Don't Graduate into a Recession

Hannes Schwandt and Till M. von Wachter. "Socioeconomic Decline and Death: Midlife Impacts of Graduating in a Recession." National Bureau of Economic Research Working Paper No. 26638, January 2020.

Much of the previous economic research on the aftermath of recessions has focused on their short-term effects on earnings and jobs. These effects were thought to disappear after around 10 years. New research suggests that the picture is worse, with longer-term consequences not only for workers' earnings, but also for their health and family outcomes.

In a National Bureau of Economic Research working paper earlier this year, Hannes Schwandt of Northwestern University and Till von Wachter of the University of California, Los Angeles analyzed the effects of graduating into the 1982 recession on mortality and socioeconomic status at midlife.

In their analysis, the authors estimated the size of the midlife effects through a new approach. First, they compiled health information from U.S. vital statistics of the Centers for Disease Control and Prevention, population and socioeconomic data from the U.S. Census Bureau and other related surveys, and state-level unemployment data in order to connect unemployment rates at graduation to lifetime outcomes. Then they created a novel measure of unemployment that accounted for variation in economic conditions and mitigated potentially confounding effects from interstate migration or individuals' choices about when to enter the labor market.

The authors arrived at five major findings. First, by comparing the mortality rates of those graduating into a recession to those graduating into standard or booming economic conditions, they found that the 1982 recession graduates exhibited higher mortality starting in their late 30s and increasing through age 50. Specifically, every 1 percent increase in the unemployment rate at graduation was associated with an increase in the mortality rate at age 49 of one death per 10,000. The authors also looked at how mortality effects compound with age, estimating average life expectancy loss from age 50 until death. The 1982 recession graduates, who faced a reweighted unemployment rate that was 3.9 percentage points higher than what average non-recession graduates faced, lost six to nine months in life expectancy.

Second, by regressing causes of death for the 1982 graduates against the most common causes of death in the United States, the authors found that the primary factor in higher midlife mortality was related to increases in heart disease, liver disease, and lung cancer, which are strongly linked to unhealthy habits like smoking, drinking, and inactivity. A secondary factor was "deaths of despair," which include deaths due to suicide, drug overdoses, and liver disease. Overall, disease-related causes and deaths of despair accounted for nearly two-thirds of the increased mortality. The authors suggested that the stress of graduating into a recession likely encouraged unhealthy behaviors, which contributed to the negative long-term health outcomes.

Third, while examining the effects of graduating into a recession on socioeconomic measures such as earnings and family outcomes, the authors confirmed that initial incomes were reduced. They then demonstrated the unexpected persistence of these negative effects into midlife — every 1 percent increase in the unemployment rate at graduation was associated with a 1 percent reduction in midlife earnings.

Fourth, the authors quantified the impact of graduating into a recession on marriage, divorce, and childlessness rates. Although they recognized the difficulty in separating causality, they found that while recession graduates were initially more likely to marry, by middle age this trend reversed, with higher divorce and childlessness rates.

Finally, by comparing their findings across four demographic subgroups — male and female non-Hispanic whites and male and female nonwhites and Hispanics — the authors found that although the overall mortality effects were similar across races, deaths of despair were more prevalent among whites while disease-related deaths were more frequent among nonwhites. The authors noted that epidemics during the sample period (such as the HIV and crack epidemics) may account for some of this variation. In terms of socioeconomic effects, the negative effects on both earnings and family outcomes were worse for whites. Of the four groups, white men experienced the most significant losses in long-term earnings, with consistent losses in their 30s compounding during their 40s.

Schwandt and von Wachter's unique approach allowed them to present evidence that the long-term effects of graduating into a recession are costlier than previously believed. In fact, they suggested that their findings may underestimate the true impact on mortality and socioeconomic status. Their results highlighted that white males may have the most to lose by graduating into a recession and reinforced the link between economic conditions and morbidity and mortality, one that worsens with age.