Document Type

Working Paper

Publication Date


Working Paper Number

WP #03008, December 2005


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 incentives 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 Manski and Pepper's (2000) monotone instrumental variable (MIV) bound.

Our empirical results suggest that conclusions derived from conventional latent variable report- ing error models are being driven largely by ad hoc distributional and functional form restrictions. Moreover, under relatively weak assumptions, we find that an assumption of unbiased reporting is not supported. Nonworkers appear to overreport work limitations.

Publication Status

This is a working paper of an article from Journal of the American Statistical Association 102, no. 478 (2007): 432-441. doi: 10.1198/016214506000000997.

File Format



27 pages