Box 6.2

Constitutional Interruptions

One of the central concerns in regard to political institutions and governability in Latin America is the history of interrupted governments. Until a few decades ago, coups d’état and military governments were frequent in Latin America. Most of the region’s countries have steered clear of military governments in recent years, yet constitutional interruptions (when either presidents or congress do not complete the term for which they were elected) are still a recurrent occurrence. Since the early 1980s (or the return to democracy), there have been 14 constitutional interruptions in the 18 countries analyzed in this study. These interruptions (or more precisely, their absence) could be taken as an indication of the institutionalization of the political regime.

Not surprisingly, a high percentage of constitutional interruptions is associated with poor values on measures of several of the features of public policies, especially stability.1 Also not surprisingly, constitutional interruptions are associated with low levels of some of the key variables describing the degree of institutionalization and quality of incentives in the policymaking process. Constitutional interruptions are associated with weak (and obstructive) congresses, dependent judiciaries, and weak party systems.

These findings reinforce the ideas developed in this report. Poorly functioning institutions are likely to breed the type of discontent that makes constitutional interruptions more likely. But such interruptions (and the expectation that they are likely to happen) in turn shorten the horizons of political actors, leading to more opportunistic short-term strategies and to the lack of investment in building policymaking capacities and better institutions.

Even though there are forces that make all these characteristics interdependent, there are institutional configurations that make constitutional interruptions more likely—a topic that has long preoccupied political scientists, particularly in the case of presidential democracies such as those in Latin America. Some scholars have argued that presidential systems, as opposed to parliamentary ones, pose special problems for democratic stability, given the weaker incentives inherent in presidential systems for forming and maintaining coalitions.2 Others have argued that the stability of presidential democracies may be problematic mainly in the context of multi-party systems.3 Recent work has suggested that presidential democracies in which the party in government does not control the legislature are a problem only to the extent that the president is unwilling or unable to form a majority (or near-majority) coalition.4

Following this lead, this study explored the links between constitutional interruptions and the type of government (majority or minority, single party or coalition) in a sample of 98 democratic governments in 18 Latin American countries between 1978 and 2005. The evidence, summarized in the table on the next page, is consistent with the findings of Chasquetti (1999) and Cheibub, Przeworski, and Saiegh (2004). Democratic stability is more at risk in the case of minority governments. In fact, minority governments are five times more likely to suffer constitutional interruptions than governments that have a majority or near- majority in congress. However, whether majorities or near-majorities in congress are achieved through a single party, or through a coalition, does not seem to have an impact on the incidence of constitutional interruptions. Thus, particularly in the case of multi-party systems that tend to produce fragmented congresses, the ability of presidents to form and maintain majority coalitions is a very worthwhile area of inquiry. Chapter 7 further discusses these issues of policymaking in fragmented systems, focusing on the contrasting cases of Brazil and Ecuador.

1 The value for the correlation between constitutional interruptions and stability is –0.667, and is significant at the 99 percent level. Clearly, in these cases, causality is not unidirectional; the process is interactive. Constitutional interruptions are likely to induce shorter horizons, policy volatility, and poor coordination—but poor policies are likely to induce poor economic and social outcomes, which are a breeding ground for constitutional interruptions. For the purposes of this correlation, constitutional interruptions were measured as the share of presidential periods since 1978 or since the return to democracy that have been subject to interruptions, because either the president or the legislature did not complete their mandated terms.
2 For example, Linz (1990).
3 For example, Mainwaring (1993).
4 Chasquetti (1999); Cheibub, Przeworski, and Saiegh (2004).