The effects of general policy uncertainty on trade flows and U.S. wages

Abstract

This dissertation consists of three essays at the crossroads of international trade and the labor market. We measure the degree of uncertainty using a general and well-established methodology based on Baker et al. (2016). We investigate the degree to which trade policy uncertainty (TPU) at the industry-country-year level affects the global trade flows of major importers and exporters (e.g., the U.S., Canada, China, Mexico, and the European Union). Similarly, we construct the U.S. index of economic uncertainty at the industry-year level to investigate its effects on U.S. wages. In the first essay, we use a text-mining approach to construct a general index of trade policy uncertainty (TPU) for the U.S. and some of its main trade partners. This TPU index captures uncertainty on U.S. trade policy at a very detailed level (partner and industry levels) from 2001 to 2017 based on US trade-related news information. It's general, thereby enabling us to control for uncertainty relative to the use of highly-regulated tariff barriers under the WTO, temporary trade barriers (TTB), export restrictions, and potential reinterpretations of trade-related national security concerns, among others. Results suggest that a one-standard-deviation increase in policy uncertainty tends to decrease U.S. imports by 1.14 percent. In contrast, uncertainty on the trade policy applied by U.S. trade partners tends to reduce U.S. exports only to markets where the importers display a significant market power level. The results also show that the effects of trade policy uncertainty are mitigated with the formation of preferential trade arrangements (PTAs). In the second essay, motivated by the important findings of U.S. TPU effects on U.S. trade flows, we extend the study to another four markets, namely, Canada, Mexico, China, and the European Union, and their trade partners. We construct a TPU index for each of these four markets based on their news information using the same method applied to the first essay. Again, this TPU index captures uncertainty on the trade policies of these four markets at the importer-exporter-industry level from 2001 to 2017. The primary findings of the second essay are very much in line with the previous results. Uncertainty on the trade policy implemented by Canada, Mexico, China, and the EU tends to lower their imports. Specifically, a one-standard-deviation increase in policy uncertainty is associated with a decline of 0.71 percent in their imports. Moreover, uncertainty on the trade policy applied by the trade partners of these four groups is more likely to reduce their exports. Specifically, a one-standard-deviation increase in TPU leads to a decline of 0.62 percent in these four markets’ exports. The impact of trade policy uncertainty on imports and exports for each of the four markets is also negative. In addition, PTAs tend to mitigate the negative effect of trade uncertainties on these four markets’ trade flows. In the third essay, we study the reaction of the labor market to the economic uncertainty in the U.S. We specifically construct the U.S. economic uncertainty index with the same method we used to create the TPU in the previous two chapters on wages. The economic uncertainty index is generated based on U.S. economic-related news information that captures uncertainty on U.S. economic events and policies at the industry level from 2001 to 2018. Interestingly, the increase in economic uncertainty is likely to reduce wages in the U.S. labor market. Our result shows that the total effects of the concurrent and lagged economic uncertainty indexes cause a decline in wages by 2.12 percent. We also get plausible results by constructing alternative U.S. economic uncertainty indices using 1) newspapers released by other countries and 2) other countries' economic uncertainty indexes as instruments.