Lives saved during economic downturns
Worldwide, countries have been restricting work and social activities to counter an emerging public health crisis due to the coronavirus pandemic. These measures have caused dramatic increases in unemployment in the short run, with an expected deepening of the recession in the long run. Some predict the ‘Great Lockdown’ to be as destructive as the Great Depression (Gopinath 2020). Internationally, a debate has ensued regarding the cost of lockdowns. Some commentators argue that the ‘draconian measures’ will do more harm than good due to the economic contraction itself but also due to the mental health impacts of the imposed social isolation (e.g. ABC 2020 and Benson 2020 for Australia, Giuffrida 2020 for Italy, Collins and Cox 2020 for the UK).
Our research (Atalay et al. 2020) seeks to contribute to both the domestic and international debate by adding new Australian evidence to an existing literature that finds mortality rates fall during recessions. The evidence from the past four decades suggests that in Australia, economic downturns have very little impact on mortality, except to reduce vehicle transport deaths. We find no impact on deaths by suicide. These findings suggest that the direct impact of the current economic contraction itself may indeed be associated with a decline in mortality, at least in countries with universal health care. Our estimates imply 425 fewer deaths if the Reserve Bank of Australia’s expectations of an increase in the unemployment rate from 5.1% to 10% are realised by the end of 2020. We present this estimate with caution given that the current economic crisis is unprecedented. Naturally, our findings do not preclude an impact on wellbeing from the recession or indeed on wellbeing and ultimately suicides from the lockdown itself. Yet, on the flipside, the effect of this recession on mortality especially due to transport accidents may well be more pronounced due to the impact of the imposed lockdowns on traffic congestion in conjunction with increased opportunities for working from home.
Our analysis of the relationship between mortality and macroeconomic conditions builds on a burgeoning literature that traces back to the seminal work of Ruhm (2000, 2015). Ruhm explored the possibility that macroeconomic downturns have positive consequences for the health of a nation. The argument was that even though economic downturns usually come with financial hardship, they leave people with more time to seek medical treatment, to socialise and care for their relatives, and to engage in healthier lifestyles. People are also expected to have fewer accidents, because they spend fewer hours in cars and are less likely to be exposed to hazardous workplaces if unemployed. Thus, at the aggregate level, weaker macroeconomic conditions, as reflected in higher unemployment rates, could be associated with better mental and physical health, and thus lower mortality rates. Although a controversial hypothesis, the evidence base from a broad range of countries (the US, Canada, Mexico, Germany) and regions (OECD, Asia-Pacific) is overwhelmingly in favour of the suggestion that in times of higher unemployment, the aggregate number of deaths falls (e.g. Ruhm 2000, Gerdtham and Ruhm 2006, Miller et al. 2009, Ariizumi and Schirle 2012, Lin 2009, Neumayer 2004, Gonzalez and Quast 2011). Our research contributes to this literature, providing new evidence from Australia exploiting administrative time-series data on cause-specific mortality between 1979 and 2017 that varies by state, sex, and age sourced from the Australian Institute of Health and Welfare (AIHW). The Australian experience has not been studied previously and provides important insights from an affluent OECD country with universal health care and a relatively generous social safety net.
Figure 1 shows our main findings for the relationship between unemployment rates and mortality, across all ages, and for five broad age groups: children (ages 0-14), teens and young adults (ages 15-24), adults ages 25-34, the middle-aged (ages 35-64) and the elderly (ages 65+). Overall, we find no effect of unemployment on all-cause mortality (see the left-most bar in the chart), with a 1 percentage point increase in the unemployment rate associated with a zero impact on mortality (-0.02% and statistically insignificant). Yet, we do see that mortality falls in times of high unemployment for the younger age groups, most notably (and statistically significantly) for those aged 15-34. Further analysis by gender reveals that the significant relationships found for the 15-24 and 25-34 year-age groups are driven by falls in mortality among men, rather than for women. For those aged 35-64 and 65 or older, we do not observe any significant relationship, either overall or in any age group within that age range, whether grouping in five- or ten-year age bands.
Thus, our findings based on Australian data are partially in line with the findings presented in Ruhm (2000) and Ruhm (2015) for data before 2000, and Stevens et al. (2015) in the institutional context of the US, where no universal health care exists. For instance, Ruhm (2000) finds significant procyclicality in mortality for 20-44 year olds (-1.9%), and similar to us, no effect for middle-aged or older groups. Stevens et al. (2015) also finds significant procyclicality for 15-29 year-old men (and 15-24 year-old women), which range in magnitudes between -1.1% and -1.8%. Like Stevens et al. (2015), we also find stronger impacts of unemployment on mortality for men than for women.
Our data also allow us to dig deeper by considering cause-specific mortality rates to assess whether our findings above are driven by deaths from a particular cause. By and large, we find no statistically significant effects of unemployment on any of the causes of deaths. There is one consistent and important exception. Higher unemployment is associated with fewer vehicle accident (road) deaths. An increase in the unemployment rate by 1 percentage point is significantly associated with a 6% decrease in transport accidents (p-value < 0.05). This effect is twice as large as found in the international literature and translates into 88 fewer deaths per year. The number of lives saved are five time larger for men (73 fewer deaths) than for women (15 fewer deaths). This effect is also stronger for the working age population. Thus, the finding of a significant negative relationship between mortality and unemployment rates for young and prime age adults as shown in Figure 1, is driven by lower mortality due to vehicle transport deaths. This finding of fewer deaths due to road accidents in weaker economic times, has been consistently documented in many countries including but not limited to the US, Germany, Canada and France, where studies find mortality due to vehicle accidents is reduced by around 2-3% for every 1 percentage point increase in unemployment (Ruhm 2000, Neumayer 2004, Gerdtham and Ruhm 2006, Lin 2009, Ariizumi and Schirle 2012, Brüning and Thuilliez 2019).
There are some notable differences between the US and Australian experience. First, unlike results for the US in Stevens et al. (2015), we do not find any impact of weaker economic conditions on the mortality rates of very young children (0-4 years) and older people (65-84 years). Significant reductions in mortality during recessions in these age groups are interpreted by Stevens et al. (2015) as evidence against the lifestyle hypothesis and in favour of an alternative interpretation. Quality of health care in the US, as Stevens et al. argue, relies on the quality of workers that can be hired for hospitals and nursing homes. Second, in contrast to Stevens et al. (2015), we do not find a significant relationship between unemployment and mortality due to suicide, heart disease, respiratory deaths, cerebrovascular disease, pneumonia or influenza. We suggest that a plausible source of the difference in our findings and those for the US is that in countries with universal health care like Australia, health care quality and individuals’ ability to access that care may be less affected by business cycle variations because public funding is consistently provided and constraints on household budgets in weak economic times do not constrain access to health care. Thus, universal health care in Australia may play a protective role, serving to insulate the health and wellbeing of groups that are not directly affected by fluctuations in the labour market, from variability in economic conditions. A similar point is made by Ariizumi and Schirle (2012) who also find no impact of business cycle variations on the mortality of the very young and older age groups in Canada, a country with a similar health care system to Australia. Gerdtham and Ruhm (2006) also show that among OECD countries, in those with stronger social insurance systems, the negative relationship between mortality and unemployment rates is weaker.
Atalay, K, R Edwards, S Schurer and D Ubilava (2020), “Lives Saved During Economic Downturns: Evidence from Australia”, IZA Discussion Paper 13742
Ariizumi, H and T Schirle (2012), “Are recessions really good for your health? Evidence from Canada”, Social Science & Medicine 74(8): 1224-1231.
Australian Broadcasting Corporation (ABC) (2020), “‘Extreme’ COVID-19 epidemic better than lockdown argues economist, but others strongly disagree”, 20 April.
Benson, S (2020), “Coronavirus Australia: suicide’s toll far higher than virus”, The Australian.
Brüning, M and J Thuilliez (2019), “Mortality and Macroeconomic Conditions: What can we learn from France?”, Demography 56(5): 1747–1764.
Cameron, A C and D Miller (2015), “A Practitioner’s Guide to Cluster-Robust Inference”, Journal of Human Resources 50(2): 317-372.
Collins, A and A Cox (2020), “Coronavirus: why lockdown may cost young lives over time”, The Conversation.
Gerdtham, U G and C J Ruhm (2006), “Deaths rise in good economic times: Evidence from the OECD”, Economics and Human Biology 4(3): 298–316.
Gonzalez, F and T Quast (2011), “Macroeconomic changes and mortality in Mexico”, Empirical Economics 40: 305–319.
Gopinath, G (2020), “The Great Lockdown: Worst Economic Downturn Since the Great Depression”, IMF Blog.
Giuffrida, A (2020), “Italy’s lockdown has taken heavy toll on mental health, say psychologists”, The Guardian.
Lin, S-J (2009), “Economic fluctuations and health outcome: a panel analysis of Asia-Pacific countries”, Applied Economics 41(4): 519-530.
Miller, D L, M E Page, A H Stevens and M Filipski (2009), “Why Are Recessions Good for Your Health?”, The American Economic Review Papers & Proceedings 99(2): 122-127.
Neumayer, E (2004), “Recessions lower (some) mortality rates: evidence from Germany”, Social Science & Medicine 58(6): 1037–1047.
Ruhm, C J (2000), “Are recessions good for your health?”, Quarterly Journal of Economics 115(2): 617–650.
Ruhm, C J (2015), “Recessions, healthy no more?”, Journal of Health Economics 42: 17-28.
Stevens, A H, D L Miller, M E Page and M Filipski (2015), “The Best of Times, the Worst of Times: Understanding Pro-cyclical Mortality”, American Economic Journal: Economic Policy 7(4): 279–311.