Elsevier

Journal of Public Economics

Volume 173, May 2019, Pages 1-20
Journal of Public Economics

Higher pay, worse outcomes? The impact of mayoral wages on local government quality in Peru

https://doi.org/10.1016/j.jpubeco.2019.01.005 Get rights and content

Highlights

  • I study how wages earned by local politicians affect local government quality.

  • Mayoral wages do not improve government performance.

  • There is no evidence of a positive effect on politician selection, municipal bureaucratic capacity, and political effort.

  • Wages increase political competition and reduce political support for the mayor.

  • Political changes may help explain the drop in performance.

Abstract

In this paper, I study how wages earned by local politicians affect local government quality. To identify the effects, I use caps imposed by the Peruvian central government on mayoral wages as an excluded instrument. The results show that mayoral wages do not improve government performance. In particular, there is a negative impact on public investment implementation and on performance goals set by the central government. Moreover, there is no evidence of a positive effect on politician selection, municipal bureaucratic capacity, and political effort. Wages do strongly affect the local political landscape, increasing political competition and reducing political support for the mayor. These changes may help explain the drop in performance as local authorities may face more political obstacles when implementing their policy agenda. Overall, the results show that higher politician wages need not improve local government quality.

Introduction

In recent decades, fiscal and political decentralization reforms have been a common phenomena in the developing world (Bardhan and Mookherjee, 2006). In many cases, policies aimed at strengthening local government capacity did not follow these reforms. This led to local governments failing at their newly acquired responsibilities (Bardhan, 2002, Finot, 2001, Bahl and Martinez-Vazquez, 2013, Sow and Razafimahefa, 2015).

This context raises the question of how to improve local government capacity and performance. Higher wages for local politicians may help obtain better outcomes. However, the theoretical literature on the effect of wages is not conclusive and points to this effect being context-specific (Besley, 2004; Caselli and Morelli, 2004; Messner and Polborn, 2004; Poutvaara and Takalo, 2007; Mattozzi and Merlo, 2008; Gagliarducci et al., 2010).1 The empirical evidence on the topic also provides mixed results (Ferraz and Finan, 2011b, Kotakorpi and Poutvaara, 2011, Gagliarducci and Nannicini, 2013, Braendle, 2015, Fisman et al., 2015, Hoffman and Lyons, 2015). Hence, extending the empirical analysis to different types of authorities and institutional frameworks can provide new useful insights.

This paper studies how local politician wages affect local government quality in a developing country context. In particular, using a novel database on Peruvian municipalities, I show how wages earned by full-time mayors affect a broad set of measures of municipal government quality. To obtain identification, I take advantage of a 2007 mayoral wage reform, introduced by the Peruvian central government, which established caps on mayoral wages. The reform divided local governments into 20 categories based on their electoral population. For each, it assigned a maximum wage the local mayor could earn. I use these caps as the excluded instrument for mayoral wages. Identification is obtained from the fact that wage limits are a step function of the electoral population. Therefore, insofar flexible controls for the continuous effects of the electoral population are introduced and other covariates vary smoothly, the estimates will only capture the effect of wages.

The results show that higher wages have a strong, robust negative impact on local government performance. In particular, increases in wages reduce public investment implementation, which is a key local government performance indicator in Peru. This impact is quite significant. A one standard deviation increase in wages leads to a decrease of 0.58 standard deviations in public investment implementation rates. This effect is not driven by better-paid mayors investing in higher quality, more complex projects or by greater accountability on the part of these mayors.

The negative impact on performance is not limited to public investment. Higher wages also lead to a drop in performance measures based on the Municipal Incentives Plan. This central government-run program establishes diverse performance goals for local governments and rewards municipalities for their accomplishment. I find that mayoral wages have a negative effect on the fraction and value of performance goals achieved.2 Again, the effect is quite strong as a one standard deviation increase in wages leads to a drop of around 0.37 standard deviations in the fraction of goals achieved.

Next, I examine the effect of wages on other relevant dimensions of government quality such as politician selection, bureaucratic capacity, and political effort. In general, I find no evidence of improvements. In fact, higher wages may lead to worse outcomes. Regarding politician selection, there are some negative effects on the mayor's education level and private management experience. There is also no evidence that wages improve the quality of the candidate pool as there are no effects on the fraction of candidates with tertiary and university studies. With respect to the effects on the municipal bureaucratic body, I find no significant effects on bureaucrat selection and on measures of municipal bureaucratic capacity. In particular, better-paid mayors are not able to retain more experienced municipal investment managers and their governments do not exhibit lower needs for assistance and training in municipal tasks. Finally, I do not find evidence of a positive impact on political output, measured by the production of local decrees. Hence, it cannot be concluded that wages increase political effort.

I then study how mayoral wages affect the local political landscape. I find that wages have a strong impact on the political climate the mayor works in. Higher wages increase the number of candidates running for office, lower the voter share of the winner party, and lead to a greater vote split. These effects are large. A one standard deviation increase in wages leads to 1.8 more candidates and a drop of 8.9 p.p. in the winner's vote share. In addition, there is sizable, though statistically insignificant, negative effect on the probability that the incumbent mayor gets re-elected and a positive effect on the probability that the mayor is recalled before the term ends. Hence, higher wages lead to more political competition, opposition, fragmentation and turnover.

The above results on the political landscape may partially explain the negative effects on performance. Increased electoral competition can lead to better outcomes. However, it may also have harmful consequences (Lizzeri and Persico, 2005). In particular, better wages may lead to more political obstacles for the mayor's policy agenda from more numerous, stronger political rivals. This appears to be the case in weak institutional contexts like the Peruvian one (Loayza et al., 2014, Vasquez, 2015). The detrimental effects of an obstructive political opposition may be amplified in a setting where rival politicians have increased leverage on the mayor through frequent social protests and recall voting (Welp, 2016). In addition, excessive mayoral turnover may be disruptive for certain local government functions (Nath, 2015).

A caveat when interpreting the previous results is that, in Peru, municipal council member allowances are limited to 30% of the mayoral wage. This means that the reform caps indirectly affect payments to local legislators. Hence, our estimates can be capturing the effect of an increase in the payoffs for all local politicians. However, the effect via council payoffs should be of second order as increases in allowances are small relative to changes in mayoral wages, council members are part-time politicians and keep other sources of income, and their role is relatively limited.

Overall, the results show that higher politician wages need not improve the quality of local governments. The evidence suggests that increasing wages may have drawbacks when its implemented in a context with weak institutions and weak political parties.

This study complements previous findings on the role of monetary incentives on politician performance and selection. The empirical literature on this matter has focused on legislative authorities (Ferraz and Finan, 2011b, Braendle, 2015, Fisman et al., 2015, Hoffman and Lyons, 2015).3 These have provided mixed evidence both in terms of performance and selection. For example, using a similar approach to the one in this paper, Ferraz and Finan (2011b) find a positive effect on political effort and a slight, positive impact on the competence of local Brazilian legislators. Similarly, Braendle (2015) finds a positive effect on legislative effort among Members of the European Parliament. On the other hand, Kotakorpi and Poutvaara (2011) show that while higher wages have a positive effect on the fraction of female Finish parliamentary candidates with higher education, there are no significant effects for male candidates. Moreover, Fisman et al. (2015) find that higher salaries decrease the average quality of European Parliament legislators and produce no effect on performance.

Less is known for the case of politicians at the executive level.4 These have different responsibilities and motivations for running for office compared to legislators. Hence, their response to wages may be different. Hoffman and Lyons (2015) find that wages have little impact on the performance and selection of US governors. At the local level, Dal Bó et al. (2017) find a positive correlation between wages paid to Swedish mayors and candidate quality. The only evidence on local executive politician performance is provided by Gagliarducci and Nannicini (2013). The authors find a negative impact of wages earned by Italian mayors on the size of municipal governments and red-tape, a result that is driven by better politician selection.

This study complements their findings in several ways. Unlike the Italian case, Peruvian mayors are full-time politicians. The effect of wages can be different as these politicians cannot substitute private work for public work.5 In addition, I analyze the effect of wages on a broad set of local outcomes and dimensions of government quality, which are relevant for decentralized developing countries. Most importantly, I provide evidence of an opposite, negative effect on performance and on a different mechanism based on political opposition to the mayor.

Finally, the results on the political landscape and its connection with the findings on performance contribute to the literature on the effects of political competition on policy outcomes. There is significant evidence on the positive effects of political competition on government outcomes (Keefer and Khemani, 2009; Besley et al., 2010; Ferraz and Finan, 2011a; De Janvry et al., 2012) and politician selection (Galasso and Nannicini, 2011, Dal Bó et al., 2017). However, intense competition and political fragmentation may also have detrimental effects. For example, Lizzeri and Persico (2005) provide a theoretical justification for restrictions on the number of parties. Beach and Jones (2017) find that greater ethnic diversity in California city councils leads to lower public good provision. Nath (2015) shows that lower re-election probabilities for incumbent politicians can worsen bureaucrat performance. While this study does not provide a causal link between lack of political support and worse performance, the pattern of the results supports previous evidence on the effects of political opposition on local government outcomes in contexts such as the Peruvian one (Loayza et al., 2014, Vasquez, 2015).

The rest of the paper is structured as follows. Section 2 provides an overview of the main features of Peruvian district municipalities and the mayoral wage reform. Section 3 describes the data, the identification strategy and the first stage results. Section 4 presents the effect of wages on local government performance. Section 5 shows the results for other dimensions of government quality such as politician selection, bureaucratic capacity, and the political landscape. Finally, Section 6 includes the concluding remarks.

Section snippets

Peruvian local governments

The present study focuses on district municipalities, which are the lowest tier of autonomous sub-national government in Peru. At the start of the 2011–2014 study period, there were 1639 district municipalities. These tend to be relatively small, rural, and poor. The median district municipality has around 3900 inhabitants, with 60% of these living in rural areas, 52% not having access to piped water, and 40% not having access to electricity.6

Data

This study compiles information on several dimensions of government capacity and performance for Peruvian district municipalities. The database includes detailed information on municipal accounts, performance indicators, electoral and recall results, as well as individual-level data on elected mayors, mayoral candidates, and municipal investment managers. The analysis focuses on the 2011–2014 electoral period, the only electoral cycle completed under the new cap system.15

Effect on performance

As aforementioned, Peruvian local governments have a significant role regarding public investment. Investment implementation is a key concern for the population, the central government and municipalities.38

Effect on other dimensions of government quality

In this section, I analyze the effect of mayoral wages on other relevant dimensions of government quality. This analysis can also shed light on what can be driving the main performance result. In particular, I will study the effect of wages on politician selection, bureaucrat capacity, political effort, and political competition.

Conclusion

This paper studies the effect of politician wages on local government quality. Using the caps imposed by the Peruvian central government on mayoral wages as an excluded instrument, I find that higher wages need not lead to better government quality.

The results show that higher wages can have a strong negative impact on performance. Better-paid mayors implement a lower fraction of their investment budgets and accomplish a lower fraction of performance goals. Moreover, I do not find evidence of

Acknowledgments

I am very grateful to Lori Beaman, Georgy Egorov, Timothy Feddersen and Matt Notowidigdo for their helpful advice and support. This paper has benefited from comments and suggestions from two anonymous referees, Fernando Aragon, Erika Deserranno, Seema Jayachandran, Lorena Keller, Cynthia Kinnan, Alessandro Lizzeri, Alexey Makarin, Roger Myerson, Nicola Persico, and various officials at Peru's Ministry of Economy and Finance (MEF). I am also thankful for the comments from the attendees at the

References (40)

  • P.K. Bardhan et al.

    Decentralization and Local Governance in Developing Countries: A Comparative Perspective

    (2006)
  • B. Beach et al.

    Gridlock: ethnic diversity in government and the provision of public goods

    Am. Econ. J. Econ. Pol.

    (2017)
  • A. Beath et al.

    Electoral rules and political selection: theory and evidence from a field experiment in Afghanistan

    Rev. Econ. Stud.

    (2016)
  • T. Besley

    Paying politicians: theory and evidence

    J. Eur. Econ. Assoc.

    (2004)
  • T. Besley et al.

    Does electoral accountability affect economic policy choices? Evidence from gubernatorial term limits

    Q. J. Econ.

    (1995)
  • T. Besley et al.

    Political competition, policy and growth: theory and evidence from the US

    Rev. Econ. Stud.

    (2010)
  • T. Braendle

    Does remuneration affect the discipline and the selection of politicians? Evidence from pay harmonization in the European parliament

    Public Choice

    (2015)
  • S. Calonico et al.

    Robust nonparametric confidence intervals for regression-discontinuity designs

    Econometrica

    (2014)
  • M.D. Cattaneo et al.

    Simple local polynomial density estimators

  • E. Dal Bó et al.

    Progress and perspectives in the study of political selection

    Annu. Rev. Econ.

    (2018)
  • Cited by (13)

    View all citing articles on Scopus
    View full text