Bureaucratic competition and public corruption: Evidence from transition countries

https://doi.org/10.1016/j.ejpoleco.2014.04.002Get rights and content

Highlights

  • We consider corruption at the firm level within post-communist countries.

  • We employ OLS, ordered probit, and interval regression methodologies.

  • Firms facing more centralized bureaucracies report less corruption.

  • Results are robust across different purposes for bribes.

Abstract

Some past theoretical models have predicted that bribes paid by firms to government officials are greater under a decentralized bureaucracy where a firm faces numerous officials. A “tragedy of the commons” arises where officials set bribe payments too high and so drive firms out of the industry. Other models predict the opposite as numerous officials in a decentralized bureaucracy bid down bribe payments. Using the Business Environment and Enterprise Performance Survey from the World Bank, a firm level survey covering 24 former communist countries and Turkey, we examine whether reported bribe payments by firms are higher when firms face potentially numerous demands for bribes. We find that bribe payments are higher under a more decentralized bureaucratic structure. This finding is robust across different purposes for which bribes are used but particularly strong for firms that use bribes in dealing with taxes and customs. Results are also robust across different bribe environments.

Introduction

Corruption can be viewed as the inappropriate actions of an agent to further the agent's interests at the expense (monetary or otherwise) of the principal's. The World Bank views such corruption as occurring within the public sector when public officials misuse their office for private gain. Such corruption is treated by newspapers, NGOs, civil society leaders, etc. as one of the “the greatest obstacles to economic and social development” in many developing countries.1 Beginning with Mauro (1995), many empirical studies have found that corruption lowers economic growth or investment.2 A survey of this literature can be found in Svensson (2005) and Asieudu and Freeman (2009). Using household data from Liberia, Beekman et al. (2013) find that corruption could even lower economic activity by 50% at the community level. Corruption could raise uncertainty thereby lowering investment and lower tax revenues which could hamper government financed development projects but corruption could also exert indirect costs, making other institutions less effective. Private rights for some are presumably less secure if other agents can skirt the law by paying officials to look the other way. Uncertainty regarding institutions will also increase if corruption causes a greater degree of capriciousness among government officials. Lower tax revenue might not only deter development projects but also provide less funding for institutions such as courts and other institutions that could improve regulatory quality or government efficiency.3

Given the purported importance of corruption for economic outcomes, a related line of research considers the determinants of corruption. Treisman (2000) considers various historical, cultural, and economic factors to determine as to what extent they can explain the differing degrees of corruption across countries. Using panel data for Italian regions, Del Monte and Papagni (2007) report that political and cultural factors are more important than economic ones in determining degrees of corruption. Apergis et al. (2012) find that corruption is lower in the U.S. states that score higher in economic freedom indices. Pieroni and d'Agostino (2013) obtain similar results using firm-level data across countries as more government intervention accompanies higher corruption.

Others have considered whether the bureaucratic structure influences the level of corruption. Abbink (2004) reports that regular rotation of officials (so that the same officials do not persistently interact with the same potential bribers) reduces corruption. Van Rijckeghem and Weder (2002) examine to what extent corruption declines with increased wages for civil service workers. Others consider the degree of centralization of the bureaucracy. Rose-Ackerman (1978) argues that a decentralized regime would lower bribery as competition across government officials would bid down the amount of bribes. However, Shleifer and Vishny (1993) develop a model where firms must collect several licenses and permits from (potentially) several officials. If these officials do not coordinate, then a tragedy of the commons can result in that burdensome bribe payments could drive some firms out of business and thereby lower the pool from which to extract bribes. Because a single government official acting as a monopolist would incorporate this effect into his demand for a bribe, total bribe payments under the centralized bureaucratic structures would be lower than under the decentralized, competitive ones. Subsequent research has explored these ideas more fully. Blackburn and Forgues-Puccio (2009) create a model where firms need several licenses to operate. They find that coordination among officials lowers the total amount of bribes and so is less harmful to growth.4 Drugov (2010) also considers firm behavior under centralized and decentralized bureaucracies. He finds that greater bureaucratic centralization achieves a better ex post allocation of licenses due to the reasons outlined above.

Given conflicting views on the association between the structure of the bureaucracy and the total amount of bribe payments, empirical examinations are also necessary. Unfortunately, little has been done due to a general unavailability of data on bureaucratic structure. Campos et al. (1999) investigate the role of predictability of corruption on investment. They find that the effect of corruption upon investment is less negative when firms a priori know the amount of the bribe. Presumably, centralized regimes are more predictable than under a decentralized bureaucracy where firms interact with numerous officials.5 Olken and Barron (2009) examine trucking in Aceh in Indonesia where drivers must pay bribes at checkpoints along the route. They find evidence that total bribe payments were greater under decentralized structures.6

This paper considers whether the level of bribes that firms pay is higher when they report greater centralization among government bureaucrats. We use a dataset constructed by the World Bank in cooperation with the European Bank of Reconstruction and Development. The 1999 Business Environment and Enterprise Performance Survey (BEEPS) spans 4100 firms across 24 post-communist countries and Turkey.7 Not only do they ask firms about the extent of their bribe payments, they also delve further into the types of interactions that firms have with government officials, including how likely firms believe that subsequent bribes will follow initial ones. We find that firms report lower bribe payments under more centralized bureaucratic structures. We also find that this result holds across different purposes for what the bribe is used (avoiding taxes, securing government contracts, etc.) and across different bribe environments such as the extent to which bribes are predictable and the degree to which government regulations are viewed as obstacles by the firm.

We find this sample extremely useful as problems of corruption could be of particular interest to these former communist countries. Hillman and Schnytzer (1986) describe rampant corruption within the Soviet Union. Bribery occurred throughout supply chains and bribes were also employed to procure government offices. Even political purges were a tool used to obtain or protect rents by removing rivals. Boettke (2001) describes Soviet communism as analogous to mercantile systems centuries ago where kings would gain revenues by selling monopoly rights to suppliers. Boettke argues that the move away from a market economy allowed the government to extract rents by limiting competition. Corruption then allowed leaders to extract the surplus from the few suppliers. Such high levels of corruption could also be self-reinforcing. Dabla-Norris (2002) develops a model that allows for strategic complementarities as officials are more likely to be corrupt when other officials behave similarly. Charap and Harm (2002) argue that such kleptocratic states will design bureaucratic structures so as to maximize the degree of rent that can be extracted by government officials. Although Boettke (2001) and Hillman and Schnytzer (1986) focus upon the Soviet system, such systems are unlikely to have completely changed following the fall of communism, especially within the same decade as to when the survey that we employ was conducted. Bayar (2011) finds corruption to be persistent over time. Vachudova (2009) describes corruption in the European Union's post-communist members, suggesting that such environments survived the fall of communism. Therefore, corruption has been a longstanding issue for these countries and remains an ongoing one.8

The rest of the paper is organized as follows. Section 2 considers in greater detail the association between bureaucratic competition and bribe payments. Section 3 describes our empirical methodology, including the BEEPS survey. Results are given in 4 Results, 5 Conclusion concludes the paper.

Section snippets

Background

Rose-Ackerman (1978) was one of the first to invoke bureaucratic competition as a possible remedy to corruption. Considering a model which allows clients seeking government services to reapply to another official when a bureaucrat demands a bribe, she shows that competition between a few honest and potentially many corrupt bureaucrats will reduce the amount of bribe and eventually make corruption disappear. The model suggests that competition among government officials in delivering public

Empirical methodology

The first part of this section provides the source of the data. The next part outlines the empirical model and describes the variables within the model in greater detail. The final part discusses the potential weaknesses of our approach.

Results

Table 3 presents the baseline results in panel A. The first column provides coefficient estimates from Eq. (1) given by interval regression.

Bribes increase with the time that representatives of the firm spend with government officials. This result could arise because more time with government officials could proxy for more onerous regulations leading to more bribery in hopes to circumvent these regulations. A greater time with government officials might also indicate that greater uncertainty

Conclusion

Bureaucratic structure has been suggested as an influencing factor of the level of corruption although theory does not agree upon the direction of the effect. Some argue that bribe payments are higher in decentralized systems whereas other models predict that centralized systems lower total bribe payments. In this paper we use cross-country firm level survey data to investigate the effect of bureaucratic centralization upon corruption.

Our analysis suggests that the amount of bribe paid by firms

Acknowledgments

We would like to thank the two anonymous referees, the editor, and numerous seminar participants at Southern Illinois University for their suggestions while absolving them for any possible errors.

References (52)

  • M. Drugov

    Competition in bureaucracy and corruption

    J. Dev. Econ.

    (2010)
  • P. Egger et al.

    Evidence on corruption as an incentive for foreign direct investment

    Eur. J. Polit. Econ.

    (2005)
  • R. Fisman et al.

    Decentralization and corruption: evidence across countries

    J. Public Econ.

    (2002)
  • S. Gupta et al.

    Corruption and military spending

    Eur. J. Polit. Econ.

    (2001)
  • A.L. Hillman

    Expressive behavior in economics and politics

    Eur. J. Polit. Econ.

    (2010)
  • A.L. Hillman et al.

    Illegal economic activities and purges in a Soviet-type economy: a rent seeking perspective

    Int. Rev. Law Econ.

    (1986)
  • N. Kahana et al.

    Endemic corruption

    Eur. J. Polit. Econ.

    (2010)
  • S. Ledyaeva et al.

    Birds of a feather: evidence on commonality of corruption and democracy in the origin and location of foreign investment in Russian regions

    Eur. J. Polit. Econ.

    (2013)
  • F. Mendez et al.

    Corruption, growth, and political regimes: cross country evidence

    Eur. J. Polit. Econ.

    (2006)
  • D. Oto-Peralias et al.

    Does fiscal decentralization mitigate the adverse effects of corruption on public deficits?

    Eur. J. Polit. Econ.

    (2013)
  • L. Pieroni et al.

    Corruption and the effects of economic freedom

    Eur. J. Polit. Econ.

    (2013)
  • D. Treisman

    The causes of corruption: a cross-national study

    J. Public Econ.

    (2000)
  • G. Abed et al.

    Corruption, structural reforms, and economic performance in the transition economies

  • T. Aidt

    Economic analysis of corruption: a survey

    Econ. J.

    (2003)
  • E. Asieudu et al.

    The effect of corruption on investment growth: evidence from firms in Latin America, Sub-Saharan Africa, and Transition Countries

    Rev. Dev. Econ.

    (2009)
  • P. Bardhan et al.

    Decentralization, corruption and government accountability

  • Cited by (27)

    • A vulnerable victim or a tacit participant? Extending the field of multinationals and corruption research

      2022, International Business Review
      Citation Excerpt :

      Using FDI “data archived by the US Bureau of Economic Analysis and corruption data reported by the World Bank,” Lee and Hong’s (2012, p. 949) study showed that “a lower level of corruption by host countries positively affects the profitability of the MNC subsidiaries.” Moreover, Diaby and Sylwester’s (2014, p. 75) study of transitional countries found that “bribe payments are higher under a more decentralized bureaucratic structure.” The cost of corruption to MNEs is not restricted to finance, but also affects reputation.

    • Politicians at higher levels of government are perceived as more corrupt

      2021, European Journal of Political Economy
      Citation Excerpt :

      The impact of decentralization on corruption is a priori ambiguous (Fan et al., 2009; Diaby and Sylwester, 2014).

    • Does corruption hinder investment? Evidence from Russian regions

      2019, European Journal of Political Economy
    • Reducing bureaucratic corruption: Interdisciplinary perspectives on what works

      2018, World Development
      Citation Excerpt :

      Other studies provide further support for the notion that uncoordinated bribe taking in decentralized political systems leads to higher levels of corruption. Based on firm-level survey data for 24 transition countries, Diaby and Sylwester (2014) find that bribe payments are higher under a more decentralized bureaucratic structure, as measured by firms’ beliefs that even if they pay bribes they will be asked by other public officials to make additional payments. Olken and Barron’s (2009) study of more than 6000 bribes paid at checkpoints by Indonesian truckers also provides evidence that decentralized corruption can result in higher total bribe payments, based on a novel natural experiment that exploits an exogenous shift in the number of checkpoints resulting from the Indonesian military’s withdrawal from the Aceh province.

    • Corruption in Africa: What role does ICT diffusion play

      2017, Telecommunications Policy
      Citation Excerpt :

      As stated by the authors, this gives an idea about the black market activity and that of corrupt administration. The complexity of corruption in Africa is different from other regions as a result of over-regulation of private activity, expanded public sector employment, weak institutions, limited foreign direct investment, a high level of external aid and dependence on commodities and raw materials (De Jong & Bogmans, 2011;; Pieroni & d'Agostino, 2013; Diaby & Sylwester, 2014). Many African countries are caught in a corruption persistence trap where three forces tend to perpetuate corruption (Collier, 2002).

    • Corruption and growth in Africa

      2016, European Journal of Political Economy
      Citation Excerpt :

      This is particularly the case for Africa, which has many countries that have both high corruption and low HDI ratings. Corruption in Africa is systemic in many countries and the development impact is likely to be different from other regions as a result of their experience of conflicts and structural adjustment programmes and of the existence of weak institutions, high levels of external aid, limited FDI and economic freedom, dependence on commodities and raw materials and, in many cases, relatively small private industrial sectors (de Jong and Bogmans, 2011; Beekman et al., 2013; Pieroni and d'Agostino, 2013; Diaby and Sylwester, 2014). In addition, the nature of corruption can be more decentralised and disorganised than in other countries (Gyimah-Brempong, 2002) although, when endemic, it is used to finance personal advancement in the government bureaucracy.

    View all citing articles on Scopus
    View full text