Politics and Quasi-Markets
In political terms, the ‘easy phase’ of expanding provision and coverage of basic social services drew to a close in the 2000s in many middle- income countries (Stein 2005; Corrales 2005). Extending the provision of healthcare and education to new areas and target populations generally confronted a congenial task in coalition building. For example, everyone can support the building of a new school: builders get contracts, local and national politicians get credit, unions get additional members, and constituents get new services. However, shifting reform from quantity to quality requires government intervention into management and rewarding and punishing employees. The benefits are few (and the beneficiaries do not always see them, certainly not in the short run), while the losers are politically well-positioned (unions).
As a general framework, it is useful to represent the triangular relationship not only by looking at the flow of funds and services—from policymakers to providers, and from these to clients—but also at relations of political accountability, between providers and policymakers and between governments and voters. These relations highlight the existence of two superimposed principal-agent problems that complicate the provision of social services (Fig. 2.2). On the first level, the service provider is the agent and the policymaker the principal, on the second, the policymaker is the agent and the client/citizen the principal.
If voters could force politicians to act according to their preferences, and if policymakers could guarantee that providers would do as instructed— the long route of accountability—then hierarchical public provision would work fine. But, a number of problems compromise accountability at each stage. For various reasons, many political systems fail to hold politicians accountable to voters regarding the quality of services. In principle, in a well-functioning democracy a majority of voters who are poor and/or poorly served by existing social welfare systems should be able to vote in favor of new politicians who will improve service delivery. In practice, in most political systems, several factors weaken this long route. For one, many poor voters are effectively disenfranchised by low voter turnout, lack of information, or manipulation through vote buying and clientelism (Stokes et al. 2013).
At the high end of the income distribution, many families have exited the public system and rely exclusively on private schools and healthcare (as well as private security and sometimes even water and electricity). These more politically active voters then have little interest in supporting parties and politicians that prioritize improving the public provision of social services (Rothstein 1998). Moreover, systemic features of the political system such as weak parties and overrepresented rural areas weaken the accountability of the legislature to all voters.
Broadening the analysis to include other power resources besides votes such as campaign financing, lobbying, collective action, and strikes and street demonstrations suggests further reasons for politicians to discount poorer citizens. There are two groups in particular that attract politicians’ attention away from the long route of accountability, namely unions and lobbies of private providers. Unions in health and education are often among the largest in developing countries, especially after the contraction of the manufacturing sector in recent decades. In some polities, these unions are closely allied with particular parties, while in others they can elect ‘their’ legislators (who are subsequently in a position to logroll and bargain with other legislators). In other words, union organization can amplify the voices of providers relative to other voters.23 In more immediate terms, unions in Latin America have often used their disruptive power to pressure governments through strikes, street demonstrations, work to rule operations, and so forth (see Stein et al. 2005; Mizala and Schneider 2014a).
Private providers, because of their dependence on public funds, usually conform to standard Olsonian expectations, and organize to guarantee the flow of resources. Even if they do not seek rents, private providers have strong incentives to invest in politics to defend their interests. These incentives are especially strong because the assets are specific and cannot be easily deployed to other uses; schools, hospitals, electric grids, sanitation systems, and other public utilities cannot be converted to alternative uses if governments end or reduce payments. The assets of construction companies are less dedicated to public works, but they have high fixed costs and are subject to volatile shifts in public spending. So, firms should be expected to invest in collective action to build strong associations and lobbying operations. They are also likely to set aside funds for electoral campaigns. As with large unions, large firms in proportional representation (PR) electoral systems of the sort that predominate in Latin America can invest in funding ‘their’ deputies whose power in national legislatures is again amplified by their ability to negotiate votes (Schneider 2013).
The overall point is that numerous additional political factors weaken or distort the long route of accountability from consumers of social services to their elected representatives. Much of this is standard Olsonian fare and hardly surprising, but it rarely comes up in discussions of quasi-markets and expanding the scope of private provision.24 However, the collateral damage to politics of creating new private lobbies is quite predictable and should be part of any basic cost/benefit analysis of quasi-market reforms.
In response, some compensatory political measures can work to improve both long and short routes of accountability. Decentralization considerably shortens the long route of accountability. Decentralization cannot by itself redress endemic problems of disenfranchisement of poorer voters or political mobilization by providers, but it can eliminate some of the problems of malapportionment and weak parties at the national level, and provide consumers with greater incentives to participate, especially if decentralization is pursued all the way to the municipal level. In general, decentralization in Latin America increases the likelihood that citizens will make demands on local governments (Montalvo 2009), though Sanchez and Pachon (in this volume) provide a cautionary tale of how national politics (especially national patrons of local politicians) can disrupt the shortened, local route of accountability.
Organized groups of consumers may enhance both long and short routes of accountability. Such associations can partially counter lobbies of providers and unions, or work with them when interests coincide. Moreover, they can provide governments with essential information. Associations of parents, doctors, and patients are much better positioned to collect information on the quality of educational and healthcare services, especially if they have support from, or access to, policymakers. Such support and access, as in putting association representatives on oversight boards, is a further policy measure that can ‘artificially’ strengthen con-
Fig. 2.2 Agency problems in publicly financed social service provision
sumer associations. On the short route, many reform proposals seek to elicit greater participation by organized groups that can serve as additional sources of monitoring and accountability for independent service providers. For example, giving schools and/or parent associations more control over their budgets might increase parent involvement.