Research

Working Abstract:

Taken together, falling fertility rates and rising lifespans are causing many of the world's wealthiest countries to face steep increases in age-related long-term care (LTC) needs that must be born by a comparatively shrinking native labor force. This paper exploits rich administrative records to provide novel evidence on the scope and distribution of those costs. Beyond their well-studied disemployment effects, we find that health shocks to the elderly cause large families to grow larger: adult children are more likely to get married, and they have more children of their own. These responses are persistent over time, and they are consistent with the idea that LTC needs induce family members to substitute from formal employment into a mix of informal caregiving and home production.  Members of smaller families, on the other hand, experience sharp increases in mortality risk, which is consistent with caregiving-related ``deaths of despair."  Leveraging quasi-experimental variation in access to international caregiver markets for individual families, we find that formal caregiver hiring amplifies positive fertility responses, whereas mortality responses vanish. These results suggest that immigration policy can causally improve native-born well-being and sustain longer-term native-born demographics. 

Abstract:

While the U.S. Disability Insurance (DI) system is large in terms of fiscal costs and beneficiary rolls, its interactions with household self-insurance mechanisms are not well-understood. In addition to crowding out the insurance value of DI, household self-insurance may drive negative selection into DI by reducing the implicit costs that stem from its lengthy application process. Allowing for such interactions in a rich life cycle model, this paper finds that revenue-neutral expansionary DI reforms do not necessarily improve welfare, as benefits often accrue to negatively-selected individuals with stronger self-insurance capacity. However, an asset test can reduce negative selection on the basis of household self-insurance capacity, so much so that expansionary reforms become welfare-improving. Finally while household self-insurance crowds out the value of different kinds of DI expansions,  abstracting away from insurance value can still lead to erroneous conclusions regarding their overall and relative welfare effects.

Abstract

Although much research has explored the rise in young disability insurance (DI) receipt, there has been much less work explaining the large geographic differences in DI rates across cities and states. We explore the drivers of this heterogeneity using administrative tax data that allows us to link young adults (age 24-34) to their parents. Our findings are threefold. First, children from low income families display sharply varying probabilities of receiving DI depending on the place where they grew up, while those from rich families show no similar differences. Second, we show that DI take-up of children from low income families exhibits heterogeneity both over time (cyclically) and over place which is not apparent for children from richer families. Third, we show that places where poor children grow up to have the highest rate of DI receipt tend to be “good” areas based on many standard characteristics, including lower inequality, lower segregation, higher school quality, and higher social capital. State level tax policies are also predictive of DI rates; states with more generous EITCs, lower tax rates, and less progressive tax structures each tend to have higher DI take-up. These are also the characteristics of places that tend to produce higher income mobility. We show that the relationship between child outcomes in terms of DI take-up and income mobility across places is mixed, but the places that tend to generate good outcomes on both measures are more rural. By comparison, this appears to be less true for the places generating particularly bad outcomes on both measures.

Abstract

The synthetic control method is widely used in comparative case studies to adjust for differences in pre-treatment characteristics. A major attraction of the method is that it limits extrapolation bias that can occur when untreated units with different pre-treatment characteristics are combined using a traditional adjustment, such as a linear regression. Instead, the SC estimator is susceptible to interpolation bias because it uses a convex weighted average of the untreated units to create a synthetic untreated unit with pre-treatment characteristics similar to those of the treated unit. More traditional matching estimators exhibit the opposite behavior: they limit interpolation bias at the potential expense of extrapolation bias. We propose combining the matching and synthetic control estimators through model averaging to create an estimator called MASC. We show how to use a rolling-origin cross-validation procedure to train the MASC to resolve trade-offs between interpolation and extrapolation bias.