Hyeok Yong Kwon, "Unemployment, Partisan Issue Ownership, and Vote Switching: Evidence from South Korea," Party Politics, 16 (July, 2010), 497-521. [Available at http://ppq.sagepub.com/content/vol16/issue4/ ]

First paragraph:
What are the political impacts of issue salience of the employment issue? How do voters respond to individual worries about the employment situation in elections? Despite a strong tradition of research on economic voting, little is known about the link between issue salience of the employment situation and the dynamics of voter choice. Furthermore, what is seldom studied is an individual heterogeneity in the relationship between the employment issue and voter choice.

Figures and Tables:

Table 1. Issue salience in the 2002 Korean presidential election

Table 2. Voter transition in the 2002 presidential election

Table 3. Employment issue and left vote in the 2002 Korean presidential election

Table 4. Robustness check

Table 5. Voter heterogeneity, employment issue and left vote in the 2002 Korean presidential election

Table 6. Effects of the employment issue among middle income voters

Table 7. Multinomial logit analysis of employment issue and left vote

Table 8. Multinomial logit analysis: abstention and voter transition

Last Paragraph:
The findings of this article suggest several implications for political science and public policy. First, the voter transition pattern found in this analysis may be considered as providing an electoral mechanism in line with the continued popularity of larger welfare states in an era of fierce economic competition across countries. When the issues regarding the employment situation and economic vulnerability become salient, electorates tend to demand a greater public provision of social welfare and social protection from the government.

Second, this article suggests a particular pattern of the politics of the middle class. Middle income groups are less likely to be strongly partisan. Instead, it is well recognized that middle income groups tend to be less partisan, more independent voters. Consistent with this fact, this analysis has found that middle income votes are likely to swing contingent upon the salient issue in the election. Middle income voters are probably more susceptible and sensitive to a rapidly changing employment situation than other groups are. As Evans and Andersen (2006) have suggested, when politics are less polarized and partisan identification is not firmly established, issues such as the economy might have 'greater latitude to cause swings in vote choice' (2006: 204). At the income group level, this logic applies to the middle class. To the extent that middle income voters tend to be swing voters critical to electoral competition, the electoral consequences of the issue salience of the employment situation are not trivial at all.