Are the gains from trade unequally distributed in society? This paper presents new evidence on the distributional effects of trade on education groups in the U.S. through both consumer prices (expenditure channel) and wages (earnings channel). Our analysis, guided by a simple quantitative trade model, leverages linked datasets that cover the entire U.S. economy and include detailed spending data on consumer packaged goods and automobiles. First, we show that the share of spending on imports is similar across education groups, leading to a neutral expenditure channel. This finding stems from offsetting forces: college graduates spend more on services, which are largely non-traded, but their spending on goods is biased towards industries and brands with higher import content. Second, on the earnings side, we find that college graduates work in industries that (1) are less exposed to import competition, (2) export more, (3) are more income-elastic, and (4) use fewer imported inputs. The first three forces cause trade liberalizations to favor college graduates; the fourth has the opposite effect. Finally, we combine and quantify the expenditure and earnings channels using the model. A 10% reduction of all import and export barriers generates a modest increase in inequality between education groups, primarily due to the earnings channel. Welfare gains are 18% higher for college graduates, whose real income increases by 2.03% compared to 1.69% for individuals without a college degree. Reductions of import barriers with China have qualitatively similar implications.
A growing number of empirical studies use identification strategies similar the one proposed by Bartik (1991), where a demand shock in a geographic location is constructed by combining the local composition of industries in the pre-period with national growth rates of industries. We provide a set of conditions on randomness of industry growth rates that ensure consistency of Bartik estimators in the many-location many-industry asymptotic. We highlight the importance of clustering standard errors to account for the correlation of residuals in locations with similar industrial composition and provide a computationally efficient quasi-maximum likelihood estimator.
A broad empirical literature uses "event study" research designs for treatment effect estimation, a setting in which all units in the panel receive treatment but at random times. We make four novel points about identification and estimation of causal effects in this setting and show their practical relevance. First, we show that in the presence of unit and time fixed effects, it is impossible to identify the linear component of the path of pre-trends and dynamic treatment effects. Second, we propose graphical and statistical tests for pre-trends. Third, we consider commonly-used "static" regressions, with a treatment dummy instead of a full set of leads and lags around the treatment event, and we show that OLS does not recover a weighted average of the treatment effects: long-term effects are weighted negatively, and we introduce a different estimator that is robust to this issue. Fourth, we show that equivalent problems of under-identification and negative weighting arise in difference-in-differences settings when the control group is allowed to be on a different time trend or in the presence of unit-specific time trends. Finally, we show the practical relevance of these issues in a series of examples from the existing literature, with a focus on the estimation of the marginal propensity to consume out of tax rebates.
Are diversified firms mere collections of independent assets, or is there anything that glues together different businesses? We explore this question by looking at segment level growth of multi-segment (i.e., producing in several 6-digit industries at the same time) manufacturing firms in Japan. We find substantial co-movement between such segments and show that it can be driven by plant-wide but not firm-wide shocks. Our findings suggest that inputs that are shared firm-wide, such as brand and organizational routines, are not too important for production.
The fraction of a population that is affected by a treatment (the “responders”) may be as important to identify as the average magnitude of the treatment effect. I show that if the distributions of potential outcomes with and without treatment are identified, then the total variation distance between them serves as the sharp lower bound on the share of responders. It can be computed for randomized control trials, instrumental variables, and other empirical designs. I demonstrate the usefulness of the approach in three examples of economic interest, related to behavioral biases in retirement savings, electoral fraud, and student cheating.