Advances in Empirical Economic Research: 2022 International Conference on Applied Economics (ICOAE), Madrid, Spain, July 7-9, 2022 (Springer Proceedings in Business and Economics) [1st ed. 2023] 3031227484, 9783031227486

Citation preview

Springer Proceedings in Business and Economics

Nicholas Tsounis Aspasia Vlachvei   Editors

Advances in Empirical Economic Research 2022 International Conference on Applied Economics (ICOAE), Madrid, Spain, July 7-9, 2022

Springer Proceedings in Business and Economics

Springer Proceedings in Business and Economics brings the most current research presented at conferences and workshops to a global readership. The series features volumes (in electronic and print formats) of selected contributions from conferences in all areas of economics, business, management, and finance. In addition to an overall evaluation by the publisher of the topical interest, scientific quality, and timeliness of each volume, each contribution is refereed to standards comparable to those of leading journals, resulting in authoritative contributions to the respective fields. Springer’s production and distribution infrastructure ensures rapid publication and wide circulation of the latest developments in the most compelling and promising areas of research today. The editorial development of volumes may be managed using Springer Nature’s innovative EquinOCS, a proven online conference proceedings submission, management and review system. This system is designed to ensure an efficient timeline for your publication, making Springer Proceedings in Business and Economics the premier series to publish your workshop or conference volume. This book series is indexed in SCOPUS.

Nicholas Tsounis • Aspasia Vlachvei Editors

Advances in Empirical Economic Research 2022 International Conference on Applied Economics (ICOAE), Madrid, Spain, July 7-9, 2022

Editors Nicholas Tsounis Department of Economics University of Western Macedonia Kastoria, Greece

Aspasia Vlachvei Department of Economics University of Western Macedonia Kastoria, Greece

ISSN 2198-7246 ISSN 2198-7254 (electronic) Springer Proceedings in Business and Economics ISBN 978-3-031-22748-6 ISBN 978-3-031-22749-3 (eBook) https://doi.org/10.1007/978-3-031-22749-3 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023, corrected publication 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Paper in this product is recyclable.

Preface

This year’s conference is co-organised by the Camilo José Cela University in Madrid, Spain, and the Department of Economics of the University of Western Macedonia, Greece, after the kind invitation by Prof. Thomas Baumert who is also a co-chair of the conference. The aim of the conference is to bring together economists from different fields of Applied Economic Research in order to share methods and ideas. The topics covered include: • • • • • • • • •

Applied Macroeconomics Applied International Economics Applied Microeconomics including Industrial Organisations Applied work on International Trade Theory including European Integration Applied Financial Economics Applied Agricultural Economics Applied Labour and Demographic Economics Applied Health Economics Applied Education Economic

All papers presented in ICOAE 2022 and published in the conference proceedings were peer reviewed by at least two anonymous referees. In total, 96 works were submitted from 17 countries while 68 papers were accepted for publication in the conference proceedings. The acceptance rate for ICOAE 2022 was 70%. The full-text articles will be published online by Springer in the series Springer Proceedings in Business and Economics, and they will be included in the SCOPUS database for indexing. The organisers of ICOAE 2022 would like to thank: • The Scientific Committee of the conference for their help and their important support for carrying out the tremendous work load organising and synchronising the peer-reviewing process of the submitted papers in a very specific short period of time. v

vi

Preface

• The anonymous reviewers for accepting to referee the submitted to the conference papers and submit their reviews on time for the finalisation of the conference programme. • Prof. Thomas Baumert, for accepting to host the conference at the Camilo José Cela University and provide the required resources. • The local organising committee and the volunteering students for their help for the success of the conference. • Mr. Gerassimos Bertsatos for running the reception desk of the conference and Mr. Lazaros Markopoulos and Mr. Stelios Angelis from the Department of Economics and Informatics, of the University of Western Macedonia, respectively, for technical support. Kastoria, Greece

Nicholas Tsounis Aspasia Vlachvei

Contents

1

2

3

4

5

6

Strategic Agility and Success Perception of Polish SMEs: An Alternative Operationalization to Pre-COVID-19 and COVID-19 Business Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tomasz Sikora and Ewa Baranowska-Prokop R&D Cooperation of Firms and Product Market Competition: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jacek Prokop Market Restrictions of Contracting Out the Public Service at the Municipal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Beáta Mikušová Meriˇcková, Daniela Mališová, and Kristína Murínová Description and Categorisation of Agriculture Holdings in the Region of Western Macedonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theodoros Siogkas, Katerina Melfou, Athanasios Ragkos, and Apostolos Polymeros

1

27

37

55

Debt and Taxes as Value-Added Factors for Multinational Enterprises: International Policy Implications . . . . . . . . . . . . . . . . . . . . . . . . Juan José Durán-Herrera and Prosper Lamothe-Fernández

69

Modelling and Forecasting GDP of Greece with a Modified Exponential Smoothing State Space Framework . . . . . . . . . . . . . . . . . . . . . Melina Dritsaki and Chaido Dritsaki

89

7

Last Mile and Blockchain: Opportunities and Challenges. . . . . . . . . . . Rafael Villa, Marta Serrano, and Tomás García

8

Total Factor Productivity and Entrepreneurship: Creative Self-Destruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jose Maria Sevilla Llewellyn-Jones

111

133

vii

viii

9

10

Contents

Generation Z Intention to Comply with Non-mandatory Government Measures for Self-protection of COVID-19 and SARS-CoV-2 Variants After Restriction Withdrawals . . . . . . . . . . Irene (Eirini) Kamenidou, Aikaterini Stavrianea, Spyridon Mamalis, Evangelia-Zoe Bara, Ifigeneia Mylona, and Stavros Pavlidis

151

The Spatial Distribution of the Population in Peninsular Spain: An Evolution of a Permanent Nature . . . . . . . . . . . . . . . . . . . . . . . . . . Daniel del Castillo Soto and Thomas Baumert

167

11

Food Waste in Greece: An Empirical Study . . . . . . . . . . . . . . . . . . . . . . . . . . . Electra Pitoska and Grana Vaya

12

Success Factors in Public-Private Partnership of High-Speed Railway Infrastructures: Elements for Improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mario González-Medrano, Tomás García Martín, and José-María Rotellar-García

13

Holistic Evaluation of Technology Transfer Extension Programmes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evropi-Sofia Dalampira, Ioannis Tsoukalidis, Dimitra Lazaridou, Smaragda Nikouli, Anastasios Livadiotis, and Anastasios Michailidis

185

195

205

14

E-Banking Loyalty and Its Background: A Bibliometric Analysis . Natacha López-Hernando, Cristina Loranca-Valle, and Pedro Cuesta-Valiño

15

Evaluating European Climate Policy Impact on the CO2 Emissions Per Capita Convergence Process in the European Union Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dionisio Ramírez-Carrera, Gemma Durán-Romero, Ana M. López, and José Antonio Negrín de la Peña

235

The Economics of Civil Orders and Medals in Spain: An Update After Ten Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thomas Baumert and Beatriz Luceño-Ramos

257

‘Ctrl+Supr’ Versus ‘Shift+Comp’ Industrialization and Business Services as Engine of Economic Growth . . . . . . . . . . . . . . . . . . . . Andrés Maroto Sánchez

275

Measuring Stakeholders’ Influence on Business Performance: Case of Slovakia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yaroslav Lysenko, Juraj Medzihorsky, and Zdenka Musova

301

16

17

18

19

Testing for Sequences and Reversals on Bitcoin Series . . . . . . . . . . . . . . . Prodromos Tsinaslanidis and Francisco Guijarro

215

317

Contents

20

21

22

23

24

25

26

27

28

29

30

CHAID Algorithm Applied in a Post-rating and Managing the Business-Risk Insurance Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Assia Khenoussi Management of Working Capital and Capital Structure in Relation to the Economic Value Added of Selected Companies in the Czech Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Markéta Šeligová

ix

327

341

Implementation of Facial Biometric Technologies: A Business Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Janka Taborecka-Petrovicova, Michal Budinsky, and Marta Sipulova

357

Bitcoin and Corporate Balance Sheets: Strategic Reserve Asset or a New Business Model? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ivan Sedliaˇcik and Michal Ištok

375

Lufthansa Airlines: The Microeconomic and Macroeconomic Environment of the Company and the Industry in 2020 and Its Readiness Against Crisis (Case Study) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spyros Zervas

385

The Difficulties of Access to Bank Financing by SMEs: A Questionnaire Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yasmine Derradj and Hanya Kherchi Medjden

405

Impact of Integration of Technology on Teaching and Learning in the Primary Schools Classroom on Economic Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stavros Kalogiannidis, Stavroula Savvidou, George Konteos, and Olympia Papaevangelou

417

The Rationale of Integrated Planning and Sustainable Development Strategies in the Development of Local Government Regional Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stavros Kalogiannidis, Stamatis Kontsas, Eirini Eleni Nikolaou, and Fotios Chatzitheodoridis

433

Quantifying the Effects of Recent Economic and Fiscal Crises on Income Inequality in Greece . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . George Petrakos, Konstantinos Rontos, Chara Vavoura, and Ioannis Vavouras

455

Ranking Stock Markets Informational (In)Efficiency During the COVID-19 Pandemic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joanna Olbrys and Elzbieta Majewska

473

Income Inequality, Economic Freedom, and Economic Growth in Greece: A Multivariate Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . Antonis Tsitouras and Harry Papapanagos

485

x

Contents

31

Gender Differences in Social Discounting of Monetary Losses . . . . . . Bartlomiej Wi´snicki and Adam Karbowski

505

32

Commodity Prices and Dry Bulk Shipping Stock Returns . . . . . . . . . . Nektarios A. Michail and Konstantinos D. Melas

517

33

Invasion of Ukraine and Effects on the German Economy: A CGE Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gerassimos Bertsatos and Nicholas Tsounis

527

From Competition to Concentration: The Case of the Media Sector in Greece . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Athanasios Papathanasopoulos

545

Staging Though the Pandemic. Evaluation and Communication in the University . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Angela Besana, Annamaria Esposito, and Chiara Fisichella

561

Business Model Canvas Design for Solar-Assisted Thermal Air Conditioning Using CO2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saad Dilshad, Naeem Abas, Qadeer ul Hasan, and Ali Raza Kalair

573

34

35

36

37

Energy Audit in Buildings for Sustainable Economic Development Ali Raza Kalair, Mehdi Seyedmahmoudian, Naeem Abas, Muhammad Shoaib Saleem, Alex Stojcevski, Saad Mekhilef, and Kek Koh

587

38

Construal Levels and Social Discounting of Monetary Losses . . . . . . Adam Karbowski, Bartłomiej Wi´snicki, and Jerzy Osi´nski

613

39

Sport NFTs in the Chinese Market: A Qualitative Study . . . . . . . . . . . . Roberto C. Sandulli Saldaña

625

40

The Impact of the COVID-19 Pandemic on Culture and the Creative Industries in a Selected Region in the Slovak Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tatiana Corejova, Anna Padourova, Slavka Pitonakova, and Maria Rostasova

641

41

Philanthropy and Partnerships for the Growth of Italian Culture . Angela Besana and Annamaria Bagnasco

42

Impacts of the Implemented COVID-19 Measures on the Sectoral Structure of Companies in Slovakia. . . . . . . . . . . . . . . . . . Hussam Musa, Stanislava Honzová, and Peter Pisár

665

Financial Performance of Slovak Banks and Insurance Companies: COVID-19 Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Janka Grofˇcíková, Katarína Izáková, and Dagmar Škvareninová

685

43

653

Contents

44

45

The Relationship Between Entrepreneurship and Economic Growth in the Pandemic Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Miguel-Ángel Galindo-Martín, María-Soledad Castaño-Martínez, and María-Teresa Méndez-Picazo The Impact of the COVID-19 Pandemic on the Automotive Industry in Visegrad Four Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hussam Musa, Frederik Rech, Zdenka Musova, and Chen Yan

xi

711

727

46

Social Protection, Poverty, and Inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saby Giannina Romero Medina

47

Socioeconomic Factors That Stimulate Entrepreneurship and Economic Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rafael-Sergio Pérez Pujol

763

The Foresight of Explicit and Valuation of Tacit Synergies in International Alliance by Real Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ˇ Andrejs Cirjevskis

783

The Impact of the Macroeconomic Environment on the Development of Business Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tomáš Pražák

803

The Effect of the Length of the Inquiry Form on the Conversion of the Corporate Website . . . . . . . . . . . . . . . . . . . . . . . . . . Tereza Ikášová

813

Link Between Women, Business, and the Law Index and Countries’ Governance and Risk Indicators . . . . . . . . . . . . . . . . . . . . . Nihal Bayraktar

825

MM Model with Taxes and Its Verification: Suitable Methodology to Avoid “Automatic” Negative Relations Between Leverage and Profitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Juraj Medzihorský, Hussam Musa, Peter Krištofík, and Yaroslav Lysenko

849

48

49

50

51

52

53

54

Leverage of Non-financial Firms and Its Determinants in the Pre-pandemic and Pandemic Period—Some Evidence from CEE Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peter Krištofík, Hussam Musa, and Juraj Medzihorský Relationship Between Innovation and Market Performance: The Chilean Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geoffrey Ditta, Cristián Gutiérrez Rojas, and Jerome Smith

743

857

871

xii

55

56

57

58

59

60

61

62

Contents

Segmenting Generation Z Based on Organic Food Decision-Making: A Preliminary Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spyridon Mamalis, Irene (Eirini) Kamenidou, Stergios Gkitsas, Aikaterini Stavrianea, Despoina Gkagkani, and Stavros Pavlidis

891

Is There Really a Gender Gap That Disfavors Female Painters? An Experimental Study in Spain. . . . . . . . . . . . . . . . . . . . . . . . . . . . Thomas Baumert, Pedro Galván-Lamet, and Esther Valbuena-García

905

Determinants of Marketing Activity by Family Business Owners: A Generational Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Irene Samanta

919

Modelling Tourist Arrivals with the Use of Differential Equations: The Case of Alternative Tourism in Greece . . . . . . . . . . . . . . Gerassimos Bertsatos, Zacharoula Kalogiratou, Theodoros Monovasilis, and Nicholas Tsounis Managing Public Sector in the Digital Reform Era: Organizational Factors and Their Impact on the Digital Transformation at the Greek Public Administration . . . . . . . . . . . . . . . . . Panagiota Xanthopoulou, Ioannis Antoniadis, and Sotiria Triantari

937

947

Shaping the Online Customer Experience Through Website Elements: An Integrated Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eirini Koronaki, Aspasia Vlachvei, and Anastasios Panopoulos

963

Social Media Engagement: What Matters? An Empirical Study on Greek Agri-Food Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Afroditi Kitta, Ourania Notta, and Aspasia Vlachvei

977

A Model of the Social Pillar of Internal CSR for a Manufacturing Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jan Slavík

993

Correction to: Gender Differences in Social Discounting of Monetary Losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

C1

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1009

Chapter 1

Strategic Agility and Success Perception of Polish SMEs: An Alternative Operationalization to Pre-COVID-19 and COVID-19 Business Conditions Tomasz Sikora and Ewa Baranowska-Prokop

Abstract The main purpose of this chapter is to present and discuss two elements influencing the success of Polish SMEs: perceived environmental hostility and strategic agility. An alternative operationalization of strategic agility (compared to our two previous studies: Sikora, Baranowska-Prokop, The role of strategic agility and economic environment’s friendliness-hostility in explaining success of polish SMEs. In: Tsounis N, Vlachvei A (eds) Advances in longitudinal data methods in applied economic research. Springer proceedings in business and economics. Springer, pp 267–284, 2021; Sikora, Baranowska-Prokop, Strategic agility and economic environment’s friendliness-hostility in explaining success of polish SMEs in the phase of COVID-19 pandemic. In: Tsounis N, Vlachvei A (eds) Advances in quantitative economic research. Springer proceedings in business and economics, pp 477–496, 2022) is proposed, based on the EMICO (or entrepreneurial marketing) scale (Fiore et al, J Mar Dev Comp 7(4):63–86, 2013). Environmental friendliness-hostility has been operationalized as the properties of the external (or macro) environment in which firms operate as perceived by respondents. Data analysis is based on two samples of Polish SMEs: a “pre-COVID-19” sample (data collected in 2019) and a “COVID-19” sample (data collected at the turn of 2020 and 2021). Results show that (when total samples are taken into account), firstly, strategic agility was significantly and positively related to two out of three market performance measures for the “pre-COVID-19” sample and to one out of three measures for the “COVID-19” sample, secondly, external environment friendliness – hostility was not significantly related to any of the measures of market performance, however, it played a moderating role between strategic agility and market performance. This hypothesized moderating role implied that the more hostile the external environment the stronger the positive relationship between strategic agility and market performance. Results for the “pre-COVID-19” sample

T. Sikora () · E. Baranowska-Prokop SGH Warsaw School of Economics, Warsaw, Poland e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_1

1

2

T. Sikora and E. Baranowska-Prokop

are not in line with this hypothesis, because positive and significant relationships between strategic agility and most of the market performance measures have been only found in the case of companies operating in a friendly environment, but for those doing business in neutral or hostile environments. Results for the “COVID-19” sample seem to support the moderation hypothesis. Weak positive and significant relationships between strategic agility and most of the market performance measures have been found for the companies operating in a friendly-to-neutral environment, and not for those doing business in a very friendly environment. In general, strategic agility is in all but one case either positively related to market performance measures or positive relationships are not significant which means that, in the absence of significant negative relationships, being strategically agile either helps or – “at worst cases” – causes no harm to enterprises’ performance. Keywords Strategic agility · Environmental hostility · Competitive strategies · Market performance · SMEs

JEL Code L11, M10, M16, M31

1.1 Introduction Evidence on the impact of the COVID-19 crisis on SMEs from numerous business surveys indicates severe disruptions and concerns among small firms (OECD, 2020). The effect on SMEs is especially harsh, particularly because of higher levels of vulnerability and lower resilience related to their size. SMEs may find it difficult to gain information not only on measures to adopt possible business strategies in order to diminish the shock but also about government initiatives that provide support and are available for them. The conclusions of Adian et al. (2020) clearly indicate that SMEs are more than 8% more likely to temporarily close their operations due to COVID-19 than larger firms, across all countries and sectors in the sample. According to a survey carried out among SMEs in 132 countries by the International Trade Centre, two-thirds of micro and small firms reported that the crisis strongly affected their business operations, and one-fifth of them indicated the risk of shutting down permanently within 3 months (ITC, 2020, p. 5). Such a contemporary global environment is often characterized as VUCA (volatility, uncertainty, complexity, and ambiguity) conditions and has been forcing enterprises to remain competitive by showing not only resilience but also by exploring new opportunities. Undeniably, strategic agility is an answer to a dynamic business environment that is loaded with fluctuations and unpredictability (Goldman et al., 1995). The concept of strategic (or organizational) agility has been coined first in the mid1980s and then discussed and explored by many researchers. According to the definition of Yusuf et al. (1999) organizational agility is a successful exploitation of competitive bases (speed, flexibility, innovation, proactiveness, quality, and

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

3

profitability) through the integration of reconfigurable resources and best practices in a knowledge-rich environment to provide customer-driven products and services in a fast-changing market environment (Yusuf et al., 1999, p. 37). Long (2000) is the first author to perceive this concept in the strategic management sense. He defines it as “not only maintaining the flexibility to respond quickly to changing circumstances and emerging opportunities but also concentrating on a clear strategic purpose” (p. 38). Another definition coined by Sambamurthy et al. (2003) points out that strategic agility can be put simply as the capability of a firm to take advantage of market opportunities quickly and by surprise. The key question is to detect how well a firm can sense a forthcoming change and make a fast adjustment of processes, actions, and resources. The dynamic approach to strategic agility was offered by Eisenhardt and Martin (2000) and attempts to indicate how a firm builds, leverages, and reconfigures capabilities to adopt to environmental changes. Strategic agility is also perceived as an important determinant of firm’s success in complex and turbulent environments: Heakel (1999), Zaheer and Zaheer (1997). According to Roberts and Grover (2012), the most important capabilities that build strategic agility are: sensing and responding quickly to environmental threats and opportunities. The best-known definition of strategic agility concept has been offered by Weber and Tarba (2014) and it goes as follows: “the ability of management to constantly and rapidly sense and respond to a changing environment by intentionally making strategic moves and consequently adapting the necessary organizational configuration for successful implementation”. Kumkale (2016) argues that strategic agility is an instrument for creating company’s competitive advantage. According to Widjajani and Nurjaman (2020), strategic agility is a Meta capability that involves not only allocating sufficient resources for development and deployment but also staying agile by balancing dynamic capabilities over time. The application of strategic agility concept could also be an effective instrument for survival and better performance in the turbulent COVID-19 business environment for SMEs in Poland. According to PARP (2021) report, throughout the pre-COVID-19 period, the number of newly formed Polish SMEs was outgrowing the number of eradicated. The total count of SMEs reached 2.1 million and constituted 99.8% of all enterprises in Poland. SMEs are generating nearly half of the Polish GDP (49.1%). The biggest share goes to micro-firms which create nearly one-third of GDP (about 29%). The COVID-19 pandemic was perceived as a shock. At the beginning of this epidemic crisis, nearly 90% of SMEs in Poland expected a decrease in turnover and 75% were anticipating problems with liquidity (Statista, 2021). In our previous articles (Sikora & Baranowska-Prokop, 2021, 2022) strategic agility has been operationalized through distinction between firms applying monostrategy (less agile firms, doing business with customers on similar terms) vs. multistrategy (more agile firms, differentiating conditions depending on the customer) in two aspects: price level (setting one similar price level for all customers vs. having multiple price setting patterns depending on customer) and product quality level (producing all goods at a given quality level vs. producing the same goods at

4

T. Sikora and E. Baranowska-Prokop

different quality levels depending on customers’ requirements and/or their ability to detect quality variations). Questionnaires sent to Polish SMEs in 2019 and 2020/2021 also raised questions about other aspects of their strategies, like entrepreneurial marketing capabilities based on the EMICO scale elaborated by Fiore et al. (2013). Comparing the content of items of this scale and the above-mentioned definitions of strategic agility suggests that three statements: – We regularly pursue untapped market opportunities regardless of budgetary or staff constraints. – My business excels at identifying marketing opportunities. – When new market opportunities arise, my business very quickly acts on them, may be considered as an operationalization of strategic agility. These three statements form a separate dimension of the EMICO scale, a subdimension “opportunity-driven” within a dimension “opportunity-vigilance.” The second sub-dimension, not used in this research, is called “proactive orientation and other dimensions of this scale are: “consumer-centric innovation” and “value creation” (Fiore et al., 2013). The difference between operationalization of strategic agility based on monovs. multi-strategy and the one based on EMICO scale is that the first one is more proactive (solutions for facing diverse market conditions are prepared in advance), and the second one is more reactive (focused on seizing opportunities and quickly acting on them).

1.2 Research Method and Hypotheses Our findings are based on surveys conducted among company’s owners and managers from two samples of Polish SMEs: a “pre-COVID-19” (first sample) and a “COVID-19” (second sample).1 Data related to the first sample were collected in May–July 2019 by AMS (certified research company) through questionnaires sent to a representative sample of Polish SMEs established after 2004 (various branches of manufacturing industries, services not included). The total sample size was 240 firms randomly selected from the database of 2969 Polish SMEs. Random selection was made within two strata of non-exporters and exporters with exports being at least 25% of their sales being sent abroad, and two strata of small and medium-sized companies. The data collection was conducted through the Internet questionnaire (CAWI) and telephone interviews (CATI). The share of exporters vs. non-exporters in the sample is 50% vs. 50%. The share of medium-sized vs. small enterprises is 33.3% vs. 67.7%.

1 Both surveys have been financed by statutory funds obtained from SGH Warsaw School of Economics.

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

5

Data related to the second sample were collected in December 2020 to January 2021 by Indicator certified research company through questionnaires sent to a representative sample of Polish SMEs established after 1995 (various branches of industry, services not included). The total sample size was 219 firms selected from the database of 1395 Polish SMEs which met the sampling criteria. Random selection was made within two strata of small and medium-sized companies, and two strata of non-exporters and exporters with at least 15% of their sales being sold abroad. Data were obtained through the Internet questionnaire (CAWI) and telephone interviews (CATI). The share of exporters is 49.8% and the share of nonexporters in the sample is 50.2%. The share of medium-sized and small enterprises constitutes 37.9% and 62.1%, respectively. Both samples differ in terms of the threshold for exports share in total sales (25% vs. 15%), age (inception after 2004 vs. 1995), and in the fact that in the “COVID-19” sample only managers of companies from medium-tech and high-tech industries, serving either B2B clients only or B2B and B2C have been interviewed (the “preCOVID-19” sample also included firms selling goods to B2C customers only). Two hypotheses, based on theory and results of previous research, imply direct relationships between market performance and separately, strategic agility and environment friendliness – hostility.2 The first hypothesis assuming positive relationship between strategic agility and measures of enterprises’ performance is based on, for example, Roberts and Grover who argued that “in today’s hypercompetitive environment, firms that are agile tend to be more successful” and found positive relationships between agility measures they used and enterprises’ performance (Roberts & Grover, 2012, p. 579). H1: There is a positive relationship between strategic agility and firms’ results (performance). Oleksiuk and Ple´sniak (2022) found in the case of Polish SMEs that perceiving the environment as hostile hindered firms’ internationalization. As far as market performance is concerned two studies (Sikora & Baranowska-Prokop, 2021, 2022) found no direct relationship between external environment friendliness-hostility scale and market performance measures. However, for the sake of clarity and completeness, we re-test the hypothesis implying a positive relationship between environmental friendliness and firms’ performance. The hypothesis is based on an assumption that hostility of external environment reduces resources, decreases profit margins, and handicaps maneuverability (Miller and Friesen (1983). H2: There is a positive relationship between the degree of external environment friendliness and market performance (alternatively, the more hostile external environment the worse economic results).

2 Hypotheses are in line with two previous studies (Sikora & Baranowska-Prokop, 2021, 2022), but they are reformulated taking into account another operationalization of strategic agility. Previously, strategic agility was a dichotomous variable, now it is an interval variable with values ranging from 3 to 21 (see the section Independent variables).

6

T. Sikora and E. Baranowska-Prokop

Strategic agility should be particularly useful in hostile, risky, and turbulent environment and less useful in friendly environment (this implies a moderating role of external environment friendliness – hostility). Therefore, it could be expected to observe the strongest positive relationship between strategic agility and firms’ results in hostile environment, weaker ones in neutral environment and the weakest ones in friendly environment. H3: Positive relationship between strategic agility and firms’ results is stronger in hostile environment compared to neutral environment and is stronger in neutral environment compared to friendly environment.

1.3 Research Results 1.3.1 Independent Variables: Firms’ Characteristics, Strategic Agility, Environment Friendliness – Hostility, Balance of COVID-19 Consequences The first set of independent variables is related to firms’ characteristics such as size (small vs. medium-sized) and exporting activities.3 The second set of independent variables is related to: – External environment friendliness – hostility – Strategic agility – Balance of COVID-19 pandemic consequences for enterprises (for the “COVID” sample only) Detailed description of properties of the environment friendliness-hostility scale have been given in Sikora and Baranowska-Prokop (2021) in tables 7 and 8 (p. 275) for the “pre-COVID-19” sample, and in Sikora and Baranowska-Prokop (2022) in tables 33.3 (p. 483) and 33.5 (p. 484) for the “COVID-19” sample. It is to be emphasized that, unexpectedly, respondents from the second sample evaluated external environment during COVID-19 pandemic as more friendly than respondents from the first sample. Therefore, for the “pre-COVID-19” sample it is possible to analyze the external environment in three forms: hostile (20.0% of enterprises), neutral (53.8% of enterprises), and friendly (26.3% of enterprises). For the “COVID-19” sample, however, only two forms of environment could be distinguished: very friendly (50.7% of enterprises) and friendly-to-neutral (49.3% of enterprises). The reasons for such a paradox are unknown. Maybe, the fact that respondents experienced the first, very severe period of lockdown changed their perception of what hostile environment is and, after lockdown suspension during the

3 Percentages

related to those characteristics for both samples have been given in the Sect. 1.2.

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

7

Summer of 2020 and the availability of large-scale government assistance programs (Sikora & Baranowska-Prokop, 2022), the second, more selective wave of lockdown has been perceived as not particularly difficult or disturbing. Strategic agility scale is based on three items from the EMICO scale (Fiore et al., 2013): – We regularly pursue untapped market opportunities regardless of budgetary or staff constraints. – My business excels at identifying marketing opportunities. – When new market opportunities arise, my business very quickly acts on them. For the “pre-COVID-19” sample reliability of this scale is high (Cronbach’s alpha coefficient is above 0.8 (0.876)). Therefore, values of the items have been added and the final single scale has values between 3 (very low strategic agility) and 21 (very high strategic agility). Descriptive statistics for items and the scale are presented in Table 1.1. For the “COVID-19” sample reliability of this scale is very high – the Cronbach’s alpha coefficient is above 0.9 (0.919). Also in this case values of the items have been added and the final single scale has values between 3 (very low strategic agility) and 21 (very high strategic agility). Descriptive statistics for items and the scale are presented in Table 1.2. It seems that strategic agility slightly improved during the COVID-19 pandemic (the mean value increased by 1 point on the scale from 12.69 to 13.68). The third independent variable included in the analysis is the balance of the COVID-19 pandemic consequences for enterprises. Among four types of negative consequences, most of the respondents indicated: supply chain disruptions (53.4%), and an increase in losses or decrease in profits (35.6%); They have also pointed out at customers’ bankruptcy/insolvability (14.2%) and the lack of suppliers (6.8%). Some respondents declared that their firms had experienced also positive consequences of the pandemic, reporting three elements such as: bankruptcy of some competitors (8.2%), increase in demand for the firm’s products (1.4%), and increase of profits or decrease of losses (1.4%). An index of the strength and directions (negative to positive) of COVID-19 pandemic consequences has been created in the way that each of the negative consequences was marked as “−1” and each of the positive consequences as “1” (distribution of this index is skewed toward the predominance of negative consequences with the balance of “−3” for 1.4% of companies in the “COVID19” sample, “−2” for 13.2%, “−1” for 76.3%, “0” for 3.2%, “1” for 4.6%, “2” for 0.9%, and “3” for 0.5%).

1.3.2 Dependent Variables: Indicators of Firms’ Performance Due to lack of precise figures about profits and sales for Polish SMEs, descriptive questions about market performance had to be applied.

Source: own elaboration

We regularly pursue untapped market opportunities regardless of budgetary or staff constraints When new market opportunities arise my business very quickly acts on them My business excels at identifying marketing opportunities Strategic agility scale: (range of values: 3–21) Valid N (listwise)

Descriptive statistics

4.41 4.31 12.69

240 240 240

Mean Statistic 3.97

240

N Statistic 240

1.505 4.223

1.595

Std. deviation Statistic 1.617

Table 1.1 Descriptive statistics for the scale of strategic agility (“pre-COVID-19” sample)

−.420 .157 .157 .157

−.415 −.544 −.494

−.348 −.284

Kurtosis Statistic −.828

Skewness Statistic Std. error −.252 .157

.313 .313

.313

Std. error .313

8 T. Sikora and E. Baranowska-Prokop

Source: own elaboration

We regularly pursue untapped market opportunities regardless of budgetary or staff constraints When new market opportunities arise my business very quickly acts on them My business excels at identifying marketing opportunities Strategic agility scale: (range of values: 3–21) Valid N (listwise)

Descriptive statistics

4.86 4.68 13.68

219 219 219

Mean Statistic 4.14

219

N Statistic 219

1.000 2.894

1.085

Std. deviation Statistic 1.033

Table 1.2 Descriptive statistics for the scale of strategic agility (“COVID-19” sample)

.164 .164 .164

.002 −.396 −.205

Skewness Statistic Std. error .091 .164

.327 .327

.327

−.093 −.428 −.522

Std. error .327

Kurtosis Statistic −.488

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . . 9

10

T. Sikora and E. Baranowska-Prokop

For the “pre-COVID-19” sample three types of measures of market performance have been taken into account: – General declarations about profits or losses for 2018 – Declarations about sales evolution in 2018 compared to 2017 – Composite variable reflecting two-year general market performance (based on a factor obtained from four variables: general declarations about profits or losses for 2018, analogous declarations for 2017, declarations about sales evolution in 2018 compared to 2017, and analogous declarations for sales evolution in 2017 compared to 2016) Respondents from companies evaluated, in general terms, the level of profit/loss and sales dynamics on a 5-point scale. The degrees of the profit/loss scale include substantial loss (1), small loss (2), result close to zero (3), small profit (4), and substantial profit (5). The degrees of a scale measuring sales dynamics include substantial decrease by two-digit percent (1), decrease by one-digit percent (2), no change (3), increase by 1-digit percent (4), substantial increase by two-digit percent (5). Since correlation between market performance measures is high (correlation coefficients around 0.8), a composite variable “Market performance 2017–2018” was obtained as a factor via factorial analysis on the above-mentioned four variables (one factor was obtained explaining 68.8% of variance). Table 1.3 presents descriptive statistics related to the measures of firms’ performance for the “pre-COVID-19” sample. Distribution of variables related to profits and sales shows skewness towards higher ends of scales reflecting good firms’ performance (i.e. substantial majority of firms declared to be profitable and reported increasing sales). Measures of market performance for the “COVID-19” sample are similar to those used for the “pre-COVID-19” sample. Also, three types of measures of market performance have been taken into account: – General declarations about profits or losses for 2020. – Declarations about sales evolution in 2020 compared to 2019. – Composite variable reflecting two-year general market performance based on a factor obtained from four variables: general declarations about profits or losses for 2020, analogous declarations for 2019, declarations about sales evolution in 2020 compared to 2019, and analogous declarations for sales evolution in 2019 compared to 2018. Table 1.4 presents descriptive statistics for declarations about profits or losses and sales evolution for the “COVID-19” sample. The period of data collection, that is, December 2020 and January 2021 is characterized by uncertainty about 2020 results, therefore they have been referred to as “estimated” in the questionnaire. The composite variable “Market performance 2020 (estimated)–2019” was obtained as a factor via factorial analysis on the above-mentioned four variables (one factor was obtained explaining 81% of variance).

Source: own elaboration

Financial results in 2018 Sales dynamics in 2018 compared to 2017 Factor Market performance 2017–2018 Valid N (listwise)

Descriptive statistics N Statistic 240 240 240 240

Mean Statistic 3.92 3.71 .00

Std. deviation Statistic .782 .881 1.00

Table 1.3 Descriptive statistics for measures of market performance (“pre-COVID-19” sample) Skewness Statistic −.593 −.134 −.179

Std. error .157 .157 .157

Kurtosis Statistic .527 −.560 −.304

Std. error .313 .313 .313

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . . 11

Source: own elaboration

Financial results in 2020 (estimated) Sales dynamics in 2020 (estimated) compared to 2019 Market performance 2019–2020 Valid N (listwise)

Descriptive statistics N Statistic 219 219 219 219

Mean Statistic 3.36 2.99 .00

Std. deviation Statistic 0.945 .995 1.00

Table 1.4 Descriptive statistics for measures of market performance (“COVID-19” sample) Skewness Statistic −1.042 .525 −.339

Std. error .164 .164 .164

Kurtosis Statistic .685 −.602 −.389

Std. error .327 .327 .327

12 T. Sikora and E. Baranowska-Prokop

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

13

Comparison of means between financial results and sales dynamics in Tables 1.3 and 1.4 shows deterioration of both indices of market performance due to COVID19 pandemic. Although the majority of companies still declared profits in 2020 (reflected by the mean above 3 and negative skewness for this measure of market performance), sales dynamics seem to have been severely, negatively hit by the COVID-19 pandemic (unlike the “pre-COVID-19” sample, skewness turned from negative to positive and mean decreased substantially from 3.71 to 2.99). Explanation of such a result may be traced to the Polish government’s policy of helping enterprises via “anti-COVID19 shields.” This financial and fiscal support made it possible for firms to remain profitable in spite of declining sales.

1.3.3 Verification of Hypotheses Related to the Relationship Between Strategic Agility, Environmental Friendliness – Hostility and Firms’ Performance Hypothesis 1 assumes a positive relationship between the market performance of enterprises and strategic agility. Table 1.5 presents correlations between measures of market performance and strategic agility scale for the total “pre-COVID-19” sample and separately for enterprises operating in friendly, neutral, and hostile external environments. For the total sample correlations between financial results in 2018 and strategic agility are close to zero. For sales dynamics and the factor reflecting two-year market performance correlations, although positive and significant, are weak or very weak (correlations related to particular forms of external environment will be commented in the section related to hypothesis 3). Table 1.6 presents correlations between measures of market performance and strategic agility scale for the total “COVID-19” sample and separately for enterprises operating in very friendly and friendly-to-neutral external environment. In the case of the “COVID-19” sample correlations between all measures of market performance are positive and significant, albeit they are very weak or weak (correlations related to particular forms of external environment will be commented in the section related to hypothesis 3). Correlations give very weak support for hypothesis 1, but a more rigorous verification is needed with other variables taken into account. Hypothesis 2 assumes positive relationship between market performance and environmental friendliness. Table 1.7 presents descriptive statistics and results of comparison of means related to measures of market performance for the “pre-COVID-19” sample. Support for hypothesis 2 is partial (for two out of three measures of market performance) and very weak (differences between two extreme groups of firms are significant at “mild” range of “p” values 0.05 and 0.1).

Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N

a Correlation

Source: own elaboration is significant at the 0.05 level (two-tailed) b Correlation is significant at the 0.01 level (two-tailed)

Factor Market performance 2017–2018 (nine outliers removed)

Factor Market performance 2017–2018

Sales dynamics in 2018 compared to 2017 (three outliers removed)

Sales dynamics in 2018 compared to 2017

Financial results in 2018 (four outliers removed)

Financial results in 2018

Strategic agility (total sample) 0.067 0.302 240 0.083 0.204 236 .172b 0.007 240 .208b 0.001 237 .140a 0.030 240 .176b 0.007 231

Strategic agility (friendly environment) 0.079 0.536 63 0.007 0.957 62 .400b 0.001 63 .487b 0.000 62 .280a 0.026 63 .386b 0.002 60

Table 1.5 Correlations between measures of market performance and strategic agility (“pre-COVID-19” sample) Strategic agility (neutral environment) −0.029 0.748 129 0.012 0.896 128 0.076 0.392 129 0.076 0.392 129 0.039 0.659 129 0.067 0.459 126

Strategic agility (hostile environment) 0.267 0.067 48 .292a 0.049 46 0.111 0.451 48 0.185 0.219 46 0.213 0.146 48 0.250 0.098 45

14 T. Sikora and E. Baranowska-Prokop

Pearson Correlation Sig. (two-tailed) N Pearson Correlation Financial results in 2020 (six outliers removed) Sig. (two-tailed) N Sales dynamics in 2020 Pearson Correlation (estimated) compared to 2019 Sig. (two-tailed) N Sales dynamics in 2020 compared Pearson Correlation to 2019 (nine outliers removed) Sig. (two-tailed) N Pearson Correlation Market performance 2019–2020 Sig. (two-tailed) N Market performance 2019–2020 Pearson Correlation (six outliers removed) Sig. (two-tailed) N

a Correlation

Source: own elaboration is significant at the 0.05 level (two-tailed) b Correlation is significant at the 0.01 level (two-tailed)

Financial results in 2020 (estimated)

Strategic agility (total sample) .147a 0.03 219 .223b 0.001 213 .171a 0.011 219 .273b 0.001 210 .172a 0.011 219 .223b 0.001 213

Strategic agility (very friendly environment) 0.161 0.091 111 .214a 0.026 109 0.101 0.29 111 .247a 0.011 106 0.125 0.192 111 0.153 0.11 110

Table 1.6 Correlations between measures of market performance and strategic agility (“COVID-19” sample) Strategic agility (friendly-to-neutral environment) 0.133 0.168 108 .235a 0.016 104 .244a 0.011 108 .295b 0.002 104 .220a 0.022 108 .306b 0.002 103

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . . 15

16

T. Sikora and E. Baranowska-Prokop

Table 1.7 Descriptive statistics and results of comparison of means related to measures of market performance for the “pre-COVID-19” sample depending on the environment Financial results in 2018

Sales dynamics in 2018 compared to 2017

Factor Market performance 2017–2018

Friendly environment Neutral environment Hostile environment Total Friendly environment Neutral environment Hostile environment Total Friendly environment Neutral environment Hostile environment Total

N 63 129 48 240 63 129 48 240 63 129 48 240

Mean 4.11 3.83 3.90 3.92 3.87 3.64 3.67 3.71 .243 −.082 −.100 .000

Std. deviation .743 .762 .857 .782 .871 .864 .930 .881 1.0275 .9646 1.0246 1.000

Std. error .094 .067 .124 .051 .110 .076 .134 .057 .1295 .0849 .1479 .0646

Statistica 2.996 2.681 1.530 1.463 2.445 2.482

df1 2 2 2 2 2 2

df2 107.777 150.828 108.473 157.680 107.904 160.020

Sig. .054 .072 .221 .235 .091 .087

Robust tests of equality of means Financial results in 2018 Sales dynamics in 2018 compared to 2017 Factor Market performance 2017–2018

Welch Brown-Forsythe Welch Brown-Forsythe Welch Brown-Forsythe

Source: own elaboration F distributed

a Asymptotically

Table 1.8 presents descriptive statistics and results of comparison of means related to measures of market performance for the “COVID-19” sample. In the case of “COVID-19” sample, the hypothesis 2 is not confirmed. For all three market performance measures differences between firms operating in very friendly and friendly-to-neutral environments are very small and not significant. Analysis of variance including firms’ characteristics, like firm’s size (small vs medium-sized) and relation to exports (exporters vs non-exporters) makes it possible to verify hypotheses 1 and 2 in a broader context. For the “COVID-19” sample another variable is included in analyses: the balance of COVID-19 consequences. Data are first presented for the entire “pre-COVID-19” and “COVID-19” samples to test hypotheses 1 and 2, and later they are split into sub-samples depending on the form of the environment to test hypothesis 3. For the “pre-COVID-19” entire sample neither strategic agility nor environmental hostility-friendliness show a significant relationship with financial results in

a Asymptotically

Source: own elaboration F distributed

Market performance 2019–2020

Sales dynamics in 2020 (estimated) compared to 2019

Financial results in 2020 (estimated)

Robust tests of equality of means

Market performance 2019–2020

Sales dynamics in 2020 (estimated) compared to 2019

Financial results in 2020 (estimated)

Welch Brown-Forsythe Welch Brown-Forsythe Welch Brown-Forsythe

Very friendly environment Friendly-to-neutral environment Total Very friendly environment Friendly-to-neutral environment Total Very friendly environment Friendly-to-neutral environment Total Statistica .019 .019 .463 .463 .167 .167

N 111 108 219 111 108 219 111 108 219 df1 1 1 1 1 1 1

Mean 3.37 3.35 3.36 3.04 2.94 2.99 .0273 −.0280 .000 df2 214.065 214.065 216.949 216.949 216.686 216.686

Std. deviation .904 .989 .945 1.017 .975 .995 .9967 1.007 1.000

Sig. .891 .891 .497 .497 .683 .683

Std. error .086 .095 .064 .097 .094 .067 .09460 .09697 .06757

Table 1.8 Descriptive statistics and results of comparison of means related to measures of market performance for the “pre-COVID-19” sample depending on the environment

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . . 17

18

T. Sikora and E. Baranowska-Prokop

Table 1.9 Results of analysis of variance related to sales dynamics in 2018 compared to 2017 for the “pre-COVID-19” sample (three outliers removed) Tests of between-subjects effects Dependent variable: Sales dynamics in 2018 compared to 2017 (three outliers removed) Type III sum Source df Mean square F of squares a Corrected model 14.056 12 1.171 1.644 Intercept 164.467 1 164.467 230.883 Strategic agility 5.339 1 5.339 7.495 Firm size (small vs. medium-sized) .000 1 .000 .001 Exporters vs. non-exporters 1.314 1 1.314 1.845 Environ. friendl.- hostil. .567 2 .284 .398 firmsize * expnonexp .287 1 .287 .403 firmsize * envirhost .332 2 .166 .233 expnonexp * envirhost .796 2 .398 .559 firmsize * expnonexp * envirhost .187 2 .093 .131 Error 159.564 224 .712 Total 3456.000 237 Corrected total 173.620 236

Sig. .081 .000 .007 .980 .176 .672 .526 .792 .573 .877

Source: own elaboration Squared = .081 (Adjusted R Squared = .032)

aR

2018, which is in line with close-to-zero correlations in Table 1.5 for strategic agility and makes a non-significant very weak difference in Table 1.7 for environmental friendliness-hostility. Table 1.9 shows the results of the analysis of variance related to sales dynamics in 2018 compared to 2017 for the “pre-COVID-19” sample. It appears that the only significant variable positively correlated (Table 1.5) with sales dynamics is strategic agility. Table 1.10 shows the results of analysis of variance related to the factor representing market performance 2017–2018 for the “pre-COVID-19” sample. Strategic agility also maintained a status of a variable significantly and positively related to the factor representing market performance 2017–2018 with the another variable reflecting higher market performance of exporters compared to nonexporters. Environmental hostility-friendliness has not been significantly related to any of the measures of market performance for the “pre-COVID-19” sample. For the “COVID-19” sample neither strategic agility nor environmental friendliness-hostility is significantly related to estimated financial results for 2020. Table 1.11 shows the results of analysis of variance related to sales dynamics in 2020 compared to 2019 for the “COVID-19” sample. For sales dynamics between 2020 and 2019 strategic agility shows significant relationship confirming results of correlation analysis from Table 1.6, but the role of external environment remains non-significant.

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

19

Table 1.10 Results of analysis of variance related to the factor representing market performance 2017–2018 for the “pre-COVID-19” sample (nine outliers removed) Tests of between-subjects effects Dependent variable: Factor Market performance 2017–2018 (nine outliers removed) Type III sum Source df Mean square F of squares a Corrected model 24.753 12 2.063 2.657 Intercept 3.205 1 3,205 4.128 Strategic agility 4.106 1 4.106 5.288 Firm size (small vs. medium-sized) .043 1 .043 .056 Exporters vs. non-exporters 4.210 1 4.210 5.422 Environ. friendl.- hostil. 2.816 2 1.408 1.813 firmsize * expnonexp .152 1 .152 .195 firmsize * envirhost .273 2 .137 .176 expnonexp * envirhost 2.192 2 1.096 1.411 firmsize * expnonexp * envirhost .948 2 .474 .611 Error 169.268 218 .776 Total 194.367 231 Corrected total 194.020 230

Sig. .002 .043 .022 .814 .021 .166 .659 .839 .246 .544

Source: own elaboration Squared = .128 (Adjusted R Squared = .080)

aR

As far as the factor representing market performance 2019–2020 for the “COVID-19” sample is concerned neither strategic agility nor environmental friendliness-hostility is significantly related to this variable. Summing up the results of analyses based on entire “pre-COVID-19” and “COVID-19” samples we may conclude that strategic agility was significantly and positively related to two out of three market performance measures for the “preCOVID-19” sample and one out of three measures for the “COVID-19” sample, what can be considered as a partial and weak support for hypothesis 1. External environment friendliness – hostility was not significantly related to any of the measures of market performance, thus not confirming hypothesis 2. According to hypothesis 3 the more hostile environment the stronger the positive relationships between strategic agility and market performance. For the “pre-COVID-19” sample there are nine cases resulting from three types of environment (friendly, neutral, and hostile) and three market performance measures (financial results, sales dynamics, and a measure of two-year market performance). For the “pre-COVID 19” sample and financial results in 2018, there are: – No significant relationship with strategic agility in both friendly and neutral environments (in line with close-to-zero correlations in the second and third columns of Table 1.5). – Weak positive relationship in the case of firms operating in hostile environment.

20

T. Sikora and E. Baranowska-Prokop

Table 1.11 Results of analysis of variance related to sales dynamics in 2020 compared to 2019 for the “COVID-19” sample (nine outliers removed) Tests of between-subjects effects Dependent variable: Sales dynamics in 2020 (estimated) compared to 2019 (nine outliers removed) Type III sum of Source df Mean square F Sig. squares Corrected model 35.965a 9 3.996 5.596 .000 Intercept 17.159 1 17.159 24.029 .000 Balance of COVID-19 conseq. .229 1 .229 .320 .572 Strategic agility 3.681 1 3.681 5.155 .024 Firm size (small vs. medium-sized) .889 1 .889 1.244 .266 Exporters vs. non-exporters 2.996 1 2.996 4.195 .042 Environ. friendl.- hostil. .358 1 .358 .502 .480 firmsize * expnonexp 8.434 1 8.434 11.811 .001 firmsize * envir .031 1 .031 .043 .836 expnonexp * envir 1.275 1 1.275 1.785 .183 firmsize * expnonexp * envir .250 1 .250 .350 .555 Error 142.816 200 .714 Total 1974.000 210 Corrected total 178.781 209 Source: own elaboration Squared = .201 (Adjusted R Squared = .165)

aR

Table 1.12 shows the results of analysis of variance related to financial results in 2018 for the “pre-COVID-19” sample, hostile environmental conditions. Correlation analysis in Table 1.5, column 4, shows a positive relationship at a more significant level, but when other variables are included it deteriorates to the “p”-values range between 0.05 and 0.1. It is the only case in line with hypothesis 3 for the “pre-COVID-19” sample. For the “pre-COVID 19” sample and sales dynamics in 2018 compared to 2017 there is: – No significant relationship with strategic agility in both neutral and hostile environments (in line with close-to-zero correlations in the third and fourth columns of Table 1.5). – Positive relationship in the case of firms operating in a friendly environment. Table 1.13 shows the results of analysis of variance related to sales dynamics in 2018 compared to 2017 for the “pre-COVID-19” sample, friendly environment conditions. Strategic agility is the only significant variable related to sales dynamics, confirming the positive correlation from the second column of Table 1.5. For the “pre-COVID 19” sample and factor representing market performance 2017–2018 there is:

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

21

Table 1.12 Results of analysis of variance related to financial results in 2018 for the “pre-COVID19” sample, hostile environment conditions (two outliers removed) Tests of between-subjects effectsa Dependent variable: Financial results in 2018 (two outliers removed) Type III sum Source df Mean square of squares b Corrected model 10.621 4 2.655 Intercept 18.542 1 18.542 Strategic agility 1.239 1 1.239 Firm size (small vs. medium-sized) .151 1 .151 Exporters vs. non-exporters 3.578 1 3.578 firmsize * expnonexp .060 1 .060 Error 16.357 41 .399 Total 755.000 46 Corrected total 26.978 45

F 6.656 46.478 3.106 .379 8.968 .152

Sig. .000 .000 .085 .542 .005 .699

Source: own elaboration hostility-friendliness scale = hostile environment b R Squared = .394 (Adjusted R Squared = .335) a Environment

Table 1.13 Results of analysis of variance related to sales dynamics in 2018 compared to 2017 for the “pre-COVID-19” sample, friendly environment conditions (one outlier removed) Tests of between-subjects effectsa Dependent variable: Sales dynamics in 2018 compared to 2017 (one outlier removed) Type III sum Source Df Mean square F of squares Corrected model 11.159b 4 2.790 4.605 Intercept 22.132 1 22.132 36.529 Strategic agility 9.935 1 9.935 16.399 Firm size (small vs. medium-sized) .146 1 .146 .241 Exporters vs. non-exporters .016 1 .016 .027 firmsize * expnonexp .015 1 .015 .024 Error 34.534 57 .606 Total 967.000 62 Corrected total 45.694 61

Sig. .003 .000 .000 .625 .870 .877

Source: own elaboration hostility-friendliness scale = friendly environment b R Squared = .244 (Adjusted R Squared = .191) a Environment

– No significant relationship with strategic agility in both neutral and hostile environments (in line with close-to-zero correlations in the third and fourth columns of Table 1.5). – Positive relationship in the case of firms operating in friendly environments. Table 1.14 shows the results of analysis of variance related to the factor representing market performance 2017–2018 for the “pre-COVID-19” sample, friendly environment conditions.

22

T. Sikora and E. Baranowska-Prokop

Table 1.14 Results of analysis of variance related to the factor representing market performance 2017–2018 for the “pre-COVID-19” sample, friendly environment conditions (one outlier removed) Tests of between-subjects effectsa Dependent variable: Factor market performance 2017–2018 (one outlier removed) Type III sum Source df Mean square F of squares Corrected model 9.099b 4 2.275 2.883 Intercept 3.654 1 3.654 4.631 Strategic agility 7.282 1 7.282 9.229 Firm size (small vs. medium-sized) .165 1 .165 .209 Exporters vs. nonexporters .126 1 .126 .160 firmsize * expnonexp .065 1 .065 .082 Error 43.398 55 .789 Total 57.579 60 Corrected total 52.497 59

Sig. .031 .036 .004 .649 .690 .775

Source: own elaboration hostility-friendliness scale = friendly environment b R Squared = .173 (Adjusted R Squared = .113) a Environment

Again, strategic agility is the only variable significantly related to the market performance of firms operating in a friendly external environment, confirming the positive correlation from Table 1.5 (second column). For the “pre-COVID-19” sample only one out of nine cases confirm, although weakly, hypothesis 3 about the greatest usefulness of strategic agility in the most disadvantageous environment. Positive and significant relationships in the friendly environment confirm hypothesis 1 for just this group of firms but disconfirm hypothesis 3. For the “COVID-19” sample there are six cases resulting from two types of environment (very friendly and friendly-to-neutral) and three market performance measures (financial results, sales dynamics, and a measure of two-year market performance). For the “COVID-19” sample and financial results in 2020, there are no significant relationships with strategic agility in both very friendly and friendly-to-neutral environments (positive and significant correlations in the second and third columns of Table 1.6 have been not strong enough to maintain significant relationship when other variables are included into analysis). For the “COVID-19” sample and sales dynamics in 2020 compared to 2019, there is: – No significant relationship with strategic agility in a very friendly environment (in spite of a significant correlation, after the removal of outliers, in the second column of Table 1.6). – Positive relationship in the case of firms operating in a friendly-to-neutral environment.

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

23

Table 1.15 Results of analysis of variance related to sales dynamics in 2020 compared to 2019 for the “COVID-19” sample, friendly-to-neutral environment conditions (four outliers removed) Tests of between-subjects effectsa Dependent variable: Sales dynamics in 2020 (estimated) compared to 2019 (four outliers removed) Type III sum Source df Mean square F Sig. of squares Corrected model 11.698b 5 2.340 3.166 .011 Intercept 5.727 1 5.727 7.750 .006 Balance of COVID-19 conseq. .310 1 .310 .420 .518 Strategic agility 2.448 1 2.448 3.313 .072 Firm size (small vs. medium-sized) .460 1 .460 .623 .432 Exporters vs. non-exporters .140 1 .140 .189 .664 firmsize * expnonexp 3.130 1 3.130 4.236 .042 Error 72.417 98 .739 Total 938.000 104 Corrected total 84.115 103 Source: own elaboration hostility-friendliness scale (2 categ.) = friendly-to-neutral environment b R Squared = .139 (Adjusted R Squared = .095) a Environment

Table 1.15 shows the results of analysis of variance related to sales dynamics in 2020 compared to 2019 for the “COVID-19” sample, friendly-to-neutral environment conditions. There are only two variables significantly related to sales dynamics in the “COVID-19” sample: interaction between size of firms and the fact of exporting or not and, to a lesser extent, strategic agility, confirming positive correlation from Table 1.6 (column three). For the “COVID 19” sample and factor representing market performance 2020– 2019 there is: – No significant relationship with strategic agility in very friendly environment (which confirms a non-significant correlation in the second column of Table 1.6). – Positive relationship in the case of firms operating in friendly-to-neutral environment. Table 1.16 shows the results of analysis of variance related to the factor representing market performance 2020–2021 for the “COVID-19” sample, friendlyto-neutral environment conditions. There are only two variables significantly related to market performance 2019– 2020, albeit at a “p”-values range between 0.05 and 0.1. In general, results for the “COVID-19” sample are more in line with hypothesis 3, stronger correlations are in less friendly environments, but relationships are much weaker compared to the “pre-COVID-19” sample.

24

T. Sikora and E. Baranowska-Prokop

Table 1.16 Results of analysis of variance related to the factor representing market performance 2020–2021 for the “COVID-19” sample, friendly-to-neutral environment conditions (five outliers removed) Tests of between-subjects effectsa Dependent variable: Market performance 2019–2020 (five outliers removed) Type III sum Source df Mean square of squares Corrected model 13.952b 5 2.790 Intercept 1.147 1 1.147 Balance of COVID-19 conseq. 2.396 1 2.396 Strategic agility 2.291 1 2.291 Firm size (small vs. medium-sized) .490 1 .490 Exporters vs. non-exporters 1.171 1 1.171 firmsize * expnonexp .319 1 .319 Error 72.879 97 .751 Total 87.310 103 Corrected total 86.831 102

F 3.714 1.526 3.189 3.050 .653 1.558 .425

Sig. .004 .220 .077 .084 .421 .215 .516

Source: own elaboration hostility-friendliness scale (2 categ.) = friendly-to-neutral environment b R Squared = .161 (Adjusted R Squared = .117) a Environment

1.4 Conclusions Results of data analysis show that, firstly, strategic agility was significantly and positively related to two out of three market performance measures for the “preCOVID-19” sample and to one out of three measures for the “COVID-19” sample, secondly, external environment friendliness – hostility was not significantly related to any of the measures of market performance, however, it played a moderating role between strategic agility and market performance. This hypothesized moderating role (hypothesis 3) implied that the more hostile the external environment the stronger the positive relationship between strategic agility and market performance. Findings for the “pre-COVID-19” sample are against this hypothesis, because most of positive and significant relationships between strategic agility and market performance measures have been found in the case of companies operating in a friendly environment, but not in neutral or hostile environments (however, a positive relationship between financial results in 2018 and strategic agility for firms operating in hostile environment confirmed this hypothesis). Results for the “COVID-19” sample seem to support this hypothesis, because positive and significant relationships between strategic agility and most of the market performance measures have been found for the companies operating in a friendly-to-neutral environment, and not for those doing business in a very friendly environment.

1 Strategic Agility and Success Perception of Polish SMEs: An Alternative. . .

25

In general, strategic agility is in all but one case either positively related to market performance measures or positive relationships are not significant which means that, in the absence of significant negative relationships, being strategically agile either helps or – “at worst cases” – causes no harm to enterprises’ performance. It is worth combining in one concise study the effects of the two operationalizations of strategic agility we offered: the one based on mono- versus multi-strategy and the one based on a dimension from the EMICO scale. Research designs should also include more in-depth analyses of more and less strategically agile companies inside particular branches of industry and services, so that periodical crises or periods of prosperity in particular branches do not interfere with the effects of agility.

Bibliography Adian, I., Doumbia, D., Gregory, N., Ragoussis, A., Reddy, A., & Timmis, J. (2020). Small and medium enterprises in the pandemic: Impact, responses and the role of development finance. Policy Research Working Paper No. 9414. World Bank, Washington, DC. Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: What are they? Strategic Management Journal, 21(10–11), 1105–1122. Fiore, A. M., Niehm, L. S., Hurst, J. L., Son, J., & Sadachar, A. (2013). Entrepreneurial marketing: Scale validation with small, independently-owned businesses. Journal of Marketing Development & Competitiveness, 7(4), 63–86. Goldman, S. L., Nagel, R. N., & Preiss, K. (1995). Agile competitors and virtual organizations: Strategies for enriching the customer. Van Nostrand Reinhold. Heakel, S. (1999). Adaptive enterprise: Creating and leading sens-and- respond organizations. Harvard Business School Press. Kumkale, I. (2016). Organization’s tool for creating competitive advantage: Strategic agility. Balkan and Near Eastern Journal of Social Sciences, 2(03), 118–124. Long, C. (2000). You don’t have a strategic plan? – Good! Consulting to Management, 11(1), 35–42. Miller, D., & Friesen, P. (1983). Strategy-making and environment: The third link. Strategic Management Journal, 4(3), 221–235. OECD. (2020). Coronavirus (COVID-19): SME policy responses. https://www.oecd.org/ coronavirus/policy-responses/coronavirus-covid-19-sme-policy-responses-04440101/#backendnotea0z6. Retrieved on May 20th, 2022. Oleksiuk, A., & Ple´sniak, A. (2022). Environment characteristics and internationalization of SMEs – Insights from polish and Finnish sample. Central European Management Journal, 30, 175–194. PARP. (2021). Raport o stanie sektora małych i s´rednich przedsi˛ebiorstw w Polsce 2021. https://www.parp.gov.pl/component/publications/publication/raport-o-stanie-sektora-malychi-srednich-przedsiebiorstw-w-polsce-2021. Retrieved on 20th May, 2022. Reed, J. (2021). Strategic agility in the SME: Use it before you lose it. Journal of Small Business Strategy, 31(3), 3. Roberts, N., & Grover, V. (2012). Investigating firm’s customer agility and firm performance: The importance of aligning sense and respond capabilities. Journal of Business Research, 65, 579– 585. Sambamurthy, V., Bharadwaj, A., & Grover, V. (2003). Shaping agility through digital options: Reconceptualising the role of information technology in contemporary firms. MIS Quarterly, 27(2), 237–263.

26

T. Sikora and E. Baranowska-Prokop

Sikora, T., & Baranowska-Prokop, E. (2021). The role of strategic agility and economic environment’s friendliness-hostility in explaining success of polish SMEs. In N. Tsounis & A. Vlachvei (Eds.), Advances in longitudinal data methods in applied economic research. Springer proceedings in business and economics (pp. 267–284). Springer. Sikora T., & Baranowska-Prokop, E. (2022, in print). Strategic agility and economic environment’s friendliness-hostility in explaining success of polish SMEs in the phase of COVID-19 pandemic. In N. Tsounis & A. Vlachvei (Eds.), Advances in quantitative economic research. Springer proceedings in business and economics (pp. 496–497). Springer. SME competitiveness outlook (2020). Executive Summary: COVID-19: The Great Lockdown and its Impact on Small Business, International Trade Centre (ITC), https://www.google.com/ url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwiOzafh3LHAhUIyosKHbj5AlAQFnoECBQQAQ&url=https%3A%2F%2Fintracen.org%2Fmedia%2Ffile %2F2467&usg=AOvVaw1OjWTMjV1FVIcAwiDwCszM. Retrieved on the May 29th, 2022. Statista. (2021). Coronavirus (COVID-19) in Poland. Statista Dossier. https://www.statista.com/ study/71486/coronavirus-covid-19-in-poland/#professional. Retrieved on the May 29th, 2022. Weber, Y., & Tarba, S. Y. (2014). Strategic agility: A State of the Art. California Management Review, 53(2), 271–286. Widjajani, & Nurjaman, R. (2020). The framework of strategic agility in small and medium enterprise. Journal of Physics: Conference Series, 1477, 052034. Yusuf, Y., Sarhadi, M., & Gunasekaran, A. (1999). Agile manufacturing: The drivers, concepts and attributes. International Journal of Production Economics, 62(1–2), 33–43. Zaheer, A., & Zaheer, S. (1997). Catching the wave: Alertness, responsiveness and market influence in global electronic networks. Management Science, 43(11), 1493–1505.

Chapter 2

R&D Cooperation of Firms and Product Market Competition: An Overview Jacek Prokop

Abstract The objective of this chapter is to provide an overview of research on the relationship between the cooperation of firms at the R&D stage and the competition on the final product market. We consider the R&D investments aimed at reducing unit manufacturing costs. Such investments generate positive externalities for industry participants. The analysis is conducted in a two-stage game with companies as players. In the first stage, the firms simultaneously decide about the R&D expenditures, and in the second stage, they meet in the final product market. Several possible scenarios for the behavior of firms in the final product market are considered: (1) the Cournot competition, (2) the Stackelberg leaderfollower scheme, and (3) the price-leadership setting. The comparison is made between the case of a duopolistic competition and the situation of a cartelized industry under various assumptions regarding the cost functions and the level of product differentiation. The role of the extent of research spillovers in this context is considered. Numerical analysis shows that a closer cooperation between rivals at the R&D stage in most of the cases strengthens the incentives to create a cartel in the final product market. Thus, significant antitrust issues emerge. Keywords R&D investments · Research externalities · Process innovation · Duopolistic competition · Industry cartelization

JEL Classification L11, L13, L41, O31, O32

J. Prokop () Department of Business Economics, SGH Warsaw School of Economics, Warsaw, Poland e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_2

27

28

J. Prokop

2.1 Introduction Research investments of firms may help reduce the manufacturing costs of the final product supplied to the customers. In this context, an important role of technological externalities has been identified, that is, the cost-saving R&D activities of one company contribute to the cost reduction of competitors. The existence of research spillovers, however, is a cause of investment coordination among firms to limit the extent of free riding (see, e.g., Martin, 2002). The R&D stage of the value chain has to be considered in the combination with the firms’ behavior in the final product market. In the case of oligopolistic market structure, several different types of competition between firms may be considered, but most of the initial literature assumed Cournot-type behavior (see, e.g., d’Aspremont & Jacquemin, 1988; Kamien et al., 1992; Suzumura, 1992; Amir, 2000). Also, it should be stressed that the firms may find it profitable to cooperate in their pursuit of profit maximization (see, e.g., Lipczynski et al., 2017: 195–215). The objective of this chapter is to provide an overview of research on the relationship between the cooperation of firms at the R&D stage and the competition on the final product market. In this context, the role of various degrees of exogenous product differentiation and research spillovers as well as various assumptions regarding the cost functions has been assessed. The relationship between the R&D investments of firms and their behavior in the final product market has constituted an important focal point of research for more than three decades. Both cooperative and noncooperative behavior of industry participants has been analyzed to better understand the market dynamics (see, e.g., Kamien & Zang, 2000; Cellini & Lambertini, 2009; Shibata, 2014). The analysis has been typically conducted in a two-stage game with the companies as players. In the first stage, the companies simultaneously choose the level of R&D investments, and in the second stage, they meet in the market for the final good. Several possible scenarios for the behavior of firms in the final product market are considered: (1) the Cournot competition, (2) the Stackelberg leader-follower scheme, and (3) the price-leadership setting. The comparison is made between the case of duopolistic competition and the situation of a cartelized industry under various assumptions regarding the cost functions and the level of product differentiation. The role of the extent of research spillovers in this context is also considered. Numerical analysis shows that closer cooperation between rivals at the R&D stage in most of cases strengthens the incentives to create a cartel in the final product market which leads to a reduction of consumer surplus and total welfare. Thus, significant antitrust issues emerge. This chapter is organized as follows. In Sect. 2.2, the basic market setup is developed. Section 2.3 describes the non-cooperative behavior of firms in a duopolistic industry of Cournot type and evaluates the incentives of market participants to create an industry cartel. In Sect. 2.4, the firms are assumed to follow the Stackelberg model of competition with the homogenous product, and their profits are compared

2 R&D Cooperation of Firms and Product Market Competition: An Overview

29

to the performance of cartel members. That is followed by the consideration of differentiated products in Sect. 2.5. Section 2.6 presents the comparison of profits in a price-leadership duopoly versus a fully cartelized industry. The last section contains the conclusions.

2.2 Market Setup We consider an industry composed of two firms, denoted 1 and 2. Firm 1 manufactures q1 units and firm 2 manufactures q2 units of a product that could be differentiated between producers. The inverse demand function for the product is given as a linear price function: pi = a − qi − sqj ,

.

(2.1)

where pi denotes the market price, qi is the volume produced by firm i, a is the demand intercept, and s is the substitutability parameter. Observe that both goods are homogenous (perfectly substitutable) when s = 1, and they are becoming more differentiated with the parameter s declining (when s = 0 each firm becomes a monopolist on its market). The cost function of each company i depends on its production level qi and on the size of R&D investments as well as on the parameters of initial efficiency (c) and of the extent of research spillovers (β):   Ci qi , xi , xj ; c, β ,

.

(2.2)

where c (c < a) is a given parameter of an initial efficiency of firm i, xi denotes the amount of research investments of firm i, and xj denotes the research investments of firm j. Parameter β (0 ≤ β ≤ 1) describes the extent of R&D spillovers, that is, the benefits for a given company obtained as a result of research undertaken by the competitor. A higher level of parameter β means that the research investments made by one firm have a stronger impact on the cost reduction of the rival firm. In further analysis, there will be two types of manufacturing cost functions considered: a linear one, and a quadratic one. The linear function will be assumed to take the form of:     Ci qi , xi , xj ; c, β = c − xi − βxj qi ,

.

(2.2a)

or   Ci qi , xi , xj ; c, β =

.

qi . c + xi + βxj

(2.2aa)

30

J. Prokop

The manufacturing cost function of the quadratic type will be given by the following formula:   Ci qi , xi , xj ; c, β =

.

qi 2 . c + xi + βxj

(2.2b)

Also, the costs of the R&D investments have to be modeled. Typically, in most of the research models discussed in this context, the costs of research investments take the form of a quadratic function: γ

.

xi2 , 2

(2.3)

where γ (γ > 0) is a given parameter, identical for both firms. It is assumed that the number of firms in the industry is constant. Each firm i must decide first about its level of R&D investments (xi ) and second, about the output of the final product (qi ). Decisions regarding research investments are assumed to be made by firms simultaneously and independently. These decisions affect the costs of manufacturing the final product offered by firms to the customers. We consider several possible schemes for the choice of the production level made by firms. Three of them capture noncooperative behavior. The first possibility is a Cournot-type of competition in the final product market. The second type of behavior is the Stackelberg leadership model. The third possibility is a priceleadership setting. Finally, there is also an opportunity for the firms to create a total industry cartel, that is, the decisions about research investments as well as about the final product output are made jointly to maximize the total profit of both firms taken together. We consider each of the above cases of competition and compares them with the performance of firms in a cartel.

2.3 The Cournot-Based Model In the case of Cournot-type competition in the final product market, the profits of firms were shown to be smaller in comparison to the fully cartelized industry. That conclusion is true no matter whether the manufacturing costs take the form of a linear or a quadratic cost function. It is illustrated by Tables 2.1 and 2.2 for the following values of parameters: a = 100, b = 1, c = 1, γ = 2, and β ∈ [0, 1]. Table 2.1 shows the equilibrium profits of firms and the total welfare under the Cournot competition and in the cartel for various levels of research externalities when the cost functions are linear, given by the Formula (2.2aa). Table 2.2 illustrates the equilibrium profits of firms and the total welfare under the Cournot competition and in the cartel for various levels of R&D spillovers when the cost functions are quadratic, described by (2.2b).

2 R&D Cooperation of Firms and Product Market Competition: An Overview Table 2.1 Cournot and cartel equilibrium profits and total welfare for linear cost functions and various levels of spillovers

β 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Cournot

Cournot

Cartel−C

31 Cartel−C

.π i

.TW

.π i

.TW

1099.37 1100.14 1100.76 1101.27 1101.69 1102.04 1102.32 1102.55 1102.73 1102.86 1102.95

4407.033 4408.746 4410.119 4411.218 4412.112 4412.838 4413.372 4413.778 4414.059 4414.186 4414.193

1237.87 1238.41 1238.90 1239.34 1239.74 1240.10 1240.43 1240.73 1241.01 1241.27 1241.51

3716.50 3718.09 3719.48 3720.72 3721.86 3722.88 3723.81 3724.66 3725.45 3726.18 3726.85

Source: own elaboration based on tables 1 and 3 in Prokop and Wi´snicki (2015) Table 2.2 Cournot and cartel equilibrium profits and total welfare for quadratic cost functions and various levels of spillovers

β

. πi

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

1001.83 1013.42 1022.71 1030.23 1036.34 1041.28 1045.22 1048.24 1050.35 1051.49 1051.45

Cournot

 .TW

. πi

 .T W

3940.19 3964.89 3983.68 3997.63 4007.41 4013.38 4015.61 4013.91 4007.73 3995.95 3976.51

1130.75 1137.07 1142.51 1147.50 1151.86 1155.77 1159.31 1162.54 1165.48 1168.19 1170.68

3346.66 3368.40 3387.37 3404.11 3419.02 3432.39 3444.47 3455.44 3465.47 3474.67 3483.15

Cournot

Cartel−C

Cartel−C

Source: own elaboration based on tables 4 and 6 in Prokop and Wi´snicki (2015)

The numerical example demonstrated that the functional form of the manufacturing costs has a qualitative impact on welfare when the firms simultaneously and independently choose the size of research investments, as well as the level of output. Under the linear production costs, the total welfare is monotonically increasing with the rising level of spillovers (β). When the production costs are given by quadratic functions, the total welfare is initially increasing with a higher size of research spillovers, but after reaching a maximum for β = 0.6, it starts declining afterward. However, when firms operate in a fully cartelized industry, the welfare effects are independent of the type of the cost functions, that is, the total welfare in both cases is monotonically increasing with the rising level of research externalities (β). It should, also, be stressed that the total welfare in the case of competition always dominates the social benefits from full cartelization of the industry for both types of cost functions and for any level of R&D spillovers.

32

J. Prokop

The fact that the profits of each firm in a fully cartelized industry are expected to be higher than their profit in a Cournot-type competition leads to the conclusion that the firms have strong incentives to enter a collusive agreement no matter the shape of the cost function and the size of research spillovers.

2.4 The Stackelberg-Based Model In the case of duopolists competing according to the Stackelberg scheme in the final product market, and when the production costs are given by a linear function, the profits of the leader are shown to be sometimes higher than the profits of firms in a fully cartelized industry. However, when the manufacturing costs take the form of a quadratic function, the profit of a firm in a fully cartelized industry is bigger than the profit of a Stackelberg duopolist, independent of the size of R&D spillovers. It is illustrated by Tables 2.3 and 2.4 for the following values of parameters: a = 100, b = 1, c = 1, γ = 200, and β ∈ [0, 1]. Table 2.3 shows the equilibrium profits of firms under the Stackelberg competition and in the cartel for various levels of research externalities when the cost functions are linear, given by the formula (2.2a). Table 2.4 illustrates the equilibrium profits of firms under the Stackelberg competition and in the cartel for various levels of R&D spillovers when the cost functions are quadratic, described by (2.2b). The numerical analysis demonstrated that the functional form of the manufacturing costs has a qualitative impact on the incentives of firms to cartelize the industry when the competition in the final product market follows the Stackelberg behavior. Under the linear manufacturing costs, when the research spillovers are relatively low (β < 0.5), the profits of the leader (the second column in Table 2.3) are higher than the firm profits in a fully cartelized industry (the last column in Table 2.3). Thus, when the level of spillovers is not too high, the Stackelberg leader will not Table 2.3 Stackelberg and cartel equilibrium profits for linear cost functions and various levels of spillovers

Leader

β

.π

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

1226.67 1227.23 1227.71 1228.09 1228.38 1228.58 1228.69 1228.70 1228.63 1228.46 1228.20

.π

Follower

609.86 610.60 611.28 611.90 612.45 612.94 613.38 613.74 614.05 614.30 614.48

Source: own calculations

Cartel−S

.π i

1226.66 1226.98 1227.33 1227.72 1228.13 1228.58 1229.06 1229.57 1230.11 1230.68 1231.28

2 R&D Cooperation of Firms and Product Market Competition: An Overview Table 2.4 Stackelberg and cartel equilibrium profits for quadratic cost functions and various levels of spillovers

β

.π ˆ

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

865.78 879.90 891.08 899.75 906.14 910.43 912.65 912.74 910.47 905.38 896.53

Leader

.π ˆ

Follower

749.66 764.19 776.60 787.34 796.75 805.09 812.58 819.46 825.94 832.37 839.27

33 Cartel−S

.π ˆi

917.09 925.74 933.92 941.66 948.99 955.93 962.49 968.72 974.63 980.24 985.58

Source: own calculations

be interested in creating a cartel. However, when the research externalities are large enough (β > 0.5), the profits of Stackelberg competitors are lower than the profits of cartel members. Thus, the cartelization of the industry can be expected only for higher levels of research spillovers. The incentives of firms to cartelize the industry are different when the production costs are given by the quadratic function. Independent of the size of R&D spillovers, the profit of a firm in a fully cartelized industry (the last column of Table 2.4) is higher than the profit of the leader (the second column of Table 2.4), as well as the profit of the follower (the third column of Table 2.4). Thus, we may expect that a larger extent of research externalities encourage the duopolists to cartelize the industry in the case of Stackelberg competition, independent of the functional form of the manufacturing costs. However, weak research spillovers may prevent cartel formation, but only in the case of the linear form of production costs.

2.5 Heterogeneous Products The analysis in the previous two sections was conducted under the assumption of homogenous products supplied by the duopolists, that is, s = 1 in the price function (1). When the final products are differentiated (s < 1), it is always better for the Stackelberg competitors to form a cartel in order to maximize profits, independent of the shape of the cost functions. Tables 2.5 and 2.6 illustrate this observation for the following values of parameters: a = 100, c = 10, γ = 20, s = 0.5, and β ∈ [0,1]. Table 2.5 demonstrates the equilibrium profits of firms under the Stackelberg competition and in the cartel for various levels of research externalities when the cost functions are linear, given by the formula (2.2a), and the final product is differentiated.

34 Table 2.5 Stackelberg and cartel equilibrium profits for linear cost functions, various levels of research externalities, and differentiated product

J. Prokop Leader

Follower

Cartel

β

.π h

.π h

.π h

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

1320.94 1326.94 1332.61 1337.94 1342.94 1347.59 1351.90 1355.85 1359.44 1362.68 1365.55

1266.65 1272.61 1278.27 1283.62 1288.67 1293.40 1297.82 1301.91 1305.67 1309.11 1312.21

1372.88 1377.79 1383.20 1389.13 1395.59 1402.60 1410.17 1418.32 1427.06 1436.42 1446.43

Source: own elaboration based on tables 1 and 3 in Prokop and Karbowski (2018) Table 2.6 Stackelberg and cartel equilibrium profits for quadratic cost functions, various levels of research spillovers, and differentiated product

β

.π ˆh

.π ˆh

.π ˆh

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

1514.62 1515.27 1515.84 1516.33 1516.74 1517.08 1517.35 1517.55 1517.69 1517.75 1517.75

1469.69 1470.33 1470.90 1471.41 1471.85 1472.25 1472.59 1472.88 1473.12 1473.31 1473.46

1564.69 1565.11 1565.55 1566.02 1566.51 1567.01 1567.53 1568.06 1568.60 1569.16 1569.72

Leader

Follower

Cartel

Source: own elaboration based on tables 10.1 and 10.3 in Prokop (2018)

Table 2.6 illustrates the equilibrium profits of firms under the Stackelberg competition and in the cartel for various levels of R&D spillovers when the cost functions are quadratic, described by (2.2b), and the final product is differentiated. The numerical analysis shows that the functional form of manufacturing costs has no impact on the incentives of firms to cartelize the industry when the competition in the final product market follows the Stackelberg behavior and the products supplied by the duopolists are differentiated. No matter the level of research externalities, the profit of a Stackelberg duopolist (columns 2 and 3 in Tables 2.5 and 2.6, respectively) is always lower than the profit of a cartel participant (last column of Tables 2.5 and 2.6, respectively) for linear as well as for the quadratic functional form of the manufacturing costs.

2 R&D Cooperation of Firms and Product Market Competition: An Overview Table 2.7 Price-leadership and cartel equilibrium profits for quadratic cost functions and various levels of research externalities

β

.π ˆ

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

639.11 642.98 646.28 649.23 651.92 654.38 656.62 658.69 660.82 663.71 669.55

P−Leader

.π ˆ

P−Follower

634.29 652.77 668.53 682.20 694.14 704.46 713.09 719.74 723.89 724.47 720.47

35 Cartel−P

.π ˆi

930.27 939.40 947.96 956.00 963.55 970.65 977.33 983.63 989.58 995.21 1000.55

Source: own elaboration based on tables 2 and 3 in Karbowski and Prokop (2018)

2.6 Price Leadership Next, we consider the price-leadership model of competition among duopolists under the assumption of homogenous products and the quadratic form of the manufacturing cost functions. Similar to other cases of duopolistic competition under the quadratic cost functions, the profits of firms following the price-leadership model are shown to be lower than the profits obtained by them in a fully cartelized industry, no matter the size of research externalities. Table 2.7 illustrates the equilibrium profits of firms under the price-leadership competition and in the cartel for the cost functions described by (2.2b) and the following values of parameters: a = 100, c = 1, γ = 150, and β ∈ [0, 1]. The numerical analysis showed that for any size of R&D spillover, the profit of a cartel member is higher than the profit of any firm in the non-cartelized industry characterized by price leadership. Thus, it can be expected that there are sufficient incentives for the firms to fully cartelize the industry.

2.7 Conclusions In this chapter, we provided an overview of research on the relationship between the cooperation of firms at the R&D stage and their competition in the final product market. On the basis of formal analyses presented here, we can conclude that in most cases the best performance of companies (in terms of profits) is achieved under collusive behavior. This result supports the hypothesis that closer cooperation in R&D between market participants strengthens the incentives for them to fully cartelize the industry.

36

J. Prokop

In particular, when the competition in the final product market is of Cournot type, the profits of firms are smaller in comparison to the fully cartelized industry for a linear as well as for a quadratic function of manufacturing costs. The same relationship is true when the duopolists compete according to the Stackelberg scheme or the price-leadership model but under the assumption of a quadratic cost function. Also, when the firms offer a differentiated product it is more profitable for the Stackelberg competitors to collude, independent of the shape of the cost functions and of the size of research externalities. However, in the case of linear production costs and homogenous products, the profit of the Stackelberg leader is shown to be higher than the profit of firms in a fully cartelized industry, for relatively low levels of R&D spillovers. Many results presented here are based on a limited numerical analysis of the basic model. Clearly, it is necessary to further analyze the robustness of the conclusions to the changes in the key parameters. Acknowledgment This research was supported by National Science Centre, Poland (grant number UMO-2017/25/B/HS4/01632).

References Amir, R. (2000). Modelling imperfectly appropriable R&D via spillovers. International Journal of Industrial Organization, 18(7), 1013–1032. Cellini, R., & Lambertini, L. (2009). Dynamic R&D with spillovers: Competition vs. cooperation. Journal of Economic Dynamics and Control, 33(3), 568–582. d’Aspremont, C., & Jacquemin, A. (1988). Cooperative and noncooperative R&D in duopoly with spillovers. American Economic Review, 78(5), 1133–1137. Kamien, M., & Zang, I. (2000). Meet me halfway: Research joint ventures and absorptive capacity. International Journal of Industrial Organization, 18, 995–1012. Kamien, M., Muller, E., & Zang, I. (1992). Research joint ventures and R&D cartels. American Economic Review, 82, 1293–1306. Karbowski, A., & Prokop, J. (2018). R&D activities of enterprises, product market leadership, and collusion. Zbornik radova Ekonomskog fakulteta u Rijeci, 36(2), 735–753. Lipczynski, J., Wilson, J. O., & Goddard, J. (2017). Industrial organization: Competition, strategy and policy. Pearson. Martin, S. (2002). Spillovers, appropriability, and R&D. Journal of Economics, 75(1), 1–32. Prokop, J. (2018). R&D activities in a differentiated goods duopoly with quadratic cost function. In N. Tsounis & A. Vlachvei (Eds.), Advances in time series data methods in applied economic research (Springer proceedings in business and economics) (pp. 135–145). https://doi.org/ 10.1007/978-3-030-02194-8_10 Prokop, J., & Karbowski, A. (2018). R&D spillovers and cartelization of industries with differentiated products. Journal of International Studies, 11(3), 44–56. https://doi.org/10.14254/20718330.2018/11-3/4 Prokop, J., & Wi´snicki, B. (2015). R&D activities in oligopoly and social welfare. International Journal of Management and Economics, 46(1), 134–146. https://doi.org/10.1515/ijme-20150025 Shibata, T. (2014). Market structure and R&D investment spillovers. Economic Modelling, 43(C), 321–329. Suzumura, K. (1992). Cooperative and Noncooperative R&D with spillovers in oligopoly. American Economic Review, 82, 1307–1320.

Chapter 3

Market Restrictions of Contracting Out the Public Service at the Municipal Level Beáta Mikušová Meriˇcková, Daniela Mališová, and Kristína Murínová

Abstract Under contracting out as the alternative service delivery arrangement, a government retains responsibility for providing a service, but it hires private firms to produce and deliver it. The theory of contracting suggests that, provided certain conditions are met, contracting out has the potential to improve efficiency without sacrificing quality, compared to direct public production. In developed countries, contracting out can sometimes improve the performance of the public sector. In countries making the transition from planned to market-based economies, the situation is much more complicated. Outcomes of contracting out public service are determined by the quality of contract management, and also by the character of the market of contracted service. The aim of this chapter is to identify the scale of local public service contracting out and market constraints affecting its application. The study examines the experience with contracting out service of collection and disposal of municipal solid waste among 195 municipalities in Slovakia by primary research. The outcome of quantitative analysis suggests contracting out is a frequently used service delivery arrangement (88.26% of municipalities use contracting out waste collection and disposal services), however, the outcomes of contracting out can be affected by the high concentration of service market measured by indexes concentration ratio (CR) and Herfindahl–Hirshman index (HHI). Keywords Alternative service delivery arrangements · Contracting out public services · Municipalities · Market concentration

B. Mikušová Meriˇcková () Faculty of Economics, Matej Bel University, Banská Bystrica, Slovakia Faculty of Economics and Administration, Pardubice University, Pardubice, Czech Republic e-mail: [email protected] D. Mališová · K. Murínová Faculty of Economics, Matej Bel University, Banská Bystrica, Slovakia © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_3

37

38

B. Mikušová Meriˇcková et al.

3.1 Introduction The relationship between the public, private and third sectors has evolved in the context of public administration reforms, particularly from the end of the last century to the present day. Since neither a bureaucratic system based on a managerial relationship between the public sector to the private sector nor a system based on a competitive relationship between the sectors has been successful in delivering public services (Pollitt & Bouckaert, 2000; Green, 2002; Warner & Hefetz, 2012), ‘new public governance’ and the partnership of public, private (including non-profit sector) and active participation of citizens in the process of service provision have become a new trend (Andersen et al., 2008; Sørensen, 2014). Reform steps to reduce public spending (especially in Western economies) and to transform the centralised public service system (especially in transition economies and newly transformed economies) in order to make service delivery more efficient have led to the demonopolisation, decentralisation and deregulation of the public service system (Walsh, 1995; Sclar, 2000; Péteri & Horváth, 2001) and the implementation of new Alternative Service Delivery Arrangements (ASDAs) as different forms of public-private partnerships in the public sector (Green, 2002; Andersen et al., 2008; Warner & Hefetz, 2012; Sørensen, 2014). In alternative service delivery systems in the public sector, the division of the functions of ‘service provider’ and ‘service producer’ changes compared to traditional/internal forms. The service provider is a public authority (also referred to as a public institution – a public administration entity), which bears the political responsibility for ensuring the provision of a service of adequate scope and quality in relation to public needs (the obligation to provide a public service is imposed on it by legislation, and the fulfilment of this obligation is linked to the economic functioning of the public organisation itself, i.e. the efficient provision of ancillary services). A public institution has several options for fulfilling this responsibility – several forms of service provision, which differ from each other in terms of the type of producer and the sources of funding for the service. Contracting out is one of the most frequently used alternative forms of providing services by ASDA in the public sector (Miranda & Andersen, 1994; Seidenstat, 1999; Wright & Nemec, 2002; Petersen et al., 2018). Contracting out a service in the public sector means that the public institution is responsible for the provision of the public service (policy decision) and the production of the public service is outsourced to a private, nonprofit, or other public organisation independent of the public institution. The advantages of contracting out public services can be seen in the greater degree of transparency in the spending of public funds, which results from the clear identification of the purposes for which the funds are used, the creation of a wider space for public scrutiny and, consequently, the growth of the accountability of elected bodies to the electorate (Bailey, 1995). However, the establishment of a system of control presupposes the existence of a corresponding legal framework that would define the principles and mechanisms for the operation of control, determine responsibilities and methods of reporting, the accounting mechanism

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

39

and the audit of public expenditure. Contracting out public services can go some way to addressing the problem of the lack of public resources for service delivery by using private sector resources. Contracting out public services also initiates the improvement of public decision-making processes through the definition of performance indicators for public organisations (Siegel, 1999; Green, 2002; Digler et al., 1997). The outcome or success of contracting out depends on the quality of contract management. These factors are linked to the contracting organisation: the degree of competition in being awarded a public contract – it can be simplistically assessed according to the procurement procedure used (Bel et al., 2014; Brown et al., 2016); ex-ante evaluation of a bidder for a public contract (Yescombe & Farquharson, 2018); a clearly defined subject of procurement (Martin, 1999; Brown et al., 2016); the extent and intensity of monitoring of external production (Marlin, 1984; Brown & Potoski, 2003; Warner & Hefetz, 2012); penalties for non-compliance with contract terms (Brown et al., 2016; Johansson et al., 2016); the contracting authority’s knowledge and experience in contract management (Yescombe & Farquharson, 2018; Placek et al., 2020); the expertise of the contracting authority on the technical parameters of the procured service (Kettl, 1993; Yescombe & Farquharson, 2018). These are factors derived from the quality of the management of the contracting out process. Based on these starting points the aim of this chapter is to identify the scale of local public service contracting out and market constraints affecting its application. This research pushes the boundaries of knowledge of the application of ASDA at the local level.

3.2 Market Restrictions on Contracting Out Local Public Services: Literature Review The interconnection of the public and private sectors can, in theory, be seen as the interconnection of two isomorphic sets that function in their own environment. For this reason, several market constraints are emerging, which contrast this interconnection. These disturbances include high transaction costs (Brown & Potoski, 2003; Johansson, 2008), hidden costs (Porcher, 2017) or an unstructured market that manifests itself in market concentration (Bel & Costas, 2006; Gradus et al., 2018). We also identify disruptions in the public sector in connection with the municipality’s additional resources (Wassenaar et al., 2010) or the municipality’s debt (Zafra-Gómez et al., 2014; Brown et al., 2016). We present a brief overview of the main research conclusions of the studies in the following text. Wassenaar et al. (2010) map the effects of the introduction of the VAT compensation fund on a sample of Dutch local governments. The aim of the fund is to create a level playing field for local governments and to eliminate the impact of the tax gap that arises between the traditional provision of local services and the alternative local service delivery. From the authors’ conclusions, we find that the

40

B. Mikušová Meriˇcková et al.

VAT compensation fund has little effect because it does not motivate Dutch local and regional authorities to provide alternative local services. Most local and regional authorities are also critical of this fund. The reason is, in addition to the institutional point of view, the fact that the market for the provision of these services is not sufficiently structured, which results in insufficient competition. Porcher (2017) draws attention to the financial implications of contracting local public services. On the basis of the evidence, it formulates the initial assumption that the private provision of a service is associated with a higher price for the service under the same conditions as in the traditional provision of a service. From a financial point of view, it is important to take into account the hidden costs that are typical of traditional service provision. Thus, in the traditional provision of local public services, we may at first glance perceive a lower price for the service, but over time, the price will increase by the cost of government debt. On the contrary, when contracting local public services, the price is fixed. The contractually determined price for the service ensures full coverage of costs without additional increases. The relationship between public debt and the rate of contracting was investigated by Brown et al. (2016). For example, in the Dutch municipalities, the authors find that high-debt municipalities have an increased will to provide local public services by contracting. Thus, municipalities at risk of financial instability and debt are more inclined to market solutions. Zafra-Gómez et al. (2014) attribute this fact to the consequences of financial stress. Fiscal stress is often caused by the effects of economic crises on public administration, or as a result of tighter budget and deficit controls. The solution to eliminate fiscal stress is to reduce the cost of providing services. For this reason, the application of alternative methods of service provision is a frequent choice. The paradigm of the effect of debt on the contracting rate is also verified in their study by Schoute et al. (2017). Based on the verification of the hypothesis by a logical model, they state that increasing the standard deviation in the financial situation of the municipality increases on average the probability of providing the service on its own or within the entities established by the municipality. Johansson (2008) explains the provision of local public services and the administrative management of their contracting in Swedish local governments. He also enriched his study by examining the impact of transaction costs (more on transaction costs in Williamson, 2014). It states that the size of transaction costs influences the choice of form of service provision. The study also finds that if vertical integration, that is, traditional forms of public service provision, dominates the municipality, performance deteriorates and service costs increase. On the contrary, municipalities that deintegrate services, that is, provide them by contracting, have lower service costs. However, contracting services, despite positive results, may not always be the key to cost-effective public service delivery. Johansson (2008) draws attention to the homogeneity of services and the need to take into account factors such as the characteristics of the local market, monopoly market positions or opportunistic political behaviour of local government administration. Based on these approaches, the authors identify market concentration as a significant market disruption in the provision of public services by contracting.

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

41

Studies (Gradus et al., 2018) increasingly point to the fact that a more concentrated market may lead to higher market prices when it comes to contracting. This can be justified (Johansson, 2008) by the fact that the provision of public services to municipalities is a specific area in which there may be weak competition and high concentrations (as evidenced by the high HHI index) due to the nature of the service. The competitive environment as a factor in the efficiency of service provision can have an impact on higher municipal expenditures or on economies of scale. According to some authors, the competitive environment is an even more important factor than the type of ownership. We agree with the results of the research (Soukopová & Vaceková, 2015) that states competitiveness has a high impact on local public service prices. Therefore, when providing a public service by contracting, it is necessary to pay more attention to the public procurement process (Horehájová et al., 2021) in order to ensure maximum participation of market players.

3.3 Impact of Market Concentration on the Provision of Local Public Services We can look at the competitive environment (Common et al., 1992) dichotomously. First, as competition for the market, where it is assessed how the market develops over time, what conditions are set for entry into the industry, or whether barriers are set (administrative or legislative regulation by the state) for the entry of new entities into the industry. Second, how competition in the market place affects the internal processes of the business. Within this view, we can monitor the volume of total production or marginal market shares of entities. Both internal and external factors are result-oriented in terms of the transformation process. However, in the case of services, it is important to take into account in particular the quality of the provision of services by individual entities. If we monitor the provision of local public services at the municipal level, it is necessary to take into account the serviceability of the territory. In terms of area service (Di Foggia & Beccarello, 2021), high market concentration may be present as the area shrinks. This factor is taken into account in its calculation of the Herfindahl–Hirshman index (HHI). The effect of market concentration on the price of the service and increasing costs is evidenced by the growing HHI index. The application of this approach in practice was the goal of the authors Gradus et al. (2018). The authors use special indices to measure the concentration of the Dutch market in the provision of a local public municipal solid waste collection service. The study was carried out in the period 2002–2014 and shows that the waste collection market is very concentrated, which increases the cost of private collection. In the monitored period, 25, later 16 entities operated on this market. In the initial period, the largest entity had a contractual relationship on the provision of waste collection services with 87 municipalities, by 2014 the share of municipalities

42

B. Mikušová Meriˇcková et al.

decreased to 37. The price for providing waste collection services increased over time, since 2006 mainly due to the introduction of a compensation fund. From the results of HHI measurements, we find that if the market had an area of 50 km, the market concentration was 4400 in 2014; if the area was 30 km, the market concentration was 6660 in 2014. In the case of expanding the territorial area to 70 km, the concentration was 3700 in 2014. Another approach to market concentration in the provision of local public services is offered by Demuth et al. (2018). This approach is based on reverse privatisation, which aims to take control of the provision of services by municipalities. The analysis of reverse privatisation was observed in the study in the period 2003–2015 in German municipalities, with the authors applying the concentration ratio index (CR) and the HHI index. From the results, we find that municipalities in response to the concentrated market more often switch to the traditional provision of local public services (insourcing). As a result, municipalities also copy the provision of services according to the model of their neighbour, that is, if a neighbouring municipality applies insourcing, it is likely that it will be used by its surrounding municipalities. The authors state that this is an induced vertical connection between municipalities. However, such behaviour may lead to a reverse privatisation spiral, which is associated with a high market share of municipalities, the exit of small private entities and an increase in the dominance of large private entities, further increasing market concentration. Demuth et al. (2018) point out that the effective outcome of such cyclicality is highly debatable.

3.4 Materials and Methods The aim of this chapter is to identify the scale of local public service contracting out and market constraints affecting its application. Based on the creation of a theoretical framework, which consists of a search of published studies (Russell & Bvuma, 2001; Wright & Nemec, 2002; Brown & Potoski, 2003; Bel & Costas, 2006; Johansson, 2008; Athias, 2013; Zafra-Gómez et al., 2014; Petersen et al., 2015; Brown et al., 2016; Bel & Rosell, 2016; Soukopová & Bakoš, 2017; Porcher, 2017; Gradus et al., 2018; Mikušová Meriˇcková, 2020; Osborne, 2020; Di Foggia & Beccarello, 2021) can be described as a significant market distortion in the alternative public service delivery arrangements by market concentration. Based on this approach, we formulate the research question as follows: RQ1: What type of collection and disposal of municipal solid waste service delivery arrangement is preferred by the surveyed municipalities? RQ2: If market concentration exists, how does it affect the prices of the contracted local public service? Understanding market concentration is an important management tool that can be used by any public or private entity interested in entering a particular market. While in the private sector, measuring market concentration is a common practice,

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

43

Table 3.1 Market concentration: Herfindahl–Hirshman Index (HHI) value (Evren et al., 2021) Concentration Low concentration Moderate concentration High concentration

Value 1000–1499 1500–2499 2500–10,000

Market type Efficient competition, part of monopolistic competition Part of monopolistic competition, loose oligopoly Tight oligopoly, dominant player

in the public sector, it is rather cautious. It is for this reason that there is a wide penetration in the alternative service delivery arrangements. However, there are several approaches (Capobianco, 2018) to measuring market concentration. The best-known methods used to analyse market structure include statistical and econometric methods, game theory and special indices. The traditional HerfindahlHirshman Index (HHI) is one of the traditional and relatively easy-to-use tools for measuring market concentration (Evren et al., 2021). The origins of this index date back to 1968, when it was used in the United States to assess horizontal mergers. The HHI was followed by the Hungarian economist J. Horvath, who designed a special index for measuring the complex concentration (CCI), which takes into account the relative size of the largest entity and the variance between all entities in the industry with respect to size. Other special indices suitable for measuring market concentration include the Hannaha-Kaya index (Bajo & Salas, 2002), which aims to provide insight into the link between concentration and inequality in the distribution of market power. Quite often, when measuring market concentration, we can also encounter the Lorenzo curve capturing the actual distribution of market shares, or the Gini index, which is used to measure the uneven distribution of market shares: n  CR = si (3.1) i=1

HHI =

n

s2 i=1 i

(3.2)

where si – market share of the market player, i – order of the market player, where i = 1, 2, 3, 4, ... n, n – number of market players. Based on the HHI value, the market is divided (Table 3.1) into low to highly concentrated. However, it is important to note (Evren et al., 2021) that the whole numerical percentage value is used to measure market share, not in decimal. For the purposes of this chapter, we will organise the data obtained on municipalities into larger units. For this procedure, we will use their territorial affiliation and regional division of the Slovak Republic into NUTS4 districts. This step will allow us to meet two important conditions. The market concentration in the provision of the selected local public service will be measured on a territorial area of 50–70 km.

44

B. Mikušová Meriˇcková et al.

On the other hand, more private entities will be represented, contracting the service to municipalities We chose the service of collection and removal of municipal solid waste as a suitable local public service, which is provided in the monitored 195 municipalities of central Slovakia. We were led to choose this local public service by foreign research (Soukopová & Malý, 2013; Soukopová & Vaceková, 2015; Pavel & Slavík, 2017; Gradus et al., 2018; Soukopová & Sládeˇcek, 2018; Bel & Sebo, 2020; Di Foggia & Beccarello, 2021), but also the fact that this service is the responsibility of all municipalities. The data we use in this chapter are data obtained by primary research, which was carried out in the form of a questionnaire survey. To verify the second research question, we focus on the cost of the contracted service, which we monitor through the cost of service per capita and the cost of service per performance indicator (one ton of waste). We used causal analysis to examine the interdependencies (causes and outcomes), that is, as private producers, they set their prices in the case of monopolies and in case the market concentration is lower. We used the comparison when comparing the results relating to individual districts, taking into account the origin of private producers (purely private companies or organisations belonging to other cities). The last part of the analysis focused on defining the research conclusions and the assumptions.

3.5 Results of Market Concentration Analysis for Contracting Out the Service of Collection and Removal of Municipal Solid Waste in the Slovak Municipalities In the first part of the evaluation of empirical results, we deal with the degree of contracting the collection and removal of municipal solid waste in the municipalities. We use the NUTS4 distribution for this observation. We present the obtained results graphically in Fig. 3.1. Based on the results, it is clear that the monitored municipalities prefer alternative provision of collection and disposal of municipal solid waste. Only 11.28% use traditional forms of providing this service. In terms of connecting the public and private sectors, we positively perceive that the other 173 municipalities use contracting for this purpose. In the traditional provision of the service of collection and removal of municipal solid waste, 45.45% of municipalities use technical services established within the municipality or city. 31.81% use the services of an independent organisation established by the municipality and 22.72% of municipalities use the services of public benefit. In terms of quality management, 64% of municipalities regularly check the performance of the service. The remaining 36% of municipalities control the performance of the service on an ongoing basis, as needed. In the event of a mistake and non-compliance with obligations, 63.63% of municipalities require redress and 13.63% of municipalities warn in writing and, in justified cases, withdraw from the contract. The other 22.72%

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

45

Forms of Public Service Delivery: a Regional Perspective Traditional Delivery of Public Service (by Public Sector Organization) Alternative Delivery of Public Service (Contracting Out)

Poudiva Bing © Microsoft, TomTom

Fig. 3.1 NUTS 4 distribution according to the form of provision of municipal solid waste collection and disposal services (n = 195)

of municipalities that provide the local public service internally stated that there was no case of error. From a managerial point of view, however, we are critical of such a vague approach. There is a need for a control mechanism in place for public administration organisations and for a protocol to be drawn up in the event of a mistake. Public sector organisations are responsible for providing public services, and in doing so they have to deal with a dynamic and turbulent environment or budgetary constraints (Felício et al., 2021). Only in this way is it possible to ensure the efficient, effective, and economical provision of public services. In the alternative provision of the service of collection and removal of municipal solid waste, we find that in terms of the method of public procurement, municipalities prefer selection on the basis of a public tender (61 municipalities). On the contrary, only one municipality uses closer competition as a public procurement method (Fig. 3.2). This method is characterised by the fact that the municipality declares it for an unlimited number of economic operators who can submit the documents required to meet the conditions for participation. The contracting authority may, on the basis of objective and non-discriminatory rules, limit the number of candidates it invites to submit a tender. However, it can limit the applicants to at least five entities. The criteria for the selection of private entities ensuring the collection and removal of municipal solid waste are determined by the municipalities. It is important to note that municipalities often take all these criteria into account. However, we asked which criteria the municipality considered to be the most fundamental. Based on this, weights were assigned. From the results, we find that 48.85% of municipalities consider the lowest price to be the most important criterion. Subsequently, 40.80% of municipalities decide on the basis of the most economically advantageous offer. From a professional point of view, we perceive economic advantage more positively than the criterion of the lowest price. Of course, we are aware of the pressure that is being put on drawing on public finances.

46

B. Mikušová Meriˇcková et al.

Used Public Procurement Methods in Municipalities 32%

35%

1% 16%

Price Offer Prime Contractor Offers Negotiated Procedure Narrower Competition Public Tender

16%

Fig. 3.2 Public procurement methods in municipalities applying contracting out (n = 174)

However, municipalities should also take into account the return and quality of this investment. Other selection criteria recorded only a low number of designations (each other criterion marked up to 10 municipalities). We consider it important that 4.02% of municipalities consider the availability of the company, long-term satisfaction, life-cycle costs or the conclusion of a contract due to a collective decision within the micro-region as selection criteria. The latter criterion is just an example of inter-municipal cooperation. Intermunicipal cooperation is becoming a widespread phenomenon (Allers & De Greef, 2017), especially intensively in small municipalities. The area of collection and disposal of municipal solid waste is one of the services with the highest local government expenditures. The shared use of the service by several municipalities reduces the cost of the local public service, ensures efficient provision and generates economies of scale. Research shows (Soukopová & Klimovský, 2016) that small municipalities can benefit from inter-municipal cooperation more than large local governments. The availability of landfills has also become another selection criterion in public procurement, and 3.45% of municipalities have indicated that selection is not possible due to a monopoly on the market. We perceive this factor as alarming due to the confirmation of the negative development of the market structure in the provision of local public services. Municipalities have stated that they often do not have a choice because only one player operating in the market will apply for public procurement. For this reason, in the following section, we move on to the interpretation of the results of measuring market concentration in municipalities on the basis of their affiliation to NUTS4 districts. When evaluating the results of market concentration according to the HHI index (Fig. 3.3), we find that we did not record a value lower than 2500 in any of the districts. The average value of the HHI set is 7260 (statistical deviation 2,15361) and is indicated by a line. Based on this result, we rule out the existence of perfect competition in the market or the possible existence of monopolistic competition. On the contrary, we recorded the presence of a natural monopoly (HHI = 10,000) in the districts of Revúca, Žiar and Hronom, Martin, Tvrdošín and Žilina. There is only one private player in most of these municipalities. For example, for the municipalities

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

HHI

10 000

10 000 9 000 8 000 7 000 6 000 5 000 4 000 3 000 2 000 1 000 -

10 000

7 222 7 551

10 000

7 222

7 654

47

10 000 10 000

7 551

6 800

5 988 5 450 4 694 3 889

5 000 4 400

DISTRICTS

Fig. 3.3 The result of market concentration in districts according to the HHI index

of Revúca and Martin, it is Brantner s.r.o. The only private producer for the city of Žilina is T + T a. s. company which was created after 1991 by the abolition of technical services of the city. Today in the Žilina Region, it provides waste collection for 365,000 inhabitants and annually processes 40,000 tons of municipal waste, from which alternative fuel for cement plants is created. HHI index in the range is reported by the districts Detva, Krupina, Vel’ký Kritíš, Dolný Kubín, Liptovský Mikuláš and Ružomberok. There are two to three market players in these districts. The dominant company is Marius Pedersen, a. s., which has been the largest provider of waste management services in Slovakia since 2003. It currently works with more than 400 municipalities, providing its services to almost 1,000,000 inhabitants and 4000 companies. HHI index in the range is reported by the districts Banská Bystrica, Brezno, Luˇcenec, Rimavská Sobota, Zvolen and Žarnovica. The lowest concentration of the market in the monitored group shows the district of Rimavská Sobota. The reason is that there are four organisations in the district providing collection and disposal of municipal solid waste. Companies that serve the area in an area of 40 km2 include Marius Pedersen, a. s., Brantner Gemer, p. r. o., Profax, MEPOS s.r.o Based on the values of the HHI index, we can clearly confirm that the market for the provision of municipal solid waste collection and removal services in municipalities in central Slovakia is highly concentrated and characterised by a monopoly or oligopoly with a dominant player. The following table (Table 3.2) shows the resulting values of the indicator’s cost per capita and cost per tonne of waste. For comparison, the table also contains the individual HHI and CR indexes. The lowest values of the first indicator of costs per capita were recorded in the district of Tvrdošín at the level of 14.32 EUR, which is also a district with a monopoly. It is here that we observe the paradox that higher market concentration is not reflected in higher service prices. The reason for this situation may be the appropriate setting of public procurement, price capping, or supervision of

48

B. Mikušová Meriˇcková et al.

Table 3.2 Cost per capita and cost per tonne of waste in the monitored districts Districts Banská Bystrica Brezno Detva Krupina Luˇcenec Revúca Rimavská Sobota Vel’ký Krtíš Zvolen Žarnovica Žiar nad Hronom Dolný Kubín Liptovský Mikuláš Martin Ružomberok Tvrdošín Žilina

Cost per capita (in EUR) 26.57 23.62 26.07 23.40 22.25 24.41 24.27 23.42 23.89 26.78 22.74 18.96 23.32 26.34 22.72 14.32 24.09

Cost per tonne of waste (in EUR) 144.84 119.76 202.12 160.41 123.05 178.81 161.10 148.98 192.89 127.03 130.14 168.04 121.50 152.02 102.42 137.27 127.52

CR 77.38 62.36 84.98 86.90 73.82 100.00 68.51 82.46 66.33 70.71 100.00 84.98 87.49 100.00 86.90 100.00 100.00

HHI 5987.65 3888.89 7222.22 7551.02 5450.00 10000.00 4693.88 6800.00 4400.00 5000.00 10000.00 7222.22 7654.32 10000.00 7551.02 10000.00 10000.00

compliance with the conditions by the founder. Other low values of the first indicator were measured in the districts Luˇcenec, Dolný Kubín, Ružomberok and Žiar nad Hronom, in which a monopoly was also recorded. In the district of Luˇcenec, two companies competing among the largest players in central Slovakia compete, and therefore we assume that their efforts to gain the upper hand led to lower prices for the district. The indicator reached the highest values in the districts of Banská Bystrica, Detva, Žarnovica and Martin. Only in the last district was the monopoly of Brantner s.r.o. In the remaining three districts, Brantner s.r.o. or Marius Pedersen a.s. Based on the values found, we observe that the highest prices were recorded in districts where the service was provided by purely private players. The negative impact of market concentration on the price of the service is thus confirmed. Looking at the lowest cost value per 1 ton of waste in the district of Ružomberok with the amount of 102.42 EUR. In addition, low values were recorded in the districts of Luˇcenec, Liptovský Mikuláš and Brezno. However, Brezno is the district with the lowest HHI value and, as in the Luˇcenec district, the largest part of the market is shared by two private players – Profax and Brantner s.r.o. The highest values of the indicator costs per tonne of waste were recorded in the district of Detva, which also had high values for the first indicator. In the district of Zvolen, the market is dominated by the private organisation Marius Pedersen a.s. In the case of central Slovakia, we perceive as a problem area that, despite the different HHI values, all the districts examined are highly concentrated and therefore it is difficult to estimate how the situation would change in less concentrated markets. Likewise, the market for municipal solid waste collection and disposal services has

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

49

large players. Thus, there is no room for new players, or this space is only in the district (local) market, which can lead to demotivation and price reduction at the expense of service quality.

3.6 Discussion and Conclusion Alternative service delivery arrangements in the form of service contracting out are an ideal way for small municipalities with low economic performance to provide local public services. In this context, it can be assumed that contracting services will bring cost savings together with a higher quality of provided services compared to traditional forms of service provision. The impact of market conditions in the public sector environment needs to be further explored. The aim of this chapter is to identify the scale of local public service contracting out and market constraints affecting its application. The contribution builds on current studies, which are increasingly trying to capture and analyse the market failures that affect the concept of alternative service delivery arrangements in the public sector. In this chapter, we verify two research questions. Answer to the first research question RQ1: What type of collection and disposal of municipal solid waste service delivery arrangement is preferred by the surveyed municipalities? Is that in 174 surveyed municipalities the contracting out the collection of municipal solid waste? At the same time, the most frequently used form of public procurement is public tender, while the main selection criterion is the lowest price. However, Petersen et al. (2019) are critical of such a selection of criteria, emphasising the need to take into account the specification of each tenderer. For 40.80% of municipalities, the selection criterion is the most economically advantageous offer. We believe that this criterion is set correctly because it takes into account long-term financial possibilities and transaction costs. Decisions motivated by transaction costs may seem uneconomical at first glance because transaction costs increase the price without directly increasing the volume or quality of service, however, transaction costs mitigate risk and ensure that contracting the service brings value (Williamson, 2014). In the article, we also reveal that in Slovak municipalities, decision-making on a private entity providing the service of collection and removal of municipal solid waste within the framework of inter-municipal cooperation – the micro-region – is outlined. The creation of intermunicipal cooperation can be beneficial, especially for small municipalities, for which the provision of collection and removal of municipal solid waste is an expense that places an excessive burden on the local budget. Reducing the price while maintaining the quality of service by creating a shared service is therefore more than welcome in these municipalities (Soukopová & Klimovský, 2016; Allers & De Greef, 2017). The second research question is: RQ2: If market concentration exists, how does it affect the prices of the contracted local public service? To answer this question, we first analysed the degree of market concentration using the concentration ratio (CR) and Herfindahl–Hirshman index (HHI). We find that the market for the provision of

50

B. Mikušová Meriˇcková et al.

municipal solid waste collection and removal services in the surveyed municipalities is a highly concentrated market and characterised by a monopoly or oligopoly with a dominant player. The analysis also briefly introduced the dominant players in the market for the collection and removal of municipal solid waste. We state that the result is in line with the results of foreign studies, which often pointed to the high concentration of the market in the provision of these services. Our research confirms the findings of the study (Di Foggia & Beccarello, 2021), where the authors found that in only 4 regions out of 20 were the HHI index values lower than 2000. 174 municipalities in 17 districts report that the value of the HHI index is higher (the average of the set is at the level of 7260). In terms of examining the impact of market concentration on price, we find that the highest prices were recorded in districts where the service was provided by purely private companies that were not established in connection with only one municipality (local player). For strategic planning, it is therefore necessary to identify whether the producer is a regional or local player. These results are in line with foreign studies, which in several cases have found that in a highly concentrated market – especially in the case of a monopoly – higher prices for service may occur (Di Foggia & Beccarello, 2021; Pavel & Slavík, 2017; Soukopová & Vaceková, 2015). We agree with the results of research by Gradus et al. (2018) that if the market is very concentrated, the cost of private collection increases. However, we add that if the municipality wants to contract this service, it often has no choice of other subjects, only large players and therefore agrees with them the proposed price. The authors Soukopová and Vaceková (2015) state that competitiveness has a high impact on the prices of collection and removal of municipal solid waste. Therefore, when providing a local public service by contracting out, it is necessary to pay more attention to the criteria of public procurement (Horehájová et al., 2021), in order to ensure maximum participation of competitors. We recommend that municipalities analyse market concentration in advance when considering the possibility of providing local public services by contracting out. Acknowledgement This chapter was prepared within the project VEGA No. 1/0029/21 Management of contracting services in the public sector.

References Allers, A. M., & De Greef, A. J. (2017). Intermunicipal cooperation, public spending and service levels. Local Government Studies, 44(1), 127–150. https://doi.org/10.1080/ 03003930.2017.1380630 Andersen, B., Spjelkavik, I., & Henriksen, B. (2008). Benchmarking applications in public sector principal-agent relationships. International Journal of Process Management and Benchmarking, 15(6), 723–741. https://doi.org/10.1108/14635770810915913 Athias, L. (2013). Local public-services provision under public–private partnerships: contractual design and contracting parties incentives. Local Government Studies, 39(3), 312–331. https:// doi.org/10.1080/03003930.2013.782859

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

51

Bajo, O., & Salas, R. (2002). Inequality foundations of concentration measures: An application to the Hannah-Kay indices. Spanish Economic Review, 4(4), 311–316. Bailey, S. J. (1995). Public sector economics. Macmillan. Bel, G., & Costas, A. (2006). Do public sector reforms get rusty? Local privatization in Spain. The Journal of Policy Reform, 9(1), 1–24. Bel, G., & Rosell, J. (2016). Public and private production in a mixed delivery system: Regulation, competition and costs. Journal of Policy Analysis and Management, 35(3), 533–558. https:// doi.org/10.1002/pam.21906 Bel, G., & Sebo, M. (2020). Introducing and enhancing competition to improve solid waste management in Barcelona. Institute de Recerca en Economia Aplicada Regional i Publica., 4, 1–35. Bel, G., Fageda, X., & Mur, M. (2014). Does cooperation reduce service delivery costs? Evidence from residential solid waste services. Journal of Public Administration Research and Theory, 24(1), 85–107. https://doi.org/10.1093/jopart/mus059 Brown, T. L., & Potoski, M. (2003). Contract-management capacity in municipal and county governments. Public Administration Review, 63, 153–164. Brown, T. L., Potoski, M., & Slyke, D. (2016). Managing complex contracts: A theoretical approach. Journal of Public Administration Research and Theory., 26(2), 294–308. https:// doi.org/10.1093/jopart/muv004 Capobianco, A. (2018). Market concentration. OECD Databaze. https://one.oecd.org/document/ DAF/COMP/WD(2018)46/en/pdf. Accessed 14 May 2022. Common, R., Flynn, N., & Mellon, E. (1992). Managing public services: Competition and decentralization (1st ed.). Billings & Sons. Demuth, J., et al. (2018). Reverse privatization as a reaction to the competitive environment: Evidence from solid waste collection in Germany. ESTM Working Paper, 18(02), 1–38. Di Foggia, G., & Beccarello, M. (2021). Market structure of urban waste treatment and disposal: Empirical evidence from the italian industry. Sustainability, 13(13), 7412–7424. Digler, R. J., Moffet, R. R., & Struyk, L. (1997). Privatization of municipal services in America’s largest cities. Public Administration Review, 57(1), 21–26. https://doi.org/10.2307/976688 Evren, A., et al. (2021). Some dominance indices to determine market concentration. Journal of Applied Statistics, 48(13–15), 2755–2775. Felício, T., Samagaio, A., & Rodrigues, R. (2021). Adoption of management control systems and performance in public sector organizations. Journal of Business Research, 124(1), 593–602. https://doi.org/10.1016/j.jbusres.2020.10.069 Gradus, R., Schoute, M., & Dijkgraaf, E. (2018). The effects of market concentration on costs of local public services: Empirical evidence from Dutch waste collection. Local Government Studies, 44(1), 86–104. Green, J. D. (2002). Cities and privatization: Prospects for the new century. Prentice Hall. Horehájová, M., Kˇrápek, M., & Nemec, J. (2021). Konkurencia vo verejnom obstarávaní v cˇ eskom ˇ zdravotníctve a jej dopad na finálnu cenu. Ceský finanˇcní a úˇcetní cˇ asopis, 16(1), 23–35. Johansson, T. (2008). Municipal contracting out: Governance choices, misalignment and performance in Swedish local government. Financial Accountability & Management, 24(3), 243–264. https://doi.org/10.1111/j.1468-0408.2008.00452.x Johansson, T., Siverbo, S., & Camén, C. (2016). Managing cooperation, coordination, and legitimacy: Control of contracted public services. Journal of Accounting, Auditing & Finance, 29(6), 1012–1037. https://doi.org/10.1108/AAAJ-08-2014-1805 Kettl, D. F. (1993). Sharing power – Public governance and private markets. The Brookings Institution. Marlin, J. T. (1984). Contracting municipal services: A guide for purchase from the private sector. Wiley. Martin, L. L. (1999). Contracting for service delivery: Local government choices. International City/County Management Association. Mikušová Meriˇcková, B. (2020). Kontrahovanie služieb vo verejnom sektore – skúsenosti v ˇ Slovenskej a v Ceskej republike. Masaryk University.

52

B. Mikušová Meriˇcková et al.

Miranda, R., & Andersen, K. (1994). Alternative service delivery in local government, 1982–1992. ICMA. Osborne, S. P. (2020). Public service logic: Creating value for public service users, citizens, and society through public service delivery. Routledge. Pavel, J., & Slavík, J. (2017). The relationship between competition and efficiency of wastecollection services in the Czech Republic. Local Government Studies, 44(2), 275–296. Péteri, G., & Horváth, T. M. (2001). Navigation to the market: Regulation and competition in local utilities sector in central and eastern Europe. Open Society Institute. Petersen, O. H., Houlberg, K., & Christensen, L. R. (2015). Contracting out local services: A tale of technical and social services. Public Administration Review, 75(4), 560–570. https://doi.org/ 10.1111/puar.12367 Petersen, O. H., Hjelmar, U., & Vrangbaek, K. (2018). Is contracting out of public services still the great panacea? A systematic review of studies on economic and quality effects from 2000 to 2014. Social Policy and Administration, 52(1), 130–157. https://doi.org/10.1111/spol.12297 Petersen, H. O., Baekkeskov, E., Potoski, M., & Brown, L. T. (2019). Measuring and managing ex ante transaction costs in public sector contracting. Public Administration Review, 79(5), 641– 650. https://doi.org/10.1111/puar.13048 Placek, M., Nemec, J., Ochrana, F., Schmidt, M., & Pucek, M. (2020). Analysis of factors of overpricing in public procurement: A study for low-performing EU countries. International Journal of Public Administration, 43(4), 350–360. https://doi.org/10.1080/01900692.2019.1636393 Pollitt, C., & Bouckaert, G. (2000). Public management reform a comparative analysis. Oxford University Press. Porcher, S. (2017). The hidden costs’ of water provision: New evidence from the relationship between contracting-out and price in French water public services. Utilities Policy, 48, 166– 175. https://doi.org/10.1016/j.jup.2017.08.002 Russell, W. E., & Bvuma, G. D. (2001). Alternative service delivery and public service transformation in South Africa. International Journal of Public Sector Management, 14(3), 241–265. https://doi.org/10.1108/09513550110390819 Schoute, M., Budding, T., & Gradus, R. (2017). Municipalities’ choices of service delivery modes: The influence of service, political, governance, and financial characteristics. International Journal of Public Administration, 21(4), 502–532. https://doi.org/10.1080/10967494.2017.1297337 Sclar, E. D. (2000). You don’t always get what you pay for: The economics of privatization. Cornell University Press. Seidenstat, P. (1999). Contracting out government services. Praeger. Siegel, G. B. (1999). Where are we on local government service contracting? Public Management Review, 22(3), 365–388. https://doi.org/10.2307/3380709 Sørensen, P. B. (2014). Public sector economics and the need for reforms. The MIT Press. Soukopová, J., & Bakoš, E. (2017). Nekalá konkurence a její vpliv na výši výdaju obcí a nákládaní s odpady. Scientific Papers of the University of Pardubice. Series D, Faculty of Economics and Administration, 39(1), 189–199. Soukopová, J., & Klimovský, D. (2016). Intermunicipal cooperation and local cost efficiency: The case of waste management services in the Czech Republic. In D. Špalková & L. Matˇejová (Eds.), Proceedings of the 20th international conference current trends in public sector research 2016 (pp. 398–405). Masaryk University, Brno. Soukopová, J., & Malý T. (2013). Municipal expenditures efficiency with emphasis on the competitiveness and type of company: Case study on waste management expenditures in the South Moravian Region. Available at SSRN: https://ssrn.com/abstract=2260945 or http:// dx.doi.org/10.2139/ssrn.2260945 Soukopová, J., & Sládeˇcek, T. (2018). Intermunicipal cooperation in municipal waste management and its effects on cost-effectiveness. Waste Management, 2018(1), 17–25. Soukopová, J., & Vaceková, G. (2015). Competition and municipal waste management expenditure: Evidence from the Czech Republic Olomouc Region. Scientific Papers of the University of Pardubice. Series D, Faculty of Economics and Administration, 2015(35), 128–138. Walsh, K. (1995). Public services and market mechanisms. MacMillan & Co.

3 Market Restrictions of Contracting Out the Public Service at the Municipal Level

53

Warner, M. E., & Hefetz, A. (2012). Insourcing and outsourcing: The dynamics of privatization among U. S. municipalities 2002–2007. Journal of the American Planning Association., 78(3), 313–327. https://doi.org/10.1080/01944363.2012.715552 Wassenaar, M. C., Dijkgraaf, E., & Gradus, R. H. J. M. (2010). Contracting out: Dutch municipalities reject the solution for the VAT distortion. Local Government Studies, 26(5), 617–636. https://doi.org/10.1080/03003930.2010.506976 Williamson, B. (2014). Knowing public services: Cross-sector intermediaries and algorithmic governance in public sector reform. Public Policy and Administration, 29(4), 292–312. https:// doi.org/10.1177/0952076714529139 Wright, G., & Nemec, J. (2002). Public management in the central and eastern European transition: Concepts and cases. NISPAcee. Yescombe, E. R., & Farquharson, E. (2018). Public-private partnerships for infrastructure: Principles of policy & finance. Butterworth-Heinemann. Zafra-Gómez, L. J., Pedauga, E. L., Plata-Díaz, M. A., et al. (2014). Do local authorities use NPM delivery forms to overcome problems of fiscal stress? Revista Española de Financiación y Contabilidad, 43(1), 21–46. https://doi.org/10.1080/02102412.2014.890823

Chapter 4

Description and Categorisation of Agriculture Holdings in the Region of Western Macedonia Theodoros Siogkas, Katerina Melfou, Athanasios Ragkos, and Apostolos Polymeros

Abstract The Region of Western Macedonia (RWM), Greece, is at a critical turning point in its history. The cessation of lignite use for energy purposes will drastically change its productive profile. Since agriculture is one of the main economic activities in the Region, apart from energy production, the sector is expected to play a significant role during and after the transition period, in the “post-lignite” era. The objective of this chapter is to investigate the structure of farm data in the RWM, explore the interrelationships between them and underline the correlations between the variables. The intent is to identify common factors that may explain the variation in farm data. The empirical analysis is based on a sample of 121 agricultural holdings of the Region from the Farm Accountancy Data Network (FADN) of the Greek Ministry of Rural Development & Food. A Principal Component Analysis (PCA) is applied in order to reduce the original large set of variables into a smaller set of uncorrelated components while still maintaining important information. Four factors were extracted with loadings that account for a significant variation of the total variance (87.90%), namely, net farm income, net farm value, farm size in terms of utilised land and labour employed, including family labour. The derived factors can be used to analyse the structure of the farming systems as well as farm operations in RWM and can be used to characterise systems and farms. This can be particularly important for policy purposes, as each component accounts for different aspects that should be considered when designing

T. Siogkas · K. Melfou () Department of Agriculture, Faculty of Agricultural Sciences, University of Western Macedonia, Florina, Greece e-mail: [email protected] A. Ragkos Agricultural Economics Research Institute, Hellenic Agriculture Organization “DIMITRA”, Athens, Greece A. Polymeros Ministry of Rural Development and Food of Greece, Athens, Greece © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_4

55

56

T. Siogkas et al

effective and applicable policies for the development of the farming sector and of rural areas of RWM in general. Keywords Principal Component Analysis · Factor loadings · Farm typology · Rural development

4.1 Introduction The Region of Western Macedonia (RWM) is one of the 13 Regions of Greece, situated in the North-Western part of the country, in the border with Albania and North Macedonia. RWM is a predominantly mountainous region with a harsh continental climate. The main economic activity in the Region is energy production, as RWM is endowed with significant lignite deposits thus it has been producing a significant part of national electricity for decades. Besides the energy sector, agriculture has been not only traditionally of high importance in terms of GDP but also in employment. Agricultural production in RWM is characterised by the predominance of annual arable crops (cereal, fodder crops and energy crops) but also arboriculture (peaches and apples), while the livestock sector is of importance, particularly dairy cattle and intensive and semi-extensive sheep and goat dairy farms. With this structure, the primary sector maintains a particularly important role in promoting economic activity and social cohesion, especially in mountainous and remote areas, where income and employment opportunities are limited. The prolonged economic crisis of 2010 – which affected the whole country and led to a decrease of more than 20% in national GDP – heavily affected the economic growth and social development of the Region. In fact, RWM was affected more heavily than most other Regions of Greece and was led to a significant recession. During the decade 2010–2019, the regional per capita GDP of RWM decreased by 20%, falling to 45% of the European average in 2019 (14,200 A C/resident), ranking the Region 7th across the 13 Regions of Greece and 215th among EU27 Regions (Eurostat). In the period 2011–2020, the population of RWM decreased by 6.9% and is actually estimated at 264.640 inhabitants, being the smallest non-island or offshore region of the EU. 24.5% of the population of RCM in 2019 was over 65 years old (27th Region in percentage of elderly population in the EU27) and the economically active population (15–64) was 1134 thousand (67.2%) with the weighted annual unemployment rate amounting to 24.9% (Eurostat) (highest in Greece and 4th highest in the EU). Additional intractable problems, in terms of employment and income, were created after the decisions at the European and National levels, for the transition of the economy to the post-lignite era. The effort to reduce the country’s dependence on fossil fuels for energy production and to reduce greenhouse gas emissions from industries has, by far, the most negative impact on RWM, because of the particular structure of its economy. The effects of the termination of lignite activity are expected to become more adverse in the coming years with significant estimated

4 Description and Categorisation of Agriculture Holdings in the Region. . .

57

losses in Gross Added Value and employment by A C1 billion and about 10 thousand jobs by 2029, which are expected to lead to a contraction in economic activity and increased population mobility. Thus, RWM is in a long-term and painful spiral of recession and in a dire need to redefine its productive model in the post-lignite era. This situation highlights the necessity to utilise all its comparative advantages and this brings the agri-food sector in front for the modernisation and restructuring of the economic activity model of RWM. Indeed, a specific focus on agriculture and rural development in general is a strategic choice for the Region, as the sector is already adequately developed and diversified. At this transition period, a well-informed analysis of the existing situation of the structure and economic performance of the agricultural sector of RWM is required to inform the design and implementation of appropriate policies and actions aimed at its valorisation and transformation to a dynamic and extrovert sector that can lead the economic development process, focusing on product quality and competitiveness. The objective of this study is to present an initial analysis of the characteristics of agricultural holdings in RWM. The study focuses on the economic performance of farms as well as on the organisation of land, labour and capital. The analysis is based on data and variables from a sample of 121 farms derived from the Farm Accountancy Data Network (FADN) of the Greek Ministry of Rural Development & Food. The empirical analysis combines a descriptive analysis of the data with a Principal Component Analysis (PCA), by means of which the larger set of FADN variables is reduced to a smaller set of meaningful variables (components). This chapter seeks to discuss the usability of deriving such comprehensive variables (components) that account for a larger number of variables in policy design. Factor analysis – in general – and PCA are established methodologies within multivariate analyses and empirical approaches. Gelasakis et al. (2017) classified Greek dairy goat farms into representative typologies using a random sample of 103 goat farms from 8 regions of mainland Greece and the islands. The farms were classified using a multivariable statistical approach combining PCA and cluster analysis. In particular, PCA was used in order to derive a set of comprehensive variables, since the analysis considered a large number of variables that accounted for farm operation, flock characteristics and management decisions. PCA results were used for a cluster analysis that led to useful conclusions about the development strategies of different farm types. A similar approach was followed by Abas et al. (2013), who provided a typology of dairy farms in Central Macedonia, Greece, based on their environmental management practices. Data from a survey of 123 dairy farms is used in a Categorical Principal Component Analysis, by means of which the variables associated with environmental management practices are grouped into two dimensions. A two-step cluster analysis provides three farm alternative profiles and the characteristics of farmers in each profile are estimated by a Multinomial Logit Model. The results showed that dairy farmers in Central Macedonia adopt diverse farming practices in relation to environmental protection while ensuring acceptable incomes.

58

T. Siogkas et al

Siasiou et al. (2021) constructed a farm typology of transhumant sheep and goat farms according to their management practices and producers’ characteristics that could be used by farmers or decision-makers. They obtained data from 551 Greek transhumant producers and used PCA to extract five components related to the production cost of the farms, the intensity of management, the evolution of the extensive character of the system, the type of herds and the future of the system. With multivariate techniques, a farm typology was established with three categories, characterised by lower production cost and intensification rate while having a more extensive character and a fourth intensive group, with among other things, high production cost and intensification rate. PCA and Cluster Analysis were also used to identify five types of farm households and their socio-economic characteristics in Rwanda, with the main differences identified between them being gender, age, education, risk perception, risk attitude, labour availability, land tenure and income. The respective five farm types appeared to have adopted varying types of sustainable technologies (Bidogeza et al., 2009). The following section is the methodological approach and describes data sources as well as the multivariate statistical analysis in this chapter. The next section presents the empirical results and discussion, followed by the final section with concluding remarks and suggestions for future research.

4.2 Methodological Approach 4.2.1 Data Sources and Variables The analysis is based on a sample of data from 121 agricultural holdings from RCM in 2018, which were derived from the FADN of the Greek Ministry of Rural Development and Food. The sample comprised farms of the following types, categorised based on Regulation (EC) 1242/2008, according to typology TF14: • • • • • • • •

Specialist cereals, oilseeds and protein crops General field cropping Specialist vineyards Specialist orchards – fruits Sheep, goats and other grazing livestock Specialist cattle Mixed crops Mixed crops and livestock

The analysis is based on a set of 15 quantitative variables describing financial, structural and technical aspects of holdings in RWM (Table 4.1). The first category includes variables that concern the structure of holdings, the second category has variables that determine the output and costs of holdings, the third category concerns

4 Description and Categorisation of Agriculture Holdings in the Region. . .

59

Table 4.1 Variables used in the analysis S/N 1 2 3 4 5 6 7 8 9 10 11

Code SE005_Economic size SE010_Total labour SE015_Family labour SE025_Utilised Area SE030_Rented Area SE131_Total output SE132_Output/Input SE410_Gross Farm Income SE415_Farm Net Value Added SE420_Farm Net Income

12 13 14

SE425_Farm Net Value Added/AWU SE436_Total assets SE501_Net worth SE510_Farm capital

15

SE521_Net Investment

Description Economic size of holding Total labour input of holding Refers generally to unpaid labour Total utilised agricultural area of holding Utilised agricultural areas rented by the Holder Total value of output Total output/Total input Output – Intermediate consumption + Balance current subsidies & taxes Remuneration to the fixed factors of production (work, land and capital) Remuneration to fixed factors of production of the family (work, land and capital) Farm Net Value Added expressed per agricultural work unit (AWU). Fixed and current assets. Total assets – Liabilities Average value (= [opening + closing]/2) of farm capital Net investment on fixed assets

M.u A C’000 AWU AWU ha ha A C % A C A C A C A C A C A C A C A C

the variables used to determine the income of holdings and the fourth category has variables related to their development prospects. These 15 variables are chosen for the analysis because they are considered appropriate to describe in a satisfactory manner the characteristics of the agricultural holdings and the production system in RWM. The variables are described in detail below. Structure of Holding SE 005 Economic Size of Holding Economic size of holding is expressed in 1000 euro of standard output (on the basis of the Community typology). SE 010 Total Labour Total labour input of holding is expressed in annual work units which equals to fulltime person equivalents. It is expressed by the sum of family labour input and paid labour input. SE 015 Family Labour Refers to labour offered by the members of the family expressed in annual work units (AWU).

60

T. Siogkas et al

SE 025 Total Utilised Agricultural Area Total utilised agricultural area of holding does not include areas used for mushrooms, land rented for less than 1 year on an occasional basis, woodland and other farm areas (roads, ponds, nonfarmed areas, etc.). It consists of land in owner occupation, rented land and land in share-cropping (remuneration linked to output from land made available). It includes agricultural land temporarily not under cultivation for agricultural reasons or being withdrawn from production as part of agricultural policy measures. It is expressed in hectares (10,000 m2 ). SE 030 Rented Area It refers to utilised agricultural areas rented by the holder under a tenancy agreement for a period of at least 1 year (remuneration in cash or in kind). It is expressed in hectares. Output and Costs of Holding SE 131 Total Output It refers to total value of output of crops and crop products, livestock and livestock products and of other output, including that of other gainful activities of the farms. SE132_Output/Input It is calculated as the total output to costs linked to the agricultural activity of the holding. Income of Holding SE410_Gross Farm Income It is calculated as the difference between Output and Intermediate consumption plus the Balance of current subsidies and taxes. SE415_Farm Net Value Added It refers to the remuneration of the fixed factors of production (work, land and capital), whether they be external or family factors. As a result, holdings can be compared irrespective of their family/non-family nature of the factors of production employed. SE420_Farm Net Income It is the remuneration to fixed factors of production of the family (work, land and capital) and remuneration to the entrepreneur’s risks (loss/profit) in the accounting year. SE425_Farm Net Value Added/AWU Farm Net Value Added expressed per agricultural work unit (AWU). It takes into account differences in the labour force to be remunerated per holding.

4 Description and Categorisation of Agriculture Holdings in the Region. . .

61

Development prospect of holding SE436_Total assets closing valuation It is calculated as the sum of fixed assets plus current assets. Only assets in ownership are taken into account. Capital indicators are based on the value of the various assets at closing valuation. SE501_Net worth It is calculated as the difference between Total assets and Liabilities. SE510_Farm capital It is average value of farm capital except land and quotas (namely [opening + closing]/2) It is calculated as the sum of Livestock, Permanent crops, Land improvements, Buildings, Machinery, equipment and Circulating capital. Not included in the value of quotas and other prescribed rights as it cannot always be dissociated from the value of the land. It is calculated only if land capital is recorded separately from the value of buildings. SE521_Net Investment on fixed assets It is calculated as the difference between Gross Investment on fixed assets and Depreciation.

4.2.2 Methodology The analysis of the FADN variables is based on a PCA which is a mathematical process that uses an orthogonal transformation to convert a set of observations from variables that are likely correlated into a set of values of uncorrelated variables called principal components. The aim is to reduce the number of variables, by linearly transforming the original variables into a smaller set of unrelated variables, but which contain the largest percentage of the variance of the original variables (Hair et al., 1998; Jain & Shandliya, 2013). For each variable in the original data set a correlation coefficient or factor loading is estimated with each dimension (often referred to as “component loading”), which is the criterion by means of which the dimension can be identified. The uncorrelated components or dimensions are new variables that can be used in further analysis, without losing important information from the original variables (SPSS, 2002). The transformation is carried out in a way that the first principal component has the largest possible variance and each succeeding component in turn has the highest variance possible under the constraint that it be orthogonal to the preceding components (Jain & Shandliya, 2013). Among the advantages of PCA is the fact that the method has fewer restrictions compared to other methods; it does not set limits on the number of factors (components) that can be estimated and the loadings of previous factors do not change when new factors are added (Schreiber, 2021).

62

T. Siogkas et al

In order to obtain reliable results from a PCA, several prerequisites should be fulfilled. First, since PCA assumes that the correlations between variables are due solely to the existence of common factors, the Kaiser–Meyer–Olkin (KMO) statistic is used to estimate correlations among variables. In particular, KMO statistics assumes values from 0 to 1 and measures sampling adequacy for each variable in the model and the complete model (Stevens, 2002) using the following formula: 



KMO =

rj2k j =k   2   2 rj k + j = pj k j =

(4.1)

Here rjk is the correlation between the variable in question and another, pjk is the partial correlation. Second, Bartlett’s sphericity test is used in order to indicate the possibility of statistically significant correlations in the correlation matrix. Bartlett’s test examines the null hypothesis, H0 that all k population variances are equal against the alternative that at least two are different. If there are k samples with sizes ni and sample variances Si 2 then Bartlett’s test statistic is      (N − k) ln Sp2 − ki=1 (ni − 1) ln Si2 2     χ = k 1 1 1 − 1 + 3(k−1) i=1 ni −1 N −k

(4.2)

The third measure used to measure the validity of the PCA outputs is the Measure of Sampling Adequacy (MSA), which is calculated for each indicator as above and demonstrates to what extent an indicator is suitable for a factor analysis (Stevens, 2002; Tabachnick & Fidell, 2012):  MSAi = 

2 j rij

2 j rij

+



j

aij2

(4.3)

After confirming the validity of using PCA, the table of non-rotating factors is derived, this is a first indication of the number of factors that will occur. However, an orthogonal rotation of the initial table of factors with the Varimax method – based on the eigenvalues of the correlation table – is performed in order to observe which variables affect each factor. The rotation transforms the initial factors into new ones that are easier to interpret. The Kaiser–Varimax rotation maximises the sum of the variance of the squared loadings, which are the correlations between variables and factors. The result is high factor loadings for a smaller number of variables and low factor loadings for the remaining variables. Based on this Table, the Communalities between the variables are checked and their reliability is examined, with Cronbach’s reliability test (Hair et al., 2007; Jain & Shandliya, 2013; Stevens, 2002; Tabachnick & Fidell, 2012). Finally, the interpretation – characterisation of each factor (component) can be done based on

4 Description and Categorisation of Agriculture Holdings in the Region. . .

63

the component loadings of each variable in each component. In general, a value higher than 0.7 demonstrates a high correlation.

4.3 Results Table 4.2 presents the descriptive statistics of the 121 farms in the sample. The average farm utilises 22.64 ha of arable land and 1.63 labour units. The total output is 37166.37 euro, while the output/input indicator is 1.35 implying a profitable use of resources. Correlation testing between the variables to determine whether or not factor analysis is appropriate yielded the results in Tables 4.3 and 4.4. In particular, the Kaiser–Meyer–Olkin (KMO) statistic for sampling adequacy was 0.782, so the sample size is satisfactory and according to the theory allows proceeding with the factor analysis. In addition, the Bartlett sphericity test was statistically significant so there are at least two variables in correlation. The MSA of each variable separately was greater than 0.5 therefore the variables are suitable for use in the analysis. Additionally, the Communalities between the variables have been tested and found greater than 0.5 so they all contribute significantly to the measurement of the underlying factors. Table 4.5 indicates that four factors have eigenvalues greater than 1 and explain 87.91% of the total variance. The results of the orthogonal rotation are shown in Table 4.6. If the loading of a variable on a factor is greater than 0.7 it explains 50% of the variance of a variable in the factor. Such loadings are considered to be very high.

Table 4.2 Descriptive statistics SE005_Economic size SE010_Total labour SE015_Family labour SE025_Utilised Area SE030_Rented Area SE131_Total output SE132_Output/Input SE410_Gross Farm Income SE415_Farm Net Value Added SE420_Farm Net Income SE425_Farm Net Value Added/AWU SE436_Total assets SE501_Net worth SE510_Farm capital SE521_Net Investment

Mean 37.45 1.63 1.21 22.64 18.55 37166.37 1.3548 28721.47 23228.99 17662.04 14874.43 100759.83 100317.56 76875.45 −4874.86

Std. deviation 33.770 .999 .589 24.729 21.973 30487.619 .87091 25907.069 22880.296 20503.165 13537.375 87958.428 87325.714 72371.517 8219.443

64

T. Siogkas et al

Table 4.3 KMO and Bartlett’s test Kaiser–Meyer–Olkin measure of sampling adequacy Approx. chi-square Bartlett’s test of sphericity df Sig.

.782 3463.789 105 .000

Table 4.4 Communalities and MSA Variables SE005_Economic size SE010_Total labour SE015_Family labour SE025_Utilised Area SE030_Rented Area SE131_Total output SE132_Output/Input SE410_Gross Farm Income SE415_Farm Net Value Added SE420_Farm Net Income SE425_Farm Net Value Added/AWU SE436_Total assets SE501_Net worth SE510_Farm capital SE521_Net Investment

Communalities Extraction Initial 1.000 .815 .923 1.000 .767 1.000 .965 1.000 .952 1.000 1.000 .933 .681 1.000 .957 1.000 .950 1.000 1.000 .936 .857 1.000 .934 1.000 .932 1.000 .898 1.000 1.000 .686

MSA .913 .662 .563 .719 .703 .928 .829 .854 .721 .691 .811 .785 .781 .912 .860

Extraction method: Principal Component Analysis

Finally, the reliability as well as the validity of each factor, that is, the ability of the scale to measure a specific concept – dimension, has been examined with Cronbach’s reliability test and the results are shown in Table 4.7.

4.3.1 Characterisation of Factors Based on the results of the PCA, the four components are characterised as follows Factor 1 – Net income of the holding The factor explains 27.37% of the total variance and includes six variables with high loadings. All six variables account for different aspects of farm economic performance. More specifically, it shows high positive loadings in variables related to the income of holdings which are “Farm net income,” “Farm Net Value Added,” “Farm Net Value Added/AWU,” “Gross Farm Income” and “Total output.” Furthermore, it shows lower positive loading in the variable related to the efficiency of the holding which is “Output/Input.” It is worth noting that the efficiency in production as depicted by the output/input ratio has a

Component 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Initial eigenvalues % of variance Total 7.377 49.181 21.162 3.174 1.536 10.240 7.325 1.099 .561 3.743 2.966 .445 2.000 .300 1.211 .182 .145 .969 .104 .692 .329 .049 .081 .012 .073 .011 .004 .026 .002 .000

Cumulative % 49.181 70.343 80.583 87.908 91.651 94.617 96.618 97.829 98.797 99.489 99.819 99.899 99.972 99.998 100.000

Extraction sums of squared loadings Total % of variance Cumulative % 7.377 49.181 49.181 3.174 21.162 70.343 1.536 10.240 80.583 1.099 7.325 87.908

Table 4.5 Factors extracted from the PCA and total variance explained Rotation sums of squared loadings Total % of variance Cumulative % 4.105 27.366 27.366 3.780 25.201 52.568 2.887 19.246 71.814 2.414 16.095 87.908

4 Description and Categorisation of Agriculture Holdings in the Region. . . 65

66

T. Siogkas et al

Table 4.6 Rotated component matrix

SE420_Farm Net Income SE415_Farm Net Value Added SE425_Farm Net Value Added/AWU SE410_Gross Farm Income SE132_Output/Input SE131_Total output SE501_Net worth SE436_Total assets SE510_Farm capital SE521_Net Investment SE030_Rented Area SE025_Utilised Area SE005_Economic size SE010_Total labour SE015_Family labour

Componenta 1 2 .918 .161 .247 .867 .051 .850 .412 .794 .748 −.293 .445 .675 .876 .159 .873 .155 .841 .092 −.035 −.802 .234 −.018 .323 .034 .308 .188 .273 .207 .043 .171

3

4

.004 .153 .140 .188 −.181 .158 .337 .352 .370 −.092 .946 .925 .762 .136 .138

.257 .339 −.334 .348 −.053 .504 .161 .158 .210 −.184 .052 .065 .322 .887 .846

Extraction method: Principal Component Analysis Rotation method: Varimax with Kaiser normalisation a Rotation converged in 5 iterations Table 4.7 Cronbach’s alpha test

Factors Factor 1 Factor 2 Factor 3 Factor 4

Cronbach’s alpha .903 .847 .909 .793

No of items 6 4 3 2

positive loading yet somewhat smaller (0.67). This may be an indication of a less efficient use of the factors of production at farm level in RWM. Factor 2 – Net value of the holding and investment The factor explains 25.20% of the total variance and includes four variables which account for the current financial situation and the development prospect of the farms. High positive loadings appear in the variables “Net worth,” “Total assets,” “Farm capital” and high negative loadings in the variable “Net Investment.” Factor 3 – Farm size The factor comprises three variables which account for the cultivated area of farms as well as their economic size and explains 19.25% of the total variation. It appears to have high positive loadings in the variables “Rented Area” and “Utilised Area,” and lower positive loadings in the variable “Economic size.” Factor 4 – Total labour of the holding The factor explains 16.09% of the total variance and includes two variables that account for labour used in farms. It appears to have high positive loadings in the variables “Total labour” and “Family labour.”

4 Description and Categorisation of Agriculture Holdings in the Region. . .

67

4.4 Conclusion RWM in Greece is undergoing an “aggressive” transformation of its productive base due to the decision to cease without delay the use of lignite for energy production, which has brought to the fore a debate on its future prospects and the possible role of the agricultural sector. In this context, this chapter examined the main characteristics of farms in RWM using FADN data and derived four sets of variables with common characteristics that can describe different aspects of the performance and prospects of a large variety of farms. In particular, the first factor “net income of the holding” reflects the productive capacity and the level of income that the farm holdings in the sample can achieve. The second factor “net value of the holding and investment” describes the financial state of holdings and their positive prospects for the future, assuming they will proceed with the required investments and technological innovations adoption to become more competitive. The third factor “Farm size” accounts for the cultivated area as well as farm outputs and explains how size can be an asset for farm productivity. Finally, the fourth factor “Total labour of the holding” accounts for variables that explain how labour is organised in farms and how they are dependent on hired labour or if they maintain a predominantly family character. Multivariate statistical techniques like PCA may be used in a broad array of circumstances related to farm typology definition. Therefore, the next step in this analysis is to use these results in order to detect homogeneous farm types (clysters) that share common characteristics. A detailed examination of such clusters of farms will account for the practices, as well as will reveal strengths and weaknesses and will highlight domains that require policy interventions. Therefore, this methodological approach can be very helpful for effective policy design – “tailor-made” to the specific characteristics, needs and challenges of each type – as well as the implementation of specific policy measures. Since uniform approaches for all farm types can be proven ineffective, a more focused approach in policy designed – as part, however, of a larger strategic plan – can be proven beneficial not only for the farming sector of RWM but for the whole economy of the Region, in these times of restructuring.

References Abas, Z., Ragkos, A., Mitsopoulos, I., & Theodoridis, A. (2013). The environmental profile of dairy farms in Central Macedonia (Greece). Procedia Technology, 8, 378–386. Bidogeza, J. C., Berentsen, P. B. M., De Graaff, J., et al. (2009). A typology of farm households for the Umutara Province in Rwanda. Food Security, 1, 321–335. https://doi.org/10.1007/s12571009-0029-8 Gelasakis, A. I., Rose, G., Giannakou, R., Valergakis, G. E., Theodoridis, A., Fortomaris, P., & Arsenos, G. (2017). Typology and characteristics of dairy goat production systems in Greece. Livestock Science, 197, 22–29.

68

T. Siogkas et al

Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). Macmillan. Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, L. R. (2007). Multivariate data analysis (6th ed.). Pearson Prentice Hall. Jain, P. M., & Shandliya, V. K. (2013). A survey paper on comparative study between Principal Component Analysis (PCA) and Exploratory Factor Analysis (EFA). International Journal of Computer Science and Applications, 6(2), 373–375. Siasiou, A., Karelakis, C., Galanopoulos, K., Mitsopoulos, I., & Lagka, V. (2021). Typology of management of transhumant sheep and goat farms in Greece: Proposals for the system continuity. European Journal of Agriculture and Food Sciences, 3(1), 84–89. SPSS. (2002). Categories 11.5. A software package, version 11.0. SPSS. Stevens, J. P. (2002). Applied multivariate statistics for the social sciences. Mahwah, NJ. Tabachnick, B. G., & Fidell, L. S. (2012). Using multivariate statistics (6th ed.). Harper & Row. Schreiber, J. B. (2021). Issues and recommendations for exploratory factor analysis and principal component analysis. Research in Social and Administrative Pharmacy, 17, 1004–1011.

Chapter 5

Debt and Taxes as Value-Added Factors for Multinational Enterprises: International Policy Implications Juan José Durán-Herrera and Prosper Lamothe-Fernández

Abstract The international heterogeneity of tax regimes among countries, where multinational firms operate, and the generally accepted principle that interest on debt is treated as a deductible cost for tax purposes, reduces corporate benefits, and consequently, the amount of taxes paid has consequences on the allocation of resources: the firm obtained a lower cost of capital (and potentially could invest more), the shareholders return increases (as well as executives paid), the income of Tax Authorities diminished, and a greater inequality results (given the high concentration of wealth and MNE capital ownership). A global economy needs a simpler and more transparent and equivalent corporate tax system among countries that only can be reached through international cooperation. Globalization as well as the process (evolution) of the digital economy have contributed to tax competition (About 60% of profits of MNE are the results of tax competition (IMF, World Economic Outlook, 2022)) that reduces income taxes of countries. Keywords Tax benefits of debt · Inequality · Taxes on multinational firms · Tax shield

JEL F23, F38, G32, H25, H32

5.1 Introduction The international heterogeneity of the corporate tax environment creates tax incentives to adopt aggressive tax planning worldwide by multinational enterprises (MNE). The differences among national tax systems and the firm’s objective

J. J. Durán-Herrera () · P. Lamothe-Fernández Universidad Autónoma de Madrid, Madrid, Spain e-mail: [email protected]; [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_5

69

70

J. J. Durán-Herrera and P. Lamothe-Fernández

towards maximization of shareholder wealth and cash flows generated are the conditions towards aggressive tax planning. Globally, multinational enterprises in 2017 had shifted about 40% of its profits to tax heavens, equivalent according to OCDE and other institutions such as Tax Justice Network, and Global Alliance for Tax Justice, to 200,000 million dollars. One dollar gained by a tax heaven implies a loss of 5 dollars for the country eluded (Zucman, 2018).1 Recent research by Saez and Zuckman (2019) showed that European Union tax havens – Luxembourg, Belgium, Cyprus, the Netherlands, Malta and Ireland – are responsible for around 90% of all tax avoidance within the union.2 The European Union has made feasible tax-reform proposals and the Organization for Economic Cooperation and Development has developed proposals for corporate-tax reforms for the world, through its ‘base-erosion and profit-shifting’ process. OECD reached a global agreement (October 2021) with 136 countries that accounts for more than 90% of the world GDP to assume a 15% tax on multinational’s profits from 2023 on. This will apply to MNE with sales above 20,000 million dollars and with a return over 10%.3 Debt financing by MNEs has important tax implications; firms are allowed partially not only to avoid paying taxes to the countries where they operate but also to be a determinant factor of the global financial structure of MNE and its subsidiaries. This is due to the accepted principle that interest on debt paid to lenders is considered as financial expenses to be deducted from business profits before tax. The firm pays to the providers of funds (to finance the assets of the firm) in the following ways: (i) interest to lenders (banks and bondholders); (ii) corporate tax to the Tax Authorities; (iii) the tax authorities receive less taxes, equivalent to the corporate tax rate multiplied by the debt contracted; it is an indirect subsidy from taxpayers; and (iv) the benefits available to remunerate shareholders, via dividends and capital gains increase. The tax shield is an incentive to increase the leverage ratio of firms and consequently is a determinant factor of its capital structure. It reduces the cost of capital, which may increase the propensity to invest. With tax shield the return on equity increases and contributes to a high concentration of capital worldwide: the inequality in society increases. In terms of opportunity cost, the tax shields reduce the income of Tax Authorities; this can be interpreted as an indirect subsidy by the taxpayer of countries to reduce the cost of capital of leveraged firms and thus increase the wealth of shareholders. Tax shield has efficiency and inequality implications. As we have mentioned above the tax benefit of the debt is materialized in the deduction of interest from the taxable base of corporation tax. These can be diluted

1 The

richest people elude 25% of their income using tax heavens.

2 The proportion of corporate tax forgone amounts to 15, 22, and 26% in Italy, France and Germany

respectively; Spain lost 14% of its corporate tax revenue. 3 The real evidence says that the international agreement towards harmonisation of fiscal regimens is far from getting an effective path.

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

71

in the assessment of the tax and observed that the effective tax paid is significantly less than the nominal one. That is, the marginal rate of the effective tax paid, in addition to the deduction of interest paid to the lenders, also incorporates other deductible items (amortisation of establishment expenses, goodwill, tax credits, tax deductions for certain commercial transactions and reasons of economic policy). Companies have the possibility of using a variety of items that constitute havens for tax deduction in the liquidation of corporation tax. The use of these items influences the volume of debt. The tax department of a corporation can be seen as a profit centre that optimizes tax disparities between countries. In the following sections of this article, we will discuss the theoretical foundations of the value of the company and its capital structure. We will analyze the levels of global debt in the world and of multinational companies. We will end by considering the economic and political implications of the tax deduction for interest on the debt.

5.2 Determinant Factors of Debt Financing Since the 2008 financial crisis, the global debt has grown significantly: from 97 billion dollars in 2007 to 169 billion dollars in 2017 (184 billion according to the FMI, about 225% of GDP; about 320% nowadays). Corporate debt went from 37 billion to 66 billion; with similar figures in the case of government debt) also according to the balance sheet of the global economy, debt has grown most significantly relative to GDP, by 178 percentage points in the 50-year time frame (1970–2020).On average world, debt represents about 86,000 dollars per capita, 2.5 times the income per capita. The three main borrowers in the world are the USA, China and Japan, which accounts for about half of the total debt; a greater participation compared to their participation in world production, The debt problem is not associated with the developing countries as it was in the past; it influences the policies of the advance countries. The debt of multinational firms is over two-thirds of total world debt (McKinsey country debt database). The provision of better information to the market is associated with the improvement of access to capital markets and to the cost of debt (Ferrer et al., 2019). The primary determinant of the debt source is the credit quality of the issuer. In this respect ‘firms with the highest quality borrow from public sources (capital markets), firms with medium credit quality borrow from banks, and firms with the lowest credit quality borrow from non-bank private lenders’ (Denis & Mihov, 2003). The cost of borrowing from the bank may be lower than from capital markets because there is less information asymmetry. Bank debt may be a positive sign to the financial markets (James, 1987). However, after the crisis of 2008 many large corporations have shifted towards bond financing; commercial bank lending has decreased due to the banking crisis. About 20% of total global corporate debt is in the form of bonds, nearly double the share in 2007.

72

J. J. Durán-Herrera and P. Lamothe-Fernández

Fig. 5.1 Historic highs. Note: *Includes households, non-financial corporations, and government debt; excludes financial sector debt. Estimated from the bottom up using data for 43 countries from the Bank for International Settlements (BIS) and data for eight countries from the McKinsey’s Country Debt Database. Figures may not add to 100% due to rounding. (Source: Bank for International Settlements (BIS), McKinsey Country Debt Database y McKinsey Global Institute Analysis)

Firms that have a smaller information gap with the market are therefore more likely to borrow from capital markets because the benefits of bank financing are smaller, and there is a cost associated with this monitoring (Denis & Mihov, 2003). Banks are able to extract rents from the firm because of its privilege relations; this can be considered as an incentive to diversify banks dependence. The top three borrowers in the world are the USA, China and Japan, which up to now account for about half of total debt; a higher share compared to its share in world production. The debt problem is not associated with developing countries as it was in the past; the countries that have been borrowing the most in recent years are the advanced countries. Corporate debt exceeds 39% of world debt (see Fig. 5.1). It is also true that the reduction of interest rates to very low or even negative levels has influenced the greater propensity to borrow on the part of companies. Thus, in Fig. 5.2 we can see how the costs of bank financing in the euro zone have been reduced by 60% in recent years.

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

73

Fig. 5.2 Interest rates of bank loans granted to non-financial corporations and households. (Source: European Central Bank (ECB))

The capital structure of MNE is influenced by country risk and foreign exchange risk, the industrial and geographical diversification, agency cost and the nature of assets (the intangibility of assets) (Burgman, 1996; Lee & Kwok, 1988; Doukas & Pantzalis, 2003; Singh & Hodder, 2000). The main multinationals showed a very high ratio of debt, which can be appreciated in the tables included in the annex. At the aggregate level of the Fortune 500 by country of origin of the MNE, we observe the high level of debt in the 3 years considered (2018–2020) as can be seen in the three tables. In some cases, MNE from 20 countries registers a debt ratio of over 70%: being 87% of the total for the 500 firms. In addition, during the period 2018–2020, the financial expenses represent a high proportion of earnings before taxes; there are strong differences in company’s leverage among different countries. Earnings before taxes and financial expenses have increased every year (see Annex). The non-financial corporation’s debt-to-surplus ratio can be seen as indicative of the capacity of non-financial corporations to assume the cost of interest and debt repayments with the operational cash flow generated (see Fig. 5.3). Debt is calculated according to OECD as the sum of the following liability categories: currency and deposits, debt securities, loans, insurance, pensions and standardised guarantee schemes, and other accounts payable. Gross operating surplus (GOS) is the value added generated by production activities after the deduction of compensation of employees. The sector of non-financial corporations includes all private and public enterprises that produce goods and non-financial services for the markets. If the ratio debt to operating surplus of a non-financial corporation is 3.5; this means that the debt outstanding is 3.5 times larger than the annual gross operating surplus. For example, in Canada, the ratio is more than three times higher than the ratio in Poland. Factors like the characteristics of financial systems, the tax level and the international diversification of companies could explain these differences.

74

J. J. Durán-Herrera and P. Lamothe-Fernández

Fig. 5.3 Non-financial corporations’ debt-to-surplus ratio. (Source: OECD)

5.3 Taxes and the Value of the Firm Under the perfect capital markets assumptions, the structure of the firm’s debt (short or long term, fixed or floating rates, currency of denomination) is irrelevant; then the capital structure of firms is irrelevant too (Modigliani & Miller, 1958). With full information available at no cost, all providers of capital can charge a fair price for the transaction. Under this type of environment, there are no taxes, no transaction costs and costs of financial distress or bankruptcy; investors and markets are rational; there is no asymmetry of information; objectives of managers and shareholders are compatible. Under imperfections of capital markets, relaxing the above assumptions, the financial structure of firms is relevant. Profits are a signal of the economic responsibility of firms. The net profit after corporate tax calculation implies that the firm has generated enough income to pay for all the economic factors of production (wages, energy, raw materials and intermediate goods, overhead costs and taxes). The net profit after taxes over assets (investments) is the economic return of the firm. The relevant question is what to do with the return on assets, who are the contractual owners of the firm’s net profits? The answer is to compensate the savings invested in the firm. In the first place, with zero debt, the return on equity will be exactly the return on assets and thus the value of unlevered firms will be: Vs =

X (1 − τ ) ke

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

75

where X = average annual profits before taxes t = tax rate ke = cost (return on) equity Vs = Shareholder value = E The interconnections between the real and financial economy of the firm take place through the average cost of capital: k = kd

E D + ke V V

V =E+D The value of the firm will increase by the present value of the tax savings. The lender (banks or investors in bonds) receives the interest paid by the borrower (the firm) and they will pay their own income taxes. The shareholders will receive dividends and capital gains and for that, they will pay their personal or business taxes. Under this situation, the value and the capital structure of firms are not affected directly by taxes on the firm’s payments to financial investors (shareholders, banks or bondholders). When there exist corporate taxes on profits the capital structure (debt and equity) matters; in theory, the decisions on debt financing are relevant as it is the financial capital structure of firms. In order to debt structure to matter, one or more of the following circumstances need to be real: taxes, transaction costs, agency costs and the possibility of distress costs. Debt financing can also lead to costs of financial distress correlated with the level of risk taken. The combination of tax benefits and financial distress is one of the major advantages of debt financing that it can reduce the firm’s tax bill. Factors affecting the decision about the level of debt undertaken. If the firm takes, more risk than originally anticipated by the potential lender this may lead to a higher interest rate. It is rational for firms to limit borrowing. As a result of debt financing, the decision to issue debt in local or foreign currency has to do with its tax advantages as part of the firm’s optimisation strategy. If interest on debt is firm’s expenses for tax purposes de cost of capital will be lower, k = kd (1 − τ )

E D + ke V V

76

J. J. Durán-Herrera and P. Lamothe-Fernández

and the value of the leverage firm (D > 0) will be: Vt =

X (1 − τ ) kd Dτ + = Vs + Dτ ke kd

The tax advantage to corporate borrowing may lead to an increase in the level of debt. In addition, since Vt = Vs + Dτ the incentive for firms will be to issue debt ‘as much as possible’. However, considering the potential bankruptcy and financial distress cost (.C˜ iq ) as a random variable we can admit that debt financing has to have a limit. The costs of financial distress (bankruptcy costs) or transaction costs of liquidation or reorganisation, may discourage borrowing Vl = Vs + τ D − C˜ iq (i − τ )

D=

n  iD (1 − τ ) t=1

(1 + kd )

t

+

D (1 + kd )n

where D = Nominal value of debt4 i = contractual interest rate l τ = tax rate Kd = Cost of debt taxes When the cost associated with debt is inferior to the return on assets, the financial return for shareholders will be greater than the economic return of the firm: E(Y ) = E(r) (A = E + D ) − kd D (1 − τ ) = E(r)E + D [E(r) − kd (1 − τ )] A decrease in the average cost of capital due to the tax shield generated may facilitate an increase in investment then it will influence the growth rate. A high level of debt may increase the level of risk, and this may lead to suboptimal future investment strategy (an agency cost). Then the optimal strategy of debt issued is a tradeoff between tax advantages of debt and the level of risk of debt. In the absence of taxes, the optimal strategy is to issue no risky debt (Myers, 1977). It is rational for firms to limit borrowing. The incremental tax advantage of borrowing may decline as debt increases and tax shields become less certain. In Fig. 5.4, we show graphically this discussion.

4 In this case, we suppose that the firm obtain the nominal of the loans (binds). Obviously, the market value of debt may not be identical to the face value.

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

77

Fig. 5.4 Debt, taxes and equity valuation

5.4 Statistical Analysis of the Relationship Between Debt and the Value of Firms. Does in Reality the Debt Financing Affects the Value of the Firm? As we have shown in Tables 5.2, 5.3, and 5.4 (see Annex) the 500 Fortune ranking is very highly debt financing. On average total assets measured by book values are nearly 90% financed by debt which allows us to ask the following question: does the long-term debt financing have a positive effect on the market value of multinational firms? To answer to this question, we analyse the statistical relation between the ratio price to book value, as a dependent variable, and to different variables related to the profitability of the multinational ranking of Fortune in the period 2018–2020. We exclude from the sample firms not listed in capital markets. In our analysis, we have excluded those firms with a very low or negative equity value due to purchasing own equity or having obtained losses. As a result of these circumstances, we have cancelled out 28 companies plus another 68 firms that are not listed in the capital market which led to a sample of 404 multinational firms. After testing different models, the best equation that explained the relation between the ratio price/book value and the variables showing the leverage of firms is shown in Table 5.1. In this table we show the explanatory variables of the two econometric models tested that only are different in terms of the variable indicating the level of debt; the ratio tax shield/net equity (model 1) or the ratio debt/total assets (model 2). The explicative variables of the model that offers the best econometric result are the following: (i) Financial return measured as a ratio of profit after taxes and net

Table 5.1 Regression data 2018 Regression analysis results Model 1 Model 2 Price/book value Price/book value Return on equity 18.350*** 23.120*** (0.000) (0.000) Tax shield/equity value 1.578*** – (0.000) – Sector −3.534 −7.657 (−0.393) (−0.131) Debt/total assets – 0.0001 – (−0.543) Constant −3.985** −0.196 (−0.040) (−0.934) R2 0.641 0.464 R2 adjusted 0.639 0.460 F 239.67 115.61 Note: *Significant to 90%, **Significant to 95%, ***Significant to 99% 2019 Regression analysis results Model 1 Model 2 Price/book value Price/book value Return on equity 20.222*** 1.899e+01*** (0.000) (0.000) Tax shield/equity value −0.31002*** – (0.000) – Sector 12.054 1791 (0.349) (0.192) Debt/total assets – −1.896e−05 – (0.156) Constant 0.61461 9.307e−03 (0.317) (0.989) R2 0.9422 0.9342 R2 adjusted 0.9417 0.9335 F 1648 1433 Source: Own production Note: *Significant to 90%, **Significant to 95%, ***Significant to 99% 2020 Regression analysis results Model 1 Model 2 Price/book value Price/book value Return on equity 7.3042*** 4.989*** (0.000) (0.000) Tax shield/equity value −0.27868*** – (0.000) – Sector 0.5816 8.707e−01 (0.271) (0.1279) Debt/total assets – −3.323e−05 – (0.936) (continued)

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

79

Table 5.1 (continued) Constant

1.338*** (0.00) 0.5911 0.5873 156.1

1.110*** (0.000) 0.5218 0.5173 117.80

R2 R2 adjusted F Source: Own production Note: *Significant to 90%, **Significant to 95%, ***Significant to 99%

capital. (ii) The ratio Tax shield/net capital as a result of multiplying the level of debt by effective tax rate and the accounting net capital. To account for the sector of activity, we use a dummy variable that takes 1 for those sectors with a high investment in intangibles such as technology and telecommunications and 0 for traditional sectors. With this variable, we show the fact that sectors more intensive in intellectual capital must have a greater ratio of price/book value. According to the first model, as it could be expected, the return on equity always has a positive effect on the ratio price/accounting value.5 We tested the models with an alternative return measure, the ROA (return on assets) but the new variable coefficients are not statistically significant. This is an important result because it assumes that the strong statistical relationship between equity valuation and ROE is not due to the impact of leverage on ROE. Our analysis shows that debt financing added value to the firm due to the present value of tax savings for the year 2018. In the following 2 years, the coefficient for this variable changed significantly, which implies that the fiscal savings of corporate leverage negatively influences the value of equity. These results that must be tested with more empirical studies indicate that in terms of valuation by the stock market the financial leverage does not seem to provide much value in environments with low-interest rates and also with clear differentials in favour of Offshore financial centres. Also possibly more leverage means that investors perceive greater risk in companies and they adjust the cost of capital globally offsetting the tax advantages. On the other hand, the sector of activity does not show to have statistical significance when we use return on equity and the tax shield as explanatory variables. However, when we consider other debt proxies, their estimated coefficients are not statistically significant. In summary, in the analysis carried out for large multinationals, there is evidence that leverage does not positively influence the value of a firm’s equity or that it can even have a negative effect. What is important for share value is the profitability obtained over equity invested (where the international arbitrage of debt may have important effects). 5 Price

to Book Value = Per * Expected Long term ROE.

80

J. J. Durán-Herrera and P. Lamothe-Fernández

5.5 Some Conclusions and International Policy Implications The tax advantage (tax shield) is a factor of the effective cost of debt and consequently is a determinant factor of the capital structure of firms. The value of tax shields depends on the debt policy of the company. The tax rate on profits should be financially neutral in the absence of a tax shield. The interest on debt is income for banks, and for bondholders, and consequently, they will pay income taxes. At the same time, shareholders will pay taxes for dividends and capital gains. The firm should pay corporate taxes for the profits obtained throughout its economic activity. If tax advantage (tax shield) to corporate borrowing disappears internationally, the allocation of international business should be more efficient; tax elution6 for this matter will not apply. However, through other means it is possible to minimize the tax burden, for example using transfer prices). At the same time, the fiscal basis (the net operating income) for corporate taxes will be greater and for the same tax rate the Fiscal Administration will have more recourses available for public expenses (education, healthcare and infrastructure) and the inequality could be reduced. A tax advantage to corporate borrowing is an incentive to increase the debt of firms and is a factor of inequality, given the high concentration of wealth and income worldwide and within a country level. In parallel concentration of production in a few companies continues to grow. In addition, the international fiscal optimization of MNEs is an increasing phenomenon: to assign the debt to subsidiaries locate in a high-tax country and assign profits to subsidiaries with a low-tax environment. The tax environment implies that the effective tax rate that the corporation pays in all countries is much less than the nominal rate.7 The international heterogeneity of the corporate tax environment creates tax incentives to adopt aggressive tax planning worldwide. The tax shield is an incentive to increase the leverage ratio of firms and consequently is a determinant factor of its capital structure. It reduces the cost of capital, which may increase the propensity to invest. With a tax shield, the return on equity increases and contributes to a high concentration of capital worldwide: the inequality in society increases. Empirical observations said that the effective cost of debt increases the propensity to assume high ratios of leverage that allow firms to buy their own shares which mostly increases the capital gains of the main shareholders increasing the inequality. In terms of opportunity cost, the tax shields reduce the income of Tax Authorities; this can be interpreted as an indirect subsidy by the taxpayer of countries to reduce the cost of capital of leveraged firms and thus to an increase of the wealth of shareholders. Tax shield has efficiency and inequality implications. Corporate political activity may be defined as ‘any deliberate firm action intended to influence government policy or process’. Economic policies of governments

6 The fiscal elusion in the EU is estimated to be around a billion euros per year the EUS and about 200,000 million globally. 7 The effective tax rate on MNE’s profits has been declining along time; nowadays the effective tax rates are between 4% and 8.5%.

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

81

make provisions for subsidies, price controls, entry barriers, and other interventions. Consequently, firms have the incentive to develop ‘domain management’ strategies – that is, to use governmental interventions in a way that supports their own strategic objectives. The alternative reaction of MNEs to institutional change is avoidance, adaptation, or reaction. The MNE can design entrepreneurial actions to modify, create or eliminate institutional changes as a group (business organisations). Government interventions may increase environmental uncertainty and consequently, firms have the incentive to develop management strategies to face the effects of such interventions. The elimination of tax advantages to corporate borrowing will delete a location advantage of tax heaven and offshore financial centres. In addition, an important barrier to a tendency to a fiscal harmonisation will disappear. Alternatively, and as an intermediate measure fiscal competition among countries to reach a similar rate on corporate profits will be more realistic and transparent and more efficient otherwise. Taxes in a global and digital world must be considered from an international perspective and not from a local and one-sided point of view. If this is not done, the tax collection for the corporate profit tax will be reduced, generating problems of sustainability of the welfare state as well as increasing economic inequality between individuals. Taxes in a global and digital world have to be considered from an international perspective and not from a local and unilateral point of view. The real evidence says that the international agreement towards the harmonization of tax regimes over decades is far from paving an effective path. However, The G20 countries G20 and the OECD Base Erosion and Profit Shifting (BEPS) has established what it is called BEPS PILLAR II. With the Participation of 141 jurisdictions agreement has been reached on the introduction of a global minimum tax of 15% on the foreign profits of large MNEs (revenues above 750 million euros). The implementation is planned for 2023. Pillar II will increase the corporate income tax faced by MNEs on their foreign profits. Both developed and developing economies are expected to benefit from increased revenue collection. The aims of BEPS Pillar II are: (i) to discourage MNEs from sifting profits to low-tax countries and to reduce tax competition between countries8 ; (ii) to stabilise international tax rules and reduce tax uncertainty; (iii) to prevent the proliferation of unilateral measures that will lead to a deterioration of the investment climate. Fiscal incentives are widely used for investment promotion (to attract FDI): Tax holidays and exemptions will lose all or most of their attraction for investors and as UNCTAD (2022) pointed out ‘More than one third of fiscal incentives were profitbased (mainly tax holidays and reduced CIT). Expenditure-based incentives, which tend to reward reinvestment (e.g., allowances or tax credits) constituted just over 1 in 10 new tax incentives’. 8 The

tax competition to promote investment has led to declining corporate income tax (CIT) rates in all geographical regions and in most economies since the 1980s. The worldwide CIT rate more than halved, from 40 per cent in 1980 to 23 per cent in 2021 (UNCTAD, 2022).

82

J. J. Durán-Herrera and P. Lamothe-Fernández

Annex: Debt Finance of Main Multinational Firms by Country of Origin 2018–2020 Table 5.2 Regression analysis results for 2018 Debt financing of the main multinationals by country of origin Country EBIT Financial expenses % Total debt Australia 61,667 25,761 41.77% 2,547,954 Belgium 13,125 4728 36.02% 187,042 Brazil 80,465 51,875 64.47% 1,371,485 Britain 94,508 24,940 26.39% 7,539,475 Britain/ 8494 616 7.25% 42,265 Netherlands Canada 47,813 6236 13.04% 3,357,397 China 681,019 274,482 40.30% 25,771,907 Denmark 2394 3237 135.21% 29,860 Finland −126 273 −216.67% 26,162 France 125,095 20,289 16.22% 9,006,739 Germany 121,716 19,727 16.21% 4,931,575 India 37,532 23,366 62.26% 649,038 Indonesia 5582 637 11.41% 25,369 Ireland 18,649 4931 26.44% 118,574 Israel 4921 546 11.10% 59,553 Italy 28,174 13,327 47.30% 2,065,520 Japan 194,697 15,968 8.20% 12,907,876 Luxembourg 3099 1172 37.82% 45,007 Malaysia 5203 57 1.10% 49,728 Mexico 12,351 7135 57.77% 236,436 Netherlands 45,210 16,267 35.98% 2,595,645 Norway 79 258 326.58% 69,458 Russia 44,139 5964 13.51% 620,922 Saudi Arabia 5763 640 11.11% 41,018 Singapore 1950 259 13.28% 32,545 South Korea 82,065 26,323 32.08% 811,169 Spain 60,223 24,693 41.00% 2,393,666 Sweden 6765 370 5.47% 53,197 Switzerland 86,388 24,369 28.21% 2,656,707 Taiwan 24,126 2750 11.40% 255,318 Thailand 5180 779 15.03% 41,040 Turkey 2452 652 26.59% 17,702 U.A.E 344 10 2.86% 23,701 USA 1,136,304 249,562 21.96% 25,030,434 Venezuela 3300 22 0.67% 141,109 Total 500 3,050,665 852,220 27.94% 105,752,593

Total assets 2,798,397 258,381 1,593,834 8,242,175 59,512

% 91.05% 72.39% 86.05% 91.47% 71.02%

3,610,564 29,105,831 61,118 47,354 9,895,683 5,638,634 756,628 47,233 216,964 92,890 2,980,529 14,303,204 75,142 134,528 186,670 2,954,995 104,530 955,927 84,489 49,625 1,268,472 2,673,119 85,804 3,144,504 354,187 62,349 25,047 33,096 29,766,063 148,659 121,816,137

92.99% 88.55% 48.86% 55.25% 91.02% 87.46% 85.78% 53.71% 54.65% 64.11% 69.30% 90.24% 59.90% 36.96% 126.66% 87.84% 66.45% 64.95% 48.55% 65.58% 63.95% 89.55% 62.00% 84.49% 72.09% 65.82% 70.68% 71.61% 84.09% 94.92% 86.81%

2019 Country Australia Belgium Brazil Britain Britain/ Netherlands Canada China Denmark Finland France Germany India Indonesia Ireland Italy Japan

EBIT 102,649.95 A C 17,000.00 A C 108,993.58 A C 148,838.00 A C 15,490.00 A C 116,323.00 A C 1,085,038.33 A C 1829.95 A C 92.10 A C 388,372.37 A C 159,266.05 A C 68,146.49 A C 9219.48 A C 17,007.00 A C 71,489.10 A C 243,153.85 A C

Table 5.3 Regression analysis results for 2019 Financial expenses 47,640.77 A C 9,259.00 A C 66,243.40 A C 44,475.13 A C 900.00 A C 60,580.00 A C 295,461.58 A C 804.12 A C 332.99 A C 95,381.06 A C 35,774.90 A C 32,691.82 A C 3489.89 A C 2010.00 A C 19,885.80 A C 8325.04 A C 46.41% 54.46% 60.78% 29.88% 5.81% 52.08% 27.23% 43.94% −361.55% 24.56% 22.46% 47.97% 37.85% 11.82% 27.82% 3.42%

%

Total debt 2,647,565.90 A C 167,617.00 A C 1,534,939.70 A C 5,979,453.90 A C 54,732.20 A C 4,464,687.20 A C 32,208,496.60 A C 24,015.00 A C 27,692.90 A C 10,138,428.30 A C 5,213,404.20 A C 741,287.90 A C 37,120.50 A C 104,272.80 A C 2,923,183.70 A C 13,446,321.80 A C

Total assets 2,820,898.90 A C 232,103.00 A C 1,771,705.70 A C 6,741,484.90 A C 67,958.20 A C 4,851,977.20 A C 36,219,385.60 A C 56,636.00 A C 45,167.90 A C 11,040,826.30 A C 6,071,039.20 A C 894,314.90 A C 64,718.50 A C 204,841.80 A C 3,193,509.70 A C 15,016,149.80 A C

(continued)

% 93.86% 72.22% 86.64% 88.70% 80.54% 92.02% 88.93% 42.40% 61.31% 91.83% 85.87% 82.89% 57.36% 50.90% 91.54% 89.55%

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . . 83

Luxembourg Malaysia Mexico Netherlands Norway Poland Russia Saudi Arabia Singapore South Korea Spain Sweden Switzerland Taiwan Thailand Turkey U.A.E USA Total 500

5668.00 A C 1429.00 A C 39,289.17 A C 106,487.40 A C 19,450.00 A C 1998.64 A C 89,410.00 A C 222,577.18 A C 4598.69 A C 129,292.93 A C 89,909.86 A C 3895.00 A C 103,243.09 A C 30,147.47 A C 7591.00 A C 2634.00 A C 1171.16 A C 1,116,048.36 A C 4,527,565.99 A C

Table 5.3 (continued) 687.00 A C 3.00 A C 9010.04 A C 39,801.55 A C 580.00 A C 29.09 A C 17,430.00 A C 290.64 A C 1498.90 A C 5521.79 A C 43,814.67 A C 191.00 A C 29,229.40 A C 3742.12 A C 820.00 A C 785.00 A C 304.69 A C 220,555.75 A C 1,097,550.13 A C 12.12% 0.21% 22.93% 37.38% 2.98% 1.46% 19.49% 0.13% 32.59% 4.27% 48.73% 4.90% 28.31% 12.41% 10.80% 29.80% 26.02% 19.76% 24.24%

49,163.00 A C 61,948.20 A C 278,857.40 A C 2,132,059.80 A C 69,538.00 A C 7566.20 A C 658,614.70 A C 173,042.10 A C 88,039.30 A C 1,335,620.80 A C 2,652,940.90 A C 39,637.60 A C 2,716,972.60 A C 595,088.20 A C 45,470.40 A C 17,308.70 A C 34,687.80 A C 26,503,711.10 A C 117,173,486.40 A C

91,249.00 A C 154,071.20 A C 291,469.40 A C 2,454,481.80 A C 112,508.00 A C 17,079.20 A C 1,023,339.70 A C 444,104.10 A C 112,980.30 A C 1,942,766.80 A C 2,957,861.90 A C 53,558.60 A C 3,206,822.60 A C 739,713.20 A C 72,348.40 A C 23,570.70 A C 34,687.80 A C 31,167,628.10 A C 134,192,958.40 A C

53.88% 40.21% 95.67% 86.86% 61.81% 44.30% 64.36% 38.96% 77.92% 68.75% 89.69% 74.01% 84.72% 80.45% 62.85% 73.43% 100.00% 85.04% 87.32%

84 J. J. Durán-Herrera and P. Lamothe-Fernández

2020 Country Australia Austria Belgium Bermuda Brazil Britain Canada China Denmark Finland France Germany Hong Kong India Ireland Italy Japan

EBIT 57,940.03 A C 2345.06 A C 17,670.00 A C 6079.00 A C 66,212.13 A C 19,516.00 A C 129,185.26 A C 1,071,464.08 A C 4128.91 A C 519.00 A C 204,025.89 A C 136,934.64 A C 2076.47 A C 31,095.65 A C 17,450.11A C 59,073.53 A C 217,617.92 A C

Table 5.4 Regression analysis results for 2020 Financial expenses 25,278.78 A C 956.75 A C 4,890.00 A C 648.00 A C 32,314.21 A C 4643.00 A C 63,269.76 A C 267,390.98 A C 867.01 A C 345.00 A C 78,538.33 A C 38,954.96 A C 259.92 A C 13,982.82 A C 2278.57 A C 31,948.78 A C 53,714.10 A C

% 43.63% 40.80% 27.67% 10.66% 48.80% 23.79% 48.98% 24.96% 21.00% 66.47% 38.49% 28.45% 12.52% 44.97% 13.06% 54.08% 24.68%

Total debt 703,677.60 A C –A C 160,926.00 A C 66,677.00 A C 1,550,078.20 A C 190,640.70 A C 5,009,604.40 A C 31,991,405.80 A C 27,301.00 A C 26,717.30 A C 10,359,138.60 A C 5,290,416.80 A C 255,555.20 A C 738,485.90 A C 77,847.60 A C 2,967,484.10 A C 14,502,011.40 A C

Total assets 1,420,653.50 A C 45,316.90 A C 236,648.00 A C 97,028.00 A C 1,785,309.20 A C 347,950.60 A C 5,480,435.00 A C 36,513,039.00 A C 55,399.00 A C 43,917.30 A C 11,352,648.00 A C 6,228,456.00 A C 316,260.20 A C 886,196.90 A C 203,647.60 A C 3,246,038.10 A C 16,162,874.10 A C

(continued)

% 49.53% 0.00% 68.00% 68.72% 86.82% 54.79% 91.41% 87.62% 49.28% 60.84% 91.25% 84.94% 80.81% 83.33% 38.23% 91.42% 89.72%

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . . 85

Source: Fortune

Luxembourg Malaysia Mexico Netherlands Norway Poland Russia Saudi Arabia Singapore South Korea Spain Sweden Switzerland Taiwan Thailand Turkey United Kingdom USA Total 500

Table 5.4 (continued)

1237.00 A C 14,414.79 A C 18,639.25 A C 62,004.52 A C 10,280.00 A C 5365.00 A C 89,500.00 A C 179.97 A C 3613.10 A C 61,135.52 A C 81,978.22 A C 5135.00 A C 103,848.52 A C 36,377.63 A C 5842.00 A C 1974.00 A C 119,120.36 A C 3,052,770.04 A C 5,714,274.60 A C

695.00 A C 655.71 A C 4247.26 A C 6638.77 A C 988.00 A C 13.00 A C 19,130.00 A C 2.17 A C 1047.30 A C 15,543.38 A C 43,776.58 A C 177.00 A C 32,178.60 A C 7898.74 A C 849.00 A C 995.68 A C 45,346.13 A C 1,048,778.37 A C 1,849,241.67 A C

−56.18% 4.55% 22.79% 10.71% 9.61% 0.24% 21.37% 1.20% 28.99% 25.42% 53.40% 3.45% 30.99% 21.71% 14.53% 50.44% 38.07% 34.35% 32.36%

49,387.00 A C 57,062.40 A C 394,824.50 A C 1,954,324.30 A C 76,924.00 A C –A C 703,316.70 A C 122,390.10 A C 77,931.00 A C 1,354,551.00 A C 2,713,816.30 A C 41,268.10 A C 2,727,026.80 A C 740,830.20 A C 53,617.00 A C 19,352.00 A C 7,925,989.40 A C 27,743,254.10 A C 120,673,832.50 A C

87,908.00 A C 152,218.40 A C 332,246.00 A C 2,344,454.70 A C 118,063.00 A C 18,804.20 A C 1,139,183.70 A C 398,348.60 A C 101,200.00 A C 1,922,665.50 A C 3,031,372.60 A C 56,077.10 A C 3,201,752.70 A C 897,840.50 A C 82,952.00 A C 25,453.00 A C 8,690,822.40 A C 32,744,319.10 A C 139,767,498.90 A C

56.18% 37.49% 118.83% 83.36% 65.16% 0.00% 61.74% 30.72% 77.01% 70.45% 89.52% 73.59% 85.17% 82.51% 64.64% 76.03% 91.20% 84.73% 86.34%

86 J. J. Durán-Herrera and P. Lamothe-Fernández

5 Debt and Taxes as Value-Added Factors for Multinational Enterprises. . .

87

References Burgman, T. A. (1996). An empirical examination of multinational corporate capital structure. Journal of International Business Studies, 553–570. Denis, D. J., & Mihov, V. T. (2003). The choice among bank debt, non-bank private debt, and public debt. Evidence from new borrowing. Journal of Financial Economics, 70, 3–28. Doukas, J. A., & Pantzalis, C. (2003). Geographic diversification and agency cost of debt of multinational firms. Journal of Corporate Finance, 9, 59–92. Ferrer, E., Santamaria, R., & Suárez, N. (2019). Does analyst information influence the cost of debt? Some international evidence. International Review of Economics and Finance, 64, 323– 342. James, C. M. (1987). Some evidence on the uniqueness of bank loans. Journal of Political Economics, 19, 217–235. Lee, K. C., & Kwok, C. C. (1988, Summer). Multinational corporations vs. domestic corporations: International environment factors and determinants of capital structure. Journal of International Business Studies, 195–217. Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. The American Economic Review, 48(3), 261–297. Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 147–175. OCDE. (2007). Fundamental reform of corporate income tax. Saez, E., & Zuckman, G. (2019). The Triumph of injustice. W W Norton. Singh, K., & Hodder, J. E. (2000). Multinational capital structure and financial flexibility. Journal of International Money and Finance, 9(6), 853–885. UNCTAD. (2022). World investment report 2022. International tax reforms and sustainable investment. Naciones Unidas. Zucman, G. (2018). The missing profit of nations. NBER.

Chapter 6

Modelling and Forecasting GDP of Greece with a Modified Exponential Smoothing State Space Framework Melina Dritsaki and Chaido Dritsaki

Abstract The comparison between models and their predictions has been a big topic in the literature in recent years. There are two main methods used to compare the prediction quality; the average square error and the distance in time. In order to evaluate the forecast performance, as well as order the forecasts, researchers have developed several measures of accuracy. In the current paper to predict the GDP of Greece, we use non-linear dynamic models ETS, using state-space-based likelihood calculations in order to choose models and calculate the forecast standard errors. The estimation of the models is made with the function of maximum likelihood, while the choice between models with additive and multiplier errors is made with the use of the Akaike (AIC) criterion based on likelihood and not on one-step ahead forecasting. The process is completed by clearly defined methods for the assessment, the likelihood evaluation with the BFGS algorithm and the analytical derivation of the forecasted points and intervals under a Gaussian error assumption. The results of the work showed that the model with a multiplier error, with a multiplier tendency to depreciation and additional seasonality, is the most appropriate for the period under examination. Moreover, the results of the forecast showed a fall in GDP for the coming quarters with this decline to depreciate. Keywords Non-linear dynamic models ETS · State space models · Akaike information criterion (AIC) · GDP Greece.

M. Dritsaki () University of Western Macedonia, Department of Economics, Kastoria, Greece e-mail: [email protected] C. Dritsaki Department of Accounting and Finance, University of Western Macedonia, Kozani, Greece © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_6

89

90

M. Dritsaki and C. Dritsaki

6.1 Introduction One of the key measures of the state of the economy is gross domestic product (GDP). Gross domestic product (GDP) is a monetary measure of the market value of all final goods and services produced in a country over a given period of time (annual or quarterly). There are two ways to measure GDP, the nominal and the real. GDP is usually collected at the nominal level but in order to compare it in different periods, nominal GDP needs to be adjusted in relation to inflation, in which case we can determine whether GDP growth is due to increased output or price increases. GDP calculations follow international standards developed by the International Monetary Fund, the European Commission, the International Monetary Fund (IMF), the European Commission (EC), the Organization for Economic Cooperation and Development (OECD), as well as the World Bank. GDP is often used as a measure of economic progress, as it is seen as the strongest statistical indicator for national growth and progress, but also as a measure for international comparisons (Dritsaki & Dritsaki, 2021b). However, many argue that with GDP we cannot measure progress since we let some external characteristics, such as resource extraction, environmental impacts and unpaid domestic work, not participate (Raworth, 2017). One way to deal with this problem is to look at GDP alongside another measure of economic growth, such as the Human Development Index (HDI). GDP enables central banks to judge whether the economy is shrinking or expanding and whether a threat such as recession or inflation is on the horizon. Countries with a higher GDP will have a higher volume of goods and services and will generally have a higher standard of living. For this reason, many political leaders consider GDP growth as an important measure of national success. However, many economists have argued that GDP should not be used as a proxy for overall economic success, but it is more accurate to use purchasing power parity (PPP) as a measure of national wealth. The GDP forecast is based on an assessment of the economic climate in the individual countries and the world economy and is calculated with an indicator that measures the growth rate compared to the previous year. GDP forecasting is a vital economic element in central bank decision-making for all countries. The selection of models and the evaluation of projected returns (forecast performance), with various precision measures, are of significant theoretical and practical importance for the objectives of economic growth (Dritsaki, 2015). In recent years, the literature on forecasting with exponential smoothing has expanded into a model approach that has been formulated by Hyndman et al. (2008). The ETS (Error-Trend-Seasonal or Exponential Smoothing) framework defines an extended class of exponential smoothing methods that includes the standard models of exponential smoothing (ES), additives and multipliers of Holt-Winters, but also offers a variety of new methods. In addition, the smoothing of the ETS provides a theoretical basis for the analysis of these models using state-spacebased probability calculations with support for the selection of models and the

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

91

calculation of forecast standard errors. The predicted values are calculated based on existing (historical) values using the version of exponential smoothing (ETS) algorithm as formulated by many researchers including Makridakis and Hibon (2000). The algorithm extensively examines the accuracy of the out-of-sample prediction compared to the other methods in these competitions and the traditional exponential smoothing framework. The proposed algorithm can be used as an alternative to forecasting processes for modelling seasonal time series. In addition, it provides the possibility of automatic modelling of multiple seasonal time series that cannot be addressed by other forecasting procedures. The rest of the chapter is organised as follows: Section 6.2 describes forecasts accuracy in the literature while in Sects. 6.3 and 6.4 a theoretical background is given. Data are presented in Sect. 6.5. In Sect. 6.6, the empirical results are shown, and finally, conclusions are provided in Sect. 6.7.

6.2 Forecasts Accuracy in the Literature Assessing the accuracy of predictions has been an important issue in the scientific community in recent years. It has been observed that the magnitude of residuals is not a reliable predictor indication. The accuracy of the forecast is determined primarily by the correct specialisation of the model in the data. In common practice, the available data are divided into two sections; the training and the control data. Training data is used to estimate the parameters, while control data is used to evaluate the accuracy of the forecast. Fair (1986) proposes the three most common precision prediction measures, namely root mean squared error (RMSE), mean absolute error (MAE) and the inequality factor of Theil (Theil’s inequality coefficient U). These measures assess the accuracy of ex-post and ex-ante forecasts. In the ex-post prediction, the actual values of the exogenous variables are used, and in the ex-ante prediction, those that guessed values in the exogenous variables are used. Zhuo Chen and Yuhong Yang (2004) distinguish prediction measures between stand-alone and relative accuracy. Autonomous precision measures are those that can be obtained without additional reference forecasts. These are the root mean squared error (RMSE), mean absolute error (MAE), mean squared forecast error (MSE) and mean absolute percentage error (MAPE). The relevant precision measures are those that take into account the benchmark and can eliminate bias trends, seasonal recommendations and outliers. A relative measure of accurate predictions is Theil’s U2 statistic which has as a benchmark the value of the last observation of the data of education. The stability of the precision measures is another issue that we should take into account. Clements and Hendry (1998), proposed the linear transformation of the original series, as an important factor for the stability of the precision measures. The measures of the forecast may produce large numbers, due to outliers, or unsuitability of the model, in which case the comparison of forecasts is unreliable.

92

M. Dritsaki and C. Dritsaki

While all measures are related to loss functions that are minimised, they all have some limitations. All measures do not provide meaningful information by themselves. For example, the RMSE is affected by outliers that are usually found in economic data. According to Fildes (1992), none of the measures are independent unless applied to percentage changes. Also, the issue of scaling is a critical one when analysing a number of data in the series. Clements and Hendry (1998) as mentioned above proposed not only linear transformations but also a generalisation of the RMSE that takes into account intercorrelations between errors when analysing more than one macroseries. The practicality of this measure was questioned by Armstrong and Fildes (1995) because forecast evaluations were based on small samples. Armstrong and Fildes (1995) show that the purpose of measuring forecast error is to provide information about the shape of error distribution and proposed a loss function to measure forecast error. Moreover, they show that it is not sufficient to use only one measure of accuracy. Fildes and Steckler (2000) analysed the problem of accuracy using statistics, pointing to milestones in the literature. For comparison between MSE indices of forecasts, Granger and Newbold (1977) propose a new statistic. Another statistic is presented by Diebold and Mariano in order to compare quantitative error measurements. Diebold and Mariano (1995) examine a test comparison to accurately two predictions below the zero hypothesis that states that there is no difference between the two predictions (that states the lack of difference). The proposed test of them was later improved by Ashley (2003), who developed a new statistic based on a bootstrap conclusion. Mariano (2002) presents the most important checks on the accuracy of predictions, including the control of Diebold–Mariano (DM). Since normal distribution is a poor approximation of low-volume data series, Harvey et al. (1997) improved the properties of finite data sets by applying some corrections, such as changing DM statistics in order to eliminate bias and compare not with the normal distribution, but to the t-Student. Clark (1999) evaluates the power of certain controls of equal predictive accuracy, such as modified versions of the DM control or those of Newey and West (1987), which are based on the Bartlett kernel and with a fixed length of data series. Finally, Hyndman (2014) suggests for prediction accuracy measurements the following: • We use MAE or RMSE if all predictions are on the same scale. • We use MAPE if we need to compare the forecast accuracy in different series with different scales unless the data contains zeros or small values or does not measure a quantity. • We use MASE if we need to compare forecast accuracy in different series with different scales, especially when MAPE is inappropriate. • We use series cross-validation where possible, instead of a simple trainingcontrol set split. In order to forecast GDP increase in India, Gupta and Minai (2019) evaluate the accuracy of a forecast based on the properties of the forecast error. For this

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

93

purpose, they have calculated three measures of forecast accuracy: mean absolute error (MAE), root mean square error (RMSE) and Theil’s U statistic. To evaluate the performance of the forecasts, the authors have compared them with naive forecasts and common rules of thumb, using moving averages. The results of the work showed that the findings of the errors suggest that the performance of the Reserve Bank of India (RBI) forecasts is favourable compared to other organisations. Abdi´c et al. (2020), explore the possibility of creating an econometric model for Bosnia and Herzegovina’s short-term GDP forecast. Their aim is to identify the most effective model for predicting GDP. For this purpose, they compare ARIMA models, bridge models and factor models for three different periods. The results showed that bridge models are more effective in predicting GDP. Maccarrone et al. (2021) compare the predictive power of different models for the US GDP forecast. Using quarterly data from 1976 to 2020, they found that the machine learning K-Nearest Neighbour (KNN) model gives better results in the long-term GDP forecast than the time series models.

6.3 A Short Description of Exponential Smoothing (ES) Exponential smoothing was first proposed by Brown in (1956) and then expanded by Holt in (1957). All exponential linear and non-linear smoothing methods have been shown to make the best predictions from state space model innovations (Hyndman et al., 2002). The classification of exponential smoothing methods carried out by Hyndman et al. (2002) and expanded by Taylor (2003) provides fifteen methods of exponential smoothing. However, a subset of these methods is better known by other names, such as N,N is the method of exponential smoothing without trend and seasonality, that is, it is the well-known method of simple exponential smoothing (Dritsaki & Dritsaki, 2021a). A,N is the method of exponential smoothing with additive trend and without seasonality, that is, the linear method of Holt, while (Ad,N) is the method of exponential smoothing with additive tendency to damp the damped trend method and without seasonality. Similarly, A,A is the method of exponential normalisation with additive tendency and additive seasonality, that is, the additive Holt-Winters method and A,M is the method of exponential smoothing with additive tendency and multiplier seasonality, that is, the multiplicative Holt– Winters’ method. Forecasting time series data containing trend patterns, seasonal patterns or both at the same time can be predicted by the smoothing method. Smoothing takes the mean value for several years to estimate the value of one year. The smoothing method is classified into two parts, the smoothing method and the exponential smoothing method. The prediction of data affected by the trend or seasonal patterns is carried out using the method of exponential smoothing using different weights for previous data and these weights have exponentially decreasing characteristics (Makridakis et al., 1999).

94

M. Dritsaki and C. Dritsaki

6.3.1 Winters’ Exponential Smoothing Method If the data pattern is not only influenced by trend patterns but also by seasonal patterns, Holt’s (1957) exponential smoothing method is not suitable for prediction because it cannot detect seasonal patterns. Thus, Winters (1960) extended the exponential smoothing of Holt by adding a parameter to overcome the problem with prediction in the seasonal patterns of the data, and the method is called the Holt-Winters seasonal method. The Holt–Winters method describes two separate models that depend on the behaviour of the initial series. This technique is based on the analysis of linear trend and the seasonal component. This method of exponential smoothing is divided into two models, the additive and the multiplier models. The analysis of the additive model is carried out if the original diagram of the data shows stable seasonal fluctuations, while the multiplier model is used if in the original diagram of the data the seasonal variations vary. Each model examines three smoothing equations. One comprises the level, one the trend and the other the seasonality. When an increase in seasonal amplitude is required, the range shows a difference between the highest and the lowest demand point in the cycles grows with time, and the multiplier model becomes adequate. When the seasonal width is constant, this means that the largest and smallest points of the cycles are independent of the temporal variation, and the additive model must be used (Holt, 2004a, 2004b).

6.3.2 Additive Seasonal Method Hot–Winters The model of the additive seasonal method of Hot–Winters is given as follows: Level function : Lt = α (yt − st−m ) + (1 − α) (Lt−1 + bt−1 ) Trend function : bt = β (Lt − Lt−1 ) + (1 − β) bt−1

(6.1)

(6.2)

Seasonality component : st = γ (yt − Lt−1 − bt−1 ) + (1 − γ ) st−m

(6.3)

Prediction function : yˆt+h|t = Lt + hbt + st+h−m(k+1)

(6.4)

where k is the integral part of the relationship h−1 m , which ensures that estimates of seasonal indicators used for the forecast come from the last year of the sample. The level function shows a weighted average among the seasonally adjusted observation (yt − st − m ) and the non-seasonal forecasts (Lt − 1 + bt − 1 ) for time t. Trend function is the same as Holt’s linear method. The seasonal function shows a

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

95

weighted average between the current seasonal index, (yt − Lt − 1 − bt − 1 ) and the seasonal index of the same period of the previous year, that is, m periods of time before.

6.3.3 Multiplicative Seasonal Method Hot–Winters The model of the multiplicative seasonal method Hot–Winters is given as follows: Prediction function : yˆt+h|t = (Lt + hbt ) st+h−m(k+1) Level function : Lt = α

yt st−m

+ (1 − α) (Lt−1 + bt−1 )

Trend function : bt = β (Lt − Lt−1 ) + (1 − β) bt−1 Seasonality component : st = γ

yt + (1 − γ ) st−m (Lt−1 + bt−1 )

where k is the integral part of the relationship frequency.

h−1 m ,

(6.5)

(6.6)

(6.7)

(6.8)

and m is the seasonality

6.4 Error-Trend-Seasonal (ETS) Forecasting Methods Exponential smoothing methods (ES) have been used by researchers for many decades. Recent methodological developments have incorporated these models into a dynamic non-linear model framework (see Chatfield et al., 2001). ETS models are a family of time series with an underlying state space consisting of a layer element, a trend element (T), a seasonal component (S) and an error term (E). The trend reflects the long-term motion of the series, the seasonality corresponds to a pattern with known periodicity, and the error is uncertain and unpredictable component of the series. Hyndman et al. (2002) describe the ETS (error-trend-seasonal) framework that defines an extended class of exponential smoothing methods (ES) and provides a theoretical basis for the analysis of these models using state-space models based on likelihood calculations, with support for model selection and calculation of standard forecast errors. ETS calculates a weighted average on all observations in the time series and uses it as an input to predict the series. Weights are exponentially decreasing over time, rather than the constant weights in simple moving average

96

M. Dritsaki and C. Dritsaki

methods. Weights depend on a fixed parameter, which is known as the smoothing parameter. The ETS model is one that takes into account error (E), trend (T), and seasonal components (S) in the data of a time series. These three components can be combined in various additive and multiplier combinations of the time series. The specification of an ETS template can be created by the combinations of the individual components: E {A,M} T {N,A,M,AD,MD} S {N,A,M} where N = none, A = additive, M = multiplicative, AD = additive dampened and MD = multiplicative dampened (damping uses an extra parameter to reduce the impact of the trend over time). There are a total of 2*5*3 = 30 possible ETS models associated with these choices. The optimal exponential ETS smoothing model is selected based on the Akaike criterion (AIC). To distinguish the templates with additive and multiplying errors, we add an additional letter on the front of the method indicator. The triplet (E,T,S) refers to the three components: error, trend and seasonality. Thus, the ETS model (A,A,N) has additive errors, additive tendency and no seasonality. In other words, we would say that it is Holt’s linear method with additive errors. Similarly, ETS (M,Md,M) refers to the model with multiplicative errors, a damped multiplicative trend and multiplicative seasonality. The simplest of the ETS templates is known as simple exponential smoothing. In ETS terms, it corresponds to the model (A,N,N), that is, a model with additional errors, no trend and no seasonality. The state space formulation of Holt’s method is: yt = yt−1 + et

(6.9)

Lt = Lt−1 + αet

(6.10)

This state space formulation can be turned into a different formulation, a forecast and a smoothing equation (as can be done with all ETS models): Prediction function : yˆt|t−1 = Lt−1

(6.11)

Level function : Lt = αyt + (1 − α) Lt−1

(6.12)

where yˆt|t−1 is the prediction of yt given the information the former step. In the simple exponential smoothing model, the prediction corresponds to the previous level. The second equation (smoothing equation) calculates the next level as a weighted average of the previous level and the previous observation. The exponential method of smoothing can be modified to incorporate a trend and a seasonal component. In the Holt-Winters additive method, the seasonal component is added to the rest. This

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

97

template corresponds to the model ETS (A,A,A) and has the following state space formulation: yt = Lt−1 + bt−1 + St−m + et

(6.13)

Lt = Lt−1 + bt−1 + αet

(6.14)

bt = bt−1 + βet

(6.15)

St = St−m + γ et

(6.16)

6.4.1 Additive Error Model: ETS (A,Ad,N) Let μt = yˆt = Lt−1 + bt−1 be the prediction of a step ahead of yt assuming that we know all the values of the parameters. If et = yt − μt is the error of predicting a step ahead of time t then: yt = Lt−1 + φbt−1 + et

(6.17)

Lt = Lt−1 + φbt−1 + αet

(6.18)

bt = φbt−1 + β ∗ (Lt − Lt−1 − φbt−1 ) = φbt−1 + αβ ∗ et

(6.19)

In the last function, we can put β = αβ ∗ . The functions above consist of a state space underlying the damped Holt’s method model. This is an innovation of the state space model because the same term of the error appears in all functions (see Aoki, 1987). Therefore, we can write in standard state space by specifying the state  vector as xt = (Lt , bt ) so the above functions can be defined as follows: yt = [1φ] xt−1 + et    1φ α xt = xt−1 + et 0φ β

(6.20)



(6.21)

The above model is fully defined when we declare the distribution of the error term et . We usually assume that these errors are independent and identically

98

M. Dritsaki and C. Dritsaki

distributed, following the normal distribution with mean 0 and variance σ 2 . ηλαδη´ et → NID(0, σ 2 ) (see Hyndman et al., 2008).

6.4.2 Multiplicative Error Model: ETS(M,Ad,N) If in a multiplier model the et = time t, then:

yt −μt μt

is the forecasting error one step ahead of

yt = (Lt−1 + φbt−1 ) (1 + et )

(6.22)

Lt = (Lt−1 + φbt−1 ) (1 + αet )

(6.23)

bt = φbt−1 + β (Lt−1 + φbt−1 ) et

(6.24)

yt = [1φ] xt−1 + (1 + et )

(6.25)

   α 1φ et xt−1 + [1 φ] xt−1 β 0φ

(6.26)

or

 xt =

We also assume here that et → NID(0, σ 2 ). It should be noted here that the above model is a non-linear model state space, which is difficult to handle the estimation and prediction. However, this is an advantage of the innovations that state space models have that can calculate predictions, probability and prediction intervals, as is the case with additive models (see Hyndman et al., 2008).

6.4.3 State-Space Models for All Exponential Smoothing Methods According to Hyndman et al. (2008), there are state space models for all exponential smoothing methods. The general model consists of a state vector of the form:  xt = (Lt , bt , st , st − 1 , . . . , st − m + 1 ) while the equations in the state space template are as follows: yt = ω (xt−1 ) + r (xt−1 ) et

(6.27)

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

xt = f (xt−1 ) + g (xt−1 ) et

99

(6.28)

where et is a Gaussian white noise process with an average zero and variance σ 2 and μt = ω(xt − 1 ). The model with additive errors has r(xt − 1 ) = 1 so that yt = μt + et . The model with multiplicative errors has r(xt − 1 ) = μt so that yt = μt (1 + et ). t Hence et = yt μ−μ is what relevant error for the multiplicative model. t For the non-linear dynamic models of the ETS that we examine here, we consider that the smoothing equations are weighted averages of a term that depends on the current prediction error and on a term that depends on the previous states. Equations according to Hyndman et al. (2002) have the following form: Lt = αP (xt−1 , et ) + (1 − α) Q (xt−1 ) φ

(6.29)

2 bt = βR (xt−1 , et ) + (1 − β) φ1 bt−1

(6.30)

St = γ T (xt−1 , et ) + (1 − γ ) St−m

(6.31)

where Pt ≡ P(xt − 1 , et ), Rt and Tt are functions of the forecast error and lagged states, and Qt = Q(xt − 1 ) is a function of lagged states. φ 1 is the damping parameter for linear trend models and φ 2 is the damping parameter for multiplicative trend models. In the absence of damping, parameters are defined as 1. The exact forms of all these equations depend on the ETS specification. The cases corresponding to the 30 possible specifications are listed in Hyndman et al., 2008 (Tables 2.2 and 2.3, pp. 21–22).

6.4.4 Estimation ETS Model Based on the assumption of the normality of the error term, et → NID(0, σ 2 ), the ETS model can be estimated (estimation of parameters) by likelihood maximisation, which is equivalent to minimising the average squared prediction error et . The likelihood is the probability of the data arising from the specified model. Thus, a high probability is associated with a good model. The maximum probabilities estimator identifies the parameters and initial states that maximise the log-likelihood function. The Gaussian log-likelihood for ETS specifications can be written in terms of the prediction errors as:

100

M. Dritsaki and C. Dritsaki n    1  2 et2 log L ϑ, x0 , σ 2 |y = − log 2π σ 2 − T 2 σ2

(6.32)

t=1

For a given set of parameters and initial states of the ETS models, probabilities are recursively evaluated using the state equations by resolving the forecast error using the appropriate prediction decomposition equation. Parameters and initial conditions are obtained maximising the probabilities of the above function with respect to ϑ x0 and σ 2 and use of Broyden, Fletcher, Goldfarb and Shanno (BFGS) algorithm.

6.4.5 Model Selection Accurate measures of the forecast, such as the mean square error (MSE), can be used to select a template in a particular data set, provided that the errors are calculated from data in a hold-out set and not from the data used in the estimate of the template. However, there are often some out-of-sample errors. Therefore, in order to have reliable results, a penalised method based on the application of the sample (on the in-sample fit) is better. Since both probabilities and predictive error can be calculated for each ETS model, they can be compared on the basis of probability with an information criterion. Likelihood-based comparisons can be performed using the standard likelihood-based criteria: information criterion Akaike (AIC), information criterion Schwarz (BIC) or criterion Hannan–Quinn (HQ). More specifically: AIC = −2 log L (ϑ, x0 ) + 2q

(6.33)

BIC = −2 log L (ϑ, x0 ) + log(T )q

(6.34)

HQ = −2 log L (ϑ, x0 ) + 2 log (log T ) q

(6.35)

where (ϑ, x0 ) are the maximum values, q is the number of parameters in ϑ and the number of free states in x0 . The model that minimises the AIC, BIC or HQ across all available models is adopted. The above criteria are used to select between models with additive and multiplicative errors. The point prediction between the two masters (additive and multiplicative) is identical, resulting in standard precision measures, but the average square error (MSE) or the average absolute error rate (MAPE) cannot be chosen between error types. Therefore, we use the criteria of AIC, BIC and HQ because they are based on probability and not on one-step predictions. Here we should mention that caution is needed to distinguish exponential methods of smoothing from underlying state space models. An exponential method of smoothing is an

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

101

Table 6.1 Descriptive statistics of GDP Mean 49571.04

Max. 64451.30

Min. 36093.10

Std. dev. 6951.458

Skew. 0.430309

Kur. 2.302693

J-B 5.316568

Obser. 104

algorithm for the production of point forecasts only. The underlying stochastic state space model gives the same point predictions, but it also provides a framework for calculating computing prediction intervals and other properties.

6.5 Data For the current work, we have used quarterly GDP data for Greece at the 2010 reference levels (ECU/EURO) from the first quarter of 1995Q1 to the fourth quarter of 2020Q4. Based on these data, the seasonal models of the Holt–Winters are developed and then applied to predict GDP. Quarterly GDP data are taken from the World Bank’s development indicators. Table 6.1, presents the descriptive statistics of GDP. From the above table, we see that the average GDP in Greece for the period we examine is 49571.04 euros (with a standard deviation of 6951.45 euros). GDP in this period follows the normal distribution, it has a positive asymmetry (therefore, there is the presence of some large value in GDP resulting in the average being distant from the median), and is wide-curved (less than 3) so GDP values are far from the average. In the following Diagram 6.1, the histogram together with the graph of normal distribution is depicted. Most of the data series follows a normal distribution as it is shown in the above diagram.

6.6 Empirical Results Time Series Plots In Fig. 6.1, the progression of GDP for Greece is shown for the examined period. From the above figure, we observe that there is a trend (upward and downward) of GDP, but also seasonality. As we have mentioned before, a series can be decomposed into three components, trend (T), seasonality (S) and error (E) where the term trend characterises the long-term movement of the series, the seasonal term corresponds to a pattern with known periodicity and the term error is the irregular, unpredictable component of the series. These three components can be combined into various additive and multiplier combinations. The specialisation of an ETS template can be created by the combinations of the individual components that are a

102

M. Dritsaki and C. Dritsaki

Diagram 6.1 Histogram and graph of normal distribution function

Fig. 6.1 Time series plot of original series GDP

total of 30, so are the possible ETS models associated with these options. Using the function of maximisation of probability, the AIC criterion, the algorithm (BFGS) and the analytical derivation of the predicted points and space under the Gaussian error hypothesis, we get the following estimates.

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

103

Table 6.2 ETS results of model (M,MD,A) ETS Smoothing Original series: GDPGR Sample: 1995Q1 2020Q4 Included observations: 104 Model: M,MD,A - Multiplicative Error, Multiplicative Dampened Trend. Additive Season (Auto E=*, T=*, S=*) Model selection: Akaike Information Criterion Convergency achieved on boundaries. Parameters Alpha: 1.000000 Beta: 0.146281 Gamma: 0.177541 Phi: 0.922911 Initial Parameters Initial level: 38577.22 Initial trend: 1.008369 Initial state 1: 626.7672 Initial state 2: 1801.973 Initial state 3: 20.49437 Initial state 4: −2449.235 Compact Log-likelihood −944.5963 Log-likelihood −850.6576 Akaike Information Criterion 1907.193 Schwarz Criterion 1930.992 Hannan-Quinn Criterion 1916.835 Sum of Squared Residuals 0.032064 Root Mean Squared Error 0.017559 Average Mean Squared Error 2154161.

The results of the above Table 6.2 show the specificity of the evaluated smoothing, the initial parameters and the summary statistics. The Akaike information criterion selected ETS model is an (M,MD,A), using data from 1995Q1 until 2020Q4, with smoothing level parameter estimation α = 1.000, smoothing trend estimation β = 0.146, smoothing seasonal estimation γ = 0.177, damping trend  estimation ϕ = 0.922, of the initial states x0 = (L0 , b0 , s0 , s−1 , . . . , s−4 ) . The summary statistics of the model ETS (M,MD,A) estimation are presented to compare the model assessed with other model specifications. Diagram 6.2 shows all the models based on the Akaike criterion. From the results of the above histogram, we observe that the model (M,MD,A) is the most suitable of the models (A,A,A) with additive error, additive trend and additive seasonal component, as well as the model (M,M,M) with multiplicative error, multiplicative trend and multiplicative seasonal component. Figure 6.2 shows the comparison of the forecasts for all the models created. The graph above shows both the latest observations of the in-sampled forecasts and the out-of-sample forecasts for each of the possible ETS specifications.

104

M. Dritsaki and C. Dritsaki

Diagram 6.2 Comparison of models with AIC

In addition, our chosen ETS display settings produced both the likelihood table which contains the actual likelihood and Akaike values for each specification and the forecast comparison table, which presents a subset of the values displayed in Table 6.3. The diagram below contains a multiple graph containing the actual and predicted values of GDP during the estimation and forecast period, along with the decomposition of the time series at the level, the trend and the seasonal components (Fig. 6.3). The following table gives the predicted GDP values of Greece for the quarters from 2021Q1 to 2024Q4 with the model ETS(M,MD,A). The results of Table 6.4 show a decline in Greece’s GDP for the next quarters 2021Q1–2024Q4. A big advantage of non-linear dynamic models ETS is that prediction intervals can be created. For most ETS templates, a forecast interval can be written as: yT + h|T ± cσ h where c depends on the coverage probability, and σh2 is the prediction variance. Table 6.5 shows the results of the prediction interval. The results of the forecast interval show a wider range as the forecast period increases.

6.7 Conclusion GDP refers to the market value of all final goods and services produced in a country over a certain period of time. GDP is an important indicator of wealth and is closely linked to the level of employment and living standard of a country. Therefore, the

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

105

Fig. 6.2 Forecast comparison graph

GDP forecast may be useful for the country in future macroeconomic arrangement and control, as well as in the calculation of the international balance of payments. The forecast involves the creation of models with inputs of various key variables or indicators, usually in an effort to achieve a future growth rate of gross domestic product (GDP). However, it is often difficult to analyse and predict GDP using traditional methods. Therefore, the use of the most appropriate model for predicting GDP can have a satisfactory result. In this paper, we use quarterly GDP data from 1995:1 to 2020:4 with the research objective to construct exponential smoothing models using probability calculations of errors-trends-seasonality based on statespace-based likelihood calculations for the selection of models and the calculation of standard forecasting errors according to Hyndman et al. (2008). The application of algorithms for predicting macroeconomic variables is a new and growing field and has proven inherently appropriate in the field of economics. In

106

M. Dritsaki and C. Dritsaki

Table 6.3 Compact actual likelihood and Akaike values Model M,MD,A M,AD,A M,M,A M,A,A M,MD,M M,AD,M M,A,M M,N,A* A,AD,M A,MD,M A,M,M A,A,M A,AD,A A,MD,A M,N,M* A,M,A A,A,A M,M,M A,N,A A,N,M M,AD,N M,MD... A,AD,N A,MD,N M.M,N M,A,N A,M,N A,A,N M,N,N A,N,N

Compact LL −944.596 −944.634 −946.376 −946.727 −946.200 −946.235 −948.265 −951.150 −949.130 −949.188 −950.691 −950.888 −951.070 −951.138 −954.462 −952.536 −952.601 −954.548 −958.000 −958.019 −1060.21 −1060.34 −1061.48 −1061.76 −1063.71 −1064.30 −1064.32 −1064.66 −1067.61 −1068.96

Likelihood −850.658 −850.695 −852.438 −852.789 −8.52.262 −852.296 −854.327 −8.57.212 −855.191 −855.249 −856.752 −856.950 −857.131 −857.199 −860.524 −858.597 −858.662 −860.609 −864.061 −864.080 −966.268 −966.397 −967.545 −967.825 −969.768 −970.361 −970.377 −970.718 −973.671 −975.025

AIC 1907.19 1907.27 1908.75 1909.45 1910.40 1910.47 1912.53 1914.30 1916.26 1916.38 1917.38 1917.78 1920.14 1920.28 1920.92 1921.07 1921.20 1925.10 1928.00 1928.04 2130.41 2130.67 2132.97 2133.53 2135.41 2136.60 2136.63 2137.31 2139.22 2141.93

BIC 1930.99 1931.07 1929.91 1930.61 1934.20 1934.27 1933.69 1930.17 1940.06 1940.17 1938.54 1938.93 1943.94 1944.07 1936.79 1942.23 1942.36 1946.25 1943.87 1943.90 2143.63 2143.89 2146.19 2146.75 2145.99 2147.18 2147.21 2147.89 2144.51 2147.22

HQ 1916.83 1916.91 1917.32 1918.03 1920.04 1920.11 1921.10 1920.73 1925.90 1926.02 1925.95 1926.35 1929.78 1929.92 1927.35 1929.64 1929.77 1933.67 1934.43 1934.47 2135.77 2136.03 2138.32 2138.88 2139.70 2140.89 2140.92 2141.60 2141.36 2144.07

AMSE 2154161 1.E+100 2294318 2297096 2093787 1.E+100 2251634 2664669 1.E+100 2080659 2219168 2231954 1.E+100 2137025 2723181 2266980 2264756 2318338 2656788 2718597 1.E+100 7238593 1.E+100 7223669 7863698 7936588 7883310 7943980 9600787 9593639

Note: In two models (M,N,A) and (M,N,M) the estimators do not converge

the work, we use the version of the BFGS algorithm of exponential smoothing (ETS) to predict GDP. This algorithm examines extensively the accuracy of the out-ofsample prediction compared to other methods. In addition, it provides the possibility of automatic modelling of multiple seasonal time series that cannot be addressed by other forecasting procedures. Based on a good model fitting effect, this paper forecasts the future trend of GDP. The results of the work show that the model with a multiplier error, with a multiplier tendency to depreciation and additional seasonality, is the most appropriate for the period of GDP we are considering. Moreover, the results of the forecast showed that as depreciation declines there is a fall in the GDP for the next few quarters.

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

107

Fig. 6.3 Decomposition GDP from 1995Q1 to 2024Q4 Table 6.4 Predicted values for GDP 2021Q1–2024Q4 Year 2021Q1 2021Q2 2021Q3 2021Q4

GDP 38246.25 40287.07 43106.53 42617.61

Year 2022Q1 2022Q2 2022Q3 2022Q4

GDP 36705.11 38877.61 41816.60 41436.30

Year 2023Q1 2023Q2 2023Q3 2023Q4

GDP 35622.63 37885.14 40906.20 40600.78

Year 2024Q1 2024Q2 2024Q3 2024Q4

GDP 34855.51 37180.54 40258.78 40005.70

Predicting future economic outcomes is a vital component of central bank decision-making for all countries. Monetary policy decisions affect the economy with a delay, so monetary policy principles must be forward-looking, that is, they need to know what is likely to happen in the future. Gross domestic product (GDP) is one of the most important indicators of economic activities for countries. The scientific prediction of the index has important theoretical and practical significance for the development of economic growth goals. In this paper, we evaluate models of exponential normalisation for the forecast of Greece’s GDP. Covid-19 and the war in Ukraine have caused a major humanitarian crisis affecting millions of people. Economic shocks, and their impact on global commodities, trade and financial markets, have a significant impact on the economic outcomes and livelihoods of all countries.

108 Table 6.5 Prediction intervals of model ETS(M,MD,A)

M. Dritsaki and C. Dritsaki Year 2021Q1 2021Q2 2021Q3 2021Q4 2022Q1 2022Q2 2022Q3 2022Q4 2023Q1 2023Q2 2023Q3 2023Q4 2024Q1 2024Q2 2024Q3 2024Q4

Lower 38,067 38,731 40,161 40,085 34,691 35,283 36,585 36,516 31,602 32,141 33,327 33,264 28,788 29,279 31,546 31,204

Mean 38246.25 40287.07 43106.53 42617.61 36705.11 38877.61 41816.60 41436.30 35622.63 37885.14 40906.20 40600.78 34855.51 37180.54 40258.78 40005.70

Upper 45,443 44,900 45,125 44,153 43,434 42,957 45,125 44,152 41,904 42,634 44,203 44,153 41,300 42,634 43,276 42,876

Before the outbreak of war, the outlook for Greece’s growth and inflation seemed generally favourable, with the result that the economy returned to normality, as the Covid-19 pandemic and supply-side restrictions weakened. The growth of the world’s GDP and of course of Greece is now projected to slow down sharply from what was predicted. Commodity prices have risen significantly, reflecting the importance of supply from Russia and Ukraine in many markets, increasing inflationary pressures and hitting real incomes and spending, particularly for the most vulnerable households. The uncertainty surrounding this prospect is high and there are some significant risks. The abrupt interruption of gas flows across Europe from Russia, and further increases in commodity prices are bringing inflationary pressures that could prove stronger than expected, with higher inflation expectations at risk, moving away from the central bank’s targets and reflected in faster wage growth amid tight labour markets. Sharp increases in interest rates could also slow growth further than projected.

References Abdi´c, A., Resi´c, E., Abdi´c, A., & Rovˇcanin, A. (2020). Nowcasting GDP of Bosnia and Herzegovina: A comparison of forecast accuracy models. South East European Journal of Economics and Business, 15(2), 1–14. Aoki, M. (1987). State space modeling of time series. Springer Verlag. Armstrong, J. S., & Fildes, R. (1995). On the selection of error measures for comparisons among forecasting methods. Journal of Forecasting, 14, 67–71.

6 Modelling and Forecasting GDP of Greece with a Modified Exponential. . .

109

Ashley, R. (2003). Statistically significant forecasting improvements: How much out-of-sample data is likely necessary? International Journal of Forecasting, 19, 229–239. Brown, R. G. (1956). Exponential smoothing for predicting demand. Arthur D. Little Inc. Chatfield, C., Koehler, A. B., Ord, J. K., & Snyder, R. D. (2001). A new look at models for exponential smoothing. Journal of the Royal Statistical Society, Series D, 50, 147–159. Chen, Z., & Yang, Y. (2004). Assessing forecast accuracy measures. Iowa State University. Clark, T. E. (1999). Finite-sample properties of tests for equal forecast accuracy. Journal of Forecasting, 18, 489–504. Clements, M. P., & Hendry, D. F. (1998). Forecasting economic time series. Cambridge University Press. Diebold, F. X., & Mariano, R. (1995). Comparing predictive accuracy. Journal of Business, Economic Statistics, 13, 253–265. Dritsaki, C. (2015). Forecasting real GDP rate through econometric models: An empirical study from Greece. Journal of International Business and Economics, 3(1), 13–19. Dritsaki, M., & Dritsaki, C. (2021a). Comparison of the Holt-Winters exponential smoothing method with ARIMA models: Forecasting of GDP per capita in five Balkan countries members of European Union (EU) post COVID. Modern Economy, 12, 1972–1998. Dritsaki, C., & Dritsaki, M. (2021b). Forecasting Greek real GDP based on ARIMA modeling. In Modeling economic growth in contemporary Greece (pp. 45–60). Fair, R. C. (1986). In Z. Griliches & M. D. Intriligator (Eds.), Handbook of econometrics (Vol. III). Elsevier Science Publishers BV. Fildes, R. (1992). The evaluation of extrapolative forecasting methods. International Journal of Forecasting, 8(1), 81–98. Fildes, R., and Steckler H. (2000). The state of macroeconomic forecasting, Lancaster University EC3/99, George Washington University, Center for Economic Research, Discussion Paper No. 99-04. Granger, C., & Newbold, P. (1977). Forecasting economic time series. Academic Press. Gupta, M., & Minai, M. H. (2019). An empirical analysis of forecast performance of the GDP growth in India. Global Business Review, 20(2), 368–386. Harvey, D., Leybourne, S., & Newbold, P. (1997). Testing the equality of prediction mean squared errors. International Journal of Forecasting, 13, 281–291. Holt, C. C. (1957). Forecasting seasonals and trends by exponentially weighted moving averages, ONR Memorandum (Vol. 52). Carnegie Institute of Technology. Available from the Engineering Library, University of Texas at Austin. Holt, C. C. (2004a). Forecasting seasonals and trends by exponentially weighted moving averages. International Journal of Forecasting, 20, 5–10. Holt, C. C. (2004b). Author’s retrospective on ‘forecasting seasonals and trends by exponentially weighted moving averages’. International Journal of Forecasting, 20, 11–13. Hyndman, R. J. (2014). Measuring forecast accuracy. Working Paper. Hyndman, R. J., Koehler, A. B., Snyder, R. D., & Grose, S. (2002). A state space framework for automatic forecasting using exponential smoothing methods. International Journal of Forecasting, 18, 439–454. Hyndman, R. J., Koehler, A. B., Ord, J. K., & Snyder, R. D. (2008). Forecasting with exponential smoothing: The state space approach. Springer-Verlag. Maccarrone, G., Morelli, G., & Spadaccini, S. (2021). GDP forecasting: Machine learning, linear or autoregression? Frontiers in Artificial Intelligence. https://doi.org/10.3389/frai.2021.757864 Makridakis, S., & Hibon, M. (2000). The M3-competition: Results, conclusions and implications. International Journal of Forecasting, 16(4), 451–476. Makridakis, S., Wheelwright, S. C., Hyndman, R. J., & McGee, V. E. (1999). Forecasting methods and application. Willey. Mariano, R. S. (2002). Testing forecast accuracy. In M. P. Clements & D. F. Hendry (Eds.), A companion to economic forecasting (pp. 284–298). Wiley-Blackwell. Newey, W. K., & West, K. D. (1987). A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix. Econometrica, 55, 703–708.

110

M. Dritsaki and C. Dritsaki

Raworth, K. (2017). Doughnut economics: Seven ways to think like a 21st-century, . Taylor, J. W. (2003). Exponential smoothing with a damped multiplicative trend. International Journal of Forecasting, 19, 715–725. Winters, P. R. (1960). Forecasting sales by exponentially weighted moving averages. Management Science, 6, 324–342.

Chapter 7

Last Mile and Blockchain: Opportunities and Challenges Rafael Villa, Marta Serrano, and Tomás García

Abstract The last mile (LM) has become one of the main problems to be solved in large cities. The growing population, the rise of e-commerce and the sustainability aspects demanded by smart cities have brought the urban distribution of goods to the brink of collapse. In this context, where accurate and fast information is critical for correct performance, blockchain technology can offer opportunities to improve urban logistics in cities. This paper conducts a systematic literature review (SLR) of the areas of implementation of blockchain technology in urban logistics. In order to check the degree of maturity, the leading real projects currently being developed are also analysed. This study aims to help practitioners and researchers better understand the issues related to blockchain application in urban logistics, considering the potential benefits and possible limitations of this technology. Keywords Last mile · Blockchain · Urban logistics · Urban delivery innovations.

7.1 Introduction 7.1.1 Last-Mile Issue Last-mile delivery refers to all logistical activities related to delivering consignments to private customers’ homes in urban areas (Boysen et al., 2021). Its relevance is a consequence of the significant challenges facing last-mile logistics. Last-mile logistics is undoubtedly a critical challenge for society, businesses and public authorities living in, operating and managing large cities.

R. Villa () · M. Serrano · T. García School of Technology and Science, Camilo José Cela University, Madrid, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_7

111

112

R. Villa et al.

Firstly, the increase in demand for e-commerce results from the increased urbanisation of cities and changing consumer habits (i). 55% of the world’s people already live in cities, and the urbanisation rate will grow to 68% by 2050 (United Nations, 2018). Urban dwellers are increasingly demanding goods and services online. This has been exacerbated by the Covid-19 pandemic, which has pushed consumers who only shopped in traditional retail outlets to join the majority who already shop online. According to Eurostat (Lone et al., 2021) already, 75% of residents are e-shoppers in EU-27 countries, reaching A C757 billion in 2020, up 10% from A C690 billion in 2019. Concerning environmental sustainability (ii), last-mile delivery already accounts for 15–20% of total urban traffic and a quarter of total CO2 emissions (Cattaruza et al., 2017). Without stakeholder interventions, carbon emissions from urban delivery traffic are expected to increase by 32% by 2030 (World Economic Forum, 2020). Air pollution mortality in large cities is 20 times higher than traffic fatalities (Khomenko et al., 2021). Congestion (iii) caused by the increase in urban parcels implies a much higher number of delivery vans entering city centres, putting additional strain on the existing infrastructure: loading-unloading and storage. Congestion already accounts for 15–20% of road traffic in a typical city (Dablanc, 2011; DGT, 2020) and this percentage is expected to grow in the coming years due to the growth of ecommerce. In London, the most congested city globally, the average driver loses 148 hours per year to traffic (INRIX, 2021). The logistics efficiency and costs associated with urban distribution (iv) erode the profitability of many companies in the transport sector to marginal levels (Fraselle et al., 2021). DUM delivery accounts for 53% of total shipping costs and 41% of total supply chain costs (WEF, 2020). In Spain, around 80% of transport companies are micro-enterprises with very tight margins (Deloitte, 2020). Finally, q-commerce or fast commerce (v) has emerged as a new model within e-commerce where speed, convenience and customer service are paramount. Users value ultra-fast delivery, being able to choose between different delivery methods and being informed of the status of their orders at all times. Q-commerce accentuates the difficulties in urban e-commerce distribution: smaller volumes, more delivery addresses, higher replenishment frequencies, lower inventory levels, low vehicle load optimisation and just-in-time deliveries (Lebeau & Macharis, 2014). All these challenges mean that urban logistics is facing one of the key transformation moments that will shape the way cities function in the coming years.

7.1.2 Blockchain and Applications in the SLR Blockchain can be summarised as a set of technologies that enable a secure, decentralised, synchronised and distributed record of digital transactions without third-party intermediation (Di Pierro, 2017). The blockchain concept was first introduced in 2008 when a person (or group of people) under the pseudonym Satoshi

7 Last Mile and Blockchain: Opportunities and Challenges

113

Nakamoto described a project to create a digital currency (Bitcoin) that would serve to account for and transfer value (Nakamoto, 2009). In his famous article, the author(s) defined the first decentralised ledger: a database that can be shared by a large number of users on a peer-to-peer basis and that allows information to be stored in an immutable and orderly way, to which anyone can write, and which is not controlled by a single or a conglomerate of identities. Blockchain technology proposes a new model in which authenticity is not verified by a third party but by the network of nodes through computers connected to the network within the blockchain. Hence, any transfer of value (be it money or any other asset that has some value) is carried out through a consensus and not through an intermediary, allowing information to be stored transparently (Lim et al., 2021). The blockchain can be classified as private (with permission) and public (without permission). In the private blockchain, only authorised participants can obtain the right to join the network, create transactions, verify blocks and receive transactions. In the public blockchain, everyone has the right to join the network, create, verify and receive transactions. All transactions are publicly disseminated (Aggarwal et al., 2019). Blockchain technology offers significant and diverse benefits for businesses and government institutions. Firstly, security (i), where blockchain allows the creation of a record that cannot be modified and has end-to-end encryption that helps prevent fraud and unauthorised activity. The information is stored on a network of computers rather than on a single server, making it more difficult for hackers to see the information. On the other hand, there is greater transparency (ii) because blockchain uses a distributed log, and transactions and data are recorded identically in multiple locations. All network participants with authorised access see the same information simultaneously, providing complete transparency. Traceability (iii) is a crucial aspect, where the blockchain creates an audit trail that documents the provenance of an asset at every step of its journey. This fact is beneficial in industries where consumers are concerned about environmental or human rights issues surrounding a product or in an industry with counterfeiting and fraud problems. Another advantage is speed and greater efficiency (iv), where transactions can be completed more quickly and efficiently and avoid traditional processes which, apart from being labour-intensive, involve much time-consuming paperwork and sometimes require intermediaries. Finally, automation (v) means that transactions can even be automated with “smart contracts”, which increase their efficiency and further speed up the transaction process. Smart contracts are simply programs stored on a blockchain executed when predetermined conditions are met. They are typically used to automate the execution of an agreement so that all participants can be immediately sure of the outcome without the involvement of any intermediary (Varfolomeev et al., 2021). The beginning of the blockchain is linked to the emergence of the bitcoin cryptocurrency, where its potential in the financial and payment sector has been extensively studied (Alladi et al., 2019). However, this technology allows for diverse applications in different fields (Chen et al., 2022). Among these applications, the following stand out: cloud storage, which allows the creation of nodes in different

114

R. Villa et al.

geographical locations that can withstand the crash of any server (Shah et al., 2020); digital identities where the blockchain provides a single, secure and immutable system that is the optimal solution to the problem of identity theft (Dunphy et al., 2018); voting systems as a new way of approaching democracy, with blockchain technology providing a new framework on which to regulate the voting system (Mohammedali & Al-Sherbaz, 2019) and the management of smart contracts that can be applied to different areas: insurance, payment automation, real estate market and mortgages (Hewa et al., 2021). In general, smart contracts have applications in all areas where traditional contracts are currently signed between two or more parties and where intermediaries can be dispensed. One of the most exciting applications of blockchain can be found in the supply chain (SC), which has become a critical element for the survival and growth of companies worldwide (Halldorsson et al., 2007). In today’s global trade and intense competition, companies seek competitive advantage and compete with each other through highly complex global supply chains (Christopher & Holweg, 2017). Due to this complexity and the lack of transparency of current supply chains, there is much interest in how blockchains could transform the supply chain and logistics industry. The main blockchain applications in transport, logistics and the supply chain can be summarised in the 4Ts (trust, technology, commerce and traceability/transparency; Pournader et al., 2019), and the possibilities are numerous. Within supplier management, technology enables secure and efficient systems for invoices and payments to suppliers without bank intervention or penalties through smart contracts. For example, Tallysticks (http://tallysticks.io/) has developed a blockchain-based platform that handles invoices and payments for logistics and other business activities. On the other hand, in industries such as food, pharmaceuticals or the luxury goods industry, where traceability and originality provide high value to customers, the blockchain makes it possible to know the exact provenance and trace each product back to its origin (Singh & Sharma, 2022). The same is understood by companies such as De Beers Group, which tracks, records and manages the extraction, production and distribution of diamonds around the world; Unilever, Nestlé, Walmart in the field of food traceability or Pfizer, Amgen and Sanofi, which are collaborating in the adoption of this technology to accelerate clinical trials of new drugs (Marr, 2018; Uddin, 2021). Blockchain also offers the possibility to track CO2 emissions and neutralise the carbon footprint within the supply chain: it enables tamper-proof and unalterable reporting of carbon emissions, including what is known as Scope 3 emissions (Upadhyay et al., 2021; Scott et al., 2021). In less digitised industries such as shipping and where container transport is one of the most important aspects of the supply chain, blockchain technology could reduce paperwork and administrative costs (Heilig & Voß, 2018; Laaper et al., 2017; Jabbar & Bjørn, 2018). In the physical flow of goods between exporter and importer, shipping organisations, vessels and officials are forced to manage more than 20 documents (Czachorowski et al., 2019). Blockchain technology could eliminate the remission of paper documents, speed up cargo transfer logistics operations, minimise the risk of customs compliance penalties imposed on customers, and save

7 Last Mile and Blockchain: Opportunities and Challenges

115

the shipping industry hundreds of millions of dollars annually (Heilig & Voß, 2018; Irannezhad, 2020). Within the field of SC, one of the most minor studied aspects in the scientific literature is the possible application of blockchain technology in the final part of the chain, urban logistics. This paper tries to fill this gap and studies the applications that can be given to solve the last-mile problem. Specifically, the main contributions of the study are: • A systematic review of the literature relevant to the research on the areas of implementation of blockchain technology in the field of urban logistics. • Exploration of existing projects, proofs-of-concept and tools being developed to implement blockchain in urban logistics. • Identification of limitations and future research directions. The rest of the paper is structured as follows: in Chap. 2, the methodology used is discussed, followed by a comprehensive literature review to identify the possibilities offered by the blockchain to the last-mile problem, and the real cases that are already in use today. Finally, conclusions are drawn and limitations and possible further studies are pointed out.

7.2 Methodology The main objective of this study is to analyse the possible solutions that blockchain technology can offer to provide solutions to the current problems of urban distribution of goods. Furthermore, to corroborate the degree of actual implementation of the technology, we analyse the blockchain solutions that are being applied in the field of city logistics. Accordingly, two research questions are formulated: H1: What are the main applications and areas of implementation of blockchain technology in urban logistics from a scientific point of view? H2: What are the main real-world applications offered by companies in last mile and blockchain technology?

7.2.1 Review Methodology To answer the first question (H1), an extensive scientific literature review was conducted to identify the leading blockchain solutions researched within the field of city logistics. To answer the second question (H2), an ad hoc search of companies’ solutions in this field was conducted. The review methodology consists of several steps.

116

R. Villa et al.

Fig. 7.1 Review process methodological framework

7.2.1.1

Review of Scientific Papers

Firstly, a systematic literature review (SLR) was carried out. The review methodology consists of several stages (see Fig. 7.1).

Search Strings The following sources were considered for the search for relevant articles: Science Direct, Scopus, Springer, Web of Science and MDPI. The keywords and search strings were as follows: (“Blockchain” AND (“last-mile” OR “city logistics” OR “urban delivery” OR “last-mile delivery” OR “urban logistics” OR “last-mile logistics” OR “urban goods distribution” OR “urban freight”)).

Selection Criteria In order to find the most relevant articles for the review, the following selection criteria were applied. On the one hand, we included all articles that met the following conditions: 1. Articles published in peer-reviewed journals, conference proceedings and articles published in reputable journals 2. Articles written in English. 3. Articles that were published in the period 2018–2022

7 Last Mile and Blockchain: Opportunities and Challenges

117

On the other hand, we excluded posters, prefaces, summaries, book reviews, editorials, readers’ letters, panels, other languages than English and articles prior to 2018.

Quality Assessment of Studies The authors conducted a qualitative review once the inclusion and exclusion criteria were applied in the selected databases. As a first step, the article’s title was reviewed, and all duplicated articles and those not related to the research topic were excluded. Secondly, the abstract was reviewed with the articles that passed the first qualitative filter. Those that did not specifically relate blockchain technology to the concepts of the last mile, city logistics or urban distribution were discarded. Finally, the last filter was to perform a complete reading of all the articles to select those focused on the object of the research. The literature review process and the number of articles selected at each stage are shown in Fig. 7.1.

7.2.1.2

Ad Hoc Search of Companies’ Blockchain Solutions

As blockchain technology is in an early adoption phase in all sectors, the search for the leading solutions of this technology in city logistics focuses on the most relevant startups that have real projects implemented at this article’s publication date. The search was done through Google, Microsoft Bing, Baidu and Yahoo.

7.3 Literature Review on Blockchain City Logistics The literature review focuses on possible applications of blockchain in the last mile (LM). From the total raw result shown by the databases at an initial stage (4907), 21 articles were identified that met all the filters applied. These articles fall into one of the following six application areas (Fig. 7.2): access to funding LM stakeholders, LM stakeholder collaboration, Reduced LM pollution, Traceability LM shipments, Improving the efficiency of LM logistics activities and Customer satisfaction LM. LM Stakeholder Collaboration this application area refers to the different benefits obtained when different stakeholders share information through blockchain technology in different areas of the last mile. Collaboration is usually horizontal and smart contracts are used to store and self-execute actions according to a set of pre-programmed parameters. Possible applications are varied: collaboration between last-mile logistics companies when sharing operations they perform in microhubs (Hribernik et al., 2020); a system involving both drivers and municipal administrations for better management of the traffic system based on LCV data in

118

R. Villa et al.

ARTICLES AND APPLICATION AREAS

8

6

4

2

LM

r

n io ct fa tis sa er m to us

C

to Ac

ce

ss

nc ie fic Im

pr

ov

e

th

e

ef

ol eh ak st LM g in nd fu

of y

ac Tr

de

n po LM uc ed

LM

bi ea

R

lit

lo

y

gi

e

st

LM

ic

s

sh

ip

ac

m

tiv

llu

iti

en

tio

ts

n io ra t bo lla co er ld ho es ak LM

st

es

0

Fig. 7.2 Articles and application areas in blockchain last-mile SLR

the city (Astarita et al., 2020); or to reduce the commission of the last-mile logistics companies (Hribernik et al., 2020), or reducing the commission taken by food delivery companies by creating a peer-to-peer transaction between the restaurant, the customer and the delivery person (Rani & Vishali, 2021). Traceability LM Shipments this section includes studies on traceability and tracking of logistics orders that focus on some actors involved in the urban distribution of goods: Authorities, Retailers, Logistics enterprises or Consumers. The solutions focus on effectively guaranteeing the traceability and transparency of the orders that reach the end customer using blockchain technology. In general terms, traceability adds value to the product and improves the image of the entities, making them more transparent. In addition, it can be supported by IoT-based systems and smart contracts to take into account all necessary data, including the specific case of urban areas, with an open data platform available to shareholders, authorities and consumers (El Midaoui et al., 2021; Cai et al., 2021; Demir et al., ˇ 2019; Cerný et al., 2021). Another promising solution is proof of delivery in online commerce, where Ethereum smart contracts can be used to prove the delivery of a shipped item between a seller and a buyer, regardless of the number of intermediate carriers required (Hasan & Salah, 2018).

7 Last Mile and Blockchain: Opportunities and Challenges

119

Improve the Efficiency of LM Logistics Activities Blockchain technology can help improve freight networks by using radio frequency (UHF-RFID), Internet of Things (IoT) sensors and smart contracts. This provides a fast shipment management architecture that ensures security between parties (Baygin et al., 2022). Other possible applications within logistics activities include traffic planning in city streets considering environmental aspects (Luo et al., 2022) and the possibility of controlling the real-time driving of truck/van drivers in urban environments (Bagloee et al., 2021). For these applications to be feasible and valuable, the information stored needs to be in real time. Reduce LM Pollution Different blockchain applications (https://www.carbonx. ca/; RecycleToCoin) aim to reduce the carbon footprint by promoting green behaviour through the design and use of specialised tokens (Esmaeilian et al., 2020). Blockchain can also be used to securely optimise fuel savings of fleets performing parcel pick-up and delivery (P&D) in urban areas (Anwar et al., 2019) or reduce greenhouse gases in cities through smart contracts and a transportation platform (Li et al., 2019). Access to Funding LM Stakeholders The world of crowdfunding is extensive, and for it to be called a collaborative economy, it must share platforms owned by the people who use it. Blockchain technology automatically distributes power to the entire community of stakeholders. It can enable access to finance for retailers (Li et al., 2020) or small logistics operators in the last mile (Rachana Harish et al., 2021). Customer Satisfaction LM Blockchain-based evaluation approach for customer satisfaction in the context of urban logistics. Four criteria affecting customer satisfaction in urban logistics are identified. A short-term memory machine learning algorithm is adopted to predict customer satisfaction in the future period. A smart contract is designed to compensate and/or refund customers when their satisfaction with delivery services is low (Tian et al., 2021). Table 7.1 summarises the papers and classifies them according to three layers of application of the technology to urban logistics: blockchain, application and user. From a blockchain layer perspective, we have highlighted, in Fig. 7.2, those feature(s) or solution(s) within the blockchain domain that are most prominent in each article. Other features may be present (or even directly related) in the article, but we focus on the most prominent ones. Within this layer, half of the articles refer to the shared ledger as features to be implemented within the blockchain. A shared ledger is a type of database that is shared, replicated and synchronised between members of a decentralised network. A shared ledger within the last-mile records transactions between network participants, such as the exchange of assets or data. Smart contracts are related to the shared ledger, which appears in 9 of the selected articles. Smart contracts are proposed as a possible solution to the lastmile problems, as they are flowcharts of lifelong operational processes that can be programmed, automated and executed without human intervention for the different actors involved in urban logistics. Smart contracts make it possible to work through executable codes on the blockchain to facilitate, effect and enforce an agreement.

Year 2022

2020

2020

2019

2021

2020

2021

2022

2020

2021

Reference (authors) Baygin, et al.

Hribernik et al.

Li et al.

Li et al.

Bagloee et al.

Esmaeilian et al.

ˇ Cerný et al.

Li et al.

Astarita et al.

Rachana Harish et al..

Description of the solution Improving local cargo networks Horizontal collaboration LM logistics operators Access to finance for retailers and small logistics operators Platform to reduce greenhouse gases Real-time driving monitoring of truck/van drivers Promoting green behaviour through designing specialized tokens Tracking a consignment and identifying the origin of the goods Horizontal cooperation between companies to reduce emissions Cooperation between drivers and Authorities to improve traffic flow Access to finance for small logistics operators

Table 7.1 Overview of papers by different types of layers

Smart contracts Tokenization

IoT Shared Ledger

Smart contracts Tokenization

Smart contracts

Cryptography

IoT

Smart contracts P2P

Blockchain layer Smart contracts Tokenization P2P Shared ledger Smart contracts Tokenization

Logistic management system

Logistic management system

Sustainable Logistics

Customer satisfaction management

Sustainable Logistics

Logistic management system

Sustainable Logistics

Logistic application layer Logistic management system Logistic management system Logistic management system

Logistics enterprises

Logistics enterprises Authorities

Logistics enterprises

Retailers Consumers

Retailers Consumers

Logistics enterprises Authorities

Logistics enterprises

Retailers

Logistics enterprises

User layer Logistics enterprises

120 R. Villa et al.

2021

2022

2021

2018

2021

2019 2019

2021

2022

2022

2021

Xia et al.

Li et al.

Rani, & Vishali

Hasan et al.

Tian et al.

Anwar et al. Demir et al.

El Midaoui et al.

Luo et al.

Raj et al.

Cai et al.

Traceability of LM orders

Organisation of logistical activities in LM LM stakeholder collaboration

Traceability of LM orders

Urban logistics customer satisfaction Reduce pollution from LM Traceability of LM orders

Horizontal cooperation between companies to share a drone fleet platform Cooperation between sender, retailer and receiver Cooperation between sender, retailer and receiver Traceability of LM orders

Smart contract Shared Ledger

Shared Ledger IoT P2P Smart contract

Shared Ledger IoT

Shared Ledger Shared Ledger

Shared Ledger

Shared Ledger

Smart contracts P2P

IoT Shared Ledger

IoT Shared Ledger

Customer satisfaction management Customer satisfaction management Sustainable Logistics Logistic management system Customer satisfaction management Logistic management system Customer satisfaction management Logistic management system Logistic management system Customer satisfaction management Logistic management system Customer satisfaction management

Logistic management system

Logistic management system

Logistic management system

Retailers Logistics enterprises Consumers Retailers Logistics enterprises Consumers

Authorities Retailers Logistics enterprises Consumers Authorities

Retailers Logistics enterprises Consumers Retailers Logistics enterprises Consumers Retailers Consumers Logistics enterprises Consumers Logistics enterprises Retailers Logistics enterprises Consumers

Logistics enterprises

7 Last Mile and Blockchain: Opportunities and Challenges 121

122

R. Villa et al.

Fig. 7.3 Most used solutions in the field of blockchain–last mile

They are closely related to peer-to-peer transactions (P2P), where communication between applications allows individuals to communicate and share information with other individuals without the need for a central server to facilitate communication (Fig. 7.3). Considering the logistic application layer, it is worth noting that the logistic management system is addressed in 14 of the 21 articles studied. It is logical to think that the first application sought is how to improve the efficiency of logistics processes in the last mile. In a sector with low margins, there is much pressure to reduce costs or to be more productive in deliveries in the city. In this context, blockchain technology can improve collaboration between LM players and improve access to finance for smaller operators. Similarly, customer satisfaction is another critical aspect for logistics operators operating in the last mile. Blockchain can substantially improve the traceability of products and the tracking of online orders produced by city residents. With the right service design and other technologies such as IoT, supply chains should focus on obtaining instant proof of delivery (PoD) confirmations. Achieving a transparent supply chain that provides real-time information has become an opportunity for retailers and logistics companies. Finally, from an LM stakeholder perspective, logistics enterprises would be the main beneficiaries of the possible applications of blockchain technology. These companies, as explained above, seek to improve their processes and make them more efficient. At the LM end, consumers and retailers also appear as potential beneficiaries. Lastly, there are the Authorities that focus their possible applications on sustainability and making cities more liveable (Fig. 7.4). It is important to note that all the analysed articles except (Esmaeilian et al., 2020) study the possibilities of blockchain technology from a theoretical perspective. This is why it is necessary to complement the systematic analysis of the literature with a search for real technology projects in urban distribution.

7 Last Mile and Blockchain: Opportunities and Challenges

123

Fig. 7.4 Beneficiaries of blockchain in last-mile solutions

7.4 Real-World Applications and Use Cases of the Blockchain in the Last-Mile Problem The scientific literature has extensively addressed the many possibilities of blockchain technology in the supply chain domain. In this paper, we have focused on its possible applications in one part of SCM, LM, where scientific articles are, on the contrary, very scarce. In order to determine the degree of actual implementation, a search for relevant business cases where blockchain technology is used in LM has been carried out.

7.4.1 Real-World Use Cases of Blockchain in Last-Mile Logistics Very few real cases of blockchain implementation in LM have been found. The results in the different search engines were the same or very similar (in the case of Baidu, the results were more difficult to find as it is a Chinese search engine). Table 7.2 shows the results found that meet the following criteria: (i) they are real projects and are relevant, (ii) they are within the scope of LM and (iii) they started at least in 2020 and are active in 2022. In Table 7.2, the project name, scope, promoter, stakeholder involved, functionality and link are detailed. As can be seen in Table 7.2, the actual projects found are scarce. Despite the significant number of results displayed by the search engines, only 8 projects were found that meet the defined criteria. In addition, a search was carried out in Google trends, and the search did not provide enough results, which is a clear indicator

Scope Fast delivery

Secure payments to urban carriers

Financing logistics operators and retailers

Postal services

Name VOLT

DL Freight

AntChain

Casemail

USPS

Alibaba

Walmart Canada

Promoter VOLT

Postal operators Consumers

Retailers Carriers

Retailers Carriers

Stakeholders involved Retailers Carriers Consumers

Table 7.2 Overview of real-world use case of blockchain last mile Functionalities The company’s core values are “decentralisation”, “transparency” and “security”. Volt enables same-day deliveries in cities quickly and reliably with blockchain-based cloud computing, AI and smart contracts. The system tracks deliveries, verifies transactions, and automates invoices in real time. Walmart and its fleet of carriers are all working off the same information and calculations by way of smart contracts. The app simplifies tax, customs and shipping, and allows banks to complete payment instantly, reducing audit costs and the risk of non-payment. Nearly 20 global banks financing through the platform. Producer of ePostage labels, generated on blockchain, which use NFT tokens to seal, encrypt and record documents on the blockchain in a unique and highly secure way

https://casemail.us/blockchain-postage/

https://antchain.net/

https://www.dltlabs.com

Link https://volttech.io

124 R. Villa et al.

Delivery management software

Freight Tracking

QUASA

Vehicle-tovehicle (V2V) communications Reducing borrowing costs in LM

Yojee

Sweetbridge

Peloton

QUASA

Yojee

Sweetbridge

Deadpooled

Carriers Retailers

Small logistics companies

Carriers

Carriers

Enable multiple freight vehicles to platoon and communicate, improving system efficiency and safety Solution to many of the existing problems in the freight and trucking industry, such as supply chain liquidity, supply chain operations and supply chain flexibility. Help logistics companies coordinate fleet coordination, track cargo, billing, manage jobs, and evaluate driver qualifications Develops an open blockchain platform for cargo transportation and provides a tracking system that customers use to see the status and position of the shipments

https://quasa.io/

https://yojee.com/

https://sweetbridge.com/

http://pelotonblockchain.com/

7 Last Mile and Blockchain: Opportunities and Challenges 125

126

R. Villa et al.

of the current lack of relevance when both concepts (blockchain + last mile) are searched together. Among the results found are numerous blockchain projects in the testing phase or other areas of the supply chain. These projects generally combine other types of technology (IoT and AI) with blockchain. The main players in this market are Amazon, Accenture, Capgemini, Infosys, IBM, Microsoft, Oracle, SAP, Tata Services and Wipro (Yahoo Finance., 2022). The key market trend of blockchain technology in the transportation and logistics industry market is the advent of blockchain as a service: BaaS (Song et al., 2021). Within the LM arena, Amazon has recently filed a patent for blockchain-based technology that tracks goods as they move through the supply chain. Amazon’s patent describes a “distributed ledger certification” system. The patent document explains that the blockchain-based tracking solution would ensure that consumer goods sold on its e-commerce site are authentic (https://www.logisticsinsider.in/ amazon-patents-blockchain-based-technology-to-track-its-shipments/). In the public sphere, the European H2020 project TOKEN assesses the impact of blockchain technologies on short food chain distribution in the context of smart cities (https:// token-project.eu/). This project aims to analyse and test the failures and opportunities that can be solved before blockchain is used and disseminated socially in different services of public interest.

7.5 Conclusions Blockchain is a concept that poses a huge revolution and will transform the global economy. Its main application is the elimination of intermediaries, decentralising all management in a secure and synchronised way. It is vital because of the enormous amount of transactions (information) managed in all areas of our society today. Nevertheless, this revolution has applications not only in our economy but in a multitude of sectors. One of the most promising is the supply chain, where blockchain technology can help solve problems of traceability, lack of trust between operators, high bureaucratic costs and traceability of global supply chains. The back end of the supply chain, the last mile of city logistics, is becoming critical to absorb the growth of e-commerce that has been increasing since its emergence and has been further accelerated by the pandemic. The main objective of last-mile logistics is to deliver the order to the customer as quickly as possible and at the lowest possible cost. The problem is that, although customers want fast deliveries at no cost, this process is highly costly, amounting to as much as 53% of the total cost of shipments (WEF, 2020). Therefore, every inefficiency in the process significantly increases delivery costs and reduces the business’s profitability. The main challenges facing the last mile are high delivery costs, lack of transparency throughout the process, inefficient route optimisation, fragmented and low-value customer orders, improving delivery traceability, and providing tools that favour collaboration between the different stakeholders. In the face of these challenges, a

7 Last Mile and Blockchain: Opportunities and Challenges

127

priori, the theoretical benefits of blockchain technology can help solve, totally or partially, most of the problems faced by distribution in the last mile. This article has conducted a systematic review of the scientific literature and searched for real applications and use cases of blockchain technology applied to the last-mile problem. The results are scarce both in the scientific and ad hoc reviews. This fact shows that, although the technology is seen as promising for the future, it is currently discussed in the scientific arena in a testimonial way, and few real applications are being developed. The authors believe that, like other applications of blockchain in the global supply chain (elimination of administrative procedures, traceability, originality of products or cooperation between agents) mature, it could be transferred to the field of urban distribution of goods. Responding to H1, in the scientific field, of all the articles selected, the application area with the greatest coverage is LM stakeholder collaboration, which is considered essential in the field of LM due to the multitude of stakeholders involved and having different interests (retailers, couriers, public authorities, consumers). Order traceability appears as the second most studied area of application. It arises due to the need for the entire supply chain (from the factory to delivery) to be visible to the consumer. Added to this aspect is the importance of the certificate of originality that blockchain technology can bring to marketed products. From another perspective, although it has little coverage in the scientific literature, the possible applications that blockchain can offer in LM sustainability (carbon footprint certifications per delivered package, for example) seem promising. Considering the different stakeholders, logistics companies are the primary beneficiaries of blockchain solutions in LM. This is followed by consumers, retailers and, finally, public authorities. This fact may condition who will be the future leading promoters of these innovations. However, the technology allows for a part of a collaboration that can foster the inclusion of the different stakeholders in LM. In the blockchain layer, shared ledger and smart contracts are the most studied solutions in the field of LM. Both solutions make it possible to store and use data that can be decentralised (stored in several places) and distributed (connected and therefore able to communicate) privately or publicly. They are closely related to the large amount of information generated and the possibility of favouring transactions without organisations that centralise this information. Considering the different stakeholders, logistics companies are the primary beneficiaries of blockchain solutions in LM. This is followed by consumers, retailers and, finally, public authorities. This fact may condition who will be the future leading promoters of these innovations. However, the technology allows for a part of a collaboration that can foster the inclusion of the different stakeholders in LM. In the blockchain layer, shared ledger and smart contracts are the most studied solutions in the field of LM. Both solutions make it possible to store and use data that can be decentralised (stored in several places) and distributed (connected and therefore able to communicate) privately or publicly. They are closely related to the large amount of information generated and the possibility of favouring transactions without organisations that centralise this information.

128

R. Villa et al.

In response to H2, concerning real use cases already implemented, it is challenging to find consolidated and relevant projects. In this research, to the best of our knowledge, only eight real projects have been found that are currently in operation. The areas of application are similar to those already discussed above in SLR, with the addition of access to finance as another possible area of application for logistics operators. These data are in line with the results of the SLR, which corroborates the limited application of blockchain technology in the field of LM. Therefore, blockchain technology is considered at an early adoption stage, and there is still a long way to go. Blockchain applications are expected to first focus on the global supply chain and eventually reach all links and stakeholders in the supply chain at later stages. In the last phase of SCM, due to its complexity and the significant challenges faced by LM, there are many possibilities for blockchain technology to help improve the performance of LM in its three main areas: economic, social and environmental, and the multiple actors and intermediaries involved: manufacturers, wholesalers, distributors, suppliers, retailers, vendors and consumers.

References Aggarwal, S., Chaudhary, R., Aujla, G. S., Kumar, N., Choo, K.-K. R., & Zomaya, A. Y. (2019). Blockchain for smart communities: Applications, challenges and opportunities. Journal of Network and Computer Applications, 144, 13–48. https://doi.org/10.1016/j.jnca.2019.06.018 Alladi, T., Chamola, V., Parizi, R. M., & Choo, K.-K. R. (2019). Blockchain applications for industry 4.0 and industrial iot: A review. IEEE Access, 7, 176935–176951. https://doi.org/ 10.1109/ACCESS.2019.2956748 Anwar, H., Arasu, M., & Ahmed, Q. (2019). Ensuring fuel economy performance of commercial vehicle fleets using blockchain technology. 2019-01-1078. https://doi.org/10.4271/2019-011078. Astarita, V., Pasquale Giofrè, V., Guido, G., & Vitale, A. (2020). The use of a blockchainbased system in traffic operations to promote cooperation among connected vehicles. Procedia Computer Science, 177, 220–226. https://doi.org/10.1016/j.procs.2020.10.031 Bagloee, S. A., Heshmati, M., Dia, H., Ghaderi, H., Pettit, C., & Asadi, M. (2021). Blockchain: The operating system of smart cities. Cities, 112, 103104. https://doi.org/10.1016/ j.cities.2021.103104 Baygin, M., Yaman, O., Baygin, N., & Karakose, M. (2022). A blockchain-based approach to smart cargo transportation using UHF RFID. Expert Systems with Applications, 188, 116030. https:// doi.org/10.1016/j.eswa.2021.116030 Boysen, N., Fedtke, S., & Schwerdfeger, S. (2021). Last-mile delivery concepts: A survey from an operational research perspective. OR Spectrum, 43, 1–58. https://doi.org/10.1007/s00291-02000607-8 Cai, Z., Yang, M., Cui, H., Li, X., & Xu, M. (2021). Research and design of blockchain system for logistic tracing. 2021 The 4th international conference on information science and systems, pp. 28–34. https://doi.org/10.1145/3459955.3460596 Cattaruza, D., Abse, N., Feillet, D., & Gonzalez-Feliu, J. (2017). Vehicle routing problems for city logistics. Euro Journal on Transportation and Logistics, 6(1), 51–79. https://doi.org/10.1007/ s13676-014-0074-0 Chen, Y., Lu, Y., Bulysheva, L., & Kataev, M. Y. (2022). Applications of blockchain in industry 4. 0: A review. Information Systems Frontiers.https://doi.org/10.1007/s10796-022-10248-7

7 Last Mile and Blockchain: Opportunities and Challenges

129

ˇ Cerný, M., Gogola, M., Kubal’ák, S., & Ondruš, J. (2021). Blockchain technology as a new driver in supply chain. Transportation Research Procedia, 55, 299–306. https://doi.org/10.1016/ j.trpro.2021.06.034 Christopher, M., & Holweg, M. (2017). Supply chain 2.0 revisited: A framework for managing volatility induced risk in the supply chain. International Journal of Physical Distribution and Logistics Management, 47(1), 2–17. Czachorowski, K., Solesvik, M., & Kondratenko, Y. (2019). The application of Blockchain technology in the maritime industry. In V. Kharchenko et al. (Eds.), Green IT engineering: Social, business and industrial applications, studies in systems, decision and control 171 (pp. 561–577). Springer Nature Switzerland. Dablanc, L. (2011). City distribution, a key element of the urban economy: Guidelines for practitioners. In C. Macharis & S. Melo (Eds.), City distribution and urban freight transport (pp. 13–36). Edward Elgar. Deloitte. (2020). https://www2.deloitte.com/content/dam/Deloitte/es/Documents/operaciones/ deloitte-es-operations-last-mile.pdf Demir, M., Turetken, O., & Ferwom, A. (2019). Blockchain and iot for delivery assurance on supply chain(Bidas). 2019 IEEE international conference on Big Data (Big Data), pp. 5213– 5222. https://doi.org/10.1109/BigData47090.2019.9006277. DGT. (2020). Evolución de tráfico por el efecto de COVID-19. Evolucion-Intensidades-0706-2020.pdf (dgt.es)DHL Express inicia entregas con drones urbanos automatizados en China. (2019, 17 de mayo). El Mercantil.https://elmercantil.com/2019/05/17/dhl-expresslanza-entregas-con-drones-urbanos-automatizados-en-china Di Pierro, M. (2017). What is the blockchain? Computing in Science & Engineering, 19(5), 92–95. https://doi.org/10.1109/MCSE.2017.3421554 Dunphy, P., Garratt, L., & Petitcolas, F. (2018). Decentralizing digital identity: Open challenges for distributed ledgers. 2018 IEEE European symposium on security and privacy workshops (EuroS&PW), pp. 75–78. https://doi.org/10.1109/EuroSPW.2018.00016. Esmaeilian, B., Sarkis, J., Lewis, K., & Behdad, S. (2020). Blockchain for the future of sustainable supply chain management in Industry 4.0. Resources, Conservation and Recycling, 163, 105064. https://doi.org/10.1016/j.resconrec.2020.105064 El Midaoui, M., Ben Laoula, E. M., Qbadou, M., & Mansouri, K. (2021). Logistics tracking system based on decentralized IoT and blockchain platform. Indonesian Journal of Electrical Engineering and Computer Science, 23(1), 421. https://doi.org/10.11591/ijeecs.v23.i1.pp421430 Fraselle, J., Limbourg, S. L., & Vidal, L. (2021). Cost and environmental impacts of a mixed fleet of vehicles. Sustainability, 13(16), 9413. https://doi.org/10.3390/su13169413 Halldorsson, A., Kotzab, H., Mikkola, J., & Skott-Larsen, T. (2007). Complementary theories to supply chain management. Supply Chain Management: An International Journal, 12(4), 284– 296. https://doi.org/10.1108/13598540710759808 Author, F., Author, S., Author, T.: Book title. 2nd edn. Publisher, Location (1999). Hasan, H. R., & Salah, K. (2018). Blockchain-based proof of delivery of physical assets with single and multiple transporters. IEEE Access, 6, 46781–46793. https://doi.org/10.1109/ ACCESS.2018.2866512 Heilig, L., & Voß, S. (2018). The intelligent supply chain: From vision to reality. Port Technology, 78, 80–82. Hewa, T., Ylianttila, M., & Liyanage, M. (2021). Survey on blockchain based smart contracts: Applications, opportunities and challenges. Journal of Network and Computer Applications, 177, 102857. https://doi.org/10.1016/j.jnca.2020.102857 Hribernik, M., Zero, K., Kummer, S., & Herold, D. M. (2020). City logistics: Towards a blockchain decision framework for collaborative parcel deliveries in micro-hubs. Transportation Research Interdisciplinary Perspectives, 8, 100274. https://doi.org/10.1016/j.trip.2020.100274 Irannezhad, E. (2020). Is blockchain a solution for logistics and freight transportation problems? Transportation Research Procedia, 48(2020), 290–306.

130

R. Villa et al.

INRIX. (s. f.). INRIX 2021 Global Traffic Scorecard: As lockdowns ease UK city centres show signs of return to 2019 levels of congestion. Inrix. Recuperado 13 de junio de 2022, de https:// inrix.com/press-releases/2021-traffic-scorecard-uk/ Jabbar, K., & Bjørn, P. (2018). Infrastructural grind: Introducing Blockchain Technology in the Shipping Domain. In Proceedings of the 2018 ACM conference on supporting Groupwork (GROUP ’18) (pp. 297–308). Association for Computing Machinery. https://doi.org/10.1145/ 3148330.3148345 Khomenko, S., Cirach, M., Pereira-Barboza, E., Mueller, N., Barrera-Gómez, J., Rojas-Rueda, D., de Hoogh, K., Hoek, G., & Nieuwenhuijsen, M. (2021). Premature mortality due to air pollution in European cities: A health impact assessment. The Lancet Planetary Health, 5(3), e121–e134. https://doi.org/10.1016/S2542-5196(20)30272-2 Laaper, S., Fitzgerald, J., Quansney, E., Yeh, W. & Basir, M. (2017). Using blockchain to drive supply chain innovation. Deloitte New York. Retrieved from: https://www2.deloitte.com/ content/dam/Deloitte/us/Documents/process-and-operations/us-blockchainto-drive-supplychain-innovation.pdf. Last Accessed 1 May 2022. Lebeau, P., & Macharis, C. (2014). Freight transport in Brussels and its impact on road traffic? Brussels Studies, 80. https://doi.org/10.4000/brussels.1239 Li, Y., Yang, W., He, P., Chen, C., & Wang, X. (2019). Design and management of a distributed hybrid energy system through smart contract and blockchain. Applied Energy, 248, 390–405. https://doi.org/10.1016/j.apenergy.2019.04.132 Li, M., Shao, S., Ye, Q., Xu, G., & Huang, G. Q. (2020). Blockchain-enabled logistics finance execution platform for capital-constrained E-commerce retail. Robotics and Computer-Integrated Manufacturing, 65, 101962. https://doi.org/10.1016/j.rcim.2020.101962 Lim, M. K., Li, Y., Wang, C., & Tseng, M.-L. (2021). A literature review of blockchain technology applications in supply chains: A comprehensive analysis of themes, methodologies and industries. Computers & Industrial Engineering, 154, 107133. https://doi.org/10.1016/ j.cie.2021.107133 Lone, S., Harboul, N., & Weltevreden, J. W. J. (2021). 2021 European E-commerce report. Amsterdam University of Applied Sciences & Ecommerce Europe. Luo, X., Liao, R., Wang, Y., & Zhan, W. (2022). Dynamic distribution and planning for traffic flow of the urban ecological road network based on blockchain technology. International Journal of Information Technology and Management, 21(1), 63. https://doi.org/10.1504/ IJITM.2022.121338 Marr, B. (2018). How blockchain will transform the supply chain and logistics industry. Forbes. Recuperado 13 de junio de 2022, de https://www.forbes.com/sites/bernardmarr/2018/03/23/ how-blockchain-will-transform-the-supply-chain-and-logistics-industry/ Mohammedali, N., & Al-Sherbaz, A. (2019). Election system based on blockchain technology. International Journal of Computer Science and Information Technology, 11(5), 13–31. https:// doi.org/10.5121/ijcsit.2019.11502 Nakamoto, S. (2009). Bitcoin: A peer-to-peer electronic cash system. http://www.bitcoin.org/ bitcoin.pdf Pournader, M., Shi, Y., Seuring, S., & Koh, S. L. (2019). Blockchain applications in supply chains, transport and logistics: A systematic review of the literature. International Journal of Production Research, 58(7), 2063–2081. Rachana Harish, A., Liu, X. L., Zhong, R. Y., & Huang, G. Q. (2021). Log-flock: A blockchain-enabled platform for digital asset valuation and risk assessment in E-commerce logistics financing. Computers & Industrial Engineering, 151, 107001. https://doi.org/10.1016/ j.cie.2020.107001 Rani, K. S., & Vishali, S. (2021). Blockchain driven iot based delish2go decentralized food delivery application. 2021 international conference on artificial intelligence and smart systems (ICAIS), pp. 1727–1732. https://doi.org/10.1109/ICAIS50930.2021.9395873 Shah, M., Shaikh, M., Mishra, V., & Tuscano, G. (2020). Decentralized cloud storage using Blockchain. 2020 4th international conference on trends in electronics and informatics (ICOEI)(48184), pp. 384–389. https://doi.org/10.1109/ICOEI48184.2020.9143004

7 Last Mile and Blockchain: Opportunities and Challenges

131

Scott, D. J., Broyd, T., & Ma, L. (2021). Exploratory literature review of blockchain in the construction industry. Automation in Construction, 132, 103914. https://doi.org/10.1016/ j.autcon.2021.103914 Singh, V., & Sharma, S. K. (2022). Application of blockchain technology in shaping the future of food industry based on transparency and consumer trust. Journal of Food Science and Technology.https://doi.org/10.1007/s13197-022-05360-0 Song, J., Zhang, P., Alkubati, M., Bao, Y., & Yu, G. (2021). Research advances on blockchain-asa-service: Architectures, applications and challenges. Digital Communications and Networks, S2352864821000092. https://doi.org/10.1016/j.dcan.2021.02.001 Tian, Z., Zhong, R. Y., Vatankhah Barenji, A., Wang, Y. T., Li, Z., & Rong, Y. (2021). A blockchain-based evaluation approach for customer delivery satisfaction in sustainable urban logistics. International Journal of Production Research, 59(7), 2229–2249. https://doi.org/ 10.1080/00207543.2020.1809733 Uddin, M. (2021). Blockchain Medledger: Hyperledger fabric enabled drug traceability system for counterfeit drugs in pharmaceutical industry. International Journal of Pharmaceutics, 597, 120235. https://doi.org/10.1016/j.ijpharm.2021.120235 United Nations. (2018). 68% of the world population projected to live in urban areas by 2050. News of Department of Economic and Social Affairs in 16th may 2018 at https://www.un.org/ development/desa/en/news/population/2018-revision-of-world-urbanization-prospects.html Upadhyay, A., Mukhuty, S., Kumar, V., & Kazancoglu, Y. (2021). Blockchain technology and the circular economy: Implications for sustainability and social responsibility. Journal of Cleaner Production, 293, 126130. https://doi.org/10.1016/j.jclepro.2021.126130 Varfolomeev, A. A., Alfarhani, L. H., & Oleiwi, Z. C. (2021). Secure-reliable smart contract applications based blockchain technology in smart cities environment. Procedia Computer Science, 186, 669–676. https://doi.org/10.1016/j.procs.2021.04.188 Yahoo finance. (2022). Blockchain technology market in transportation and logistics industry: North America to occupy 55% market share | sea segment to be significant for revenue generation | technavio. (s. f.). Recuperado 14 de junio de 2022, de https://finance.yahoo.com/ news/blockchain-technology-market-transportation-logistics-070000887.html WEF. (2020). The future of the last-mile ecosystem. https://www3.weforum.org/docs/ WEF_Future_of_the_last_mile_ecosystem.pdf

Chapter 8

Total Factor Productivity and Entrepreneurship: Creative Self-Destruction Jose Maria Sevilla Llewellyn-Jones

Abstract The lack of evidence in literature of the relationship between entrepreneurship and economic growth is a known fact as data is limited. The present paper addresses this absence by taking into consideration the economic assumptions from Schumpeter’s theory of creative destruction into a new approach to the relationship between entrepreneurship and total factor productivity (TFP). Using the data from a single country, the United States, a comparative time period analysis is done by adding a business establishment exit rate to a TFP growth regression based on Solow’s neoclassical Cobb-Douglas production function. A significant and positive relationship is found between TFP growth and a business establishment exit rate in a stagnated innovation period (2009–2019), whilst the effects are significant and negative in a highly disruptive innovative one (1995– 2005). Keywords Entrepreneurship · Schumpeter · Productivity · Panel data

8.1 Introduction One of the fundamental reasons why the study of the relationship between entrepreneurship and economic growth and development is not abundant (Carree et al., 2007; van Praag & van Stel, 2013; Prieger et al., 2016; amongst others) is due to the limitation of available data (Bleaney & Nishiyama, 2002; Erken et al., 2018). Considering this situation, literature has studied productivity as one of the most important explanatory components of economic growth with some relevant theoretical and empirical research (Solow, 1957; Lucas, 1988; Romer, 1986, 1990;

J. M. S. Llewellyn-Jones () CEU International Doctoral School (CEINDO), Doctorate Program in Law and Economics, Madrid, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_8

133

134

J. M. S. Llewellyn-Jones

Grossman & Helpman, 1991; Klenow & Rodríguez-Clare, 1997; Hulten & Isaksson, 2007; Yalçınkaya et al., 2017; amongst others). One of the aims of the present paper is to study productivity using a variable known as total factor productivity (TFP). By taking the Solow-Swan model as a Cobb-Douglas production function, TFP is understood as a residual that refers to the output growth that goes beyond the combined effect of the production factors, traditionally related to technology (Solow, 1957). When TFP is transformed into a growth variable, it is acknowledged as ‘the best available measure of the underlying pace of innovation and technological change’ (Gordon, 2015; p. 54). Hence, the idea of a direct relationship with the Schumpeterian school of thought is assumed for our research. By considering the Schumpeterian’s entrepreneur as the vehicle of introduction of new innovations into the economy (Schumpeter, 1934), we suggest the idea of the process of creative destruction (Schumpeter, 1942) as a TFP growth driver. However, there is a limitation when explaining innovative entrepreneurship, as the definition itself is wide and so are its measures. The main consideration to take into account is that innovations are introduced into the economy by Mark 1 and Mark 2 types of Schumpeterian entrepreneurship. This means that not only new innovative businesses or start-ups do this process, Mark 1; but so do incumbents when they apply intrapreneurship, Mark 2 (Malerba & Orsenigo, 1996). Therefore, a precise measurement is needed from a macro perspective to assess innovative entrepreneurship within an economy. As a solution, we propose an innovative methodology by a comparative economic analysis between two time periods. By following Gordon’s (2015) approach to TFP growth, we are considering a period of technological disruption by a high Average Annual Rate of TFP growth. For comparative purposes to support our analysis, we are going to describe a stagnated innovation period too. Assuming that technological innovation improves capital productivity, we expect to observe this effect in the Capital Services growth as an explanatory variable of TFP growth.1 Taking into consideration, as previously noted, that innovations are introduced using entrepreneurship as a vehicle, an improved capital productivity would mean that the introduction of new technological innovations happened within a period as a consequence of an intense entrepreneurial process. As a result, we would have proven the existence of Schumpeterian innovative entrepreneurship measured by capital productivity when a period of technological change occurs. Hence, we establish our first hypothesis: Hypothesis 1: An increase in the productivity of Capital Services improves TFP growth through a period of technological change.

1 Capital

Services growth data is offered by the Bureau of Labor Statistics (BLS) (2022) as an explanatory variable of TFP growth. For more detail about the relationship between Capital Services and TFP growth within a period of technological change, please read Jorgenson and Stiroh (1999).

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

135

If our first hypothesis is confirmed, we need to know what happens from a firmlevel perspective to study if the business fabric from the supply side improves in terms of productivity. To solve this problem, we propose to follow Schumpeter’s creative destruction theoretical concept, understanding that the least productive companies exit the economy when productivity is increased (Schumpeter, 1942). Accordingly, we assume that the improvement of the industrial productive fabric can be measured with a business establishment exit rate in relation to TFP growth. Therefore, we hypothesise that: Hypothesis 2: An increase in the business establishment’s exit rate leads to an increase in TFP growth in a period of technological change. To prove the above, we consider an econometric study with a TFP growth regression model based on Solow’s work (1957) and following Baily et al.’s (1992) approach. Within the proposed regression, an exit rate of business establishments is contemplated as an entrepreneurship variable to explain TFP growth, as suggested by Erken et al. (2018). Conceiving the proposed methodology as a pioneer, a further step is taken by using the data from just a single country, the United States (USA). This will allow us to analyse and indicate some specific public policy advice for the developed and entrepreneurial economies. Therefore, not only new theoretical but also new empirical contributions are presented to the current TFP research by the present paper, with our main goal being to shed some light to the nexus between economic growth and entrepreneurship from a productivity perspective. The following section introduces the theory that underpins the present work. Section 8.3 describes the used methodology and data. In Sect. 8.4 the regression model is explained for the statistical analysis. Empirical results and suggested economic interpretations are presented in Sect. 8.5, whilst Sect. 8.6 covers the conclusion.

8.2 Literature Review 8.2.1 Entrepreneurship and Economic Growth The Schumpeterian entrepreneurial process is conceived as a continuous search of new combinations in the pursuit of a profit achievement that originates disruption changes within the market structure (Schumpeter, 1934; Baumol, 2002). The process itself considers that the caused disturbances of the economic equilibrium are originated by the destruction of the business fabric whilst innovations enter the economic system (Schumpeter, 1942). Related theoretical and empirical research is scarce as, from a macroeconomic perspective, the wide dimension of the concept makes it difficult to obtain entrepreneurship data at the national level to be linked to national growth in terms of output, productivity or wealth (Wong et al., 2005).

136

J. M. S. Llewellyn-Jones

Table 8.1 List of authors related to the Schumpeterian growth theory Related research to the Schumpeterian growth theory

Aghion and Howitt (1992), Wennekers and Thurik (1999), Audretsch (1997), Audretsch and Fritsch (2002), Carree et al. (2002), Agarwal et al. (2005), Sternberg and Wennekers (2005), Wong et al. (2005), Agarwal et al. (2007), Carree et al. (2007), Carree and Thurik (2008), Dabkowski (2011), van Oort and Bosma (2013), Audretsch et al. (2015), Aubry et al. (2015), Castaño-Martinez, Méndez-Picazo and Galindo-Martín (2015), Low and Isserman (2015), Aparicio, Urbano and Gómez (2016), Erken et al. (2018), Lafuente et al. (2020)

Source: Elaborated from Urbano et al. (2019, p. 34)

Considering this problem, an important step forward was done by endogenising economic growth to innovation by the Schumpeterian growth theory (Aghion & Howitt, 1992). The fundamentals were based on the neoclassical publications of Romer (1986), Lucas (1988) and Judd (1985), who observed the endogenisation of knowledge to explain economic growth. As obsolescence and uncertainty was not studied in the former research, both economists Aghion and Howitt had to complement their findings with other investigations. In this way, it was through the work carried out by King and Rebelo (1988) that they managed to elaborate an endogenous growth model to analyse capital accumulation. Their findings made us to consider three important assumptions for this paper: (i) economic growth is a result of innovation; (ii) innovation is the result of entrepreneurial investments with the aim of having a dominant or monopolistic position in the market; and (iii) innovation replaces old technologies in a direct reference to creative destruction (Aghion et al., 2013). As a consequence of the research initiated by Aghion and Howitt in 1992, followup investigations continued with this new theoretical basis, taking into consideration the inclusion of innovation into endogenous economic growth models. A list of authors suggested by Urbano et al. (2019, p. 34) is presented in Table 8.1.

8.2.2 TFP and Entrepreneurship For the industrial analysis purpose of this study, technical change can be acknowledged by the neoclassical Cobb-Douglas production function approach initially made by Solow (1957). In a similar manner, the following equation explains the relationship between technological change, inputs and output as per the literature review work done by Syverson (2011; p. 331): f (L, K, M) = Y = ALα K β M γ

(8.1)

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

137

where Y denotes Total Production, K is Capital Input, L is Labour Input and M is Intermediate Materials. The output elasticity of the Labour Input is represented by α, β is the elasticity of the Capital Input, and γ is the one for the Intermediate Materials. The term A is a Hicks-neutral shifter conceived as the portion of output not explained by the inputs used in production, a residual that can be understood as a productivity variable associated with the level of technical change that is interpreted as TFP. As in Prescott’s (1997) paper, entitled Needed: A Theory of Total Factor Productivity, when trying to explain the difference in TFP rates between countries, the author concluded that TFP is not only described by technological knowledge. Since then, there has been some literature trying to understand TFP and its explanatory variables (Isaksson, 2007). Taking into account that our aim is to study the relationship of a wide concept of entrepreneurship and TFP, our review focuses on this aspect. Based on methodology, previous research generally considers crosscountry regression analysis for this purpose (Dabkowski, 2011; Erken et al., 2018; amongst others), whilst there is a lack of studying a single country’s economic data. As mentioned, the reason is that there is a special difficulty in formally finding variables to properly elaborate a statistical analysis. Special attention is devoted to Baily et al.’s (1992) and Foster et al.’s (2008) work, as they consider studying demand and price as drivers of TFP in the US economy (Hall, 1988, 1989). The former research comprised the analysis of the US manufacturing industry productivity’s dynamics by computing TFP growth. This investigation is important for us, as the empirical findings of these economists regard plant-level disparities between price and marginal cost in the US industry that are consistent with monopolistic competition. In the latter mentioned paper, we can find how demand factors affect TFP dynamics, where producer-level quantities and prices drive market selection.

8.3 Methodology and Data In this section we identify TFP and TFP growth following Baily, Hulten and Campbell’s approach. Furthermore, firm-level data is considered as proposed by Erken et al. (2018). The purpose is to suggest a TFP growth relationship to the process of creative destruction by the comparison between two time periods, one characterised by a notorious technological change with high TFP growth and another one stagnated with low TFP growth.

8.3.1 Total Factor Productivity Growth In the first place, we are explaining technological change of an industry i at a time t following Eq. 8.1:

138

J. M. S. Llewellyn-Jones

f (Lit , Kit ) = Yit = Ait Lit αit Kit βit Mit γit

(8.2)

For both mathematical calculation and economic interpretation of both TFP and TFP growth, we transform Eq. 8.1 to a log-log linear function: ln Yit = ln Ait + αit ln Lit + βit ln Kit + γit ln Mit

(8.3)

clearing ln Ait : ln TFPit = ln Ait = ln Yit − αit ln Lit − βit ln Kit − γit ln Mit

(8.4)

Hence, we define TFP growth of an industry i over the period t − τ as ln TFPGrowthit−iτ =

ln TFPit − ln TFPiτ

(8.5)

Our dataset is structured using the total number of industries of the US economy following the North American Industry Classification System (NAICS); the full list of industries and reference numbers is included in Table 8.4 in the Appendix. Growth rates are computed in indexes where the base year is 2012. As proposed by the BLS, TFPGrowthit stands for Sectoral TFP growth, Yit for Sectoral Output growth, Kit for Capital Services growth, Lit for Labour growth and Mit for Intermediate Inputs growth for a given industry i and time t.

8.3.2 Firm-Level Analysis As Information and Communications Technologies (ICT) are considered to be the origin of the Third Industrial Revolution (Gordon & Sayed, 2020), we assume that the entrance of innovations can be measured by a higher capital productivity within a period of disruptive innovation. For this purpose, we are comparing the Capital Services growth between a period of technological change and a stagnated one. On the other hand, we are studying Schumpeter’s creative destruction. As mentioned for this purpose, we suggest analysing the business fabric productivity by using a firm-level based variable, an exit rate of business establishments. The theoretical foundation presented for this approach is grounded on the concept that innovative companies force out those that do not adapt to the new set of productivity levels within an industry in a time of technical change (Aghion & Bessonova, 2006). Therefore, we introduce the concept of business establishment as a variable that measures any economic activity at the lowest level using employment. As an active establishment is the one that has at least one employee, an establishment exit occurs when an existing establishment reaches zero employees (US Census Bureau, 2021). Consequently, by the estimation of an Exit Rate of Establishments

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

139

that accounts for the total exit of establishments divided by the existing number of active ones, we consider the ability of comparing this variable between different time periods. Hence, for our panel we are using data provided by the Business Dynamics Statistics (BDS) from the US Census Bureau (USCB) for the total number of NAICS industries. We propose the following exit rate considering the USCB suggestions when elaborating an Exit Rate of Establishments2 :

ExitRateit =

estabsexityearit  × 100  0.5 × estabsyearit + estabsyearit−1

(8.6)

In this equation, the exit number of establishments in a year t is divided by the average number of active establishments in year t and year t − 1 for an industry i.

8.3.3 Time Periods Considering that the Third Industrial Revolution started in the 1970s, the annual growth rate of productivity in the United States stagnated from 1972 to 1995 (Gordon, 2013), reviving in the decade of 1995–2005 (Gordon & Sayed, 2020). The main reason for this growth path is substantiated by the deferred effect between technological discovery and technological application within an economy (David, 1990; Gordon, 2015). For a more detailed analysis, we are subdividing the 1995– 2005 period into two 5-year terms. The first period from 1995 to 2000 was characterised by a significant investment in ICT where pre-existing companies had to adapt their business models and important companies such as Google (1998) emerged (Jorgenson & Stiroh, 2000; Oliner & Sichel, 2000). Although the highest growth in productivity is observed in industries related to ICT, services and industrial machinery sectors linked to the production of computers also experienced a noticeable growth (Gordon & Sayed, 2020). These investments in technology and spillovers transferred from ICT producer industries to ICT-intensive industries from 2000 to 2005 (Jorgenson et al., 2005). Interestingly after 2005, a productivity slowdown period started with a noticeable stagnation in the decade of 2009–2019 (Gordon, 2021). One of the

2 The

USCB (2021) recommends to ensure that the count of year t active establishments is longitudinally consistent with the count of active establishments in year t − 1. For this purpose, the same scope of establishments needs to be applied for both considered years. For year t, a longitudinally consistent count of employment active establishments is made in year t − 1 with this specific calculation: estabs_year t − 1 = estabs year t + estabs_exit year t – estabs_entry year t.

140

J. M. S. Llewellyn-Jones

Table 8.2 Average Annual Rate of TFP Growth for the United States Average Annual Rate of TFP Growth

1972–1995 0.38

1995–2005 1.14

2009–2019 0.56

Source: Cells are computed using data from University of Groningen and University of California (n.d.), US Census Bureau. (n.d.)

main suggested economic interpretations would be the end of the Third Industrial Revolution, a possible cusp before a new innovation period to come originated by robots and artificial intelligence (Gordon, 2015, 2021). Considering that the Third Industrial Revolution is ICT driven, we propose to measure the technological pace of this revolution with an Average Annual Rate of TFP growth:  Average Annual Rate of TFPGrowtht−τ =

1 TFPt ln t −τ TFPτ

 × 100

(8.7)

The presented Average Annual Rate of TFP growth formula considers the annual average growth of TFP over a period t − τ . As seen in Table 8.2, the Average Annual Rate of TFP growth between 1995 and 2005 is highly disruptive compared to the 1972–1995 one. As the data provided by BLS comprises just from the period 1987 to 2020, we are choosing the decade between 2009 and 2019 as representative of a stagnation period as its rate is similar to the 1972–1995 one. We have also considered that it is not directly affected by any economic recession.

8.4 Regression Model In line with the methodology suggested in Sect. 8.3 to conceive TFP and TFP growth, a linear regression is proposed by adding the Exit Rate of Establishments as one more of the composite inputs of TFP growth. We are also contemplating the contribution of innovative entrepreneurship to TFP growth by the analysis of the Capital Services growth’s coefficient within the proposed technological change period. Consequently, with the obtained results, we are going to be able to analyse the magnitude, symbol and significance of the Exit Rate of Establishments coefficient in relation to TFP growth to measure creative destruction. Labour and Intermediate Inputs’ coefficients have an economic meaning related to how they are affected by ICT: LogTFPGrowthit = αi + β1 LogYit − β2 LogK it − β3 LogL − β4 LogM it −β5 LogExitRateit + εit

(8.8)

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

141

The proposed regression models in natural logarithms the Sectoral TFP’s growth as a function of the Sectoral Output’s growth, Capital Services’ growth, Labour Input’s growth, Intermediate Inputs’ growth and the Exit Rate of Establishments for an industry i at a time t. β j are the coefficients for each of the variables, α 0 is the intercept, and εit is the overall error term for the proposed industry i and a time t. To compute parameters, we need to consider that any data not explained by the proposed variables implies differences from a transversal perspective. These differences are collected in the error and determine the right methodology to be chosen. In the first place, we estimated our model using least squares random effects. To find out if it was the best choice from a statistical point of view, the parametric contrast of the Hausman test (1978) was run. After applying the Hausman test to the regression model obtained in both periods 1995–2005 and 2009–2019, we observed that group-specific errors were correlated with the regressor. Consequently, we considered that the regressor was biassed. Hence, we estimate our regression model using the least squares fixed effects methodology. We accept the results as appropriate for our analysis’ purposes as it is preferable to prioritise unbiasedness, meaning that the mean of the sample distribution is equal to the population parameter, over efficiency, where the estimator is inefficient due to the existence of multicollinearity.

8.5 Empirical Results and Suggested Economic Interpretations The estimations obtained, when running regressions in both time periods according to the above-specified model, are presented in Table 8.3 in the Appendix. The main objective is to compare coefficients as elasticities in terms of standard deviations to be economically interpreted. All coefficients are statistically significant, and their signs are the same as in the proposed Eq. 8.4. We first need to notice that the Sectoral Output contribution to the Sectoral TFP growth barely changes. Therefore, there is no noticeable change effect on the relationship between Sectoral Output growth and TFP growth when technological change occurs. After this first consideration, we can now focus on the variables of our interest to explain the entrepreneurial innovative process contribution to TFP growth. Since in the 1995–2005 period of technological change, there was a notorious investment in ICT Capital, we are going to compare this period’s Capital Services growth’s coefficient to the one from the stagnated period of 2009–2019. Following the results presented in Table 8.3, a 1 percent increase of the Capital Services makes Sectoral TFP growth decrease by a −0.074 percent in the decade of 1995– 2005 and by a −0.3 percent in the one of 2009–2019. Therefore, Sectoral TFP growth decreased in a lesser quantity if Capital Services growth increased in

142

J. M. S. Llewellyn-Jones

the technological change period, meaning that the newly introduced capital by innovative entrepreneurship is more productive in the technological change period. Consequently, we conclude that the results shown by our model initially support the Schumpeterian approach of our research. We do confirm Hypothesis 1. Once confirmed that innovative entrepreneurship happened with more intensity in the period of 1995–2005, we now want to study the effects of creative destruction from a firm-level perspective comparing the obtained results to the stagnated period of 2009–2019. In the latter decade, the coefficient of the independent variable Exit Rate of Establishments is a positive 0.043. Therefore, a 1 percent increase in the Exit Rate of Establishments leads to a 0.043 percent increase in the Sectoral TFP growth. We economically interpret that there was an exit of unproductive establishments that created a more productive business fabric. Consequently, we confirm Schumpeter’s creative destruction theory happens when the technological change is slow, as it is for the chosen period. On the other hand, there was an interesting effect on the technological change period, as there is a negative correlation between the Exit Rate of Establishments and Sectoral TFP. We understand from the data presented in Table 8.3 that when the Exit Rate of Establishments coefficient increases 1 percent, the Sectoral TFP growth’s coefficient decreases −0.032 percent. Therefore, from the aggregated perspective coming from our research, we can conclude that, as the process of creative destruction happens with more intensity in the selected period of technological change, a negative effect on productivity growth occurs when productive companies are forced to leave. Hence, we do not confirm Hypothesis 2. From the perspective of the Labour growth’s relation to the Sectoral TFP growth, there is not an important change in coefficients between periods. These results highlight an important characteristic from the ICT implemented in the Third Industrial Revolution after 2005: there are no noticeable implications in human labour productivity from the implementation of ICT. Finally, slight changes are shown in the Intermediate Inputs coefficients from our model, as the one between 1995 and 2005 is −0.44 and the related elasticity is −0.23 in the period 2009–2019. Economic intuition leads us to think that there was a possible delay in transmission of the technology associated with the disruptive period (Tables 8.4, 8.5, and 8.6).

8.6 Conclusions According to Acs and Audretsch (1990), Christensen and Raynor (2003) and Gordon (2016), some innovations create larger economic growth, whilst others are more trivial. This theoretical approach was taken into consideration to study the implications from innovative entrepreneurship and the process of creative destruction as productivity enhancers in periods of technological change.

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

143

As developed countries like the United States are regarded as innovationdriven economies, some policy implications can be drawn from the current study. By acknowledging that our analysis has an aggregated perspective, our results suggested that although innovative entrepreneurship is beneficial for productivity, there is an exit of productive companies in periods of technological change. As there was no industry comparison for the regression model, we find a limitation that could be considered as an opportunity for future research. Therefore, a Sectoral TFP growth analysis for each industry would make it easier to consequently encourage technological transfer within specific disruptive industries. We also recommend further research related to an optimal level of an Exit Rate of Establishments in relation to Sectoral TFP growth. This is, as we consider that the reversal in the correlation between business establishments and productivity presented in this paper makes a complementary advance to prior studies concerning entrepreneurship and competition (Aghion et al., 2005; Aghion & Griffith, 2006; Aghion et al., 2015), economic development (Carree et al., 2002, 2007) and economic growth (Prieger et al., 2016). In conclusion, given the proposed methodology, we hope to have set a research track that will open and enrich new theoretical and empirical findings for the Schumpeterian school of thought.

Appendix

Table 8.3 Results from panel estimations Fixed-effects model ln(Sectoral Output) ln(Capital Services) ln(Labour Input) ln(Intermediate Materials) ln(ExitRate) α0 R-squared Adjusted R-squared Prob (F-statistic) Durbin-Watson Stat

TFP growth (1995–2005) 0.8534(29.93)*0.0000 −0.0740(−4.78)*0.0000 −0.2819(−16.79)*0.0000 −0.4408(−30.56)*0.0000 −0.0327(−2.62)*0.0094 4.4163(61.03)*0.000 0.9647 0.9604 0 0.4917

Values in brackets are t-statistic values *Indicates significance of the coefficient

TFP growth (2009–2019) 0.8454(28.2)*0.0000 −0.3006(−14.67)*0.0000 −0.2842(−11.29)*0.0000 −0.2344(−14.09)*0.0000 0.0431(4.55)*0.0000 4.3855(45)*0.000 0.9176 0.9073 0 0.5903

144

J. M. S. Llewellyn-Jones

Table 8.4 North American Industry Classification System (NAICS) Sector 11 21 22 23 31–33 42 44–45 48–49 51 52 53 54 55 56 61 62 71 72 81 92

Definition Agriculture, forestry, fishing and hunting Mining, quarrying and oil and gas extraction Utilities Construction Manufacturing Wholesale trade Retail trade Transportation and warehousing Information Finance and insurance Real estate and rental and leasing Professional, scientific and technical services Management of companies and enterprises Administrative and support and waste management and remediation services Educational services Healthcare and social assistance Arts, entertainment and recreation Accommodation and food services Other services (except public administration) Public administration

Variable ln(Sectoral Output) ln(Capital Services) ln(Labour Input) ln(Intermediate Materials) ln(ExitRate) α0

90% CI Low 0.806265 −0.099611 −0.30971 −0.46469 −0.053373 4.296734

Coefficient 0.853403 −0.074004 −0.281945 −0.440849

−0.032755 4.416351

−0.012137 4.535968

High 0.900542 −0.048397 −0.254181 −0.417009

Table 8.5 Coefficient intervals for the estimated regression in 1995–2005

−0.057361 4.273598

95% CI Low 0.797148 −0.104564 −0.31508 −0.469301 −0.008149 4.559104

High 0.909659 −0.043444 −0.24881 −0.412398

−0.065216 4.228028

99% CI Low 0.77919 −0.11432 −0.325658 −0.478383

−0.000295 4.604674

High 0.927617 −0.033688 −0.238233 −0.403316

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction 145

ln(Sectoral Output) ln(Capital Services) ln(Labour Input) ln(Intermediate Materials) ln(ExitRate) α0

Coefficient 0.845366 −0.300582 −0.284205 −0.234388 0.043159 4.385567

90% CI Low 0.795854 −0.334451 −0.325792 −0.261882 0.027504 4.224456 High 0.894879 −0.266713 −0.242618 −0.206895 0.058814 4.546679

Table 8.6 Coefficient intervals for the estimated regression in 2009–2019 95% CI Low 0.786277 −0.341002 −0.333835 −0.267199 0.024476 4.193295 High 0.904456 −0.260163 −0.234574 −0.201577 0.061841 4.57784

99% CI Low 0.767414 −0.353905 −0.349679 −0.277673 0.018512 4.131916

High 0.923319 −0.24726 −0.218731 −0.191103 0.067805 4.639219

146 J. M. S. Llewellyn-Jones

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

147

References Acs, Z. J., & Audretsch, D. B. (1990). Innovation and small firms. The MIT Press. Agarwal, R., Bayus, B. L., & Tripsas, M. (2005). Abandoning innovation in emerging industries. Voice, 8(4), 356–370. Agarwal, R., Audretsch, D., & Sarkar, M. B. (2007). The process of creative construction: Knowledge spillovers, entrepreneurship, and economic growth. 1(3–4), 263–286. https:// doi.org/10.1002/sej.36 Aghion, P., Akcigit, U., & Howitt, P. (2013). What do we learn from Schumpeterian growth theory?. PIER Working Paper No. 13-026. https://doi.org/10.2139/ssrn.2274704 Aghion, P., & Bessonova, E. (2006). On entry and growth: Theory and evidence. Revue de l’OFCE, 97(bis (5)), 259–278. https://doi.org/10.3917/reof.073.0259 Aghion, P., Bloom, N., Blundell, R., Griffith, R., & Howitt, P. (2005). Competition and innovation: An inverted-U relationship. The Quarterly Journal of Economics, 120(2), 701–728. https:// doi.org/10.1093/qje/120.2.701 Aghion, P., & Griffith, R. (2006). Competition and growth: Reconciling theory and evidence. The MIT Press. Aghion, P., & Howitt, P. (1992). A model of growth through creative destruction. Econometrica, 60(2), 323–351. https://doi.org/10.2307/2951599 Aghion, P., Akcigit, U., & Howitt, P. (2015). The Schumpeterian growth paradigm. Annual Review of Economics, 7, 557–575. https://doi.org/10.1146/annurev-economics-080614-115412 Aparicio, S., Urbano, D., & Gómez, D. (2016). The role of innovative entrepreneurship within Colombian business cycle scenarios: A system dynamics approach. Futures, 81, 130–147. https://doi.org/10.1016/j.futures.2016.02.004 Aubry, M., Bonnet, J., & Renou-Maissant, P. (2015). Entrepreneurship and the business cycle: The “Schumpeter” effect versus the “refugee” effect—A French appraisal based on regional data. The Annals of Regional Science, 54, 23–55. https://doi.org/10.1007/s00168-014-0645-x Audretsch, D. B. (1997). Economics. The Annals of the American Academy of Political and Social Science, 549(1), 210–211. https://doi.org/10.1177/0002716297549001036 Audretsch, D. B., Belitski, M., & Desai, S. (2015). Entrepreneurship and economic development in cities. The Annals of Regional Science, 55, 33–60. https://doi.org/10.1007/s00168-015-0685-x Audretsch, D. B., & Fritsch, M. (2002). Growth regimes over time and space. Regional Studies, 36(2), 113–124. https://doi.org/10.1080/00343400220121909 Baumol, W. J. (2002). The free-market innovation machine. Princeton University Press. Baily, M. N., Hulten, C., & Campbell, D. (1992). Productivity dynamics in manufacturing plants. Brookings Papers on Economic Activity, 23, 187–267. https://doi.org/10.2307/2534764 Bleaney, M., & Nishiyama, A. (2002). Explaining growth: A contest between models. Journal of Economic Growth, 7(1), 43–56. https://doi.org/10.1023/A:1013466526642 Bureau of Labor Statistics. (2022). Major industries. https://www.bls.gov/productivity/tables/ major-industry-total-factor-productivity-klems.xlsx. Accessed 15 May 2022. Carree, M. A., & Thurik, A. R. (2008). The lag structure of the impact of business ownership on economic performance in OECD countries. Small Business Economics, 30(1), 101–110. https:/ /doi.org/10.1007/s11187-006-9007-0 Carree, M. A., Van Stel, A. J., Thurik, A. R., & Wennekers, A. R. M. (2002). Economic development and business ownership: An analysis using data of 23 OECD countries in the period 1976-1996. Small Business Economics, 19(3), 271–290. https://doi.org/10.1023/ A:1019604426387 Carree, M. A., Van Stel, A. J., Thurik, A. R., & Wennekers, A. R. M. (2007). The relationship between economic development and business ownership revisited. Entrepreneurship and Regional Development, 19(3), 281–291. https://doi.org/10.1080/08985620701296318 Castaño-Martínez, M. S., Méndez-Picazo, M. T., & Galindo-Martín, M. Á. (2015). Policies to promote entrepreneurial activity and economic performance. Management Decision, 53(9), 2073–2087. https://doi.org/10.1108/MD-06-2014-0393

148

J. M. S. Llewellyn-Jones

Christensen, C. M., & Raynor, M. E. (2003). The Innovator’s solution: Creating and sustaining successful growth. Harvard Business School Press. Dabkowski, A. P. (2011). Entrepreneurship and economic growth: An investigation into the relationship between entrepreneurship and total factor productivity growth in the EU. CASE Network Studies and Analyses, 427. https://doi.org/10.2139/ssrn.1883546 David, P. A. (1990). The dynamo and the computer: An historical perspective on the modern productivity paradox. American Economic Review Papers and Proceedings, 80(5), 55–61. Erken, H., Donselaar, P., & Thurik, R. (2018). Total factor productivity and the role of entrepreneurship. The Journal of Technology Transfer, 43, 1493–1521. https://doi.org/10.1007/s10961-0169504-5 Foster, L., Haltiwanger, J., & Syverson, C. (2008). Reallocation, firm turnover, and efficiency: Selection on productivity or profitability? American Economic Review, 98(1), 394–425. https:/ /doi.org/10.1257/aer.98.1.394 Gordon, R. J. (2013). U.S. productivity growth: The slowdown has returned after a temporary revival. International Productivity Monitor, Centre for the Study of Living Standards, 25, 13– 19. Gordon, R. J. (2015). Secular stagnation: A supply-side view. American Economic Review, 105(5), 54–59. https://doi.org/10.1257/aer.p20151102 Gordon, R. J. (2016). The rise and fall of American growth: The U.S. standard of living since the civil war (REV). Princeton University Press. https://doi.org/10.2307/j.ctvc77bwm Gordon, R. J. (2021). Productivity and growth over the years at BPEA. Brookings Papers on Economic Activity, 2021(1), 213–229. https://doi.org/10.1353/eca.2021.0003 Gordon, R. J., & Sayed, H. (2020). Transatlantic technologies: The role of ICT in the evolution of U.S. and European productivity growth. International Productivity Monitor, Centre for the Study of Living Standards, 38, 50–80. https://doi.org/10.3386/w27425 Grossman, G., & Helpman, E. (1991). Innovation and growth in the global economy. MIT Press. Hall, R. E. (1988). The relation between price and marginal cost in U.S. industry. Journal of Political Economy, 96(5), 921–947. Hall, R. E. (1989). Invariance properties of Solow’s productivity residual. National Bureau of Economic Research. Hausman, J. (1978). Specification tests in econometrics. Econometrica, 46, 1251–1271. https:// doi.org/10.2307/1913827 Hulten, C. R., & Isaksson, A. (2007). Why development levels differ: The sources of differential economic growth in a panel of high and low income countries. NBER Working Paper No 13469. https://doi.org/10.3386/w134619 Isaksson, A. (2007). Determinants of total factor productivity: A literature review. Research and Statistics Branch UNIDO, 1, 101. Jorgenson, D. W., Ho, M. S., & Stiroh, K. J. (2005). Growth of US industries and investments in information technology and higher education. In NBER chapters, measuring capital in the new economy (pp. 403–478). https://doi.org/10.1080/0953531032000111781 Jorgenson, D. W., & Stiroh, K. J. (2000). U.S. economic growth at the industry level. American Economic Review, 90(2), 161–167. https://doi.org/10.1257/aer.90.2.161 Jorgenson, D. W., & Stiroh, K. J. (1999). Information technology and growth. American Economic Review, 89(2), 109–115. https://doi.org/10.1257/aer.89.2.109 Judd, K. L. (1985). On the performance of patents. Econometrica, 53, 567–585. King, R. G., & Rebelo, S. T. (1988). Business cycles with endogenous growth. Unpublished. University of Rochester. Klenow, P. J., & Rodríguez-Clare, A. (1997). The neoclassical revival in growth economics: Has it gone too far? NBER Macroeconomics Annual, 12, 73–103. https://doi.org/10.2307/3585220 Lafuente, E., Acs, Z. J., Sanders, M., & Szerb, L. (2020). The global technology frontier: Productivity growth and the relevance of Kirznerian and Schumpeterian entrepreneurship. Small Business Economics, 55, 153–178. https://doi.org/10.1007/s11187-019-00140-1 Low, S. A., & Isserman, A. M. (2015). Where are the innovative entrepreneurs? Identifying innovative industries and measuring innovative entrepreneurship. International Regional Science Review, 38(2), 171–201. https://doi.org/10.1177/0160017613484926

8 Total Factor Productivity and Entrepreneurship: Creative Self-Destruction

149

Lucas, R. E. (1988). On the mechanisms of economic development. Journal of Monetary Economics, 22(1), 3–42. https://doi.org/10.1016/0304-3932(88)90168-7 Malerba, F., & Orsenigo, L. (1996). Schumpeterian patterns of innovation are technology-specific. Research Policy, 25, 451–478. https://doi.org/10.1016/0048-7333(95)00840-3 Oliner, S. D., & Sichel, D. E. (2000). The resurgence of growth in the late 1990s: Is information technology the story? Journal of Economic Perspectives, 14(4), 3–22. https://doi.org/10.1257/ jep.14.4.3 Prescott, E. C. (1997). Needed: a theory of total factor productivity. Federal Reserve Bank of Minneapolis, Staff Report 242. Prieger, J. E., Bampoky, C., Blanco, L. R., & Liu, A. (2016). Economic growth and the optimal level of entrepreneurship. World Development, 82, 95–109. https://doi.org/10.1016/ j.worlddev.2016.01.013 Romer, P. M. (1986). Increasing returns and long-run growth. Journal of Political Economy, 94(5), 1002–1037. https://doi.org/10.1086/261420 Romer, P. M. (1990). Endogenous technological change. Journal of Political Economy, 98(5), S71– S102. https://doi.org/10.1086/261725 Schumpeter, J. A. (1934). The theory of economic development: An inquiry into profits, capital, credit, interest and the business cycle. Harvard University Press. Schumpeter, J. A. (1942). Capitalism, socialism and democracy. Harvard University Press. Solow, R. M. (1957). Technical change and the aggregate production function. Review of Economics and Statistics, 39, 312–320. https://doi.org/10.2307/1926047 Sternberg, R., & Wennekers, S. (2005). Determinants and effects of new business creation using global entrepreneurship monitor data. Small Business Economics, 24, 193–203. s11187-0051974-z. Syverson, C. (2011). What determines productivity? Journal of Economic Literature American Economic Association, 49(2), 326–365. https://doi.org/10.1257/jel.49.2.326 Thurik, A. R., Carree, M. A., Van Stel, A. J., & Audretsch, D. B. (2008). Does self-employment reduce unemployment? Journal of Business Venturing, 23(6), 673–686. https://doi.org/10.1016/ j.jbusvent.2008.01.007 University of Groningen and University of California, Davis. (n.d.). Total Factor Productivity at Constant National Prices for United States [RTFPNAUSA632NRUG]. Retrieved from FRED, Federal Reserve Bank of St. Louis. https://fred.stlouisfed.org/series/RTFPNAUSA632NRUG. Accessed 15 May 2022. Urbano, D., Aparicio, S., & Audretsch, D. B. (2019). Institutions, entrepreneurship, and economic performance. International Studies in Entrepreneurship, Springer. US Census Bureau. (2021). BDS FAQ. https://www.census.gov/programs-surveys/bds/ documentation/faq.html. Accessed 15 May 2022. US Census Bureau. (n.d.). BDS Explorer. https://bds.explorer.ces.census.gov. Accessed 15 May 2022. van Oort, F. G., & Bosma, N. S. (2013). Agglomeration economies, inventors and entrepreneurs as engines of European regional economic development. The Annals of Regional Science, 51, 213–244. https://doi.org/10.1007/s00168-012-0547-8 van Praag, M., & van Stel, A. (2013). The more business owners, the merrier? The role of tertiary education. Small Business Economics, 41, 445–557. https://doi.org/10.1007/s11187012-9436-x Wennekers, S., & Thurik, R. (1999). Linking entrepreneurship and economic growth. Small Business Economics, 13, 27–56. https://doi.org/10.1023/A:1008063200484 Wong, P., Ho, Y. P., & Autio, E. (2005). Entrepreneurship, innovation and economic growth: Evidence from GEM data. Small Business Economics, 24, 335–350. https://doi.org/10.1007/ s11187-005-2000-1 ˙ & Çelik, A. K. (2017). The impact of total factor productivity Yalçınkaya, Ö., Hüseyni, I., on economic growth for developed and emerging countries: A second-generation panel data analysis. Margin: The Journal of Applied Economic Research, 11(4), 404–417. https://doi.org/ 10.1177/0973801017722266

Chapter 9

Generation Z Intention to Comply with Non-mandatory Government Measures for Self-protection of COVID-19 and SARS-CoV-2 Variants After Restriction Withdrawals Irene (Eirini) Kamenidou, Aikaterini Stavrianea, Spyridon Mamalis, Evangelia-Zoe Bara, Ifigeneia Mylona, and Stavros Pavlidis

Abstract Vaccines for the COVID-19 pandemic initial virus, the “severe acute respiratory syndrome coronavirus 2” (SARS-CoV-2), and its variances have been distributed in the developed countries to the majority of the citizens. However, new variants constantly emerge, and even though mortality rates have been decreased, morbidity (including those that are asymptomatic) is still high. This brings to the surface the fact that individual precautious measures should be continuously applied, even though they might not be mandatory based on government regulations, not only for individuals’ safety but also for the safety of the older people that surround them. Under this perspective, this paper has its aims to explore the generation Z cohort’s intention to comply with non-mandatory government measures in order to protect themselves from infection but also to not be a “moving threat” to the people in contact. The research was undertaken for this purpose with an online questionnaire in 2021 from May to the end of July, resulting in a sample of 1086 valid questionnaires. Gender differences were tested with an independent sample t-test, which revealed that females have a higher intention to comply with non-mandatory protective measures. Based on these differences marketing communication is discussed to increase non-mandatory self-protection measures.

I. (E). Kamenidou () · S. Mamalis · E.-Z. Bara · I. Mylona · S. Pavlidis Department of Management Science and Technology, School of Business and Economics, International Hellenic University, Agios Loukas, Kavala, Greece e-mail: [email protected]; [email protected]; [email protected] A. Stavrianea Department of Communication and Media Studies, National and Kapodistrian University of Athens, Athens, Greece e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_9

151

152

I. (E). Kamenidou et al.

Keywords COVID-19 · Generation Z cohort · Self-protection · Restrictions withdrawals · Communication marketing

JEL Codes M30, M31, M37, M39, I10, I18

9.1 Introduction The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) which caused the COVID-19 pandemic, on June 7, 2022, resulted globally in 53,235,583,5 cases and 63,000,60 deaths (JHU, 2022). The three countries with the most cases confirmed are the United States (84,882,287 cases and 1,008,857 deaths); India (43,185,049 cases and 524,708 deaths) and Brazil (31,195,118 cases and 667,041 deaths). Worldwide due to the aggressive nature of the SARS-CoV-2 leading to the COVID-19 pandemic, measures have been adopted by countries ranging from simple two-meter distancing and mask-wearing to countries lockdowns (Fu et al., 2022; Gerace et al., 2022; Immel et al., 2022; Or et al., 2022; Wu et al., 2022; Wee et al., 2021; Houvèssou et al., 2021; Aquino et al., 2020; Anderson et al., 2020; Kamenidou et al., 2020a; Haider et al., 2020; Van Rooij et al., 2020). Additionally, different vaccines were developed to combat the virus (Ndwandwe & Wiysonge, 2021; Rosenblum et al., 2021; Vitiello & Ferrara, 2021). Though variants arose (Arif, 2022; Takashita et al., 2022; Rubin, 2021; Kunal et al., 2021; Roy et al., 2021; Torjesen, 2021) leading to new doses of vaccines (Burki, 2022; Munro et al., 2022) or countries lockdowns (e.g., the ongoing Shanghai China lockdown). The main obstacle to tackling the SARS-CoV-2 virus is its continuous emerging new mutants ranging from weak to strong infectious results. By 2021, there were eight known noble variants: Alfa, Beta, Gamma, Delta, Eta, Iota, Kappa, and Lambda (Mahase, 2021), while others are added increasing the number of variants to at least twelve (Fernandes et al., 2022; Yang et al., 2022). WHO (2022) has classified the variants into variants of concern (VOC), VOC lineages under monitoring (VOCLUM), Variants of interest (VOI) and Variants under monitoring (VUM), which in total exceed 20 variants, some of which have caused deaths of people that had been vaccinated even with the booster vaccine. These new developments in the pandemic, combined with the high level of contagiousness and transmissibility, the asymptomatic variants and asymptomatic contagion (Liu & Cao, 2022; Gusev et al., 2022), indicate that the bottom line is that individuals need to take self-protective measures regardless of if the government imposes measures or not. Earlier research reveals that asymptomatic contagion is a significant reason for COVID-19 transmission (Ralli et al., 2021; Bastos et al., 2021). Previous research also reveals that asymptomatic carriers are usually young people (Xiao et al., 2021; Day, 2020). Therefore, it is very important to know the protective measures

9 Generation Z Intention to Comply with Non-mandatory Government. . .

153

undertaken by them after lockdown and curfew were withdrawn since they are usually asymptomatic and highly contribute to the contagion of the virus. Regarding Greece, and starting from the 26th of April 2021, restrictions were one by one minimized to be lifted by the middle of May in order to get to the tourism mode. The government allowed the reopening of all entertaining services, with mandatory conditions vaccination certificate, masks, and two-meter social distance. Conditions for entry into the country to limit the spread of COVID-19 coronavirus were named by the Government Gazette 1969B (2021). Building upon the above discussion, this paper has as its purpose to describe generation Z cohorts’ intention to comply with proactive preventive behaviour after the government withdrew the major restrictions (lockdown and curfew). Secondly, it has an objective to explore gender differences in their intention to comply with preventive non-mandatory self-protective measures. Concerning the perspective of this paper, it uses the adult members of the Greek generation Z (Gen Z) cohort as a sample unit, that is, Gen Zers born between 1994 and 2003 at the time of the research (Strauss & Howe, 2020). Consequently, this study adds to the previous work in the following ways: 1. It studies Gen Zers’ intention to comply with protective measures from the COVID-19 disease and the SARS-CoV-2 variants after a government has withdrawn the majority of the restrictions implemented to control the spread of the disease. 2. It examines if males and females of the Gen Z cohort have an equal intention to comply with the non-mandatory protective measures, a subject which has not been studied up to now to our knowledge.

9.2 Literature Review Since the initial outbreak of COVID-19 there has been an abundance of studies that focus on different aspects of it, with the main axes of interest focusing on: • The initial virus pathophysiology (Osuchowski et al., 2021; Yuki et al., 2020; Amawi et al., 2020; Shereen et al., 2020). • The pandemic, including infection symptoms and transmission (Samui et al., 2020; Li et al., 2020; Madabhavi et al., 2020; Shereen et al., 2020). • Vaccination and vaccines of the initial virus (Calina et al., 2020; Lurie et al., 2020; Lipsitch & Dean, 2020). • Variants and vaccinations (Ciotti et al., 2022; Tregoning et al., 2021; Rubin, 2021). • Government measures (Hale et al., 2021; Haug et al., 2020; Fang et al., 2020; Shah et al., 2020). • Citizens’ attitudes and behaviour (Cheng et al., 2022; Jørgensen et al., 2021; Shaw et al., 2020; Kamenidou et al., 2020a).

154

I. (E). Kamenidou et al.

• Impact of COVID-19 on different aspects of life such as the economy, mental issues, education, etc. (Ibn-Mohammed et al., 2021; Pham et al., 2021; Chaturvedi et al., 2021; Iglesias-Pradas et al., 2021). So far, one subject that is understudied is the behaviour of the Gen Zer cohort as regards compliance with protective measures. Multiple articles have been found that concern Gen Zers behaviour towards COVID-19 which mainly falls into the elearning/education category, tourism, and travel, purchasing and eating behaviour, and workplace. Fewer have focused on their prevention measures or coping with the pandemic. Since we did not find research directly associated with our aim and objective, published work that was closely related is presented. Simic and Pap (2022), explored among others the behavioural intentions towards the prescribed measures of the COVID-19 pandemic (N = 442, Croatia) and found that Gen Zers consider the most challenging to adopt the measures related to social distancing, but not the ones that refer to wearing face masks, measuring body temperature and applying higher hygiene standards. Gender differences were not studied. Prati et al. (2022) explored gender differences in risk perception, attitudes towards quarantine measures and adoption of precautionary behaviours during the COVID-19 pandemic (N = 1569, Italy) and found that females had “higher scores on perceived severity, worry, precautionary behaviors, and attitudes toward quarantine restrictions. Gender differences in the perceived likelihood of infection with SARS-CoV-2 were not significant.” This study did not refer to Gen Zers. Kamenidou et al. (2022) examined the information sources (providers: N = 8 and communication channels: N = 6) Gen Zers trust (N = 1411; Greece) as well as gender differences based on trust in COVID-19 information dissemination and found seven statistically significant gender differences for information providers and one for communication channels. Karabay et al. (2022) explored the psychological consequences of the COVID19 pandemic on the current attitudes, behaviours and future expectations of Gen Zers (N = 299) and found that the “female respondents feel less secure and more pessimistic in terms of the new normal conditions”. Jose (2022), identified the factors that affect Gen Zers behaviour in the adoption of the COVID-19 Vaccine (N = 30; qualitative research, thematic analysis) and found that the key variables that expedite the adoption process among Gen Zers are “Observability, Country of Origin, Brand and Word of mouth”. Gender differences were not studied. Truong et al. (2022) explored how the media influenced preventive behaviours against COVID-19 (N = 609; Vietnam) among different generational cohorts (N = 4), including the Gen Zers. They found that there were no significant differences between genders and their preventive behaviours towards COVID-19. Yet, they did find between cohorts. They also found that Gen Zers were affected by social media more than by mass media.

9 Generation Z Intention to Comply with Non-mandatory Government. . .

155

Ramayanti et al. (2021) examined the association between Gen Zers’ knowledge and attitudes towards the use of masks, keeping their distance, and washing hands as a means of preventing COVID-19 (N = 147) and found that there is a significant association between knowledge and attitudes towards COVID-19 infection prevention. Gender differences were not studied. Deckman et al. (2020) measured (among others) Gen Zers attitudes about the impact of the coronavirus on health, work, and finance, and found that “women are more likely to be concerned that COVID-19 will harm their ability to keep their jobs” and “..Gen Z men are significantly more likely to express concerns about their own personal health with respect to COVID-19”. Kamenidou et al. (2020b) studied Gen Zers COVID-19 symptom knowledge as well as gender differences in symptom knowledge (N = 762; Greece) and found that for overall knowledge of COVID-19 symptoms, as well as COVID-19 symptom knowledge regarding fever and fatigue, myalgia (muscle pain), pharyngodynia, nausea–vomitus, hemoptysis and abdominal pain, there were statistical differences between men and women. Galasso et al. (2020) explored gender differences in COVID-19−related beliefs and behaviours (N = 21,649; eight Organisation for Economic Co-operation and Development countries) and found that “women are more likely to perceive COVID19 as a very serious health problem, to agree with restraining public policy measures, and to comply with them. Gender differences in attitudes and behavior are sizable in all countries. They are accounted for neither by sociodemographic and employment characteristics nor by psychological and behavioral factors.” A study by Bulotait˙e et al. (2021) explored the intention to follow various preventive recommendations for COVID-19 (N = 472) in the post-pandemic period (N = 472) and found that “In postpandemic period 10 percent less repondents intend to follow preventive behavior recommendations.” They did not focus on Gen Zers, nor did they explore gender differences.

9.3 Materials and Methods The subject explored in this study deals with the non-mandatory precautions taken by the adult members of the Greek Gen Z cohort, to prevent infection by the SARSCoV-2 virus variants and its potential spread. This topic was examined via one multiitem question: “Outbreaks of the virus mutations have already been reported, and the government has not taken strict action such as a lockdown or movement restrictions, while we are in the tourism season mode. Please state whether this summer you are willing to comply with the following non-mandatory (by the government) practices as a mean to protect yourself from getting infected or spreading the SARS-CoV2 variants and COVID-19.” This question included 24 non-mandatory preventive actions, which were presented on a seven-point Likert-type scale (1 = very low intention to comply, 2 = low intention to comply, 3 = somewhat low intention to comply, 4 = neither high nor low intention (neutral) to comply, 5 = somewhat

156

I. (E). Kamenidou et al.

high intension to comply, 6 = high intension to comply, 7 = very high intention to comply). The above measures/items were adopted from the study of Kamenidou et al. (2020a). Since non-mandatory protective measures were rated, excluded were the use of masks in external settings and 14-day self-isolation if coming from abroad (non-Europe country). In both of these cases, fines were given if not complied with. The research took part from May to the end of July 2021, solely via an online questionnaire and returned a sample of 1086 valid answers. This sample size was regarded as sufficient for the statistical analysis employed (Lehmann et al., 1998) and verified by power analysis with G*Power 3.1 software for power analysis. Data analysis (IBM SPSS ver. 27) involves descriptive statistics, that is, frequencies, percentages (%) and mean values (MV) and independent sample t-test for gender differences (a = 0.05; p < 0.05). The validity of the multi-item question reported in this study was assessed (content validity and face validity). To guarantee the content validity of the items, these were adopted from previous studies, even though, most were withdrawn passing time. The questionnaire’s face validity was confirmed by the pilot test (N = 111 generation Zers) examining its readability and understanding (Kent, 1993). The reliability of the scale provided Cronbach a = 0.960, which is regarded as acceptable (Spector, 1992).

9.4 Results 9.4.1 Sample Profile The total sample size was N = 1086 Greek Gen Zers born from 1995 to 2003. Female subjects were slightly underrepresented (N = 534; 49.2%) compared to male subjects (N = 552; 50.8%). In relation to age, the younger generation Z members (age 18–21) consisted of 57.3% of the sample (N = 622), while the older generation Z members (age 22–26) consisted of 42.7% (N = 464). The vast majority of the sample (95.8%; N = 1039) were single; university students (56.7%) and resided in a city (64.6%).

9.4.2 Self-Reported Prevention Behaviour Contributors were asked to inform about their intention to comply with a proactive protection behaviour against the COVID-19 disease and against increasing the probability of spreading the SARS-CoV-2 virus variants (rated on a 7-point Likert type scale). Table 9.1 presents participants’ answers in percentages (%) and MV in descending order. Table 9.1 reveals that, overall, participants for 11/24 measures will somewhat comply with these measures (4.50 < MV < 5.50), and only towards

9 Generation Z Intention to Comply with Non-mandatory Government. . .

157

Table 9.1 Self-reported willingness to comply with non-mandatory protective measures Non-mandatory prevention measures Use of hand sanitiser (that contains at least 65% alcohol) after touching objects and surfaces in public Avoid touching the face (in particular eyes, nose and mouth) with hands Avoid crowded and overcrowded public areas Clean and disinfect objects and surfaces that are frequently touched by many people Avoid contact with individuals who have acute respiratory illness Social distancing by home isolation as much as possible Always maintain a distance of at least 2 meters from others Very good hand washing after touching objects and surfaces in public Strict compliance with hygiene standards regarding shared toilets (evidence of fecal-oral transmission) Self-quarantine at home for 14 days following the last exposure with suspected infected individuals Avoid all non-mandatory transportation and travel Self-isolation for at least 14 days after contact with people who have come from abroad Strict adherence to hygiene rules at home Avoid contact with individuals with a high risk for severe illness (vulnerable groups) Using single-use hand gloves in public settings Clean and disinfect packaged products Limit transportation only to the necessary Movement to public services, organisations, and areas wearing hand gloves

1 4.9

2 7.4

3 7.0

4 10.0

5 12.1

6 19.2

7 39.4

MV 5.32

4.1

8.7

8.4

12.5

15.2

20.5

30.6

5.10

10.7

11.0

10.0

17.4

14.0

16.8

20.1

4.44

4.2

8.3

8.8

13.5

14.6

18.6

31.9

5.09

2.4

6.9

6.6

11.0

13.8

20.1

39.2

5.44

20.3

32.5

44.6

59.7

71.8

84.6

11.7

12.0

12.0

15.2

16.4

15.0

17.8

4.29

4.5

8.4

7.7

10.9

15.7

19.2

33.7

5.17

3.1

6.4

6.4

11.2

11.5

16.5

44.9

5.51

4.6

6.9

11.2

12.6

19.3

37.0

5.26

15.3

13.1

9.4

14.5

14.5

14.6

18.6

4.18

6.0

8.6

7.7

12.4

13.2

18.0

34.1

5.09

3.7

8.0

7.5

11.9

18.4

21.4

29.2

5.14

3.5

6.8

7.2

12.6

15.8

20.3

33.8

5.26

33.2

13.8

9.4

11.4

10.3

9.9

12.0

3.29

18.9

13.7

12.3

13.7

13.3

12.1

16.0

3.89

21.1

14.7

10.7

13.9

12.8

11.6

15.2

3.78

32.8

13.3

9.5

11.0

10.4

11.4

11.7

3.34

8.3

3.87

(continued)

158

I. (E). Kamenidou et al.

Table 9.1 (continued) Wash thoroughly fruits and vegetables Regular updates of the COVID-19 disease outbreak and precaution measures that should be implemented Respiratory hygiene (covering the cough or sneeze drops by wearing a face mask and washing my hands often.) Checking body temperature monitoring for fever, cough, or dyspnea daily Nonparticipation in parties and friendly gatherings or social events Regular updates on new symptoms related to the virus

6.3

7.2

9.0

14.2

14.7

18.5

30.1

5.00

9.4

11.1

11.6

16.6

16.2

15.2

19.9

4.44

3.8

6.6

7.5

9.8

15.0

20.3

37.1

5.35

25.2

14.1

12.2

13.0

12.1

11.0

12.3

3.55

16.1

12.5

11.1

15.7

13.3

14.0

17.3

4.09

10.7

11.6

10.7

15.4

14.3

15.3

22.1

4.45

Source: the authors

one measure did they borderline tend to comply with the measure. Specifically, they tend to comply (borderline) with the measure “Strict compliance with hygiene standards regarding shared toilets (evidence of the fecal-oral transmission)” with MV = 5.51. Excluding this measure, the next top-rated measures were “Avoid contact with individuals who have acute respiratory illness” (MV = 5.44) and “Respiratory hygiene (covering the cough or sneeze drops by wearing a face mask and washing my hands often.)” with MV = 5.35. On the other hand, the three least measures that Gen Zers comply with are: “Using single-use hand gloves in public settings” (MV = 3.29); “Movement to public services, organizations, and areas wearing hand gloves” (MV = 3.34), and “Checking body temperature, monitoring for fever, cough, or dyspnea daily” (MV = 3.55).

9.4.3 Gender Differences In order to not lose valuable information regarding the measures that the Gen Zers are willing to comply with, gender differences were explored by using independent samples t-test, for all statements without preceding factor analysis. The main hypothesis tested was that there are no differences in MV between genders towards the specific measure. Therefore, this one hypothesis incorporated 24 subhypotheses, one for each measure evaluated. Table 9.2 presents the results of the independent sample t-tests (Df = 1084 equal variations assumed; Sig. two-tailed), as well as the MV of the male and female subjects towards the statement.

9 Generation Z Intention to Comply with Non-mandatory Government. . .

159

Table 9.2 Gender differences in the non-mandatory prevention measures Non-mandatory prevention measures Use of hand sanitiser (that contains at least 65% alcohol) after touching objects and surfaces in public Avoid touching the face (in particular eyes, nose and mouth) with hands Avoid crowded and overcrowded public areas Clean and disinfect objects and surfaces that are frequently touched by many people Avoid contact with individuals who have acute respiratory illness Social distancing by home isolation as much as possible Always maintain a distance of at least 2 meters from others Very good hand washing after touching objects and surfaces in public Strict compliance with hygiene standards regarding shared toilets (evidence of the fecal-oral transmission) Self-quarantine at home for 14 days following the last exposure with suspected infected individuals Avoid all non-mandatory transportation and travel Self-isolation for at least 14 days after contact with people who have come from abroad Strict adherence to hygiene rules at home Avoid contact with individuals with a high risk for severe illness (vulnerable groups) Using single-use hand gloves in public settings Clean and disinfect packaged products Limit transportation only to the necessary Movement to public services, organisations and areas wearing hand gloves Wash thoroughly fruits and vegetables Regular updates of the COVID-19 disease outbreak and precaution measures that should be implemented Respiratory hygiene (covering the cough or sneeze drops by wearing a face mask and washing my hands often.) Checking body temperature monitoring for fever, cough, or dyspnea daily Not participating in parties and friendly gatherings or social events Regular updates on new symptoms related to the virus Source: the authors

F 0.686

p .001

Males MV 5.14

Female MV 5.52

1.575

.002

4.93

5.28

0.720 1.400

.265 .002

4.37 4.92

4.50 5.27

3.409

.000

5.26

5.63

3.619

.890

3.86

3.87

0.092

.068

4.18

4.40

0.002

.003

5.01

5.34

4.347

.000

5.30

5.72

0.202

.002

5.09

5.44

0.326 0.033

.113 .007

4.08 4.93

4.28 5.25

4.271 0.174

.403 .025

5.10 5.15

5.19 5.39

0.512 0.045 0.048 0.494

.523 .675 .942 .285

3.33 3.86 3.78 3.41

3.25 3.92 3.79 3.27

0.126 0.065

.041 .575

4.88 4.41

5.12 4.48

1.153

.000

5.13

5.57

0.453

.864

3.54

3.56

1.226

.775

4.07

4.10

0.126

.662

4.43

4.48

160

I. (E). Kamenidou et al.

As table 9.2 reveals, the t-tests showed that for eleven out of 24 cases gender differences do exist. It also reveals that in all cases females are more inclined to comply with self-protective measures than their male counterparts, as their MV shows.

9.5 Discussion and Conclusions The results of this study cannot be directly compared with previous studies, but only indirectly since most of the research found was during the first wave of the pandemic and lockdowns and curfews. Research by Bulotait˙e et al. (2021) in measuring people’s intention to comply with COVID-19 preventive measures in a post-pandemic period found that about 10% of the participants reported not following the main preventive measures such as wearing a mask, and distance in communication. This research though did not refer to Gen Zers nor did it include gender differences. Additionally, in Greece, wearing a mask was a mandatory measure after restrictions were withdrawn. The social distancing of two meters when communicating is indifferent by Gen Zers (MV = 4.29; neither do nor do not intend to comply). In general, measures that deal with limiting social life seem not to be handled very well by Gen Zers. On the other hand, measures that deal with hygiene are scored higher and Gen Zers somewhat intend to follow these preventive measures. Gender differences revealed that in all cases where statistically significant differences exist, females have a higher intention to comply with non-mandatory self-protective measures. This is in line with previous studies such as Prati et al. (2022) who found that females had higher precautionary behaviours and attitudes towards quarantine restrictions. Yet, their study was not focused on Gen Zers. Also, our study is opposite to the findings of Truong et al. (2022) who found that there were no significant differences between male and female Gen Zers and their preventive behaviours towards COVID-19. However, these comparisons are made with extreme caution since they do not measure similar to this study’s issues (for example Prati et al., 2022 and Truong et al., 2022, did not measure after restriction removals intention to comply with selfpreventive non-mandatory measures).

9.6 Conclusions and Limitations This study provided an understanding of the Greek Gen Z cohort on the topic of self-reported intention to comply with non-mandatory self-protective measures of COVID-19 variants. This research adds important results which may help government officials in developing a social marketing campaign targeting the specific group of the Gen Z cohort based on their gender differences. For example, males are

9 Generation Z Intention to Comply with Non-mandatory Government. . .

161

more reluctant than females to comply with preventive measures, therefore there should be a continuous reminder that self-protective measures are significant in order for them to be accepted by their female counterparts. Also, these messages should be distributed by channels and providers trusted by Gen Zers and by males and females separately as research by Kamenidou et al. (2022) revealed. The first obvious conclusion is that intention to comply (on a 7-point Likerttype scale) with non-mandatory measures overall is small, with Gen Zers tending to agree (borderline) to only one statement- measure and tending to somewhat agree towards eleven measures. The second conclusion drawn is that where differences between male and female subjects exist, female subjects are more inclined to comply than male ones. Therefore, even though both groups should be targeted, a focus should be given to males which are riskier in behaviour than females. Based on these gender differences, focus on message distribution ought to be applied to boost Gen Zers’ intention to comply. As Gen Zers are raised in a high-tech environment, message dissemination cannot be made through the traditional information paths, such as mass media but through new types of information distribution such as digital platforms and social media (Truong et al. 2022; Kaplan, 2012). Some unavoidable limitations characterise this study which may trigger future research. This research was exploratory, employed a non-probability sampling method, and data were collected from one country and one cohort. Although it has these limitations, its contribution to the academic research is significant, offering important information about a generational cohort that has been understudied regarding COVID-19.

References Amawi, H., Abu Deiab, G. A. I., Aljabali, A., Dua, K., & Tambuwala, M. M. (2020). COVID-19 pandemic: An overview of epidemiology, pathogenesis, diagnostics and potential vaccines and therapeutics. Therapeutic Delivery, 11(4), 245–268. Anderson, R. M., Heesterbeek, H., Klinkenberg, D., & Hollingsworth, T. D. (2020). How will country-based mitigation measures influence the course of the COVID-19 epidemic? The Lancet, 395(10228), 931–934. Aquino, E. M., Silveira, I. H., Pescarini, J. M., Aquino, R., Souza-Filho, J. A. D., Rocha, A. D. S., Ferreira, A., Victor, A., Teixeira, C., Machado, D. B., & Paixão, E. (2020). Social distancing measures to control the COVID-19 pandemic: Potential impacts and challenges in Brazil. Ciencia & Saude Coletiva, 25, 2423–2446. Arif, T. B. (2022). The 501. V2 and B. 1.1. 7 variants of coronavirus disease 2019 (COVID-19): A new time-bomb in the making? Infection Control & Hospital Epidemiology, 43(3), 415–416. Bastos, S. B., Morato, M. M., Cajueiro, D. O., & Normey-Rico, J. E. (2021). The COVID-19 (SARS-CoV-2) uncertainty tripod in Brazil: Assessments on model-based predictions with large under-reporting. Alexandria Engineering Journal, 60(5), 4363–4380. Bulotait˙e, L., Žiedelis, A., Bagdži¯unien˙e, D., & Sargautyt˙e, R. (2021). Are we ready to follow COVID-19 preventive measures in postpandemic period: Predictors of preventive behavior. In Prevention in a COVID recovery society: Challenges of planning and implementing prevention approaches in the new normality: 12th EUSPR conference and members’ meeting, 29th September-1st October 2021: Online event (pp. 19–20). EUSPR.

162

I. (E). Kamenidou et al.

Burki, T. K. (2022). Omicron variant and booster COVID-19 vaccines. The Lancet Respiratory Medicine, 10(2), 17. Calina, D., Docea, A. O., Petrakis, D., Egorov, A. M., Ishmukhametov, A. A., Gabibov, A. G., Shtilman, M. I., Kostoff, R., Carvalho, F., Vinceti, M., & Spandidos, D. A. (2020). Towards effective COVID-19 vaccines: Updates, perspectives and challenges. International Journal of Molecular Medicine, 46(1), 3–16. Chaturvedi, K., Vishwakarma, D. K., & Singh, N. (2021). COVID-19 and its impact on education, social life and mental health of students: A survey. Children and Youth Services Review, 121, 105866. Cheng, K. K., Lam, T. H., & Leung, C. C. (2022). Wearing face masks in the community during the COVID-19 pandemic: Altruism and solidarity. The Lancet, 399(10336), 39–e40. Ciotti, M., Ciccozzi, M., Pieri, M., & Bernardini, S. (2022). The COVID-19 pandemic: Viral variants and vaccine efficacy. Critical Reviews in Clinical Laboratory Sciences, 59(1), 66–75. Day, M. (2020). Covid-19: Identifying and isolating asymptomatic people helped eliminate virus in Italian village. BMJ: British Medical Journal (Online), 368:m1165. https://doi.org/10.1136/ bmj.m1165 (Published 23 March 2020). Deckman, M., McDonald, J., Rouse, S., & Kromer, M. (2020). Gen Z, Gender, and COVID-19. Politics & Gender, 16(4), 1019–1027. Fang, Y., Nie, Y., & Penny, M. (2020). Transmission dynamics of the COVID-19 outbreak and effectiveness of government interventions: A data-driven analysis. Journal of Medical Virology, 92(6), 645–659. Fernandes, Q., Inchakalody, V. P., Merhi, M., Mestiri, S., Taib, N., Moustafa Abo El-Ella, D., et al. (2022). Emerging COVID-19 variants and their impact on SARS-CoV-2 diagnosis, therapeutics and vaccines. Annals of Medicine, 54(1), 524–540. Fu, Y., Jin, H., Xiang, H., & Wang, N. (2022). Optimal lockdown policy for vaccination during COVID-19 pandemic. Finance Research Letters, 45, 102123. Galasso, V., Pons, V., Profeta, P., Becher, M., Brouard, S., & Foucault, M. (2020). Gender differences in COVID-19 attitudes and behavior: Panel evidence from eight countries. Proceedings of the National Academy of Sciences, 117(44), 27285–27291. Gerace, A., Rigney, G., & Anderson, J. R. (2022). Predicting attitudes towards easing COVID19 restrictions in The United States of America: The role of health concerns, demographic, political, and individual difference factors. PLoS One, 17(2), e0263128. Government Gazette 1969. (2021). Conditions for entry into the country to limit the spread of COVID-19 coronavirus. https://www.elinyae.gr/ethniki-nomothesia/ya-d1agpoik-299252021fek-1969b-1452021 Gusev, E., Sarapultsev, A., Solomatina, L., & Chereshnev, V. (2022). SARS-CoV-2-specific immune response and the pathogenesis of COVID-19. International Journal of Molecular Sciences, 23(3), 1716. Haider, N., Osman, A. Y., Gadzekpo, A., Akipede, G. O., Asogun, D., Ansumana, R., Lessells, R. J., Khan, P., Hamid, M. M. A., Yeboah-Manu, D., & Mboera, L. (2020). Lockdown measures in response to COVID-19 in nine sub-Saharan African countries. BMJ Global Health, 5(10), e003319. Hale, T., Angrist, N., Goldszmidt, R., Kira, B., Petherick, A., Phillips, T., Webster, S., CameronBlake, E., Hallas, L., Majumdar, S., & Tatlow, H. (2021). A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker). Nature Human Behaviour, 5(4), 529–538. Haug, N., Geyrhofer, L., Londei, A., Dervic, E., Desvars-Larrive, A., Loreto, V., Pinior, B., Thurner, S., & Klimek, P. (2020). Ranking the effectiveness of worldwide COVID-19 government interventions. Nature Human Behaviour, 4(12), 1303–1312. Houvèssou, G. M., Souza, T. P. D., & Silveira, M. F. D. (2021). Lockdown-type containment measures for COVID-19 prevention and control: a descriptive ecological study with data from South Africa, Germany, Brazil, Spain, United States, Italy and New Zealand, February–August 2020. Epidemiologia e Serviços de Saúde, 30. Ibn-Mohammed, T., Mustapha, K. B., Godsell, J., Adamu, Z., Babatunde, K. A., Akintade, D. D., et al. (2021). A critical analysis of the impacts of COVID-19 on the global economy and

9 Generation Z Intention to Comply with Non-mandatory Government. . .

163

ecosystems and opportunities for circular economy strategies. Resources, Conservation and Recycling, 164, 105169. Iglesias-Pradas, S., Hernández-García, Á., Chaparro-Peláez, J., & Prieto, J. L. (2021). Emergency remote teaching and students’ academic performance in higher education during the COVID-19 pandemic: A case study. Computers in Human Behavior, 119, 106713. Immel, L., Neumeier, F., & Peichl, A. (2022). The unequal consequences of the Covid-19 pandemic: Evidence from a large representative German population survey. Review of Income and Wealth. JHU. (2022). COVID-19 Dashboard by the Center for Systems Science and Engineering (CSSE) at John Hopkins University. Available online: http://coronavirus.jhu.edu/map.html. Accessed on 7 June 2022. Jørgensen, F., Bor, A., & Petersen, M. B. (2021). Compliance without fear: Individual-level protective behaviour during the first wave of the COVID-19 pandemic. British Journal of Health Psychology, 26(2), 679–696. Jose, S. (2022). COVID vaccine and generation Z – A study of factors influencing adoption. Young Consumers, 23(1), 16–32. https://doi.org/10.1108/YC-01-2021-1276 Kamenidou, I. E., Mamalis, S., Stavrianea, A., Bara, E. Z., Mylona, I., & Pavlidis, S. (2022). Risk communication based on gender differences of COVID-19 related trusted information sources. Medical Research Archives, 10(4). Kamenidou, I. E., Stavrianea, A., & Liava, C. (2020a). Achieving a Covid-19 free country: Citizens preventive measures and communication pathways. International Journal of Environmental Research and Public Health, 17(13), 4633. https://doi.org/10.3390/ijerph17134633 Kamenidou, I. E., Stavrianea, A., Mamalis, S., & Mylona, I. (2020b). Knowledge assessment of COVID-19 symptoms: Gender differences and communication routes for the generation Z cohort. International Journal of Environmental Research and Public Health, 17(19), 6964. Kaplan, A. M. (2012). If you love something, let it go mobile: Mobile marketing and mobile social media 4x4. Business Horizons, 55(2), 129–139. Karabay, M. E., Ensari, M. S., ¸ & Yurtkoru, E. S. (2022). An empirical investigation of the impact of COVID-19 pandemic on university students. International Journal of Education and Social Science Research, 5(2), 284–307. Kent, R. (1993). Marketing research in action. Routledge. Kunal, S., Gupta, K., & Ish, P. (2021). COVID-19 variants in India: Potential role in second wave and impact on vaccination. Heart & Lung: The Journal of Cardiopulmonary and Acute Care, 50(6), 784–787. Lehmann, D. R., Cupta, S., & Steckel, J. H. (1998). Marketing research. Addison Wesley Longman. Li, W., Zhang, B., Lu, J., Liu, S., Chang, Z., Peng, C., Liu, X., Zhang, P., Ling, Y., Tao, K., & Chen, J. (2020). Characteristics of household transmission of COVID-19. Clinical Infectious Diseases, 71(8), 1943–1946. Lipsitch, M., & Dean, N. E. (2020). Understanding COVID-19 vaccine efficacy. Science, 370(6518), 763–765. Liu, Q., & Cao, L. (2022). Modeling time evolving COVID-19 uncertainties with density dependent asymptomatic infections and social reinforcement. Scientific Reports, 12(1), 1–14. Lurie, N., Saville, M., Hatchett, R., & Halton, J. (2020). Developing Covid-19 vaccines at pandemic speed. New England Journal of Medicine, 382(21), 1969–1973. Madabhavi, I., Sarkar, M., & Kadakol, N. (2020). COVID-19: A review. Monaldi Archives for Chest Disease, 90(2). https://doi.org/10.4081/monaldi.2020.1298 Mahase, E. (2021). Covid-19: How many variants are there, and what do we know about them? BMJ, 2021, 374. https://doi.org/10.1136/bmj.n1971. (Published 19 August 2021). Munro, A. P., Feng, S., Janani, L., Cornelius, V., Aley, P. K., Babbage, G., Baxter, D., Bula, M., Cathie, K., Chatterjee, K., & Dodd, K. (2022). Safety, immunogenicity, and reactogenicity of BNT162b2 and mRNA-1273 COVID-19 vaccines given as fourth-dose boosters following two doses of ChAdOx1 nCoV-19 or BNT162b2 and a third dose of BNT162b2 (COV-BOOST): A multicentre, blinded, phase 2, randomised trial. The Lancet Infectious Diseases, 22, 1131.

164

I. (E). Kamenidou et al.

Ndwandwe, D., & Wiysonge, C. S. (2021). COVID-19 vaccines. Current Opinion in Immunology, 71, 111–116. Or, Z., Gandré, C., Zaleski, I. D., & Steffen, M. (2022). France’s response to the Covid-19 pandemic: Between a rock and a hard place. Health Economics, Policy and Law, 17(1), 14– 26. Osuchowski, M. F., Winkler, M. S., Skirecki, T., Cajander, S., Shankar-Hari, M., Lachmann, G., Monneret, G., Venet, F., Bauer, M., Brunkhorst, F. M., & Weis, S. (2021). The COVID-19 puzzle: Deciphering pathophysiology and phenotypes of a new disease entity. The Lancet Respiratory Medicine, 9(6), 622–642. Pham, T. D., Dwyer, L., Su, J. J., & Ngo, T. (2021). COVID-19 impacts of inbound tourism on Australian economy. Annals of Tourism Research, 88, 103179. Prati, G., Stefani, S., & Barbieri, I. (2022). Women tend to perceive greater risks associated with the COVID-19 outbreak and are more likely to follow precautionary measures. European Journal of Health Psychology, 29(2), 99–106. https://doi.org/10.1027/2512-8442/a000089 Ralli, M., Morrone, A., Arcangeli, A., & Ercoli, L. (2021). Asymptomatic patients as a source of transmission of COVID-19 in homeless shelters. International Journal of Infectious Diseases, 103, 243–245. Ramayanti, I., Anggraini, W., Qonitah, F. F., Ghiffari, A., & Prameswarie, T. (2021). COVID-19 health protocol and religious activities: Knowledge, attitude, and compliance among generation Z. Bioscientia Medicina: Journal of Biomedicine and Translational Research, 5(7), 685–692. Rosenblum, H. G., Hadler, S. C., Moulia, D., Shimabukuro, T. T., Su, J. R., Tepper, N. K., Ess, K. C., Woo, E. J., Mba-Jonas, A., Alimchandani, M., & Nair, N. (2021). Use of COVID-19 vaccines after reports of adverse events among adult recipients of Janssen (Johnson & Johnson) and mRNA COVID-19 vaccines (Pfizer-BioNTech and Moderna): Update from the Advisory Committee on Immunization Practices—United States, July 2021. Morbidity and Mortality Weekly Report, 70(32), 1094. Roy, B., Dhillon, J. K., Habib, N., & Pugazhandhi, B. (2021). Global variants of COVID-19: Current understanding. Journal of Biomedical Science, 8(1), 8–11. Rubin, R. (2021). COVID-19 vaccines vs variants—Determining how much immunity is enough. JAMA, 325(13), 1241–1243. Samui, P., Mondal, J., & Khajanchi, S. (2020). A mathematical model for COVID-19 transmission dynamics with a case study of India. Chaos, Solitons & Fractals, 140, 110173. Shah, A. U. M., Safri, S. N. A., Thevadas, R., Noordin, N. K., Abd Rahman, A., Sekawi, Z., Ideris, A., & Sultan, M. T. H. (2020). COVID-19 outbreak in Malaysia: Actions taken by the Malaysian government. International Journal of Infectious Diseases, 97, 108–116. Shaw, R., Kim, Y. K., & Hua, J. (2020). Governance, technology and citizen behavior in pandemic: Lessons from COVID-19 in East Asia. Progress in Disaster Science, 6, 100090. Shereen, M. A., Khan, S., Kazmi, A., Bashir, N., & Siddique, R. (2020). COVID-19 infection: Emergence, transmission, and characteristics of human coronaviruses. Journal of Advanced Research, 24, 91–98. Simic, M. L., & Pap, A. (2022). Generation Z and Covid-19 pandemic: Perceptions and prosocial behavior analysis. In Economic and social development (book of proceedings), 79th international scientific conference on economic and social (p. 129). Spector, P. (1992). Summated rating scale construction. SAGE Publications. Strauss, W., & Howe, N. (2020). Generation, Z. Available online: http://incomeresult.com/ generation-z/#cite_note-McCrindleAU-27. Accessed on 7 Aug 2020. Takashita, E., Kinoshita, N., Yamayoshi, S., Sakai-Tagawa, Y., Fujisaki, S., Ito, M., et al. (2022). Efficacy of antibodies and antiviral drugs against Covid-19 omicron variant. New England Journal of Medicine, 386, 995. Torjesen, I. (2021). Covid-19: Delta variant is now UK’s most dominant strain and spreading through schools. Tregoning, J. S., Flight, K. E., Higham, S. L., Wang, Z., & Pierce, B. F. (2021). Progress of the COVID-19 vaccine effort: Viruses, vaccines and variants versus efficacy, effectiveness and escape. Nature Reviews Immunology, 21(10), 626–636.

9 Generation Z Intention to Comply with Non-mandatory Government. . .

165

Truong, N. X., Ngoc, B. H., & Ha, N. T. (2022). The impacts of media exposure on COVID19 preventive behaviors among Vietnamese people: Evidence using expanded protection motivation theory. SAGE Open, 12(2), 21582440221096129. Van Rooij, B., de Bruijn, A. L., Reinders Folmer, C., Kooistra, E. B., Kuiper, M. E., Brownlee, M., Olthuis, E., & Fine, A. (2020). Compliance with COVID-19 mitigation measures in the United States. Amsterdam law school research paper. Vitiello, A., & Ferrara, F. (2021). Brief review of the mRNA vaccines COVID-19. Inflammopharmacology, 29(3), 645–649. Wee, L. E. I., Conceicao, E. P., Tan, J. Y., Magesparan, K. D., Amin, I. B. M., Ismail, B. B. S., Toh, H. X., Jin, P., Zhang, J., Wee, E. G. L., & Ong, S. J. M. (2021). Unintended consequences of infection prevention and control measures during COVID-19 pandemic. American Journal of Infection Control, 49(4), 469–477. WHO. (2022). Tracking SARS-CoV-2 variants.https://www.who.int/activities/tracking-SARSCoV-2-variants. Accessed on 7 Aug 2020. Wu, J. S., Font, X., & McCamley, C. (2022). COVID-19 social distancing compliance mechanisms: UK evidence. Environmental Research, 205, 112528. Xiao, T., Wang, Y., Yuan, J., Ye, H., Wei, L., Liao, X., Wang, H., Qian, S., Wang, Z., Liu, L., & Zhang, Z. (2021). Early viral clearance and antibody kinetics of COVID-19 among asymptomatic carriers. Frontiers in Medicine, 8, 595773. Yang, Z., Zhang, S., Tang, Y. P., Zhang, S., Xu, D. Q., Yue, S. J., & Liu, Q. L. (2022). Clinical characteristics, transmissibility, pathogenicity, susceptible populations, and re-infectivity of prominent COVID-19 variants. Aging and Disease, 13(2), 402. Yuki, K., Fujiogi, M., & Koutsogiannaki, S. (2020). COVID-19 pathophysiology: A review. Clinical Immunology, 215, 108427.

Chapter 10

The Spatial Distribution of the Population in Peninsular Spain: An Evolution of a Permanent Nature Daniel del Castillo Soto and Thomas Baumert

Ignorance of the structure of our population is a source of errors in our economic policy. (Perpiñá Grau, 1954)

Abstract In the mid-1930s, the young Spanish economist Román Perpiñá Grau began his research into the spatial distribution of the Spanish population. Considered as one of the pioneers in studies of the economic structure of Spain, he was also recognised as a master by some of the most prominent figures in Spain in this field. Two works will be the basis of our study: De economía hispana (1935) and Corología (1954). In this paper we will analyse how his research on the behaviour of the population has stood the test of time, comparing the results he presented with those obtained today, almost 90 and 70 years after the dates of their respective publications. Keywords Population density · Spatial distribution · Economic structure · Spain · Román Perpiñá Grau

10.1 Introduction The purpose of this research is the analysis and evaluation of the permanence – read, current validity – of the research carried out more than 80 years agoby

D. del Castillo Soto () Universidad Camilo José Cela, Madrid, Spain T. Baumert Universidad Nebrija, Madrid, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_10

167

168

D. del Castillo Soto and T. Baumert

Román Perpiñá Grau1 regarding the evolution and spatial distribution of the Spanish population. Perpiñá continued working on this topic during the 1950s and 1960s, deserving him to be considered one of the pioneers of population studies in Spain. At the beginning of 2022, the population census in Spain reached 47,435,597 inhabitants – 50 years ago this figure was just under 35 million. The economic growth experienced by Spain since the early 1960s has led to an increase in life expectancy which, in turn, has resulted in significant population growth. This positive relationship between economic growth and increases in population and life expectancy is not exclusive to Spain, but rather the relationship between these variables is common throughout the world (Deaton, 2015, pp. 47–63). Another factor contributing to the increase in population is the transition experienced by Spain from being a source of emigration to becoming a net receiver of immigrants.2 According to Perpiñá, the population of a country forms part of its economic structure (Perpiñá Grau, 1954) although not necessarily exclusively, as it will also form part of other structures. Elsewhere he will state that “the ignorance of the structure of our population is a source of errors of [Spanish] economic policy” (Perpiñá Grau, 1954, p. 15). Based on this observation, in this paper, we will study the nature of the Spanish population in relation to its geographical configuration, in such a way that, together with knowledge of other characteristics, it will allow to obtain relevant information that may help economic and political agents in their decision-making.3 Spain is a country with a very varied geography. Perpiñá was one of the first to attach importance to natural conditions, arguing that their location and characteristics clearly influence the spatial distribution of its population (Velarde Fuertes, 2022). It is not an exclusive characteristic of Spain that its population is distributed in a heterogeneous way, nor that its densities by regions have very significant differences, so the interest of this study does not lie in the singularity

1 Román Perpiñá Grau (1902–1991). Born in Reus, this Catalan was one of the pioneers in the development of economic science in Spain. Focusing on studies on economic structure, he had a very prolific activity in terms of economic publications, not only with research on economic structure. His research showed a great philosophical influence of Greek thinkers and he was also a fervent Catholic. His research on the economic development of Spain, his defence of free international trade and his criticism of state protectionism of products stand out. Among other recognitions and merits, he was awarded the Prince of Asturias prize for social sciences in the first edition of this award in 1981 and was also named Doctor Honoris Causa by the Universities of Valencia and Barcelona. 2 As stated in the publication Memoria gráfica de emigración española “Although today Spain is mainly known (especially in Latin America) as a destination country for immigrants from all over the world, we cannot forget that Spain has been and continues to be a “country of emigration”. Spanish emigration has been a political, social and economic phenomenon that has characterised our history, accentuated from the second half of the 19th century until beyond the mid-20th century” (Rivas & Rodríguez, 2009, p. 138). 3 One might think that population distribution is not a field of research exclusive to economics, and that it would fit better in sociology, but we find economic variables that affect it to such a degree as to be considered prevalent.

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

169

of the Spanish case (cf. Tamames & Rueda, 2022, chapter 4), but in the possible application of its study procedure and conclusions to other countries. One of the problems currently facing Spain in relation to the characteristics of its population is the concern about the ageing of its population. This interest, far from being a novelty, has already given rise to detailed studies on the relationships between the different levels of the population pyramid, taking into account the interregional aspect, that is, applying a spatial criterion (De Miguel, 1992; Tamames & Rueda, 2022). Regarding the spatial imbalances, the concept of an “Empty or Emptied Spain” has also recently gained momentum, reflecting the growing political and social concern about the imbalances in population densities between different Spanish regions. In the following pages, we will try to show that this phenomenon, far from being recent, is part of a process of structural development that has taken place steadily over the last six decades, thus modifying the characteristics of the spatial distribution of the population that had hardly changed between 1900 and 1960.

10.2 Perpiñá Grau and His Population Studies There are two main works in which Perpiñá deals with population studies as part of the economic structure in Spain, namely, De economía hispana and Corología. Teoría estructural y estructurante de la población de España.1900–1950. The former had a great influence on the development of economic science in Spain, while the latter – in which he continued the research begun in the first – presents a more in-depth analysis with variations on the territorial divisions with respect to the first.

10.2.1 De economía hispana Let us briefly summarise the origin and the issues addressed by Perpiñá in his most famous publication, and then focus on the analysis of the points related to the distribution of the population in the national peninsular territory. The origin of De economía hispana can be found in two moments and in two cities. The first was Barcelona, at the time of the Universal Exhibition held in the Catalonian capital in 1929. Perpiñá, who at that time was head of the economic

170

D. del Castillo Soto and T. Baumert

studies department at CHADE-SOFINA4 – a company run by Francisco Cambó5 – attended the event. There he carefully observed a model of the Iberian Peninsula, presented by his company at the international exhibition, which clearly showed its orographic composition. This view led Perpiñá to consider how orography could influence Spanish economic development and the differences that this could generate between peripheral and inland Spain. The second key moment took place outside Spain. At the beginning of the 1930s, Perpiñá visited Germany for the second time (which he already knew from his time as a doctoral student, when he visited the German country to gather information for the preparation of his thesis in the second half of the 1920s). After this first stage, he returned as a doctoral student and an established economist during the summers of 1933 and 1934 to the headquarters of the Institut für Weltwirtschaft (Institute for World Economics) in the northern German city of Kiel. This made it possible for him to work alongside, among others, professors of the stature of Harms6 and Preadhol.7 ,8 In addition to writing De economía hispana, as a result of his stays in Germany, he also translated Harbeler’s work on the theory of international trade, which was published in its Spanish version and presented as an appendix to this work for the first time in 1936. It was Professor Harms who pioneered the application of the term “structure” to economics. Harms asked Perpiñá different questions about the Spanish economy. Perpiñá’s answers, and the information that he had carried out a more exhaustive and systematic analysis of the subject at the Centro de Estudios Económicos Valencianos,9 led the German scientist to show interest in Perpiñá’s

4 CHADE

(Compañía Hispano Americana de Electricidad) SOFINA (Societé Financière de Transports et d’Enterprises Industrielles). After the Great War, German shareholders were forced by armistice conditions to sell their shares in neutral countries. This is why CATE (Compañía Alemana Transatlántica de Alemania) became the property of SOFINA, a Spanish consortium based in Brussels. 5 Francisco Cambó Batlle (1876–1947) Catalan lawyer and politician of a conservative and Catalanist persuasion. He was one of the founders of the Regionalist League party and a member of parliament on several occasions. His postulates could be summarised as an attempt to make Catalonia an influential region in the future of Spanish politics as a whole. Defender of protectionism and director of CHADE when Román Perpiñá, who was hired by Cambó himself, was in charge of the company’s studies section. 6 Bernard Harms (1876–1939). German economist who founded the Institute of World Economics in Kiel in 1914. This institution has been a world leader in the field of economic and business research for more than 100 years. 7 Andreas Preadhol (1892–1974) was a German economist who headed the Kiel Institute, replacing Harms, and later became rector of the University of Kiel. He managed to maintain the research environment in Kiel in the difficult years between 1935 and 1945. 8 In 1935 he had to cancel another planned trip when he had to go to London as an advisor to a commission negotiating a trade treaty between Spain and Great Britain. 9 Perpiñá was part of its foundation and the first person to direct it in 1929. Here the German influence on Perpignà was also present in his attempt to replicate the generation of data files for the production of statistics, similar to those he had used in Germany for his thesis.

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

171

work (Perpiña Grau, 1972, p. 27). Harms then suggested him that he might write a more extensive monograph on the subject on the economic structure of Spain, an activity that Perpiñá would start as soon as he returned to Valencia. The first edition (in German) was published in January 1935 in the journal Weltwirtschaftlisches Archiv. Perpiñá’s conclusions on the factors holding back Spain’s economic development in comparison with other Western countries were that the main problems facing Spain were of a structural nature and not of a conjunctural nature. As most economic research in the international scientific community at that time was based on short-term studies, Perpiñá’s contributions went against the mainstream and were therefore certainly original. Perpiñá defined the economic structure of a country as “the result of internal natural factors and economic policy, as well as the external influences produced by both factors” (Perpiña Grau, 1972, p. 37). In relation to natural factors (that Perpiñá termed as “infrastructure”), these have different characteristics, depending on the territory. He exposes the problem of the difficult orography and climate, which means that Spain is divided into a central and a peripheral Spain. The former was less populated according to its size, lacked industry and had low agricultural productivity due to its altitude. The latter was more industrialised, had more productive land and higher export capacity, a larger population and a better standard of living. In short, Perpiñá argues in his work that the different characteristics of the natural infrastructure determine to a large extent the differences in development between these “two Spains”, a conception that he also extended to the differences between different nations. The great level of detail that the author presents with data on production and population density was certainly novel for the time and empirically supported the division between the coast and the interior. Regarding economic policy, as the second component of the structure, he describes it as the means or systems for the exploitation and management of wealth, with the aim of preserving and increasing the well-being and civilisation of a territory (Perpiña Grau, 1972, p. 32). Perpiñá points to protectionism as the main obstacle to economic development, even more important than the problems caused by an unfavourable infrastructure. According to Perpiñá, the autarkic policy in Spain was not motivated by scientific arguments, but by the individual interests of some businessmen and politicians who were positively influenced by this system of tariff protection in some cases, or by import monopolies – such as that of oil – in others. These economic guidelines had caused a major brake on the development of an economy and did not benefit the bulk of the Spanish population, but only a few (Perpiña Grau, 1972, pp. 72–89). For Perpiñá, protectionism encouraged the lower productivity of domestic agriculture, while at the same time preventing the potential of the industry from being developed. The solution proposed by the author consisted basically in opening up the economy. Another mistake Perpiñá pointed out was the aim for production to be national. The important thing was that the product should be Spanish rather than competitive, which reduced the possibility of achieving better results in terms of development and growth (Perpiña Grau, 1972, p. 64).

172

D. del Castillo Soto and T. Baumert

In his research, Perpiñá also explored issues related to international trade. Faced with the question of whether the degree of well-being achieved by a country can be something permanent and progressive, achieved individually, or whether in order to maintain or increase this level it must be at the expense of third countries, the author advocates international cooperation and openness to foreign markets.10 But let us return to population, which Perpiñá includes among the natural factors, and to the study of its spatial distribution the author devotes a chapter of De economía hispana, analysing its distribution. For this purpose, in the creation of the two regions mentioned, he uses the provincial limits, adding the corresponding provinces to each of the two zones. The division of peninsular Spain is very elementary: the peripheral region is made up of the coastal provinces, including two inland provinces, Orense and Vitoria. The rest of the provinces form part of the inland region. The interior region occupies 69% of the territory of mainland Spain, leaving the periphery with the remaining 31%. The proportions of the population in the 1930s were maintained with minimal variations since the beginning of the century in the following proportion: 52% of the population resided in the peripheral area and 48% in the interior provinces. This distribution remained practically unchanged until 1960. Today, the differences are even greater, with 58% of the population living in the coastal area and 42% in the inland region. Looking at the evolution of these proportions by decades, it can be seen that the increase in the gap began in the 1960s over a period of 20 years and from then on, the rate of growth of the gap decreased more slowly but steadily until present day, reaching an increase in the difference between regions of 300% since the date of Perpiñá’s analysis (Figs. 10.1 and 10.2). 0.7

peripheral ratio 0.6 0.5 0.4

center ratio

0.3 0.2 0.1 0

1900 1910 1920 1930 1940 1950 1960 1970 1981 1991 2001 2011 2021

Fig. 10.1 Evolution of the distribution ratio of the population in the two areas. (Source: Own elaboration) 10 Perpiñá’s

concern for the future of the world economy was a constant element of his research, and became a topic to which he would return throughout his life.

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

180

173

peripheral density

160 140 120 100

80 60 40 20

center density

0 1900 1910 1920 1930 1940 1950 1960 1970 1981 1991 2001 2011 2021

Fig. 10.2 Evolution of population density in the two areas. (Source: Own elaboration)

10.2.2 Chorology. Structural and Structuring Theory of the Population in Spain. 1900–1950 Published in 1954, this work presents us with a Perpiñá at the height of his intellectual maturity. Many vicissitudes would have happened in these almost 20 years of difference between the two works. A civil war in between and a move from Valencia to Madrid due to professional changes. On the verge of opting, in 1955, for the first professorship in economic structure, which he finally did not obtain – something that would be detrimental to spatial studies in Spanish universities (Del Río Disdier, 2015, 2016, 2017) – Perpiñá culminates with this work a trajectory that had begun in the 1940s. Perpiñá’s idea of division into two areas of different population density had passed the test of permanence when in 1950 the same proportions of population distribution as described in the previous section were met. However, it did not seem sufficient for the study to stop at that point and the author himself referred to the insufficiency of the model, assuming that a second step had to be taken to get closer to the structural reality of the Spanish population. In this second step, Perpiñá looked for the influence of the large population centres on the configuration of densities in the neighbouring provinces. To begin his study, Perpiñá divides peninsular Spain in the following way: the first space will be a central circle with a radius of approximately 13 km, which will contain the municipal limits of the city of Madrid. The following divisions are obtained as a result of projecting circles from the centre of the initial circle with radii of 100, 200, 300 and 400 km, resulting in five “rings” (to use the term employed by Perpiñá). The first one, with a width of 87 km after deducting the central circle of the capital. The second ring will be between the distances of 100 km and 200 km. Both the central circumference and the first two rings would be entirely on national

174 Table 10.1 Description of surface and density characteristics of the concentric divisions of Román Perpiñá

D. del Castillo Soto and T. Baumert

Centre circle Ring 1 Ring 2 Ring 3 Ring 4 Ring 5

Surface area km2 576 30,840 94,248 140,600 158,000 67,000

Area % 0,1% 6% 19% 29% 32% 14%

inhab/km2 2810 33 25 37 59 108

Source: Own elaboration

territory. The third ring, between 200 and 300 km, is practically integrated into Spain, except for a minimal part that would enter Portuguese territory. The last two rings would contain Spanish and Portuguese territories and maritime areas. The fourth is 100 km wide, between 300 and 400 km, and the fifth ring, whose limits go from the upper limit of the previous ring, 400 km from the centre, to the Spanish peninsular surface furthest from the concentric point of the rings. The total and relative surface areas and population densities of the different areas are as follows (Table 10.1)11 : The centre-periphery division did not show the behaviour of population density within the two areas indicated. When each of the zones is studied in more detail, a sequence is observed in which there is no increase in density from the centre towards the coasts. The higher density marked by the urban nucleus of Madrid gives way to a lower density in the first ring and passes to the lower density of the second ring. That is to say, the maximum density of the central population centre is surrounded by an area of intermediate density, to continue to the areas of lower density in the second ring. From this the densities of the third, fourth and fifth rings increase. The second corona (200–300 km) has the lowest density. Starting from the high densities of the central core and the peripheral settlements, these densities decrease towards the intermediate point of the second ring. This second ring would represent the frontier space between the attraction of the population of the centre and the space of attraction of the population of the periphery. In Perpiñá’s own words: The Peninsula is presented to us theoretically, with the comprehensive simplification of the phenomena essential to the real Spanish structure, as a circle divided into seven spaces or coras, a central hexagonal cora and six semi-hexagons regularly tangent to its sides; each cora has a hexagonal centre, a dasicora, and each dasicora is surrounded by its respective areocora, with increasingly smaller densities (Perpiñá Grau, 1954, p. 33).

Based on the data and density behaviour according to distances from the city centre of Madrid, the next step of the research is to divide the national and Portuguese territory into seven areas, resulting in the so-called Perpignan hexagon. Based on the presence of six markets of influence for each area, six national and one Portuguese, as well as six aligned with the vertices of the hexagon and the sea and

11 The

calculations are presented by Perpiñá using a map from the Instituto Geográfico y Catastral and the INE census densities (Perpiñá Grau, 1954, p. 25).

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

175

Map 1 Representation of the Perpignan hexagon and the concentric circles. (Source: Own elaboration) (In the presentation of his hexagon, Perpiñá showed a theoretical model in which the peninsula was represented in a circular form, to show the division made (Perpiñá Grau, 1954, p. 32). We have considered it interesting to transfer the division proposed by the author to the map of Spain)

one in the centre of the hexagon. Perpignan names each area with the word “cora” or jora (from the Greek χωρα) ´ which means zone, region or field. It is very common for the Spanish economist to use his knowledge of Greek to name new concepts in his research. Therefore, the first result is one Portuguese cora, and six Spanish cora. Each Spanish cora will be formed by the grouping of provinces, which he will differentiate into two types according to their density, naming as dasicoras those of high density and areocoras those of lower density. Again, Perpiñá uses Greek for the creation of nouns, since dasys in Greek means dense and araiós means sparse, sparsely populated. Thus, by combining dasys and araiós with jora we get dasicoras and areocoras. Perpiñá named each cora after a seaport of reference in the coastal coras, so that the five coastal coras would be Vigo, Bilbao, Barcelona, Valencia and Cadiz. The central cora will be called the cora of Madrid, as this is the only dasicora in the central cora. In each cora there will be one or two dasicoras. Perpiñá explains that the use of the names of important ports in each area is due to the fact that these ports are a representation of the main national markets. In what follows, and in order to avoid redundancies in the names of the coras with the names of some of the

176

D. del Castillo Soto and T. Baumert

NW

N

S

NE

E

C

30.00% 25.00% 20.00% 15.00% 10.00% 5.00% 0.00%

Fig. 10.3 Percentage distribution of the territory by coras. (Source: Own elaboration)

dasicoras integrated in them, we will rename the coras according to their cardinal location in the peninsula, instead of the names of ports and cities. Therefore, the cora of Vigo will become the northwest cora, the cora of Bilbao the north cora, the cora of Barcelona the northeast cora, the cora of Valencia the east cora, the cora of Cadiz the south cora and finally the cora of Madrid the central cora. The following is the composition of the six coras with their respective dasicoras and areocoras (Table 10.2). The territorial division of the coras in relation to the Spanish mainland territory they occupy in percentage terms is as follows: Central Cora 29%; South Cora 18%; East Cora 17%; Northwest Cora 13%; Northeast Cora 13% and North Cora 10% (it should be remembered that Perpiñá’s criteria for dividing areas only took into account aspects of population density of the Spanish provinces and the proximity between them). Hence, when Perpiñá explains that the different zones were each influenced by the main market, it could be said that this division was mainly for economic reasons. The political and cultural aspects of the different areas, which do seem to have conditioned to a certain extent the division of the territories of peninsular Spain into the fifteen autonomous communities present in it, would not therefore have had any influence on his model. The following two illustrations show the differences in the proportion of size of the coras between them and those of the autonomous communities (Figs. 10.3 and 10.4). The growth of the Spanish population implied a logical increase in the population density of all the coras. As can be seen, these increases have had a similar behaviour in terms of their rate of growth in five of them, forming a channel in which they are all channelled, except for the cora of Barcelona, which will grow at a faster rate,

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

177

25.00%

20.00% Andalucía Aragón Asturias 15.00%

Cantabria Castilla y León Castilla - La Mancha

Cataluña

10.00%

Comunidad Valenciana Extremadura Galicia 5.00% Madrid Murcia Navarra

La Rioja

País Vasco

Murcia

Navarra

Galicia

Madrid

Extremadura

Comunidad Valenciana

Cataluña

Castilla y León

Castilla - La Mancha

Cantabria

Aragón

Asturias

Andalucía

0.00%

País Vasco La Rioja

Fig. 10.4 Percentage distribution of territory by Autonomous Community. (Source: Prepared by the authors)

leaving the channel in the 1970s and increasing the differences with the rest until the present day. The other five have maintained a similar growth rate over the more than 120 years analysed, which shows that the dynamics of change is structural in nature and gives validity and permanence to Perpiñá’s conclusions in his study in the mid-twentieth century (Figs. 10.5 and 10.6).

10.2.3 The Development of the Dasicoras Thoughout Time As described above, the coras are divided according to their population density into dasicoras and areocoras. The dasicoras are, therefore, the ten provinces with the

178

D. del Castillo Soto and T. Baumert

Table 10.2 Distribution of the coras by provinces, differentiating the dasicoras and areocoras in each of them Cora Cora de Vigo/Northwest

Dasicoras PontevedraLa Coruña

Cora de Bilbao/North

VizcayaGuipúzcoa

Cora de Barcelona/Northeast

Barcelona

Cora de Valencia/East

ValenciaAlicante

Cora de Cádiz/South

CadizMalaga

Cora de Madrid/Central

Madrid

Areocoras Lugo Orense Asturias Leon Zamora Álava Santander La Rioja Palencia Burgos Navarre Tarragona Lérida Gerona Huesca Zaragoza Castellón Teruel Basin Albacete Murcia Almeria Grenada Jaén Cordoba Seville Huelva Avila Segovia Soria Guadalajara Valladolid Salamanca Cáceres Badajoz

Source: Own elaboration

highest population density in the mid-1950s, the same ten as at the beginning of the 20th century and the same today. If we compare them with each other, we can see that they have experienced alternations in their positions, but they have always maintained the highest density within the Spanish provinces as a whole. As can be seen in the following tables, not all of them have evolved in the same way. The dasicora of Madrid has grown more than the others and is currently the one with the highest density, starting at the beginning of the last century in the fifth position and being in the third place when Perpiñá presented his research. The dasicora of Barcelona was the one with the highest density in 1950, having been the second in

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

1900

1920

1940

1960

1980

2000

179

2020

140

120

Central

100

Northwest North

80 South 60

Northeast

40

East

20 0

Fig. 10.5 Evolution of population density per coras from 1900 to 2021. (Source: Own elaboration)

1900 1910 140

1920

120

1930

100 80 60 40 20 0

1940 2021 2011 2001 1991 1981 1970 1960 1950 1940 1930 1920 1910 1900

1950 1960 1970 1981 1991 2001 2011 2021

Fig. 10.6 Evolution of population density per coras from 1900 to 2021. (Source: Own elaboration)

1900 and also at present. The Basque dasicoras have remained among those with the highest density throughout the entire period studied. On the one hand, Biscay is the third with the highest density nowadays, being the one with the highest density at the beginning of the twentieth century, to be in second place in the middle of the twentieth century. On the other hand, the Guipuzcoa region is in fourth place in all periods. The dasicoras of Levante, Valencia and Alicante – which in 2021 follow the Basque dasicoras in density – have also experienced growth relative to the Galician dasicoras, Pontevedra and La Coruña. In 1900, Valencia was in eighth place and is now two positions higher, and Alicante went from seventh to fifth place in these

180

D. del Castillo Soto and T. Baumert

Density of dasicoras 1900 Vizcaya Barcelona Pontevedra Guipuzcoa Madrid La Coruña Alicante Valencia Málaga Cádiz Territorio Nacional 0

20

40

60

80

100

120

140

160

Fig. 10.7 Coras ordered by population density in 1900. (Source: Own elaboration)

Density of dasicoras 1950 Barcelona Vizcaya Madrid Guipuzcoa Pontevedra Valencia La Coruña Alicante Málaga Cádiz Territorio Nacional 0

50

100

150

200

250

300

350

Fig. 10.8 Coras ordered by population density in 1950. (Source: Own elaboration)

121 years. Pontevedra started in third place and is now in eighth and La Coruña has gone from sixth to tenth and last. The two Galician dasicoras are the ones that have increased their density at a slower rate than the rest. Finally, the two Andalusian dasicoras, which had the lowest densities in 1900 and 1950, ninth for Malaga and tenth for Cadiz, are currently seventh and ninth, respectively (Figs. 10.7, 10.8, and 10.9). As a summary of the behaviour of the population and in order to follow the evolution of the population density of the dasicoras, Perpiñá presents his dasicoric

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

181

Density of dasicoras 2021 Madrid Barcelona Vizcaya Guipuzcoa Alicante Valencia Málaga Pontevedra Cádiz La Coruña Territorio Nacional 0

100

200

300

400

500

600

700

800

900

Fig. 10.9 Coras ranked by population density in 2021. (Source: Own elaboration)

coefficient or spatial concentration of the population formula (Perpiñá Grau, 1954, p. 134). This coefficient is the result of dividing the density of the dasicoras by the total population density, and the results showed that the density in 1950 was 2.7 times higher in the dasicoras than in the national territory as a whole. Perpiñá’s forecast for the year 2000 was that the dasicore coefficient would be set at 3.3, making a population projection of 41,557,000 inhabitants, with a dasicore density of 279 and 82 in the peninsular territory. (Perpiñá Grau, 1954, p. 142). His prediction regarding the total population resulted quite accurate, since according to the official population data of 2001, the Spanish population reached 41,116,842 inhabitants, coinciding with 99% of the prediction made by Perpiñá almost 70 years earlier. For the period 1950, the dasicoric coefficient gives 2.89 for the year 1950 and 3.9 for the year 2000. The increase predicted by Perpiñá was 22.2% and the result for the year 2000 was 35.6%. Although the size of the increase is not exactly the same as the one predicted by Perpiñá, it can be affirmed that the dynamics of change is the one he advocated, as the densities of dasicoras increase in relation to the total population density, with the consequent increase in the dasicoric coefficient.

10.3 Roman Hispania and the Hispania of Román Perpiñá It is interesting to compare the Perpignan model with the division of the Roman province of Hispania at the end of the third century AD. The reason for this choice is that it is the result of almost five centuries of Roman administration and its

182

D. del Castillo Soto and T. Baumert

borders evolved into five regions12 : Gallaecia, Lusitania, Baetica, Carthaginensis and Tarraconensis. One difference is that Perpiñá does not take Portuguese territory into account in his study, although he does point out that a Portuguese cora with the Lisbon province as dasicora would fit perfectly into the model. On the other hand, the political centre of the Roman government was on the other side of the Mediterranean, so that the centre of the peninsula was not important as a market, while in Perpignan’s model the capital located in Madrid exerts an influence on the rest of the country that this area did not have in ancient times. Even so, there are some spatial coincidences between the two models or divisions, one of which was real and practical in its time, and the other only theoretical. The Roman provinces of Gallaecia and Baetica coincide almost completely with the coras of Vigo and Cadiz, respectively. The province Tarraconensis conforms to the sum of the coras of Barcelona and Bilbao with minor differences. The Carthaginensis contains the cora of Valencia except for practically the whole of the province of Castellón, and the northern half of the province of Teruel, both included in the Roman Tarraconensis, and the eastern half of the territory of the cora of Madrid and the eastern territories of the provinces of Jaén, Granada and Almería, which are part of the Cádiz cora. The western half of the Madrid cora would form part of the province of Lusitania. The central cora of Perpignan would be the most widely distributed among the Roman provinces. Its territory would be located in important parts in Lusitania, Tarraconensis, Carthaginensis and slightly in Baetica. This is because the Roman administrative distribution did not take into account the division of the current Iberian Peninsula, with the presence of two countries, and the Portuguese territory in its majority, the Roman Lusitania, conditions the distribution of coras made by Perpignan.

10.4 Conclusions Perpiñá’s interest in the distribution of the Spanish population appeared in his first relevant research and remained with him throughout his career. His first model differentiating the centre from the periphery already evidences a steadiness in the tendency not only of maintaining but also of increasing the differences between the

12 The first division made by Republican Rome in 197 BC was Hispania Citerior and Hispania Ulterior, into two parts of the peninsula as it advanced inland. The first emperor Augustus took control of the whole of Hispania and divided it into three regions, Lusitania, Baetica and Tarraconensis in 27 BC. Two centuries later Caracalla in 214 AD created the province of Gallaecia in the northwestern territories of Tarraconensis. The last division dates from 293 AD and is the one we have taken to compare with the Perpignan model. In this division of the Iberian Peninsula, the emperor Diocletian created the province of Carthaginensis, leaving the peninsula with five large continental regions, the result of the presence of an administration like the Roman one for almost five centuries.

10 The Spatial Distribution of the Population in Peninsular Spain: An. . .

Northwest

183

North Northeast

Central

East

South

Map 2 Comparison of the Roman provinces created in 293, whose limits are represented by the black line, with the Roman Perpignan choruses of 1954, identified with different colours. (Source: Own elaboration)

densities of the two areas. This behaviour is slow but steady over the years, which is how changes in the structures occur. In the second step of the analysis, division into coras made up of dasicoras and areocoras, we see that the evolution of these almost 70 years shows a very similar behaviour, and the conditions of higher density in the dasicoras in relation to their neighbouring areocoras and regarding the average density of the country remain. Once again, the changes would be within a structural dynamic in terms of rhythm and period variations. The division of the peninsular territory into coras could well be a model to be taken into account for the management of public administrations, as this model is much more balanced in terms of surface area, population and, therefore, density. Efficiency in the use of public resources could be increased by a reduction in political representation, such as a smaller number of regional governments. In relation to health, it would be possible to achieve shorter average distances to health centres for the population, and there would be greater convergence in matters related to education. It is true that the model of division into coras clashes with the existence of historical regions, but on the other hand it would not have the presence of autonomous communities of a single province which in our opinion shows an excessive division that does not add any advantage to the nation as a whole. In fact, in an attempt to compare Perpiña’s model with other divisions of the peninsular

184

D. del Castillo Soto and T. Baumert

territory, we find ourselves with the division carried out by the Roman Empire almost three centuries after the total control of the Iberian Peninsula, a division that was established without any historical-cultural conditioning, but taking into account above all aspects related to geography, natural resources, seeking efficiency in relation to the administration carried out in the different provinces. Hence, we conclude that the knowledge of the population structure, being one of the fundamental parts of the economic structure, is necessary for correct economic decision-making. Therefore, its lack of study can lead to imbalances that will affect the efficiency and the best development of a country. Román Perpiñá Grau made great and pioneering contributions in this field, detecting the importance and influence of population distribution in the economy. His studies and conclusions have clearly stood the test of time, showing a permanence in the behaviour of the spatial distribution of the Spanish population.

References De Miguel, A. (1992). La sociedad española 1992–1993. Alianza Editorial. Deaton, A. (2015). The Great Escape. Health, wealth and the origins of inequality. Fondo de Cultura Económica de España. Del Río Disdier, J. P. (2015). Guidelines and options for studies of spatial economic structure in Spain: the intellectual crossroads of 1955 (I). In Anuario Jurídico y Económico Escurialense, XLVIII (pp. 311–334). Real Centro Universitario Escorial-María Cristina. Del Río Disdier, J. P. (2016). Guidelines and options for studies of spatial economic structure in Spain: the intellectual crossroads of 1955 (II). In Anuario Jurídico y Económico Escurialense, XLIV (pp. 475–500). Real Centro Universitario Escorial-María Cristina. Del Río Disdier, J. P. (2017). Guidelines and options for studies of spatial economic structure in Spain: the intellectual crossroads of 1955 (and III). In Anuario Jurídico y Económico Escurialense, XLIV (pp. 401–432). Real Centro Universitario Escorial-María Cristina. Perpiñá Grau, R. (1954). Corología. Teoría estructural y estructurante de la población de España (1900–1950). Instituto de economia Sancho de Moncada. Perpiña Grau, R. (1972). De economía hispana, infraestructura, historia. Ariel. Rivas, A., & Rodríguez, J. J. (2009). Graphic memory of Spanish emigration. Ministry of Labour and Immigration. Tamames, R., & Rueda, A. (2022). Estructura económica de España- 2022. JdeJ Editores. Velarde Fuertes, J. (May 21, 2022). Raíces y actualidad de una obra fundamental sobre la economía española. El Debate, https://www.eldebate.com/economia/20220521/raices-actualidad-obrafundamental-sobre-economia-espanola.html.

Chapter 11

Food Waste in Greece: An Empirical Study Electra Pitoska and Grana Vaya

Abstract The increasing number of food markets has been accompanied by the evolutionary effect of food waste and loss and is associated with irrational food management. The problem of food waste has taken on environmental and economic dimensions and has been closely related to the social and moral conditions affecting consumers. For the last few decades, food waste has reached alarming proportions with economic, social, and environmental implications. The precise reasons for wasting food are varied and are largely contingent upon country- and area-specific circumstances. Consumer habits are shaped by social factors, specific consumer attitudes and views about food, and, finally, by consumers’ poor information, knowledge and skills. Reducing food waste by at least 50% by 2030 is among the primary goals set by the United Nations (UN) for Sustainable Development. Greece, among all European Union member countries, is committed to reducing food waste, as waste prevention is a key measure suggested by Circular Economy policies and European Green Agreement requirements. The present empirical study demonstrated that Greek consumers seem to be aware of the problem of food waste and its impact on both the environment and their income. They are also concerned about the amount and cost of wasted food. Keywords Food waste · Consumers · Environment · Greece

E. Pitoska () · G. Vaya Department of Accounting & Finance, University of Western Macedonia, Kozani, Greece e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_11

185

186

E. Pitoska and G. Vaya

11.1 Food Waste and Consumer Behaviour The increasing number of food markets has been accompanied by the evolutionary effect of food waste and loss and is associated with irrational food management. The problem of food waste has taken on environmental and economic dimensions and has been closely related to the social and moral conditions affecting consumers. For the last few decades, food waste has reached alarming proportions with economic, social, and environmental implications. Regarding the impact of food waste on the environment, it is estimated that 8% of greenhouse gas emissions from human activities are produced by food waste. Wasted food in landfills releases methane, which is 25 times as harmful as CO2 (Garnett, 2008; Hall et al., 2009). The problem of food waste is mainly caused by consumers, whose tendency to buy food in larger quantities than necessary has now been a regular habit. Remarkably, 53% of all food waste is household waste, that is, it is produced during consumption (Gustavsson et al., 2011). Vast amounts of food waste are observed in middle- and high-income countries, which implies that food is thrown away despite being still safe for consumption. According to the Working Group of Rational Use of Food at the Federation of Polish Food Banks, there is a clear distinction between food loss and food waste. Food loss is the measure of the reduction in the edible mass of food resulting from wrong management, errors and irregularities during production, distribution, and trade, whereas food waste implies the irrational financial management of food in the hospitality sector and consumer households (Wrzosek et al., 2012). The reasons driving consumers to waste food are varied and are largely contingent upon country- and area-specific circumstances. Consumer habits are shaped by social factors, specific consumer attitudes and views about food, and, finally, by consumers’ poor information, knowledge, and skills. Remarkably, the psychographic profile of consumers has received increasing attention in the relevant literature about waste issues (Russell et al., 2017). As regards food waste in Greece, ranked third among the worst European Union countries in terms of understanding the “best before” label, the extant research has demonstrated that only 22% of the Greek citizens fully understand the specific warning about food (WWF, 2020). In addition, the number of Greek consumers who would throw away food after the “best before” date, irrespective of whether it is safe to consume, is twice the average in the European Union (WWF, 2020). Greece is fourth among the worst EU countries in terms of food waste, which accounts for 196 kg for each Greek citizen compared to 173 kg of annual food waste for each European (WWF, 2020). In addition, Greek households waste 98.9 kg of food surplus annually, and about 1.4 million people, namely, 12.9% of the population, were food insecure in 2015, whereas 5.1% of food production for human consumption is wasted per year, more than double the European average (WWF, 2020).

11 Food Waste in Greece: An Empirical Study

187

11.2 Factors that are Reducing Food Waste and Consumer Attitudes Food waste and loss occur during all stages of the food supply chain, from food storage, transport and processing to food shops and restaurant kitchens, hotels and households (Lundqvist et al., 2008). It is a global phenomenon observed at various stages of the food supply chain. With a view to achieving more efficient food management, the European Union drafted a report in 2016, “Estimates of European food waste levels”, in the EU-28. Data collection and analysis from all 28 EU members demonstrated that food waste accounts for 88 million tonnes (Stenmarck et al., 2016). In recent years, the international literature has focused on food waste, particularly in households. It has also explored the drivers for consumer attitudes towards food waste and emphasised the social, environmental and psychological factors underlying individuals’ behaviour towards food waste (Russell et al., 2017). Due to the complex attitudes affecting the amount and occurrence of household food waste, it is rather hard to make any relevant predictions (Quested et al., 2013). In addition, actions aimed at reducing household food waste are characterised as rather poor and inefficient. Despite the fact that the psychological underlying mechanisms are not primarily explored, there is a significant focus on the methods to identify motivation and/or barriers to prevent household food waste reduction (GrahamRowe et al., 2015). Consumers who positively evaluate the effectiveness of waste management activities are in favour of sustainability-related behaviours (Swami et al., 2011), mainly adopted by people who are convinced that their welfare is threatened. Subjective-personal rules have also been found to play a significant role in shaping waste management attitudes (Barr, 2007). Research on the environmental factors involved in food waste (Dunlap et al., 2000) has shown that people who are more concerned about their bodies than the environment display a less environmentally friendly attitude (Schultz & Zelezny, 1999). On the other hand, other studies highlight that there is a rather weak or moderate correlation between environmental concern and environmentally friendly attitudes (Bamberg, 2003). In the framework of the European project “Fusions Drivers” (Canali et al., 2014), the results revealed the following groups of factors typically affecting consumer behaviour towards food waste: • Social factors, such as household type, stage of family life and relevant lifestyle • Consumers’ attitudes and views about food • Consumers’ poor information, knowledge and skills The European Parliament report “Tackling food waste: The EU’s contribution to a global issue” underlines that social urbanisation trends and changes in diet, as well as consumers’ overall culture, are key determinants of consumer behaviour (EPRS, 2016).

188

E. Pitoska and G. Vaya

An interesting model of the factors affecting consumer behaviour towards food waste has been proposed by Aschemann-Witzel et al. (2015). The model includes two groups of factors, socio-demographic and psychological, which are fundamental to explain consumers’ attitudes towards food waste. Activities aimed at reducing food waste can be also affected by financial factors. The financial crisis may trigger some alternative or new patterns of behaviour called “freeganism” or “dumpster diving”. In addition, the tendency to avoid wasting food can be part of a lifestyle and a consumer’s identity (Aschemann-Witzel et al., 2015). It has also been demonstrated that attitudes towards food waste depend on consumers’ understanding of the information on product packaging, in particular information about the date of minimum durability and expiry date (Newsome et al., 2014). However, the correct use of the terms “expiry date/use by” and “best before” is not sufficiently understood by consumers, who interpret them differently depending on food type (Van Boxstael et al., 2014). Consumers are aware of the principles of rational household food management. Rational storage management in warehouses and markets usually requires a schedule compliant with reducing food waste methods. Educational activities focused on waste-minimising behaviours, which should be carried out at an early age, play a significant role in reducing food waste (Radzyminska et al., 2016). In the extant literature, consumers’ attitudes and behaviours towards food waste are affected by various factors. Consumer knowledge and awareness do not fully reflect the activities of anti-food waste supporters (Radzyminska et al., 2016). Food waste typically occurs during consumption. Households produce the greatest amount of wasted food; thus, food waste is mainly consumer-related. The major concern of all those involved in food waste is, first, to inform consumers about the significance of the specific problem and then provide valuable information on more efficient food management. In addition to consumers’ benefits, social and economic advantages are also perceived among the major goals of actions aimed at raising awareness of food waste. Various methods to reduce food waste have been suggested: shopping lists, checking expiry dates, budget control, refrigerator maintenance, cooking meals with leftovers, freezing food and composting.

11.3 The Empirical Study The research was carried out to investigate food waste and, more specifically, the amount of food waste in Greece, as well as consumer attitudes and behaviour. Due to the special conditions of COVID-19 and contact/communication problems, data collection was based on convenience sampling. The survey employed a structured questionnaire with 15 questions, to be answered via Google Forms from 9 November 2021 to 30 November 2021. A total number of 403 questionnaires were collected. Data analysis demonstrated that only half of the survey participants (50.6%) are aware of the problem of “Food Waste”. It appears that 41.2% of the respondents

11 Food Waste in Greece: An Empirical Study

I feed animals

189

3%

I put it in the freezer

10%

I give it away to a home for the old/children’s institution

2%

I offer it to another person

3%

I recycle it

26%

I use it straight away

17%

I throw it away

39% 0%

5% 10% 15% 20% 25% 30% 35% 40% 45%

Fig. 11.1 Management practices for “use-by” food

are totally ignorant and a smaller number (8.2%) have only got a vague and rather general idea. The main source of information is the Internet (40%), the media (34%) as well as social institutions (13.2%), whereas it was also stated that 11.2% of the relevant information is based on personal experience and 1.7% on other sources. Almost all consumers agree that food waste has got a direct impact on income. In addition, 42% of the respondents answered that they had knowingly and intentionally consumed food after the expiry date, whereas 11% stated that they always check expiry dates and 47% that they may have consumed food after expiry, since they do not check expiry dates. 46% and 13.6% of all survey participants “often” and “always” check expiry dates, respectively. Finally, the subjects stated that they “sometimes” (29.5%) or “rarely” (7.4%) check expiry dates and only 3% they never check dates. Food management practices for “use-by” food demonstrate that the subjects do not adopt sustainability-related attitudes, as the number of answers to questions about whether they are willing to offer food either to another person or vulnerable groups (i.e. a home for the old) is too low, whereas 10% of the respondents answered that they store food in a freezer and 17% consume food straight away. Only 26% of them use recycling methods and 39% tend to throw away food close to their expiry date (Fig. 11.1). The subjects appear to distinguish between “expiry date/use by” and “best before” on food products (71%). Most of the subjects answered that they “sometimes” (37.7%) or “often” (40%) buy food in damaged packaging provided that the content has not been affected.

190

E. Pitoska and G. Vaya

54.10%

60.00%

52.40%

50.00% 38.50% 40.00% 30.00%

19.60%

20.30%

20.00% 10.00%

6.20%

6.60%

0.00%

Fig. 11.2 Most frequently wasted type of food Table 11.1 Reasons for throwing away food Reasons for throwing away food It has expired Food I really don’t need Nasty appearance/condition We haven’t finally eaten leftovers We forgot the food was in or out of the fridge It didn’t taste good (but not rotten) Rotten food Ruined cooked food (burnt, overcooked, too salty, etc.)

Percentage 96.8 4.2 2 3 4.5 3.7 90.1 88.6

As regards the ten food categories surveyed, the participants stated they often waste lunch on meat (66%), followed by pasta/rice (54.10%), fruits/vegetables (52.4%), meat/chicken (38.5%), bread/bakery (20.3%) and dairy products (19.6%), whereas soft drinks and juice are wasted by 6.2% of the consumers, and legumes by 1%. Finally, none of the surveyed consumers stated that they do not waste food (Fig. 11.2). The participants appear to make efforts to adopt management practices for efficient purchase planning and organisation, as the majority organise meal plans for the next few days, use a shopping list, and try to buy only food list items. The reasons for “throwing away” food are shown in the Table 11.1. The analysis demonstrated that consumers make real efforts for proper food management and that no food is wasted. A great majority answered, “I feel sorry when I throw away food”, “I try to throw away as little food as possible,” “I offer

11 Food Waste in Greece: An Empirical Study

191

Table 11.2 Reasons for reducing food waste Reasons To save money Feeling guilty when throwing away food that can be consumed by other people A shift to healthy diet Sustainability aware – I don’t want to affect the environment Concern about the global food problem Better household organisation Table 11.3 Average weekly expenditure on food purchase

A C Up to 50 51–100 101–150 151–200 201+ Total:

Percentage 98 4.7 28 10.2 36.7 53.3 Percentage 41.9 51.6 6.7 0 0 100.0

food instead of throwing it away” and finally, “I store, protect & preserve food properly.” The participants consider that “Food packaging that ends up in the trash is an environmental problem” (56.57%), they pay attention to the expiry dates of the products they buy (65.76%) and state that food organisations provide relevant information (50.62%). They are also concerned about the amount (56.33%) and cost (55.33%) of food they throw away. They are confident that food waste is a global economy-related problem, and a huge social and moral issue, given food insecurity and hunger, as about 800 million people suffer from chronic malnutrition symptoms, affecting almost one in three people on the planet (OHE). Finally, as regards whether they efficiently organise shopping in order to minimise food waste, the subjects answered that they agree, they do as much as possible, they often go to supermarket, and, to better manage household food, they buy only necessary food and food bargains. The analysis demonstrated that participants would be encouraged to reduce the amount of food they throw away, first, for money-saving reasons (98%), followed by better organising households (53.3%), concern about the global food problem (36.7%), which reflects consumers’ guilt when food is wasted, a shift to healthy diet (28%), sustainability awareness (10.2%) and, finally a shift to a healthy lifestyle (4.7%) (Table 11.2). As regards composting, the survey demonstrated that only half of the subjects seem to be aware of the specific issue. The average weekly expenditure on food is listed in the Table 11.3.

192

E. Pitoska and G. Vaya

11.4 Conclusions The problem of food waste has taken on alarming proportions in recent decades with economic, social, and environmental implications. About one-third or 1.3 billion tonnes a year of the world’s food production for human consumption is lost or wasted. The problem of food waste is mainly caused by consumers, whose tendency to buy food in larger quantities than necessary has now been a regular habit. Food waste and loss occur along the entire food supply chain, from food storage, transport, and processing to food shops and restaurant kitchens, hotels and households. The international research has focused on food waste, and, more specifically, household food waste, thus, examining consumer attitudes towards the problem of food waste (consumer food waste behaviour). Reducing food waste by at least half by 2030 is one of the major goals set by the United Nations (UN) for Sustainable Development. Greece, among all European Union member countries, is committed to reducing food waste, as waste prevention is a key measure suggested by Circular Economy policies and European Green Agreement requirements. The research results demonstrated that Greek consumers are aware of the problem of food waste, and the impact both on the environment and their income. The analysis also found that they tend to organise purchases by making shopping lists and are concerned about the amount and cost of wasted food.

References Aschemann-Witzel, J., de Hooge, I., Amani, P., Bech-Larsen, T., & Oostindjer, M. (2015). Consumer-related food waste: Causes and potential for action. Sustainability (Switzerland), 7(6), 6457–6477. https://doi.org/10.3390/su7066457 Bamberg, S. (2003). How does environmental concern influence specific environmentally related behaviors? A new answer to an old question. Journal of Environmental Psychology, 23(1), 21– 32. Barr, S. (2007). Factors influencing environmental attitudes and behaviors: A UK case study of household waste management. Environment and Behavior, 39(4), 435–473. Canali, M., Östergren, K., Amani, P., Aramyan, L., Sijtsema, S., Korhonen, O., et al. (2014). Drivers of current food waste generation, threats of future increase and opportunities for reduction. FUSIONS. Dunlap, R. E., Van Liere, K. D., Mertig, A. G., & Jones, R. E. (2000). New trends in measuring environmental attitudes: Measuring endorsement of the new ecological paradigm: A revised NEP scale. Journal of Social Issues, 56(3), 425–442. EPRS. (2016). Tackling food waste: The EU’s contribution to a global issue. Briefing. Garnett, T. (2008). Cooking up a storm: Food, greenhouse gas emissions and our changing climate. Food Climate Research Network, Center for Environmental Strategy. http://www.fcrn.org.uk/ sites/default/files/CuaS_web.pdf

11 Food Waste in Greece: An Empirical Study

193

Graham-Rowe, E., Jessop, D. C., & Sparks, P. (2015). Predicting household food waste reduction using an extended theory of planned behaviour. Resources, Conservation and Recycling, 101, 194–202. Gustavsson, J., Cederberg, C., Sonesson, U., van Otterdijk, R., & Meybeck, A. (2011). Global food losses and food waste: Extent causes and prevention (pp. 1–37). Food and Agriculture Organization of the United Nations (FAO). http://www.fao.org/3/mb060e/mb060e.pdf Hall, K., Guo, J., Dore, M., & Chow, C. (2009). The progressive increase of food waste in America and its environmental impact. PLoS One, 4, e7940. https://doi.org/10.1371/ journal.pone.0007940 Lundqvist, J., de Fraiture, C., & Molden, D. (2008). Saving water: From field to fork: Curbing losses and wastage in the food chain. Stockholm International Water Institute. Newsome, R., Balestrini, C., Baum, M. D., Corby, J., Fisher, W., Goodburn, K., Labuza, T., Prince, G., Thesmar, H. S., & Yiannas, F. (2014). Applications and perceptions of date labeling of food. Comprehensive Reviews in Food Science and Food Safety, 13(4), 347–769. Quested, T., Marsh, E., Stunell, D., & Parry, A. (2013). Spaghetti soup: The complex world of food waste behaviours. Resources, Conservation and Recycling, 79, 43–51. Radzyminska, M., Jakubowska, D., & Staniewska, K. (2016). Consumer attitude and behaviour towards food waste. Journal of Agribusiness and Rural Development, 1(39), 175–181. Russell, S., Young, C., Hardin, K., & Robinson, C. (2017). Bringing habits and emotions into food waste behaviour. Resources, Conservation and Recycling, 125, 107. Schultz, P. W., & Zelezny, L. (1999). Values as predictors of environmental attitudes: Evidence for consistency across 14 countries. Journal of Environmental Psychology, 19(3), 255–265. Stenmarck, A. S., Jensen, C., Quested, T., Moates, G., Buksti, M., Cseh, B. Z., & Redlingshofer, B. (2016). Estimates of European food waste levels. IVL Swedish Environmental Research Institute. Swami, V., Chamorro-Premuzic, T., Snelgar, R., & Furnham, A. (2011). Personality, individual differences, and demographic antecedents of self-reported household waste management behaviours. Journal of Environmental Psychology, 31(1), 21–26. Van Boxstael, S., Devlieghere, F., Berkvens, D., Vermeulen, A., & Uyttendaele, M. (2014). Understanding and attitude regarding the shelf life labels and dates on pre-packed food products by Belgian consumers. Food Control, 37, 85–92. World Wide Fund for Nature (WWF). (2020). Measures to reduce food losses. http:// contentarchive.wwf.gr/images/pdfs/food-waste-measures.pdf Wrzosek, M., Koło¨zyn-Krajewska, D., & Krajewski, K. (2012). Nieracjonalne wykorzystanie z¨ ywno´sci perspektywa globalna i odpowiedzialno´sci społecznej, Prace Studentów i Młodych Pracowników Nauki. Teoria i praktyka zarzadzania ˛ przedsi˛ebiorstwem. Wybrane zagadnienia, IV, 59–72.

Chapter 12

Success Factors in Public–Private Partnership of High-Speed Railway Infrastructures: Elements for Improvement Mario González-Medrano, Tomás García Martín, and José-María Rotellar-García

Abstract European railway transportation is not as efficient as it could be. Several factors contribute to this absence of efficiency, including a lack of private investment and exclusively public railway administrators and enterprises. Public–private partnerships can yield higher efficiency in railway transport. Within the high-speed railway sector, there are a few implemented infrastructures through public–private partnerships, and in many of them, the result has not been the expected one. Hence, the purpose of this research is to develop a list of recommendations and good practices that allow governments, private investors and railway stakeholders to take better and more efficient decisions on the implementation of new highspeed rail lines. Consequently, this research has analysed through seven case studies, all the high-speed lines designed through public–private partnerships in Europe. The research methodology is based on the exploratory case study and on the identification of critical success factors. This article has made it possible to develop the following list of recommendations and good practices, cross-border cooperation for international sections, substructure and superstructure work in the same contract, a separate contract for signalling and communications systems, availability risk transferred to the private party, proven purchase of rolling stock, savings in travel times, conservation of existing stations in the conventional rail network, responsibility of the rail infrastructure manager the interfaces between the new infrastructure and the existing network, and the inclusion of the activities of maintenance in the public–private partnership contract.

M. González-Medrano () · T. G. Martín Universidad Camilo José Cela, Madrid, Spain e-mail: [email protected] J.-M. Rotellar-García Universidad Francisco de Vitoria, Madrid, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_12

195

196

M. González-Medrano et al.

Keywords Public–private partnership (PPP) · High-speed line (HSL) · High-speed train (HST) · Investment · Finance

12.1 Introduction High-speed rail infrastructures are critical elements that produce the following benefits in society: greater economic development, reduced poverty and inequality, increased job creation, and environmental sustainability. In addition, railway infrastructures generate a high social return and improve the well-being of the population. The governments of each region are responsible for the provision of public services and the infrastructures necessary to operate them. For this reason, investments in railway infrastructure are often part of the social pact between governments and citizens (World Bank, 2017). Exclusively public railway managers and operators, an absence of cost-reducing competition and a lack of private investment can be classified as the historical factors that have not allowed railway transport to be as efficient as it might be. The European Union initiated railway transport liberalization in 2006 with international freight transport, in 2007 with domestic freight transport, continued in 2010 with international passenger transport and finished on 14 December 2020 with domestic passenger services. These measures represent the initial actions taken toward an efficient railway system in Europe. For the development of high-speed railway infrastructures, it is necessary to have both the economic investment and the knowledge to be able to carry them out. When a public administration does not have either of these two resources, investment, or knowledge, or wishes to execute the infrastructure in a more efficient way than a traditional contracting, public administrations have the option of carrying out these infrastructure developments through a public–private partnership (PPP). In this way, it is a possible and achievable objective to achieve greater efficiency in high-speed rail transport, which can be obtained through the construction and operation of the infrastructure through PPPs, by achieving a reduction in public funding. Thanks to the mobilization of private investment, it allows maintaining the sustainability of public finances, obtaining a better capacity for project management and promoting optimization and innovation provided by the private sector. At all times, but especially at these times when public sector spending is very high, the collection is affected by economic cycles, which hinder the public financial sustainability of a long-term project and the high existing indebtedness, which is an important limit to the financing these projects can seek on the markets. As such, PPP is especially relevant, given these commented limitations of fiscal policy, and the fact that rail transportation is a sector with such intensive capital needs in order to build infrastructure. This will contribute to efficiency gains in the provision of services, lower costs and free the public sector from the pressure of investment spending in a few years, being able to structure it more appropriately over time thanks to PPP. In this way, it will contribute to the fulfilment of the stability

12 Success Factors in Public–Private Partnership of High-Speed Railway. . .

197

objectives of the country in which the action is carried out, it will ensure the necessary investment to carry out the infrastructures, and it will increase efficiency in the economy as a whole since these works will not be seen delayed due to lack of funds derived from a compromised public budget restriction (European PPP Expertise Centre [EPEC], 2015; Rotellar-García, 2019). The PPP model has been successfully applied in different public services, including road transport. However, very few high-speed lines have been built through PPPs and in many of them, the result has been contrary to expectations. Therefore, based on these experiences, there is a need to study the reasons why PPPs are not successful in the field of high-speed rail infrastructures, and how failures, mainly attributed to the decisions made by public administrations and private investors, can be avoided. Therefore, the objective of this research is to develop a series of elements for improvement that serve as a tool for governments, private investors and interest groups in decision-making, before the construction of new high-speed rail lines, and thus be able to obtain all the socio-economic benefits provided by investments in high-speed rail infrastructure implemented through the PPP model.

12.2 Research Methodology The applied research methodology has been the exploratory and collective case study theory defined by the most influential authors in this field, Yin and Stake, as well as the identification of critical success factors (CSF) defined by Rockart (Stake, 1995; Yin, 2009; Rockart, 1982). The study has consisted of the following stages: (a) the collection of a wide range of data from the railway infrastructure concessionaires, railway infrastructure managers, railway undertakings, public authorities and railway-specific publications; (b) the classification of the data into six areas: project, infrastructure, transport service, contract, corporate structure and investment; (c) comparison of the case studies; (d) identification and analysis of the CSFs and (e) development of the elements for improvement and good practices for governments, private investors and railway stakeholders.

12.3 Case Studies The research analyses, through seven case studies, all the high-speed lines designed through PPPs in Europe. The first of these is the Channel Tunnel Rail Link (CTRL), which was later renamed High Speed 1 (HS1), is located in England and links London to the British side of the Channel Tunnel and connects through an underwater rail tunnel to France.

198

M. González-Medrano et al.

The second case is the Hogesnelheidslijn Zuid (HSL-Zuid) high-speed line, which in English means south, which is located in the Netherlands and connects Amsterdam with the Belgian border through the town of Breda, to connect with the rest of the European high-speed network. The following three case studies are from French lines. France is the country in Europe that has most implemented the PPP model in high-speed rail infrastructures. The third case is the French line Bretagne-Pays de la Loire (BPL) that connects the towns of Le Mans and Rennes. The fourth case is the French line Sud-Europe Atlantique (SEA) that connects the towns of Tours and Bordeaux. The fifth case is the Contournement Nîmes-Montpellier (CNM) a by-pass between the French cities of Nîmes and Montpellier. The sixth case is the Figueras-Perpignan high-speed international section. This section has made it possible to connect Spain with France at high speed using Union Internationale des Chemins de fer (UIC) gauge. The last case study is the Rede Ferroviária de Alta Velocidade (RAVE), the Portuguese high-speed rail network. Portugal designed a network through six PPP contracts whose purpose was to connect with the Spanish high-speed network and with the rest of Europe through the UIC gauge. Five PPP contracts were defined for the development of the substructure and superstructure. In addition, a PPP contract was designed whose scope was the design, supply, installation, financing and maintenance of the European Rail Traffic Management System (ERTMS) signalling system and the Global System for Mobile Communications-Railway (GSM-R) communications system, for the entire network. Within this project, only the Lisbon-Poceirão section was awarded, which was later terminated and the rest of the sections were not awarded, since the project was abandoned due to the international financial crisis of 2008.

12.4 Analysis and Discussion of the CSFs Once the seven case studies have been compared, the following list of elements for improvement has been concluded.

12.4.1 Cross-Border Cooperation for International Sections In the case of the international section Figueras-Perpignan and the Portuguese high-speed network, both projects were defined through bilateral Spanish-French and Spanish-Portuguese summits. At these summits, important points were defined such as the layout and the track gauge to be implemented, which was the UIC. However, we have other opposing cases, such as that of the Dutch Zuid line, for which Belgium requested the payment of financial compensation to accept the interconnection route between the two countries (Boletín Oficial del Estado, 1998).

12 Success Factors in Public–Private Partnership of High-Speed Railway. . .

199

Therefore, with these experiences, a good understanding between neighbouring countries is essential for the development of common infrastructures and it is recommended that the cross-border section be the responsibility of both countries. For this alliance to be solid, it is necessary to ensure the participation of stakeholders, guarantee a coherent objective for all participants and ensure that the results of the cooperation bring similar benefits to both sides of the border (Galko & Volodin, 2016).

12.4.2 Substructure and Superstructure Work in the Same Contract The separation of substructure and superstructure works means that it is almost impossible not to make mistakes in engineering designs since a modification in the substructure also means modifying the superstructure. On the Dutch Zuid line, these works were contracted separately, which involved engineering rework due to the lack of coordination between the awardees of the substructure and superstructure (Von der Heidt et al., 2009). Firstly, it is necessary to indicate that the substructure supports the superstructure and transmits the loads to the foundation. Secondly, the superstructure is the area above ground level that receives the loads from the trains which are then transferred to the substructure. So, it is clear that there is an interaction between the substructure and superstructure. Therefore, it is recommended that the scope of design and construction works for the substructure and superstructure always be carried out by the same contractor (Giannakos, 2010).

12.4.3 A Separate Contract for Signalling and Communications Systems ERTMS signalling and GSM-R communication systems represent the greatest technological risk in railway infrastructures, due to the constant evolution of technical specifications and their obsolescence. Since the first technical specification of the ERTMS system in 2000, 12 new versions of this document have been published. With regard to the GSM-R communication system, it will be replaced by a new one called Future Railway Mobile Communication System (FRMCS). In addition, the separate tendering of this work means that there may be greater competition in the substructure and superstructure tenders, as the number of ERTMS and GSM-R technologists is very limited and would mean that only a very small number of partnerships could be created. Therefore, it is advisable to treat signalling and communication systems independently due to their importance and

200

M. González-Medrano et al.

continuous updating, as well as to facilitate competition and reduce costs (European Commission, 2020).

12.4.4 Availability Risk Transferred to the Private Party Of the different PPP contracts in some of them, the winning company assumed the risk of availability, as in the case of the HSL-Zuid line, the BPL line and the CNM by-pass. In other cases, the successful bidder assumed the traffic risk, such as the HS1 line or the Figueras-Perpignan international section. And for the Portuguese high-speed rail network, a mixed formula was chosen based on availability and traffic. After the commissioning of these infrastructures and the start of transport services, lower than expected economic income has been observed, leading to the renegotiation of the contract and subsequently, becoming publicly administered, only in those infrastructures in which the traffic risk was assumed by the winning company. This is the case of the HS1 lines and the international section FiguerasPerpignan. Therefore, it is much safer, not only for the private investor but also for the public administrator, to carry out PPP contracts in which the risk of availability is transferred to the concessionaire (Lawther & Martin, 2014).

12.4.5 Proven Purchase of Rolling Stock With the commissioning of the new infrastructures, new high-speed rolling stock was also put into service by rail operators in the cases of the French SEA and Dutch HSL-Zuid lines. Regarding the rolling material to operate, HSL-Zuid launched a joint purchase with the Netherlands, to acquire new high-speed trains to operate in the Netherlands and carry out international routes with Belgium. For the award of this contract, offers were presented from the railway constructors Siemens, Bombardier and Alstom, whose offers were based on modifications to vehicle models already in operation; however, the technologist Ansaldo Breda was the only manufacturer that fulfilled all the technical requirements based on a train model that still had to be developed. In 2004, Ansaldo Breda won the contract. According to planning, the trains were to be available in 2006, in order to carry out tests and start operating in 2007. The infrastructure was completed and received in 2007, however, until March 2012, the new vehicles were not available. In December 2012, international services connecting Amsterdam with Brussels began. In January 2013, the vehicles were discontinued for commercial service due to failures that occurred during operation. For this reason, it is very important for a railway operator to buy already tested railway vehicles or make modifications on models already in service because their reliability is known, and also the delivery plan will be more realistic and more achievable (Geluk, 2007).

12 Success Factors in Public–Private Partnership of High-Speed Railway. . .

201

12.4.6 Savings in Travel Times In the case studies, the reductions in travel times compared to the conventional railway ranged from 8 minutes for the Paris-Nantes connection on the BPL line, to 7 hours and 21 minutes for the Lisbon-Madrid connection within the project of the Portuguese High-Speed Network. As we can see, the time savings are quite significant, so when implementing these new infrastructures, it is very important to improve travel times compared to conventional rail or other ways of transport (Rede Ferroviária de Alta Velocidade [RAVE], 2006; SNCF Réseau, 2016).

12.4.7 Conservations of Existing Stations in the Conventional Rail Network In reference to the commercial stops on the lines under study, there is a tendency to continue with services at existing stations on the conventional rail network. For the BPL, HSL-Zuid and SEA lines, the existing stations remain on the new route, and only adaptation works are carried out to the new infrastructure. In the case of the CTRL infrastructure, a mixed model is adopted, in which the St. Pancras in London, the platforms of Ashford station are adapted and those of Stratford and Ebbsfleet are built. In the case of the Portuguese High-Speed Network, a mixed model is also adopted, with the construction of the Évora station in the Poceirão-Caia section, the Leiria station in the Lisbon-Pombal section and the Averio station in the PrombalOporto section. In the case of the international section Figueras-Perpignan, due to the characteristics of the infrastructure, there are no stations. For only the CNM by-pass, all the stations, in this case, those of Montpellier Sud de France and NîmesManduel-Redessan, are newly built. Therefore, the tendency to maintain existing stations in the designs of new high-speed lines has several advantages, such as favouring inter-modality in the city centre, the creation of exchanges with other ways of transport, and quick access to final destinations, and the reduction in the necessary investment (National Audit Office [NAO], 2012).

12.4.8 Responsibility of the Rail Infrastructure Manager for the Interfaces Between the New Infrastructure and the Existing Network In the case of the French lines, the integration between the new lines and the existing network was carried out through a contract independent from that of PPP and for which the railway infrastructure manager, in this case, SNCF Réseau, was responsible. In the case of the Dutch HSL-Zuid line, a traditional contract independent from that of the PPP of the superstructure was made to connect the

202

M. González-Medrano et al.

substructure of the new high-speed line with the conventional rail network (Priemus, 2011). In the connection between a new high-speed rail line with the existing rail network or another high-speed line, the most significant risks lie in the signalling system. To connect lines equipped with the ERTMS Level 1 or 2 signalling systems of different Technologists, it is necessary to develop an interface for the interlocking or the RBC. This need leads to a separate contract for these interfaces to be necessary and for them to be managed by the infrastructure manager for lines within the same country or by the States for cross-border interfaces.

12.4.9 Inclusion of the Activities of Maintenance in the PPP Contract The PPP contracts include infrastructure maintenance activities in their scope, which translates into a double advantage for the contracting administration. The successful bidders will carry out a better design and choice of materials and components to reduce corrective maintenance needs and will also finalize and comply with the planning since they mainly receive the most significant payments during the maintenance phase (Engel et al., 2010).

12.5 Conclusion This research was analysed through seven case studies, all the high-speed lines designed through PPPs in Europe. This study has required the application of an exploratory approach and the identification of CSFs. The purpose of this research was to develop a list of elements for improvements that will allow the governments, private investors and stakeholders to be able to achieve all the socio-economic benefits that PPPs bring in the field of high-speed rail infrastructures. This article has made it possible to develop the following list of recommendations and good practices, cross-border cooperation for international sections, substructure and superstructure work in the same contract, a separate contract for signalling and communications systems, availability risk transferred to the private party, proven purchase of rolling stock, savings in travel times, conservation of existing stations in the conventional rail network, the responsibility of the rail infrastructure manager the interfaces between the new infrastructure and the existing network, and the inclusion of the activities of maintenance in the PPP contract.

12 Success Factors in Public–Private Partnership of High-Speed Railway. . .

203

References Boletín Oficial del Estado. (1998). Acuerdo entre el Gobierno del Reino de España y el Gobierno de la República Francesa para la construcción y explotación de la sección internacional de una línea ferroviaria de alta velocidad entre España y Francia (vertiente mediterránea). Boletín Oficial del Estado, 25, 1–3. https://www.boe.es/boe/dias/1998/01/29/pdfs/A03055-03057.pdf Engel, E. M., Fischer, R. D., & Galetovic, A. (2010). The economics of infrastructure finance: Public-private partnerships versus public provision. EIB Papers, 15(1), 40–69. https:// www.econstor.eu/bitstream/10419/45373/1/657028975.pdf European Commission. (2020). ERTMS: First work plan of the European coordinator.https:// ec.europa.eu/transport/sites/transport/files/work_plan_ertms_2020.pdf European PPP Expertise Centre (EPEC). (2015). PPP motivations and challenges for the public sector: Why (not) and how. European Investment Bank. https://www.eib.org/attachments/epec/ epec_ppp_motivations_and_challenges_en.pdf?f=search&media=search Galko, S. V., & Volodin, D. S. (2016). Outcomes of cross-border cooperation: Infrastructure development aspect. £ÍÕÖÃÎßÐi ÒÓÑÄÎÈÏË ÈÍÑÐÑÏiÍË, 2(176), 32–40. http://irbis-nbuv.gov.ua/cgi-bin/irbis_nbuv/cgiirbis_64.exe?C21COM=2&I21DBN=UJRN&P2 1DBN=UJRN&IMAGE_FILE_DOWNLOAD=1&Image_file_name=PDF/ape_2016_2_6.pdf Geluk, J. (2007, February). HSL-Zuid: Ready, steady, go! Global Railway Review. https:// www.globalrailwayreview.com/article/1029/hsl-zuid-ready-steady-go/ Giannakos, K. (2010, May). Interaction between superstructure and substructure in railways. In Proceeding of the Fifth International Conference on International Conferences on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, San Diego, California. https://scholarsmine.mst.edu/cgi/viewcontent.cgi?article=2855&context=icrageesd Lawther, W. C., & Martin, L. (2014). Availability payments and key performance indicators: Challenges for effective implementation of performance management systems in transportation public-private partnerships. Public Works Management & Policy, 19(3), 219–234. National Audit Office (NAO). (2012). The completion and sale of High Speed 1. https:// www.nao.org.uk/wp-content/uploads/2012/03/10121834.pdf Priemus, H. (2011). Contracting public transport infrastructure: Recent experience with the Dutch high speed line and the Amsterdam north-south metro line. In 11th International Thredbo Conference on Competition and Ownership in Land Passenger Transport, Delft, The Netherlands. Rede Ferroviária de Alta Velocidade (RAVE). (2006). Relatório e contas 2006. https:// www.infraestruturasdeportugal.pt/sites/default/files/rc_2006.pdf Rockart, J. F. (1982). The changing role of the information systems executive: A critical success factors perspective. Sloan Management Review, 24(1), 3–13. Rotellar-García, J. M. (2019). El milagro económico de Madrid. Think-Tank Civismo. https:// civismo.org/wp-content/uploads/2019/02/El-milagro-economico-de-Madrid.pdf SNCF Réseau. (2016). Ligne á grande vitesse Bretagne-Pays de la Loire. Comité de suivi départemental. Le Mans, le 8 novembre 2016. http://www.sarthe.gouv.fr/IMG/pdf/ comite_suivi_foncier_-_sarthe_-_8_nov_2016_v1-1.pdf Stake, R. E. (1995). The art of case study research. Sage. Von der Heidt, T., Gillett, P., Charles, M. B., & Ryan, N. (2009). Contractual arrangements and their implications for the provision of an Australian HSR system. Simposio llevado a cabo en next generation infrastructures conference, Chennai, India. World Bank. (2017). Public-private partnerships: Reference guide version 3. https:// library.pppknowledgelab.org/documents/4699/download Yin, R. K. (2009). Case study research: Design and methods. Sage.

Chapter 13

Holistic Evaluation of Technology Transfer Extension Programmes Evropi-Sofia Dalampira, Ioannis Tsoukalidis, Dimitra Lazaridou, Smaragda Nikouli, Anastasios Livadiotis, and Anastasios Michailidis

Abstract Technology transfer is one of the core elements in agricultural economics and agricultural extension offers educational programmes for technology transfer. Although, the booming of agricultural technology and innovation is not followed by the generation of methodological tools able to diffuse innovation in farmers and other stakeholders and lead agribusinesses to economic stability. Evaluating agricultural extension programmes has multidimensional challenges and limitations. In this research, we propose a methodological and conceptual framework assembled from the strengths of various evaluation methodologies. This methodology can be useful to policymakers managers or researchers, in order to construct, implement and evaluate an FFS agricultural programme. The hybrid FFS strategy describes how agricultural education approaches of the past can create educational environments of the future and lead to learning accelerators in the agricultural sector. Keywords CIPP model · Agricultural extension · Evaluation

13.1 Introduction Technology transfer is the diffusion of agricultural innovations from centres of discovery to end users and is the basic element of agricultural extension (Brown et al., 2018). Technology transfer attempts to train and influence farmers’ practices

E.-S. Dalampira () · A. Michailidis Aristotle University of Thessaloniki, Agricultural Economics, Thessaloniki, Greece I. Tsoukalidis DOMI KOINEP, Kavala, Greece D. Lazaridou Aristotle University of Thessaloniki, Thessaloniki, Greece S. Nikouli · A. Livadiotis Development Association of Halkidiki S.A., Polygyros, Greece © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_13

205

206

E.-S. Dalampira et al.

via the introduction of new technologies (Cook et al., 2021). Nowadays, agricultural extension research is booming and is using various technology transfer methodologies (Cook et al., 2021). But when experts transfer technology to farmers, they often neglect social, ecological or other issues (Charatsari & Lioutas, 2020; Cook et al., 2021). Methodological challenges of technology transfer are creating gaps between science and practice. Evaluation of agricultural extension learning (Liossi, 2019), plays a critical role: it highlights what needs to change for a better quality of learning (Charatsari & Lioutas, 2020). Experts have the experience to give feedback for changes in extension programmes but are often unaware of real challenges of farmers (van den Berg et al., 2021). On the other hand, co-creating and evaluating extension systems with farmers can strengthen their effectiveness (Lioutas et al., 2019). Hence, there is a need for holistic evaluation tools able to close the gap of perspectives between experts and farmers (Salehi et al., 2021) or other stakeholders (Dalampira & Nastis, 2020b) and simplify the assessment (Dalampira et al., 2018, 2019; Dalampira & Nastis, 2020a, b). All in all, an innovative approach is needed to promote the harmonization of complementary attributes of different FFS forms (Osumba et al., 2021). The aim of the present paper is to support research into evaluation of technology transfer educational programmes. More specifically we analyse and critically present the current evaluation and develop a new conceptual and methodological strategy which attempts to build on strengths and surpass limitations. First, a literature review of evaluation in technology transfer is presented. Afterwards, a meta-analysis describes the limitations of these methods and explores the possibility of a more comprehensive model. Then, a holistic evaluation model is proposed, as a more dynamic and systematic approach of an agricultural extension learning programme. This is an attempt to fuse different evaluation methods in order to cover technology transfer methodological gaps in future agricultural extension programmes.

13.2 Literature Review: Evaluation for Technology Transfer Farmers Field School (FFS) is an agricultural approach developed in the late 1980s by the Food and Agricultural Organization of United Nations (FAO). Field School is a group-based extension concept based on the principles of adult learning (Osumba et al., 2021). In all its forms, FFS methodology is usually based on technology transfer and co-generation, infused by agricultural extension. In this section, different forms of FFS are presented. These forms are according to differentiation found in the literature but also recognized by the FAO, the organization behind the creation of the FFS approach. In this way, we aim to reveal the main characteristics of each form, as well as its strengths and weakness. Therefore, this section is structured according to this observation of different forms of FFS and their evaluation.

13 Holistic Evaluation of Technology Transfer Extension Programmes

207

13.2.1 Bennett’s Hierarchy: Training Needs Evaluation Bennett’s hierarchy of needs is a simple framework which can be used in any project. It has seven steps which guide the evaluation to the implementation of the analysis (Radhakrishna & Bowen, 2010). It basically creates a connection between programme development and evaluation by highlighting the criteria of programme performance (Diaz, 2018). Bennett’s hierarchy of seven steps provides a framework for developing a methodology to evaluate the performance, design, and future approaches of extension programmes (Alyaarbi et al., 2019). Focusing on the fifth step, knowledge, attitude, skills, and aspirations (KASA) of farmers and other stakeholders, training needs can be identified for the design of agricultural extension programmes.

13.2.2 CIPP Impact Evaluation One of the oldest and most thoroughly tested approaches to evaluation is the CIPP model developed by theorist Dan Stufflebeam and his colleagues over the years. CIPP stands for context, input, process, and product evaluation. It has been used for a large number of evaluation studies of various disciplines (Kellaghan & Stufflebeam, 2003). Although it was originally formulated for educational programmes, it was developed in consumer-oriented evaluation, and project development (Meehan et al., 2013; Salehi et al., 2021). The CIPP methodology is valuable for executing formative and summative evaluations (Toosi et al., 2021). The CIPP model is particularly useful for securing accountability data for large-scale nationwide educational programmes supplied by public funds. Requirements for its use are needs assessment, evidence-based programmes, promoting organizational culture, adequacy of resources, identification of stakeholders and comprehensive cooperation and existence of an appropriate evaluation system (Vali et al., 2021).

13.3 Meta-Analysis: Covering Gaps of Current Evaluation Strategies in Technology Transfer FFS is based on experimental adult learning, hence one of its core principles is a targeted, fully organized evaluation (Charatsari & Lioutas, 2020). For FFS evaluation it is important to consider achieving the practicality and usefulness of the evaluation methodology (Brown et al., 2018). In this way, the next cycle of learning will be revised and corrected for better results. Before establishing FFS content, an assessment should be performed to identify educational needs (FAO, 2016). Formal surveys alone cannot provide in-depth analysis to understand farmers’ empowerment (David, 2007). Qualitative studies using diffusion and

208

E.-S. Dalampira et al.

social network mapping, focus groups and participant observation are needed to complement formal surveys (Bogner et al., 2009). The key indicators (group profile, plants, good management, group experimentation, etc.) of successful FFS will give the appropriate information on whether the learning cycle is working between participant farmers and facilitator (FAO, 2016). Widely used evaluation methods and models, such as Bennett’s Hierarchy and CIPP model can be applied in different FFS forms, depending on the aim and objectives of the learning programme (Salehi et al., 2021; Sasidhar et al., 2011). Bennett’s hierarchy has been used for DFFS and revealed a major impact on developing awareness (Sasidhar et al., 2011). A wide range of evaluation concepts and models exists, but the framework of the original CIPP model (Owen, 2004), is still successfully used in FFS research (Salehi et al., 2021). According to FAO, combination of Bennett and CIPP evaluation models in FFS programmes focuses on rural empowerment, innovation, and agricultural development (Amanah et al., 2021). CIPP has been proposed from key theorists as a basis to build a metamodel of evaluation (Owen, 2004). The CIPP evaluation model approaches the components of context, input, process, and product (CIPP). The context evaluation stage sets the programme goals and helps the design. At the input evaluation stage, resources and inputs such as personnel, facilities, and budgets are identified. The purpose of the process evaluation stage was to identify potential problems and threats that cause failures in the programme and required changes in the project. At the product evaluation stage, it is determined the extent to which the goals accomplished (Salehi et al., 2021).

13.4 Towards a New Framework: Holistic Evaluation for Counting the Degree of the Achievement Evaluation is the final but the closing step of the Hybrid FFS strategy cycle. Evaluation strategically analyses the resources that should be allocated to accomplish the mission, of the value network. Without evaluation, Hybrid FFS cannot be finished. We created a metamodel of evaluation, based on the CIPP model (Owen, 2004) and combined two generic forms of evaluation with detailed information (Table 13.1). CIPP FFS evaluation model is based on a holistic view to assessing FFS and closes the gap between experts and farmers perspectives (Charatsari et al., 2018; Salehi et al., 2021). We propose a multiperspective evaluation, with the inclusion of programme development (context stage) and impact assessment (input-processproduct-output-programme reengineering stages), in order to determine before and after implementation, the worth of the learning programme (Owen, 2004). Level of the programme evaluation can be from regional, national to multi-national, depending on the data collection.

Approach (Owen’s evaluation) Programme development/proactive (Bennett’s evaluation) Impact assessment/learning accountability (CIPP evaluation)

Stage Focus Planning Goals/objectives (programme synthesis) Summative/formative Evaluation Outcomes/delivery (settled/finished programme)

Audience benefited Type Internal Summative

Participants & experts External

Evaluator Possible participants

Table 13.1 Holistic FFS evaluation metamodel (inspired by Owen (2004), Salehi et al. (2021))

Goalbased/needsbased

Method Needs-based

After implementation

Time Before implementation

13 Holistic Evaluation of Technology Transfer Extension Programmes 209

210

E.-S. Dalampira et al.

Regarding the type of evaluation, CIPP FFS uses a two-step mixed (formative and summative) evaluation of the programme. In programme development, the summative part reports on the programme. The summative part is needs-based and justifies the learning needs of participants (Charatsari & Lioutas, 2020). Afterwards, we used the results from CIPP FFS context, to Input, Process and Output (Dalampira et al., 2020; Salehi et al., 2021). The formative step is goal-based (goals of the modules) and reports to the programme for its improvement (Salehi et al., 2021). We combined summative and formative (mixt type) evaluation. Summative offered the knowledge of what extent the goals of the programme have been reached. Formative offered the knowledge of programme’s imapct to stakeholders and society (Charatsari et al., 2020; Salehi et al., 2021). The first step, in the context of the evaluation, is programme development. Programme development evaluation inquiries values and expectations of stakeholders (Charatsari & Lioutas, 2020). It will be a formative evaluation for the improvement of the programme at the stage of planning, during the synthesis of the programme. Programme development evaluation reports to the programme. Hence, this evaluation benefits the internal audience (participants, experts and staff of the programme). Evaluators implementing programme development should be external and can vary (experts, farmers and other stakeholders) in order to check the effectiveness of basic elements of FFS, such as experimentation and learning (Salehi et al., 2021). In this early stage of planning, it is important to include participants (i.e. farmers) and external experts, in order to investigate the probable gap between their views (Salehi et al., 2021), values and expectations (Charatsari & Lioutas, 2020). Programme development evaluation focus on goals and objectives (context), derived from training needs assessment (Charatsari & Lioutas, 2020; Dalampira et al., 2018) and other rapid participatory rural appraisal methods, which change their value system and lead to empowerment and transformation. As a tool for collecting data for needs assessment, a structured online or face-to-face questionnaire (Charatsari & Lioutas, 2020), provide appropriate quantitative data for statistical analysis. In other words, participants and experts should plan the training needs, goals, objectives at this stage, and succeed in a bottom-up and top-down evaluation of Hybrid FFS. The second step includes the following stages: process, product, output, and programme reengineering, which are the impact assessment of the programme (Dalampira & Nastis, 2019). Likewise, evaluators should be external (participants and external experts), for the reasons described in programme development evaluation. As a tool, we propose a structured questionnaire using quantitative data with statistical analysis. Firstly, external experts review and modify the validity of the questionnaire and afterwards, the questionnaire is used by the participants. The focus is on the outcomes and the delivery of the program, but the method is mixed (goal-based/needs-based) and incorporated with rapid participatory rural appraisal methods. Participants and experts (internal audience) implement impact assessment (Dalampira & Nastis, 2020b), but this evaluation benefits the external audience (i.e. decision makers). Overall, the proposed CIPP FFS evaluation is depicted in Fig. 13.1.

13 Holistic Evaluation of Technology Transfer Extension Programmes

211

Fig. 13.1 Incorporated Hybrid FFS strategy and evaluation metamodel (inspired by Kellaghan and Stufflebeam (2003), Salehi et al. (2021))

13.5 Conclusion In a rapidly changing rural environment, there is an urgent need for modern technology transfer methodologies. Even though different forms of FFS are ideal approaches for co-generation and technology transfer, there are methodological and practical gaps. There is a demand for a methodological tool able to work at an international level, which will offer a holistic approach for closing gaps between science, technology and practice. We propose an original conceptual and methodological. Finally, this strategy should count the degree of achievement by a holistic evaluation. Holistic evaluation strategy is a proposed methodological tool trying to surpass limitations and discovering weaknesses of different FFS forms. This newly induced framework could work as the first step of organizing technology transfer programmes from the beginning or investigate whether ongoing projects are right on track. This work may provide a new perspective for future planning as well as the evaluation of extension/education projects.

References Alyaarbi, Y., Camkin, J., Neto, S., & Wegener, P. (2019). Knowledge, attitudes, skills, and aspirations of farmers in Abu Dhabi and Western Australia on groundwater management: A comparison study. World Water Policy, 5(2), 161–178. https://doi.org/10.1002/WWP2.12012 Amanah, S., Seprehatin, S., Iskandar, E., Eugenia, L., & Chaidirsyah, R. M. (2021). Investing in farmers through public-private-producer partnerships – Rural Empowerment and Agricultural Development Scaling-up Initiative in Indonesia. FAO Investment Centre Country Highlights, No. 7. FAO and IFPRI. https://doi.org/10.4060/cb7126en

212

E.-S. Dalampira et al.

Bogner, A., Littig, B., & Menz, W. (2009). Introduction: Expert interviews—An introduction to a new methodological debate. In A. Bogner, B. Littig, & W. Menz (Eds.), Interviewing experts (pp. 1–13). Palgrave Macmillan. https://doi.org/10.1057/9780230244276_1 Brown, B., Llewellyn, R., & Nuberg, I. (2018). Why do information gaps persist in African smallholder agriculture? Perspectives from farmers lacking exposure to conservation agriculture. The Journal of Agricultural Education and Extension, 24(2), 191–208. Routledge. https://doi.org/ 10.1080/1389224X.2018.1429283 Charatsari, C., & Lioutas, E. D. (2020). Evaluating agricultural extension and education projects: The VELVET approach. Development in Practice, 30(4), 548–557. https://doi.org/10.1080/ 09614524.2020.1727856 Charatsari, C., Koutsouris, A., Lioutas, E. D., et al. (2018). Promoting lifelong learning and satisfying farmers’ social and psychological needs through farmer field schools: Views from rural Greece. Journal of Agricultural and Food Information, 19(1), 66–74. https://doi.org/ 10.1080/10496505.2017.1325742 Charatsari, C., Lioutas, E. D., & Koutsouris, A. (2020). Farmer field schools and the co-creation of knowledge and innovation: The mediating role of social capital. Agriculture and Human Values, 37(4), 1139–1154. https://doi.org/10.1007/s10460-020-10115-8 Cook, B. R., Satizábal, P., & Curnow, J. (2021). Humanising agricultural extension: A review. World Development, 140, 105337. https://doi.org/10.1016/j.worlddev.2020.105337 Dalampira, E., & Nastis, S. A. (2019). Mapping sustainable development goals: A network analysis framework. Sustainable Development, 28, 46–55. https://doi.org/10.1002/sd.1964 Dalampira, E.-S., & Nastis, S. A. (2020a). Back to the future: Simplifying sustainable development goals based on three pillars of sustainability. International Journal of Sustainable Agricultural Management and Informatics, 6(3), 226–240. https://doi.org/10.1504/IJSAMI.2020.112089 Dalampira, E.-S., & Nastis, S. A. (2020b). Mapping sustainable development goals: A network analysis framework. Sustainable Development, 28(1), 46. https://doi.org/10.1002/sd.1964 Dalampira, E.-S., Papadaki-Klavdianou, A., Michailidis, A., et al. (2018). Multifunctional farms in Greece: A new business model need assessment tool for their activities. Agricultural Economics Review, 19(1), 14–25. Available at: https://www.scopus.com/inward/record.uri?eid=2-s2.085122610494&partnerID=40&md5=26f0496812613ab7408224025c833574 Dalampira, E. S., Papadaki-Klavdianou, A., Nastis, S., et al. (2019). Food for thought: An assessment tool for environmental food identities. International Journal of Sustainable Development & World Ecology, 27, 80–95. https://doi.org/10.1080/13504509.2019.1695231 Dalampira, E. S., Papadaki-Klavdianou, A., Nastis, S., et al. (2020). Food for thought: An assessment tool for environmental food identities. International Journal of Sustainable Development and World Ecology, 27(1), 80. https://doi.org/10.1080/13504509.2019.1695231 David, S. (2007). Learning to think for ourselves: Knowledge improvement and social benefits among farmer field school participants in Cameroon. Journal of International Agricultural and Extension Education, 14, 35–49. Diaz, J., Warner, L., & Webb, S. (2018). Outcome framework for school garden program development and evaluation: A delphi approach. Journal of Agricultural Education, 59, 143– 166. https://doi.org/10.5032/jae.2018.02143 FAO (2016). Farmer field school guidance document: Planning for quality programmes. FAO. ISBN 978-92-5-109126-5. Kellaghan, T., & Stufflebeam, D. (2003). International handbook of educational evaluation. Springer. https://doi.org/10.1007/978-94-010-0309-4 Liossi, F. (2019). Validation of non-formal and informal learning: From European policies to ´ της μη-τυπικης ´ και ατυπης ´ μαθησης: ´ Aπ´o τις national implementation | Eπικυρωση ευρωπα¨ικς ´ πoλιτικ´ς στις εθνικ´ ς εϕαρμoγ´ ς. Academia (Greece), 2019(14), 93–118. ˇ Lioutas, E. D., Charatsari, C., Cerniˇ c Isteniˇc, M., et al. (2019). The challenges of setting up the evaluation of extension systems by using a systems approach: The case of Greece, Italy and Slovenia. The Journal of Agricultural Education and Extension, 25, 139–160. Routledge. https://doi.org/10.1080/1389224X.2019.1583818

13 Holistic Evaluation of Technology Transfer Extension Programmes

213

Meehan, T., King, R., & Kavanagh, D. J. (2013). Programme evaluation and benchmarking. In Manual of psychosocial rehabilitation (pp. 229–239). Wiley-Blackwell. https://doi.org/ 10.1002/9781118702703.ch18 Osumba, J. J. L., Recha, J. W., & Oroma, G. W. (2021). Transforming agricultural extension service delivery through innovative bottom–up climate-resilient agribusiness farmer field schools. Sustainability, 13(7), 3938. https://doi.org/10.3390/su13073938 Owen, J. (2004). Evaluation roots. SAGE Publications, Inc. https://doi.org/10.4135/ 9781412984157 Radhakrishna, R., & Bowen, C. F. (2010). Viewing Bennett’s hierarchy from a different lens: Implications for extension program evaluation. Journal of Extension, 48(6), 1–5. Salehi, M., Abbasi, E., Bijani, M., et al. (2021). Evaluation of agricultural extension model sites approach in Iran. Journal of the Saudi Society of Agricultural Sciences, 20(8), 506–518. https:/ /doi.org/10.1016/j.jssas.2021.06.002 Sasidhar, P. V. K., Suvedi, M., Vijayaraghavan, K., et al. (2011). Evaluation of a distance education radio farm school programme in India: Implications for scaling up. Outlook on Agriculture, 40(1), 89–96. https://doi.org/10.5367/oa.2011.0033 Toosi, M., Modarres, M., Amini, M., et al. (2021). Context, input, process, and product evaluation model in medical education: A systematic review. Journal of Education and Health Promotion, 10(1), 199. https://doi.org/10.4103/jehp.jehp_1115_20 Vali, L., Ataollahi, F., Amiresmaili, M., et al. (2021). The requirements of developing programs for the management of non-communicable diseases in Iran based on the CIPP model: A qualitative study. Journal of Health Research, 36, 1068–1077. https://doi.org/10.1108/JHR-11-2020-0546 van den Berg, H., Phillips, S., Poisot, A.-S., et al. (2021). Leading issues in implementation of farmer field schools: A global survey. Journal of Agricultural Education and Extension, 27(3), 341–353. https://doi.org/10.1080/1389224X.2020.1858891

Chapter 14

E-Banking Loyalty and Its Background: A Bibliometric Analysis Natacha López-Hernando, Cristina Loranca-Valle, and Pedro Cuesta-Valiño

Abstract Digital banking has become essential from the point of view of both the client and financial institutions. The restructuring of the banking sector is causing the closure of bank branches and the reduction of employees in search of greater profitability for the entities, which has caused an acceleration in client-entity relations through e-banking, being this type of non-face-to-face service vital for banking entities. This document analyzes a detailed review of the scientific literature on one of the most significant variables in the current field of banking marketing, digital loyalty, due to its contribution to obtaining competitive advantages and financial results. A bibliometric analysis and mapping of 209 publications on customer loyalty and e-banking from 1999 to 2022 have been executed. The Web of Science database was the one used for the conducted search. The results obtained group the most outstanding data of the systematic review of the loyalty variable within the field of digital banking: the antecedents of loyalty, the concepts with the greatest presence in the literature, and the most relevant sources and authors in this environment. This chapter presents the most important factors related to digital loyalty: satisfaction, quality of service, and trust. The work shows that these variables, which have been widely studied in a general way in the marketing field, also appear in the e-banking loyalty literature. However, there are few articles that relate all the variables in the same study, which indicates that it would be recommendable to continue investigating this topic in future research. Keywords Loyalty · E-banking · Satisfaction · Trust · Service quality · Bibliometric analysis

N. López-Hernando () Department of Economics and Business Administration, University of Camilo José Cela, Madrid, Spain C. Loranca-Valle · P. Cuesta-Valiño Department of Economics and Business Administration, Universidad de Alcalá, Alcalá de Henares, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_14

215

216

N. López-Hernando et al.

14.1 Introduction Customer loyalty is one of the most debated topics in the marketing literature in recent years. From the point of view of the banking business, customer loyalty is considered a key route to profitability (Ehigie, 2006; Reichheld, 1993). In online environments, loyalty to virtual establishments (e-loyalty) has been viewed as a favorable attitude of the customer towards an entity with a repeated behavior of use/purchase (Anderson & Srinivasan, 2003). Nonetheless, this exclusively behavioral conceptualization has been criticized (Dick & Basu, 1994; Shankar et al., 2003), since a user can constantly repeat the purchase of a service (thus behaviorally faithful), without having a favorable attitude towards that behavior and therefore it will continue to look for other alternatives or will not recommend the service to other potential users. Global commercial offers are just a few clicks away, which causes high competition between entities, making it necessary to analyze the factors that can increase customer loyalty and satisfaction in virtual environments (Bhattacherjee, 2002; Sharma & Sheth, 2004; Srinivasan et al., 2002). Various authors show that trust in the selling company in virtual environments is a primary factor in the interactive purchase decision, substantially when the buyer does not have much information about that company (Flavián & Guinalíu, 2006, 2007; Jarvenpaa et al., 1999; Wang & Emurian, 2005). In the context of online banking, trust also appears as a precedent with special relevance in online banking services (Ba, 2001; Kassim & Abdulla, 2006; Suh & Han, 2002). Loyalty or fostering consumer loyalty is a key factor in all types of organizations, which is why there is a need to analyze e-banking loyalty in this study and whether the variables that explain loyalty in other organizations are applicable to this field. This is the motivation for the following questions: Q1: Does e-banking loyalty behave in the same way as consumer loyalty in any other organizations? Q2: Is service quality a key variable for online banking consumer loyalty? Q3: Is satisfaction the most noteworthy variable of e-banking loyalty? Q4: Is customer loyalty affected by the trust variable in e-banking? Next, a review of the existing literature on loyalty from an e-banking point of view is presented, first analyzing the different concepts of loyalty and their main antecedents (satisfaction, service quality, and trust), to later specify which search criteria have been applied. In the bibliometric analysis, the obtained results related to e-banking loyalty are analyzed, achieved from the systematic analysis of the existing literature, ending with the conclusions section of this work.

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

217

14.2 Concept of Loyalty and Its Main Background The loyalty of service companies has been conceptualized by various authors and has also been studied in the area of commercial banking. Customer loyalty has been shown to be critical to the long-term financial performance of business organizations (Jones & Sasser, 1998; Teeroovengadum, 2022). This aspect has also been verified with respect to retail banking (Belás & Gabˇcová, 2016; Mcdougall & Levesque, 2000; Teeroovengadum, 2022). It has been observed that the relationship between customer loyalty variables and customer intention to continue using mobile banking applications is significant (Shahid et al., 2022), as well as that this proximity to the customer has a direct and indirect influence on loyalty (Mulia et al., 2020). Many authors agree that loyalty comprises all purchase behaviors of the same brand repeatedly, both now and in the future (Cuesta-Valiño et al., 2021b, 2022; Dick & Basu, 1994; Hennig-Thurau et al., 2002; Holbrook & Chaudhuri, 2001; Loranca-Valle et al., 2021; Oliver, 1999) so retail banks focus on developing strategies to improve customer satisfaction, which has been shown to be a variable strongly related to loyalty (Anderson & Fornell, 2000). The concept of customer loyalty has been fundamental for any business organization (Bhat et al., 2018), although it is a challenge for a service company to create and retain loyal customers (Mainardes et al., 2020) since the cost of attracting new customers is relatively higher than retaining existing ones (Richard & Zhang, 2012), making customer retention crucial for a service business. Loyal customers maintain a positive perception of the company, purchase frequently and repeatedly and suggest that other customers buy there, too (Levy & Hino, 2016). In the case of the banking sector, it has been confirmed that the levels of electronic trust and electronic loyalty depend on the structure of electronic banking. In another study, it is indicated that in banks following a click-and-mortar business model, this model referring to a combination of physical and digital business strategy (Turban et al., 2000), users showed more trust and loyalty in their online banks than users of banks that followed a pure E-Commerce strategy, with only a digital dimension (Kaabachi et al., 2020).

14.2.1 Satisfaction The improvement in customer satisfaction leads to increased customer loyalty (Heskett et al., 1990; Kashif et al., 2015; Kaur & Soch, 2018). It has been shown that there is a relationship between quality and satisfaction, satisfaction and loyalty, trust and loyalty, among other variables, and the use of mobile banking, the most significant being the relationship between loyalty and the use of mobile banking (Alonso-Dos-Santos et al., 2020). Various authors state that the following:

218

N. López-Hernando et al.

• Customer satisfaction has a positive and significant effect on loyalty (Coker, 2013; Esmaeili et al., 2021; Nguyen et al., 2020; Thakur, 2014). • Customer satisfaction is a direct antecedent of consumer loyalty (Eid, 2011) • Higher levels of customer satisfaction lead to higher levels of individual loyalty among customers (Yoon & Kim, 2000). With the electronic banking service, customers do not need to go to the bank in person and all transactions are confirmed immediately, anywhere, in a fast and profitable way, so the elements of the quality of the banking service become an important factor to increase customer satisfaction and strengthen customer loyalty (Wang et al., 2003). Satisfaction is a comparison between perceived reality and customer expectations, so the level of satisfaction depends on the differences between what is actually perceived and what is expected (Krampf et al., 2003). Various studies support that the quality of the electronic banking service has a positive correlation with customer satisfaction, so that commercial banks can increase the satisfaction of their customers by increasing the quality of their online service (Jun & Cai, 2001; Olsen & Johnson, 2003; Parasuraman et al., 1985; Pikkarainen et al., 2006; Siu & Mou, 2005).

14.2.2 Service Quality Quality perceived as satisfactory generates customer satisfaction, which ultimately translates into customer loyalty (Singla, 2012). Quality has been defined in various ways: • • • • • •

Value (Abbott, 1955; Feigenbaum, 1951). Compliance with the required requirements (Crosby, 1979). The aptitude or suitability for the use of the product (Juran, 1996). Absence of flaws (Crosby, 1979). The opportunity cost for the client (Ross, 1989). The result of meeting and exceeding consumer expectations (Gronroos, 1983; Parasuraman et al., 1985).

However, one of the most agreed definitions is that of discrepancy between the prior expectations of a consumer about a certain service and his perception about the purchase or the services (Boulding et al., 1993; Hennig-Thurau et al., 2002; Parasuraman et al., 1988). This means that, based on the same perception, the higher the consumer’s expectations, the lower the perceived quality (Parasuraman et al., 1988). Models have been studied in which service quality has been related to loyalty satisfaction, trust, and the different dimensions of commitment (affective, conative, and cognitive) (Cuesta-Valiño et al., 2021a). The quality of a website-based service is different from that of traditional services. Therefore, it is important to investigate service quality in the Internet banking industry (Choudhury & Beis, 2013; George & Kumar, 2014; Ho &

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

219

Lin, 2010; Kaura et al., 2015; Ranaweera & Sigala, 2015). Bank customers are increasingly more open to competitive developments, so Internet service quality alone may not be enough to ensure a long-term relationship between customers and banks (Brun et al., 2014). Consequently, customer satisfaction and loyalty have been identified as important factors in building and maintaining the relationship with customers to reduce the perceived risk of using Internet banking (Aldás Manzano et al., 2011; Amin, 2016; Chen et al., 2012; Chen, 2013; Dahlstrom et al., 2014). In general, the quality of the electronic service has a great impact on customer loyalty and satisfaction (Ganguli & Roy, 2011; Puriwat & Tripopsakul, 2017).

14.2.3 Trust The mobile banking-related literature by Burrell and Morgan (1994) suggests that customer trust has a positive effect on loyalty. Trust is the belief on the side of one of the parties that the actions carried out by the other party will necessarily satisfy their needs (Anderson & Mansi, 2009; Anderson & Weitz, 1989; Cuesta-Valiño et al., 2020; Loranca-Valle et al., 2021). Trust is one of the basic ingredients to establish successful relationships (Cuesta-Valiño et al., 2019). In virtual banking, it is essential to increase consumer confidence, since the risk of losing customers due to the effectiveness of online banking is greater than that in traditional banking (Gerrard & Barton Cunningham, 2003; Hewer & Howcroft, 1999; Lee & Hwan, 2005; Lee et al., 2005; Polatoglu & Ekin, 2001; Sadiq Sohail & Shanmugham, 2003). Customers not trusting a virtual establishment will not be loyal, even if they are generally satisfied with it (Anderson & Srinivasan, 2003; Shannon, 1998). In the virtual environment, as a consequence of the mistrust it generates, one of the most effective means of communication is word of mouth, especially if the recommendation comes from close individuals whom the consumer trusts (Ennew et al., 2000). Customer loyalty towards a virtual establishment is related to their levels of trust (Flavián et al., 2004; Flavián & Guinalíu, 2006, 2007; Lee et al., 2000). In the specific context of online banking, the study by Floh and Treiblmaier (2015), shows the direct and positive influence of customer trust in a given online entity on the loyalty towards it.

14.3 Methods Following other examples (Aguirre Montero & López-Sánchez, 2021; LorancaValle et al., 2021; Södergren, 2021), to identify, catalog, and outline the existing publications on the loyalty variable in the context of e-banking and m-banking, a systematic review of the scientific literature was analyzed. The procedure was carried out in three steps: identifying the literature, extracting the data, and presenting the results. In the search phase, the potentially most suitable articles for

220

N. López-Hernando et al.

this research were identified using the Web of Science (WoS) data source. WoS was used as the main source because it is one of the most important scientific academic databases which contains a large collection of scientific publications with a high impact rate. Likewise, its advanced search feature allows for evaluating, analyzing, and classifying the studies examined. The concept of loyalty has been extensively investigated in various fields, so very limited search criteria have been used. The keywords used in the search were: “loyalty and e banking”, “loyalty and e-banking”, “loyalty and satisfaction and e banking”, “loyalty and satisfaction and e-banking”, “loyalty and service quality and e banking”, “loyalty and service quality and e-banking”, “loyalty and trust and e banking” and “loyalty and trust and e-banking”. The variables that mainly affect loyalty (satisfaction, service quality, and trust) were linked to e-banking and ebanking. The time range of the search was established between 1999 and 2021. During this period, the banking service has evolved experiencing a significant change, with the appearance of online banking, which has become one of the most frequently used channels by financial services customers (Acosta et al., 2006). With the increasing use of smartphones, it is crucial for banks to offer mobile banking to allow their customers to conduct bank transactions at their convenience. Banks need to understand what factors play a role in encouraging or discouraging customers from using mobile banking (Changchit et al., 2020). In the first classifying phase, 1597 articles were collected which, once duplicates had been withdrawn, totaled 613. The keywords and summaries of the classified articles were analyzed, excluding those not related to the banking sector, leaving 381 articles. Subsequently, those which did not measure the loyalty variable were eliminated, resulting in 209 articles that measured the loyalty variable in a digital banking context. An Excel spreadsheet was used to analyze the articles, including all phases of classification, categorization, and subtraction. In the next section, the data extracted using the WoS database is analyzed.

14.4 Result of the Bibliometric Analysis The results of this review indicate that there are several authors who have extensive experience in e-banking loyalty. Table 14.1 shows the authors with the highest number of published articles and the most cited authors among the analyzed articles. It is remarkable or noticeable that none of the authors found with the most publications regarding the search for e-banking is among the most cited articles. In the first position in the ranking of the most cited authors Chebat, J.C. can be found, who has published two articles related to digital banking. It is also surprising that among the most scholarly authors on the subject with the largest number of published articles, none appears among the most cited. For example, Aali et al. (2020), who has published four articles, maintains 92 citations, while Chebat, J.C.

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis Table 14.1 Number of papers and citations per author

Author Aali, S Ekers, JE Marimon, F Nehowing, KR Sharma, P Author Chebat & Slusarczyk (2005) de Jong & de Ruyter (2004) Baabdullah et al. (2019) Perez, AC Kantsperger & Kunz (2010)

221 Number of publications 4 4 4 4 4 Number of citations 345 132 122 113 105

Source: Authors’ creation Table 14.2 Number of publications and citations per journal Journal International Journal of Bank Marketing Sustainability Total Quality Management & Business Excellence Journal of Business Research Journal of Islamic Marketing Journal Information Systems Research Journal of Business Research Expert Systems with Applications Online Information Review Decision Sciences

Number of publications 18 10 10 9 9 Number of citations 371 355 193 157 133

Source: Authors’ creation

with two published articles maintains 345 citations. It is also remarkable that the maximum number of articles regarding “loyalty and e-banking” published by the same author is only four. This allows us to envisage that there is still a long way to go in the academic study of digital banking. There are numerous journals which study banking services and the new technologies applied to them. In addition, articles on gaining the loyalty of the bank customer/digital banking user are also published. The journals which publish the most articles on the studied subject are mainly specific economics and business journals and environmental sciences. The important weight of the study of digital banking within the subject of environmental sciences must be highlighted, being the relationship between online banking and the care for the environment a significant one. Table 14.2 shows the five journals that publish the most journal as well articles on this topic. However, the most cited magazines on gaining customer loyalty in electronic banking cover the category of management, with no magazine specializing in banking appearing in the top positions.

222

N. López-Hernando et al.

Table 14.3 Most cited articles by most cited authors Articles How emotions mediate the effects of perceived justice on loyalty in service recovery situations: An empirical study Adaptive versus proactive behavior in service recovery: The role of self-managing teams Consumer use of mobile banking (M-Banking) in Saudi Arabia: Towards an integrated model Corporate social responsibility and customer loyalty: Exploring the role of identification, satisfaction and type of company Consumer trust in service companies: A multiple mediating analysis

Cites Authors 345 Chebat & Slusarczyk (2005) 132 de Jong & de Ruyter (2004) 122 Baabdullah, AM 113

Perez, A

105

Kantsperger & Kunz (2010)

Source: Authors’ creation Table 14.4 Geographic distribution of loyalty and e-banking

Country India United States PR China Spain Taiwan

Number of publications 58 49 45 35 35

Source: Authors’ creation

Table 14.2 shows the number of publications and citations per journal. Regarding the most cited articles, the following stand out, which in turn coincide with the most cited authors (Table 14.3): Regarding the geographical area where more content is published on gaining consumer loyalty in electronic banking, the following countries stand out: India with 58 articles, the United States with 49, Peoples R China with 45, followed by Spain and Taiwan, both with 35. Table 14.4 shows the countries that study the most about e-banking loyalty, with the darkest shade of blue being the country that publishes the most articles, the shade fading according to the number of articles published, and the gray countries being the ones that do not post anything about it (Fig. 14.1). Within the 613 articles analyzed, after the search for “loyalty and e banking” and “loyalty and e-banking” and after the result obtained when searching for “loyalty and satisfaction and e banking”, “loyalty and satisfaction and e-banking” out of 369 articles, it stands out that there are only 154 which relate both variables in their study. Furthermore, analyzing the result of searching for “loyalty and service quality and e banking”, “loyalty and service quality and e-banking”, where 273 articles have been obtained, the figure is reduced to 120 when we select those which deal with the relationship between both variables. In the case of the search for “loyalty and trust and e banking” and “loyalty and trust and e-banking”, corresponding to the variable least related to loyalty of the three studied, 99 items are obtained. Therefore, it is concluded that in the literature that appears in the search for loyalty in online financial services, the variables that are most related to it in the studies – ranked

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

223

Fig. 14.1 Geographic distribution of loyalty and e-banking. (Source: Authors’ creation)

articles LOYALTY and TRUST 27%

LOYALTY and SATISFACCION 41%

LOYALTY and QUALITY 32%

LOYALTY and SATISFACCION

LOYALTY and QUALITY

LOYALTY and TRUST

Source: authors´ creation

Fig. 14.2 Most cited variables. (Source: Authors’ creation)

from the highest to the lowest- are satisfaction, quality of service, and trust (Figs. 14.2 and 14.3). In total, there are only 12 articles that relate the four variables together in their study. Table 14.5 shows the articles that relate all the variables and the number of citations. However, none of them are among the most cited. This makes us intuit that

224

N. López-Hernando et al.

articles LOYALTY and TRUST 27%

LOYALTY and SATISFACCION 41%

LOYALTY and QUALITY 32%

LOYALTY and SATISFACCION

LOYALTY and QUALITY

LOYALTY and TRUST

Source: authors´ creation

Fig. 14.3 Most cited articles. (Source: Authors’ creation)

there is still a lot of field to analyze in the relationship between the four variables loyalty, satisfaction, quality of service, and trust in the field of e-banking, being an important study to take into account, since the use of online banking has become an essential service in the financial sector, with banks depending on electronic banking to enhance their productivity and increase the accessibility of their customers to their web pages at low cost (Gupta & Mittal, 2013).

14.5 Discussion and Conclusion 14.5.1 Theoretical and Managerial Discussions The loyalty construct appears generalized in the economic academic literature and particularly in the context of digital banking, which makes it difficult to locate the publications that may interest us the most. That is why this bibliometric review of more than 200 references is proposed to highlight and organize the most important information on digital loyalty in banking and detect the different elements which define this concept, as well as to compare whether these elements are similar to the variables analyzed in the generic study of loyalty. Therefore, the variables participating in loyalty have been examined in depth. Satisfaction is the variable that appears the most in literature along with digital loyalty, since it is considered to be the antecedent of loyalty and implies that the satisfied user perceives a higher value of the service provided (Amin, 2016;

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

225

Table 14.5 Articles that include all the variables References Chu et al. (2012) Kao and Lin (2016) Kingshott et al. (2018) Do Thanh Nguyen et al. (2020) Mulia et al. (2020) Esmaeili et al. (2021) Chao et al. (2009) Garepasha et al. (2020) Sharma and Goel (2021) Arjmand and Boldaji (2018)

Gong et al. (2014) Terzides et al. (2011)

Title Service quality, customer satisfaction, customer trust, and loyalty in an e-banking context The relationship between perceived e-service quality and brand equity: A simultaneous equations system approach The impact of relational versus technological resources on e-loyalty: A comparative study between local, national and foreign branded banks Impact of service quality, customer satisfaction and switching costs on customer loyalty The role of customer intimacy in increasing Islamic bank customer loyalty in using e-banking and m-banking Customer loyalty in mobile banking: Evaluation of perceived risk, relative advantages, and usability factors Customer loyalty in virtual environments: An empirical study in e-bank Dynamics of online relationship marketing: Relationship quality and customer loyalty in Iranian banks I can live without banks, but not without banking: Role of trust on loyalty and evangelism A study on the role and status of customers’ trust and economic risk in E-businesses with an emphasis on customers’ behaviors (Case study: Tehran Province Ayandeh Bank E-banking customers) Applying the IS success model and service quality on e-banking in China Innovative strategies of electronic communication in the greek banking sector

Times cited 45 34 20

16

7 3 3 2 0 0

0 0

Source: Authors’ creation

Shankar et al., 2003). In the banking sector, there are several studies that deduce a positive relationship between satisfaction and service quality (Hansen & Sand, 2008; Ladhari et al., 2011; Levesque & Mcdougall, 1996; Olsen & Johnson, 2003; Wang et al., 2003). It is also noteworthy that when consumers trust an entity, it positively affects their feelings, improving its reputation (Walsh et al., 2009). However, despite being a poorly understood variable, online trust is critical to loyalty in financial marketing (Das & Teng, 2004; Harridge-March, 2006; Sultan & Mooraj, 2001; Walczuch & Lundgren, 2004). Among the studied elements related to loyalty, online trust stands out, which according to Pavlou (2003), is the most notorious factor in consumer behavior when conducting internet transactions (Pavlou, 2003). Nevertheless, it must be taken into account that, although it has been one of the most studied concepts in the literature, it has also been the least understood variable (Das & Teng, 2004; Gefen et al., 2003; Gefen & Straub, 1997; Grabner-Kräuter & Faullant, 2008; Najafi, 2014; Pavlou

226

N. López-Hernando et al.

& Fygenson, 2006; Riegelsberger et al., 2005). Lack of trust is positioned as the most important reason when related to consumers’ distrust of using the Internet to carry out financial operations (Grabner-Kräuter & Faullant, 2008), even more so when the entity and its website are unknown or the consumer is unfamiliar with this technology (Cheung & Lee, 2006). Regarding the existing relationship between satisfaction and loyalty, Bhattacherjee (2002), has studied it in the context of virtual environments, being (Anderson & Srinivasan, 2003) the ones who regard the study of satisfaction as a central component of the user’s decision in banking virtual environments. Regarding the quality of the online service, several authors appear in the literature mentioning it. Montoya-Weiss et al. (2003) consider that consumers’ perception of the quality of the online service can be measured considering the quality of the information provided, the way they navigate on their website, and its design, while (Amin, 2016) establishes that an increase in the quality of e-banking is related to loyalty, customer satisfaction and its purpose of continuing to deal with the bank. Other authors indicate that a high quality of service is frequently associated with the loyalty, trust, and satisfaction variable (Bitner, 1990; Caruana, 2002; Headley & Miller, 1993; Kumar et al., 2009; Zeithaml et al., 1996) and direct and indirectly with loyalty (Cronin & Taylor, 1992; Loranca-Valle et al., 2019, 2021; Mandhachitara & Poolthong, 2011). The percentage of presence of these two variables together in the e-banking literature covers almost 74%, somewhat higher than the presence of this set of variables in the publications of the literature on sports management analyzed in the article by Loranca-Valle et al. (2021), which is close to 53%. The presence of the variables loyalty and service quality together is also greater in the distance banking literature compared to that of sports management: 58% compared to 38%, with the same trend existing in the joint comparison of the loyalty and confidence variables: 48% vs. 11%. It can be concluded that the study of the variables loyalty, satisfaction, quality of service, and trust as a whole in the context of electronic banking is superior to that of the sports management sector. If we compare the most cited variables related to e-banking loyalty, satisfaction appears first, 41%, followed by service quality, 32%, and trust, 27%, which coincides with the proportion of articles that include these relationships (Fig. 14.3). Therefore, it is proposed to expand the study of these variables to improve the loyalty of online banking customers, since it has been shown that keeping a customer can be up to ten times cheaper than gaining a new one (Heskett et al., 1990).

14.5.2 Limitations and Future Research After this research, a wide variety of studies on loyalty in digital banking services are verified. Still, of the 209 scrutinized, only 5.74% raise the relationship between all the analyzed variables: loyalty, quality of service, satisfaction, and trust. Therefore, the design of new research in the analysis of models which explain the

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

227

loyalty variable in this context is proposed, even introducing more variables that may possibly impact on the loyalty of the customer of the online service.

14.5.3 Conclusion Online banking arouses great environmental, social and economic inclusion interest which generates expectations of study in the academic literature demonstrated by the finding of a large number of articles focused on the topic chosen for study. However, there are not many articles that relate at the same time the loyalty variable with satisfaction, service quality, and trust in their study in the context of digital banking. The research tries to glimpse which fields of the scientific study of e-banking are still to be developed in the analyzed literature. Data Availability Statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Author Contributions All authors contributed to the conception and design of the research, reviewed the theoretical background, organized the database, performed the bibliometric analysis, wrote all sections of the manuscript, contributed to the review of the manuscript, read and approved the submitted version.

References Aali, S., Ibrahimi, A., Mirabi, V. R., & Zare, S. (2020). The role of organizational leadership in customer loyalty through managing conflict between employees and customers. International Journal of Organizational Leadership, 8(3), 1–16. https://doi.org/10.33844/ijol.2020.60479 Abbott, L. (1955). Quality and competition. Columbia University Press. https://doi.org/10.7312/ abbo92492 Acosta, M., Sainz, J., & Salvador, B. (2006). Hago click y opero a tu lado: Estrategia de la banca online en España. Cuadernos de Gestion, 6(1), 101–110. Aguirre Montero, A., & López-Sánchez, J. A. (2021). Intersection of data science and smart destinations: A systematic review. Frontiers in Psychology, 12, 712610. https://doi.org/10.3389/ fpsyg.2021.712610 Aldás Manzano, J., Lassala Navarré, C., Ruiz Mafé, C., & Sanz Blas, S. (2011). Análisis de los factores determinantes de la lealtad hacia los servicios bancarios online. Cuadernos de Economía y Dirección de la Empresa, 14(1), 26–39. https://doi.org/10.1016/j.cede.2011.01.003 Alonso-Dos-Santos, M., Soto-Fuentes, Y., & Valderrama-Palma, V. A. (2020). Determinants of mobile banking users’ loyalty. Journal of Promotion Management, 26(5), 615. https://doi.org/ 10.1080/10496491.2020.1729312 Amin, M. (2016). Internet banking service quality and its implication on e-customer satisfaction and e-customer loyalty. International Journal of Bank Marketing, 34(3), 280. https://doi.org/ 10.1108/IJBM-10-2014-0139 Anderson, E. W., & Fornell, C. (2000). Foundations of the American customer satisfaction index. Total Quality Management, 11(7), 869. https://doi.org/10.1080/09544120050135425

228

N. López-Hernando et al.

Anderson, E. W., & Mansi, S. A. (2009). Does customer satisfaction matter to investors? Findings from the bond market. Journal of Marketing Research, 46(5), 703. https://doi.org/10.1509/ jmkr.46.5.703 Anderson, R. E., & Srinivasan, S. S. (2003). E-satisfaction and E-loyalty: A contingency framework. Psychology and Marketing, 20(2), 123–138. https://doi.org/10.1002/mar.10063 Anderson, E., & Weitz, B. (1989). Determinants of continuity in conventional industrial channel dyads. Marketing Science, 8(4), 310–323. https://doi.org/10.1287/mksc.8.4.310 Arjmand, P., & Boldaji, A. T. (2018). A study on the role and status of customers’ trust and economic risk in E-businesses with an emphasis on customers’ behaviors (Case study: Tehran Province Ayandeh Bank E-banking customers). International Journal of Computer Science and Network Security, 18(1), 173–178. Ba, S. (2001). Establishing online trust through a community responsibility system. Decision Support Systems, 31(3), 323–336. https://doi.org/10.1016/S0167-9236(00)00144-5 Baabdullah, A. M., Alalwan, A. A., Rana, N. P., Kizgin, H., & Patil, P. (2019). Consumer use of mobile banking (M-banking) in Saudi Arabia: Towards an integrated model. International Journal of Information Management, 44, 38–52. https://doi.org/10.1016/j.ijinfomgt.2018.09.002 Belás, J., & Gabˇcová, L. (2016). The relationship among customer satisfaction, loyalty and financial performance of commercial banks. E a M: Ekonomie a Management, 19(1), 132. https://doi.org/10.15240/tul/001/2016-1-010 Bhat, S. A., Darzi, M. A., & Parrey, S. H. (2018). Antecedents of customer loyalty in banking sector: A mediational study. Vikalpa, 43(2), 92. https://doi.org/10.1177/0256090918774697 Bhattacherjee, A. (2002). Individual trust in online firms: Scale development and initial test. Journal of Management Information Systems, 19(1), 211–241. M E Sharpe Inc. Bitner, M. J. (1990). Evaluating service encounters: The effects of physical surroundings and employee responses. Journal of Marketing, 54(2), 69. https://doi.org/10.1177/ 002224299005400206 Boulding, W., Kalra, A., Staelin, R., & Zeithaml, V. A. (1993). A dynamic process model of service quality: From expectations to behavioral intentions. Journal of Marketing Research, 30(1), 7. https://doi.org/10.1177/002224379303000102 Brun, I., Rajaobelina, L., & Ricard, L. (2014). Online relationship quality: Scale development and initial testing. International Journal of Bank Marketing, 32(1), 5–27. https://doi.org/10.1108/ IJBM-02-2013-0022 Burrell, G., & Morgan, G. (1994). Sociological paradigms and organisational analysis. Ashgate. Caruana, A. (2002). Service loyalty: The effects of service quality and the mediating role of customer satisfaction. European Journal of Marketing, 36, 811. https://doi.org/10.1108/ 03090560210430818 Changchit, C., Klaus, T., Lonkani, R., & Sampet, J. (2020). A cultural comparative study of mobile banking adoption factors. Journal of Computer Information Systems, 60(5), 484. https://doi.org/ 10.1080/08874417.2018.1541724 Chao, Y., Lee, G. Y., & Ho, Y. C. (2009). Customer loyalty in virtual environments: An empirical study in e-bank. AIP Conference Proceedings, 1148(1), 497. https://doi.org/10.1063/1.3225357 Chebat, J. C., & Slusarczyk, W. (2005). How emotions mediate the effects of perceived justice on loyalty in service recovery situations: An empirical study. Journal of Business Research, 58(5), 664. https://doi.org/10.1016/j.jbusres.2003.09.005 Chen, M. K. (2013). The effect of language on economic behavior: Evidence from savings rates, health behaviors, and retirement assets. American Economic Review, 103(2), 690–731. https:// doi.org/10.1257/aer.103.2.690 Chen, C.-Y., Lin, Y.-H., & Claussen, C. L. (2012). Celebrity endorsement for sporting events using classical conditioning. International Journal of Sports Marketing and Sponsorship, 13(3), 46– 56. https://doi.org/10.1108/IJSMS-13-03-2012-B005 Cheung, C. M. K., & Lee, M. K. O. (2006). Understanding consumer trust in Internet shopping: A multidisciplinary approach. Journal of the American Society for Information Science and Technology, 57(4), 479–492. https://doi.org/10.1002/asi.20312

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

229

Choudhury, H. G., & Beis, K. (2013). The dimeric form of the unphosphorylated response regulator BaeR. Protein Science, 22(9), 1287–1293. https://doi.org/10.1002/pro.2311 Chu, P. Y., Lee, G. Y., & Chao, Y. (2012). Service quality, customer satisfaction, customer trust, and loyalty in an e-banking context. Social Behavior and Personality, 40(8), 1271. https://doi.org/ 10.2224/sbp.2012.40.8.1271 Coker, B. (2013). Antecedents to website satisfaction, loyalty, and word-of-mouth. Journal of Information Systems and Technology Management, 10(2), 209. https://doi.org/10.4301/s180717752013000200001 Cronin, J. J., & Taylor, S. A. (1992). Measuring service quality: A reexamination and extension. Journal of Marketing, 56(3), 55. https://doi.org/10.2307/1252296 Crosby, P. B. (1979). Quality is free. The art of making quality certain. McGraw-Hill. Corpus ID: 109686740. Accessed 22 May 2022. Cuesta-Valiño, P., Rodríguez, P. G., & Núñez-Barriopedro, E. (2019). The impact of corporate social responsibility on customer loyalty in hypermarkets: A new socially responsible strategy. Corporate Social Responsibility and Environmental Management, 26(4), 761. https://doi.org/ 10.1002/csr.1718 Cuesta-Valiño, P., Loranca-Valle, C., & Nuñez-Barriopedro, E. (2020). La Promoción del Deporte a través de la Felicidad del Deportista Federado en Kárate. aDResearch ESIC International Journal of Communication Research, 21(21), 48–69. https://doi.org/10.7263/adresic-021-03 Cuesta-Valiño, P., Gutiérrez-Rodríguez, P., & Loranca-Valle, C. (2021a). Sustainable Management of Sports Federations: The indirect effects of perceived service on member’s loyalty. Sustainability, 13(2), 458. https://doi.org/10.3390/su13020458 Cuesta-Valiño, P., Gutiérrez-Rodríguez, P., & Núnez-Barriopedro, E. (2021b). The role of consumer happiness in brand loyalty: A model of the satisfaction and brand image in fashion. Corporate Governance (Bingley), 22, 458. https://doi.org/10.1108/CG-03-2021-0099 Cuesta-Valiño, P., Gutiérrez-Rodríguez, P., & Loranca-Valle, C. (2022). Sponsorship image and value creation in E-sports. Journal of Business Research, 145, 198–209. https://doi.org/ 10.1016/j.jbusres.2022.02.084 Dahlstrom, E. A., Brooks, D. A., & Bichsel, J. (2014). The current ecosystem of learning management systems in higher education: Student, faculty, and IT perspectives. Research report. ECAR. Das, T. K., & Teng, B.-S. (2004). The risk-based view of trust: A conceptual framework. Journal of Business and Psychology, 19(1), 85–116. https://doi.org/10.1023/ B:JOBU.0000040274.23551.1b de Jong, A., & de Ruyter, K. (2004). Adaptive versus proactive behavior in service recovery: The role of self-managing teams. Decision Sciences, 35(3), 457. https://doi.org/10.1111/j.00117315.2004.02513.x Dick, A. S., & Basu, K. (1994). Customer loyalty: Toward an integrated conceptual framework. Journal of the Academy of Marketing Science, 22(2), 99. https://doi.org/10.1177/ 0092070394222001 Ehigie, B. O. (2006). Correlates of customer loyalty to their bank: A case study in Nigeria. International Journal of Bank Marketing, 24(7), 494. https://doi.org/10.1108/02652320610712102 Eid, M. I. (2011). Determinants of e-commerce customer satisfaction, trust, and loyalty in Saudi Arabia. Journal of Electronic Commerce Research, 12(1), 78–93. Ennew, C. T., Banerjee, A. K., & Li, D. (2000). Managing word of mouth communication: Empirical evidence from India. International Journal of Bank Marketing, 18(2), 75. https:// doi.org/10.1108/02652320010322985 Esmaeili, A., Haghgoo, I., Davidaviciene, V., & Meidute-Kavaliauskiene, I. (2021). Customer loyalty in mobile banking: Evaluation of perceived risk, relative advantages, and usability factors. Engineering Economics, 32(1), 70. https://doi.org/10.5755/j01.ee.32.1.25286 Feigenbaum, A. V. (1951). Quality control: Principles, practice and administration: An industrial management tool for improving product quality and design and for reducing operating costs and losses. McGraw-Hill. Corpus ID: 106472306. Accessed 22 May 2022.

230

N. López-Hernando et al.

Flavián, C., & Guinalíu, M. (2006). Consumer trust, perceived security and privacy policy: Three basic elements of loyalty to a web site. Industrial Management & Data Systems, 106(5), 601– 620. https://doi.org/10.1108/02635570610666403 Flavián, C., & Guinalíu, M. (2007). Un análisis de la influencia de la confianza y del riesgo percibido sobre la lealtad a un sitio web. El caso de la distribución de servicios gratuitos. Revista Europea de Dirección y Economía de la Empresa, 16(1), 159–178. Flavián, C., Guinalíu, M., & Gurrea, R. (2004). Análisis empírico de la influencia ejercida por la usabilidad percibida, la satisfacción y la confianza del consumidor sobre la lealtad a un sitio web. In XVI Encuentros de profesores universitarios de marketing (pp. 209–226). ESIC Editorial. Floh, A., & Treiblmaier, H. (2015). What keeps the e-banking customer loyal? A multigroup analysis of the moderating role of consumer characteristics on e-loyalty in the financial service industry. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2585491 Ganguli, S., & Roy, S. K. (2011). Generic technology-based service quality dimensions in banking. International Journal of Bank Marketing, 29(2), 168. https://doi.org/10.1108/ 02652321111107648 Garepasha, A., Aali, S., Zendeh, A. B., & Iranzadeh, S. (2020). Dynamics of online relationship marketing: Relationship quality and customer loyalty in Iranian banks. Revista Brasileira de Gestao de Negocios, 22(1), 140. https://doi.org/10.7819/rbgn.v22i1.4043 Gefen, D., & Straub, D. W. (1997). Gender differences in the perception and use of e-mail: An extension to the technology acceptance model. MIS Quarterly: Management Information Systems, 21(4), 389. https://doi.org/10.2307/249720 Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27(1), 51. https://doi.org/10.2307/30036519 George, A., & Kumar, G. S. G. (2014). Impact of service quality dimensions in internet banking on customer satisfaction. Decision, 41(1), 73–85. https://doi.org/10.1007/s40622-014-0028-2 Gerrard, P., & Barton Cunningham, J. (2003). The diffusion of Internet banking among Singapore consumers. International Journal of Bank Marketing, 21(1), 16. https://doi.org/10.1108/ 02652320310457776 Gong, J., Lee, U.-K., & Yoo, I.-S. (2014). Applying the IS success model and service quality on e-banking in China. Journal of Product Research, 32(3), 63–76. https://doi.org/10.36345/ kacst.2014.32.3.005 Grabner-Kräuter, S., & Faullant, R. (2008). Consumer acceptance of internet banking: The influence of internet trust. International Journal of Bank Marketing, 26(7), 483. https://doi.org/ 10.1108/02652320810913855 Gronroos, C. (1983). Strategic management and marketing in the service sector. Marketing Science Institute. Gupta, S., & Mittal, A. (2013). Emerging role of information technology in banking sector’s development of India. ACME International Journal of Multidisciplinary Research, 1, 61–65. Hansen, H., & Sand, J. A. (2008). Antecedents to customer satisfaction with financial services: The moderating effects of the need to evaluate. Journal of Financial Services Marketing, 13(3), 234. https://doi.org/10.1057/fsm.2008.18 Harridge-March, S. (2006). Can the building of trust overcome consumer perceived risk online? Marketing Intelligence and Planning, 24, 746. https://doi.org/10.1108/02634500610711897 Headley, D. E., & Miller, S. J. (1993). Measuring service quality and its relationship to future consumer behavior. Journal of Health Care Marketing, 13(4), 32–41. Hennig-Thurau, T., Gwinner, K. P., & Gremler, D. D. (2002). Understanding relationship marketing outcomes: An integration of relational benefits and relationship quality. Journal of Service Research, 4(3), 230. https://doi.org/10.1177/1094670502004003006 Heskett, J. L., Sasser, W. E., & Hart, C. W. L. (1990). Service break-throughs: Changing the rules of the game. The Free Press. Hewer, P., & Howcroft, B. (1999). Consumers distribution channel adoption and usage in the financial services industry: A review of existing approaches. Journal of Financial Services Marketing, 3, 344–358.

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

231

Ho, C. B., & Lin, W. (2010). Measuring the service quality of internet banking: Scale development and validation. European Business Review, 22(1), 5–24. https://doi.org/10.1108/ 09555341011008981 Holbrook, B., & Chaudhuri, A. (2001). The chain of effects from brand trust and brand affect to brand performance: The role of brand loyalty. Journal of Marketing, 65(2), 81–93. Jarvenpaa, S. L., Tractinsky, N., Saarinen, L., & Vitale, M. (1999). Consumer trust in an internet store: A cross-cultural validation. Journal of Computer-Mediated Communication, 5, JCMC526. https://doi.org/10.1111/j.1083-6101.1999.tb00337.x Jones, T. O., & Sasser, W. E. (1998). Why satisfied customers defect. IEEE Engineering Management Review, 26(3), 11. https://doi.org/10.1061/(asce)0742-597x(1996)12:6(11.2) Jun, M., & Cai, S. (2001). The key determinants of Internet banking service quality: A content analysis. International Journal of Bank Marketing, 19(7), 276–291. https://doi.org/10.1108/ 02652320110409825 Juran, J. M. (1996). La calidad y el diseño (Díaz de Santos, Ed.). España. Kaabachi, S., Ben Mrad, S., & Fiedler, A. (2020). The moderating effect of e-bank structure on French consumers’ trust. International Journal of Bank Marketing, 38(2), 501. https://doi.org/ 10.1108/IJBM-04-2019-0119 Kantsperger, R., & Kunz, W. H. (2010). Consumer trust in service companies: A multiple mediating analysis. Managing Service Quality: An International Journal, 20(1), 4. https:// doi.org/10.1108/09604521011011603 Kao, T. W., & Lin, W. T. (2016). The relationship between perceived e-service quality and brand equity: A simultaneous equations system approach. Computers in Human Behavior, 57, 208– 218. https://doi.org/10.1016/j.chb.2015.12.006 Kashif, M., Wan Shukran, S. S., Rehman, M. A., & Sarifuddin, S. (2015). Customer satisfaction and loyalty in Malaysian Islamic banks: A PAKSERV investigation. International Journal of Bank Marketing, 33(1), 23–40. https://doi.org/10.1108/IJBM-08-2013-0084 Kassim, N. M., & Abdulla, A. K. M. A. (2006). The influence of attraction on internet banking: An extension to the trust-relationship commitment model. International Journal of Bank Marketing, 24, 424. https://doi.org/10.1108/02652320610701744 Kaur, H., & Soch, H. (2018). Satisfaction, trust and loyalty: Investigating the mediating effects of commitment, switching costs and corporate image. Journal of Asia Business Studies, 12(4), 361–380. https://doi.org/10.1108/JABS-08-2015-0119 Kaura, V., Durga Prasad, C. S., & Sharma, S. (2015). Service quality, service convenience, price and fairness, customer loyalty, and the mediating role of customer satisfaction. International Journal of Bank Marketing, 33(4), 404–422. https://doi.org/10.1108/IJBM-04-2014-0048 Kingshott, R. P. J., Sharma, P., & Chung, H. F. L. (2018). The impact of relational versus technological resources on e-loyalty: A comparative study between local, national and foreign branded banks. Industrial Marketing Management, 72, 48–58. https://doi.org/10.1016/ j.indmarman.2018.02.011 Krampf, R., Ueltschy, L., & D’Amico, M. (2003). The contribution of emotion to consumer satisfaction in the service setting. Marketing Management Journal, 13(1), 32–52. Kumar, M., Tat Kee, F., & Taap Manshor, A. (2009). Determining the relative importance of critical factors in delivering service quality of banks. Managing Service Quality: An International Journal, 19(2), 211. https://doi.org/10.1108/09604520910943198 Ladhari, R., Ladhari, I., & Morales, M. (2011). Bank service quality: Comparing Canadian and Tunisian customer perceptions. International Journal of Bank Marketing, 29(3), 224. https:// doi.org/10.1108/02652321111117502 Lee, M.-C., & Hwan, I. S. (2005). Relationships among service quality, customer satisfaction and profitability in the Taiwanese banking industry. International Journal of Management, 22(4), 635–648. Lee, J., Kim, J., & Moon, J. Y. (2000). What makes Internet users visit cyber stores again? Key design factors for customer loyalty. In Proceedings of the SIGCHI conference on Human factors in computing systems – CHI ’00 (pp. 305–312). ACM Press. https://doi.org/10.1145/ 332040.332448

232

N. López-Hernando et al.

Lee, E. J., Kwon, K. N., & Schumann, D. W. (2005). Segmenting the non-adopter category in the diffusion of internet banking. International Journal of Bank Marketing, 23(5), 414. https:// doi.org/10.1108/02652320510612483 Levesque, T., & Mcdougall, G. H. G. (1996). Determinants of customer satisfaction in retail banking. International Journal of Bank Marketing, 14(7), 12. https://doi.org/10.1108/ 02652329610151340 Levy, S., & Hino, H. (2016). Emotional brand attachment: A factor in customer-bank relationships. International Journal of Bank Marketing, 34(2), 136. https://doi.org/10.1108/IJBM-06-20150092 Loranca-Valle, C., Cuesta-Valiño, P., & Núñez-Barriopedro, E. (2019). Gestión de calidad como estrategia clave de la felicidad en el deporte federado. Retos, 9(18), 203–218. https://doi.org/ 10.17163/ret.n18.2019.02 Loranca-Valle, C., Cuesta-Valiño, P., Núnez-Barriopedro, E., & Gutiérrez-Rodríguez, P. (2021). Management of loyalty and its main antecedents in sport organizations: A systematic analysis review. Frontiers in Psychology, 12, 1–11. https://doi.org/10.3389/fpsyg.2021.783781 Mainardes, E. W., Rosa, C. A. M., & Nossa, S. N. (2020). Omnichannel strategy and customer loyalty in banking. International Journal of Bank Marketing, 38(4), 799. https://doi.org/ 10.1108/IJBM-07-2019-0272 Mandhachitara, R., & Poolthong, Y. (2011). A model of customer loyalty and corporate social responsibility. Journal of Services Marketing, 25(2), 122. https://doi.org/10.1108/ 08876041111119840 Mcdougall, G. H. G., & Levesque, T. (2000). Customer satisfaction with services: Putting perceived value into the equation. Journal of Services Marketing, 14(5), 392. https://doi.org/ 10.1108/08876040010340937 Montoya-Weiss, M. M., Voss, G. B., & Grewal, D. (2003). Determinants of online channel use and overall satisfaction with a relational, multichannel service provider. Journal of the Academy of Marketing Science, 31(4), 448. https://doi.org/10.1177/0092070303254408 Mulia, D., Usman, H., & Parwanto, N. B. (2020). The role of customer intimacy in increasing Islamic bank customer loyalty in using e-banking and m-banking. Journal of Islamic Marketing, 12(6), 1097. https://doi.org/10.1108/JIMA-09-2019-0190 Najafi, I. (2014). Identify effective factors for improving E-trust of E-transactions in the context of E-commerce and E-government. International Journal of Computer Trends and Technology, 17(6), 281–299. https://doi.org/10.14445/22312803/IJCTT-V17P152 Nguyen, D. T., Pham, V. T., Tran, D. M., & Pham, D. B. T. (2020). Impact of service quality, customer satisfaction and switching costs on customer loyalty. Journal of Asian Finance, Economics and Business, 7(8), 395. https://doi.org/10.13106/JAFEB.2020.VOL7.NO8.395 Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63(4_suppl1), 33–44. https://doi.org/10.1177/00222429990634s105 Olsen, L. L., & Johnson, M. D. (2003). Service equity, satisfaction, and loyalty: From transactionspecific to cumulative evaluations. Journal of Service Research, 5(3), 184. https://doi.org/ 10.1177/1094670502238914 Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1985). A conceptual model of service quality and its implications for future research. Journal of Marketing, 49(4), 41. https://doi.org/10.1177/ 002224298504900403 Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). Servqual: A multiple-item scale for measuring consumer perceptions of service quality. Journal of Retailing, 64, 12–40. Spring. Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust and risk with the technology acceptance model. International Journal of Electronic Commerce, 7(3), 101. https://doi.org/10.1080/10864415.2003.11044275 Pavlou, P. A., & Fygenson, M. (2006). Understanding and predicting electronic commerce adoption: An extension of the theory of planned behavior. MIS Quarterly: Management Information Systems, 30(1), 115. https://doi.org/10.2307/25148720

14 E-Banking Loyalty and Its Background: A Bibliometric Analysis

233

Pikkarainen, K., Pikkarainen, T., Karjaluoto, H., & Pahnila, S. (2006). The measurement of enduser computing satisfaction of online banking services: Empirical evidence from Finland. International Journal of Bank Marketing, 24(3), 158. https://doi.org/10.1108/02652320610659012 Polatoglu, V. N., & Ekin, S. (2001). An empirical investigation of the Turkish consumers’ acceptance of Internet banking services. International Journal of Bank Marketing, 19(4), 156. https://doi.org/10.1108/02652320110392527 Puriwat, W., & Tripopsakul, S. (2017). The impact of e-service quality on customer satisfaction and loyalty in mobile banking usage: Case study of Thailand. Polish Journal of Management Studies, 15(2), 183. https://doi.org/10.17512/pjms.2017.15.2.17 Ranaweera, C., & Sigala, M. (2015). From service quality to service theory and practice. Journal of Service Theory and Practice, 25(1), 2–9. https://doi.org/10.1108/JSTP-11-2014-0248 Reichheld, F. F. (1993). Loyalty-based management. Harvard Business Review, 71(2), 64–73. Richard, J. E., & Zhang, A. (2012). Corporate image, loyalty, and commitment in the consumer travel industry. Journal of Marketing Management, 28(5–6), 568. https://doi.org/10.1080/ 0267257X.2010.549195 Riegelsberger, J., Sasse, M. A., & McCarthy, J. D. (2005). The mechanics of trust: A framework for research and design. International Journal of Human Computer Studies, 62(3), 381. https:/ /doi.org/10.1016/j.ijhcs.2005.01.001 Ross, P. J. (1989). Taguchi techniques for quality engineering. McGraw Hill. Sadiq Sohail, M., & Shanmugham, B. (2003). E-banking and customer preferences in Malaysia: An empirical investigation. Information Sciences, 150(3–4), 207–217. https://doi.org/10.1016/ S0020-0255(02)00378-X Shahid, S., Islam, J. U., Malik, S., & Hasan, U. (2022). Examining consumer experience in using m-banking apps: A study of its antecedents and outcomes. Journal of Retailing and Consumer Services, 65, 102870. https://doi.org/10.1016/j.jretconser.2021.102870 Shankar, V., Smith, A. K., & Rangaswamy, A. (2003). Customer satisfaction and loyalty in online and offline environments. International Journal of Research in Marketing, 20(2), 153–175. https://doi.org/10.1016/S0167-8116(03)00016-8 Shannon, J. (1998). Net nightmares thwart potential. Marketing Week, 21, 25. Sharma, N., & Goel, P. (2021). “I can live without banks, but not without banking”: Role of trust on loyalty and evangelism. International Journal of Electronic Government Research, 17(3), 1–20. https://doi.org/10.4018/IJEGR.2021070101 Sharma, A., & Sheth, J. N. (2004). Web-based marketing: The coming revolution in marketing thought and strategy. Journal of Business Research, 57(7), 696. https://doi.org/10.1016/S01482963(02)00350-8 Singla, V. (2012). Role of service quality and customer satisfaction in predicting customer loyalty. Journal for Global Business Advancement, 5(3), 248. https://doi.org/10.1504/ JGBA.2012.052268 Siu, N. Y. M., & Mou, J. C. W. (2005). Measuring service quality in internet banking: The case of Hong Kong. Journal of International Consumer Marketing, 17(4), 99. https://doi.org/10.1300/ J046v17n04_06 Södergren, J. (2021). Brand authenticity: 25 years of research. International Journal of Consumer Studies, 45(4), 645–663. https://doi.org/10.1111/ijcs.12651 Srinivasan, S. S., Anderson, R., & Ponnavolu, K. (2002). Customer loyalty in e-commerce: An exploration of its antecedents and consequences. Journal of Retailing, 78(1), 41. https://doi.org/ 10.1016/S0022-4359(01)00065-3 Suh, B., & Han, I. (2002). Effect of trust on customer acceptance of Internet banking. Electronic Commerce Research and Applications, 1(3–4), 247–263. https://doi.org/10.1016/S15674223(02)00017-0 Sultan, F., & Mooraj, H. A. (2001). Designing a trusted based a e-business strategy. Marketing Management, American Marketing Association, 1, 40–45. Teeroovengadum, V. (2022). Service quality dimensions as predictors of customer satisfaction and loyalty in the banking industry: Moderating effects of gender. European Business Review, 34(1), 1. https://doi.org/10.1108/EBR-10-2019-0270

234

N. López-Hernando et al.

Terzides, K., Papadopoulou, E., & Kosmidis, K. (2011). Innovative strategies of electronic communication in the Greek banking sector. In D. Vrontis, Y. Weber, H. Kaufmann, S. Tarba, & E. Tsoukatos (Eds.), Conference readings book proceedings of the 4th annual Euromed conference of the Euromed Academy of Business: Business research challenges in a turbulent era, Elounda, Greece (pp. 1839–1853). Thakur, R. (2014). What keeps mobile banking customers loyal? International Journal of Bank Marketing, 32(7), 628. https://doi.org/10.1108/IJBM-07-2013-0062 Turban, E., McLean, E., & Wetherbe, J. (2000). Information Technology for Management: Transforming organizations in the digital economy. Information Technology, 2(4), 210. https:// doi.org/10.1108/09593840010377644 Walczuch, R., & Lundgren, H. (2004). Psychological antecedents of institution-based consumer trust in e-retailing. Information & Management, 42(1), 159–177. https://doi.org/10.1016/ j.im.2003.12.009 Walsh, G., Mitchell, V. W., Jackson, P. R., & Beatty, S. E. (2009). Examining the antecedents and consequences of corporate reputation: A customer perspective. British Journal of Management, 20(2), 187. https://doi.org/10.1111/j.1467-8551.2007.00557.x Wang, Y. D., & Emurian, H. H. (2005). An overview of online trust: Concepts, elements, and implications. Computers in Human Behavior, 21(1), 105. https://doi.org/10.1016/j.chb.2003.11.008 Wang, Y., Lo, H. P., & Hui, Y. V. (2003). The antecedents of service quality and product quality and their influences on bank reputation: Evidence from the banking industry in China. Managing Service Quality: An International Journal, 13(1), 72. https://doi.org/10.1108/ 09604520310456726 Yoon, S. J., & Kim, J. H. (2000). An empirical validation of a loyalty model based on expectation disconfirmation. Journal of Consumer Marketing, 17(2), 120. https://doi.org/10.1108/ 07363760010317196 Zeithaml, V. A., Berry, L. L., & Parasuraman, A. (1996). The behavioral consequences of service quality. Journal of Marketing, 60(2), 31. https://doi.org/10.2307/1251929

Chapter 15

Evaluating European Climate Policy Impact on the CO2 Emissions Per Capita Convergence Process in the European Union Countries Dionisio Ramírez-Carrera, Gemma Durán-Romero, Ana M. López, and José Antonio Negrín de la Peña

Abstract Carbon dioxide (CO2 ) emission is considered the main reason for global warming and climate change. The European Union (EU) Environment Council recognized the need of limiting greenhouse gas emissions and considered that CO2 emissions should be significantly and immediately reduced. In this chapter, the effectiveness of European climate policies in reducing per capita CO2 emissions in European countries is evaluated using a convergence approach. We will examine the existence of stochastic convergence for CO2 emissions per capita in European Union countries by conducting a set of time series unit root tests between the years 1990 and 2018. We expect that, if European environmental policy has been successful, there will be a convergence process in the CO2 per capita emissions of all European countries. Our results suggest that there has been a convergence process in all the countries of the EU. In addition to this, applying the multiple breakpoint tests of Bai and Perron, we determine that the European climate policy has had a significant impact on the convergence process in most of the countries analyzed.

D. Ramírez-Carrera () University of Castilla-La Mancha (UCLM), Spanish and International Economics (Applied Economics), Facultad de Derecho y Ciencias Sociales, Ciudad Real, Spain e-mail: [email protected] G. Durán-Romero Dpto. Estructura Económica y Economía del Desarrollo, Autonomous University of Madrid, Instituto Complutense de Estudios Internacionales (ICEI), Madrid, Spain e-mail: [email protected] A. M. López Instituto L.R. Klein, Autonomous University of Madrid, Madrid, Spain e-mail: [email protected] J. A. Negrín de la Peña University of Castilla-La Mancha (UCLM), Facultad de Derecho y Ciencias Sociales, Ciudad Real, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_15

235

236

D. Ramirez-Carrera et al.

Keywords Carbon dioxide emissions · Stochastic convergence · European Union · Climate policy · Structural change

15.1 Introduction Nowadays, climate change is one of the most worrying environmental problems that mankind faces. Although greenhouse gas emissions (GHGs) do not directly affect human health, it causes undesirable effects in the atmosphere, land, and oceans in all regions of the world that will last for centuries. This GHG is mainly linked to the use of fossil fuels that have been used massively since the industrial revolution started in Europe. One of the main effects of these emissions has been the rise in the average temperature of the planet. Although in certain periods, the global temperature has remained stable or even it has decreased, it cannot be denied that, in general, the planet is warming. However, the change in global mean surface temperature is only a small part of a much bigger problem of unpredictable consequences such as extreme weather events or the disappearance of animal and marine species, and hundreds of kilometers of coastline and islands since it has been observed that the sea level is rising due to the expansion of the oceans as they warm and the additional water coming from the melting of the glaciers. If no efforts are made to reduce GHG, the sea could be almost ice-free before the middle of the twenty-first century (Delbeke & Vis, 2015). Thus, many industrialized countries, including most of the EU Member States, have implemented national policies and measures aimed at achieving the proposed objectives for the reduction of carbon dioxide emissions. The EU is the region of the world where the most climate policies have been implemented, and where practical policy experimentation in the field of the environment and climate change has been taking place at a rapid pace over the last 25 years. This has led to considerable success in reducing CO2 emissions (Fig. 15.1). However, this aggregate evolution can hide significant disparities. To prepare an EU climate policy after the end of the Kyoto Protocol, the European Commission defined new and ambitious commitments for carbon emissions reductions in 2007, the so-called 20/20/20.1 The historical Member States should make greater efforts to reduce their carbon emissions while the emissions of the new Member States could increase. Nevertheless, the largest declines in primary energy use occurred in Eastern European countries while the largest increases occurred in the historical Member States (Benjamin et al., 2015). To know if European climate policies in reducing per capita CO2 emissions in all the European countries have been successful, we will examine the existence 1 A reduction in EU greenhouse gas emissions of at least 20% below 1990 levels, 20% of EU energy production to come from renewable resources and a 20% reduction in primary energy use.

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

237

9

8.5

Metric tons

8

7.5

7

6.5

6

Fig. 15.1 EU CO2 per capita emissions from 1990 to 2018

of stochastic convergence of CO2 emissions from 1990 to 2018 in the 27 UE countries by conducting a set of time series unit root tests. It would be expected that if European climate policy had been effective, convergence for CO2 per capita emissions of all European countries would be observed. On the other hand, lack of convergence would be interpreted as the EU’s long-term objectives for climate policy not being correctly founded. Therefore, the one-size-fits-all energy and environmental policy in the EU may be deemed unreasonable and the environmental policies applied so far have not achieved the expected results.

15.2 Evolution of the UE Climate Policy EU climate policy initially started as part of environmental policy, established by the Single European Act, which entered into force in 1987. Through this Act, new provisions were incorporated into the EU treaty which referred to the environment. Today, 25 years later, there is a whole series of EU legislative measures dealing with the environmental protection of air quality, water, waste, and biodiversity. On the other hand, in the 1980s and 1990s, the development of the EU’s single market for goods and services raised questions about short-term distortions of competition because of environmental policy measures developed at the Member State level.

238

D. Ramirez-Carrera et al.

The institutional measure was embodied in the Single European Act which adopts rules to protect the environment at the European level, thereby minimizing the risk of distortion of competition in the internal market of the EU. During the years following the application of the new institutional arrangements focus on whether the policy mix should be broadened to include economic instruments and how can apply a pricing system for economic externality (Delbeke & Vis, 2015). In the 1990s, the European Commission proposed a combined carbon and energy tax. However, this generated controversy in the Council and Parliament. Finally, after almost 10 years of negotiations, due to the reservations of some Member States, the tax route was abandoned by the EU. Economic instruments at the EU level moved on to cap-setting and emissions trading, following in the footsteps of the United States. Although at first, the European Union was reluctant to implement this system, capping the total number of emissions is an environmental benefit and, in addition to this, the EU realized that was easier to reach agreements by a qualified majority in the Council, against the resistance inherent to the carbon and energy tax. Based on a Commission proposal, in March 2007, the European Council made an independent commitment to reduce global GHG by 20% compared to 1990 levels, which meant that the EU would apply the reduction commitment regardless of what other countries did in terms of reducing GHG. This commitment was accompanied by specific energy targets, in particular a binding target of increasing the share of renewable energy in final energy consumption and an energy consumption reduction target. This transition to a low-carbon economy required a Low Carbon Economy Roadmap and an Energy Roadmap, created by the European Commission in 2011.

15.3 The Convergence of Carbon Dioxide Emissions in the UE: Concepts and Empirical Review The convergence approach, since the seminal work of Strazicich and List (2003), has been used as an alternative to measuring the policy’s success in reducing carbon emissions. Convergence is an imprecise concept, with many interpretations, based on the assumption that the countries are initially in disequilibrium (Acaravci & Erdogan, 2016). Payne (2020) points out that the literature on income convergence (see Barro & Sala-i-Martin, 1992) has served as a basis for the empirical development of the analysis of the convergence of CO2 emissions, which can be divided into three fundamental concepts: beta (β), sigma (σ), and stochastic convergence, and further divided into absolute and conditional convergence (Pettersson et al., 2014). β-convergence occurs when countries with high initial levels of per capita CO2 emissions have a lower emission growth rate than countries with low initial levels of per capita CO2 emissions.

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

239

However, Quah (1993a, b), Evans (1996), and Evans and Karras (1996) noted that both approaches, the absolute and conditional β-convergence, do not consider the dynamics of the growth process and, using panel data, there may be a bias if there is insufficient time-series data. To solve this problem, Quah (1993b, 1996a, b, 1997) proposes a different concept known as σ-convergence which evaluates the decrease over time in the cross-section variance of per capita CO2 emissions. Thus, β-convergence is a necessary, but not a sufficient condition for σ-convergence. Based on Carlino and Mills (1993) and Bernard and Durlauf’s (1995, 1996) works, convergence is present in the country’s per capita CO2 emissions if relative per capita CO2 emissions, defined as per capita CO2 emissions for country i relative to another country (or the average of the sample of countries) is trend stationery using unit root or stationarity tests. This kind of convergence is known as stochastic convergence. More recently, Phillips and Sul (2007, 2009) developed a club convergence approach, which does not depend on a unit root or cointegration tests, examining the convergence process as a conditional σ-convergence test within a panel setting. Although there is a huge amount of work about CO2 convergence (Payne, 2020), we only will point out those which include States Members of the EU. Jobert et al. (2010) apply the Bayesian approach of Maddala et al. (1997) to test, from 1971 to 2006, both absolute and conditional β-convergence for 22 European Union countries in their per capita GHG emissions. Although they found support for absolute β-convergence, the results from conditional β-convergence vary somewhat depending on the variables included. In Herrerias (2013), per capita CO2 emissions convergence is examined from 1920 to 2007 for 25 European Union countries using the approach of Quah (1993b, 1996b). The sample is divided into two periods, preand post-Second World War, finding the strongest convergence after the 1970s.

15.4 Data and Methodology We analyze the stochastic convergence in the European Union CO2 emissions of the 27 countries: Austria (AUT), Belgium (BEL), Bulgaria (BGR), Croatia (HRV), Cyprus (CYP), Czech Republic (CZE), Denmark (DNK), Estonia (EST), Finland (FIN), France (FRA), Germany (DEU), Greece (GRC), the Netherlands (NLD), Hungary (HUN), Ireland (IRL), Italy (ITA), Latvia (LVA), Lithuania (LTU), Luxembourg (LUX), Malta (MLT), Portugal (PRT), Poland (POL), Romania (ROU), Slovak Republic (SVK), Slovenia (SVN), Spain (ESP), and Sweden (SWE). Metric tons of CO2 per capita is used to represent emissions, and the data is taken from the World Bank (WB)2 with annual frequency for the period 1990 to 2018.

2 https://data.worldbank.org/indicator/EN.ATM.CO2E.PC?end=2018&locations=EU&name_

desc=false&start=1990.

240

D. Ramirez-Carrera et al.

A visual inspection of the series (Fig. 15.2) shows significant differences in the evolution of CO2 per capita emissions between the EU countries which justifies our analysis. Stochastic convergence analysis requires variables expressed in relative terms to another country or the average of the sample of countries (see Ramírez-Carrera & Rodríguez, 2012). Relative CO2 emissions can be calculated as the difference in levels among the metric tons of CO2 per capita emissions of each country—coi, t — and the reference variable that will be the metric tons of CO2 per capita emissions of the whole European Union—eu27, t —. So, relative CO2 emissions—rcoi, t —for the country i can be calculated as: rcoi,t = coi,t − eu27,t

.

(15.1)

According to Carlino and Mills (1993), there will exist convergence if stochastic convergence is verified. This type means that shocks only have a temporary effect. Using regional US data, they find no evidence of stochastic convergence without including a break in the trend of the series. Doing that, they show that three of eight US regions display stochastic convergence, indicating that at least part of the USA is converging. The latter type means that poorer provinces are on average catching up to the national average. Finally, they add that the bulk of the US convergence took place before World War II. In related research, Loewy and Papell (1996) have extended these findings by testing for a unit root allowing for an unknown break date. They find evidence in support of stochastic convergence in seven out of eight US regions. Recently, Tomljanovich and Vogelsang (2002) contribute to this debate by expanding the findings of Carlino and Mills (1993) and Loewy and Papell (1996). Their approach consists of using the econometric tools suggested by Vogelsang (1998) and Bunzel (1998), which allow the researcher to estimate and perform inference on the parameters related to the trend function of the series. The most important fact of these econometric tools is that these statistics are robust to the presence of a unit root in the noise function of the time series. Empirically, stochastic convergence (long-term convergence) can be proven using Bernard and Durlauf (1995, 1996) approach, which considers that convergence exists between two regions—in this case between country i and the total per capita of the whole European Union—if the long-term forecasts of the CO2 emissions for both regions are equal at a fixed time t: .

  lim E coi,t+T − eu27,t+T |ζt = 0

T →∞

(15.2)

where coi,t > eu27,t (coi,t < eu27, t ) and ζ t is the information available at t. According to Oxley and Greasley (1997), this kind of convergence implies that such differences will have a transitory character when long-term forecasts

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

241

16

14

12

10

Metric tons

DEU AUT 8

BEL BGR CYP

6

HRV DNK

4

2

0

25

20

Metric tons

15

SVN ESP EST FIN GRC

10

FRA HUN

5

0

Fig. 15.2 (a–d) CO2 per capita emissions in EU countries from 1990 to 2018

242

D. Ramirez-Carrera et al. 35

30

25

Metric tons

IRL 20

ITA LVA LTU

15

LUX MLT NLD

10

5

0

16

14

12

Metric tons

10 PRT POL 8

CZE SVK ROU

6

4

2

0

Fig. 15.2 (continued)

SWE

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

243

of the difference among any couple of countries converge toward zero as the forecast horizon increases. Hence, convergence requires that differences in the CO2 emissions between two regions or countries follow a stationary process, and, therefore, they should not contain a unit root or deterministic trend throughout the time if the regions have reached the convergence, which would imply that CO2 emissions should be cointegrated. While Bernard and Durlauf (1995) define stochastic convergence as the cointegration between two (or more) time series, an alternative definition of stochastic convergence can be found in Carlino and Mills (1993). If there is no stationarity, shocks will have permanent effects on relative CO2 emissions and there will not be convergence between time series. Carlino and Mills (1993) and Evans and Karras (1996) demonstrate that stochastic convergence can be analyzed through unit root tests.

15.5 Empirical Results We have used a set of time series unit root tests to verify the existence of stochastic convergence.3 According to the results of the Augmented Dickey and Fuller (1979)—ADF—and Phillips and Perron (1988)—PP—tests, the relative CO2 emissions per capita of Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, France, Germany, the Netherlands, Latvia, Lithuania, Luxembourg, Malta, Portugal, Romania, Slovak Republic, Slovenia, and Spain reject the null hypothesis of a unit root, while the rest (Austria, Finland, Greece, Hungary, Ireland, Italy, Poland, and Sweden) could not be considered stationary, and, therefore, there will not be stochastic convergence (Table 15.1). However, it is well known4 that the failure of unit root tests can be caused by an erroneous specification of the deterministic trend. The presence of structural changes in the time series can cause a spurious non-rejection of the null hypothesis. As consequence, it is necessary to keep in mind the inclusion of, at least, one break inside the unit root tests. We have applied Zivot and Andrews (1992)—ZA— unit root tests with one structural change5 on the non-stationary relative CO2 per capita emissions (Austria, Finland, Greece, Hungary, Ireland, Italy, Poland, and Sweden). According to the results of the unit root test, Austria, Finland, Hungary, Italy, and Poland reject the null hypothesis of a unit root, and it can be considered that there exists stochastic convergence. On the contrary, there is no stochastic convergence in the case of Greece, Ireland, and Sweden (Table 15.2).

3 EViews

12 and GAUSS 6.0 have been used. Perron (1989), Campbell and Perron (1991) or Montañés and Reyes (1998). 5 In addition to this, it would be possible to apply other unit root tests with one structural change: Perron and Rodríguez (2003) —PR— and Lee and Strazicich (2013) —LS1— tests. 4 See

244

D. Ramirez-Carrera et al.

Table 15.1 Standard unit root tests Kmax = 6 ADF—Ho:I(1)— Constant None −0.981 −2.244 AUT −1.030 −0.297 BEL −0.718 −2.305 BGR −2.278b −1.142 HRV −3.566a −4.013a CYP −2.232b −3.047b CZE 0.113 −1.163 DNK EST −1.023 −2.767c −0.529 −0.818 FIN −0.902 −3.627b FRA −1.847c −1.620 DEU −1.331 −1.613 GRC HUN −0.444 −1.249 IRL −0.563 −1.936 −1.049 −1.638 ITA LVA −2.109b −1.183 LTU −0.354 −3.000b LUX −1.667c −2.452 −0.375 −2.011 MLT NLD 0.124 −3.953a −0.093 −0.732 POL −2.131b −2.719c PRT 0.2425 −1.212 ROU −2.609b −5.713a SVK SVN −2.476b −2.081 −2.208b −2.070 ESP 0.292 −0.739 SWE

Constant & Trend −1.660 −3.536c −4.055b −2.939 −3.500c −4.697a −4.990a −2.540 −1.529 −3.470c −1.677 −0.968 −1.497 −1.758 −1.294 −4.571a −5.248a −4.376a −1.950 −3.463c −1.077 −2.889 −3.751b −1.569 −1.909 −1.746 −3.161

PP—Ho:I(1)— None Constant −0.992 −2.289 −0.652 −0.781 −0.762 −2.446 −1.699c −0.525 −3.566a −3.959a −2.481b −4.223a −0.916 −0.891 −2.299b −4.142a −0.775 −2.271 −1.118 −3.627b −1.606 −2.409 −1.503 −1.773 −0.444 −1.416 −0.484 −1.713 −1.193 −1.615 −0.194 −2.223 −0.415 −2.990b −1.544 −1.594 −0.155 −1.850 0.249 −3.961a 0.052 −0.688 −2.131b −2.719c 0.143 −4.056a −2.478b −5.995a −2.494b −2.147 −2.193b −2.286 0.287 −0.861

Constant & Trend −1.601 −3.585b −4.049b −3.130 −3.476c −4.697a −4.990a −5.247a −2.523 −3.470c −1.980 −0.933 −1.497 −1.492 −1.275 −5.025 −4.962a −1.846 −1.963 −3.461c −0.779 −2.852 −3.733b −4.056b −1.886 −1.765 −3.162

ADF test: Lag length selection based on AIC; PP test: Bandwidth: Newey-West using Bartlett kernel Kmax maximum lag length a, b, c Denote the levels of significance at 1%, 5%, and 10%, respectively

Results for Greece, Ireland, and Sweden may be surprising. Nevertheless, Lumsdaine and Papell (1997) point out that even within the class of endogenous break models, such as the ZA test, results regarding unit root tests are sensitive to the number of breaks in the alternative specification. In consequence, allowing for the possibility of two endogenous breakpoints, it is possible to find more evidence against the unit-root hypothesis. Nevertheless, augmented Dickey-Fuller (ADF)type endogenous break unit root tests, such as the Zivot and Andrews (1992) and Lumsdaine and Papell (1997) tests, show a limitation because the critical values are derived while assuming no break(s) under the null. Nunes et al. (1997) and Lee and Strazicich (2003) show that this assumption leads to size distortions in the presence

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

245

Table 15.2 Unit root test with one structural change of Zivot and Andrews (1992) Kmax = 4 AUT FIN GRC HUN IRL ITA POL SWE

Model A (Constant) Tb ZA Min. −3.277 2001 2001 −3.215 2010 −3.830 2014 −3.906 −3.795 1997 1997 −3.002 1998 −5.621a 2000 −4.023

Model B (Trend) ZA Min. Tb −4.139c 2005 −3.544 2007 −3.714 2007 −5.908a 2014 −3.470 2005 −4.477b 2005 −4.215c 2003 −3.422 2008

Model C (Constant & Trend) ZA Min. Tb −4.949c 2003 −5.044c 2003 −3.751 2010 −5.686a 2013 −3.541 2004 −4.287 2004 −5.136b 1998 −3.629 2010

Kmax, Maximum lag length, Lag length selection criteria: General to Specific de Ng and Perron (1995) a, b, c Denote the levels of significance at 1%, 5%, and 10%, respectively

of a unit root with one or two breaks which can lead to spurious rejections may occur, rejecting too often the null hypothesis of a unit root, as well as they tend to estimate breakpoints imprecisely. Lee and Strazicich (2003)—hereafter LS— proposed an alternative unit root test with two endogenous breaks that are unaffected by structural breaks under the null. Two structural breaks can be considered as follows: Model A allows for two shifts in level and Model C includes two changes in level and trend. Breakpoints (Tb ) are determined endogenously using the minimum LM unit root test. According to Vougas (2003), the studentized version (τ ) of the LM test is more powerful and takes, into account, the variability of the estimated coefficients. The breakpoints are determined to be where the test statistic is minimized. Critical values are tabulated in Lee and Strazicich (2003), where λ = Tb /T, and T is the sample size. As expected in the endogenous break test, a trimming region of (0.1T, 0.9T) is used to eliminate endpoints. Therefore, we proceed to apply the LS unit root test with two structural changes to Greece, Ireland, and Sweden. In all countries, it is possible to reject the null hypothesis of a unit root which can be interpreted as proof of the existence of a stochastic convergence process (Table 15.3). These results are consistent with previous studies focused on the EU. However, in the case of Sweden, it is only possible to reject the unit root hypothesis at a 10% significance level, so the unit root hypothesis is not very sensitive to the introduction of a structural change. Such a weak rejection can be due to idiosyncratic factors in Sweden (Ramírez-Carrera et al., 2022). This country has a high gross national product (GNP) per person, a large industry, long transport distances, and cold winters. However, while these factors are generally associated with high greenhouse gas emissions, Sweden’s domestic greenhouse gas emissions are relatively low. We can consider two key factors in this situation. First, Swedish electricity production is largely based on hydroelectric power and nuclear power, but both the installed capacity and annual electricity production from wind power are increasing steadily, as the use of bioenergy in combined heat and

246

D. Ramirez-Carrera et al.

Table 15.3 Unit root test with two structural changes of Lee and Strazicich (2003) Kmax = 4 Model A (Constant) Tb2 LMτ k Tb1 3 2006 2008 −1.821 GRC 2 2004 2014 −1.987 IRL 4 1998 2005 −2.387 SWE

Model C (Constant & Trend) k Tb1 Tb2 LMτ λ used for critical values 3 2005 2011 −7.095a λ = (0.6, 0.8) 4 1997 2007 −6.031b λ = (0.2, 0.6) 1 1996 2007 −5.320c λ = (0.2, 0.6)

Kmax maximum lag length the levels of significance at 1%, 5%, and 10%, respectively

a, b, c Denote

power plants. Second, the high extension of Swedish forests and land remove CO2 , and the total removals remain at a high level. In consequence, emissions have been moving away from the EU average (Ministry of the Environment, 2020).

15.6 Evaluating European Climate Policy Impact on the Convergence Process To determine which European climate policy measures have affected the convergence process in each of the countries that are part of the UE, the Bai-Perron (BP) procedures have been used. The methodology developed by Bai and Perron (1998) aims to test multiple structural breaks in time series. Some of the procedures developed by Bai and Perron are not valid when trending regressors are allowed; therefore, it is possible to apply such procedures to this work because, in all countries, relative CO2 per capita emissions do not show a trending behavior as we showed in the previous section. Using Bai and Perron methodology, we can identify the dates when took place significant changes in the evolution of the relative CO2 per capita emissions. We assume that these break dates are caused by the climate policy measures implemented by the EU that have managed to affect the convergence process analyzed above. The BP procedure is based on the principle of global minimizers of the sum squared of the residuals (SSR) and can determine consistently the number of structural changes. Given a maximum number of breakpoints (mmax ), it consists in estimating their position for each mi ≤ mmax , i = 1,..., N, testing for the significance of the breaks. Once the dates have been estimated, the point is to select a suitable number of structural breaks. For that purpose, Bai and Perron (1998) use two different procedures. In short, the first procedure relies on the use of information criteria—the Bayesian information criterion (BIC) and the modified Schwarz information criterion (LWZ) of Liu et al. (1997). The second one, called sequential procedure (SP), applies pseudo-F-type test statistics on the sequential computation and detection of structural breaks. Following Esteve García et al. (2017), first, we determine if at least one breakpoint is present within the time series through the UDmax and WDmax tests.

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

247

Table 15.4 Results of the Bai and Perron tests WDmax (1%) Country UDmax AUT 137.405a 249.333a 56.116a 66.183a BEL 22.254a 39.678a BGR 15.585a 15.585a CYP 36.681a 49.940a CZE 24.584a 44.611a DEU 53.177a 72.399a DNK 62.324a 94.727a ESP EST 10.705b 14.575a 16.506a 25.839a FIN 10.468b 18.996a FRA 38.934a 70.650a GRC 146.367a 265.597a HRV HUN 15.175a 20.661a a, b, c Denote

BIC 2 2 2 1 2 3 2 2 2 2 3 3 3 2

LWZ 2 2 2 1 2 3 2 2 0 1 0 3 3 1

Country IRL ITA LTU LUX LVA MLT NLD POL PRT ROU SVK SNV SWE

UDmax 59.842a 30.496a 13.339a 64.895a 31.668a 13.715a 6.534 45.956a 43.170a 14.309a 37.546a 67.180a 32.590a

WDmax (1%) 81.472a 48.394a 18.356a 96.110a 50.408a 18.673a 11.330c 63.665a 52.810a 21.455a 53.510a 91.463a 45.236a

BIC 2 3 2 3 2 2 1 3 1 2 2 2 2

LWZ 2 2 2 3 2 1 1 2 1 2 2 2 1

the levels of significance at 1%, 2.5%, and 5%, respectively

Second, we use the information criteria to choose the number of breaks, and dates are selected according to the LWZ criterion because Bai and Perron (2003) conclude that it performs better than the BIC. In addition to this, following the practical recommendations of Bai and Perron (2003, 2006), it has been allowed the existence of correlation in the errors and different variances in the residuals and the data between segments or subsamples because these specifications improve the power of the tests and the precision in the selection of the number of breaks. A trimming ε = 0.20 has been used to allow a maximum of 3 breaks because of the short period of the sample. Our estimations show that, in all countries there exists, at least, structural change in the process of convergence (Table 15.4). In those cases, in which the LWZ information criterion is contrary to the results of the UDmax and WDmax tests, we have used the BIC criterion to select the dates. Now, the estimated dates of the change in the convergence process (Table 15.5) are compared with the main facts of the European climate policy. EU climate policy instruments and measures from 1990 to 2018 have been structured in stages, although the stages can overlap each other depending on instruments, strategies, and/or objectives: • Pre-Kyoto Stage (1990–1997): Although emissions were declining at the beginning of the 1990s, this was mainly due to effects resulting from mainly national policies and we can consider that there was not a real common European climate policy. • Stage I (1998–2007): Except for voluntary agreements with car producers on emissions reductions in 1998 and the Landfill Directive to reduce methane in 1999, there was no significant progress regarding European climate policies in the late 1990s. However, in the year 2000, the European Climate Change

AUT CI 95% CI 90% BEL CI 95% CI 90% BGR CI 95% CI 90% CYP CI 95% CI 90% CZE CI 95% CI 90% DEU CI 95% CI 90% DNK CI 95% CI 90%

2000 (1999, 2001) (1999, 2001) 2010 (2009, 2012) (2009, 2012) 2006 (2004, 2008) (2005, 2007)

1994 (1993, 1995) (1993, 1995) 2004 (2003, 2008) (2003, 2007) 1997 (1993, 2003) (1994, 2002) 1997 (1994, 2006) (1995, 2004) 1994 (1993, 1999) (1993, 1997) 1996 (1995, 2000) (1995, 1999) 1999 (1998, 2004) (1998, 2003)

2010 (2009, 2014) (2009, 2013) 2002 (2000, 2005) (2001, 2004) 2010 (2008, 2011) (2009, 2011)

.TˆB2

.TˆB1

2010 (2007, 2011) (2008, 2011)

.TˆB3

Table 15.5 Dating of breakpoints and confidence intervals HUN CI 95% CI 90% IRL CI 95% CI 90% ITA CI 95% CI 90% LTU CI 95% CI 90% LUX CI 95% CI 90% LVA CI 95% CI 90% MLT CI 95% CI 90%

1994 (1992, 2002) (1992, 2000) 1997 (1996, 1998) (1996, 1998) 1997 (1995, 1999) (1996, 1999) 1994 (1994, 2006) (1994, 2002) 1994 (1993, 1995) (1993, 1995) 1994 (1994, 1997) (1994, 1996) 2006 (2005, 2012) (2005, 2011)

.TˆB1

2013 (2009, 2015) (2010, 2015) 2008 (2007, 2009) (2007, 2009) 2011 (2010, 2013) (2010, 2012) 2008 (2007, 2009) (2007, 2009) 2001 (1997, 2003) (1998, 2003) 2008 (2008, 2008) (2008, 2008) 2013 (2004, 2017) (2006, 2016)

.TˆB2

2012 (2010, 2014) (2011, 2013)

.TˆB3

248 D. Ramirez-Carrera et al.

1998 (1997, 2000) (1997, 1999) 1994 (1993, 2005) (1994, 2002) 2000 (1987, 2009) (1990, 2006) 1997 (1996, 2002) (1996, 2001) 1994 (1992, 1998) (1993, 1997) 1996 (1994, 1998) (1995, 1997)

CI confidence interval

ESP CI 95% CI 90% EST CI 95% CI 90% FIN CI 95% CI 90% FRA CI 95% CI 90% GRC CI 95% CI 90% HRV CI 95% CI 90%

2008 (2007, 2012) (2007, 2011) 2006 (2004, 2007) (2004, 2007) 2011 (2010, 2016) (2010, 2015) 2002 (2001, 2003) (2001, 2003) 1999 (1997, 2000) (1998, 2000) 2001 (2000, 2003) (2000, 2003) 2008 (2006, 2010) (2006, 2009) 2011 (2010, 2013) (2010, 2012) 2006 (2003, 2007) (2004, 2007)

NLD CI 95% CI 90% POL CI 95% CI 90% PRT CI 95% CI 90% ROU CI 95% CI 90% SVK CI 95% CI 90% SNV CI 95% CI 90% SWE CI 95% CI 90%

1994 (1989, 2007) (1991, 2004) 1997 (1996, 1998) (1996, 1998) 1997 (1996, 2001) (1996, 2000) 1997 (1996, 2002) (1996, 2001) 1994 (1994, 13) (1994, 11) 1994 (1993, 1996) (1993, 1995) 2003 (2002, 2008) (2002, 2007) 2006 (2003, 2009) (2004, 2008) 1999 (1995, 2000) (1996, 2000) 2005 (2001, 2007) (2002, 2006)

2008 (2007, 2009) (2007, 2009)

2013 (2010, 2015) (2010, 2015)

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . . 249

250

D. Ramirez-Carrera et al.

Programme (ECCP) was launched. It was part of the Lisbon and Sustainable Development Strategy. It examined a wide range of policy sectors and instruments to reduce GHG emissions and developed common and coordinated strategies to fulfil the Kyoto targets. • Stage II (2008–2015): In March 2007, EU Heads of State agreed on a set of three targets referred to as 20-20-20 by 2020. This agreement is called the Strategy Europe 2020 or the 3x20 policy on GHG emissions, renewable energies, and energy efficiency. To achieve the new targets, the European Commission introduced the Climate and Energy Package (CEP) in 2008 which was implemented in 2010. • Stage III (2016–2019): In 2016, the European Commission published a major set of legislative proposals and new policy initiatives on climate and energy, known as the “Winter Package,” to put in place the policies and measures that the EU needs to deliver on its climate and energy targets for 2030. It was part of the Strategy for an Energy Union in 2015 and includes its climate mitigation target, presented as its contribution to last year’s Paris Agreement to reduce its GHG emissions by −40% by 2030 and −80% by 2050. According to these stages, the convergence process behavior suffered a change in 21 of the UE countries before Stage I, except in Belgium, Denmark, Spain, Finland, Malta, and Sweden, which can mean that national policies had already been implemented affecting significantly to the process in those countries. Regarding the effect of the common climate policies, the ECCP would have had a slightly bigger impact on the convergence process than the CEP, since a greater number of countries have seen how such a process would have changed in its evolution in the first stage versus the second. Thus, the ECCP would have affected 16 countries (Austria, Belgium, Bulgaria, Croatia, Denmark, Finland, France, Germany, Greece, Latvia, Luxembourg, Malta, Romania, Slovak Republic, Slovenia, Spain, and Sweden) while the CEP would have an influence on 15 countries (Belgium, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Portugal, Poland, and Spain). Since the sample finishes in 2018, we have not seen any change caused by the Winter Package because it would be necessary to wider time span to detect an additional break, in case there were. It is remarkable to note that the convergence process has not suffered any change in its evolution because of the European climate policy in three countries (Cyprus, the Netherlands, and Portugal). Finally, to measure the impact of European climate policy on the convergence process, we calculated the average relative emission for every subperiods into which the sample is divided based on the points of structural change estimated with the Bai procedure and Perron in each EU country, and following, the variation, pre- and post breakpoint, of the average relative emissions is calculated. In cases where the variation has a negative sign, it will be considered that the difference between the CO2 emissions per capita of the country—coi, t — and the European mean—eu27, t —has been reduced, that is, the average relative emission is lower, and, therefore, the European climate policy has made an effective

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

251

Table 15.6 Impact of pre-Kyoto stage on the convergence process Pre-Kyoto period (1990–1997) Average relative emission .TˆB AUT Variation BEL Variation BGR Variation CYP Variation CZE Variation DEU Variation DNK Variation ESP Variation EST Variation FIN Variation FRA Variation GRC Variation HRV Variation

−0.62

1994 −86.5%

Average relative emission −0.08

No break −1.29 −1.38 5.18 3.09

1997 59.4% 1997 −67.2% 1994 −25.1% 1996 −19.4%

−2.06 −0.45 3.88 2.49

No break No break Variation 9.35

1994 −60.4%

28.6% 3.70

No break −2.01 −0.76 −4.67

1997 −12.6% 1994 −81.1% 1996 −12.5%

−1.76 −0.14 −4.09

HUN Variation IRL Variation ITA Variation LTU Variation LUX Variation LVA Variation MLT Variation NLD POL Variation PRT Variation ROU Variation SVK Variation SNV Variation SWE Variation

Average relative emission −2.21 0.97 −0.99 −1.28 21.47 −2.33

.TˆB

1994 9.6% 1997 202.2% 1997 −80.6% 1994 233.9% 1994 −40.5% 1994 94.0%

Average relative emission −2.42 2.94 −0.19 −4.29 12.78 −4.51

No break 1.96

1994

2.52

0.91

1997 −94.6% 1997 −40.8% 1997 51.7% 1994 −81.3% 1994 −76.3%

0.05

−3.49 −2.36 0.90 −1.60

−2.07 −3.59 −0.17 −0.38

No break

contribution to the convergence process. On the contrary, if the variation has a positive sign, the difference between the country’s emissions—coi, t —and the European Union average—eu27, t —has increased, so that, in this case, the European climate policy would not have had the desired effects on the process of convergence. According to the variation of the average relative CO2 per capita emissions, pre, and post breakpoints, pre-Kyoto climate measures had a positive effect on the convergence process in 13 countries (Austria, Cyprus, Czech Republic, Germany, Estonia, France, Croatia, Italy, Luxembourg, Poland, Portugal, Slovak Republic, and Slovenia) versus seven countries (Bulgaria, Hungary, Ireland, Lithuania, Latvia, Romania, and the Netherlands) in which the average relative emission increased (Table 15.6).

252

D. Ramirez-Carrera et al.

Table 15.7 Stage I impact on the convergence process European common climate policy (1998–2007) Average Average relative relative emissions emissions .TˆB −0.08 2000 0.88 AUT 950.6% Variation BEL 2004 2.35 3.27 −28.2% Variation BGR −2.06 2006 −0.83 −59.9% Variation CYP No break Variation CZE No break Variation 2002 1.89 DEU 2.49 −24.2% Variation DNK 1999 2.02 3.65 −44.8% Variation 1998 −0.54 −2.28 ESP Variation −76.4% EST 2006 5.92 3.70 Variation 59.9% 3.11 2000 3.96 FIN Variation 27.1% FRA 2002 −2.07 −1.76 Variation 17.9% GRC 1999 1.01 −0.14 604.5% Variation −4.09 HRV 2001 −3.16 Variation −22.7% 2006 −2.57 −3.16 HRV −18.9% Variation

HUN Variation IRL Variation ITA Variation LTU Variation LUX Variation LVA Variation MLT Variation NLD Variation POL Variation PRT Variation ROU Variation SVK Variation SNV Variation SWE Variation

Average relative emissions No break

.TˆB

Average relative emissions

No break No break No break 12.78

2001 18.2%

15.10

2006 −39.9%

−1.02

2006 −16.3% 1999 351.4% 2005 −20.1% 2003 55.6%

−3.00

No break −1.70 No break No break No break −3.59 −0.17 −0.38 −1.59

−0.76 0.30 −2.48

Once the common European climate policy begins, the actions introduced in Stage I had fewer clear effects on the overall convergence process since, the same number of countries moved away from the average emissions of the EU as those that came closer. Thus, nine countries (Belgium, Bulgaria, Germany, Denmark, Spain, Croatia, and Malta) reduced their differences with the EU mean, but eight countries (Austria, Estonia, Finland, France, Greece, Luxembourg, Slovak Republic, and Slovenia) experienced a variation with a positive sign (Table 15.7). However, the common climate policy’ Stage II clearly promoted the convergence process, reducing relative emissions in 11 countries (Belgium, Czech Republic, Denmark, Finland, France, Greece, Hungary, Ireland, Lithuania, Luxembourg, and

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

253

Table 15.8 Stage II impact on the convergence process European common climate policy (2008–2015) Average Average relative relative emissions emissions .TˆB No break AUT Variation 2010 1.73 2.35 BEL −26.3% Variation No break BGR Variation No break CYP Variation 2010 3.24 CZE 3.88 −16.5% Variation 2010 2.45 DEU 1.89 29.8% Variation 2010 −0.19 2.02 DNK −90.7% Variation 2008 −1.10 −0.54 ESP Variation 104.1% EST No break Variation FIN 3.96 2011 1.98 Variation −49.9% FRA 2008 −1.83 −2.07 Variation −11.8% GRC 2011 −0.15 1.01 −85.0% Variation HRV No break Variation 2013 −2.08 −2.42 HUN −14.0% Variation

IRL Variation ITA Variation LTU Variation LUX Variation LVA Variation MLT Variation NLD Variation POL Variation POL Variation PRT Variation ROU Variation SVK Variation SNV Variation SWE Variation

Average relative emissions 2.94 −0.19 −4.29 15.10 −4.51 −1.02

.TˆB

2008 −55.0% 2011 363.4% 2008 −31.9% 2012 −32.1% 2008 −31.3% 2013 145.4%

Average relative emissions 1.32 −0.89 −2.92 10.25 −3.10 −2.51

No break 0.05 0.84 −3.49

2008 1601.8% 2013 62.4% 1997 −40.8%

0.84 1.36 −2.07

No break No break No break No break

Latvia), while only in five countries (Germany, Spain, Italy, Malta, and Poland) the average relative emission grew up, which can be interpreted as a success of the measures adopted by the EU during this period to increase the convergence (Table 15.8). Despite these quite different results from one stage to another, it can be said that the common European climate policy has contributed positively to the convergence process of the CO2 per capita emissions among the EU countries, and it could be interpreted that the measures implemented by the EU to reduce CO2 emissions are increasingly successful.

254

D. Ramirez-Carrera et al.

15.7 Conclusion Applying a wide pool of unit root test, all countries analyzed reject the existence of a unit root in their relative CO2 per capita emissions. This can be interpreted as a proof of the existence of convergence in all the EU countries and that the climate policy of the European Union has been effective, thereby showing that the implementation over time of environmental and energy policies has been effective, even in the case of Sweden, which shows a distance from the EU average due to the faster reduction in its per capita CO2 emissions. Those responsible for European climate policy should analyze such differences with the aim of implementing these elements as a way of reducing GHG emissions, especially in those member countries that still show levels of per capita emissions above the European average. Comparing the estimated dates of the change in the convergence process, obtained through the Bai and Perron tests, with the timeline of the European climate policy, we can affirm that the process has been affected by the measures taken by the UE to reduce the GHG in most of the countries except three countries. Previously to the Kyoto protocol, national policies had already been implemented significantly affecting the convergence process. Although the two common policy stages in climate policy affect a similar number of countries, the importance of Stage II in the convergence of the CO2 per capita emission is much bigger, although it has not had the same impact in all countries. We consider that these differences require further investigation being necessary to analyze the drivers of climate change in every country to explain the reasons for the lopsided impact of the climate policy in the EU countries.

References Acaravci, A., & Erdogan, S. (2016). The convergence behavior of CO2 emissions in seven regions under multiple structural breaks. International Journal of Energy Economics and Policy, 6(3), 575–567. Bai, J., & Perron, P. (1998). Estimating and testing linear models with multiple structural changes. Econometrica, 66(1), 47–78. Bai, J., & Perron, P. (2003). Computation and analysis of multiple structural change models. Journal of Applied Econometrics, 18, 1–22. Bai, J., & Perron, P. (2006). Multiple structural change models: A simulation analysis. In D. Corbea, S. Durlauf, & B. E. Hansen (Eds.), Econometric theory and practice: Frontiers of analysis and applied research (pp. 212–237). Cambridge University Press. Barro, R., & Sala-i-Martin, X. (1992). Convergence. Journal of Political Economy, 100(2), 223– 251. Benjamin, C., Cadoret, I., & Hubert, M. (2015). The European climate policy is ambitious: Myth or reality? Revue D’économie Politique, 125(5), 731–753. Bernard, A., & Durlauf, S. (1995). Convergence in international output. Journal of Applied Econometrics, 10(2), 97–108. Bernard, A., & Durlauf, S. (1996). Interpreting tests of the convergence hypothesis. Journal of Econometrics, 71, 161–173.

15 Evaluating European Climate Policy Impact on the CO2 Emissions Per. . .

255

Bunzel. (1998). Robust inference in models of cointegration. Center of Analytical Economics. Working paper no. 98-13, Cornell University. Campbell, J. Y., & Perron, P. (1991). Pitfalls and opportunities: What macroeconomics should know about unit roots. National Bureau of Economic Research (NBER). Technical working paper no. 100, from http://www.nber.org/papers/t0100 Carlino, G. A., & Mills, L. O. (1993). Are US regional incomes converging?: A time series analysis. Journal of Monetary Economics, 32(2), 335–346. Delbeke, J., & Vis, P. (2015). EU climate policy explained. Routledge. Dickey, D., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74(366a), 427–431. Esteve García, V., Navarro Ibáñez, M., & Prats Albentosa, M. A. (2017). The present value model of US stock prices revisited: Long-run evidence with structural breaks, 1871–2012. Economics discussion papers no. 2017-93, from http://www.economics-ejournal.org/economics/ discussionpapers/2017-93/ Evans, P. (1996). Using cross-country variances to evaluate growth theories. Journal of Economic Dynamics and Control, 20(6–7), 1027–1049. Evans, P., & Karras, G. (1996). Convergence revisited. Journal of Monetary Economics, 37, 245– 248. Herrerias, M. J. (2013). The environmental convergence hypothesis: Carbon dioxide emissions according to the source of energy. Energy Policy, 61, 1140–1150. Jobert, T., Karanfil, F., & Tykhonenko, A. (2010). Convergence of per capita carbon dioxide emissions in the EU: Legend or reality? Energy Economics, 32(6), 1364–1373. Lee, J., & Strazicich, M. C. (2003). Minimum Lagrange multiplier unit root test with two structural breaks. The Review of Economics and Statistics, 85(4), 1082–1089. Lee, J., & Strazicich, M. C. (2013). Minimum LM unit root test with one structural break. Economics Bulletin, 33(4), 2483–2492. Liu, J., Wu, S., & Zidek, J. V. (1997). On segmented multivariate regressions. Statistics Sinica, 7, 497–525. Loewy, M. B., & Papell, D. H. (1996). Are US regional incomes converging? Some further evidence. Journal of Monetary Economics, 38(3), 587–598. Lumsdaine, R. L., & Papell, D. H. (1997). Multiple trends breaks and the unit-root hypothesis. The Review of Economics and Statistics, 79(2), 212–218. Maddala, G. S., Trost, R. P., Li, H., & Joutz, F. (1997). Estimation of short-run and long-run elasticities of energy demand from panel data using shrinkage estimators. Journal of Business and Economic Statistics, 15(1), 90–100. Ministry of the Environment, Sweden. (2020). Sweden’s long-term strategy for reducing greenhouse gas emissions. United Nations Framework Convention on Climate Change (UNFCCC), from https://unfccc.int/documents/267243 Montañés, A., & Reyes, M. (1998). The asymptotic behaviour of the Dickey-Fuller unit root tests under a shift in the trend function. Econometric Theory, 14, 355–363. Ng, S., & Perron, P. (1995). Unit Root Tests in ARMA models with data-dependent methods for the selection of the truncation lag. Journal of the American Statistical Association, 90, 268–281. Nunes, L. C., Newbold, P., & Kuan, C. M. (1997). Testing for unit roots with breaks: Evidence on the great crash and the unit root hypothesis reconsidered. Oxford Bulletin of Economics and Statistics, 59(4), 435–448. Oxley, L., & Greasley, D. (1997). Convergence in GDP per capita and real wages: Some results for Australia and the UK. Mathematics and Computers in Simulation, 43, 429–436. Payne, J. E. (2020). The convergence of carbon dioxide emissions: A survey of the empirical literature. Journal of Economic Studies, 47(7), 1757–1785. Perron, P. (1989). The great crash, the oil-price shock, and the unit-root hypothesis. Econometrica, 57, 1361–1401. Perron, P., & Rodríguez, G. (2003). GLS detrending, efficient unit root tests and structural change. Journal of Econometrics, 115, 1–27.

256

D. Ramirez-Carrera et al.

Pettersson, F., Maddison, D., Acar, S., & Soderholm, P. (2014). Convergence of carbon dioxide emissions: A review of the literature. International Review of Environmental and Resource Economics, 7(2), 141–178. Phillips, P. C. B., & Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75, 335–346. Phillips, P. C. B., & Sul, D. (2007). Transition modeling and econometric convergence tests. Econometrica, 75(6), 1771–1855. Phillips, P. C. B., & Sul, D. (2009). Economic transition and growth. Journal of Applied Econometrics, 24(7), 1153–1185. Quah, D. (1993a). Galton’s fallacy and the convergence hypothesis. The Scandinavian Journal of Economics, 95(4), 427–443. Quah, D. (1993b). Empirical cross-section dynamics in economic growth. European Economic Review, 37(2–3), 426–434. Quah, D. (1996a). Twin peaks: Growth and convergence in models of distribution dynamics. Economic Journal, 106(437), 1045–1055. Quah, D. (1996b). Empirics for economic growth and convergence. European Economic Review, 40(6), 1353–1375. Quah, D. (1997). Empirics for growth and distribution: Stratification, polarization and convergence clubs. Journal of Economic Growth, 2(1), 27–59. Ramírez-Carrera, D., & Rodríguez, G. (2012). Have European unemployment rates converged? Zeszyty Naukowe, 10, 135–156. Ramírez-Carrera, D., Durán-Romero, G., & de la Peña, J. A. N. (2022). European climate policy effectiveness: Convergence analysis for per capita CO2 emissions. International Advances in Economic Research. https://doi.org/10.1007/s11294-022-09845-8 Strazicich, M. C., & List, J. A. (2003). Are CO2 emission levels converging among industrial countries? Environmental and Resource Economics, 24, 263–271. Tomljanovich, M., & Vogelsang, T. J. (2002). Are U.S. regions converging? Using new econometric methods to examine old issues. Empirical Economics, 27(1), 49–62. Vogelsang, T. J. (1998). Trend function hypothesis testing in the presence of serial correlations. Econometrica, 66(1), 123–148. Vougas, D. V. (2003). Reconsidering LM unit root testing. Journal of Applied Statistics, 30(7), 727–741. Zivot, E., & Andrews, W. K. (1992). Further evidence on the great crash, the oil-price shock, and the unit-root hypothesis. Journal of Business and Economic Statistics, 10(3), 251–270.

Chapter 16

The Economics of Civil Orders and Medals in Spain: An Update After Ten Years Thomas Baumert and Beatriz Luceño-Ramos

Abstract The present paper picks up on previous research presented by one of the authors ten years ago at the ICOAE 2012 in Uppsala (Sweden), in which the bestowing of civil orders and medals by different Spanish governments—depending on their political orientation—was analysed using an economics approach. For this purpose, it equated the former to monetary series and establishing an inverse relationship between the number of awards granted and their “value” in terms of honour and merit. The decade since elapsed seems a period long enough to look back and check whether the conclusions reached in 2012 have stood the test of time or if, on the contrary, they do not proof right any longer. Our results show that while the differences in the bestowing between government of different orientations have nearly vanished, the important gender gap still persists. Keywords Premial law · Economic analysis · Orders and medals · Merit

Imagine, then, that political merit is a kind of game in which you are appointed to direct; and consider, that if you grant the prizes to a few, and those the most worthy, and on such conditions as the laws prescribe, you will have many champions in this contest of merit. However, if you gratify any man that pleases, or those who can secure the strongest interest, you will be the means of corrupting the very best natural dispositions. Aeschines, Against Ctesifonte

T. Baumert () Universidad Antonio de Nebrija, Madrid, Spain e-mail: [email protected] B. Luceño-Ramos Universidad Camilo José Cela, Madrid, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 N. Tsounis, A. Vlachvei (eds.), Advances in Empirical Economic Research, Springer Proceedings in Business and Economics, https://doi.org/10.1007/978-3-031-22749-3_16

257

258

T. Baumert and B. Luceño-Ramos

16.1 Introduction This chapter picks up on a previous research presented by one of the authors 10 years ago at the ICOAE 2012 in Uppsala (Sweden), entitled “On the Devaluation of Merit” (Baumert & Roldán, 2012) in which the bestowing of civil orders and medals by different Spanish governments – depending on their political orientation – was analysed using an economics approach. For this purpose, it equated the former to monetary series and assumed an inverse relationship between the number of awards granted and their “value” in terms of the underlying honour and merit. The decade since elapsed seems a period long enough to look back and check whether the conclusions reached in 2012 have stood the test of time or if, on the contrary, they do not prove right any longer. To start with, it should be mentioned that the original 2012 paper was much longer, as the economic analysis of Premial Law was then a complete novelty, thus requiring an extensive presentation aimed to set and explain the adequate theoretical frame in which the study was embedded. In this sense, important advances have been made since then, not only in terms of scholarly research (Frey & Gallus, 2017; García-Mercadal, 2019, Baumert & Valbuena, 2020, 2022, to cite just some of the more recent works) but even of academic recognition, as the Spanish Royal Academy of Jurisprudence and Legislation created in 2014 a specific Premial Law’s section. Special interest has been put on the strong gender bias in the bestowing, a fact first pointed out by Baumert and Roldán (2012) and which has been recently confirmed in a much broader study (Baumert & Valbuena, 2020). Also, the Spanish government finally started to compilate and publish – although irregularly – the number of bestowing (more on this later on). Hence, the aim of this paper will be (again) to check whether there are statistically significant differences in the bestowing of civil orders in Spain depending on the political orientation of the governing party, although now employing longer time series.

16.2 Theoretical Background “Yet another method of preventing crimes is, to reward virtue” (Beccaria, 1764/1991, p. 83). This sentence from the last part of Beccaria’s crucial work Dei delitti e delle pene sums up the essence of what was to become Premial – also Praemial, or Laudative – Law, (and, from the economic point of view, the art of public incentives): the awarding of honours – including orders, decorations and medals – as a form of publicly recognising merit through exteriorisation, thus acting both as a reward and as a stimulus and incentive for future excellence and virtuosity (cf. Fehr & Falk, 2002).

16 The Economics of Civil Orders and Medals in Spain: An Update After Ten Years

259

Since the times of Ancient Egypt1 it is custom that the state (or the sovereign) distinguishes those citizens and subjects who have achieved outstanding merits – either military or civil – with orders, medals, titles and honours. The underlying idea is to point out these persons from the rest allowing them to exteriorise their merit through the right of wearing an object that is widely recognised as a distinction.2 Up to our days, medals and orders are widespread through nearly all nations,3 independently from their form of government, size or cultural tradition:4 from republics like France and the United States,5 to monarchies like Great Britain and Spain. From dictatorships – both far right like Nazi Germany and socialist like the former GDR and the USSR – to democracies like Ancient Athens. Even the Vatican State has an important number of awards in form of orders and decorations.6 Thus, it might not wonder that the German philosopher Arthur Schopenhauer, who generally was very pessimistic about human nature and institutions, had a quite positive attitude towards decorations. Regarding the Prussian order Pour le Mérite (established in 1740 by Frederic the Great),7 he wrote – using a comparison taken out of economics – in his Parerga and Paralipomena: Orders, it may be said, are bills of exchange drawn on public opinion, and the measure of their value is the credit of the drawer. Of course, as a substitute for pensions,8 they save the State a good deal of money; and, besides, they serve a very useful purpose, if they are distributed with discrimination and judgment. For people in general have eyes and

1 Pharaohs

rewarded both their brave soldiers and their outstanding civil servants with a collar decorated with anthropomorph pendants. 2 This occurs directly with orders and medals which are worn on the uniform or gala-dress (or as miniature or small ribbon on a civil-dress), but also indirectly which other awards, such as honors. For example, it is common to see members from nobility wear a signet ring with the family crest, another object that is widely recognized as a distinction. 3 One exception might be underlined: the Swiss Confederation does not bestow any orders at all (although their military personnel is allowed to accept and exhibit foreign awards). 4 For an overview of orders and decorations and their changes though history see, among others, Hieronymussen et al. (1966), Gritzner (1893), Klietmann and Neubecker (1984), Honig (1986), Perrot (1988), Damien (1991), and Bander van Duren (1995). An —although incomplete— recompilation of orders by nations can be found in Frey (2005), pp. 39–68 (Appendix). 5 Unlike what is commonly believed, the United States have bestowed a quite important number of civil awards, among others the Presidential Medal of Freedom, the Presidential Citizens Award, the Congressional Gold (and Silver) Medal, etc. 6 Among others, the Vatican State awards the Order of Christ, the Order of the Golden Spur, the Order of Pius, the Order of St. Gregory the Great and the Order of St. Sylvester. 7 The order wears the French inscription Pour le Mérite, as this was the official court-language in Prussia during the reign of Frederic the Great. Although originally established as a strictly military order, in 1842 a parallel civil version of the order was bestowed. It seems that Schopenhauer refers to the latter. The last Knight of the military Pour le Mérite, the illustrious writer Ernst Jünger, died in 1998 aged 102. 8 Thus, for example, the founding for the Kaiser Wilhelm Gesellschaft —the forerunner of the current Max-Planck Society— was achieved by the Emperor Wilhelm’s initiative of awarding the Grand Cross of the Prussian Black Eagle Order to all those businessmen who donated more than 100,000 marks to its financing.

260

T. Baumert and B. Luceño-Ramos

ears, it is true; but not much else, very little judgment indeed, or even memory. There are many services of the State quite beyond the range of their understanding; others, again, are appreciated and made much of for a time, and then soon forgotten. It seems to me, therefore, very proper, that a cross or a star should proclaim to the mass of people always and everywhere, “This man is not like you; he has done something”. [ . . . Though] it is a pleonasm to inscribe on any order pour le mérite; for every order ought to be for merit, —ça va sans dire—. (Schopenhauer, 1851/1909, vol. IV, pp. 425–426)

Schopenhauer’s concluding ça va sans dire shows that what seemed clear to him – that an order only made sense if it rewarded real merit – was not necessarily a commonplace, as in the mid-nineteenth century social status and inherited privileges could still be more relevant than merits. As Ceballos-Escalera and García-Mercadal (2003, p. 26) point out, “if we set aside the achievements inspired in this matter [orders] by Bonapartism, many times awarding did not aim to publicly recognising relevant merits and outstanding services, but satisfying the futility of the oligarchy and political clientele of the moment”. Thus the philosopher’s final comment should be understood as an ironical wink (Fuhrmann, 1992, p. 8). But Schopenhauer also draws his attention to this risk of “inflation” of orders, by stating that: [ . . . ] Orders lose their value when they are distributed unjustly, or without due selection, or in too great numbers: a prince should be as careful in conferring them as a man of business is in signing a bill. (Schopenhauer, 1851/1909, vol. IV, p. 426)

It is worth noticing, that the monetary comparison of orders and awards was not only used by Schopenhauer – but also by the aforementioned Beccaria, when affirming that: The coin of honour is always inexhaustible and fruitful in the hands of a wise distributor. (Beccaria, 1764/1991, p. 83)

And a similar concept is to be found in Jiménez de Asúa’s statement regarding this fact: But today honorary awarding suffer the worst discredit. Disdain, this big destructor of prestige, is ruining the value of distinctions. And we must confess that the reasons for this are true. Today decorations are no longer a reminder of a virtuous or heroic action [ . . . ] but something that decorates with its glance, which completes a gala suit. Uniforms, with their magnificence and their splendour, don’t represent for many the distinctive of a [governmental or military] body, but simply an ornament, in which orders, collars and medals are just another decorative element. Many times, their function has nothing to do with rewarding a merit: when a Sovereign visits a foreign nation, the Head of the visited country distributes orders among the whole entourage, following the ancient tradition of giving presents as a font a friendship, but without any laudatory significance at all. The abuse and denaturalisation of orders has brought along their discredit. (Jiménez de Asúa, 1915, p. 55)

16 The Economics of Civil Orders and Medals in Spain: An Update After Ten Years

261

And the “monetary” comparison of orders9 can even be found in ancient Greeks authors such as Aristotle (cf Baumert & Roldán, 2012) and Aeschines, as can be seen in the opening quote to this work, which continues as follows: At what time did our citizens display the greatest merit then or now? They were then eminent; now, much less distinguished. At what time were rewards, crowns, proclamations, and public honors of every kind most frequent—then or now? Then they were rare and truly valuable; then the name of merit bore the highest lustre; but now it is tarnished and effaced; while your honors are conferred by course and custom, not with judgment and distinction. (Aeschines 330 BC/1961)

This has been accompanied by an (at least apparent) arbitrary concession, many times, by “course and custom”.10 It seems, though, that de La Grasserie’s (1900, p. 395) assertation “this simple idea, that is, that as punishment can only be the result of a court judgement, it can also be only a court judgement who bestows a reward, has not yet germinated in society’s mind” might be still plain valid. Hence, we can summarise that, in a quite perfect analogy to economics, where an increase in the amount of money in circulation conduces to a reduction in the value of money, a raise in the number of orders awarded will imply a diminishing in the perceived value of the orders and of the underlying merits, thus perverting its original function. Thus, it is necessary to “behave avaricious when awarding orders, as their prestige depends on their scarcity” (Jiménez de Asúa, 1915: 57). A rise in the number of bestowing – and their subsequent devaluation – might be due to three possible reasons (for an overview of the role of orders and medals as incentives, see Baumert & Roldán, 2012; Frey, 2005; Frey & Gallus, 2017): (a) An increase in the meritorious people as a result of a more virtuous society. (b) A relaxation of the required merits or, even worse, awarding “by routine” (in fact the problem pointed out by Aeschines). (c) A less restricted control of the awarding in order to minimise the number of false non-awardings (“false negative” or type II error).

9 Parthey

(1907, pp. 450–451) indicates a curious form of intended inflation on an award: in order to degrade the Légion d’Honneur bestowed by Napoleon in the eyes of the French, Louis XVI devised the remedy of massively awarding this order —usually earned for bravery on the field of battle or for extraordinary merits— as a prize for schoolchildren and students. Additionally, a new Order of Saint Louis should be bestowed, that would have been worn simultaneously by all knights of the Légion d’Honneur. Nevertheless, the sense of honor of the French reacted so aversely, that the King was forced to completely abandon this idea. And, in fact, the order maintained its prestige, and was again worn proudly, as was soon later noticed even by foreign travelers (Schack, 1888). The prestige of the order continued to be so high, that von Bismarck (1898, vol. I, pp. 81–82) recorded in his Memories how after a public incident the police officers ceased their violence against the mob after detecting among the masses a Monsieur décore. 10 The custom of awarding orders to foreign visitors —from where derives the still in use practice among diplomats of “exchanging” awards during official visits— has its origin in the Middle Age. This gave birth to the rule that the most prestigious orders, like the Order of the Golden Fleece, were incompatible with any other award (Keen, 1984, pp. 278–280).

262

T. Baumert and B. Luceño-Ramos

Regarding point (c), it might be noted that giving awards not only acts as positive incentive to the person awarded (or to whom aspires to receive it) but may also have a negative external effect on the person disappointed by not receiving it while believing he should have (type II error). This question was already a matter of reflection by the father of modern Premial Law, Jeremy Bentham, when affirming that: Concerning persons in one line of merit it is not possible to honour some of them with a grade of elevation without relatively stooping down the others. (Bentham, 1818, pp. 34–35)

More than a century later the matter was still acute, as demonstrated by Winston Churchill’s speech in the House of Commons on August 22, 1944: A medal glitters, but it also casts a shadow. The task of drawing up regulations for such awards is one which does not admit of a perfect solution. It is not possible to satisfy everybody without running the risk of satisfying nobody. All that is possible is to give the greatest satisfaction to the greatest number and to hurt the feeling of the fewest. (Quoted in Frey, 2005, pp. 25–26)

Nevertheless, when trying to minimise type II errors, it should be borne in mind that both – errors and arbitrariness – are much less grievous in matters of rewards than in matter of punishment (Holbach, 1904, p. 6).

16.3 Empirical Analysis Regarding the current Spanish Premial System,11 it has to be stressed that according to article 62.f) of the Spanish Constitution of 1978, the awarding of all sorts of honours and distinctions is exclusively reserved to HM the King, thus confirming an ancestral tradition (García-Mercadal, 2010, pp. 223–230, 2019). Nevertheless, and notwithstanding this clear principle, in fact, it is the executive power that confers decorations. More precisely, the awarding is always made in name of the Head of State, but it is the executive power to decide the concession: the government for the higher grades (Grand Crosses and Collars) and the corresponding Ministry in the lower ones. Traditionally it has been believed that these awarding were completely discretional; nevertheless, recent interpretations of article 106.1 of the Constitution and of the Ley de la Jurisdicción Contencioso-Administrativa tend to understand that at least all objective criterions regarding the award could in fact be reviewed by courts. This refers mainly to questions like assuring that a person does not receive a higher category of an award than he is allowed to, etc. Unfortunately, this seldom

11 Several works, some of them in very broadly, have studied the Spanish Orders and their statutes. Among them we might point out the following: Gil Gorregaray (1864–1865), Silva Jiménez (1906), Sosa (1913–1915), de la Puente (1953), Calvó Pascual (1987), Grávalos González and Calvo Pérez (1988), Lorente Aznar (1999), Pérez Guerra (2000), as well as the already cited study by CeballosEscalera and García-Mercadal (2003).

16 The Economics of Civil Orders and Medals in Spain: An Update After Ten Years

• • • • • • • • • • • • • • • • •

263

Insigne Orden del Toisón de Oro (bestowed 1496). Real y Distinguida Orden Española de Carlos III (bestowed 1771, reformed in 2002), the most important among Spanish orders Real y Americana Orden de Isabel la Católica (bestowed 1815; reformed in 1998). Orden Civil del Mérito Agrícola (bestowed 1905; reformed in 1987). Orden al Mérito en el Trabajo (bestowed 1926; reformed in 1982). Orden de Alfonso X El Sabio (bestowed 1939). This order substituted the previous Orden de Alfonso XII (established in 1902). Orden Civil de Sanidad (bestowed 1943). Orden Imperial del Yugo y las Flechas (bestowed 1937). Orden de San Raimundo de Peñafort (bestowed 1944). Orden de Cisneros (bestowed 1944). Orden del Mérito Civil (bestowed 1926; reformed in 1998). Orden del Mérito Deportivo (bestowed 1982). Orden de la Solidaridad Social (bestowed 1988). Orden del Mérito Constitucional (bestowed 1988). Orden al Mérito del Plan Nacional sobre Drogas (bestowed 1995). Real Orden de Reconocimiento Civil a las Víctimas del Terrorismo (bestowed 1995). Order of Merit in Commerce (bestowed 2010).

Fig. 16.1 Civil orders that constitute the current Spanish Premial System. (Source: own elaboration: The year of the reform is always that of the most recent one)

occurs, and irregularities were – and unfortunately still are – no exception (CeballosEscalera & García-Mercadal, 2003, p. 74). Figure 16.1 gives an overview of the different civil orders that constitute the current Spanish Premial System. It should be noted that there are also four “historical” orders who derive directly from the medieval Knight Orders—Orden de Santiago, Orden de Alcántara, Orden de Calatrava and Orden de Montesa— which, anyhow, have degenerated into a sort of “nobiliary club” and have thus been excluded of our study. A couple of points have to be explained regarding these orders: first, the Insigne Orden del Toisón de Oro (The Order of the Golden Flees) is not properly a Spanish order, but a dynastic order, linked to the Borbón family. For their part, the Orden Imperial del Yugo y las Flechas and the Orden de Cisneros have not been awarded since 1978 – the former due to its strong symbolism related to the Franco government, the later for no evident reason – although there are still many bearers alive (García-Mercadal, 2010, p. 232). Finally, the last six awards, and more specifically the last four ones, are not properly orders – even if they erroneously have bear that name – but decorations, that is, medals (cf. Ceballos-Escalera & GarcíaMercadal, 2003, p. 65), and may therefore be ignored for the purpose of our study.

264

T. Baumert and B. Luceño-Ramos

16.4 Empirical Analysis Unlike the situation in 2012 – when the original version of this study was presented – the Spanish government now publishes, although unregularly and with a certain lag, the number of civil orders and medals bestowed classified by the year of the awarding12 ; according to the classes of the award (yet simplified in only two categories: highest class, that is, Great Crosses and Collars, and lower classes, which encompasses all the remaining ones); and indicating the gender (binary classification) of the recipient. By doing so, it accomplished one of the points demanded in the already mentioned 2012 paper. However, despite the important improvement that this means, the data series present some important shortcomings: for example, the aggregation by natural years makes it difficult to do estimation by the legislature (as in the present paper), especially for the lower classes, as their details, including the exact day of bestowing, are not made public.13 Also, no information has been published since the new Spanish government took over control in 2019. Hence, we will limit our study to the seven civil orders above listed, and for a time span of 37 years (1982–2019) differentiating eleven different legislatures with three different political orientation of the Spanish government: Center (UCD), Socialist (PSOE) and Popular party (PP).14 Following the example of the original paper, we then run a simple ANOVA model to check whether there are statistically significant differences in the number of bestowing between governments of different coleurs, according to the following null and alterative hypothesis: H 0 : µij = µij  H 1 : µij = µij  where i represents the order and j the legislature Table 16.1 shows the result of the Levene test which checks for homoscedasticity as can be observed, the null hypothesis of equal variances is only rejected at the 5% level in two cases (Mérito Agrario and Mérito Constitucional), implying that in those two we will have to use post-hoc analysis assuming heterogeneous variances.

12 https://www.mpr.gob.es/prencom/documentos/Paginas/RegistroOrdenesCondecoraciones.aspx 13 As has already been explained, the highest classes are bestowed by the Government, and hence need to be published in the Official State Bulletin, while all other classes, which are awarded by their corresponding Ministry, are not. 14 Obviously, this implies a certain simplification of the complex Spanish political reality, as several of this governments where formed as coalitions that often encompassed parties with apparently contradictory political positions. And also, insofar as some legislatures lasted less than the originally expected 4 years, thus making the comparison less homogenous. Hence, the ultimate criterion has been the party adscription of the President of the Government. The complete list of the legislatures can be consulted at: https://www.lamoncloa.gob.es/gobierno/gobiernosporlegislaturas/ Paginas/index.aspx

16 The Economics of Civil Orders and Medals in Spain: An Update After Ten Years

265

Table 16.1 Test of homogeneity of variances Order Mérito Agrario (all sections) Alfonso X el Sabio Mérito constitucional Carlos III Mérito civil Isabel la Católica San Raimundo de Peñafort

Levene statistic 3.979 1.760 210.374 1.956 .791 1.554 2.152

df1 8 8 9 10 10 10 10

df2 24 24 25 29 29 29 29

Sig. .004 .136