Selective college admissions are fundamentally a question of tradeoffs: given capacity, admitting one student means rejecting another. Research to date has generally estimated average effects of college selectivity, and has been unable to distinguish between the effects on students gaining access and on those losing access. Black, Denning, and Rothstein use the introduction of the Top Ten Percent rule and administrative data from the state of Texas to estimate the effect of access to a selective college on student graduation and earnings outcomes. They estimate separate effects on two groups of students. The first, who are highly ranked at schools that previously sent few students to the flagship university, gain access due to the policy, the second, students outside the top tier at traditional “feeder” high schools, tend to lose access. The researchers find that students in the first group see increases in college enrollment and graduation with some evidence of positive earnings gains 7-9 years after college. In contrast, students in the second group do not see declines in overall college enrollment, graduation, or earnings. The Top Ten Percent rule, introduced for equity reasons, thus also seems to have improved efficiency.
Farronato, Fradkin, Larsen, and Brynjolfsson study the effects of occupational licensing on consumer choices and market outcomes in a large online platform for residential home services. They exploit exogenous variation in the time at which licenses are displayed on the platform to identify the causal effects of licensing information on consumer choices. They find that the platform-verified licensing status of a professional is unimportant for consumer decisions relative to review ratings and prices. The researchers confirm this result in an independent consumer survey. They also use variation in regulation stringency across states and occupations to measure the effects of licensing on aggregate market outcomes on the platform. The results show that more stringent licensing regulations are associated with less competition and higher prices but not with any improvement in customer satisfaction as measured by review ratings or the propensity to use the platform again.
Official poverty statistics and even the extreme poverty literature largely ignore the homeless. Meyer, Wyse, Grunwaldt, Medalia, and Wu examine the labor market attachment, earnings, safety net utilization, demographic characteristics, and mobility patterns of individuals experiencing homelessness in order to understand their economic well-being. This project is part of the development of the Comprehensive Income Dataset, which combines household survey data with administrative records to improve estimates of income. Specifically, the researchers use restricted microdata from the 2010 Decennial Census, which enumerates both the sheltered and unsheltered homeless, the 2006-2016 American Community Survey (ACS) which surveys the sheltered homeless, and longitudinal shelter use data from several major US cities. The researchers link these data to longitudinal administrative tax data as well as data on Supplemental Nutrition Assistance Program, Temporary Assistance to Needy Families), Medicare, Medicaid, housing assistance, and mortality. They document the patterns of transitions between homeless situations and other housing statuses, as well as factors associated with these transitions. The approach benefits from large samples that offer a guide to national homelessness patterns and allow for the comparison of estimates between data sources, including the Department of Housing and Urban Development's point-in-time counts. By shedding light on issues of data linkage and survey coverage among the homeless, Meyer, Wyse, Grunwaldt, Medalia, and Wu contribute to efforts to better incorporate this hard-to-survey population into income and poverty estimates.
Optimal insurance benefit design requires understanding how coverage generosity impacts individual behavior and insured costs. Cabral and Dillender explore the impact of workers compensation wage replacement benefit generosity on individual behavior, program costs, and welfare. Workers' compensation income benefit schedules are set by the state, where the weekly benefit amount paid for time out of work is a linear function of an injured worker's prior average weekly wage, up to a maximum weekly benefit. Using unique comprehensive administrative data from Texas, the researchers leverage a sharp increase in the maximum weekly benefit in a difference-in-differences research design by comparing outcomes for workers differentially exposed to the initial maximum benefit who were injured either just before or after the new maximum benefit was implemented. They find that increasing the generosity of wage replacement benefits does not impact the number of claims but has a large impact on claimant behavior, leading to longer income benefit durations and increased medical spending. The estimates indicate that behavioral responses along these two margins -- income benefit duration and medical spending -- are equally important drivers of increased program costs, where collectively these behavioral responses imply an increase in insured costs that is nearly 1.5 times the mechanical increase in insured costs when benefits are expanded. Drawing on these estimates along with an estimate of the consumption drop experienced by injured workers, the researchers calibrate a model to estimate the marginal welfare impact of increasing the generosity of workers' compensation wage replacement benefits. The estimates suggest that increasing benefit generosity does not improve welfare, with much of the projected welfare loss attributable to the previously unexplored impact of income benefit generosity on medical spending.
In addition to the conference paper, the research was distributed as NBER Working Paper w26976, which may be a more recent version.
Please download our paper here:
Care Prices on Wages
Over 150 million Americans receive health insurance benefits from an employer as a form of compensation. In recent years, health care costs have grown rapidly, raising concerns that increased health care spending crowds-out wage increases. Arnold and Whaley leverage geographic variation in health care price growth caused by changes in hospital and physician market structure to test the impact of health care prices on wages and benefit design. They use changes to hospital and physician market structure as a source of exogenous variation. The reduced form results find that that hospital concentration is associated with a 2.5% reduction in wages, changes to physician market structure do not have a strong impact on wages. Using this variation, the researchers find markets that experience 10% higher price growth than the national average experience 4.1% slower wage growth. Distribution impacts. This effect is concentrated among workers without a college degree. The researchers also find that a 10% increase in health care prices leads to a 9.5% increase in the the growth of high-deductible health plans and a 6% increase in health care costs paid by patients. Overall, the results show how rising health care costs are passed to workers in the form of lower wages and less generous benefits.
We develop a dynamic extension of the canonical two-way fixed effects model of Abowd, Kramarz, and Margolis (1999) and use it to study the evolution of hiring wage inequality in a panel of Italian workers that allows for accurate measurement of "poaching wages." The proposed framework is shown to nest a reduced form of the sequential auction model of Postel-Vinay and Robin (2002), which allows hiring wages to depend both on the identity of the hiring firm (``where you're at'') and the identity of the labor market state or firm from which the worker was hired (``where you're from''). Conditions are provided under which these origin and destination effects are separately identified by worker mobility. Bias correcting a variance decomposition of the growth in hiring wages, we find that the change in destination effects explains roughly 4-5 times as much wage growth variation as the change in origin effects. Contrary to the prediction of models where destination effects represent compensating differentials for future wage growth, we find that a firm's origin and destination effects are strongly positively correlated. The wage penalty associated with being hired from non-employment averages roughly 4\% and covaries strongly with the unemployment rate. Men are found to enter the labor market at higher paying employers than women, and this gap in employer pay widens with additional job to job moves. Differences in the identity of hiring firms explain roughly a third of the age-adjusted gender gap in hiring wages, while differences in the states from which men and women are hired explain less than 1\% of the gap.
Governments go to great lengths to attract foreign multinational enterprises because these enterprises are thought to raise the wages paid to their employees (direct effects) and to improve outcomes at incumbent local firms (indirect effects). Setzler and Tintelnot construct the first US employer-employee dataset with foreign ownership information from tax records to measure these direct and indirect effects. They find the average direct effect of a foreign multinational firm on its US workers is a 7 percent increase in wages. This premium is larger for higher skilled workers and for the employees of firms from high GDP per capita countries. The researchers leverage the past spatial clustering of foreign-owned firms by country of ownership to identify the indirect effects. An expansion in the foreign multinational share of commuting zone employment substantially increases the employment, value added, and -- for higher earning workers -- wages at local domestic-owned firms. Per job created by a foreign multinational, the estimates suggest annual gains of 16,000 USD to the aggregate wages of local incumbents, of which about two-thirds is due to indirect effects. The researchers compare their findings to the value of subsidy deals received by foreign multinationals.
Italy and Germany have similar geographical differences in productivity - North more productive than South in Italy; West more productive than East in Germany - but have adopted different models of wage bargaining. Italy sets wages based on nationwide contracts that allow for limited local wage adjustments, while Germany has moved toward a more flexible system that allows for local bargaining. The Italian system has significant costs in terms of forgone aggregate earnings and employment because it generates a spatial equilibrium where workers queue for jobs in the South and remain unemployed while waiting. Boeri, Ichino, Moretti, and Posch's findings are relevant for other European countries.
In addition to the conference paper, the research was distributed as NBER Working Paper w25612, which may be a more recent version.