Research

Working Papers

I develop a multi-country heterogeneous firm model to study the aggregate effects of multilateral trade policy over time. The model captures the slow evolution of production and trade networks in response to trade policy as firms make durable and irreversible investments in source-specific productive capacity and destination-specific exporting capacity. It also incorporates capital, international assets, firms, and endogenous labor supply while still matching world geography. The model is calibrated to match size and trade flows of the US and its major trade partners as well as the split of trade between consumption, capital, and material goods. I find that the short run fluctuations in the economy following a policy change are a key determinant of the overall gains from trade and that transitions are not simply represented by gradual convergence to a new steady state. Futhermore, I find that the long-run effects of trade are poorly approximated by quantitative models without dynamics. While all the model features are important, the behavior of the domestic economy in the short- and long-run relies most on the semi-fixed trade networks and intertemporal trade incentives. The model is used to evaluate the effects on the US of being left out permanently or temporarily from a world trade liberalization. Being left out is quite costly, with reductions in utility concentrated in the initial periods of the liberalization.

  • "Trade Policy is Real News: An Analysis of Past, Present, and Future Trade Costs" with George Alessandria, 2018

We evaluate the aggregate effects of changes in current and future trade barriers in a dynamic quantitative two-country model in which trade responds gradually to changes in trade policy and trade policy changes are gradual. We capture the growth and trade factors driving the economy with movements in productivity, investment efficiency, the labor wedge, and trade costs. Our model offers insights into how changing trade barriers affects the economy and how business cycle shocks can affect trade. We find that a fall in current trade barriers has an expansionary effect while a decline in future trade costs can be recessionary on impact. Conversely, cancelling agreed upon declines in barriers is expansionary in the short-run but substantially lowers growth over the next ten years. We use Bayesian estimation to match the time series on trade integration and business cycles since 1980. Our estimation yields a path for current and expected future trade barriers and allows us to decompose the source of aggregate fluctuations. We find that trade barriers have been expected to decline but that these declines have been repeatedly delayed. Aside from these delays that the outlook on future trade barriers did not change much between 2012 and 2017. Repeating the estimation in a static model yields different paths of trade costs and a different decomposition of growth. Our model is well-suited to consider the impact of alternative unilateral and bilateral changes in trade barriers being contemplated.

Work in Progress

  • "Multilateral Trade Costs and the Global Trade Integration"
  • "The Trade Collapse and the Structure of Global Trade"