Retirement in the 20th Century

Britain has aged during the twentieth century. The age structure of the British population has shifted toward older age groups; high youth unemployment has increased the pressure on these age groups to withdraw from the labour force. With financial support from the ESRC, CEPR Research Fellow Pat Thane organized a workshop on 11 January to discuss the institution of retirement during the present century in the light of historical experience.

The morning session commence with an overview of the economic aspects of retirement by Paul Johnson (LSE and CEPR). In 'The Economics of Old Age – a Long Run View 1881 – 1981’, Johnson contrasted three' interpretations of retirement. The functionalist approach suggests that the old retire when they can no longer effectively compete within the labour market. The liberal view interprets retirement and pensions as the distribution of an increased prosperity by an enlightened society. There is also a radical critique of these views, which argues that poverty and dependency in old age are socially created, and that retirement and pensions are a means of social control.

Johnson argued that none of these views adequately describes or explains the dramatic fall over the last hundred years in the labour force participation rates of the elderly in Britain. Data on the employment of males aged 65 and above, drawn from the Census returns from 1881 to 1981, show that changes in retirement rates do not clearly coincide with the introduction or extension of the state old age pension scheme, nor with more direct efforts by government either to encourage older workers to stay in the labour force or to promote premature withdrawal from work. A cross-sectional analysis by occupation of elderly employment since 1881 shows that old workers have been consistently forced into marginal, low-wage occupations such as labouring and agricultural work, suggesting that they stay in the labour market because of a lack of alternative resources. Even among those who do retire, economic insufficiency is the norm, with the majority of elderly people today, as in Edwardian times, living in poverty; for this group the life-cycle theory of savings seems inapplicable. Johnson concluded that the scope for individual economic provision for old age, either through more private saving or continued employment, was still slight, and the elderly were likely to be a growing financial burden on the state. He went on to advocate more cross-sectional work on census data in order to increase our understanding of changes in retirement behaviour.

Workshop participants agreed that class differences were significant. Working-class retirees were faced with problems of poverty, while the middle classes often viewed retirement as a valuable period of leisure. In addition, those with important skills were often able to negotiate a convenient time of retirement, which added another important dimension to the analysis. It was also suggested that any economic model should include the influence of the state on the attitudes of the labour force, though this might prove difficult to measure.

In the discussion, Jon Stern (DHSS) questioned Johnson's criticisms of the life-cycle savings model and suggested that the applicability of this model may have changed over time as income has increased. Stern also thought that the level of disaggregation of the data - by broad occupational group -was too limited. It was agreed that a major obstacle to research in this area is the absence of a longitudinal sample of personal income and wealth, such as is available in the US. Nicholas Bosanquet (York and CEPR) stressed the need for more research at the micro-economic level in order to show the impact of non-governmental factors like personal saving and occupational pensions on retirement behaviour.

Leslie Hannah (LSE and CEPR), in presenting his paper ‘The Role of Occupational Pensions – 1900-1980’, stressed the wealth of data available In the records of occupational pension funds, which researchers have neglected. Hannah traced the development of these occupational pension schemes. Their growth was associated with an increasing bureaucratization of employment, and the development of internal labour markets within firms. He maintained that the institution of fixed retirement age could be attributed to the structure of occupational pension schemes. Salaries in bureaucratic enterprises were often steeply graded according to seniority, and this gave rise to a gap between wages and the productivity of older workers. Hannah argued that the final retirement age represented the point at which the enterprise no longer found the gap acceptable. The growth of occupational pensions was also encouraged by insurance companies which perceived economies of scale in dealing with large organizations, as well as the favourable tax treatment accorded to pension funds. Hannah noted that until 1960 only one-sixth of contributors actually collected their pension -the rest either died or left the organization before retirement! Occupational pensions thus seemed to be a form of general saving, though not necessarily for old age. Hannah concluded his paper with a comparison of final salary and index-linked pension schemes. Despite the obvious benefits to the retired of index-linking, final salary schemes are much more common. This was a further indication, Hannah argued, that the needs of the employer have been dominant in the establishment of current retirement patterns.

Discussants agreed that the possible taxation of pension funds and the growth of individual insurance schemes might force organizations to abandon occupational pensions. Tom Schuller (Institute of Community Studies), however, questioned Hannah's view of decision-making by large organizations. Schuller's own research into the retirement policies of companies had revealed that employers had little knowledge of the costs of pension arrangements, and this ignorance did not accord with Hannah's view of the economic rationality of employers. Hannah agreed that this was a crucial issue. He suggested that where employers required a formal retirement policy merely to secure acceptance of a mandatory retirement age, they may have selected pension schemes in a rather arbitrary fashion. Dulcie Groves (Lancaster) turned the debate towards pension provisions for female workers. The current increase in the rate of divorce had made this issue very topical. She noted that the right of divorced women to share in the pension of their former husbands was currently under review; it was complicated and required further research. The morning session concluded with a brief discussion on the relationship between the age of an industry, the age of the workforce, and the introduction of a formal pension scheme. This area also needed further investigation.

Vanishing Spinsters

In ‘The Vanishing Spinsters: Women and Pensions 1937-59’, Dulcie Groves (Lancaster) discussed pension provisions for women, a subject largely neglected in the morning session. This years 1937 to 1959 are of particular Interest. During this period the number of single women of working age declined sharply. Groves argued that changes In the proportion of spinsters in the labour force significantly influenced the recognition of female pension rights. The first World War had created a group of single, surplus women who had been obliged to enter the labour force. During the 1930s these women campaigned for the reform of women's pension rights. As single women, they were perceived by society to have a 'legitimate' reason for taking paid employment. This legitimacy assisted their campaign, and a state pension for women at age 60 was introduced in 1940. After the Second World War, however, the numbers of men and women were more evenly balanced, and the social climate favoured early female marriage. This encouraged women to enter the labour force before marriage and to return to it after they had completed their families. In addition, many married women chose not to pay full National Insurance contributions, and this helped maintain the perception of women as financial dependants. During the 1950s, although the Labour Party's superannuation proposals considered female pension provision, women themselves appeared more concerned with improving widow's benefits than pension rights. Groves argued that these attitudes were consistent with the low proportion of single career women in the workforce.

During the discussion, it was agreed that the crucial factor affecting female pensions was women's part-time employment, which arose from family obligations and social pressures. Current evidence suggests, however, that even if there were no obstacles to pension rights for part-time workers, most women would still opt out of such schemes. Female perceptions of themselves thus enforce their already dependant status.

Pat Thane (Goldsmiths College and CEPR), in her paper' Ageing and the Economy -1948-1964', outlined four subjects for investigation: why the elderly leave paid employment, how they make the transition from full-time employment, what positive contribution they make to the economy when retired, and what effects changes in the age structure of society have on the economy. After the Second World War it was recognised that the population structure of Britain was shifting towards older age groups. This led to intense interest and research into the potential social, economic, and political effects of such a shift. The labour shortage in the post-War period had led these researchers to advocate later retirement ages, and indeed the Ministry of Labour had called for delayed retirement. By the early 19605, the calls had subsided. The trend towards earlier retirement continued, aided by the labour boom of the early 1960s and the extension of state pensions. Thane concluded that the pattern of retirement was influenced by the state, by employers, and by individual preferences, and of these three factors, government seemed to have the weakest effect. Similarly, survey evidence from the 1950s indicated that individuals did not object to the concept of retirement, but to mandatory retirement at a pre-determined age. They also preferred a period of lighter or part-time work prior to full-time withdrawal from the labour force. Thane suggested that the continuing trend towards early retirement was influenced by structural changes in industry and pressures from management. These resulted in an increasing pace of work and less control by workers over that pace. Older workers found this particularly stressful and were often forced to abandon such employment before the mandatory age. In addition, there were fewer jobs available involving only light labour which had formerly provided a transition to full retirement.

Chris Phillipson (Keele) argued that the 1950s survey research had employed a naive methodology. The State played a major role, he argued, in retirement patterns through its control over employers. Other participants felt that while historical data were useful, their interpretation was problematical. British evidence was cited suggesting that a transitional period prior to full retirement created social and economic difficulties for the individual. Lynes produced evidence from France and Sweden, however, illustrating the popularity of such schemes. There was thus a need for further research into the benefits of gradual or 'progressive' retirement.

The final paper, 'Managing the Transition from Employment: Questions for Today', was presented by Tom Schuller (Institute for Community Studies). He suggested that the transition from full-time employment to retirement was a complicated process which varied in length and nature and moved through specific phases before the Individual attained stability. The roles played by the State, the employer, unions, and the individual should be analyzed simultaneously. Some measure of the prevalent attitudes and expectations about retirement should be included in the analysis. Schuller pointed out that individuals often had difficulties in forming coherent ideas about retirement, both because of short planning horizons and because of a lack of experience of the nature of retirement. Since the retired have little contact with the unemployed and retirement is generally irreversible, many people enter retirement with little idea of how they will be socially and psychologically affected by the absence of daily paid work.

Jon Stern questioned the benefits of combining within the same model economic and statistical data with attitudinal, ideological and behavioural evidence. He doubted whether the latter material was relevant for policy research. While Schuller conceded the difficulty of measuring workers' attitudes, he maintained that both the transition period and the changing perceptions of retirement contributed an important dimension to the understanding of the aggregate data. Individuals clearly experienced difficulties in restructuring their lives outside the routine of employment; further research into individuals' 'time management' would be beneficial.

The workshop concluded with an assessment of the proceedings. It was agreed that the integration of historical, social and economic data, though difficult, was important and stimulating. Participants expressed fears that pension schemes and retirement were about to be restructured without sufficient knowledge of the effects of either the current or the proposed arrangements. The papers revealed that while there are important social, economic and ideological forces and constraints affecting retirement, we know little about their interaction at either the individual or the aggregate level. The present century offers an opportunity for such analysis. Despite changing social and economic conditions, there has been a gradual but steady increase in the withdrawal from the labour market of men and women still able to contribute to the economy.