The issue
Multiple job holding has generated a growing body of literature examining who is working in more than one job and why, especially whether it is due to constraints or opportunities. Hence, whether multiple job holding is regarded as a widespread or small-scale phenomenon in contemporary labour markets depends on how it is defined and measured, and the data sources used.
Researchers who work with large-scale secondary quantitative data often have to adapt their research questions and variable definitions to the available data. However, when it comes to multiple job holding, critical reflections on definitions, data capture and methodological limitations remain surprisingly limited. We argue that these issues warrant greater attention, particularly given that the policy implications of multiple job holding remain inconsistent and sometimes contradictory. This lack of clarity may explain why multiple job holding has not been included in standard labour market analysis.
How many, why and who?
Estimates of the prevalence of multiple job holding vary greatly across studies, leading to divergent answers to the question of ‘how many’ workers multi job. For example, official labour market statistics in the USA and Australia suggest that around five to six per cent of the workforce hold more than one job. Another study instead finds that a much higher level (18 per cent) of US households have members with second earnings.
Findings on ‘why’, on the motivations of multiple job holding, also differ substantially from not having a choice, due to bad jobs that do not provide enough hours and pay, to more positive motivations including enjoyment, skill development or career enhancement. There is also disagreement about ‘who’ engages in multiple job holding. Some studies point to a greater need among women to take on additional jobs, while others find that in many countries in Europe, young men are more likely than young women to work in more than one job at the same time.
The differing interpretations of how many people engage in multiple job holding, why they do so and who is involved are all closely linked to how multiple job holding is measured, specifically who is included and who is not. Often also described as ‘moonlighting’ or ‘side hustles’, such definitions of multiple job holding usually include some form of self-employed activity. Using a definition of multiple job holding that includes self-employed activity is more likely to capture people motivated by enjoyment or dissatisfaction with their first job, similar to the motivations of becoming self-employed in general. In this context, concepts such as ‘portfolio’ or ‘mosaic’ careers, emphasising the more positive connotations of agency over one’s working career, may be more suitable for the ‘side hustler’ or ‘hybrid entrepreneurs’ who experiment with self-employment activities while retaining a less fulfilling job, at least in the short term.
Second or multiple job holding is also defined as having more than one concurrent job as an employee. When multi job holding is conceptualised this way, we are more likely to find people who take on additional jobs because they cannot find a suitable full-time job, which is especially relevant for women.
Beyond these definitional variations, we know little about how people respond to survey questions on whether they held a second job during a given reference period (typically the previous week). Measurement may be affected by under-reporting as some workers might be reluctant to report certain ‘side hustles’, for example due to concerns about disclosure in governmental surveys. They may be more likely to report more stable additional jobs such as regular part-time work. There are likely to be group biases in who reports an additional job or not.
Is administrative data useful for multiple job holding?
In 2024, the UK Office for National Statistics added to their data resources a linked administrative dataset combining the Annual Survey of Hours and Earnings (ASHE) with real-time payroll data from the tax authorities (HM Revenue and Customs). ASHE is an annual employer-completed survey of employees, providing highly accurate information on pay and working hours. However, personal information is limited (to gender and age) and there is the caveat that large employers are more likely than small employers to respond to the survey. In addition, as with many social surveys, ASHE has a reference period so that we don’t know anything about the employees outside its narrow window of time.
The payroll data that can be linked to ASHE are a collection of employees’ payslips from HMRC’s Pay As You Earn (PAYE) system. Employers are required to use PAYE for workers above a certain earnings threshold, as well as for those receiving state benefits or holding a second job. On their own, these administrative payslip data contain limited information beyond earnings. However, when linked with ASHE, they become more meaningful, allowing, for example, a gender analysis of detailed pay and employment.
All payslips per person in a tax year should be in the data, allowing the identification of multiple payslips on a weekly basis. In our analysis, we exclude multiple payslips from the same employer (these could be bonus payments) and any payslips that are related to pensions or a zero amount. Concurrent payslips may not be, due to distinctly different jobs, such as the earnings academics receive from external examining. Among high-skilled occupations, it is therefore likely that multiple payslips are found more often in administrative than in survey data. We therefore propose the term ‘multi-payroll’ to describe this situation. This measure captures formal, recorded employment activities, but administrative data do not capture informal work. In addition, since self-employment activities are not covered in PAYE data, this measure is not suitable for studying the gig economy.
Despite limitations, payroll administrative data offer a valuable tool for examining multiple earnings from paid employment, reducing concerns about group biases in recording employment status (for example, gender differences). However, a key downside is the absence of subjective information, specifically, how workers think about their activities. Are these viewed as two jobs, for example?
Our estimation and conclusions
We estimate that the prevalence of multi-payroll work among 16–64-year-olds in Great Britain was between 18 and 20 per cent for women and between 14 and 16 per cent for men in 2018–19. These figures are much higher than those reported in official labour market statistics but notably close to Scott et al.’s (2020) estimate of second earnings in the USA (18 per cent), despite our exclusion of the ‘hidden economy’.
Whether including the hidden economy or all formal employment activities, however short in duration and without individual response bias, multiple jobbing seems much better captured here than in official social surveys. While our multi-payroll measure is not without limitations, it strengthens the case for recognising multiple job holding as a significant feature of contemporary labour markets.
Data acknowledgement: Office for National Statistics; His Majesty’s Revenue and Customs, released 01 August 2024, ONS SRS Metadata Catalogue, dataset, Annual Survey of Hours and Earnings linked to PAYE and Self-Assessment data – GB, https://doi.org/10.57906/566k-5q15
Funding: This work is supported by ADR UK (Administrative Data Research UK), an investment of the UK Economic and Social Research Council (ESRC) (part of UK Research and Innovation). [Grant number: ES/Z502406/1].
Disclaimer: This work was undertaken in the Office for National Statistics Secure Research Service using data from ONS and other owners and does not imply the endorsement of the ONS or other data owners.
Darja Reuschke is an Associate Professor in the Department of Strategy and International Business at the University of Birmingham. Her research concerns changes in local labour markets and new forms and locations of work and businesses that are emerging through new technologies, economic restructuring, crisis and social change. She has recently been working on multiple jobholding in the UK using payroll data.
Tracey Warren Is Professor of Sociology at the Nottingham University Business School and internationally recognised for her research on working lives. Her areas of research expertise include job quality, underemployment, work time, work-life balance, atypical working, paid and unpaid labour, equality, diversity and inclusion in work, and employment in/equalities.
Image credit: Kounotori via Unsplash