ArticlesExcess mortality of unemployed men and women during a period of rapidly increasing unemployment
Introduction
Until the late 1980s the unemployment rate in Finland remained low (about 5%). However, by late 1990 the unemployment rate began to increase and by the end of 1992 had climbed to 15%, reaching a peak in early 1994 of 19%.1 Unemployment also grew in other European countries, although not as rapidly as in Finland. In 1994, the average European Union unemployment rate was 12%. Long-term joblessness has become more common in Finland and redundancy now also affects the white-collar occupations.2 Research into the possible health effects of unemployment is thus more timely than ever.
Previous studies of individuals have shown that mortality rates are higher in the unemployed than the employed persons.38 The causal effect of unemployment and selection bias have been suggested as two possible mechanisms to explain this finding.9
Becoming jobless and long-term unemployment may have adverse effects on health and increase the risk of premature death. These effects are largely assumed to be mediated through psychosocial stress, which affects both mental and physical health. An increase in tobacco, alcohol, and drug use may also be indicative of attempts to cope with this stress. Other outcomes of unemployment such as loss of income, material deprivation, loss of networks, and social stigma may have independent effects on health and mortality but may also increase the stress experienced by the unemployed.
Selection bias is encountered when unemployed people or those who have difficulty in becoming re-employed have pre-existing ill-health. Selection bias may also be due to socioeconomic factors such as social class and housing tenure; lifestyle risks like tobacco and alcohol consumption and poor diet or personal characteristics such as increased age, sex, physical weakness, and psychological dysfunction that increase the risk of premature mortality. Although the direct selection of people with pre-existing ill-health may not be of paramount importance,6, 7 lack of data about personal characteristics and lifestyle make it difficult to control for all possible selection factors.
Aggregate-level studies that use regional populations or occupational groups as their units of analysis, have generally not shown a relation between changes in unemployment and mortality rates.911 These results suggest that selection may account for a large part of the excess mortality of unemployed persons.
The aim of this study was to estimate the strength of the association between unemployment, re-employment, and mortality during rapidly increasing unemployment in Finland. The results for the total Finnish population of men and women are presented. As in previous individual-level studies, we sought to reduce the possible effects of selection by controlling for the relevant confounding factors for which data were available. Because we cannot account for all potential confounders, we applied a different method to assess the effects of selection and analysed the association between annual changes in individual employment and mortality during a period of a rapid increase in national unemployment rate. If selection were a major factor explaining the excess mortality among the unemployed, then this excess should be smaller when the general unemployment rate is high than when it is low. At a time of high unemployment, becoming and remaining redundant can be assumed to be less dependent on individual characteristics that increase the risk of ill-health.
Section snippets
Methods
This study linked the 1990 census records of all 25 to 59-year-old Finnish men and women with their death certificates for the period 1991–93. Data on annual employment and unemployment from 1987 to 1992 were obtained from Statistics Finland's labour force data files.12 These files cover all employment and unemployment episodes of individuals and are updated continuously. Statistics Finland established the linkage of data sets by means of personal identification codes.
Unemployed individuals
Unemployment before 1991 and mortality in 1991–93
Table 1 describes the relations between unemployment in the periods 1987–89 and 1990 and mortality in the period 1991–93.
Age-standardised mortality was clearly higher among unemployed than among employed men. Men who had experienced unemployment for the first time in 1990 (ie, just before the recession) had 141% (mortality ratio 2–41 [95% CI 2–18-2–68]) higher mortality than men employed both in 1987–89 and 1990. In women the corresponding excess mortality was only 35% (mortality ratio 1–35
Discussion
Both Finnish men and women who experienced unemployment in the 1987–92 period had greater mortality than did employed men and women, respectively, after control for age, education, occupational class, and marital status. Studies attempting to control for the effects of confounding have also found that the confounders do not fully explain the excess mortality of the unemployed.38 In the British Regional Heart Study,7 involving 6191 men aged 45 to 64 years, the investigators controlled for age,
References (11)
- Statistics Finland, Työvoimatilasto (Labour force statistics; in Finnish), Official Statistics of Finland, Labour...
- Statistics Finland, Työvoiman koulutus ja ammatit 1984-1992/1993 (Labour force by education and occupation status...
Unemployment and mortality among Finnish men, 1981-85
BMJ
(1990)- et al.
Loss of employment and mortality
BMJ
(1994) Economic activity and mortality of the 1981 Census cohort in the OPCS Longitudinal Study
Popul Trends
(1996)
Cited by (273)
Economic complexity and health outcomes: A global perspective
2020, Social Science and MedicineCitation Excerpt :By contrast, people living in a complex economy that produces a variety of sophisticated products are subject to more occupational choices. Importantly, it is well established that an individual's occupational status is a crucial determinant of health conditions (see, for instance, Martikainen and Valkonen, 1996; Morris et al., 1994; Strully, 2009). Specifically, unemployment affects health outcomes through several factors, such as social status and self-esteem, the consumption of alcohol and tobacco and material deprivation (Herzer, 2017).
The Great Recession and adverse birth outcomes: Evidence from California, USA
2019, SSM - Population HealthTrends in gender differences in health at working ages among West and East Germans
2019, SSM - Population HealthCitation Excerpt :This association may be causal, in that employment improves health, or it could result from the selection of healthy individuals into the labor market (Wilkinson & Marmot, 2003). A number of studies have found that unemployment increases the risk of morbidity and mortality (Bartley, Ferrie, & Montgomery, 2006; Cohen et al., 2007; Leist, Glymour, Mackenbach, van Lenthe, & Avendano, 2013; Martikainen & Valkonen, 1996; McKee-Ryan, Song, Wanberg, & Kinicki, 2005; Roelfs, Shor, Davidson, & Schwartz, 2011). The role of gender has also been discussed in the literature on employment and health.
Precarious employment at a young age and labor-market marginalization during middle-adulthood: A register-linked cohort study
2023, Scandinavian Journal of Work, Environment and Health