Things have got better – labour market performance 1992-2002

Issue date

07 Aug 2003

Click here to download the pdf version of this report

Since 1997 the policy agenda has challenged the previous orthodoxy by introducing a National Minimum wage, strengthening employment protection and introducing a new right for trade union recognition. According to the Government’s critics, these policies should have meant fewer people in work and higher unemployment. The Government has also vigorously pursued tax-benefit reform to make work pay and through the New Deal revitalised labour market programmes to help people back to work. The critics have also attacked these policies as expensive, over-complicated and ineffective.

But as we show below, the actual outcome has been very different. Indeed, we are now closer to full employment that at any point in the past 25 years.

There is no official definition of full employment. However, one conventional indicator of an economy full employment would be an ILO unemployment rate of between 3 and 4 per cent. This is roughly the modern day equivalent of the historical claimant count unemployment rate experienced in the period before the mid 1970s. This compares with today’s national unemployment rate of 5 per cent. Using the same benchmark, we estimate that up to 40 per cent of the UK labour market is already at or near to 'full employment'.

Of course, this can only be a partial measure. We are also concerned with measures of labour market exclusion, such as the inactive that say they want a job and the number of workless households. In overall numbers these both now pose even bigger challenges than open measured unemployment.

But to tell a complete story we also need to look behind the overall numbers at the quality and fairness of employment. We need to look at to what extent people are taking 'flexible' forms of employment such as temporary or part time jobs out of choice. We also want to see if wage inequality - the gap between the top and the bottom and the overall position of low paid workers - has improved. We also need to know whether inequalities between men and women and black and white workers are getting worse or better.

This report compares two five-year periods - 1992 to 1997 and 1997 to 2002. The year 1992 is a convenient starting point for several reasons.

The most important is that it allows us to compare the labour market impact of five years of consistent economic growth up to 1997 with five years of consistent economic growth after 1997. We thus avoid including the recession years of 1990-1991 that would have otherwise severely depress the pre 1997 record.

1992 was also an Election year, so we can compare a period in which one set of labour market policies operated with another in which a quite different set of polices were in place.

In addition, ongoing revisions to labour market statistics in the light of the 2001 Census means that several measures do not at present go back beyond 1992 on a consistent basis.

Of course, no choice of comparative period is perfect. Our 2002 end date means we miss out on some of the most up to date indicators of labour market change into the first few months of 2003. However, we do not feel this significantly distorts the picture we paint.

Employment

The rate of job creation has sharply increased since 1997. Between 1992 and 1997 total employment under the Labour Force Survey definition went up by 740,000, compared with 1,290,000 between 1997 and 2002. In other words, the economy has created 75 per cent more jobs in the five years since 1997 compared with the previous five years.

The increase in employment since 1997 has been driven by higher employee employment. Self-employment fell between 1997 and 2002 (although latest figures for 2002-2003 show this has now been reversed). In contrast, there was a modest increase in self-employment between 1992 and 1997 following a major reduction in the recession of the early 1990s.

In both periods there were equally large falls in both unpaid family workers and of those on government training schemes without a contract of employment. The latter in part reflects changes in employment status. More people have been undergoing training supported by government funding as employees with a contract of employment rather than as government trainees without one. But in addition, actual spending and provision of government supported training was cut back between 1992 and 1997.

Since 1997 all of the increase has been in permanent jobs, nearly two thirds of them full-time. In contrast, the number of people in temporary jobs fell.

This is in sharp contrast to the trend in job creation before 1997. Between 1992 and 1997 job creation was dominated by part time and temporary job growth.

Part time work often dominates in the early years of a labour market recovery, so the economic cycle provides some explanation for the balance between full and part time jobs before 1997. However, this cannot explain the extraordinary 35 per cent growth of temporary work in the five years before 1997. Temporary jobs accounted for over 50 per cent of the net growth in employee jobs in this period.

The jobs record compared

	1992-1997		1997-2002
Total in work	+738	+ 2.9%	+1292	+ 4.9%
Employees	+875	+ 4.0%	+1574	+ 6.9%
Self-employed	+82	+ 2.6%	-145	- 4.4%
Other	-217	-39.4%	-137	-41.0%
Employees
Full time	+234	+ 1.4%	+1165	+ 6.9%
Part time	+641	+12.5%	+ 409	+ 7.1%
Permanent	+410	+ 2.0%	+1787	+ 8.5%
Temporary	+465	+35.9%	- 213	-12.1%

Notes: 'other' is unpaid family workers and people on government training schemes without a contract of employment. Permanent employees estimated by subtracting temporary employees from total employees. All figures UK, seasonally unadjusted, Spring.

Source: Office for National Statistics, TUC estimates.

Involuntary temporary and part time working

Much of the growth in part time and temporary work before 1997 was involuntary. Nearly 45 per cent of the growth in temporary work in this period was involuntary: people took the jobs because no permanent work was available. This increased the already high level of involuntary temporary working from 37 per cent of all temporary workers in 1992 to 39 per cent in 1997.

Part time work has generally been more popular, with relatively low levels of involuntary working. Even so, the share of part time workers (including the self-employed) who said they wanted a full time job increased from just over 11 per cent to nearly 13 per cent between 1992 and 1997. About 21 per cent of the overall increase in part time work (including self-employment) in this period was involuntary.

In contrast, involuntary working has fallen very sharply since 1997. The number of people who said they took temporary work because no permanent jobs were available fell by nearly 40 per cent between 1997 and 2002, while the number of involuntary part time workers (including self-employed) dropped by 29 per cent. The share of temporary workers who said they could not find a permanent job is still high, but in 2002 was down to 28 per cent. The share of part time workers (including self-employed) saying they could not find a full time jobs has also fallen, to 8.4 per cent in 2002.

Involuntary Temporary And Part Time Working

UK, adjusted	1992-1997		1997-2002
Change	000s	%	000s	%
Temporary	+207	+44.5%	-251	-37.3%
Part time	+165	+25.7%	-231	-28.7%
All involuntary	+372	+33.6%	-482	-32.6%

Note: temporary employees who said they took a temporary job because they could not find a permanent job; part time (employees and self-employed) who said they took a part time job because no full time jobs were available.

Source: Labour Force Survey

Unemployment

Unemployment by the ILO definition (all those who looked for work in the 4 weeks before the survey and are able to start a job in 2 weeks time) has fallen significantly since 1997, from 7.2 per cent in Spring 1997 to 5.2 per cent in Spring 2002. The actual number unemployed by ILO definitions fell by just over 500,000 or nearly 26 per cent. Latest figures show that unemployment and the unemployment rate continued to fall over the year Spring 2002-2003.

The period 1992 to 1997 also saw big reductions in unemployment. The ILO unemployment rate fell from 9.7 per cent to 7.2 per cent. Over this period unemployment fell by nearly 750,000, or 27 per cent. However, as we show below, job creation can only explain part of this fall.

Long-term unemployment

A key concern in the labour market has been long -term unemployment. Between 1992 and 1997 reductions in long term unemployed lagged behind the overall improvement in unemployment. Although the number of ILO unemployed out of work for more than 12 months fell by just under 22 per cent, the share of long-term unemployment in total unemployment went up from just under 36 per cent to just under 39 per cent.

Since 1997, long-term unemployment has fallen significantly and faster than unemployment as a whole. Between 1997 and 2002, those out of work for more than twelve months fell by over 56 per. The share of long -term unemployment fell from 37.5 per cent in 1997 to 21.7 percent in 2002.

Some of this may reflect the economic cycle, with long- term unemployment typically falling with a lag behind short-term unemployment in the early years of a labour market recovery. But it will also reflect the effectiveness of New Deal policies in targeting the long-term claimant unemployed, especially the young.

Long-term unemployment 1992-1997

UK, adjusted	1992-1997		1997-2002
	000s	%	000s	%
All ILO	-758	- 27.1%	-512	-25.1%
Over 12 mnths	-220	-22.3%	-433	-56.7%
Over 24 mnths	+ 25	+5.5%	-307	-63.6%

Note: ILO unemployed includes all who looked for work in 4 weeks before survey and were able to start a job in 2 weeks.

Source: Labour Force Survey

Labour market exclusion

There is a major discrepancy between the rise in jobs between 1992 and 1997 and the fall in unemployment. We would normally expect the increase in employment to be bigger than the fall in ILO unemployment, as some people will move out of economic inactivity to either take new jobs or actively start looking for work. However, between 1992 and 1997 the reverse happened - the fall in ILO unemployment was slightly greater than the increase in total employment. Clearly, other factors must also be responsible to account for such a big fall in ILO unemployment in the period before 1997.

unemployment benefit reform: a side-effect of the introduction of Jobseekers Allowance and the overall priority to reduce the numbers on the highly visible claimant count meant many people (often male older workers) were shifted out of the active labour market and moved from unemployment related benefit on to longer term sickness and disability benefit, so they no longer counted as claimant unemployed;

fewer young people: demographic changes meant that the population of under 25s fell by nearly 900,000 or just over 12 percent between 1992 and 1997. Under 25s typically have much higher unemployment rates than prime age workers;

exclusion of lone parents: the rising number of lone parents found it hard to enter the active labour market because of tax-benefit disincentives, lack of affordable child-care and weak or non-existent labour market support programmes. Much of the rise in workless households since 1992 was in lone parent households.

Since 1997 these trends have gone into reverse. The number of working age inactive who want a job has fallen by just over 130,00, the number of working age people in workless households has dropped by just over 300,000. In addition, the fall in the population of under 25s has halted and is now starting to reverse.

If all those who want work are taken into account the balance between the two periods changes. Between 1992 and 1997 there was still a significant fall in the total that wanted work, down by nearly 12 per cent. However, the fall has been greater since 1997, with a decline of just under 15 per cent.

Similarly the TUC’s estimate of the 'want work' rate which takes account of the inactive who say they wanted a job fell from 16.3 per cent in 1992 to 14.3 per cent in 1997. In comparison, the want work rate fell from 14.3 per cent to 12 per cent between 1997 and 2002.

Unemployment and labour market

exclusion compared

UK, unadjusted	1992-1997			1997-2002
Spring each year	000s	%		000s		%
ILO unemployed	- 758	- 27.1%		- 512		- 25.1%
Inactive, want work	+180*	+ 8.2%		- 126		- 5.3%
All who want work	- 578	- 11.6%		- 638		-14.5%

Workless households	+274	+ 6.2%		- 307		- 6.5%

Rates (%)	1992		1997		2002
ILO unemployment	9.8%		7.2%		5.2%
Want work rate	16.3%		14.3%		12.0%
Worklessness	12.6%		13.2%		11.6%

Notes: *April-June 1992 to Mar-May 1997. Inactive who want work are all those of working age not in work or defined as ILO unemployed who still said they wanted a job 'Workless households' are total number of people in households where no working age adult has a job. Want work rate is ILO unemployed plus inactive who want a job as percentage of active labour market plus inactive who want a job. Workless rate is working age living in workless households as share of all of working age.

Source: Office for National Statistics, TUC estimate.

Industrial structure

The statistics for this section come from the Employer Survey, which is generally regarded as giving a more accurate picture of employment by industrial sector. The totals however do differ slightly from those shown by the household Labour Force Survey. The Employer Survey results show that between 1997 and 2002 the number of employees increased as follows:

Employee employment in services went up by nearly 2,100,000 or just over 11 per cent;

Employee employment also went up in construction by 144,000 or just under 15 per cent;

In contrast, manufacturing saw employment fall by 562,000 or nearly just over 13 per cent;

There were also losses in agriculture, down by nearly 80,000 or 21 per cent, and in energy and water, down 13,000 or 6 per cent.

Latest figures show the loss of manufacturing jobs carried through into 2003. Employee employment in agriculture and energy and water has also declined. There was further modest growth in service sector employment.

Service sector employment was also the biggest job provider before 1997, with services expanding by just over 1.2 million. However, between 1992 and 1997 manufacturing recorded modest job gains, albeit following massive job losses in the recession of the early 1990s.

Other sectors had mixed fortunes before 1997. Agricultural employment was stable, but employee employment in energy and water (including coal-mining) declined very sharply. In both periods the changes in construction employee employment will have been affected by people moving back into and out of self-employment.

More jobs in services and construction, fewer in manufacturing, agriculture, and energy

UK, June	1992-1997		1997-2002
Change	000s	%	000s	%
Services	+1227	+ 7.1%	+2072	+11.2%
Construction	- 75	- 7.1%	+ 144	+14.6%
Manufacturing	+ 36	+ 0.9%	- 562	-13.4%
Agriculture	+ 3	+ 1.0%	- 67	-21.2%
Energy, water	- 99	-31.0%	- 13	- 5.9%
All industries	+1092	+ 4.7%	+1575	+ 6.5%

Note: all figures employees, seasonally adjusted

Source: Employer Survey, Office for National Statistics

A more detailed look at job change by industry is presented below. The figures use the standard statistical industrial classification. They can only show the 'average' change in jobs across each industry as whole. Within each industry group there will be sub-sectors doing much worse and others better than the industry average. Some classifications also bring together different industries under the same heading which may also be experiencing divergent trends.

Services

In both periods the biggest source of new jobs was business services, up 22 per cent between 1992 and 1997 and by another 20 per cent between 1992 and 1997. Large numbers of jobs were also created in the distribution and hospitality industries in both periods.

In contrast, the finance sector saw hardly any change in employment levels in both periods. However, in the past few years finance has seen more job losses following the stock-market collapse.

Important differences were in transport and telecommunications, which saw no net job creation between 1992 and 1997. In contrast, these sectors saw significant job growth since 1997.

There were also major transformations across the public sector based industries of public administration, education and health care. Education and health care had seen modest job growth before 1992, but this greatly accelerated after 1997. Public administration went from a net job shedder to a job gainer.

However, the Employer Survey makes no distinction between the public and private sectors. Both education and health care include significant numbers of jobs that are classified to the private sector. By national accounts definitions, for example, between 30 and 40 per cent of jobs shown as education services or health care services are classified as private sector employees. Estimates of public and private sector employment based on official estimates from 1992 onwards are set out below.

Service sector job growth 1992-2002

UK, June, employees	92-97	97-02
	000s	000s
Business services, real estate	+554	+ 594
Education services	+ 41	+ 321
Distribution	+370	+ 246
Health and social work services	+ 89	+ 196
Hotels and restaurants	+133	+ 188
Other personal services	+107	+ 192
Transport	- 18	+ 99
Telecommunications, post	- 4	+ 77
Public administration	- 102	+ 76
Finance	- 12	+ 9
All service industries	+1227	+2072

Note: other personal services includes refuse collection, hairdressing, dry-cleaning, recreational, cultural and sports activities.

Source: Employer Survey, Office for National Statistics.

Manufacturing

The big turn-round in manufacturing job trends is in part because of accelerated job losses in some sectors already under pressure, (eg textiles metals and ceramics, industrial machinery) and because other sectors moved from being significant net generators of new jobs (eg rubber and plastics and higher tech electrical goods) to significant job shedders. External factors linked to exchange rate movements and changes in global markets have been important. These include:

The sharp fall in the euro against the pound in 1997, making it harder to sell in Europe at a profit;

The 1998 Asian Crisis triggered devaluations in the region, increasing competitive pressures from low cost imports;

The collapse in 2000 in global markets for some high tech products and components.

This helps explain why job loss has been so highly concentrated. About two thirds of all manufacturing job losses since 1997 have come from just three groups of industries - textiles and clothing, metals and ceramics, and electrical equipment.

However, external pressures cannot be the sole explanation. The UK manufacturing sector has been vulnerable to increased competitive pressures because of a failure to address long-standing weaknesses in skills, investment and innovation. In addition, the relative ease and cheapness of plant closure makes UK employment vulnerable at a time of increased uncertainty across the global economy.

Manufacturing job losses 1992-2002

UK, employees, seasonally adjusted	92-97	97-02
June each year	000s	000s
Clothing, textiles, footwear	- 54	-171
Ceramics, metals, metal products	- 16	-131
Electrical and optical	+63	- 84
Machinery and equipment	- 25	- 52
Food, drink tobacco	nc	- 36
Rubber and plastics	+54	- 32
Paper, printing, media	+12	- 29
Chemicals	- 19	- 20
Transport equipment	- 14	- 17
Other manufacturing	+36	- 9
All manufacturing	+36	-562

Source: Office for National Statistics

Job growth in public and private sectors

The impression is sometimes given that the jobs record since 1997 depends on public sector job creation. It is true that public sector employment has grown faster than private sector employment since 1997, but more jobs have been created in the private sector than in the public sector.

Unfortunately, the poor state of the official statistics in distinguishing between public and private sector employment mean it is hard to pin down by exactly how much employment has gone up in each sector. This weakness was highlighted in the 2002 Treasury cross-cutting review of the public sector labour market.

As shown above, the Employer Survey makes no distinction between the public and private sectors. The only other sources are a direct estimate based on returns from public sector employers using national account definitions, and more recently estimates from the household Labour Force Survey. Both have weaknesses.

The national account definitions have become increasingly unrealistic as the boundaries between public and private sectors have become more blurred. They exclude many employees - including most people working in the higher education sector - who on a reasonable definition would be part of the public sector. Jobs in the old Universities had always been classified to the private sector, but between 1989 and 1993 all jobs in former polytechnics, higher and further education and sixth form colleges were also reclassified to the private sector. At the time of transfer this represented at least 173,000 jobs.

However, the Labour Force Survey may well overstate public sector employment because some individuals who said they worked in the public sector in reality worked for a private employer supplying services to the public sector. As an example of the potential discrepancy, national account definitions give a public sector workforce of around 5 million while the wider Labour Force Survey definition gives a figure in excess of 6 million. The actual figure corresponding to industrial reality is somewhere in between these two estimates.

Although the Labour Force Survey measure and the national accounts measure of public sector employment show very different levels, they have moved in line since the mid 1990s. So by using the national accounts definition we are unlikely to seriously mislead about the overall growth in public sector employment. However, there are still difficulties at the industrial level, especially for education services. The following results are based on the National Statistics estimates for public and private sector employment as published in July 2003 and using national account definitions.

Public sector jobs went up by 344,000 between 1997 and 2002 or 6.9 per cent, while private sector employment) went up by 944,000 or 4.3 per cent (workforce definition, including self-employed, HM Forces, government trainees and second jobs);

Public sector employee jobs in health and social work went up by 125,000 or 7.8 per cent; private sector employee jobs went up by 78,000 or 8.0 per cent;

Education employee jobs classified to the public sector went up by 169,000 or 14.1 per cent, while education jobs classified to the private sector went up by 56,000 or 7.2 per cent;

Public administration and social security employee jobs went up by 77,000 or 5.6 per cent;

Police jobs (including civilians) went up by 12,000 or 5.8 per cent;

HM forces went down by 6,000 or 2.9 per cent;

Employee jobs in construction, transport, energy and water classified to the public sector went down by 9,000 or 2.1 per cent, while jobs classified to the private sector went down by 221,000 or 3.5 per cent.

The increase in private sector employment in education since 1997 may reflect more people working in private education services. But it could just as easily reflect more jobs in the publicly funded higher education sector, most of which is classified to the private sector under national account definitions. Both the education and health care totals for the private sector would also include agency and supply workers working within the NHS and in state schools.

There is a huge contrast between these changes and those before 1997, when the official policy was to reduce the relative share of the public sector in the economy. Between 1992 and 1997 the public sector headcount went down by 829,000.

However, only some of this resulted in a net loss of jobs to the economy. The fall also reflects statistical reclassifications of jobs to the private sector, especially in education, as well as direct transfers of people and jobs through privatisation and contracting out. In addition, parts of the public sector became more reliant on agency and other temporary staff employed by the private sector as an alternative to filling vacancies within the directly employed workforce.

We therefore have to take account of changes in total employment in areas such as health and education. The figures show that overall job growth in education and health care services was much weaker in the five years before 1997 compared with the five years after. So even we take account of jobs moving from the public to the private sector before 1997, it is clear that the overall supply of education and healthcare services has increased much faster after 1997 compared with the five years previously. These figures are on a slightly different basis to those set out above in the section on industrial employment change, but the overall trend is the same.

Employee employment in health care and social work went up by 3.5 per cent in total between 1992 and 1997, but increased by 7.9 per cent between 1997 and 2002;

Employment in education services went up by 2.5 per cent across both sectors between 1992 and 1997, compared with a rise of 11.4 per cent between 1997 and 2002;

There are some sectors for which we have no private sector equivalent, so we cannot say if the falls in public sector employment where offset by increased private sector employment (for example, by contracting out) or vice versa (for example, services being brought in-house or replacing agency work with direct employment). The official statistics show that:

Employment classified to NHS/NHS Trusts went down by just under 3 per cent between 1992 and 1997 but went up by over 13 per cent between 1997 and 2002;

Employment in the police service (including civilians) went up by 1 per cent between 1992 and 1997, and by 5.8 per cent between 1997 and 2002;

Employment in public administration fell by 7 per cent between 1992 and 1997, but increased by 5.6 per cent between 1997 and 2002;

Employment in HM Forces fell by 27.6 per cent between 1992 and 1997 compared with a fall of 2.9 per cent between 1997 and 2002.

Employment classified to the public sector

	1992-1997		1997-2002
	000s	%	000s	%
Education*	- 197	-12.0%	+ 169	+14.1%
Healthcare	- 39	- 2.4%	+ 125	+ 6.9%
NHS	- 32	- 2.6%	+161	+13.4%
Police	+ 2	+ 0.9%	+ 12	+ 5.8%
Public admin	- 103	- 7.4%	+ 77	+ 5.6%
HM forces	- 80	-27.6%	- 6	- 2.9%
Production	- 224	-34.0%	- 9	- 2.1%
All public	- 829	-14.3%	+ 344	+ 6.9%

Notes: *most of fall between 1992-1997 is re-classification of previously local authority controlled higher education institutions to private sector in 1993. Health care includes social work. NHS includes NHS Trusts. Police is local government, including civilians. Production includes construction, transport, energy and water and other production. Some of these jobs were reclassified to the private sector by privatisation prior to 1997.

Source: Office for National Statistics.

Working time

Since 1997 fewer people are working in either very long hour or very short-term hour jobs. All the increase in employment has been in middle range part and full time jobs involving usual weekly hours of between 16 and 45 hours. In 1997 about 66 per cent of all jobs involved hours of between 16 and 45 a week. By 2002 this had increased to just over 70 per cent. In contrast, jobs with usual hours of more than 45 a week fell from 23.9 per cent of the total to 21.5 per cent.

This reverses the previous trend where most of the increase in jobs was coming from those with very long hours in excess of 45 hours a week. Between 1992 and 1997 the number of long hour jobs went up by 16 per cent, increasing the share of such jobs from 21.3 per cent to 23.9 per cent.

The rise of 'middle hour jobs' since 1997

Share of employees	1992	1997	2002
Less than 6 hours	1.6%	1.6%	1.2%
6 to 15 hours	8.3%	8.2%	7.1%
16 to 30 hours	13.6%	15.4%	17.0%
31-45 hours	55.1%	50.9%	53.2%
Over 45 hours	21.3%	23.9%	21.5%

Note: UK employees, seasonally adjusted, Spring Source: Labour Force Survey

Wages and wage inequality

Since 1997 most workers have seen significant increases in average earnings. Average earnings have increased by just over 25 per cent comparing May 1997 and May 2002. Average earnings in the private sector have gone up by slightly more (just over 26 per cent) and average earnings in the public sector slightly less at just under 22 per cent. The biggest rise was in private services, up 27 per cent. Latest figures suggest that the gap between the private and public sectors has narrowed further in the year to May 2003, but has not yet closed.

Before 1997 wage growth was less robust. Between 1992 and 1997 average earnings went up by 18 per cent compared with 25 per cent between 1997 and 2002. However, the difference between the two periods was very large for public services, where average earnings s increased by only 12 per cent between 1992 and 1997 compared with average increases of 20 per cent in the private sector.

Average earnings growing faster after 1997

Change in index	1992-1997	1997-2002
Average earnings	+18.5%	+25.2%
Public sector	+12.6%	+21.9%
Private sector	+20.0%	+26.1%
Private Services	+18.4%	+27.0%
Manufacturing	+25.2%	+22.4%

Note: All figures change in average earnings index, may of each year.

Source: Office for National Statistics

The UK labour market has overall become more unequal in terms of wage distribution for full time employees since 1997. A simple measure of wage equality is to compare earnings for those in the bottom 10 per cent with the earnings of those in the top 10 per cent. The New Earnings Survey shows hourly earnings including overtime for full time employees. In 1997 the top 10 per cent of full time employees had hourly earnings 3.55 times greater than the hourly earnings of the bottom 10 per cent. By 2002 the gap had increased to 3.66 times.

The widening gap was primarily driven by faster wage growth at the top. Low wage employees improved their position slightly against the median wage (when exactly 50 per cent get more and exactly 50 per cent get less). So overall, the bottom half of the earnings distribution became a bit more equal while the top half became more unequal.

Manual and non-manual earnings moved in different directions. Manual earnings became more equal, with the gap between the top and bottom 10 per cent borrowing. The opposite happened for non-manual wages, with the top ten per cent securing bigger pay increases than the bottom ten per cent. Indeed, it has been big pay rises for non-manual men at the top of the wage distribution that has driven the overall distribution apart since 1997.

Overall, there was a similar increase in wage inequality between the top and the bottom in the period 1992-1997. However, there was a crucially important difference. Over this period the low paid were also falling behind the median - so the bottom half of the wage distribution was also becoming more unequal before 1997.

We can therefore conclude that the low paid have done better in the five years since 1997 than in the five years before. The National Minimum Wage has clearly been one factor in improving pay at the bottom relative to the median. However, there is no corresponding policy lever to control pay at the top, hence the widening gap between the median wage and the top.

Wage inequality compared

Ratios top 10% to bottom 10%	1992	1997	2002
Manual full time	2.42	2.48	2.39
Non-Manual full time	3.62	3.67	3.82
All full time employees	3.41	3.55	3.66
Ratios median to bottom 10%
Manuals	1.56	1.57	1.52
Non-manuals	1.83	1.86	1.86
All full time employees	1.69	1.75	1.73

Note: all figures hourly earnings of adult full time employees, including overtime, April each year. Ratios are hourly earnings of top 10 per cent divided by earnings of bottom 10 per cent and median earnings divided by bottom 10 per cent.

Source: New Earnings Survey

There is little reliable earnings data over time for part time workers, mainly because the New Earnings Survey excludes many part time workers with low weekly wages. The figures we have suggest that the wage gap continued to widen for female part time workers between 1997 and 2002 comparing the top 10 per cent of part time workers with the bottom 10 per cent of part time workers. However, the growth in wage inequality between part time workers was greater between 1992 and 1997.

Men and women in the labour market

In both periods employment grew faster for women than for men. However, male ILO unemployment fell much faster than for women in both periods. As a result, the gap between male and female unemployment rates has narrowed significantly in both periods.

Between 1992 and 1997 significant numbers of men dropped out of active labour market, so a modest increase in employment produced a big fall in unemployment. Over this period male employment went up by just over 200,000, but male ILO unemployment fell by nearly 580,000. The number of men classified as economically inactive who said they wanted a job went up by just over 190,000.

In contrast, between 1997 and 2002 most of the fall in male unemployment was due to faster job growth. However, there has been no progress in reducing the number of men classified as economically inactive who said they wanted a job. Indeed, there was a slight rise in the total over this period.

The more modest fall in female unemployment has been because many of the new jobs were being taken by women re-entering the labour market. This was especially true after 1997, when big increases in female employment produced relatively small falls in female unemployment. Between 1997 and 2002 the number of women in work went up by over 740,000 but the number of women classified as ILO unemployed went down by only 143,000. However, the number of women classified as inactive but who said they wanted a job fell significantly, down by just over 150,000.

Jobs, unemployment and inactivity by gender

	1992-1997	1997-2002
Men	000s	000s
In work	+205	+549
ILO unemployed	- 579	-368
Inactive, want work	+193	+ 25
Women
In work	+533	+743
ILO unemployed	- 180	- 143
Inactive, want work	+ 60	- 151

Source: Labour force Survey

Men work fewer hours, women more hours

All of the fall in long hour jobs of over 45 hours a week was for men, with a fall of over 9 per cent. There was a slight rise in women in long hour jobs. In contrast, all of the fall in short hour jobs of less than 16 hours a week was for women, down 7 per cent, while short hour jobs held by men went up.

This will partly reflect shifts in labour demand. But it is also evidence that the government’s efforts to address the 'poverty trap' through tax-benefit reform have had some success (the poverty trap occurs when people taking on more hours are no better off due to loss of means tested benefits and higher tax). The reforms to date - such as the introduction of the Working Family Tax Credit - are likely to have had most impact on low paid women with families.

Gender wage gap continues to narrow

Overall, the wage gap between men and women in full time work has narrowed slightly. In 2002 women’s median hourly earnings (the hourly rate where exactly 50 per cent earn more and 50 per cent earn less) were 85 per cent of male median hourly earnings. This compares with 83 per cent in 1997. However, a similar rate of progress was recorded between 1992 and 1997. In 1992 female median earnings were just under 80 per cent of male median earnings.

Black workers

In this report the term black workers refers to all ethnic minorities as defined by the Census and the Labour Force Survey.

We can say relatively little about either current position of black workers or how that position has changed until new statistical estimates are available in the Autumn taking into account the results of the 2001 Census. In addition, the Labour Force Survey introduced a new system of ethnic classification from 2001 onwards consistent with those used in the Census that makes direct comparisons with earlier periods more difficult.

What we can say for certain is that unemployment among ethnic minorities has fallen significantly since 1997. Between 1997 and 2000 on the old definitions unemployment among ethnic minorities fell from 15 per cent to 12.2 per cent. Between Spring 2001 and Spring 2002 under the new definitions the black unemployment rate dropped from 11.1 to 10.7 per cent.

We have no evidence to suggest that the gap between white and ethnic minority unemployment has closed since 1997. The ratio shows some volatility from year to year, reflecting the small sample size for black workers within the Labour Force Survey. Since 1997 the black unemployment rate has varied between 2.3 and 2.5 times the white unemployment rate.

Between 1992 and 1997 the unemployment rate for black workers also fell, from 17.5 per cent to 15 per cent, but the white unemployment rate fell faster. As a result, the ratio widened somewhat from just under 2 to 1 in 1992 to just over 2 to 1 by 1997. The table below sets out the evidence available at the time of writing from the old and new classifications of ethnicity.

Unemployment by ethnic origin 1997-2002

Old series	White	Ethnic minorities	Ratio
Spring 1992	9.2%	17.5%	1.90
1997	6.6%	15.0%	2.27
2000	5.1%	12.2%	2.39
New series
Spring 2001	4.4%	11.1%	2.52
Spring 2002	4.7%	10.7%	2.28

Note: main groupings under the new Census 2001 classifications are White, Mixed, Asian or Asian British; Black or Black British; Chinese; other ethnic groups.

Sources: Labour market participation of ethnic groups, Labour Market Trends, January 2001; Labour Force Survey Quarterly Supplement, August 2002.

Latest figures for Spring 2003 suggest little change over the past year. The all-ethnic minority unemployment rate edged up slightly, but given the small sample size this could easily be a blip. The ratio between black and white unemployment rates was just under 2.5 to 1, much the same as in 2001 under the new series.

The statistical gaps mean our conclusions on this section are more tentative. In both periods unemployment rates for black workers fell significantly. The relative position of black workers worsened between 1992 and 1997 comparing the ratio between white and black unemployment rates. The relative position appears to have stabilised since 1997 but there is no sign the gap between black and white unemployment rates is closing.

Conclusions

Our key conclusion is straight-forward - on both quantity and quality measures the labour market has done significantly better in the five years since 1997 compared with the previous five years. We have seen many more jobs created and serious inroads being made into long-term unemployment and in bringing the excluded back into the labour market. Wage growth has been faster and the gap between the public and private sectors has significantly closed.

We have also seen significant increases in employment in key areas such as the NHS and education. This is in stark contrast to the pre 1997 policy agenda.

There are areas were progress has been more limited, notably wage inequalities between the top and the bottom and between men and women and in the relative unemployment position of black workers. Not all groups of excluded workers have been successfully brought back into the active labour market. Yet even here there has for the most part been an improvement over the 1992-1997 period. The numbers may not be moving as fast as we would like them to, but at least they have either stopped getting worse or are moving in the right direction.

There are some parts of labour market performance were there has been no improvement. The biggest single weakness since 1997 has been the almost continuous attrition of manufacturing jobs. The concentration of job loss in particular industries suggests that external factors such as unfavourable exchange rate movements and changes in global markets have played a significant role. But this cannot be a complete answer and there is clearly a major policy challenge here that the government needs to address.

Stay Updated

Want to hear about our latest news and blogs?
Sign up now to get it straight to your inbox

Join a union