What Works in Latin American Municipalities?: Assessing Local Government Performance 1803929065, 9781803929064

Citation preview

What Works in Latin American Municipalities?

I gratefully dedicate this book to my parents, Carmen and Pedro Avellaneda, who planted in me the spirit of self-growth, learning, and continuous improvement; as well to all my “earth angels,” who have supported me during my personal and academic journeys.

What Works in Latin American Municipalities?

Assessing Local Government Performance

Edited by

Claudia N. Avellaneda O’Neill School of Public and Environmental Affairs, Indiana University, USA

Cheltenham, UK • Northampton, MA, USA

© Claudia N. Avellaneda 2023

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library Library of Congress Control Number: 2022950541 This book is available electronically in the Political Science and Public Policy subject collection http://dx.doi.org/10.4337/9781803929071

ISBN 978 1 80392 906 4 (cased) ISBN 978 1 80392 907 1 (eBook)

EEP BoX

Contents List of figuresvi List of tablesvii List of contributorsix Acknowledgementsx Introduction to What Works in Latin American Municipalities?1 Claudia N. Avellaneda 1

Assessing the influence of Brazilian mayors’ human capital and political context on fiscal inputs Claudia N. Avellaneda and Marco Antonio Catussi Paschoalotto

2

Administrative capacity and Chilean local governmental effectiveness55 Gabriel Piña and Claudia N. Avellaneda

3

Colombian education quality: political, managerial, or bureaucratic quality? Claudia N. Avellaneda

4

Mayor’s gender and task-specific education influences on Ecuadorian municipal financial efficiency Julio C. Zambrano and Claudia N. Avellaneda

107

5

Explaining Mexican mayors’ preferences for participatory decision-making: an experimental analysis Claudia N. Avellaneda and Johabed Olvera

128

6

Determinants of property value reappraisals: municipal responsiveness to urban changes Claudia N. Avellaneda and Gabriel Piña

149

7

Conclusion to What Works in Latin American Municipalities?202 Claudia N. Avellaneda

16

81

Index219 v

Figures 1.1

Brazil (São Paulo State highlighted) and São Paulo State (645 municipalities)

26

1.2

Dependent variable: logged municipal tax collection

26

3.1

Department of Norte de Santander, Colombia

90

4.1

The marginal effect of public administration education on financial self-sufficiency as mayor gender changes

119

4.2

The marginal effect of public administration education on current expenses as mayor gender changes

120

6.1

Number and percentage of municipalities updating their cadaster by year (2007–2014)

154

6.2

Percentage of municipalities by department updating their cadaster (2007–2014)

155

6.3

Percentage of eligible municipalities updating their cadaster by year (2007–2014) 

155

A6.1 Municipalities updating cadaster in the study period (2007–2014)201

vi

Tables 1.1a Description and source of the independent variables

28

1.1b Summary statistics

30

1.2

Panel data estimates for the ratio of property tax collected over total municipal revenues, 2009–2016

33

1.3

Panel data estimates for the total property tax collected (reais) in São Paulo municipalities, 2009–2016

35

1.4

Panel data estimates for the ratio of actual property tax collected over estimated property tax collected, 2009–2016

37

1.5

Panel data estimates with interaction effects for the ratio of property tax collected over total municipal revenues, 2009–2016

39

1.6

Panel data estimates with interaction effects for the total property tax collected (reais) in São Paulo municipalities, 2009–2016

42

1.7

Panel data estimates with interaction effects for the ratio of actual property tax collected over estimated property tax collected, 2009–2016

44

2.1

Descriptive statistics

69

2.2

Effectiveness in infrastructure grants approved (projects approved/projects requested)

70

2.3

Effectiveness in infrastructure grant approved (money acquired/money requested)

73

3.1

Summary statistics

92

3.2

Ordered logit and mixed-effect estimates on the determinants of school performance, 2007

99

4.1

Top ten fields of study of Ecuadorian mayors between the 2005–2009 and 2009–2013 administrative periods vii

115

viii

What works in Latin American municipalities?

4.2

Descriptive statistics

117

4.3

The effect of public administration education and gender on financial self-sufficiency

118

4.4

The effect of public administration education and gender on the composition of financial self-sufficiency (own income/ current expenses)

121

5.1

Descriptive statistics

137

5.2

Distribution of mayor’s decisions by hypothetical options

140

5.3

OLS (LPM) estimates

142

5.4

Logit estimates

143

6.1

Descriptive statistics

170

6.2

Determinants of cadaster update in Colombian municipalities (2007–2014)172

A6.1 Municipalities updating cadaster in less than five years since last update

188

A6.2 Total number of urban properties and growth (2007–2014)

191

A6.3.1 Correlation matrix

192

A6.3.2 Correlation matrix

194

A6.3.3 Correlation matrix

195

A6.4 Determinants of cadaster update for municipalities with five or more years since last update, including mayors’ education variable (2007–2014)

196

A6.5 Determinants of cadaster update: municipalities with five or more years since last update, including mayors’ experience variable (2007–2011)

198

7.1

Assessed performance dimension across countries

205

7.2

Performance effects of assessed drivers across countries

206

7.3

Significant performance effects of control variables across countries208

Contributors Claudia N. Avellaneda, O’Neill School of Public and Environmental Affairs, Indiana University, USA. Johabed Olvera, School of Public Policy, Penn State University, USA. Marco Antonio Catussi Paschoalotto, São Paulo School of Business and Administration, Getulio Vargas Foundation, Brazil. Gabriel Piña, Chield Trends, USA. Julio C. Zambrano, TUM School of Social Science and Technology, Technical University of Munich, Germany.

ix

Acknowledgements This book would not have been possible without the strong efforts, patience, and valuable insights of my graduate research assistants, who also are contributors to this endeavor. My deep gratitude also goes to many Latin American mayors and public servants. Their cooperation and willingness to share data, insights, and personal experiences have helped me build a research agenda about which I am passionate.

x

Introduction to What Works in Latin American Municipalities? Claudia N. Avellaneda The constant search for improving governmental performance never ceases. Across all government levels, chief executives, public managers, consultants, and advisors pursue strategies and governance arrangements expected to increase governmental performance. The search for government improvement has generated calls for documenting and demonstrating governmental performance (Kelman 2007, Marr 2009, Radin 2006, Van Dooren and Van de Walle 2008, Sanger 2013, Behn 2003, 2006). Measuring performance implies “documenting whether an organization is reaching its goals” (Sanger 2013, 185). Administrative reforms have encouraged the use of performance assessment instruments in the public sector (van Thiel and Leeuv 2002, Heinrich 2002, Berman 2008). As a result, worldwide governments started documenting activities, inputs, costs, complaints, procedures, beneficiaries, among other things and producing reports that are used for accounting, auditing, budgeting, and management purposes, as well as for complying with legislative mandates (Berman 2008). Consequently, governments created administrative data on inputs and outputs. These data permit governments to produce statistics, which can be used for internal management, budget purposes, and reports available to stakeholders external to the agency. In sum, performance indicators become necessary for any well-run government (Berman 2008). Despite a widespread trend for assessing government performance, uncertainty still exists over whether measuring performance improves government performance (Sanger 2013). As Hatry (2002, 358) puts it, “performance measurement by itself is only one of a number of needed tools for effective governing for results.” What appears certain is that enhanced governmental performance has the potential to boost and/or regain citizens’ satisfaction and trust in government (Keele 2007, Welch et al. 2005, Bouckaert and Van de Walle 2001, Wu et al. 2021), political leaders (Greasley and John 2011), and regime type (Norris 2011). However, measuring government performance proves to be a complex task (Boyne 2002, 2003, Berman 2008, Behn 2003, 2006, 2008, Brewer and Selden 2000). Organizational performance is a socially constructed phenomenon and 1

2

What works in Latin American municipalities?

hard to measure in the public sector (Brewer and Selden 2000). Governments carry out numerous activities including delivering services, regulating behaviors, managing disasters, and collecting taxes. Moreover, organizational performance can be assessed through various dimensions and criteria (Boyne 2002, 2003, Carini et al. 2020. Bayo-Moriones et al. 2020, Walker and Andrews 2015), and these dimensions, in turn, can be captured with objective and subjective (internal and external) measures (Boyne et al. 2006, Andrews et al. 2006b, Meier and O’Toole 2012, 2013, Singh et al. 2016, Cheon et al. 2021). Assessing performance in government begins with public executives’ identification of public purposes, specifying the main goals, and selection of the means to measure the accomplishments of those goals (Marr 2009, Berman 2008). Unlike in the private sector, “[g]overnment performance has no single criterion of success” (Berman 2008, 3). While private organizations survive and grow by meeting customers’ needs and expectations, a comparable, compelling survival requirement does not exist for governments to consult with their constituents (Berman 2008, 3). Public hearings, though, offer opportunities for citizens to express their views regarding programs, legislation, and government performance (Berman 2008). Complexity in measuring government performance should not deter people from doing. Boyne (2002, 2003) offers a practical set of five dimensions of organizational performance to assess government actions and goal achievement. While the outputs dimension involves both quantity and quality, the service outcome dimension includes formal effectiveness, impact, and equity/ fairness. The efficiency dimension pertains to cost per unit of output, and the responsiveness dimension can be captured with consumers’ satisfaction, citizens’ satisfaction, and employees’ satisfaction. Finally, the democratic dimension seeks to assess participation, accountability, and probity (Boyne 2002, 2003). Despite the variety of dimensions and criteria to assess government performance, preferences exist. Data availability, timing, managers’ and auditors’ choice, and policy area may explain the preference of certain dimensions over others. Systematic reviews of empirical organizational performance studies reveal a greater number of studies focusing on outputs and outcomes (Boyne 2003, Carini et al. 2020, Walker and Andrews 2015). However, the measurement and reporting of governmental outcomes are still uncommon in government reports (Berman 2008) despite documented citizens’ preferences for outcomes (Berman 2008). This mismatch in dimension reporting has led Berman (2008) to argue that although the citizens elect government leaders, support government efforts through taxes, and receive government services, their preferences in terms of performance reporting are unmet. Grizzle (2018, 329–340) also calls for outcomes and accountability measures by develop-

Introduction

3

ing the concept of joint responsibility and focusing on program outcomes rather than activities. Heinrich (2002), however, notes that requirements for outcomes-based performance management are increasing performance evaluation activities at all governmental levels. In sum, reliance on certain performance dimensions calls for further studies, including several dimensions captured with heterogeneous subjective and objective indicators across multiple policy areas, countries, regions, and variant levels of democracy and economic development. Focusing on understudied settings has the potential to add knowledge by identifying moderators, mediators, rejecting existing causal mechanisms and/or uncovering new ones. In fact, Andrews, Boyne and Walker (2011) point toward the need for research that includes a variety of performance measures, or as Behn (2003, 586) puts it, “[d]ifferent purposes require different measures.” It also becomes important to recognize the problems with performance measures. As Hatry (2002, 358) states, “performance measurement has many limitations, as well as uses.” These limitations have led some scholars to challenge the performance measurement movement (Radin 2006). No measure or indicator is perfect in producing accurate estimates. Behn (2008), for example, lists the seven big errors of PerformanceStat, the name Behn uses to denote all but similar performance strategies. Common method bias, recall issues, and cross-contamination problems that characterize subjective performance measures are some of the measurement errors. These errors should be reduced by using a combination of subjective and objective measures (Andrews et al. 2010). However, subjective and objective measures should be correlated only when they assess the same dimension of performance (Andrews et al. 2010). Moreover, performance assessment should consider the unique characteristics of the public sector by developing systems with multiple performance indicators and striking a balance between “measure pressure” and minimizing dysfunctional effects (van Thiel and Leeuw 2002, 267–281). Despite the problems, government performance measures do some good. For instance, Heinrich (2002, 712–725) reports that the results of empirical analyses confirm that the use of administrative data in performance management is unlikely to produce accurate estimates of true program impacts. However, Heinrich’s (2002) analysis also suggests administrative data can still generate useful information for public managers about policy tools that can be manipulated to improve organizational performance. Likewise, Newcomer (2007) identifies challenges and opportunities public managers face with these performance measures. Therefore, public managers and government agencies should not be discouraged from ratcheting up performance measures (Behn 2006). Assuming public managers overcome the performance measures challenges and take on their informational and guiding opportunities, the focus shifts

4

What works in Latin American municipalities?

toward factors explaining governmental performance. What works isn’t as simple in national and state/provincial agencies or in Latin American municipalities. Public management scholarship has dedicated considerable attention on identifying what works and does not work. It would take several pages to list all the works and colleagues who have undertaken this research question at all government levels and whose research I have admired, relied on, and from which I have learned a lot. What works seems to depend on the context (Meier, Rutherford and Avellaneda 2017, O’Toole and Meier 2017), policy area and/ or issue salience (e.g., Meier et al. 2017), level of government (e.g., Bouckaert and Halligan 2011), politics (e.g., Avellaneda 2012), managerial gender (e.g., Meier et al. 2006, Avellaneda et al. 2022), turnover (e.g., Bello-Gomez and Rutherford 2021), qualifications (e.g., Lynn 1987, Avellaneda 2009a, 2009b, 2012, 2016, Olvera and Avellaneda 2019), bureaucratic factors (gender, discretion, representativeness, motivation, incentives, red tape, qualifications, tenure, etc., e.g., Keiser et al. 2002, Christensen et al. 2017, Marvel and Resh 2015), organizational size (e.g., Avellaneda and Correa Gomes 2015) and structure (e.g., Whetsell et al., 2021), past performance (e.g., Nicholson-Crotty and O’Toole 2004), leadership (e.g., Fernandez 2005), timing (Desai and Madsen 2021, Short and Payne 2008), networking (e.g., Hicklin et al. 2008, Van den Bekerom et al. 2017), regulations (e.g., Zambrano and Avellaneda 2022), market competition (e.g., Walker et al. 2011), resources (human and material, e.g., Carmeli and Tishler 2004, Bradley et al. 2001), strategic planning (e.g., Elbanna et al. 2016, Hendrick 2003), strategy content (e.g., Andrews et al. 2006a), innovation (e.g., Damanpour et al. 2009, Teodoro 2009, Walker et al. 2011), implementation strategies (e.g., O’Toole 2000), external actors (e.g., Brudney and Hebert 1987), exogenous historical factors (e.g., La Porta et al. 1999), institutions (formal and informal, e.g., Helmke and Levitsky 2006, Robinson and Acemoglu 2012), sociodemographic features (e.g., Lineberry 1976, Marvel and Resh 2015), intergovernmental relations (e.g., Agranoff and Radin 2015), social capital (e.g., Keele 2007), governance arrangements (contracting out, at-home production, hybrid modes, e.g. Damanpour et al. 2022), decision-making processes (e.g., Rainey et al. 2010), use of performance measures (e.g., Sanger 2013), performance management, (e.g., Moynihan 2008), and benchmarks (Ammons 2014), among others. As with performance measures, preferences exist in studying certain drivers over others. In 2003, Boyne’s systematic reviews of 65 empirical studies showed five predominant theoretical perspectives: resources, regulations, management variables, market competition, and organizational variables (Boyne 2003). However, Boyne’s systematic analysis found empirical support only for resources and management variables. In 2015, Walker and Andrews presented another systematic review of 86 empirical studies on local government performance and found that greater attention has been given to

Introduction

5

organization size, strategy content, planning, staff quality, personnel stability, representative bureaucracy, and networking. The same systematic review reveals strong positive performance effects from staff quality, personnel stability, and planning, and moderate support for the benefits of networking, representative bureaucracy, and strategy content (Walker and Andrew 2015). Walker and Andrews’ (2015) subanalyses reveal different relationships across dimensions of performance, and that British and American scholars have dominated these studies (Walker and Andrews 2015). This book seeks to reduce the regional dominance in existing research by testing different theoretical perspectives in local settings in the understudied Latin American region whose countries are characterized by varying levels of democracy and economic development. Each of the six empirical chapters assesses local government performance across different dimensions and policy areas: education outcome quality, financial efficiency, effectiveness in infrastructure grant acquisitions, responsiveness to climate changes, participatory decision-making in economic development, and fiscal effectiveness in resource inputs. The next section provides more details about the book’s goals and chapters’ specific cases and analyses.

THIS BOOK This book focuses on explaining government performance at the local level. Since the adoption of fiscal, political, and administrative decentralization local governments have become responsible for implementing policies enacted by national, state, and local legislatures (Rondinelli et al. 1983, Cheema and Rondinelli 2007, Smoke 2015, Rondinelli 2017). Although having differing responsibilities, local governments all supply basic services. Citizens depend on these services being delivered well and predictably, and they pay local taxes to receive outputs and outcomes (Berman 2008). Yet in certain settings, information about how local governments are performing tends to be unavailable or available inconsistently (Berman 2008). Moreover, identifying the drivers of local implementation and subsequently performance becomes a priority because what is implemented may deviate from what was enacted. Localities directly contribute to human development through policy implementation. Therefore, local decisions have significant social and economic implications. The book seeks to assess local government performance across different dimensions and across several policy areas. Despite data availability challenges, we were able to compile six empirical studies, each one covering a different Latin American country. To our knowledge, this is the first effort to consolidate several quantitative studies of local governments in the region. Compiling data derived from the almost 2,000 municipalities analyzed across

6

What works in Latin American municipalities?

six countries took about five years. The studied localities and countries include: 645 (out of 5,550) in Brazil, 900 (out of 1,103) in Colombia, 340 (out of 346) in Chile, 221 (out of 221) in Ecuador, and 50 (out of 2,454) in Mexico. The book’s first goal was assessing municipal performance objectively across different countries, several performance dimensions, different policy areas, and distinct estimated techniques. Each of the studied countries evaluates a different dimension of performance. The Brazilian case centers on fiscal inputs; Colombia on education output quality; Chile on effectiveness in grant acquisition; Ecuador on administrative cost efficiency; Mexico on participatory decision-making; and Colombia on responsiveness to urban structural changes. The book’s second goal involves identifying common drivers, if any, across (a) six local settings, (b) six performance dimensions, and (c) policy areas. The tendency of international financial institutions to advocate for standard tools, reforms, practices, and governance indicators, among others (Buduru and Pal 2010), encourages continued research into unveiling what works, when it works, and under what conditions after considering context (O’Toole and Meier 2017, Meier, Rutherford and Avellaneda 2017). Although findings’ generalization continues to be desirable, context-specific research has the potential to uncover contingences, indirect drivers, and overlooked factors. The third goal of this volume consists of testing the performance effects of diverse potential drivers. Because no single factor explains organizational performance, testing the explanations for individual, organizational, and contextual factors should increase our understanding of government performance. In doing so, each of the empirical cases focuses on a particular driver, while controlling for standard political, demographic, and socioeconomic factors. Among the specific drivers, the studies test the multi-dimensional performance effects of individual factors, such as mayors’ gender, qualifications (education and experience), political support, political ambition, and managerial quality. At the organizational level, the studies also test the performance effects of municipal capacity (personnel qualifications, expertise, and number), institutional capacity (e.g., existence of plans and committees), and bureaucratic quality. Among the contextual factors, one of the empirical studies assesses the municipal performance effect of guerrilla presence in Colombia. The inclusion of drivers at different levels should help us understand how cross-level drivers interact to influence municipal performance. This book aims to contribute in several ways. First, the six empirical studies focus on an understudied region whose democracies illustrate the third wave of democratization (Hagopian and Mainwaring 2005, Mainwaring and Torcal 2006, Haggard and Kaufman 2016). The drivers of local government performance in these non-consolidated democracies (Diamond 1994, Schedler 1998, 2001) may differ from those identified in consolidated democracies, such as

Introduction

7

the United States and United Kingdom. Second, considerable public management research heavily has focused on the outputs and outcomes performance dimensions when assessing organizational performance (e.g., see Boyne 2003 for a 65-article review), leaving other dimensions unexplored. This book helps close this gap by dedicating equal attention to six different performance dimensions: output quality, responsiveness, effectiveness, efficiency, participative decision-making, and fiscal inputs. In doing so, the book adds to existing literature by including diverse performance dimensions captured with various objective indicators across different policy areas. Thirdly, this books also contributes to local governance literature by presenting six empirical studies in developing and transitional economies. Due to data availability, great emphasis has been devoted to developed economies. Conducting research in developing and transitional economies is not just necessary but also a moral duty. Research in understudied settings has the potential for testing models and frameworks developed in high-income units. In these underdeveloped and transitional economies, government performance may more likely be affected by personalistic, electoral, and/or political factors, rather than institutional and managerial sources. Therefore, the empirical cases presented here control for political, electoral, and contextual factors. Fourthly, the empirical studies rely on several policy areas: education (Colombia), fiscal policy (Brazil), public finances (Ecuador), infrastructure grants (Chile), urban planning (Colombia), and decision-making in development programs (Mexico). While many assessments of government performance have overemphasized some policy areas more than others (see Boyne’s 2003 review of 65 empirical studies), this book presents a range of policy areas, thus allowing for additional tests of existing explanations. Finally, this research endeavor has facilitated the training of five doctoral students. The five additional contributors to this volume are all from the Latin American region. This book has given me the opportunity and honor to mentor, collaborate, and co-author with them. Their Latin American background and passion for the region have enriched the presented studies. We started working on their respective country’s chapter during their doctoral program. Time has passed and now they are successful professionals, flying high and dedicated to exploring the working of Latin American local governments. In fairness to readers, we want to reveal that author’s and contributors’ country-specific knowledge determined case selection. When choosing between randomness in case selection and the author’s and contributors’ knowledge, personal contacts, and data collection availability, the latter triumphed. Nevertheless, research in each country was conducted objectively, focusing on administrative data and surveys directly administered by the authors.

8

What works in Latin American municipalities?

WHAT IS AHEAD The case studies follow in countries’ alphabetic order. The first empirical chapter corresponds to Brazil. Marco Antonio and I suggest an integrative framework to examine the direct and nonlinear (moderating) performance effects of political support and managerial quality. In the Brazilian context, municipal performance is measured as fiscal inputs (property tax collected) and effectiveness in property tax collection (ratio of estimated tax to be collected to the actual tax collected). The study relies on a data set derived from the 645 municipalities that make up the state of São Paulo during the 2009–2016 period, covering two mayoral administrations. Mayoral party alignment with governor and president, as well as margin of victory and political support at the city council level, capture political support. Mayoral educational attainment, as well as years of public sector experience, capture managerial quality. After controlling for other political, organizational, and economic factors, results suggest neither mayoral human capital nor political support correlates with property tax collection. Moreover, none of the indicators capturing political support moderates the mayoral/managerial quality-property tax collection relationship. The second empirical chapter features the entire population of Chilean municipalities. The authors highlight the importance of organizational capacity for a well-functioning government, as well as the scarce empirical analysis on the organizational capacity-government performance relationship. To address these gaps, the study objectively measures organizational capacity across three dimensions – capability, expertise, and human resources – and tests the effect of organizational capacity on local government effectiveness in securing infrastructure grants. The study relies on a data set of approximately 54,000 infrastructural grant proposals submitted by 340 (out of 345) Chilean municipalities during a nine-year period covering three mayoral administrations (2005–2013). After controlling for past performance and other grant and municipal features, results suggest municipal effectiveness in securing grants is positively influenced by both administrative capacity and political factors. Findings are robust across alternative model specifications. The third research chapter presents the Colombian case. Here, the author explores the determinants of public school performance in terms of education outcome quality. The study suggests that political, managerial, and bureaucratic quality influence school performance. These propositions are tested using data from the 88 public schools located in the 39 municipalities that comprise one of the Colombian states. Education outcome quality refers to the school ranking based on student performance in the standardized national test for 11th graders. Mayoral education captures political quality, while princi-

Introduction

9

pals’ professional ranking assesses managerial quality and teachers’ education years and professional ranking measure bureaucratic quality. After controlling for demographic, socioeconomic, contextual variables, and school prior year performance, findings show that school size, teachers’ quality, and principals’ quality positively influence school performance through improved ranking. This study reports a comprehensive test of the expected drivers influencing organizational performance. The following chapter highlights Ecuadorian municipalities. In doing so, authors recognize that considerable studies have explored gender effects at the state and national levels (Dollar, Fisman, and Gatti 2001, Michelle Heath, Schwindt-Bayer, and Taylor-Robinson 2005, Lawless 2004, Wängnerud 2009). However, less is known about whether those findings extrapolate to the local level. To address this gap, this empirical chapter assesses the link between women representation as local chief executives and municipal financial self-sufficiency (ratio of expenditures to own local revenues). Beside the gender effects, the chapter also examines the performance effects of mayors’ job-specific training/knowledge (Hall 2001, Holland 1973, Wiersema and Bantel 1992, Bertrand and Schoar 2003, Avellaneda 2012). The authors argue that knowledge or education specific to a manager’s task also should contribute to organizational performance across specific dimensions. That is, the study suggests a direct and indirect (through gender) relationship between task-specific education and municipal financial efficiency. The authors test the propositions using a data set derived from the 221 Ecuadorian municipalities from 2005 to 2013, thus covering two mayoral administrations. After controlling for potential confounding factors, findings show that municipalities led by women mayors are not statistically different in terms of financial self-sufficiency. Municipalities led by mayors with a background education in public administration exhibit more financial self-sufficiency than those peers led by mayors without that educational background. Moreover, the positive effect of mayors’ public administration education increases in municipalities governed by women. That is, while gender does not have a direct effect on municipal self-sufficiency, women do contribute to municipal finance when they possess public administration education. The chapter dedicated to Mexican municipalities presents a field experiment analysis. The authors test the effect of chief executives’ political ambition and qualifications (education and public sector experience) on their preferences to delegate in a collaborative arrangement through participatory decision-making. The field experiment generates a hypothetical municipal collaborative arrangement involving three partners to test mayors’ willingness to delegate the selection of project beneficiaries to one of the collaboration partners. The authors suggest mayors with re-election ambition prefer less participatory decision-making or are less inclined to delegate the selection of

10

What works in Latin American municipalities?

project beneficiaries, while mayors with political ambition for higher levels of government are expected to be more inclined to delegate. The authors also expect elected public officials with higher education levels and previous public sector experience to be less likely to delegate because of self-capacity reasons. These expectations are tested in a survey of 50 Mexican mayors. Results provide empirical support of the propositions: while mayors with re-election ambition are less likely to delegate selection of beneficiaries, mayors with political ambition for higher levels of government are more likely to delegate. More experienced and educated mayors are less likely to delegate the selection of beneficiaries. Finally, the last empirical chapter features the Colombian municipalities. The authors investigate the determinants of municipal responses to urban structural changes. Specifically, the authors examine the determinants of property value reappraisals by testing the effect of political, contextual, and managerial factors. The study relies on a data set derived from 916 Colombian municipalities for an eight-year period (2007–2014). After implementing robustness checks, results consistently show that political factors, such as the electoral cycle and a mayor’s city council ideological alignment, influence the likelihood of a municipality updating its property values. Municipal contextual factors also matter: while royalties from natural resources seem to discourage recipient municipalities from conducting reappraisals, on the other hand urbanization appears to promote cadaster updates. Small municipalities (fewer than 20,000 inhabitants) tend to be influenced by other contextual factors such as having the necessary cartographic tools. Managerial factors are also strong predictors: mayoral qualifications in terms of education level positively contribute to municipal update of cadaster. In small municipalities mayoral public sector experience also promotes reappraisals of property values.

REFERENCES Agranoff, R. and Radin, B.A. 2015. “Deil Wright’s overlapping model of intergovernmental relations: the basis for contemporary intergovernmental relationships.” Publius: The Journal of Federalism, 45(1): 139–159. Ammons, D. 2014. Municipal Benchmarks: Assessing Local Perfomance and Establishing Community Standards. Routledge. Andrews, R., Boyne, G.A. and Walker, R.M. 2006a. “Strategy content and organizational performance: an empirical analysis.” Public Administration Review, 66(1): 52–63. Andrews, Rhys, Boyne, G. and Walker, R. 2006b. “Subjective and objective measures of organizational performance: an empirical exploration.” In Public Service Performance: Perspectives on Measurement and Management. Cambridge University Press, 14–34.

Introduction

11

Andrews, Rhys, Boyne, G.A. and Walker, R.M. 2011. “Dimensions of publicness and organizational performance: a review of the evidence.” Journal of Public Administration Research and Theory, 21(3): 301–319. Andrews, R., Boyne, G.A., Moon, M.J. and Walker, R.M. 2010. “Assessing organizational performance: exploring differences between internal and external measures.” International Public Management Journal, 13(2): 105–129. Avellaneda, Claudia N. 2009a. “Mayoral quality and local public finance.” Public Administration Review, May/June: 469–486. Avellaneda, Claudia N. 2009b. “Municipal performance: does mayoral quality matter?” Journal of Public Administration Research and Theory, 19(2): 285–312. Avellaneda, Claudia N. 2012. “Do politics or mayors’ demographics matter for municipal revenue expansion?” Public Management Review, 14(8): 1061–1086. Avellaneda, Claudia N. 2016. “Government performance and chief executives’ intangible assets: motives, networking, and/or capacity?” Public Management Review, 18(6): 918–947. Avellaneda, Claudia N. and Correa Gomes, R. 2015. “Is small beautiful? Testing the effects of size on Brazilian municipal performance.” Public Administration Review, 75(1): 137–149. Avellaneda, Claudia, Bello-Gomez, R.A. and Correa Gomes, R. 2022. “Municipal fiscal performance: mayors’ gender and organizational human resources.” Journal of Policy Studies, 37(3). Bayo-Moriones, A., Galdon-Sanchez, J.E. and Martinez-de-Morentin, S. 2020. “Performance appraisal: dimensions and determinants.” The International Journal of Human Resource Management, 31(15): 1984–2015. Behn, R.D. 2003. “Why measure performance? Different purposes require different measures.” Public Administration Review, 63: 586–606. Behn, R.D. 2006. Performance Leadership: 11 Better Practices that can Ratchet Up Performance. Washington, DC: IBM Center for the Business of Government. Behn, R.D. 2008. The Seven Big Errors of Performance Stat. Cambridge, MA: John F. Kennedy School of Government, Harvard University. Bello-Gomez, R.A. and Rutherford, A. 2021. “Predicting executive vacancies: an organizational approach.” Public Administration. https://​doi​.org/​10​.1111/​padm​ .12770 Berman, Barbara J. Cohn. 2008. “Involving the public in measuring and reporting local government performance.” National Civic Review, 97(1): 3–11. Bertrand, M. and Schoar, A. 2003. “Managing with style: the effect of managers on firm policies.” Quarterly Journal of Economics, 118: 1169–1208. Bouckaert, G. and Halligan, J. 2011. “Managing performance across levels of government: lessons learned or reproducing disconnects?” In Policy, Performance and Management in Governance and Intergovernmental Relations, 236–254. Bouckaert, Geert and Van de Walle, S. 2001. “Government performance and trust in government.” In Ponencia presentada en la annual conference of the European group on public administration, vaasa (Finlandia), 2: 9–42. Boyne, G.A. 2002. “Theme: local government: concepts and indicators of local authority performance: an evaluation of the statutory frameworks in England and Wales.” Public Money & Management, 22(2): 17–24. Boyne, G.A. 2003. “Sources of public service improvement: a critical review and research agenda.” Journal of Public Administration Research and Theory, 13(3): 367–394.

12

What works in Latin American municipalities?

Boyne, G.A., Meier, K.J., Meier, K.J., O’Toole Jr, L.J. and Walker, R.M. eds. 2006. Public Service Performance: Perspectives on Measurement and Management. Cambridge University Press. Bradley, S., Jones, G. and Millington, J. 2001. “The effect of competition on the efficiency of secondary schools in England.” European Journal of Operational Research, 13(5): 545–568. Brewer, Gene and Coleman Selden, Sally. 2000. “Why elephants gallop: assessing and predicting organizational performance in federal agencies.” Journal of Public Administration Research and Theory, 10(4): 685–712. Brudney, J.L. and Hebert, F.T. 1987. “State agencies and their environments: examining the influence of important external actors.” The Journal of Politics, 49(1): 186–206. Buduru B. and Pal, Leslie A. 2010. “The globalized state: measuring and monitoring governance.” European Journal of Cultural Studies, 13(4): 511–530. Carini, E., Gabutti, I., Frisicale, E.M. et al. 2020. “Assessing hospital performance indicators. What dimensions? Evidence from an umbrella review.” BMC Health Services Research, 20(1038), 2–13. Carmeli, A. and Tishler, A., 2004. “The relationships between intangible organizational elements and organizational performance.” Strategic Management Journal, 25(13): 1257–1278. Cheema, G. Shabbir and Rondinelli, Dennis A. eds. 2007. Decentralizing Governance: Emerging Concepts and Practices. Brookings Institution Press. Cheon, O., Song, M., Mccrea, A.M. and Meier, K.J. 2021. “Health care in America: the relationship between subjective and objective assessments of hospitals.” International Public Management Journal, 24(5): 596–622. Christensen, R.K., Paarlberg, L. and Perry, J.L. 2017. “Public service motivation research: lessons for practice.” Public Administration Review, 77(4): 529–542. Damanpour, Fariborz, Walker, R. and Avellaneda, C.N. 2009. “Combinative effects of innovation types on organizational performance: a longitudinal study of public services.” Journal of Management Studies, 46(4): 650–675. Damanpour, Fariborz, Sanchez, Fernando and Avellaneda, Claudia N. 2022 “Environmental and organizational antecedents of plural sourcing of public services.” Public Administration Review, 82(2): 325–337. Desai, Vinit and Madsen, Peter. 2021. “Take your time? How activity timing affects organizational learning and performance outcomes.” Organization Science. https://​ doi​.org/​10​.1287/​orsc​.2021​.1490 Diamond, Larry. 1994. “Rethinking civil society: toward democratic consolidation.” Journal of Democracy, 5(3): 4–17. Dollar, D., Fisman, R. and Gatti, R. 2001. “Are women really the ‘fairer’ sex? Corruption and women in government.” Journal of Economic Behavior & Organization, 46(4): 423–429. Elbanna, S., Andrews, R. and Pollanen, R. 2016. “Strategic planning and implementation success in public service organizations: evidence from Canada.” Public Management Review, 18(7): 1017–1042. Fernandez, Sergio. 2005. “Developing and testing an integrative framework of public sector leadership: evidence from the public education arena.” Journal of Public Administration Research and Theory, 15: 197–217. Greasley, Stephen and John, Peter. 2011. “Does stronger political leadership have a performance payoff? Citizen satisfaction in the reform of subcentral Governments in England.” Journal of Public Administration Research and Theory, 21(2): 239–256.

Introduction

13

Grizzle, Gloria A. 2018. “Measuring state and local government performance: issues to resolve before implementing a performance measurement system.” In Public Sector Performance. Routledge, 329–340. Haggard, Stephan and Kaufman, Robert R. 2016. “Democratization during the third wave.” Annual Review of Political Science, 19: 125–144. Hagopian, Frances and Mainwaring, Scott P., eds. 2005. The Third Wave of Democratization in Latin America: Advances and Setbacks. Cambridge University Press. Hall, D.T. 2001. Careers In and Out of Organizations. Thousand Oaks, CA: Sage Publications. Hatry, Harry P. 2002. “Performance management: fashion and fallacies.” Public Performance & Management Review, 25(4): 352–358. Heinrich, Carolyn J. 2002. “Outcomes-based performance management in the public sector: implications for government accountability and effectiveness.” Public Administration Review, 62: 712–725. Helmke, G. and Levitsky, S., eds. 2006. Informal Institutions and Democracy: Lessons from Latin America. jhu Press. Hendrick, R. 2003. “Strategic planning environment, process, and performance in public agencies: a comparative study of departments in Milwaukee.” Journal of Public Administration Research and Theory, 13(4): 491–519. Hicklin, A., O’Toole Jr, L.J. and Meier, K.J. 2008. “Serpents in the sand: managerial networking and nonlinear influences on organizational performance.” Journal of Public Administration Research and Theory, 18(2): 253–273. Holland, J.L. 1973. Making Vocational Choices: A Theory of Careers. Englewood Cliffs, NJ: Prentice Hall. Keele, L. 2007. “Social capital and the dynamics of trust in government.” American Journal of Political Science, 51: 241–254. Keiser, L.R., Wilkins, V.M., Meier, K.J. and Holland, C.A. 2002. “Lipstick and logarithms: gender, institutional context, and representative bureaucracy.” American Political Science Review, 96(3): 553–564. Kelman, S. 2007. “The transformation of government in the decade ahead.” In D.F. Kettl and S. Kelman (eds.), Reflections on 21st Century Government Management. Washington, DC: IBM Center for the Business of Government. La Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R. 1999. “The quality of government.” The Journal of Law, Economics, and Organization, 15(1): 222–279. Lawless, J.L. 2004. “Politics of presence? Congresswomen and symbolic representation.” Political Research Quarterly, 57(1): 81–99. Lineberry, Robert L. 1976. “Equality, public policy and public services: the underclass hypothesis and the limits to equality.” Policy and Politics, 4: 67–84. Lynn, Laurence E., Jr. 1987. Managing Public Policy. Boston, MA: Little Brown. Mainwaring, Scott and Torcal, Mariano. 2006. “Party system institutionalization and party system theory after the third wave of democratization.” Handbook of Party Politics, 11(6): 204–227. Marr, B. 2009. Managing and Delivering Performance: How Government, Public Sector, and Not-for-profit Organizations Can Measure and Manage what Really Matters. Oxford: Butterworth-Heinemann. Marvel, J.D. and Resh, W.G. 2015. “Bureaucratic discretion, client demographics, and representative bureaucracy.” The American Review of Public Administration, 45(3): 281–310.

14

What works in Latin American municipalities?

Meier, Kenneth J. and O’Toole, Laurence J. 2012. “Subjective organizational performance and measurement error: common source bias and spurious relationships.” Journal of Public Administration Research and Theory, 23(2): 429–456. Meier, K.J. and O’Toole, L.J. 2013. “I think (I am doing well), therefore I am: assessing the validity of administrators’ self-assessments of performance.” International Public Management Journal, 16(1): 1–27. Meier, K.J., O’Toole Jr, L.J. and Goerdel, H.T. 2006. “Management activity and program performance: gender as management capital.” Public Administration Review, 66(1): 24–36. Meier, Kenneth J., Rutherford, Amanda and Avellaneda, Claudia N., eds. 2017. Comparative Public Management: Why National, Environmental, and Organizational Context Matters. Washington, DC: Georgetown University Press. Michelle Heath, R., Schwindt‐Bayer, L.A. and Taylor‐Robinson, M.M. 2005. “Women on the sidelines: women’s representation on committees in Latin American legislatures.” American Journal of Political Science, 49(2): 420–436. Moynihan, D.P. 2008. The Dynamics of Performance Management: Constructing Information and Reform. Georgetown University Press. Newcomer, Kathryn E. 2007. “Measuring government performance.” International Journal of Public Administration, 30(3): 307–329. Nicholson-Crotty, S. and O’Toole Jr, L.J. 2004. “Public management and organizational performance: the case of law enforcement agencies.” Journal of Public Administration Research and Theory, 14(1): 1–18. Norris, Pippa. 2011. “Does democratic satisfaction reflect regime performance?” In How Democracy Works: Political Representation and Policy Congruence in Modern Societies, 115–135. Olvera, Johabed and Avellaneda, Claudia N. 2019. “Subnational government performance: testing multiple dimensions of governors’ experience.” Public Administration Review, 79(3): 383–398. O’Toole Jr, L.J. 2000. “Research on policy implementation: assessment and prospects.” Journal of Public Administration Research and Theory, 10(2): 263–288. O’Toole, Laurence J., Jr. and Meier, Kenneth J. 2017. “Introduction: comparative public management: a framework for analysis.” In Kenneth J. Meier, Amanda Rutherford, and Claudia N. Avellaneda (eds.), Comparative Public Management: Why National, Environmental, and Organizational Context Matters. Washington, DC: Georgetown University Press, 1–26. Radin, B.A. 2006. Challenging the Performance Movement: Accountability, Complexity and Democratic Values. Washington, DC: Georgetown University Press. Rainey, Hal G., Ronquillo, John and Avellaneda, Claudia N. 2010. “Decision making in public organizations.” In Paul C. Nutt and David Wilson (eds.), The Blackwell Handbook of Decision Making. Wiley-Blackwell, 349–377. Robinson, James A. and Acemoglu, Daron. 2012. Why Nations Fail: The Origins of Power, Prosperity and Poverty. London: Profile. Rondinelli, Dennis A. 2017. “Decentralization and development.” In International Development Governance. Routledge, 391–404. Rondinelli, Dennis A., Nellis, John R. and Shabbir Cheema, G. 1983. “Decentralization in developing countries.” World Bank Staff Working Paper 581: 13–28. Sanger, Mary Bryna. 2013. “Does measuring performance lead to better performance?” Journal of Policy Analysis and Management, 185–203. https://​doi​.org/​10​.1002/​pam​ .21657

Introduction

15

Schedler, Andreas. 1998. “What is democratic consolidation?” Journal of Democracy, 9(2): 91–107. Schedler, Andreas. 2001. “Measuring democratic consolidation.” Studies in Comparative International Development, 36(1): 66–92. Short, J.C. and Payne, G.T. 2008. “First movers and performance: timing is everything.” Academy of Management Review, 33(1): 267–269. Singh, S., Darwish, T.K. and Potočnik, K. (2016). “Measuring organizational performance: a case for subjective measures.” British Journal of Management, 27(1): 214–224. Smoke, Paul. 2015. “Managing public sector decentralization in developing countries: moving beyond conventional recipes.” Public Administration and Development, 35: 250–262. Teodoro, Manuel P. 2009. “Bureaucratic job mobility and the diffusion of innovations.” American Journal of Political Science, 53(1): 175–189. Van den Bekerom, P., Schalk, J. and Torenvlied, R. 2017. “Transforming input into output: how downward networking mediates the effect of external networking on organizational performance.” Public Performance & Management Review, 40(4): 625–651. Van Dooren, W. and Van de Walle, S., eds. 2008. Performance Information in the Public Sector: How it is Used. Basingstoke: Palgrave Macmillan. Van Dooren, W., Bouckaert, G. and Halligan, J. 2010. Performance Management in the Public Sector. New York: Routledge. van Thiel, Sandra and Leeuw, Frans L. 2002. “The performance paradox in the public sector.” Public Performance & Management Review, 25(3): 267–281. Walker, R.M. and Andrews, R. 2015. “Local government management and performance: a review of evidence.” Journal of Public Administration Research and Theory, 25(1): 101–133. Walker, Richard, Brewer, Gene, Boyne, George and Avellaneda, Claudia N. 2011. “Market orientation and public service performance: NPM gone mad?” Public Administration Review, 71(5): 707–717. Wängnerud, L. 2009. “Women in parliaments: descriptive and substantive representation.” Annual Review of Political Science, 12: 51–69. Welch, Eric W., Hinnant, Charles C. and Jae Moon, M. 2005. “Linking citizen satisfaction with e-government and trust in government.” Journal of Public Administration Research and Theory, 15(3): 371–391. Whetsell, T.A., Kroll, A. and DeHart-Davis, L. 2021. “Formal hierarchies and informal networks: how organizational structure shapes information search in local government.” Journal of Public Administration Research and Theory, 31(4): 653–669. Wiersema, M.F. and Bantel, K.A. 1992. “Top management team demography and corporate strategic change.” Academy of Management Journal, 35(1): 91–121. Wu, Cary, Shi, Zhilei, Wilkes, Rima, Wu, Jiaji, Gong, Zhiwen, He, Nengkun and Zang Xiao et al. 2021. “Chinese citizen satisfaction with government performance during COVID-19.” Journal of Contemporary China, 30(132): 930–944. Zambrano, Julio and Avellaneda, Claudia N. 2022. “The (un)intended effects of regulatory burden on policy effectiveness: the case of regulations targeting NGOs.” Journal of Policy Studies, 37(1): 15–25.

1. Assessing the influence of Brazilian mayors’ human capital and political context on fiscal inputs Claudia N. Avellaneda and Marco Antonio Catussi Paschoalotto INTRODUCTION Worldwide federalism and decentralization have made local governments more relevant. The occurrence of natural disasters, health challenges, demographic shifts, security threats, among other things, have made local government performance essential. International financial institutions highlight the role of local governments in promoting accountability, responsiveness (Shah, 2006) and local economic development (World Bank Group, 2016), especially in developing countries. For local governments to handle both the unexpected challenges and to fulfill the assigned set of responsibilities, financial and human resources become crucial. In terms of financial resources, local governments rely extensively on central governmental transfers, making them heavily dependent on intergovernmental handovers (De Mello, 2000, Bahl and Johannes, 1994, Bergvall et al., 2006, Bradford and Oates, 1971, Hines and Thaler, 1995, Bailey and Connolly, 1998, Treisman, 1996, Masaki, 2018). Numerous calls have encouraged local governments to become more financially autonomous by increasing their own-collected monies through fees, foreign aid, as well as where legally allowed, sales and property tax collection (Freire and Garzόn, 2014). Despite these general calls, success in boosting local revenues through property tax collection varies considerably across localities, countries, and regions. This variation motivates this study to explore the drivers of municipal performance in terms of property tax collection or fiscal inputs. The existing literature on tax collection can be grouped into three major determinants: political factors (Buchanan and Lee, 1982, Alt and Lowry, 1994, Clingermayer and Wood, 1995, Lowry et al., 1998, Rubin, 2000, Amorim-Neto et al., 2001), socio-economic conditions (Smolka and De Cesare, 2006), and 16

Brazilian mayors’ human capital and political context on fiscal inputs

17

social and cultural norms (Hofstede, 1980, Smolka and De Cesare, 2006, Tsakumis et al., 2007, Bergman, 2009). Without denying their explicatory power, and as Petrovsky and Avellaneda (2014) suggest, these three sets of determinants of tax collection performance make up the organizational environment in which organizational and/or governmental leaders, managers, and chief executives operate when collecting taxes. Recent studies have started considering the role of individuals on financial and fiscal performance (Hayo and Neumeier, 2014, Von Hagen and Harden, 1995, Galasso and Nannicini, 2011). Therefore, in line with Petrovsky and Avellaneda (2014), this study adds to the understanding of tax collection and fiscal inputs by testing the role of chief executives’ human capital. That is, we draw insight from public management and organizational performance (Lynn, 1981, Meier and O’Toole, 2002, Hill and Lynn, 2003, Boyne, 2004, Andrews and Boyne, 2010, Walker et al., 2010) to investigate whether chief executives’ human capital contributes to explaining property tax collection. However, we deviate from Petrovsky and Avellaneda (2014) by arguing that the potential tax collection effects of chief executives’ human capital are conditioned by the political environment in which elected leaders operate for the following reasons. First, in most settings, chief executives must obtain the approval of the legislature to make any changes to the property tax rate, exceptions, extensions of deadlines, discounts for prompt payment, forgiveness of interest due, payment arrangements for past due taxes, and lottery drawings for punctual taxpayers (Avellaneda, 2009b, p. 473, see also Petrovsky and Avellaneda, 2014). Second, in addition to the legislature support, chief executives also rely heavily on political support from above. In fact, state/provincial and national political support may translate into financial benefits which, in turn, may condition a chief executive’s motivations to increase, maintain, or reduce a property tax rate and enforcement mechanisms. Third, in enforcing property tax collection, chief executives also need political support from below, that is, from citizens. Electoral support for a chief executive should condition a chief executive’s motivation to boost and enforce property tax collection. Under low margins of victory, a chief executive would hesitate to increase or enforce property taxes, but under a large margin of victory, the same chief executive may feel more compelled to do so. In summary, political support from above and below and the legislature moderates the property tax collection effects of leaders’ human capital. To test the direct and indirect effects of chief executives’ human capital on property tax collection, this chapter relies on data from the municipalities of the Brazilian state of São Paulo during 2009–2016, covering two municipal administrations. Brazilian municipalities constitute an interesting case. They enjoy a substantial degree of decentralization and make decisions over local taxes, and their mayors exercise both political and executive leadership in

18

What works in Latin American municipalities?

local government. In the analyzed data, results suggest neither mayoral human capital nor political support correlates with property tax collection. Moreover, none of the indicators capturing political support moderates the mayoral human capital-property tax collection relationship. Our study aims to make three contributions to research on fiscal performance at the local level. First, previous studies on fiscal performance have mainly relied on the direct effect of political and ideological factors (Alesina et al., 1996, Rogoff and Sibert, 1988, Barro, 1979, Buchanan and Wagner, 1977, Hibbs Jr., 1977, Roubini and Sachs, 1989). We test the direct influence of key political factors – mayor’s margin of victory, and party alignment between governors, president, councilors and the mayor. In addition, we also apply insights from human resource theory and public management performance (Lynn, 1981, Meier and O’Toole, 2002, Hill and Lynn, 2003, Boyne, 2004, Andrews and Boyne, 2010, Walker et al., 2010) to examine how mayors’ human capital influences property tax collection. That is, property tax collection is also explained with personal attributes of leaders. Second, as elected leaders govern within highly political environments, it becomes relevant to assess the conditioning effects of leaders’ political environment on property tax collection. The political environment of local governments is dynamic and complex. Political support for the mayor varies from below, above, and in legislatures. So, the interaction between a leader’s human capital and political environment should be accounted for to explain property tax collection. In this sense, our study contributes by applying an interactive model to predict property tax collection performance, while controlling for ideological, individual, and contextual factors. Third, as central government assigns more responsibilities to local governments, understanding the drivers of local government performance becomes crucial. This is true especially in settings where jurisdictional fragmentation has been the trend, increasing the number of local governments in the last two decades. That is the case of Brazil, as municipalities have increased in number from 4,491 in 1991, to 5,565 in 2010, and to 5,570 in 2020 (IBGE, 1991, IBGE, 2010, IBGE, 2020a). By focusing on Brazilian municipalities, this study also contributes to understanding property tax collection in a transitional economy with a non-consolidated democracy. The reminder of this chapter is structured as follows. In the next section, we briefly review the literature of municipal performance. Next, we address the role of leaders’ human capital and political factors to derive two sets of propositions to predict property tax collection performance. This is followed by the study’s research design, methodology, and statistical results. Then, we discuss the implication of our study for research on tax collection, list the study’s limitations, and conclude.

Brazilian mayors’ human capital and political context on fiscal inputs

19

LITERATURE REVIEW Assessing Municipal Performance and its Drivers in the Latin American Region With the adoption of fiscal, political, and administrative decentralization, local governments in unitary, and even federal, countries have become responsible for electing their leaders, as well as for implementing national legislation and delivering services. Their increasing role, along with governments’ greater willingness and capacity to collect administrative data, triggered interest and allowed for research on the drivers of municipal variation in performance. Assessing performance comes with challenges, especially in the public sector (Kellough, 2002, Murphy and Cleveland, 1995) and in developing settings (Avellaneda, 2009a). Some studies capture performance assessments with perceptive or subjective measures derived from citizens, employees, and/ or managerial evaluations. While some scholars advocate in favor of subjective measures (Singh et al., 2016), others raise concerns about their susceptibility to common source bias leading to spurious results (Meier and O’Toole, 2013). Moreover, in certain settings, managers and administrators are hesitant to document and report performance measures (Behn, 2002). In addition, not all objective measures are desirable. Given data availability, researchers may rely on expenditure measures. However, as Boyne (2002) posits, expenditures by themselves tell us little about performance unless they are linked to either quantity or quality of output. Therefore, this chapter contributes to the call for adopting objective rather than subjective measurements and by moving from expenditures measures to capture governmental performance through collecting property taxes and enforcing collection regulations. In Latin America, subnational government performance has been assessed in terms of tax collection (Avellaneda and Gomes, 2015, Petrovsky and Avellaneda, 2014); financial indicators (Avellaneda 2009b, 2012, Gomes et al., 2013); health care scores (Dantas et al., 2017, Olvera and Avellaneda, 2019, Ferraresi, 2021, De Oliveira et al., 2016, Paschoalotto et al., 2018, Ribeiro et al., 2018, Russo et al., 2019, Vargas et al., 2015); electricity and water services, educational achievements (Avellaneda, 2009a, Caetano et al. 2017, Queiroz et al., 2020, Rosa et al., 2019, Paschoalotto et al., 2020; Santos et al., 2020); and environmental indicators (Dantas et al., 2017, De la Riva 2021). Findings from existing empirical studies reveal that no single factor matters for subnational governmental performance. At the risk of simplification, the performance drivers can be grouped into political, contextual, individual, external, and socio-economic categories. Among the political factors, studies

20

What works in Latin American municipalities?

have tested the performance effects of partisan alignment, electoral cycle, margin of victory, and party ideology (Avellaneda 2009a, 2009b, 2012, 2014, 2016, Petrovsky and Avellaneda, 2014, Gomes and Avellaneda, 2017, Avellaneda and Gomes, 2015). Among them, electoral cycle has shown explicatory power because local governments tend to increase performance, especially in terms of service delivery in years prior to elections (Avellaneda 2009a, 2016). In the Latin American context, other studies have explored the role of party alignment in explaining subnational governmental performance (Avellaneda 2009a, 2009b, Avellaneda and Gomes, 2015, 2017, Gomes et al., 2013). For instance, party alignment between local government and upper-level governments appears to contribute to higher subnational revenues (Avellaneda, 2012, 2016). On the contrary, party ideology fails to explain municipal performance. This null result may be due to the multi-party systems that characterize many countries in the region. Among the contextual drivers, several studies have tested the performance effects of socio-demographic factors (e.g., Avellaneda, 2012; Paschoalotto et al., 2020; Puppim de Oliveira, 2017; Rosa et al., 2019). Some of the existing works suggest that population size, rurality, violence, and population density matter for social service delivery (Avellaneda 2009a, 2009b, Avellaneda and Gomes, 2015, Ferraz et al., 2012, Puppim de Oliveira, 2017, Santos et al., 2020, Zucco, 2013, Queiroz et al., 2020). The effects of population density should be analyzed more carefully because municipal investment in electrical towers, sewage services, schools and health centers locations are greatly determined by economies of scale. In addition to the above drivers, considerable attention has been devoted to assess the performance effects of chief executives’ human capital, qualifications, or background (e.g., Avellaneda 2009a, 2009b, 2012, 201, 2016, Petrovsky and Avellaneda, 2014, Gomes and Avellaneda, 2017). The considerable variation in elected leaders’ professions and backgrounds explains scholars’ interest. In fact, mayors come into office with a wide range of educational and background experiences. Some have been farmers, teachers, drivers, artists, priests, dentists, physicians, business persons, engineers, and other occupations. The performance effects of mayoral background have been tested across several policy areas, such as education, health, environment, finances, and infrastructure (electricity and sewage). In the Brazilian public health sector, for example, studies report the background of health workforce is negatively correlated with child and infant mortality rate (Dantas et al., 2017, De Oliveira et al., 2016, Moosa et al., 2017, Paschoalotto et al., 2020, Paschoalotto et al., 2018, Vargas et al., 2015). The positive impact of leaders’ education and experience on health indicators (e.g., primary health care and infant mortality rate) seems greater in small municipalities (Moosa et al., 2017, Olvera and Avellaneda, 2019, Russo et al., 2019).

Brazilian mayors’ human capital and political context on fiscal inputs

21

Other studies have looked specifically at the role of leaders’ education and experience on expenditures (Avellaneda, 2009b) and tax collection (Petrovsky and Avellaneda, 2014, Avellaneda, 2009b). According to these studies, leaders’ education and job-related experience positively correlate with property tax collection (Avellaneda, 2009b, Petrovsky and Avellaneda, 2014, Avellaneda and Gomes, 2015), holding all other factors constant. However, these studies assume leaders’ qualifications operate in a vacuum when, in reality, leaders must navigate the political environment of their respective jurisdictions in order to put into action their skills and knowledge. For this reason, this analysis focuses on the direct role of leaders’ human capital, as well as on the mediating role of political environment on the human capital-performance relationship. Human Capital Attributes and Tax Collection Performance If the identity of leading politicians matters, “selecting good politicians becomes ever more crucial” (Galasso and Nannicini, 2011, p. 79).1 The notion is that “voters prefer good and clean policies so that candidates of higher quality have higher chances of election than candidates of lower quality” (Caselli and Morelli, 2004, p. 759). Moreover, “a central role of (successful) political institutions is to ensure the selection of the right (honest, competent, motivated) politician” (Acemoglu et al., 2010, p. 1511). Since politicians’ attributes are expected to influence their electoral success, it becomes relevant to assess whether the individual characteristics of leading politicians affect their policy choices, financial decisions, and fiscal performance. Caselli and Morelli (2004) suggest the quality of public officials involves two dimensions – competence and honesty – which vary considerably across countries’ leaders. Competence refers to the expertise in doing something well and the skills and knowledge specific to a domain (Chi et al., 2014, Ericsson and Smith, 1991). Competence permits politicians to recognize proper policy objectives and achieve lower social costs (Caselli and Morelli, 2004, p. 759). We propose that education, professional background, and job-related experience should proxy individual competence and, in turn, affect fiscal performance. Several theoretical underpinnings drive our propositions. First, human capital theory (Mincer, 1958) postulates that investments in higher formal education go along with higher productivity on the job. Second, the signaling effect hypothesis suggests that formal education signals higher ability (Spence, 1973). Bowman and Myers (1967) suggested that schooling and work experience are determinant factors of human capital. Barney (1991, p. 101) indicates that human capital contributes to firms’ competitive advantages. Moreover, the human capital theory also has been applied to the mayoral ability to increase

22

What works in Latin American municipalities?

fiscal capacity and independence in Colombian municipalities to achieve fiscal sustainability (Avellaneda, 2009a, 2009b). Other theoretical underpinnings also justify our propositions. Hambrick and Mason’s (1984) upper echelons theory assumes “[o]rganizational outcomes are viewed as reflections of the values and cognitive bases of powerful actors in the organization” (Hambrick and Mason, 1984, p. 193). Therefore, executives’ behavior and interpretations of situations are a function of their values, experiences, and personalities (Finkelstein et al., 1996, Hambrick, 2007, Hambrick and Mason, 1984). That is, when leading politicians are presented with the same problem, they define the problem and recommend solutions largely in terms of goals and activities in their areas (Dearborn and Simon, 1958) and industry (Hambrick and Mason, 1984). For instance, public managers from the private sector may adopt more business-oriented strategies (Avellaneda, 2016). Miles et al.’s (1978) typology of strategies also draws on the professional background of managers. Finally, Fiedler's (1986, p. 32) cognitive resource theory helps us understand the mechanisms through which job-relevant experience contributes to leader/manager performance by: (1) providing job-related knowledge, (2) enhancing managers’ ability to cope with stressful conditions, and (3) enhancing greater self-confidence and control of leadership situations. Several studies have documented the correlation between leaders’ education and professional background, and governmental performance. For instance, Besley et al. (2011) and Congleton and Zhang (2013) show that more educated political leaders can stimulate economic growth. Dreher et al. (2009) and Freier and Thomasius (2016) provide evidence that the professional background of the head of government influences the implementation of market-liberalizing reforms and the level of expenditures and public debt. Similarly, Göhlmann and Vaubel (2007), Farvaque et al. (2009), and Farvaque et al. (2011) demonstrate that decision-makers’ educational and professional background is relevant to inflation and fiscal performance. Likewise, Avellaneda and Gomes (2015) and Avellaneda (2009a) found a positive relationship between mayoral experience and financial performance regarding local tax collection. Like education and professional background, specific job-related experience, that is, expertise, also should add to leading politicians’ competence in managing fiscal policy. For instance, Jochimsen and Thomasius (2014) argue that long-serving finance ministers achieve lower deficits and provide empirical evidence for German states. Likewise, Freier and Thomasius (2016) found that mayors with prior experience in office, that is with some level of expertise, tend to reduce the level of local public debt, lower total municipal expenditures, and decrease local taxes. Yet, these results are contingent on model specification. The above discussion leads to:

Brazilian mayors’ human capital and political context on fiscal inputs

23

H1: Municipal tax property collection is positively associated with mayoral human capital (e.g., education and job-related experience).

POLITICAL CONTEXT AS A MODERATOR OF THE PERFORMANCE-HUMAN CAPITAL RELATIONSHIP Lynn et al. (2000) and Forbes and Lynn (2005) highlight the influence that external factors have on organizational performance. Likewise, O’Toole and Meier (2017) offer a framework that takes into account the moderating role of context on performance. By context, O’Toole and Meier (2017) refer to the political, environmental, and internal context. Their framework seeks to settle inconsistent results across and within countries in terms of the effects of management factors on performance. The moderating effect of some of these contextual variables was illustrated with several country analyses and across different policy areas (Meier, Rutherford and Avellaneda, 2017). In doing so, they identify a set of variables that condition the impact of management in an interactive model. Without denying the role of environmental and internal moderating factors, here we seek to assess the conditioning effect of the political context on the mayoral human capital-performance relationship. Within the political context, O’Toole and Meier’s (2017) framework includes whether a governing system is unitary versus shared powers, singleor multiple-level, corporatist versus adversarial, with or without a formal performance appraisal system. However, other variables count as political context, such as leaders’ political support from (a) above (higher levels of government), (b) below (citizens through margin of victory), and from (c) the respective jurisdictional legislators. Therefore, in addition to managerial human capital, the political context in which managers or chief executives perform is expected to influence their organizations’ performance. Political support from citizens, for example, potentially conditions leaders’ human capital effect on performance. Specifically, a mayor’s experience may suggest a leader who supports the benefits and need for increasing property tax rates. However, the electorally competitive context that led to the mayor’s narrow victory also indicates that increasing property tax rates may result in losing supporters. In this case, it makes sense rationally to increase the tax rate, but strategically it entails political risk. The same logic applies to enjoying partisan alignment in higher levels of governments. For one, intergovernmental partisan support may result in more resources and money transfers, discouraging local leaders from enforcing tax collection or increasing tax rates. Two working hypotheses address the co-partisan-intergovernmental grants relationship. One considers the positive political effect that grant acquisition may have on jurisdictions with a high number of swing voters (Lindbeck and Weibull, 1987, Dixit and Londregan, 1998). The alternative hypothesis sug-

24

What works in Latin American municipalities?

gests that, due to risk aversion, grants tend to be allocated to politically aligned jurisdictions (Cox and McCubbins, 1986). Even if a mayor enjoys high voter support and no partisan alignment at the national or provincial level, the mayor’s plans for enforcing tax collection or increasing tax rates must be approved by local legislators. Consequently, partisan alignment with the city council should moderate mayors’ human capital effects on performance. This discussion leads us to the following interactive hypotheses: H2: Political context moderates the performance-mayoral human capital relationship. H2a: The higher the mayor’s margin of victory, the higher the effect of mayoral human capital on property tax collection. H2b: The effect of mayoral human capital on property tax collection declines under mayoral partisan alignment with either or both governor and president. H2c: The higher the percentage of legislators party-aligned with the mayor’s party, the higher the effect of mayoral human capital on property tax collection.

CASE SELECTION Brazilian Municipalities: The Case of São Paulo As a federation, Brazil is organized into three independent and autonomous government levels: municipalities, states, and the federal government. The three levels have executive, legislative, and judiciary branches, except for municipalities, where the judiciary is shared with state governments. Brazil has 26 states, a federal district, and 5,570 municipalities. Until January 2021, the Superior Electoral Tribunal (TSE)2 reported 33 political parties registered across the three levels of government. The multi-party system encourages political parties to create coalitions to promote a candidacy regardless of ideology (Cheibub et al., 2004). With more than 200 million people (IBGE, 2010), Brazil exhibits large income, health, educational, and social disparities (Haddad et al., 2017; Paschoalotto et al., 2018), making it the Latin American country with the largest GINI coefficient (IBGE, 2020b). The mayor is elected for four years and is eligible for one direct re-election, after which the mayor must wait one term before running again (Federal Constitution of Brazil, 1988).

Brazilian mayors’ human capital and political context on fiscal inputs

25

Tax Collection in Brazilian Municipalities According to the 1988 Federal Constitution, municipalities are entitled to pass legislation on local taxes, namely property, services, and property sale taxes. Small municipalities, however, lack power to leverage revenues, making about 70% of Brazilian municipalities heavily dependent on state and federal transfers (De Mello, 2000, 2002, Valle and Gomes, 2014). In fact, Brazil has 4,897 (88% of the total of 5,570) cities with less than 50,000 inhabitants (IBGE, 2013). Local property tax is unpopular and characterized by low enforcement and tax avoidance. Besides intergovernmental transfers stipulated by the National Tax Code (IBGE, 2012), municipalities derive revenues from: (a) property tax collection; (b) royalties derived from the extraction of natural resources, such as minerals, oil, and gas; (c) other services (e.g., cemetery space, public transportation, and public spaces for commercial ends); and (d) fines on residents and businesses due to violations. Municipalities also may receive block grants and matching grants from upper government levels (Martell, 2008). Brazilian mayors come to office with diverse (if any) professional training and various sectorial backgrounds. A strong mayoral system makes mayors the political and administrative leaders, as the office of city manager does not exist. Municipalities with more than 200,000 inhabitants may hold a run-off election if no candidate wins an outright majority. Local elections take place in November, and the winners take office on the next January 1, meaning that yearly statistics match a mayoral administration year. Brazilian localities are multi-service providers. Municipalities are responsible for providing basic education, health services (in conjunction with the state and federal governments), recreational, and developmental services. Mayors do so with scarce resources. That is why identifying the drivers of property tax collections may shed light on how to boost municipalities’ coffers.

DATA: VARIABLE DEFINITION AND OPERATIONALIZATION This analysis relies on the 645 municipalities of the state of São Paulo. São Paulo is the Brazilian state with the wealthiest economy and second-highest number of municipalities (IBGE, 2010) (see Figures 1.1 and 1.2). São Paulo has an organization named SEADE Foundation, which gathers different data at the municipal level, including mayors’ background information (SEADE Foundation, 2016). Our dataset includes eight years, covering two mayoral administrations from 2009 to 2012 and 2013 to 2016. Therefore, our final dataset consists of a balanced panel data with 5,160 observations. The unit of analysis is municipality year.

26

What works in Latin American municipalities?

Source: SEADE Foundation (2016).

Figure 1.1

Brazil (São Paulo State highlighted) and São Paulo State (645 municipalities)

Source: Authors.

Figure 1.2

Dependent variable: logged municipal tax collection

DEPENDENT VARIABLE – MUNICIPAL PROPERTY TAX COLLECTION We capture municipal performance through municipal property tax collection per year, reported in 1,000 Brazilian reais. The Brazilian National Treasury

Brazilian mayors’ human capital and political context on fiscal inputs

27

Secretary (SICONFI) collects annual data for all the Brazilian municipalities. Actual values are logged to reduce skewness in data values (see Figure 1.2). However, three different measures assess municipal performance in collecting property tax collection. First, we calculated the ratio of property tax collected over the total municipal revenues. In other words, this measure captures the proportion of municipal revenue that derives from property tax collection. This measure makes comparison across units more uniform. It is a continuous variable. The values for the total municipal revenue come from the Brazilian National Treasury Secretary (SICONFI). The second measure assessing property tax collection refers to the total property tax collected per municipal-year reported in 1,000 Brazilian reais. We do not divide this value by population, but we do control for population in the analysis (see below control variables). The third indicator to capture property tax collection is the ratio between the estimated property tax to be collected and the actual property tax collected. When crafting the annual budget, municipal officials calculate the estimated monies to be collected out of property taxes. Therefore, the ratio between the estimated and the actual values also should capture municipal performance in property tax collection. The three dependent variables are continuous. See Tables 1.1a and 1.1b for description and summary statistics of all variables. Independent Variables Data for our key independent and control variables come from several sources, such as (a) Brazilian and São Paulo state official databases: DATASUS – SUS Informatics Department; (b) the IBGE, the Brazilian Institute of Geography and Statistics; (c) municipalities’ websites; (d) the SEADE Foundation; and (e) Brazilian National Treasury Secretary. Two indicators capture mayors’ human capital – years of education and years of local government experience. These two individual-level variables are continuous and were collected from the SEADE Foundation. In our database, 0.5% of mayors have achieved only four years of education, the minimum value, while 0.2% of mayors have achieved 18 years of education, the maximum value. Likewise, 44.2% of mayors came to office without experience in local government. The second set of independent variables captures political support that mayors have from above, below, and the local council. The continuous variable labeled margin of victory refers to the difference in the proportion of votes between the mayor and the second runner-up. These values come from the Superior Electoral Tribunal (TSE). Two dummy variables reveal whether the mayor’s party is aligned to the governor’s or the president’s party. Finally, a fourth variable measures the proportion of council members aligned with

Which traditional political party the mayor aligns with? Arena or MDB

IBGE – Brazilian Institute of Geography and Statistics

Mayor’s gender (dummy) Number of residents in the municipality during the calculation period (logged)

Population (logged)

SEADE Foundation

Mayor’s private sector experience? (dummy)

Female mayor

SEADE Foundation

TSE – Supreme Electoral Court

TSE – Supreme Electoral Court

Municipalities’ councils’ websites

TSE – Supreme Electoral Court

TSE – Supreme Electoral Court

TSE – Supreme Electoral Court

SEADE Foundation

SEADE Foundation

Source

Mayor’s private sector experience

(social) (dummy)

Did mayor get re-elected? (dummy)

Mayor’s Social Democratic Party

Proportion of municipality´s councils aligned with mayor’s party (ratio)

(dummy)

Has the mayor aligned with the president’s (PT) political party?

(dummy)

Has the mayor aligned with the governor’s (PSDB) political party?

percentage of votes

Mayor’s winning percentage of votes less the second candidate’s

Number of years working in the local level (municipality)

master-18)

Number of years studying (elementary-9; high school-12; college-16;

Re-elected mayor

Controls

Mayor-council party alignment

Mayor-president party alignment

Mayor-governor party alignment

Mayor’s margin of victory

Mayors’ political power

Mayor’s years of local experience

Mayor’s years of education

Description

Description and source of the independent variables

Mayors’ human capital

Variables

Table 1.1a

28 What works in Latin American municipalities?

Number of urban residents in the municipality during calculation period

Urbanization rate

Percentage of added value in the GDP by the public administration (%) Child mortality rate for each 1,000 inhabitants

Municipality contribution to GDP

Child mortality rate

calculation period (ratio)

divided by the number of residents in the municipality during the

Description

Variables

Department

DATASUS – SUS Informatics

National Treasury Secretary

SEADE Foundation

Source

Brazilian mayors’ human capital and political context on fiscal inputs 29

423336.2 3.154766

Ratio estimate property tax collected/actual property tax collected

2.72955

Mayor’s years of local gov. experience

.2941292 .1052838 22.45207

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Proportion of council members aligned with mayor’s party .5152642 .090411 .3432485 .2757339 .446927 85.09436 25090.65 11.6075

Mayor’s private sector experience

Female mayor

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Controls

24.9174

Margin of victory

Political support

14.17808

Mayor’s years of education

Mayors’ human capital

.8874656

Mean

Total property tax collected

Summary statistics

Ratio of property tax collected/total revenues

Dependent variables

Table 1.1b

12.60148

23196.98

14.04939

460739.3

.446927

.4748401

.2867977

.4998159

12.64205

.3069487

.4556949

25.7356

2.747701

2.619915

18.34145

4141022

.1200723

S.D.

0

4856.374

24.57

803

0

0

0

0

0

0

0

0

0

4

.0001103

30.30443

0

Min.

214.3

401303.9

100

1.16e+07

1

1

1

1

100

1

1

100

12

18

419.1317

1.17e+08

1

Max.

30 What works in Latin American municipalities?

Brazilian mayors’ human capital and political context on fiscal inputs

31

the mayor’s party. Data on council members’ party affiliation come from the Superior Electoral Tribunal (TSE).

CONTROL VARIABLES The study controls for other individual and contextual factors expected to influence property-tax collection. At the mayor’s level, we control for four individual factors. First, a dummy variable receives “1” whether the mayor has any private-sector experience prior to being elected. Second, a dummy variable captures the gender of the mayor, getting “1” if female; otherwise, “0.” Third, re-elected mayors are likely to have more knowledge and experience in terms of municipal finances. Consequently, a dummy variable gets “1” if the mayor is re-elected; otherwise, “0.” Finally, the analysis also controls for the mayor’s party. Given Brazil’s multi-party system, party categorization on the left-right continuum is difficult. We rely on Carreirão’s (2006) party classification, used by Avellaneda and Gomes (2015) as well. According to Carreirão (2006), the following political parties DEM, PR, PP, PFL, PRN, PDC, PL, PTB, PSC, PSP, PRP, PSL, PSDB, and PRONA represent the right wing. PMDB and PSDB are regarded as central, but because of their conservative orientation, we also considered them part of the rightist category. The PHS, PPS, PRB, PRTB, PSD, PSDC, PT DO B, PTC, and PTN are considered centrist parties. The PT (President Lula’s Party), PDT, PPS, PCdoB, PSB, PV, PSTU, PCO, and PMN are regarded as left-wing parties. For our analysis, we created a dummy variable labeled mayor’s leftist party, receiving “1” if the party of the mayor belongs to one of the leftist parties; otherwise, “0.” Finally, the analysis also controls for contextual variables at the municipal level, such as (a) population, which is logged to reduce skewness; (b) urbanization rate reported in percentages, (c) GDP/capita in Brazilian Reais, and (d) infant mortality rate per 1,000 inhabitants to capture municipal development. All four contextual variables are continuous.

ANALYSIS AND RESULTS The VIF for the regression models ranges from 1.23 (for the lineal models) to 1.31 (for the models with interaction terms). Although some of the predictor variables are multiplied to create interaction terms, the main effects are not closely correlated with their interaction terms, mainly because the linear predictors were centered. Given the data structure, a balanced panel dataset, we report random-effect estimation to explain across group variation, a fixed-effect estimation to explain in-group variation and control for time-invariant factors, and an Arellano-Bond estimation to include a lag of the dependent variable as property tax collection in year t-1 may determine

32

What works in Latin American municipalities?

enforcement of tax collection in year t. Consistency of findings across models should serve as robustness checks. Table 1.2 reports the random-effect, fixed-effect, and Arellano-Bond estimates for the lineal influences of mayor’s human capital and political support on the ratio of property tax collected over total municipal revenues. For each model, we ran the influence and leverage diagnostics, leading us to remove the municipality-year observations that influence estimations. After removing influential observations and missing values, the total number of observations decreases to 5,110. According to Table 1.2, results are consistent across the three lineal estimations. None of the two indicators of mayor’s human capital shows statistical significance. Consequently, H1 receives no empirical support with the data analyzed here. Although we did not propose lineal effect for the indicators of partisan support, the inclusion of their lineal coefficients is required to test their interactive effects. Only one of the four variables capturing political support reports statistical significance both in the random-effect and fixed-effect model. Specifically, the coefficient on mayor’s party alignment with governor’s party is negative and statistically significant at the 5% level. That is, in municipalities whose mayor’s party is aligned with the governor’s party, the ratio of property tax collected over the total municipal revenues is lower. From the control variables, the coefficient on mayor’s private sector experience is negative and statistically significant at the 5% level, but only at the fixed-effect model. In addition, the coefficient on urbanization rate is negative and statistically significant in the random-effect (at 5% level) and fixed-effect (at 1% level) model. Based on Table 1.3, the absence of significant results is consistent across the three lineal estimations that explain the second dependent variables: total property tax collected (Brazilian reais). The two indicators of mayor’s human capital fail to gain statistical significance. Consequently, H1 receives no empirical support with the data analyzed here. None of the four variables capturing political support reports statistical significance. From the control variables, the coefficient on GDP/capita is positive and statistically significant at the 5% level in the random- and fixed-effect model. In explaining our third dependent variable, the ratio of actual property tax collected over estimated property tax collected, results from Table 1.4 offer no significant effects across the three models. The two indicators of mayor’s human capital show no statistical significance. Consequently, H1 receives no empirical support with the data analyzed here. None of the four variables capturing political support reports statistical significance. From the control variables, the coefficient on urbanization rate is negative and statistically significant in the random-effect (at 5% level) and fixed-effect (at 1% level) model.

-.000 (000)

Mayor’s years of local gov. experience

-.011 (.005)** .005 (.008) -.000 (.000)

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Council members aligned with mayor’s party .004 (.004) -.000 (.009) -.004 (.004) -.007 (.005) .002 (.002) -.000 (.000)** 4.13e-07 (7.98e-08)*** -.000 (.000) .905 (.025)*** 5,110 645 No

Mayor’s private sector experience

Female mayor

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Constant

Number of observations

Number of municipalities

Year dummy

Controls

.000 (.000)

Margin of victory

Political Support

.000 (.000)

Mayor’s years of education

Mayors’ human capital

Random-effect Model

Yes

645

5,110

.708 (.110)***

-.000 (.000)

3.99e-08 (5.55e-08)

.002 (.001)*

-.004 (.006)

-.004 (.006)

-.002 (.004)

.003 (.012)

.010 (.004)**

-.000 (000)

-.002 (.011)

-.013 (.005)**

.000 (.000)

-.000 (000)

.001 (.001)

Fixed-effect Model

Yes

641

3,800

6.825 (5.084)

-.000 (.001)

-1.39e-06 (1.19e-06)

-.017 (.018)

-.454 (.362)

-.015 (.013)

.030 (.018)

.000 (.026)

.037 (.026)

.000 (.000)

.014 (.023)

.003 (.017)

-.000 (.000)

-.003 (.003)

-.000 (.003)

Arellano-Bond Model

Panel data estimates for the ratio of property tax collected over total municipal revenues, 2009–2016

Name of the independent variables

Table 1.2

Brazilian mayors’ human capital and political context on fiscal inputs 33

No 1.23

Municipality dummy

Variance inflation factor

Notes: ***p < .001; **p < .05; *p < .01 (one-tailed test).

Random-effect Model

Name of the independent variables

Fixed-effect Model 1.23

Yes

Arellano-Bond Model 1.23

Yes

34 What works in Latin American municipalities?

-14979.17 (13290.24)

Mayor’s years of local gov. experience

227007.3 (243535.9) 137.034 (1155.926)

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Council members aligned with mayor’s party -18052.48 (29717.08) -9836.921 (22000.47) 32748.27 (50243.79) 58975.47 (50924.89) 42164.38 (40931.77) 3052.138 (6019.294) 6.186 (2.240)** -22.885 (148.370) -380301.7 (733575) 5,110 645 No

Mayor’s private sector experience

Female mayor

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Constant

Number of observations

Number of municipalities

Year dummy

Controls

-154.625 (457.643) 37933.87 (49195.88)

Margin of victory

Political support

-2260.057 (2360.399)

Mayor’s years of education

Mayors’ human capital

Random-effect Model

Yes

645

5,110

1350117 (638862.2)**

-28.767 (141.664)

6.308 (2.304)**

-10302.3 (7164.169)

-18375.14 (16134.26)

63515.63 (53774.46)

26508.59 (51033.45)

-5839.733 (21282.88)

-28038.28 (29901.41)

15.94919 (1112.883)

218604.7 (239003.1)

43446.3 (53052.33)

-65.256 (432.297)

-13720.09 (13109.9)

-2765.998 (2378.849)

Fixed-effect Model

Yes

641

3,800

376881.7 (344446.5)

76.434 (45.930)*

1.429589 (.9748857)

-4878.991 (3891.584)

12207.85 (15466.43)

-13326.51 (12153.84)

4456.285 (6746.271)

-10058.73 (9571.142)

2361.018 (8448.866)

314.135 (251.168)

22377.57 (21340.86)

2936.32 (8599.268)

-11.761 (154.326)

-1487.407 (1024.124)

1164.754 (683.451)*

Arellano-Bond Model

Panel data estimates for the total property tax collected (reais) in São Paulo municipalities, 2009–2016

Name of the independent variables

Table 1.3

Brazilian mayors’ human capital and political context on fiscal inputs 35

No 1.23

Municipality dummy

Variance inflation factor

Note: ***p < .001; **p < .05; *p < .01 (one-tailed test).

Random-effect Model

Name of the independent variables

Fixed-effect Model 1.23

Yes

Arellano-Bond Model 1.23

Yes

36 What works in Latin American municipalities?

-.038 (.057)

Mayor’s years of local gov. experience

.370 (.542) -.280 (.400) .018 (.035)

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Council members aligned with mayor’s party .654 (.557) -1.250 (1.583) .581 (.485) .221 (.545) 2.815 (1.829) -.308 (.145)** 8.11e-06 (6.60e-06) -.001 (.002) .960 (9.283) 5,110 645

Mayor’s private sector experience

Female mayor

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Constant

Number of observations

Number of municipalities

Controls

-.006 (.004)

Margin of victory

Political support

.028 (.149)

Mayor’s years of education

Mayors’ human capital

Random-effect Model

645

5,110

9.135 (12.303)

-.000 (.002)

9.95e-06 (6.65e-06)

-.453 (.264)*

3.183 (2.298)

.219 (.549)

.509 (.517)

-1.409 (1.683)

.707 (.600)

.015 (.036)

-.079 (.399)

.369 (.534)

-.005 (.004)

-.032 (.062)

.063 (.161)

Fixed-effect Model

641

3,800

-45.592 (25.889)*

.000 (.002)

-3.11e-06 (6.95e-06)

-.240 (.152)

7.310 (3.952)

-.429 (.680)

.811 (.585)

.266 (.416)

.679 (.528)

.000(.023)

.296 (.553)

.074 (.339)

-.004 (.003)

-.068 (.055)

-.146 (.164)

Arellano-Bond Model

Panel data estimates for the ratio of actual property tax collected over estimated property tax collected, 2009–2016

Name of the independent variables

Table 1.4

Brazilian mayors’ human capital and political context on fiscal inputs 37

no no 1.23

Year dummy

Municipality dummy

Variance inflation factor

Note: ***p < .001; **p < .05; *p < .01 (one-tailed test).

Random-effect Model

Name of the independent variables

Fixed-effect Model

1.23

yes

yes

Arellano-Bond Model

1.23

yes

yes

38 What works in Latin American municipalities?

-.000 (.000)

Mayor’s years of local gov. experience

-.011 (.005)** .008 (.008) .000 (.000)

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Council members aligned with mayor’s party

.001 (.002) -.001 (.002) 8.41e-06 (.000) .000 (.000)* .000 (.001) .001 (.002) .000 (.000)

Education*aligned with governor’s party

Education*aligned with president’s party

Education*aligned with council members’ party

Local exp*margin of victory

Local exp*aligned with governor’s party

Local exp*aligned with president’s party

Local exp*aligned with council members’ party

.005 (.004) -.000 (.009)

Mayor’s private sector experience

Female mayor

Controls

8.64e-06 (.000)

Education*margin of victory

Interaction effects

.000 (.000)

Margin of victory

Political support

-.000 (.000)

Mayor’s years of education

Mayors’ human capital

Random-effect Model

.004 (.012)

.010 (.004)**

.000 (.000)

.001 (.002)

-.000 (.001)

.000 (.000)

.000 (.000)

-.004 (.003)

.000 (.002)

4.73e-07 (.000)

-.000 (.000)

.002 (.012)

-.013 (.005)**

.000 (.000)

-.000 (.000)

.001 (.001)

Fixed-effect Model

-.001 (.028)

.036 (.026)

.000 (.000)

.001 (.009)

.001 (.005)

.000 (.000)

-.000 (.000)

.003 (.010)

-.004 (.006)

-2.27e-06 (.000)

-.000 (.000)

.009 (.029)

.006 (.019)

-.000 (.000)

-.003 (.003)

.002 (.003)

Arellano-Bond Model

Panel data estimates with interaction effects for the ratio of property tax collected over total municipal revenues, 2009–2016

Name of the independent variables

Table 1.5

Brazilian mayors’ human capital and political context on fiscal inputs 39

-.004 (.004) -.007 (.005) .002 (.002) -.000 (.000)** 4.10e-07 (8.09e-08)*** -.000 (.000) .907 (.022)*** 5,110 645 No No 1.31

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Constant

Number of observations

Number of municipalities

Year dummy

Municipality dummy

Variance inflation factor

Note: ***p < .001; **p < .05; *p < .01 (one-tailed test).

Random-effect Model

Name of the independent variables

Fixed-effect Model

1.31

Yes

Yes

645

5,110

.694 (.111)***

-.000 (.000)

3.88e-08 (5.65e-08)

.002 (.001)**

-.003 (.006)

-.005 (.007)

-.003 (.004)

Arellano-Bond Model

1.31

Yes

Yes

641

3,800

6.895 (5.160)

-.000 (.001)

-1.39e-06 (1.21e-06)

-.018 (.020)

-.454 (.363)

-.014 (.014)

.035 (.023)

40 What works in Latin American municipalities?

Brazilian mayors’ human capital and political context on fiscal inputs

41

To test the hypothesized interaction effects (H2, H2a, H2b and H2c), Table 1.5 presents the three sets of estimates for the first dependent variable: the ratio of property tax collected over the total municipal revenues. In line with results from the lineal model (Table 1.2), the two indicators of mayor’s human capital report no statistical significance. Consequently, H1 receives no empirical support with the data analyzed here. Likewise, only one of the four variables capturing political support reports statistical significance both in the random-effect and fixed-effect model. Specifically, the coefficient on mayor’s party alignment with governor’s party is negative and statistically significant at the 5% level. That is, in municipalities whose mayor’s party is aligned with their governor’s party, the ratio of property tax collected over total municipal revenues is lower. None of the interaction terms shows statistical significance, except for the coefficient on the interactive terms between local experience* margin of victory, which is statistically significant but only at the 10% level in the random-effect model. The lack of statistical significance across the interactive terms leads us to reject H2, H2a, H2b, and H2c. From the control variables, the coefficient on mayor’s private sector experience is negative and statistically significant at the 5% level but only at the fixed-effect model. In addition, the coefficient on urbanization rate is negative and statistically significant at the 5% level both in the random-effect and fixed-effect model. Based on Table 1.6, absence of significant results exists across the three nonlinear estimations that explain the second dependent variables – total property tax collected (Brazilian reais). The two indicators of mayor’s human capital fail to gain statistical significance. Consequently, H1 receives no empirical support with the data analyzed here. None of the four variables capturing political support nor any of the eight interactive terms reports statistical significance. Consequently, H2, H2a, H2b, and H2c fail to receive empirical evidence. From the control variables, the coefficient on GDP/capita is positive and statistically significant at the 5% level in the random- and fixed-effect model. In explaining our third dependent variable, the ratio of actual property tax collected over estimated property tax collected, results from Table 1.7 offer no significant effects across the three nonlinear models. The two indicators of mayor’s human capital show no statistical significance. Consequently, H1 receives no empirical support with the data analyzed here. Similarly, none of the four variables capturing political support or any of the eight interaction terms reports statistical significance. Therefore, H2, H2a, H2b, and H2c receive no empirical evidence. From the control variables, the coefficient on urbanization rate is negative and statistically significant in the random-effect (at 5% level) and fixed-effect (at 1% level) model.

-9126.051 (11590.14)

Mayor’s years of local gov. experience

33434.65 (45810.11) 149049.6 (173988.5) 1450.85 (2014.794)

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Council members aligned with mayor’s party

2477.545 (12465.93) 61678.6 (71275.99) 791.981 (604.222) 21.773 (195.041) -10726.91 (6016.509)* -61119.1 (52162.84) 396.853 (317.327)

Education*aligned with governor’s party

Education*aligned with president’s party

Education*aligned with council members’ party

Local exp*margin of victory

Local exp*aligned with governor’s party

Local exp*aligned with president’s party

Local exp*aligned with councilmen’s party

-32533.59 (31969.43) -3668.535 (24100.1)

Mayor’s private sector experience

Female mayor

Controls

-149.120 (129.323)

Education*margin of victory

Interaction effects

-373.077 (729.438)

Margin of victory

Political support

-10438.18 (10010.96)

Mayor’s years of education

Mayors’ human capital

Random-effect Model

701.413 (23456.57)

-42191.44 (34356.99)

409.761 (322.416)

-58747.09 (50444.03)

-8996.728 (6025.779)

8.355 (187.400)

788.226 (612.995)

61300.18 (71763.71)

2865.627 (13059.24)

-152.653 (130.352)

1523.186 (2027.966)

142789.3 (171154.4)

38416.03 (49497)

-285.563 (695.373)

-8669.263 (11747.89)

-11647.6 (10787.5)

Fixed-effect Model

-12784.84 (8830.263)

711.592 (8637.535)

12.262 (59.783)

-4727.812 (7807.135)

1376.485 (3770.99)

-53.317 (39.162)

94.972 (80.019)

10403.27 (6293.986)*

2931.098 (2720.205)

38.862 (35.390)

294.011 (316.260)

13579.14 (21074.55)

2283.847 (8235.263)

19.578 (170.957)

-2069.462 (1344.311)

197.952 (1200.844)

Arellano-Bond Model

Panel data estimates with interaction effects for the total property tax collected (reais) in São Paulo municipalities, 2009–2016

Name of the independent variables

Table 1.6

42 What works in Latin American municipalities?

44058.82 (59911.15) 59395.03 (53046.65) 40197.84 (38980.5) 3547.173 (6013.466) 6.251 (2.228)** 24.282 (156.803) -474458.2 (727391) 5,110 645 No No 1.31

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Constant

Number of observations

Number of municipalities

Year dummy

Municipality dummy

Variance inflation factor

Note: ***p < .001; **p < .05; *p < .01 (one-tailed test).

Random-effect Model

Name of the independent variables

1.31

Yes

Yes

645

5,110

1222554 (652875.9)*

10.43593 (148.1005)

6.362239 (2.29341)**

-9518.423 (7413.834)

-19841.07 (15930.62)

63395.62 (55571.59)

37135.98 (60493.76)

Fixed-effect Model

1.31

Yes

Yes

641

3,800

409232.2 (347153.8)

95.923 (49.012)*

1.428478 (.970)

-5013.536 (3893.049)

12158.61 (15493.29)

-14250.49 (12297.6)

7205.408 (6953.884)

Arellano-Bond Model

Brazilian mayors’ human capital and political context on fiscal inputs 43

-.027 (.068)

Mayor’s years of local gov. experience

.335 (.540) -.222 (.431) .022 (.039)

Mayor’s party alignment with governor’s

Mayor’s party alignment with president’s

Council members aligned with mayor’s party

.146 (.147) -.072 (.135) .003 (.007) -.000 (.001) -.042 (.151) -.111 (.085) -.006 (.009)

Education*aligned with governor’s party

Education*aligned with president’s party

Education*aligned with councilmen’s party

Local exp* margin of victory

Local exp* aligned with governor’s party

Local exp* aligned with president’s party

Local exp* aligned with councilmen’s party

.688 (.581) -1.289 (1.602)

Mayor’s private sector experience

Female mayor

Controls

-.002 (.002)

Education*margin of victory

Interaction effects

-.007 (.005)

Margin of victory

Political support

-.012 (.159)

Mayor’s years of education

Mayors’ human capital

Random-effect Model

-1.450 (1.700)

.744 (.628)

-.006 (.009)

-.113 (.091)

-.024 (.143)

-.000 (.001)

.002 (.007)

-.085 (.148)

.151 (.139)

-.002 (.002)

.020 (.041)

-.023 (.442)

.326 (.527)

-.007 (.006)

-.024 (.077)

.019 (.170)

Fixed-effect Model

.216 (.420)

.670 (.535)

.003 (.004)

-.148 (.122)

-.110 (.135)

-.000 (.001)

.002 (.006)

.157 (.168)

.216 (.234)

.000 (.000)

-.003 (.032)

.231 (.508)

.060 (.340)

-.003 (.003)

-.042 (.038)

-.204 (.217)

Arellano-Bond Model

Panel data estimates with interaction effects for the ratio of actual property tax collected over estimated property tax collected, 2009–2016

Name of the independent variables

Table 1.7

44 What works in Latin American municipalities?

.630 (.486) .206 (.567) 2.801 (1.818) -.307 (.144)** 7.68e-06 (6.67e-06) -.000 (.002) 1.547 (9.754) 5,110 645 No No 1.31

Re-elected mayor

Mayor’s leftist party

Population

Urbanization rate

GDP/capita (Brazilian reais)

Infant mortality rate/1,000

Constant

Number of observations

Number of municipalities

Year dummy

Municipality dummy

Variance inflation factor

Note: ***p < .001; **p < .05; *p < .01 (one-tailed test).

Random-effect Model

Name of the independent variables

1.31

Yes

Yes

645

5,110

9.729 (14.159)

-.000 (.002)

9.50e-06 (6.54e-06)

-.445 (.254)*

3.163 (2.290)

.207 (.577)

.545 (.515)

Fixed-effect Model

1.31

Yes

Yes

641

3,800

-48.273 (28.500)*

.001 (.002)

-3.17e-06 (6.92e-06)

-.233 (.143)

7.278 (3.917)*

-.479 (.702)

.931 (.669)

Arellano-Bond Model

Brazilian mayors’ human capital and political context on fiscal inputs 45

46

What works in Latin American municipalities?

DISCUSSIONS AND CONCLUSIONS This study explored (1) whether a mayor’s human capital directly affects municipal collection of property tax collection, and (2) whether political support nonlinearly affects property tax collection by moderating the human capital-tax collection relationship. We tested the linear and nonlinear effect using data from two mayoral administrations, 2009–2016, in the 645 municipalities of the Brazilian state of São Paulo. We assessed property tax collection with three indicators: the ratio of property tax collected over the total municipal revenues; total property tax collected (Brazilian reais); and the ratio of actual property tax collected over estimated property tax collected. Mayors’ years of educational attainment and years of local governmental experience capture mayors’ human capital. Margin of victory and mayor’s party alignment with (a) governor’s, (b) president’s and (c) council members assess political support. In general, our findings report null results, as none of the indicators of mayor’s human capital and not one of the interaction terms explains property tax collection in the municipalities of São Paulo from 2009–2016. These results deviate from other studies reporting (a) a positive mayoral education-property tax collection relationship both in Brazilian and Colombian municipalities (Avellaneda and Gomes, 2015, Avellaneda, 2009b), (b) a positive mayoral local experience-property tax collection relationship (Avellaneda and Gomes, 2015, Avellaneda, 2009b, Petrovsky and Avellaneda, 2014) in Brazilian and Colombian localities, and (c) a negative relationship between mayoral education and public sector experience and property tax collection in the Brazilian municipalities of the state of Minas Gerais (Avellaneda and Gomes, 2017). The null findings also are inconsistent with studies linking managerial teams’ human capital to (a) municipal performance in Israel (Carmeli, 2004), (b) grant-revenue expansion in Colombian municipalities (Avellaneda, 2012), and (c) school districts’ performance in the USA (Johansen, 2012, Fernandez, 2005, and Meier and O’Toole, 2002). Future research, however, should test the robustness of these results in other local settings and across other municipal activities and services (e.g., schools, health, and trash collection). Our analyses also fail to demonstrate that the political support a mayor enjoys neither directly nor indirectly influences, by moderating, the mayor’s human capital-property tax collection relationship. This finding also deviates from Avellaneda and Gomes’ (2017) findings which show that political factors (legislature support and electoral cycle) are strongly correlated with municipal property tax collection in the Brazilian municipalities of Minas Gerais during the 2005–2010 period. Specifically, municipalities in which the mayor enjoys more partisan support of the city council tend to collect more property

Brazilian mayors’ human capital and political context on fiscal inputs

47

taxes. More studies are needed to investigate these inconsistent results across Brazilian states. Future studies also should explore whether political support directly and/or indirectly influences other fiscal indicators, such as grant awards, intergovernmental transfers, and spending allocation. We offer some potential explanations for the lack of findings in our study. First, São Paulo is the richest state in Brazil. São Paulo’s GDP surpasses that of other Latin American countries, with the exception of Mexico. In fact, São Paulo is the third economic power in the region, behind Brazil and Mexico (IBGE, 2020b). This economic status may explain the considerable homogeneity in terms of property tax collection across its municipalities. Future studies should replicate this study in the poorest states, such as Piauí, Maranhão, and Alagoas. Second, São Paulo’s economic status also may also lead to greater homogeneity in mayoral human capital, as citizens, and local leaders, enjoy greater access to higher education and employment opportunities. This again calls for replication studies in other less advantageous states. Our study has some limitations. Lack of data impeded covering more administrative periods. A panel dataset over several terms would add to the analysis. More robust results also could be obtained by testing the model on other financial indicators, on different dimensions of municipal performance, and on other services delivered by the municipalities. What does not work in one policy area may work in another. Also, lack of data hampers controlling for the number and appraisal of municipal properties. Finally, exploring the tax-enforcing mechanisms that municipalities have in place would help explain tax collection levels. This study explored the effect of chief executives’ human capital on tax collection in a transitional economy. In Brazil, the number of small municipalities has proliferated in the last two decades, suggesting the need to understand municipal performance, as key service deliverers. Additionally, despite growing interest in the relationship between human capital and government performance, limited attention has been paid to the contextual conditions that moderate this relationship. Our study contributed to this gap by testing the moderating effect of political support. Despite the null findings, further research should explore the performance effects of other potential moderating factors.

NOTES 1.

2.

Another set of studies addresses the quality of political candidates based on potentially available reward in office, campaign costs, veto power, the flexibility of institutions, among other factors (Caselli and Morelli 2004, Galasso and Nannicini 2011, Poutvaara and Takalo 2007). https://​www​.tse​.jus​.br/​partidos/​partidos​-politicos

48

What works in Latin American municipalities?

REFERENCES Acemoglu, D., Egorov, G., and Sonin, K. (2010). Political selection and persistence of bad governments. The Quarterly Journal of Economics, 125(4): 1511–1575. Alesina, A. et al. (1996). Political instability and economic growth. Journal of Economic Growth, 1: 189–211. Alt, J. E. and Lowry, R. C. (1994). Divided government, fiscal institutions, and budget deficits: evidence from the States. The American Political Science Review, 88(4): 811–828. https://​doi​.org/​10​.2307/​2082709 Amorim-Neto et al. (2001). The political determinants of public deficits in Latin America (1980–1998). Annual of the Latin American Studies Association, Washington D.C. Andrews, Rhys and Boyne, George A. (2010). Capacity, leadership, and organizational performance: testing the black box model of public management. Public Administration Review, 70(3): 443–454. https://​doi​.org/​10​.1111/​j​.1540​-6210​.2010​ .02158​.x Avellaneda, C. N. (2009a). Municipal performance: does mayoral quality matter? Journal of Public Administration Research and Theory, 19(2): 285–312. https://​doi​ .org/​10​.1093/​jopart/​mun001 Avellaneda, C. N. (2009b). Municipal performance: mayoral quality and local public finance. Public Administration Review, 69(3): 469–486. https://​doi​.org/​10​.1111/​j​ .1540​-6210​.2009​.01993​.x Avellaneda, C. N. (2012). Do politics or mayors’ demographics matter for municipal revenue expansion? Public Management Review, 14(8): 1061–1086. https://​doi​.org/​ 10​.1080/​14719037​.2012​.662442 Avellaneda, C. N. (2014). Decentralization and reform in Latin America: improving intergovernmental relations. Publius: The Journal of Federalism, 44(3): e10. https://​ doi​.org/​10​.1093/​publius/​pju015 Avellaneda, C. N. (2016). Government performance and chief executives’ intangible assets: motives, networking, and/or capacity? Public Management Review, 18(6): 918–947. https://​doi​.org/​10​.1080/​14719037​.2015​.1051574 Avellaneda, C. N. and Gomes, R. C. (2015). Is small beautiful? Testing the direct and nonlinear effects of size on municipal performance. Public Administration Review, 75(1): 137–149. https://​doi​.org/​10​.1111/​puar​.12307 Avellaneda, Claudia N. and Gomes, Ricardo Correa. (2017). Mayoral quality and municipal performance in Brazilian local governments. Organizacoes & Sociedade, 24(83): 555–579. Bahl, Roy and Linn, Johannes. (1994). Fiscal decentralization and intergovernmental transfers in less developed countries. Publius: The Journal of Federalism, 24(1): 1–20. Bailey, S. J. and Connolly, S. (1998). The flypaper effect: identifying areas for further research. Public Choice, 95: 335–361. Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1). https://​doi​.org/​10​.1177/​014920639101700108 Barro, Robert J. (1979). On the determination of the public debt. Journal of Political Economy, 87(5): 940–971. https://​doi​.org/​10​.1086/​260807 Behn, R. (2002). The psychological barriers to performance management: or why isn’t everyone jumping on the performance-management bandwagon? Public Performance & Management Review, 26(1): 5–25. https://​doi​.org/​10​.1080/​15309576​ .2002​.11643684

Brazilian mayors’ human capital and political context on fiscal inputs

49

Bergman, M. (2009). Tax Evasion and the Rule of Law in Latin America: The Political Culture of Cheating and Compliance in Argentina and Chile. Penn State University Press. Bergvall, Daniel, Charbit, Claire, Kraan, Dirk-Jan, and Merk, Olaf. (2006). Intergovernmental transfers and decentralised public spending. OECD Journal on Budgeting, 5(4): 111–158. Besley, T. et al. (2011). Do educated leaders matter? The Economic Journal, 121(554): F205–227. https://​doi​.org/​10​.1111/​j​.1468​-0297​.2011​.02448​.x Bowman, M. J. and Myers, R. G. (1967). Schooling, experience, and gains and losses in human capital through migration. Journal of the American Statistical Association, 62(319): 875–898. Boyne, George. (2002). Public and private management: what is the difference? Journal of Management Studies, 39: 97–122. Boyne, George. (2004). Explaining public service performance: does management matter? Public Policy and Administration, 19(4). https://​doi​.org/​10​.1177/​ 095207670401900406 Bradford, D. and Oates, W. (1971). The analysis of revenue sharing in a new approach to collective fiscal decisions. The Quarterly Journal of Economics, 85(3): 416–439. Buchanan, J. M. and Lee, D. R. (1982). Tax rates and tax revenues in political equilibrium: some simple analytics. Economic Inquiry, 20(1): 344–354. https://​doi​.org/​10​ .1111/​j​.1465​-7295​.1982​.tb00352​.x Buchanan, J. M. and Wagner, R. E. (1977). Democracy in Deficit: The Political Legacy of Lord Keynes. New York: Academic Press. Caetano, C. et al. (2017). A relação entre as transferências governamentais, a arrecadação tributária própria e o índice de educação dos municípios do estado de Minas Gerais. Revista de Administracao Publica, 51(5): 897–916. 10.1590/0034–7612174433 Carmeli, Abraham. (2004). Strategic human capital and the performance of public sector organizations. Scandinavian Journal of Management, 20(4): 375–392. Carreirão, Yan de Souza. (2006). Ideologia e Partidos Políticos: Um Estudo Sobre Coligações em Santa Catarina [Ideology and political parties: an essay about coalitions]. Opinião Pública, 12(1): 136–163. Caselli, F. and Morelli, M. (2004). Bad politicians. Journal of Public Economics, 88(3–4): 759–782. https://​doi​.org/​10​.1016/​S0047​-2727(03)00023​-9 Cheibub, J. A., Przeworski, A., and Saiegh, S. M. (2004). Government coalitions and legislative success under presidentialism and parliamentarism. British Journal of Political Science, 34(4): 565–587. Chi, M. T., Glaser, R., and Farr, M. J. (2014). The Nature of Expertise. Psychology Press. Clingermayer, J. C. and Wood, B. D. (1995). Disentangling patterns of state debt financing. American Political Science Review, 89(1): 108–120. https://​doi​.org/​10​ .2307/​2083078 Congleton, R. and Zhang, Y. (2013). Is it all about competence? The human capital of U.S. presidents and economic performance. Constitutional Political Economy, 24(2): 108–124. Cox, G. W. and McCubbins, M. D. (1986). Electoral politics as a redistributive game. The Journal of Politics, 48(2). https://​doi​.org/​10​.2307/​2131098 Dantas, M. K., de Oliveira, L. R., Ferolla, L. M., Paschoalotto, M. A. C., Lopes, J. E. F., Passador, J. L., and Passador, C. S. (2017). Cross-sectoral assessment of public policies in health and the environment: scenario of the municipalities in the state of Sao Paulo. Evaluation and Program Planning, 65: 30–39.

50

What works in Latin American municipalities?

De la Riva, Agüero Renzo. (2021). Do co-governance and CSOs supplement municipal capacity for service delivery? An assessment of differences in simple versus complex services. Journal of Public Administration Research and Theory. https://​ doi​.org/​10​.1093/​jopart/​muab026 De Mello, Luiz R. (2000). Fiscal decentralization and intergovernmental fiscal relations: a cross-country analysis. World Development, 28(2): 365–380. De Mello, L. R. (2002). Public finance, government spending and economic growth: the case of local governments in Brazil. Applied Economics, 34(15): 1871–1883. De Oliveira, C. et al. (2016). Infant mortality: temporal trend and contribution of death. ACTA Paulista de Enfermagem, 29(3): 282–290. 10.1590/1982–0194201600040 Dearborn, DeWitt C. and Herbert A. Simon. (1958). Selective perception: a note on the departmental identifications of executives. Sociometry, 21(2): 140–144. Dixit, A. and Londregan, J. (1998). Ideology, tactics, and efficiency in redistributive politics. The Quarterly Journal of Economics, 113(2): 497–529. https://​doi​.org/​10​ .1162/​003355398555667 Dreher, A. et al. (2009). The impact of political leaders’ profession and education on reforms. Journal of Comparative Economics, 37(1): 169–193. https://​doi​.org/​10​ .1016/​j​.jce​.2008​.08​.005 Ericsson, K. A. and Smith, J. (1991). Toward a General Theory of Expertise: Prospects and Limits. Cambridge University Press. Farvaque, E. et al. (2009). How monetary policy committees impact the volatility of policy rates. Journal of Macroeconomics, 31(4): 534–546. https://​doi​.org/​10​.1016/​j​ .jmacro​.2009​.01​.002 Farvaque, E. et al. (2011). Corporate disclosure: a review of its (direct and indirect) benefits and costs. International Economics, 128: 5–31. Federal Constitution of Brazil. (1988). Testimony of Brazilian Federal Government. http://​www​.planalto​.gov​.br/​ccivil​_03/​constituicao/​constituicao​.htm Fernandez, Sergio. (2005). Developing and testing an integrative framework of public sector leadership: evidence from the public education arena. Journal of Public Administration Research and Theory, 15(2): 197–217. Ferraresi, M. (2021). Political budget cycle, tax collection, and yardstick competition. B.E. Journal of Economic Analysis and Policy, 21(3): 1149–1161. Ferraz, C., Finan, F., and Moreira, D. B. (2012). Corrupting learning. Evidence from missing federal education funds in Brazil. Journal of Public Economics, 96(9–10): 712–726. https://​doi​.org/​10​.1016/​j​.jpubeco​.2012​.05​.012 Fiedler, F. E. (1986). The contribution of cognitive resources and leader behavior to organizational performance. Journal of Applied Social Psychology, 16(6): 532–548. https://​doi​.org/​10​.1111/​j​.1559​-1816​.1986​.tb01157​.x Finkelstein, S. et al. (1996). Strategic Leadership: Theory and Research on Executives, Top Management Teams, and Boards. St. Paul, MN: West Publishing. Forbes, M. and Lynn, L. E. (2005). How does public management affect government performance? Findings from international research. Journal of Public Administration Research and Theory, 15(4): 559–584. Freier, R. and Thomasius, S. (2016). Voters prefer more qualified mayors, but does it matter for public finances? Evidence for Germany. International Tax and Public Finance, 23: 875–910. https://​doi​.org/​10​.1007/​s10797​-015​-9382​-z Freire, M. E. and Garzόn, H. (2014). Managing local revenues. In Farvacque-Vitkovic, C. D. and Kopanyi, M. (eds.), Municipal Finances: A Handbook for Local Governments. Washington, D.C.: World Bank Publications.

Brazilian mayors’ human capital and political context on fiscal inputs

51

Galasso, V. and Nannicini, T. (2011). Competing on good politicians. American Political Science Review, 79–99. Göhlmann, S. and Vaubel, R. (2007). The educational and occupational background of central bankers and its effect on inflation: an empirical analysis. European Economic Review, 51(4): 925–941. https://​doi​.org/​10​.1016/​j​.euroecorev​.2006​.05​.001 Gomes, R. C. and Avellaneda, C. N. (2017). Mayoral quality and municipal performance in Brazilian local governments. Organizações & Sociedade, 24(83). http://​dx​ .doi​.org/​10​.1590/​1984​-9240831 Gomes, R. C. et al. (2013). Analyzing local government financial performance: evidence from Brazilian municipalities 2005–2008. Revista de Administração Contemporânea, 17(6). https://​doi​.org/​10​.1590/​S1415​-65552013000600005 Haddad, M. A., Freguglia, R., and Gomes, C. (2017). Public spending and quality of education in Brazil. The Journal of Development Studies, 53(10): 1679–1696. Hambrick, D. C. (2007). Upper echelons theory: an update. Academy of Management Review, 32(2). https://​doi​.org/​10​.5465/​amr​.2007​.24345254 Hambrick, D. C. and Mason, P. A. (1984). Upper echelons: the organization as a reflection of its top managers. Academy of Management Review, 9(2). https://​doi​.org/​10​ .5465/​amr​.1984​.4277628 Hayo, B. and Neumeier, F. (2014). Political leaders’ socioeconomic background and fiscal performance in Germany. European Journal of Political Economy, 34: 184–205. https://​doi​.org/​https://​doi​.org/​10​.1016/​j​.ejpoleco​.2014​.01​.009 Hibbs Jr., D. A. (1977). Political parties and macroeconomic policy. American Political Science Review, 71(December): 1467–1487. Hill, C. and Lynn, L. (2003). Producing human services – why do agencies collaborate? Public Management Review, 5(1): 63–81. https://​doi​.org/​10​.1080/​ 1461667022000028861 Hines, J. R. and Thaler, R. H. (1995). The flypaper effect. Journal of Economic Perspectives, 9(4): 217–226. Hofstede, G. (1980). Culture and organizations. International Studies of Management & Organization, 10(4): 15–41. https://​doi​.org/​10​.1080/​00208825​.1980​.11656300 IBGE. (1991). Brazilian Institute of Geography and Statistics. Demographic Census 1991. Available at: www​.ibge​.gov​.br IBGE. (2010). Brazilian Institute of Geography and Statistics. Demographic Census 2010. Available at: www​.ibge​.gov​.br IBGE. (2012). Código Tributário Nacional (National Tax Code, CTN). IBGE. (2013). Brazilian Institute of Geography and Statistics. Demographic Census 2010. Brasil 2013. Available at: www​.ibge​.gov​.br IBGE. (2020a). Brazilian Institute of Geography and Statistics. Brazilian characteristics projection. Available at: www​.ibge​.gov​.br IBGE. (2020b). Brazilian Institute of Geography and Statistics. Síntese de indicadores sociais: uma análise das condições de vida da população brasileira. Available at: https://​biblioteca​.ibge​.gov​.br/​visualizacao/​livros/​liv101760​.pdf Jochimsen, B. and Thomasius, S. (2014). The perfect finance minister: whom to appoint as finance minister to balance the budget. European Journal of Political Economy, 34(C): 390–408. https://​doi​.org/​10​.1016/​j​.ejpoleco​.2013​.11​.002 Johansen, Morgen S. (2012). The direct and interactive effects of middle and upper managerial capacity on organizational performance. Administration & Society, 44(4): 383–411.

52

What works in Latin American municipalities?

Kellough, J. E. (2002). Performance appraisal and merit pay in public agencies: a critical examination. In C. Ban and N. M. Ricucci (eds.), Public Personnel Management: Current Concerns, Future Challenges, 3rd edition, 181–193. New York: Longman. Lindbeck, A. and Weibull, J. W. (1987). Balanced-budget redistribution as the outcome of political competition. Public Choice, 52: 273–297. https://​doi​.org/​10​.1007/​ BF00116710 Lowry, R. C. et al. (1998). Fiscal policy outcomes and electoral accountability in American states. American Political Science Review, 92(4): 759–774. https://​doi​ .org/​10​.2307/​2586302 Lynn, B. (1981). Productivity and the San Diego Police Patrol Bureau. Public Productivity Review, 5(1): 77–80. https://​doi​.org/​10​.2307/​3380172 Lynn, L. E. et al. (2000). Studying governance and public management: challenges and prospects. Journal of Public Administration Research and Theory: J-PART, 10(2): 233–261. Martell, C. R. (2008). Fiscal institutions of Brazilian municipal borrowing. Public Administration and Development, 28: 30–41. Masaki, Takaaki. (2018). The impact of intergovernmental transfers on local revenue generation in sub-Saharan Africa: evidence from Tanzania. World Development, 106: 173–186. Meier, K. J. and O’Toole, L. J. (2002). Public management and organizational performance: the effect of managerial quality. Journal of Policy Analysis and Management, 21(4): 629–643. https://​doi​.org/​10​.1002/​pam​.10078 Meier, K. J. and O’Toole, L. J. (2013). Subjective organizational performance and measurement error: common source bias and spurious relationships. Journal of Public Administration Research and Theory, 23(2): 429–456. https://​doi​.org/​10​ .1093/​jopart/​mus057 Meier, K. J., Rutherford, A. and Avellaneda, C. N. (2017). Comparative Public Management: Why National, Environmental, and Organizational Context Matters. Washington, D. C.: Georgetown University Press. Miles, R. E. et al. (1978). Organizational strategy, structure, and process. The Academy of Management Review, 3(3): 546–562. https://​doi​.org/​10​.2307/​257544 Mincer, J. (1958). Investment in human capital and personal income distribution. Journal of Political Economy, 66(4): 281–302. Moosa, S., Derese, A., and Peersman, W. (2017). Insights of health district managers on the implementation of primary health care outreach teams in Johannesburg, South Africa: a descriptive study with focus group discussions. Human Resources for Health, 15(1): 1–9. https://​doi​.org/​10​.1186/​s12960​-017​-0183​-6 Murphy, K. R. and Cleveland, J. (1995). Understanding Performance Appraisal: Social, Organizational and Goal-based Perspectives. Thousand Oaks, CA: Sage. Olvera, J. G. and Avellaneda, C. N. (2019). Subnational government performance: testing multiple dimensions of governors’ experience. Public Administration Review, 79(3): 383–398. https://​doi​.org/​10​.1111/​puar​.13035 O’Toole, Laurence J. and Meier, Kenneth J. (2017). Public management, context, and performance: in quest of a more general theory. Journal of Public Administration Research and Theory, 25(1): 237–256. Paschoalotto, M. A. C. et al. (2018). The regionalization of the Brazilian National Health System: proposition for performance evaluation of the Regional Health Departments of São Paulo state, Brazil. Saúde & Sociedade, 27(1). https://​doi​.org/​10​ .1590/​S0104​-12902018170095

Brazilian mayors’ human capital and political context on fiscal inputs

53

Paschoalotto, M. A. C., Passador, J. L., Passador, C. S., and de Oliveira, P. H. (2020). Local government performance: evaluating efficiency, efficacy, and effectiveness at the basic education level. BAR – Brazilian Administration Review, 17(4): 1–17. https://​doi​.org/​10​.1590/​1807​-7692bar2020190149 Petrovsky, Nicolai and Avellaneda, Claudia N. (2014). Mayoral public sector work experience and tax collection performance in Colombian local governments. International Public Management Journal, 17(2): 145–173. Poutvaara, P. and Takalo, T. (2007). Candidate quality. International Tax and Public Finance, 14(1): 7–27. http://​dx​.doi​.org/​10​.1007/​s10797​-006​-6691​-2 Puppim de Oliveira, J. A. (2017). Brazilian public administration: shaping and being shaped by governance and development. Chinese Political Science Review, 2(1): 7–21. https://​doi​.org/​10​.1007/​s41111​-017​-0052​-4 Queiroz, M. V. A. B., Sampaio, R. M. B., and Sampaio, L. M. B. (2020). Dynamic efficiency of primary education in Brazil: socioeconomic and infrastructure influence on school performance. Socio-Economic Planning Sciences, 70(August 2019): 100738. https://​doi​.org/​10​.1016/​j​.seps​.2019​.100738 Ribeiro, F. G., Braun, G., Carraro, A., Teixeira, G. da S., and Gigante, D. P. (2018). An empirical assessment of the healthy early childhood program in Rio Grande do Sul state, Brazil. Cadernos de Saude Publica, 34(4). https://​doi​.org/​10​.1590/​0102​ -311x00027917 Rogoff, K. and Sibert, A. (1988). Elections and macroeconomic policy cycles. The Review of Economic Studies, 55(1): 1–16. Rosa, L., Martins, M., and Carnoy, M. (2019). Achievement gains from reconfiguring early schooling: the case of Brazil’s primary education reform. Economics of Education Review, 68(July 2017): 1–12. https://​doi​.org/​10​.1016/​j​.econedurev​.2018​ .10​.010 Roubini, N. and Sachs, J. (1989). Government spending and budget deficits in the industrial countries. Economic Policy, 4(8): 100–132. Rubin, B. R. (2000). The political economy of war and peace in Afghanistan. World Development, 28(10): 1789–1803. https://​doi​.org/​10​.1016/​S0305​-750X(00)00054​-1 Russo, L. X., Scott, A., Sivey, P., and Dias, J. (2019). Primary care physicians and infant mortality: evidence from Brazil. PLoS ONE, 14(5): 1–16. https://​doi​.org/​10​ .1371/​journal​.pone​.0217614 Santos, R. R., de Freitas, M. M., and Flach, L. (2020). Avaliação da Eficiência dos Gastos Públicos com Educação dos Municípios de Santa Catarina. Administração Pública e Gestão Social, 12. https://​doi​.org/​10​.21118/​apgs​.v12i2​.5755 SEADE Foundation. (2016). São Paulo State Mayor´s Electoral Informations. Https://​ Produtos​.Seade​.Gov​.Br/​Produtos/​Eleicoes/​ Shah, Anwar. (2006). Local Governance in Developing Countries. Public Sector Governance and Accountability. Washington, D.C.: World Bank. https://​ openknowledge​.worldbank​.org/​handle/​10986/​7192 License: CC BY 3.0 IGO Singh, S., Darwish, T. K. and Potočnik, K. (2016). Measuring organizational performance: a case for subjective measures. British Journal of Management, 27: 214–224. Smolka, M. O. and De Cesare, C. (2006). Property taxation and informality: challenges for Latin America. Land Lines, 18(2): 14–19. Spence, M. (1973). Job market signaling. The Quarterly Journal of Economics, 87(3): 355–374. https://​doi​.org/​10​.2307/​1882010 Treisman, D. (1996). The politics of intergovernmental transfers in post-Soviet Russia. British Journal of Political Science, 26(3): 299–335. doi:​ 10​ .1017/​ S0007123400007481

54

What works in Latin American municipalities?

Tsakumis, G. T. et al. (2007). The relation between national cultural dimensions and tax evasion. Journal of International Accounting Auditing and Taxation, 16(2): 131–147. 10.1016/j.intaccaudtax.2007.06.004 Valle, A. G. R. M. and Gomes, R. C. (2014). Analyzing the importance of financial resources for educational effectiveness: the case of Brazil. International Journal of Productivity and Performance Management, 63(1): 4–21. Vargas, I. et al. (2015). Regional-based integrated healthcare network policy in Brazil: from formulation to practice. Health Policy and Planning, 30(6): 705–717. 10.1093/ heapol/czu048 Von Hagen, J. and Harden, I. J. (1995). Budget processes and commitment to fiscal discipline. European Economic Review, 39(3–4): 771–779. Walker, R. M. et al. (2010). Public Management and Performance: Research Directions. Cambridge University Press. https://​doi​.org/​10​.1017/​CBO9780511760587 World Bank Group. (2016). The Role of Local Governments in Promoting Local Economic Development in Uganda. World Bank, Washington, D.C. https://​ openknowledge​.worldbank​.org/​handle/​10986/​25359 License: CC BY 3.0 IGO Zucco, C. (2013). When payouts pay off: conditional cash transfers and voting behavior in Brazil, 2002–10. American Journal of Political Science, 57(4): 810–822. https://​ doi​.org/​10​.1111/​ajps​.12026

2. Administrative capacity and Chilean local governmental effectiveness Gabriel Piña and Claudia N. Avellaneda INTRODUCTION Subnational organizational capacity is crucial in the effective functioning of government across the world (World Bank 2001, United Nations 2009). In developing and centralized unitary countries, the recent adoption of political, fiscal, and administrative decentralization has provoked a lively debate about the capacity of local and subnational governments to manage, finance, and plan for their new set of responsibilities. Despite the generalized understanding of both the importance of administrative capacity and its contributing role in organizational production, public management literature has overlooked capacity and its relationship to public organizational effectiveness, especially in developing settings. Still fewer studies have examined the link between capacity and government effectiveness (Andrews and Boyne 2010, Hall 2008, Terman and Feiock 2014, Wimpy et al. 2017). While a growing body of studies explores management capacity (e.g., Ingraham 2007, Ingraham et al. 2003), its determinants (e.g., Knack 2002), and its effect on stakeholders’ assessment of performance (Andrews et al. 2010), few studies measure organizational capacity with objective indicators and/or test its influence on organizational performance (e.g., Andrews and Boyne 2010, Hall 2008, Terman and Feiock 2014). The available studies provide scarce consideration of the impact of capacity in contexts other than U.S. state governments (Bowman and Kearney 1988), tend to rely on subjective measures of capacity and performance, and are based on cross-sectional rather than longitudinal analyses (Andrews and Brewer 2013, Andrews and Boyne 2010). Reliance on subjective measures may be due to unclear concept definition. Here we adopt a multi-dimensional perspective of organizational capacity that embraces three dimensions: (1) possession of human resources to perform a task, (2) capability to deploy resources, and (3) expertise acquired to perform a particular task. Although organizational capacity is desirable at all levels of governments, local governments are more likely targeted as having 55

56

What works in Latin American municipalities?

insufficient organizational capacity to perform their tasks (Brown and Potoski 2003, Hall 2008). Therefore, more research is needed at the local government level to better understand whether and how capacity translates into greater government effectiveness. In the present study, we explore whether administrative capacity affects local government effectiveness in acquiring and implementing funds for infrastructure projects. We compiled a data set of infrastructure grant proposals submitted by 340 (out of 345) Chilean municipalities over a nine-year period (2005–2013), covering three municipal administrations. We also drew on data from interviews with local government administrators, grant reviewers, and regional authorities in an effort to better understand the causal mechanisms behind municipal effectiveness in securing grants. Government effectiveness is operationalized through percentage of municipal grant projects approved, measured with two indicators: (1) percentage of grants obtained in relation to the total number of grant proposals submitted, and (2) percentage of money secured in relation to total amount requested. Organization capacity is measured across three dimensions: (1) human resources (total administrative personnel), (2) capability (inter-organizational cooperation for grant submission), and (3) expertise (middle-level managers’ grant-related expertise). After controlling for the municipal political context, past performance, and other grant proposal and municipal features, results suggest administrative capacity boosts government effectiveness in acquiring grants. Political factors and the electoral cycle also appear to influence municipal grant acquisition. This study contributes to the currently limited body of research on the role of capacity in government effectiveness. It does so by addressing four research needs. First, this study measures organizational capacity across three dimensions: capability, expertise, and human resources. Second, as studies on performance largely have focused on educational measures of effectiveness (e.g., percentage of pupils passing an exam), this study expands in terms of policy area by examining government effectiveness in securing infrastructure grants. Increasing grant revenues has become a primary local strategy for counterbalancing declining intergovernmental transfers and locally collected revenues (Avellaneda 2015). Third, this study shifts the research focus on capacity and effectiveness from U.S. states and English local governments to Latin American localities, jurisdictions struggling to improve capacity. Finally, our longitudinal data set, which covers nearly all Chilean localities, allows for dynamic rather than static connections. The first section of this chapter draws on existing literature about effectiveness and performance. The second section defines organizational capacity and discusses its role in government performance in order to develop the testable hypotheses. Subsequently, case selection, units of analysis, data, and measures

Administrative capacity and Chilean local governmental effectiveness

57

are outlined. We then present the multivariate statistical results from the panel data, highlighting the key findings.

PERFORMANCE IN TERMS OF EFFECTIVENESS Defining effectiveness is not without controversy (Mitchell 2012). Two main approaches have been used to study effectiveness: the goal-attainment approach and the system resources approach (Forbes 1998). Organizational effectiveness is usually defined as the extent to which an organization achieves its objectives (Miles 1980, Price 1972). For example, a widely used effectiveness measure in public management research is the percentage of pupils passing a specified exam. This approach provides two advantages: the researcher can directly measure the degree of attainment of a particular objective (Rainey 2009), by assessing the level of outputs and outcomes accomplished (Daft 2010). The relationship between effectiveness and performance is so strong that scholars often use the terms “effectiveness” and “performance” interchangeably (Selden and Sowa 2004). Effectiveness is a fundamental dimension measure of performance (Boyne 2002). A considerable number of performance models draw upon the “3Es” model of economy, efficiency, and effectiveness of services and the “IOO” model examining the sequence of inputs, outputs, and outcomes (Boyne 2002, Walker et al. 2010). In the public management literature, effectiveness has been used to assess schools (Meier and O’Toole 2001), job training programs (Heinrich 1999), public bureaucracies, state governments (Selden and Sowa 2004, Ingraham and Moynihan 2001), and local governments (Avellaneda 2009, Petrovsky and Avellaneda 2014). The systems resource approach, on the other hand, defines effectiveness as the ability of organizations to exploit resources from their environments (Forbes 1998). That is, government effectiveness also can be assessed in terms of acquiring resources, or inputs, to support the organizational survival. As Seashore and Yuchtman (1967) assert, good performance involves “the ability to exploit [the organization’s] environment in the acquisition of scarce and valued resources to sustain its functioning” (393). To sustain the organizational functioning, an organization’s inputs can be more critical than its outputs because the organization must have the resources required to operate, regardless of its production. Therefore, effectiveness in acquiring resources (e.g., funds) could be the most important indicator of performance (Selden and Sowa 2004).

58

What works in Latin American municipalities?

ADMINISTRATIVE CAPACITY AND ORGANIZATIONAL PERFORMANCE In 2010, Andrews and Boyne lamented the status of the evidence linking performance and organizational capacity, noting that studies have mainly focused on explaining policy adoption rather than organizational performance in terms of service delivery. The scarcity in this line of research is in part due to the variety of organizational capacities addressed. While some studies refer to “organizational/government capacity” (Berman and Wang 2000), others opt to focus on “administrative capacity” (Wimpy et al. 2017), and a few others center on “management capacity” (Andrews and Boyne 2010, Andrews and Brewer 2013, Wang et al. 2015). Along with this variation in terminology, the empirical studies also vary in concept operationalization. For example, Berman and Wang (2000) assess government capacity for implementing performance management systems by operationalizing it with counties’ stakeholder support and technical infrastructure. Wimpy et al. (2017) examine administrative capacity in African countries using the World Bank’s quality-of-government indicators. Wang et al. (2015) assess management capacity with a survey of elite opinion assessments of three components – the managing of the government’s operations, insuring quality in policy implementation, and coordinating human resource management outside of the core government administration. On the other hand, Andrews and Brewer (2013) and Andrews and Boyne (2010) assess management capacity across five management systems: financial management, human resource management, information technology, capital management, and leadership. Even among studies using common terminology, their concept operationalization varies. Gargan (1981), however, contends that “capacity should not be viewed exclusively from a management perspective” because government capacity “is a function of expectations, resources, and problems” (1981, 651–652). Accordingly, this study focuses on general organizational capacity, following Honadle’s (1981) characterization of capacity as “organizations’ ability to achieve their aims” (quoted in Berman and Wang 2000, 410). In the literature examining intergovernmental grant acquisition and grant performance, some empirical studies explore the role of administrative capacity in grant programs. Some of these studies specifically examine how political and administrative capacity influences grant program and policy outcomes in federal systems (Hall 2008, Terman and Feiock 2014). This literature centers on the United States, studying how the relationship between grantor and grantee influences the likelihood of obtaining and implementing grants (Nicholson-Crotty 2012). Other scholars have investigated how capacity impacts the selection of recipients for competitive grant awards (Collins and Gerber 2006, 2008, Hall 2008),

Administrative capacity and Chilean local governmental effectiveness

59

and some studies have considered political factors and local governments’ capacities in concert (Collins and Gerber 2008, Terman and Feiock 2014). In the existing literature, however, the terms of organizational capacity, capability, and competence have been used interchangeably (Andrews et al. 2016, Avellaneda 2012). This practice led Kolar Bryan (2011) to describe the different definitions of organizational capacity, identifying three different perspectives prevalent in the literature: capacity as resources, capacity as organizational capabilities, and capacity as organizational competency. Capacity as Resources Resources are the inputs into an organization’s production process (Honadle 1981, Ingraham et al. 2003). The ability of an organization to realize its goals is a function of its capacity to obtain resources. This notion derives from open-system organizational theories, which stress the importance of obtaining resources from the environment for organizational survival (Pfeffer and Salancik 2003). Organizational resources can be tangible (financial) or intangible (human capital: reputation, experience, expertise, knowledge, connections) (Avellaneda 2015, Burgess 1975). According to the resource-based view, an organization’s set of tangible and intangible resources constitutes its competitive advantage (Rumelt 1984, Penrose, 1959). The organizational performance literature takes the view that resources positively affect performance. Some scholars, however, suggest both accountability and managerial/bureaucratic capacity condition the resource-performance relationship. Empirical analyses testing the resource-performance relationship abound. Hence, in 2003, Boyne identified 18 studies testing the effect of financial resources on service performance, and 26 studies testing the influence of human resources (staff quality and quantity) on different dimensions of performance. All these studies report inconsistent results. Other studies have related administrative capacity to employee stability, as bureaucratic permanence is considered an intangible resource. Wimpy et al. (2017) use a measure of organizational capacity determined through public administrative resources that assesses how civilian central government staff is structured to design and implement policy/programs and deliver services effectively. Despite the diverse indicators used to operationalize capacity, most of the studies rely on a measure of human resources, specifically the size of the administrative staff (Huber and McCarty 2004, Hall 2008). Consequently, we propose that: H1: The more human resource capacity an organization has, the higher its effectiveness.

60

What works in Latin American municipalities?

Capacity as Organizational Capabilities As Piening writes, “[w]hile resource refers to an input to production that a firm owns ... a capability describes the firm’s capacity to deploy resources to achieve a desired outcome” (Piening 2013, 212, see also Helfat and Peteraf 2003). In other words, resources alone do not constitute capacity (Kolar Bryan 2011, 12) because organizations also must have access to the skills and processes needed to convert inputs into outputs (Dess et al. 2007) by managing resources effectively (Honadle 1981, Ingraham et al. 2003). According to Ingraham et al. (2003), administrative “know how” constitutes managerial capacity. This perspective of managerial capacity is also reinforced by Helfat et al., who assert that capability is “the ability of an organization to perform a coordinated set of tasks, utilizing organizational resources, for the purpose of achieving a particular end result” (2007, 999). Likewise, Harvey et al., state that capabilities “emphasize the key role of strategic management in adapting, integrating, and reconfiguring internal and external skills, resources, and functional competences to match requirements with the changing environment” (2010, 83). For others, such as Andrews et al. (2015), organizational capability is associated with structural configuration, including department size, structural complexity, agencification, personnel stability, and use of temporary employees. In their qualitative comparative analysis of U.K. central government departments, Andrews et al. (2016) find that high-capability departments exhibit two organizational configurations – low structural complexity and personnel stability – while low-capability departments are characterized by having personnel instability, structural complexity, and departmental agencification. In the public sector, intergovernmental cooperation/collaboration can add to organizational capability since “alliances strengthen a firm’s asset position by gaining access to new, external resources and capabilities” (Piening 2013, 212, see also Eisenhardt and Martin 2000 and Keil 2004). Indeed, Kolar Bryan and Roussin Isett’s (2013) study, which included 56 interviews in four states, finds the capability to collaborate with other organizations is critical to an organization’s perceived ability to carry out its mission and agenda. In resource-scarce contexts, intergovernmental cooperation through technical assistance, for example, should contribute to resource acquisition and, in turn, to government performance. Collaborating with other organizations leads to organizational access to knowledge (Kelman et al. 2012) and complementary skills, new technologies, and the ability to provide a wider range of products and services beyond its organizational boundaries. Moreover, there are certain policy areas whose implementation of programs demand collaboration with other units and governmental or nongovernmental agencies. In complex policy areas, intergovernmental collaboration becomes

Administrative capacity and Chilean local governmental effectiveness

61

necessary (Agranoff and McGuire 1998). Wang et al. (2015) illustrate the need for collaboration and coordination in implementing local green economic strategies, as an “effort to build human resource management capacity and practices” (6). Therefore, H2: The more an organization engages in intergovernmental cooperation, the higher its effectiveness. Capacity as Organizational Competency In addition to resources and capabilities, organizational capacity is also defined in terms of competency (Kolar Bryan 2011). This perspective understands capacity as those organizational resources and capabilities that are related to organizational effectiveness (Kolar Bryan 2011, 13). According to Bryson, “a competency is a capability, set of actions, or strategy that help an organization perform well on its key success factors. In other words, an organization may have a competency, but if it does not help the organization do well on a key success factor, it is not much of a competency” (2004, 126). In sum, competency refers to the ability to do something well. For Hroník, managerial competence is a “bunch of knowledge, skills, experience” that supports the achievement of organizational objectives (2007, in Krajcovicova et al. 2012, 1120). Similarly, Krontorád and Trčka define competence as “a combination of knowledge, skills, abilities, and behaviors that an employee uses in carrying out [his or her] work” (2005, in Krajcovicova et al. 2012, 1120, see also Kolar Bryan 2011). These definitions characterize knowledge, experience, skills, and expertise as key managerial competencies. Expertise, according to Ericsson, Krampe, and Tesch-Römer (1993), refers to “domain-specific skills and knowledge, which are important to attainment of expert performance” (365), and “is acquired slowly over a very long time as a result of practice” (366). They also argue that “[e]xperts are faster and more accurate … and their memory for representative stimuli from their domain is vastly superior to that of lesser experts, especially for briefly presented stimuli” (Ericsson, Krampe, and Tesch-Römer 1993, 365). Empirical research linking expertise and expert performance (Chi, Glaser, and Farr 2014, Ericsson and Smith 1991) has shown that experts’ superior performance is acquired through long experience, and the effect of practice on performance is large (Ericsson, Krampe, and Tesch-Römer 1993, 365–368). Likewise, Wang et al. (2015), referencing the work of Donaldson (2001) and Mintzberg (1979), contend that “if a local government has a dedicated staff whose main task is to coordinate and manage certain efforts and strategies, it will enable the government to achieve the expected outputs/outcomes by

62

What works in Latin American municipalities?

gaining the benefit of specialization” (5). In sum, expertise is important for all policy and management areas (Wang et al. 2015, 5). Therefore, H3: The higher an organization’s specific task-related expertise/competence, the higher its effectiveness in that specific task.

CASE SELECTION: CHILEAN MUNICIPALITIES If local governments’ actions influence state and national outcomes, then more understanding on local capacity’s effects is needed. Following this view, Andrews and Boyne (2010, 444) suggest that “to build a strong evidence base on the relationship between management capacity and performance, it is essential to investigate its presence in these [local] organizations.” Moreover, the existing studies exploring the capacity-performance relationship rely on cross-sectional data sets. Longitudinal data would allow researchers to better explore cause-and-effect mechanisms (Wooldridge 2010). To address these gaps, we test our hypotheses using data from 340 (out of 345) Chilean municipalities, over a nine-year period (2005–2013). Chile is formally a unitary country organized into 15 regions and 345 municipalities. The majority of Chilean municipalities are relatively small: the municipal average population is about 48,000 residents, but its median population is 18,000 inhabitants. The most populous municipalities (generally more than 100,000 residents) are concentrated in the capital (Santiago) and in a few regional capitals. Of all the municipalities, 75% have a population of less than 50,000 people. Equivalent in scope and structure to U.S. counties, Chilean municipalities enjoy extensive, constitutionally granted fiscal and political autonomy, including the authority to design, fund, and implement policies and programs. Chile, like most Latin American countries, has a particular form of local leadership – a “strong, elected mayor.” Mayors are elected for four-year terms and may continuously serve consecutive terms if re-elected. The Chilean Constitutional Law of Municipalities stipulates a legislative body oversees a directly elected mayor. This municipal council is elected concurrently with the mayor for a four-year period and consists of six to ten members, depending on the number of eligible voters in the municipality. Similar to the U.S., but unlike other Latin American countries, most Chilean municipal spending is financed through municipally collected funds. Municipal direct revenues – collected primarily from royalties, service provision, property tax, and sales of their own assets – can be spent in any sector. On average, approximately 60% of the municipal budget comes from local taxes, and the other 40% comes from transfers from a small number of rich municipalities to poorer ones (Bravo Rodríguez 2014). This transfer system is

Administrative capacity and Chilean local governmental effectiveness

63

known as the Common Municipal Fund (Fondo Común Municipal, or FCM), used to redistribute revenue. The primary responsibilities of local governments are social programs, such as public education (elementary schools and high schools) and public health. The law determines two types of functions for municipalities: “exclusive” (privativas) and “shared” (compartidas). Exclusive activities are those specific to the municipality without participation of any other agency or organization. Examples of such activities include enforcement of transportation rules, garbage collection, creation of local development plans, and enforcement of building codes. The shared activities involve other public and private organizations and include education, healthcare, social welfare, and recreation. The primary source of revenue directed to fund infrastructure projects comes from the National Public Investment System, the most consolidated investment appraisal system in Latin America (Gómez-Lobo 2012). In Chile, by law, all public bodies, such as ministries, regional governments, municipalities, publicly owned companies, or public services wishing to undertake an investment project or program using funding from the central government must apply to the National Public Investment System. Only initiatives that have been evaluated through this system can be undertaken within the public sector. Depending on the type of project, an evaluation consists of either a cost-benefit analysis or a cost-effectiveness analysis. Once the project is approved, the regional authorities can prioritize the project, allocating the resources in the first or second year following approval.1 The literature on performance of Chilean municipalities is extremely scarce. Ormeño’s (2011) study and Pribble’s (2015) study are, to our knowledge, the only two studies focused on factors associated with municipal performance in Chile. However, both rely on a cross-sectional analysis. We therefore aim to provide the first longitudinal analysis of Chilean municipal performance. Research Design The unit of analysis in this study is the municipality-year. Data availability limited the study to a nine-year period (2005–2013), which covers the four years of the 2005–2008 mayoral administration, the four years of the 2009–2012 administration, and one year of the 2013–2016 administration. Mayoral inauguration normally occurs in December of the year before the administration commences. Because the beginning of the mayoral administration nearly coincides with the beginning of the calendar year, it is possible to associate annual municipal indicators with a specific mayoral administration. Data were obtained from several sources. The National System of Municipal Indicators (Sistema Nacional de Indicadores Municipales, SINIM),2 a centralized data warehouse for municipalities run by the central government’s

64

What works in Latin American municipalities?

Integrated Projects Bank (Banco Integrado de Proyectos, BIP),3 provided information on municipal applications. The Transparency System, a system similar to the Freedom of Information Act in the U.S., which applies to almost all public organizations in Chile, provided additional information requested by the authors. Additionally, data on political variables were collected through the National Electoral Service (SE). Variable Definition and Operationalization We assess organizational performance through municipal effectiveness in securing infrastructure projects. Evidence indicates that acquiring funds through external grants is among mayors’ and managers’ most pursued local government strategies (Stevens and McGowan 1983, Avellaneda 2015). We measure municipal performance in terms of effectiveness with respect to projects implemented (number of projects awarded relative to total number of applications) and effectiveness in monetary terms (value of money awarded relative to total amount requested). As explained in the previous section, any public organization in Chile interested in carrying out an infrastructure project must first present the project to the central government. Municipalities may use their own funds, but, on average, one infrastructure project is equivalent to about 10% of a municipality’s annual revenues, so locally funding infrastructure projects are rarely feasible. Most municipalities must therefore apply to the central government for funds to invest in local projects, such as building a park, repairing a classroom in a local school, or paving a street. In theory, municipalities must reach agreement with their regional government about the projects they plan to propose each year. Once municipalities and the regional government agree on the projects to fund for the next year, each municipality must send its application to the Ministry of Social Development (MSD), which evaluates proposals based on technical and economic merit. Once the MSD approves a project, it is generally implemented one or two years later. Some municipalities lack the capacity to design and present these projects because they have neither the technical knowledge nor the access to resources to carry out thorough economic analysis, evaluation, and development of their projects (Espinoza 2014). According to some of the interviewees, this deficiency is the main barrier to obtaining funding, particularly for small municipalities. For example, whereas in the period under analysis (2005–2013) the MSD project approval rate for all public organizations was 65% (including older projects that were renewed during the study period), projects proposed by municipalities had an average approval rate of 55%. We collected data from the MSD regarding the status of approximately 54,000 municipal project applications since 2005. We limited our data set to

Administrative capacity and Chilean local governmental effectiveness

65

applications for new projects, excluding continuations of previous projects and funds that replicated previously funded projects within the same municipality, as these funding decisions likely were based on precedent and therefore related to previous, rather than current administrative conditions. To measure effectiveness in funds acquisition, we calculated the number of projects presented by each municipality and the number of these projects that were approved and implemented, to construct a ratio of projects approved over projects presented in each year.4 We also calculated the amount of funding received relative to the total amount requested. Table 2.1 lists the descriptive statistics for all of the variables. The five excluded municipalities are the richest in Chile, possessing sufficient resources to fund their projects without the need for assistance from the central government. Independent Variables Although some studies have assessed the effect of organizational capacity on policy outputs and/or organizational performance, most capacity measures have relied on subjective and external assessments, which have limitations. For instance, when capacity and performance are measured by the same rater, subjective measures can suffer from common-method bias, caused by the tendency of respondents to give similar responses to distinctive questions (Walker and Boyne 2006, Walker et al. 2010). Subjective measures also suffer from recall issues, as respondents may lack a comprehensive understanding of the organizational issues asked in a survey (Golden 1992). Objective measures are traditionally viewed as the “gold standard” of public management research (Walker et al. 2010). Nevertheless, to our knowledge, scarce research has used objective measures of both administrative capacity and performance. We measure capacity in three different ways, in line with our hypotheses. Capacity is measured as resources (administrative personnel), capabilities (collaboration), and competence (expertise). Resources are operationalized as administrative personnel. Mayors and middle-level municipal managers have employees working directly under them. In one interview, a planning manager complained that he had insufficient staff to develop projects, as his five employees spent most of their time on previously approved projects, leaving little time to apply for new funds.5 Because we do not have data on specific municipal teams’ characteristics (i.e., number of proposal writers), for each year we use the total number of employees per municipality as a measure of municipal human resources, similar to approaches used in previous research (e.g., Hall 2008). We do not include a measure of financial resources as “capacity” since greater financial resources could be positively associated with funds acquisition by providing material inputs to the application process, yet more resources also could make the need

66

What works in Latin American municipalities?

for funds less pressing, reducing the incentive to apply for external resources. We instead use financial resources (e.g., revenues) as a control variable. Organizational capability is operationalized as municipal-regional/central collaboration. As a proxy for inter-organizational collaboration, we measure the percentage of annual applications for projects within a municipality that are submitted by an employee of the regional or central government, rather than the municipal government. This variable does not directly measure collaboration, but we expect that it correlated with it, because in order for regional or central government employees to review and submit an application pertaining to a particular municipality, they must have some degree of coordination with the staff or leadership of that municipality. The willingness of actors outside of the municipality to undertake these activities also signifies a collaborative relationship. The MSD is the only government agency with the authority to approve municipal projects. Neither the regional government nor other central government agencies participate in the official approval process, but, in general, both central government agencies and regional governments possess greater technical knowledge and experience in applications than municipalities (Espinoza 2014) owing to their scope and human resources. For example, regional governments develop large, complex projects involving multiple localities (e.g., highways), and some have specialized project development teams.6 Municipalities that form collaborative relationships with these agencies therefore have access to greater organizational capabilities than those that do not. Organizational competence is operationalized as municipal projects-related expertise. The data from the Integrated Projects Bank includes the name of the person submitting the final version of each project application. We use this data to construct a proxy for local projects-related expertise. We counted the number of times within a study period that a submitter’s name appears on successful past applications. In other words, this variable measures the number of times the employees of a given municipality have previously participated in effective project designs, development, and submission. For example, if three employees work on several projects in a given municipality-year, we counted the number of projects these three employees submitted and received approvals for in prior years (since 2000), which could be five, ten, or 12 projects, respectively. We then calculated the average expertise of these three employees by dividing the total number of successful projects submitted by the total number of submitters ((5+10+12)/3) = 9.7 For a given municipality-year, we calculated the average expertise of all employees appearing as project submitters. This measure does not distinguish between middle-level managers’ expertise and other employees’ expertise, since the database does not clearly describe the position of the person sending the application, but the data do allow us to see that at least 35% of the submit-

Administrative capacity and Chilean local governmental effectiveness

67

ters are middle-level managers in charge of planning. The advantage of this variable is that it is available for the entire sample. Control Variables The political context in which an organization operates is also expected to influence its performance. There is evidence that regional authorities, particularly regional council members, can have a strong influence in the application process (Espinoza 2014). Regional authorities can assist municipalities with the preparation and design of projects before they are sent to the central government for final approval.8 Regional government authorities are chosen by the president, therefore, the variable “partisan alignment” measures party alignment between the mayor and regional governments. This dummy variable is coded “1” when the mayor and the governor belong to the same political coalition; otherwise “0.”9 In addition, government organizations operating under a divided government may perform differently from those operating under a unified government. To account for this, we also measure mayor and municipal council partisan alignment. Council members also are elected at the local level. We include this variable to address the possibility that mayors performing in politically divided governments may face difficulty in obtaining the council’s support for their projects. Under these circumstances, mayors may be more aggressive in seeking national and/or state funding, as these funds may not require the council’s oversight. This measure is a continuous variable reporting the percentage of council members that are politically aligned with the mayor. Electoral competitiveness is assessed in terms of the margin of victory, given as a percentage, between the winner and the runner-up in mayoral elections. To avoid misattributing municipality effects to any of the independent variables, and to avoid omitted variables, the study also controls for mayors’ ideology (coalition), environmental characteristics, such as total population, percentage of population under the poverty line, percentage of population living in rural areas, and total revenues, all obtained from SINIM. We also control for the impact of the February 2010 earthquake, the fifth-largest recorded earthquake in world history, which leveled many buildings and infrastructure in the Chilean fifth, sixth, seventh, and eighth regions. This catastrophe destroyed significant public infrastructure, such as schools, roads, and bridges, creating the need for larger investment in the affected municipalities. We construct a dummy variable taking the value of “1” in municipalities where the intensity was beyond seven on an MSK-64 scale that measures damage and destruction for years 2010, 2011, and 2012. We also control for several project characteristics, such as (a) number of projects submitted in the previous year; (b) whether the project relates to

68

What works in Latin American municipalities?

a design project (first phase) or execution project (second phase); (c) average cost of all municipal grants applied for; and (d) the number of grant applications related to education, justice, or sports policy (because the central government requires regional governments to devote at least 2% of their budgets to these three areas). Finally, because projects may be contingent on the electoral cycle, we control for administration year. This is a categorical variable, and the model includes the second, third and fourth administration years, which are compared to the first administration year.

MODEL AND VARIABLE DEFINITION Tables 2.2 and 2.3 provide the estimations for two dependent variables: effectiveness in funds acquisition as number of projects approved relative to total number of applications, and total money awarded relative to total money requested.10 The unit of analysis is the municipality-year. The same independent variables are used in both models since we want to test for differences in factors that influence the percentage of the number of projects approved and the percentage of money awarded. For each dependent variable, we use fixed-effects, random-effects, and Arellano-Bond estimations. The variance inflation factor (VIF) suggests that multicollinearity is not an issue. Because we used a panel data set, our preferred estimation model is fixed-effects, which allows us to control for time-invariant unobserved characteristics at the municipal level.11 The Arellano-Bond estimates allow us to control for the “stickiness” in the process, to address the possibility that project preparation in a given year can build upon previous years’ work. All regressions use cluster-consistent standard errors to correct for heteroskedasticity and serial correlation within clusters. Effectiveness in Funds Acquisition (Projects Approved over Total Number of Applications) Our administrative capacity hypothesis receives strong support with respect to local effectiveness in funds acquisition. Models 1, 2, and 3 in Table 2.2 report the estimations for the effectiveness in infrastructure funds acquisition as total number of projects approved in a given year. Results are consistent across the three models. All measures of administrative capacity are significant at the 1% level and with the expected signs. For instance, holding all else constant, one additional employee increases funds acquisition effectiveness by 0.1%, whereas one additional unit of expertise (one more previously funded project) increases effectiveness by 2%. Similarly, collaborating with the regional and central governments increases municipal effectiveness: a 10% increase in the percentage of projects on which there is inter-organizational cooperation

Administrative capacity and Chilean local governmental effectiveness

Table 2.1

69

Descriptive statistics

Variables

(1) N

(2) Mean

(3) SD

(4) Min

(5) Max

Effectiveness (number)

2,893

0.418

0.299

0

100

Effectiveness (money)

2,893

0.438

0.354

0

100

Administrative personnel

3,048

105.5

158.7

2

1,952

Expertise

2,887

5.217

11.64

0

210

Collaboration-Regional

2,893

0.093

0.187

0

100

Collaboration-Central

2,893

0.012

0.0845

0

100

Dependent variables

Administrative capacity

Political factors Electoral competitiveness

3,105

0.164

0.138

-3.2*

82.7

Party alignment

3,105

0.488

0.500

0

100

Legislative support

3,105

0.466

0.194

0

100

Population (log)

3,105

9.906

1.378

5.493

13.74

Poverty

3,001

17.15

8.764

0**

58.33

Earthquake

3,105

0.113

0.317

0

1

Rurality

3,105

37.89

30.03

0

100

Average cost (million Ch$)

2,893

428.5

736.9

2.375

26,930

Execution phase

2,893

0.763

0.254

0

1

Central funding

2,893

0.095

0.187

0

1

Self-funding

2,893

0.026

0.098

0

1

Specific sectors

2,893

0.354

0.277

0

1

Second administration year

3,105

0.222

0.416

0

1

Third administration year

3,105

0.222

0.416

0

1

Fourth administration year

3,105

0.222

0.416

0

1

Revenues (billion Ch$)

3,091

5.928

11.59

0.095

172.6

Total number of applications

2,893

7.510

6.501

1

69

Right’s ideology

3,105

0.358

0.480

0

1

Controls

Notes: * In one municipality, the mayor died after being elected. The second runner took his place, having in practice a negative margin of victory. ** Two municipalities have an effective poverty rate of zero.

increases the effectiveness by 6% (regional government) and 4% (central government), all other things being equal. Therefore H1, H2, and H3 receive empirical support. Political factors seem to play an important role with respect to the party alignment. On average and all else equal, when the mayor and the governor

70

Table 2.2

What works in Latin American municipalities?

Effectiveness in infrastructure grants approved (projects approved/projects requested)

Variables

(1) Fixed-effects

(2) Random-effects

Effectiveness lagged

(3) Arellano-Bond 0.069** (0.030)

Administrative capacity Administrative personnel Expertise Collaboration-Regional Collaboration-Central

0.10***

0.01***

0.01***

(0.00)

(0.00)

(0.00)

1.10***

1.01***

0.013***

(0.10)

(0.10)

(0.10)

0.685***

0.672***

0.719***

(0.035)

(0.030)

(0.043)

0.407***

0.344***

0.262**

(0.112)

(0.111)

(0.120)

3.61***

2.80**

1.90

(1.20)

(1.10)

(1.70)

-0.043

-0.038

0.041

(0.045)

(0.032)

(0.057)

-0.054

0.005

0.052

(0.056)

(0.046)

(0.078)

-0.149

0.001

-0.495**

(0.166)

(0.009)

(0.244)

0.002*

0.000

0.004**

(0.001)

(0.001)

(0.002)

-0.083***

-0.082***

-0.088***

(0.020)

(0.017)

(0.025)

0.000

0.000

0.004

(0.003)

(0.000)

(0.005)

-0.001

-0.003**

0.006

(0.003)

(0.001)

(0.004)

0.000

-0.000

0.003**

(0.001)

(0.001)

(0.001)

-0.000***

-0.000***

-0.000***

(0.000)

(0.000)

(0.000)

0.110***

0.103***

0.149***

(0.026)

(0.024)

(0.035)

0.227***

0.228***

0.225***

(0.040)

(0.038)

(0.044)

Political factors Party alignment Legislative support Electoral competitiveness Controls Population (log) Poverty Earthquake Rurality Revenues Total number of applications (lag) Average cost Execution phase Central funding

Administrative capacity and Chilean local governmental effectiveness

Variables

(1) Fixed-effects

(2) Random-effects

(3) Arellano-Bond

Self-funding

-0.033

0.020

-0.055

(0.062)

(0.058)

(0.109)

0.001

0.011

0.013

(0.023)

(0.022)

(0.026)

-0.022

-0.010

-0.044*

(0.016)

(0.012)

(0.024)

0.056***

0.055***

0.062***

(0.014)

(0.014)

(0.015)

0.071***

0.073***

0.065***

(0.014)

(0.014)

(0.015)

0.011

0.012

0.001

(0.013)

(0.013)

(0.014)

1.593

0.166

4.758*

(1.683)

(0.103)

(2.472)

Observations

2,631

2,631

2,217

R-squared

0.325

Number of municipalities

340

340

334

Specific sectors Right’s ideology Second administration year Third administration year Fourth administration year Constant

71

Notes: Robust standard errors in parentheses. *** p