/0149206306294405 Luhmann, N. (1979). Trust and power:Two works. Wiley. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​. 19​​95​.​95​​08080​​335 McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38, 24–59. https://doi​.org​/ 10​.5465​/256727 McCord, M. A., Joseph, D. L., Dhanani, L. Y., & Beus, J. M. (2017). A meta-analysis of sex and race differences in perceived workplace mistreatment. Journal of Applied Psychology, 103(2), 137. https://doi​.org​/10​.1037​/apl0000250 McEvily, B., & Tortoriello, M. (2011). Measuring trust in organizational research: Review and recommendations. Journal of Trust Research, 1(1), 23–63. https​:/​/do​​i​.org​​/10​.1​​080​/ 2​​15155​​81​.20​​​11​.55​​2424 Meyerson, D., Weick, K. E., & Kramer, R. M. (1996). Swift trust and temporary groups. In R. M. Kramer, & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 166–196). Thousand Oaks, CA: SAGE. Morgeson, F. P., Mitchell, T. R., & Liu, D. (2015). Event system theory: An event-oriented approach to the organizational sciences. Academy of Management Review, 40(4), 515–537. https://doi​.org​/10​.5465​/amr​.2012​.0099 Moscoso, S., García-Izquierdo, A. L., & Bastida, M. (2010). Reactions toward affirmative action measures for women. Revista de Psicología del Trabajo y de las Organizaciones, 26(3), 211–221. https://doi​.org​/10​.5093​/tr2010v26n3a5 Mulligan, J. R., & Bull Schaefer, R. A. (2011). A new hope for rank and yank. Journal of Leadership & Organisational Studies, 18(3), 385–396. https://doi​.org​/10​.1177​/ 1548051811405206 Nam, J., & Lee, H. (2018). High commitment human resource practices and employee behavior: A multi-level analysis. International Journal of Manpower, 39(5), 674–686. https://doi​.org​/10​.1108​/IJM​-09​-2016​-0171

304  Rosalind Searle and Rami Al-Sharif

Nerstad, C. G., Searle, R., Černe, M., Dysvik, A., Škerlavaj, M., & Scherer, R. (2018). Perceived mastery climate, felt trust, and knowledge sharing. Journal of Organizational Behavior, 39(4), 429–447. https://doi​.org​/10​.1002​/job​.2241 Ng,T.W. H. (2015).The incremental validity of organizational commitment, organizational trust, and organizational identification. Journal of Vocational Behavior, 88, 154–163. https​:/​/ do​​i​.org​​/10​.1​​016​/j​​.jvb.​​201​5.​​03​.00​3 Nienaber, A.-M., Hofeditz, M., & Romeike, P. D. (2015).Vulnerability and trust in leaderfollower relationships. Personnel Review, 44(4), 567–591. https://doi​.org​/10​.1108​/PR​09​-2013​-0162 Patent, V., & Searle, R.H. (2019). Qualitative meta-analysis of propensity to trust measurement. Journal ofTrust Research, 9(2), 136–163. https​:/​/do​​i​.org​​/10​.1​​080​/2​​15155​​81​. 20​​​19​.16​​75074​ Prieto, I. M., & Pilar Pérez Santana, M. (2012). Building ambidexterity: The role of human resource practices in the performance of firms from Spain. Human Resource Management, 51(2), 189–211. https://doi​.org​/10​.1002​/hrm​.21463 Rezvani, A., Khosravi, P., & Ashkanasy, N. M. (2018). Examining the interdependencies among emotional intelligence, trust, and performance in infrastructure projects: A multilevel study. International Journal of Project Management, 36(8), 1034–1046. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.ijpr​​oman.​​20​18.​​08​.00​2 Robinson, S. L. (1996).Trust and breach of the psychological contract. Administrative Science Quarterly, 41(4), 574–599. 10.2307/2393868 Robinson, S. L., & Rousseau, D. M. (1994). Violating the psychological contract: Not the exception but the norm. Journal of Organizational Behavior, 15(3), 245–259. https:// doi​.org​/10​.1002​/job​.4030150306 Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Salamon, S. D., & Robinson, S. L. (2008).Trust that binds:The impact of collective felt trust on organizational performance. Journal of Applied Psychology, 93(3), 593–601. https​:/​/do​​i​. org​​/10​.1​​037​/0​​021​-9​​010​​.9​​3​.3​.5​​93 Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 224–253. https://doi​.org​/10​. 2307​/2392563 Sanders, K., Shipton, H., & Gomes, J. F. S. (2014). Guest editors’ introduction: Is the HRM process important? Past, current, and future challenges. Human Resource Management, 53(4), 489–503. https://doi​.org​/10​.1002​/hrm​.21644 Sapegina, A., & Weibel, A. (2017). The good, the not so bad, and the ugly of competitive human resource practices: A multidisciplinary conceptual framework. Group & Organization Management, 42(5), 707–747. https://doi​.org​/10​.1177​/1059601117730238 Saunders, M. N., Dietz, G., & Thornhill, A. (2014). Trust and distrust: Polar opposites, or independent but co-existing? Human Relations, 67(6), 639–665. https://doi​.org​/10​. 1177​/0018726713500831 Searle, R. H. (2018). Trust and HRM. In R. H. Searle, A.-M. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge companion to trust (pp. 483–505). New York, NY: Routledge. Searle, R. H., & Ball, K. S. (2004). The development of trust and distrust in a merger. Journal of Managerial Psychology, 19(7), 708–772. https​:/​/do​​i​.org​​/10​.1​​108​/0​​26839​​404​10​​55939​2 Searle, R. H., Den Hartog, D., Weibel, A., Gillespie, N., Six, F., Hatzakis, T., & Skinner, D. (2011). Trust in the employer: The role of high involvement work practices and

Multilevel Trust and HRM  305

procedural justice in European organizations. International Journal of Human Resource Management., 22(5), 1068–1091. https​:/​/do​​i​.org​​/10​.1​​080​/0​​95851​​92​.20​​​11​.55​​6782 Searle, R. H., & Dietz, G. (2012). Trust and HRM: Current insights and future directions. Human Resource Management Journal, 22(4), 333–342. https://doi​.org​/10​.1111​/1748​8583​.12006 Searle, R. H., Nienaber, A., Price, D., & Holtgrave, M. (2018). Lone star or team player? The interrelationship of different identification foci and the role of self-presentation concerns. Human Resource Management., 57(2), 529–547. https://doi​.org​/10​.1002​/hrm​.21868 Searle, R. H., & Skinner, D. (2011). Trust in the context of performance appraisal. In R. Searle & D. Skinner (Eds.), Trust and HRM (pp. 177–197). Cheltenham, UK: Edward Elgar. Siebert, S., & Czarniawska, B. (2018). Distrust: Not only in secret service organizations. Journal of Management Inquiry, 29(3), 286–298. https://doi​.org​/10​.1177​/1056492618798939 Sitkin, S. B., & Roth, N. L. (1993). Explaining the limited effectiveness of legalistic “remedies” for trust/distrust. Organization Science, 4(3), 367–392. 10.1287/orsc.4.3.367 Skiba, T., & Wildman, J. L. (2019). Uncertainty reducer, exchange deepener, or selfdetermination enhancer? Feeling trust versus feeling trusted in supervisor-subordinate relationships. Journal of Business and Psychology, 34(2), 219–235. https​:/​/do​​i​.org​​/10​.1​​007​/ s​​10869​​-01​8-​​9537-​x Storey, J. (2007). Human resource management today: An assessment. In J. Storey (Ed.), Human resource management: A critical text (pp. 3–20). London: Thompson Learning. Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel, & W. Austin (Eds.), Psychology of intergroup relations (2nd ed., pp. 7–24). Chicago, IL: Nelson-Hall. Tan, H. H., & Tan, C. S. F. (2000). Toward the differentiation of trust in supervisor and trust in organization. Genetic, Social, and General Psychology Monographs, 126(2), 241–260. Truxillo, D. M., & Bauer, T. N. (1999). Applicant reactions to test scores banding in entrylevel and promotional contexts. Journal of Applied Psychology, 84(3), 322–339. https​:/​/ do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.8​​4​.3​.3​​22 Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group behavior. In E. J. Lawler (Ed.), Advances in group processes (pp. 77–122). Greenwich, CT: JAI Press. Tzafrir, S. S. (2005).The relationship between trust, HRM practices and firm performance. International Journal of Human Resource Management, 16, 1600–1622. https://doi​.org​/0​. 1080​/09585190500239135 van der Werff, L., & Buckley, F. (2014). Getting to know you: A longitudinal examination of trust cues and trust development during socialization. Journal of Management, 43(3), 742–770. https://doi​.org​/10​.1177​/0149206314543475 Van Laar, C., Derks, B., & Ellemers, N. (2013). Motivation for education and work in young Muslim women: The importance of value for ingroup domains. Basic and Applied Social Psychology, 35(1), 64–74. https​:/​/do​​i​.org​​/10​.1​​080​/0​​19735​​33​.20​​​12​.74​​6609 Vlaar, P. W. L.,Van den Bosch, F. A. J., & Volberda, H. W. (2007). On the evolution of trust, distrust, and formal coordination and control in interorganizational relationships: Toward an integrative framework. Group & Organization Management, 32(4), 407–428. https://doi​.org​/10​.1177​/1059601106294215 Wehrle, K., Klehe, U.-C., Kira, M., & Zikic, J. (2018). Can I come as I am? Refugees’ vocational identity threats, coping, and growth. Journal of Vocational Behavior, 105, 83– 101. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jvb.​​201​7.​​10​.01​0

306  Rosalind Searle and Rami Al-Sharif

Whitener, E. M. (2001). Do “high commitment” human resource practices affect employee commitment? A cross-level analysis using hierarchical linear modeling. Journal of Management, 27(5), 515–535. https​:/​/do​​i​.org​​/10​.1​​177​/0​​14920​​6301​0​​27005​​02 Whitener, E. M., Brodt, S. E., Korsgaard, M. A., & Werner, J. M. (1998). Managers as initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23(3), 513–530. https://doi​.org​/ 10​.5465​/amr​.1998​.926624 Wong, Y. T. (2018). Trust, job security and subordinate–supervisor guanxi: Chinese employees in joint ventures and state-owned enterprises. Asia Pacific Business Review, 24(5), 638–655. https​:/​/do​​i​.org​​/10​.1​​080​/1​​36023​​81​.20​​​17​.13​​84207​ Wyatt, M., & Silvester, J. (2015). Reflections on the labyrinth: Investigating black and minority ethnic leaders’ career experiences. Human Relations, 68(8), 1243–1269. https://doi​.org​/10​.1177​/0018726714550890 Zhu, J., Lau, D., & Lam, L. (2021).Trust me or us? A multilevel model of individual and team felt trust by supervisors. In N. Gillespie, A. Fulmer, & R. Lewicki (Eds.), Understanding trust in organizations: A multilevel perspective. New York, NY: Routledge.

13 TRUST CUES IN ARTIFICIAL INTELLIGENCE A Multilevel Case Study in a Service Organization Lisa van der Werff, Kirsimarja Blomqvist, and Sirpa Koskinen

Introduction Our daily routines and experiences are increasingly impacted by human–computer interactions and intelligent technology (Andras et al., 2018). Indeed, digital transformation over the last number of decades has created a world where vulnerability to, reliance on, and cooperation with technology is commonplace. A growing body of literature confirms that trust can fulfill many important functions for social systems including reducing social complexity (Luhmann, 1979) and acting as an “organizing principle” (McEvily et al., 2003) to enhance efficient and effective coordination of expectations and interactions (Bachmann, 2001). As such, the successful development and adoption of artificial intelligence (AI) are likely to require trust from stakeholders at varying levels of society. As intelligent technologies become more prevalent, the potential is greater for either misplaced or low trust in technology to create costly and destructive consequences (de Visser et al., 2018; Lee & See, 2004). In this chapter, we aim to enhance understanding of organizational trust as a multilevel and cross-level phenomenon (Fulmer & Gelfand, 2012) by developing our understanding of the factors driving trust processes in the context of AI.To do so, we examine a case where a new service based on machine learning and artificial intelligence (AI) is developed and adopted in an organization. Trust has the potential to have a considerable impact on organizing systems (von Krogh, 2018), and as Fulmer and Gelfand (2012) note, “trust within one level does not occur in a vacuum and needs to be considered in the context of trust and related factors at other levels” (p. 1204). Our case study provides an opportunity to consider how adopting AI can be influenced by trust and trust-related cues at multiple organizational levels. Understanding how people at different levels of analysis in an DOI:10.4324/9780429449185-13

308  L. van der Werff, K. Blomqvist, and S. Koskinen

organization develop trust when designing and adopting AI may inform research not only on trust in AI but also more broadly on cross-level trust dynamics. For example, do people trust AI because of features of the service itself or because they trust those who develop or introduce AI services? How do people trust AI when issues with information asymmetry and transparency mean that it is not clearly a technology or human? The remainder of this chapter is organized as follows. First, we will discuss AI and the elements of these technological artifacts that make them a particularly interesting context in which to study trust. Second, we will review the theoretical literature on micro- and macro-level cues for trust.Third, we outline the small but growing body of empirical evidence that has examined the antecedents of trust in automation including AI and machine learning. Fourth, we will briefly outline the mini case study which we have conducted and take an abductive approach to our exploration of the multilevel cues influencing trust in AI and in our case study. Thereafter, we discuss the potential for cross-level influences on trust in AI and conclude with a summary and discussion of further research avenues.

Trusting AI AI has repeatedly been identified as a disruptive technology that is likely to have profound consequences for organizations and society more broadly (Sousa & Rocha, 2019; von Krogh, 2018). AI can be defined as “a new generation of technologies capable of interacting with the environment and aiming to simulate human intelligence” (Glikson & Woolley, 2020, p. 2) or “a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments” (OECD, 2019, p. 15). The predictions of disaster scenarios confront the possibilities of a powerful AI that reshapes our world in a way that is not compatible with human values or preferences (Armstrong et al., 2016). As computer scientists engage with ethicists and philosophers to debate and attempt to protect against these disaster scenarios, current automation systems operate at a far lower level of capability. Nevertheless, the potential for AI to disrupt organizations and work has been debated across a range of fields from law (Alarie et al., 2018), finance (Hendershott et al., 2011), and medicine (Thompson et al., 2018) to mining (Hyder et al., 2018), and industry surveys repeatedly highlight AI as a priority for corporate strategic investment (Accenture, 2017). There are several facets of AI that are particularly interesting to the study of trust. In the first instance, trust scholars have long argued that risk and interdependence are necessary conditions for trust to arise (see Lewis & Weigert, 1985; Rousseau et al., 1998). Perceptions of risk in the context of trust are inextricably intertwined with an uncertainty of how the other party will behave. In some respects, at least early attempts at AI have relatively low levels of uncertainty in that they are relatively predictable in their behavior (at least for technology experts)

Trust Cues in Artificial Intelligence  309

and decision-making is governed by the restraints of the algorithm. However, increasing levels of machine learning as the AI field matures introduce more considerable levels of uncertainty (Shrestha et al., 2019). Interdependence between individuals and AI can also be expected to increase exponentially if academic and industry predictions for its impact on everyone’s lives are realized (Accenture, 2017; de Visser et al., 2018). As a result, trust is likely to become increasingly salient and critical in the AI landscape. AI also highlights a number of interesting questions about the antecedents of a willingness to be vulnerable and trust another party. The study of the antecedents of trust in technology artifacts is not new; however, AI differs from other technological artifacts in two ways that are important to the formation of trust. First, the provision of AI services typically relies on a supply chain of technology providers, many of whom remain somewhat unseen to a consumer despite their influence on service provision. Second, increasing levels of sophistication and opacity in AI raise a question of agency and the extent to which this technology can be thought of as acting with intention in accordance with beliefs and an ability to plan (Shank, 2014). Previous considerations of the antecedents of trust in technology have discounted the possibility of technological agency and argued that the motives or intentions of the technology are thus not of concern when making trustworthiness judgments (e.g., Söllner et al., 2016). However, if technology is capable of agentic thought, this introduces the potential to act in a way that is benevolent (or malevolent) (Shank & DeSanti, 2018).While true agency remains out of reach for current AI technology, public perceptions of agency in these systems are noticeably further advanced. This discrepancy in actual versus perceived capacity for agency is driven by two complementary forces. First, there is a significant lack of transparency surrounding AI services that is related to our capacity to understand the complexity of AI processes (Lindebaum et al., 2020). Within the computer science field, researchers are striving to address some of these issues and increase willingness to use AI through the study and development of interpretable machine learning (Yin et al., 2019). Second, there is considerable popular media hype about the topic of AI where software is often anthropomorphized and portrayed as aware and intentional (Shank & DeSanti, 2018). In their experimental study of MTurk participants, Shank and DeSanti (2018) argue that the general population considers AI to be aware, intentional, and capable of morally unjust action. Whether or not this is the case, the perceptions of users interacting with AI may be more important than the truth of current AI abilities.

Theoretical Background In considering the potential antecedents of trust in AI, we look to the wider theoretical literature on trust cues and predictors across multiple levels of analysis.

310  L. van der Werff, K. Blomqvist, and S. Koskinen

Although the referent of trust itself can also exist at various levels, our focus in this chapter is on the AI service as the central trust referent and the trust cues and information that influence decisions to trust the AI service. We organize our discussion below into cues that originate from micro and macro sources.

Micro Trust Cues We use the term micro-level trust cues to refer to cues that arise from information about the trust referent itself or about the trustor. Typically, these cues accrue from interpersonal interaction and evaluation and are directly or indirectly social in nature. When the referent of trust is an IT artifact such as AI, interactions are of course not interpersonal but micro-level cues can nonetheless be directly perceived by the trustor through experiences of interacting with the technology. In this instance, we use micro-level trust to refer to a personal relationship an individual might develop with an IT artifact such as a particular piece of hardware or software. The seminal conceptualization of micro-level cues is the Mayer et al. (1995) model, which proposes that trust as a psychological state is based on an individual’s trust propensity and their perceptions of the referent’s trustworthiness. In this model, trustworthiness consists of an aggregate evaluation of ability, benevolence, and integrity informed by previous experiences. The second micro-level cue incorporated in the model refers to a dispositional tendency in one’s propensity to trust other parties in general and is typically thought of as a relatively stable individual difference. This model has been translated to consider how we might judge the trustworthiness of technology (see van der Werff et al., 2018 for a full discussion) and is argued to be robust for referents across levels of analysis (Schoorman et al., 2007). Recent reviews of trust antecedents (e.g., Baer & Colquitt, 2018) confirm that while trustworthiness has traditionally been viewed as the primary determinant of trust, other micro-level cues can play an important role in influencing trust. In particular, Lyu and Ferrin (2018) review evidence that interpersonal trust is driven by trustor factors, trustee factors including trustworthiness and behavior, and more contextual influences including aspects of the relationship and the context in which individuals are interacting. Meanwhile, Baer and Colquitt (2018) argue for affective and heuristic influences alongside trustworthiness and individual differences of the trustor. While empirical research on some of these micro-level cues is relatively nascent, our qualitative case study approach offers an important opportunity to explore whether and how these types of trust cues play a role in trust in AI. Several trust theorists have highlighted the critical role that emotion and affective processes play in trust (e.g., Jones & George, 1998; Williams, 2001). However, empirical studies have treated trust as an almost exclusively cognitive phenomenon and the role of emotion and mood in trusting has largely been ignored (van

Trust Cues in Artificial Intelligence  311

Knippenberg, 2018). Unlike the cognitive cues we have discussed above, affect as a cue for trust does not necessarily require time to develop. Emotions in particular are known to be characterized as a response to a stimulus that is typically intense and short in duration (Kelly & Barsade, 2001). Recent theoretical developments suggest that emotional responses and emotion regulation play an important role in trust in automation and technology artifacts (de Visser et al., 2019). The final category of micro-level trust cues that have been highlighted in the literature can collectively be referred to as heuristic cues. Heuristic trust cues is a term used to describe judgmental rules or cognitive shortcuts that have been stored in memory and can influence decisions including trust (Baer & Colquitt, 2018). Heuristics cues operate at a less conscious, systematic level and are likely to be particularly influential in situations where clear cues for trustworthiness are not (or not yet) available. One of the earliest discussions of heuristic processing in the trust literature can be seen in McKnight et al.’s (1998) conceptualization of situational normality as a driver of trust. Situational normality is a term that refers to a belief that things are normal, customary, and in their proper order, which allows a trustor to proceed in interacting with the trustee without the need for a conscious weighing up of other trust cues. Within the trust in technology space, situational normality has been demonstrated to be an important antecedent of trust in technologies such as e-commerce (Gefen et al., 2003) and mobile banking (Gu et al., 2009). Other heuristic processes that appear to play an important role in the trust literature include aesthetics. Although the role of aesthetics in trust has received limited attention in the organizational sciences (for notable exceptions, see Baer et al., 2018; Todorov et al., 2005), their influence has long been recognized in information sciences (e.g., Cyr et al., 2010; Kim & Moon, 1998; Tuch et al., 2010).

Macro-Level Trust Cues Given the nature of AI services as systemic, embedded in supply chains and complex networks of technologies, organizations, and professionals, a consideration of trust in AI must also take into account the wider structure of the contextual environment. We use the term macro-level trust cues to encompass information that arises from trust referents at a higher level than the AI service itself including service providers, complementary technologies, and regulatory standards. These impersonal cues are not typically perceived through direct personal contact with the other party. As such, trust built on this basis has been considered as a type of indirect social relationship, where information about organizations, systems, or technology can shape trusting attitudes, decisions, and behavior regarding another trust referent (Bachmann, 2018; Kosonen et al., 2008; Shapiro, 1987). Furthermore, impersonal trust in an institutional system such as an organization can transfer to trust in a related referent (Stewart, 2003). This form of trust cue often consists of institutional safeguards providing control and supporting coordination (Bachman

312  L. van der Werff, K. Blomqvist, and S. Koskinen

& Inkpen, 2011) at meso and micro levels. In essence, the impact of these cues on trust in an AI service represents a cross-level relationship in itself. Literature on impersonal, system-based trust stems mainly from the sociology (Giddens, 1990; Luhmann, 1979; Shapiro, 1987; Zucker, 1986) and information systems (Gefen, 2000; Lankton et al., 2015; McKnight et al., 2011) fields. In sociology, Luhmann (1979) was one of the first authors to elaborate the concept of system trust, arguing that aspects of social systems such as money, power, and truth can also become objects of trust (Kroeger, 2019). Relatedly, Giddens (1990) discusses expert systems as “systems of technical accomplishment or professional expertise that organize large areas of the material and social environment in which we live today” (p. 80). Even with social or institutional systems, “expert systems” (Giddens, 1990) trust can be based on the systemic principles created by humans who have created such systems (Bachmann, 1998; Kroeger, 2015; Sydow, 1998). Thus, humans can trust the systems such as AI without personally knowing the engineers who created the code and system, but by placing trust in the well-functioning system (Kroeger, 2015; Luhmann, 1979). Then trust in abstract systems is vested in their abstract capacities (money or artificial system) and the correctness of principles involved, rather than in people and their good intentions (Kroeger, 2019).

Empirical Evidence for the Antecedents of AI While popular media discussion of AI has been abundant, empirical evidence for the factors that drive trust in AI is relatively sparse. In reviewing this literature, we have cast our net slightly wider than strict definitions of AI to consider the antecedents of trust in automation more generally, including algorithms, machine learning, and robots. Different types of automation can be related to each other in terms of the degree of automation and intelligence they entail; for a useful taxonomy, see Yagoda and Gillian (2012). In their meta-analysis of the factors affecting trust in human–robot interaction, Hancock and colleagues outline three key categories: human-related, robotrelated, and environmental factors such as physical environment, task type, and culture (Hancock et al., 2011). Although their sample of studies was relatively small (21 papers), their analysis indicated that the most influential of these are robot-related factors, which include performance-based indicators including reliability and level of automation, and attribute-based factors including judgments of robot personality and level of anthropomorphism. Using a similar framework, Schaefer and colleagues conducted a meta-analysis focused on trust in automation and found 30 papers that demonstrated similar effect sizes for automation-related indicators overall, with a slightly smaller effect size for attribute-based factors (Schaefer et al., 2016). In this second study, however, human-related factors such as attitudes towards technology and demographics also played an important role in driving trust.

Trust Cues in Artificial Intelligence  313

Factors that information systems researchers have termed automation or robot-related indicators are essentially perceptions of the characteristics of the system, which in the trust literature would typically be labeled trustworthiness. As such, these perceptions fall under the theoretical umbrella of what we have called micro-trust cues in our discussion of theory above. Within this, performance-based indicators have been the primary focus of studies investigating the antecedents of trust in automation. In particular, the importance of system reliability has been demonstrated repeatedly (e.g., de Vries et al., 2003). In addition, there is increasing evidence that transparency around automated decisionmaking and explanation of any errors made is useful in building trust (Verberne et al., 2012) although this is likely to be context-dependent (Baker et al., 2018). The study of attribute-based indicators is closely related to our discussion of agency above and the extent to which people are inclined to attribute human characteristics like personality, intent, morality, and even intelligence to AI systems. This process of anthropomorphism is increasingly likely when AI systems are opaque (Andras et al., 2018). Empirical research on human–computer interaction suggests that people are more inclined to trust AI if it is capable of emotional expression (Melo et al., 2016) and if it behaves in line with social norms for communication (Pejsa et al., 2015). However, over-anthropomorphism can lead to misunderstandings and overestimation of AI ability, which may be more damaging to trust in the longer term (Baker et al., 2018). The antecedents of trust that information systems researchers have termed human-related factors refer to dispositional tendencies such as trust propensity (Gefen, 2000) as well as factors such as expertise, understanding, confidence, and comfort in dealing with the technology (Schaefer et al., 2016). The individual difference discussed most frequently in this literature is trust propensity, which has been contextualized to a technology context. For instance, McKnight and colleagues have differentiated between faith in general technology – a belief that IT is generally reliable, functional, and helpful – and a technology-trusting stance – a belief that interacting with technology is likely to lead to positive outcomes. Information sciences scholars have highlighted the scarcity of empirical research investigating the role of so-called human factors in influencing trust in automation systems (Baker et al., 2018; Hancock et al., 2011). The small body of research that has accumulated indicates that extraversion is positively correlated with trust in automation (Merritt & Ilgen, 2008) and that dispositional differences in cognitive style influence people’s responses to anthropomorphic cues like humanvoiced computers (Lee, 2010). Attempts to correlate trust in automation with demographics such as age, gender, and previous experience have been somewhat inconclusive (Molnar et al., 2018). One thing that is clear from the accumulating literature on the antecedents of trust in automation and AI is that the focus thus far has been on cues that are directly related to the technology itself, or individual differences of the trustor. This is somewhat surprising given the multilevel nature of the systems that

314  L. van der Werff, K. Blomqvist, and S. Koskinen

support AI, which are influenced by people and institutions including development teams and organizations, third-party technology providers, and regulatory bodies. Understanding of what automation researchers have termed environmental factors, and what trust literature might refer to as macro-level cues, has been significantly underdeveloped. Trust literature outside of the AI field (Bachmann, 2018; Hong & Cho, 2011; Jarvenpaa & Teigland, 2017; Kramer & Lewicki, 2010) would suggest that macro cues would be more influential in building trust in AI than the empirical work we have reviewed above indicates. At the very least, as we have noted in our introduction, trust in AI cannot operate within a vacuum and should be studied in the context of trust and trust-related cues that originate at more macro levels. In studying trust in AI, we are interested in developing a more holistic understanding of the kinds of trust cues individuals consider when evaluating their willingness to be vulnerable to an AI service.

Empirical Case Study Given the relative scarcity of empirical evidence for how trust attitudes are formed in this novel context (Glikson & Woolley, 2020), we adopt a theoretically interesting multilevel case study approach to illustrate our discussion of the cues driving stakeholder trust. In particular, we focus our discussion on the micro- and macro-level cues that stakeholders perceive when evaluating AI-based services and making the decision to trust. Case company Gamma (pseudonym) comes from a highly regulated and competitive service industry where new digital technologies, automation, and AI create both a major threat and an opportunity. Within Gamma, AI is seen as a source for competitive advantage and differentiation in their regulated service business. In this specific consumer business area, Gamma’s services are almost completely based on digital customer encounters. With the AI pilot project, Gamma aims to enhance the quality of the digital customer experience and also build new organizational capabilities related to AI. Specifically, our case focuses on the development of an AI-based service for dealing with customer queries as a replacement/ supplement to traditional phone-based customer interactions. The AI-service development project included the proof of concept and socalled third-generation voice user interface (VUI) that was able to understand natural language and even different dialects. The VUI service architecture comprises a mobile phone application, related personalized services, authentication, business services, and the platform providing speech to text, text to speech, natural language processing, dialogue machine, and sentiments. It was designed to operate with different terminals and built by integrating Google Speech, Microsoft Bing Speech cloud service, and IBM Watson for the dialogue motor.The project started in summer 2018 as a closed ‘Sandbox’ pilot to test the functionality and dialogue with 40 internal users’ test accounts. Internal users were Gamma employees, who also had the service user role in their private capacity, thus they were a suitable

Trust Cues in Artificial Intelligence  315

audience for early user testing of the service. In early 2019, the proof of concept was designed and tested with 80 internal users’ real service transactions to study the functionality of the VUI service and how customers used the service in practice. To understand the stakeholders’ trust-related perceptions and expectations, we interviewed eight informants, including the AI project team members at strategic and operative levels, and internal end-users. Gamma’s chief technology officer (CTO) represented the strategic organizational level, while three experts of the development team represented the operative level, and four internal end-users represented potential consumers. We interviewed the key informants during the AI service development project about their earlier experiences on AI and about developing applications that utilize AI. We also asked them to share their expectations for AI services they were building considering the opportunities and risks involved. In the sections that follow, we use the abbreviations PT and IU to refer to participants from the wider project team (including the CTO) and the internal end-users respectively. We used NVivo12 software to organize our data and performed content analysis for coding data into categories and themes of microversus macro-level trust. We employed an abductive approach to data analysis (Dubois & Gadde, 2002; Ketokivi & Choi, 2014) iterating between past research, data, and theory building. This allowed us to provide a thorough and systematic description of the data giving the interviewees voice and demonstrating connections among the data and the theoretical concepts.

Micro-Level Trust Cues Individual Differences In our interviews, propensity to trust was mentioned or alluded to by a number of participants. Some participants spoke about their trust propensity in a more general sense (“I generally trust people, somehow I’m not usually worried about that kind of thing” [PT3]), while other participants highlighted the role of technologyspecific propensity (“I don’t see any direct risks in utilizing technology….I have a positive attitude on this so I can’t see them” [PT2]). Our data also highlighted the role of other individual differences, particularly those related to educational and industry backgrounds that created significant levels of information asymmetry. In line with the senior management who commented “usually the management in these expert organizations is rarely the one who knows best how to do everything,” other members of the project team were quick to emphasize the extent of their own technical expertise (for instance, “AI has been quite a foreign concept for me before as I don’t strictly have any kind of technical background” [PT3]) and that of their team members (“there are many people like myself since I haven’t really studied in the field of AI….PT3 isn’t a programmer either” [PT1]). The team recognized the benefits of this multidisciplinary collaboration (“that

316  L. van der Werff, K. Blomqvist, and S. Koskinen

kind of thing, you understand it a lot better when you have a background in languages” [PT3]) as well as the considerable challenges: when we are trying to solve a problem regarding language technology it is sometimes very challenging because we don’t understand each other at all. I’ve learnt a lot of new vocabulary and I’m sure they have learnt a lot of new vocabulary too. (PT3)

Trustworthiness In our case study, and in line with previous research (Baker et al., 2018; Hancock et al., 2011), aspects of ability and competence of the AI service were considered critical to trust and appeared in every interview.The project team spoke at length about the potential abilities (and limitations) of AI in general, admitting that “usually these AI services are quite limited in their operating capability” (PT2). In particular, the project team felt AI abilities are far less than what is often depicted in the popular media: “you read about how robots will become self-aware and take over the world and something like that…it is not very realistic after all” (PT3). In relation to this specific AI service, the Gamma project team expected some deficiencies in performance at least initially:“of course there is that possibility that the information isn’t correct.We are only approaching the pilot now, and now for the first time it is doing things in real life, so definitely there will be a lot of mistakes” (PT3). In contrast, internal user expectations for AI ability and performance were considerably higher:“the standards are really high…so that one would find added value in [using] it” (IU1); “I see a lot of possibilities and I’m hoping for a lot” (IU2). A considerable amount of discussion from both groups focused on ability in relation to the customer service experience including issues such as accuracy, response time, and ease of use. Discussion of the benevolence of the AI service and corresponding concepts such as helpfulness (McKnight et al., 2011) were less central to the trust expectations of our case study participants. The project team generally thought about the benevolence of AI primarily as an issue of customer experience. For instance, “we always think about it from the end user’s point of view” (PT1). In a similar vein, internal users mentioned that they would like if the service could “make my life easier and save me time” (IU2). However, they also highlighted the importance of more personalized cues including a wish to experience “some signals or characteristics that actually tell me it is communicating with me” (IU1) and for the service to be acting entirely in their interests: “it would be really exciting if it actually evolved into making your feel like you have an assistant” (IU1). In general, both groups had reservations about allowing the AI service to make decisions alone without human monitoring. Members of the project team suggested that “decision making should always be done by people…but AI could be a huge help if you used it as a support”

Trust Cues in Artificial Intelligence  317

(PT2). Similarly, the end-users mentioned the need to retain a “kind of control” and overview of the process behind decisions and recommendations (IU4). The third sub-dimension of trustworthiness, integrity, is typically translated in the technology environment as perceptions of the reliability and predictability of a particular system (van der Werff et al., 2018). In our interviews, reliability appeared as a very influential driver of trust in the AI service and was identified as critical by project team members and internal users alike. The project team appeared to have given considerable thought to the importance of signaling integrity to users through transparency: “the user can see straight away what our system has understood of his speech” (PT1). Interestingly, some members of the team had also explored the more value-based aspects of integrity: “there’s also the kind of moral aspect to it in that we also take care that if our system misbehaves somehow, we haven’t taken any shortcuts regarding those things” (PT1).

Affect Affective processes emerged as an important theme in many of the interviews in our case. Interestingly, the emotions reported by the project team and the endusers were considerably different at this stage of the Gamma project. In anticipation of the product release, the project team expressed a level of anxiety related to their responsibility for the project. For instance, one team member commented “I’m actually nervous about whether it will match my expectations” (PT2), while another expressed a combination of enjoyment and a kind of optimistic anxiety: “it is a lot of fun and its wild to think that in a way everything it says goes through me so I think there is a lot of responsibility…there is also a kind of an emotional bond that when you’re using it I’m like I hope it succeeds” (PT3). Overall these team emotions seem to reflect the level of uncertainty that is still present at this stage of the project as the team prepares for the launch and their trust is based on limited evidence regarding the ability or performance of the AI. The Gamma team also voiced their expectation of fear from end-users on the basis of lack of understanding: “it can be very scary and some people may not always understand it” (PT3). However, the team’s concerns were not shared at this stage by the internal end-users. In fact, Gamma internal users repeatedly stressed their enthusiasm and interest for this and other AI products: “I always pick services that are automated or use AI, because I find that interesting” (IU2).

Heuristics Evidence from our own case study emphasizes the role of situational normality in building trust in AI.The project team in particular has considered how user familiarity with other technologies might help build trust in the AI service: “we abandoned that idea because avatars are a bit…we had bad experiences of them…we

318  L. van der Werff, K. Blomqvist, and S. Koskinen

were thinking that OK people use a chat bot and such, so maybe visually showing the text would be a cool thing here” (PT2). They also voiced the importance of removing the feeling of mystery and exoticism that might be associated with AI: “we’re trying to maintain the company Gamma value that this application isn’t too mystical but it feels like a reliable application” (PT1). While the project team spoke at length about the importance of customer experience and system design as heuristics for trust in AI, they tended to focus predominantly on the technical capabilities, accuracy, and response time of the AI service. For example, PT1 comments, “the response time that is, how quickly it can complete the task. That can give you an idea of how the AI algorithm has been built and how good results it will give.” The end-users had, however, considered more aesthetic influences, and when asked about their hopes and expectations for the service, one internal user replied, “somehow I really hope that it would have a nice tone of voice” (IU1).

Cross-Level Effects: Macro-Level Trust Cues The second category of cues that emerged from our data represents cross-level topdown effects whereby information about a more macro-level entity influenced trust in the AI service. Based on relatively scarce past research (Bachmann, 2018; Kroeger, 2015), we approach our discussion of macro-level trust cues according to three subcategories or lenses: (1) cues about organizations providing and developing the AI service, (2) cues about societal institutions such as regulation, culture, and norms controlling the AI service, and (3) cues about AI services as a technological system, i.e., the different technological components required to provide AI-based services.

Organizations Developing and Providing AI Services Macro-level cues related to the trustworthiness of the organization providing or developing an AI service provide information to trustors forming perceptions of an AI service. Macro-level organizational cues for ability consisted of Gamma’s AI-service development, data analytics–related business competencies, service functionality, and system design: “we have this sort of, we call AI competence unit, or a team, I think there’s about twenty people now” (CTO). At the same time, the project team was very aware of the importance of developing enough competency and understanding among the more general staff members: If a customer walks into the office and says that I have this VUI application and now it did this and that, it doesn’t give a very good picture of the company, if the staff members first has [sic] to say, [chuckles] what is this? (PT1)

Trust Cues in Artificial Intelligence  319

In the context of AI, the type and amount of data available also become a critical resource: “In many things we’ve encountered the fact that we could have good AI engines, but we don’t have the data” (PT1). The information asymmetry related to organizational ability was visible between more knowledgeable project team members and internal users, who were mainly projecting the company’s ability based on its past references: “But my understanding is that it has to do with managing everyday personal services. And well, we have quite often had positive surprises happen in the company so [they] have my full trust that they handled it very well” (IU1). While benevolence of the AI service itself played a relatively minor role as a micro-level cue, at an organizational level Gamma’s benevolence emerged as an important driver of trust in AI. For the project team and internal users alike, benevolence was about taking a customer perspective and creating a good customer experience that reflected Gamma’s good intentions for the benefit of its customers. It was also comparable with helpfulness (Lankton et al., 2015; McKnight et al., 2011): “this is a new user interface that will make it easier for people to do business” (PT3). “We want to trust that AI has been produced in a way that it can help the customer manage things” (IU3). In the interviews, macro-level cues for organizational integrity were abundant and very salient to both the project team and internal users. Cues of Gamma’s integrity were predominantly concerned with two issues: the extent to which organizational values were shared with the trustor, and how those values were turned into ethical behavior. One internal user commented, I think acting responsibly should come as a gut reaction for everyone. But of course if there are some questions which, or some new things that haven’t been encountered before, completely new topics and new questions, then there just needs to be some policy regarding them of course, a responsible one in that case I’m sure. But I do think that those core values should be familiar to everyone involved. (IU1) When we asked about the ethical principles related to AI, interviewees considered them similar to the ethical principles in traditional business. They are largely based on our values and, they are similar to, how we encounter customers also face to face, so of course being appreciative and using the customer’s information responsibly and, never with the wrong intentions. I don’t think that we have to think about the ethical instructions for AI as a sort of separate thing right now. (PT2)

320  L. van der Werff, K. Blomqvist, and S. Koskinen

Open communication in setting expectations and being transparent was closely related to the perceived organizational integrity: That has to do with trust too….For example, when it comes to communicating about this VUI application, we don’t want to give our customers too great promises on behalf of the company…but in general we say that this is a proof of concept or pilot experiment. I personally expect that kind of open communication from the company. (PT1) Definitely a certain kind of transparency, what comes to mind first is all the data or information they have about me and storing that securely…I guess knowing what is going on or why something is happening and I’m told about issues honestly, for example if there is a problem they tell you about it and don’t try to cover it up or hide it. (PT3) In addition to trust cues based on the ABI-model, organizational predictability was also emphasized as a cue for trust in the AI service: “So that’s how I would see a company. If they make decisions that are all over the place, at least for me it would mean that this is inconsistent, I wouldn’t trust it anymore” (PT3). “The predictability. It’s a very important thing.The ability to understand the big picture” (PT2). Related to the issues of situational normality discussed in the micro cues section, familiarity with the service provider was considered as a source of cues for trust in AI.The project team understood the value in familiarity that was both a source of good reputation and an asset to building customer trust in new AI-based services: “We haven’t tried to make this too mystical even in the use cases. So all the business processes that we execute, they partly come from the existing background systems, so they’re familiar to the customer” (PT1). Internal users referred to Gamma’s reputation as an important trust cue: It probably comes back to this issue of trust. I’m part of a generation who have learned to use this kind of services and trust companies behind it. Company Gamma is a big actor in Finland, and I can’t remember anything shocking, like a scandal or something that would have made me lose trust in Gamma. (IU2) Both project team and internal users had learned to trust Gamma and believed that its good reputation would also function as a factor why the company would be very careful in bringing new AI-based services to the market:

Trust Cues in Artificial Intelligence  321

We in Gamma are well known and appreciated brand in Finland, whose general manager is often on TV and, and we have long history. There is this huge, foundation of value that we have from there and, maybe for us it helps a little bit when launching these new services, that the customers, at least if I think of myself as a customer I trust more in something that, Gamma launches that [sic] some brand that is completely unknown to me. (CTO) Company reputation was also perceived by internal users as a trust cue for the correct use of customer data: “Yeah I’d believe that, that it’s quite like that and maybe it’s also otherwise related to the kind of image, an image of reliability of an organization, that the information is used correctly” (IU3). In addition to being a reputable organization with a long history, the case company’s image of themselves as a trailblazer in applying new technologies and business models influenced the team’s motivation to launch a trustworthy AI service: Reliability and everything else have been the basic values, and we want to still maintain them, but still one factor that one of the managers keeps reminding us about is that Gamma has in the past been the first, that is, a trailblazer in many things.We want to bring that forward too, to restore that role of a trailblazer. (PT1)

Institutions Cues and information regarding higher-level institutions also informed trust in the AI service. In our case, the end-users’ awareness of the specific industry selfregulation (Shapiro, 1987) was an important cue for trust in the AI service: It would bring some risks personally but that of course has been largely about how people’s information [is used]…well there’s so much of all sorts of regulation here as well, that I think that the information gotten here doesn’t go elsewhere that easily. (IU4) However again, information asymmetry plays a critical role and the Gamma project team understands the limits of industry-regulation: if I think about it from the point of view of data-processing, there is a lot of talk at the moment about how my data is used and where they are and

FIGURE 13.1 The

TRUSTOR

App

Mobile phone App

Voice User Interface (VUI)

Interface Other organizations, i.e. complementary suppliers necessary to provide AI-based services, e.g. cloud service providers

– IBM Watson for dialogue motor – Microsoft Bing Speech cloud service – Google Speech

System(s)

Proof-of-Concept (POC)

– Authentication Platform

– Personalized Services & Business Services

VUI Service architecture

Services

opaqueness of AI-based service trustworthiness to the user.

Micro-level trust cues

Macro-level trust cues

FOCAL ORGANIZATION

Institutional regulation, culture and norms

Trust Cues in Artificial Intelligence  323

all that. So those elements exist in this as well because that data goes, at least some of it goes who knows where and they may be used for, nobody knows, not even our team knows that properly. (PT3) In our case, we see a lack of trust or even distrust in third parties and cloud service providers handling end-users’ data having an impact on the perception of trustworthiness of the AI services.That is, trust in a familiar service provider based on its reputation, brand, and familiarity suffers from a perceived lack of trust or distrust in third parties required for AI-based service provision.

AI Services as a System IT platforms and related services are based on modular technological systems (de Reuver et al., 2018). For example, case company Gamma’s AI-based services involve a voice user interface (VUI), mobile phone applications, a VUI service architecture platform (personalized services, authentication), and systems such as cloud services by complementary providers. For an end-user, much of the technological complexity is hidden and the potential related trust referents become opaque, exacerbating the considerable levels of information asymmetry (see Figure 13.1). In contrast, project team members see the system supporting AI-based services as an important cue for trust in the AI service itself: “AI will work just as well as it’s modelled. Expertise and its maintenance is the thing” (PT1). The project team members had a good understanding of how the overall design of the system would affect user trust: Of course we have our basic principles regarding the design of the system, related to risk management, information security, security of supply, and clarity. Then there’s the cost and performance efficiency. If you can get a good score in those sub-parts…then the system starts to become trusted by users. (PT1) At the more technical level, the decentralized information management techniques were seen to build trust in AI-based services. Also, the opaque nature of AI systems (Andras et al., 2018) created a clear lack of trust in cloud service providers involved in AI service provision. For instance: Maybe one difficult, practical operative risk for us is now that we don’t necessarily have sufficient visibility when we’re transferring the data to the cloud as to whether the data is used by somebody else. For example, if we

324  L. van der Werff, K. Blomqvist, and S. Koskinen

send our discussion to be transcribed in some cloud that we have an agreement for, we can’t perfectly guarantee what Google is ultimately doing with it. (PT1)

Cross-Level Effects: Bottom-Up Influences We found evidence of bottom-up effects influencing trust in the AI-based services at the heart of our Gamma case. One of the internal users in particular was acutely aware of the influence they wielded in trust signaling and convincing others to use technology in general: I want to create a sort of story and create that conception through that….I want to target it more that I will show this VUI application to my parents… .I think that by raising this sort of positive mental image makes other people want to try it as well. (IU3) This theme in our case study is in line with empirical work more generally in the information sciences where trust in new technologies is recognized to propagate through social networks from early adopters (Andras et al., 2018). In addition, members of the Gamma project team spoke of the influence they had had in challenging the norms of the organization in trusting AI services. In particular, PT1 speaks of the influence he has had in bringing a personal interest in technology and robots into the company and comments, “so the idea sort of lingered on, how Gamma could really benefit from these robots and technology.” He also spoke of the core team’s influence on senior management more directly: “I can tell you that during the investment meeting when we presented this…he turned to me and said ‘PT1 is this the thing of the future’ and PT2 said that it came from me straight away that yes, it is” (PT1). While our focus in this chapter is on trust in the AI service, we also saw considerable evidence that the project team felt evidence of typical bottom-up transfer, where trust in a lower level referent was considered to be important in impacting trust in higher-level referents, included project teams’ consideration that the design of the AI service would impact trust in the organization itself: “the company image is maybe such that we didn’t want to start making this VUI application into the kind of silly thing that would chat back at you” (PT1).

Discussion Realizing the value of any automated system relies on our capacity to build and facilitate human–system collaboration (Schaefer et al., 2016). Understanding how trust in AI is formed and the cues that are influential in this process will be crucial

Trust Cues in Artificial Intelligence  325

for building AI that can realize its potential to augment and complement human performance. Thus far, empirical research examining the ways in which individuals trust AI-based services is scarce, and much of this research is quantitative and inappropriate for studying trust processes in this relatively novel context. Our mini case study allows us to begin the exploration of the array of multilevel cues that are likely to play a role in the process. We will use this discussion section to summarize what we believe are the most important findings of our case, to discuss the implications of these findings for our understanding of the multilevel phenomenon of trust, and to highlight what we believe are promising avenues for future research. Our case study confirmed the importance of traditional micro-level trust cues such as ability and integrity in exerting influence on stakeholder trust in AI. Benevolence arose less regularly as a theme in our case, although stakeholders in the senior management, developer, and user categories did express an awareness of the importance of benevolence in their reluctance to allow the technology to operate with complete agency and without sufficient oversight. Similarly, concepts such as trust propensity and, more specifically, propensity to trust technology were evident in our data. Perhaps more interestingly, our case revealed several microlevel cues that are less regularly discussed in the literature. First, our stakeholders felt that divergences in educational and industry backgrounds created a level of information asymmetry that influenced propensity to trust the technology and the trust process more generally. In line with this, although trust propensity has typically been considered a stable construct, researchers have begun to consider the possibility of fluctuation and change in this propensity over time (e.g., van der Werff et al., 2019). Our case study suggests that context-relevant knowledge might play an important role in the ongoing development of trust propensity, particularly in novel and unfamiliar referents such as AI. Second, the case study demonstrated interesting differences between stakeholders in the types of affect they experienced at this phase of the project and that anticipating the affect of other stakeholders might have a role in building trustworthy technology. The wider field of organizational decision-making is increasingly cognizant of the role of affect in informing and influencing our decisions (e.g., van Kleef, 2014), and our results offer the first step in investigating this process in the context of decisions to trust AI-based services. Our case underscores the role of affect in both decisions to trust and in intentions to signal trustworthiness to others. While our focal trust referent (the AI service) was not autonomous enough to consider its impact on emotion, those designing it were very cognizant of this and its potential to influence user trust. We argue that the role of emotion has been underplayed in empirical research both in the interpersonal trust literature and in the study of trust in technology and deserves further investigation. Finally, our results indicate the potential for heuristic processes to influence our attitudes toward AI-based services. Specifically, interviews with stakeholders

326  L. van der Werff, K. Blomqvist, and S. Koskinen

highlighted the perceptions related to situational normality, technology aesthetics, and a host of macro-level information and cues as having potential heuristic influences on subsequent trust in AI. Our case study lends support to recent arguments that heuristics and affect play a more important role in trusting than that portrayed by traditional models of interpersonal trust.While heuristics have appeared as antecedents of trust in a small body of previous information systems literature (e.g., Li & Yeh, 2010), the concept of heuristic influences on trust has been very much a side act to the main show of ability, benevolence, and integrity or their corresponding conceptualizations. In the context of novel and unfamiliar trust referents like AI, fast responses like heuristics and affect are likely to play a key role in the initial trust decisions that may then influence the subsequent processing of information about the technology through confirmation bias. Research on cross-level effects in trust research is still nascent yet there is some evidence that these effects are an important determinant of trust. For example, top-down effects where trust in leaders was impacted by organizational form (Ambrose & Schminke, 2003), individual trust was affected by system values (Shamir & Lapidot, 2003), and organizational trust by market condition (Hodson, 2004).There seems to be even less research on bottom-up effects such as trust within teams having an impact on trust at the organizational level. Further, different trust referents may be interrelated, and trust in one referent may substitute or complement other trust referents (Fulmer & Gelfand, 2012) in a process known as trust transfer. Macro-level cues for trustworthiness provided important information to trustors in our case study. For instance, organizational ability to build AI services, benevolence in creating services for their customer needs (Lankton et al., 2015; McKnight et al., 2011), and integrity in taking care of data privacy were considered important in influencing trust in the AI service. However, here stakeholders’ trust perceptions differed so that the project team members were more knowledgeable and critical with respect to Gamma’s competencies and industry partners’ integrity, whereas internal users’ trust was based on Gamma’s reputation and past performance. These heuristic macro cues appeared to provide a proxy for the ability and predictability of the AI service. Although predictability only rarely appears in traditional models of trust, its potential relevance has been highlighted in the context of technology (van der Werff et al., 2018) and in sociological views of trust in institutions (Bachmann & Inkpen, 2011). In line with previous research (Bachman & Inkpen, 2011; Cook et al., 2009; Lane & Bachman, 1997; Luhmann, 1979), value congruence between Gamma and its stakeholders created trust across organizational stakeholders, whereas project team members were more critical of the system’s perceived ability to put values and norms in practice. Macro- or organizational-level integrity was also signaled through open communication, privacy protection, and fair information practices, which are thought to build trust by providing more transparency and potential control over their information (Tang, Hu, & Smith, 2008). In addition to macro cues at the organizational level, several industry-level cues played a role in influencing trust in the AI service. Industry self-regulation (Gefen

Trust Cues in Artificial Intelligence  327

& Pavlou, 2006; Shapiro, 1987) was an important cue for internal end-users’ trust in the AI service, whereas project team members had a more critical view related to their perceptions of complementary technology providers’ integrity. Furthermore, institutional trust itself may also function as a macro-level cue when trust in the institutional trust referent is transferred to trust in the service provider (Fulmer & Gelfand, 2012). In past research from the e-commerce context, consumer trust in third-party platforms and intermediaries has been relatively more important than their trust in individual sellers on the platforms (Hong & Cho, 2011; Jarvenpaa & Teigland, 2017). In our context, we see a lack of trust or even distrust in third parties and cloud service providers handling end-users’ data, thus having an impact on the perception of trustworthiness of the AI services. That is, the trust in familiar service providers based on their reputation, brand, and familiarity suffers from the perceived lack of trust or distrust in third parties required for AI-based service provision. Instead, the focal organizations’ past strong reputation becomes an important source for trustworthiness expectations for its AI-based services across organizational stakeholders. Thus, a highly established and reputable organization provides familiarity with its traditional ‘bricks-and-mortar’ service business, brand, and reputation transferred to the AI-based service context (Stewart, 2003). In exploring bottom-up effects, Lumineau and Schilke (2018) outline three mechanisms through which individuals or lower-level trust is likely to influence trust in a higher-level entity. Specifically, individuals can disseminate their own personal beliefs and experiences, challenge the norms of their social grouping, or advocate for control mechanisms and changes to structures. We found evidence in our case that the first two of these processes were relevant in influencing trust in the AI service. We did not come across evidence in this project for individuals advocating for control mechanisms or changes to the structures that govern AI use although this is likely to have been influenced by the stage of the project we were investigating. Indeed, it appeared that individual members of the project team indicated a level of frustration at the controls that were impeding progress. At this stage, it is possible that the need for individual advocacy in increasing control was not yet salient to either the team or end-users.

Conclusion and Future Research Directions As AI-based services become increasingly ubiquitous in our lives, the societal and economic importance of trustworthy AI is becoming critical. Trust has been identified as a key driver of technology adoption and is likely to be even more crucial in technologies wherein there is a perceived (or actual) level of agency in the actions of the technology. While we recognize the limitations related to generalizability in our case study, our findings suggest that AI developers need to consider a wide range of cognitive and affective influences on trust across a range of individual, organizational, and societal levels. More specifically, our mini case indicated that alongside traditional micro cues like trustworthiness, developers

328  L. van der Werff, K. Blomqvist, and S. Koskinen

should be aware of the potential for heuristic cues (appearance, sound, etc.) that are apparent when interacting with AI-based services. These cues are likely to have swift, automatic consequences for trustors before there is time to gather more systematic information about the ability or integrity of the service. Trust cues related to shared values, norms, and culture are likely to function across stakeholders and trust referents when building trust in AI. Furthermore, the cross-level effects uncovered in this study suggest that the perceived ‘weakest link’ in a supply chain or ecosystem could have spillover effects on consumer trust in a particular AI-based service. As such, building trustworthy AI will require input and effort from stakeholders across organizational, political, and regulatory boundaries. Our chapter offers insights beyond those directly relevant to the AI context. In general, trust theory has focused on trust cues influencing trust in referents at a similar level with relatively little consideration of context or cues at other levels. Our case expands understanding of the potential for cues at different levels of analysis to have an important impact. In essence, all of the cues examined in our macro-level cues section represent a cross-level effect. Recent empirical work highlights the potential for trust cues to operate across levels and for trust to transfer up or down through an organizational system (e.g., De Cremer et al., 2018; Fulmer & Ostroff, 2017; Lipponen et al., 2020). Building on these insights, future work will need to further investigate the systemic nature of AI-based services and the role of boundary conditions such as time and knowledge in influencing trust.

References Accenture (2017). Report prepared by M. Purdy & P. Daugherty. Why artificial intelligence is the future of growth. Accenture Institute for High Performance. https​:/​/ww​​w​.acc​​entur​​ e​.com​​/t201​​70524​​T0554​​35_​_w​:_​/ca​​-en/_​​acnme​​dia​/P​​DF​-52​​/Acce​​nture​​-Why-​​AI​-is​​the​-​​Futur​​e​-of-​​Growt​​h​.pdf​ Alarie, B., Niblett, A., & Yoon, A. H. (2018). How artificial intelligence will affect the practice of law. University of Toronto Law Journal, 68, 106–124. https://doi​.org​/10​.2139​/ ssrn​.3066816 Ambrose, M., & Schminke, M. (2003). Organization structure as a moderator of the relationship between procedural justice, interactional justice, perceived organizational support, and supervisory trust. Journal of Applied Psychology, 88, 295–305. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​021​-9​​010​​.8​​8​.2​.2​​95 Andras, P., Esterle, L., Guckert, M., Han, T. A., Lewis, P. R., Milanovic, K., Payne, T., Perret, C., Pitt, J., Powers, S. T., & Urquhart, N. (2018). Trusting intelligent machines: Deepening trust within socio-technical systems. IEEE Technology and Society Magazine, 37, 76–83. https://doi​.org​/10​.1109​/MTS​.2018​.2876107 Armstrong, S., Bostrom, N., & Shulman, C. (2016). Racing to the precipice: A model of artificial intelligence development. AI & Society, 31, 201–206. https​:/​/do​​i​.org​​/10​.1​​007​/ s​​00146​​-01​5-​​0590-​y Bachmann, R. (1998). Trust - conceptual aspects of a complex phenomenon. In C. Lane & R. Bachman (Eds), Trust Within and Between Organizations (pp. 298–322). Oxford: Oxford University Press.

Trust Cues in Artificial Intelligence  329

Bachmann, R. (2001). Trust, power and control in trans-organizational relations. Organization Studies, 22, 337–365. https://doi​.org​/10​.1177​/0170840601222007 Bachmann, R. (2018). Institutions and trust. In The Routledge Companion to Trust (pp. 218– 227). Abingdon: Routledge. https://doi​.org​/10​.4324​/9781315745572​-15 Bachmann, R., & Inkpen, A. C. (2011). Understanding institutional-based trust building processes in inter-organizational relationships. Organization Studies, 32, 281–301. https://doi​.org​/10​.1177​/0170840610397477 Baer, M. D., & Colquitt, J. A. (2018). Moving toward a more comprehensive consideration of the antecedents of trust. In R. H. Searle, A. M. Neinaber, & S. B. Sitkin (Eds.), Routledge Companion to Trust (pp. 163–182). Abingdon: Routledge. Baer, M. D., Van Der Werff, L., Colquitt, J. A., Rodell, J. B., Zipay, K. P., & Buckley, F. (2018).Trusting the “look and feel”: Situational normality, situational aesthetics, and the perceived trustworthiness of organizations. Academy of Management Journal, 61, 1718– 1740. https://doi​.org​/10​.5465​/amj​.2016​.0248 Baker, A. L., Phillips, E. K., Ullman, D., & Keebler, J. R. (2018).Toward an understanding of trust repair in human-robot interaction: Current research and future directions. ACM Transactions on Interactive Intelligent Systems, 8, 30. https://doi​.org​/10​.1145​/3181671 Cook, K. S., Levi, M., & Hardin, R. (Eds.). (2009). Whom can we trust?: How groups, networks, and institutions make trust possible. New York, NY: Russell Sage Foundation. Cyr, D., Head, M., & Larios, H. (2010). Colour appeal in website design within and across cultures: A multi-method evaluation. International Journal of Human-Computer Studies, 68, 1–21. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.ijhc​​s​.20​0​​9​.08.​​005 De Cremer, D., Van Dijke, M., Schminke, M., De Schutter, L., & Stouten, J. (2018). The trickle-down effects of perceived trustworthiness on subordinate performance. Journal of Applied Psychology, 103, 1335. https://doi​.org​/10​.1037​/apl0000339 de Reuver, M., Sørensen, C., & Basole, R. C. (2018). The digital platform: A research agenda. Journal of Information Technology, 33, 124–135. https​:/​/do​​i​.org​​/10​.1​​057​/s​​41265​​01​6-​​0033-​3 de Visser, E. J., Beatty, P. J., Estepp, J. R., Kohn, S., Abubshait, A., Fedota, J. R., & McDonald, C. G. (2018). Learning from the slips of others: Neural correlates of trust in automated agents. Frontiers in Human Neuroscience, 12, 309. https://doi​.org​/10​.3389​/fnhum​.2018​. 00309 de Visser, E. J., Peeters, M. M., Jung, M. F., Kohn, S., Shaw, T. H., Pak, R., & Neerincx, M. A. (2019). Towards a theory of longitudinal trust calibration in human–robot teams. International Journal of Social Robotics, 1–20. https​:/​/do​​i​.org​​/10​.1​​007​/s​​12369​​-019​-​​00596​-x de Vries, P., Midden, C., & Bouwhuis, D. (2003). The effects of errors on system trust, self-confidence, and the allocation of control in route planning. International Journal of Human-Computer Studies, 58, 719–735. https​:/​/do​​i​.org​​/10​.1​​007​/s​​12369​​-019​-​​00596​-x Dubois, A., & Gadde, L. E. (2002). Systematic combining: An abductive approach to case research. Journal of Business Research, 55, 553–560. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​148-​​ 2963(​00)00​195-8​ Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38, 1167–1230. https://doi​.org​/10​. 1177​/0149206312439327 Fulmer, C. A., & Ostroff, C. (2017). Trust in direct leaders and top leaders: A trickle-up model. Journal of Applied Psychology, 102, 648. https://doi​.org​/10​.1037​/apl0000189 Gefen, D. (2000). E-commerce: The role of familiarity and trust. Omega, 28, 725–737. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​305-​​0483(​00)00​021-9​

330  L. van der Werff, K. Blomqvist, and S. Koskinen

Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27, 51–90. https://doi​.org​/10​.2307​/30036519 Gefen, D., & Pavlou, P. (2006).The moderating role of perceived regulatory effectiveness of online marketplaces on the role of trust and risk on transaction intentions. ICIS 2006 Proceedings, 81. Giddens, A. (1990). The consequences of modernity. Stanford: Stanford University Press. Glikson, E., & Woolley, A. W. (2020). Human trust in artificial intelligence: Review of empirical research. Academy of Management Annals. https://doi​.org​/10​.5465​/annals​. 2018​.0057 Gu, J. C., Lee, S. C., & Suh, Y. H. (2009). Determinants of behavioral intention to mobile banking. Expert Systems with Applications, 36, 11605–11616. https​:/​/do​​i​.org​​/10​.1​​016​/ j​​.eswa​​.200​9​​.03​.0​​24 Hancock, P. A., Billings, D. R., Schaefer, K. E., Chen, J. Y., De Visser, E. J., & Parasuraman, R. (2011). A meta-analysis of factors affecting trust in human-robot interaction. Human Factors, 53, 517–527. https://doi​.org​/10​.1177​/0018720811417254 Hendershott, T., Jones, C. M., & Menkveld, A. J. (2011). Does algorithmic trading improve liquidity? The Journal of Finance, 66, 1–33. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1540​​-6261​​.201​0​​. 0162​​4.x Hodson, R. (2004). Organizational trustworthiness: Findings from the population of organizational ethnographies. Organization Science, 15, 432–445. https://doi​.org​/ 10​.1287​/orsc​.1040​.0077 Hong, I. B., & Cho, H. (2011). The impact of consumer trust on attitudinal loyalty and purchase intentions in B2C e-marketplaces: Intermediary trust vs. seller trust. International Journal of Information Management, 31, 469–479. https​:/​/do​​i​.org​​/10​.1​​016​/ j​​.ijin​​fomgt​​.20​11​​.02​.0​​01 Hyder, Z., Siau, K., & Nah, F. F. (2018). Use of artificial intelligence, machine learning, and autonomous technologies in the mining industry. MWAIS 2018 Proceedings, 43. Jarvenpaa, S., & Teigland, R. (2017). Introduction to trust, identity, and trusted systems in digital environments minitrack. In Proceedings of the 50th Hawaii international conference on system sciences. https://doi​.org​/10​.24251​/HICSS​.2017​.700 Jones, G. R., & George, J. M. (1998). The experience and evolution of trust: Implications for cooperation and teamwork. Academy of Management Review, 23, 531–546. https:// doi​.org​/10​.5465​/amr​.1998​.926625 Kelly, J. R., & Barsade, S. G. (2001). Mood and emotions in small groups and work teams. Organizational Behavior and Human Decision Processes, 86, 99–130. https://doi​.org​/ 10​.1006​/obhd​.2001​.2974 Ketokivi, M., & Choi,T. (2014). Renaissance of case research as a scientific method. Journal of Operations Management, 32, 232–240. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jom.​​201​4.​​03​.00​4 Kim, J., & Moon, J.Y. (1998). Designing towards emotional usability in customer interfaces— trustworthiness of cyber-banking system interfaces. Interacting with Computers, 10, 1–29. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​953-​​5438(​97)00​037-4​ Kosonen, M., Blomqvist, K., & Ellonen, R. (2008). Trust and its impersonal nature. In Encyclopedia of Networked and Virtual Organizations (pp. 1683–1690). IGI Global. https​:/​/ do​​i​.org​​/10​.4​​018​/9​​78​-1-​​59904​​-8​85-​​7​.ch2​​22 Kramer, R. M., & Lewicki, R. J. (2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annals, 4, 245–277. https​:/​/do​​i​.org​​/10​.5​​465​/1​​94165​​20​.20​​​10​.48​​7403

Trust Cues in Artificial Intelligence  331

Kroeger, F. (2015). The development, escalation and collapse of system trust: From the financial crisis to society at large. European Management Journal, 33, 431–437. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.emj.​​201​5.​​08​.00​1 Kroeger, F. (2019). Unlocking the treasure trove: How can Luhmann’s theory of trust enrich trust research? Journal of Trust Research, 9, 110–124. https​:/​/do​​i​.org​​/10​.1​​080​/ 2​​15155​​81​.20​​​18​.15​​52592​ Lane,C.,& Bachmann,R.(1997).Co-operation in inter-firm relations in Britain and Germany: The role of social institutions. British Journal of Sociology, 226–254. https://doi​.org​/ 10​.2307​/591750 Lankton, N. K., McKnight, D. H., & Tripp, J. (2015). Technology, humanness, and trust: Rethinking trust in technology. Journal of the Association for Information Systems, 16, 1. https://doi​.org​/10​.17705​/1jais​.00411 Lee, E. J. (2010). The more humanlike, the better? How speech type and users’ cognitive style affect social responses to computers. Computers in Human Behavior, 26, 665–672. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.chb.​​201​0.​​01​.00​3 Lee, J. D., & See, K. A. (2004). Trust in automation: Designing for appropriate reliance. Human Factors, 46(1), 50–80. https​:/​/do​​i​.org​​/10​.1​​518​/h​​fes​.4​​6​.​1​.5​​0​_303​​92 Lewis, J. D., & Weigert, A. (1985). Trust as a social reality. Social Forces, 63, 967–985. https:// doi​.org​/10​.2307​/2578601 Li,Y. M., & Yeh,Y. S. (2010). Increasing trust in mobile commerce through design aesthetics. Computers in Human Behavior, 26, 673–684. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.chb.​​201​0.​​01​.00​4 Lindebaum, D., Vesa, M., & den Hond, F. (2020). Insights from “The Machine Stops” to better understand rational assumptions in algorithmic decision making and its implications for organizations. Academy of Management Review, 45, 247–263. https://doi​. org​/10​.5465​/amr​.2018​.0181 Lipponen, J., Kaltiainen, J., van der Werff, L., & Steffens, N. K. (2020). Merger-specific trust cues in the development of trust in new supervisors during an organizational merger: A naturally occurring quasi-experiment. The Leadership Quarterly, 31, 101365. https​:/​/ do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​19​.10​​1365 Luhmann, N. (1979). Trust and power. Chichester: Wiley. Lumineau, F., & Schilke, O. (2018). Trust development across levels of analysis: An embedded-agency perspective. Journal of Trust Research, 8, 238–248. https​:/​/do​​i​.org​​/10​. 1​​080​/2​​15155​​81​.20​​​18​.15​​31766​ Lyu, S. C., & Ferrin, D. L. (2018). Determinants, consequences and functions of interpersonal trust within organizations: What is the empirical evidence? In R. H. Searle, A. M. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 65–104). Abingdon: Routledge. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20, 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.19​​95​.​ 95​​08080​​335 McEvily, B., Perrone,V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, 14, 91–103. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​.​1​.9​​1​.128​​14 McKnight, D. H., Carter, M., Thatcher, J. B., & Clay, P. F. (2011). Trust in a specific technology: An investigation of its components and measures. ACM Transactions on Management Information Systems, 2, 12. https://doi​.org​/10​.1145​/1985347​.1985353 McKnight, D. H., Cummings, L. L., & Chervany, N. L. (1998). Initial trust formation in new organizational relationships. Academy of Management Review, 23, 473–490. https:// doi​.org​/10​.5465​/amr​.1998​.926622

332  L. van der Werff, K. Blomqvist, and S. Koskinen

Melo, C. D., Marsella, S., & Gratch, J. (2016). People do not feel guilty about exploiting machines. ACM Transactions on Computer-Human Interaction, 23, 8. https://doi​.org​/10​. 1145​/2890495 Merritt, S. M., & Ilgen, D. R. (2008). Not all trust is created equal: Dispositional and history-based trust in human-automation interactions. Human Factors, 50, 194–210. https://doi​.org​/10​.1518​/001872008X288574 Molnar, L. J., Ryan, L. H., Pradhan,A. K., Eby, D.W., Louis, R. M. S., & Zakrajsek, J. S. (2018). Understanding trust and acceptance of automated vehicles: An exploratory simulator study of transfer of control between automated and manual driving. Transportation Research Part F: Traffic Psychology and Behaviour, 58, 319–328. https​:/​/do​​i​.org​​/10​.1​​016​/ j​​.trf.​​201​8.​​06​.00​4 OECD (2019). Artificial intelligence in society. Paris: OECD Publishing. Retrieved from https://www​.oecd​-ilibrary​.org Pejsa, T., Andrist, S., Gleicher, M., & Mutlu, B. (2015). Gaze and attention management for embodied conversational agents. ACM Transactions on Interactive Intelligent Systems, 5, 3. https://doi​.org​/10​.1145​/2724731 Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23, 393–404. https:// doi​.org​/10​.5465​/amr​.1998​.926617 Schaefer, K. E., Chen, J.Y., Szalma, J. L., & Hancock, P. A. (2016). A meta-analysis of factors influencing the development of trust in automation: Implications for understanding autonomy in future systems. Human Factors, 58, 377–400. https://doi​.org​/10​.1177​/ 0018720816634228 Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007).An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32, 344–354. https://doi​. org​/10​.2307​/20159304 Shamir, B., & Lapidot, Y. (2003). Trust in organizational superior: Systemic and collective considerations. Organization Studies, 24, 463–491. https​:/​/do​​i​.org​​/10​.1​​177​/ 0​​17084​​06030​​​24003​​912 Shank, D. B. (2014). Impressions of computer and human agents after interaction: Computer identity weakens power but not goodness impressions. International Journal of HumanComputer Studies, 72, 747–756. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.ijhc​​s​.20​1​​4​.05.​​002 Shank, D. B., & DeSanti,A. (2018).Attributions of morality and mind to artificial intelligence after real-world moral violations. Computers in Human Behavior, 86, 401–411. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.chb.​​201​8.​​05​.01​4 Shapiro, S. P. (1987). The social control of impersonal trust. American Journal of Sociology, 93, 623–658. https://doi​.org​/10​.1086​/228791 Shrestha, Y. R., Ben-Menahem, S. M., & Von Krogh, G. (2019). Organizational decisionmaking structures in the age of artificial intelligence. California Management Review, 61, 66–83. https://doi​.org​/10​.1177​/0008125619862257 Söllner, M., Hoffmann, A., & Leimeister, J. M. (2016). Why different trust relationships matter for information systems users. European Journal of Information Systems, 25(3), 274– 287. https://doi​.org​/10​.1057​/ejis​.2015​.17 Sousa, M. J., & Rocha, Á. (2019). Skills for disruptive digital business. Journal of Business Research, 94, 257–263. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jbus​​res​.2​​​017​.1​​2​.051​ Stewart, K. J. (2003). Trust transfer on the world wide web. Organization Science, 14, 5–17. https​:/​/do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​4​​.1​.5​​.1281​0

Trust Cues in Artificial Intelligence  333

Sydow, J. (1998). Understanding the constitution of interorganizational trust. Trust within and between organizations: Conceptual issues and empirical applications. In C. Lane & R. Bachman (Eds.), Trust Within and Between Organizations (pp. 31–63). Oxford: Oxford University Press. Tang, Z., Hu, Y., & Smith, M. D. (2008). Gaining trust through online privacy protection: Self-regulation, mandatory standards, or caveat emptor. Journal of Management Information Systems, 24(4), 153–173. https​:/​/do​​i​.org​​/10​.2​​753​/M​​IS074​​2​-1​22​​22404​​06 Thompson, R. F.,Valdes, G., Fuller, C. D., Carpenter, C. M., Morin, O., Aneja, S., Lindsay,W. D., Aerts, H. J., Agrimson, B., Deville Jr, & Rosenthal, S. A. (2018). Artificial intelligence in radiation oncology: A specialty-wide disruptive transformation? Radiotherapy and Oncology, 129, 421–426. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.rado​​nc​.20​​​18​.05​​.030 Todorov, A., Mandisodza, A. N., Goren, A., & Hall, C. C. (2005). Inferences of competence from faces predict election outcomes. Science, 308, 1623–1626. https://doi​.org​/10​.1126​/ science​.1110589 Tuch, A. N., Bargas-Avila, J. A., & Opwis, K. (2010). Symmetry and aesthetics in website design: It’s a man’s business. Computers in Human Behavior, 26, 1831–1837. https​:/​/do​​i​. org​​/10​.1​​016​/j​​.chb.​​201​0.​​07​.01​6 van der Werff, L., Freeney, Y., Lance, C. E., & Buckley, F. (2019). A trait-state model of trust propensity: Evidence from two career transitions. Frontiers in Psychology, 10, 2490. https://doi​.org​/10​.3389​/fpsyg​.2019​.02490 van der Werff, L., Real, C., & Lynn, T. (2018). Individual trust and the internet. In R. H. Searle, A. M. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 39–407). Abingdon: Routledge. https://doi​.org​/10​.4324​/9781315745572​-27 van Kleef, G. A. (2014). Understanding the positive and negative effects of emotional expressions in organizations: EASI does it. Human Relations, 67, 1145–1164. https://doi​. org​/10​.1177​/0018726713510329 van Knippenberg, D. (2018). Reconsidering affect-based trust: A new research agenda. In R. H. Searle, A.-M. I. Nienaber, & S. B. Sitkin (Eds.), The Routledge Companion to Trust (pp. 3–13). Abingdon: Routledge. Verberne, F. M., Ham, J., & Midden, C. J. (2012).Trust in smart systems: Sharing driving goals and giving information to increase trustworthiness and acceptability of smart systems in cars. Human Factors, 54, 799–810. https://doi​.org​/10​.1177​/0018720812443825 von Krogh, G. (2018). Artificial intelligence in organizations: New opportunities for phenomenon-based theorizing. Academy of Management Discoveries, 4, 404–409. https:// doi​.org​/10​.5465​/amd​.2018​.0084 Williams, M. (2001). In whom we trust: Group membership as an affective context for trust development. Academy of Management Review, 26, 377–396. https://doi​.org​/10​.5465​/ amr​.2001​.4845794 Yagoda, R. E., & Gillan, D. J. (2012).You want me to trust a ROBOT? The development of a human–robot interaction trust scale. International Journal of Social Robotics, 4, 235–248. https​:/​/do​​i​.org​​/10​.1​​007​/s​​12369​​-01​2-​​0144-​0 Yin, M.,Vaughan, J.W., & Wallach, H. (2019). Understanding the effect of accuracy on trust in machine learning models. CHI Conference on Human Factors in Computing Systems Proceedings (CHI 2019). https://doi​.org​/10​.1145​/3290605​.3300509n Zucker, L. G. (1986). Production of trust. Institutional sources of economic structure, 1840–1920. Research in Organizational Behavior, 8, 53–111.

14 EMPLOYEE TRUST IN ORGANIZATIONS ACROSS CULTURES A Multilevel Model S. Arzu Wasti and Çetin Önder

Introduction A central concern of organizational studies is to identify the determinants of intra-organizational cooperation, coordination, and control; and the role of trust in this quest is well established (Fulmer & Gelfand, 2012). Yet, how people in different cultures understand, build, and sustain trust is still an underexplored terrain (Dietz et al., 2010). In this chapter, we present a multilevel model that seeks to explain how the societal context affects organizations and individuals in the development of employee trust in the organization. As an individual-level variable, employee trust in an organization is defined as “a psychological state comprising of the intention to accept vulnerability based on positive expectations of an organization” (Fulmer & Gelfand, 2012, p. 1174). Trust in one’s organization is argued to be of greater importance than interpersonal trust in view of the realities of contemporary workplaces such as the increased complexity of the operational environments facing many organizations, the decreased opportunities for face-to-face communication, and the increased use of temporary and virtual teams (Vanhala & Dietz, 2015). However, compared to trust in supervisors, research on employee trust in organizations is limited. Although evidence converges on the importance of management competence, organizational support, and fairness as antecedents across a variety of countries (e.g., Hodson, 2004; Pillai et al., 2001; Searle et al., 2011), for the most part, this literature has not problematized the role of the societal context on employee trust. The few studies that have done so appear to represent two distinct and disconnected approaches to cross-national research, which we label as the cultural versus the institutional approach. DOI: 10.4324/9780429449185-14

Trust in Organizations Across Cultures  335

The cultural approach has explored the moderating influence of cultural values on the antecedents of trust. For instance, Caldwell and Clapham (2003) proposed organizational ability (operationalized as competence, financial balance, and quality assurance) to be more important in individualistic compared to collectivist cultures. Drawing on North American and East Asian samples, they supported this prediction for competence, but not for financial balance or quality assurance. They further predicted interactional courtesy, akin to interactional justice, to be more important for collectivists than individualists – an argument also put forth by Pillai and colleagues (2001) regarding Indians, another collectivist culture. However, neither study was able to empirically substantiate this expectation; in fact, Caldwell and Clapham (2003) found interactional courtesy to be more important to North Americans than East Asians. In a similar comparison, Li and Cropanzano (2009) argued that the relationship between justice perceptions and organizational trust would be stronger in individualistic North America. Drawing on high power distance, they further proposed that employees in East Asia would be less attuned to unfairness committed by authority figures. Their meta-analysis, however, did not yield significant differences across the groups. In contrast, the institutional approach has challenged the relevance, or at least, the importance of employee trust in organizations in certain contexts. For instance, Zhang et al. (2008) argued that under volatile or weak political-economic contexts, such as those typically found in ex-communist or developing countries (Pearce et al., 2000), employees are unlikely to expect a stable relationship with their employers and tend to rely on personal relationships. This renders trust in supervisors the main currency in the employment relationship and results in a generalized distrust of organizations (Child & Möllering, 2003). This brief review suggests the literature on the influence of the societal context on employee trust in organizations appears to be not only sparse but also sporadic. The institutional stream is limited to the implications of weak institutions regarding the prevalence and consequences of particularistic networks, where particularistic refers to the primacy of personal relationships as opposed to impersonal, merit-based appointments (Pearce et al., 2000). How institutional characteristics such as the strength of the legal system shape organizational or individual attributes pertinent to trust is not directly addressed. Relatedly, as this perspective focuses on aggregate entities, it lacks micro-foundations that could help explain how organizations or employees respond to (e.g., ignore or defy) institutional influences.The cultural stream, which has yielded inconsistent results, solely focuses on values, typically by reference to national scores on large-scale survey studies (e.g., Hofstede, 2001). As noted by Peterson and Barreto (2018), the assumption underlying such research is that individuals in a particular society embrace the societal values (e.g., all individuals in a collectivist culture value societal collectivism), which has been empirically shown to be unfounded (e.g., Fischer & Schwartz, 2011). Research on values has also treated culture as an ingrained, general, or stable reaction to stimuli, which is inconsistent with the

336  S. Arzu Wasti and Çetin Önder

observed adaptation of individuals to context-specific cues, such as those exhibited by expatriates or bicultural individuals (Leung & Morris, 2015). With multicultural life experiences becoming common (Chao & Moon, 2005), a useful model of trust in organizations should elucidate the experience of, for example, a Turkish person who completes his education at a German high school in Istanbul and a university in the United States (US) and works for a Dutch multinational enterprise (MNE) back in Turkey. In this chapter, we present an overarching multilevel model comprised of the societal, organizational, individual, and event levels to explicate the mechanisms of employee trust in organizations (see Figure 14.1). We adopt the framework by Gillespie and Dietz (2009) and hold that at the center of the model is the organization, described by its internal system components such as strategy, leadership, climate, and human resource (HR) practices. Each of these components generates events, which send cues to employees about the organization’s trustworthiness. To develop this framework into a multilevel model, we integrate the research on comparative institutions, international business, and cultural psychology (e.g.,

Societal level

Socio-institutional system Ideologies, values, norms, beliefs

Political system

Financial system

Human capital development system

Corporate governance system

Organizational level

P1a

Organizational situationscape Leadership and management practice

Organizational culture and climate

Strategy

P2a, P2b, P2c

Structures, policies, and practices

Individual level

P3

Individual carriers of culture Schemas

Social axioms

Norm perceptions

Values

P2a, P2b

Trust in organization

P1b

Event level

P4a, P4b

Events

FIGURE 14.1 A

P4c

Episodic trustworthiness assessment

multilevel model of employee trust in organizations across cultures.

Trust in Organizations Across Cultures  337

Gelfand et al., 2008; Leung & Morris, 2015; Whitley, 1999) to offer four sets of propositions. First, regarding the influence of the societal level on organizations, we argue that every organization is embedded in a particular socio-institutional context, which shapes and binds its internal system components (e.g., leadership, climate) in predictable ways (P1a). The consequent constellation of the internal system components, which we call the organizational situationscape, in turn will determine the content and coherence of the events generated (P1b). Second, we delineate the cross-level influences that relate to the societal and individual-level variables. We hold that the socio-institutional context will have cross-level effects on employee values and social axioms (beliefs) associated with generalized trust in organizations (P2a–b). We further argue that in each socio-institutional context, employees will have norms and schemas as to typical workways (P2c), which refer to workplace beliefs, mental models, and practices that embody a society’s ideas about what is true, good, and efficient within the domain of work (Sanchez-Burks & Lee, 2007). Third, we advance cross-level propositions between the organizational and individual levels, whereby we discuss the role of organizational situationscapes in shaping individual-level attributes (P3). Our fourth and final set of propositions address how these carriers of culture at the individual level, i.e., norms, schemas, and values will influence the assessment of organization-generated events in terms of their trustworthiness to ultimately determine employee trust (P4a–c). In what follows, we first provide an overview of the theoretical model by defining every concept at each level. To explain the theoretical constructs and their relationships, we identify two prototypical cases at the societal level, namely, modern versus neotraditional socio-institutional systems, and explicate the propositions outlined above around these cases. Thus, the propositions we offer in this chapter only exemplify the model and it is possible to advance a broader or an alternative set of propositions to examine other socio-institutional systems. The definitions of our theoretical constructs, the depictions of the exemplar cases at each level, as well as the corresponding propositions that posit cross-level relationships are presented in Table 14.1. Upon detailing these propositions, we conclude with a discussion of our contributions and offer fruitful future research directions.

A Multilevel Model of Employee Trust in Organizations: An Overview To provide a better understanding of the premises the model is built on, we begin by presenting each level.We first explain how we conceptualize the societal level, and then briefly summarize the framework by Gillespie and Dietz (2009) to describe the organizational level. Because trustworthiness cues are generated from organizational events, we proceed to a discussion of the event level and conclude with a description of the individual level.

Definition

Exemplification

Cross-level effects

Configurations of cultural ideas Modern systems (strong state, reliable legal system, On the organizational (P1a) and (ideologies, beliefs, norms, and stable financial system, collaborative human individual (P2a, P2b, and P2c) values) and societal institutions capital development system, professional levels (political system, financial system, control over organizations) informed by human capital development individualism and low power-distance ideas system, and corporate versus neotraditional systems (weak state, governance system) ineffectual rule of law, unstable financial system, weak human capital development system, family-controlled businesses) informed by in-group collectivism and high power distance Organizational level Constellations of leadership Constellations that uphold collaboration, On the event (P1b) and individual (organizational and management practices, participation, and meritocratic universalism (P3) levels situationscape) organizational culture and versus constellations that uphold climate, organizational strategy, opportunism, paternalism, and particularism and structures, policies, and processes Individual level Psychological culture (individuals’ Social axiom of low social cynicism, On the event–trustworthiness (individual carriers of own endorsement of cultural individualism, and low power distance values, relationship (P4a, P4b) and culture) values, descriptive and prescriptive with norms and schemas that are built on trustworthiness–trust in norms, beliefs/social axioms, and egalitarianism, meritocratic universalism, and organization relationship (P4c) schemas) and subjective cultural regulatory compliance versus social axiom of press (individuals’ perceptions of high social cynicism, in-group collectivism, cultural values, descriptive and and high power distance values, with norms prescriptive norms, and beliefs/ and schemas that are built on hierarchism, social axioms) particularism, and opportunism

Societal level (socio-institutional context)

Level

TABLE 14.1 Concept definitions, examples, and a summary of propositions

338  S. Arzu Wasti and Çetin Önder

Event level (events)

Events emanating from Events signaling high trustworthiness (e.g., high task On employee trust in organization organizational situationscapes performance or ethical behaviors of leaders, such as task or team-maintenance transparent leader communication, consistent performance of leaders, promises application of rewarding mechanisms, made by leaders, HR practices, successful implementation of coherent and organizational rituals, changes long-term strategies, consistent compliance in organizational strategy or with external regulations) versus events structure signaling low trustworthiness (e.g., incompetent or unethical leader behavior, cheating and exploitation, violation of procedural justice, incoherent or ineffective strategy implementation, erratic compliance with external regulations) Trust in Organizations Across Cultures  339

340  S. Arzu Wasti and Çetin Önder

Societal Level Drawing from recent discussions in comparative institutional (Fainshmidt et al., 2018) and international business research (Beugelsdijk et al., 2017), we conceptualize the societal level as configurations or sets of characteristics that hang together, rather than treating cultural dimensions and societal institutions separately. Accordingly, we embrace the definition of culture as the explicit and implicit patterns of historically derived and selected ideas (i.e., ideologies, values, norms, beliefs), which are embodied in societal institutions and practices, and inform interactions and, ultimately, individuals (Hamedani & Markus, 2019). Given our focus on employee trust in business organizations, the societal institutions of particular relevance are the political, financial, human capital development, and corporate governance systems (Fainshmidt et al., 2018) and we distinguish between modern and neotraditional configurations of these systems. A modern system, typically informed by ideas related to individualism and low power distance (Hofstede, 2001; House et al., 2004), comprises societal institutions that are rational-legal, that is, formally constituted and impersonally functioning (i.e., universalistic; Pearce et al., 2000). An example would be the Anglo-Saxon cultural zone, including the US and the United Kingdom among others. In contrast, neotraditional systems typically feature practices that are particularistic, based on who the person in question is, which are mutually reinforcing with ideas of ingroup collectivism and high power distance (Basabe & Ros, 2005). These systems, such as those observed in Greece, Mexico, South Korea, and Turkey (Hotho, 2014), are labeled neotraditional because as opposed to traditional systems, they do exhibit characteristics of modern systems but these characteristics are only superficially related to the usual functioning of these systems.

Organizational Level At the organizational level, we refer to the four internal system components outlined by Gillespie and Dietz (2009), which include (1) leadership and management practices, (2) organizational culture and climate, (3) organizational strategy, and (4) structures (e.g., formalization), policies, and processes (i.e., procedures governing decision-making, communication, employee conduct, and HR management). Although leaders are nested in organizations, given there are four levels in the model and our priority is to explicate societal effects, we have opted to represent leaders as organizational agents (Levinson, 1965) at the organizational level.We refer to the constellations of the four internal system components as organizational situationscapes.

Event Level We posit that organizational situationscapes generate events (e.g., Fehr et al., 2017) that signal the organization’s trustworthiness. For instance, explicit promises made

Trust in Organizations Across Cultures  341

by organizational leaders or their perceived fulfillment of functions critical to task and team maintenance (Burke et al., 2007; Gillespie & Dietz, 2009) can constitute events regarding organizational trustworthiness. Employees will form further assessments based on events generated by the other system components (i.e., culture, climate, strategy, structure, policies, and procedures). In particular, HR practices tend to be interpreted as structural signals regarding an organization’s intentions towards its workers (Vanhala & Dietz, 2015).These events will breed or impede employee trust as mediated by their episodic assessments of organizational trustworthiness – more specifically, its ability, benevolence, and integrity (Mayer et al., 1995). Gillespie and Dietz (2009) defined organizational ability as an organization’s competencies and characteristics that enable it to function effectively to achieve its goals and meet its responsibilities. Organizational integrity refers to an organization’s adherence to legal, ethical, and moral principles, and codes of conduct acceptable to stakeholders, such as honesty and fairness. Benevolence at the organizational level entails systems, policies, and practices that signal the organization’s genuine care and concern for the well-being of its stakeholders.

Individual Level At the individual level, in addition to employee trust in organizations, we consider psychological culture, which refers to individuals’ own endorsement of cultural values, norms, and beliefs as well as subjective cultural press, which refers to individuals’ perceptions of cultural values, norms, and beliefs (Gelfand et al., 2008). In what follows, sometimes we specify which individual-level variable drives the proposition, but often we do not make such a differentiation because these variables are expected to interact. For instance, values as desirable goals may activate norms, or norms may render particular schemas, i.e., cognitive templates that guide individual interpretations, to be more salient (Leung & Morris, 2015). Nonetheless, we keep the constructs distinct because emergent research suggests that their influence on behavior varies depending on the characteristics of the situation, such as its ambiguity or privacy (Leung & Morris, 2015). We return to this point in the discussion and, for now, we turn to our propositions.

Propositions Cross-Level Effects of Societal Context on Organizations As mentioned, at the societal level we differentiate between modern versus neotraditional configurations of societal institutions. Modern political systems exhibit strong states and reliable legal systems that function as third-party guarantors (Bachmann & Inkpen, 2011). In these systems, authority derives from formally credentialed expertise, rather than personal characteristics (Whitley, 2005). Countries with modern financial systems additionally experience less uncertainty

342  S. Arzu Wasti and Çetin Önder

in their financial markets. Some of these countries (e.g., Germany) also boast corporate governance systems that mandate codetermination and human capital development systems based on collaboration between the state, employers, and labor unions, reflecting a more egalitarian culture (Sorge, 2005; Whitley, 2008). In contrast, neotraditional political systems comprise the mutually reinforcing characteristics of weaker states, ineffectual rule of law, and in-group collectivism (Pearce et al., 2000; Whitley, 2005). In these systems, states are more likely to be predatory (i.e., failing to provide welfare or rather destroying it to enrich a greedy ruling elite) and authority is not necessarily based on formal credentials or fair elections (Hotho, 2014; Whitley, 2005). Neotraditional financial systems are typically underdeveloped and fragile, undergoing frequent crises. These systems do not have collaborative human capital development arrangements or codetermination. Regarding corporate governance, neotraditional systems are typically based on direct control and coordination by shareholding family members (Whitley, 1999). Considering that business organizations are open systems that embody and reinforce the dominant societal culture, we argue that there are cross-level societal effects on organizational situationscapes (P1a). Generally speaking, the configuration of strong states, reliable legal systems, and stable financial markets coupled with ideas representing individualism and low power distance makes it easier for managers to assess competence and ensure observance of managerial directives or expectations by employees through impersonal means (Whitley, 1999). Along with empowering leadership and meritocratic policies and processes (Fitzsimmons & Stamper, 2014; Pearce et al., 2009; Whitley, 2005) modern systems tend to afford organizational cultures characterized by egalitarianism and participation. Collaborative human capital development and corporate governance systems that entail codetermination further facilitate delegation of power to employees and practices such as job enrichment and vocational training, as well as policies such as organizational careers (Sorge, 2005; Whitley, 2008). Where these systems prevail, competitive strategies based on flexible and high-quality production that require employee participation are more probable (Thelen, 2010; Whitley, 2008). Therefore, collaborative systems also support organizational cultures that foster organizational learning and innovation. In contrast, the social institutions that typify neotraditional systems coupled with ideas of high power distance and in-group collectivism tend to discourage meritocratic policies and processes. In neotraditional systems, business organizations are more likely to display paternalistic leadership characterized by leader protection in return for follower loyalty in relationships between ownermanagers and salaried top managers (Whitley, 2008) as well as the immediate supervisor–subordinate relationships (Fitzsimmons & Stamper, 2014). Due to the absence of a collectively governed human capital development system, business organizations typically disregard strategies based on specialized employee skills, opting for opportunistic diversification or quick disbanding of preexisting businesses that require little input from employees for success (Whitley, 2008). Thus,

Trust in Organizations Across Cultures  343

organizational cultures and policies in neotraditional societies are unlikely to uphold egalitarianism, employee development, or voice. These arguments lead us to our first proposition regarding organizational situationscapes that prevail in modern versus neotraditional institutional systems: Proposition 1a: Modern socio-institutional systems mutually reinforced by societal individualism and low power distance will generally afford organizational situationscapes characterized by long-term collaborative strategies, participative leadership, and meritocratic universalism whereas neotraditional socio-institutional systems mutually reinforced by societal in-group collectivism and high power distance are more likely to afford organizational situationscapes with opportunistic strategies, paternalistic leadership, and particularism. In the model, organizational situationscapes are argued to generate events (e.g., revision of business strategies, organizational rituals, enactment of HR policies, and task-, relationship-, or change-oriented behaviors by leaders; Burke et al., 2007; Gillespie & Dietz, 2009) that may provide input for episodic assessment of organizational trustworthiness (P1b). Gillespie and Dietz (2009) have suggested that events that signal incoherence or ineffectiveness of organizational strategy, disregard for certain stakeholders, absence, inconsistent application, or abandonment of certain policies or processes toward restricting incompetent, and unfair or unethical behaviors are likely to be considered untrustworthy. Building on the arguments that led to our first proposition, we posit that such events are more likely to be produced by organizational situationscapes that develop under neotraditional systems. Indeed, Pearce and her colleagues (2000) argued that under particularistic systems, perceptions of procedural justice are undermined and work relationships are characterized by pervasive distrust resulting in dysfunctional behaviors such as withholding of information, exploitation of others, and cheating (Pearce et al., 2009). In contrast, events embodying transparent and two-way communication, compliance with external regulatory codes of conduct, consistent application of rewarding and sanctioning mechanisms, or persistent pursuit of long-term strategies (e.g., Gillespie & Dietz, 2009; Pearce et al., 2009) are more likely to be generated under modern systems and to signal greater trustworthiness. Thus, we propose the following: Proposition 1b: Organizational situationscapes that develop under neotraditional and modern systems are respectively more likely to generate events that signal lower and higher trustworthiness.

Cross-Level Effects of Societal Context on Individuals Having discussed the influence of the societal context on organizations, which in turn has implications for organizational events, we now consider cross-level

344  S. Arzu Wasti and Çetin Önder

linkages that relate the societal level with the individual level. We begin with the proposition explicating the link between societal context, the social axiom of social cynicism, and trust in organizations (P2a). Social axioms are a person’s beliefs about their social world and life goals (Leung & Bond, 2004). Leung and Bond (2004) found that economically, politically, and socially unstable societies tended to be higher in social cynicism, a social axiom defined as a negative view of people and a mistrust of social institutions. Li and Leung (2012) argued that undesirable societal conditions may lead individuals to feel a psychological contract breach with their society in view of their natural right to live a decent life. Indeed, they found that the subjective evaluation of societal conditions had a lagged effect on social cynicism over time. This is in line with the trust literature, which suggests that in contexts characterized by economic instability or a failure of legal or regulatory enforcement, generalized trust may be eroded (Gillespie & Dietz, 2009), breeding distrust in organizations (Child & Möllering, 2003). In contrast, individuals in countries characterized by economic development, broad access to education and information, and well-functioning democratic institutions tend to endorse higher levels of generalized trust (Almakaeva et al., 2018), including greater trust in business organizations (Pearce et al., 2009). Therefore, we offer the following: Proposition 2a: The societal context will have cross-level effects on the social axiom of social cynicism at the individual level. Individuals in societies with modern (versus neotraditional) socio-institutional systems will generally have lower (versus higher) social cynicism, which in turn will foster (versus hamper) employee trust in organizations. Research also suggests that high societal in-group collectivism may be unconducive to trust in organizations (P2b). Yamagishi (1998) reasoned that the strong norms of identification with one’s in-groups and the ensuing in-group favoritism create societies characterized by low generalized trust, which in turn impedes individuals’ belief in human benevolence. In contrast, individualistic societies’ emphasis on independence renders group memberships less consequential to trust formation, thereby facilitating the establishment of relationships between relative strangers (Huff & Kelley, 2003). Similarly, Bohnet et al. (2010) argued that individuals in socio-institutional contexts that foster tight-knit, collectivist relationships rather than featuring formal trust-generating mechanisms such as a strong, regulatory state (i.e., relationship- versus rule-based trust) have higher levels of betrayal aversion. This betrayal aversion is conducive to sustaining trust among individuals with mutual personal attachment, rather than fostering confidence in institutions or office-holders. Proposition 2b: Societal in-group collectivism and individualism will have cross-level effects at the individual level. Individuals in in-group collectivist (versus individualist)

Trust in Organizations Across Cultures  345

societies will generally endorse social axioms, values, and norms that hamper (versus foster) the development of trust in organizations. The above two propositions addressed how societal effects might shape individuals’ generalized distrust in organizations. We further posit the societal context to have other cross-level effects at the individual level (P2c), which will influence the interpretation of specific organizational events (P4). At the societal level, widespread regularities in a group’s beliefs, values, and behaviors constitute norms, and individuals develop subjective expectancies around these norms (Morris et al., 2015). These subjective expectancies, which are called descriptive norms, comprise perceptions as to what is typically felt, thought, or done in a cultural group; prescriptive norms are perceptions as to what the cultural group approves or disapproves of. Similarly, individuals hold culturally conferred schemas, which reflect generalized collections of knowledge deriving from experience. Norms and schemas not only guide one’s own behavior but also form the basis for predicting or interpreting others’ behaviors, particularly in ambiguous or social situations. Of note, there appears to be greater agreement within societies with respect to cultural norms and schemas as opposed to cultural values (Leung & Morris, 2015). Research from different areas speaks to the pervasiveness of cultural norms and schemas regarding workways. For instance, Sanchez-Burks and Lee (2007) described how modern-day American workplace professionalism norms in terms of limited socializing at work could be traced to Calvinist theology of the 17th century. More generally, Rousseau (2001) argued that psychological contracts, which are subjective beliefs that comprise the exchange agreement between an employee and the employing firm start forming at the pre-employment stage as a result of societal socialization and prior work experiences including those of close others. Kramer (2010) noted that unlike interpersonal trust, which benefits from personal knowledge about others, trust in organizations is built on schematic knowledge and stereotypic beliefs regarding the organization and what membership in it signals about other members’ trustworthiness. We propose that the socio-institutional context (e.g., weak legal system) and the characteristics of the dominant organizational situationscapes it affords (e.g., opportunism) will shape employees’ norms and schemas regarding organizations, their working conditions, workforces, and workways. To develop this proposition, rather than detail specific practices or policies, we focus on the principles that may underlie descriptive norms and schemas and offer the following: Proposition 2c:The societal context will have cross-level effects on norms and schemas at the individual level. Individuals in societies with neotraditional socio-institutional systems will generally hold norms and schemas of hierarchism, particularism, and opportunism whereas individuals in societies with modern socio-institutional systems will generally hold norms and schemas of egalitarianism, meritocratic universalism, and regulatory compliance.

346  S. Arzu Wasti and Çetin Önder

Cross-Level Effects of Organizational Situationscapes on Individuals So far, our discussion has been limited to a mono-cultural world, as we have focused on explaining how a particular societal context would shape organizations and individuals.With this backdrop, we now turn to the scenario of a multicultural setting, which is a more realistic depiction of any particular work context today, albeit at different rates. In the first proposition (P1a), we argued that in each societal context, a dominant organizational situationscape is likely to prevail and indeed, research evidence substantiates this notion. For instance, paternalism is commonly observed in the collectivist and power distant cultures of Asia, the Middle East, Latin America, and Africa (Aycan, 2006). Nonetheless, in each socio-institutional context, there is variance that originates from the presence of MNEs, joint ventures, or emergent organizational forms (e.g., high-tech startups) as well as the possibility that some local organizations will be more likely to circumvent institutions, enjoy niches, be guided by supra-national institutional mandates, or import foreign practices (e.g., kaizen; Hotho, 2014). Depending on the extent of variance and their own exposure, employees may embrace values, norms, and schemas beyond their heritage. For instance, employees working in an organization with strong cooperative rewards, such as profit-sharing or team-based performance appraisals, may adopt collectivist norms and values irrespective of societal tendencies (Gelfand et al., 2008). In fact, Wan and Lu (2014) argued that individual-level culture is sometimes shaped more strongly by the proximal context of organizational situationscapes than by the distal societal culture. Thus, we posit: Proposition 3: In any given societal context, there will be variance within and across organizational situationscapes with respect to organizational norms and values, which will have cross-level effects on norms, schemas, and values at the individual level by way of increasing the cultural repertoires that employees can draw from.

Cross-Level Moderation Effects of Individual Carriers of Culture Our last set of propositions address the moderating effect of culture at the individual level on trustworthiness assessments of organizational events embedded in a particular socio-institutional context. Depending on the organization the employee is working at, these events may reflect the dominant situationscape and may activate an employee’s native norms or schemas, or may signal another cultural organizational situationscape, which may activate corresponding norms, schemas, or values that the employee may have acquired with experience (Chao & Moon, 2005).Without being able to do justice, we offer three propositions that recognize the complexity of contemporary workplaces.

Trust in Organizations Across Cultures  347

With respect to psychological culture, which is individuals’ own endorsement of cultural values, norms, and beliefs (Gelfand et al., 2008), we note the compelling evidence on the importance of cultural fit or match in predicting desirable work-related outcomes. For instance, in a study across 24 countries, House et al. (2013) showed that the fit between subordinates’ cultural norms (e.g., employee participation) and leader behavior (e.g., employee empowerment) was critical for CEOs’ effectiveness, measured as employees’ commitment, effort, and team solidarity. Moreover, leaders who violated cultural norms were deemed less effective than those who conformed to these norms. Similarly, Peretz and Fried (2012) reported that using multiple raters in performance appraisal was most acceptable to employees in countries low on power distance and high on future orientation and individualism. Zhang et al. (2015) found that impersonal, merit-based HR practices imported from the West, such as pay for performance, were less effective in contexts traditionally functioning on particularistic networks, such as the Chinese guanxi. This evidence attests to the desirability of the match between the relational elements of the dominant organizational situationscape and societal culture (e.g., empowering leaders in low power distance cultures). However, as with the variance in organizational situationscapes, in any given context we conceive that some employees will have a higher likelihood of exhibiting their normative native culture (e.g., having an interdependent self in a collectivist culture or being statusconscious in a high power distance culture), whereas others (e.g., biculturals) may draw upon multiple cultural legacies (Chao & Moon, 2005). Indeed, recent studies are pointing to the variance or transition emerging in certain contexts. For example, Zhang, Huai, and Xie (2015) have questioned the viability of paternalistic leadership in China as younger generations are becoming more Westernized and getting acquainted with Western firms. In contrast, Chinese employees who are traditional appear to be more accepting of the hierarchical role relationships prescribed by Confucian social ethics (Farh et al., 2007). Therefore, underlining the need to take into account within-country variance, we advance a cultural match proposition with respect to relational events: Proposition 4a:The relationship between events and trustworthiness assessments will be moderated by employees’ schemas, norms, and values with respect to hierarchism, particularism, and in-group collectivism. To the degree to which there is a match between employees’ schemas, norms, and values and the relational elements of the events, perceptions of organizational trustworthiness will be enhanced. The arguments that led to P3 indicate that in any particular socio-institutional context employees may experience the full range of trust breeding or impeding organizational events, albeit with differential probabilities. Moreover, the literature on employee trust in organizations evokes several universal antecedents of trustworthiness. For instance, events resulting from coherent and

348  S. Arzu Wasti and Çetin Önder

effective business strategies are likely to be positively associated with organizational competence everywhere (Searle et al., 2011). Likewise, events signaling favoritism will likely be associated with negative perceptions of organizational trustworthiness across the board (Summereder et al., 2014). Even studies from contexts that represent prevalent particularism such as China suggest that although guanxi relations benefit individuals occasionally, they also convey the message that the organization lacks a common principle that underlines the HR system and that the system is unpredictable (Chen et al., 2011). Indeed, Chen et al. (2004) have found that guanxi practices decreased employees’ trust in management. Pearce and her colleagues (e.g., Pearce et al., 2000, 2009) have made parallel observations regarding ex-communist countries, where employees were more likely to trust and collaborate with other employees when rules were applied uniformly. We propose that such seemingly generalizable relationships will be moderated by employees’ subjective cultural press, that is, their descriptive norms and schemas which capture perceptions regarding the psychological characteristics that are widespread in a culture (Yamagishi, 2014) or rules, role expectancies, and reward contingencies that have been institutionalized in a context (Bachmann & Inkpen, 2011; Morris et al., 2015). This proposition draws on the so-called frog pond effects, which suggest that individuals will evaluate a circumstance by virtue of its relative standing in a particular context (Johns, 2006). For instance, Rousseau (1995) has argued that in contexts where societal guarantees are lacking such as neotraditional systems, individuals may react to favoritism with greater tolerance. On the other hand, in these contexts, where non-compliance, tax evasion, and fraud are a norm rather than an exception, employees may perceive a company that simply abides by the rules and regulations as an exceptionally responsible corporation (Jamali & Mirshak, 2007). In contrast, in welfare states like Germany and Denmark, job security, training, and voice tend to show little variance among organizations since they are regulated by the law or labor arrangements. Hence, employees may react to events regulated by external authorities, however positive, with some complacency (Wasti et al., 2016). Employees may also hold norms as to what is more discretionary and hence subject to greater scrutiny (Nohria et al., 2008). For instance, in neotraditional contexts that foster relationship- versus rule-based trust (Bohnet et al., 2010), employees may be more attentive to cues emitted by leaders’ behaviors and their discretion over HR practices. On the other hand, insurances provided by modern systems may allow employees to allocate greater attention to cues related to organizational strategy. As implied in the arguments leading to P2a, external assurances (provided by perceptions of a strong legal system or capable government) may render most signs relatively irrelevant for the employees, feeding trust by proxy (Gillespie & Dietz, 2009). Thus, drawing on employees’ societal as well as organizational norms and schemas, we propose the following mechanism:

Trust in Organizations Across Cultures  349

Proposition 4b: The relationship between organizational events and trustworthiness assessments will be moderated by employees’ descriptive norms and schemas regarding the typicality of and the discretion involved in the event within the comparative context. Finally, we advance our proposition regarding the moderating role of individual carriers of culture on the relationship between trustworthiness assessments and trust in organizations (P4c). It will be noted that until now we did not differentiate between the factors of trustworthiness (i.e., ability, benevolence, integrity) and referred simply to overall trustworthiness. This decision was guided by the evidence that many events generate cues that speak to multiple factors. For instance, Searle et al. (2011) argued that high-performance work practices not only signal the organization’s care and concern to the employees but also its ability to meet its objectives. Nonetheless, in our finishing proposition, we differentiate the perceived factors of trustworthiness and we posit that the importance of organizational ability, benevolence, and integrity may be moderated by culture at the individual level (Doney et al., 1998). Although focusing on interpersonal trust, Doney et al. (1998) argued that in collectivist cultures characterized by a high degree of social connectedness, benevolent motivations are important processes in trust formation. On the other hand, societal norms and values applaud achievement in individualist cultures. Similarly, studies building on moral foundations theory (Haidt & Kesebir, 2010) have found greater sensitivity for violations of universalist principles of fairness in individualistic cultures and for those with liberal ideologies as opposed to violations of loyalty in collectivist cultures and for those with conservative beliefs. Emphasizing the importance of culture-sensitive operationalizations of ability, benevolence, and integrity (Wasti et al., 2007), our final proposition states that: Proposition 4c: The relationship between trustworthiness assessments and trust in organizations will be moderated by employees’ prescriptive norms and values such that for employees holding in-group collectivist (versus individualist) prescriptive norms and values, benevolence (versus ability and integrity) assessments will have a greater positive effect on trust in organizations.

Discussion In this chapter, we attempted to explicate the role of culture in the development of employee trust in organizations. To this end, we distinguished between modern versus neotraditional institutional systems, which we argued are mutually reinforcing with ideas related to individualism and egalitarianism versus in-group collectivism and hierarchism. We explained that these socio-institutional constellations have differential cross-level effects on individual-level variables such as

350  S. Arzu Wasti and Çetin Önder

cynicism and in-group collectivism, which are associated with lower levels of generalized trust in organizations.We further identified the attributes of the dominant organizational situationscapes these two systems tend to afford. In particular, we argued that modern systems would be more likely to afford organizational situationscapes characterized by empowerment, meritocracy, and long-term collaboration, and we proposed such situationscapes may produce events that signal organizational trustworthiness to a greater extent. Finally, we offered that the societal as well as organizational variables would shape employees’ norms and schemas as to what constitutes typical and discretionary workways in their context, which in turn would serve as a benchmark in evaluating their particular trust-related organizational experiences. As such, we tried to put together a model that would not only alert researchers and practitioners alike to the complex set of variables involved in predicting trust in organizations but also address the experiences of the ever-increasing multicultural workforces and workplaces around the world. Albeit with simplifications, our model is based on a more holistic definition of culture, whereby culture is intertwined with institutions, practices, and individual behaviors. Furthermore, we do not operationalize culture at the individual level merely as internalized values, and we acknowledge the importance of cultural norms and schemas in explaining individual behavior as well. Broadening the carriers of culture allows the circumvention of unreasonable assumptions as to the sharedness or stability of one’s native societal culture. Rather, it recognizes that irrespective of their own psychological culture, employees may adjust their judgment and behavior according to perceived norms and culturally prevalent schemas. Furthermore, they may revise their norms and schemas as they immerse themselves in multiple cultures at the societal or organizational level. Our multilevel model advances the literature on trust in organizations in at least two ways: first, it integrates the relatively unconnected comparative institutional, cross-cultural organizational behavior/HR, and cultural psychology research to provide a comprehensive analysis of how societal and organizational levels relate to antecedents of trust at the employee level. Thus, our model allows for examining a broader set of top-down influences and cross-level interactions and unearthing new mechanisms concerning the development of employee trust in organizations. In this respect, our model carries the potential of expanding the comparative institutional perspective with micro-foundations as to how business systems are maintained, which is conspicuously missing from this literature. Likewise, we enrich the cultural psychology research by explicating the societal institutions and organizational system components that function as carriers of cultural ideas. Second, our broad operationalization of culture at the individual level that integrates the role of norms and schemas in the interpretation of organizational events brings in the context as the shaper of meaning (Johns, 2006) in understanding trust formation. For instance, Wasti, Tan, and Rodrigues (2016) observed that Turkish employees used a schema of the prototypical family-owned

Trust in Organizations Across Cultures  351

and managed firm characterized by particularistic HR and arbitrary managerial practices as a benchmark against which they compared subsequent work experiences, thereby privileging developed country MNEs with more formalized procedures in assessments of trustworthiness. Thus, our model highlights the importance of recognizing the organizational topology with respect to the prevalence and variance of certain organizational forms and practices to situate the employee’s particular psychological experience. Empirical testing of our full model requires relatively large, nested datasets that are capable of capturing variance between countries, organizations nested within countries, individuals nested within organizations, and events nested within individuals. Nevertheless, our propositions can also be tested separately. One important task in this respect is to test whether organizational situationscapes vary across societies (P1a). Fine-grained measures of societal-institutional characteristics such as state capacity (e.g., Guillén & Capron, 2016) or financial system stability (World Bank, 2020) and culture (e.g., House et al., 2004) as well as characteristics of representative samples of business organizations (e.g.,World Bank’s Enterprise Surveys) are increasingly available for a large number of countries. Thus, multilevel models testing for relationships between societal-institutional and organizational characteristics are fairly feasible. An alternative would be to sample organizations that boast different situationscapes and test for the relationship between the characteristics of these organizations and the events they generate (P1b). Publicly listed firms disclose ample information regarding their organizational system components (e.g., their strategies, policies, or practices) and the events they generate (e.g., layoffs or collective bargaining). Experimental designs that involve manipulation of simulated events (e.g., switching between events that are characteristic and uncharacteristic of a prototypical organization) and self-report measures of individual-level cultural variables could be used to understand how trustworthiness evaluations evolve and shape trust in organizations, which would be an empirical test of P4. We foresee several fruitful avenues toward the development of the model that we have proposed. First, the categorization of modern versus neotraditional configurations is a simplification that warrants future research using finer-grained distinctions. Though we assume modern (neotraditional) systems as mutually reinforcing with individualism and low power distance (in-group collectivism and high power distance), there are notable exceptions to this particular coupling of institutions and culture. For instance, our model could be expanded to incorporate countries that boast strong states with collectivism and high power distance (thus, displaying both modernist and neotraditionalist features), such as Japan. Likewise, modern systems vary along salient dimensions; while some display greater concern for collaborative relations (e.g., Germany), others prefer more arm’s length relationships between employers and employees (e.g., the US). Future research could further unearth multiple and at times conflicting pressures emanating from culture, societal institutions, and organizations, especially

352  S. Arzu Wasti and Çetin Önder

in contexts that are undergoing profound change, most notably China. Also, our rather coarse distinction obscures certain ways through which within-configuration change arises over time, such as extensive deregulation eroding the basis for generalized trust in modern systems (Bachmann, 2001). Thus, finer-grained distinctions could help incorporate additional mechanisms into explanatory models and help better understand how trust is built and sustained. Secondly, future research could directly address cross-border institutional influences on organizational situationscapes. Such influences may emanate from the diffusion of global organizational templates, transfer of practices from one country to others, or cross-border mobility of firms. While countries with indigenously developed institutions resist these cross-border influences (Höllerer, 2013) or make it hard for firms to flout national conventions (Whitley, 1999), those with underdeveloped institutions may be more welcoming of them (e.g., Amsden, 2001). In latter contexts, organizational system components may embody elements of modern as well as neotraditional systems, sending employees mixed signals. Thirdly, we have limited our model to internal systems of organizational trustworthiness; yet, employees also assess trustworthiness by reference to public reputation (Gillespie & Dietz, 2009). Societal institutions may shape what constitutes a firm reputation and organizational policies and practices geared towards developing reputation. For instance, states that emphasize national economic development may encourage firms to form interest group organizations. While firms that join these associations gain legitimacy as supporters of developmental goals, those that remain outside them may be “seen as weak and/or exploitative” (Whitley, 2005, p. 200). Fourthly, although we expanded the individual-level culture variables that have been covered in the organizational trust literature, more work needs to be done in understanding when and how different carriers of culture predict individual behavior. In their inspiring framework of situated dynamics, Leung and Morris (2015) argued that values may play a more important role in weak situations of fewer constraints, whereas norms and schemas may have more influence in situations involving social evaluation or ambiguity (Leung & Morris, 2015). Considering that organizational situationscapes in neotraditional contexts may be prone to generate ambiguous events, as hierarchical contexts present fewer opportunities for candid interaction with power holders (Rousseau & Schalk, 2000), norms and schemas may explain greater variance than they do in modern socio-institutional contexts. Leung and Morris (2015) further suggested that societal norms may be more relevant in the early phases of an employment relationship, and be replaced by organizational norms subsequently. Testing these propositions will greatly enhance our understanding of the role of culture in organizational trust. Finally, our model in its current form fails to incorporate the temporal dimension. Fulmer and Gelfand (2013) have proposed that cultures are likely to differ in terms of how quickly they form, dissolve, and repair trust. Trust tends to fluctuate widely over the course of a relationship, which is more likely to be true when the

Trust in Organizations Across Cultures  353

referent is the organization.Thus, a fruitful direction is to capture the evolution of and changes in trust in an employment relationship in different cultures.

Conclusions The study of trust in organizations across cultures necessitates a comparative analysis of the concepts of work, employment, and organization, which are products of the particular sociological, political, legal, and economic environment. It is time to seek out multidisciplinary collaborations to further our understanding of the complex relationship between culture and organizational trust. We hope that our chapter will serve as an inspiration to continue this journey.

References Almakaeva, A., Welzel, C., & Ponarin, E. (2018). Human empowerment and trust in strangers:The multilevel evidence. Social Indicators Research, 139(3), 923–962. https​:/​/do​​i​. org​​/10​.1​​007​/s​​11205​​-01​7-​​1724-​z Amsden, A. H. (2001). The rise of “Rest”: Challenges to the west from late-industrializing economies. New York: Oxford University Press. Aycan, Z. (2006). Paternalism. In U. Kim, K-S.Yang, & K-K. Hwang (Eds.), Indigenous and cultural psychology (pp. 445–466). Boston, MA: Springer. Bachmann, R. (2001). Trust, power and control in trans-organizational relations. Organization Studies, 22, 337–365. https://doi​.org​/10​.1177​/0170840601222007 Bachmann, R., & Inkpen, A. C. (2011). Understanding institutional-based trust building processes in inter-organizational relationships. Organization Studies,  32(2), 281–301. https://doi​.org​/10​.1177​/0170840610397477 Basabe, N., & Ros, M. (2005). Cultural dimensions and social behavior correlates: Individualism-collectivism and power distance. International Review of Social Psychology, 18(1), 189–225. Beugelsdijk, S., Kostova, T., & Roth, K. (2017). An overview of hofstede-inspired countrylevel culture research in international business since 2006. Journal of International Business Studies, 48(1), 30–47. https​:/​/do​​i​.org​​/10​.1​​057​/s​​41267​​-01​6-​​0038-​8 Bohnet, I., Herrmann, B., & Zeckhauser, R. (2010). Trust and the reference points for trustworthiness in Gulf and Western countries. Quarterly Journal of Economics, 125(2), 811–828. https​:/​/do​​i​.org​​/10​.1​​162​/q​​jec​.2​​010​​.1​​25​.2.​​811 Burke,C.S.,Sims,D.E.,Lazzara,E.H.,& Salas,E.(2007).Trust in leadership:A multi-level review and integration. The Leadership Quarterly, 18(6), 606–632. https​:/​/do​​i​.org​​/10​.1​​016​/j​​. leaq​​ua​.20​​​07​.09​​.006 Caldwell, C., & Clapham, S. E. (2003). Organizational trustworthiness: An international perspective. Journal of Business Ethics, 47(4), 349–364. https://doi​.org​/10​.1023​/A​: 1027370104302 Chao, G. T., & Moon, H. (2005). The cultural mosaic: A metatheory for understanding the complexity of culture. Journal of Applied Psychology, 90(6), 1128–1140. https​:/​/do​​i​.org​​/ 10​.1​​037​/0​​021​-9​​010​​.9​​0​.6​.1​​128 Chen, C. C., Chen,Y. R., & Xin, K. (2004). Guanxi practices and trust in management: A procedural justice perspective. Organization Science, 15(2), 200–209. https://doi​.org​/10​. 1287​/orsc​.1030​.0047

354  S. Arzu Wasti and Çetin Önder

Chen,Y., Friedman, R.,Yu, E., & Sun, F. (2011). Examining the positive and negative effects of guanxi practices: A multi-level analysis of guanxi practices and procedural justice perceptions. Asia Pacific Journal of Management, 28(4), 715–735. https​:/​/do​​i​.org​​/10​.1​​007​/ s​​10490​​-00​9-​​9176-​x Child, J., & Möllering, G. (2003). Contextual confidence and active trust development in the Chinese business environment. Organization Science, 14(1), 69–80. https​:/​/do​​i​.org​​/ 10​.1​​287​/o​​rsc​.1​​4​.​1​.6​​9​.128​​13 Dietz, G., Gillespie, N., & Chao, G.T. (2010). Unravelling the complexities of trust and culture. In M. Saunders, D. Skinner, N. Gillespie, G. Dietz, & R. J. Lewicki (Eds.), Organizational trust: A cultural perspective (pp. 3–41). Cambridge University Press. https​ :/​ /do​​ i​ . org​​/10​.1​​017​/C​​BO978​​05117​​​63106​​.002 Doney, P. M., Cannon, J. P., & Mullen, M. R. (1998). Understanding the influence of national culture on the development of trust. Academy of Management Review, 23(3), 601–620. https://doi​.org​/10​.5465​/amr​.1998​.926629 Fainshmidt, S., Judge, W. Q., Aguilera, R. V., & Smith, A. (2018). Varieties of institutional systems: A contextual taxonomy of understudied countries. Journal of World Business, 53(3), 307–322. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.jwb.​​201​6.​​05​.00​3 Farh, J. L., Hackett, R. D., & Liang, J. (2007). Individual-level cultural values as moderators of perceived organizational support–employee outcome relationships in China: Comparing the effects of power distance and traditionality. Academy of Management Journal, 50(3), 715–729. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mj​.20​​07​​.25​​53086​6 Fehr, R., Fulmer, A., Awtrey, E., & Miller, J. A. (2017). The grateful workplace: A multilevel model of gratitude in organizations. Academy of Management Review, 42(2), 361–381. https://doi​.org​/10​.5465​/amr​.2014​.0374 Fischer, R., & Schwartz, S. (2011). Whence differences in value priorities? Individual, cultural, or artifactual sources. Journal of Cross-Cultural Psychology, 42(7), 1127–1144. https://doi​.org​/10​.1177​/0022022110381429 Fitzsimmons, S. R., & Stamper, C. L. (2014). How societal culture influences friction in the employee–organization relationship. Human Resource Management Review, 24(1), 80–94. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.hrmr​​.201​3​​.07​.0​​01 Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels.Journal of Management,38(4),1167–1230.https://doi​.org​/10​. 1177​/0149206312439327 Fulmer, C. A., & Gelfand, M. J. (2013). How do I trust thee? Dynamic trust patterns and their individual and social contextual determinants. In K. Sycara, M. Gelfand, & A. Abbe (Eds.), Models for intercultural collaboration and negotiation: Advances in group decision and negotiation (Vol. 6). Dordrecht: Springer. https​:/​/do​​i​.org​​/10​.1​​007​/9​​78​-94​​-007​-​​ 5574-​​1_5 Gelfand, M. J., Leslie, L. M., & Fehr, R. (2008). To prosper, organizational psychology should… adopt a global perspective. Journal of Organizational Behavior, 29(4): 493–517. https://doi​.org​/10​.1002​/job​.530 Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34(1), 127–145. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​09​​.35​​71331​9 Guillén, M. F., & Capron, L. (2016). State capacity, minority shareholder protections, and stock market development. Administrative Science Quarterly, 61(1), 125–160. https://doi​.org​/ 10​.1177​/0001839215601459 Haidt, J., & Kesebir, S. (2010). Morality. In S. T. Fiske, D. T. Gilbert, & G. Lindzey (Eds.), Handbook of social psychology (Vol. 2, pp. 797–832). Hoboken, NJ:Wiley. https​:/​/do​​i​.org​​/ 10​.1​​002​/9​​78047​​05611​​19​.so​​​cpsy0​​02022​

Trust in Organizations Across Cultures  355

Hamedani, M. G., & Markus, H. R. (2019). Understanding culture clashes and catalyzing change: A culture cycle approach. Frontiers in Psychology, 10, 1–7. https://doi​.org​/10​. 3389​/fpsyg​.2019​.00700 Hodson, R. (2004). Organizational trustworthiness: Findings from the population of organizational ethnographies. Organization Sciences, 15(4), 432–445. https://doi​.org​/10​. 1287​/orsc​.1040​.0077 Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions and organizations across nations. Thousand Oaks, CA: SAGE. Höllerer, M.A. (2013). From taken-for-granted to explicit commitment:The rise of CSR in a corporatist country. Journal of Management Studies, 50(4), 573–606. https://doi​.org​/10​. 1111​/joms​.12029 Hotho, J. J. (2014). From typology to taxonomy: A configurational analysis of national business systems and their explanatory power. Organization Studies, 35(5), 671–702. https://doi​.org​/10​.1177​/0170840613502767 House, R. J., Dorfman, P. W., Javidan, M., Hanges, P. J., & de Luque, M. F. S. (2013). Strategic leadership across cultures: GLOBE study of CEO leadership behavior and effectiveness in 24 countries. Thousand Oaks, CA: SAGE. House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta,V. (Eds.). (2004). Culture, leadership, and organizations: The GLOBE study of 62 societies. Thousand Oaks, CA: SAGE. Huff, L., & Kelley, L. (2003). Levels of organizational trust in individualist versus collectivist societies: A seven-nation study. Organization Science, 14(1), 81–90. https​:/​/do​​i​.org​​/10​. 1​​287​/o​​rsc​.1​​4​.​1​.8​​1​.128​​07 Jamali, D., & Mirshak, R. (2007). Corporate social responsibility (CSR):Theory and practice in a developing country context. Journal of Business Ethics, 72(3), 243–262. https​:/​/ do​​i​.org​​/10​.1​​007​/s​​10551​​-00​6-​​9168-​4 Johns, G. (2006). The essential impact of context on organizational behavior. Academy of Management Review, 31(2), 386–408. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​.20​​06​​.20​​20868​7 Kramer, R. M. (2010). Collective trust within organizations: Conceptual foundations and empirical insights. Corporate Reputation Review, 13(2), 82–97. https://doi​.org​/10​.1057​/ crr​.2010.9 Leung, K., & Bond, M. H. (2004). Social axioms: A model for social beliefs in multicultural perspective. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 36, pp. 119–197). San Diego, CA: Elsevier Academic Press. https​:/​/do​​i​.org​​/10​.1​​016​/S​​0​065-​​ 2601(​04)36​003-X​ Leung, K., & Morris, M. W. (2015). Values, schemas, and norms in the culture–behavior nexus: A situated dynamics framework. Journal of International Business Studies, 46(9), 1028–1050. https://doi​.org​/10​.1057​/jibs​.2014​.66 Levinson, H. (1965). The relationship between man and organization. Administrative Science Quarterly, 9, 370–390. https://doi​.org​/10​.2307​/2391032 Li,A., & Cropanzano, R. (2009). Do East Asians respond more/less strongly to organizational justice than North Americans? A meta-analysis. Journal of Management Studies, 46(5), 787–805. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1467​​-6486​​.200​9​​.0082​​5.x Li, F., & Leung, K. (2012). Effects of evaluation of societal conditions and work–family conflict on social cynicism and distress: A longitudinal analysis. Journal of Applied Social Psychology, 42(3), 717–734. https​:/​/do​​i​.org​​/10​.1​​111​/j​​.1559​​-1816​​.201​1​​.0081​​6.x Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https​:/​/do​​i​.org​​/10​.5​​465​/a​​mr​. 19​​95​.​95​​08080​​335

356  S. Arzu Wasti and Çetin Önder

Morris, M. W., Hong, Y. Y., Chiu, C. Y., & Liu, Z. (2015). Normology: Integrating insights about social norms to understand cultural dynamics. Organizational Behavior and Human Decision Processes, 129, 1–13. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.obhd​​p​.20​1​​5​.03.​​001 Nohria, N., Groysberg, B., & Lee, L. (2008). Employee motivation: A powerful new model. Harvard Business Review, 86(7/8), 78. Pearce, J. L., Branyiczki, I., & Bigley, G. A. (2000). Insufficient bureaucracy: Trust and commitment in particularistic organizations. Organization Science, 11(2),148–162.https​:/​/ do​​i​.org​​/10​.1​​287​/o​​rsc​.1​​1​.2​​.1​​48​.12​​508 Pearce, J. L., Dibble, R., & Klein, K. (2009).The effects of governments on management and organization. Academy of Management Annals, 3(1), 503–541. https​:/​/do​​i​.org​​/10​.5​​465​/ 1​​94165​​209​03​​04754​1 Peretz, H., & Fried, Y. (2012). National cultures, performance appraisal practices, and organizational absenteeism and turnover: A study across 21 countries. Journal of Applied Psychology, 97(2), 448–459. https://doi​.org​/10​.1037​/a0026011 Peterson, M. F., & Barreto, T. S. (2018). Interpreting societal culture value dimensions. Journal of International Business Studies, 49(9), 1190–1207. https​:/​/do​​i​.org​​/10​.1​​057​/ s​​41267​​-01​8-​​0185-​1 Pillai, R., Williams, E. S., & Justin Tan, J. (2001). Are the scales tipped in favor of procedural or distributive justice? An investigation of the US, India, Germany, and Hong Kong (China). International Journal of Conflict Management, 12(4), 312–332. https://doi​.org​/10​. 1108​/eb022861 Rousseau, D. (1995). Psychological contracts in organizations: Understanding written and unwritten agreements. Thousand Oaks, CA: SAGE. Rousseau, D. (2001). Schema, promise and mutuality: The building blocks of the psychological contract. Journal of Occupational and Organizational Psychology, 74(4), 511– 541. https://doi​.org​/10​.1348​/096317901167505 Rousseau, D., & Schalk, R. (Eds.). (2000). Psychological contracts in employment: Cross-national perspectives. Thousand Oaks, CA: SAGE. Sanchez-Burks, J., & Lee, F. (2007). Cultural psychology of workways. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (Vol. 1, pp. 346–369). New York: Guilford. Searle, R., Den Hartog, D. N., Weibel, A., Gillespie, N., Six, F., Hatzakis, T., & Skinner, D. (2011). Trust in the employer: The role of high-involvement work practices and procedural justice in European organizations. The International Journal of Human Resource Management, 22(5), 1069–1092. https​:/​/do​​i​.org​​/10​.1​​080​/0​​95851​​92​.20​​​11​.55​​6782 Sorge, A. (2005). Systemic perspectives on business practices and institutions. A plea beyond comparative statics. In G. Morgan, R. Whitley, & E. Moen (Eds.), Changing capitalisms? Internationalization, institutional change, and systems of economic organization (pp. 110–136). Oxford: Oxford University Press. Summereder, S., Streicher, B., & Batinic, B. (2014).Voice or consistency? What you perceive as procedurally fair depends on your level of power distance. Journal of Cross-Cultural Psychology, 45(2), 192–212. https://doi​.org​/10​.1177​/0022022113505356 Thelen, K. (2010). Beyond comparative statics: Historical institutional approaches to stability and change in the political economy of labor. In G. Morgan, J. L. Campbell, C. Crouch, O. K. Pedersen, & R. Whitley (Eds.), The Oxford handbook of comparative institutional analysis (pp.41–62). Oxford University Press. https​:/​/do​​i​.org​​/10​.1​​093​/ o​​xford​​hb​/97​​80199​​2337​6​​2​.003​​.0003​ Vanhala, M., & Dietz, G. (2015). HRM, trust in employer and organizational performance. Knowledge and Process Management, 22(4), 270–287. https://doi​.org​/10​.1002​/kpm​.1491

Trust in Organizations Across Cultures  357

Wan, C., & Lu, C. (2014). Implications for cultural sharedness when sharedness is not assumed.  Journal of Cross-Cultural Psychology, 45(1), 25–29. https://doi​.org​/10​.1177​/ 0022022113513399 Wasti, S. A., Peterson, M. F., Breisohl, H., Cohen, A., Jørgensen, F., Rodrigues, A., Weng, Q., & Xu, X. (2016). Location, location, location: Contextualizing workplace commitment. Journal of Organizational Behavior, 37, 613–632. https://doi​.org​/10​.1002​/job​.2094 Wasti, S. A., Tan, H. H., Brower, H., & Önder, Ç. (2007). Cross-cultural measurement of supervisor trustworthiness: An assessment of measurement invariance across three cultures. Leadership Quarterly, 18, 477–489. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​07​. 07​​.004 Wasti, S. A.,Tan, H. H., & Rodrigues, A. (2016, July). Organizational trustworthiness: A journey through Brazil, China, India, the Netherlands, and Turkey. Paper presented at the 32nd EGOS Colloquim, Naples, Italy. Whitley, R. (1999). Divergent capitalisms: The social structuring and change of business systems. Oxford: Oxford University Press. Whitley, R. (2005). How national are business systems? The role of states and complementary institutions in standardizing systems of economic coordination and control at the national level. In G. Morgan, R. Whitley, & E. Moen (Eds.), Changing capitalisms? Internationalization, institutional change, and systems of economic organization (pp. 190–231). Oxford: Oxford University Press. Whitley, R. (2008). Business systems and organizational capabilities: The institutional structuring of competitive competences. New York: Oxford University Press. World Bank (2020). World development indicators. Washington, DC: World Bank. Yamagishi, T. (1998). The structure of trust: The evolutionary games of mind and society. Tokyo: University of Tokyo Press. Yamagishi, T. (2014). From a measurement model to a dynamic causal model commentary on Schwartz. Journal of Cross-Cultural Psychology, 45(1), 30–36. https://doi​.org​/10​.1177​/ 0022022113513105 Zhang,Y., Huai, M.Y., & Xie,Y. H. (2015). Paternalistic leadership and employee voice in China: A dual process model. The Leadership Quarterly, 26(1), 25–36. https​:/​/do​​i​.org​​/10​. 1​​016​/j​​.leaq​​ua​.20​​​14​.01​​.002 Zhang, Y., Long, L., Wu, T. Y., & Huang, X. (2015). When is pay for performance related to employee creativity in the Chinese context? The role of guanxi HRM practice, trust in management, and intrinsic motivation. Journal of Organizational Behavior, 36(5), 698–719. https://doi​.org​/10​.1002​/job​.2012 Zhang,Y.,Tsui, A. S., Song, L. J., Li, C., & Jia, L. (2008). How do I trust thee? The employeeorganization relationship, supervisory support, and middle manager trust in the organization. Human Resource Management, 47(1), 111–132. https://doi​.org​/10​.1002​/ hrm​.20200

PART IV

Conclusion and Way Forward

15 MULTILEVEL TRUST Reflections, Insights, and a Future Research Agenda Guido Möllering, Nicole Gillespie, and Roy Lewicki

Chapter Introduction and Overview The essays in this volume have presented multiple disciplinary, theoretical, and empirical perspectives on multilevel trust in organizations. As noted in the introduction, while research on trust has burgeoned over the last several decades, the focus has dominantly been at a single level of analysis, often the individual level. Moreover, researchers have often extrapolated trust concepts, theories, and findings from the individual level to group and organizational levels, without serious examination of the appropriateness of this approach and the underlying assumption of isomorphism across levels. While many theoretical and empirical papers on trust draw ‘implications’ for a multilevel perspective, these works rarely spell out these implications or suggest the appropriate conceptual and empirical tools for their examination. Recognizing this gap in the research literature, the editors challenged a group of internationally renowned researchers to break new ground in our understanding of multilevel trust in the workplace.They invited experts who had already made contributions to a multilevel perspective on trust, and who would approach these questions from different disciplinary and theoretical viewpoints. We are proud of and excited about the contributions they have provided. The papers clearly offer new perspectives on multilevel dynamics around trust and identify significant gaps in our understanding of how trust operates, particularly at the within-group, betweengroup, and organizational levels. Moreover, while the author teams of each chapter largely worked independently of each other, many of their papers complement each other as building blocks to a deeper understanding of multilevel trust. The 13 substantive chapters are diverse in the way they have approached their specific trust foci using a multilevel perspective. Some have focused predominately DOI:10.4324/9780429449185-15

362  G. Möllering, N. Gillespie, and R. Lewicki

on top-down or bottom-up influences, others on multiple cross-level effects, some on issues of convergence and divergence within units, and others on trust targets or referents residing at multiple levels.The large majority make conceptual advances grounded in a review and synthesis of the existing relevant research. These contributions include novel theoretical models and frameworks (e.g., Gillespie, Lockey, Hornsey, & Okimoto, Chapter 7; Korsgaard & Bliese, Chapter 3; Zhu, Lau, & Lam, Chapter 6), novel concepts (e.g., McEvily, Zaheer, & Soda, Chapter 8) and propositions (e.g., Long, Chapter 5; Wasti & Önder, Chapter 14), conceptual clarifications (e.g., Fulmer & Ostroff, Chapter 2; Jones & Shah, Chapter 9), as well as novel application of existing concepts from other fields (e.g., Mayer & Williams, Chapter 11) to deepen insight on multilevel trust. Two chapters (Chapter 10 by Hye Jung Eun and colleagues and Chapter 13 by Lisa van der Werff and colleagues) also employ novel empirical work to inform insights. The contributions to this volume make for valuable insights together because they apply various ways of looking at how the multiple levels of trust interact. Below we draw out reflections and insights on the key emergent themes from across the contributions.The resulting picture of multilevel trust shows how the contributions complement each other and reveal rich opportunities for future research in this area.

Reflections and Insights: Multiple Ways to Look at Multiple Levels Top-Down and Bottom-Up Processes Across Levels First of all, the chapters can be distinguished by whether they mainly assume multilevel and cross-level effects to occur top-down (whereby higher-level factors influence trust at the lower levels) versus bottom-up (the impact of lower-level forces working upwards to influence trust at a higher level). While it is easy to acknowledge in principle that both directions can be relevant, it is important to understand top-down and bottom-up effects as such and the conditions that make them salient. A closer look at the chapters reveals important insight in this respect. In Chris Long’s chapter on trust and control (Chapter 5) the top-down “cascade” from the strategic to the managerial to the subordinate hierarchical levels within organizations is very pronounced. His insightful review of research on trust and control sees the higher level as the one that deploys trust and control downward, depending on assessments of the lower-level needs and capabilities, and makes choices about different forms, degrees, and combinations of trust and control. Future work can extend Long’s analysis by examining the dynamics of the reverse direction, that is, how trust and also control move upward in ultimately recursive power relations across levels. Another chapter that examines trust dynamics primarily from a top-down view is Roger C. Mayer and Michele Williams’s conceptualization of “Organization Dissociative Identity Disorder” (Chapter 11). This chapter powerfully lays out

Reflections and Future Research  363

what may happen when a disconnect in trust evolves, and the problematic repercussions this disconnect can have on the organization and its people. Coherence in an organization is lost when higher-level issues cause frustration at the lower levels. Again, we can ask whether similar processes may also work bottom-up. Similarly, but less bleakly, Chapter 6 by Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam focuses on “felt trust” at lower levels. A top-down effect is assumed when lower-level members of the organization look up to see if they are trusted, which is shaped, for example, by the levels of delegation and monitoring they experience from above. Supervisors are the sense-givers. Yet team members’ interactions are critical in shaping how employees feel trusted by their leader, and these authors adopt the social identity perspective to discuss both top-down and bottom-up processes between the individual and team levels. Future work could further consider how followers’ perceptions of trust matter for supervisors, most commonly because the latter depend on the performance of the former. As these examples illustrate, a rich area for future research is the examination of whether and how both top-down and bottom-up processes may operate in a particular context and around a specific form or dynamic of trust.

Vertical and Horizontal Analyses of Multilevel Trust As noted above, the question of top-down versus bottom-up cross-level effects highlights vertical perspectives in multilevel trust. These ‘organizational levels’ can be viewed as hierarchically ordered (with one level above or below the other), but the levels might also be viewed as embedded or nested, one inside the other (see Chapter 2 by C. Ashley Fulmer and Cheri Ostroff). However, many chapters in this book also show a strong horizontal orientation to multilevel trust and discuss effects at the same level that, in turn, may influence how this level affects other levels. It is not uncommon for a vertically higher-level dynamic to be the result of lower-level horizontal dynamics. After all, the notion of emergence entails that a higher-level entity is the distinct result of convergence from lower-level elements (Fulmer & Ostroff, 2016). For example, we need to understand how the horizontal process at the lower level – i.e., trust among team members – affects how much the team as a collective trusts other teams. Note that this involves trust referents at two levels that must not be confounded (see Chapter 2). Such dynamics are also discussed in Chapter 3 by Audrey Korsgaard and Paul Bliese, showing how horizontal effects are key because collective trust as a vertical variable is framed around a horizontal comparison of individual members’ trust in the group. Specifically, individual-level divergence impacts the group-level trust on which the members converge (more or less), and trust in a collective is not necessarily a simple aggregation or average of trust by individuals that make up the collective (see also Chapter 2). Hence, it is necessary to understand horizontal interactions between the individuals to explain, vertically, the strength of their collective trust.

364  G. Möllering, N. Gillespie, and R. Lewicki

Along the same lines but with a different theoretical lens, Edward Tomlinson and Luke Langlinais (Chapter 4) take an attribution-theoretical approach and emphasize social influence processes to shows how actors’ trust attributions are horizontally networked. It implies the simple but important insight that some actors will trust in a higher-level entity only if and when others do so as well. Vertical trust attributions may thus be “borrowed,” “transferred,” or shared horizontally, which is very different from the idea that actors develop trust independently. Group trust is not just a simple, static aggregation of attributions by individuals. The need to consider vertical and horizontal effects together is also evident in Chapter 6 where Julie N.Y. Zhu, Dora C. Lau, and Long W. Lam discuss the effects of vertical felt trust by supervisors on horizontal outcomes, such as team performance and team trust. These authors point to the need to study non-hierarchical relations (among co-workers at the same level) in addition to the trust that is felt vertically because one individual’s feeling of being trusted is not detached from their peers’ felt trust. Perceived collective trust may spill over when one feels trusted because one senses that others feel trusted. Or there may be competition as to why others seem to be more trusted than we are. This line of research is in its infancy and much remains to be done. Probably more than any other chapters in this volume, the two chapters that apply a network perspective to multilevel trust (Chapters 8 and 9; to some extent also Chapter 12) reinforce the need to study horizontal effects. Bill McEvily, Akbar (Aks) Zaheer, and Giuseppe Soda (Chapter 8) make a very strong case for an additional form of trust that is distinct from, for example, simple dyadic interpersonal trust, institution-based trust, or trust in whole networks. Their “network trust” flows horizontally through indirect network connections. They point out that trust can be shaped by the way network ties are formed.The idea of an active formation of trust through network design, as opposed to simple ‘emergent trust,’ is strongly embraced by these authors. Stephen Jones and Priti Shah (Chapter 9) also highlight how different trust relationships at the same level produce an overall trust effect. Their concept of “networked trust” is less flat than McEvily et al.’s “network trust” and recognizes that trust shapes networks, not just that networks shape trust. They frame the questions of how networks should be conceived at different nested levels, such as a network of individuals in a team, a network of teams in a department, a network of departments in an organization, and so on up the vertical ladder. This includes the question of what dynamics, if any, emerge from networks viewed vertically. The network perspective is a strong reminder that each level within multilevel trust holds complex structures and dynamics that may influence cross-level effects. The problem of considering only one level horizontally is also pointed out by Nicole Gillespie, Steve Lockey, Matthew Hornsey, and Tyler Okimoto (Chapter 7) who find that the research on trust repair mostly assumes that interpersonal trust repair mechanisms occur at higher organizational levels. Their arguments

Reflections and Future Research  365

highlight the need to recognize the added complexity of higher (i.e., collective) level trust repair, compared to the lower level (i.e., individual or inter-personal), and argue why trust repair is not isomorphic across these levels. They give special attention to mechanisms that involve trust repair at multiple levels (e.g., individual transgressor to collective victims; collective transgressor to multiple collective victims that may reside at different levels) and specific horizontal levels (e.g., intergroup trust repair). In this respect, it is interesting to consider which mechanisms actually connect levels vertically and drive trust development horizontally at the same time. One important example featured in this book is human resource management (HRM). Rosalind Searle and Rami Al-Sharif present a systematic review of HRM and trust with reference to multilevel issues (Chapter 12). HRM policies provide some level of standardization and convergence, while at the same time being tailored to meet the needs at different levels and in different parts of the organization through their implementation. HRM, thus, is both a vertical, often top-town force for organizational trust, but also one that responds to horizontal dynamics (such as “felt trust,” Chapter 6). As an example, for performance appraisals to be considered fair, they need to strike a balance between treating all employees the same and being responsive to individual conditions and needs. Any general policy will thus be interpreted locally, by individuals and groups, as to whether it instills trust or not. Overall, the vertical perspective of levels above and below each other remains a key focus of this volume. The idea of nested or embedded levels also expresses a vertical, hierarchical order of layered levels. In Chapters 10, 13, and 14 we learn about those levels that work like contextual layers for trust. Chapter 10 by Hye Jung Eun, Roy Chua, and Mengzi Jin shows gender as a societal force working through all levels. Non-human systems of artificial intelligence impact trust at micro and macro levels, as Chapter 13 by Lisa van der Werff, Kirsimarja Blomqvist, and Sirpa Koskinen demonstrates empirically. And S. Arzu Wasti and Çetin Önder (Chapter 14) discuss the impact of societal culture, institutions, and events on employee trust. These chapters highlight that cross-level trust requires a good understanding of trust at each level with its inherent dynamics and tensions.

Stable and Dynamic Relationships Across Levels Taking the contributions in this volume together, we observe further that the relationships between the levels are often seen as relatively stable and fixed, i.e., one level continuously having an impact on another level. However, there are also several indications that cross-level effects can vary over time, which means that we need to identify the triggers that initiate new or renewed effects from one level to another, leading to changes over time, as well as the conditions where the levels stabilize each other. Examples of these dynamics can be found in some of the chapters.

366  G. Möllering, N. Gillespie, and R. Lewicki

The time perspective required to understand multilevel trust dynamics is evident in Chapter 3 by Audrey Korsgaard and Paul Bliese because their discussion of collective trust convergence and divergence implies that convergence takes time. Specifically, they propose that convergence takes longer after negative events. Hence, whenever we find divergence in collective trust, we need to ask whether or not the trajectory over time goes toward convergence, and at what speed. Also, if divergence seems to be rather stable, it will be interesting to investigate what actions or events inhibit the dynamics toward convergence, such as the composition of a group that does not change due to high exit barriers (cf. Korsgaard & Bliese, Chapter 3). Some recent empirical work has started to tease out these factors (e.g., De Jong, Gillespie, Williamson, & Gill, 2020). In their discussion of dissociation in organizations, Roger C. Mayer and Michele Williams (Chapter 11) also imply dynamics over time, especially destabilizing ones, which often happen gradually rather than suddenly. This insight begs the question of how such developments might be stopped so that a beneficial level of trust can be stabilized across the organization.

Convergent and Divergent Trust Across Levels Related to the insight that relationships across levels are not always fixed and that there may be tensions causing a multilevel system to evolve over time, it is important to note that several chapters in this volume point out that movement is not always toward convergence (i.e., that individual group members’ attitudes, judgments, or behaviors converge toward similarity over time), but could go toward divergence (i.e., members can viably hold different trust judgments of trust within a group) and even “dissociation” (Mayer & Williams, Chapter 11) as well. Note that divergence can occur at the same level, too, and impact the kind of effect observed at the level above or below.This also suggests, for example, that the divergence at one level may reflect the divergence at another level (and the same for convergence). Imagine this in contexts where trust repair (Chapter 7) is supposed to happen at all levels throughout an organization, but different dynamics result in divergence of post-repair trust between levels. In other words, although trust might have been repaired adequately at one level (e.g., co-workers), the fact that another level (e.g., top management) may still actively distrust or be lagging behind in repair, may damage the repaired trust. As already mentioned above, the question of convergence and divergence is a central issue in the chapter by Audrey Korsgaard and Paul Bliese (Chapter 3). They argue, overall, that members’ trust in their team tends to converge, but divergent individual trust attitudes in groups appear to lead to lower rather than higher collective trust. Especially after destabilizing events, members trust their groups less; higher levels of trust held by some individuals may be pulled down as collective low trust converges. It will be important to understand better why this happens and how it might also be possible to bring collective trust up, rather

Reflections and Future Research  367

than down, when trust is a crucial resource for successful collaboration within and between groups. The question of convergence or divergence of trust at multiple levels takes yet another turn when considering Rosalind Searle and Rami Al-Sharif ’s (Chapter 12) important insights regarding the relationship between diversity and trust. As organizations become more diverse at every level and across several dimensions (identity, gender, race, etc.), what is multilevel trust going to look like? If anything, it will be important to actively take diversity into account, and not treat people at different hierarchical levels as if they are literally all the same by their designation as ‘workers,’ ‘managers,’ ‘directors,’ etc. This added complexity will not make trust easier, but diversity might also make levels more permeable and enhance crosslevel collaboration rather than dissociation.

Revisiting the Question of Cross-Level Isomorphic Trust Altogether, what do the distinctions between top-down and bottom-up, vertical and horizontal, stable and dynamic, convergent and divergent relationships tell us? Most importantly, they enable a fresh look at the question of how isomorphic trust is across levels (see Chapters 2 and 7). Consider the overall proposition that trust is not isomorphic between levels and that this result can be explained by within-level conditions that result in differential effects of trust at other levels. For example, Gillespie and colleagues (Chapter 7) argue trust repair is qualitatively different at the individual and collective levels, due to the influence at the collective level of multiple transgressors (who may vary in their remorse and desire to repair trust), diverse trustors (with differing understandings of the violation and expectations for repair), as well as (inter)group dynamics and processes, thirdparty influences, and a greater repertoire of repair strategies that occur at the collective level. Consider also the possibility that any isomorphism (or a lack thereof) in trust across levels might be only a temporary state. The insights from the chapters in this book confirm that these are important considerations and argue for a strong note of caution against assuming isomorphism in trust across levels. Instead, the multilevel approach itself helps us to explain the trust idiosyncrasies, variations, and dynamics that we observe in organizations in reality.

A Future Research Agenda on Multilevel Trust ‘The Web of Trust’: A Social Networks Approach to Multilevel Trust A key insight from Chapters 8 and 9 is the value of taking a social network approach to understand the embedded nature and formation of trust within and across levels. We observe that there are multiple levels ‘around’ networks, begging

368  G. Möllering, N. Gillespie, and R. Lewicki

the question of how trust within networks can be integrated with, and shaped by, other levels. Much remains to be done to understand how trust emerges within networks as a bottom-up process (i.e., from lower-level elements), as well as how top-down influences can influence or shift network trust (e.g., broader trust embedded at the institutional or professional level). Networks are a rich arena to deepen understanding of distrust dynamics, and how they develop and spread across levels over time, as well as for studying the dark sides of trust (e.g., corruptive practices in organizational networks, obligation, or coercion) and how networks and influences within levels can prevent appropriate scrutiny and oversight at higher levels. Finally, we observe an opportunity for closer integration of network theory with multilevel theory, rather than the predominant application of just one or the other, to deepen understanding of trust. Related to the network approach, future research on multilevel trust could address the role of third-parties from new angles. The third-party literature on trust has been drawn upon in this volume, particularly in Chapters 7, 8, and 9. We can ask if the ‘third party’ is located at the same level as the parties it is connected to or influencing. Given the insights from this volume, it should make a difference, for example, if the third party is an individual co-worker bridging a structural hole between other co-workers, or whether those co-workers are referring to a higher-level institution or authority as third party, or if there are third parties as mere observers whose views might still matter for the trusting parties. Such distinctions should be examined carefully in future research.

‘Seeing’ Contextual Influences on Trust Through a Multilevel Approach As shown in this volume, multilevel theory provides a set of conceptual tools that can help us to see and theorize more precisely the influence of context on trust. ‘Context’ – whether culture, institutional arrangements, senior leadership turnover, or stability/disruption – is embedded and integrated into a level, which can then be theorized and empirically examined, enabling greater specificity in understanding its impact.This was particularly evident in Chapter 14 by Wasti and Önder and Chapter 11 by Mayer and Williams. We note that there is a perceptual element to levels, and which levels we actually pay attention to is likely to be influenced by a range of factors including context. For example, the COVID-19 pandemic has made salient the influence of certain levels, notably the government and local community levels, on citizen trust. How responses to the pandemic across multiple levels influence trust is a question ripe for multilevel theorizing. Understanding how the relevance and salience of trust at various levels can dynamically change requires further research. Another important question to explore is how contextual factors such as organizational culture may influence the trust cues people pay attention to and the motivation to trust at different levels.

Reflections and Future Research  369

Understanding Emerging Trust Challenges from a Multilevel Perspective There are many pressing trust challenges facing organizations and society more broadly that are ripe for tackling using a multilevel approach. What roles do trust and distrust play at the individual, group, network, and institutional levels in driving the increasing polarization we are witnessing in society? How are group dynamics, together with ‘echo chambers’ within social media and online networks, influencing trust and distrust in information sources, and what is judged to be credible information as opposed to disinformation, misinformation, or conspiracy views? How do an individual’s ideology and personal and social identity (e.g., in relation to political parties) influence these trust and distrust dynamics? A related emerging trust challenge is posed by the increasingly ubiquitous use of intelligent technology in the workplace, and in society more broadly. Indeed, it was artificial intelligence technology that enabled the personalized tailoring of news feeds and advertisements, which created the conditions for the voter manipulation scandal by Cambridge Analytica–Facebook through targeted political advertising in the lead-up to the 2016 US presidential election. This raises the important question of whether and how our understanding of trust in human relationships translates to non-human intelligent agents. There are two faces to the trust–technology relationship: technology-mediated trust and trust in technology itself. The latter raises the question of which level(s) we should be using to examine and understand trust in technology. Van der Werff and colleagues (Chapter 13) suggest that a multilevel perspective is required, with trust influenced by multiple referents including the developers, the organization, and the technology-enabled service itself.The role of advanced technology in disrupting and altering the trust–control relationship (e.g., intelligent agents monitoring and giving feedback to humans in the workplace) is another ripe area for multilevel theorizing and study (Schafheitle et al., 2020).

Deepening a Processual View of Trust Through a Multilevel Perspective We believe the multilevel perspective is well suited to facilitate a more processual view of trust over time. To really understand and tease out multilevel effects and their causes requires the examination of their development over time, and in return, a multilevel approach can extend the process views of trust that have been outlined, for example, by Möllering (2013), who proposed that “the multi-level perspective of trusting as constituting is also a longitudinal perspective” (p. 296). This volume has shown that trust processes and dynamics may function at different speeds across levels, as we know also from the literature on swift trust (e.g., Meyerson, Weick, & Kramer, 1996).Yet again, cross-level effects are relevant here – for example, that a rapid yet fragile form of trust between group members is enabled

370  G. Möllering, N. Gillespie, and R. Lewicki

by higher-order identification with the collective group, or a strong top-down group climate and norms. In other contexts, higher-order influences, such as an inter-group competitive frame, can undermine and slow trust development between groups. Given evidence that trust emerges at different rates across levels (Korsgaard & Bliese, Chapter 3), and that top-down and cross-level effects take time, as do the processes of emergence and consensus/divergence of trust within groups (e.g., De Jong et al., 2020), we recommend that future research adopt a longitudinal approach to examine multilevel trust.

Enablers of Multilevel Trust Research This volume has highlighted how a broad range of theories and methods can be fruitfully used to advance understanding of trust from a multilevel perspective. For example, Korsgaard and Bliese (see Chapter 3) helpfully review the strengths and weaknesses of a range of techniques that can be used to model convergence and divergence in trust, including the qualitative approach (e.g., ethnography, observation, and interviews), computational modeling, social network analysis, and the consensus emergence model (CEM), as well as highlighting best practice approaches. Chapter 2 by C. Ashley Fulmer and Cheri Ostroff serves as a valuable reference for achieving clarity around levels of analysis and trust referents. We hope that this volume will inform fellow researchers’ choices and decisions about which levels to focus on and which to leave out in their investigations, a question that is equally as important as the statistical technique used for multilevel analysis. We are not advocating, nor do we think it wise, to attempt an overly complex study of all levels at once. Rather, focused questions on specific levels, conducted with a sensitivity for multilevel effects, have the most potential to extend boundaries of knowledge and insight on trust. This volume provides a rich tapestry of conceptual theorizing and frameworks to guide future research.The challenge now is to move from theorizing to empirical examination. In doing so, trust researchers can look to other fields where the empirical examination of phenomena across multiple levels is further advanced to draw insight and inspiration, such as the literatures on organizational climate, teams, and leadership (e.g., Braun, Peus,Weisweiler, & Frey, 2013; Chen, Kirkman, Kanfer, Allen, & Rosen, 2007; Naumann & Bennett, 2000;Wallace & Chen, 2006; Zohar & Luria, 2005).We also see value in trust scholars partnering with theorists and empiricists working at different levels to their own foci as a facilitating tool. In this spirit, we look forward to a blossoming of novel insights on multilevel trust.

References Braun, S., Peus, C., Weisweiler, S., & Frey, D. (2013). Transformational leadership, job satisfaction, and team performance:A multilevel mediation model of trust. The Leadership Quarterly, 24(1), 270–283. https​:/​/do​​i​.org​​/10​.1​​016​/j​​.leaq​​ua​.20​​​12​.11​​.006

Reflections and Future Research  371

Chen, G., Kirkman, B. L., Kanfer, R., Allen, D., & Rosen, B. (2007). A multilevel study of leadership, empowerment, and performance in teams. Journal of Applied Psychology, 92(2), 331–346. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​2​.2​.3​​31 De Jong, B. A., Gillespie, N., Williamson, I., & Gill, C. (2020). Trust consensus within culturally diverse teams: A multistudy investigation. Journal of Management (Online First). https://doi​.org​/10​.1177​/0149206320943658 Fulmer, C. A., & Ostroff, C. (2016). Convergence and emergence in organizations: An integrative framework and review. Journal of Organizational Behavior, 37, S122– S145. https://doi​.org​/10​.1002​/job​.1987 Meyerson, D., Weick, K. E., & Kramer, R. M. (1996). Swift trust and temporary groups. In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 166–195). Thousand Oaks, CA: Sage. Möllering, G. (2013). Process views of trusting and crises. In R. Bachmann & A. Zaheer (Eds.), Handbook of advances in trust research (pp. 285–305). Cheltenham: Edward Elgar. https://doi​.org​/10​.2139​/ssrn​.2109376 Naumann, S. E., & Bennett, N. (2000). A case for procedural justice climate: Development and test of a multilevel model. Academy of Management Journal, 43(5), 881–889. https:// doi​.org​/10​.5465​/1556416 Schafheitle, S.,Weibel, A., Ebert, I., Kasper, G., Schank, C., & Leicht-Deobald, U. (2020). No stone left unturned? Toward a framework for the impact of datafication technologies on organizational control. Academy of Management Discoveries, 6(3), 455–487. https://doi​. org​/10​.5465​/amd​.2019​.0002 Wallace, C., & Chen, G. (2006).A multilevel integration of personality, climate, self-regulation, and performance. Personnel Psychology, 59(3), 529–557. https​:/​/do​​i​.org​​/10​.1​​111​/j​​. 1744​​-6570​​.200​6​​.0004​​6.x Zohar, D., & Luria, G. (2005). A multilevel model of safety climate: Cross-level relationships between organization and group-level climates. Journal of Applied Psychology, 90(4), 616– 628. https​:/​/do​​i​.org​​/10​.1​​037​/0​​021​-9​​010​​.9​​0​.4​.6​​16

INDEX

Page numbers in bold denote tables, in italic denote figures ability, benevolence, and integrity (ABI) 67–68, 103, 132, 185, 239, 257, 260, 262, 286, 292, 310, 320, 325, 341, 349 affective trust 209, 215, 218, 281 aggregation 15, 16, 23, 82, 122, 196, 267, 268, 290, 363, 364 agreement 7, 15, 23, 28, 29, 57, 72, 80, 83, 245, 258, 324, 345 Al-Sharif, R. 10, 297, 300, 365, 367 Alesina, A. 236 Apfelbaum, E. P. 297 artificial intelligence (AI) 10–11, 307–328, 322, 365, 369 Ashford, S. J. 103, 259 attributions 8, 66–83, 69, 70–71, 90–91, 95–96, 145, 147, 149, 150–152, 165, 213–214, 263, 265, 364 Bachmann, R. 5, 87, 143–145, 148, 150, 157, 162, 167–168, 307, 311–312, 314, 318, 326, 341, 348, 352 Bachrach, D. G. 74 Baer, M. D. 6, 28, 30, 36, 48, 121, 124, 126, 129, 133, 215, 218, 310–311 Bandura, A. 76, 103, 105, 132, 160 Barczak, G. 234 Barreto, T. S. 335 behavioral trust model 135 Beijer, S. 71, 78–79 Bennett, S. H. 144, 161

Benoit, W. L. 148, 154 Bevelander, D. 239 Bijlsma-Frankema, K. 58, 87, 89, 101, 108, 168, 287, 295 Blake-Beard, S. 109 Bliese, P. 6–8, 16, 36, 46–47, 55, 58–60, 362–363, 366, 370 Blomqvist, K. 10, 365 Bobko, P. 103, 259, 270 Boh, W. F. 209, 216–217 Bohnet, L. 344, 348 Bond, M. H. 344 Boon, C. 278, 289–290 bottom-up 5, 11, 131–132, 134–135, 324–326, 362–363, 367–368 Bowen, D. E. 72, 80–81 Bradley, B. H. 245 Brower, H. H. 7, 14, 56, 109, 122–124, 126, 209, 218, 239, 298 Buckley, F. 60, 282, 284, 288, 292, 295, 297, 299 Bunker, B. B. 50, 207, 214, 259, 285, 295 Burke, C. M. 260, 264 Burt, R. S. 4, 14, 66, 87, 144, 180, 184, 187, 189, 208–212, 214, 216, 217, 219, 221, 223, 262, 285 Bush, G. W. 154 Caldwell, C. 335 Capell, B. 280, 284, 292, 296, 298, 300

374 Index

Carter, M. 31, 48 Casual Dimension Scale for Teams (CDS-T) 73, 82 Chaudhuri, A. 236 Chen J. 48 Cheng, B. 109 Cherpitel, C. J. 71, 75–76 Chow, G. M. 71, 73, 76–77, 79, 82 Chowdhury, S. 234 chronic state 8, 51, 56 Chua, R. 10, 180, 209, 218–219, 234, 239, 241, 243, 247, 365 Clapham, S. E. 335 cognitive trust 209, 218, 281 co-worker 6, 10, 26, 146, 153, 155, 199, 222, 279, 281–282, 288–289, 364, 366, 368 cognitive social structure (CSS) 210, 210, 212–213, 215, 220 Coleman, J. S. 180, 186, 189, 195, 195, 200, 216 collective trust 7, 28, 35, 45–46, 49–51, 56, 58, 60–61, 146, 148, 150, 153–154, 156, 158, 160, 165, 363–364, 366 collective-to-collective 146, 150–151, 156–158, 160, 162, 163, 166 Colquitt, J. A. 47, 49, 121, 132, 134, 207, 260, 285–286, 310–311 community levels 368 compilation 7, 15, 18, 21–22, 23–24, 26, 28, 30–33, 35, 47, 193 composition 7, 10, 15, 18, 19, 23–24, 26, 28–31, 33, 47, 56, 125, 247, 366 consensus emergence model (CEM) 58–59, 61, 370 contextual influence 4, 11, 51, 110, 206, 261, 263–264, 266, 310, 368 contingencies 9, 96, 130, 182, 186–187, 186, 191, 193, 348 control activities 8, 96, 99, 102–103 control–trust dynamic 8, 87–88, 89n1, 90n1, 91, 93, 94, 95, 106–108, 110–111 convergence 8, 46, 49–53, 55–61, 362–363, 365–367, 370 Cooper, C. D. 7, 55, 69, 143, 151, 154–155, 213, 280, 284, 285, 288, 290, 294, 299 Corley, K. G. 131, 271 corporate social responsibility (CSR) 70, 74, 77–78, 81 Costa, A. C. 5, 14–15, 23, 26, 28–29, 46, 87, 89, 108, 110, 236, 281, 284, 288, 299 COVID-19 264–265, 271, 368 Cropanzano, R. 48, 155, 335

cross-level 8, 11, 28, 67, 70, 72, 76, 122–123, 130–131, 153, 155, 165, 246, 279, 288–289, 307–308, 312, 318, 324, 325, 328, 337, 338, 341–346, 349–350, 362–365, 367, 369–370 cultural values or cultural dimensions 32, 335, 338, 340, 341, 345, 347 Davis, J. H. 4, 14, 50, 67–68, 87, 109, 124–126, 134, 143–144, 179, 207, 239, 259–260, 279 De Cremer, D. 105, 149, 151–152, 328 De Jong, B. A. 5, 7, 14, 16, 24, 28, 32, 35, 49–50, 70, 127, 366, 370 delegation 47, 51, 123, 126–128, 342, 363 de Reuver, R. 72 DeSanti, A. 309 destabilizing events 8, 51, 54–55, 55, 59–60, 366 Deutsch, M. 67, 129 Dietz, G. 4–5, 7, 18, 26, 70, 91, 100–101, 103–104, 143–144, 148, 150, 156–157, 167, 200, 281, 284, 285–288, 295–296, 299, 334, 336–337, 340–341, 343–344, 348, 352 Dirks, K. T. 5, 7, 14, 26, 28, 32, 35, 47, 49, 55, 69–70, 99, 121, 124, 127, 134, 143–146, 149, 150–155, 157, 166, 180, 205, 207–208, 211, 213, 218, 236, 260, 262 dispersion 7, 15, 18, 21–22, 23–24, 26, 28–33, 35, 47, 50 distrust 10, 55, 58, 145, 168, 184, 197, 199, 209, 287, 294–296, 299, 300, 323, 327, 335, 343–345, 366, 368–369 Dithurbide, L. 70, 73, 79 divergence 8, 46–49, 51–54, 53, 56–61, 57, 163, 296, 324, 362–363, 366–367, 370 diversity 50, 52, 56, 60, 157, 164, 168, 192, 234, 236, 242–245, 294, 296–297, 299, 300, 367 Doney, P. M. 349 Dorenbosch, L. 72 Drescher, M. A. 6, 58 dyad 14, 18, 19–22, 28, 45, 124, 179, 184, 189–190, 193, 195, 201, 205–210, 206, 212, 218–220, 222, 235, 239–240, 241, 247 dyadic level 6, 16, 24, 34, 193, 201, 224, 235, 236, 239, 240, 246 Eagly, A. H. 237, 241–242 embeddedness 5, 57, 144

Index  375

emotion 11, 152, 310, 311, 325 emotional intelligence (EI) 282, 284, 293 Emsley, D. 87, 108 Eun, H. J. 10, 362, 365 event level 298, 336–337, 336, 339, 340 Fang, C. 52, 58 feedback 3, 68, 77, 94, 98, 103, 106, 238, 296, 369 felt trust 7–8, 36, 48, 60, 121–135, 123, 215, 218, 220, 282, 287, 363–365 Feltz, D. 23, 71, 73, 76–77, 79, 82 Ferguson, A. J. 6, 52, 56 Ferrin, D. 26, 35, 50–51, 60, 66, 69, 92, 99, 104, 121, 124, 134, 143, 146, 148–149, 151, 154–157, 180, 200, 205, 207–210, 213, 221, 223, 260, 262, 310 Frank, E. L. 6, 36, 124, 215 Fried,Y. 347 Fuller, M. A. 28, 51, 287 Fulmer, C. A. 4–7, 14–18, 26–29, 27n1, 35–36, 45–46, 66, 70, 82, 90, 104, 121, 134, 143, 162, 164, 205, 207, 236, 239, 259–260, 270, 279, 288, 307, 325–328, 334, 352, 362–363, 370 Galizzi, M. M. 168 Gangadharan, L. 236 Gavin, M. B. 6, 17, 32, 134, 259–260 Gelfand, M. J. 4–6, 14–17, 26, 28–29, 36, 45, 66, 70, 90, 104, 134, 143, 160, 162, 164–165, 205, 207, 236, 239, 259–260, 270, 307, 325–326, 334, 337, 341, 346–347, 352 General Social Survey (GSS) 236–237 Giddens, A. 312 Gill, C. 7, 14, 50, 366 Gillespie, N. 4–7, 9, 14, 18, 26, 32, 50, 70, 91, 100–101, 103–104, 126, 135, 143–144, 146, 148, 150, 156–158, 167–168, 180, 200, 207, 222, 234, 281, 284, 285–286, 288, 294–296, 299, 336–337, 340–341, 343–344, 348, 352, 362, 364, 366–367 Gillian, D. J. 312 Gould, R.V. 184 Granovetter, M. 183–184, 186–187, 205, 210 group level 10, 49, 70, 73, 144, 153, 235, 236, 244–245, 248, 279n1, 361, 363 group trust 3, 16, 52, 56, 60, 156, 364 Grunberg, R. 297 Gustafsson, S. 278

Hackman, J. R. 271 Haesevoets, T. 152 Halevy, N. 210, 297 Hatzakis, T. 77 HC-HRM see high-performance work systems Heider, F. 67, 79, 187n1, 220 Helper, S. 157 Hewstone, M. 70, 76, 80, 82 high-commitment HRM (HCHRM) 72, 282–283, 288–289 high-performance work systems (HPWS) 78–79 Hornsey, M. J. 9, 144, 148, 159–161, 167, 362, 364 House, R. J. 261, 340, 347, 351 Huai, M.Y. 347 Huang, M. 109 human resource management (HRM) 10, 70–71, 72, 78–81, 277–279, 280–284, 285–300, 365 Hurley, R. 157, 167, 200 ICC 46 indicators 9, 70, 74, 81, 97–100, 128, 182, 186–187, 186, 187n2, 191–192, 312–313 individual level 4, 6, 15–18, 24, 25, 28–30, 32–34, 47, 59, 66–67, 69, 70–71, 72–74, 77–79, 82, 122–123, 123, 125, 127, 143, 151–152, 162, 165–166, 180, 196, 236–238, 235, 246, 279, 279n1, 282, 286–289, 293, 334, 336, 337, 338, 341, 344–346, 349–352, 361, 363 individual performance 123, 128–130, 134, 282, 300 individual-to-collective 151, 153–155, 161, 163, 166 individual-to-individual 146, 150–151, 153–154, 158, 161–162, 163, 166 Ingram, P. 180, 209, 219, 234, 240, 243 institutional level 90, 97, 144–145, 198, 369 instrumental performance 94, 105 intergroup trust 150, 159–160, 365 interpersonal organizational citizenship behaviors (OCBIs) 222 interpersonal trust 4–5, 9–10, 15, 20, 23, 26, 28–29, 31–32, 34, 36, 121, 124, 134, 146, 149, 151–153, 155, 165, 168, 180, 205–207, 212, 218, 220–222, 224–225, 262, 272, 289, 291–293, 310, 325, 334, 345, 349, 364 isomorphism 23, 361, 367

376 Index

Jin, M. 10, 241, 247, 365 Jones, S. 6, 9, 180, 205, 209, 214, 223, 262, 270, 362, 364 Kang, S. 297 Kelley, H. H. 67, 72, 79–81, 213 Kent, R. 73–74, 79 Khodyakov, D. M. 91 Kidon, F. 87, 108 Kilduff, M. 185, 194, 209, 211–212, 217, 243 Kim, H. S. 71, 74–75, 79 Kim, P. H. 7, 55, 69, 143–144, 148–149, 149, 151, 153–155, 158, 162 Kim, T.Y. 48, 135 Knez, M. 179, 184, 189, 208, 210, 216–217, 221, 262 Korsgaard, A. 6–8, 14, 18, 36, 50, 55–56, 58, 90n1, 102, 109, 121–122, 209, 218, 239, 278, 298, 362–363, 366, 370 Koskinen, S. 10, 365 Kramer, R. M. 5, 50, 55, 67, 122, 143–144, 197, 207, 218, 259, 294–295, 314, 345, 369 Kulik, C. T. 155 La Ferrara, E. 236 Lam, L. 8, 31, 36, 55, 121, 124, 215, 218, 362–364 Langlinais, L. 8, 263, 364 Lau, D. 8, 30, 36, 55, 121–122, 124–126, 129, 134–135, 180, 209–210, 215–216, 218, 234, 242, 287, 362–364 leader 6, 16, 18, 29, 30, 35, 36, 49, 107, 122–124, 153, 155, 160, 165–167, 224, 261, 262, 271, 342, 347, 363 leadership 4, 26, 46, 47, 49, 58, 59, 61, 77, 78, 92, 124, 133, 153, 155, 157, 234, 260, 270, 281, 293, 299, 336, 340, 342, 347, 368, 370 Lee, C. 103, 105, 259 Lee, F. 337, 345 Lester, S. W. 7, 14, 56, 109, 122, 124, 209, 218, 239, 298 Leung, K. 336–337, 341, 344–345, 352 levels-of-analysis 3–8, 10–11, 14–18, 23–24, 25, 26–29, 27n1, 33–34, 36, 45, 66–72, 70–71, 75–77, 90, 90n1, 131, 144–146, 162, 163, 224, 236, 246, 272, 279, 279n1, 281, 307, 310, 328, 370 Lewicki, R. J. 5, 7, 26, 36, 50, 55, 109, 143– 144, 151–152, 166, 168, 192, 197, 207, 209, 214, 259, 285, 287, 294–295, 314

Lewin, K. 257 Li, A. 335 Li, F. 335 Lirtzman, S. I. 261 Lockey, S. 9, 143, 295, 362, 364 Long, C. 8, 18, 25, 33, 87, 89–93, 95, 99–110, 362 Lount, R. B. 48, 151, 213–214 Lu, C. 346 Luciano, M. M. 6, 36, 124, 215 Luhmann, N. 286, 307, 312, 326 Lumineau, F. 6, 14, 60, 87, 91, 93, 99, 104, 107–108, 326 Lyu, S. C. 310 McAllister, D. J. 7, 26, 106, 121, 127, 179–180, 207, 209, 234, 236–238, 240, 259, 285–286, 293 MacDuffie, J. 157 McEvily, B. 9, 88, 100, 104, 143, 148, 154, 179, 186–187, 189, 191–192, 199, 211, 220–223, 286, 307, 362, 364 McKnight, D. H. 50, 198, 205, 207, 214, 311–313, 316, 319, 325 managerial activities 94, 102 managerial assessments 94, 97n4, 100 Martinko, M. J. 73–74, 79 Matta, F. K. 6, 36, 124, 215 Mayer, R. C. 4, 6–7, 10, 14, 16, 26, 28, 50–51, 66–69, 77, 87, 90, 103–104, 109, 123, 125–126, 129, 134, 143–144, 148, 153, 179, 207, 214–215, 218, 239, 259–260, 262–263, 265, 270, 279, 285–286, 310, 341, 362, 366, 368 Meyerson, D. 197, 294, 369 mistrust 159, 160, 344 Möllering, G. 11, 167–168, 182, 335, 344, 369 monitoring 47, 68, 87, 92, 110, 123, 126–128, 146, 147–148, 211, 217, 222, 316, 363, 369 Morgeson, 74 Morris, M. W. 180, 209, 219, 234, 240, 243, 247, 336–337, 341, 345, 348, 352 Mossholder, K. 31, 48, 121 multilevel assessment 4 multilevel perspective 3, 5, 7, 14, 17, 108, 144, 146, 168, 279, 361, 369–370 multilevel trust 3, 5–7, 10–11, 45, 61, 70, 144, 162, 164, 166–168, 259, 299, 361–364, 366–368, 370 multiple levels 3–4, 7, 9–11, 45, 66, 88, 143, 147, 205, 207, 235, 236, 246,

Index  377

259, 261, 264, 272, 281, 309, 362, 365, 367–368, 370 Murrell, A. J. 109 mutual trust 7, 54, 56, 124, 186–187, 212, 218 Nam, J. 282, 284, 288–289 Navarro-Martinez, D. 168 network centrality 211, 235, 242 network density 216, 235, 243 network diversity 243–244 network level 180, 182, 192–193, 201, 235, 242 network theory 9, 182, 184–185, 191, 206–207, 368 network trust 9, 180, 182–185, 183, 186, 194–200, 195, 225, 364, 368 Okimoto, T. 9, 144, 148, 152–153, 155–156, 159–160, 165, 167, 362, 364 Önder, Ç. 11, 362, 365, 368 Organization Dissociative Identity Disorder (ODID) 10, 257–259, 261–266, 262, 269–273, 362 organization-focused assessments 96, 101, 104 organizational climate 17, 23, 32, 35, 72, 165, 370 organizational context 4, 14–15, 60, 131, 144, 153–154, 157, 198, 260–263, 297 organizational identities (OI) 257, 259, 266–267, 269, 271, 297 organizational level 3, 6, 8, 28, 88, 93, 97, 110, 143, 161, 167, 247, 278–279, 279n1, 282, 286, 288–289, 295, 307, 315, 319, 325–326, 336, 337, 338, 340–341, 350, 361, 363 organizational performance 78, 98, 246 organizational trust 3–6, 11, 28, 45, 100, 151, 157–158, 168, 259, 262n2, 273, 278, 281, 288, 292, 307, 325, 335, 341, 343, 347–348, 350, 352–353, 365 Ostroff, C. 6–7, 17, 27, 35, 72, 80–81, 105, 328, 362–363, 370 Page, M. J. 239 Panagopoulos, N. G. 74 Pearce, J. L. 110, 335, 340, 342–344, 348 Peretz, H. 347 Perkins-Williamson, A. 109 Perrone,V. 88, 100, 104, 143, 179, 220 Peterson, M. F. 335 Peterson, R. S. 6, 52, 56

Petriglieri, J. L. 168 Pettit, N. C. 213–214 Picot, A. 6, 58 Pillai, R. 334–335 Podolny, J. M. 184, 192, 217 Porter, D. M. 109 Pratt, M. G. 259, 267–269, 271 predictors 35, 59, 71–73, 70–71, 76, 160, 309 propositions 9–10, 46, 50–51, 53–54, 56–57, 72, 80, 95–96, 107, 125, 128–130, 132–135, 150, 160, 162, 166, 221–223, 236–239, 241–247, 337, 338, 341, 343–349, 351–352, 362, 367 prototrust 9, 182–183, 183, 185, 186, 190–196, 195, 198, 200 Proudfoot, D. 234, 246 referents 4, 27, 88, 90, 122, 123, 125, 130–132, 135, 143, 218, 259, 260, 279, 286–288, 291, 292, 295, 297, 299–300, 310, 311, 323–327, 353, 362, 363, 369, 370 relational component 206, 206, 208 relational performance 94, 100, 105–106 relational trust 9, 156, 180–183, 180, 183, 186–187, 187n1, 188–189, 195–198 Revised Causal Dimension Scale 82 Rezvani, A. 282, 284, 288, 293, 299 Riolli, L. 70, 73, 79 risk 15. 47, 50, 54, 67, 68, 122, 126, 181, 187, 196, 197, 238, 285, 308, 323 risk-taking 51, 109, 195 Rizzo, J. R. 261 Rodrigues, A. 350 Rousseau, D. M. 4, 6, 14, 16, 23, 33, 66, 87, 89, 91, 144, 179–180, 187, 214, 261, 285, 294, 308, 345, 348, 352 Sanchez-Burks, J. 337, 345 Sanders, K. 70–71, 72, 79–81, 278 Schilke, O. 5–6, 14, 60, 88, 326 Schoorman, F. D. 4, 14, 50, 67–68, 87, 109, 122–123, 125–126, 143–144, 179, 207, 239, 259–260, 279, 310 Searle, R. 3–4, 10, 87, 90, 270, 278, 285– 286, 288–289, 296–297, 334, 348–349, 365, 367 secondhand trust 9, 180, 182–183, 185– 190, 186, 187n2, 188–189, 194–198, 195, 200 self-focused assessments 96, 101, 104 Serva, M. A. 28, 51, 287

378 Index

Shah, P. P. 6, 9, 134, 146, 180, 205, 207–209, 211, 214, 219, 223, 262, 270, 362, 364 Shank, D. B. 309 Shazi, R. 180, 222–223, 234, 247 Sherman, D. K. 71, 74–75, 79 Simmel, G. 186, 210 Simmelian ties 183, 210–211, 210, 217, 219 Simpson, J. A. 239 Sitkin, S. B. 3–4, 14, 18, 25, 33, 58, 66, 87, 89–93, 95, 99, 102–103, 105–108, 144, 148, 168, 179–180, 214, 285, 287, 295 Skarlicki, D. P. 153, 155 Skinner, D. 278, 296 Sniezek, J. A. 207, 214 social identity 73, 122–123, 130–131, 133, 158–159, 185, 194, 244, 363, 369 social network 9–11, 22, 24, 58, 198, 205–208, 211–215, 217–221, 223–225, 236, 235, 239, 240n2, 242–245, 248, 294, 300, 323, 367, 370 societal level 11, 198, 327, 336, 337, 338, 340–341, 344–345 Soda, G. 9, 154, 179, 212, 220, 362, 364 Sommer, S. M. 70, 73, 79 Spector, M. D. 236, 239, 241 Steen, J. 180, 222, 234 Steffen,V. J. 238 stereotype content model (SCM) 237 Stevens, M. 148, 157 strategic activities 71, 78, 94, 98 strategic assessments 94, 97–98 strategy 152, 164, 289, 340, 341, 343, 348 Strickland, L. H. 68, 91 structural component 9, 206, 206, 209 subordinate activities 94, 101, 105 subordinate assessments 94, 103 Sullivan, P. 73 Tajfel, H. 131, 158, 244, 294 Tan, H. H. 350 Tasselli, S. 209, 212, 217 team level 6, 8, 14, 17, 21, 30, 66, 70–71, 73–76, 123, 123, 125, 127–128, 134, 279, 281, 288–289, 293, 299, 363 team performance 8, 24, 73, 128, 129, 293 team trust 5, 15, 17, 23, 29–31, 33, 71, 75, 293, 364 temporal perspective 215 third-party 9, 134, 146, 148, 154–155, 164–165, 183, 189–190, 210, 216–217, 219–221, 314, 326, 341, 368

Tomlinson, E. 8, 26, 66, 68–69, 69, 77, 81, 109, 148, 153, 168, 263, 265, 294, 364 top-down 5, 8, 11, 14, 131, 134, 270, 289, 325, 350, 362–363, 367–368, 370 trust concepts 4, 7, 16, 24, 29, 88–89, 287, 361 trust conceptualizations 7, 15, 28, 34, 207, 280, 282, 293 trust cues 10, 292–293, 309–311, 313–315, 318, 320–321, 322, 324, 328, 368 trust development phases 8, 52, 53, 59 trust levels 20, 22, 27–28, 35, 46, 49–50, 52–54, 59, 61, 294 trust repair 3, 9, 77, 143–146, 147–150, 150–162, 163, 164–168, 281, 295–296, 364–367 trust violation 55, 77, 144–145, 147, 150–157, 159–160, 162, 164–165, 167 trust-building activities 97, 100–103, 107–109 trustee 4, 7, 15, 17, 19, 21, 25, 26–36, 27n1, 47, 50, 52, 54–55, 67–69, 89n1, 90–91, 90n1, 121–122, 124, 129, 134–135, 150, 179, 181, 183, 185–188, 190, 205–210, 206, 210, 212–221, 224–225, 235, 239–242, 263, 270, 310–311 trustor 4, 7, 9, 15, 17–18, 19, 21, 23, 25–36, 25, 27n1, 50, 53–55, 67–68, 89n1, 90, 90n1, 121, 124, 126, 129, 134–135, 144–145, 150, 152–154, 162, 163, 164, 179, 181, 183, 185–188, 187n2, 190, 197, 205–210, 206, 210, 212–215, 217–221, 224–225, 239–241, 270, 279n1, 310–311, 313, 318–319, 322, 325, 328, 367 trustworthiness 10–11, 18, 50–52, 55, 67–69, 69, 71, 73, 76, 78, 90–91, 93, 103–104, 106, 128, 133, 145–146, 148, 156–157, 163, 207–209, 213–214, 216–218, 220, 223–224, 234, 235, 237, 239, 257, 260, 262–267, 262, 262n2, 269–271, 279, 281, 285–286, 289, 292, 309–311, 313, 316–318, 322, 323, 325– 328, 336–337, 336, 338–339, 340–341, 343, 345–352 unit level 6–7, 14–16, 21, 24, 25, 27–33, 35, 70–71, 74, 77–79, 224 Van de Voorde, K. 71, 78–80 van der Werff, L. 10, 60, 282, 284, 288, 292, 295, 297, 299, 310, 317, 324, 326, 362, 365, 369 Van Hiel, A. 152

Index  379

Van Overwalle, F. 152 Van Swol, L. M. 207, 214 Vlachos, P. A. 70–71, 74, 77–81, 83 voice user interface (VUI) 314–315, 318, 320, 322, 323–324 vulnerability 54, 56, 89, 109, 122, 125, 126, 129, 214, 285, 287, 291, 293, 300, 307, 334

Wigand, R. T. 6, 58 Williams, M. 10, 60, 262–263, 265, 310, 362, 366, 368 Williamson, I. 7, 14, 34, 50, 366 Wohl, M. J. A. 144, 148, 160–161 Wong, S. 209, 216–217 Wong,Y. T. 283–284, 288, 291, 297–299

Wan, C. 346 Wang, A. 109 Wang, J. 48 Wang, T. 109 Wasti, A. 11, 348–350, 362, 365, 368 Weibel, A. 26, 28, 58, 88, 90n1, 92–93, 103–105, 110, 168, 287, 296 Weick, K. E. 106, 197, 294, 369 Weiner, B. 68, 73, 76, 79, 263 Wellman, N. 6, 36, 124, 215 Welpe, I. M. 6, 58 Whitener, E. M. 91, 102–103, 105, 121, 124, 239, 278, 283–284, 289, 299

Xie,Y. H. 347 Yagoda, R. E. 312 Yamagishi, M. 198 Yamagishi, T. 198, 344, 348 Yang, H. 71, 79–81 Zaheer, A. 9, 88, 100, 104, 109, 143, 154, 157, 179, 191, 212, 220, 362, 364 Zhang, X. 234, 240 Zhang,Y. 335, 347 Zhou, J. 234, 240 Zhu, J. 8, 36, 135, 287, 362–364