How the changing shape of the UK workforce has started to drag down wages

It’s clichéd to say that the most certainty you can take from any economist’s prediction is that it will be mistaken, but many got their forecasts for 2014 horribly wrong. At the beginning of this year, commentators were fairly united in hailing the ‘year of the pay rise’, with most expecting the UK’s enduring real wage squeeze to finally turn the corner, and some predicting that pay would bounce back quite rapidly over the course of the year. Two thirds of the way through 2014 this has not come to pass and the outlook has been repeatedly downgraded. For example the Bank of England now doesn’t expect a return to real wage growth until the middle of next year. Looking back, many of us are left wondering why we were all far too optimistic.

In answering this question some have pointed to the fact that the official data looks worse than other pay surveys, with pay settlements, for example, running at or above inflation during 2014. A major difference between the official Average Weekly Earnings (AWE) series and other surveys is that AWE, as well as capturing year-on-year changes in employees’ pay, will be affected by changes in the overall shape of the workforce. Do such ‘compositional’ factors shed light on what’s been happening to pay of late?

New analysis by the Resolution Foundation comprehensively assesses the impact of the changing make-up of the workforce on wages since the mid-2000s. It looks across a number of job and employee characteristics, including sector, occupation, age, sex, qualification level and job tenure, in order to understand how much of pay growth is down to pay changes within groups, and how much is compositional – owing to a shift in the proportion of employees across groups. For example, as others have highlighted previously, recent increases in employment in lower-paid sectors and continued contraction in the finance industry are likely to have dragged down on average pay growth across all employees.

While changes in the industrial mix look to be putting downward pressure on average pay, other factors, such as strong full-time employment growth in the last 12 months, are likely to be pulling in the opposite direction. To disentangle the pushes and pulls our analysis looks at all factors together, controlling for the overlap between each, to understand the overall direction in which compositional changes are driving wages and the relative importance of different job and employee characteristics to this. The overall compositional effect is shown in the red bars in the the following chart.

We find that between 2006 and 2013 the compositional effect was always positive, acting as a boost to pay growth. This is to be expected due to long-run shifts in the workforce such as rising qualification levels – a compositional boost to pay could be regarded as a normal state of affairs. It’s sobering to think that, had it not been for these changes, Britain’s already huge pay squeeze would have been a third deeper.

However, recent months mark a departure from this trend, with the overall compositional effect turning negative for the first time in the period we’ve looked at. This is due to negative effects from factors including occupations, age and job tenure outweighing the positive impact of rising hours and qualification levels, as the chart below shows.

Some of these drag effects are probably part of a good news story – the entry and re-entry of younger and less experienced workers reflecting rapid employment growth and falling youth unemployment this year. With the UK’s employment rate having returned to its pre-recession peak these effects are likely to fall out next year – we may be observing the temporary ‘growing pains’ of recovery. By contrast, the drag effect of the shift towards lower-paid occupations – the largest single compositional factor affecting pay growth – is more worrying as it could represent a longer-term change in the labour market.

To return to our opening question, this analysis does suggest that compositional factors help explain why pay growth in 2014 has looked quite so bad. We find that without the reversal in compositional effects between 2013 and 2014, real pay in the first half of 2014 would have grown by a very modest 0.1%, rather than the 0.8% fall we’ve experienced. So there may be reasons to give those economists who called 2014 wrongly a break this time round, although 0.1% real growth hardly represents a boom.

This analysis shows the important role that the changing make-up of the workforce plays in our understanding of average pay growth, for which reason we’ll be keeping a close eye on these trends as more data becomes available. But it mustn’t hide the fact that a generalised and dramatic slowdown in pay growth within sectors and different groups of workers has been by far the biggest factor explaining falling UK wages since the financial crisis (shown by the decreasing size of the grey bars on the first chart). Ultimately, prospects for ending the pay squeeze rest on a return to productivity growth and the willingness of employers to ensure that workers obtain a fair share of these gains.

This blog originally appeared on the OECD insights blog