This chapter examines trends in labour market participation and the transition from work to retirement. Labour market activity is important in the context of pensions because earnings from employment help people to build savings and to make private pension contributions to fund their retirement. Employment also builds entitlement to state pension provision. Questions about pressures on pension adequacy and whether people are extending their working lives in response to increasing life expectancy is one reason why the transition from work to retirement is of interest. The chapter begins by looking at how the economic support ratio, a measure of the economic burden on working people, might be affected by the ageing of the UK population. It then examines changes in labour force participation over time, considers labour force participation of older age groups and examines trends in the average age of withdrawal from the labour market.
Currently, the State Pension Age (SPA) for men is 65. SPA for women was 60 until 5 April 2010 but, under the Pensions Act 1995, the SPA for women began to increase on 6 April 2010 (see Pension Trends Chapter 1). The increases are being implemented in stages, with the exact age at which women become entitled to a state pension depending on their date of birth. Under the Pensions Act 1995, women’s SPA would have reached 65 (in line with men’s) by 2020. However, under the Pensions Act 2011, women’s SPA will increase more quickly: to 65 between April 2016 and November 2018.
Under the Pensions Act 2007, the SPA for both men and women would increase to 68 by 2046 (with increases to 66 between 2024 and 2026, to 67 between 2034 and 2036 and 68 between 2044 and 2046). However, the Pensions Act 2011 brings forward the increase from 65 to 66 to between December 2018 and October 2020.
The Government’s April 2011 Green Paper, “A state pension for the 21st century”, indicated that further changes to SPA might be proposed. In the Chancellors 2011 Autumn Statement and within the Government’s January 2013 White Paper, “The single-tier pension: a simple foundation for saving”1, it was announced that SPA would increase to 67 between 2026 and 2028. However, this change is not yet law and will require the approval of Parliament.
Rates of employment, unemployment, economic activity and inactivity can be presented for any group of people and are the proportion of that group who have that particular labour market status.
The labour market can be divided into two groups: the economically active and inactive. The economically active are defined as those aged 16 or over who are either in employment or unemployed. Economic inactivity refers to people who are neither ‘in employment’ nor ‘unemployed’, for example, people who are looking after a home or family, or who have retired or are long-term sick. The Office for National Statistics (ONS) Labour Force Survey categorises all economically inactive people over 69 years of age as retired.
According to the International Labour Organization (ILO) definition, people in employment may be employees, self-employed, on a government-supported training programme, or unpaid family workers (e.g. working without pay in a family-owned shop). The employment rate is the percentage of people in a given age group who are in one or more hours of paid employment a week.
The ILO definition of unemployment specifies that it applies to people who are without a job, have actively sought work in the last four weeks and are available to start work in the next two weeks, or to people who are out of work and have found a job which they will start in the next two weeks. The unemployment rate is the percentage of economically active people in a given age group who are unemployed.
Historically, ONS reported employment, unemployment and economic inactivity rates for the working age population, which was defined as age 16 to SPA. Until April 2010, SPA was 65 for men and 60 for women. However, SPA for women is now increasing, and it will increase for both sexes in future (see Definitions: Changes to State Pension Age).
ONS is reporting headline rates for employment and inactivity for the 16-64 age group. The headline unemployment rate is based on the economically active population aged 16 and over. The employment, unemployment and economic inactivity rates reported in this chapter are on the same basis as the ONS headline rates.
For the Chancellor’s 2011 Autumn Statement; for the April 2011 Green Paper, "A state pension for the 21st century"; for the January 2013 White Paper, “The single-tier pension: a simple foundation for saving”.
Pension Trends Chapter 2 discusses changes in the structure of the UK population. It shows that the old age dependency ratio, the number of people of State Pension Age (SPA) and over for every 1,000 people of working age, is projected to rise in the next few decades despite the planned increases to SPA. This is important because, in the UK, taxes paid by current workers fund the state pensions of current pensioners. If the old age dependency ratio rises, the system will have proportionately more pensioners depending on fewer workers.
However, the old age dependency ratio does not take into account the fact that working-age people who are not in employment and children below working age also receive state support funded by workers’ taxes. The economic support ratio, which is related to the old age dependency ratio, attempts to address this issue. The economic support ratio is found by dividing the estimate of the number of working people by the difference between the UK population and this estimate. In the economic support ratio, the population as a whole is considered and this enables the burden placed on workers to be better understood. Figure 4.1 shows that this ratio rose from 0.78 in 1972 to a peak of 0.93 in 2008, although there were dips in the recessions of the early 1980s and 1990s due to increases in unemployment.
Previous editions of Pension Trends have calculated that the economic support ratio for the UK would fall in the long term, reflecting the projected ‘ageing’ of the population (see Pension Trends Chapter 2). However, such projections are now subject to too much uncertainty to be presented here. There are several reasons for this, including the fact that further changes to SPA are not yet law (see Definitions: Changes to State Pension Age). In addition to this, there are sources of variability relating to long-term population projections and changes in employment and unemployment rates over economic cycles.
All other things being equal, a decline in the economic support ratio means a greater burden on the employed as there would be more people dependent on fewer workers. However, although the economic support ratio takes into account the number of working people in the population who are available to support others, it still has limitations. For instance, it gives equal weight to each person supported whereas, in reality, different age groups, children, working age people and those over SPA, attract different levels of government support in the form of benefits and tax credits. They also differ in the extent to which they draw upon government-funded services such as health and education. Moreover, many older people who are economically inactive care for grandchildren, thus enabling parents to work.
A second limitation of the economic support ratio is that comparisons of the ratio over time do not take into account changes in productivity in the economy. As the ability of the working population to support pensioners partly depends on the broader context of government spending and productivity, neither the old age dependency ratio (Pension Trends Chapter 2) nor the economic support ratio give a full picture and can only be taken as broadly indicative.
In terms of pensions, analysing the labour market provides an insight into the ability of people to save for their retirement. Labour force participation is important as working people accumulate state pension rights and may build up private pensions and other savings.
Analysis on data from the Labour Force Survey (LFS) shows that unemployment among economically active people aged 16 and over (see Definitions: Definitions of labour market status), fell between 1993 and 2004. In the early months of 1993 the number of unemployed people was 3.0 million and this fell to 1.4 million in mid-2004. It then rose gradually until mid-2008 (around 1.7 million) and increased sharply during the 2008 recession. Since this time there has been little change in those recorded as unemployed, with only small fluctuations occurring around the 2.5 million level1.
Figure 4.2 shows Office for National Statistics (ONS) headline proportional rates of employment and unemployment for three-month rolling quarters from April-June 1971 to July-September 2012. These rates have remained fairly stable except for around periods of recession during the 1980s, the 1990s and 2008.
It should be noted that employment rates include part-time workers. The highest recorded increase in part-time workers since comparable records began in 1992 occurred between the three months to April 2010 and the three months to July 2010. The three months to August 2011 saw the largest quarterly fall in the number of part-time workers, but the numbers have recovered somewhat since then1. In the three months to September 2012 there were an estimated 8.1 million part-time workers.
The trends shown in Figure 4.2 mask important differences between men and women. Figure 4.3 presents the employment, unemployment and inactivity rates for men and women for three-month rolling quarters between April-June 1971 and July-September 2012.
The decrease in the male employment rate since the early 1970s has not been matched by equivalent increases in unemployment, except temporarily as a result of the recessions of the early 1980s, early 1990s and 2008-09. Instead, over this period there has been an increase in the number of men classified as economically inactive (see Definitions: Definitions of labour market status). Between the three months to March 1971 and the three months to September 2012, male unemployment rose by 5 percentage points while male inactivity rose by 11.6 percentage points.
This change is widely understood as following from the decline in manufacturing. Manufacturing activity fell sharply through the 1974–75 and 1980–81 recessions and manufacturing jobs were lost as output fell. Official statistics for jobs by sector only began in 1978, but show a decline of manufacturing jobs from 7.1 million at the end of 1978 to 5.3 million by the end of 1983. A full set of data is not available, but the evidence suggests that job losses were disproportionately concentrated among older age groups. The older workers who lost their jobs had more difficulty finding new work and many ended up, more or less permanently, on invalidity or incapacity benefits and were, therefore, likely not to be recorded as unemployed.
For most of the 1970s and early 1980s, the employment rate for women aged 16 to 64 was below 57%. It rose steadily from the beginning of 1983, reaching 63.5% in 1990. The female employment rate fell less than that of men in the early 1990s recession; and it increased from a low of 61.6% at the beginning of 1993 to 65.6% at the start of the 21st century. It then remained fairly stable over the next decade between 65.3% and 67.1%. In the three months to September 2012 it was 66.0%. Meanwhile, the female inactivity rate has fallen from over 40% in the early 1970s to 28.6% in the three months to September 2012.
As the male employment rate has fallen and the female employment rate has risen, the gap between the two has narrowed. In the three months to April 1971, the gap was 39.2 percentage points; in the three months to November 2011 it was 9.8 percentage points. In the three months to September 2012 the difference was 10.4 percentage points. However, a higher proportion of women than men work part-time. In the three months to September 2012, 43.7% of women in employment worked part-time, compared with 13.5% of men.
Figure 4.4 shows that in April to June 2012, the employment rate for men was highest in the 35 to 39 age group, at 88.7%, compared to 75.4% for those aged 55 to 59 and to 55.3% for those aged 60 to 64, the age group that is nearing their State Pension Age (SPA) of 65. For women, the employment rate was highest for those aged 45 to 49 (77.8%). For women aged 55 to 59, who are nearing their SPA, the employment rate was 66.3%. It fell to 35.8% for those in the 60 to 64 age group, which mainly comprises women who are over SPA1.
In April to June 2012, employment rates for women were higher than for men in the 16 to 19 age group. For all other age groups, women’s employment rates were lower than those of men. For people below age 60, the largest difference between men’s and women’s employment rates was in the main age groups for child bearing, peaking in the 25 to 39 age range. Within this age group the largest difference between men and women was within the 30 to 34 age range where the proportion in employment was 15.9 percentage points higher for men than women. The other age group where women’s employment rates were much lower than men’s in April-June 2012 was in the 60 to 64 age group, where men were below SPA and most women were above SPA. The employment rate for women in this age group was 19.5 percentage points lower than the employment rate for men.
Figure 4.5 shows how the age structure of the labour market has changed since 1992. It should be noted that it has been produced using data that is not seasonally adjusted. This has been done so that we can divide the upper age bands at SPA as seasonally adjusted data for these age bands are not available. For men at present, SPA is simply age 65; but for women, from 6 April 2010 their date of birth must be used to divide those aged 60 and 61 into ‘under SPA’ and ‘SPA and over’ categories. This analysis should shed some light on whether changes in SPA are having an impact on employment rates in older age groups (see below). Other factors, such as changes in the economic climate or women’s employment patterns, may also affect employment rates at older ages. Therefore, caution is advised in attributing any changes to a single factor.
Since 1992, there has been an increase in employment rates of older people and a decrease in those of younger people. The former has been driven mainly by rising employment rates for older women. The fall in the employment rate in the 16 to 17 age group since 1992 is largely due to increased participation in further education. The 2008-09 recession caused a large fall in employment rates in the 16 to 17 age group and rates have continued to fall since the recession.
Male employment rates for the age groups 18 to 24, 25 to 34 and 35 to 49 were fairly stable (except for seasonal variations) for the decade to mid-2008, but they fell in the 2008-09 recession and have struggled to recover since. The employment rate for men of SPA and over was relatively stable in the 1990s, at around 7-8% for most of the decade. In the second quarter of 2011, it reached 11.9%, the highest level since 1992. It has remained stable around this point since this time with the employment rate for men in the third quarter of 2012 in this age group being 12%.
Employment rates for women in all age groups over the age of 25 have risen in recent years, with the greatest increases in the rates for women aged between 50 and SPA. In this age group, female employment rates rose from 58.0% in the third quarter of 1993 to 70.9% in the third quarter of 2011. In the ‘SPA and over’ group, employment rates rose from 7.8% in the third quarter of 1993 to 11.6% in the third quarter of 2012.
By contrast, the employment rate for women aged 16 to 17 has fallen since the fourth quarter of 1997, when it reached 53.0%, to 24.8% in the third quarter of 2011. There has been some element of recovery for the employment rates for women in this age group with a rate of 27.4% recorded in the third quarter of 2012. Employment rates for women aged 18 to 24 were fairly stable in the ten years to mid-2008 (except for seasonal variations), but fell in the 2008-09 recession and have continued to fall since then.
It is not yet clear whether the rise in women’s SPA since April 2010 is having an impact on employment rates. There are some signs of a change in employment rates in the ‘SPA and over’ group, with quarter-on-quarter falls in employment rates of this group in four (out of six) quarters since April 2010. It is not clear whether this is attributable to the SPA increases. In 2012, this trend continues to be seen within the employment rates for women in the SPA and over category. Of greater policy interest is the outcome for women who are nearing SPA (age 50 to SPA). However, detailed analysis of this topic is beyond the scope of Pension Trends.
The transition to retirement can be examined more closely by focusing on the employment status of the older workforce on either side of State Pension Age (SPA). Due to the increases in women’s SPA from 6 April 2010, it is now more difficult to do such analyses. But it is still possible using data that has not been seasonally adjusted from the Labour Force Survey, as explained in the previous section.
Table 4.6 shows that in April-June 2012 fewer women, 11.6%, aged SPA and over were in employment than men. For men in April-June 2012, 12.2% were recorded as being in employment in the SPA and over age group (the sum of full-time and part-time workers presented here is different to Table 4.6 because of rounding). This trend is different to what has been historically reported within this chapter, as previously a higher proportion of women than men have continued to work beyond SPA. It is thought that the numbers and proportions of those recorded as in employment above this age will be affected as the SPA for women gradually increases to become in line with men.
For those who are employed after their SPA there is a preference for part-time work, while for those below it the majority work full-time. In April-June 2012, around 67.2% of those who were in employment and working after SPA did so part-time. This is in comparison to 32.8% who were employed full-time. For those who were aged between 16 and SPA, 74.5% were employed full-time and 25.5% worked part-time.
From Table 4.6 it can be seen that, for men who were aged between 50 and SPA, there were around six times as many working full-time as there were part-time. This ratio is reduced to two-thirds after SPA and might be indicative of male pensioners deciding to work part-time past their SPA so that they can supplement their pension income1.
For women who are aged between 50 and SPA and SPA and over, the difference in the proportion of those working full or part-time is less pronounced. In April-June 2012, there were around 1.4 full-time workers for every woman that worked part-time in the 50-SPA age group. After SPA this ratio falls to 0.4 and again this might be representative of people deciding to supplement their pension income with part-time work1.
|50-SPA||SPA and over||50-SPA||SPA and over|
The transition to part-time employment over a worker’s life time can be seen more clearly in Figure 4.7. It shows that in April-June 2012, full-time work dominated for men until the 60 to SPA age group (40.7% of men in this age group were in full-time work). As men approach SPA, more of them work part-time. In April-June 2012, 14.3% of men in the 60 to SPA group worked part-time, compared with 9.2% in the 55 to 59 age group and 6.6% in the 50 to 54 age group. For women, there is less change in the proportion working part-time as they approach retirement: 28.6% in the 55 to SPA age group in April-June 2012, compared with 30.9% in the 50 to 54 age group.
Figure 4.7 also shows that the employment rate falls significantly at SPA for both men and women as the majority retire. However, it can be seen that a larger proportion of men than women take early retirement. In April-June 2012, 8.8% of women in the 55 to SPA age group were classified as retired. In comparison, 20.8% of men had already retired in the 60 to SPA age group.
From Figure 4.7 it can also be seen that before SPA a large proportion of people who are economically inactive give sickness or disability as the reason. From SPA, the proportion of sick and disabled falls for both men and women, as many consider themselves retired rather than sick or disabled.
For those who are in employment and are at their SPA or above, Figure 4.8 shows that more people are classified as self-employed when compared to their younger counterparts. In April-June 2012, 31.3% of workers at their SPA and above were self-employed. For those aged between 16 and SPA, self-employment accounted for only 13.5% of the total workforce.
The types of jobs that men and women carry out after their SPA were also different in April-June 2012, with around two thirds of men being classified as working in high skilled jobs. For women, a similar proportion were estimated to be working in lower skilled jobs.
The most common jobs that women undertake at or after their SPA include elementary occupations (such as cleaners), administrative positions and work within the professional occupations. For men the most common jobs include: managers, directors and senior officials; professional occupations; and the skilled trades.
Figure 4.8 shows that in April-June 2012 a significant proportion of those who continue to work after their SPA are still classified as employed. For these people, it can be seen in Figure 4.9 that they are more likely to have worked for the same employer for a significant amount of time. It shows that around two-thirds of people who are working after their SPA have remained with the same employer for 10 years or more. It also shows that around 8 in 10 who are at or above the SPA have worked for the same employer for five years or more.
The age at which people retire is difficult to estimate, in part because older people who have become economically inactive may give different reasons for their inactivity – retirement, ill-health, inability to find suitable work – even though their situations are the same. An alternative, more useful approach in the context of pensions is to examine the ages at which older people leave the labour market. To estimate these ages various methodologies can be used and the 'static' and 'duration of working (DWL)' estimates are presented here. Details of the methodologies used to calculate these indicators can be found within a recent article1 and within the Average Age of Exit Assumptions section of this article.
In most cases2, the point when people leave the labour market is when they cease to generate earnings to build up private pensions or to make National Insurance contributions. Therefore, these people may start to draw on pensions and other savings.
Figure 4.10 shows the trends in average age of withdrawal from the labour market using the DWL-based indicator for 2004 to 2010, the earliest and latest years for which this indicator can be calculated. For men, the estimate of average age of withdrawal increased from 63.8 years in 2004 to 64.6 years in 2010. For women, it increased from 61.2 years in 2004 to 62.3 years in 2010.
The DWL estimates underlying the calculation of average age of withdrawal from the labour market throw more light on the differences between men and women. Figure 4.11 shows working life expectancy (or DWL) for people aged 51 to 75. It also shows their non-working life expectancy, which is the difference between life expectancy and working life expectancy. In 2010, men had higher working life expectancy and lower non-working life expectancy than women at every age. For instance, men aged 58 in 2010 had some two years more working life expectancy and some five years less non-working life expectancy than women.
An additional benefit of using the DWL method is the production of estimates on duration of employment. For example, a man aged 51 in 2010 could expect to be in employment for another 11.0 years, while a woman of the same age could expect to be in employment for 8.8 years3.
Although the DWL-based measure is now our preferred measure of average age of withdrawal from the labour market, the estimates based on DWL are produced later in the year than those for the static indicator due to the DWL calculation’s reliance on life expectancy data. Also, they can be produced only from 2004 onwards, whereas results are available using the static method for 1984 onwards. For these reasons, this chapter also presents results using the static method.
It is possible to compare the results produced using the static indicator for the period 2004 to 2010 in Figure 4.12 with those based on the DWL-based indicator (Figure 4.11 above). For men, the results using the static indicator are similar to those using the DWL-based indicator, with the average age of withdrawal rising from 63.8 years in 2004 to 64.7 years in 2010. The static estimate suggests that the average age of exit from the labour market in 2012 for men was 64.8 years. For women, the results are slightly different: average age of withdrawal based on the static indicator rose from 61.6 years to 62.5 years between 2004 and 2010. In 2012, the average age of exit of women from the labour market as predicted by the static estimate was 62.6 years.
The long-term trends, which can be seen in Figure 4.12, are different for men and women. The estimated average age of withdrawal from the labour market for women has shown a rising trend overall since 1984, and a particularly sharp rise over the past decade. For men, on the other hand, there was a reduction in average age of withdrawal in the early 1990s, after which it remained stable until 2002 and then rose to 64.7 years in 2010. In 2011, it fell slightly. However, in 2012 the average age of withdrawal of men from the labour market, using the static estimate, rose again to 64.8 years.
An analysis of the average age of labour market withdrawal will be particularly important over the next few decades as the State Pension Age (SPA) in the UK increases. Raising the SPA is designed to lower the pension burden on Government and private pension providers by reducing the average number of years spent in retirement and increasing years spent in employment and saving for retirement4.
The calculations made to estimate average age of withdrawal from the labour market using the static method can also be used to estimate the proportion of men and women aged between 51 and 75 who leave the labour market at each age. Figure 4.13 shows the results for April-June 2012; for example, 13.7% of women in this age group withdrew from the labour market at the age of 61.
For men, the age of labour market withdrawal peaks between 64 and 66 years. For women, the peak is between 60 and 62 years. Thus, labour market exits peak around SPA for both sexes. However, many people retire before SPA, and others work beyond SPA. Figure 4.13 shows that in April-June 2012, 37.6% of men and 22.4% of women stopped working after age 65. It also shows that 51.9% of men aged 51 to 64 and 26.3% of women aged 51 to 59 had already left the labour market.
There are signs that the pattern of early retirement may be changing for women as SPA changes. In April-June 2011, the proportion of women aged 51 to 59 who had already left the labour market was higher, at 28.1%. In April-June 2010, when the SPA changes were introduced, the proportion of women who had left the labour market was 32.2%. However, as noted in The age structure of employment, caution is advised in attributing such changes to a single factor, as there may be more than one reason for them.
In 2002, the Barcelona European Council agreed to set an EU target to raise the average age at which people stop working by five years by 2010. In order to monitor this target, the EU adapted an The Organisation for Economic Co-operation and Development (OECD) methodology developed by Peter Scherer1 for estimating the average age of withdrawal from the labour market of people aged over 50. This is known as the dynamic exit age indicator. However, the Office for National Statistics (ONS) preferred to use the static indicator developed by the International Labour Organization (ILO) for measuring average age of withdrawal from the labour market2 because the dynamic approach produces more volatile results than the static approach.
Both the static and dynamic indicators are probability-based measures using economic activity rates from labour force surveys. Both methods suffer from weaknesses relating to the simplifying assumptions that need to be made in order to produce the indicators. As a result of concerns about these weaknesses, the EU recommended the use of a new approach based on a duration of working life (DWL) or ‘working life expectancy’ indicator, which combines information on life expectancy with economic activity rates. A methodology article published by ONS in 2010 entitled Average age of withdrawal from the labour market: A methodology update 3 explained how this recommendation was developed into a feasible methodology for the UK and why ONS now prefers the DWL-based approach to the static exit age indicator. The main reasons for preferring the DWL-based measure are that it is more robust and that it is not affected by unrealistic assumptions about:
everyone being economically active or inactive below or above certain ages;
people remaining economically active once they become so (until they finally leave the labour market);
and the same factors affecting the activity decisions of adjacent cohorts.
All of these assumptions need to be made in the case of the static indicator. The assumption that once someone is economically active they remain so until they leave the labour market is particularly tenuous in the case of women, who are likely to experience temporary periods of economic inactivity in the form of breaks to care for children and elderly relatives. As a result, the static method inflates the probability of remaining in the labour market, particularly for women, resulting in an overestimate of the average age of withdrawal.
Latulippe, D (1996): ‘Effective retirement age and duration of retirement in the industrial countries between 1950 and 1990’, Issues in social protection. Discussion Paper 2, International Labour Organisation, Geneva.
Banks J, Breeze E, Lessof C, Nazroo J (Editors) (2006) Retirement, health and relationships of the older population in England: The 2004 English Longitudinal Study of Ageing, Institute for Fiscal Studies.
Dini, E. (2009) ‘Older workers in the UK: variations in economic activity status by socio-demographic characteristics, household and caring commitments’, Population Trends No. 137, Office for National Statistics.
Latulippe, D (1996): ‘Effective retirement age and duration of retirement in the industrial countries between 1950 and 1990’, Issues in social protection. Discussion Paper 2, International Labour Organisation, Geneva.
Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: firstname.lastname@example.org
These National Statistics are produced to high professional standards and released according to the arrangements approved by the UK Statistics Authority.