Retirement age is a complex subject due to its dependence on a variety of factors of which health, wealth, involuntary redundancy and lifestyle are the most prominent. Economic resources in terms of state and occupational pension income, savings and assets have a direct bearing on whether and when individuals retire. The existence of an early retirement option in defined benefit (DB) pension schemes was seen as an important determinant of retirement among scheme members in the 1980s and early 1990s. This is now more or less defunct.

Two key developments in recent years have altered how an individual makes decisions about when to retire. The first was Adair Turner’s report on behalf of the Pension’s Commission (2005), which recommended an increase of retirement age for the state pension scheme. Increased life expectancy and financial pressures on pension funds are some of the key contributory factors that lie at the core of government’s pension reform in the UK and its philosophy of keeping people working. The second was the European Employment Directive prohibiting age discrimination in employment, since October 2006. Under this legislation, employers’ mandatory retirement ages will be considered age discrimination unless they can be justified (DTI, 2005).

Our recent study (Gough and Arkani 2007) adds to the debate on retirement age by focusing on the relationship between occupational pensions and retirement age. It uses the waves of Labour Force Survey (LFS) (1984-1991) and Quarterly Labour Force Survey (QLFS) (1992-2003) to review the changes in the actual average retirement age in the UK during the period 1984-2003 by gender and ethnicity and further investigates the link between occupational pension schemes and the actual retirement age of men and women. It also explores the impact of pension type on employees’ expected retirement age and the decision to take early retirement using the English Longitudinal Survey of Ageing (2002-2003).

The current debate surrounding the extension of retirement age to beyond 65 and possibly 70 is dominated by two diverse arguments. On the one hand, the government claims that it wants to enable those who are willing and able to work longer to do so, in order to make better financial provisions for their retirement (DTI, 2003). On the other hand, there is opposition among working people, unions and other interest groups on the grounds that extension of retirement age forces poorer employees to work longer because they do not have pension rights or savings put aside for retirement.

The study attempts to create a more realistic picture of retirement behaviour based on the actual retirement age of previous labour market participants. It uses the LFS and QLFS data to evaluate the actual retirement age of participants who have left the labour market in the previous 12 months and their pension status controlling for ethnicity. It then uses data from English Longitudinal Survey of Ageing to correlate retirement age to pension type.

There is a need to examine the change in the proportion of retirees and whether or not they have a pension. Computed and existing variables from the QLFS (1992-99) provided us with data about the percentage of retired respondents with occupational pensions, with state pensions, with no occupational pension, and with no state pension. The time series of the data indicated an increase in the proportion of retired men with occupational pensions from 1992-1999, which corresponded to a 1.7 year drop in the actual mean retirement age of men during the same period.

A similar situation was portrayed with respect to retired women. The proportion of those with occupational pensions rose between 1992 and 1999, with a slight reduction between 1997 and 1999. This corresponds to a drop of 1.2 years in the actual mean retirement age of women during the period. In summary, the evidence indicated that the rise in the proportion of men and women with occupational pensions between 1992-1999 paralleled a drop in the actual mean retirement age of men and women in the same period. It also showed that the highest proportion of retired men and women with occupational pensions were in the 60-64 and 65-plus (in the case of men) age groups.

The change in the retirement age trend of the 1990s and 2000s can be attributed to a number of factors. First, increased pressure on pension funds dissuaded firms from offering generous early retirement incentives to employees on DB schemes. Second, falling annuity rates and less generous income streams discouraged employees in DC schemes to take early retirement). Third, the normal retirement age (NRA) of private sector schemes rose between 1987 and 2000 in an attempt to reduce the liabilities of schemes. This is reflected in the reduction of active members in private pension schemes with a NRA of 60 and the increase of active members in schemes with a NRA of 65.

There is consensus among policy makers and academics about the relationship between occupational pensions and retirement age. Both groups believe that people with occupational pensions retire earlier than those without, and DB pension schemes reduce retirement age while defined contribution (DC) schemes increase it. However, analyses of the retirement surveys 1988-89 and 1994 have indicated that only 10% of early retirement is attributed to membership of occupational pensions compared with 30% ascribed to ill-health.

Our review of the average actual retirement age time series in the UK, as well as updating the existing literature, indicates that the changes in male and female retirement age trends occurred in the early rather than the mid or late 1990s. This is consistent with the Pensions Commission (2004) findings in terms of presenting a change in the direction of the trend in mean retirement age. The discrepancy in the actual point in time at which the change occurred may be ascribed to the commission’s use of data at five-year intervals.

The evidence presented in relation to membership of occupational pensions and retirement age suggests two correlations. First, the increase in the proportion of people with occupational pensions relates to a decrease in the actual mean retirement age from which we conclude that people with occupational pensions retire earlier. Second, the age groups of the majority of retirees coincide with the normal retirement ages of pension schemes with the greatest active members, namely 60 and 65. From this we conclude that occupational pensions have a direct impact on the decision to retire.

People’s expectations of retirement age have now completely altered. The general consensus is that pension schemes are no longer offering enhanced early retirement packages. For those aged 50 plus there is an acknowledgement that their pensions will not maintain their expected lifestyles. The main category of employees who are most keen to build up their pensions rights are women who left the labour market at some point in their careers and now want to build up their pension rights.

The picture that emerges is that in the long term the retirement age of both
men and women will increase, mainly due to two factors. First, there is an acknowledgment of increased life expectancy whereby individuals realise they may have to work longer than previous generations, albeit it in a more flexible way. Second, the declining membership of defined benefit pension schemes has in the past tended to encourage early retirement. The government’s recent implementation of the prohibition of age discrimination directive now facilitates and enables employees over the age of 50 to obtain employment more easily.

Pension trustees, as a result, not only have to take on the rigours of understanding the investments of their pension schemes but also need to have a familiarity with the changing landscape of an individual’s expectations of retirement age, retirement policy and retirement benefits of their sponsored company schemes.

Orla Gough is head of department of finance and business law at the University of Westminster and director of the Pensions Investment Academy