Tuesday, 2 February 2016

SWOB 2: Is the Rise in Low Skilled Jobs holding back Wage Growth?

Some commentators have suggested that the latest upswing has been characterised by a greater share of low skilled jobs in the recovery compared to previous upturns. If so then we can all blame low skilled firms/the low skilled for poor productivity growth

Look at the graph below. It shows the share of all the net employment change between 2011 & 2015 accounted for by high and low wage jobs - (strictly by each wage decile ie splitting employees by the average wage in each occupation and ranking the occupations by average wage paid in 2011). The graph does indeed show that low skilled jobs or at least low paying jobs  (say those in the bottom two pay deciles) account for quite a lot of net job growth over this period  (about 30% of the total).  However there has been more growth in employment in higher paying occupations. The 8th and 9th deciles account for 50% of all net job growth in this period.
Now look at the black line. This shows the average real wage growth by each occupation ranking. It is consistently negative. In other words over this period, real wages have fallen - by around 3-5% - all across the board , high paying and low paying. So it is hard to argue that lots of low paying jobs are dragging down real wage growth on this basis. Rather low pay growth has been a feature of most jobs.

Employment and Wage Growth 2011-2015 by Pay Ranking of Job          Source; LFS
The next graph repeats the analysis for the previous recovery 1994-1998. Once again the growth in net new jobs was skewed to both low paying and high paying jobs - though not quite as many high paying jobs as in 2011-2015 . The pattern of wage growth was different however. Wage growth in higher paying occupations then was much less subdued than wage growth at the bottom.
Employment and Wage Growth 1994-1998 by Pay Ranking of Job          Source; LFS

What are we conclude? On the evidence of the above it seems that low skilled (low paid) job creation was likely to have been a much greater drag on wage growth in the 1990s than now

Jonathan Wadsworth

ps. for those of a technical bent we can decompose the change in average wages between any 2 periods as the weighted average of the wage in each decile, which in turn is given by the change in the shares plus the change in the wages for each decile

Using this it is apparent that wages in the bottom deciles are not primarily responsible for overall sluggish real wage growth now. They were much more likely to hold wage growth back in the 1990s recovery.

No comments: