John V. Pepper

Disability and Employment: Reevaluating the Evidence in Light of Reporting Errors

Measurement error in health and disability status has been widely accepted as a central problem for social science research. Long-standing debates about the prevalence of disability, the role of health in labor market outcomes, and the influence of federal disability policy on declining employment rates have all emphasized issues regarding the reliability of self-reported disability. In addition to random error, inaccuracy in survey datasets may be produced by a host of economic, social, and psychological factors that can lead respondents to misreport work capacity. We develop a nonparametric foundation for assessing how assumptions on the reporting error process affect inferences on the employment gap between the disabled and nondisabled. Rather than imposing the strong assumptions required to obtain point identification, we derive sets of bounds that formalize the identifying power of primitive nonparametric assumptions that appear to share broad consensus in the literature. Within this framework, we introduce a finite-sample correction for the analog estimator of the monotone instrumental variable (MIV) bound. Our empirical results suggest that conclusions derived from conventional latent variable reporting error models may be driven largely by ad hoc distributional and functional form restrictions. Under relatively weak nonparametric assumptions, nonworkers appear to systematically overreport disability.

The impact of the National School Lunch Program on child health: A nonparametric bounds analysis

Children in households reporting the receipt of free or reduced-price school meals through the National School Lunch Program (NSLP) are more likely to have negative health outcomes than observationally similar nonparticipants. Assessing causal effects of the program is made difficult, however, by missing counterfactuals and systematic underreporting of program participation. Combining survey data with auxiliary administrative information on the size of the NSLP caseload, we extend nonparametric partial identification methods that account for endogenous selection and nonrandom classification error in a single framework. Similar to a regression discontinuity design, we introduce a new way to conceptualize the monotone instrumental variable (MIV) assumption using eligibility criteria as monotone instruments. Under relatively weak assumptions, we find evidence that the receipt of free and reduced-price lunches improves the health outcomes of children.

Disability and Employment

Measurement error in health and disability status has been widely accepted as a central problem in social science research. Long-standing debates about the prevalence of disability, the role of health in labor market outcomes, and the influence of federal disability policy on declining employment rates have all emphasized issues regarding the reliability of self-reported disability. In addition to random error, inaccuracy in survey datasets may be produced by a host of economic, social, and psychological factors that can lead respondents to misreport work capacity. We develop a nonparametric foundation for assessing how assumptions on the reporting error process affect inferences on the employment gap between the disabled and nondisabled. Rather than imposing the strong assumptions required to obtain point identification, we derive sets of bounds that formalize the identifying power of primitive nonparametric assumptions that appear to share broad consensus in the literature. Within this framework, we introduce a finite-sample correction for the analog estimator of the monotone instrumental variable (MIV) bound. Our empirical results suggest that conclusions derived from conventional latent variable reporting error models may be driven largely by ad hoc distributional and functional form restrictions. We also find that under relatively weak nonparametric assumptions, nonworkers appear to systematically overreport disability.

Robust inferences from random clustered samples: an application using data from the panel study of income dynamics

Abstract I examine the implications of clustered samples on inference. Important differences are revealed in comparisons between the estimated asymptotic variances derived assuming random and clustered sampling, even when there are only a few observations per cluster.

The Intergenerational Transmission Of Welfare Receipt: A Nonparametric Bounds Analysis

Using a nonparametric bounding method and data from the Panel Study of Income Dynamics, I examine the effect that growing up in a household that receives Aid to Families with Dependent Children (AFDC) has on welfare participation as a young adult. In light of the ambiguities created by the selection problem, a number of alternative assumptions and estimates are presented. While the data alone cannot be conclusive, the results generally strengthen the evidence that being exposed to AFDC as a child increases both the probability and the expected duration of future welfare participation.

More on Monotone Instrumental Variables

Econometric analyses of treatment response often use instrumental variable (IV) assumptions to identify treatment effects. The traditional IV assumption holds that mean response is constant across the sub-populations of persons with different values of an observed covariate. Manski and Pepper (2000) introduced monotone instrumental variable assumptions, which replace equalities with weak inequalities. This paper presents further analysis of the monotone instrumental variable (MIV) idea. We use an explicit response model to enhance the understanding of the content of MIV and traditional IV assumptions. We study the identifying power of MIV assumptions when combined with the homogeneous linear response assumption maintained in many studies of treatment response. We also consider estimation of MIV bounds, with particular attention to finite-sample bias. Copyright (C) The Author(s). Journal compilation (C) Royal Economic Society 2009

Slippery When Wet: The Effects of Local Alcohol Access Laws on Highway Safety

Using detailed panel data on local alcohol policy changes in Texas, this paper tests whether the effect of these changes on alcohol-related accidents depends on whether the policy change involves where the alcohol is consumed and the type of alcohol consumed. After controlling for both county and year fixed effects, we find evidence that: (i) the sale of beer and wine may actually decrease expected accidents; and (ii) the sale of higher alcohol-content liquor may present greater risk to highway safety than the sale of just beer and wine.

The effect of alcohol prohibition on illicit-drug-related crimes

We evaluate the effect of alcohol access on drug‐related crime and mortality using detailed information on access laws in Texas between 1978 and 1996. Counties with alcohol access have higher average levels of drug‐related crimes. However, after controlling for both county and year fixed effects, we find that having local alcohol access decreases crime associated with illicit drugs. This basic finding is replicated in two alternative analyses. First, we find that prohibiting the sale of beer to persons under 21, which arguably increases the implicit price of liquor more for juveniles in wet counties than for those in dry counties, increases the fraction of drug‐related arrests involving juveniles more in wet counties than in dry counties. Second, we find that after controlling for both county and year fixed effects, local alcohol access decreases mortality associated with illicit drugs. Alcohol access and illicit‐drug‐related outcomes appear to be substitutes.

Dental care coverage and retirement

OBJECTIVES To examine the convergence of an aging population and a decreased availability of dental care coverage using data from the Health and Retirement Study (HRS). METHODS We calculate national estimates of the number and characteristics of those persons age 51 years and above covered by dental insurance by labor force, retirement status, and source of coverage. We also estimate a multivariate model controlling for potentially confounding variables. RESULTS We show that being in the labor force is a strong predictor of having dental coverage. For older retired adults not in the labor force, the only source for dental coverage is either a postretirement health benefit or spousal coverage. CONCLUSIONS Dental care, generally not covered in Medicare, is an important factor in the decision to seek dental care. It is important to understand the relationship between retirement and dental coverage in order to identify the best ways of improving oral health and access to care among older Americans.

Wealth Effect and Dental Care Utilization in the United States

OBJECTIVE The purpose of this article is to examine the relationship of wealth and income and the relative impact of each on dental utilization in a population of older Americans, using data from the Health and Retirement Study (HRS). METHODS Data from the HRS were analyzed for US individuals aged 51 years and older during the 2008 wave of the HRS. The primary focus of the analysis is the relationship between wealth, income, and dental utilization. We estimate a multivariable model of dental use controlling for wealth, income, and other potentially confounding covariates. RESULTS We find that both wealth and income each have a strong and independent positive effect on dental care use of older Americans (P < 0.05). A test of the interaction between income and wealth in our model failed to show that the impact on dental care utilization as wealth increases depends on a persons income level or, alternatively, that the impact on dental use as income increases depends on a persons household wealth status (P > 0.05). CONCLUSIONS Relative to those living in the wealthiest US households, the likelihood of utilizing dental care appears to decrease with a decline in wealth. The likelihood of utilizing dental care also appears to decrease with a decline in income as well.