Lifecycle effects of a recession on health behaviors: Boom, bust, and recovery in Iceland
Introduction
Pioneering work by Ruhm (2000) found that although there is considerable evidence that long-term economic growth promotes population health, short-term downturns in economic activity counter-intuitively lead to reduced mortality. That study was based on data from the U.S. but appears to reflect a more global phenomenon; e.g., Gerdtham and Ruhm (2006) found that mortality increased during high employment or strong economic conditions in 23 Organization for Economic Cooperation and Development countries. The increasingly robust finding of lower mortality rates during economic downturns spawned a wave of studies investigating the relationships between business cycles and health that was no doubt fueled by the Great Recession in the U.S. and many other developed countries. A number of studies have examined effects on health behaviors, an important potential pathway.
As pointed out by Burgard et al. (2013), the Great Recession was noteworthy in terms of its severity, and the effects of such a deep downturn may differ from those of milder past recessions. Overall, the pre-Great Recession literature indicates that mortality and morbidity are lower during economic downturns, but that mental health appears to deteriorate during periods of higher unemployment (see Catalano et al., 2011, Modrek et al., 2013), and there appears to be little consensus on the directional effects of business cycles on any health behavior, including smoking, binge drinking, and food choices (see Ásgeirsdóttir et al., 2014). However, studies of the Great Recession in the U.S. and other developed countries have revealed no effects (Barrett and O'Sullivan, 2013, Tekin et al., 2013) or procyclical effects (Urbanos-Garrido and López-Valcárcel, 2015) on self-assessed health, and mixed effects on mortality. Although van Gool and Pearson (2014) continued to find procyclical effects of poor economic conditions in countries on mortality in the context of the Great Recession, others found that this relationship no longer appeared to hold in the U.S. (Ruhm, 2013) or across countries (Toffolutti and Suhrcke, 2014), and yet another study from the U.S. found a strong association between unemployment and mortality before the Great Recession, but a marginally positive relationship during the last decade of the 2000s, among individuals age 65 and over (McInerney and Mellor, 2012). Finally, studies have continued to find a procyclical relationship between economic conditions and mental health in the context of the Great Recession (e.g., Ruhm, 2013, Toffolutti and Suhrcke, 2014, Urbanos-Garrido and López-Valcárcel, 2015, van Gool and Pearson, 2014).
Studies of the effects of the Great Recession on health behaviors represent a loose patchwork. In the U.S., Compton et al. (2014) found positive associations between individuals’ employment status and substance use (alcohol, tobacco, and illicit drugs) during the Great Recession; Antillón et al. (2014) found that higher state unemployment rates during the Great Recession were associated with better sleep (less sleeplessness); and Smith et al. (2014) found that state unemployment rates (before and during the Great Recession) had no association with individuals’ time spent cooking or eating away from home. Examining 23 European countries, Toffolutti and Suhrcke (2014) found no aggregate effects of unemployment during the Great Recession on alcohol consumption, although in countries with a low level of “social protection”, high unemployment was related to lower alcohol consumption. In Ireland, Barrett and O'Sullivan (2013) found that smoking was lower during the Great Recession than in the pre-crisis years (2006–2007) among adults age 50+. All of these studies were based on repeat cross-sectional data; that is, they did not observe the same individuals over time.
It is useful to distinguish between individual-specific and more general effects of recessions. For example, individuals may experience job loss, asset loss, or income loss, which may affect their health behaviors. However, more broad-based aspects of recessions--such as changes in prices or availability of goods, public expenditures on healthcare, pollution, and congestion--that affect the population at large can also lead to changes in health behaviors or health. The most straightforward strategy for disentangling the two effects is the use of individual level longitudinal data that spans a time period during which a clearly defined recession occurred.
A few studies examined the unemployment-health relationship using individual-level longitudinal data, considering associations between changes in employment status and changes in health status. Böckerman and Ilmakunnas (2009), using pre-Great Recession Finnish data, found that although those who are not employed had worse self-assessed health, transitions into and out of employment were not associated with that outcome. Granados et al. (2014), using a U.S. panel from 1979 to 1997 (pre-Great Recession), found that high state unemployment rates were associated with lower mortality risk, confirming the findings of most previous studies.
Three recent studies used individual-level longitudinal data to study the effects of labor-market conditions or own unemployment on health behaviors during long time periods that included the Great Recession. Colman and Dave (2014), used two longitudinal datasets from the U.S. (PSID 1999–2009 and NLSY 1998–2010) in fixed- and random-effects estimations of the effects of own job loss on body-weight related behaviors (including cigarette smoking); the authors found evidence that job loss increased smoking among women (but not men) but decreased the number of cigarettes per day among smokers, and that job loss decreased fast food consumption. Latif (2014), using Canadian panel data from 1994 to 2009, found that higher provincial unemployment rates were associated with more binge drinking, but had no impact on the probability of being a smoker. However, none of these studies explicitly investigated effects of the Great Recession.
Ásgeirsdóttir et al. (2014) used individual-level longitudinal data to estimate effects of the 2008 economic crisis in Iceland on a large set of health-compromising and health-promoting behaviors. Observing the same individuals one year before and one year after the crisis, they found that the macroeconomic shock led to reductions in seven different health-compromising behaviors (including binge drinking and smoking), as well as a number of health-promoting behaviors such as consumption of fruit and vegetables. The authors also found that for binge drinking, smoking, consumption of sugared soft drinks, fruit and vegetable consumption, and getting recommended sleep, the broad context of the macroeconomic downturn (such as price changes) appeared to be more important than changes in individual-level factors, including work hours, mortgage debt, real income, real assets, and mental health. Using the same data and research design, similar results were found by Ólafsdóttir and Ásgeirsdóttir (2015) for other measures of alcohol consumption (including frequency of drinking) and by Ólafsdóttir et al. (2015) for other measures of smoking (including smoking intensity). For sweets, indoor tanning, fast food consumption, and consumption of fish oil, Ásgeirsdóttir et al. (2014) found that the individual factors explained half (for fish oil) or more of the change in consumption between the pre-crisis and crisis periods.
As far as we know, no research to date has investigated the effects of economic recession on health or health behaviors beyond the immediate aftermath of the recession or after the decline has ended or reversed. In particular, it is not known whether the effects of a recession on individuals’ health behaviors are short-lived or enduring. If the effects represent mere “blips” in people's health behavior trajectories, the long-term effects of recessions on health – at least through the health behavior channel – are unlikely to be consequential. On the other hand, if the effects on health behaviors persist even after economic recovery, the long-term effects of recessions on health through individuals’ health behaviors could be more substantial.
In this study, we address this knowledge gap using longitudinal data from Iceland, a country that has experienced a partial recovery from a severe economic crisis, and examining a number of different health behaviors of adults at three points in time—during the boom period leading up to the crisis, during the depths of the crisis, and after the economy had to a significant extent recovered. In particular, we expand the Ásgeirsdóttir et al. (2014) study by examining essentially the same set of health behaviors but adding a third important time point, 2012, occurring after partial but significant economic recovery. As far as we know, this is the first study to investigate effects of recovery from deep recession on health behaviors or to explicitly consider whether the effects of recessions on health behaviors are short-lived or lingering. It is necessary to understand whether shifts in behaviors due to recession are temporary or persistent in order to understand the implications for longer-term mortality and morbidity.
Section snippets
The Icelandic context
The effects of the Great Recession on health and health behaviors have occurred in different structural, cultural, geographic, and policy contexts. For example, van Gool and Pearson (2014) considered the role of changes in health expenditures across countries including the U.S., France, Germany, Ireland, Iceland and Spain; Toffolutti and Suhrcke (2014) considered the role of overall social protection expenditures across countries; and Ásgeirsdóttir et al. (2014) explored the role of
Data
The data used for this study come from a health and lifestyle survey “Heilsa og líðan” (Health and Wellbeing) carried out by the Directorate of Health in Iceland (previously the Public Health Institute of Iceland) in 2007, 2009, and 2012. The survey contained questions about health, illnesses, use of drugs, smoking and drinking, diet, health care, height and weight, accidents, exercise, sleep, quality of life, and other lifestyle related issues, as well as demographics and work-related factors
Measures
We consider 10 different health behaviors for which information was collected at all three survey waves (2007, 2009, and 2012). These include measures of substance use, diet, indoor tanning, and sleep, and largely correspond to those considered by Ásgeirsdóttir et al. (2014) for 2009 compared to 2007, with certain exceptions as noted.
For substance use, we coded individuals as smokers if they answered the question “Do you smoke?” with a yes, and as heavy drinkers if they provided information
Methods
We exploit the economic crisis of 2008 in Iceland and the subsequent swift and substantial reversal that followed to investigate the lifecycle effects of a deep recession on health behaviors. Using 2007, 2009 and 2012 data, we estimate individual fixed-effects models, which implicitly control for all unobserved time-invariant individual-level characteristics and account for cross-period correlation in standard errors. In these models, we estimate each of the 10 health behaviors, with the key
Results
Health behaviors and other characteristics (weighted means) of our full analysis sample (those participating in all three survey waves and who had non-missing data on key analysis variables) and the subsample of working-age (25–61 years old in 2007) are presented in Table 1 for 2007, 2009, and 2012. Except for daily sweets, all health-compromising behaviors declined during the crisis (between 2007 and 2009) and continued to decline during the recovery (between 2009 and 2012). Fruit consumption
Effects in context
We found little evidence that the observed effects of the crisis on most health behaviors operated through individual-level changes in hours worked (or alternatively, employment), real household income, or mental health, although all of these were strongly affected by the crisis. We therefore consider the potential role of more general aspects of the Icelandic economic crisis that may underlie the observed effects. A very prominent aspect of this particular crisis was prices. Currency
Conclusion
In this study, we used individual-level longitudinal data from Iceland, a country that has experienced a partial recovery from a severe economic crisis, to investigate the extent to which the effects of a recession on a large set of health behaviors are lingering or short-lived. That is, we explored trajectories in health behaviors from pre-crisis boom, to crisis, to recovery. As far as we know, this is the first study to explicitly consider the life cycle of a recession on health behaviors—a
Acknowledgements
The authors are grateful to Victoria Halenda for excellent research assistance and to the Icelandic Research Fund (grant number 130611-051) and The Directorate of Health in Iceland (formerly the Public Health Institute of Iceland), which funded the data collection and permitted its usage for this project. We also thank Lysi Ltd. for the data on fish-oil prices. The authors are grateful for helpful comments from participants at the Canadian Center for Health Economics and Temple University
References (45)
- et al.
Sleep behavior and unemployment conditions
Econ. Hum. Biol.
(2014) - et al.
Was the economic crisis of 2008 good for Icelanders? Impact on health behaviors
Econ. Hum. Biol.
(2014) - et al.
Unemployment and substance outcomes in the United States 2002–2010
Drug Alcohol Depend.
(2014) - et al.
Deaths rise in good economic times: evidence from the OECD
Econ. Hum. Biol.
(2006) - et al.
Trends in sugar-sweetened beverage consumption among youth and adults in the United States: 1999–2010
Am. J. Clin. Nutr.
(2013) The impact of recession on drinking and smoking behaviours in Canada
Econ. Model.
(2014)- et al.
Recessions and seniors’ health, health behaviors, and healthcare use: Analysis of the Medicare Current Beneficiary Survey
J. Health Econ.
(2012) - et al.
Assessing the short term health impact of the Great Recession in the European Union: a cross-country panel analysis
Prev. Med.
(2014) Long-term damage from the Great Recession in OECD countries
European Journal of Economics and Economic Policies: Intervention
(2014)- et al.
A short note on the wealth, health and wellbeing of Ireland's older people before and during the economic crisis
SHARE Working Paper (15-2013)
(2013)
Unemployment and self-assessed health: evidence from panel data
Health Econ.
The Great Recession and health: People, populations, and disparities
Ann. Am. Acad. Pol. Soc. Sci.
The health effects of economic decline
Annu. Rev. Public Health
Unemployment and Health Behaviors Over the Business Cycle: a Longitudinal View (No. w20748)
Fast Food Still Major Part of U.S. Diet
Iceland's Central Bank Sees Economic Recovery but Flags Risks Retrieved from the Wall Street Journal
Individual joblessness, contextual unemployment, and mortality risk
Am. J. Epidemiol.
Iceland's Stabilized Economy is a Surprising Success Story
Baltic and Icelandic experiences of capital flows and capital flow measures
International Monetary Fund: Working Paper
Heilsa og líðan Íslendinga 2012: Framkvæmdaskýrsla
Iceland's Crisis and Recovery and Current Challenges
Schools in two communities weather the crash
Cited by (43)
Bad lighting: Effects of youth indoor tanning prohibitions
2023, Journal of Health EconomicsDrinking during downturn: New evidence from the housing market fluctuations in the United States during the Great Recession
2021, Economics and Human BiologyLong-term effects of job loss on male health: BMI and health behaviors
2021, Economics and Human BiologyNew evidence on the impact of the Great Recession on health-compromising behaviours
2021, Economics and Human BiologyCitation Excerpt :For comparison purposes, we also estimate a static version of Eq. (1), which does not control for initial conditions and correlated individual effects. This type of specification has been widely used in the literature studying the determinants of unhealthy behaviours using longitudinal data (e.g. Ásgeirsdóttir et al., 2014, 2016). First, we check whether our findings are robust to using a conditional fixed-effects logit model (Chamberlain, 1980) and a linear fixed-effects model.
The effect of the economic collapse in Iceland on the probability of cardiovascular events
2020, Economics and Human Biology