Essays on Disability, Food Insecurity, Assistance Programs, and Employment

Abstract

Populations with disabilities are at higher risks of food insecurity and low employment than those without a disability which can lead to poor nutritional outcomes and decreased quality of life. The objective of this dissertation is to examine the effects of participation in assistance programs for households with disabled members and to analyze the effects of policy changes that designed to improve employment outcomes for people with disabilities. This dissertation consists of three essays and in the analyses, we consider three programs that include the Supplemental Security Income (SSI) and Supplemental Nutrition Assistance Program (SNAP) in the U.S. and the Employment Quota System (EQS) in South Korea. The objective of the first essay is to understand the underlying relationships between food insecurity and various disability characteristics of household members and look at how the relationship is affected by participation in assistance programs. Using data from the 2011–2016 National Health Interview Survey and by applying ordered probit and local polynomial regression models, we find that food insecurity is not only affected by type, severity, and multiplicity of disability of a household member but also affected by who in the household has a disability. Results suggest that participation in assistance programs may shield food security from a household member’s disability. The objective of the second essay is to examine the effects of SNAP participation and the 2013 SNAP benefit changes on food insecurity for households with disabled members. We make use of the public- and restricted-access National Health Interview Survey (NHIS) 2011–2015, in which two different indicators of disability are used: the presence of member(s) with disabilities and who in the household has a disability. To obtain more efficient and consistent estimates, copula distribution functions are incorporated into in the maximum likelihood function of the switching regression model in which state-specific SNAP policy variables serve as instrumental variables to satisfy exclusion restrictions. Main results suggest that SNAP is more effective in reducing food insecurity for households with disabled members than for those without disabled members, and the effects of SNAP vary with a household head’s, spouse/partner’s and children’s disabilities. Additionally, we find that the decrease in SNAP benefits that occurred in 2013 weakens the program’s effectiveness. The objective of the third essay is to examine a set of changes in the employment quota system for people with disabilities that was implemented in 2010 in South Korea. Using data from the Panel Survey of Employment of the Disabled (PSED) from South Korea and ordered probit models with sample selection, we estimate the extent to which these exogenous policy changes have desired employment outcomes for people with disabilities. Results suggest that policy changes bring about improved employment for only men with disabilities; for women with disabilities, no improved employment outcomes are found, and that they are significantly disadvantaged in the labor-market.