How has the UK labour market fared in recessions and recoveries since 1870?

A look back at the past 150 years of booms and slumps in the UK economy reveals shifts in how the labour market responds. While the government’s furlough scheme has averted much higher joblessness, it is likely to hamper recovery in output and reduction in the unemployment rate.

The current rate of unemployment in the UK is 4.9%, which is a four-year high. With the Coronavirus Job Retention Scheme scheduled to end in September, there are concerns that unemployment may increase further if firms are unable to keep their furloughed workers, who currently number 4.7 million people. A longer-term view can provide perspective on current rates and help us understand how best to respond to the potential for rising joblessness.

Unemployment over 150 years

Figure 1 shows annual unemployment rates for the UK on a comparable basis back to 1870. The chart picks out the high unemployment rates between the two world wars and the strikingly low rates from the late 1940s to the mid-1970s. Since then, unemployment rates have been significantly higher and more volatile – closer to the average experience from 1870 to 1940.

Figure 1: UK unemployment rate, 1870-2020

Sources: 1870-1978: Boyer and Hatton, 2002; 1979-2020: Office of National Statistics, series LF2Q.

It is sometimes suggested that while recessions raise unemployment sharply, recoveries are long and drawn out. There is some support for this idea. Over a century and a half, year-to-year decreases in unemployment have outnumbered year-to-year increases (53% compared with 47% of years). Accordingly, the average rise in the unemployment percentage has exceeded the average drop. But from 1974 to 2019, a greater percentage of years have been spent in recovery: 59% versus 41%.

Unemployment and output

Unemployment fluctuates with economic output – the production of goods and services. We can see this by looking at the relationship between changes in the unemployment rate and percentage changes in GDP per capita – what is known as Okun’s law (Ball et al, 2017). The larger (more negative) is the Okun’s Law coefficient, the more unemployment responds to a change in output.

As Table 1 shows, the Okun’s law coefficient is highest when labour supply constraints are least binding. In the interwar period when unemployment was high, it responded strongly to variations in output; in contrast, in the early post-war years, unemployment responded much less.

Table 1: Okun’s law coefficients

	1871-1913	1921-1939	1947-1973	1974-2019
Coefficient	-0.30	-0.58	-0.14	-0.33
Standard error	(0.04)	(0.14)	(0.05)	(0.06)

Note: Author calculations using the unemployment series in Figure 1 and GDP per capita from the Maddison database.

The Okun’s Law coefficient for the years since 1973 is similar to that before 1914. But there is also some evidence that from the 1920s onwards, unemployment has been more responsive to changes in GDP when there is a recession (when GDP growth falls below its long-term trend) than in recovery (when it rises above trend).

In the period from 1974 to 2019, the coefficient is -0.45 in recessions and -0.27 in recoveries; whereas before 1914, the difference is much smaller: -0.34 compared to -0.28.

Labour markets in recoveries

There are three main reasons why employment (and unemployment) is less responsive to output in economic recoveries:

The industries and regions that grow rapidly during recoveries are not those that lost most jobs during the recession, leading to excess demand in sectors that are on a growth trend and excess supply in those undergoing long-term decline.
Workers laid off in deep recessions lose job-specific skills (and often the motivation to regain employment), making it harder to match unemployed workers with the new vacancies on offer in the recovery.
In downturns, firms tend to retain more of their skilled workers than are strictly required so that in the recovery, this excess capacity declines and output grows faster than employment.

Changes in the structure of the economy and the labour force

The most dramatic structural change in the economy over the long run has been the decline in heavy industries that are intensive in manual labour and the rise of the services sector, which includes information and communications, tourism and hospitality. For example, manufacturing, mining and construction fell from 43.9% of total output in 1950 to just 18.5% in 2010.

As employment in services – and especially in the ever-expanding government sector – tends to fluctuate less over the business cycle, this should have made the economy more stable. But since the 1970s, it has not done so.

On the supply side, there are three main features:

First, the share of women in the labour force increased from 30% in 1911 to 47% a century later.
Second, over the century, the share of part-time workers grew from a small (but unknown) proportion to one-third.
Third, what economists call the ‘human capital’ of workers – their skills and experience – has grown in importance, especially in the form of general skills. This is reflected in the increase in average years of education from about seven to 12.

These trends have implications for labour market responses in booms and slumps.

The incidence of unemployment in major recessions

Looking at the structure of the economy and the composition of the labour force together reveals significant changes in the pattern of recoveries. While the interwar period was characterised by the entry of more women into work and the growth of part-time employment, in the years after 1973, the key features have been part-time work and the increasing importance of workers’ skills.

The interwar recessions were felt most severely by the textiles, mining and heavy industries that were located in the north where unemployment rates were double those in the south. Less skilled and adult men, whose unemployment rates were far higher than those of women and young workers, were also significantly affected.

The fall in unemployment during the 1930s was slowed by a mismatch between the type and location of new jobs, which left a legacy of long-term unemployment. By the mid-1930s, more than a quarter of unemployed men had been jobless for more than a year (Crafts, 1987).

In contrast, in the aftermath of the global financial crisis of 2007-09, the shock to employment fell somewhat more evenly across the country. But as in other recessions since the 1970s, unemployment fell most heavily on younger workers aged 16-24, as employers held on to more experienced staff in order to preserve their stocks of human capital (UK Commission on Employment and Skills, 2014).

The advance of public welfare

Before 1911, income support for the unemployed was negligible. The nascent welfare state of the interwar period came just in time to avert extreme poverty among many households where the main breadwinner was unemployed, and it has provided more comprehensive protection since then (Boyer, 2019). But partly as a result, the labour market became less flexible, especially compared with before the First World War (Hatton, 2007).

In more recent recessions, the pattern has been different with unemployment falling particularly hard on those aged 16-24 and ethnic minorities, while a proliferation of security programmes have protected families with children more than single adults and younger people. Since the 1980s, a series of programmes have aimed to help young workers and other marginalised groups. This has comprised a mixture of carrots and sticks, including training and employment subsidies (Vaitilingam, 2009).

This time is different

The coronavirus shock is very different from past experiences. From 2019 to 2020, GDP fell by even more than between 1929 and 1931. But while Okun’s law would predict a rise in the unemployment rate of nearly five percentage points, in fact, it increased by less than one percentage point.

On this reckoning, the UK government’s scheme of wage subsidies to keep workers on furlough or on reduced hours connected to their employers has averted a four percentage point rise in unemployment.

Because the current economic crisis is so different from those that preceded it, history is an uncertain guide. But the following outcomes seem likely:

First, as firms have retained so many of their workers while output has fallen so steeply, output growth should be especially fast in the early stages of recovery.
Second, because the furlough scheme has largely frozen the links between employers and employees, this may lead to a later bulge in the mismatch between workers and firms, which could hamper the recovery.
And third, those who have lost their jobs or failed to get into the labour market, mainly younger and part-time workers, may face an even tougher struggle than in the past (Wadsworth, 2021).

Where can I find out more?

The Great Recession was not so Great: This VoxEU column by Jan van Ours looks at the characteristics of unemployment in booms and slumps since 1970.
Returning to growth in the UK: Policy lessons from history: This VoxEU column by Nicholas Crafts provides a historical perspective on UK macroeconomic policies.
Long-term trends in UK employment: 1861 to 2018: The Office for National Statistics provides a summary of long-term trends in the UK labour force and employment.