Elsevier

Economics & Human Biology

Volume 4, Issue 3, December 2006, Pages 298-316
Economics & Human Biology

Deaths rise in good economic times: Evidence from the OECD

https://doi.org/10.1016/j.ehb.2006.04.001Get rights and content

Abstract

This study uses aggregate data for 23 Organization for Economic Cooperation and Development (OECD) countries over the 1960–1997 period to examine the relationship between macroeconomic conditions and deaths. The main finding is that total mortality and deaths from several common causes rise when labor markets strengthen. For instance, controlling for year effects, location fixed-effects (FE), country-specific time trends and demographic characteristics, a 1% point decrease in the national unemployment rate is associated with growth of 0.4% in total mortality and the following increases in cause-specific mortality: 0.4% for cardiovascular disease, 1.1% for influenza/pneumonia, 1.8% for liver disease, 2.1% for motor vehicle deaths, and 0.8% for other accidents. These effects are particularly pronounced for countries with weak social insurance systems, as proxied by public social expenditure as a share of GDP. The findings are consistent with evidence provided by other recent research and cast doubt on the hypothesis that economic downturns have negative effects on physical health.

Introduction

Widely cited analyses of aggregate time-series data by Harvey Brenner, 1973, Brenner, 1975, Brenner, 1979 reveal a countercyclical variation in admissions to mental hospitals, infant mortality rates, and deaths due to cardiovascular disease, cirrhosis, suicide, and homicide.1 However, this research suffers from serious technical flaws (Gravelle et al., 1981, Stern, 1983, Wagstaff, 1985, Cook and Zarkin, 1986) and studies correcting the problems (Forbes and McGregor, 1984, McAvinchey, 1988, Joyce and Mocan, 1993) fail to uncover a consistent relationship between the macroeconomy and health.2 Instead, the results are sensitive to the choice of countries, time periods, and outcomes, with falling unemployment frequently being correlated with worse rather than better health. This lack of robustness is not surprising since any lengthy time-series may contain confounding factors that are spuriously correlated with economic conditions.3

Ruhm (2000) has recently addressed the omitted variables bias issue by estimating fixed-effect (FE) models for a panel of the 50 states and District of Columbia over a 20-year period (1972–1991). These specifications exploit within-state changes and so automatically control for time-invariant factors that are spuriously correlated with economic conditions across locations. Evidence is provided that mortality increases when labor markets are tight. Unemployment is negatively and significantly related to total mortality and 8 of 10 specific causes of death, with suicides the important exception. For instance, a 1% point fall in the state unemployment rate is associated with 0.5, 3.0, 0.7, and 0.4% increases in deaths from all causes, motor vehicle deaths, influenza/pneumonia, and heart disease.

Using similar methods and data from German states over the 1980–2000 time span, Neumayer (2004) finds that a 1% point decline in unemployment is predicted to raise mortality from all causes by between 0.7 and 1.1%.4 Increases in deaths from cardiovascular diseases, pneumonia/influenza and motor vehicle accidents are also obtained and, interestingly, the estimates also suggest a procyclical variation in suicides.

The analysis below extends on the previous research by examining whether deaths rise when labor market conditions improve using information for 23 member nations of the Organization for Economic Cooperation and Development (OECD) over the 1960–1997 period. Specifically, countries are the unit of observation and we investigate how total mortality and nine causes of deaths vary with national unemployment rates (or the percentage of the population employed), after controlling for time-invariant country-specific factors, general time effects, demographic characteristics and (usually) country-specific time trends.

Thus, this research can be viewed as a test of whether the finding that death rates increase when labor markets strengthen, obtained using aggregate national data, is generalizable across industrialized countries that often have quite different economic institutions, lifestyles, and systems of medical care. However, this same heterogeneity may raise concerns if the results may vary with the institutional features of the specific nations. For instance, it seems plausible that the macroeconomic effects differ in countries with strong social insurance systems (e.g. income replacement policies) or labor protection laws (defined by worker termination policies and maximum allowable hours or days per week of work), when compared to those with weaker safety nets or constraints on employers. A full examination of these issues is beyond the scope of the current analysis but we do provide an exploratory investigation of how the results vary with the comprehensiveness of social insurance systems.

Our major finding is that robust economic conditions are associated with increases in total mortality rates and several important sources of death. In the preferred model, a 1% point decrease in the national unemployment rate is estimated to raise total mortality by 0.4% and deaths from cardiovascular disease, influenza/pneumonia, liver disease, motor vehicle deaths and other accidents by 0.4, 1.1, 1.8, 2.1 and 0.8%, respectively. Conversely, deaths from cancer are largely unaffected and there is some evidence that suicides and homicides decline in good economic times. We also uncover evidence of larger procyclical fluctuations in mortality for countries with relatively weak social insurance protections, as proxied by public social spending as a share of GDP, compared to those with more extensive programs.

These results suggest that physical health deteriorates when labor market conditions improve and generally accord with the findings of recent research using similar methods and data (e.g. Ruhm, 2000, Tapia Granados, 2004, Neumayer, 2004). The estimated effects of unemployment rates on mortality are of comparable absolute size to previous results based on the U.S. states (Ruhm, 2000); however, the mechanisms for these effects may be complicated and vary across sources of death. For instance, higher incomes are predicted to reduce some but raise other types of deaths, and the dynamics of the adjustment to a sustained change in unemployment differs with the cause of mortality.5

Although we use standardized unemployment rates as our main proxy of labor market conditions, it is important to emphasize that reductions in mortality during bad times need not be restricted to or even concentrated among those becoming newly unemployed. To the contrary, job loss could induce stress that counteracts other beneficial effects and thus raise death rates among jobless individuals, even while overall mortality declines.6 There is also no reason to believe that all types of health respond in the same way. For instance, increasing stress provides one reason why mental health might deteriorate despite gains in physical well-being.7 Similarly, we expect to see larger fluctuations in deaths from sources such as cardiovascular disease that may be strongly and rapidly affected by changes in lifestyles, environmental factors and medical interventions, than from those like cancer that probably are not.

Section snippets

Why might deaths increase in good economic times?

Many researchers hypothesize that cyclical upturns benefit health by reducing the stress associated with economic insecurity (e.g. Brenner and Mooney, 1983, Catalano and Dooley, 1983, Fenwick and Tausig, 1994). However, there are at least three reasons why health might instead worsen. First, non-market “leisure” time decreases, making it more costly for individuals to undertake time-intensive health-producing activities such as exercise. Data from the Behavioral Risk Factor Surveillance System

Estimation strategy

We use linear regression to estimate the relationship between labor market conditions and death rates. Using the subscripts j and t to index the country and year, the basic specification is:Mjt=αt+Xjtβ+Ejtγ+Cj+εjt,where M is the natural log of the mortality rate, E the unemployment rate, X a vector of regressors controlling for the age and sex distribution of population, α year-specific intercept, C a country fixed-effect, and ɛ a disturbance term.

The year effect holds constant determinants of

Data

Annual information on death rates, country characteristics, and per capita disposable income are obtained for 23 OECD countries over the 1960–1997 period from the OECD Health Data 2000 (OECD, 2000).11 The outcomes examined are the total mortality rate and deaths from nine leading causes: malignant neoplasms (cancer), major cardiovascular (heart) disease, influenza/pneumonia, liver disease and cirrhosis, motor vehicle accidents, other

Mortality rises in good times

Table 2 summarizes the results of a variety of econometric models, all showing that deaths rise when labor markets are strong. Here and below, the tables display the estimated effect of a 1% point increase in the standardized unemployment rate. Specification (a) shows the results of regressing the natural log of the total death rate on unemployment, country demographic characteristics (sex and age), fixed-effects, year effects, and country-specific linear time trends. The statistically

Cause-specific mortality

Table 4 demonstrates that deaths from a variety of causes increase when labor markets strengthen. The econometric models again control for location fixed-effects, country-specific time trends, general year effects, national demographic characteristics, and (in specification b) per capita incomes. A one point fall in the standardized unemployment rate is estimated to raise deaths from cardiovascular disease, influenza/pneumonia, or liver disease by a statistically significant 0.4, 1.1, and 1.8%,

Dynamics

Economic conditions have been assumed to have only a contemporaneous impact on mortality until now. Information on the dynamics of the adjustment process was obtained by estimating models that included 4-year lags of the national unemployment rate and calculating the impact of a 1% point rise in unemployment persisting for k years as n=0kβˆtn, for βˆtn the regression coefficient on the n-year lag of unemployment.

The adjustment patterns vary substantially across causes of death.29

Differences across social insurance systems

The macroeconomic effects could be mitigated or accentuated by institutional factors such as the type and availability of health insurance or, more generally, the comprehensiveness of the social insurance system. Stronger procyclical fluctuations might occur in countries with relatively weak social protections if individuals have incentives to work particularly hard during good economic times to offset the effects of reduced incomes during downturns. Conversely, an employment-based system of

Discussion

We provide evidence that deaths increase when labor markets strengthen. In the preferred specification, a 1% point fall in the national unemployment rate is estimated to raise total mortality by 0.4% and deaths due to cardiovascular diseases, influenza/pneumonia, liver ailments, and vehicle accidents by 0.4, 1.1, 1.8, and 1.9%, respectively. These patterns are similar to, although generally somewhat weaker than, those obtained in Ruhm's (2000) study of the United States using comparable methods

Acknowledgements

We are grateful to the editor and four anonymous referees for careful and thoughtful comments. Financial support from the Swedish National Institute of Public Health and the Swedish Council for Working Life and Social Research (Contract 2002-0376) to Gerdtham and the National Science Foundation (SES9876511) to Ruhm is gratefully acknowledged. The opinions, findings, conclusions, or recommendations expressed are those of the authors and do not necessarily reflect the views of the funding

References (66)

  • P.B. Mortensen et al.

    Psychiatric illness and risk factors for suicide in Denmark

    Lancet

    (2000)
  • E. Neumayer

    Recessions lower (some) mortality rates: evidence from Germany

    Soc. Sci. Med.

    (2004)
  • C.J. Ruhm

    Economic conditions and alcohol problems

    J. Health Econ.

    (1995)
  • C.J. Ruhm

    Good times make you sick

    J. Health Econ.

    (2003)
  • C.J. Ruhm

    Healthy living in hard times

    J. Health Econ.

    (2005)
  • C.J. Ruhm et al.

    Does drinking really decrease in bad times?

    J. Health Econ.

    (2002)
  • A. Wagstaff

    Time series analysis of the relationship between unemployment and mortality: a survey of econometric critiques and replications of Brenner's studies

    Soc. Sci. Med.

    (1985)
  • Arai, M., Thoursie, S.P., January 2001. Incentives and selection in cyclical absenteeism, Mimeo, Stockholm...
  • D.B. Baker

    The study of stress at work

    Annu. Rev. Public Health

    (1985)
  • M.H. Brenner

    Mental Illness and the Economy

    (1973)
  • M.H. Brenner

    Trends in alcohol consumption and associated illnesses: some effects of economic changes

    Am. J. Public Health

    (1975)
  • M.H. Brenner

    Mortality and the national economy

    Lancet

    (1979)
  • G. Calvo

    Quasi-Walrasian theories of unemployment

    Am. Econ. Rev.

    (1979)
  • R. Catalano

    The health effects of economic insecurity

    Am. J. Public Health

    (1991)
  • R. Catalano et al.

    Health effects of economic instability: a test of the economic stress hypothesis

    J. Health Soc. Behav.

    (1983)
  • K.Y. Chay et al.

    The impact of air pollution on infant mortality: evidence from geographic variation in pollution shocks induced by a recession

    Q. J. Econ.

    (2003)
  • K.B. Clark et al.

    The dynamics of youth unemployment

  • P.J. Cook et al.

    Homicide and economic conditions: a replication and critique of M. Harvey Brenner's new report to congress

    J. Quant. Criminol.

    (1986)
  • R. Dehejia et al.

    Booms, busts, and babies’ health

    Q. J. Econ.

    (2004)
  • D. Dooley et al.

    Personal and aggregate unemployment and psychological problems

    J. Soc. Issues

    (1988)
  • R.D. Edwards

    Commentary: work, well-being, and a new calling for countercyclical policy

    Int. J. Epidemiol.

    (2005)
  • G. Esping-Anderson

    The Three Worlds of Welfare Capitalism

    (1990)
  • S.L. Ettner

    The relationship between labor market outcomes and physical and mental health: exogenous human capital or endogenous health production

  • Cited by (401)

    • Economic development and health status in Iran: A panel data analysis

      2023, Environmental and Sustainability Indicators
    View all citing articles on Scopus
    View full text