Do the latest employment figures really suggest labour market recovery?
At first glance, the June labour market statistics tell a positive story: the headline unemployment figure fell by 88,000 in the quarter to April, while employment grew by 80,000. So far, the increase in private sector jobs has been more than sufficient to offset the fall in public sector employment.
However, as several commentators have pointed out, including the TUC’s Richard Exell, one swallow doesn’t make a summer. Some indicators reveal a more gloomy picture, including the claimant count figure which increased strongly in May, and job vacancies which fell again in the quarter to April, almost back to the lowest level experienced during the recession.
Nevertheless, there is something unusual going on.
On the face of it, the strong recent employment performance is difficult to square with the weak figures for GDP growth. It’s possible that the GDP figures are simply wrong, seriously underestimating the recovery in output, and will be retrospectively revised upwards. If we take them at face value, however, alongside the employment figures they do imply a rather implausible fall in labour productivity, as the BBC’s Stephanie Flanders has noted.
My view is that it’s important not to get too excited about these short-term employment data, for two reasons. The first is that, in the grand scheme of things, these are small changes and, if looking over a slightly longer timescale, there is really no evidence of a sustained labour market recovery. To see this, we need only look at Figure 1 which shows the ‘Beveridge curve’, plotting total vacancies against total (LFS) unemployment since 2008. Throughout the first year of the recession the economy tracked down the curve, with vacancies steadily falling from 700,000 to just over 400,000 while unemployment grew from 1.6m to nearly 2.5m. Since mid-2009, there has been no real change, and the Beveridge curve has barely moved, with unemployment stuck at 2.4-2.5m and vacancies fluctuating in the 450-500,000 range.
This looks to me like a labour market experiencing chronic demand-deficiency. Until we see the figures moving back up the Beveridge curve in a sustainable way, it is too early to talk of labour market recovery, not least because, as Richard points out, the bulk of public sector job loss is still to come.
Figure 1: Vacancies and Unemployment (2008-11)
The second reason is that we get a rather different picture if we look at trends not in total employment (headcount) but in total hours worked, arguably a better indicator of underlying labour demand. The recession itself was more severe than previous recessions, with output falling by over 6%, and so far, it has recovered only to a level still 4% below pre-recession level. In contrast, employment fell much less, by only around 2% (although this still meant 660,000 fewer jobs), and has since recovered to within 1% of pre-recession levels. The weaker response of employment to a severe recession has been attributed to factors such as employers retaining staff through use of short-time working, shifts to part-time work, and a climate of wage restraint in (until recently) a low inflation environment.
Consistent with this, as Figure 2 shows, total hours worked in the economy fell rather more than total employment, by an amount closer to (although still not as severe as) the fall in GDP. But we should also note the surprising and sharp fall in hours worked in 2011; following recovery in 2010, indeed total hours are now almost back down at their 2009 level. ONS attributes this recent fall to a greater than usual number of bank holidays, but these are unlikely to fully explain it. Alongside the continued low (and recently falling) levels of vacancies, it raises the question of whether we are already facing another slackening of labour demand, to be further reinforced as the full effects of public sector cuts bite.
Figure 2: Trends in employment and hours worked (2008-11)