Who works beyond the 'standard' retirement age and why?

One of the challenges facing Australia's labour supply today is its ageing population. As observed in the Australian Government's 2010 intergenerational report, in the future there will be fewer workers to support retirees and young dependents. With life expectancy increasing, and with adjustments to public policy aimed at encouraging older workers to stay in the workforce, the decisions people make about retirement are changing. It is therefore useful to consider who works beyond the 'standard' retirement age—defined by the authors as 65 years—and why.

This report describes the characteristics of those who continue to work beyond the age of 65 and examines how the skill requirements of their jobs and their remuneration change over time.

Key messages

• Two groups of workers, in terms of their educational qualifications, are more likely to remain working beyond the age of 65 years—the most-educated and the least-educated.
• Workers with vocational skills tend to retire around the standard retirement age. After standard retirement age the labour force participation rate of those with vocational qualifications declines more than for workers with higher and lower educational levels.
• Educational attainment is a strong predictor of wage levels among older workers, more so than document literacy and literacy use.

The finding that the most-educated and the least-educated are the ones to remain working beyond age 65 years probably reflects two of the components that make work attractive— remuneration and job satisfaction. The most-educated tend to have more interesting jobs with high wage levels. By contrast, the least-educated have significantly lower lifetime incomes and may well be working to maintain living standards. The policy implications are that retirement income policies matter, as do policies such as flexible work arrangements that make work more satisfying.

Tom Karmel
Managing Director, NCVER

Australia will experience substantial demographic change over the coming decades as a consequence of low fertility rates in the recent past and increasing life expectancy. The extent to which these changes will affect economic and social conditions will depend, among other factors, on the ability of older workers to continue working. As a result of the increasing life expectancy and changes to public policies that extend the age at which public pensions are available, individuals may find they need to work to older ages than they had envisaged. Against this background, this study investigates which types of workers, in the last ten years, continued in employment beyond the 'standard' retirement age and explores some of the reasons why they might have been motivated to do so. The analysis uses data from the 1996 Survey of Aspects of Literacy (SAL) and the 2006 Adult Literacy and Life Skills (ALLS) Survey.

Although the definition of the 'standard' retirement age varies substantially across occupations and industries, our analysis compares workers younger than age 65 (observed in 1996) with workers who remain employed beyond the age of 65 (observed in 2006). This allows us to see which groups of workers remain employed in terms of their skill and education levels and how the skill requirements of their jobs and their remuneration change over time. The findings give us some sense of whether older workers remain in the labour force through choice or necessity and could be useful for informing policies that may be necessary to encourage labour force participation at older ages.

This study consists of two broad components. One component provides information about skills, education levels, remuneration and other relevant characteristics (such as occupation and size of employer) of workers who continue to work beyond the standard retirement age. The second component involves regression analysis, which allows us to examine the factors that are relevant in explaining labour force participation. Particular attention is paid to the skills of workers who work beyond the standard retirement age and the skills that are required in their jobs.

The comparison of the two adult literacy surveys permits an examination of changes over time for specific birth cohorts, allowing us to estimate ageing effects on their labour force participation. Both include information on skill requirements, employment status, individual literacy skills and earnings of people up to age 74 years. Although the two surveys are based on different samples and do not allow longitudinal analysis, they enable us to compare two cohorts of workers younger than age 65 in 1996 (aged 45–54 and 55–64) with a survey undertaken ten years later (where the cohorts are aged 55–64 and 65–74, respectively) to investigate the types of workers who remain employed in terms of skills and skill requirements at work, their education levels and remuneration and other factors.

Our analysis pays particular attention to the extent to which worker skills and the skill requirements in their jobs affect their retirement decisions. In addition, we investigate where in the 1996 wage distribution the workers who remained employed in 2006 came from (that is, whether they were predominantly from the top of the distribution or from lower down the distribution). Using estimates from a wage regression model, we can predict the wages of workers in 1996, given the characteristics of persons who work beyond the standard retirement age in 2006. Specifically, we may compare the resulting counterfactual distribution with the actual wage distribution in 1996 to draw inferences about the original position of workers in the wage distribution, given that they have the same characteristics as persons who work beyond age 65 years in 2006. The findings derived from this analysis provide information about the self-selection process of workers into employment beyond the 'standard' retirement age.

The major findings and their implications are highlighted in the points below:

Labour supply determinants:

• Labour force participation rates were around 80% for the population aged 15–54 years and substantially lower for older individuals. The overall labour force participation rate of 65 to 74-year-olds in 2006 was 14%.
• The increase in overall labour force participation between 1996 and 2006 may be attributed predominantly to an increase among women.
• A substantial part of the labour force population without post-school qualifications remains employed beyond age 65 years. The share of workers with vocational qualifications within the birth cohort 1932–41 declined between 1996 and 2006, while the proportion of employed persons with a university degree increased.
• Older workers appear to be relatively well matched to their jobs. Their skills are slightly below average and they tend to work in jobs with relatively low skill requirements.
• Self-assessed skills may matter more for retirement decisions than actual skills. This was especially apparent in females, as those who remain employed consider themselves as highskilled even though this might not actually be the case.
• Differences in individual skills and skill requirements over time and across birth cohorts are rather small.
• An increase in individual literacy increases the propensity to participate in the labour force, although the increase slows down at the higher literacy levels.
• After controlling for relevant characteristics, gender differences in labour force participation remain highly significant.
• Educational attainment is positively associated with labour force participation.

Wage determinants:

• Older workers are less likely to be full-time employed than younger workers, which affects
  weekly wages:
  • The share of male workers employed full-time in the birth cohorts 1932–41 declined from 87.0% in 1996 to 58.9% in 2006.
  • The corresponding share of female workers employed full-time dropped from 37.2% to 23.1% over the same period.
• Wages of highly educated workers who remain employed beyond age 65 years are higher than those of younger age groups.
• The regression results provide evidence for a strong relationship between education and wages. In addition, even after controlling for relevant observable characteristics, a gender wage gap of about 30% may be observed.

Self-selection of older workers into employment:

• Positive self-selection of full-time employed workers aged 55–64 years into full-time employment beyond age 65 is evident in the comparison of wage distributions from 1996 and 2006. Workers who continued in employment had higher wages.
• The most-educated and least-educated groups of workers are more likely to remain employed beyond age 65 years than other groups. Nevertheless, in general there is evidence of positive selection of workers beyond age 65 years into part-time employment as well as full-time employment.