Social, economic, commercial and environmental conditions are the strongest determinants of people’s health. This includes people’s access to homes that are safe, stable and warm; the availability of an adequate financial safety net; access to healthy, affordable food; and the quality of their work.

This long read presents new analysis exploring changes in one of these wider determinants of health: the labour market. This includes the quantity and quality of employment over the last 10 years, the implications of these changes for health, and what they mean for public policy. It is the first in a series addressing changes in the wider determinants of health and what they mean for health inequalities.

Since the 2008 recession, the UK labour market has performed well in terms of generating employment. Employment is at record high levels. Unemployment poses a range of risks for health and wellbeing, so on the face of it, high employment rates are good news.

However, this interpretation overlooks the levels of low-quality work experienced by workers in the UK, which can also pose a risk to health. And while much of the debate on low-quality jobs has focused on insecurity and zero-hours contracts, low-quality employment goes beyond insecurity.

Recent political developments have thrown this debate on low-quality jobs into sharper relief: given the UK has left the EU, and the most recent withdrawal agreement bill did not include clauses protecting EU-derived employment rights. Instead, the Queen’s Speech has promised an employment bill, promising to ‘protect and enhance workers’ rights as the UK leaves the EU’. This presents an opportunity to take a fresh look at what we mean by employment and how the quality of work can affect health.

The presence, adequacy and quality of work matter for our health directly, as well as underpinning other determinants of health such as income or social networks.

Unemployment and health

Research has often focused on the role of unemployment as a hazard for people’s health, driven by the high unemployment crises of the past.

Unemployment harms health in many ways:

• It is a source of stress and can harm mental health.
• It can result in unhealthy coping behaviours, such as smoking and drinking (although it also reduces the resources available to spend on these).
• It can cause poverty, which is in itself damaging to health.
• A spell of unemployment can affect future employment prospects.

These effects become more pronounced as the duration of unemployment increases. People also experience a loss of the health-promoting aspects that good work can offer, such as social connections or a sense of structure and purpose.

Low-quality work and health

Concerns about the health risks of low-quality work are not new. Historically, these have tended to focus on potential hazards in agricultural and industrial workplaces. These areas are now subject to substantial regulation establishing health and safety standards. This means the risks of fatal and non-fatal injury today are far lower than they were 20 or 30 years ago. There are still physical risks linked to work in the UK, experienced by 31% of workers in 2015, but these fall into broader categories, including prolonged sitting and working with screens.

Aside from physical risks, an important aspect of low-quality work is the psychosocial environment. Broadly, this relates to how aspects of our lives can act as sources of stress and eventually manifest themselves physiologically.

Some of these findings about work and health stem from the Whitehall II Study. This found a social gradient in health outcomes even among civil servants, with the highest grades having the best health outcomes. The two main theoretical models for understanding this are:

• Demand control: While the demands on more senior workers can be higher, the amount of autonomy and resource they have to respond to these demands is also higher, reducing the stress burden.
• Effort–reward imbalance: Those with higher quality jobs receive more in return for their efforts in terms of satisfaction, promotion or remuneration. For those with low-quality jobs, there are other potential stressors at play that can contribute to worse health outcomes, such as insecurity and unpredictability of work.

The longer the time spent in low-quality work, the higher the risk of accumulating health problems, as the damage from this as a source of stress accumulates.

Identifying causality in this area is difficult. Does low-quality work lead to poor health, or are people more likely to enter low-quality work because their health prevents them from acquiring better work? It may also be the case that a third factor influences both, for example, individual short-termism leading to worse health and lower quality work. The relationship is complicated and likely to interact in both directions over time.

Studies that use longitudinal data and use external shocks as part of their methods, which partially account for some of these problems, provide evidence of some impact of work quality on health outcomes. While the precise extent and direction of causality is difficult to pin down, for policymakers the fact people with poorer health are more likely to be in low-quality jobs should be a key a concern.

As it stands, and providing that the economy does not imminently enter recession, unemployment is no longer the labour market problem that should be giving policymakers the most restless nights. Concern about the quality of jobs has taken its place.

We can see these concerns playing out in the media and through the political attention on job quality, rising up the policy agenda in recent years: zero-hours contracts, the gig economy (short-term and freelance work), in-work poverty, and a lack of progression in employment. These changes – or at least shifts in relative importance to unemployment – may have consequences for health and for how government might intervene to address the wider determinants of health.

In some ways, the UK labour market is unrecognisable compared with 10 years ago. In the June-to-August of 2019, the working-age employment rate was 75.9% – close to the record employment rate going back to 1971. 5 years ago, the rate was around 73%. Even that – which was above average for the pre-recession 2000s – would mean around 1.4 million fewer people in work if it applied today. 
This is quite remarkable given the major recession at the end of the last decade and prolonged period of underwhelming growth in productivity and the wider economy.

However, this success focuses on one aspect of employment: the quantity of employment. Alongside the reduction in unemployment there has been considerably less progress in two other areas:

• Rate of pay: There has been a historically prolonged period of weak pay growth, which has barely recovered in real terms.
• Underemployment: As Figure 1 indicates, unemployment is below the levels of the pre-recession boom, at around 2.5% of those aged 16 and older (around 1.3 million people). However, it is a different picture for underemployment. Underemployment – where people are in work but not working as many hours as they would like – remains above pre-recession levels, at nearly 5% of the adult population (around 2.4 million people).

Even given the current high levels of employment, there is still a public policy problem related to unemployment – especially for people in long-term unemployment. The proportion who are long-term unemployed has fallen, but of those who are unemployed, around 41% have still been out of work for more than 6 months.

There are also still inequalities in the experience of unemployment. The maps in Figure 2 show that people’s risk of unemployment still varies by where they live (data shown by local authority area). Inequalities by characteristics such as ethnicity or age have narrowed, but they are still often considerable. Unemployment is defined as when someone is actively seeking and available to start working. Here it is expressed as a proportion of the working-age population (rather than the economically active population), sometimes called an unemployment ratio.

Although the difference between the local authority with the lowest and highest unemployment has reduced, it is still considerable: the local authority with the lowest proportion of working-age adults who unemployed has a figure of 1.8%, whereas the council with the highest has 7.4% of its working-age population unemployed. What is more, low unemployment cannot be taken for granted. The next recession may not play out in the same way as 2008/09, when there was a relatively small increase in unemployment given the size of the reduction in the economy

As unemployment as an issue has receded in recent political debate, the question of work quality has risen. This is normally focused on low pay or insecurity, such as insecure types of employment contracts like zero-hours contracts or self-employment. Trends in the number of workers with some of these contract types are shown in Figure 3.

While there have been increases in self-employment and potentially zero-hours contract (although there are concerns with how much of the increase in the statistics was due to increased awareness), not all these contracts can be considered ‘insecure’. These findings are explained in more detail in Box 2.

Note: These figures are not exclusive and cannot be added together. For example, many people on a zero-hours contract are also on a temporary contract.

Overall, there does not seem to be a step change in the number of people with potentially insecure contracts in recent years. The rise occurred earlier, in the period immediately after the recession, but has since levelled off. But insecurity is only one aspect of low job quality, even if this is the most often discussed.

Insecurity and low pay have taken the limelight in discussions of job quality. A sharply rising minimum wage and strong political commitments across the political divide have helped to address low pay. But these do not reflect the totality of job quality and how job quality can influence health.

What comprises good-quality work is not an easy question to answer, and is even harder to measure. There have been a few attempts at quantifying what comprises a decent job. The recent Taylor review of modern working practices used the EU QuInnE (Quality of jobs and Innovation Generated Employment Outcomes) measure. However, it also recommended that the UK develop a new definition and measure it.

This was attempted by the Carnegie Trust and the RSA. They covered various domains of work quality, such as employment contract types, support at work, security and stability, and work-life balance. These categories indicate an important insight: that a job can be secure and still low quality. Poor-quality work is not confined to zero-hours contracts.

Impact of different dimensions of employment

The Carnegie Trust/RSA measure is limited due to lack of available data, so we use a slightly modified version of measures of work quality using the measure developed by Chandola and Zhang. This draws on self-reported aspects of employment: autonomy, job wellbeing, pay, security and satisfaction. These are defined Box 3.

These aspects are associated with people’s health. Figure 4 shows the relationship between self-rated health for those employees who do not have ‘good’ health (meaning they rated their health as poor or fair, rather than good, very good or excellent) – known to be strongly associated with objective health outcomes. The proportion of people rating their health as ‘non-good’ is presented for each aspect of low job quality.

Figure 4 shows that 17% of employees with low job security do not have good health – almost twice the share for all employees (11%). For any individual aspect of low job quality, a higher proportion report a lack of good health – from 12% for those with low pay to 19% for those with low job satisfaction. Overall, the prevalence of less than good health is twice as high for those with two or more negative job aspects (15%) than for those with no negative job aspects (7%).

Figure 4 shows the correlation between poor health and poor-quality work, and does not directly account for various other factors, such as age or household income. Chandola and Zhang’s study does take such factors into account. It also looked at biomarkers of health rather than just self-rated health, indicators of possible health problems such as blood pressure. The study looked at the changes in certain biomarkers of health when unemployed people either remained unemployed, entered good-quality work or entered low-quality work. The study found that predicted levels of allostatic load – wear and tear on the body from chronic stress – were lowest for those who moved into work with no negative job aspects, or only one, and higher for those who moved into work with multiple adverse job aspects than those who remained unemployed. The changes in health associated with these changes do not conclusively demonstrate causality but suggest that low-quality work might actually be worse for health than unemployment.

How common is low-quality work? 

A higher number of negative job aspects experienced in one job are associated with worse self-rated health.

Given the evidence that these measures of job quality are associated with health outcomes, it is concerning that in 2016/17: 

• only around 27% of UK employees reported no negative aspect to their job in line with the measures discussed in Figure 4 
• 36% reported two or more negative aspects – 24% had two negative aspects, and 13% had three negative aspects.

Figure 5 looks at how many employees in the UK experience each of these dimensions.

The least common dimension of low job quality, at 7%, was job security (based on a subjective assessment of the likelihood of losing employment within 12 months), but job security is most associated with poor health (Figure 4). Job security fell more than other dimensions of job quality between 2010/11 and 2016/17: from 12% to 7%. This may be because job security is the most obvious aspect that would change with higher employment figures over this period, although there did seem to be slight improvements across all dimensions up to 2014/15. 

In contrast, low autonomy is the most common dimension of low job quality, at half of employees. Between 2010 and 2016 there was hardly any change in prevalence across the workforce, and this dimension is least associated with poor health (Figure 4).

Overall, the prevalence of low-quality work has fallen somewhat as a proportion of all employees – from 40% with multiple negative aspects in 2010/11 to 36%, driven mostly by job security.

There are also inequalities in exposure to low-quality work (Figure 6). Certain groups are more likely to be in low-quality work: younger adults, people in more routine occupations, and members of black and minority ethnic groups generally, but particularly those of mixed ethnicity, Pakistani, Bangladeshi and African Caribbean heritage. There are also significant geographical variations with Northern Ireland, Wales, the north east, and west midlands all having high levels of low-quality work.

Overall, from this analysis we can draw two key insights:

• The extent of low-quality work is wider than one would expect based on more conventional measures such as low pay or type of employment contract.
• Labour market success should be measured by job quality as well as levels of employment.

Duration of low-quality work 

It is also important to take into account how long people spend in low-quality work. The health damage caused by stress – such as from low-quality work – accumulates over time, leading to worse health outcomes. The duration for which people are exposed to stressors may be an important indicator of health outcomes.

Figure 7 examines this issue in relation to the aspects of work quality described above, such as autonomy and wellbeing. It indicates the proportion of people who were in low-quality work in 2010/11 that were still in low-quality work in 2016/17. It also shows those who ‘escape’ low-quality work and the transitions experienced by those in high-quality work.

Figure 7

Figure 7 shows that 51% of those in low-quality work in the first period (2010/11) were still in low-quality work 6 years later. This category includes people who were ‘stuck’ (in other words, in low-quality work for the entire period) and people who may have moved in and out of low-quality work. A further 27% of people who were in higher quality work in the first period entered low-quality work, or the quality of their job deteriorated.

Because this analysis looks at those in work in both periods, it excludes some other groups, such as people who left poor quality work for health reasons in between the two reference years.

Overall, then, low-quality work – as perceived by employees themselves – is more common than discussions of job insecurity might suggest. And it is unequally distributed across society, both in terms of geography and demography, reflecting broader inequalities. The strong labour market over this period has coincided with a small decrease in the proportion of jobs that are low quality, though in the most recent year of data appears to have partially reversed. Job quality is also quite stable: those in good-quality work tend to stay in good-quality work, and around half of those in low-quality work tend to stay there too.

The extent of low-quality work, the inequalities embodied within it, and the duration people spend in it are all likely to contribute to worse health outcomes overall, and to inequalities in health outcomes.

In recent years, unemployment has receded as a problem for the UK labour market as a whole, though it still affects some population groups disproportionately. The quality of work, however, remains a significant, widespread issue. Despite the apparent strength of the labour market in recent years, employment growth has not dramatically increased the proportion of people acquiring good-quality work and the improvement looks to have stalled. The focus of policy should now be on the challenge low-quality work presents for health.

In recent years, analysis of low-quality work has focused particularly on job insecurity, for example zero-hours contracts and the gig economy. It is clear from this new analysis that this focus does not reflect the breadth of the problem.

Our recent report Creating healthy lives makes the case for the whole of government and other sectors to take action to create the conditions for good health. A strategy on the quality– not just quantity – of employment should be one of the cornerstones of such an approach. A focus on quality of work poses a challenge to government at different levels, which is more familiar with measures to reduce unemployment than enabling high-quality work but there is much that could be done.

As the UK reviews its post-EU employment law, there is an opportunity to ensure that employment rights protect and improve health by improving the quality of work. It is also important for employers to consider job design such as improving in-work progression to prevent people becoming stuck in low-quality work. Broader solutions might be found in rethinking aspects of corporate governance or industrial strategy. The forthcoming report The Marmot Review 10 years on, supported by the Health Foundation, will also consider how to improve work quality.