This dissertation examines the aggregate and sectoral implications of the self-selection of workers with different skills into sectors and occupations.
The first chapter concerns the low growth rate of labor productivity in services in the United States that accompanied the large expansion of the sector. The slow growth of labor productivity in this sector may be caused by the selection of less productive
workers into services as the sector expands. Using panel data for US workers, I document that workers moving into professional services are more productive than an average worker in the sector. On the other hand, workers moving into education, health, and the public sector are less productive than incumbents. I build a generalized
quantitative multi-sector Roy model that allows for these rich patterns of selection. Compared to a conventional selection model where new workers arriving at all services sectors are negatively selected by assumption, the estimated model in
this paper leads to an effect of selection on labor productivity for professional services that is 14 percentage points higher, and a weaker negative effect for education, health, and public services in the U.S. between 1989 and 2019. Overall, selection leads to little effect on labor productivity in aggregate services, contrasting sharply with the prediction of a conventional selection model.
The second chapter, joint with Siddharth George, studies the potential gains in aggregate productivity from mitigating the intergenerational persistence in occupations. Workers are much more likely to enter the occupation of their parents around
the world. In this chapter, we develop a measure of this dynastic bias for occupations – the odds ratio of the probability of choosing an occupation conditional on whether one’s father is in that occupation – for over 90 countries with data from 275 censuses or national surveys. At the occupation level, we document that the dynastic bias is increasing in the eliteness of occupation but decreasing in the average years of schooling. At the aggregate level, we document that the dynastic bias first rises and then falls with both GDP per capita and the average years of schooling of an economy. To evaluate the aggregate implication of the dynastic bias on productivity, we build a quantitative Roy model of occupational choice with entry barriers into occupations that depend on whether one’s parent works in that occupation or not. We calibrate the model and perform counterfactual exercises to infer the potential gains in productivity for all economies in our sample. The entry barriers of occupations are set to match the dynastic bias of each occupation. We then remove all entry barriers and solve for the new equilibrium so that workers pursue their occupations entirely based on their comparative advantage without any income penalty. This frictionless counterfactual leads to labor productivity gains at the aggregate level that average
at 8.6% for all censuses in our sample. Focusing on the latest censuses of the 51 non-high-income countries with data after 2000, we find that the average potential gains in labor productivity remain considerable at 8.7%, with 20 countries having
potential gains above 10%.
| Identifer | oai:union.ndltd.org:bu.edu/oai:open.bu.edu:2144/47015 |
| Date | 22 September 2023 |
| Creators | Shu, Martin |
| Contributors | Lagakos, David, Restrepo, Pascual |
| Source Sets | Boston University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
| Rights | Attribution-NonCommercial-NoDerivatives 4.0 International, http://creativecommons.org/licenses/by-nc-nd/4.0/ |
Page generated in 0.0022 seconds