Sidelined in the labour market
While the unemployment rate is an important indicator of the state of the economy, it is only one piece of the puzzle. Another unemployment statistic, the duration of job search, is an essential indicator of economic well-being.1 It is important to distinguish between long-term unemployment and medium- and short-term unemployment. While the latter two are associated with normal labour turnover, long-term unemployment is related to structural rigidities in the labour market.
Long-term unemployment has always garnered attention because of its high costs and pernicious nature. In most industrialized countries, a negative relationship exists between the duration of unemployment and the probability of returning to work (see Long-term unemployment internationally). On a personal level, long-term unemployment is associated with the loss of present and future opportunities, financial problems, social exclusion, loss of self-esteem, and health problems. In economic terms, it leads to a decrease in tax revenues, lessened productivity because of loss of acquired skills, and an increase in the costs of social and health care programs. In fact, the very efficiency of the labour market is adversely affected by high levels of long-term unemployment because of the structural adjustment costs it entails.2
This article seeks to shed light on long-term unemployment in Canada for the period 1976 to 2003 (see Data source and definitions). It looks at how the incidence of long-term unemployment (the long-term unemployed as a percentage of all unemployed)3 has changed over time. Next, it identifies the most affected groups, since total time unemployed is not distributed uniformly (see Are the long-term unemployed different?).
Recession and long-term unemployment
In 1976, following the end of the 1975 recession, of the 738,000 persons experiencing a spell of unemployment, 29,000 were unemployed for 12 months or more, representing a long-term unemployment incidence of 3.9%. This increased gradually until the beginning of the 1980s, accelerating with the 1981-82 recession. By 1985, nearly 165,000 persons were unemployed for a year or more, an incidence of 11.9%. As the job recovery gathered steam, the incidence gradually declined to around 7% in 1990 (81,000 persons). Following the recession of the early 1990s, it rebounded sharply, reaching a new peak of 17.3% (nearly 263,000) in 1994. Remaining high for much of the 1990s, the incidence fell substantially starting in 1998. In 2003, 9.7% of unemployed persons, or 126,000, were on long-term unemployment. Despite a fairly comparable overall unemployment rate (approximately 7.5%), the incidence of long-term unemployment in 2003 was 39% higher than in 1990, and more than double (+120%) that in 1977. The question arises whether certain cyclical factors may have raised the 'equilibrium' level of long-term unemployment—a phenomenon labour economists call the hysteresis effect.4
Much of the variation in long-term unemployment appears related to cyclical fluctuations in the economy (Chart). The overall unemployment rate and long-term unemployment are strongly correlated (Wong, Henson and Roy 1999), but with a lag between a rise in the unemployment rate and an increase in long-term unemployment. Similarly, long-term unemployment generally remains high for several years during economic recoveries, even though the unemployment rate rapidly adjusts downward. For example, after the recession of the early 1990s, Canada's unemployment rate peaked in 1993 (11.4%), whereas the highest incidences of long-term unemployment were observed in 1994 (17.3%) and in 1996 (16.3%). This suggests that the last workers laid off are generally the first to return to work when the economic situation improves. By contrast, persons who have been unemployed for some time, along with less skilled workers, tend to represent a larger proportion of the unemployed population.
In considering the duration of unemployment, differentiating between cyclical and structural causes is generally difficult. The model most often used by labour economists assumes that once individuals become unemployed, the duration of unemployment will depend on the probability of their receiving and accepting a job offer. The probability of receiving a job offer is determined by factors such as education or work experience (structural aspects of the labour supply) and the economic context in which the jobseeker is operating (cyclical aspect of labour demand). Similarly, the probability of accepting the offer is determined by the expected wage, that is, the lowest wage package (including benefits and working conditions) for which the person is willing to work, which in turn depends on personal characteristics and economic conditions.
Structural causes of long-term unemployment are many and varied. These may include industrial restructurings and reorganizations that arise from trade liberalization, low labour mobility, regional disparities, and skill obsolescence resulting from technological change. Furthermore, long-term unemployment may also be influenced by organizational and institutional policy changes affecting wage flexibility. For example, cutbacks in provincial social assistance during the 1990s encouraged recipients to look for work. These jobless persons then saw themselves as unemployed rather than as not in the labour force (Bédard, Bertrand and Grignon 2001).
Some are harder hit
Although strong increases in long-term unemployment resulted from the recessions of the early 1980s and 1990s, some groups and regions were hit harder than others.
For more than 20 years, unemployed men have had a considerably higher incidence of long-term unemployment than women (Table 1). This gap has continued despite the growing presence of women in the labour force. In 2003, the incidence of long-term unemployment for men was 11% compared with 8% for women, a gap of almost 40%.
The gap may be due in part to the greater participation of men in the labour market, but it may also be due to differences in industry and the type of work. For example, labour turnover is greater for women than for men (Blau, Ferber and Winkler 2002), and women are more heavily represented in services and in part-time work, both characterized by higher turnover.
Older unemployed persons (45 and over) consistently posted the highest incidence of long-term unemployment. That incidence was 17% in 2003, compared with 10% for persons 25 to 44, and 3% for those 15 to 24 (Table 2). These figures indicate a positive relationship between age and the risk of long-term unemployment—the opposite of the relationship between age and risk of being unemployed, as expressed by the unemployment rate. In other words, the probability of job loss appears to be lower among older workers, but once unemployed, they seem to have greater difficulty finding work.
The higher incidence of long-term unemployment among older persons may be explained by a number of factors, including lower mobility (related to higher relocation costs), a lower education level than among those aged 25 to 44, a lower capacity for job-hunting, and a certain amount of discrimination against them (HRDC 1997; Hutchens 1988). Also, it is generally harder to find a new position after having had the same job for a number of years and accumulated non-transferable skills. Furthermore, since they have more occupational experience and higher net worth, they may be more selective—which lengthens their job-search period. They may also involuntarily withdraw from the labour force, often through early retirement, which amounts to hidden unemployment. Hence, long-term unemployment among older workers may be underestimated.
On the other hand, the lower incidence of long-term unemployment among younger persons may be related to their high turnover on the labour market. They may be more inclined to accept jobs that are part-time, unstable or less well-paying, or to go back to school after an unsuccessful job search. However, even though they are proportionally less affected by long-term unemployment, they may experience its consequences more acutely. For example, many have no real experience related to their training, have very few ties to the labour market, and have not accumulated the hours needed to be eligible for Employment Insurance. Moreover, since they have the lowest net worth, they would likely be more vulnerable when faced with a prolonged absence of income.
The gap in the incidence of long-term unemployment between older and younger persons has widened over the past two decades. A comparison of 1980 and 2003 shows that the incidence of long-term unemployment remained relatively stable (3%) for those aged 15 to 24 while almost doubling for those aged 45 and over, rising from 9% to 17%. The growth of the 45-and-over unemployed group in the labour force may have resulted in increased competition among jobseekers in that group. On the other hand, unemployed persons aged 45 and over in 2003 may differ from their 1980 counterparts. For example, they may have socioeconomic characteristics that enable them to be more selective about the jobs available—such as more accumulated wealth or belonging more frequently to a two-income family.
The less educated
Unemployed persons with a low level of education generally have a higher incidence of long-term unemployment than other groups (Table 3). In 2003, those with less than grade 9 had an incidence of nearly 16%, compared with 9% for those with between grade 9 and university, and 12% for those with a university degree. This is consistent with the unemployment rate, indicating that education has a positive influence on the search for work.
However, the relationship between the incidence of long-term unemployment and education is not completely linear. For example, in 2003, those in the highest education level (university degree) had a higher incidence of long-term unemployment than those at the intermediate education level (between grade 9 and university degree). This may reflect their aversion to jobs that do not interest them. They may try harder to obtain the job (and wage) they are looking for, even if it means a longer search. The least educated face greater job instability. They would therefore be more likely to accept whatever jobs are available, even ones that are part-time, temporary or poorly paid.
Quebec and British Columbia
The incidence of long-term unemployment varies greatly by region, from 13% in British Columbia to 4% in the Prairies (Table 4). The ranking is similar to that for regional unemployment rates, except for the Atlantic region, which had the highest unemployment rate in 2003. This is not surprising, given the importance of seasonal unemployment, which is of short or medium duration.
Quebec was hardest hit by long-term unemployment, followed by British Columbia. British Columbia came out of the recession of the early 1990s in better shape than the other regions; in 1994, it posted the lowest incidence of long-term unemployment (12%), compared with Ontario's nearly 20%.
Long-term unemployment affected less than 4% of all unemployed persons in 1976, but grew substantially during the recessions of the early 1980s and 1990s. It reached a peak in 1994, when more than one unemployed person in six (17%) was affected. Despite a significant drop since then, the incidence still stood at nearly 10% in 2003. Men, older workers, persons with less education, and those residing in Quebec and British Columbia exhibited higher rates than other groups.
Data source and definitions
The monthly Labour Force Survey is the source for this study. Persons unemployed at the time of the survey are asked how many weeks they have been actively looking for work.
The duration of unemployment is an uninterrupted period during which the person was unemployed. This concept does not measure time spent not working (which includes periods when the respondent was not part of the labour force). In addition, because it includes only spells of unemployment that continue up to the time of the survey, it is not a complete measure of the duration of unemployment. The duration of unemployment is a lagging indicator (or a lagging cyclical indicator).
The unemployment figures contained in this article do not include persons who were not looking for work because they had a job that was to begin at a later date. Persons not looking for work are not asked about the duration of job search.
The labour force is the civilian population aged 15 and over (excluding institutional residents) who, during the survey's reference week, were employed or unemployed.
The unemployed are persons who, during the reference week, were available for work and had been laid off temporarily, had looked for work during the past four weeks, or were to start a job during the next four weeks.
For this article, short-term unemployment is 26 consecutive weeks or less. Since unemployed persons whose duration of unemployment is unknown are those who were not looking for work because of a job that they were to start at a later date, it is probable that the incidence of short-term unemployment is slightly underestimated. Medium-term unemployment is more than 26 but less than 52 weeks. Long-term unemployment is 52 weeks or more.
The incidence of long-term (short-term, medium-term) is the proportion of unemployed persons on long-term (short-term, medium-term) unemployment in relation to all unemployed persons.
The unemployment rate is the number of unemployed persons in a group, expressed as a percentage of the persons in the labour force within that group.
The duration of unemployment is the number of consecutive weeks during which a person has been temporarily laid off, or has been without work and is looking for work.
Structural unemployment refers to the situation in which workers cannot occupy the positions available because they do not have the desired skills, do not live where the positions are offered, or are not willing to work at the market wage.
Discouraged workers are jobless persons who want to work but do not look for work because, for various reasons, they do not believe that they can find a satisfactory job. Since these individuals are not actively looking for work, they are not included among the unemployed.
Long-term unemployment internationally
The incidence of long-term unemployment varies considerably from one country to another. The incidence is generally much lower in North America than in most industrialized countries. Among the G-7 countries, for example, Canada ranked second in 2002, just behind the United States (8.5%). Among the 30 OECD countries, Canada ranked fifth after Mexico (first) and the United States (fourth) (OECD 2003).
By definition, the incidence of long-term unemployment is based on the time spent unemployed. The greater the labour turnover in a given country, the larger the proportion of short spells of unemployment and the lower the incidence of long-term unemployment. Since North American labour turnover rates are among the highest in the world, it is not surprising that incidences are among the lowest. However, a low incidence can also mask another, almost identical phenomenon: Longer episodes of unemployment may be replaced by a greater number of shorter episodes. When all the unemployment spells experienced by one person over the course of a given year are added up, the total duration of unemployment may be similar to that of a person on long-term unemployment.
In addition, the large gaps in incidence between countries may be due, in part, to differences in economic cycles. However, a higher incidence does not result solely from an increase in overall unemployment caused by difficult economic conditions. This is especially apparent when Canada's unemployment rate (7.7%) is compared with that of the United Kingdom (5.1%) in 2002. Thus, the differences observed from one country to another are longstanding and do not appear to be due to either disparities or changes in unemployment rates (OECD 1987). On the other hand, differences in institutional policies affect the observed disparities. Some aspects of national employment insurance programs or the presence of specific measures to combat long-term unemployment (for example, the use of wage subsidies) are most often cited in this regard. Table.
Are the long-term unemployed different?
Nearly four unemployed persons in five (79.1%) were short-term unemployed in 2003. A high number of temporary layoffs and a high level of seasonal unemployment in some sectors were probably major factors. Because of its magnitude, short-term unemployment has characteristics that most closely resemble those of overall unemployment: a higher incidence among women (80.5%), younger workers (89.2%), persons whose education level lies between grade 9 and a university degree (80.7%), and residents of the Prairies (85.1%).
Medium-term unemployment was the least frequent, accounting for 7.0% of all unemployed in 2003. Overall, the incidence of medium-term unemployment is higher for men (7.3%), older persons (9.2%), persons with a university degree (9.4%), and Ontario (7.6%). In general, the medium-term unemployed appear to have more in common with the long-term unemployed than with the short-term unemployed. However, some differences between the two are evident, notably education. This may be because the most educated are more selective in the medium term in their job search, partly because they have higher wage expectations and also because they generally consider themselves more likely to receive a better job offer in the future. Table.
- Of interest in this regard is another indicator published by Statistics Canada: the average duration of unemployment. However, this indicator says nothing about how the duration of unemployment is distributed. Yet, for a given average duration of unemployment, it makes considerable difference whether all workers were unemployed for one month in a year or only one-twelfth of workers were unemployed for the entire year.
- For a thorough review of the consequences of long-term unemployment, see OECD (1993), chapter 3.
- The incidence of long-term unemployment is not a function of the unemployment rate of the group. For example, a group may have a high unemployment rate but a low incidence of long-term unemployment. This would indicate that while the members of this group have a strong probability of being unemployed, the probability that they will remain unemployed for a year or more is low. The long-term unemployment rate, which would be the probability of members of the group (both working and unemployed) being on long-term unemployment, is not dealt with in this article.
- Simply put: An increase in unemployment generally has the effect of increasing the proportion of persons on long-term unemployment. As these persons remain unemployed, they gradually become sidelined in the labour market. They then have a diminishing influence on the wage-setting process. As a result, wages remain high. All else being equal, this situation represents an impediment to job creation and thereby contributes to a further worsening of the overall unemployment situation.
- Bédard, Marcel, Jean-Francois Bertrand and Louis Grignon. 2001. The unemployed without recent employment. Strategic Policy Series, Working Paper no. W-00-4E. Applied Research Branch, Human Resources Development Canada. Ottawa.
- Blau, Francine D., Marianne A. Ferber and Anne E. Winkler. 2002. The economics of women, men and work. Fourth edition. Upper Saddle River, N.J.: Prentice Hall.
- Human Resources Development Canada. 1997. Lessons learned: A review of older worker adjustment programs. Technical report no. 1. Evaluation and Data Development, Human Resources Development Canada.
- Hutchens, Robert M. 1988. "Do job opportunities decline with age." Industrial and Labor Relations Review 42, no. 1 (October): 89-99.
- Organisation for Economic Co-operation and Development (OECD). 2003. OECD Employment Outlook. Paris.
- ---. 1993. Employment Outlook. Paris.
- ---. 1987. Employment Outlook. Paris.
- Wong, Ging, Harold Henson and Arun Roy. 1999. Long-term unemployment, worker profiling and program evaluation issues. Evaluation and Data Development, Human Resources Development Canada.
Full article in PDF
Vincent Dubé is with the Transportation Division. He can be reached at (613) 951-7031 or firstname.lastname@example.org.
You need to use the free Adobe Reader to view PDF documents. To view (open) these files, simply click on the link. To download (save) them, right-click on the link. Note that if you are using Internet Explorer or AOL, PDF documents sometimes do not open properly. See Troubleshooting PDFs. PDF documents may not be accessible by some devices. For more information, visit the Adobe website or contact us for assistance.