Session 15: The Labor Market Experience of Vulnerable Populations of Workers

People with disabilities are underemployed across the world. To increase their representation, more than 100 countries have established quota regulations requiring firms to hire people with disabilities. This paper studies the labor market consequences of enforcing modest disability hiring quotas. Using the introduction of a reform in Brazil that enhanced enforcement of a new hiring quota regulation, my market-level analysis finds that people with disabilities in local labor markets more exposed to the reform experienced larger increases in employment and earnings. Leveraging variation in enforcement across firms, I document three key margins along which firms respond to the quota scheme. First, firms hire more workers with disabilities into low-paying jobs. Second, workers with disabilities experience reduced wage growth. Third, the quota also does not come at a cost to workers without disabilities in terms of wages or employment, or to firms in terms of closure. Through the lens of a simple model, I show that the policy generates aggregate welfare gains. My findings support that, in labor markets characterized by discrimination in hiring, mandating modest increases in employment for the disadvantaged can promote redistribution and improve welfare.

In this paper, we revisit the evidence on the employment effects of the Americans with Disabilities Act (ADA) of 1990. The existing literature has assessed the impact of the policy by comparing the labor market outcomes of individuals who report limitations to their ability to work (work limitations) and the labor market outcomes of individuals who do not. Since the ADA applies to all disabled individuals, not just those with work limitations, we rely on rich health and limitation information from the Survey of Income and Program Participation to draw a distinction between individuals with work limitations and individuals with other types of limitations that do not necessarily impact their ability to work (functional limitations). Consistently with the literature, over a longer sample period than examined in previous work, we find that the ADA has had a negative effect on the employment and wages of individuals with work limitations. However, we also find that the policy has had a substantial and significant positive effect on the employment of individuals with physical or mental limitations that are not work-impacting even when severe, with virtually no effect on their wages. To interpret this evidence, we develop a search and matching model of the labor market in which a worker’s productivity and value of non-market time vary with a worker’s disability status and firms face different costs to hire, accommodate, and separate from different types of workers, with and without disabilities, which can account for all these patterns. We then use the model to evaluate the heterogeneous impacts of the policy on the employment and wages  of work disabled and non-work disabled workers, to examine the relative importance of the different components of the policy on these outcomes, and to compare alternative designs of it.

Using linked employer-employee data from Brazil, we document a large gender pay gap due to women working at lower-paying employers with better nonpay attributes. To interpret these facts, we develop an equilibrium search model with endogenous firm pay, amenities, and hiring. We provide a constructive proof of identification of all model parameters. The estimated model suggests that amenities are important for both men and women, that compensating differentials explain half of the gender pay gap, and that there are significant output and welfare gains from eliminating gender differences. However, equal-treatment policies fail to achieve those gains.

We analyze monthly earnings volatility using administrative payroll data. While it is well-documented that wages are largely stable, we find that this wage stability does not translate into earnings stability for most U.S. workers. Even within stable employment relationships, and even when wages are constant, many workers nevertheless face substantial monthly earnings volatility. The standard deviation of monthly earnings changes is 28%, while the standard deviation of base wage changes is only 2%. There is substantial heterogeneity in this volatility, with much higher volatility for hourly workers than for salaried workers. This degree of volatility is far higher than what is implied by benchmark models of earnings processes which are calibrated to previously-available annual data and used as inputs for leading macro models. To understand the welfare consequences of pay volatility, we estimate the amenity value of volatility using worker quits in a model of a frictional labor market. We find that workers have a high willingness to pay to reduce earnings volatility. Overall, this analysis shows that high-frequency labor market shocks are an important source of risk and fragility which has been masked by past studies of annual earnings.

Existing research documents that more generous unemployment insurance (UI) delays job finding with limited effects on job match quality. Using employer-employee matched data from Washington state and a fuzzy regression discontinuity design, this paper exploits the eligibility threshold to examine how UI receipt impacts job search for marginally attached workers who earn lower incomes. We find that UI access only minimally delays re-employment, but improvements in job match quality lead to longer hours and higher earnings. When scaled by our first-stage, we estimate UI receipt increases cumulative hours worked by approximately 14 full-time weeks and earnings by $14,000 in the two years following job loss. Earnings effects are driven by longer hours at one's next employer in the medium-run and higher hourly wages in the longer-run. Individuals eligible for UI work for significantly fewer firms after job loss. Our calculations suggest lowering the eligibility threshold is among the most effective UI policy changes because it supports particularly low-income workers and its cost is partially offset by higher earnings and taxes upon re-employment.

Children of married couples in the United States have better academic, physical health, and mental health outcomes than children of unmarried couples. It has been suggested in the economics and sociology literature that this correlation is causal, and that incentivizing marriage over cohabitation or single parenting may improve child development. In this paper, we quantify the effects of family structure on the outcomes of children. We create a structural dynamic model of mothers' relationship choices, labor market choices, and child development, and estimate it using measurements on children's outcomes and parents' relationships drawn from the Fragile Families panel dataset. Our counterfactuals consider the effects on children of financially incentivizing marriage between their birth parents, and we find that even large increases in marriage rates (from 23% in our baseline to 47% in the counterfactual) are associated with only small benefits in child outcomes. Our findings suggest observed differences between child outcomes across family structures are driven by selection on child and relationship quality, and that policies targeting parental relationships may not have large eects.

The United States has both one of the highest incarceration rates and the largest prison population worldwide. Furthermore, the U.S.'s reliance on incarceration in criminal justice policy is not only a modern phenomenon but dates back to the late 19th century. An active literature seeks to understand the social and economic impacts of this policy, including the labor market effects of incarceration, reaching mixed conclusions. In this paper, we trace the labor market trajectories of incarcerated vs. non-incarcerated individuals in the U.S. during the first major increase in incarceration, from 1870-1940. Using data on millions of historical inmates from census and state prison ledgers, we provide new descriptive statistics on the evolution of incarceration. We then provide estimates of the incarceration penalty, the difference in labor market outcomes such as occupation, earnings, employment, and labor force participation between incarcerated and non-incarcerated men across this full time period. We find a penalty in occupation-based income scores ranging from 5-15% over the period, with the penalty increasing over time as incarceration rates increased. Using double-linked samples and a linked sample of brothers, we show the penalty remains even after controlling for observed and unobserved measures of family background. Finally, we estimate a "pseudo'' event study around year of prison admission for state prisoners linked to census and show that labor market trajectories prior to incarceration cannot explain the penalty.

We study the role of employment protection legislation (EPL) in boosting employment among older workers. We do so by conducting a comprehensive analysis of the sharp and complete elimination of EPL that occurs at age 67 in Sweden, as well as reform- driven shifts in this age cutoff. First, focusing on direct separation effects, we find that 8% of jobs separate in response to the elimination of EPL. Effects stem from jobs with stronger initial EPL (long-tenure, firms subject to “last in, first out” rules), and those in the public sector. Separations appear involuntary to workers, with firms targeting plausibly unproductive workers (sick leave users). Second, we focus on effects on continuing jobs. While wages appear rigid to EPL elimination, we uncover novel, sizable intensive-margin hours reductions among continuing jobs, and an 8% drop in earnings conditional on staying on the job. Third, we estimate total equilibrium effects at the cohort level, where separations fully pass through into employment to population rate effects, with no offsetting effect from hiring. On a per-capita basis, total earnings of older workers causally drop by 21.5% due to EPL elimination. We validate these local effects by leveraging a reform-driven shift in the age cutoff from 67 to 68. EPL is therefore a potential tool to prop up labor income among older workers, by prolonging the duration of their final job.