Corruption and Democracy in Latin America

Chapter6: Local Accountability and the Peruvian Vaso de Leche Program

111 Decentralization has become a dominant mantra in many development programs throughout the world. The reasoning seems sound enough. The larger the government unit, the more remote it is from popular control, the less accountable it will become. To solve the problem, decentralization is advocated by many policy analysts as the best way of putting control (back?) into the hands of the people , where public officials will be held accountable, and where public funds have the greatest chance of responding to local needs and local conditions. Moreover, at the local level, where citizens can observe the actions of public officials firsthand, corruption can be difficult to hide and relatively easy to control. This logic goes even further: take public programs out of the hands of public officials and turn them over to local civil society organizations, which, having a popular base of support, will be the most efficient, transparent , and noncorrupt administrators of public services.1 In many developing countries in recent years, as states have shrunk under pressure from neoliberal restructuring (as well as the already noted thesis that local is better), civil society organizations José R. Lópe z -Cálix , M itc hell A . Seligson, and Lorena Alcáz ar 6 Local Accountability and the Peruvian Vaso de Leche Program We would like to acknowledge the work of Erik Wachtenheim for his important contributions to the research that led to this study. 112   Lópe z-Cálix, Seligson, and Alcázar have become the repository of many services that were once run by the state. The classic work by Putnam (1993), which is often cited as the key study to demonstrate the importance of cultural values in promoting democracy , is equally a landmark work about the importance of local government (in the particular case of Italy, regional government) and the role of civil society organizations in “making democracy work.” Those who promote decentralization and civil society organizations are not without their critics. Two main lines of criticism have emerged. First, local governments have fewer resources to institute controls over public spending and to carry out effective audits (López-Cálix and Melo 2004). Second, civil society organizations can often be highly undemocratic and promote the worst form of discrimination, as has been illustrated by Armony (2004) with the examples of civil society organizations promoting the rise of Nazi control in Germany as well as segregationist groups in the American South promoting lynchings of Blacks. In contemporary Guatemala, local civil society organizations have been linked to the wave of vigilante attacks that have become regular occurrences in that country (Seligson 2005). One of the problems in the literature attempting to determine which side of this debate is closer to the truth is that much of it is qualitative in nature, where anecdotal illustrations of the pluses and minuses are deployed as evidence. On the other hand, the quantitatively based literature suffers from serious limitations of scientific control. That is, much of that work involves cross-sectional studies in which the level of centralization /decentralization is contrasted. The problem is that the controls are often too limited to rule out alternative explanations. Longitudinal studies suffer from the same problem; the macroeconomic and other conditions under which the newly decentralized government or civil society organizations are operating differ in many ways from their more centralized predecessor arrangements. As a result, one cannot be sure that it is decentralization rather than some other variable that is responsible for producing the outcome. This study seeks to avoid the weaknesses of prior work. Rather than comparing across space or across time, with all of the inherent limitations in establishing effective control variables, this study examines a single country, Peru, and a single time period, 2002, and focuses on a single program, the Vaso de Leche (glass of milk). It draws upon a detailed study of public expenditures that enable us to trace the use of public funds from their inception in the budget process down to the consumption of the Local Accountabilit y and the Vaso de Lec he Program   113 glass of milk inside households, and in so doing take note of the points in the system in which the loss of the milk occurs. We do not need to make any questionable assumptions about control variables, since we are not varying the place or the time in which the study is being conducted. The focus of the...