1350.0 - Australian Economic Indicators, 1994  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 01/06/1994   
   Page tools: Print Print Page

1994 Feature Article - The Dynamics of Long-Term Unemployment
This article was published in Australian Economic Indicators June 1994 issue on 1 June 1994.


Introduction

This article has been extracted from Australia’s Long-Term Unemployed - A Statistical Profile (cat. no. 6255.0), and discusses the dynamics of long-term unemployment. The article uses published and unpublished ABS data, principally from the monthly ABS Labour Force Survey and associated supplementary surveys.

The last two decades have seen the emergence and growth of long-term unemployment as a persistent labour market phenomenon. This has resulted in an increasing number of Australians being directly affected by long-term unemployment or being members of households with one or more long-term unemployed. The difficulty experienced by the long-term unemployed in obtaining employment has been referred to as “the long-term unemployment trap”.

At the macroeconomic level, it has been argued that continuing high levels of unemployment make macro-economic management more difficult, making it less likely for the economy to deliver low levels of inflation and unemployment in the future. This arises essentially because a growing proportion of the unemployment pool becomes less relevant to employers and find it progressively more and more difficult to re-enter the workforce (Chapman 1993). The lower escape rate from unemployment of the long-term unemployed can result from a loss of skills and on the job training, a reduced intensity of job search and a reluctance of employers to hire the long-term unemployed.

For an individual, the personal and social costs of long-term unemployment are significant. They arise because of the centrality of work to people’s lives. Employment is the main source of income and thus of material welfare for most people. In addition, employment provides people with a sense of identity, participation and order to daily life (McClelland 1993).

Measurement of Unemployment

The Australian Bureau of Statistics (ABS) conducts a monthly population survey which provides official estimates of unemployment and is the major source of statistics discussed in this article. In line with the recommendations of the International Labour Organisation, the ABS classifies a person as unemployed where they are:

  • without work during the reference week;
  • currently available for work;
  • actively looking for work.
    A person who has been unemployed for 52 weeks or more is defined as long-term unemployed. Current duration of unemployment is the period from the time a person began looking for work or was laid off to the end of the reference week. Thus the survey measures unemployment duration based on current (and continuing) periods of unemployment rather than completed spells (footnote 1).

    The following summary measures are used in this article.
    • The unemployment rate; for any group, the number of unemployed persons expressed as a percentage of the labour force in the same group.
    • The incidence of long-term unemployment for any group, the proportion of unemployed persons who are long-term unemployed. This term is used by the OECD in Employment Outlook.

    In March 1994 the trend estimate of the unemployment rate was 10.4 per cent, which was the sixth successive monthly fall in this series. There were 351,000 people in long-term unemployment, which represented 37.0 per cent of the total number of unemployed persons.

    Inflows to and outflows from unemployment

    The number of unemployed persons at any given time will depend directly on the number of people entering the unemployment pool (inflows to unemployment), and the number of people leaving the pool (outflows from unemployment). The relationship between movements in the number of unemployed persons, and inflows to and outflows from unemployment can be expressed as:

    UT = UT-1 + lT -OT

    where UT is the number of unemployed persons in month T, lT is the inflow to unemployment between months T-1 and T, and OT is the outflow from unemployment between months T-1 and T.

    The long-term unemployed in month T are a subset of the unemployed in month T-1 . The incidence of long-term unemployment (i.e. the proportion of unemployed persons who are long-term unemployed) will depend on the relative magnitude of the inflows and outflows and the duration of unemployment of the outflows.

    Business cycle and long-term unemployment

    Initially during an economic downturn, the increased inflow of new entrants into unemployment will cause the number of short-term unemployed to rise at a faster rate than the number of long-term unemployed and hence reduce the incidence of long-term Unemployment. Between 1989-90 and 1990-91 there was a rapid inflow into unemployment and a sharp rise in the unemployment rate from 6.2 per cent to 8.4 per cent. However, during this period the incidence of long-term unemployment declined from 22.6 per cent to 21.1 percent

    As an economic downturn proceeds, even though the unemployment inflows may remain high, a proportion of these new entrants will move into long-term unemployment, causing an increase in the incidence of long term unemployment. In 1991-92 the incidence of long-term unemployment rose sharply to 29.1 per cent, as a high proportion of the 1990-91 new entrants to unemployment moved into long-term unemployment.

    The impact of the unemployment outflows on the incidence of long-term unemployment will depend on two opposing influences. During an economic downturn, outflows from unemployment to employment will decline, probably more significantly for the long-term unemployed, which will act to increase the incidence of long-term unemployment. However, outflows from unemployment to not in the labour force will generally rise as unemployed jobseekers become discouraged. These outflows will probably be greater for the long-term unemployed and will act to reduce the incidence of long-term unemployment.

    During an economic recovery, the incidence of long-term unemployment will initially rise due to a combination of a reduced inflow of new entrants into unemployment and higher outflows from unemployment to employment, mainly from the short-term unemployed. Between 1983-84 and 1984-85 the unemployment rate fell substantially from 9.6 per cent to 8.6 per cent but the incidence of long-term unemployment continued to rise from 29.8 per cent to 31.1 percent.

    Only as an economic recovery proceeds are outflows to employment likely to make inroads into the long-term unemployed. Inflows to unemployment from not in the labour force will rise with the re-entry of discouraged jobseekers, which will act to reduce the incidence of long-term unemployment. In 1985-86, the incidence of long-term unemployment fell to 29.1 per cent, as an increasing number of the long-term unemployed found employment.
    GRAPH 1. UNEMPLOYMENT RATE AND INCIDENCE OF LONG-TERM UNEMPLOYMENT: ANNUAL AVERAGE
    YEAR ENDED JUNE 1981 TO YEAR ENDED JUNE 1993
    Graph 1 shows the unemployment rate and incidence of long-term unemployment as an annual average from the year ended June 1981 to the year ended June 1993.


    Persistence of long-term unemployment

    It has been argued that when the unemployment rate falls during an economic recovery, the incidence of long-term unemployment does not fall to the level prior to the downturn (OECD, 1987). In Australia, this effect was evident after the 1982-83 recession with the incidence of long-term unemployment not returning to the 1980-81 level. An important issue, especially for policy makers, is assessing the major factors impacting on the lower escape rate from unemployment of the long-term unemployed.

    These factors are often subdivided by labour market analysts into two broad categories. The first category, termed “duration dependence”, includes those factors relating to duration of unemployment, such as:
    • a reduction in the effectiveness and intensity of job search which can be caused by demoralisation or loss of contact with the world of paid work;
    • loss of skills and on the job training; and
    • employer screening (i.e. the reluctance of employers to hire long-term unemployed people).

    The second category, termed “heterogeneity”, includes those factors relating to the specific characteristics of people. Those people remaining unemployed have an inherently lower probability of finding employment, because of personal characteristics. The relative influence of duration dependence and heterogeneity on the long-term unemployed is not readily separated. People with inherently lower probabilities of finding employment will also tend to suffer from duration dependence.

    Exit to employment

    The probability of finding employment for various groups of unemployed people provides some insight into how personal characteristics affect unemployment duration and hence indicate the type of people most susceptible to the long-term unemployment trap. One method of deriving these probabilities is to use the proportion of unemployed people in a particular month who had found employment by the following month (footnote 2).

    Employment probability and duration

    The average monthly percentage of unemployed people gaining employment during the year ended June 1993 generally decreased as the duration of unemployment increased. This implies that the longer the period of time spent in unemployment, the lower a person’s chance of gaining employment. The decreases in the percentage of unemployed people gaining employment were greater between the two shorter durations of unemployment but these decreases tapered off between the longer durations of unemployment. Unemployed people with the shortest duration of unemployment of less than 13 weeks had more than four times the chance of gaining employment (24.9 per cent) as unemployed people with the longest duration of unemployment of 156 weeks and over (6.0 per cent).

    GRAPH 2. PERCENTAGE OF UNEMPLOYED PERSONS GAINING EMPLOYMENT BETWEEN SUCCESSIVE MONTHS:
    DURATION OF UNEMPLOYMENT, ANNUAL AVERAGE, YEAR ENDED JUNE 1993
    Graph 2 shows the percentage of unemployed persons gaining employment between successive months classified by the duration of unemployment in weeks, as an annual average for the year ended June 1993.



    Employment probability and age

    The long-term unemployed have significantly lower percentages gaining employment than the short-term unemployed, for both males and females across all age groups. The percentage of long-term unemployed males gaining employment generally decreased as age increased. In contrast, the percentages for short-term unemployed males were reasonably constant for those aged less than 55 years, but then decreased substantially for those aged 55 years and over.

    GRAPH 3. PERCENTAGE OF UNEMPLOYED MALES GAINING EMPLOYMENT BETWEEN SUCCESSIVE MONTHS:
    AGE, ANNUAL AVERAGE, YEAR ENDED JUNE 1993
    Graph 3 shows the percentage of unemployed males gaining employment between successive months, classified by age group, as an annual average for the year ended June 1993.


    While the percentages of short-term unemployed females gaining employment were reasonably constant across all age groups, the percentages for long-term unemployed females were reasonably constant for those aged less than 45 years, but then decreased substantially for those aged 45 years and over.

    GRAPH 4. PERCENTAGE OF UNEMPLOYED FEMALES GAINING EMPLOYMENT BETWEEN SUCCESSIVE MONTHS:
    AGE, ANNUAL AVERAGE, YEAR ENDED JUNE 1993
    Graph 4 shows the percentage of unemployed females gaining employment between successive months, classified by age group, as an annual average for the year ended June 1993.



    Employment probability and birthplace

    Both the short-term and long-term unemployed migrants from Non-English Speaking Background Countries had lower percentages gaining employment than people born in Australia and migrants from Main English Speaking Background Countries.

    While migrants from Main English Speaking Background Countries had slightly higher percentages of short-term unemployed people gaining employment than people born in Australia, they had slightly lower percentages of long-term unemployed people gaining employment.

    GRAPH 5. PERCENTAGE OF UNEMPLOYED PERSONS GAINING EMPLOYMENT BETWEEN SUCCESSIVE MONTHS:
    BIRTHPLACE, ANNUAL AVERAGE, YEAR ENDED JUNE 1993
    Graph 5 shows the percentage of unemployed persons gaining employment between successive months, classified by birthplace, as an annual average for the year ended June 1993.



    Employment probability and educational attainment

    The chances of short-term unemployed migrants from Main English Speaking Background Countries gaining employment may be better than people born in Australia because they have higher levels of educational attainment. However, long-term unemployed migrants from Main English Speaking Background Countries, may have less chance of gaining employment because they are generally much older than people born in Australia, which may outweigh the usual advantage associated with having higher levels of educational attainment.

    The percentage of short-term unemployed people gaining employment was much greater for people with post-school qualifications than for those without post-school qualifications. People with degrees had the highest percentage of short-term unemployed people gaining employment (35.7 per cent), while people who did not attend the highest level of secondary school had the lowest (19.5 per cent).

    Although there was little difference between the percentage of long-term unemployed people gaining employment for those with and without post-school qualifications, the chances of gaining employment varied substantially with the different levels of educational attainment. While people with degrees had the highest percentage of long-term unemployed gaining employment (15.3 per cent), people with trade qualifications (9.0 per cent) and certificates and diplomas (7.8 per cent) had much lower percentages of gaining employment than those who only had attended the highest level of secondary school (13.1 per cent).

    GRAPH 6. PERCENTAGE OF UNEMPLOYED PERSONS GAINING EMPLOYMENT BETWEEN SUCCESSIVE MONTHS:
    EDUCATIONAL ATTAINMENT, FEBRUARY 1993 TO MARCH 1993
    Graph 6 shows the percentage of unemployed persons gaining employment between successive months, classified by educational attainment, from February 1993 to March 1993.


    This apparent anomaly is due in part to the concentration of unemployed people with trade qualifications, certificates or diplomas in older age groups, which have lower than average percentages of long-term unemployed people gaining employment. Whereas those who have attended only the highest level of secondary school are concentrated in younger age groups, which have higher than average percentages of long-term unemployed people gaining employment. Indeed, 31.0 per cent of unemployed people with trade qualifications and 14.9 per cent of unemployed people with certificates or diploma were aged 45 to 69, compared with only 6.6 per cent of unemployed people who have attended only the highest level of secondary school.

    When the age structure of the unemployment groups by qualification are taken into account, the unemployed with a degree have a lower than average incidence of long term unemployment for both males and females. Unemployed males, with a trade qualification, have a lower than average incidence of long term unemployment but unemployed females with trade qualifications have a higher than average incidence of long term unemployment. For both males and females, those unemployed who attended highest level of secondary school available had a lower than average incidence of long term unemployment.

    Conclusion

    The article has illustrated the lagged effect that the business cycle has on the incidence of long-term unemployment and indicates that sustained economic growth can reduce the incidence of long-term unemployment. Further, characteristics such as age, birthplace and educational attainment can influence an individual’s susceptibility to the long term unemployment trap.

    A more detailed description of the long-term unemployed, in terms of their demographic and other characteristics, such as geographic location, family structure and previous industry and occupation can be found in “Australia’s Long-Term Unemployed - A Statistical Profile” (cat. no. 6255.0).


    This feature article was contributed by John Preston and Judy Harwood, Labour Statistics Analysis Unit, ABS.

    Footnotes

    1 . The definition of unemployment adopted by the ABS does not necessarily correspond to joblessness. Persons who were without work during the reference week and wanted to work, but did not satisfy the other conditions relating to unemployment, form part of a group of persons not in the labour force who are described as marginally attached to the labour force. This group includes discouraged job seekers. While those marginally attached to the labour force do not satisfy the official criteria to be classified as unemployed, they are experiencing joblessness. As a result, official estimates of long-term unemployment may understate the severity of the problem of persons experiencing prolonged periods without employment. < Back

    2. These probabilities are derived using estimates of flows between different labour market states from the monthly ABS Labour Force Survey. These are calculated by matching the characteristics of persons included in the survey from one month to the next. < Back

    References

    Chapman, B.J. 1993, “Long-Term Unemployment: The Dimensions of the Problem”, The Australian Economic Review, Melbourne, Second Quarter 1993.

    McClelland, A. 1993, “Long-Term Unemployment: Costs and Responses”, The Australian Economic Review, Melbourne, Second Quarter 1993.

    OECD (Organisation for Economic Co-operation and Development) OECD Employment Outlook 1987.