A prominent literature on voting behavior during democratic transition in Mexico focuses upon two considerations: voters' attitudes toward the dominant party and their uncertainty about the consequences of opposition government. These two considerations are said to form the first step of a "two-step" process that voters use to determine which party to support. In this article, I examine the evidence for this argument, using data from both Mexico and Taiwan and conclude that voters in nations with hegemonic parties give greater weight to the public policies of the dominant party than the two-step model argues. These findings have important implications for the opposition's behavior in trying to complete a transition from a one-party dominant government to a multi-party democracy.
The process by which nations transform their governments from authoritarian regimes to competitive democracies is a terribly important problem for political scientists. In the past 20 years, this gradual transformation in Eastern Europe, Asia, and Latin America has created the promise of a more peaceful and prosperous world. At the same time, the transition has moved much more slowly in other parts of the world, most notably the Middle East and Africa. Democratization is particularly important in these areas because the continued presence of authoritarian governments constitutes a great danger to peace.
In nations undergoing the transition from one-party rule to a multi-party democracy, defeating the dominant party is often a key to advancing democratic reforms. In these cases, significant questions concern how voters approach these elections. To what extent are these elections primarily referenda on the dominant party's performance? How willing are voters to support an opposition party with little or no record of governing? What role do policies play in voters' decisions?
Prominent research on voters' behavior during democratic transition in Mexico focuses upon these questions and examines voters' decisionmaking in such a situation. In their analysis of the 1988 and 1991 Mexican elections, Dominguez and McCann (1995,1996: 11) argue for a "twostep model" of voters' behavior:
Voters decide, first, on their view of the ruling party. For those open to the possibility of being governed by another party, but only for them, there is a second step. They support an opposition party, and they choose among such parties motivated by policy preferences and social cleavage attachments.
This implies that attitudes toward the dominant party are the most important consideration for voters in such elections and that democratic transition to multi-party democracy occurs only when enough people are dissatisfied with the governing party because they otherwise do not even consider supporting an opposition party. Policies are relevant only for voters who reject the dominant party. Their hypothesis is foremost a theory consistent with simple retrospective voting theories (e.g., Key 1966) that evaluations of the incumbent influence voter behavior. They also introduce a second condition that voters must meet before considering the opposition; overcoming qualms about supporting a party without any record to indicate how it would have governed had it been in office (Downs 1957).
Dominguez and McCann's (1995, 1996) studies have greatly influenced later studies of electoral behavior in Mexico. For instance, Lawson (1997: 19) refers to this model in citing polls showing "many among the 40 percent of Mexico City residents who say that they would never vote for the PRI remain undecided about which opposition party they prefer" to bolster his argument about calculations the opposition used in determining how best to unseat the PRI. Other studies of Mexican elections build upon their study by examining the role of voter uncertainty about the opposition (e.g., Cinta 1999, Morgenstern and Zechmeister 2001, Poiré 1999).
While this research focuses upon electoral politics in Mexico, the theory may apply to other developing multiparty democracies with dominant parties where voters have an inability to evaluate the opposition's ability to govern. …