How the exclusion of the self-employed might obscure the 2015 earnings recovery

Over the past few years growth in self-employment has been a big story – the self-employed account for two thirds of the net change in employment since the pre-recession peak. But this may now be turning a corner; self-employment peaked back in June and has fallen steadily by 79,000 over the last two quarters, while employee numbers have been growing strongly.

Both the Resolution Foundation and the ONS have theorised that some of the post-recession growth in self-employment was a reaction to weak employee vacancies. As the number of vacancies increase we could see more people moving from self-employment into employee jobs.
Many self-employed people will find this an attractive prospect. The earnings of a typical self-employed worker have been squeezed harder than those of employees during the downturn – by more than one fifth since 2008. We estimate that more than four in ten self-employed people earned below £220 a week in 2012-13 – less than a full-time worker on the minimum wage. For these workers, any full-time employee job would mean a pay rise.

But there’s a catch. Self-employed workers are not included in average weekly earnings (AWE) figures. A shift from low-earning self-employment to employee roles would mean AWE capturing a new cohort of low-paid staff – dragging down official pay growth, even if it means a pay rise for individuals. Just as the exclusion of self-employed workers from official data has failed to capture the size of the wage squeeze since 2008, it could now under-state the pace of the recovery.

For this reason, the Resolution Foundation has developed an ‘all-worker’ earnings measure that captures the self-employed as well as employees. This provides a true picture of what’s happening to average earnings across the workforce, rather than exclude around 4.5 million workers from official pay figures.

Our estimate for ‘all worker’ earnings, now updated to include the reported earnings from self-employment in the 2012-13 Family Resources Survey, shows that that the wage squeeze would be 20-30 per cent deeper than our official data shows were the self-employed captured. Average earnings actually fell by 13 per cent by late-2014, rather than the 10 per cent fall captured by AWE. Whether that remains the case in 2015 is a key question in our understanding of the labour market and the recovery.

We have called for an improved official earnings measure that captures all workers and not just employees. Sir Andrew Dilnot confirmed that the ONS was considering this issue in September, but it is not yet clear whether any action will be taken. While we continue to push for better official data it will keep its best guess indicator updated so that we are not completely blind to earnings across the workforce.

Watch Sarah O’Connor, employment correspondent for the Financial Times, cite our findings in this video: