Co-authored with Vladimir Tyazhelnikov
We introduce a two-country quantifiable model of international production of a complex good that consists of multiple parts. Production of any pair of parts in the same country exhibits pair-specific cost complementarity and, consequently, decisions on production locations are interrelated, which generates rich patterns of clustering. We use the World Input-Output Database to estimate the model for the case of US production and find that these complementarity costs are highly heterogeneous between industries and, on average, account for 1.94% of total production costs. Gains from trade associated with the decrease in these costs, which we interpret as non-tariff trade liberalization, are 3-6 times larger compared to the case of a decrease in tariffs. Finally, using our model, we construct an index of international integration, which, unlike standard measures, accounts for not only total costs of parts produced abroad but also what parts are produced and how similar these parts are between each other. We find that this index is a better predictor of the welfare consequences of trade liberalization than conventional measures of offshoring.
I develop a multi-country general equilibrium model on global sourcing which considers individual firm’s decisions on outsourcing as well as offshoring. These decisions are closely connected as more extensive offshoring provides incentives for further integration of inputs. The firm-level decisions aggregate to produce gravity style equations of trade flows between countries, and intra-firm transactions.
Close and Up: Pricing Responses to Temporary Closures in Retail Gasoline Market
Using station-level fuel price data in New South Wales, Australia, I identify service stations that were temporarily closed and find that surrounding stations tend to increase their markup by over 6% during the closure period. At the same time, inter- and intra-temporal price dispersion and price cycle characteristics showed no significant changes. Furthermore, detailed FuelCheck usage data shows consumer search intensity remained stable during closure periods and likely didn't contribute to the observed change in markup.