Search results
home Home navigate_next Search results

Search results

Abstract

This study examines the shifts in voter turnout during Taiwan’s two referenda in 2018 and 2021, resulting from the transition from concurrent elections to separate ones. The research reveals a relationship between referendum turnout and the percentage of young voters when they are no longer held concurrently with general elections. In 2018, regions with a higher proportion of young voters experienced higher participation in referenda, whereas in 2021, areas with a lower percentage of young voters displayed increased turnout. Furthermore, even after controlling for the level of higher education among the population and the political party in power at the town levels, an inverse relationship between the proportion of young voters and turnout remained consistent. When referenda were decoupled from general elections, a higher percentage of young voters was associated with lower overall turnout, and the relationship between higher education and turnout was found to be insignificant. When comparing changes in turnout between cities not governed by the Democratic Progressive Party (DPP), a significant increase in turnout was observed in DPP-led areas in 2021. This suggests that when referenda were decoupled from general elections, the ruling party became more proactive in mobilizing voters. Furthermore, an analysis of approval votes indicates that higher turnout encourages more approval votes. However, in the context of reduced turnout during nonconcurrent elections, the proportion of approval votes also decreased accordingly.

Abstract

This explorative study employs a large-scale analysis of individual turnout data from the 2016 and 2020 presidential and legislative elections, alongside the 2018 local elections and referenda in Taiwan, to advance our understanding of Indigenous voting behaviour in three ways. Firstly, we challenge the conventional wisdom that Indigenous voters exhibit lower turnout rates compared to the non-Indigenous electorate. Secondly, we further compare turnout differences among Highland and Lowland voters. Lastly, we analyse turnout variations across different types of ballots: national and local elections and referendum
bills. Contrary to prevailing international and domestic norms, we provide strong empirical evidence that Indigenous voters are diverse and dynamic in their voting behaviours, and this diversity extends not only in comparison to the Han voters but also within the Indigenous population, interacting with the different types of elections.

Abstract

Arguments for and against referendum are often based on certain theoretical hypotheses of its practice. However, very little empirical work, with the exception of a few case studies, has been done to test these hypotheses. As a result, the merits and defects of referendum are still in dispute. In this paper, I compare referendum experiences of Switzland, France, Italy, Britain, and the United States. Based on these experiences, I examine various hypotheses of referendum. I find that referendums in these five countries are often mediated by governments or parties, which help to contain the populist impulses of the masses. In addition, I argue that the impact of referendums often extend beyond their legal premises. As a result, I suggest that more attention should be paid to the agenda setting effect of referendum.

Abstract

When opinions are polarized in a bimodal pattern, parties will occupy the two poles of the political issue, retaining the support of the extremists. According to rational choice theory, on the other hand, voters have inadequate time to collect political information, thus they need a clear shortcut to reach decisions. Our first proposition is that issue polarizat ion will happen when political parties have long-term stances on certain issues. Our second proposition is that individual's issue position will decide one's voting behavior when polarization happens. To prove these two propositions, we examine how national identity and unification/independence issue transformed to the referendum issue in the 2004 presidential election. Our finding shows that whether or not picking up the referendum ballots indeed turned a polarized issue and it influenced voting behavior.

Abstract

This paper analyzes the extent to which performances of the central governments could shape Legislative Yuan by-election outcomes. Based on the retrospective voting theory and referendum voting theory, this research used the survey data of Taiwan's Elections and Democratization Studies (TEDS) to explore whether one's assessments of the central government's performances affect his/her choice in by-elections. There are some findings from this research: first, the result showed Miaoli's first district voters have differing voting behavior compared to voters in Taipei City's sixth district and Yunlin's second district. The assessment on central government performance has no significant influence on Miaoli's voters voting behavior; second, for the voters in Taipei City sixth district and Yunlin second district, the assessment of the central government performance have mixed effects due to the election context. As a result, we may conclude that the by-elections in Taipei City sixth district and Yunlin second district can be regarded as referenda on the central government, but the by-election in Miaoli's first district was different from the others.

Abstract

Taiwan's ruling party, Kuomintang (KMT), handed a loss to the opposition party, Democratic Progressive Party (DPP), in the 2009 Magistrate Elections. The conventional wisdom suggests two competing arguments to interpret the election results-while some argued that the KMT's loss was mainly due to the lack of partisan mobilization within the KMT, some posited that it was because a significant proportion of voters changed their voting preferences from the KMT to the DPP between the 2008 national and 2009 local elections.In fact, both arguments share the same premise: there exists a linkage between the performance of the ruling party (in the central government) and its electoral prospect in local elections. Specifically, the performance of President Ma has a substantial impact on the 2009 local election. This study utilizes survey data to verify such premise. Our data analysis shows that a voter's assessments on President Ma's job performance as well as on general and personal economic conditions affect his/her voting intention. In other words, we found that the ”referendum voting model,” in which voters caste their votes in midterm elections on a basis of their evaluations of the ruling party, properly explains voting behavior in Taiwan's midterm elections such as the 2009 local elections.

Abstract

Political parties enlist voter support from electoral mobilization. It not only helps the party’s candidate win the election but also cultivates political talents and builds the party’s strength for sustainable development. In the election of local public officials in 2018, the Democratic Progressive Party (DPP) led to social controversy due to the implementation of a few reform policies in the central government, which led to a significant reduction in the seats of mayors and county magistrates and a decline in the share of votes. On the contrary, the Kuomintang (KMT) benefited from the atmosphere of “Dislike DPP,” and the “Han Wave (韓流)” originated in Kaohsiung city, which made a huge gain in the mayor and county magistrate elections and got higher vote shares. From the perspective of party politics, beneath the surface of one party’s gain and another party’s loss in the local-level elections, one topic that deserves attention is whether the political power of the political parties in local politics has changed after Taiwan has experienced three times party alternations? And from the perspective of the analysis of the aggregate data, what is the political meaning of the topic of removing the party label in the local election?
This study mainly took the 2018 local public officers’ elections in Taiwan as an example. We collects the data of the local elections in the past 10 years and crosschecks the results of the national referendum held at the same time as the 2018 election. The methods which are used in this study to examine difference and correlation of votes gained by the DPP, KMT, and independent candidates in various levels of local elections included descriptive statistics, data visualization, analysis of (partial) correlation, ecological regression, and principal component analysis. The findings including: (1) the phenomenon of de-labeling of political parties in the lower-level elections is more obvious in local elections, especially the KMT label; (2) while the party label will still reduce the votes of independent candidates, but the party label may also be hidden in the election; (3) the voter’s attitude towards referendum issues can assist in judging their political affiliation and establishing an “alternative” political map; (4) in addition to mayoral elections, the more support the KMT’s position, the less vote for independent candidates.