Long-run Effects of Severe Economic Recessions: Drastic Changes in Working Hours and Male BMI Trajectories

The recent economic downturn has inspired researchers to revisit earlier literature [started by Ruhm (2000)] documenting that mortality decreases during economic recessions. Ruhm (2005) demonstrates that during the hard times people smoke less and lose weight. Moreover, these changes in behavior disproportionately occur among heavy smokers and severely obese individuals. At the same time, those who were completely physically inactive become more likely to exercise.

Applying the conclusions from this literature to the collapse of the Soviet Union implies that the severe economic downturn (of about 40%) that happened over the course of 5-6 years should have led to a tremendous improvement in health outcomes. Yet, since the beginning of transition period Ukraine, Russia and other post-Soviet countries have also experienced dramatic health deterioration. Males in those countries suffered the most health decline with Ukrainian male life expectancy at birth being the lowest among European countries (62.2 years in 2011). So, perhaps a modest economic downturn is good for one’s health, but too much of it can lead to deadly outcomes.

Ukraine has the fastest rate of depopulation in Europe. Premature mortality of Ukrainian prime-age males is one of the major contributors to this pattern. The leading cause of death and disability is cardio-vascular disease (67% of all deaths in 2009), which is closely related to hypertension, obesity, and tobacco use. A recent health survey shows that hypertension prevalence among males is 30 percent and almost 50 percent of them are unaware of having this condition, 12 percent of males are obese and 50 percent are overweight. At the same time more than 60 percent of them are daily smokers.

This paper’s goal is to investigate to what extent labor market fluctuations—both through individual experience and local conditions—are responsible for changes in obesity prevalence and related health behaviors. We focus our attention on three questions. The first two use data aggregated to the regional level. First, do sizeable changes in working hours change BMI trajectories. Second, can these changes in BMI trajectories be explained by changes in physical exercise related to more or less time availability, or by changes in tobacco and alcohol consumption due to more or less stressful lifestyle. Finally, we study whether an individual’s history of labor market shocks during the 1990s (a time of considerable economic downturn) affects their individual BMI-age trajectory.

We use growth curve approach to model the trajectory of BMI by age. These models can help control for selection by allowing the differences in the initial levels of BMI (intercept) and in slopes of the BMI-age trajectories to vary with individual future working hours: those who significantly increased working hours, those who significantly decreased working hours, and those who preserved same number of working hours. To address the issue of endogeneity, we use local labor market conditions as instruments for changes in individual working hours and the history of local labor market conditions as an instrument for the history of individual labor market shocks.