Research
Job Market Paper
The Rise of Non-Employment Among Prime-Aged Men: Trends, Household Support, and Persistence
(with Bruce D. Meyer and Derek Wu)
Abstract
The non-employment rate of prime-aged men (ages 25-54) nearly quadrupled from 2.9% in 1965 to 11.1% in 2023. Using more than five decades of survey and Census data, we examine the nature and determinants of this rise and what it implies for welfare. The increase occurred during recessions, as non-employment rose by 11.1 percentage points during recession years and fell by only 2.8 percentage points during non-recession years. About two-thirds of the long-run increase was accommodated by support from other household members, most commonly family, and increases in household-supported non-employment persisted through expansions. Household-supported non-employed men were also more likely than other non-employed men to be out of work two years later. Most of the rise in household-supported non-employment was among men who did not report illness or disability as their reason for not working, and non-employed men spent little more time than employed men on caregiving or home production. Chinese import competition raised non-employment mainly among men not supported by their households, while oxycodone exposure raised it mainly among household-supported men, but neither accounted for most of the long-run increase. The results also provide limited support for explanations in which the rise mainly reflected short-term family insurance or a shift into caregiving or home production. Instead, our evidence suggests that adverse employment and health conditions played an important role and that temporary shocks had lasting effects on employment. As a result, recessions may be more costly than their immediate effects on employment suggest, with other household members financing much of the non-employment that persists.Working Papers
The Geography of Disadvantage: Do Cost-of-Living Adjustments Improve Poverty Measurement and Program Targeting?
(with Bruce D. Meyer and Derek Wu)
Abstract
Individuals are commonly assumed to be more disadvantaged in higher-cost areas. As a result, geographic adjustments for local prices are embedded in many federal payments and have been proposed or implemented for various poverty measures. We assess whether such adjustments achieve a central objective of a poverty measure: identifying the least advantaged population. We examine ten domains of well-being – material hardships, appliances owned, home quality, food security, public services, health, education, assets, permanent income, and mortality – drawing on 71 indicators from linked survey and administrative records. Within each domain, we contrast those classified as poor under the Supplemental Poverty Measure and the new Comprehensive Income Poverty Measure, with and without a geographic adjustment. In nine of ten domains, we find that incorporating a geographic adjustment instead identifies a less deprived poor population. This result reflects local prices being sufficiently correlated with public goods and locational amenities that those with low incomes appear less disadvantaged in higher-cost areas. Because poverty thresholds underpin eligibility for many transfer programs and the geographic allocation of federal funding, geographic adjustments may direct resources away from the most disadvantaged.Misleading Estimates from Nonlinear Models with a Binary Outcome
(with Bruce D. Meyer and Derek Wu)
Under review
Abstract
When estimating nonlinear models for binary outcomes, such as probit and logit models, researchers often rely on average partial effects (APEs) to summarize the effect of a regressor. Because the marginal effect of a variable in these models depends on the values of all other variables, the value of an APE hinges on the portion of the sample used for the calculations. When averaged over parts of the sample drawn from a subpopulation not used to define the object of interest, the APE may be misleading. This paper highlights common situations, such as differences-in-means with a secondary group and difference-in-differences designs, where APEs calculated for the full sample deviate from marginal effects for the appropriate part of the sample. We propose a simple and costless solution in specific cases and demonstrate through simulations that recalculating APEs over the appropriate subsample yields unbiased results. Reexamining published results from multiple papers, we find statistically significant discrepancies between the reported estimates and the appropriately calculated APEs.Innovation and Health Behavior: Evidence from the American Public Bath Movement
(with Michael McKelligott)
