This thesis is comprised of four main chapters. Although the chapters are distinct works, they are related by their focus on housing and land markets and their reliance on urban economic theory and methods. They aim to contribute to the understanding of how these spatial markets function in order to work towards an improved implementation of urban policy. In particular this thesis tries to understand how house prices are determined by demand- and supply-side factors across different scales. It provides support for the idea that at a local level prices are determined by demand, in that they compensate for differences in locational amenities. It also investigates some of the consequences of price determination such as displacement of original residents from gentrifying neighbourhoods and welfare losses as a result of planning restrictions to development. The overall message that emerges from the body of work is that urban policy should pay close attention to the way that supply and demand interact to determine prices in markets for housing and land.

The thesis consists of three independent chapters on environmental and urban economics. A central theme explored in this thesis is what determines the distribution of economic activities across space. My exploration in this direction begins with the roles of industrial pollution and transportation infrastructure in shaping the spatial distribution of skills, and extends to evaluate the spatial allocation efficiency of renewable energy projects. The first chapter,“The Long Shadow of Industrial Pollution: Environmental Amenities and the Distribution of Skills”, investigates the role of industrial pollution in determining the competitiveness of post-industrial cities, with a focus on their ability to attract skilled workers and shift to a modern service economy. I assemble a rich database at a fine spatial resolution, which allows me to track pollution from the 1970s to the present and to examine its impacts on a whole range of outcomes related to productivity and amenity, including house prices, employment, wages, and crime. I find that census tracts downwind of highly polluted 1970s industrial sites are associated with lower housing prices and a smaller share of skilled employment three decades later, a pattern which became evermore prominent between 1980 and 2000. These findings indicate that pollution in the 1970s affected the ability of parts of cities to attract skills, which in turn drove the process of agglomeration based on modern services. To quantify the contribution of different mechanisms, I build and estimate a multi-sector spatial equilibrium framework that introduces heterogeneity in local productivity and workers’ valuation of local amenities across sectors and allows the initial sorting to be magnified by production and residential externalities. Structural estimation suggests that historical pollution is associated with lower current productivity and amenity; the magnitudes are higher for productivity, more skilled sectors and central tracts. I then use the framework to evaluate the impact of counterfactual pollution cuts in different parts of cities on nationwide welfare and cross-city skill distribution. The second chapter, “Travel Costs and Urban Specialization: Evidence from China’s High Speed Railway” examines how improvements in passenger transportation affect the spatial distribution of skills, exploiting the expansion of high speed railway (HSR) project in China. This natural experiment is unique because as a passenger-dedicated transportation device that aims at improving the speed and convenience of intercity travel, HSR mostly affects urban specialization through encouraging more frequent intercity trips and face-to-face interactions. I find that an HSR connection increases city-wide passenger flows by 10% and employment by 7%. To further deal with the issues of endogenous railway placement and simultaneous public investments accompanying HSR connections, I examine the impact of a city’s market access changes purely driven by the HSR connection of other cities. The estimates suggest that HSR-induced expansion in market access increases urban employment with an elasticity between 2 and 2.5. The differential impacts of HSR on employment across sectors suggest that industries benefiting more from enhanced market access are the ones intensive in nonroutine cognitive skills, such as finance, IT and business services. These findings highlight the role of improved passenger travel infrastructure in promoting the delivery of services, facilitating labour sourcing and knowledge exchange across cities, and ultimately shifting the specialization pattern of connected cities towards skilled and communication intensive sectors. In the last chapter, “Where does the Wind Blow? Green Preferences and Spatial Misallocation in the Renewable Energy Sector” , I focus on the spatial allocation efficiency of renewable energy projects. How efficiently are renewable energy projects distributed across the US? Are “greener” investors worse at picking sites? Using extensive information on wind resources, transmission, electricity prices and other restrictions that are relevant to the siting choices of wind farms, I calculate the predicted profitability of wind power projects for all possible locations across the contiguous US, use this distribution of this profitability as a counterfactual for profit-maximizing wind power investments and compare it to the actual placement of wind farms. The average predicted profit of wind projects would have risen by 47.1% had the 1770 current projects in the continental US been moved to the best 1770 sites. I also show that 80% and 42% respectively of this observed deviation can be accounted for by within-state and within-county distortions. I provide further evidence that a large proportion of the observed within-state spatial misallocation is related to green investors’ tendency of invest locally and sub-optimally. Wind farms in more environmentally-friendly counties are more likely to be financed by local and non-profit investors, are closer to cities, are much less responsive to local fundamentals and have worse performance ex-post. The implementation of state policies such as Renewable Portfolio Standard (RPS) and price-based subsidies are related to better within-state locational choices through attracting more for-profit investments to the “brown" counties, while lump-sum subsidies have the opposite or no effects. My findings have salient implications for environmental and energy policy. Policy makers should take account of the non-monetary incentives of renewable investors when determining the allocative efficiency of policies.

This thesis is composed of four chapters. The first one investigates the impact of immigration on housing markets. The rest study the effects of transport policy on economic outcomes. Chapter 1 provides causal estimates of the effects of an increase of foreign-born population on house prices. I use data for the Spanish provinces between 2001 and 2010. In order to infer causality I construct an instrument based on past location patterns by immigrant nationality. I find positive effects of the increase in the share of foreign-born population on both rental and purchase prices. The estimated elasticities are 0.6% for rental prices and 2% for purchase prices. I also investigate the relationship between immigration and native location (native displacement) and I find that immigrants attract natives to the same regions they locate. When I re-estimate the effects using solely the variation on population growth which is due to exogenous location of foreign-born, I find that estimates are around 30-40% smaller than if we ignored the relationship between immigration and native location decisions. Chapters 2 to 4 investigate the effects of road improvements on aggregate and individual economic outcomes, using data for Great Britain during the period 1998-2008. Chapter 2 develops the methodology to estimate the economic impacts of transport improvements. We summarise the existing evidence and the theoretical channels through which transport policy can impact firm, worker and aggregate economic outcomes. To capture the effect of road improvements, we construct a measure of accessibility to employment through the road network. For this purpose, we collect novel data on 31 major road improvement projects and combine this information with the trunk road network in Great Britain in 2008. This information is used to calculate optimal travel times between locations at each point in time, which are used in the computation of the accessibility measures. The last two chapters discuss the empirical results, for ward and firm outcomes (chapter 3) and for individual labour market outcomes (chapter 4). I find positive effects of accessibility on ward employment and number of plants, a limited effect on plant employment and no effect on productivity. Accessibility from workplace has substantial impacts on individual wages and total hours worked, while accessibility from home only seems to have an effect on reducing the travel time to work.