How do economic downturns affect population health? In several groundbreaking papers Chris Ruhm showed that some health outcomes actually improved during the recessions of the previous three decades: people smoked and drank less, they stayed off the roads leading to fewer accidents, and the smokestacks from some factories stopped, lowering air pollution and perhaps lung disease. Although few people would advocate slowing the economy down, it would be a comforting silver lining if recessions made us healthier.
These optimistic assessments seem less realistic during the economic slump that began in 2007 and has never fully ended. First, the current recession is both longer and deeper than any of the previous recessions since World War II. Second, the cuts to the safety net and social services – including income assistance, job training, and health programs has been especially deep in spite of a significant federal stimulus. Third, the recession has hit homeowners particularly acutely, leading to a shrinkage of home credit, a precipitous drop in home purchasing, and ultimately a wave of home foreclosures.
There is now an excellent working paper by Janet Currie and Erdal Tekin that examines the consequences of the home foreclosure crisis on rates of hospitalizations at the level of zip codes in four states, Arizona, California, Florida, and New Jersey, that were hardest hit by the foreclosure crisis. (Note: zip codes are geographical areas that vary in size, the average size in the sample is ~30,000 people). Currie and Erdal link hospital discharge data includes a specific diagnosis with quarterly home foreclosure data from the individual’s zip code from 2005 to 2009. This allows them to estimate a time series model with fixed effects for zip codes and for time period.
The authors find that foreclosures are associated with a substantial increase in hospitalizations for non-elective treatments, but not for cancer treatments and elective treatments (treatments that are discretionary and are typically scheduled in advance such as a hip replacement). For example, among the non-elderly adults, a quarterly increase of 100 home foreclosures is associated with an additional 12 hospital visits for anxiety disorders. They find somewhat smaller effects for physical conditions: such as an 8 percent increase for diabetes. Within subgroups, the largest effects are concentrated among blacks and Hispanics, and the effects tend to be lowest among non-elderly individuals.
The results from this study are plausible and carefully estimated. If home foreclosures are a stressful event – and surely suddenly losing one’s home equity and facing eviction is extremely wrenching – then we would expect that the effect would be largest for those most directly affected. Racial minorities and younger people were among the hardest hit by the foreclosure crisis, and they experience the greatest negative effects in the Currie and Erdal study. From a policy standpoint, this study speaks to the substantial health externalities of not doing more to help distressed homeowners. Not only did home foreclosures disrupt the social and economic fabric of communities, they probably created many negative health effects both for the individuals facing foreclosure and for their neighbors.
There are a few conceptual points that this study raises that I think warrant further exploration. One is pinning down the causal mechanisms. How does a mental or physical health problem escalate and lead to a hospitalization? One story is that where home foreclosures are the worst, the resources to effectively screen and treat people in primary care are greatly diminished. This is likely to be especially true because the four states in the study all massively cut their safety net programs, including public health insurance during the study period. A complementary story is that lower social support and more acute stressors limit the ability of people to effectively manage their health, leading to health emergencies that could have been better managed. Third, there is the wealth effect of home foreclosures – being foreclosed on obliterates a person’s financial resources, making it more likely that she will gamble that a problem will fix itself (sometimes it might, but when it doesn’t it ends up with a trip to the emergency room). Finally, there is a trickier issue which is simply regression to the mean – places that experienced major foreclosures were actually probably “poorer” in terms of education and resources in 2005 than they looked on paper – residents in those areas were probably also under their expected levels of certain acute medical problems, because they were riding the prosperity of the economic bubble. When the housing foreclosure crisis hit, these places therefore experienced a double whammy: both the shock effect of the economic downturn, and the predictable effect of regressing back to the level of sickness and poverty that would have prevailed in the absence of a bubble. The housing bubble may have made some areas look both richer and healthier than they actually were.