Recession Report June 2009 - Comparison of previous recessions

PART ONE: This month's figures

The latest employment figures cover the three months to April 2009, and show:

• 29,108,000 people in work, 271,000 fewer than the previous 3 months and 399,000 fewer than the same period a year earlier. This quarterly fall is the largest on record (with data starting in 1971). The reduction in male employment was 156,000, while the reduction in female employment was 117,000.
• A working age employment rate of 73.3 per cent, down 0.8 percentage points on the previous quarter and 1.5 points on the same period in 2008. This rate is comparable with those from mid-1998.
• Unemployment at 2,261,000 (1,376,000 men and 885,000 women), up 232,000 on the quarter and 605,000 on the year.
• An unemployment rate of 7.2 per cent, up 0.7 points on the previous quarter, 1.9 on the year. Rates were last this high in mid-1997.
• Figures to the end of May 2009 show:
• ·424,000 job vacancies, 18,000 down on the previous month and 235,000 down on the year.
• ·1,544,800 people claiming JSA, an increase of 39,300 on the month and 726,100 on the year.

The rate of increase in unemployment may be beginning to fall: ILO unemployment increased by just over 2 per cent between February and March 2009, which is one of the lowest monthly rates of increase since Summer of last year. Lessons from previous recessions suggest it is too soon to be optimistic. In the 1990s, the rate of monthly unemployment increase peaked one year into the downturn, but unemployment continued to rise for close to two years. In the 1980s, it took longer for rates of unemployment increases to peak (18 months), after which levels rose for over three years.

Pay and earnings

The latest earnings data are for the three months to April; the annual increase (excluding bonuses) is 2.7 per cent - the lowest increase since the figures started distinguishing between increases with and without bonuses in 2001.

Average earnings (excluding bonuses), Sept 2001- April 09[1]

Last month's data for earnings including bonuses showed a fall (0.3 per cent), compared with March 2008, the only time this has been recorded since 1991. This month's figures show an improvement, but this year's figures are all substantially lower than the figures for 2008:

Earnings (inc. bonuses) Jan 2008 - April 09

Average earnings growth is now significantly stronger in services than in manufacturing:

Earnings growth in services and manufacturing, May 04 - Feb 09

Average earnings growth is also higher in the public sector (3.5 per cent, excluding bonuses) than the private sector generally (2.6 per cent) or private sector services specifically (2.9 per cent).

With inflation low and jobs scarce, it should not surprise us that, on average, pay deals are lower than the increase in earnings and on a downwards trajectory. Incomes Data Services report that the median for settlements in the three months to April was 2 per cent, down from 3 per cent the previous month. The Labour Research Department looks at figures slightly differently, and their database shows the median increase for lowest basic rates rather higher, at 2.5%; IRS on the other hand has a median figure of 1.5% for the whole economy.

In other words, no one should be surprised that pay deals are coming down, but media claims that most workers face a wage freeze or even a cut are still inaccurate.

Young people

• Analysis shows that during February and March of this year young people's unemployment rates increased more quickly than the national rate. This pattern may continue into future months, but given the volatility in the monthly rates of unemployment increase it is too early to tell.

Monthly rates of increase in unemployment levels, Q1 2008 - Q1 2009

	Overall rate of monthly increase	18-24 monthly rate of increase
2008 Qtr1
Jan	0.4	0.0
Feb	0.6	2.2
Mar	2.0	0.2
2008 Qtr2
Apr	-1.7	-2.1
May	3.5	4.8
Jun	2.5	3.2
2008 Qtr3
Jul	3.8	2.9
Aug	1.8	3.6
Sep	2.1	3.1
2008 Qtr4
Oct	3.2	2.8
Nov	2.5	0.3
Dec	2.9	0.8
2009 Qtr1
Jan	3.5	1.6
Feb	5.5	7.1
Mar	2.1	2.8

Self-employment

One of the interesting developments in the labour market since the mid-1990s has been the growth of self-employment. The number of self-employed workers and the share of all employees they represented were very stable until 2003. In that year self-employment grew very quickly for about six months and has been growing more slowly ever since:

Self-employment, May 1992 - Jan 2009

If we concentrate our attention on what has been happening in the last year, we can see that the number of people in self-employment fell in the first half of 2008, but has risen for most of the last six months. The change has been small, however, and the number of self-employed workers has still not regained the level of 12 months ago.

Self-employment in the recession, May 08 - Apr 09

Temporary and part-time work

The number of temporary workers has been declining since 1997; at the same time, the proportion of temporary workers who are working in temporary jobs because they could not find permanent jobs has been falling, a process that began even earlier:

Temporary work, May 1992 - Jan 2009

These processes have reversed in the recession - the number of temporary workers and the proportion who are involuntary are both rising:

Temporary work in the recession, Jan 08 - Apr 09

The number of part-time workers has been growing since 1997; the overall labour force has been growing as well, so the proportion of workers who are part-time peaked in 2003.

Part-time employment, Mar - May 1997 - April 09

If we look at what has been happening over the last year, we can see that the number of part-time workers has been falling for 6 months. This fall has been slower than the decline in overall employment, so the proportion of employees who are working part-time has been rising:

Part-time employment in the recession, Mar - May 08 -

Unsurprisingly, the recession has seen an increase in involuntary part-time work:

Proportion of part-time workers who could not find a full-time job, Jan - Dec 2008

Longer term unemployment

We are starting to see large increases in the number of people who are unemployed for six months or more. The following tables show the levels and rates of increase in unemployment of up to 6 months, 6-12 months and over 12 months since the first quarter of 2008. They demonstrate that rates of increase in unemployment of over 6 months are significantly higher than the rates of increase in unemployment of 6 months or less, and that the rate of quarterly increase in unemployment of over 12 months is rapidly rising.

Levels of increase (000s) and rates of increase in unemployment by up to 6 months and 6-12 months unemployed, Q1 2008 - Q1 2009

Date	Up to 6 months unemployed		6-12 months unemployed
	Levels	Rate of quarterly increase	Levels	Rate of quarterly increase
2008 Q1	959	-0.5	266	4.7
Q2	994	3.6	276	3.8
Q3	1106	11.3	284	2.9
Q4	1202	8.7	316	11.3
2009 Q1	1333	10.9	376	19.0

Levels of increase (000s) and rates of increase in unemployment by over 12 months unemployed, Q1 2008 - Q1 2009

Date	Over 12 months unemployed
	Levels	Rate of quarterly increase
2008 Q1	399	3.9
Q2	415	4.0
Q3	435	4.8
Q4	453	4.1
2009 Q1	506	11.7

Recently released research[2] from the Institute or Leadership and Management found that a quarter (28 per cent) of employers said they would be less likely to hire someone who had been unemployed for six months or more.

PART TWO: Comparing recessions

With the recession having now run for at least a year this Recession Report is probably a good point to take stock and reflect on the shape and progress of the downturn.

A common definition says that a recession begins at the start of two successive quarters of 'negative growth'[3]. In January, there was a lot of press coverage for the news that the UK had 'officially' been experiencing recession, as a result of the release of the data for the third and fourth quarters of 2008. In fact, a revision in the data for the second quarter showed a slight decline in GDP (less than 0.1 per cent), but most commentators have decided to stick with a third quarter starting point. We have based our analysis of the recession on the data from the second quarter of 2008, as this was the point at which unemployment[4] started to rise.

This means that we now have a year's data for the current recession and it is possible for us to ask how this recession compares with previous recessions. In particular, how will the 'Labour recession' of the 2000s compare with the 'Tory recessions' of the 1980s and 1990s?

In this section of the report we start by considering GDP during previous recessions before looking at how past downturns have affected both employment and unemployment rates as well as levels of economic inactivity. Our analysis then looks at unemployment by gender and by region, and at how different industrial sectors have fared in recessionary periods. We consider secondary evidence on the impacts that recessions have for particular groups, and how pay and working time can be affected. We then conclude with an assessment of what the historical data tell us about the possible impacts that the current recession will have.

GDP and previous recessions

In the charts that follow, we look at the recessions of the 1980s and 1990s and the current downturn. We show what happened from the last quarter before the unemployment rate for people of working age started to rise; this does not always follow the 'two successive quarters' definition, but it does help us to get a more complete picture of economic change during that recession.

We have used the same GDP index in each chart - each shows the value of GDP as a proportion of GDP in 2003, uprated to take account of inflation. The line (measured against the left-hand scale) shows what the GDP index was for that quarter; the bars (measured against the right-hand scale) show how much it had changed, compared with the previous quarter. When the bars are above zero and the line is rising, GDP is growing; when the bars are below zero and the line is falling, then GDP is declining. Where no bar is visible GDP quarterly change was negligible.

The charts for the 1980s and 1990s both show that, using the 'two successive quarters' definition, both recessions lasted about 18 months, but with some stuttering before the economy was fully recovered. On the other hand, if we ask how long it took for the economy to recover to the level of output it had achieved at the start of the recession, we get a timescale of about three years. This is an important point: 'green shoots', a.k.a. rising GDP, can be seen in the third quarter of 1981 and the first quarter of 1992, but in each case this was more like a half-way mark than the beginning of the end.

GDP in the 1980s recession, Q3 1979 - Q4 1984

GDP in the 1990s recession, Q2 1990 - Q3 1993

What happened in the past is not a way to predict the future, but we can see that, although today's recession is well-established, it has not yet lasted as long as its two predecessors.

GDP in the current recession, Q1 2008 - Q1 2009

According to the Bank of England, output loss in the 1980s recession was 4.6 percent of GDP, and in the 1990s, it was 2.5 percent.[5] ONS data shows that GDP decreased by 0.7 per cent in the third quarter of 2008, 1.6 per cent in the fourth quarter and 1.9 per cent in the first quarter of 2009. The Treasury's round-up of new GDP forecasts shows that, on average, independent forecasters predict that, for 2009 as a whole, GDP growth will be -3.8 per cent. [6] This depressing picture is confirmed if we take each recession as starting in Q1 and follow what happened to GDP growth in succeeding quarters, as is shown below.

GDP across three recessions, analysis from the start of each recession

This is a very severe recession; if previous recessions are anything to go by it will be some time before output returns to the level the economy had achieved in the first quarter of 2008.

Unemployment, employment and the recessions

In the charts in this section, we continue to use the 2003 index for GDP, with the scale on the left hand side. But we also show unemployment rates, with the scale on the right hand side.

In both the 1980s and 1990s recessions, unemployment went on rising after GDP had begun to recover, but there was a very significant difference between these two experiences. In the 1990s, unemployment had begun to come down even before GDP had reached its pre-recession level. In the 1980s, by contrast, unemployment went on rising for over a year after GDP had completely recovered.

Unemployment and GDP in the 1980s recession, Q3 1979 - Q4 1984

Unemployment and GDP in the 1990s recession, Q2 1990 - Q3 1993

ONS analysis[7] shows that after the 1980s recession, unemployment levels and rates did not return to their pre-recession position at any point before the next recession in the 1990s. And it was seven years after the start of the 1990s recession that unemployment levels returned to pre-recession levels.

There was a significant difference in the scale of unemployment in the two recessions. In the first quarter of 1980, unemployment stood at 5.9 per cent, it peaked at 12 per cent (more than twice as high) in the first and second quarters of 1984. In the third quarter of 1990, the unemployment rate was 7.2 per cent; it peaked in the first quarter of 1993 at 10.8 per cent.

In the second quarter of 2008, unemployment was lower than at the start of either of the other recessions considered: 5.6 per cent.

Unemployment and GDP in the current recession, Q1 2008 - Q1 2009

In the first four quarters of rising unemployment, the current recession saw a larger increase in the unemployment rate than either of the preceding recessions. This can be seen in the following table.

Percentage increases in the first four quarters of rising unemployment during recessions

First quarter	Unemp. Rate	Fourth Quarter	Unemp. Rate	% increase
1979 Q4	5.6	1980 Q3	7.2	29%
1990 Q3	7.2	1991 Q2	8.8	22%
2008 Q2	5.6	2009 Q1	7.3	30%

Claimant count in and out flow data are only available from the first quarter of 1989, so a comparison can only be drawn between the current recession and the 1990s downturn. Analysis shows that while people are still moving off benefits and into work, the gap between the number of new claimants and the number of claimants moving into jobs is wider than at any point during the 1990s: between Q4 1990 and Q1 1991 the gap between the in and out flow was 78,700, but between Q4 2008 and Q1 2009 the gap was 91,900.

UK claimant count, standardised in and out flows Q1 1989 - Q1 2009 (000s)

A rather more encouraging comparison, however, is of employment rates. In the table below, we start in the second quarter of 1979, the second quarter of 1990 and the first quarter of 2008, and compare what happened to employment rates. It is still an early point in the current recession, but so far, employment rates have held up a little better in the current recession than they did in the previous two:

Employment in three recessions, analysis from the start of each recession

Inactivity rates

Changes in 'economic inactivity' (people who are neither employed, nor unemployed, such as students, people with caring responsibilities and disabled people) in response to recessions are less noticeable than changes to employment and unemployment. Historically, women have been more likely to be economically inactive than men, but the gap has been narrowing for a long time:

Economic inactivity rates, Mar- May1992 - Jan - Mar 2009

If we magnify the scale, we can see that inactivity levels changed very differently in the past two recessions:

Economic inactivity in three recessions, analysis from the start of each recession

As the ONS's study of The Impact of the Recession on the Labour Market noted, the spike in inactivity in the 1980s recession came more than a year after GDP stopped falling. The 90s recession, by contrast, saw a long-term increase in inactivity. The ONS judgement is that 'the first three quarters of this recession have been more consistent with the 1980s recession than the 1990s.'

Gender and previous recessions

During the 1980s recession, women entered the downturn with a higher unemployment rate than men (6.2 per cent for women during Q3 1979 compared to 4.8 per cent for men during the same period). But as the downturn progressed, male unemployment peaked slightly higher than female unemployment (12.2 per cent in Q3 1983 for men compared to 11.8 per cent in Q2 1984 for women) and by the first quarter of 1990, when unemployment was at its lowest rate since the end of the 1980s recession, male unemployment was at 7.1 per cent while women were experiencing lower unemployment rates of 6.7 per cent. This can be seen in the following chart.

Unemployment rates for women and men, Q3 1979 - Q4 1984

EHRC, DWP and GEO analysis[8] suggests that this trend is likely to be a result of men's concentration in industries that were more severely affected by the 1980s recession, such as manufacturing, while women were concentrated in sectors such as education and health that are more resilient to economic cycles.

Women's unemployment rates have remained below those of men ever since, and the 1990s recession had a much greater impact on men than on women. While women's unemployment rates rose 1.3 percentage points over the entire downturn, with unemployment peaking at 8 per cent, unemployment rates for men rose by 5.6 percentage points, and male unemployment peaked at 12.7 per cent. This can be seen below:

Unemployment rates for women and men, Q2 1990 - Q4 1993

This time around the gender impact looks likely to be different from both previous recessions. More men than women are losing their jobs, and male unemployment is also increasing at a faster rate than female unemployment. This can be seen below.

Unemployment rates for women and men, Q1 2008 - Q1 2009

But there are also indications that both men and women will be hit hard by the downturn as over the first five quarters of rising unemployment the increases for women have been slightly greater than those experienced in either the 1990s or the 1980s. The increases for men have been comparable with the 1980s.

Percentage point quarterly increases in unemployment rates for women and men in the first four quarters of rising unemployment during each recession

	1980s women	1990s women	2000s women
Q1-Q2	0.1	0	0.1
Q2-Q3	0.2	0.2	0.3
Q3-Q4	0.3	0.3	0.3
Q4-Q5	0.5	0.3	0.6
	1980s men	1990s men	2000s men
Q1-Q2	0.2	0.3	0.2
Q2-Q3	0.3	0.5	0.5
Q3-Q4	0.7	0.7	0.6
Q4-Q5	0.9	1	0.9

During previous recent recessions women's employment has recovered much more quickly than male employment.

Employment levels by gender, Q1 1971 - Q1 2009

While women and men's employment levels are both being affected by this downturn, male employment has fallen by a greater amount than female employment. But so far employment rates have been affected less, for women and for men, than in the 1990s. Men's employment has been affected relatively slightly less than in the 1980s, while women's employment rates are affected slightly more.

Levels of reduction (000s) in employment for women and men in the first four quarters of rising unemployment during each recession

	1980s women	1990s women	2000s women
Q1-Q2	61	8	16
Q2-Q3	20	-25	-23
Q3-Q4	0	-60	-13
Q4-Q5	-10	-57	-33
	1980s men	1990s men	2000s men
Q1-Q2	-10	-42	-10
Q2-Q3	-50	-125	-76
Q3-Q4	-60	-160	-33
Q4-Q5	-130	-197	-123

Particular groups and previous recessions

Data are not readily available to enable unemployment and employment rates to be compared across recessions by characteristics including age, disability, ethnicity and parenthood. For example, Labour Force Statistics can only be disaggregated by age from 1992. In this section we therefore rely on secondary evidence to consider how particular groups are affected across recessions.

A recent DWP literature review[9] concluded that disabled people, BME workers and less skilled workers all experienced an increase in, and longer duration of, unemployment during previous recessions. The review also found that while in the 1980s lone parents' decisions about entering employment were influenced by non-recessionary factors (for example the age of their children) recent increases in the lone parent employment rates may mean that this group are now more likely to be affected by economic downturns.

But joint DWP, EHRC and GEO analysis[10] suggests that so far unemployment rates for BME groups and disabled people have increased by less than the national average, and that employment rates for BME groups have increased. While these groups entered the recession at an employment disadvantage, these findings provide some hope that employment gaps will not significantly widen during this downturn.

The report also provides analysis of the position of lone parents: over the last twelve months this group have experienced rising employment rates (2.5 percentage points), but also sharp rises (1.6 percentage points) in unemployment.

Recent IFS analysis[11] has considered the impact of previous recessions on living standards. A key finding is that those with less education appear to have suffered slightly more during previous recessions: while those who left education at 21 or older saw their incomes stagnate or increase slightly over the course of the 1980s and 1990s recessions, during the same period those who left earlier saw their incomes fall or stagnate.

IFS's analysis of this recession indicates that we may see a similar pattern, with workers who have the lowest levels of qualifications currently experiencing the highest rates of increase in unemployment: in October 2008 unemployment rates for this group were two percentage points higher than they were in October the previous year, while for graduates the rate of unemployment growth was less than half as fast.

Research also demonstrates that across recessions, the most deprived areas, which have higher than average concentrations of residents experiencing additional disadvantages, are affected more than average. For example, the DWP, EHRC and GEO note that the employment rate in deprived areas now stands at 63.8 per cent, almost ten percentage points below the overall rate. The ILO unemployment rate in deprived areas is now at 11.3 per cent, up from 8.7 per cent a year earlier. This is a rise of 2.6 percentage points, compared to a rise of 1.9 percentage points among the overall population.

Pay and working time across recessions

ONS have undertaken analysis of real earnings growth across recessions[12], deflating earnings by the RPI to chart real earnings over time. They show that across recessions real earnings growth has generally been negative during the first two quarters of the recessionary period before returning to positive growth. Their analysis shows that the current recession is also following this pattern, as is shown below in their reproduced chart (recessionary periods are highlighted with the shaded vertical bars).

IDS[13] have found that a key feature of the current recession, in contrast to the 1990s, is that few organisations have spare layers or capacity to cut. Their analysis indicates that this recession has seen an increase in use of short-time working and sabbaticals (accompanied by pay freezes) as a means to retain skills and limit redundancies.

Regional unemployment rates across recessions

Unemployment data are only available by regions from 1992, but regional claimant count data are available from the mid-1970s. While these data are not accurate indicators of unemployment levels over the last three recessions, and do not accurately reflect trends by gender, they do provide an important means to consider how regions were affected during previous downturns.

Previous recessions have had extremely different impacts across the UK regions. In the 1980s, the national claimant count rate peaked at 10.3 per cent. But in many regions the rate was far higher. For example, the highest rate of 15.7 per cent was in the North East. This can be seen in the following table.

Claimant count rates across UK regions in Q3 1979 and Q4 1984

Regions	1979 Q2	1984 Q4	Percentage point increase
UK rate	3.7	10.3	6.6
London	2.4	8.1	5.7
North West	4.5	12.7	8.2
South East	2.2	6.7	4.5
East	2.5	7.5	5.0
Yorkshire and Humber	3.7	11.3	7.6
North East	6.4	15.7	9.3
West Midlands	3.7	12.1	8.4
East Midlands	3	9.5	6.5
South West	3.6	8.6	5
Wales	4.8	12.9	8.1
Scotland	5.4	12.1	6.7
N. Ireland	7.8	15.4	7.6

During the 1990s regional differences remained, but were not as extreme. There were also differences in which regions were most badly affected, with London, the South West and the South East seeing the largest percentage point increases in claimant count rates. This was mainly because rates in these regions had shown a greater recovery since the 1980s downturn, whereas in areas that were badly hit in the 1980s (for example the North East and Northern Ireland) claimant count rates were already relatively high at the beginning of the 1990s recession.

Claimant count rates across UK regions from Q2 1990 to Q2 1993

	Q2 1990	Q2 1993	Percentage point increase
UK rate	5.2	9.8	4.6
London	4.4	11	6.6
North West	6.8	10.1	3.3
South East	2.5	8.3	5.8
East	3.0	9.0	6.0
Yorkshire and Humber	6.1	9.8	3.7
North East	8.9	12.2	3.3
West Midlands	5.2	10.5	5.3
East Midlands	4.6	9.1	4.5
South West	3.7	9.1	5.4
Wales	6.2	9.8	3.6
Scotland	7.7	9.2	1.5
N. Ireland	12.8	13.7	0.9

Claimant count data from this recession demonstrate that regional differences are beginning to emerge again, with areas including the North East, the West Midlands and Yorkshire and Humber seeing relatively large percentage point increases in claimant rates.

Claimant count rates across UK regions from Q1 2008 - Q1 2009

	2008 Q1	2009 Q1	Percentage point increase
UK rate	2.4	4.2	1.8
London	2.7	3.8	1.1
North West	3	4.9	1.9
South East	1.4	2.9	1.5
East	1.9	3.6	1.7
Yorkshire and Humber	2.9	5.1	2.2
North East	3.9	6.4	2.5
West Midlands	3.4	5.7	2.3
East Midlands	2.4	4.4	2
South West	1.3	3	1.7
Wales	2.7	5	2.3
Scotland	2.5	4	1.5
N. Ireland	2.7	4.8	2.1

The following charts shows claimant count rates across the UK regions and nations from the mid-70s until the last calendar quarter. They demonstrate the large regional differences in rates that the UK has previously experienced. For example, in 1986 Q4 claimant unemployment reached 17.1 per cent in Northern Ireland, but was only 6.6 per cent during the same period in the South East.

Claimant count rates across the four UK nations, Q2 1974 to Q1 2009

Claimant count rates across Yorkshire and Humber, North East, North West and London, Q2 1974 to Q1 2009

Claimant count rates across East Midlands, West Midlands, South East, East and South West, Q2 1974 to Q1 2009

The charts also show that regional differences in claimant rates had fallen dramatically during the 1990s and early 2000s, and that we entered this recession with small regional claimant rate differences that were comparable to the mid-70s. Indications that this recession may lead to the re-establishment of extreme regional variations in unemployment is therefore concerning, as it may suggest that the recent reductions in regional claimant count differentials are being reversed.

Industrial sectors and previous recessions

National vacancy and redundancy data are only available by sector from 1995 and 2001 respectively. But statistics on workforce jobs by industry are available from 1978. We therefore consider the impact of previous recessions by industrial sector using these data.

Recent recessions have had very different impacts across industrial sectors. During the 1980s the number of jobs in sectors including finance and business services and education and health and public admin remained stable, and returned to growth relatively quickly. In others there were sharper falls (for example in distribution, hotels and restaurants) followed by sharp recoveries. Manufacturing was affected far more than other sectors: the recession accelerated the decline of manufacturing jobs, which did not stabilise (briefly) until the late 1980s.

During the 1990s, manufacturing jobs showed another steep decline, and construction was more badly affected than in the 1980s. There were also falls in workforce jobs in finance and business services and distribution, hotels and restaurants. Again, by the late 1990s all sectors except manufacturing were showing jobs growth.

During the current recession emerging trends for distribution, hotels and restaurants, transport and construction appear similar to previous recent downturns. Finance and business services are however showing steeper declines than previously, suggesting that the overall size of the sector may shrink as a result of the recession. The ongoing decline in manufacturing jobs has been accelerated by the downturn. While education, health and public admin jobs have remained relatively stable, future public spending cuts could put jobs in this sector in jeopardy. These trends can be seen in the following charts.

UK workforce jobs by industry, Q2 1974 to Q1 2009

Manufacturing workforce jobs across recessions, analysis from the start of each recession

Finance and business services workforce jobs across recessions, analysis from the start of each recession

Distribution, hotels and restaurants workforce jobs across recessions, analysis from the start of each recession

Conclusion

Comparing the current downturn with those of the 1980s and 1990s shows us that we are in a severe recession, with initial falls in output steeper than in those in either previous recessionary period. The current recession is more like the deep 1980s recession than the shallower 1990s downturn. Evidence from the past also suggests that even when the economy begins to grow again it could be several years before we return to pre-recession levels of growth.

Unemployment increases were far greater in the 1980s than the 1990s, but in both recessions unemployment levels continued rising for at least a year after GDP started to increase and remained above pre-recessionary levels for years to come. During the first year of this recession unemployment rates have risen at a comparable speed to the 1980s. Even if the rate of increase begins to slow, it therefore looks set to remain high in the medium as well as the short term.

Comparisons with previous recessions provide a mixed picture of the relative impacts for groups facing labour market disadvantage. As we have previously reported, men are losing their jobs faster than women, but women look set to experience greater rises in unemployment than in the 1990s. While detailed analysis of unemployment data by ethnicity and disability across recessions is not possible, there may be some hope that existing disadvantages these groups experience will not be as badly exacerbated by this downturn as they have been previously. But emerging evidence does suggest that, in common with previous recessions, the lowest skilled and those in the most deprived areas are likely to feel the effects of this downturn the most. Regional impacts are also beginning to emerge, with concerning signs that we could be heading for a return to large regional variations in unemployment rates.

So far the impacts of this recession across industrial sectors are comparable with previous downturns. This is particularly bad news for manufacturing, which looks set to experience another sharp decline in jobs. The exception to this trend is finance and business services, which is experiencing steeper jobs reductions than the 1980s or 1990s.

Overall, the key lesson from the past is that the effects of recessions are felt for years, through economic restructuring, significant changes in unemployment rates and changes in regional distributions of jobs. Historical comparisons do not provide a basis for accurate predictions, but do suggest that whenever it ends, this recession will continue to impact on our labour market for many years to come.

Notes

---

[1] LFS data, seasonally adjusted, excluding bonuses. Data are annual increase (%) up to month in question.

[2] ILM (2009) Bouncing Back: attitudes towards unemployment London: ILM.

[3] There is no consensus about how to define a recession. In the USA, recessions have 'official' starts and ends, decided by the National Bureau of Economic Research's Business Cycle Dating Committee. The Committee makes a judgement based on a range of economic data; this is very useful for authoritative historical comparisons, but it will usually take time, leading to uncertainty for policy-makers. On the other hand, the two quarters definition depends on the measurement of GDP and revisions of official data can move an economy into or out of recession. Even more difficulty can be caused when (as often happens) there is a quarter of weak growth in the midst of a recession - economists commonly get round this difficulty by ignoring it.

[4] Using the ILO definition.

[5] Inflation Report, Bank of England, Feb 2009, p 20.

[6] Forecasts for the UK Economy, HMT, May 2009, p 3.

[7] ONS (2009) The impact of the recession on the labour market Newport: ONS.

[8] EHRC, DWP and GEO (2009) Monitoring the impact of the recession on various demographic groups.

[9] Stafford B and Duffy D (2009) Review of evidence on the impact of economic downturn on disadvantaged groups London: DWP.

[10] EHRC, DWP and GEO, (2009) Ibid.

[11] Muriel A and Sibieta L (2009) Living Standards During Previous Recessions: IFS Briefing Note BN85 London: IFS.

[12] ONS (2009) Ibid.

[13] IDS (2009) IDS Focus on Recession 2009 London: IDS.