| 6 | The Fight against Idleness |
| Unemployment and Discouraged Workers | |
| with Yaojun Li and Elisabeth Garratt |
Introduction
Idleness has a rather pejorative connotation nowadays. It has overtones of laziness and implies that the idle person is not bothering. I am sure that this is not actually what Beveridge would have had in mind. I suspect he was thinking of the Great Depression of the 1930s when jobs simply dried up for many industrial workers leaving them with no possibility of work. It would have been enforced idleness that concerned Beveridge—the unemployment of people who would have preferred to work if work had been available. Indeed, Beveridge’s next major piece of work, after his 1942 Report, was his 1944 book Full Employment in a Free Society, in which he wrote about possible ways to prevent ‘Idleness enforced by mass unemployment’.1
In Full Employment in a Free Society Beveridge argued that full employment in essence involved an unemployment rate of no higher than 3 per cent, since there would always be some frictional unemployment as workers left one job and looked for another. The crucial point, in his view, was that there should be more vacancies than there were workers looking for jobs so that the time people spent between jobs was kept to a minimum. The government’s role in a free society, he argued, should be to stimulate demand along the lines outlined by John Maynard Keynes in his celebrated The General Theory of Employment, Interest and Money (published in 1936),2 rather than to direct labour as the wartime government had done.
Ever since the Great Depression, enforced idleness has been recognized to have wide ramifications both for the unemployed themselves and for their families. A classic ethnographic study showed the devastating social impact of mass unemployment in the Austrian town of Marienthal after the town’s principal factory and main employer closed down during the Depression. A team of social scientists showed not only the poverty and material hardship which resulted but also the demoralization and the decline of community life which followed.3 Beveridge himself was associated with a somewhat similar British study, Men without Work, which was conducted in six British towns in the 1930s and documented similar consequences of long-term unemployment.4 In Full Employment in a Free Society he wrote that, although the purely material loss due to mass unemployment was serious, ‘the main evil of unemployment is in its social and human effects upon the persons unemployed and upon the relations between citizens’.5 He concluded ‘Idleness is not the same as Want, but a separate evil, which men do not escape by having an income.’6
Subsequent scientific research has supported Beveridge. Unemployment does indeed have direct and immediate economic effects in the form of reduced income and lost output, but it has harmful effects over and above the loss of income involved. It affects one’s longer-term economic prospects since time out of the labour market means that one’s work skills are becoming rusty, while people who have kept their jobs are acquiring further job skills and experience, with major implications for their future earning power. Moreover, job loss and unemployment have a much wider range of non-economic consequences in addition to the longer-term financial implications. Job loss and unemployment affect one’s psychological well-being and physical health as well as one’s pocket. It is associated with increased risk of alcohol and drug problems, with increased risk of suicide and with reduced civic engagement and social participation. It also has spillover consequences for other people such as one’s family, increasing for example the risks of divorce, of domestic violence, and even mental health problems for one’s partner.7 There are some tricky technical problems in demonstrating that these risks are directly caused by unemployment.8 But my reading of the evidence suggests that we should certainly take these wider social and psychological effects of unemployment very seriously indeed. I am sure Beveridge was right to see Idleness as a serious evil in its own right and not simply as an aspect of Want.
Economists David Bell and David Blanchflower have also emphasized that unemployment when one is young, especially if it lasts for a long time, causes enduring scars rather than temporary blemishes. ‘For the young, a spell of unemployment does not end with that spell; it raises the probability of being unemployed in later years and has a wage penalty. These effects are much larger than for older people.’9 The worry is that young people, especially those with few educational qualifications, have disproportionately borne the brunt of unemployment in recent decades.
In the 1950s young people could leave school at the earliest opportunity, with no qualifications to their name, and could still expect to get reasonable jobs, for example in the shipyards or factories. The long-term decline of Britain’s manufacturing industry and the growing emphasis on educational qualifications may mean that young men and women without qualifications in the twenty-first century have become increasingly at risk of unemployment. As more and more young people are acquiring qualifications, are the ones without good grades at GCSE increasingly marginalized? Has the giant of Idleness come to pick on young people, especially young men with low qualifications, much more than he did in the past? Beveridge had argued that ‘Failure to find any use for adaptable youth is one of the worst blots on the record of the period between the wars.’10 Has youth unemployment once again become one of the worst blots on the record of contemporary governments?
We also need to think a bit about whether the giant of Idleness has changed character. The strict modern definition of unemployment from the International Labour Office defines the unemployment rate as ‘the number of jobless people who want to work, are available to work and are actively seeking employment’.11 However, long periods of unemployment or lack of suitable jobs means that increasing numbers of out-of-work people may stop seeking employment altogether, and may thus drop out of the unemployment statistics. This phenomenon of what I will term discouraged workers may be particularly prevalent among older workers who lost their manufacturing or mining jobs during the period of industrial restructuring in the 1980s and have been unable either to retrain or to move away in search of work. So I should not limit my focus in this chapter solely to unemployment as defined by government statisticians. Beveridge would certainly have wanted to include discouraged workers within his concept of Idleness.
An analogous problem, which has indeed been widely recognized in government circles, is that of young people who are neither in education, employment, nor training (often described as NEET). This again goes wider than the strict definition of unemployment, since it covers young people who maybe do not want to work, or who are not actively looking for work, as well as those who fulfil the strict International Labour Office definition. Again, there has been substantial research showing the damaging effects of being NEET on people’s futures.
Before turning to these more detailed questions, I will look first at the long-term trends in unemployment, setting them in an international context. A central question for this section of the chapter is whether Margaret Thatcher’s reforms of the labour market in the 1980s—reforms which aimed to make the labour market more flexible—achieved their purpose. Has Britain with its flexible labour market succeeded better than the more regulated labour markets of continental Europe in keeping unemployment closer to Beveridge’s 3 per cent full employment target?
I will then move on to a more detailed examination of who is most at risk of unemployment, focussing on ethnic and gender differences as well as on youth. Has the dark side of educational progress led to greater vulnerability for young people, especially young men, who have failed to keep up in the race for educational qualifications? I will then try to supplement this with data on NEET and on discouraged workers.
Long-Run Trends in Unemployment: How Does Britain Compare?
I begin by charting unemployment trends in Britain and our seven peer countries. My central question is how Britain compares with our peer countries in protecting workers from the risks of unemployment, and whether this has changed over time. In particular, did Britain move up or down the rankings after Margaret Thatcher’s reforms which deregulated the labour market in the 1980s? There were a large number of different reforms and a variety of different aims behind these reforms. Some reforms were designed to weaken the power of the trade unions to call strikes, to make work pay by reducing the relative value of unemployment benefits, to reduce government interference in the labour market (for example by abolishing wages councils), and to make self-employment more attractive (for example through the enterprise allowance scheme).12 The overall thrust was to move Britain from the kind of highly regulated labour market typical of European countries such as France and Germany and in the direction of the much more deregulated US labour market. As a result we might expect to see Britain converging with (low) US rates of unemployment and diverging from (high) European rates.
Before turning to the actual data on long-term trends, I need to take a brief methodological digression. Basically, there are two different ways of measuring unemployment. The first relies on administrative data such as the government’s records of the number of people claiming unemployment-related benefits. Administrative records like this used to be the only source of data. However, in somewhat similar fashion to education statistics, there were numerous changes over time in how unemployment was recorded. Perhaps the most important change was in 1982 when official statistics moved from measuring unemployment on the basis of the number of people registered at employment exchanges (the register count) to the number claiming unemployment-related benefits (the claimant count). This move had the effect of substantially reducing the numbers recorded as unemployed.13 So exactly as with education and housing, there are major problems of comparability both over time and between countries if we just rely on government administrative data.14
The alternative method, and the one that is now generally used in developed countries for international comparisons, relies on representative sample surveys of the general population. Respondents to these surveys (in Britain the Labour Force Survey is the main one used) are asked a series of questions to determine whether they meet the International Labour Office’s definition of being currently out of work, available for work, and actively seeking work. These surveys are now seen as superior to administrative data.
The OECD has constructed harmonized measures of unemployment from 1955 onwards, based on these survey measures, in order to establish what proportion of the workforce meet this formal definition of unemployment. Unfortunately, however, not all of our countries are included for the whole period because of various limitations of the data available. In the case of Britain, for example, the OECD only publishes harmonized data from 1985, whereas they publish data for the USA and Canada going back to 1955. In Figure 6.1 I have therefore added—for Britain but not for the other countries—earlier administrative data based for example on register counts.15 This is clearly less than ideal but the margin of error is probably only a couple of percentage points.
Figure 6.1. The UK’s unemployment rate was relatively low in the 1950s, 1960s, and 1970s but became relatively high in the recessions of the early 1980s and early 1990s
Source: OECD harmonized unemployment rates, https://data.oecd.org/unemp/harmonized-unemployment-rate-hur.htm; register counts and modelled data for the UK 1955–8216
Still, the broad outlines of the picture are clear enough. In the 1950s and 1960s, the four countries for which data are available (Canada, Japan, the UK, and the USA) had relatively low unemployment rates. These were just over Beveridge’s target of 3 per cent in the cases of Canada and USA and below 3 per cent in Japan and the UK. My guess is that some of the other peer countries would have had equally low rates of unemployment at this time. For example, in the 1950s, a number of Western European countries such as France and Germany began programmes of recruiting foreign workers from outside the country in order to fill chronic labour shortages. There is no reason to doubt that there was pretty full employment immediately after the war in these countries, although I suspect that Italy would have been an exception and remained a country of net emigration, with more people leaving the country than entering it. Belgium, for example, recruited a large number of Italian labour migrants around this time.17
Unemployment rates then began to climb in the 1970s, after the oil shock of 1973/4 which pushed up the price of crude oil and led to rising inflation in many developed countries. (In a nutshell the Organization of Arab Petroleum Exporting Countries raised the price of oil and cut production.) However, unemployment remained in single figures throughout the 1970s in the four countries for which we have data, only a few points higher than it had been in the 1950s and 1960s.
Since 1980, however, it has been rather a bumpy ride, with surges in unemployment in almost all countries in the early 1980s, the early 1990s, and again after the financial crash and the great recession of 2007/8. The same global forces affected all the major economies in rather similar ways, although to slightly varying degrees.
On the other hand, it is very clear that there have been long-standing differences between countries, too. The impact of global recessions is only half the story. Thus Japan’s unemployment rate remained throughout well below that of Canada or the USA—in 2015 it was still just 3.4 per cent, close to Beveridge’s full employment target of 3 per cent. This is rather similar to education, where we found evidence that Japan had long been distinctive from other peer countries. In contrast, Italy had one of the highest unemployment rates of the eight countries from the earliest years for which we have harmonized data. These national differences do not appear to be new phenomena. My impression is that there are enduring institutional differences which predate, and survive, the succession of global recessions.
However, there are one or two rather interesting exceptions to this picture of stable differences between countries. The most noticeable exception is that of Sweden, a country which is often viewed as a model social democratic welfare state. Sweden did indeed use to be a model with a low unemployment rate, very similar in this respect to Japan. But this changed drastically in the 1990s, when Sweden saw its unemployment rate soar from Japanese levels to British levels and even higher. I do not claim to understand precisely why this happened, but the change coincided with a move towards a more market-oriented economy and the adoption of neoliberal, free-market policies. My interpretation is that changed political priorities account for Sweden losing its distinctively low unemployment rate and becoming a relatively high unemployment country.
What about Britain? The story is pretty clear with three distinct phases. In the first phase, which lasted up until 1980, British unemployment rates were lower than those of Canada and the USA, though not quite as low as Japan’s. But then in the second phase, in every single year from 1981 to 2001, British rates of unemployment were higher than the US and Japanese rates and were among the highest of all eight countries during the two recessions of the early 1980s and early 1990s. In this second phase, a high proportion of the unemployed in Britain were long-term unemployed (that is, unemployed for twelve months or more). In 1983, for example, the proportion of the unemployed who had been out of work for twelve months or more reached 46 per cent—a considerably higher proportion than in any other peer country apart from Italy. In the third phase, since 2001, the British rate of unemployment became relatively low once again and tended to oscillate around the US rate. In this third phase the proportion who were unemployed long term also fell—in 2015 it was down to 31 per cent of the unemployed, a proportion below that of Japan, Germany, France, Italy, and Spain.18
It is hard to avoid the conclusion that the major change beginning in 1981 had something to do with the free-market economic policies followed by Margaret Thatcher’s governments in Britain. However, the overall result of these policies seems to have been to raise the level of unemployment rather than to reduce it. Detailed analysis by the economist David Blanchflower and his colleagues suggests that the policies designed to make the British labour market more flexible did not have the intended result of reducing the rate of unemployment. Other aspects of the labour market reforms, like curbing the power of the trade unions or increasing British competitiveness, might well have been successful in achieving their goals. But the move to a flexible labour market did not achieve the goal of reducing unemployment. The economists concluded: ‘The “flexibility” explanation of unemployment is wrong.’ Instead ‘It is the fall in demand for labour that is the culprit.’19 Sustaining the level of demand for labour so that full employment could be achieved was of course precisely what Beveridge had been advocating in 1944.
Doubtless other economists could be found who might advance a different interpretation, but my guess is that any effects, for good or ill, of Margaret Thatcher’s labour market reforms were swamped by the adverse effects of her government’s monetary policy and high interest rates, designed to tackle inflation. Thousands of manual jobs in the manufacturing industry were lost during the early 1980s recession. Some might argue that this was a price worth paying in order to tackle the evil of inflation. In a famous speech in 1979 Margaret Thatcher argued that ‘The evil of inflation is still with us. We are a long way from restoring honest money…We should not underestimate the enormity of the task which lies ahead. But little can be achieved without sound money. It is the bedrock of sound government.’20 In effect, the Conservative government prioritized control of inflation over the maintenance of employment. This is not the place to get into an argument about whether or not this was the right thing to do at the time, but it is clear that Margaret Thatcher’s programme, taken as a whole, did not at the time reduce unemployment but increased it.
All the same in the early twenty-first century Britain’s overall unemployment rate once again compares quite well with that of peer countries—and it is not at the time of writing very much higher than the full employment target of 3 per cent. Part of the explanation for this will I suspect be found to be Britain’s low interest rates, which will have kept demand for labour high. Another factor might be the rise of self-employment in Britain, which had indeed been one of Margaret Thatcher’s policy objectives. Self-employment provides a potential way out for people faced with unemployment. It can be very insecure and may not generate much income, but it probably avoids some of the harmful social and psychological effects of idleness that Beveridge worried about. At all events, back in 1980 at 8.1 per cent Britain had one of the lowest self-employment rates of our eight peer countries. By 2015 this had increased to 14.9 per cent according to OECD figures, one of the highest among our peer countries.21 It would not be unreasonable therefore to see Britain’s currently low unemployment rate as in part a legacy of Margaret Thatcher’s policies of making self-employment easier and more attractive in the way it was treated for tax purposes.22
Who Is Most Affected by Unemployment? Has This Changed over Time?
While overall levels of unemployment in Britain since 2000 have been relatively low compared with rates in most of our peer countries, the risks of unemployment strike some groups much more than others. For some people, risks of unemployment were still very high even in 2015. It has probably always been the case that the young, migrants, and the least educated were more exposed to unemployment. In the 1950s with unemployment around 3 per cent or less, the differences in risk would have been in practice very small. However, as unemployment rose—particularly in the recessions of the early 1980s and early 1990s and after the financial crash in 2007/8—inequalities in risks of unemployment will have risen, too.
An important question is whether these increases were temporary, like the recessions, or whether something more fundamental changed. The increase in income inequality which occurred in the 1980s has become a permanent new feature of the British labour market. Do we find a parallel more or less permanent increase in risks of unemployment for vulnerable groups of workers? I raised in Chapter 4 the question of the dark side of educational expansion—what happens to those left behind in the race to acquire more and better qualifications? With a growing number of highly qualified peers to compete against, one guesses that the economic prospects of the poorly qualified will have deteriorated relative to better-qualified workers.
So I shall now have a closer look at unemployment in Britain. Who is most at risk, and how have those risks changed over time? Is the giant of Idleness now more selective about whom he picks on? I will start with gender differences and then move on to differences between people with different qualification levels, ages, and ethnic backgrounds. Moreover, are these different risks compounded so that some sections of society suffer a double or treble burden?
Gender Differences in Unemployment Rates
Britain, like most of the peer countries, saw major changes in the post-war period in the proportion of women going out to work. When I was growing up in the 1950s, it was rather rare for married women with children at home to take paid employment. Most women worked between leaving school and having their first baby, and many returned to work after their children had grown up (exactly as my mother did). But one of the major transformations of British society in the post-war period was the increased participation of mothers in the labour market (although often taking part-time jobs while their children were young).
The kinds of job that women enter have also slowly changed. In my mother’s day, jobs were highly gendered—women like my mother would take jobs such as a typist or secretary which were almost totally a female preserve at the time, while there were a large number of jobs, particularly in the manufacturing, mining, and construction industries which were equally exclusively male preserves. Gender segregation is not yet a thing of the past, but it has been gradually changing alongside Britain’s shift from being an industrial society to a post-industrial or service-based economy. So what have been the implications for men and women’s risks of unemployment?
Figure 6.2 uses statistics from representative government surveys (not administrative data) to trace gender differentials in risks of unemployment from the early 1970s onwards.23 (The main surveys which I draw on here only started in the 1970s, and I have not been able to find earlier administrative data which distinguish male from female unemployment rates.)
Figure 6.2. Men experienced higher rates of unemployment than women from the 1980s onwards in Great Britain
Note: samples include all economically active respondents (men aged 16–64 and women aged 16–59)
Source: pooled General Household Surveys (1972–2005) and Labour Force Surveys (1983–2015)
Figure 6.2 tells a striking story. In the 1970s men and women had rather similar, and fairly low, risks of unemployment.24 This changed in the recession of the early 1980s, when male unemployment rose distinctly higher than female unemployment did. This gender differential remained evident for the following thirty years. The gender gap was particularly wide at the peak of each recession, reaching four percentage points in the recession of the early 1990s. More recently, the gap narrowed and male and female risks of unemployment more or less converged after 2012. However, I would not be surprised if the gap were to rise again during the next recession. It is probably too early to say whether the convergence of male and female unemployment rates in 2015 is the new normal.
The natural interpretation of the 1980s shift in men’s and women’s relative risks of unemployment is that it reflected the changing structure of employment in the last decades of the twentieth century. There was a long-term decline of the predominantly male industries of manufacturing, shipbuilding, and the like, while increasing numbers of more female-oriented white-collar and service jobs were being created in the second half of the twentieth century.
The shift in the gender gap could also reflect in part the way in which women gradually overtook men in education, as I described in Chapter 4. Better educational qualifications reduce the risks of unemployment, since they provide access to higher-skilled and more secure employment. In contrast, unskilled work, which is predominantly all that is left for people with low qualifications, is much more precarious and brings higher risks of unemployment.25 If men continue to lag behind women in educational qualifications, the likelihood is that they will also continue to lag behind in their chances of finding work in the contemporary service economy.
Educational Inequalities
Higher qualifications give access to better jobs, not just ones with higher earnings but also ones with greater job security and more favourable contractual conditions (such as the period of notice required) and employment rights. The reason for this is that highly skilled workers are more valuable to firms, and are not so easy to replace, so firms have a greater incentive to keep them on when times become hard. The importance of educational level for risks of unemployment can clearly be seen in Figure 6.3.
Figure 6.3. Men and women with higher qualifications have lower risks of unemployment, with little change over time in Great Britain in their relative risks
Source: pooled General Household Surveys (1972–2005) and Labour Force Surveys (1983–2015)
Figure 6.3 shows clearly that, throughout the period covered by the data sources, the higher one’s level of education, the lower one’s risk of unemployment. Among both men and women graduates in 2015 the risk of unemployment was around 3 per cent, rather similar to the risk facing graduates back in the 1970s. The risks then increase steadily as one descends the educational ladder, reaching 10 per cent or more for people with no or low qualifications (that is, qualifications lower than GCSE or its equivalent).
Moreover, the vulnerability of people with low or no qualifications increases markedly during a recession. The gaps narrow when there is high demand for labour but then widen when there is a slack labour market at the height of the recession. As Figure 6.3 shows, the risks facing men with no or low qualifications were particularly large in the recessions of the early 1980s and early 1990s.
Leaving the recessions aside, has there been a long-term increase in the vulnerability of people with low qualifications? This is not an entirely straightforward question to answer, but if we compare the non-recession years of 1978, 1989, 2001, and 2015, it is hard to discern any clear long-term trends. Thus, in 1978 the gap between the best- and least-educated men was 4.1 points; in 1989 the gap was rather higher at 6.3 points; however, in 2001 the gap dropped slightly to 5.9 points; and in 2015 the gap was 5.7 points. I think it is safest to conclude that there was no discernible trend over time, at least as regards male unemployment.26 In the case of women, the gap was considerably larger in 2015 than it had been earlier, so perhaps less-qualified women did become more vulnerable.
Age Differentials
A major concern nowadays is that young people may be particularly at risk of failing to find a job and that, in Beveridge’s words, ‘failure to find any use for adaptable youth’ has become once again one of the worst blots on Britain’s record. As I remarked earlier, young people in the 1950s could leave school at the earliest opportunity, with no qualifications to their name, and could still expect to get reasonable jobs in manufacturing industry. So there is a major question as to whether the giant of Idleness has come to pick on young people, especially young men, much more than he did in the past. This is surely a fundamental social problem because of the long-term scarring effects which unemployment can have, as well as its social and psychological ramifications.
Figure 6.4 shows that young people have elevated risks of unemployment compared with older workers. Thus, among men in 2015, those aged 16–24 had almost five times the risk of unemployment as those faced by men aged 36 to 50 in mid-career. Risks of unemployment facing men or women aged 25 to 35 tend to be quite close to the risks facing older workers, so it is the young who stand out. Moreover, these age differentials are distinctly larger than the educational differentials shown in Figure 6.3.
Figure 6.4. Young people under 25 have higher risks of unemployment than do older people, and the differential has been increasing over time among men in Great Britain
Source: pooled General Household Surveys (1972–2005) and Labour Force Surveys (1983–2015)
Just as with the educational differentials, the differentials between older and younger workers tend to become larger during recessions, when young people are particularly hard hit, just as the less qualified were. However, if we compare non-recession years, it does look as though young men have become increasingly vulnerable to unemployment. In the four non-recession years of 1978, 1989, 2001, and 2015, the unemployment risks of men in mid-career remained more or less constant, but for young men they rose steadily from 8.4 per cent in 1978 to 11.2 per cent in 1989, 13.7 per cent in 2001, and 15.1 per cent in 2015. So it really does look as though the employment situation deteriorated for young men. The trend is not quite so dramatic for young women, but it too suggests that the relative vulnerability of young women deteriorated somewhat.
Perhaps even more importantly, the age and educational differentials are cumulative. For example, the risks of unemployment in 2015 were almost 9 per cent for men with qualifications below GCSE, and they were around 15 per cent for young men aged 16–24. But among men who had the misfortune to be both young and to have low or no qualifications, the risks rise to over 24 per cent—fifteen times the risk faced by a middle-aged male graduate (a mere 1.5 per cent in 2015). This is social inequality of a pretty extreme kind—and one that may have even more serious consequences than the purely monetary inequalities which we saw in Chapter 2.
In Figure 6.5 I restrict the analysis to people with no or low qualifications. What we can clearly see is that the age differences were distinctly greater in the last decade of the twentieth century and the first decades of the twenty-first century than they had been in the 1970s or the non-recession years in-between. If we compare the non-recession years of 1978, 1989, 2001, and 2015, we find that at the two earlier time points the difference in the unemployment risks facing younger men with no or low qualifications were ten points greater than those facing older workers with similarly low levels of education. In contrast, at the two later time points the gaps had risen to eighteen points. The picture is not quite so clear for women (as there was a rather large gap in 1978), but in 2015 the gap had reached twenty points. I think it is safe to conclude that there had been a real deterioration over time in the labour market position of younger people with low or no qualifications to their name.
Figure 6.5. Young people with low or no qualifications have greatly elevated risks of unemployment, and the differentials increased in Great Britain
Note: analysis restricted to those with no or low (below GCSE level) qualifications
Source: pooled General Household Surveys (1972–2005) and Labour Force Surveys (1983–2015)
It is not entirely clear what has driven this change but it does coincide with the great expansion in the 1990s of younger people achieving GCSE qualifications (Figure 4.4). This is consistent with the concerns that I raised earlier about the dark side of educational expansion: as more and more young people obtain qualifications, the competitive position of young people who lack qualifications becomes weaker, and their risks of unemployment become relatively worse.
To be sure, we need to recognize that the kinds of young people who lack qualifications may be rather different in the twenty-first century from what they would have been in the 1970s or earlier, when it was still common for young people to enter the labour market with no qualifications. In the twenty-first century this is much less common and the minority without qualifications may perhaps be unusual in having learning or behavioural difficulties, histories of school exclusion, or other problems which hold them back.27 These problems in turn may reduce their chances of finding work. But all the same, lack of formal qualifications will make these young people more visible to employers in the twenty-first century, and thus perhaps make it easier for employers to reject them as unsuitable for work. In earlier days, such young people would not have been so visible—it might have been easier for them to escape unnoticed. In the twenty-first century they are much more likely to be screened out by employers.
Ethnic Differences
Ethnic minorities are another group of people who are vulnerable to unemployment. True, there are substantial differences between differing minority groups in their risks of unemployment, in part reflecting their different levels of education. People with Black Caribbean, Black African, and Bangladeshi or Pakistani origins have the highest risks, while Chinese and Indians have lower risks. Indeed, the risks facing the latter two groups are fairly similar to those of the white British. The question which I want to address next is whether these Black and Muslim groups have the same, better, or worse risks of unemployment compared with their white peers with similar levels of education. Do young Black men with low qualifications bear an additional burden over and above the risks that young white men experience in Britain? Do they suffer a triple rather than the double burden faced by young white men with low qualifications?
In Figure 6.6 I focus on the unemployment rates of young people (aged 16–24) with low or no qualifications so that we can roughly compare results with those shown in Figure 6.5. Because the sample sizes become rather small when one limits the analysis in this way, I combine Black Caribbeans with Black Africans to form a Black group, Pakistanis with Bangladeshis to form a predominantly Muslim group, and Indians with Chinese to form an Asian group. I compare these three groups with the white members of the samples. I also combine three adjacent years in order to increase sample sizes and improve the reliability of the estimates. (Even with these measures taken, some cases still have rather small sample sizes and therefore in the two graphs for men and women, results are omitted if the number of people in the sample in the relevant year and ethnic group is less than ten.)
Figure 6.6. Young Black men and women with low qualifications are even more at risk of unemployment than their white peers, especially during recessions
Note: analysis restricted to those with no or low (below GCSE level) qualifications and aged 16–24
Source: pooled General Household Surveys (1972–2005) and Labour Force Surveys (1983–2015)
The results are a bit patchy because of the small sample sizes, but the overall picture is clear enough. What we find is that, at least in the most recent decades, there is indeed an accumulation of disadvantage affecting both the Black and the Muslim groups. In 2014/15, for example, young Black men with no or low qualifications had a 50 per cent risk of unemployment, almost double that of young white men with similar qualification levels. We can therefore speak of a treble burden which these young, less-qualified, Black or Muslim men have to carry. Young Black and Muslim women also experience this treble burden.
This accumulation of disadvantage is particularly apparent during recessions, and tends to reduce when overall unemployment rates fall (as for example in 2005–7). This strongly suggests that these groups are the most marginalized in the labour market—I suspect that they are the last to be hired and the first to be fired. As we will see in Chapter 7, racial discrimination is almost certainly a part of the explanation for this third burden.
Broadening the Measure: Discouraged Workers
The statistics on unemployment which I have reported so far in this chapter adhere to the strict definition of the unemployed as people who want to work, are available for work, and are actively seeking work. While this kind of unambiguous definition makes very good sense when attempting to make comparisons between countries or over time, it is a bit too narrow if we want to take seriously Beveridge’s approach to the giant of Idleness.
The key problem with the official definition is that long periods of unemployment or lack of suitable jobs may demoralize out-of-work people. They may give up hope of finding a job—and this would not be irrational. The longer someone is out of work, the poorer their chances are of finding a job since they will have fallen behind their in-work peers in gaining skills and work experience. Employers may also question the hole in their CV. So older workers may take early retirement while young people may withdraw and look for an alternative future on the streets or in the black economy.
One specific problem, which has been widely recognized in official circles, is that of young people who are neither in education, employment, nor training. Again, there has been substantial research showing the damaging effects of being NEET on people’s futures—it has been shown to lead to educational underachievement, higher risks of unemployment, a greater susceptibility to engage in crime, poor health, and feelings that life has no purpose. Low-skilled youth who become NEET find it more difficult to re-engage in employment and there is evidence that they may become trapped, leading to long-term scarring with major social and economic costs.28
The concept of NEET goes wider than the strict definition of unemployment, since it covers young people who maybe do not want to work, or who are not actively looking for work, in addition to those who fulfil the strict definition. In other words, it is a broader measure than unemployment in the strict sense defined by the International Labour Office. Basically, it includes all young people who are neither in work nor in some form of education or vocational training. However, it does include young people who have dependent children or other caring duties. The measure is implicitly based on the assumption that it is normal for young people to be either in education, training, or work rather than looking after the home. With the gradually rising age at which women have their first baby, this is perhaps not a silly assumption, but we do need to interpret the measure with some caution.
Another complication to be aware of when we attempt to measure the proportion who are NEET is that standard official measures (such as those used by government or the OECD) look at the proportion of the whole age group who are NEET, whereas official figures for unemployment exclude those members of the age group who are either in education or who are not looking for work. This means that we cannot directly compare NEET and unemployment figures. To make it a bit clearer what is going on, therefore, I present the figures in a different way from the usual official one. In Figure 6.7, I distinguish four categories:
Figure 6.7. The proportion of 16–24-year-old men who are neither in education, work, nor looking for work has been increasing; among 16–24-year-old women it has been decreasing
Source: pooled General Household Surveys (1972–2005) and Labour Force Surveys (1983–2015)33
The latter two categories are basically the ones that make up young people who are NEET. I cumulate these four categories so that they sum up to 100 per cent, with separate charts for young men and young women.29
As with Figure 6.6, I combine adjacent years in order to deal with the problem of small numbers in the samples on which the results are based. As before, there will be considerable sampling error, but the broad outlines of the story are clear enough. Over the period for which we have the survey data, the proportion of young people in some form of education or training markedly increased in the 1990s, which of course is what I reported in Chapter 4. Conversely, the proportion in employment steadily fell over time, down from 58 per cent in the early 1970s to around 40 per cent in 2014/15 among young men.
What about NEET? The unemployment element of NEET naturally parallels what we saw in Figure 6.4 with peaks in the recession years of the early 1980s, early 1990s, and after the financial crash of 2007/8. However, among young men, we can also see a gradual increase in the proportions who are inactive—that is, who are neither in education or employment nor looking for work. The proportion increases from around 2 per cent at the beginning of the period to nearly 6 per cent at the end of the period (equivalent to about 50,000 young men). We cannot be sure that these are all discouraged young people. Some, for example, may be young people from affluent backgrounds who are taking gap years between school and university—although if they have gone abroad for their gap year they should not show up in the samples anyway. The data are not conclusive but they do suggest that there is a growing potential issue of young men with nothing to do.
The long-term story for young women is rather different. As we can see, much larger proportions of young women in the 1970s and 1980s were neither in education, work, nor looking for work. Almost certainly this reflects the earlier age at which women had their first child in the earlier decades. Consider, for example, women born in 1952 and who would therefore have been young adults in the first (1972–4) round of data in Figure 6.7. Twenty per cent of these young women had had their first child by their twentieth birthday, and would therefore have been rather unlikely to be studying or looking for work. In contrast, among women born in 1995, and thus becoming young adults in the final round of data, only 7 per cent had had their first child by their twentieth birthday.30 What this in effect means is that there has been a convergence between young men and women in their rates of being NEET.
The Office for National Statistics has published its own statistics on trends in NEET, although only since 2002.31 Their detailed figures are very similar to mine for the same period, showing that NEET peaked in 2012 and has since been declining. As in Figure 6.7, the decline since 2012 largely reflects the decline in the unemployment rate, although at some point I would expect to see an additional reduction as a result of the raising of the age (in England) during which young people have to undertake some kind of training.
As one might expect, young people with no qualifications are disproportionately likely to be NEET. Pregnancy and parenthood are also unsurprisingly associated with being outside education and the labour market.32 An even greater risk of being NEET, however, is found among those with a disability (defined as current physical or mental health conditions which reduce one’s ability to carry out day-to-day activities). Young people who have been excluded or suspended from school are more likely to be NEET, as are those who have been under supervision by a youth offending team and those who have been in care. A trenchant report by the National Audit Office concluded:
The poor life experiences of too many care leavers are a longstanding problem. Without well-targeted support their deep needs will not be met, with costly consequences both for the young people and for society. While there is a clear legal framework and an inspection regime in place, the system is not working effectively to deliver good outcomes for all care leavers. On the key measure of numbers in employment, education and training the situation has deteriorated since 2007–08.34
Disability, school exclusion, offending, low qualifications, experience of the care system—this is a litany of the most marginalized and vulnerable young people in British society. International comparisons also suggest that marginalization in the form of being NEET is a particular problem in Britain. Among young people under 20, Britain compares particularly unfavourably. Only Italy has a worse record. Germany in contrast has a rate of NEET less than a third of the British rate.
In Figure 6.8 I show the trends since the turn of the millennium in the proportion of young people (men and women combined) who were NEET in the eight peer countries. Italy stands out as having the poorest record—which echoes the high overall unemployment rate in Italy reported at the beginning of the chapter in Figure 6.1. In contrast, Germany, Sweden, and Japan have consistently had levels well below the British. Even France, with its relatively high overall unemployment rate, fares slightly better than Britain.
Figure 6.8. The UK had one of the highest rates of NEET among young people under 20 in the twenty-first century
Note: data for 15–19-year-old men and women
Source: OECD, ‘Youth not in employment, education or training (NEET)’ (2017), https://data.oecd.org/youthinac/youth-not-in-employment-education-or-training-neet.htm
We can also see that most countries have experienced declining rates of NEET since 2010. British government interpretations have tended to credit the decline in Britain to the success of the government’s training policies, in particular to the legislation increasing participation in education or training to ages 17 and then to 18. Figure 6.8 should, however, make one a little sceptical of this optimistic interpretation. Since rates of NEET declined in most countries, it is more likely that the explanation lies in common factors—such as the generally declining overall rate of unemployment—than to specific policy initiatives in Britain.
This is not to say that the English reforms had no value—almost certainly they were a step in the right direction. But I suspect that they will not be sufficient on their own to bring British rates of NEET down to German or Japanese levels. This brings me to an important paradox. Whereas overall British unemployment rates compare rather favourably with those in peer countries (as Figure 6.1 showed), youth unemployment and participation compares rather unfavourably.
Conclusions
Has youth unemployment once again become one of the worst blots on the record of contemporary governments? Certainly, unemployment is very bad news for those affected—especially for young people—and for the wider society. It has wide adverse consequences for well-being over and above its purely financial implications. I am sure Beveridge was right to treat it as a giant evil in its own right and to draw attention to the particular plight of young people.
In the 1970s, and probably throughout the 1950s and 1960s, too, Britain tended to have a lower overall unemployment rate than its peer countries. This changed in the 1980s and 1990s, when Britain’s unemployment rate surged during the two recessions and became consistently higher than in the peer countries. However, more recently, since 2001, Britain’s unemployment rate has once again become relatively low.
The theory that greater flexibility of the labour market produces a lower unemployment rate receives scant support from detailed econometric analysis. Certainly the total package of reforms which Margaret Thatcher’s administrations undertook after 1979 was associated with higher, not lower, unemployment. My interpretation (and I suspect that of professional economists such as David Blanchflower) is that the focus on controlling the evil of inflation (and the associated high interest rates) in the 1980s reduced demand for labour. As the period of high interest rates came to an end, so demand recovered and unemployment fell. The policies of the European Central Bank failed to respond as effectively, and so Eurozone demand remained lower and unemployment rates in a number of European countries remained higher than in Britain.
Despite the fact that overall British unemployment rates fell, the evidence available suggests that inequalities in the risks of unemployment tended to become larger, not smaller. True, there have long been differences in who is most at risk of unemployment. A long-standing feature of most advanced economies is that the young, those with less education, and migrants or the native-born descendants of migrants have higher risks of unemployment. These groups also appear to be harder hit whenever there is a recession (as in the early 1980s, early 1990s, and after the financial crash).
These inequalities in the risks of unemployment have remained, and indeed the relative position of young men with low qualifications seems to have deteriorated after 1990. Moreover, young British men appear to be particularly at a disadvantage compared with their contemporaries in Germany, Sweden, Canada, and the USA.
I suggested earlier that this might represent the dark side of educational expansion. Many sociologists have suggested that the educational system is a kind of sorting and labelling mechanism, dividing people into the educational equivalents of sheep and goats. Moreover, this sorting and labelling takes an earlier and more visible form in Britain, with high-stakes examinations (GCSEs to be precise) taken at ages 15 or 16. For example, in countries like the USA and Canada large proportions of the age group go on to complete high school and receive their diplomas at the age of 18. In contrast, in Germany, sorting and labelling take place much earlier (around age 11 as it used to do during the tripartite system in Britain), but there is a much closer articulation between school and work with German vocational programmes having real value in the labour market.
Britain manages to fall between these two stools. After 1965 it largely moved away from the German-style selective system towards a US-style comprehensive system, but at the same time Britain retained high-stakes examinations like GCSEs from the previous system and failed to embrace the high-participation and completion rates of the US and Canadian systems. My interpretation, then, following the conclusions of Alison Wolf’s influential report, is that the British system leaves many less able young people with fairly worthless qualifications. ‘The staple offer for between a quarter and a third of the post-16 cohort is a diet of low-level vocational qualifications, most of which have little to no labour market value.’35 The government has put in place reforms implementing Alison Wolf’s recommendations, but it is too early to determine whether they have been sufficient to close the gap with peer countries.
Given the long-term scarring effects of unemployment, and the wider social repercussions of having large numbers of young men without work, this represents a kind of time bomb ticking away at the heart of British society—and not one for which we can put the blame on foreign extremists. It is one of our own making. To paraphrase Beveridge, failure to find any use for adaptable youth is one of the worst blots on Britain’s record in the twenty-first century.