2
Introduction
Why do state government bureaucracies in Brazil exhibit different institutional
designs? Why do some states delegate policy-making powers to independent institutions
and create rigid institutional rules and procedures for policy making (some of which are
enshrined in the state’s constitutions and in legislation which requires qualified majorities
for change), while others have kept policymaking informal, flexible, and, as a consequence,
less formalized? What are the possible connections between electoral vulnerability of state
political elites, institutional design and constitutional choice at the state level? What’s the
role of political uncertainty in determining key features of the bureaucratic structure such as
autonomy, rigidity or centralization?
This article explores these issues with reference to the 22 newly created independent
regulatory bodies (IRAs), which were created at the state level in Brazil in the last 6 years
(up to 27 states in Brazil, only six of them have not created these bodies). These institutions
vary significantly in terms of the degree of autonomy they enjoy vis-à-vis the executive, as
well as other analytically relevant features. They therefore provide a quasi-natural
experiment for discussing constitutional choice. The paper looks in particular at the
incentive structure the Brazilian state governors face when they have to choose how to
regulate and what should be the institutional structure and process for the new IRAs. We
argue that the state executive is the key actor in this process since the Governors are the
truly agenda setters, who in addition to be responsible to take regulatory initiative have
many ways to compel State Legislatures to cooperate with them.
In principle, cooperation among politicians should lead to wealth enhancing
policies. However transaction costs in political exchange added to uncertainties generated
by elections can create situations where cooperation is not the likely result. When
delegating regulatory tasks, politicians would like to give the IRAs ample powers and
discretion allowing them to be able to effectively accomplish their mission. But due to the
information asymmetries and costs of monitoring inherent in these relationships, politicians
run the risk that this discretion might be used to pursue outcomes that may harm their
interests (agency costs). It may thus seem surprising that they would be willing to delegate
so often to independent regulatory agencies, since in principle the same tasks could be
accomplished by other non autonomous bureaucratic bodies, which politicians could much