Using establishment-level panel data from the Occupational Employment and Wage Statistics program, I estimate the effect of minimum wage increases implemented by 10 states in 2014 and 2015. I show that minimum wage increases lead to wage spillovers within establishments. I find little evidence that minimum wage increases induce establishments to reorganize their occupational mix. Finally, I find that minimum wage increases propagate up the management hierarchy, leading to increased wages for supervisors. Nonetheless, I find overall wage inequality decreases within establishments
after minimum wage increases.
pre-publication version: https://elizaforsythe.web.illinois.edu/wp-content/uploads/2023/07/Effect_of_Minimum_Wage_on_Establishments.pdf
Abstract: I estimate the share of eligible individuals who received unemployment insurance (UI) benefits during the first year of the COVID-19 pandemic. I use individual data on reported recipiency from the Current Population Survey Annual Social and Economic Supplement (CPS-ASEC) survey to validate a UI eligibility algorithm that I then apply to the monthly CPS data. Combined with administrative data on actual payments and adjustments for fraud, I estimate that 88 percent of eligible individuals received UI benefits. When I calculate recipiency by program, I find 98 percent of individuals who were eligible for standard UI received benefits, whereas only 76 percent of individuals who were eligible for Pandemic Unemployment Assistance received benefits.
Cortes, G. M., & Forsythe, E. (2023). Distributional impacts of the Covid-19 pandemic and the CARES Act. The Journal of Economic Inequality. https://link.springer.com/article/10.1007/s10888-022-09552-8
Abstract: Using data from the Current Population Surveys, we investigate the aggregate and distributional consequences of the Covid-19 pandemic and the associated public policy response on labor earnings and unemployment benefits in the United States up until February 2021. We find that year-on-year changes in labor earnings for employed individuals were not atypical during the pandemic months, regardless of their initial position in the earnings distribution. The incidence of job loss, however, was, and continues to be, substantially higher among low earners, leading to a dramatic increase in labor income inequality among the set of individuals who were employed prior to the onset of the pandemic. By providing very high replacement rates for individuals displaced from low-paying jobs, the initial public policy response was successful in reversing the regressive nature of the pandemic’s impacts. We estimate, however, that recipiency rates for displaced low earners were relatively low. Moreover, from September onwards, when policy changes led to a decline in benefit levels, earnings changes became much more regressive, even after factoring in benefits.
Previous version: “Impacts of the COVID-19 Pandemic and the CARES Act on Earnings and Inequality” (IZA DP No. 13643)
I present four facts about occupational mobility: (1) most movements occur within firms, (2) downward moves are frequent, (3) wage growth reflects the direction and distance of mobility, and (4) relative occupational wages before mobility predict the direction of mobility, except for non-displaced movers between firms. I show these facts are consistent with models of vertical sorting. I show that non-displaced movements between firms obscure the positive selection of upward occupational movers, likely reflecting moves up a firm-wage job ladder. Displaced workers show similar pre-displacement selection to internal movers, with pre-displacement occupational wage rank predicting the direction of occupational mobility.
Abstract: At the onset of the COVID pandemic, the U.S. economy suddenly and swiftly lost 20 million jobs. Over the next two years, the economy has been on the recovery path. We assess the labor market two years into the COVID crisis. We show that early employment dynamics were almost entirely driven by temporary layoffs and later recalls. Taking these into account, we show that the labor market remained surprisingly tight throughout the crisis, despite the dramatic job losses. By spring 2022, the labor market had largely recovered and was characterized by extremely tight markets and a slightly depressed employment-to-population ratio driven largely by retirements. Finally, we see surprisingly little evidence of excess reallocation, despite predictions that COVID would dramatically and permanently change the way we live and work. We do see that employment has reallocated somewhat away from low-skilled service jobs, and, in light of the job vacancy patterns, conclude that worker preferences or changes in job amenities are driving this shift. In addition, the retirements paved the way for movements up the job ladder, making low-skilled customer-facing jobs even less desirable.
Report prepared for the Department of Labor Chief Evaluation Office Summer Data Challenge on Equity and Underserved Communities
Blog from DOL: https://blog.dol.gov/2022/02/10/were-using-data-to-better-understand-our-work-and-create-more-equitable-programs-and-policies
Abstract: Using data from before and during the Covid-19 pandemic, we show that the expansion of benefits under the CARES Act only modestly increased self-reported UI recipiency among UI eligible workers, from 27% in 2018 to 36% in 2020/2021. We find that the same demographic groups that historically are less likely to report receiving benefits (less educated, younger, and racial and ethnic minorities) continued to be less likely to receive benefits during the pandemic. In addition we find non-heterosexual workers are also substantially less likely to report receiving benefits. The overarching reason for these disparities is differences in beliefs about eligibility, resulting in likely-eligible workers not applying for benefits. We show that union members and individuals who live in states with historically higher recipiency rates are less likely to be misinformed about eligibility, suggesting a role for policy and informational interventions to improve recipiency rates.
Abstract: We investigate the impact of computerization of white-collar jobs on wages and employment. Using online job postings from 2007 and 2010–2016 for office and administrative support (OAS) jobs, we show that when firms adopt new software at the job-title level they increase the skills required of job applicants. Furthermore, firms change the task content of such jobs, broadening them to include tasks associated with higher-skill office functions. We aggregate these patterns to the local labor-market level, instrumenting for local technology adoption with national measures. We find that a 1 standard deviation increase in OAS technology usage reduces employment in OAS occupations by about 1 percentage point and increases wages for college graduates in OAS jobs by over 3 percent. We find negative wage spillovers, with wages falling for both workers with and without a college degree. These results are consistent with technological adoption inducing a realignment in task assignment across occupations, leading office support occupations to become higher skill. We argue relative wage gains for OAS workers indicates that factor-augmenting features of OAS technological change dominate task-substituting features. In addition, while we find that total employment increases, these gains primarily accrue to college-educated women.
Upjohn Institute working paper: 19-310, NBER Working Paper 29866
U of I News Bureau: “Paper: Economy benefits when secretarial jobs require more computer skills” (December 2019)
Abstract: Recessions are known to be particularly damaging to young workers’ employment outcomes. I find that during recessions the hiring rate falls faster for young workers than for more-experienced workers. I show this cannot be explained by the composition of jobs or workers’ labour supply decisions, and I conclude that firms preferentially hire experienced workers during periods of high unemployment. I develop a new model of cyclical upgrading that relaxes the classic assumptions of exogenous firm size and rigid wages. I show this model predicts larger log wage decreases during recessions for young workers than for experienced workers, a prediction that is supported by the data. I conclude that policy makers should consider extending unemployment insurance coverage during recessions to new labour market entrants.
pre-publication version: http://publish.illinois.edu/elizaforsythe/files/2022/01/Forsythe_Youth_Hiring_prepub_version_2021.pdf
U of I News Bureau: “Paper: Young workers hit hardest by slow hiring during recessions” (May 2016)
Abstract: We study the distributional consequences of the Covid-19 pandemic’s impacts on employment, both during the onset of the pandemic and over recent months.Using cross-sectional and matched longitudinal data from the Current Population Survey, we show that the pandemic has exacerbated pre-existing inequalities.Although employment losses have been widespread, they have been substantially larger – and persistently so – in lower-paying occupations and industries. We find that Hispanics and non-white workers suffered larger increases in job losses, not only because of their over-representation in lower paying jobs, but also because of a disproportionate increase in their job displacement probability relative to non-Hispanic white workers with the same job background. Gaps in year-on-year job displacement probabilities between black and white workers have widened throughout the course of the pandemic recession, both overall and conditional on pre-displacement occupation and industry. These gaps are not explained by state-level differences in the severity of the pandemic or the associated response in terms of mitigation policies. We also find evidence that suggests that older workers have been retiring at faster rates.
Previous version: Upjohn Institute working paper 20-327
U of I News Bureau Coverage: “Paper: Pandemic-fueled job losses exacerbating preexisting inequalities among workers” (June 2020)
Abstract: It is well-known that recessions can lead to long-term scarring for young workers. I show that employers hire fewer young workers when there are few job openings per unemployed job seeker, while hiring rates for workers with more than 10 years of potential experience are much less cyclically volatile. During the COVID-19 pandemic, youth employment rates rebounded particularly quickly compared with other groups and historic patterns. I show this is consistent with the historic relationship between tightness and youth hiring rates, suggesting youth scarring from the COVID-19 pandemic may be less severe compared with previous recessions.
Abstract: We investigate employer recruiting behavior, using detailed firm-level data from a national survey of employers hiring recent college graduates. We show employers adjust recruiting effort, hiring standards, and compensation with the business cycle, beliefs about tightness, and their own hiring plans. We then show that firms expending greater recruiting effort hire more individuals per vacancy. The results suggest that when firms want to increase hires they adjust vacancies and recruiting intensity per vacancy. If true more broadly in the labor market, it may help explain the breakdown in the standard matching function during the Great Recession. For the firms in our sample, the difference in firm vacancy yields between 2011 and 2015 would have more than doubled if recruiting effort had been constant. Finally, we estimate that our measure of recruiting effort can explain 61% of the elasticity of the vacancy yield with respect to hires in our data.
Abstract: We report on the state of the labor market midway through the COVID recession, focusing particularly on measuring market tightness. As we show using a simple model, tightness is crucial for understanding the relative importance of labor supply or demand side factors in job creation. In tight markets, worker search eﬀort has a relatively larger impact on job creation, while employer proﬁtability looms larger in slack markets. We measure tightness combining job seeker information from the CPS and vacancy postings from Burning Glass Technologies. To parse the former, we develop a taxonomy of the non-employed that identiﬁes job seekers and excludes the large number of those on temporary layoﬀ who are waiting to be recalled. With this taxonomy, we ﬁnd that eﬀective tightness has declined about 50% since the onset of the epidemic to levels last seen in 2016, when labor markets generally appeared to be tight. Disaggregating market tightness, we ﬁnd mismatch has surprisingly declined in the COVID recession. Further, while markets still appear to be tight relative to other recessionary periods, this could change quickly if the large group of those who lost their jobs but are not currently searching for a range of COVID-related reasons reenter the search market.
Labour Economics, Volume 69, 2021, 101955, ISSN 0927-5371, https://doi.org/10.1016/j.labeco.2020.101955.
Abstract: We investigate the sources of heterogeneity in the levels and cyclical sensitivity of unemployment rates across demographic groups. We develop a new methodology to decompose cyclical and level differences in unemployment rates between groups into flows between three states (employment, unemployment, and out-of-the-labor-force). We find that increases in unemployment rates during recessions for young, non-white, and less-educated groups of workers are primarily driven by reductions in the job-finding rates, which can explain more than 60% of cyclical fluctuations in the unemployment rate across demographic groups, compared with under 20% driven by separations. However, separations are the most important factor in explaining the persistent gap in unemployment rates between each disadvantaged group and their respective counterpart group, with important differences between groups. For less-educated workers, separation rates explain most of the unemployment gap, with 75% of the separation rate attributable to industry and occupation. Less-educated workers also spend less time searching. For younger workers, we find separation rates explain all of the unemployment gap, while industry and occupation explain only 60% of their elevated separation rates. For non-white workers, hiring explains almost half of the unemployment gap. Non-white workers search more intensely for work than other groups, but spend less time interviewing per search time, suggesting that labor market discrimination contributes to non-white workers’ persistently high unemployment rates.
Forsythe, E., Kahn, L. B., Lange, F., & Wiczer, D. (2020). Labor demand in the time of COVID-19: Evidence from vacancy postings and UI claims. Journal of public economics, 189, 104238. (https://www.sciencedirect.com/science/article/pii/S004727272030102X)
Abstract: We use job vacancy data collected in real time by Burning Glass Technologies, as well as unemployment insurance (UI) initial claims and the more traditional Bureau of Labor Statistics (BLS) employment data to study the impact of COVID-19 on the labor market. Our job vacancy data allow us to track the economy at disaggregated geography and by detailed occupation and industry. We find that job vacancies collapsed in the second half of March. By late April, they had fallen by over 40%. To a first approximation, this collapse was broad based, hitting all U.S. states, regardless of the timing of stay-at-home policies. UI claims and BLS employment data also largely match these patterns. Nearly all industries and occupations saw contraction in postings and spikes in UI claims, with little difference depending on whether they are deemed essential and whether they have work-from-home capability. Essential retail, the “front line” job most in-demand during the current crisis, took a much smaller hit, while leisure and hospitality services and non-essential retail saw the biggest collapses. This set of facts suggests the economic collapse was not caused solely by the stay-at-home orders, and is therefore unlikely to be undone simply by lifting them.
NBER Working Paper No. 27061
Note on March labor market data.
Abstract: The occupational structure of an establishment provides a description of its production process by detailing the distribution
and relative intensity of tasks performed. In this study, the author investigates whether there are substantive differences in the
occupational structures of low- and high-wage service sector establishments. The author shows that low-wage establishments
organize production to use less labor in professional occupations compared with high-wage establishments operating in
the same local labor market and industry. In addition, low-wage establishments employ fewer individuals in information
technology occupations, employ fewer managers, and have substantially wider supervisory spans of control. These results
indicate that, despite operating in the same narrowly defined labor and product markets, low-wage establishments organize
production to less intensively use labor in skilled occupations.
Upjohn Institute working paper; 18-292
Abstract: With falling labor market dynamism in the United States, opportunities within firms take on increasing importance in young workers’ career progression. Developing a variety of occupational ranking metrics, I show that occupational mobility within firms follows a standard lifecycle pattern in which the frequency, distance, and wage return from mobility fall with age. However, when upward and downward mobility are considered separately, the distance of moves increases over the lifecycle. Thus, while young workers make the smallest distance occupational moves up and down, they have the largest wage gains and losses associated with these moves. I find that wage growth for young workers deteriorated substantially in the first decade of the 2000s, primarily driven by a reduction in wage growth within firms, whereas mid-career workers have experienced no such change. I argue this is most likely driven by the dramatic fall in employer-to-employer mobility for young workers since the early 2000s. Encouragingly, wage growth has improved markedly for young workers since 2012.
Upjohn Institute working paper; 18-286