New JRF analysis comparing online job vacancies with the profile of unemployed people paints a worrying picture of competition for jobs, especially for people laid off from lower-paid roles.
Throughout the chaotic economic storm brought on by COVID-19 it has often been people in low-paid, insecure jobs whose work is most at risk. While the frightening worst-case scenarios of unemployment have been prevented by some successful labour market interventions, most notably the Coronavirus Job Retention Scheme (CJRS), many people have still lost their jobs. The latest unemployment rate is 5%, meaning 1.7 million people are now looking for work, and most economic forecasts expect this to rise significantly in 2021 if the CJRS comes to an end this spring as currently planned.
We know that being unemployed can have severe long-term consequences, lowering employability and wage prospects and ultimately raising the risk of future unemployment. That’s why avoiding mass unemployment has been, and continues to be, one of the highest priorities in these turbulent times. Losing your job can be particularly damaging for young people and workers with few qualifications.
The combination of these two facts should worry us all. It’s a major concern that workers who are either in poverty already or close to being swept into it are both disproportionately likely to become unemployed and more likely to be negatively impacted by this experience. Supporting a swift return to work will be a priority for policy in the months ahead if we are to ensure that people struggling to keep their heads above water are not left to sink. Doing so effectively will require a good understanding of the opportunities on offer to these workers.
We’ve taken a look at what online job vacancies and the profile of unemployed people can tell us about competition for new jobs, especially among people in lower-paid occupations. It’s not easy to dig into because vacancy data is rarely split by occupation (the ONS is currently working to produce this data). We’ve analysed a dataset, provided to us by the Institute of Employment Studies, which assigns vacancies listed on Adzuna during summer/autumn 2020 to specific occupation groups, giving us the opportunity to explore this important issue.
The national context: it’s hard to find a job in an economic crisis
As COVID-19 hit there were 1.7 unemployed people for every job vacancy (ONS, January to March 2020). This rose to four people per vacancy between April and June and was still at a high 3.2 by the third quarter of the year. This simple figure illustrates the result of rising job losses and plummeting vacancies.
More competition for work is expected in any recession, but this downturn has been especially bad for people in search of new opportunities. Between March and May (2020) job vacancies fell by 60% to a very low 330,000 jobs, led by huge falls in retail and hospitality for obvious reasons. A strong summer recovery slowed through the autumn and winter, and by the end of the year vacancies were still 33% lower than at the end of 2019.
Within this high-level picture, does the competition experienced by job searchers vary? Are some roles more sought after than others?
Competition for lower-paid occupations is particularly intense
Strikingly, there is a significant difference between the occupations on offer and those being searched for.
First, a look at the types of jobs on offer online in September last year (2020) shows us a strong tilt towards higher-paid roles. More than a quarter (28%) were for professional occupations, such as teachers, lawyers and doctors. That’s a much higher proportion than for the total job market (22%). In contrast, a similar share of vacancies (27%) covers all three of the roles defined as requiring fewer skills: ‘elementary’ roles such as cleaners, security guards and labourers, sales and customer services roles, and roles operating machines or routine processes.
Second, this breakdown is compared with the roles last done by people who were unemployed between June and September, and so searching for work among these vacancies. The most obvious difference is the much smaller share who lost professional jobs, just 13%. A much greater share of them were in sales and customer services roles (13%) or ‘elementary’ occupations (23%) than for vacancies.
More competition for new jobs with lower-skilled requirements
The result is a useful insight into where competition for jobs was likely to be most intense. Three occupations stand out as having the greatest mismatch – meaning there are more unemployed people for each job opening:
- Sales and customer service roles: this comes as no surprise given restrictions have significantly limited business activity in the sectors that provide these jobs. These roles as roughly three times as competitive as the average job, in terms of the number of unemployed for each vacancy.
- ‘Elementary’ occupations: this is for a similar reason, given that many of these jobs involve face-to-face interaction in the services industry. These are the lowest-paid roles, on average, in the labour market and we find they are attracting more than twice as much competition as the average job opening.
- Administrative and secretarial roles: while this might seem less obvious, it fits with the fact that these occupations experience a big initial hit to vacancies during the first lockdown of a similar size to hospitality, so are clearly impacted by the economic restrictions due to COVID-19.
If we look back at June, July and August, the skew of vacancies was more extreme. Two of every five job openings were in professional occupations, and just 6% were in ‘elementary’ occupations. On the one hand, this means the situation presented above for September is an improvement. On the other, finding such competition for lower-paid roles even in a period of recovery is concerning. Most usefully, it helps us understand what may be to come this spring as the third UK lockdown is unwound. While we do not have current vacancies split by occupation, we expect some similarities to last year as the pattern of lockdown restrictions followed by reopening is repeated.
The story varies around the country. Competition for job vacancies is significantly higher in the North East of England, with more than twice the number of unemployed people for each role. For example, a greater share of unemployed people previously worked in ‘elementary’ occupations (than nationally), but these roles take up a smaller share of the job vacancies in the region. In contrast, there is a more balanced distribution of job vacancies in the South East and South West with fewer people per job opening than for the UK as a whole.
Accounting for new labour market entrants raises competition even higher
In addition to the unemployed people analysed in the bar chart above, there is a sizeable group who are looking for work (and so classed as unemployed) but have no listed last occupation. The vast majority of these fall into three categories: still in education and seeking work, recent education leavers, and long-term inactive or unemployed (for example due to caring for young children for many years).
As a result, they have fewer formal qualifications (on average) than the group of unemployed people analysed. In addition, some of those listed as highly qualified but who have been out of work for a long time may have seen their work skills and experience weaken due to their absence from the workforce. This may cause them to search for lower-skilled occupations than they previously held. This means we are likely underplaying the skew of those looking for work towards lower-paid occupations and overplaying the prominence of higher-paid roles, implying that competition could be even more intense than considered above.
The COVID-19 labour market story is full of uncertainty, but these findings are a reason to be concerned
Our analysis suggests that it has been harder for people losing lower-paid jobs to move back into work due to more job searchers fighting for every role. This is particularly concerning for two reasons.
First, two of the roles that seem to be most competitive are those with high poverty rates among workers. This means that many people who have lost these jobs and are waiting to find new roles were also in a difficult financial position. Many will have had little or no savings to fall back on when they found themselves unemployed because even with the job, they were struggling to make ends meet. This means any period of unemployment will be very difficult, but the prospect of longer-term unemployment due to competition for jobs will be extremely challenging.
Two of the roles with most competition also have high poverty rates amongst workers
Second, this is particularly worrying given that the long-term harm from time spent unemployed can be greater for this cohort of workers. Workers with fewer qualifications are not only more likely to be made unemployed during economic downturns, but they also tend to take longer to find new employment and when they do are more likely to take a larger pay cut than more highly qualified workers.
This research highlights just how important policies to support a labour market recovery are to people living in poverty whose work has been hit by this economic storm. Avoiding much higher levels of unemployment with an effective, careful unwinding of the CJRS will prevent competition for roles becoming more intense, and we will be publishing our recommendations for how to do this very shortly.
Job creation and employment support will improve the opportunities on offer and support people to return to work. In the meantime, the social security system is a lifeline for people who find themselves out of work and so it is essential the £20 uplift in Universal Credit and Working Tax Credits remains in place to provide this vital income support.
1) A note on the limitations of the job vacancy dataset. Not all vacancies have been successfully matched to an occupation. The coverage rate varies by month and is 43% across September. For the purposes of this discussion, we assume the distribution of vacancies by occupation for which we have assigned data is the same as the total population of vacancies, but we cannot be sure this is the case. So far, we have not found the data necessary to check this assumption, but any insights and evidence on this are very welcome. Given the importance of the issue we believe that it is still useful to show and discuss this analysis despite these drawbacks.
2) Thanks to Dafni Papoutsaki, Institute of Employment Research, for providing the dataset and analytical guidance.