Biting the Hands that Feed Them? Place-Based Policies and Decline of Local Support: Case of EU Funding in the UK
Whereas empirical research on place-based policies and pork-barrel politics often suffer from selection bias and confounding factors, I found a quasi-natural experiment in the UK to address these concerns. Leveraging the byzantine state-aid rules and discontinuous regional aid categories of the EU based on the regional GDP in 2011, I checked the support for and the evaluation of the EU in similar areas with and without significant EU funding. I merged the British Election Studies geocoded survey data with my original geo-data of the EU-funded programs. Using spatial regression discontinuity design, I showed that the high-skilled or highly-educated population in the area with better EU fund availability are more likely to support the UK’s exit from the EU and believe that the EU made the UK poorer. Then I geocoded most of the EU-funded projects in England and Wales, and I ran the Genetic Matching to check the effect of local projects on the opinion of similar people. I found that local development projects and labor market support funded by the EU made the high-skilled local residents more anti-EU, while research and development, and business support did not affect their support. I concluded that some people are hostile to place-based policies that they are unlikely to benefit directly, regarding these projects as wasteful or inappropriate. While the findings are bewildering, it is consistent with the Eurosceptic attitudes in Eastern European countries with high exposure to EU funds.
Choosing Voters? Partisan Sorting of Voters following Close Municipal Elections in France
Taking advantage of large French administrative datasets of 19 million people before and after the 2014 local elections, I analyzed if the socio-economic characteristics of individuals are related to the choice of their relocation destinations. Using the regression discontinuity design with close elections, I found that the retired people, who overwhelmingly support the Right, tend to migrate into the municipalities with the Right mayor. Curiously, the effect was greater when the Right mayors are narrowly elected. I also found that the narrowly elected Right mayors tended to set low property tax, while the comfortably elected Right mayors did not. I suggested that there could be strategic manipulation by electorally threatened mayors to increase the share of likely supporters by policy measures. As the literature of political sorting is largely theoretical, I contributed by providing rigorous empirical tests and newly introducing strategic incentives of politicians to the discussion.
Power of Immobility: Underrepresentation of Geographically Mobile Voters and its Consequences
I established that high geographic mobility is associated with a low level of political participation. I analyzed American and Swedish survey data, using Genetic Matching to have optimal covariate balance. I found that people who relocated or have the intention to move are significantly less likely to engage in various political activities, even though their political efficacy and engagement in social activities are often higher than those of the stationary people with similar demographic characteristics. With the help of Professor Sven Oskarsson at Uppsala University, I obtained the restricted-access panel data of the Swedish Level-of-Living Surveys including the locational and occupational history of the respondents and found that not only those who recently moved but also those who are to move in the near future show lower election turnout and less attendance in meetings and rallies. In addition, I participated in designing the opinion poll of the Institute of Government Studies and the Los Angeles Times and inserted the questions asking the willingness to move for better opportunities. The results confirmed that one’s readiness to move correlates with lower political participation as well, independent of their actual movement. Cross-sectional analysis using the American National Election Studies data corroborates the findings. While these results are by no means causal, mobile respondents did show a higher level of political efficacy and social capital after matching, which should lead to a higher turnout. Given the paradoxical findings, I formed a simple theory that mobile people are unlikely to voice concerns in politics by virtue of having exit options, and they are less dependent on and less certain about the policies that help them. Mobility is often an overlooked and understudied aspect of political science, yet this bias in politics may have profound consequences.
I hypothesize that the low political weight of mobile voters would lead to policy choices that hinder mobility, such as restrictive zoning. I am using convoluted municipal borders as an instrumental variable for the high rate of inter-municipal migration and looking at how different levels of residential turnover affect local policy choices on housing and development. These theories are original and fruitful fields for my future work.
Too big expectations? Big Infrastructure Projects, Rise and Fall of Incumbent Support, and Strategic Retirement Timing; Evidence from Japan and Italy
In terms of the political benefit of the place-based policies, I also have a project regarding large-scale infrastructure projects and inter-temporal changes in political support for the incumbents. It is in the early stage and will not be included in the dissertation. Using the geographic data of Italian Autostradi and Japanese Shinkansen, I analyzed the local political return of infrastructure investment on the incumbent politicians, as well as their retirement timings. I exploit different phases of the projects in different areas – from planning and assessment to partial or full opening – as well as different fates of the projects – including postponement and scaling down – to disentangle different mechanisms and to create good control areas. Preliminary findings using a simple difference-in-differences approach suggest that the incumbent support goes up in the early stages of the projects and goes down soon after completion, with net negative effects in the very long run. I also found that politicians are likely to retire around the completion of the projects in local areas.
Every unhappy family is unhappy in its own place: how place shapes the association of socioeconomic indicators and their political attitudes.
I indicated that mobile people, whose policy preferences may not depend on local issues, are politically weak. As a consequence, it is imperative to know how local peculiarity affects people’s political preferences, against the backdrop of regional economic divergence. Here, I complement traditional political economy models based on income and assets by introducing local job availability and local living cost structures. For example, urban service sector jobs and semi-rural blue-collar jobs may offer a similar level of income, but workers’ concerns about job scarcity and living costs may differ significantly depending on the location, resulting in different attitudes toward the minimum wage, subsidies, regulations, and unions. Moreover, OECD’s recent report titled “Under Pressure” reported that the rise in the cost of housing and transport far outpaced that of middle-class income across OECD countries. As housing and transport costs show significant geographical heterogeneity, living cost structure may shape people’s political preferences. For example, geographical location mostly determines transport costs, which could deeply affect people’s readiness to accept fuel tax. Such geographic divergence may help explain the current political polarization. With the help of Professor Gabriel Lenz (Berkeley), I applied for the restricted-access ANES data with detailed geocoding to address these issues.