The great recession and health spending among uninsured U.S. immigrants: implications for the Affordable Care Act implementation

Abstract

Objective: We study the association between the timing of the Great Recession (GR) and health spending among uninsured adults distinguishing by citizenship/nativity status and time of U.S. residence.

Data source: Uninsured U.S. citizens and noncitizens from the 2005-2006 and 2008-2009 Medical Expenditure Panel Survey.

Study design: The probability of reporting any health spending and the natural logarithm of health spending are our main dependent variables. We compare health spending across population categories before/during the GR. Subsequently, we implement two-part regression analyses of total and specific health-spending measures. We predict average health spending before/during the GR with a smearing estimation.

Principal findings: The probability of reporting any spending diminished for recent immigrants compared to citizens during the GR. For those with any spending, recent immigrants reported higher spending during the GR (27 percent). Average reductions in total spending were driven by the decline in the share of the population reporting any spending among citizens and noncitizens.

Conclusions: Our study findings suggest that recent immigrants could be forgoing essential care, which later translates into higher spending. It portrays the vulnerability of a population that would remain exposed to income shocks, even after the Affordable Care Act (ACA) implementation.

Keywords: Uninsured; immigrants; recession; total health care costs; two-part model.

Figure 1

Estimated Total Health Spending before and during the Great Recession among the Uninsured Population by Citizenship and Nativity StatusNote. Predicted health spending was estimated using the two-part model outcomes. By contrast with average means shown in Table1, predicted spending in Figure1 accounts for population differences. Natural log values (Ln) of health spending were first transformed to monetary values using a smearing estimation. This predicted value was multiplied by the predicted probability of reporting any health spending (Prob). All costs were adjusted using U.S. dollars of 2010. Health spending before the recession (2005–2006) was the reference group for each citizenship/nativity category. Statistically significant differences are ^§p < .001, ^‡p < .01, ^†p < .05.Source: 2005–2006 and 2008–2009 Medical Expenditure Panel Survey data.