Torun, DavidDavidTorun2023-04-132023-04-132022-09-19https://www.alexandria.unisg.ch/handle/20.500.14171/108280This dissertation theoretically and empirically investigates the international flow of goods between countries and the welfare consequences of globalization. In Chapter 1, I build a quantitative Ricardian model of international trade that explains country-pair and production zeros at the industry level. I model zeros via a bounded productivity distribution and a non-homothetic final goods assembly function. Without productivity caps, trade shares reduce to a standard gravity equation. I develop a novel calibration strategy to fit data on trade flows and production. The framework replicates the empirical distribution of bilateral zeros. Counterfactual exercises suggest that welfare changes after trade-cost shocks are typically amplified when accounting for the extensive margin of trade. Industry-level bilateral zeros change by approximately half the shock size; e.g., a 10% rise in trade costs increases the number of zeros by 5%. Chapter 2 investigates the so-called triangle inequality (TI) in international trade. The TI should, theoretically, hold for any three countries to avoid cross-border arbitrage. When trade costs change, re-routing opportunitiescaptured by the TImight arise because a shipment through an intermediary becomes cheaper under adjusted trade costs. We show that the widely-used exact hat algebra approach is unable to measure potential gains from re-routing. We show that standard empirical estimates of iceberg trade costs often violate the TI; thus being inconsistent with the theory. We propose a model-consistent estimation routine that respects the TI. The resulting estimates allow us to compute the welfare gains from re-routing; these gains are often substantial. Chapter 3 analyzes how the history of trade liberalizations affects aggregate trade flows today. We develop a dynamic general equilibrium model of international trade with heterogeneous firms. The key feature is that serving a market involves higher fixed cost upfront. As a consequence, there is less market exit and entry in response to a trade shock compared to a model without sunk fixed costs. We derive a gravity equation and show that countries that liberalized their trade relationship earlier trade more today. Our theory can help explain a home bias in international trade, and why historic events can have persistent consequences for trade. We provide supporting evidence for the underlying mechanism, exploiting the structure of our model and data on average firm sales. Product-level international trade flows are classified using the Harmonized System (HS). In Chapter 4, we inspect the regular HS classification updates and resulting intertemporal inconsistencies in trade statistics. These updates hinder the comparison across time of products covering up to 44% of world goods trade. Existing methods to standardize different HS vintages either (i) drop numerous product codes over time, or (ii) bulk adjusted codes into large synthetic groups. We develop a simple algorithm to convert trade flows between different HS vintages that avoids (i) and (ii). The conversion estimates are robust to year and country sample choices.enWelthandelGlobalisierungWohlfahrtEDIS-5255International tradeglobalizationwelfare gains from tradedoctoral thesis