Current Research

Political Opportunity Structure Moderates the Legacy of Political Violence (Draft)

Previous studies render contradictory evidence linking political repression before and political participation after democratization. This article suggests that the perceived political opportunity structure moderates the effect of political violence: victims, their predecessors, and neighbors will punish the authoritarian successor party only if their district is not dominated by it. If the authoritarian successor party is perceived to win, voters would instead vote for the party. This article examines the hypothesis through the unique political context in Taiwan, where the former authoritarian party KMT is still competitive after the democratization and keeps ruling in some districts. Analysis of a newly published White Terror Dataset including 13,206 victims during the martial law period (1949-1987) shows that, if one district has more White Terror victims, KMT receives even more votes in KMT-dominated districts and even fewer votes in districts where KMT did not dominate. The psychological mechanism of this moderation effect is then supported by a pre-registered survey experiment (n = 910) in Taiwan, showing that the White Terror priming increases KMT’s vote share when KMT already leads in the poll. The result helps reconcile previous findings in the literature and explain the resilience of authoritarian successor parties in new democracies.

Who are the Non-separable Voters? (draft)

Non-separable preference is defined as an individual’s preference on one issue being conditional to the outcome of another issue, which challenges the assumptions of democracy. Given its importance, literature does not examine who are the non-separable voters possibly because of the complexity of the measurement. Studies in political behavior suggest three theories explaining the non-separable preferences: (1) Highly cognitive capacity with strong policy preferences, (2) motivated partisan independence, and (3) non-attitude. This article exploits a new rank order question design implemented right before the 2021 Taiwan referendum, in which two power outages before the voting encouraged the formation of non-separable preferences on two referendums choosing between environmental protection and power supply. The rank order question enables researchers to distinguish different types of non-separable preference rankings. The result of this pre-registered survey (n=910) shows that the majority of the measured non-separable voters are driven by non-attitude instead of policy or partisan concerns, even though these non-separable voters can alter the referendum results at the aggregate level. Non-separable voters are, on average, lower education, lower political knowledge, and more likely to be non-partisan. Its implication to the measurement of non-separable preferences and the function of democracy is finally discussed.

Strategic Ambiguity, Strategic Clarity, and Dual Clarity (with Yao-yuan Yeh, Fang-yu Chen, and Charles Wu, draft)

The U.S. had successfully intervened and prevented the military conflict between China and Taiwan since the 1980s by the Strategic Ambiguity (SA) policy, which discourages both sides from deviating from the status quo (SQ) by not committing to defend or not to defend Taiwan. The recent US-China tensions and the increasing nationalism in China and Taiwan drew critics to SA and suggested it be replaced with strategic clarity. We argue that the choice of Dual Clarity (DC) – the US promises to defend only if Taiwan did not provoke first – is widely ignored. We develop an updated game-theoretical model incorporating the rising nationalism in China and Taiwan and examine the psychological motivations through a pre-registered within-subject survey experiment in Taiwan (n=910). The model indicates that DC’s capacity to maintain the SQ is the same as S, and the survey confirms our theoretical expectations.

Media Literacy and Partisan Convergence across Social network sites (with Yao-yuan Yeh, Fang-yu Chen, and Charles Wu, draft)

Literature highlights social network sites (SNS) as an important channel of consuming news and receiving misinformation. However, people surf multiple SNS daily, such as Facebook, Twitter, Line, etc. People also connect and diversely interact with each other, given the characteristics and settings of SNS. Driven by the partisan bias, a party and its supporters may be motivated to converge to a specific SNS for enforcing the echo chamber. In such a partisan convergence scenario, politicized SNS usage may bias the spread of fact-checking and, therefore, influence users’ level of media literacy. We examine this hypothesis in a pre-registered national survey (n=1060) of the 2020 Taiwan Presidential Election. A media literacy scale, including eight fact-checking items, was created to evaluate Taiwanese voters’ media literacy during the 2020 campaign. Results show that (1) in general, the usage of the private messaging app Line is related to a lower level of media literacy. The effect holds across political and non-political misinformation. (2) DPP supporters are much more likely to use Facebook than other supporters, and therefore (3) Facebook usage is related to a higher media literacy discerning DPP-related fake news but not others, and the effect exists beyond the DPP supporters. 

Stress and coping strategies between domestic and overseas Taiwanese people early during the COVID-19 crisis: government guiding as a stress-coping strategy (with Cheng-che Chen, Harry Wu, and MJ Yeh. Draft)

Overseas Taiwanese participants had a significantly higher level of stress than domestic counterparts (2.89 to 2.69, p<0.001). Domestic participants relied more on the government guiding as coping strategies, which explains lower stress after controlling for demographics, while the government guiding was ineffective for overseas Taiwanese. The effect of residency (p< 0.01) is mediated by the coping strategies, especially for the government’s action and supportive social network.