Life-cycle Effects of Health RiskJournal of Monetary Economics, vol. 74, Sept. 2015, pp. 67-88

Health status affects individuals' labor supply, asset accumulation and welfare through four main channels: productivity, medical expenditures, time endowments and survival probabilities. Using a life-cycle model calibrated to the U.S. for different education groups, I evaluate the relative importance of each channel and quantify their interactions. The productivity and time endowment channels dominate in importance and the risks implied by them contribute significantly to income inequality and precautionary savings. Health effects are larger for the non-college than college educated and account to a large extent for the lower labor supply and higher reliance on government transfers of the non-college group.

 Skills and the Evolution of Wage Inequality, Labour Economics, vol. 28, June 2014, pp. 41-57

This paper studies wage inequality in the United States between 1980 and 2010 in a framework that accounts for changes in the employment of physical and cognitive skills and their returns. I find that the secular rise in the employment of cognitive skills is largely accounted for by labour force composition changes in shares of gender–education groups rather than changes that occur within these groups. Average employed skills differ greatly across groups, but over time their average employed cognitive skills have remained approximately constant. Returns to cognitive skills increased very sharply for high skill levels, more gradually around mean levels, and decreased at low levels. Returns to physical skills generally declined. These trends account for approximately 63% of the increase in the college wage premium, with changes in returns to cognitive skills playing a dominant role.

 A Model of the Distribution of Talent Across Education Fields, part of PhD thesis, 2011

When students choose a field of study, their objectives are to signal their strongest talents and hide their weakest abilities. Fields of study differ in the ability to reveal students’ talents due to the fact that they test different skills. This paper uses a signalling model to analyze the sorting decisions of students with analytical and verbal skills into fields of study, where each field can reveal only one skill dimension. The first part of the paper describes a model where students’ expected wages are a function of their revealed skill level in the chosen field and the average unrevealed skill in that field. An example where agents are uniformly distributed is fully solved. Comparative statics exercises are performed with respect to the relative population averages of the two skills, their relative variances, and the degree of correlation between skills. In each case, I look at the effects on the relative sizes of the education fields. Simulation results obtained using normally distributed agents are qualitatively identical with the results found using a uniform distribution. Finally, the results of this model remain robust to a modification of the environment in which students’ payoffs depend on firms’ expectations of their talents, and firms face an inference problem about the unrevealed talent.

Working Papers

Health Shocks and the Evolution of Consumption and Income over the Life-Cycle (PDF), with Michael Keane and Shiko Maruyama 

This paper studies the effects of health on earnings dynamics and on consumption inequality over the life-cycle. We build and calibrate a life-cycle model with idiosyncratic health, earnings and survival risk where individuals make labor supply and asset accumulation decisions, adding two novel features. First, we model health as a complex multi-dimensional concept. We differentiate between functional health and underlying health risk, temporary vs. persistent health shocks, and predictable vs. unpredictable shocks. Second, we study the interactions between health and human capital accumulation (learning-by-doing). These features are important in allowing the model to capture the degree to which, and the pathways through which, health impacts earnings and consumption patterns. They are also very important in estimating the value of health insurance and social insurance. A key finding is that health shocks account for roughly half of the growth in offer wage inequality over the life cycle. Eliminating health shocks leads to a 5.5% decline in the variance of the present value of earnings across all individuals.

Long-term Care Insurance Reform in the United States, with Gary Hansen and Minchung Hsu

We build a life-cycle model of household consumption and saving decisions, where long term care (LTC) expenditures are endogenous. We use an LTC-state dependent utility function where regular consumption and LTC are valued differently. The model includes both married and single households, thus capturing important family dynamics that are important for precautionary savings and LTC decisions. Married individuals face the risk of a spouse needing LTC and quickly depleting joint assets. However, those needing LTC can benefit from the presence of a healthy spouse who provides informal care, lowering the costs of LTC given a fixed quality of care. We use the calibrated model to estimate the importance of family dynamics for savings and consumption decisions, and also to quantify the impacts of LTC policy reforms such as the provision of a universal public system that pays for a minimum level of LTC costs.

Selection in Employer Sponsored Health Insurance (PDF, Appendix A, Appendix B) (under review)

Approximately one in four workers aged 25-40 who lacked private health insurance in 2010 in the U.S. did not enroll in employer-sponsored health insurance (ESHI) that was available to them. This paper examines the extensive margin of selection in ESHI using data from the Medical Expenditures Panel Survey 2001-2010 and the National Longitudinal Survey of Youth '97 in 2010. Controlling for a large set of firm and job characteristics that proxy for plan choice and premiums, I find that workers aged 25-40 who decline ESHI and remain privately uninsured have significantly higher health risk than those who enroll. No correlation between health and insurance take-up is found in the 41-64 age group. The advantageous selection among young employees is in part explained by education, family income and Medicaid crowding out ESHI for low socioeconomic status workers who have higher health risk. The results shed light on the characteristics of uninsured workers in the U.S. and on the interaction between private and public health insurance, with implications for the evaluation of health care reform. 

Socio-economic Disparities in U.S. Healthcare Spending: The Role of Public vs. Private Insurance (PDF), with Michael Keane and Shiko Maruyama 

This paper examines socio-economic (SES) disparities in medical spending for all medical conditions at the 3-digit ICD-9 level in the US healthcare system. Using data on adult respondents from the Medical Expenditure Panel Survey 2000-14, we estimate multivariate regressions for individual medical spending (total and out-of-pocket) controlling for medical conditions, demographics, health, and insurance, separately by sex, education, and age. Within age-sex categories, we assess how spending on each condition varies with education (a proxy for SES). In the predominantly privately insured population aged 24-64, system spending for several of the most socially costly conditions is strongly increasing in education (e.g., breast cancer for women and chest symptoms for men). Disparities are not explained by differences in health or ability-to-pay, suggesting they arise due to discrimination. But we find no positive SES gradients for individuals over 64, suggesting that the public Medicare program plays an important role in improving equity.

Why a College Educated Spouse Brings Healthy Life (slides), with Xiaodong Fan 

Using data on married individuals from the Medical Expenditures Panel Survey (MEPS) and the Health and Retirement Study (HRS), we find that having a college educated spouse is strongly associated with better health, better health behaviors, better health transition, and lower mortality, after accounting for own education and demographics. Four main mechanisms could lead to this association: (1) selection into marriage with a college educated spouse, (2) effects of spousal education on household level characteristics, (3) a correlation between spousal education and his/her health, which affects own health and behaviors, and (4) cross-productivity in health production between spouses. We use controls on own and family background, household characteristics, and spouse health to empirically quantify the contribution of mechanisms (1)-(3). We find that selection into marriage with a college educated wife explains a large fraction of the association between spousal education and better health for non-college educated males only. For all other groups, causal mechanisms dominate in importance. Household level characteristics account for a small fraction of the association, while spouse's health plays a relatively larger role. Together, observables account for the entire association in the non-college male group, but for less than half of the association in all other demographic groups. The unexplained fraction is likely due to cross-productivity between spouses in health production. Our findings are relevant for theories of health capital formation, returns to education, health inequality, and marriage market decisions.

Other work in progress

Private Health Insurance Take-up Among Older Australians: the Roles of Cognition and Family, with Kaarin Anstey, Diane Hosking, Kim Kiely, and Timothy Neal (RA Robert Tiong)

Health Shocks and Occupational Mobility, with Olena Stavrunova (on hold)