Abstract
The field of decentralization studies has undergone many reviews since the most simplistic proposition of transferring “local matters to local hands”. For some scholars (and governments), it seems now clear that decentralization does not mean to “eliminate” central governments from local matters, quite on the contrary: their role in the institutional design of reforms is crucial, for example, by guaranteeing minimal national standards on service delivery or compensating territorial inequalities. This movement towards “centralization”, however, may present a potential drawback to decentralization theories as it diminishes the level of decision-making autonomy of local governments, a key reason to decentralize. If central governments are in fact centralizing the bulk of decision-making power, then what can local governments actually decide and do? Do their creative or innovative local policies make any difference in a scenario of high levels of central government regulation? The recent development of education policy in Brazil illustrates this point. It emerged in the context of a widely supported decentralization agenda during transition to democracy but has been subjected to a series of national regulations by different governments since the mid-1990s. As improvements in this policy area start to emerge, the main question becomes: are these results explained by what local governments did or by federal regulation? Introduction Many are the arguments highlighting the positive effects of decentralization reforms: a political channel in ethnically divided societies, an instrument of democratization and participation, increased government responsiveness, improved provision of public services, bringing local communities closer to governments to quote only a few. Some studies show that decentralization reforms were associated with improved public services, including universal ones, especially in developing countries (for a summary of existing studies, see SHAH, THOMPSON, & ZOU, 2004). FAGUET & SANCHEZ (2008), for example, find considerable improvement in some social and urban infrastructure policies in Colombia and Bolivia after decentralization reforms. In Bolivia’s case, political and fiscal decentralization would have allowed localities previously completely deprived of revenues to expand education services (FAGUET & SANCHEZ, 2008, pp. 1300-1). Essentially, if one departs from the main propositions in the field of decentralization studies, these results would be explained by the capacity of subnational governments to readapt and to innovate in terms of local policies. Access to public education has expanded considerably in Brazil for the last 20 years and it was largely carried out by subnational governments. From approximately 32 million students enrolled in 1991 it reached almost 44 million by 2010, considering only enrolments in public schools run by states or municipal governments. Similarly, education performance of students in public schools has shown signs of improvements in recent years, although still far from the minimal adequate levels of learning. At first sight, these results might suggest that the option of “leaving local matters to local hands” – proposing a causal relation between decentralization and improved government performance – seem to explain the transformations in the Brazilian educational system. However, a closer analytical look shows that the period under consideration is also characterized by an intense role of the federal government, since at least the mid-1990s, and, moreover, independently of the party holding the presidential office. In this sense, although local matters are still important to understand micro-variations, the Brazilian case is an example of decentralization with a strong hand of the central government. The main question guiding this paper is whether this advances could be attributed sonly to decentralization or national policies carried out by the federal government have also to be considered. Although I do not fully answer this question, my preliminary results confirm similar findings presented by (ARRETCHE, 2002; ARRETCHE, et al., 2011) and my previous work (GOMES, 2009) that the format of decentralization in education policies in Brazil is characterized by national regulations combined with subnational implementation of public services. This essentially questions the notion that once decision-making authority is fully transferred to local governments, improved public services will be