Scott Mainwaring, "Electoral Volatility in Brazil," Party Politics, 4 (October 1998), 523-545.

First Paragraph:
For decades, scholars have been interested in patterns of party system change and stability in the advanced industrial democracies (e.g. Przeworski, 1975; Pedersen, 1983; Bartolini and Mair, 1990). Recently, some scholars have started to examine this issue for Latin America (Coppedge, 1995; Mainwaring and Scully, 1995; Roberts and Wibbels, 1997). The study of electoral stability or volatility in Latin America draws attention to profound differences between most Latin America party systems, which are quite volatile, and West European party systems, which remain comparatively stable, notwithstanding a trend toward greater volatility in recent decades. The examination of Latin American party systems poses interesting questions about why they differ so markedly from West European systems along this important dimension and what the consequences are for democracy.

Figures and Tables:
Table 1: Summary of party system volatility, 1982-96
Table 2: Lower chamber vote by party, 1978-94 (% of valid vote)
Table 3: Number of state deputies by party, 1982-94
Table 4: Presidential elections, 1989-94 (% valid votes)
Table 5: Number of mayors elected by party, 1982-96
Table 6: Electoral volatility by region and federal unit
Table 7: Lower-chamber electoral volatility in 12 Latin American countries

Last Paragraph:
A more stable party system is no panacea, but it would help voters identify the parties more clearly, thereby enhancing mechanisms of accountability and representation. It would diminish the electoral prospects of political outsiders like Fernando Collor de Mello, the maverick president elected in 1989 and impeached in 1992; with a more stable party system, it is more difficult for political outsiders to sweep to power. And a more stable party system might encourage stronger ties between civil society and parties; organized interests would have greater incentives to invest in such ties.