4  Who are the low wage workers? (continued)

The picture presented above is based on people who work full-time. In the English-speaking world, there has been a substantial growth in part-time employment over the past two decades. Many of these part-time jobs pay low wages. In studies of low wage workers in Australia, Richardson (1998) and Richardson and Harding (1999) examine all workers. This work concludes that:

• the profile of low wage workers is sensitive to the level at which the low wage is set (the higher the level, the more the low wage group looks like the general workforce)
• the risks of being in low wage work are higher for people who are female, employed part-time, aged 21-24, with low education, married if a woman and single if a man, and for non-student children living at home
• the typical low wage worker is female, employed full-time, of prime working age, married and with little formal education: about one third have dependent children
• nearly all low wage men work full-time and 30% have dependent children
• one-quarter have a post-school certificate or diploma and over half are married
• they are most likely found in sales, personal service or labouring jobs, employed in wholesale and retail industry
• 27% lived at home with a parent
• the lower the wage, the more likely were the workers to be female, of prime age, employed part-time and married with dependent children.

Richardson and Harding (1999: 153) conclude as follows.

Low-wage workers are not predominantly the sons and daughters of the affluent middle class, working a few hours each week to finance their holiday in Bali while studying law and business. Nor are they predominantly hard-pressed heads of families struggling to put food on the table to feed their families. There are some each of these groups, but they are relatively small in number. The typical low-wage worker works full-time, is of prime age, with no formal education qualifications, probably married and equally likely to be male as to be female. One-third have dependent children. The men are more likely than the women to be found in low income families.

Bernstein and Hartmann (1999:29, 39) provide a very useful summary not only of the current picture for the US, but also how it has been changing. They conclude:

• Low-wage workers are disproportionately female, minority, non-college-educated, nonunion, and concentrated in retail trade.
• These characteristics notwithstanding, the low-wage workforce is becoming more male and more highly educated, which is to be expected given widespread educational upgrading and the long-term wage decline among non–college graduates.
• The likelihood of being a low-wage worker has increased, even when the wage impacts of changes in education, experience, occupation, and industry are taken into account.
• Rising education and experience levels and occupational upgrading have combined to prevent the share of female workers in low-wage jobs from rising. This has not been the case for men, even though their total share of the low-wage workforce is still below that of women.
• Like the rest of the workforce, the low-wage sector included more minorities and became older, more highly educated, and less likely to work in the manufacturing industry.
• Unlike the rest of the workforce, however, the low-wage sector included less women.
• Women made up an additional 4.3 percentage points of the total workforce, while their share in the low-wage group fell 9 percentage points.
• The “high school or less education” category declined by 13.5%.

The profiles contained in both summaries confirm that not all people who are employed in low wage jobs are young, single and have low education.

Young people are, however, more likely to be in low wage jobs than are older workers. This is entirely to be expected, as young people are making the transition from student/child to independent working adult. Indeed, if all low wage workers were young (say, under the age of 25), then there would be little reason to worry about low wage jobs and every reason to believe that they represented a widely used means into higher paying jobs. By necessity, if all low wage workers are young, then duration in a low wage job is temporary.

Of course, the proportion of low wage workers who are young will vary depending on the definition of low wage and of young. Thus the picture given by different studies varies in ways that do not necessarily reflect real differences. Using US data for the first half of the 1990s, Long (1999) finds that as many as 46% of the workers earning the federal minimum wage are teenagers (aged 15-19), with only 6% aged 55 or more years. Similar percentages are reported by Card and Krueger (1995) and by Smith and Vavricheck (1992). When a more generous measure of low wage is used (two thirds of the median) and only full-time workers are considered, Table 2 shows that only 22% of US low wage workers are aged under 25.

To illustrate the sensitivity of the characteristics of low wage workers to the level of low wage selected, Table 4 compares the profile of US workers who are low wage according to the OECD definition (and work full-time) with the profile of minimum wage workers.

Source: For minimum wages, Long 1999:496. For two-thirds earnings, Keese, Puymoyen and Swain, 1998:230

Compared with other OECD countries, the value of the US minimum wage is quite low. With this in mind, Table 4 provides a clear account of how the characteristics of low wage workers vary, depending on how low wage is defined. When the wage is set very low – as for the US minimum – then the low wage population is more female, much younger and considerably less educated than when a higher value of low wage is used. This US picture is consistent with that found for Australia (Richardson, 1998), where sensitivity of the low wage population to various levels of the low wage was explicitly examined. The idea that low wage workers are mostly teenagers who live at home, and who rapidly move to higher paying jobs (and therefore whose low wages are unproblematic) has come mainly from US research that is based on the minimum wage population.

As with the Australian and US data reported above, a number of publications provide a more nuanced picture of the characteristics of low wage workers.

Entry jobs for people without higher education are typically low paid. In addition to youth, such jobs are taken by a) people who are unemployed, and b) mothers returning to the workforce.

Gregg and Wadsworth (1998), using British evidence, find that a high proportion of people in entry jobs were previously unemployed: a quarter had had an unemployment spell of more than two years, and another quarter had been less than three months out of work. Gosling et al (1997)found that, in the UK during the early 1990s, 56% of people who moved from unemployment to a job had entry wages that were in the bottom fifth of the wage distribution. We will look later at interesting new evidence on the cycling between unemployment and low wage employment.

Mothers returning to the workforce have similar characteristics to displaced workers, with the added difficulties that their time out of the labour force is often longer and they often return to work in part-time jobs with little or no chances for promotion. The obstacles facing mothers of young children, especially sole mothers, tend to make upward mobility especially difficult. Perhaps for this reason, Kim (2000) finds that in the US low wage women tend to be never married (almost half of the women who have never married are low-paid), and have children: having three or more children under the age of six makes the risk of low pay very high. The recent changes to the US social welfare system, that prevent people remaining on social welfare payments for more than two continuous years, are likely to magnify the results reported by Kim. The sole mothers being pushed into employment are overwhelmingly going into low wage jobs. It will be of great interest to monitor what proportion are able to move on to better paid jobs after a spell. Initial indications are that there is some upward mobility in earnings, but it is modest. Savner et al (2002) report that “--in most states, earnings in the fourth quarter after exit [from the welfare system] grew by only a few hundred dollars when compared to earnings in the first quarter.” (p 3). A longer term study of welfare leavers in Wisconsin found that over three years earnings on average rose from $8,608 to $10,924, “ – still well below the US poverty line.” (p3)

In addition to new entrants to the labour force, some people end up in low wage jobs as a result of losing a higher paying job. There is consistent evidence that workers who lose their jobs suffer both an immediate and a longer term loss in pay, even when re-employed soon after they lose their jobs. (Kletzer 1998, Gregory and Jukes 1998, Fallick 1996). Furthermore, their wage rate remains well below that of the previous job for a number of years, if not permanently. Jacobson, LaLonde and Sullivan (1993) estimate that for their sample of US displaced workers, total long-term wage losses were on average 25% of initial earnings per year and persisted even six years after separation. Jacobson et al argue that these earnings losses are not due to any individual characteristics of displaced workers, since their original median age, skills, tenure and earnings growth were similar to those of staying workers. They find no indication that displaced workers” earnings ever return to their prior expected levels. Thus displaced workers’ earnings losses may be permanent.

While the evidence of wage decline for displaced workers is strong, the research appears to be silent on what proportion of displaced workers end up in low-wage jobs as defined for this paper. This is a relatively new area of research and more is yet to be known about displaced workers. More longitudinal data are needed to further our understanding of the employment dynamics of displaced workers. Theory, with some empirical support, leads us to expect that workers suffering the greatest wage fall on displacement would be those who had a high level of human capital that was specific to their employer and/or to an occupation or industry that is in decline; had a high quality job match with their last employer; had a high-paying last employer; and had a seniority component in their pay.

The characteristics of low paid workers reported above have focussed on one feature at a time. It may well be, however, that sole mothers are likely to have low wages not because they are sole mothers, but because they are female and have low levels of education, for example. Some studies have sought to identify the impact on the probability of having a low wage of a specific attribute, holding other attributes constant.

One consistent finding of such studies is that women are significantly more at risk of low wage work than are men, other things equal (Dunlop, 2000; Eardley, 1998; Asplund and Persson, 2000). So too are youth. If young people are excluded, Dunlop (2000:16) finds no relation between age and the probability of being in a low paid job. Multivariate analysis supports the earlier conclusion, that never married people are more likely to be low paid than are those who are or have been married (Dunlop, 2000; Stewart and Swaffield, 1997). For migrant countries such as Australia, the US and Canada, there is strong evidence from the migration literature that limited capacity in English is associated with low wages (and unemployment). Geography matters too. Dunlop estimates that Australians living outside the metropolitan areas were 6 percentage points more likely to be low paid than their metropolitan counterparts (though this comparison makes no allowance for differences in the cost of living).