Context

People’s earnings are rarely static and instead often vary due to hours worked, bonuses, voluntary job changes, and redundancies. Unpredictable earnings can lead to households experiencing financial stress, making it difficult to plan for the future, and increasing reliance on credit or social support. Understanding earnings volatility in the UK is important for creating fairer labour markets, effective social policies, and financial security for individuals and households.

The study

This study was funded by the Joseph Rowntree Foundation and led by the Resolution Foundation to examine high-frequency - weekly and monthly - volatility in UK employee earnings and how this differed between different groups of workers.

The study involved analysis of HM Revenue and Customs (HMRC) payroll data for a random, nationally representative (1%) sample of over 250,000 UK employees, covering April 2014 to March 2019. The data related to pre-tax earnings from HMRC’s Pay As You Earn (PAYE) Real Time Information system. For about two-thirds of these employees, their payroll data was linked to data from employers about the characteristics of employees and their jobs from the Annual Survey of Hours and Earnings (ASHE).

The analysis was limited to people aged 20-59, and for consistency, pay was converted into a monthly amount pro rata (separate analysis was undertaken on those paid weekly) and nominal pay was converted into real term pay indexed to March 2019 values.

Key findings

• Monthly earnings volatility: Among those employed in consecutive months, 43% experienced no change in monthly earning, 29% experienced a change in earnings of up to 10 per cent, a further 13% of up to 25 per cent, and a further 15% by more than 25 per cent. The average month-to-month change was 15%. Only 26 per cent had monthly earnings within 10 per cent of their average monthly wage in all 12 months of the year, while 14% had earnings that varied by at least 25% in at least four months in a year.
  • Volatility on this measure was highest for 20-24-year-olds (24%), compared with 10% of those aged 40 and over. Volatile earnings were experienced by 27% of hospitality and 23% of arts and recreation sector employees, and were most common among the bottom decile of earners (30%).
  • Regression analysis confirmed that ‘high-volatility’ characteristics were independently related to higher earnings volatility.
• Weekly earnings volatility: 12% of UK employees were paid weekly, rising to 51% of those in food processing sectors. Among employees paid in consecutive weeks, 7% experienced a fall in earnings fall of at least 25 per cent. The average week-on-week change in earnings was 19 per cent, equivalent to a change of £84 per week on an average paycheque of £400.

The report makes recommendations to Government and employers to reduce earnings and income volatility.

Points to consider

Methodological strengths or limitations:

• The report presents rigorous, quantitative evidence on high-frequency volatility in pre-tax earnings, based on a large-scale nationally representative sample of UK employees in the UK, and using descriptive and multivariate analysis.
• The authors note their grounds for considering nominal rather than real earnings, but also that inflation was low during the period covered by the data and differences between the two measures were mostly negligible.
• Note that some of the changes reported refer to arc percentage changes (the difference in earnings divided by the average value of earnings in the two periods) rather than usual percentage change (the difference in earnings divided by earnings in the first period).

Applicability:

• The findings should be of interest to policymakers and employers who seek to secure better outcomes for workers with volatile earnings and incomes.
• The report is highly technical in places but explains the findings well for a wide range of readers.
• The authors note that the data did not allow them to examine the effect of job changes, hours worked or self-employment on earnings volatility.

Relevance:

• The data covered the five-year period prior to the onset of Covid-19.

Generalisability/transferability:

• ·Some of the findings may transfer, at a high level at least, to other countries where there are similar income, employment and social security structures.