Heterogeneous groups and rent-seeking for public goods

https://doi.org/10.1016/j.ejpoleco.2007.03.004Get rights and content

Abstract

We present a model of endogenous public good provision and group rent-seeking influence. Specifically, two groups with different preferences over public policy and different sizes engage in rent-seeking or lobbying activities to influence policymaking in their preferred direction. When there is within-group cooperation in lobbying, both groups neutralize each other in the political process. Without within-group cooperation, the free-rider problem in lobbying makes the smaller group politically influent. In both cases, the total level of rent-seeking activities is shown to be increasing in taste heterogeneity but decreasing in group size asymmetry.

Introduction

Many countries or societies are divided into several groups of people, whose preferences and sizes are heterogeneous. This two-dimensional heterogeneity can arise from ethnic cleavages (as in the US or many African countries) or different geographic locations. As a consequence, economic policy is often driven by a political game contest which involves a struggle between one group that defends a certain policy and other groups that challenge it by fighting for other policies. The conflict may be peaceful, as in well functioning democracies where interest groups lobby government for preferential treatments, or may be violent, as in less developed countries in which government functions poorly or is unable to enforce the law.

Most of the literature on conflicts stresses the diversity of the groups either in terms of preferences or in term of relative sizes. There is, for example, a burgeoning empirical literature on the impacts of preference heterogeneity and ethnic divisions on various economic outcomes such as growth, corruption, trust or funding of public goods.1 But, in this literature, the groups are politically inactive and do not seek to influence policy in their preferred direction. Following Olson's (1965) contribution, there is, however, a traditional and sizable literature that analyzes political and collective action that is undertaken by a group to counter similar action by competing groups. The main focus of this literature is on the aggregate potency of a group as a function of its relative size and concludes that the free-rider problem makes the smaller group more politically effective.2 This literature, however, abstracts from the diversity of the groups in term of preferences.

The motivation of this paper is, therefore, to seek to understand how diversity in preferences and in sizes between groups affects the conflict over policy. One may intuitively think that the greater the ‘polarization’, the stronger the conflict is and the higher the level of rent-seeking and influence activities is. The precise objective of this paper is to carefully investigate the nature of ‘polarization’ between groups in terms of preference heterogeneity and group size asymmetry.

For our purpose, we model a union consisting of two groups that differ both with respect to their relative intensity of preferences for public good versus private consumption and with respect to their membership sizes. The government provides a global public good financed by general taxation. Since the two groups do not have the same preferences over public good consumption, a conflict of interest arises between members of the two groups. This conflict plays out in a political process which is assumed to take the form of a lobbying game. More precisely, the process of policy decision-making is represented as a simple two-stage game. In the first stage, each group simultaneously and independently exerts rent-seeking pressure (or lobbying) on the central government in order to influence policy and public good provision in its preferred direction. In the second stage, the government chooses the level of public good by maximizing a weighted sum of the welfare of both groups. We assume that the weight of each group in the welfare function of the central government depends on its lobbying effort or rent-seeking expenditures in the first stage of the game.

The relation between the relative sizes of groups and their political influence is a central element in our analysis.3 This is because there may be a free-rider problem in collective action or lobbying (Olson, 1965). We, therefore, consider two situations: one in which individuals within each group make their lobbying decisions non-cooperatively and the other in which they make these decisions cooperatively. In the first case, the less populated group benefits from a larger relative weight compared to its relative size in the objective function of the government. As a result, the smaller group successfully influences the policy of the government in its preferred direction.

When there is within-group cooperation in lobbying, the influence of one group exactly cancels out the influence of the other group. Aggregate influence is then equal to zero and the government supplies the same level of public good as without rent-seeking. Still, individuals in both groups are politically active and the aggregate level of influence activities depends upon differences both in sizes and in preferences for public spending. However, because lobbying and rent-seeking is a public good to the agents in a group, the social waste is lower when there is a free-rider problem in lobbying than in the case where individuals, within each group, act cooperatively.

In the two situations (with and without within-group cooperation), we have the following results. An increase in the polarization of preferences for public good consumption acts to increase rent-seeking efforts expended by both groups. The intuition is that the greater the difference in preferences, the stronger the conflict of interest is and the more each group has to spend to counteract the influence of the other. We also show that the level of influence activities of each group is always positively related to its size even though free-riding and the sub-optimality of collective action increases due to a non-cooperative behavior in lobbying. Indeed, when a group grows larger, it increases its rent-seeking effort but its antagonist counter-attacks by raising its own effort even more, which results in a weaker political influence of the growing group. Finally, for a given size of the total population, the total level of rent-seeking activities is shown to be decreasing in group size asymmetry. We can then conclude that conflicts are stronger when the society is divided into two groups of similar size but with very different preferences over policy. This is in accordance with some empirical evidence. For example, Collier and Hoeffler (1998), using a world-wide sample over the period 1960–92 show that the probability of a civil war is maximized when society is fractionalized into two ethnic groups of fairly similar size.

This paper is related to the literature on rent-seeking initiated by Krueger (1974), Posner (1975) and Tullock, 1967, Tullock, 1980.4 Most of the literature on rent-seeking assumes that groups or individuals compete directly over a rent (a public or a private good) that is produced without costs. In addition, the level of the rent is fixed and does not depend on individuals' preferences. There are two important differences between this literature and our analysis. First, interest groups compete for political influence, which in turn can affect the policy of the government and public good provision. Second, and more importantly, the source of the funds with which the rent (or the public good) is produced is present in our analysis. The consequence of these two departing assumptions is that the size of the rent is not exogenous but endogenous. This size results from the rent-seeking process and may be lower or higher than the optimal level.5

Hence, while the literature typically suggests that homogeneity among the contenders tends to intensify the competition (see Hillman and Riley, 1989, Nti, 1999) we show the contrary, namely, that heterogeneity exacerbates lobbying and rent-seeking for public goods. As for the distribution of population among groups, we find that the aggregate level of rent-seeking activities depends negatively on group size asymmetry. This is also in contradiction with earlier work. For example, Katz et al. (1990) and Ursprung (1990) show that neither the sum of individuals in both groups, nor the distribution of individuals among groups affects the total amount of rent-seeking for pure public goods. This is because, when a group becomes larger, the free-rider problem within each group – which, ceteris paribus, lowers individual effort – just counterbalances the increased number of contributors for the prize. Riaz et al. (1995) show that this neutrality result does not hold in a general model where the value of the public good depends on the whole consumption bundle, including public and private goods, and also incorporating substitution and income effects. We also show that when groups value the public good differently, both the size of the total population and its distribution among groups affect the social waste of resources but for different reasons than emphasized by Riaz et al. (1995). Indeed, in our analysis, preferences are assumed to be linear so that we abstract from income effects. The difference with the Riaz et al.'s (1995) setup is that groups have different preferences and that the size of the rent is not exogenously fixed. It is endogenous and varies with relative group sizes through the groups' rent-seeking efforts. In other words, the Katz et al.'s (1990) neutrality result does not hold in our set-up since the representative individual's effort affects not only the relative influence of the group she belongs to but also the size of the public good.

In another paper, Katz and Tokatlidu (1996) model a two-stage group rent-seeking contest and show that group size asymmetry may yield less rent-seeking whether the rent is divisible or not. However, they can not obtain, unlike in the present analysis, a monotonicity result. They indeed show that group size asymmetry reduces rent dissipation only if the total population is sufficiently unequally distributed among groups. In addition, as in Riaz et al. (1995), the level of the rent is exogenous and there is no preference heterogeneity across groups. By contrast, in our analysis, the size of the rent is endogenous and groups engage in rent-seeking only if they value public good consumption differently. The extent of rent-seeking is indeed driven by the interplay between preference heterogeneity and group size heterogeneity.

The present paper is also related to the analysis of Coughlin et al., 1990a, Coughlin et al., 1990b. In a probabilistic voting model, they study how government size responds to a change in the influence of interest groups and they show that, in general, government size increases with political influence of interest groups. We will see that the influence of interest groups on the public sector level crucially depends not only on the difference in preferences and on the relative sizes of interest groups but also on whether these groups are subject to a free-rider problem in lobbying. With cooperation in lobbying, the size of the government does not respond to political influence. Without cooperation in lobbying, this size is, actually, lower than the optimal level when the smaller group is also the low-preference group for public good consumption.

The remainder of the paper is organized as follows. Section 2 presents the model. In Section 3, we analyze the rent-seeking pressure game when individuals, within each group, act cooperatively and when they act non cooperatively. Section 4 provides further discussion and Section 5 offers some concluding remarks.

Section snippets

The economic environment

Consider a union consisting of two groups of unequal size (indexed by j = A, B). The number of individuals in each group is exogenously given as ni with nA + nB = N. Within each group, all individuals have identical endowments yj and consume a private good and a public good or service. The level of public spending G is decided by a central authority and is funded by a general and uniform lump-sum tax. The level of private consumption for an individual in group j is then xj = yj  c(G) / N where c(G) is the

Rent-seeking and within-group cooperation

We now characterize the first stage of the policy game. Let us first consider that there is cooperation in rent-seeking within both groups. In other words, we assume in this subsection that both groups have managed to solve their collective action problem. One can think of various ways by which groups can circumvent free-riding and achieve partial or full cooperation. For example, this could be done by implementing costly incentives schemes, by direct monitoring of member actions or by exerting

Discussion and extensions15

Some assumptions made in this paper deserve further discussion. First, we have assumed that the cost of public spending is shared equally among groups. In general, if groups are sufficiently different with respect to their preferences or other characteristics, then an optimal allocation may call for differentiated taxes across groups. However, on empirical grounds, most centralized systems of government appear to operate according to uniform tax rules. This is the case, as in France for

Conclusion

This paper focuses on rent-seeking activities and competition between heterogeneous groups for political influence. We consider in turn a situation where individuals in both groups cooperate in lobbying and a situation where both groups fail to solve the free-rider problem in lobbying. In the latter case, the smaller group is politically influent and the social waste is lower than in the first case. However, independent of whether there is cooperation in lobbying within both groups,

Acknowledgments

I thank Laurent Linnemer, Javier Ortega, Wilfried Sand-Zantman, the seminar participants at the Université Montpellier 1, at the fourth international conference on Public Economic Theory, PET04 (Pekin, China) and at the ASSET 2004 Meeting (Barcelona, Spain) for comments and suggestions. I also thank two anonymous referees for comments and suggestions that led to significant improvements in the paper.

References (31)

  • P. Collier et al.

    On economic causes of civil wars

    Oxford Economic Papers

    (1998)
  • P. Coughlin et al.

    Electoral politics, interest groups and the size of government

    Economic Inquiry

    (1990)
  • R. Dawes et al.

    Cooperation

    Journal of Economic Perspectives

    (1988)
  • J. Esteban et al.

    Collective action and the group size paradox

    American Political Science Review

    (2001)
  • M. Gradstein

    Rent-seeking and the provision public goods

    Economic Journal

    (1993)
  • Cited by (30)

    • Rent-seeking incentives in share contests

      2018, Journal of Public Economics
      Citation Excerpt :

      In the context of our application to contests for public funds, our model makes a substantial contribution to the related literature. While scholars have already focused on rent seeking over public policy and public funds, the majority of studies use a conventional quasi-linear utility setup and focus on questions of heterogeneity in group size and composition (Riaz et al., 1995; Katz and Tokatlidu, 1996; Cheikbossian, 2008), or on comparing rent seeking with the alternative of a market (Gradstein, 1993). By imposing such specific functional forms, however, the literature has implicitly constrained the marginal rate of substitution between public funds share and contest (or lobbying) effort: the marginal rate of substitution is assumed to be insensitive to the amount of public funds allocated.

    • Raiding the honey pot: The resource curse and weak institutions at the project level

      2017, Extractive Industries and Society
      Citation Excerpt :

      In countries where there are powerful, organized groups and a ‘honey pot’, those groups all seek their share of the rents, resulting in slower growth overall. Several studies have shown that competition for rents among pre-existing ethnic or religious groups can heighten rent seeking intensity and dampen overall wealth (Baggio and Papyrakis, 2010; Bjorvatn and Selvik, 2008; Cheikbossian, 2008). In a sample of 87 countries, Mehlum et al. (2006) looked at the effect of institutions’ friendliness towards rent seeking.

    • The theory of contest: A unified model and review of the literature

      2013, European Journal of Political Economy
      Citation Excerpt :

      A general result is that regardless of the value of β, a contestant's expected effort can never exceed the prize (Ellingsen, 1991). Cheikbossian (2008) studied two-dimensional heterogeneity in a rent-seeking game for public goods. Unlike the standard rent-seeking model, the prize is endogenous.

    • Firm characteristics and influence on government rule-making: Theory and evidence

      2013, European Journal of Political Economy
      Citation Excerpt :

      For example, rent dissipation is increasing in the number of competing rent seekers, and can be offset when competitors are organized into like-minded groups. Other important factors include how the rules determining within-group rent-distribution are set (Baik et al., 2006), and whether groups, abilities or prize valuations are asymmetric (Cheikbossian, 2008; Szymanski and Valletti, 2005; Nti, 2004). Country-level variables can also influence the extent of rent dissipation.

    • Competitive careers as a way to mediocracy

      2012, European Journal of Political Economy
      Citation Excerpt :

      Skaperdas (1996) offers a generalized and axiomatized ratio-form csf. The ratio-form model is particularly used for the discussion of collective rent seeking where different groups compete against each other, see, for example, Lee and Kang (1998), Lee and Cheong (2005), Baik et al. (2006), Baik and Lee (2007), and Cheikbossian (2008). The paper also relates to the public economics literature on political careers and bad politicians.

    View all citing articles on Scopus
    View full text