Lifetime experiences of self-employment
A hard copy of this report summary can be obtained by contacting Paul Noakes [E-Mail: Paul.Noakes@dwp.gsi.gov.uk] or by writing to him at the 'Social Research Division, Department for Work and Pensions, 4th Floor, Adelphi, 1-11 John Adam Street, London WC2N 6HT'.
Research Report No. 120
By Genevieve Knight and Stephen McKay
This report, by the Policy Studies Institute (PSI), examines the characteristics of self-employed people and the nature of, and trends in, self-employment up to 1994/95. The analysis used data from the Family and Working Lives Survey, a representative survey of more than 9,000 people aged 16-69 in 1994/95. The survey collected lifetime information on economic activity, family circumstances and pension provision.
The main findings were:
- In 1994/95 12 per cent of those in paid work were self-employed. Almost all of the self-employed had been an employee at some point during their working lives and most were employees immediately before and after a spell of self-employment. Compared to employees, those who were self-employed in 1994/95 were more likely to be male, older and married/cohabiting. They were also much more likely to work at home, work alone, and work as 'managers/administrators' and in 'craft and related' industries.
- Inflows into self-employment increased during the 1980s, but slowed down during the early 1990s. While in the past, each new generation was more likely than the previous generation to become self-employed, this no longer appeared to be the case by 1994/95. Spells of self-employment have also been getting shorter over time.
- Evidence suggests that the nature of self-employment is changing. Those who entered self-employment between 1992-1995 were more likely to: work part-time, work alone, work in 'associate professional/technical occupations', or in 'personal/protective services', and were less likely to be 'managers/administrators', than any previous cohort of self-employed.
- A mixed picture emerges with regard to retirement provision. Those who were self-employed in 1994/95 were more likely than employees to own their accommodation outright and have relatively high levels of savings. Both groups were building up rights to the basic state pension at a similar rate, and had built up a similar level of contributions to the state earning related pension scheme (SERPS) (a result of the fact that almost all of the self-employed had been employees at some point in their lives). Overall, both groups had spent a similar amount of time contributing to non-state pensions.
- However, among those who were self-employed in 1994/95 certain groups appeared to lack adequate pension provision. Compared with employees of a similar age, those self-employed aged 30-55 were particularly likely to have less than 5 years in a second tier pension. This was even more likely for those who had spent a large proportion of their working lives in self-employment.
- Only a small number of respondents in the survey who were over the state pension age had some experience of self-employment. However, there was no clear evidence to suggest that self-employment had reduced the level of their income. Experience of self-employment was associated with higher levels of savings and a greater likelihood of working after the state pension age, although it was difficult to ascertain from the survey whether the latter finding was through choice or financial necessity.
Introduction
This research analyses self-employment over the working lifetime, and the effects of self-employment on pension provision. The analysis is based on the Family and Working Lives Survey dataset (FWLS), a nationally representative survey of 9,139 people aged 16-69 years conducted in 1994 and 1995. It contains information concerning non-state pension rights; data about state pension rights has been derived.
A total of 12 per cent of those doing any paid work were self-employed, as they defined it, in their main job. This represents about 2.9 million workers whose main job was self-employed. These figures match established information from other surveys.
Self-employment was more common among men than women - 16 per cent of men were self-employed, compared with only seven per cent of women. The self-employed, as a proportion of all workers, also tend to increase with age. Only five per cent of workers aged 16-29 were self-employed, compared with 13-15 per cent of those workers aged 30-59, and 21 per cent for the over 60s. The high rates of self-employment for the over 60s mostly reflected a small number of employees in this age range - we estimate that there are only about 220,000 self-employed over 60 (and 850,000 employees), compared with 1 million in their forties (and 5½ million employees).
The proportion of men working as self-employed seems to have fallen slightly in recent years.
Personal characteristics of the self-employed
The general picture of the self-employed is one of stability: stable lives, stable employment. Large numbers of the self-employed were male (73 per cent), 'middle-aged' (34 per cent were in their forties, and only 12 per cent under 30) and married or cohabiting (82 per cent). They were also more likely to have children aged under 16 (60 per cent of self-employed compared with 50 per cent of employees). These figures relate to those working as self-employed in 1994/95. The self-employed were more likely than employees to have vocational qualifications, but less likely to have attained academic qualifications.
There were strong links between employment status, and the type of work done by a person's partner and parents. The precise reasons for these links were not entirely clear. Nevertheless, more than twice as many of the self-employed had working partners who were self-employed as did employees (34 per cent compared to 15 per cent). The rate of self-employment was twice as high for those whose father was self-employed (20 per cent), rather than an employee (10 per cent). There was also an effect for those whose mother was self-employed, though this was weaker than for the father's status.
Those with health problems affecting their work were more likely to be self-employed than the general population. The self-employment rate of those with health problems “which affected their ability to do paid work” was 19 per cent, or 1½ times that of the general working population.
The self-employed were no more likely than employees to be providing care for a disabled or elderly person. However, 52 per cent of carers working as self-employed said it affected their work in some way, compared with 30 per cent of employees who were carers. More detailed investigation found that the self-employed were more able to combine caring for someone with paid work than were employees, more of whom left paid work in the course of caring.
Job characteristics of the self-employed
There was no difference in 1994/95 between employees and the self-employed in rates of working part-time, but rates of working at home were higher among the self-employed (31 per cent of self-employed, compared to six per cent of employees). Home working was particularly important for disabled people, carers and lone parents. Almost two thirds (65 per cent) of the self-employed worked on their own, employing no other staff.
The self-employed were concentrated in particular jobs and industries. Some 25 per cent of self-employed declared themselves 'managers and administrators', (compared to 12 per cent of employees) and 22 per cent defined their occupation as 'craft and related' (compared to 10 per cent of employees): this group mainly comprises skilled manual workers with a 'trade'. Very few of the self-employed (just two per cent) worked in the kinds of clerical/secretVerdana jobs quite common among employees (16 per cent of employees).
Some 45 per cent of the self-employed did not give details of their net earnings, predominantly because they said they were unable to. This large amount of missing data means we should be particularly cautious about interpreting earnings figures for the self-employed - as all other research has found.
Among those whose net earnings were known, the self-employed were more likely than employees to appear at both the lower and upper ends of the earnings distribution. As many as 29 per cent of the self-employed earned below £3,600 per year (£69 per week) compared with 22 per cent of employees. But one in five (21 per cent) of the self-employed had net annual earnings of £18,501 (£356 per week) or more, compared with just 7 per cent of employees. Self-employed workers employing staff had the highest earnings, on average, with 39 per cent of men and 19 per cent of women earning receiving net annual earnings of £18,501 (£356 per week) or more. More than half (55 per cent) of self-employed women earned less than £69 per week, net.
There were no clear links between self-employment and being made redundant, or having experienced unemployment. However, the self-employed were marginally more likely than employees to have experienced long-term unemployment (of 6 months or longer).
Self-employment over the lifetime
The number of people of working age who had tried self-employment, but were no longer self-employed, was about equal to the number of those currently self-employed - seven per cent of the population in each case. One man in every five (20 per cent) had been self-employed at some stage (including those still self-employed), double the proportion (nine per cent) of women.
Almost all of the self-employed (94 per cent) had at least some previous experience of being an employee. Conversely, only seven per cent of current employees had any experience of being self-employed. Nearly half (48 per cent) of the currently self-employed had first been self-employed before the age of 30.
There was a great deal of diversity in the motivations for becoming self-employed. The most popular reason was “Flexibility of being own boss” (27 per cent), followed by the “Nature of the job” (23 per cent). “To Earn more money” (14 per cent) and “Self-fulfilment” (13 per cent) were also strong motivating factors for starting self-employment.
For all spells of self-employment (current and previous), the median duration was close to eight years. Spells of self-employment are getting shorter over time, with spells begun in the 1990s being particularly shorter than before. Those moving into self-employment from unemployment stayed in this status, on average, only half the time of those who were employees immediately prior to becoming self-employed.
Taking all spells of self-employment, immediately prior to becoming self-employed, nearly two thirds (65 per cent) were working as an employee, 23 per cent were inactive (not working, and not looking for work), and 11 per cent unemployed and looking for work. Among younger age groups, more of the moves into self-employment came from unemployment (26 per cent for those aged under 30). After a completed spell of self-employment 57 per cent moved into employment, 29 per cent into inactivity and 14 per cent into unemployment.
Trends in self-employment
The long-term economic trend towards a larger services sector, and a smaller production sector, seems likely to sustain a high level of self-employment. Other factors influencing the number of self-employed are rather more difficult to predict.
Analysis of the FWLS found that the trend towards greater self-employment among younger generations had come to an end. Those now aged 16 to 29 were about as likely to be self-employed, in their twenties, as were those now aged in their thirties at the same age. Previously, each generation was “more” likely than the last to be self-employed.
Analysis by year of becoming self-employed found that inflows had risen in the 1980s, but were slowing down in the 1990s. More of the entrants to self-employment in the early 1990s than previously were working part-time and did not have staff. A higher proportion in more recent years have been in 'personal/protective services', and rather fewer working as 'managers'. An increasing number have worked in the financial sector rather than hospitality.
The self-employed are a diverse group. There are no straightforward methods of classifying them into particular types. However, some of the main patterns of self-employment are linked to their age (whether 'young', prime or older), gender and working hours, and whether they employ staff.
Financial provision for retirement
The self-employed (under state pension age) were more likely than employees under state pension age to have relatively high levels of savings. Of those with savings, about one in five (19 per cent) of the self-employed had £20,000 or more saved, compared with nine per cent of employees, whilst eight per cent of the self-employed had £50,000 or more saved, compared with three per cent of employees. They were also more likely to be outright owners of their accommodation.
Three-quarters (75 per cent) of the self-employed had spent some time contributing towards a non-state pension, as had 72 per cent of employees. The self-employed had spent an average of 4½ years in occupational pensions, and 5 years in personal pensions of all types (9½ years in total). The total for employees was similar (8½ years), but with the majority of that time spent in occupational pensions (7½ years) and very little time in personal pensions (one year on average).
Analysis controlling for age found that:
- Rights to the basic state pension were being built up at very similar rates among employees and the self-employed, although self-employed women appear to be spending longer in the labour market than employed women.
- The average amounts of time spent in SERPS were rather low, reflecting the date of the survey (1994-95). Those most likely to have been in SERPS were the young (who were least likely to have opted out), and employees in their fifties (who had spent longer in the scheme).
- The average number of years spent contributing to non-state pensions was similar for each age group among employees and the self-employed. Employees were much more likely than the self-employed to have occupational pension rights, and less likely to have made personal pension arrangements.
- The ex-self-employed appeared to have built up fewer pension rights than the currently self-employed.
In 1994/95, occupational pensions which the currently self-employed had previously joined were less likely to be 'defined benefit' schemes than for employees (45 per cent of the currently self-employed, compared with 61 per cent of employees), and more likely to be 'defined contribution'.
About one quarter (23 per cent) of those who were currently self-employed and who were aged 30-55 had 'limited' second tier pension provision (defined as less than five years in a second tier pension, whether state or non-state), compared with 18 per cent of those who were currently employees and who were of the same age. However, among those who had spent at least 40 per cent of their working life self-employed, some 40 per cent had limited second tier pension coverage.
Work, incomes and savings after state pension age
A total of 129 respondents over the state pension age had some experience of self-employment. Among this small total, more than half (56 per cent) were men, compared with 31 per cent of those with no experience of working as self-employed.
There was no clear evidence that the experience of self-employment reduced the level of income after state pension age (this is consistent with the findings of Meager et al, 1994, using different data).
The main sources of income after state pension age are state and non-state pensions, earnings, savings and other benefits. Those with experience of self-employment were among the most likely to still be in paid work - eight per cent of those who had ever been self-employed were in full-time work, and a further 14 per cent were working part-time, compared with four and eight per cent of those over state pension age without experience of self employment. About half of those with some experience of self-employment had income from an occupational or personal pension - the same proportion as those who had never been self-employed.
Experience of self-employment was associated with higher levels of savings, and a slightly lower than average chance of receiving a means-tested benefit in retirement.
Publication details
Knight, G. and McKay, S. (2000) “Lifetime Experiences of Self-Employment ”(DSS Research Report No. 120) Leeds: CDS (£31.50)
Relevant publications
Other relevant publications on individuals' pension provision
McKay, S., Heaver, C. and Walker, R (2000) “Building up pension rights” (DSS Research Report No. 114) Leeds: CDS (£33.50)
Thomas, A, Pettigrew, N and Tovey, P(1999) “Increasing compulsory pension provision: attitudes of the general public and the self-employed ” (DSS In-house report No. 48)
Hedges, A. (1998) “Pensions and Retirement Planning” (DSS Research Report No. 83) Leeds: CDS (£28.00)Disney, R., Grundy, E. and Johnson, P. (1997)“The Dynamics of Retirement” (DSS Research Report No 72) London: TSO (£36.00)
Field J. and Prior G. (1996) “Women and Pensions” (DSS Research Report No. 49) London: HMSO (£35.00)
Hawkes, C. and Garman A. (1995) “Perceptions of Non State Pensions ”(DSS In-house report No.8)
Bone, M., Gregory, J., Gill, B. and Lader D. (1992) “Retirement and retirement plans” (Social Survey Division, Office of Population Censuses and Surveys) London: HMSO