From Industrial Relations to Personal Relations:
The Coercion of Society

Youth Suicide and Youth Unemployment

Barry Maley

Suicide, as a cause of death in the general population, lags a long way behind cancer and cerebrovascular disease. The overall suicide rate is about 17 per 100,000 of the population per annum. This is less than one tenth of the rate of cancer at 180 per 100,000 of population (ABS, 1994:55).

However, for younger age groups, which have low overall rates of mortality, suicide now competes with road accidents as the leading cause of death. In 1992, in the 15-24 year-old age group there were 455 suicides. Consequently, any change in the suicide rate for young people has significant effects on the overall level of youth mortality. In the 1980s and l990s youth suicide rates increased at the same time as suicide rates for older age groups were decreasing (ABS, 1994:55).

The increase for young males, in particular, has been quite dramatic. The figures have led many commentators to begin searching for social, economic and cultural changes specific for youth, and especially for young males, that might be implicated in these changes. Table 1 summarises the relevant statistics.

In terms of international comparisons, Table 2 shows that the youth suicide rate in Australia is the fourth highest in the world.

Sex Differentials

Men in general are four times more likely to commit suicide than women. In the 15-24 age groups, males commit suicide five times more frequently than females (ABS, 1994:56). One factor here may be that men bent on suicide do it more effectively than women and by more violent and certain means---such as guns, explosives and hanging. Females use poisons and sedatives which are slower and whose effects may be nullified by prompt treatment.

Urban/Rural Suicide Rates

For men and women of all ages, there is little difference in the suicide rates for urban and rural areas. But the suicide rate for young men in rural areas is markedly different, as shown in Table 3. This may have something to do with the readier availability of firearms for young men in rural areas, perhaps in synergistic interaction with intensification of some of the socio- economic factors to be discussed later, in those areas.

Why Do People Kill Themselves?

Suicide is, of course, a very private act, the precise reasons for which in individual cases can rarely be known to others. Most psychiatrists naturally emphasise mental states in talking about causation and say that it is due to such things as depression, schizophrenia, drug addiction, alcoholism, physical suffering and personality disorders. Yet there are 'studies which indicate that mental illness is not a necessary precondition of suicidal behaviour' and that 'only a small proportion of people committing suicide have been under psychiatric observation or treatment' (Ruzicka and Choi, 1993:114). As for substance-abuse suicides, the real question is: 'Why the substance abuse in the first place'? Neither mental disturbance nor substance abuse explains the overall rates of suicide or sudden or major variations in those rates (Ruzicka and Choi, 1993:114).

At the aggregated level, it is clear that we must look to broadly social factors to help explain national suicide rates and variations in those rates. Demographers and sociologists have long noted the statistical correlations between changing social conditions and changing patterns of suicide. Indeed, the examples are multiple and fascinating and this has prompted speculation about underlying causes.

Durkheim and Social Integration

Almost 100 years ago, in his classic study of suicide, the great French sociologist Emile Durkheim (1897), argued for the social causation of suicide. He showed how suicide rates within and between countries varied with social circumstances. Catholics and Jews killed themselves less frequently than Protestants, and the married killed themselves less frequently than the unmarried or the divorced. There were fewer suicides in war than in times of peace, and more in times of economic crisis; and so on.

He hypothesised that the two critical social variables were what he called 'social integration' and 'social regulation'.

There is a huge literature on these two variables and the nature of the social realities they purport to represent. For our purposes here, I intend to amalgamate the elements Durkheim identified into the single variable of social integration and to contrast two polar kinds of individual situation: the socially integrated individual; and the socially detached, isolated or unintegrated individual.

Integration, then, is the extent to which individuals are tied in to their societies and feel personal fulfilment in sustaining a variety of social and economic attachments made possible by those ties. It is the idea of social and interpersonal embeddedness. To be 'embedded' or 'integrated' in this way is also to accept the moral authority of the embedding social relationship or association. It is the regulation of the relationship by this authority which Durkheim has in mind when he identifies 'regulation' as a key variable. But, as I said, let's keep it simple by treating integration as entailing both ideas.

Integration has the further connotation of cohesiveness, or what Durkheim calls the 'common conscience' of shared sentiments and beliefs.

What today we might call detached and isolated individualism, is the opposite state of affairs, where individuals lack firm social and interpersonal ties and do not share in the common conscience, or find themselves alienated from it. They live lives only tenuously connected to the collective life or to interpersonal relations; or they find their bonds to it suddenly severed---such as in bereavement. Indeed, 'bereavement', in a metaphorical sense, is an apt term for various kinds of social and interpersonal loss that need not entail the death of a person. The important notion is the idea of social and/or interpersonal isolation in a spiritual, mental or emotional sense. You will remember that Socrates chose to swallow the hemlock rather than face exile and that kind of life.

So, in these conditions, Durkheim argues, such an individual, denied all hope of fulfilling his nature as a socially purposive and emotionally transacting being; and cut off from, or denied, a meaningful place in the social order, finds life intolerable and ends it. Insofar as social circumstances conspire to place more and more individuals in this unintegrated position, so will the suicide rate rise.

It should be noticed, incidentally, that Durkheim, in speaking of social integration, is talking about a balance between individual freedom and social attachment. The integrated state is not a state of rigid collectivism. Indeed, that state is characterised by excessive regulation and high suicide rates. The integrated state is one in which multiple individual needs and aspirations can be realised through a variety of voluntary social and interpersonal attachments. Research has shown, for example, that mental and emotional support within a family strongly reduces depression and suicidal behaviour (Bjarnason, 1994:204).

Against that theoretical background, let us look very briefly at just one example of a social correlate of suicide before examining suicide rates and unemployment statistics.

Suicide and Marriage

The available statistics from those societies that compile them, show that suicide is considerably higher among those who are not married than among those who are married (Ruzicka and Choi, 1993:108). This is certainly true of Australia. For the sake of brevity, Table 4 sets out the age-standardised ratios of suicides according to marital status, only for the years 1986-90; but the pattern for these years is similar to that for earlier years.

The number 100 stands for the average level of suicide in Australia as a whole. A ratio higher than 100 indicates an above-average suicide mortality, and one less than 100 a below-average suicide mortality.

The figures are revealing. They show, for example, that divorced men had a suicide rate twice that of the general population and three times that of married men. The female pattern is similar, except that divorce seems to be somewhat more devastating for women, and widower status more devastating for men. We can conjecture why this might be so.

However, two explanations have been suggested for the differences revealed here. One is that those who divorce are a group selected in terms of mental health status (Ruzicka and Choi, 1993:109-110). That is, they might include a preponderance of those mentally or emotionally unfitted for marriage and whose marriages are ill-starred from the start. The other explanation is essentially Durkheimian. Marriage, it seems for most, is good for your health. Marriage is a form of social integration which attaches people to an important social institution and common sentiments, and carries, too, all sorts of interpersonal satisfactions and fulfilments that serve our social and emotional needs in a health-sustaining way.

Youth Unemployment and Youth Suicide

If we are to hypothesise a causal link between youth unemployment and youth suicide, in order to verify the hypothesis we should expect to find a significant statistical correlation between the two phenomena; we should be able plausibly to suggest that the fact of unemployment generates a life-taking motive under circumstances that can be specified; we should specify those circumstances; and we should then demonstrate that they are present where the statistical correlation between suicide and unemployment is strong.

Granted that individuals who are poorly integrated in a social and interpersonal sense are more likely to suicide, is employment an important condition of the social integration of young men? There is a fair body of evidence showing the positive effects of employment on the well-being of those of working age. Research has shown that 'Earned income enhances self-esteem and a sense of mastery, which, in turn, increases overall well-being' (Ishii-Kuntz:497). The magnitude of this effect increases with the extent to which the persons concerned are identified with, and identify their own self-esteem in terms of, bread-winning roles and responsibilities.

I take it that it is still the case that male identity continues to be more deeply attached to bread-winning capacity and roles than female identity, despite the great changes for women over recent years. If that is generally true, we would expect a stronger response to unemployment, with its consequence of reduced self-esteem and well-being among young males, than among young females. There would therefore be a stronger association between unemployment and suicide for young males than for young females. This might diminish over time as young females identify themselves more strongly with non-maternal and non-domestic careers. Nevertheless, it is an important fact that maternity and domesticity remain either central, or alternative, sources of social identity and self-esteem for the great majority of women in a way that simply cannot be true for young men.

Also, we have seen that the married state protects against suicide and we must assume that the urge to marry is still strong amongst most men and women because 80 per cent of them continue to get married.

However, marriage is effectively ruled out for unemployed young males in their early twenties, but not for unemployed females of the same age. Not so long ago this was the age at which most men and women got engaged and married. Since then, age at marriage has steadily risen. De facto relationships are common, but they are about ten times more likely to break up than marriages.

What I'm suggesting is that unemployment for young men represents a serious failure of social integration in itself. But it also exacerbates the further failure of integration represented by non-marriage or an unstable substitute for marriage. It is likely that a high proportion of males in the 20-24 age group would no longer live with their parents and siblings (Morrell et al., 1993:755) and would therefore be deprived of their emotional and other supports in a stressful situation, thus reinforcing the isolation and social detachment characteristic of unemployment and the socially unintegrated state.

A corollary of what I am saying is that if employment is so important for male identity and well-being, this should be just as true of middle aged men as of young men. We would therefore expect rates of suicide amongst middle-aged men to be as sensitive to their unemployment rates as amongst young men, subject to possible modifications attributable to higher marriage rates among middle-aged men. Figure 1 shows very high rates of suicide among older men during the 1930s depression when unemployment was much higher among older men than among young men. That situation was completely reversed, however, from 1980 on.

We have already noted the quite striking differences between male and female suicide rates in general. There is no significant or consistent correlation between female unemployment and female suicide rates. The remarkable thing about female suicide rates is their steadiness, which, apart from a blip in the 1960s usually attributed to a sedative suicide epidemic, has remained around the 5 per 100,000 mark for over 80 years, and which has not been responsive to unemployment rates.

But the story is quite different for males. Morrell and others, using Australian records from 1907 to 1990, investigated trends in suicide stratified by age and gender, with an emphasis on suicide by young men, and they compared these trends with unemployment rates. The results are summarised in Figure 2.

  • Male suicide rates have fluctuated sharply over this 83-year period. The rates went down during two world wars---that is, in periods when national purpose and cohesiveness were strongest and unemployment very low. The highest rates of male suicide occurred during the Great Depression in the early 1930s when unemployment was at its worst in this century. In 1912 to 1915 and 1920-21 there were suicide peaks coinciding with peaks of unemployment.

Youth Unemployment

The unemployment rates for young men and women as at August 1993 were as shown in Table 5.

Suicide rates among 20-24 year old males reached their highest point in this century in 1990, with approximately 36.1 deaths per 100,000 in this age group. This is the most suicidal group in the country, with the exception of men over 75.

As the figures in Table 5 show, unemployment amongst 15-19 year-old males is about one-third higher than for 20-24 year-old males; yet their suicide rate is lower. Why might this be so? I have already suggested a possible answer. More 15-19 year old males are still living with their families and thus have the emotional and integrative support which research has shown to be so important. Furthermore, it is likely, for some of the reasons I have already given, that pressures for financial independence, bread-winning capability, and the establishment of social identity hinging on employment, a career and marriage, are not yet as insistent for 15-19 year olds as for 20-24 year olds, for whom the expectations to take up adult responsibilities and careers are very powerful indeed. Young men in the 20-24 age group exhibit the closest association between unemployment and suicide. The statistical correlation is high, as shown in Figure 3.

Morrell and his colleagues comment on this as follows: 'If 20-24 male unemployment rates are taken as a ratio of the overall unemployment rate, and 20-24 male suicide rates as a proportion of the male age standardised suicide rate, the R2 value increases to 80%' (Morrell, 1993:753). This is illustrated in Figure 4.

It is also worth noting that the median duration of unemployment is longer for males in the 20-24 age group than in the 15-19 year group, as shown in Figure 5, prepared for me by Mr John Logan.

There is little doubt that there is a strong association between suicide and unemployment rates for 15-24 year old males in this country, and a particularly strong one for the 20-24 year old group. This is consistent with a similar, but in most cases less strong, association in other OECD countries (Morrell, 1993:754). The crucial question is this: Are we entitled to conclude that high unemployment rates for young men will cause high suicide rates among them?

Allowing that it is virtually impossible to prove for any particular suicide what the determining conditions were, at the aggregate level we are dealing with here, I believe we are entitled to conclude that the evidence presented strongly 'supports the hypothesis that unemployment is significant as a predisposing factor for increasing the risk of suicide, especially in males' (Morrell et al., 1993:755), and especially in the 20-24 age group.

And, above all, we have strong grounds for criticising public policies which are, predictably but avoidably, exposing thousands of young men to the kinds of anguish, idleness and despair that arise from unemployment and which, for too many of them, may have suicide as one of its consummations.

Bob Day referred at this conference to the need to liberate employers and employees from the 'slavery' of the award system. That system is especially pernicious to the extent that it condemns young men to idleness and isolation. That is a particularly deadly form of servitude.


ABS (Australian Bureau of Statistics) 1994, Australian Social Trends, Catalogue No. 4102.0, ABS, Canberra.

Bjarnason, Thoroddur 1994, 'The Influence of Social Support, Suggestion and Depression on Suicidal Behaviour Among Icelandic Youth', Acta Sociologica, Vol. 37: 196-206.

Durkheim, E. 1897, Le Suicide: etude de sociologie. English translation John A. Spauld and George Simpson (1951), Suicide: A Study in Sociology, Glencoe, The Free Press.

Ishii-Kuntz, Masako 1994, 'Work and Family Life: Findings From International Research and Suggestions for Future Study', Journal of Family Issues, Vol. 15 No. 3, September: 490-506.

Morrell, Stephen, Richard Taylor, Susan Quine and Charles Kerr 1993, 'Suicide and Unemployment in Australia 1907-1990', Social Science and Medicine, Vol. 36 No. 6: 749-756. (Figures 1-4 in the text reprinted from this article with kind permission from Elsevier Science Ltd., The Boulevard, Langford Lane, Kidlington OX5 lGB, UK).

Ruzicka, L.T. and C.Y. Choi 1993, 'Suicide Mortality in Australia, 197-1991', Journal of the Australian Population Association, Vol. 10 No. 2: 101-117.

Table 1

Suicide death rates(a) by age and sex

15-24 25-34 35-44 45-54 55-64 65-74 75+ Total
rate rate rate rate rate rate rate rate
Males 1975 14.8 18.3 23.1 27.0 23.1 30.2 30.4 20.9
1980 18.0 22.7 22.6 23.9 23.9 24.3 30.6 27.5
1985 21.4 26.4 22.7 23.2 24.9 26.9 35.7 24.3
1990 26.5 29.5 25.8 24.3 23.2 26.2 35.1 26.7
Females 1975 4.6 7.2 10.8 13.9 13.4 10.0 7.0 9.2
1980 4.5 6.7 8.6 11.7 10.2 9.0 6.7 8.0
1985 4.8 6.0 7.0 9.7 8.8 7.9 7.4 7.0
1990 4.9 7.1 6.9 7.3 8.1 7.7 8.4 6.9
Persons 1975 9.8 12.9 17.1 20.6 18.1 19.3 15.3 15.2
1980 11.4 14.9 16.0 18.0 17.0 16.1 15.4 15.1
1985 13.3 16.3 15.0 16.6 16.8 16.6 17.7 15.6
1990 15.9 18.4 16.5 16.0 15.7 16.2 18.3 16.7

(a) Rates per 100,000 in the same age and sex group. Source: ABS, Social Trends, Cat. 4102.0, 1994.

Table 2

Suicide rate(a), 15-24 year olds

Country Year Males Females Persons
rate rate rate
Iceland 1991 61.0 4.9 33.3
Finland 1991 42.2 7.3 25.1
New Zealand 1989 37.9 7.0 22.6
Australia 1991 26.7 6.4 16.7
Switzerland 1991 26.0 6.2 16.2
Canada 1990 24.6 5.0 15.0
Norway 1990 22.1 6.3 14.4
Sweden 1989 19.8 8.3 14.2
USA 1989 22.2 4.2 13.3
Hungary 1991 19.6 5.3 12.6
Singapore 1990 13.3 7.7 10.6
Germany 1990 14.4 4.3 9.5
Poland 1991 15.8 2.4 9.3
Ireland 1990 14.2 4.1 9.3
France 1990 14.1 4.4 9.3
Denmark 1991 12.0 3.6 7.9
UK 1991 11.5 2.3 7.0
Japan 1991 9.1 4.7 7.0
Netherlands 1990 8.2 3.6 5.9
Spain 1989 8.4 1.9 5.2
Israel 1989 6.2 3.6 4.9
Portugal 1989 6.9 2.1 4.6
Italy 1989 5.1 1.6 3.4
Greece 1990 5.2 1.1 3.2

(a) Rates per 100,000 in the same age and sex group.

Source: World Health Organisation (1992) World Health Statistics Annual

Table 3

Urban/rural suicide rates(a), 1987-91

Urban Rural

15-24 All 15-24 All
Sex years ages years ages
rate rate rate rate
Male 25.0 21.3 36.6 22.6
Female 5.1 6.0 5.0 4.0
Total 15.2 13.6 21.6 13.5

(a) Rates per 100,000 in the same age and sex group.

Source: ABS, Social Trends, Cat. 4102.0, 1994.

Table 4

Suicide and Marital Status

Age-standardised ratios (where 100=average level of suicide)

Marital Status Male Ratio Female Ratio
Currently married 68.7 67.3
Never married 128.6 138.2
Widowed 201.2 132.1
Divorced 206.4 243.7

(Adapted from Ruzicka, 1993:109)