ESSAYS ON LABOR MARKETS IN PORTUGAL: TOP EARNINGS MOBILITY, LABOR MARKET DISCRIMINATION, AND THE EFFECTS OF LABOR MARKET CONCENTRATION

Abstract

This dissertation focuses on inequality and labor markets in Portugal. I use various administrative data sources to assess topics inequality, social mobility, discrimination against immigrants, and the effects of labor market concentration. The first essay uses the universe of tax filing data, as well as a linked employer-employee dataset covering private sector employees to (i) analyze the trends in income and wage concentration at the top of the Portuguese earnings distribution, (ii) quantify the changes in composition at the top, and (iii) estimate the persistence levels and trends at the top of the distribution. In line with (Cantante 2018), I find evidence suggestive of a glass ceiling for women at the top of the distribution. Immigrants represent a growing share of the top, though they are highly polarized within the distribution: those from non-high-income countries over-represented at the bottom 50%, those from high-income countries over-represented at the higher end, and in the top 0.1% by a factor of 10; immigrants also account for the entirety of the growth in top earnings shares since 2000. Linked employer-employee data suggest that while mobility at the top increased from 1986 until the early 90s - in line with what Cardoso (2006) found for the whole distribution - since then there has been a sharp decrease in top mobility. These results suggest a progressive entrenchment at the top that has been taking place for two and a half decades. The second essay uses the approach developed by Bartolucci (2014 to test for immigrant discrimination in the labor market. My baseline estimates for 2004-2017 suggest that immigrants suffer wage discrimination, receiving on average 6%-10% less than native workers on the same firm. However, the baseline estimate hides substantial heterogeneity: (i) high-income country (HIC) immigrants are over-represented in high-wage firms and receive a wage premium relative to natives, but the opposite happens for non-HIC immigrants; (ii) immigrant women suffer from double discrimination. Finally, discrimination decreases with tenure (without fully disappearing) and higher product market competition does not appear to mitigate it. I use the staggered rollout of a business registration reform that increased product and labor market competition to assess if increased competition reduced the immigrant-native wage gap and conclude against it. The third essay explores the effects of labor market concentration on multiple worker outcomes. Through an instrumental variable (IV) approach, I find that it decreased wages by 1.0-1.1% for new hires and by 0.2-0.4% for the entire stock of workers, during the 2002-2009 period. However, these estimates are highly heterogeneous and driven solely by the effect on male workers. Unlike Bassanini (2023)'s results for Portugal in 2010-2019, which are contingent on the wage measure used and who find no gender heterogeneity, these estimates are robust to alternative wage measures, alternative samples, alternative local labor market definitions, specifications including spell fixed effects, and using an alternative (weighted) IV. Using the pandemic as a plausibly exogenous shock, I find effects consistent with those of 2002-2009: workers facing higher labor market concentration in 2019 have significantly lower wages in the post-pandemic period, and the reduction is of similar magnitude to that found in 2002-2009.