The Department of Employment's statistics show that unemployment reached nearly 3 million 18 months ago and has since been falling, with today's figures expected to show another drop. An alternative survey by NOP, which adopts a more stable definition of workforce participation, suggests that unemployment could be over 3.3 million and may only just have begun to fall.
The growing importance of self-employment, part-time jobs, multiple employment and government-sponsored training schemes has introduced huge ambiguity into the interpretation of official statistics in these areas.
The department's two-year-old Labour Force Survey Quarterly Bulletin, which asks individuals what they are doing in the labour market, promises to fill most of the gaps in our understanding. But there are still pitfalls.
In other areas of economic statistics, extreme departures from the norm are excluded from the calculations to prevent false conclusions being drawn. An example is the retail price index, which excludes the expenditure of people in households with the top 4 per cent slice of incomes and low-income pensioners mainly dependent on state benefits.
No such process is applied to the Government's employed workforce statistics, where part-time employment is deemed to begin at just one hour per week.
Ten years ago, someone who worked fewer than eight hours a week was a rarity, unless they were looking after children. Even if such jobs had been plentiful, the structure of the benefit system meant it was barely worthwhile working for such a low level of implied weekly income. But the introduction in 1988 of the reduced rate employers' National Insurance contributions offered cost savings to firms that sub-divided work into part-time units.
Other factors that have promoted the supply of jobs offering very low weekly hours are the increasing fixed costs that attach to permanent full-time posts and the increasing desirability of workforce flexibility in the wake of a lengthy recession.
Low-hours jobs have become feasible for the growing ranks of single-parent families and for students in full-time education, whose incomes have been affected by the reduced level of financial support by the state. The introduction of Income Support in April 1988, with a maximum threshold of 15 hours a week, has removed some of the disincentives to accepting such jobs. More people are also building a living wage out of a portfolio of low-hours employment.
Perhaps the most significant recent development has been the dramatic increase in the numbers of self-employed people who have found themselves with little or no work. Lacking the two years of employed National Insurance contributions necessary to qualify for unemployment benefit, and yet disqualified from means-tested benefits by their personal assets, tens of thousands of the jobless have no incentive to register their unemployment and are still classified as being self-employed.
Others have swapped the relatively secure job status of an employee for that of an (involuntarily) self-employed contractor, with variable and unpredictable hours of work.
A fresh insight into the question of meaningful work participation is provided by a forthcoming study* using NOP's Financial Research Survey, which is designed to measure personal financial behaviour. It is based on 60,000 interviews annually, a sample size large enough to rival the Labour Force Survey for reliability.
Unlike the official labour statistics, NOP adopts the Census convention that part-time employment constitutes between eight and 29 hours worked per week. This helps to identify under-employment among part-time workers and the self-employed. The NOP definition of work participation also excludes part-time workers who are also in full-time education, those on government employment schemes and training programmes, and unpaid family workers. Finally, NOP omits workers aged 65 and over, numbering about 500,000.
Whereas the Labour Force Survey reports a rise in self-employment of 67,000 between its summer 1993 and spring 1994 surveys, NOP figures show a fall of 150,000 over a virtually identical nine- month period. Both surveys agree there has been some recovery in self-employment between winter 1993 and spring 1994. Similarly, for part-time employees the LFS indicates an increase of employment of 150,000 over the previous nine months but the NOP study shows only a very small gain.
It appears that the main explanation for the contradictory trends is the exclusion of those working less than eight hours per week.
According to the stricter definitions of the NOP survey, self-certified unemployment is estimated at an average of 3.36 million during the first half of 1994 and has risen by about 130,000 since last summer. This compares with a claimant count of 2.71 million for the same period.
On the basis of the NOP data, male unemployment stands at 17 per cent of the male workforce and female unemployment at 9 per cent. This gives a national unemployment rate of 14 per cent, compared with the official measure of barely 9 per cent.
As the Conservative Party conference bemoans the apparent failure of economic recovery to raise government popularity, delegates may care to reflect on the alternative view of labour market developments offered by the NOP survey.
*Changing Working Patterns: of concern to financial services providers?, an NOP survey in conjunction with John Gilbert.
Peter Warburton is chief economist at Robert Fleming Securities.
(Photograph omitted)Reuse content