Does Increased Education Lower Health Care Spending? Findings for Self-Managed Health Conditions

I study the association between education and health care spending for a set of health conditions amenable to self-management.  The underlying theory of the role of education in the health production process based on Grossman’s seminal work (1972a, 1972b) posits that increased educational attainment improves individual health through greater productive efficiency: more highly educated individuals are able to produce more health from a given amount of medical care inputs and their own time.  Alternatively, others have argued that greater education leads to improved health through greater allocative efficiency: more educated individuals are better informed about the true effects of inputs on health and thus are able to make more efficient input choices compared to those with less education (Rosenzweig & Schultz, 1983a, 1983b).  Both explanations suggest that human capital not only improves health outcomes, but also leads to more efficient use of health care inputs and time in the production of health.  Consequently, increased education may lead to a reduction in health care spending required to produce a given level of health. 

I address this issue by examining the cost saving effect of education on health care spending for a set of adverse health conditions – hypertension, diabetes, and asthma – for which individuals may be able to influence the amount of medical care resources used through self-management of care.  I utilize the data from the Medial Expenditure Panel Survey in order to estimate a two-part generalized linear model for health spending and predict the level of spending for each condition as a function of an individual’s educational attainment.  The full specification includes measures of the severity of the condition to capture the direct effect of education on health care spending, while controlling for the indirect effect of education on spending through its possible impact on condition severity.  Condition severity may represent a pathway through which education affects health spending and thus incorporates some of the impact of education on health spending.  I next estimate models without a severity measure in order to obtain the total effect of education on health spending capturing both the direct and indirect effects. 

My findings support the underlying theory for individuals with asthma and/or hypertension.  One of the key findings is a strong inverse relationship between education and health care spending among elderly adults with asthma.  Predicted annual health care spending by elderly adults with asthma who completed at least 12 years of schooling is about $4,000 less than those who completed fewer than 12 years of schooling.  These associations are direct effects of education excluding its saving impact by being healthier.  The total effect of education on health spending strengthens the relationship and is most evident among elderly adults with hypertension and/or asthma.  This reflects the fact that greater cost efficiency exists through the channel that increased education affects condition severity.  These findings suggest that economies in health spending may be achieved with greater education, especially with the completion of a high school education.