The U.S. health care expenditure: evidence from 2001 to 2009

ABSTRACT

This study utilizes state-level data from 2001 to 2009 to estimate the impact of the 2007 financial crisis upon health care expenditures. Higher death rates are consistently found to have a positive and statistically significant impact on health expenditures. While mental health and COPD are not generally found to impact expenditures, increases in the percentage of the population diagnosed with cholesterol and obesity tend to increase health expenditures. Increases in health expenditures slowed considerably after the financial crisis. Even though recessions (high unemployment rates) are generally found to have a positive impact on health expenditures, the post-financial crisis time period is estimated to have much lower health expenditures than in other time periods. Our results can be used to give insight into the conditions under which the slower rate of increase in health expenditures can be expected to increase. More research will be needed to be able to more completely explore not only the reasons for these changes in health expenditures but also whether they are likely to continue into the future.

KEYWORDS:

• Health care expenditure
• health production function
• financial crisis

JEL CLASSIFICATION:

• I10
• I12

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The National Vital Statistics Reports (Vol. 60, No. 3, December 29, 2011) released the detailed 2009 data on U.S. deaths, death rates and other statistics.

² There is a growing literature examining the relationship between business cycles and health. For example, Ruhm (2000, 2003, 2005, 2007), Neumayer (2004), Gertham and Ruhm (2006) and Miller et al. (2009), applying data of different recession periods, find that health conditions tend to improve during recessions but worsen during expansions.

³ As changes in illness may react in a different way to a recession than the death rate does, we include the four measures of morbidity in our study.

⁴ Both variables are endogenous theoretically (see e.g. Olsen 1993) but more importantly the Death Rate generally tests as endogenous in the empirical work presented in the next section.

⁵ The exogenous variables used to estimate death rates include Bachelor’s Degree, Personal Income per Capita, tobacco consumption, alcohol consumption, violent and property crime rates, manufacturing employment and racial and regional dummy variables.

⁶ Notice that the main impact that assumption that the Death Rate is endogenous is on the sign of the Death Rate coefficient. In OLS regressions of the models in Table 4, the impact of the Death Rate is just the opposite, with poorer health decreasing health expenditures. Notice that the Wu-Hausman test of endogeneity reported in Table 4 tends to be significant, indicating that the Death Rate is, in fact, endogenous.