Trends on remote and hybrid working, new insights into the gender pay gap, and occupational exposure to AI are all covered in the latest update from Charlie Ball, Jisc's head of labour market intelligence
This is a relatively short update as it's been a quiet couple of weeks on the labour market front. There have, though, been some interesting releases from the Office for National Statistics (ONS), first data on working arrangements.
Working patterns have been relatively stable for the last couple of years with just over 10% of the workforce work from home exclusively (although interestingly they're the only ones who don't see a big dip in work at holiday periods, suggesting they might not be taking holiday), between a quarter and a third of the workforce working hybrid, and just under half travelling to work (the rest generally reported not working in the seven day period).
Overall, in July, 36% of the workforce worked from home some or part of the time. It's more common for older workers, and for men. Men are also more likely to say that they had the option to work at home but chose not to, as did older workers (despite the rhetoric that young people might not like homeworking, they're more likely than older workers aged 50 to 69 to take up the option if it's offered). Women and younger workers are more likely to say the option to work at home is not available to them.
Meanwhile, the latest wave of the Business Insights and Conditions Survey (BICS) is here and parts of it focus on the workforce. First, here are hiring and downsizing intentions for September by industry:
Overall 4.2% of businesses expect to increase headcount, 6.7% expect to decrease. It's not specified which part of the education sector makes up the 13% of education providers expecting to shed staff at the start of a new academic year, but I think we in HE can probably hazard a guess unfortunately. It's not actually the worst sector - I've left off the hospitality industry, where one in seven businesses expect to lose staff, because there's no data on what proportion expect to increase headcount. It's worth noting that the arts figures are low because nearly 40% of businesses don't know yet what they expect from September.
The next graph is on worker shortage and recruitment difficulties:
Yes, you read that right, at the same time as 13% of education providers say they're going to lay people off, 13.9 % say they don't have enough staff and 14.3% say they're struggling to recruit. One of those sets of figures would be concerning, all of them together should be sounding alarm bells, although of course, it could represent different parts of the education sector (and I'm sure readers will have ideas). Nevertheless it shows there are serious issues in the education labour market.
Proposed redundancies are down from the levels of April and May, but are still around double the numbers last year. The labour market has clearly not fared very well in 2025 and as I've been saying for a while now, I think it's reasonable to assume this will affect new graduates and that this year's, and next year's, outcomes data should be expected to be worse than the data we've just had.
And we have new data for July on job ads, which does suggest a modest but welcome recovery might be taking place, data both for new job ads and for ads overall suggest we may have bottomed out (for the time being) earlier in 2025.
Around 54% of vacancies are at professional level. This has been pretty stable in 2025 and is up a bit on the latter half of 2024, when it fell to nearly 50/50, suggesting a specific recovery in the graduate jobs market. We can actually track some of this data at occupational level, so some of the occupations that saw a big rise last month included GPs, metalworkers, clinical psychologists, therapists, clothing and fashion designers, air-con installers (I think we all know why!), fork lift drivers and aerospace engineers.
The Recruitment & Employment Confederation's (REC) new Labour Market Tracker with Lightcast also finds the number of job ads up, but new postings down, suggesting roles are taking longer to fill.
- The number of new job postings in the UK in July 2025 was 652,733 - a decrease of 9.2% from June 2025.
- The overall number of active job postings in July 2025 was 1,527,906 - an increase of 3.4% on the number of job postings in June 2025.
- There's a seasonal element to this data as fewer jobs tend to be posted in July, but the overall figure is around 11% lower than this time last year - but it was the second half of 2024 when the jobs market started to take a palpable turn for the worse.
Has the gender pay gap been underestimated?
Underrepresentation of small businesses may have led to the UK's gender pay gap being systemically underestimated by around one percentage point for over 20 years according to a team of researchers led by John Forth at Bayes Business School at City St George's University.
The UK's gender pay gap was 7% in 2024 for full-time employees, according to official figures, but the team's analysis suggests it ought to be around 8%. This seems to have arisen because the sample used was over-representative of larger businesses, where pay tends to be higher and the gap between men and women's earnings is usually smaller.
Research explores occupational exposure to AI
A team led by Golo Henseke at UCL has been examining occupational exposure to AI. Golo's been at the forefront of academic research into graduate employment and it's always worth reading his output and this is certainly no exception.
The team uses a task-based approach and data from the long-running Skills and Employment Surveys (SES) to assess job activities where an LLM or LLM-powered system can reduce task completion time by at least 25% beyond existing productivity tools. They find that by 2023/24, nearly all UK jobs exhibited some exposure, yet only a minority were heavily affected. Aggregate exposure rose since 2017, but this is due to changes in the roles themselves rather than changes in task profiles.
The price premium for AI-exposed tasks declined relative to 2017, measuring approximately 12% lower in 2023/24. Job postings fell following the release of ChatGPT, with job postings 5.5% lower (equating to about 84,000 fewer jobs per month) in 2025-Q2 than if pre-GPT hiring patterns had persisted (although we must bear in mind also that the full public models of GPT were released into a heavily pandemic-affected labour market).
The team also find professional-level (i.e. graduate) jobs to be the most exposed, with computer programmers and R&D managers the most exposed occupations examined, and the least exposed all being manual workers.
Do SMEs utilise graduate skills effectively?
And we'll finish up with another paper from another fine researcher, Daria Luchinskaya at Strathclyde Business School. Daria has been using Futuretrack data to look at how SMEs use graduate skills, and found that contrary to some concerns that SMEs might sometimes lack the capacity to use graduate skills effectively, SME employers do not impede graduates' skill utilisation and may even enhance the use of disciplinary knowledge at work for overqualified graduates (those judged by some measures as not being in a job that needs a degree).
Luchinskaya makes the distinction between jobs where a degree is not required to get a role and the role doesn't use their degree skills, and those where a degree is not required but their degree skills do get used, and argues that the data suggests that graduates at SMEs in roles that don't formally require degree level qualifications may be more likely to, nevertheless, use their skills. This is important for a number of reasons but mainly because graduates who think that they are using their skills experience higher levels of satisfaction and well-being, even if that job is judged as not needing a degree.
Written by
Was this page useful?
Thank you for your feedback