Labour market statistics: go with the flow
The gloomy news about the economy makes it likely that we will be seeing higher unemployment in the near future, and increased media attention will be paid to the monthly Labour Market Bulletin put out by the Office for National Statistics.
The ‘headline figures’ are the level of employment, unemployment and inactivity, and their respective rates expressed as a percentage of a relevant denominator. In the case of the inactivity rate, for example, this is the number economically inactive as a percentage of those of working age, conventionally (though increasingly inappropriately) taken as 16-64.
Level variables are effectively ‘stocks’ – in this case, the average numbers of people over a given period. But in order to understand why these stocks increase or decrease we need to examine the ‘flows’ between various labour market statuses. These can shed important light on what’s going on in the labour market.
For example, a few years back there was a sustained increase in the ‘stock’ of the self-employed. Many quick opinion articles were written. From the Right, the story was how this showed a rise in entrepreneurial energy as more Brits looked to work on their own account; from the Left, workers were being made redundant and were having to scrape a living doing odd jobs.
Examination of ‘flows’, however, suggested an alternative interpretation. There was not so much a big increase in the numbers entering self-employment, but rather a decline in the numbers leaving this status. The self-employed were sticking with their businesses for longer. If outflows were reducing, the stock of the self-employed would naturally rise.
Labour market flows are complicated. We normally think of four statuses: that of an employee (working for an employer); being self-employed; being unemployed and being economically inactive. These last two statuses cause confusion. For many years we’ve used the International Labor Organization definition of unemployment, which means that people are available and willing to work, and in the last four weeks have made efforts to obtain a job by, for instance, answering adverts or contacting Jobcentre Plus. Economic inactivity involves being unable or unwilling to work – for instance by being ill or disabled, in full-time education, looking after young children or aged parents full-time, or taking early retirement. Whether you are unemployed or economically inactive does not match with claiming benefits, as many still seem to believe.
Our knowledge of the numbers in these statuses comes primarily from the Labour Force Survey, a regular household survey covering tens of thousands of households each quarter. Respondents are asked to state whether they are employees, self-employed, unemployed or inactive. The LFS is a rolling survey where ‘cohorts’ overlap, with some leaving and some entering each three-monthly period, while each household stays in the sample for five quarters. Changes in status between surveys enables us to observe ‘flows’ between the various ‘stocks’.
In any period, there are twelve flows which can be measured – employee to self-employment, unemployment or inactivity; self-employment to employee, unemployment or inactivity; unemployment to employee, self-employment or inactivity; and inactivity to each of the other three statuses. We can also track people who change jobs while remaining in employment: in the most recent three-month period for which we have data there were just shy of a million job changes.
I’m probably losing you already but let me just take this a little further. Let’s concentrate on inactivity. This has been very much in the news recently.
There are nearly nine million working-age people currently inactive. This total ‘stock’ barely changed in July-September this year from the previous quarter; however, it was by no means always the same people making up this number. Out of sight, the labour market was churning away, with many people changing their status.
572,000 previously inactive people took employee jobs, but on the other hand 536,000 previous employees became economically inactive. 109,000 inactive people became self-employed, but 86,000 previously self-employed became inactive. And while 489,000 inactive people became unemployed (in other words, they were now newly available and had started looking for work), 316,000 unemployed people, however, gave up looking for work and became inactive.
The net flows between statuses may have been relatively small, but there’s clearly a lot going on under the bonnet. It is changes in flows which lead to the more visible changes in stocks on which people pontificate.
Incidentally, anybody alert enough to check will notice that the net flows resulting from the figures just quoted do not add up to the observed change in inactivity which was published in the last Labour Market Bulletin. This is partly because the figures here are not seasonally adjusted, as the Bulletin figures are. But it is also because of a further complication in the flow picture.
The LFS flow data cover UK residents who were in the 16-64 age group in both the last period and the period before. However, in a three-month period new 16-year-olds will enter the labour market, and they can in principle enter each of the four statuses. Similarly, older workers will exit the working-age population as they turn 65 (although, of course, they may continue to work). They may do so from each of the four statuses. This means eight additional flows, some of which may be in hundreds of thousands, to complicate the picture.
Nor, unfortunately, is this all. For what about migration? In any period new migrants will enter the workforce, and they can in principle again enter any one of the four stocks. Similarly, significant numbers of working-age people will migrate from the UK, either temporarily or permanently; they can come from any of the four stocks.
If you’re keeping score, that makes twenty-eight different flows. For the last sixteen of them, we don’t have figures which precisely match into the LFS data.
So understanding exactly what is happening in the labour market is always going to be a difficult task. Simple evidence-light stories about why the economic inactivity rate is higher or the numbers of self-employed have been falling may therefore need to be treated with a degree of scepticism. Importantly, so do policy proposals based on these stories.