11.2 Unemployment, low wage jobs and mental well-being
The experience of being unemployed and its consequences for future success in the labour market will be affected by its psychological impact, as well as by the signal that it gives employers. We here report what the psychology (and some economics) literature has to say about the consequences of unemployment for mental health, and whether it is worse for mental health to be unemployed as compared with being in a low wage, low skill job.
Within the field of psychology, the vast majority of research into employment and unemployment has focussed on the psychological consequences of unemployment, rather than on the psychological predictors of subsequent employment status (Winefield et al 1993). Psychologists face the same problem as economists in finding it difficult to distinguish causation from selection, when they observe that unemployed people exhibit signs of psychological distress: does unemployment cause the distress, or do the distressed become unemployed?
11.2.1 The psychological impact of employment status
The relationship between employment status and psychological well-being is two-way. O'Brien (1986:239) states that “the majority of studies show that unemployment produces, in most people, a state of dissatisfaction and distress”. It is widely found that people who become unemployed, compared to those who are employed, tend to suffer from damaged identity formation, lower self-competence, self-esteem, happiness and life-satisfaction, and have higher symptoms of stress and depression.
In an interesting recent study, Flatau, Galea and Petridis (2000) use large scale Australian health surveys to explore the empirical link between unemployment (distinguished by duration), part-time employment, non-employment and mental health. Compared with men who are in full-time employment, men not in the labour force (and not studying) have the worst mental health—worse than those who are unemployed of any duration. The negative effect of being out of the labour force is much more muted for women, though still apparent. The relation between mental health and unemployment varies by duration of unemployment, in a similar way for women and men. Mental health is worst for those unemployed for 13-26 weeks and over 52 weeks, by a substantial amount. This suggests that there may be adaptations occurring as the duration of unemployment extends. Flatau et al show that, controlling for a wide range of socio-demographic attributes and for levels of physical health, unemployed men and men employed part-time have lower levels of mental health than men employed full-time (aged 18-64). The impact of unemployment is diminished, but still statistically significant, if the level of equivalent income of the household is controlled for. This suggests that each of the loss of income and the direct experience of unemployment have deleterious effects on the mental wellbeing of men. The unemployment result, but not the negative effects of part-time employment, applies also to women.
Flatau et al rely on cross-section data, so we cannot be sure that unemployment/part-time employment cause rather than are caused by mental illbeing. And they contrast unemployment with all full-time employment, not with employment in a low wage job.
Within the psychological literature, some important comparisons have been made between the psychological impact of unemployment and “good” and “bad” employment. There are also three psychological theories of employment which implicitly state within them the expected psychological impact of employment of different levels of quality.
Firstly, Jahoda's (1981, 1982). “Deprivation theory” states that, in addition to its obvious function of providing income, employment (even bad employment) has five latent functions that are psychologically beneficial and keep us in touch with reality.
First, it imposes a time structure on the waking day; second, it provides regular social contacts with people outside the nuclear family; third, it imposes goals and purposes that transcend those of the individual (shared goals); fourth, it defines status and identity: and finally, it enforces activity
Jahoda, 1981:188
Additionally, Jahoda (1981:189) implies that even bad jobs are preferable to unemployment by stating "even unpleasant ties to reality are preferable to their absence".
In contrast to Jahoda's deprivation theory, Fryer (1986) proposed an “Agency theory”, in which the five supposed benefits of employment stated by Jahoda are claimed to often be costs of employment rather than benefits. If Fryer is right, bad employment may be detrimental to the individual's life satisfaction, perhaps to the point that no job would be better than the bad job. Agency theory stresses the proactive and independent aspects of humans, in which people wish to plan for themselves, whereas deprivation theory assumes that people are reactive and dependent (Winefield et al 1993).
The third psychological theory of employment is Warr's (1987) “Vitamin model”, which is concerned with the effects of certain environmental features on mental health. Warr suggests that nine features of the environment have a curvilinear effect on mental health, in an analogous manner to the way vitamins influence physical health. These nine features of the environment are: 1) opportunity for control; 2) opportunity for skill use; 3) externally generated goals; 4) variety; 5) environmental clarity; 6) money; 7) physical security; 8) opportunity for interpersonal contact; and 9) valued social position.
According to Warr, some of these environmental features act like vitamins A and D, in that very high levels of them may not only stop being beneficial, but can be harmful. Others resemble vitamins C and E, in that at very high levels they stop being beneficial but are not harmful. Warr suggests that three of the environmental features, money, physical security and valued social position, are like vitamins C and E, but that the rest are like vitamins A and D, and thus can be harmful in high doses. Warr's model does not explicitly distinguish between work and non-work environments. Rather, the extent to which any environment is beneficial to mental health depends on the extent to which it provides these nine environmental features. Therefore, there can be good and bad work environments, and good and bad non-work environments. Without suggesting that employment is necessarily better than unemployment in terms of these nine features, most good jobs provide them at the beneficial levels (Winefield et al 1993). But the vitamin model makes it possible in theory for bad employment to be worse for mental health than unemployment.
We note here that social welfare systems clearly and strongly embed the view that substantial numbers of people would prefer unemployment to some forms of employment. This is implicit in the steps that are taken to ensure that people who receive unemployment benefits are “genuinely looking for work”. Active steps are taken by most (all?) welfare states to prevent people from choosing unemployment over employment. The most dramatic recent example of policy directed to preventing a choice of welfare over employment is to be seen in the two-year limits on access to welfare implemented in the United States in 1996. Welfare economics, of course, is predicated on the assumption that people know their best interests. Social welfare policy implies that governments believe that many people, given the choice, would opt for unemployment over employment in the sort of jobs that they may reasonably expect to obtain. If we accept these two propositions, it leads us to conclude that for a sizeable proportion of people at the low end of the skill distribution, no job (with some welfare income) is better than a poor job.
A more explicitly stated, stronger opinion on bad employment is that of Leim (1992), who suggests that, by taking an unsatisfactory job, the worker gives up a sense of personal control and incurs damage to his or her sense of self. He claims that the psychological costs of accepting an unsatisfactory job are often greater than those incurred by remaining unemployed.
The psychological literature that has investigated this difference between “good” jobs, “bad” jobs, and unemployment, has predominantly supported the propositions of Fryer's (1986) agency theory and Warr's (1987) vitamin model, in that people in bad employment are usually no better off psychologically than those who are unemployed.
One method of operationalising “good” and “bad” jobs has been to measure the level of job satisfaction they provide a worker with. The sources of job satisfaction are challenging and interesting work, having pleasant co-workers, adequate pay and opportunities for advancement (Weiten, 1995). These studies have found that dissatisfying work environments can lead to subsequent psychological damage (Borgen, Amundson and Harder, 1988; Burris, 1983; Landy, Quick and Kasl, 1994). O'Brien and Feather (1990) found in a longitudinal study of school leavers that the positive benefits of employment for these young people depended upon the quality of employment they had. The quality of employment (defined as good or poor) was based on the degree to which their job allowed them to utilise their skills and education. School-leavers who obtained good quality employment had lower depressive affect, higher life satisfaction, higher internal control and higher personal competence than those who were unemployed. However, there was little difference on these variables between the unemployed and the poorly employed. School leavers who took poor jobs suffered negative effects to their psychological functioning, just as those who ended up unemployed. The only—but important-- difference was that the poorly employed reported feeling significantly more positive about their lives, even though they were similar on the other variables.
Kaufman (1982) and Leana and Feldman (1995) obtained similar findings in longitudinal studies of the quality of re-employment after being laid-off. Kaufman (1982) found that unemployed professionals who became re-employed in jobs that did not require a high degree of utilisation of the person's ability, knowledge or skills (termed “underemployed”), were no better adjusted than those who remained unemployed. Further, underemployment was found to be a highly stressful experience that was comparable to unemployment. Leana and Feldman (1995) found those who ended up with unsatisfactory employment or underemployment had higher levels of psychological distress and anxiety, and lower levels of life satisfaction than those satisfactorily employed, but there was a lack of differences between the unsatisfactorily employed and the unemployed. In fact, the unsatisfactorily employed reported significantly lower levels of life satisfaction than the unemployed, which suggests that a bad job can be worse in one way than unemployment.
Winefield et al (1993) have found in all the analyses they have conducted since 1984 on a sample of school leavers, that those who ended up with jobs that they regarded as unsatisfactory, were no better off psychologically than the unemployed. Whereas there were no differences between the unsatisfactorily employed and the unemployed on levels of self-esteem, locus of control, depression, negative mood, psychiatric symptoms, psychological distress, hopelessness and social alienation, the satisfactorily employed (and full-time tertiary students) displayed superior well-being on all of these psychological measures.
It can be concluded that all of these studies consistently suggest that “bad” or unsatisfactory employment is no better for a person's psychological well-being than having no job at all, which is in line with propositions of Fryer's (1986) and Warr's (1987) psychological theories of employment. These findings have important implications for government policy and practice. However, the four major longitudinal studies comparing “good” and “bad” jobs with unemployment (Kaufman, 1982; Leana and Feldman, 1995; O'Brien and Feather, 1990; Winefield et al., 1993), are not able to throw light on the long-term consequences for psychological well-being of bad jobs in comparison to good jobs and unemployment. Winefield et al’s (1993) longitudinal study suggests that people in unsatisfactory employment are more likely to enter satisfying employment after one year than to remain dissatisfied employed or to end up unemployed. In this study, Winefield et al (1993) examined how much stability there was in employment status from one data collection point to the next. They found that on average, 55% of those who were dissatisfied employed became satisfied employed one year later, and that the satisfied employed were the most stable group as 90% remained satisfactorily employed after one year. However, about 32% of the dissatisfied employed remained dissatisfied one year later, and another 7.4% found themselves unemployed one year later. Thus it appears that for about half of the dissatisfied employed, this dissatisfaction is not a long-term, stable employment status, which suggests that for these people the negative psychological consequences of dissatisfactory employment may be short lived. The psychological consequences for the 32% who remain in dissatisfactory employment over long periods of time is yet to be determined.
