02The UK’s army of hidden unemployed people

Official unemployment figures mask substantial numbers of hidden workers in cities and large towns in the North of England, in particular. This underscores, once again, the need for a set of levelling up policies to tackle the struggles of places outside the Greater South East.

Big increases in economic inactivity – those people not searching for work – since the onset of Covid have grabbed headlines and led to the prime minister asking for a review to look at the causes.

While it is right to be concerned about recent rises, the debate has been largely blind to how this plays out across the country. This chapter shows there is a much greater, longer-term problem with inactivity that has a very clear North-South split.

Box 1: Methodology

Defining cities

Centre for Cities research focuses on the UK’s 63 largest cities and towns, defined as primary urban areas (PUAs). Unless otherwise stated, Centre for Cities uses data for PUAs in its analysis – a measure of the ‘built-up’ area of a large city or town, rather than individual local authority areas. You can find the full definitions and a methodological note at www.centreforcities.org/puas

Data used for this research

Most of the research uses data from the Annual Population Survey for 2022, which provides information on economic inactivity and unemployment at the local authority level. When this was conducted, the latest data available covered the 12 months to June 2022. Data for Belfast was not available.

Box 2: Defining labour market terms

People are generally grouped into two broad categories:

  • Those who are economically active, so individuals who are either employed (currently in paid work) or unemployed (without a job but actively seeking work within the past four weeks).
  • Those who are economically inactive, meaning they do not have a job but are not categorised as unemployed because they are not seeking work, or are not available (or both).

The unemployment rate measures the share of the economically active population that is out of work. The inactivity rate measures the share of the working-age population that is inactive.

Unemployment is at historically low levels

At first glance, the UK jobs market appeared buoyant in 2022. Unemployment was at record lows, with only 3.7 per cent of the working-age population (1.2 million people) classed as unemployed by October. That is not only less than pre-pandemic levels, it is the lowest rate Britain has seen since the early 1970s. And it is a picture that played out across the country.

There is a geography to unemployment – cities and large towns in the North or the Midlands had higher rates than those in the Greater South East, as Figure 1 shows. But even in places that fared the worst, like Birmingham and Middlesbrough, the unemployment rate was around six per cent, which is very low in the context of the last four decades.2

Figure 1: While unemployment is higher in cities and large towns in the North, it is relatively low by historical standards

Source: ONS, Annual Population Survey 2022.

The cause of such low unemployment was not a booming jobs market – the UK is one of the only countries in the OECD not to have returned to pre-Covid employment rates.3 It was the result of rising numbers of people being classed as ‘inactive’ because they were withdrawing from the jobs market despite an increase in vacancies. And this spurred the prime minister’s request for a review.

Not all forms of economic inactivity should be a worry for policymakers. Some of it is driven by individuals making positive choices, such as enrolling in higher education or retiring early if they are financially secure enough to do so. These types of inactivity are unrelated to the economic performance of the local area in which people live and work. Instead, they are more closely tied to the benefits and characteristics that different places offer, such as universities in cities, or amenities that particularly appeal to retirees. This partly determines where people who are inactive choose to live (see Box 3).

Box 3: Cities have higher levels of overall inactivity, mostly driven by students

The inactivity rate is around 1 percentage point higher in cities. In total, they are home to 5 million inactive people compared with 3.6 million in non-urban areas.

This is mostly driven by the distribution of some specific types of inactivity, and students in particular (Figure 2). In urban areas, nearly 30 per cent of all inactive people are students, compared with 24 per cent in non-urban areas. Meanwhile, about 65 per cent of inactive students live in cities, and these places also have a higher share of people looking after a family or home.

In contrast, early retirement is less of an urban issue, with retirees much more likely to live outside of cities and large towns.

Figure 2: In cities, economic inactivity is largely driven by students, while early retirement is less of an urban issue

Source: ONS, Annual Population Survey 2022.

Other forms of inactivity are less likely to be the result of positive choice. For example, people may leave the labour market and stop looking for work if they are discouraged, believe there are no jobs available (or no good positions) or cannot work because of health issues. These should all be concerns for policymakers.

There is an army of hidden workers in northern cities and large towns

By discounting those for whom inactivity is more likely to be a choice, it is possible to compute an ‘adjusted’ inactivity rate.4 This more accurately measures the share of the working-age population that is likely to be involuntarily inactive and, in principle, could work or look for a job if they had adequate support or better employment prospects.

Nationally, there are 3.5 million people who are involuntarily inactive – almost three times the number who are unemployed. Nearly two million live in cities, which is about 60 per cent of the total.

Reflecting the geography of unemployment, involuntary inactivity is greater in the North (Figure 3). In fact, eight of the 10 places with the highest rates are northern cities. More than one in eight working-age people in Barnsley and Sunderland, for example, fall into this category compared with around one in twenty in Reading or Basildon. And this pattern has persisted for many decades, as Box 4 shows.

Figure 3: Cities in the North have higher involuntary inactivity rates

Source: ONS, Annual Population Survey 2022. Note: Crawley, Cambridge, Ipswich and Worthing have been removed from the analysis because of missing data.

Table 1: Cities in the North have higher involuntary inactivity rates

Top 10 involuntary inactivity, June 2022
Barnsley 13.4%
Sunderland 13.4%
Hull 13.3%
Middlesbrough 12.9%
Blackburn 12.8%
Mansfield 12.7%
Newport 12.6%
Doncaster 12.4%
Swansea 12.3%
Birkenhead 12.3%
Bottom 10 involuntary inactivity, June 2022
Oxford 6.8%
Chatham 6.6%
Brighton 6.4%
Bristol 6.2%
York 5.8%
Edinburgh 5.6%
Gloucester 5.3%
Norwich 5.0%
Basildon 4.8%
Reading 4.7%

Source: ONS, Annual Population Survey 2022. Note: Crawley, Cambridge, Ipswich and Worthing have been removed from the analysis because of missing data.

Hidden unemployment is up to four times higher in some northern cities and towns than official figures suggest

Combining those who are registered as unemployed and those who are involuntarily inactive takes the national rate from 3.7 per cent to 12.1 per cent. That is an increase from 1.2 million to 4.7 million people, nearly the population of Manchester and Birmingham urban areas combined.

These shifts are even larger at the local level and reveal a much greater North-South divide than unemployment figures on their own suggest. Of the 10 places with the highest hidden unemployment rates, nine are in the North. In Middlesbrough, Hull and Blackburn, accounting for that group increases the rate from 5 per cent to 20 per cent – a four-fold rise. In Liverpool, it takes the total number of unemployed people from 13,000 to more than 46,000 people, which is almost enough to fill Anfield stadium. This is a strong contrast with cities like Gloucester, Reading and York where the hidden unemployment rate is around 8 per cent, which is much closer to the headline unemployment figures.

The result is much greater disparities between cities than those headline figures suggest, as Figure 4 shows. The gap between cities with the lowest and highest unemployment rate is 4 percentage points. But for hidden unemployment, this increases to 12 percentage points.

Figure 4: There is a large gap between official unemployment and hidden unemployment

Source: ONS, Annual Population Survey 2022. Note: Crawley, Cambridge, Ipswich and Worthing have been removed from the analysis because of missing data for 2022. The hidden unemployment figure here is computed as the number of people unemployed and involuntary inactive (as defined above) as a percentage of the labour force plus the involuntary inactive.

Box 4: The scars of deindustrialisation are still visible on today’s labour market

Cities and large towns in the North, like Middlesbrough, Hull and Blackburn, have long grappled with high levels of economic inactivity and this is, in part, a legacy of their industrial past. Most had a significant proportion of work in mining and manufacturing in the mid-20th century, but they experienced large-scale job losses in the 1970s, 1980s and 1990s owing to industrial decline. In those places, the main labour market response was not an increase in unemployment; instead workers – mostly men – withdrew into economic inactivity.5 The impact of this is still visible today.

Figure 5 compares involuntary inactivity now and in 1981. It shows that those cities with the greatest inactivity in 1981 are still the most affected. Of the 10 with the highest 1981 rates, six are still in the top 10 in 2022 – Barnsley, Swansea, Doncaster, Hull, Sunderland and Middlesbrough.

Figure 5: The legacy of the 1980s is still visible on today’s labour market

Source: ONS, Annual Population Survey 2022, Census 1981.
Note: the 1981 Census data does not allow to isolate the ‘looking after family or home’ factor, which likely explains why values are relatively high across all cities.

This though cannot simply be explained by high numbers of former miners or factory workers among those who are currently economically inactive, as few are still of working age. Instead, Figure 5 shows the impact of deindustrialisation on subsequent generations – the children and grandchildren of those withdrawing from the labour market in the 1970s, 1980s and 1990s. This is explored in greater detail below.

Recent increases in inactivity rates have created new problems in some southern cities, in addition to compounding existing ones in the North

The well-documented recent rises in inactivity have compounded this picture at the national level, but they have not been evenly spread across the country.

Inactivity had generally been declining since the turn of the 21st century, mostly driven by more women joining the labour force.6 But the Covid-19 pandemic has reversed this trend. Between December 2019 and October 2022, the total number of people in the jobs market shrunk by 370,000, despite low unemployment and worker shortages. This moved the inactivity rate from 20.2 per cent of the working-age population to a six-year high of 21.7 per cent, making the UK a clear international outlier.

Some of this has been driven by a rise in students who may have delayed their entry into higher education because of the pandemic. More older workers retiring early, either in response to redundancies and dismissals or as a result of post-pandemic lifestyle changes, may have played a role too. National data for October 2022 suggests these two reasons combined explain, at most, a third of the overall increase. This means, therefore, that it has largely been made up of involuntary inactivity.

Perhaps surprisingly, despite the national focus, involuntary inactivity has not gone up across the country. Figure 6 looks at the performance of cities and large towns in 2019 and changes in involuntary inactivity since then, and there are four things to note:

  • Not all cities saw a rise; in places like Derby, Edinburgh and Wakefield rates actually decreased.
  • Recent changes have not entirely followed the geography of inactivity set out above. Cities with large increases are located across the UK. For those in the top right quadrant of Figure 6, such as Barnsley and Hull, the rises have compounded high long-term rates. For mainly southern cities, such as Northampton, Swindon and Milton Keynes in the top left, this has posed a new problem for their economies.
  • Similarly, places that have seen either low increases or decreases are geographically spread. For cities like Dundee and Liverpool (bottom right), recent falls have brought some relief to their inactivity challenge. For those in the bottom left, such as London, Edinburgh and Bristol, inactivity continues not to be a great concern.
  • Despite the attention the recent increases have received, they are small relative to the much larger problem of stubbornly high levels of involuntary inactivity in many places. In a city like Swansea, the 2 percentage point rise over the past three years is a concern not because it has happened, but because it has added to the 10 per cent of working age people who were already involuntarily inactive.

Figure 6: Recent increases in inactivity have not entirely followed the North- South divide

Source: ONS Annual Population survey 2019 and 2022.

Note: this chart only includes 42 of the 62 cities as only those for which all data was available for both 2019 and 2022 were included.

The recent focus has been on worsening health as a cause of economic inactivity

In recent months, national headlines have largely focused on the role that poor health plays in economic inactivity.7 It is undoubtedly a concern, as long-term sickness is now given as the main reason why people are outside the labour force. Quarterly national data for October 2022 shows about 28 per cent of inactive individuals reported being outside the labour market for health reasons – that’s about 2.5 million people.

Health does appear to be a driver of long-run patterns of involuntary inactivity. Looking at places with the highest share of inactivity driven by ill health shows, once again, a clear North-South divide, matching the geography of hidden unemployment (Figure 7). In Newport and Sunderland, for example, more than 40 per cent of all inactive people are not seeking work because of poor health compared with less than 15 per cent in Aldershot and Norwich.

Figure 7: Cities in the North have higher shares of inactivity due to poor health

Source: ONS, Annual Population Survey 2022.

In many places, particularly in the North, poor health among the workforce and the working-age population is not a new issue. Cities like Sunderland, Newcastle and Barnsley have historically had higher than average rates of inactivity because of sickness. In the 1980s and 1990s, it was especially acute among former coal miners, steelworkers and other industrial workers. Exposure to industrial injury led many to become inactive, often claiming incapacity benefits, when their industries shut down.8

These patterns have persisted.9 ONS data for 2022 shows people who previously worked in wholesale retail and trade, transportation, construction and manufacturing were most likely to be inactive due to poor health.10 And this could be for two reasons. Firstly, many of these jobs still tend to be physically demanding and are more likely to impair health than most white-collar roles. Secondly, jobs in these industries are harder to carry out while managing long-term sickness and are less adaptable to hybrid and homeworking, which compels workers to withdraw from the labour market. What is less clear is the role of health in explaining more recent increases. While long Covid and NHS backlogs have been put forward as explanations, data on NHS waiting times doesn’t correlate very well with changes in involuntary inactivity across cities (see Box 5).

Box 5: There are spatial inequalities in access to health care and treatment across the UK

Access to health care has worsened in the UK. This is due, in part, to the pandemic but is also the result of a decade of austerity.11 Recently, the NHS backlog for routine care hit a record high of seven million people in England, and the median wait time for treatment is now 14 weeks, compared with seven weeks before the pandemic started.

This is likely to play a role in health-related inactivity. About a fifth of people aged 50-64 who have left the labour market since the beginning of the pandemic were on the NHS waiting list for medical treatment.

Health care availability is not uniform across the country; analysis of referral-to-treatment wait times at the Hospital Trust level shows large spatial disparities. The share of patients waiting less than 18 weeks goes from just above 50 per cent in cities like Telford, Birmingham and Exeter to nearly 75 per cent in Sunderland, Barnsley and Newcastle. That said, initial calculations suggest there is no clear relationship between inequalities in access to healthcare and recent increases in health-related inactivity, but it still warrants further analysis.

Economic factors have been less widely discussed, but also play a major role

Skills are an important predictor of both employment outcomes and economic inactivity. Nationally, unemployment rates are two percentage points higher for people with no qualifications compared to those who have a degree.12 About half of people with no qualifications are economically inactive (for all reasons combined), against 11 per cent of those with Level 4 and above.13

This illustrates two potential impacts on people – the ‘scarring’ of long-term worklessness on employment prospects. This can lead people, particularly those with low educational attainment, to become ‘discouraged workers’ and leave the labour force, or they don’t enter the workforce in the first place because of the lack of opportunities.14

The disadvantages faced by people with low skills or educational attainment in the jobs market help explain why cities with the highest percentage of residents with no formal qualifications also have the largest hidden unemployment rates (Figure 8). Many, like Hull, Sunderland and Barnsley, are in the North.

Figure 8: Low skills lead to higher levels of hidden unemployment

Source: ONS, Annual Population Survey 2022, Labour Force Survey 2021.

But this is only part of the explanation. As Figure 9 shows, employment rates are determined by the strength of the local economy.15 Low-skilled people are still more likely to be employed in economically stronger cities (with a high share of knowledge-intensive jobs) like Reading and Edinburgh than in Sunderland and Mansfield.

Figure 9: Employment rates for people with low skills are higher in stronger economies

Source: Annual Population Survey, 2019.

In northern cities and large towns with high hidden unemployment rates and relatively weaker labour markets, demand for workers – and job opportunities – is limited. As the likelihood of finding work is lower (leading to higher unemployment), the incentives to stay active in the labour market and look for a job are reduced too (leading to higher inactivity). Figure 10 shows the interaction between demand for labour and hidden unemployment across Britain’s cities and largest towns. Cities are broadly split into two groups.16

Figure 10: Cities in the North are more likely to face a job shortage, while those in the South face a labour shortage

Source: Indeed, 2022. ONS, Annual Population Survey, 2022.

Weaker economies – those above the dotted line – account for a larger share of Britain’s hidden unemployment relative to their share of job postings.17 This suggests they are suffering from a job shortage, with too few available compared to the size of the hidden unemployed population. There are 38 cities in this category and most are in the North or Wales. Middlesbrough, for instance, accounts for only 0.5 per cent of all job postings in Britain, but around 1 per cent of all hidden unemployed. In total, all cities in the North or Wales account for 16 per cent of vacancies but 21 per cent of hidden unemployed.

The problem in many of these places isn’t just that they have too few jobs, it’s that many are low-skilled, low-paid roles. Cities like Sunderland, Barnsley and Hull have a far higher share of those than places like Reading, Cambridge or Edinburgh. This is a symptom of the weakness of their local labour market and there are two reasons why it is likely to contribute to economic inactivity. Firstly, the prospect of low pay may deter people from entering the jobs market in the first place and secondly, many low-skilled roles tend to be physically demanding so increasing the chances of people withdrawing from the labour market for health issues.

This relationship between hidden unemployment and the strength of the economy illustrates, in part, the legacy of deindustrialisation, as Box 4 noted. Many post-industrial cities, particularly in the North, have not transitioned towards high-skilled, high-wage economies. As industry declined, policy interventions failed to provide these places with the tools to adapt to (and attract) more knowledge-based services, for instance by investing in the skills of the workforce. Instead, they have replicated the low-skilled nature of their economies.18 As a result, some places have not only lost jobs in mining and manufacturing, which were relatively well-paid, but the jobs that have subsequently been created are disproportionately lower skilled and lower paid in workplaces like distribution sheds and call centres.19 It has meant that deindustrialised cities such as Sunderland, Barnsley and Hull have been stuck in a low-skill, low-wage equilibrium.

This is in stark contrast to the stronger economy cities below the dotted line of Figure 10, which have the opposite mismatch. Their share of the UK’s total job postings is higher than their proportion of hidden unemployed. Bristol, for instance, accounts for 2.1 per cent of all UK vacancies, despite representing only 0.9 per cent of all hidden unemployed. In London, it is 23.6 per cent and 16.1 per cent respectively. This does not necessarily mean that, in absolute terms, there are more vacancies than people outside the labour force in these cities, but it suggests their jobs markets have more capacity to accommodate those currently inactive than weaker city economies.

The UK’s long-term inactivity problem is largely one of job shortages

Addressing recent increases in inactivity has tended to be framed in terms of bringing people back to the jobs market to fill open vacancies. But in many cities the bigger challenge is a long-term jobs shortage rather than a short-term worker shortage. And this is before the expected recession further worsens the picture.

If policy is to tackle the UK’s true inactivity problem, it will need to address the lack of work in many northern cities and large towns in particular, alongside health and skills issues. In cities like Middlesbrough, Hull and Blackburn, there is more slack in the jobs market than headline figures would suggest. This means their economies are unlikely to be able to absorb higher levels of employment should those outside the labour force decide to return to paid work, because currently there aren’t enough jobs for them.

The expected recession is likely to put even greater short-term emphasis on policies to deal with the fall-out from it, such as employment support schemes. But the jobs shortage problem shows how important delivering on the levelling up agenda is if we are to turn around the struggles of many places outside the Greater South East.

This requires the Government to deliver on its levelling up promises

The levelling up agenda has been derailed by a series of short-term crises since the current Government took office. Covid was the biggest of these, followed by the immediate response to the war in Ukraine and the rising cost of living. A recession and the recent increase in inactivity are likely to be the latest reasons as to why policy attention is diverted away from tackling the UK’s entrenched geographic divides.

The constant firefighting of short-term problems means there has been almost no policy delivery to address the long-term issues of poor national and local economic growth. The analysis in this chapter shows the implications of continued inaction, with hidden unemployment rates up to four times higher in places like Middlesbrough, Blackburn and Hull than official figures suggest.

There needs to be a programme of delivery set out in 2023 that tackles the reasons why there is a lack of jobs, skills and good health in a number of places outside the Greater South East, in particular.

Last February’s Levelling Up White Paper did a good job of diagnosing the problems and setting out a broad approach to addressing them. Now, 2023 needs to be the year this is followed up with policy implementation.


  • 2 While time series data on unemployment at the local level starts in 2004, claimant count data shows that the claimant rate is close to its lowest point since the data series began in 1986.
  • 3 Between 2019 and the first quarter of 2022, levels of inactivity fell by 0.7 percentage points in the median OECD nation, while rising in the UK over the same period. Also see Financial Times, ‘UK lags behind developed nations on post-Covid employment recovery,’ 9 November 2022.
  • 4 This is computed by removing students, retirees and people looking after family and home from inactivity figures.
  • 5 Beatty, C et al (2022). The real level of unemployment 2022, Sheffield Hallam University.
  • 6 In 1971, nearly 45 per cent of woman aged 16-64 were economically inactive. This fell in the decades that followed, to 24 per cent by December 2019.
  • 7 For example, see Financial Times, 7 October 2022, ‘Half a million missing workers show modern Britain’s failings’ and ONS, ‘Half a million more people are out of the labour force because of long-term sickness’ (10 November 2022).
  • 8 8 Beatty C, Fothergill S and Gore T(2019), The State of the coalfields, Sheffield: Sheffield Hallam University.
  • 9 Average life expectancy in the former coalfields is still around a year less than the national average. See Beatty C, Fothergill S and Gore T(2019), The State of the coalfields, Sheffield: Sheffield Hallam University.
  • 10 2022, ‘Half a million more people are out of the labour force because of long-term sickness’.
  • 11 In the decade leading up to the pandemic, per capita spending on health increased by just 0.4 per cent a year on average (including four years in which it actually fell). This compares with a 5.6 per cent average growth in the years between 1997 and 2010. Source: Appleby J and Gainsbury S (2022), The past, present and future of government spending on the NHS, Nuffield Trust.
  • 12 About 5 per cent of 16-64 year olds without qualifications are unemployed, against 3 per cent of those with NVQ4+ (ONS, 2022).
  • 13 This excludes people in full-time education. Source: Annual Population Survey 2022.
  • 14 See ONS, 2021: Which groups find it hardest to find a job following a period out of work?
  • 15 The fact that low-skilled people have better economic outcomes in cities where the economy is stronger is an aggregate demand effect: the presence of high-value, high-paid jobs in the most productive sectors fuels demand in the local services sectors, which indirectly creates low-skilled jobs. This has been referred to as the ‘multiplier effect’. See Magrini E (2019), Opportunity knocks? London: Centre for Cities.
  • 16 Research from the Institute for Employment Studies has shown that below the national headline, only about twofifths of local authorities had more vacancies than unemployed people. In contrast, in about a third of all UK local authorities, there were more than twice as many unemployed as there were vacancies. See Wilson, T and Williams, M (2022), Work Local: labour market analysis, Brighton: Institute for Employment Studies.
  • 17 A more direct metric would have been comparing absolute figures, but the total number of vacancies at the PUA level was unavailable.
  • 18 Swinney P and Thomas E (2015), A century of cities. Urban economic change since 1911, London: Centre for Cities.
  • 19 Salvatori, A (2018) The anatomy of job polarisation in the UK. Journal for Labour Market Research