IJE Advance Access originally published online on August 24, 2006
International Journal of Epidemiology 2006 35(5):1357; doi:10.1093/ije/dyl178
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Letter to the Editor |
Mortality effects of the economic crisis in South Korea
Institute of Labor and Industrial Relations and School of Social Work, University of Michigan, Ann Arbor, MI, USA
E-mail: jatapia{at}umich.edu
Observational epidemiology is a tricky endeavour. Establishing cause–effect relationships, though most important in science1,2 is always very difficult. The difficulty is greatest, indeed almost insurmountable, when researchers try to establish causal connections between a unique event and other events that follow it. Social scientists are aware of post hoc ergo propter hoc fallacies. Of course, we can establish that an earthquake occurring, say, in January 2004, destroyed so many residential units (the causal connections are obvious), but before-after comparisons to ascertain whether the disaster increased the rate of suicide among those living in the affected areas may lead to not very solid conclusions. Say the suicide rate was 1.5, 1.4, 1.4, and 1.3 in the 4 years before the earthquake, ‘went down’ to 1.0 in 2004 and ‘increased’ to 1.2, 1.3, and 1.3 in the next 3 years. Do we dare to say that the rate of suicide was decreased by the earthquake? Of course, numbers matter; most scientists would likely agree that annual rates such as 1.5, 1.4, 1.4, and 1.3 before the event followed by 0.5, 0.6, 1.1, and 1.2 after the event are much more suggestive of a causal short-lagged effect of the earthquake on suicides.
The paper by Khang et al.3 tries to establish mortality effects of the strong depression affecting the South Korean economy in the late 1990s. The authors were obviously looking for increments in mortality, but they hardly found any. Indeed, suicides increased strongly during the recession, but they were more than compensated by a drop in deaths due to traffic injury, one of the major killers in modern society. As presented by Khang et al.3 these two effects are probably ‘large enough’ for most researchers to accept that they are linked to the economic meltdown of the late 1990s. But what about other causes of death? The authors' answer is that the short-term mortality effects of the crisis ‘were relatively small’ because ‘any shortterm effects of the economic decline were overwhelmed by the momentum of large declines in causes of death such as stroke, stomach cancer, and liver disease’. But what is ‘the momentum’ of mortality rates? There is plenty of recent research showing how fluctuations of the economy impact on general and cause-specific death rates.4–14 Leaving aside the 1990s demographic catastrophe in transitional Eastern Europe, investigators have found with few exceptions that recessions in market economies tend to raise suicides, while reducing (probably through decreases in atmospheric pollution, traffic, occupational stress, and harmful consumption) general mortality; infant mortality; adult mortality; and mortality due to CVD, injuries, liver disease, respiratory disease, and infections. This is quite consistent with what Khang et al.3 found in South Korea. If they had used longer series, transformed them by filtering or differencing, and performed some bivariate analysis15,16 perhaps they might have detected other effects.
Conclusions based on analysis of a single event and short time series are rarely reliable, unless effects are massive.
References
1 MacIver RM. Social Causation. Gloucester, MA: Peter Smith, 1973.
2 Mackie JL. The Cement of the Universe—A Study of Causation. Oxford: Clarendon Press, 1974.
3 Khang YH, Lynch JW, Kaplan GA. Impact of economic crisis on cause-specific mortality in South Korea. Int J Epidemiol 2005;34:1291–301.
4 Ruhm CJ. Are recessions good for your health? Q J Econ 2000;115:617–50.[CrossRef][Web of Science]
5 Ruhm CJ. Good times make you sick. J Health Econ 2003;22:637–58.[CrossRef][Web of Science][Medline]
6 Neumayer E. Recessions lower (some) mortality rates—Evidence from Germany. Soc Sci Med. 2004;58:1037–47 (erratum corrigendum in: Soc Sci Med 59:1993).[CrossRef]
7 Tapia Granados JA. Recessions and mortality in Spain, 1980–1997. Eur J Popul 2005;21:393–422.[CrossRef]
8 Tapia Granados JA. Increasing mortality during the expansions of the US economy, 1900–1996. Int J Epidemiol 2005;34:1194–202.
9 Chay KY, Greenstone M. The impact of air pollution on infant mortality: evidence from geographic variation in pollution shocks induced by a recession. Q J Econ 2003;118:1121–67.[CrossRef]
10 Dehejia R, Lleras-Muney A. Booms, busts, and babies' health. Q J Econ 2004;119:1091–130.[CrossRef]
11 Abdala F, Geldstein RN, Mychaszula SM. Economic restructuring and mortality changes in Argentina—Is there any connection? In: Cornia GA, Paniccià R (eds). The Mortality Crisis in Transitional Economies. New York: Oxford University Press, 2000, pp. 328–50.
12 De Rios Neto E, Carvalho JAM. Demographic consequences of structural adjustment: the case of Brazil. In: Tapinos G, Mason A, Bravo J (eds). Demographic Responses to Economic Adjustment in Latin America. Oxford: Clarendon Press, 1997, pp. 174–98.
13 Ortega Osona JA, Reher D. Short-term economic fluctuations and demographic behaviour: Some examples from 20th-century South America. In: Tapinos G, Mason A, Bravo J (eds). Demographic Responses to Economic Adjustment in Latin America. Oxford: Clarendon Press, 1997, pp. 129–55.
14 Gerdtham U-G, Ruhm CJ. Deaths Rise in Good Economic Times: Evidence from the OECD. Boston, MA, National Bureau of Economic Research, 2002 NBER Working Paper No. 9357.
15 Glass GV, Wilson VL, Gottman JM. Design and Analysis of Time-Series Experiments. Boulder: Colorado Associated University Press, 1975.
16 Diggle P. Time Series: A Biostatistical Introduction. New York: Oxford University Press, 1989.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||