Measuring the Effectiveness of Vote-Buying: Evidence from Municipal Elections in Taiwan

This project aims to examine the validity of vote-buying in Taiwan. Specifically, can vote-buying effectivelyinfluence the electoral outcomes? Since it is difficult to monitor voters’ decisions under the secret ballot, theeffectiveness of vote-buying is severely weakened and thus merits further investigation. However, given itsillicit nature, how to capture vote-buying presents a challenging task to empirical research (Hicken, Leider,Ravanilla, and Yang, 2015). To this end, this project exploits the judicial judgments to capture thevote-buying during the 2002-2014 municipal elections in Taiwan. Although the credibility of judicial data isarguably justified by the judicial system in Taiwan, it might still suffer from sample selection bias.In this project, I consider a bivariate probit model to correct the vote-buying data collected from judicialjudgments. Based on the constructed data, I then examine the effectiveness of vote-buying from twodimensions. First, I estimate the effect of electoral competition on the vote-buying by revisiting Cheng,Santos, and Wu (2015). Secondly, I explore the effect of vote-buying on the turnout. The results can beindirectly inferred as the validity of vote-buying. Furthermore, Nichter (2008) proposed to distinguishvote-buying from the turnout buying strategies. Our result also serves to identify the vote-buying strategy inTaiwan between them.