Poverty and Psychiatric Diagnosis in the U.S.: Evidence from the Medical Expenditure Panel Survey

Background: A number of social programs are targeted at persons with psychiatric diagnosis with the intention of reducing poverty. Previous studies have shown that persons with psychiatric conditions are more likely to be poor and face disparities in education and employment outcomes. A better understanding of the severity of poverty faced by persons and families with diagnosis is necessary for better policy targeting and monitoring. Aims of the Study: This paper seeks to measure the prevalence, depth and severity of poverty for families with persons with psychiatric diagnoses in the United States using data from the 2007 Medical Expenditure Panel Survey (MEPS). We compare poverty profiles of families with diagnosis to those without. Methods: First, we calculate poverty rate, gap and severity using MEPS data for families with and without diagnosis. Second, we present results of multivariate analysis of the association between psychiatric diagnosis and poverty after controlling for a number of characteristics. Results: This paper finds that the poverty rate, depth, and severity are significantly greater for families with a working-age member who has been diagnosed. Median and mean total incomes are lower while health expenditures are higher for families with psychiatric diagnosis. In a multivariate regression, the odds that a family is poor is 1.76 times higher for a family with a diagnosis compared to a family without a diagnosis. We also identify groups who are the most disadvantaged according to severity of income poverty among families with diagnoses. These include families whose head of family has no high school education, whose head has been unemployed for the entire year, or whose head is Black or Hispanic. Families with non-married heads face greater severity of poverty, as do single persons. Families with more severe psychiatric diagnoses, including mood and psychotic disorders, are also found to face more severe poverty. Discussion: There is a statistically significant association between poverty and psychiatric diagnosis, in particular for mood and psychotic diagnoses. This result suggests that existing poverty reduction programs have not adequately reached this population. The analysis has several limitations. The MEPS is not representative of the entire working age population with psychiatric diagnoses, likely leading to underestimates of their poverty. Our study also does not attempt to answer the question of what are the causes of poverty, but has limited the analysis to highlight family and