The process of deindustrialization has swept large swaths of America's manufacturing heartland and many other advanced economies. Historically, industrial production clustered by location, creating one-industry cities and regions. In many of these places, manufacturing employment has dropped significantly as industrial production has moved elsewhere and economies have shifted toward the service sector. This has had a major impact on many cities around the world, which face high unemployment and poverty rates. Declining firms and local politicians in such places clamor for government support. Yet that support is easier to find in some settings than others. While the current literature on subsidies focuses almost exclusively on national-level price distortions, this research looks at the causes and effects of subsidies and other forms of government aid at the city level, where failing industries can have their biggest political and economic impact.
This project will bridge the gap between research on the provision and effects of government protection -- both why some firms receive support while others do not and the impact of such protection on the incentives of firms to innovate. By doing a comparative study on the provision and effects of government support, the research will show how governmental involvement works as a brake or an accelerator for economic decline in regions that house concentrated industries. These findings have important implications in a world where deindustrialization is rife and old industrial areas face the constant challenge of reinventing themselves.
The research has three parts. First, the researchers will examine the relationship between subsidies and geographically concentrated industries using a quantitative analysis of firm-level data across countries. The researchers will supplement this quantitative empirical work with more nuanced cross-national case studies. The first will be a comparative study of the auto industry in three different electoral systems: the US, the UK and Germany. The auto industry's relatively concentrated production in these three countries allows the researchers to hold economic concentration constant while varying politicians' responsiveness to the industry. The second case study will utilize the change in New Zealand's electoral system. By switching from an electoral system where politicians care about geographic interests to a system where politicians care about national constituencies, the researcher can examine how different electoral systems affect firm and politician behavior in a national context where the geography of industrial production has not changed. Finally, the last part of my research links the provision of subsidies to their effects on industries and local economies. The researchers expect firms that receive subsidies to decrease innovation at the expense of rent-seeking. They will combine an examination of the link between subsidies and levels of innovation using firm-level data from around the world with a more detailed use of micro-data within the United States.
The researchers expect this research to have important implication for our understanding of why government support is allocated to some firms and not others. Equally important, the researchers expect it to provide important insight into why some post-industrial cities and regions are able to make the transition to new, knowledge-based economies fairly easily, while others remain stuck in protracted economic decline.
While often politically popular, subsidies to ailing industries prolong the process of regional decline and the potential for economic recovery. In most cases, governmental attempts to aid declining industries fail since larger economic forces of technology, tastes, and transportation costs are simply too strong for the government to effectively combat. In these places governmental aid only prolongs the death of firms. And yet, why do many declining clustered industries receive rents while others do not? The provision of rents is not random, since all rents involve governments picking winners and losers, and the provision to fund industries is usually hotly contested by politicians, competing industries, and workers. I argue to properly understand the effect of subsidies we must first understand why some industries receive subsidies and others do not. I argue the calculus for the provision of rents lies at the intersection of economic geography and electoral systems. My theory relies upon economic geography to explain the demand for rents within countries, and then looks at electoral systems to explain variation across countries. I argue that the provision of rents is determined by the combination of economic geography and electoral systems. In systems where politicians care about their place in political parties (Party-Centered systems), they will cater to large geographically dispersed industries. In systems where politicians care about their local domestic constituents (Personalistic systems), politicians will cater to geographically concentrated industries. Few studies provide an explanation of both the supply and demand of rents; instead focusing on either within or across country variation. In contrast, my theory of demand and supply provides predictions about cross-country rents, as well as within-country rents. By focusing on the electoral impact of geographically concentrated industries, and a more fine tuned aspect of electoral systems, I clarify predictions about where rents occur and gain insight into their impact. To test my theory about the provision and effect of rents, I utilized a wide array of evidence. To show that electoral systems impact the provision of rents, I look at the case of Thailand, where the 1997 Asian financial crisis was largely responsible for the constitutional change in the same year. Before 1997, politicians only cared about their local districts, and spent most of their efforts extracting rents for important local businesses. However, after the 1997 change in the Thai constitution, political parties became much more powerful, and politicians had electoral incentives to cater to large, geographically dispersed industries. This caused a sea change in the provision of rents to the rice industry in Thailand. Rice went from a relatively neglected industry at the periphery of the government, to the third rail of politics in Thailand and, in a way, rice policy has contributed to the recent military coup. Thailand’s rice policy created a coherent, large, and unified group of voters in the north that disrupted traditional power structures in Thailand. The current coup is a reaction by traditional power brokers in Thailand (Military, Elites in Bangkok, etc.) who cannot accept the problem of increased votes in the north against their policies. To show that economic geography is fundamental to understanding the demand for rents and depressed demand for public goods, I look at the state of Bihar in India. In 2001, due to a decades long independence movement, the state of (undivided) Bihar broke into two smaller states: Bihar and Jharkhand. Before the break, undivided Bihar was economically diverse, with several important industries. During this time, there was little effort on the part of politicians to finance businesses within Bihar. Instead, there was high spending on public goods such as education. After the state broke into two smaller ones, however, this trend changed. Both Bihar and Jharkhand drastically reduced their spending on important public goods, and increased their effort to provide rents to their main industries. This shows how economic geography is pivotal in sculpting the desire for public goods and rents for industries. Each of these studies illuminates different sides of my argument. Together, this evidence builds a cohesive story about the relationship between economic geography, electoral systems, and the provision of rents. Geographically concentrated industries produce electorates with homogenous preferences in favor of supporting established local firms. In electoral systems where politicians are beholden to voters in a narrow geographic constituency, they will support efforts to prop up declining industries even as these measures stymie innovation. Conversely, in electoral systems where politicians are beholden to broad party interests, they will support nationally important and geographically dispersed industries. Concentrated industries, by contrast, are more likely to die a rapid death and leave public resources available for pioneering, new firms. Thus, the intersection between electoral and political geography provides insight into the Schumpeterian creative destruction needed to transform a city into a post-industrial economy.
