This dissertation consists of three chapters on international economics and housing economics. These chapters explore the intersection between housing economics and international macroeconomics and analyze various economic implications of the housing market. The first chapter studies the housing market's role in international business cycles. The second chapter examines the links between housing purchase and portfolio choices. The third chapter studies the effect of land supply policies on regional economies.
In Chapter 1, I analyze the role of housing market in international business cycles. Despite its distinctive features, such as large expenditure shares and inelastic supply, housing service has received scant attention in the international macroeconomics literature. To fill this gap, I examine the role of housing in international business cycles for eurozone countries. I show that housing rents exhibit larger variations than the prices of tradables and other nontradables, both in cross-country and time series. In addition, among all prices, housing rent stands out as the dominant contributor to both the Balassa-Samuelson effect and the negative Backus-Smith correlation. By simulating eurozone economies using a two-country model with a realistically calibrated housing sector, I show that the cross-country distribution of sectoral productivities, inelastic housing supply, and its interaction with the wealth effect via incomplete markets are key to understanding the empirical moments of real exchange rates. Compared with the standard model, the model with the housing sector generates larger variations of the real exchange rate, a stronger Balassa-Samuelson effect, and more realistic Backus-Smith correlations.
In Chapter 2, I study the effect of housing on households' portfolio choices. Although numerous studies have examined the crowding-out effect of housing on stock holdings via the house price risk channel and the liquidity constraint channel concurrently, separate influences on the crowding-out effect via the two channels have received less attention. In this paper, by exploiting a unique Korean housing tenure type called jeonse, which affects a household's investment decision only through the liquidity constraint channel, I study both effects separately. A calibrated life-cycle portfolio choice model with endogenous housing tenure choice and stock market participation shows that the liquidity constraint channel only affects young households and households with a low net wealth-to-income ratio and does not affect old or wealthier households. The house price risk channel, on the other hand, affects all types of households, including households with a high wealth-to-income ratio. Regressions using a household level panel survey show that the crowding-out effect of jeonse exists only for households with a low net wealth-to-income ratio and young households, whereas the crowding-out effect of homeownership affects all types of households.
In Chapter 3, we explore the effect of two different types of land use policies on regional economies by using regional level data of South Korea. We analyze the effect of conventional land-use restrictions in existing cities as well as the effect of unique land-supply policies, motivated by the South Korean government's 2nd New Town Project which built new cities from the scratch, supplying 666,000 houses near Seoul Metropolitan Area (SMA). We estimate the effect of such policies on the aggregate and regional economies, considering both the efficiency gain from the resource reallocation and externalities from regional decline. If the government introduces tighter land-use restrictions in SMA and relaxes land-use restrictions in other regions, regional population will more evenly distribute at the cost of aggregate GDP loss and universal housing price increases. Our back-of-the-envelope calculation indicates that the 2nd New Town Project was cost effective as it permanently increased steady state real aggregate GDP flow by 0.4%, for the one-time cost amounting 4.05% of GDP. It however exacerbated regional decline by decreasing overall rural population share by 4%.
