What makes presidents centralize different types of advisory units within the Presidency? The literature on institutional presidency claims that the presidents of single-party governments centralize different types of advisory units depending on the political, institutional, or contextual constraints they face in solving governance problems. In addition to the informational function, advisory units are centralized to reassure the public, ease pressures on the president in a serious situation, draw attention to a specific problem, and boost executives’ political support through publicity. We argue that also in multiparty governments the structural patterns of advisory structures are associated with the different functions that they have in governability.
Presidential advisory structures (PAS) are defined here as the structures centralized at the Presidency in charge of policy-making and policy-implementation tasks related to specific issues (i.e. racial equality, climate change, etc.). Presidents use their discretion and administrative powers to centralize these units through executive action unilaterally. The level of institutionalization of these units (low, intermediate, high) and it varies considerably regarding their duration, the number of members, and the internal specialization of their structures.
Our main argument is that in multiparty governments the presidential choice between types of PAS is associated with situations as constraints to governability, emergencies, and decline in public support to presidents' agenda . Especially, we expect that poorly institutionalized structures are mobilized mainly in face of the president's goals to coordinate the coalition government and leading collective decision-making processes. Yet, we argue that intermediate structures are centralized mostly in face of emergency situations that require immediate and unprecedented decisions in specific policy areas. Finally, we expect that highly institutionalized structures are mobilized by presidents as a strategy to expand the political support of the civil society for their agenda. For testing these arguments, our empirical strategy is an analysis of PAS mobilized by presidents of Brazil, a case of multiparty government, from 1990 and to 2017. Using an original database, we use ordinal logistic regression models to estimate these effects. Also, we include the following control variables in the analysis: honeymoon, election year, the effective number of parties in the Lower House, and cohesion of the partisan majority.