Job-filling and the recovery

4th October 2014

Paul Bivand examines recent trends in job-finding and how far the recovery has yet to go.


Over the recession and the recovery, job-finding rates for the unemployed, inactive and the employed have moved in the same ways, and are depressed below pre-recession levels. Looking across the whole period, it is striking that job finding rates have only regained half of the ground lost on the boom years of the early 2000s. In terms of job-finding, we are still a long way from recovery.


There are more clues emerging about where we are in the recovery as far as job-filling is concerned. The Bank of England has been looking at trends in job to job moves. So has the Office for National Statistics, in its October Economic Review. The Bank looked at the headlines, while the ONS has dug more deeply.


The headline finding is that the proportion of those in employment who move to new jobs each quarter is still well below pre-recession levels. In the last three months before the recession began, 2.6% of people in jobs moved to new jobs. In the latest quarter, 2.2% did. While this is a substantial recovery from the pit of the recession, when the proportion of those moving to new jobs fell to 1.7%, it still remains well below where it was.


To put this another way, the proportion of workers moving jobs fell by 35% in the recession. It has since recovered, but remains 15% below pre-recession levels.


How do these trends in job to job moves compare with trends for people out of work?


Chart 1 shows the trends in the rate at which people find new jobs for those who were already in work, were unemployed (looking for work and available for work) or were inactive (not looking for work and/ or not available for work). I have put these on a common scale with the pre-recession job-finding rate at 100, so the lines represent the per cent of the Jan-Mar 2008 rates for all three groups.


Chart 1: Rates of moving to new jobs for employed, unemployed or inactive: Jan-Mar 2008=100



The recession collapse in the rate for those moving from one job to another was larger than the change for those moving from unemployment or from inactivity into work. Perhaps unsurprisingly, all the trends share in both the recession and the recovery, with the timing of changes very similar.


The rates for the inactive are rather more jumpy (technically, volatile) than the others, which could be randomness as well as different changes affecting different groups that are inactive, such as students, the long term sick, and returners from caring for family.


However, when we go a little more deeply into the figures, the relationship between the job-finding rates for the employed and the unemployed look remarkably stable over the period since 2003 (the period ONS give us the figures).


Chart 2 shows the job-finding rates for employed (along the bottom) and for the unemployed (up the side), with the line representing the path over time. Starting from the top right of the chart (2003), job-finding rates slide down towards the bottom left (i.e. rates fell for both the employed and the unemployed), recover slightly, slide again (the recession) and then recover.


Chart 2: Comparison of job to job moves and unemployment to job moves with linear trend



A number of things stand out from this.


First, the data is almost a straight line – with a coefficient of determination of 0.97. This is almost a perfect fit, and shows that the changes over the recession are consistent with those both before and after.


Secondly, it is clear that the decline in job-finding rates began well before the recession, with rates falling between 2004 and 2006 before seeing a mini-recovery in 2007-8. Some have attributed the fall in job-finding rates among the unemployed as being due to issues inside Jobcentre Plus at the time, followed by a re-imposition of the full JSA regime with tighter conditionality.


However the third thing that stands out is that the falls in job-finding rates are remarkably similar between those in-work and those unemployed. Does the parallel fall in job to job moves disprove the theory that changes for the unemployed were being driven by Jobcentre Plus? Possibly, but some of those job to job moves may actually be people moving through unemployment (as they are only surveyed every three months).


Fourthly, finally, looking across the whole period, it is striking that job finding rates have only regained half of the ground lost on the boom years of the early 2000s. In terms of job-finding, we are still a long way from recovery.


Next steps


Given that the job-finding rates for the employed and the unemployed are closely linked (and those for the inactive slightly less so), are they linked closely to other economic indicators, and what does this mean for economic management and for expectations for programmes? Watch this space…