Essays on Urban Economics

Abstract

In this dissertation I aim at documenting costs and benefits of cities through the lenses of transport economics. In the second chapter of this dissertation, I analyse on the long-term economic effects of transportation infrastructure on the reorganization of economic activity and the growth of cities. Differently, in the third and fourth chapters, I focus on two of the negative externalities that within city transportation imposes on urban residents: traffic and noise. In chapter two, titled "Paving the way to modern growth: the Spanish Bourbon roads", I analyse the local impact of increasing accessibility through transport networks. Specifically, I aim at estimating the potential impact that the improvements in accessibility associated to the construction of the new road network had on the population growth of Spanish municipalities between 1787 and 1857. One of the main challenges of this type of analysis is endogeneity. The advantage of the Spanish road network is that it did not seek specifically for routes with high potential trade but was instead explicitly designed to directly connect Madrid to the main cities of the country. To reinforce my case, I also adopt an inconsequential-unit approach, removing from the analysis the termini and other cities targeted in the network design, and considering only the municipalities which got accessibility due to their random location in the middle of two important cities. I find that the increase in market access associated to road accessibility had a substantial effect on local population growth. The impact was substantially higher on the municipalities that had a more diversified occupational structure. By contrast, the effect of the new network on population growth was negative in municipalities close but without direct access to the roads. I interpret these findings as evidence of a process of rural-to-rural migration due to the new roads. In the third chapter of this dissertation, titled "Low emission zones and traffic congestion: evidence from Madrid Central", I exploit the implementation a low emission zone (LEZ) in Madrid to ascertain whether LEZs have an effect on traffic. LEZs are areas to which the access is restricted for the most polluting vehicles. They have been found to be effective in reducing pollution, however, there is a lack of exhaustive evidence about their effect on traffic and car use. I benefit from the exogeneity of the implementation timing to traffic dynamics to develop a pre/post panel fixed-effects analysis. Results suggest that the implementation of Madrid Central led to an overall small increase in traffic for the whole city of Madrid. Nevertheless, this average result hides important spatial patterns in terms of traffic dynamics. In fact, the implementation did reduce traffic in the restricted area. This traffic relief for the treated district is offset by an overall increase in transit in the other areas of the city, which I interpret as displacement effect. Using heterogeneity analyses, I further identify which of the city’s streets are most negatively affected by the displacement, as well as showing that the reduction in the city centre gradually decreases over time and eventually disappears seven months after the implementation. I find that the policy led to a renewal of the vehicles fleet, with a shift towards cleaner and exempted cars. Finally, I look at potential changes in commuting and I identify a switch to public transport for commutes directed to the restricted area and rerouting of trips for destinations outside Madrid Central as two of the possible mechanisms explaining these results. In the fourth chapter, titled "The price of silence", I study whether street noise is capitalised into housing prices, using housing transactions and noise data from Barcelona. To the best of my knowledge, my paper is one of the first ones estimating the capitalization of street noise into housing prices in a causal way. Specifically, I combine hedonic price estimates with a fixed-effects model, and I benefit from a peculiar architectural feature of my framework, to isolate the effect of unobservable and precisely estimate the effect of interest. My aim is to exploit the spatial variation in very granular data on street noise and housing transactions data. The grid pattern and the square blocks of the Eixample district of Barcelona provide a good setting to exploit variation in traffic noise intensity between properties belonging to the same building block but exposed to different noise levels. Specifically, I exploit variations in noise levels within 150x150 square meters area. I do so by including in my model building blocks fixed effects. I do find the price of silence to be sizable in the Eixample district of Barcelona. Specifically, my results suggest that street noise leads to a price depreciation of 1.6% for sales’ posted prices. In other words, referring to the average price in my estimation sample, I find that moving from one category to another of noise exposure (i.e., increase of 5db) induces a price reduction of about 7,250€. Differently, I do not find an effect for rents. Furthermore, looking at heterogeneous results I find that evening and night, more than daily, are driving the price depreciation. I also find the negative effect on posted prices for sales is eight times larger for recreational noise (i.e., nightlife noise) than the one related to traffic. Differently, pedestrian noise is found to have a positive effect of housing prices.