Import tariffs tend to be higher for final goods than for inputs, a phenomenon commonly referred to as tariff escalation. Yet neoclassical trade theory -- and modern Ricardian trade models, in particular --predict that welfare-maximizing tariffs are uniform across sectors. We show that tariff escalation can be rationalized on efficiency grounds in the presence of scale economies. When both downstream and upstream sectors produce under increasing returns to scale, a unilateral tariff in either sector boosts the size and productivity of that sector, raising welfare. While these forces are reinforced up the chain for final-good tariffs, input tariffs may drive final-good producers to relocate abroad, mitigating their potential productivity benefits. The welfare benefits of final-good tariffs thus tend to be larger, with the optimal degree of tariff escalation increasing in the extent of downstream returns to scale. A quantitative evaluation of the US-China trade war demonstrates that any welfare gains from the increase in US tariffs are overwhelming driven by final-good tariffs.

Multinational firms (MNEs) accounted for 42 percent of US manufacturing employment, 87 percent of US imports, and 84 of US exports in 2007. Despite their disproportionate share of global trade, MNEs' input sourcing and final-good production decisions are often studied separately. Using newly merged data on firms' trade and FDI activity by country, we show that US MNEs are more likely to import not only from the countries in which they have affiliates, but also from other countries within their affiliates' region. We rationalize these patterns in a unified framework in which firms jointly determine the countries in which to produce final goods, and the countries from which to source inputs. The model generates a new source of scale economies that arises because a firm incurs a country-specific fixed cost that allows all its assembly plants to source inputs from that country. This shared fixed cost across plants creates interdependencies between firms' assembly and sourcing locations, and leads to non-monotonic responses in third markets to bilateral trade cost changes.

This paper presents a tractable model of endogenous production networks with fixed costs associated with the formation of links between firms. The model consists of a finite number of firm types producing differentiated products. Each firm is characterized by firm-specific parameters describing its CES production function, firm-specific domestic and foreign demand shifters, and a firm-specific set of potential suppliers and buyers. We consider versions of the model where either the buyer or the supplier may initiate the formation of links, and versions in which the production network can be cyclic or acyclic. Our main theoretical result is that the closed economy equilibrium is unique if the set of feasible networks consists only of networks that are acyclic and the buyer initiates the link formation while having full bargaining power in price negotiations with the supplier. We provide examples of multiple equilibria if the supplier initiates the link formation in both cyclic and acyclic feasible networks or if the buyer initiates the link formation in a cyclic production network. We take the acyclic production network model to Belgian data on firm-level production networks and show that it approximates well the salient features of the observed production networks. The model suggests a moderate role for the endogeneity of domestic firm-to-firm linkages in shaping the aggregate response to trade shocks.

We quantify and explain the firm responses and worker impacts of foreign demand shocks to domestic production networks. To capture that firms can be indirectly exposed to such shocks by buying from or selling to domestic firms that import or export, we use Belgian data with information on both domestic firm-to-firm sales and foreign trade transactions. Our estimates of firm responses suggest that Belgian firms pass on a large share of a foreign demand shock to their domestic suppliers, face upward-sloping labor supply curves, and have sizable fixed overhead costs in labor. Motivated and guided by these findings, we develop and estimate an equilibrium model that allows us to study how idiosyncratic and aggregate changes in foreign demand propagate through a small open economy and affect firms and workers. Our results suggest that the way the labor market is typically modeled in existing research on foreign demand shocks—with no fixed costs and perfectly elastic labor supply—would grossly understate the decline in real wages due to an increase in foreign tariffs.

We examine the application of quantitative spatial models to the growing body of fine spatial data used to study economic outcomes for regions, cities, and neighborhoods. In “granular” settings where people choose from a large set of potential residence-workplace pairs, idiosyncratic choices affect equilibrium outcomes. Using both Monte Carlo simulations and event studies of neighborhood employment booms, we demonstrate that calibration procedures that equate observed shares and modeled probabilities perform very poorly in such settings. We introduce a general-equilibrium model of a granular spatial economy. Applying this model to Amazon’s proposed HQ2 in New York City reveals that the project’s predicted consequences for most neighborhoods are small relative to the idiosyncratic component of individual decisions in this setting. We propose a convenient approximation for researchers to quantify the “granular uncertainty” accompanying their counterfactual predictions.