Random Regret Minimization Models of Electoral Choice in Taiwan

• In Event: Explaining the 2024 Taiwan Presidential Election

Abstract

The 2024 national elections in Taiwan demonstrate apparent structural changes in the electorate. The traditional two-party system features the two major parties divided over the unification versus independence issue (or generally called “Tondu” issue). Most recent developments however suggest some voters, particularly the younger generations, are becoming indifferent to such issue, preferring neither options for the country’s future relationship with China. The rise of the Taiwan People Party provides a third option, winning support from voters who choose maintaining status quo, emphasizing distribution policies and envisioning a pragmatic approach in building the national identity. The “white” power received 26 percent of the popular votes in the presidential election. More notably, it is unlike the previous third parties which generally are spin-off groups from either Kuomintang (KMT) or Democratic Progressive Party (DPP) and the tag word is “neither”, representing the non-green, non-blue subscription.

In this study, a new model is proposed to estimate the multiple choices in the new party system. The random regret models (RRM) started by Chorus (2012) for analyzing consumer choice are alternatives to the classic random utility models (RUM) which have been the dominant statistical tool for analyzing discrete choices since the 1970s (McFadden, 1974). Unlike RUM's that assume utility maximization, RRM's stipulate that agents minimize the anticipated regret involved in choosing one alternative rather than another. They do this by pairwise comparisons of each attribute of each alternative with the same attribute of other alternatives in the choice set. By relaxing the stringent notion of utility maximization, analysts can construct RRMs that are congenial with some of the major stylized facts emerging from research in experimental economics and cognitive psychology.

Using a combination of Taiwan election surveys, this study will test and compare the RRM with conventional utility based models. In particular, the new model taps into the emotional loss instead of rational gains in Taiwan voters’ electoral calculus and test the hypothesis that the new disdain sentiments toward the two major parties make leeway to a multi-party system. Substantively, we will test the theory of cleavage-party system mismatch and address the question of whether the new cleavage structure will shape the new party system in Taiwan. This study will contribute the literature of cleavage politics comparing the new democracy in Asia to the Anglo-American democracies (Ford and Jennings 2020, Bornschier 2012).

References

Bornschier, Simon. 2012. "Cleavage politics in old and new democracies." Living reviews in Democracy 1 (2009): online.

Chorus, Caspar G.. Random Regret-based Discrete Choice Modeling: A Tutorial. New York: Springer.

Chorus, Caspar, G., Sander van Cranenburgh and Thijs Dekker. 2014. "Random Regret Minimization for Consumer Choice Modeling: Assessment of Empirical Evidence." Journal of Business Research 67: 2428-36.

Hensher, David A., William H. Greene and Caspar G. Chorus. 2013. "Random Regret Minimization or Random Utility Maximization: An Exploratory Analysis in the Context of Automobile Fuel Choice." Journal of Advanced Transportation. 47: 667-78.

McFadden, Daniel. 1974. "Conditional Logit Analysis of Qualitative Choice Behavior." In P. Zarembka, ed. Frontiers of Econometrics. New York: Academic Press, 1974.

Train, Kenneth, 2009. Discrete Choice Modeling with Simulation. New York: Cambridge University Press.

Author

• Karl Ho, University of Texas, Dallas