The TUC has talked of “an increasingly insecure, vulnerable workforce”, the CBI of “contracts [which] play a vital role as a way of keeping people in employment”. Who’s right? One side? Another? Neither? Both?
With UK employment at around 75%, the government has a headline figure on its side when it wants to talk about its economic record. It’s lower than it was before COVID-19, but given that the figure for January 2020 was an all-time high of 76.6%, it’s also not exactly disastrous. By this measure, the economy is healthy. Still, it’s the job of research to look beyond a headline and consider the finer details. What sorts of jobs are in the gig economy and are they good for people?
Let’s start by defining our terms. ‘Gig’ contracts are those where “people agree to be available for work as and when required, but have no guaranteed hours or times of work …[providing] employers with a pool of people who are ‘on-call’ and can be used when the need arises”. ONS estimates that there were just over a million people on such contracts at the end of 2021, or 3.2% of the workforce.
Who is on zero-hours contracts?
Thomas Keely at the University of Aberdeen used data from Understanding Society to examine whether zero-hours contracts were having an impact on people’s mental health (of which more later), but his paper is worth reading in full if you’re interested in zero-hours contracts, and precarious employment generally, for its literature review which sets out many years of findings. It also has a useful summary of who is employed in this way.
On average, he found, gig workers were “younger, more female, work fewer hours (more concentrated in part-time work) in lower-status jobs, are more private-sector focused, and have lower educational achievement.” Around a fifth of the sample were full-time students (who made up just 2% of employees in the sample who weren’t on zero-hours contracts).
Interestingly, while we might think of these jobs as being menial or low-skilled, 22% of zero-hours workers have managerial status. However, “even though one-fifth of the zero-hours sample are classed as a manager, only 4% feel they have the duties of one … [which] suggests that zero-hours jobs are … associated with lower job control”.
Do they boost employment?
The defence of zero-hours contracts usually hinges on the idea that they are good for people and the economy because they increase the number of jobs available. Let’s examine this claim in more detail.
Research by Silvia Avram at the Institute for Social and Economic Research showed that the UK has indeed “experienced strong employment growth” in the last decade (after an unsurprising dip around the time of the 2008 financial crash). But there has also been “an increase in atypical non-standard forms of employment, such as zero-hours contracts, on-call and agency work, and the gig economy … [which] have in common the fact that workers are not guaranteed a minimum level of work and pay or only a very low one”.
Her November 2020 report, Hours and pay insecurity in the UK labour market, funded by the Nuffield Foundation, used Understanding Society and Labour Force Survey (LFS) data to ask:
- do jobs with unstable hours and pay help the unemployed move into work?
- do they benefit groups that typically face barriers to work (such as lone parents, mothers with small children, the low skilled or the long-term unemployed)?
On the first question, Avram wrote that flexibility is seen as a virtue in the labour market because if employers can’t offer these jobs, “peripheral workers might simply end up unemployed”. We should expect, then, that people would be unemployed for shorter periods if they lived in areas with a greater number of unstable jobs. In fact, LFS data showed the opposite: in areas with a larger share of unstable work, people stayed unemployed longer. This could be down to local conditions – “weaker regional economies may have both higher unemployment and more unstable jobs”.
To get a more detailed answer, Avram used Understanding Society data to look at the likelihood of being unemployed for people with GCSEs or lower educational qualifications facing different levels of job instability. This chart is from a newer version of the analysis in her paper Unstable jobs and time out of work: evidence from the UK.
Her analysis shows that women with lower levels of education in areas with a higher share of unstable jobs were more likely to leave unemployment only when they had been unemployed for five years or more. For shorter spells of unemployment, the differences were minimal. For low-educated men, there was hardly any difference at all.
A further set of modelling using ‘discrete time survival’ (looking at distinct chunks of time) took into account many of the factors in Silvia’s second question about barriers to employment: education, number of children under 14 in the household, single parenthood status, poor health etc. Again, there was little or no difference between areas with larger or smaller numbers of unstable jobs.
Given that she was using a sample size of almost 4,500 men and over 8,600 women, the findings do give weight to the idea that zero-hours work may not, in fact, help people get jobs.
What does ‘zero hours’ mean for mental health?
Thomas Keely used our data to examine zero-hours contracts and mental health. Understanding Society is ideal for this, because we ask about people’s employment status and the type of contract they’re on – and use the GHQ-12 questionnaire to measure mental health in each wave of data collection.
He found “a highly significant result that the mental health of zero-hours workers is, on average, 12.2% worse than other workers”. Keely couldn’t rule out endogeneity – that is, the possibility that poor mental health was pre-existing, or had another cause – but his paper “corroborates previous findings that [zero-hours contracts] are detrimental to health”.
He was also able to test the ways in which this kind of job might affect people’s wellbeing – looking at sources of stress such as job insecurity, job (dis)satisfaction, unpredictable working patterns, and insecure income. Keely found that only income insecurity and job satisfaction could definitely be pinpointed as mechanisms by which the harm to wellbeing was happening.
There is a logic to the job security finding: although “theory and literature suggest that job insecurity is prevalent in zero-hours work and that it is a threat to wellbeing”, a job doesn’t have to be zero-hours to be insecure, so this mechanism would not be unique to this kind of work. It is surprising, though, that unpredictable hours (or “schedule insecurity”) weren’t confirmed to be a factor – but this may be because our Study asks if people have a “usual working pattern”. This may not capture unexpected changes to one’s shift, cancellations, or ‘zeroing down’ – which is where people find their work hours gradually reduced by their employer as a punishment for turning down one or more shifts, and which Keely says is “arguably one of the greatest sources of stress within [these contracts]”.
Good quality work
The evidence we’ve seen so far, then suggests that zero-hours contracts don’t help people into regular work, and aren’t good for mental health. Another paper, using Understanding Society’s biomarker data, adds a further dimension. Tarani Chandola and Nan Zhang at the Cathie Marsh Institute in Manchester looked at the idea that any work is better than no work, and found higher levels of chronic stress among adults who moved into low quality work than among those who remained unemployed.
They didn’t explicitly look at the gig economy, but it seems reasonable to assume that zero-hours work and low-quality work are at least areas with significant overlap. The work is also given credence by its use of biomarkers – objective measures such as cholesterol, proteins in the blood, blood pressure, and body fat distribution. Taken together, they measure the wear and tear on the body which can be a sign of serious physical and mental health conditions to come.
The research concluded that job quality must be a consideration when we think about people’s ‘success’ in leaving unemployment and finding a job. While they may stop claiming unemployment benefits, the quality of the job they find may have significant implications for their long-term health and wellbeing – and these also represent a cost to the state.
‘Choosing’ flexible work
One interesting point in Keely’s paper is a section on whether people with existing health conditions might choose zero-hours contracts because flexible work fits more easily around a challenging life. He called the argument “unconvincing… It is unlikely that those of poor health would self-select into zero-hours jobs out of flexibility reasons, when the flexibility offered by [them] is restricted and of a specific scope”. He went on to point out that a TUC survey in 2017 found that “43% of zero-hours staff say they are on a [zero-hours contract] because it is their only option”.
The idea of choice was also something Silvia Avram considered, this time not with our data, but through an experiment in which 301 people used a custom-made app to “choose between receiving a fixed payment, a ‘benefit’, or completing a work task for higher pay.” Participants found themselves in a control group (where work was always available, paying £1.50 for each task) or one of two treatment groups, and had to decide whether to work at the beginning of each round, before they knew if work was available. (This was decided on a coin toss, so it was a 50/50 chance.)
In the first treatment group, the money was the same, £1.50 per task – so they got the same money as the control group, but could rely on half as much work. In the second, they got £3 per task, so could expect to earn as much as the control group for working half as much.
At first, everyone could choose to receive the benefit – £1 – instead of working. In the second half of the experiment, though, they could only get this if there was no work available. Also, in this later phase, one of the treatment groups could face sanctions in the form of benefit loss if they chose not to work twice in a row.
The results showed that people “who faced uncertainty about work availability were between 15 and 30 percentage points less likely to choose to work compared to participants who could always work. This result was only in part driven by financial incentives. Participants continued to choose work less often even when their pay rate was doubled so their expected financial gain was the same as that of the no uncertainty group.”
One finding in particular is relevant for welfare policy: whether benefits were available or subject to sanctions, “participants facing uncertainty were much less likely to choose to work”.
In a blog about the research, Avram backs up Keely’s point about choice, saying: “The results suggest uncertainty is a burden for workers and contradict the view they willingly choose unstable jobs.”
An explosion in uncertain work?
Having said all this, one question we should tackle is the issue of how much of a ‘problem’ zero-hours contracts are. It seems likely, from this evidence, that they are affecting people’s mental health, and aren’t necessarily helping people into work. To what extent, though, are they a social ill?
In a piece for the Guardian earlier this year, Torsten Bell, Chief Executive of the Resolution Foundation, suggested that instability was not the problem we might think, observing: “We have seen a growth in less secure work, from zero-hours contracts to self-employment, but firms are much less likely to fire these days. In the late 1990s, 0.8% of workers would lose their jobs each quarter; immediately pre-pandemic that had halved. Modern business has nothing in common with The Apprentice.”
There was, apparently, a significant rise in the prevalence of zero-hours contracts about ten years ago, with the Labour Force Survey suggesting an increase of 132% between 2012 and 2013. The House of Commons Library’s briefing paper on the subject, though, says “this very likely reflected better awareness of zero-hours contracts as a result of extensive media coverage in 2013”. This argument that there was greater awareness and recognition of the term is supported by the fact that “around half of the reported change in zero-hours contracts between 2012 and 2013 came from people who said they had been in their job for over a year, which suggests that some respondents must have failed to report or identify that they were on a zero-hours contract in previous surveys”.
Potential policy approaches
Under Jeremy Corbyn, Labour planned to ban zero-hours contracts outright, but this, Keely said, would have been counterproductive. “Zero-hours contracts have potential to be a valuable tool used which can benefit both firms and employees”, he wrote. “Indeed, there are some groups of people who would prefer to be on this contract type, for example, students or semi-retirees.”
That does not mean that they are perfect, of course. As long as they can harm wellbeing, Keely suggests allowing zero-hours contracts but offering more secure work to people who do regular hours, or would simply prefer not to be doing zero-hours work. “This would shift the dynamic, putting more control in employees’ hands and allowing greater flexibility to those who need it and greater security to others.”
He also suggests that employers can give workers more control over their schedules to mitigate the effects of schedule insecurity – and he calls ‘zeroing down’ simply “unethical”, arguing that it is “harmful to put all the pressure of flexibility onto staff, punishing them when their schedules do not align with that of the company”.
One change which has been made to zero-hours contracts came in 2015, with regulations which ‘banned’ exclusivity clauses. These prevented people who were signed up to them from seeking extra work with another employer if they weren’t getting enough hours or salary from the one they were contracted to. (To be strictly accurate, the regulations didn’t ban this practice, but introduced measures designed to make such clauses unenforceable.) We know, then, that it is possible to introduce or amend legislation in such a way as to encourage greater cooperation and balance between employers and employees.
Avram suggests that “Employers could be mandated to offer workers a contract that reflects their regular working hours rather than a minimum. Limits on short-term scheduling and re-scheduling, as well as compensation for shifts cancelled at short notice would similarly help.
“The Government could also mandate a higher minimum pay whenever hours vary in the short-term. Finally, the welfare system can support workers in unstable jobs by promptly topping up their pay when work is unavailable. At the moment, arrears in Universal Credit payments amplify swings in income.”
What next?
The challenge for any government is to balance the competing interests of different parties – in this case, workers and employers. That might be a challenge in a polarised political landscape, but shouldn’t be impossible.
Ian Brinkley, now at The Work Foundation, wrote a blog, In defence of zero hours contracts, when he was Acting Chief Economist at the Chartered Institute of Personnel and Development. It says “Zero hours do not work for everyone. Their use should always be dependent on a robust business case”. He pointed out, too, that “Bad employers can use zero hours to coerce their employees by threatening to withhold work, but it is naïve to think that bad employers would not carry on with bad practices regardless of what the contract says”.
No legislation can make the world perfect, but the evidence suggests there are still adjustments we could make to improve the world for people on low incomes and in precarious work.
Zeroed Down: The Effects of Zero Hours Contracts on Mental Health and The Mechanisms Behind Them, Thomas Keely, University of Aberdeen Discussion Paper, September 2021
Hours and pay insecurity in the UK labour market, Silvia Avram, University of Essex, November 2020
Unstable jobs and time out of work: evidence from the UK, Silvia Avram, Socio-Economic Review, March 2022
Re-employment, job quality, health and allostatic load biomarkers: prospective evidence from the UK Household Longitudinal Study, Tarani Chandola, Nan Zhang, International Journal of Epidemiology, February 2018
More information about our biomarker data
To investigate potential research questions of your own, look at our questionnaires and our variable search to see the topics we cover.
Authors
Chris Coates
Chris is Research Impact and Project Manager at Understanding Society
