Xiang Zhang

We provide the first large-scale causal estimates of demand-side agglomeration spillovers in the non-tradable service sector, leveraging data on 851 grocery store openings across the United States between 2019 and 2022. We combine a convolutional neural network with propensity score matching to identify credible counterfactual opening sites and compare business outcomes between actual and counterfactual locations. Grocery store openings increase foot traffic to nearby businesses within 0.1 miles by 23.4 percent within 6–12 months, with spillovers decaying sharply beyond this radius. The effects are strongest for wholesale and retail establishments and hospitality services, and are substantially attenuated after COVID-19—from 42.7 to 14.6 percent—consistent with a trip-chaining mechanism. Grocery anchors also reshape the local business landscape: within 0.1 miles, establishment growth rates rise by 7.0 percentage points and employment growth rates by 8.8 percentage points, averaged over the first three post-entry years. Decomposing establishment growth rates, the effect is accounted for by higher entry and fewer exits.

We develop a tractable quantitative model of automation and non-neutral technical change with dynamic labor mobility and capital accumulation. Given data on an economy somewhere along the transition path to an unobserved steady-state, we show how to solve for this unobserved steady-state and the counterfactual transition path in response to shocks. We invert the equilibrium conditions of the model to recover shocks that exactly rationalize the observed employment and wage data. We find relatively rapid convergence to steady-state, such that initial conditions are relatively unimportant in explaining the observed employment and wage dynamics. We find that intermediate-wage occupations have experienced the greatest automation, whereas high-wage occupations have witnessed the greatest productivity growth. We find that the combination of automation and productivity growth explains most of the observed employment and wage polarization in the data. We compute an inequality-output frontier, which summarizes the impact of automation and productivity growth on wage inequality and aggregate output, and can be used to assess the effects of emergent new technologies, such as artificial intelligence.

The recent surge in trade and industry policy uncertainty has led to growing disruptions in global value chain relationships, posing additional risks for exporters. I study the effects of relationship termination risk on exporter performance, investment, and search decisions. Event-study analysis highlights that the loss of major international buyers has a long-lasting detrimental impact on exporter performance. Firms losing major buyers experience a decline in export value and an increased likelihood of exiting the export market. These firms intensify their search efforts after the loss, yet their newly formed relationships tend to be short-lived. Inspired by these facts, I build an export dynamics model with search frictions and endogenous relationship-specific investment. Calibration results indicate that exporting firms face substantial search costs, as suggested by the existing literature. Moreover, relationship termination risk, and its subsequent influence on firms' endogenous choices, are primary factors explaining the observed firm dynamics.

This paper uses two large panel data sets in China to study the effects of a health shock on household income mobility from 1991 to 2016. We compare outcomes of households with a member who receives a health shock with comparable households that do not receive any health shocks. To do so, we match on demographic and worker characteristics of household members. At the aggregated level, a health shock lowers the probability of "getting out of the low-income trap" by 8.4 percentage points. At the household level, a health shock lowers household income per capita by 13.1% and income position by 3.9 percentiles. We show that the reduced labor supply at the extensive margin and decrease in labor productivity, measured in hourly wage, explain the majority of households' income loss. Households who become poor due to a health shock continue to exhibit lower income mobility after three years.

How do the labor market dynamics of inventors affect firm innovation? We provide new empirical evidence on how research productivity changes following inventor job moves, both for the moving inventors and for their former co-inventors who remain at the firms they leave. We link patent records to inventors' employment histories and use a matched event-study design to compare similar inventors moving to young firms and large incumbents. We find that inventors moving from startups to large incumbents experience a 47% increase in patent output. However, their departures are followed by a 17% decline in patenting among former co-inventors who remain at startups. We find no comparable decline among co-inventors following inventor departures from large incumbents. These findings reveal an asymmetry in inventor mobility: moving to large incumbents is associated with higher productivity for inventors but lower patenting among their former collaborators at startups. To quantify the aggregate implications of these effects, we develop a model of creative destruction with inventor mobility and examine how the allocation of inventors across firms shapes innovation and firm dynamics.