Data Insight: Linked administrative data reveals much higher level of multiple jobholding in Britain than previously thought

ASHE and official labour market statistics substantially underreport multiple jobholding. In HMRC PAYE data, not only is the level of multiple jobholding substantially higher but its prevalence among young women also becomes visible. Read this Data Insight from ADR UK fellow Darja Reuschke to find out more.

Read publication

The proportion of employees in the Annual Survey of Hours and Earnings (ASHE) with two or more concurrent jobs is at around 3%. In the Annual Population Surveys and Labour Force Surveys, multiple jobholding is similarly low. Multiple jobholding, however, is dynamic, with people having in some months or weeks multiple jobs but not consistently over the whole year. In HM Revenue and Customs real time payroll (PAYE) data from individuals that are also observed in ASHE between 2014-2018, multiple jobholding among 16-64-year-olds is between 18% and 20% for women and 14% and 16% for men, depending on whether monthly or weekly data are used. These figures are more in line with estimates of second earnings of households in the USA (at 18%, Scott et al., 2020) than official labour market data.

Young women aged 16-29 are, accounted for job characteristics and geographical location, at the highest risk of having two or more concurrent jobs. This risk is disguised when multiple jobholding is examined with the ASHE alone calling for greater use of linked administrative data to reveal working practices of different groups of workers. Linked administrative data also reveal differences between different types of multiple employment.

What we found

For both, women and men, having two or more concurrent jobs is associated with low hourly pay and low working hours. These are indicators of necessity-driven employment choices. It is therefore not only low-paid jobs but also time-based underemployment that is driving multiple jobholding. Previous studies have suggested that women are more likely to have multiple jobs than men. My findings confirm this but, thanks to the large sample size of the linked ASHE-HMRC PAYE data, I could dig further into socio-demographic characteristics of multiple jobholding. In fact, young women aged 16-29 have the highest risk of multiple jobholding. Among young workers, the gender gap of multiple jobholding is statistically the largest. This applies to urban and rural areas, and within urban areas to large cities. Hence, large cities do not offer better jobs in terms of enough hours and pay relative to rural areas.

Mixing employee jobs with self-employment is not, as multiple jobholding, linked with low-paid jobs. However, the likelihood of taking up additional self-employment as an employee is increased for both women and men when the regional unemployment rate is higher. This is also an indicator of necessity.

Why it matters

Multiple jobholding has been mentioned as work practice of women in various studies on precarious employment including, for example, in recent research on young women transitioning from education to employment (e.g. Allen et al., 2025). However, using linked administrative data and testing a new method of identifying multiple jobholding as a dynamic work practice demonstrates that it is not a rare phenomenon. Rather than treating multiple jobholding as a ‘side’ issue, it needs to become more mainstream in research on employment and precariousness and in policy-focused research. Furthermore, research has either looked at multiple jobholding or time-based underemployment when these are clearly linked and would need to be investigated in a more integrated fashion.

What next?

I used the ASHE-HMRC linked dataset to investigate an employment phenomenon that is relatively persistent in current capitalist labour markets. For further policy-focused research, a continuation of the linkage for more recent years including the post-COVID period, is required. Data availability has hitherto prevented more detailed analysis of multiple employment in Britain. The linked ASHE-HMRC dataset is a valuable source for future research on this topic. For example, seasonal changes in multiple jobholding could be further explored.

In stakeholder engagement, the finding of multiple jobholding among young women has received particular attention. I have not worked previously on labour market issues of young women specifically. This is an important area of research I would like to work more on in the future. The stakeholder engagement embedded in the ADR UK fellowship pushed me to work proactively with stakeholders from the onset of the research. I highly value this opportunity and seek to continue working in this way.