INCOME AND DEMOCRACY IN BRAZIL: The richer you are, the more likely you are to prefer parliamentary government to a presidential system

The poorer parts of Brazilian society prefer a presidential system of government to a parliamentary one, according to research by Marcos Yamada Nakaguma, which is published in the December 2015 issue of the Economic Journal.

The study finds that this is because parliamentary systems provide too much power to legislators, which the population, especially the poor, does not trust. In contrast, presidential systems allow the population to vote for both the executive and the legislative branches of government. This increases checks and balances, and allows for more effective political accountability.

The choice of the form of democratic government is a recurrent and controversial issue in many societies, as seen recently in Egypt, Tunisia and Turkey. There are two major systems: a presidential system, in which the executive is directly elected by voters and is guaranteed to stay in office for a fixed term; and a parliamentary system, in which the executive is appointed by parliament and may be removed by a vote of no confidence.

The new study examines the main factors driving the constitutional preferences of citizens over the form of government. It analyses the case of Brazil, where a 1993 referendum allowed the population to choose between a presidential system and a parliamentary one. The research uses a novel dataset of the results of the referendum at municipal level as well as several opinion surveys conducted at that time.

Analysis of the data reveals a surprising new fact: while the Brazilian population in general voted in favour of a presidential system, there was a strong positive relationship between income per capita and the percentage of votes cast for a parliamentary regime across municipalities – that is, the poor (rich) groups in the population voted more for a presidential (parliamentary) system (see Figure 1).

The evidence that the poor citizens voted more for a presidential system is quite puzzling. Indeed, a common view among researchers in comparative politics is that presidential systems are predominant among developing countries because of the interests of local political elites in keeping power centralised in their own hands.

The study shows that the result cannot be explained by the fact that poor citizens simply lacked knowledge about the referendum and, therefore, voted more for a presidential system merely because it represented the status quo. Moreover, the relationship cannot be explained by either local income inequality or the voters'' preferences for a charismatic and populist leader, such as Lula da Silva (see Figure 2).

Among the electorate, a common reason for rejecting a parliamentary system was the absence of direct elections for the executive branch in connection with a lack of confidence in congress (see Figure 3). The general perception was that corruption was widespread among legislators and that a parliamentary system would further exacerbate the problem.

As in many developing countries, corruption was (and still is) a major concern in Brazil. The poor population is especially vulnerable to expropriation by legislators because the lower quality of their local accountability institutions (for example, the local media and judicial courts) makes it more difficult for them to monitor and hold their representatives accountable.

The traditional argument in favour of a parliamentary system is that it allows for the possibility of replacing an unpopular or incompetent government prior to elections. Indeed, many political analysts in Brazil supported it as being the most modern and flexible regime.

Nevertheless, this new analysis suggests that a parliamentary system is also the regime that is most ''intensive'' in political institutions: it requires a strong system of protection against expropriation and a class of politicians that can be trusted to represent well the interests of voters.

''Choosing the Form of Government: Theory and Evidence from Brazil'' by Marcos Yamada Nakaguma is published in the December 2015 issue of the Economic Journal. Marcos Yamada Nakaguma is in the economics department at the University of Sao Paulo.