Housing is a crucial channel through which migration affects the local economy and has been increasingly mentioned by politicians defending immigration limits or even the expulsion of immigrants. Yet, most of what we know about the effects of migration on housing focuses on the inflow of migrants. This paper empirically quantifies the impact of an outflow of immigrants on local housing by studying one of the largest ethnically motivated migration shocks in U.S. history: the United States Mexican repatriation of the 1930s. Using a novel automated matching technique to link houses across the 1930 and 1940 Censuses, we show that repatriating Mexicans during the Great Depression had negative consequences for U.S. housing. Employing an instrumental variable approach, we show that Mexican-occupied houses experienced a disproportionately large devaluation of their values and rents in cities more exposed to repatriation. We also find that U.S.-born homeowners who resided close to Mexicans also experienced a decrease in their home values. Critically, the repatriation depressed aggregate wealth and growth in U.S. cities: it decreased building permit growth rates, the median house value growth, and the median rent growth at the city level. Our results suggest that repatriations have a long-lasting negative impact on housing wealth.

We study the effect of the bilateral trade integration with China on wage inequality in Brazil. Previous studies have documented the contribution of trade opening to the decline in inequality since the 1990s, driven primarily by cross-firm pay differences. We find a sharper reduction in wage inequality over the 2000s, parallel to China’s accession to the WTO. Our reduced-form analysis of the China shock suggests that some firms are harmed by import competition, while others profit from increased exports and cheaper inputs. We rationalize these patterns by extending the theoretical framework of Helpman et al. (2017) to include sector heterogeneity in trade exposure and firm-level selection into imports. Our calibrated model indicates that the rise of China led to a reduction in cross-firm wage inequality in Brazil since the cross-sectoral effect—which tends to benefit low-wage sectors and hurt high-wage sectors—dominates the within-sector increase in inequality due to a rise in importers and exporters.

This paper studies how and why the bilateral trades in manufacturing and services differ in their response to changes in the determinants, both theoretically and empirically. We build a unified theoretical framework that incorporates a demand bias towards services and a difference in the degree of national product differentiation between the two product groups. Estimation results support the theoretical predictions. The empirical model includes, among others, two non-standard trade-cost variables: a measure of internet penetration and virtual proximity (the number of bilateral hyperlinks). An important finding is that virtual proximity—thus far ignored in most gravity models—is a strong predictor of aggregate trade in both services and manufacturing. Also, physical distance is an important determinant of bilateral trade in manufacturing and services, even while controlling for virtual proximity.

We study the role of port efficiency on international trade, estimating the impact of vessel turnaround time on Brazilian exports. The main empirical challenge is to control for non-observed local factors that determine trade flows. This paper addresses this challenge by combining detailed data of Brazilian exports with an empirical strategy that allows us to control for various unobserved local determinants of exports. We use a unique database with vessel turnaround time at each port and city-level exports, including information on the Brazilian port used, the destination country, and products. The empirical strategy relies on a difference-gravity equation to explore the variation in port procedures turnaround. This approach controls for unobserved characteristics and determinants common to geographically close cities, exporting the same product to the same destination country. The results suggest that port delays are associated with decreased volumes of exports and decreased product variety. We find that each additional hour of port procedure delay is equivalent to a reduction in relative local exports of 2%. On average, a 10% relative reduction in vessel turnaround time increases the number of exported product categories by 1%. Our findings suggest that delays in port procedures represent costs to Brazilian exporters, affecting both the intensive and extensive margins of trade.

Is ethnic segregation in Europe driven by native flight or immigrant self-isolation? If the former, which natives avoid immigrants? Which immigrants? What is the geographic scope of homophilic residential preferences? We answer these questions using a matched panel containing the universe of individuals and properties in Denmark from 1987 through 2017. We take advantage of the quasi-random nature of refugee placements and simulated exogenous Markov-chain predictions to generate experimental variation regarding local immigrant arrivals. We find strong evidence of native flight, even at the building level. Flight is stronger among the old and a reaction to the arrival of low-income immigrants. As neighborhoods become more immigrant-dense, housing prices decline, and subsequent move-ins are more likely to be other immigrants or young, low-income native citizens without children.