Does Coverage of Health Insurance Contribute to Obesity? An Empirical Study among American Young Adults

Purpose: This paper explores whether insurance status can influence preventable unhealthy behaviors (ex-ante moral hazard; see Ehrlich and Becker, 1972) and how such a behavioral change, if there is any, could contribute to weight gain.  We used body mass index (BMI) among American young adults and relevant BRFSS survey data to answer this question. 

Background: A major benefit of insurance is that the insured are more likely to avail themselves to routine physician visits and preventive services than the uninsured. For this and other reasons, most economists believe that health insurance has a positive effect on the health of the insured. Once insured, however, an individual may become less cautious about his/her health and more likely to engage in riskier activities than before. These two opposing effects of health insurance may be offsetting each other and the net effect of insurance on health may be positive or negative.

 Data and Method:  We extended the Ehrlich and Becker (1972) model of health behavior under the influence of health insurance to explain ex-ante moral hazard associated with a change in an individual’s health insurance status. We then apply the model to the BRFSS data from 2001-2011 for American young adults to estimate the relationship between health insurance and obesity. The IV regression approach and Lewbel IV techniques were used to estimate the model controlling for physician visits and other key endogenous variables such as health condition and employment status of the respondent.

 

Results: The results show that a switch from no health insurance to having insurance is associated with a decrease in BMI of 0.188 kg/m² and this suggests a weight loss of 1.33 pounds. In addition, health insurance is also associated with a decrease in 7.105 percentage points in the probability of being overweight and a reduction in the probabilities of 8.518 percentage points of being obese. However, subsample analyses for different American regions show that the influence of having insurance does not have the same effects on BMI across all U.S. regions.