Counting Heads: A Theory of Voter and Elite Behaviour in Patronage-Democracies Chapter submitted for publication in Herbert Kitschelt and Steven Wilkinson eds, Patrons, Clients and Policies (Forthcoming: Cambridge University Press)

The observation that patronage politics and expectations of ethnic favouritism tend to go together reflects a well documented consensus among scholars who study either patronage or ethnicity. There is no consensus, however, on the cause of this association. The principal variables that have been suggested to account for this association include institutional legacies that privilege ethnic identities; a presumed cultural similarity which makes patronage transactions between co-ethnics easier than transactions with non co-ethnics; and the functional superiority of ethnic networks. This chapter proposes a theory of individual voter and elite behaviour in “patronage-democracies” which identifies a distinct, cognitive, mechanism explaining the association. The voting decision in such democracies, I argue here, is shrouded in severe information constraints. Such severe information constraints produce a self-enforcing and self-reinforcing equilibrium of ethnic favouritism in patronage democracies that should exist in the absence of institutional legacies, cultural similarities, and network ties binding coethnics. Throughout, my focus is on the behaviour of individual voters and elites. In the broader project of which this chapter is a part, I relate the individual microfoundations developed here to the behaviour of aggregates such as political parties and organizations. COUNTING HEADS: A THEORY OF VOTER AND ELITE BEHAVIOUR IN PATRONAGE DEMOCRACIES1 The observation that patronage politics and expectations of ethnic favouritism go together is supported by a well-documented consensus among scholars of patronage democracies. According to Kearney, a student of Sri Lanka: “A common expectation seems to be that a person holding a public office or other position of power will use his position for the near-exclusive benefit of his “own” people, defined by kinship, community or personal loyalty.”2 According to Haroun Adamu, a student of Nigerian politics: “It is strongly believed in this country that if you do not have one of your own kin in the local, state and/or national decisionmaking bodies, nobody would care to take your troubles before the decision makers, much less find solutions to them.”3 Kenneth Post’s description of elections in Nigeria emphasizes much the same point: “It was rare for a man to stand for election in a constituency which did not contain the community in which he was born. It did not matter if he had been educated elsewhere and had his business interests outside the community in which he was born, so long as he regarded it as his home. He would still be a better representative for it than someone 1 This chapter is excerpted from Chapters 1 and 2 of Kanchan Chandra, Why Ethnic Parties Succeed: Patronage and Ethnic Headcounts in India (Cambridge University Press, 2004) 2 Robert Kearney, The Politics of Ceylon (Sri Lanka). (Ithaca: Cornell University Press, 1973), 8. 3 Cited in Richard Joseph, Democracy and Prebendal Politics in Nigeria (Cambridge: Cambridge University Press 1987), 67. who came from outside, who could not even speak in the same tongue.”4 According to Chabal, speaking of Africa in general: “All politicians, whether elected locally or nationally, are expected to act as the spokespeople and torchbearers of their community.”5 And Posner’s investigation of voter expectations in Zambia in the 1990s found that the assumption that politicians in power will favour their own ethnic group was practically “an axiom of politics.”6 This chapter proposes a theory of individual voter and elite behaviour in “patronage democracies” which explains expectations of ethnic favouritism as an outcome of the information constraints that characterize patronage transactions in such democracies. Situations in which observers have to distinguish between individuals under severe information constraints, I argue, bias them towards schemes of ethnic categorization. The voting decision in a patronage democracy is such a limited information situation. Consequently, voters are biased towards ethnic categorizations of the beneficiaries of patronage transactions. Confronted with voter biases, I show why elites are forced to favour voters from their “own” categories in their search for office. And voters, observing in turn that politician help their “own,” but unaware that their own perceptual biases drive elites to adopt such a strategy, place their trust primarily in co-ethnic politicians, leading to a self-enforcing and reinforcing 4 Kenneth Post, The Nigerian Federal Election of 1959. (London: Oxford University Press, 1963), 391. 5 Patrick Chabal and Jean-Pascal Daloz, Africa Works. (Oxford: International African Institute, 1999), 99. 6 Daniel Posner, “Institutional Origins of Ethnic Politics in Zambia.” (Ph.D Dissertation, Harvard University, 1998), 118. equilibrium of ethnic favouritism in patronage democracies. This theory is summarized in the diagram below: Figure 1.1: Self Re-inforcing Equilibrium of Ethnic Favouritism Voting Decision in a PatronageDemocracy Limited Information Ethnic Favouritism Section I elaborates upon the concept of a “patronage-democracy.” Section II lays out the theory identifying the link between limited information, patronage-democracy and a politics of ethnic favouritism. Section III identifies factors which mitigate the information constraints under which the voting decision is made in patronage-democracies and therefore reduce the likelihood of ethnic favouritism. Throughout, my focus is on the behaviour of individual voters and elites. In the broader project of which this chapter is a part, I relate the individual microfoundations developed here to the behaviour of aggregates such as political parties and organizations.7 I use the terms politician or political entrepreneur to mean any individual seeking to obtain or retain elected office. Among politicians, I distinguish between candidates (those who seek to obtain office) and incumbents (those who seek to retain office). In patronage democracies, those who have the capital to launch a political career tend to be “elites” i.e. upwardly mobile middle class individuals, better educated and better off than the voters whom they seek to mobilize. I use the term “elite” interchangeably, therefore, with the terms “politician,” “candidate” “incumbent” and “entrepreneur” in this chapter. I. The Concept of “Patronage Democracy.” I use the term “democracy” here in a minimal sense to mean simply a system in which the political leadership is chosen through competitive elections.8 By the term “patronagedemocracy,” I mean democracies in which the state has a relative monopoly on jobs and services, and in which elected officials enjoy significant discretion in the implementation of laws allocating the jobs and services at the disposal of the state. The term “patronage-democracy” may apply to a political system as a whole, or to a subsystem within it. The key aspect of a patronage democracy is not simply the size of the state but the power of elected officials to distribute the vast resources controlled by the state to voters on an 7 Kanchan Chandra, Why Ethnic Parties Succeed: Patronage and Ethnic Headcounts in India (Cambridge University Press, 2004) 8Samuel Huntington. The Third Wave: Democratization in the Late Twentieth Century. (Norman: University of Oklahoma Press, 1991), 7. individualized basis through their discretion in the implementation of state policy. This individualized distribution of resources, in conjunction with a dominant state, I will argue, makes patronage democracies a distinct family of democracies with distinct types of voter and elite behaviour. A democracy is not patronage based if the private sector is larger than the public sector as a source of jobs and provider of services, or if those who control the distribution of state resources and services cannot exercise discretion in the implementation of policy concerning their distribution. Before going further, let me clarify the relationship between the term “patronage politics” as used in this chapter and other terms which have slightly different meanings but are often used interchangeably: “rent-seeking”; “corruption”; “clientelism” and “pork-barrel politics.” The terms “rent-seeking” and “corruption” typically refer to the sale of public goods for private gain, without specifying whether that private gain takes the form of wealth or political support. I use the term “patronage politics” here to refer to that form of rent-seeking and corruption in which the returns to politicians take the form of votes rather than bribes. The term “clientelism” is often used, especially in anthropological studies, to refer to a dyadic transaction between traditional notables and their dependents bound by ties of reciprocity. While “patronage politics” as used here certainly describes dyadic transactions between voters and politicians, the definition does not require voters and politicians to be connected by traditional status roles or traditional ties of social and economic dependence. In fact, as I will show later, voters and politicians can end up in a relationship of mutual obligation to each other without such pre-existing ties. The use of the term “patronage politics,” thus, is distinct from the traditional anthropological usage of the term “clientelism.” However, in the introduction to this book, the term clientelism is differently defined to mean “a particular mode of exchange between electoral constituencies as principals and politicians as agents in democratic systems.” (Page XXX). This definition of clientelism is consistent with my use of the term “patronage politics.” Indeed, Kitschelt and Wilkinson use the terms patronage and clientelism interchangeably. Finally, the term “pork barrel politics” refers primarily to the practice of courting voter support through policy legislation (especially budgetary allocations). The term “patronage politics” as used here refers to an attempt to court support not by promising some group of voters favourable legislation but assuring them of favourable implementation. For instance, an attempt to obtain the support of farmers by enacting a law providing them with subsidies on inputs would fall into the category of pork-barrel politics. The term “patronage politics” as used here does not describe the enactment of such legislation. However, lets imagine that in order to procure such a subsidy, farmers have first to obtain a certificate of eligibility from some politician with discretionary power over the distribution of such certificates. If such a politician courts the support of some farmers rather than others by promising to selectively employ his discretionary power in their favour, the transaction would be classified as a “patronage” transaction according to this chapter. Although the term “patronage-politics” is often used interchangeably with “pork-barrel” politics,9 this distinction between the two terms is important. The collective transfer of goods to citizens through policy legislation produces different political outcomes from the individualized transfer of goods through policy implementation. II Theory of Voter and Elite Behaviour in Patronage Democracies In a patronage democracy, obtaining control of the state is the principal means of obtaining both a better livelihood and higher status. Elected office or government jobs, rather than the private sector, become the principal source of employment. And because individuals who control the state are in a position of power over the lives of others, it also brings with it higher status. Those who have the capital to launch a political career in patronagedemocracies, therefore, seek political office. And for those who do not, obtaining access to those who control the state becomes the principal source of both material and psychic benefits. Proximity to a state official increases a voter’s chances of obtaining valued state resources and services. At the same time, it affords the voter the chance to bask in the reflected glory of his patron’s power. Patronage-democracies, therefore, produce an overwhelming preoccupation with politics on the part of both elites and voters seeking both material and psychic goods.10 9 For instance, see Barry Ames, Political Survival: Politicians and Public Policy in Latin America. (Berkeley: University of California Press, 1987); Barbara Geddes, Politician’s Dilemma: Building State Capacity in Latin America. (Berkeley: University of California Press, 1994); and Allyson Lucinda Benton, “Stability of Provincial Party Systems.” Ph.D. Dissertation, University of California Los Angeles, 2001. 10 See for instance William Riordon, Plunkitt of Tammany Hall (Boston: Bedford Books, 1994); Joseph, Democracy and Prebendal Politics, 1987. The propositions in this section explain when and why these overwhelmingly politicized populations are likely to organize their struggle along ethnic lines. Propositions 1-8 explain why voters in patronage democracies should expect elites to favour co-ethnic elites rather than others in the distribution of material benefits. Proposition 9 explains why voters expect to obtain psychic benefits also from elites from their “own” ethnic group rather than elites with whom they share other bases of group affiliation. Proposition 10 shows how these expectations result in a self-enforcing and reinforcing equilibrium of ethnic favouritism in patronage democracies. II.1 Politicians in patronage democracies have an incentive to collect rents on policy