Chapter 4: The Social Significance Of Unemployment

A chapter of Globalisation and Welfare State

Keywords: Labour Studies;

Dr Richard Smith, deputy editor of the British Medical Journal, described unemployment as a `medical problem’. (1) While unemployment has been treated as an economic problem with political overtones, Smith’s description reflects a growing recognition of unemployment’s impact on the health and welfare of individuals and their social groups.

Officially unemployment is defined as the situation of being without work, wanting work, and actively seeking work. It is usually measured as a percentage of the people in the total labour force or the total for some social group. The labour force is defined as those who are employed or officially unemployed (see appendix).

This is the international definition developed to enable cross-national comparisons. The definition is not entirely satisfactory, for it excludes those who are without work, want work, but are too discouraged by the state of the labour market to seek work. It also excludes those who are in effect unemployed but are working just a few hours perhaps in a very marginal job, and would desperately like more hour work. The indications are that this broader concept would double the numbers described as unemployed.

This challenging of the official definition is not a claim that it is wrong. Its main purpose is to track unemployment’s incidence over time, and between countries and social groups. However raising the difficulties draws attention the problem of defining unemployment. Note that unemployment is measured at a point in time, but for the unemployed it is an experience through time. Special problems are the long term unemployment, and the intermittently employed, who go through a repeated cycle of finding a job, employment, and losing it.

As well as having a seasonal cycle over the year, unemployment follows the business cycle, the fluctuation in the level of economic activity. The business cycle is not regular, but typically is of a longer period; near three years. Unemployment follows it inversely, so that when business activity is high unemployment is low, and vice versa. (2) Around these fluctuations unemployment also experiences secular – or long term – trends discussed in chapter 1, which observed that the increase in quantitative measures do not fully capture the rise in the stress of being unemployed.

Today the New Zealand official unemployment rate, at around 7 percent of the labour force, is in the middle of the range of OECD countries, whereas it would have been at the bottom end thirty years ago. While most OECD countries experienced a fall in their rates of unemployment during the 1980s, New Zealand experienced a sharp rise.

Unemployment is not spread evenly through the community. As a general rule the more marginal a social group is in the community, the more likely it is to experience a higher than average rate of unemployment.

This is vividly illustrated in the regional statistics. The metropolitan centres of Auckland, Christchurch, and Wellington all experience rates below the national average, with the rest of the country above. Within Auckland the relative rates between Takapuna, Auckland City, and Manakau reflect their social rankings.

On social variables unemployment tends to be more focused among
– the young, with 45 percent of the total unemployed are under the age of 25;
– ethnic minorities, with Maori rates just over three time European rates, and Pacific Island-Polynesian rates nearly as high;
– those without educational qualifications, with an unemployment rate is only about a third higher than average;
– among those in what are considered low level occupations; and
– among those in low incomes.

The one exception to the marginality rule is that it appears that there is not much difference between the unemployment rates between the genders: they may be fractionally higher for men. This probably reflects an artefact of the official unemployment measure. Women are more likely to be in those groups willing to work but not actively seeking work, or working inadequate hours. More generally the concepts of employment, unemployment, and not-in-the-labour force were devised from a male perspective, and involve considerable ambiguity when applied to many women.

The longer someone is unemployed the more likely they are to belong to these marginal groups. Moreover the long term unemployed have been a rising proportion of total unemployment.

In summary, unemployment is a disease with a low incidence among those responsible for its analysis and treatment. It has not been a fashionable disease, nor one which has been given high priority.

The process of redeployment during a working life probably affects every worker. But while they may appear as an official unemployment statistic, for many the period of unemployment is short and far from traumatic.

Indeed it is not obvious why unemployment should be detrimental, except insofar as there is a loss of income, which today is partly covered by redundancy payments, the unemployment benefits, and personal savings. Not that the unemployment benefit is generous. However it does not appear so miserable as to explain all the afflictions associated with the disease.

What seems to happen is that unemployment lowers one’s resistance to a range of other diseases. The longer one is unemployed, the lower the resistance. It is also evident that this lowering of resistance is contagious, particularly in the close family.

The empirical evidence points to a wide range of detrimental personal and social conditions which are associated with unemployment. Most of the evidence comes from overseas, but the little research that has been done here confirms the general picture in other high income countries applies to New Zealand.

As is typical in a lot of epidemiological research, scientifically controlled experiments are not possible. Correlations from actual life situations do not, prove causality, but such is the overwhelming nature of the evidence from so many different studies using so many different approaches, one can say with confidence that unemployment is a cause of some of these conditions. The consequential associations with unemployment can be categorised into the psychological, physiological, the social, and the economic.

The most spectacular of all the psychological associations is that the rate of suicide (and parasuicide) appears to be higher among the unemployed. The one anomaly seems to be the effect is not as strong among women as men, perhaps because of the ambiguities in the definition of labour force status for women swamp the underlying relationship. This appears to occur because employment not merely a means of obtaining income, but of has a much wider role in a person’s material wellbeing. Social Psychologist, Marie Jahoda captured this well when she suggested there existed latent social functions of work.

“Firstly among them is the fact that employment imposes a time structure on the working day. Secondly, employment implies regularly shared experiences and contacts with people outside the nuclear family. Thirdly, employment links an individual to goals and purposes which transcend his or her own. Finally employment enforces activity. (3)

Conversely to be unemployed is to lose these latent functions and the related social enjoyment of work, leading to poorer mental health. Warr identifies nine ways: financial worries; restricted behaviours and environments; the loss of `traction’, the way in which the structure of work pulls you along; the smaller scope for making decisions; the satisfaction of developing new skills; an increase in threatening and humiliating experiences; anxiety about the future; reduced quality of interpersonal contacts; and a decline in social position.

Again the evidence for the physiological consequences of unemployment is strongest for the most extreme case. Unemployment is associated with mortality. A British study using a longitudinal study based on the Population Census between 1971 and 1981 found that men between 15 and 64 had a standardised mortality ratio, adjusted for age and class, of 21 percent above the average if they were unemployed and seeking work. (4) There are numerous studies which associate various medical conditions with the unemployed, including them posing greater pressures on the medical services.

While it is not difficult to describe a process in which unemployment causes psychological distress it is less clear how mortality and physical morbidity is elevated by it, excepting the violence related events which may be accounted for by spare time, and alcoholism. Presumably that the unemployed are poorer may affect their physical health. But the evidence also provides tantalising support for the hypothesis that psychological health affects physical health, for even such biochemical conditions such as neoplasms (cancer) are higher among the unemployed. That the mechanisms are not solely physical but are operating through psychological and social processes is supported by studies which suggest that mortality and morbidity rates of the wives of unemployed men are also elevated.

The social consequences of unemployment are less well explored and more fragmentary. The scientific findings are generally less compelling as a result. Some of the research is supportive of the sorts of things to be expected given the psychological and physical evidence – family stress, including violence and marriage breakdown, and children with unemployed parents suffering from a deterioration in health, behaviour and educational attainment.

It is easy to blame rising crime and delinquency on unemployment, but the evidence of association rarely deals with causality. Boredom and availability of time may also explain the association of unemployment and the misuse of drugs – both alcohol abuse and the use of illegal drugs. The unemployed probably also smoke more. In the area of drug use, again causality (and a shortage of convincing studies) is a problem.

The economic consequence of unemployment is the loss of production. It has been calculated that reducing registered unemployment by 105,000 would increase GDP by 18 percent. In addition unemployment generates increased pressures on the use of resources. The list above indicates that demand medical resources rise with increasing unemployment.

These fiscal concerns are but a backdrop to the real tragedy that unemployment diminishes the health and welfare of the unemployed, and their families. Even were there sufficient resources to treat the resulting conditions, we should aim to pursue preventative rather than curative programs.

Unfortunately some unemployment is inevitable, as a byproduct of a modern dynamic economy, driven by the process of redeployment, in which workers move from low productivity and obsolete jobs to high productivity jobs in new and expanding activities. Sometimes the redeployment will involve moving to new worksites, firms, industries, and even regions. On such occasions the transition may involve a period of unemployment.

That does not mean that a nation must tolerate the high levels of unemployment where the unemployed are stressed to the point that they suffer the psychological, physiological and social consequences listed above. Full employment, defined as a state where the unemployed are not stressed, is a major economic policy objective.

Next Chapter Ch5: The Great Diversification: 1966-1984

1. R. Smith (1988)
2. See B.H. Easton, In Stormy Seas (1997).
3. M. Jahoda (1979).
4. C. A. Moser et al (1984).


Social behaviour never conforms to the unambiguous analytic categories of social scientists, and even less to the definitions of social statisticians. The labour market is no exception. The adult population tends to be divided into three categories: `employed’, `unemployed’, and `not-in-the-labour force’ (NILF). with the employed plus the unemployed making up `the labour force’. The concepts typically apply to a point in time, practically a week, so that in a slightly longer period (a month or a year) a person may be employed, unemployed, and NILF.

For statistical purposes a person is classified as `employed’, if they work for one hour in the week. That means that many people who one would think are practically unemployed but are doing a few hours work (perhaps mowing the neighbour’s lawns for a few dollars), may be classified as employed. To be `unemployed’ a person has to (for statistical purposes) be not at all employed (i.e. not working even one hour a week), and actively seeking work (which has a specific definition). Anyone else is NILF, even though they may want a job and be willing to accept one if offered (but are not actively seeking one, perhaps because in despair they think active search a waste of effort). This is an international definition. Its advantages is that it allows international comparisons, and is not subject to political interference. For many purposes it is not ideal, but usually sufficient data is available to provide a more refined measure if it is desired.

Note that the previous paragraph has avoided what is meant by `work’. If a person is paid for their economic activity this is classified as work, while a relative who assists the family in a financial enterprise is also so classified.

Usually when we think of employment and unemployment, we have some concept of a traditional male worker desiring a full time job. However this does not apply to many women (and increasingly to many men, even if we exclude the obviously problematic categories of students and the retired). It is easy to ignore women who do not conform to the implicit cultural definition that we impose upon labour force behaviour. But the same problem applies to Maori males. Proportionally fewer of their adults are employed, but also fewer are in the labour force (employed and unemployed). One suspects that a higher proportion have despaired at finding a job, and so are NILF (although if one were offered they may well accept it with alacrity). (1)

The options for persons on the unemployment benefit illustrates the potential confusion. They may be in any one of the three categories.
* They may work a few hours a week (as they are encouraged to by the Department of Social Welfare) in which case they are defined as employed;
* They may not have any work but be actively seeking, so they are defined as unemployed.
* But it is also recognized that some unemployed beneficiaries are so unlikely to get work (because of their or local circumstances) so they need not actively seek work (although they will be required to take it if it arise), and so they will be NILF.

There are four broad means of measuring unemployment in New Zealand. In historical order they are:

Census Unemployment: Since 1895 (1926 for Maori) census respondents have been asked to classify their labour market status. It is a self response, and so we are not sure how individuals report themselves, although in principle it should be close to the official unemployment definition. (A further data limitation is that population censuses occur only quinquennially.)

Unemployed Beneficiaries In principle the number of unemployed beneficiaries ought to give some indication of the state of the labour market, but as already been mentioned the benefit definition does not match to any standard definition of unemployment. In recent years administrative changes (as when the stand-down period was changed to affect entitlement), while some of those we might call unemployed have shifted to other better paying benefits (domestic purposes, invalids, and sickness).

Registered Unemployment: People may register with the labour force as a means of seeking work. This data series is our longest and is monthly. Unfortunately it is not very reliable, because it is subject to administrative definitions, while it has not always been up to date (so it contained individuals who were employed but whose registration had not lapsed). Moreover not everyone who is unemployed registered. (Between 1976 and 1981, the proportion of unemployed who registered increased dramatically from about a fifth to four-fifths, because registration was necessary to get work program jobs. Thus the registered unemployment series seriously mislead us to the actual state of unemployment in the period. The data has been misleading for other periods too. (2))

The last two definitions raise the important issue that while the government often collects data as a part of its administration, such data is at best affected by the administrative purpose for which it is used and at worst may be manipulated by politicians to give some favourable outcome. for these reasons social scientists should always be cautious when using administratively derived data as general indicators, If the data is sufficiently politically important a sort of Heisenberg reaction occurs, whereby the politicians will manipulate the administrative rules in a way which contaminates the usefulness of the data.

Surveyed Unemployment Since 1986 the Statistics New Zealand has surveyed quarterly (and monthly for a period until the Government cut their funds), a sample of New Zealanders asking them specific questions about their labour force status. The statistical results of the Household Labour Force Survey are internationally comparable, and can also be used to calculate other measures (such as those who would work more hours if offered, which covers those who are employed for too few hours, and those who are not actively seeking work). There was a notable divergence between the registered and surveyed unemployment figures after 1987, indicating how unreliable the registered figures are (although the latter are still used publicly because they come out monthly rather than quarterly).

1. B.H. Easton, Maori in the Labour Force (1995).
2. B.H. Easton “Unemployment: 1976-1981” (1981).