Essays in International Trade

The first chapter quantitatively examines the impact of exporting countries' reputations for product quality on aggregate trade flows. I introduce a novel data set in which recall incidences retrieved from the Consumer Product Safety Commission are matched to U.S. import data from 1990-2009. Using a model of learning I construct a measure for exporter reputation where consumers internalize product recalls as bad signals. Structural estimation of the model finds that reputation is important and especially impactful for products used by children. The market share elasticity of exporter's reputation is around 1.49 across products, similar in magnitude to the average price elasticity, which is around 1.51. Improving reputation can increase export value, but reputation is sluggish: increasing reputation by 10\% can take decades for most exporters. Counterfactual exercises confirm that quality inspection institutions are welfare improving, and quality inspection is especially important for consumers of toys.

The second chapter summarizes the correlation between export decisions of Chinese firms and product recalls for Chinese products. I use a new data set where I link recall data scraped from CPSC to monthly Chinese Customs Data. I found that recalls from previous months correlates negatively with the decision of export participation, but not with export value.

The third chapter, coauthored with Kendall Kennedy and Xuan Jiang, analyzes how China's industrialization and the immediate export growth due to the Open Door Policy change Chinese teenagers' education decisions, which explains the education decline. We find that, middle school completion rates increased and high school completion rates decreased in response to export growth. This suggests a tradeoff between education and labor market opportunities in China. These education effects are more prominent for cohorts who were younger when China's Open Door Policy began, even though these teenagers also faced a stronger education system compared to the earlier cohorts.