First, some good news. When a recent UK Statistics Authority survey asked a representative sample of the public, 85 per cent of people who expressed a view said they trusted the statistics produced by the ONS. Moreover, nearly three in four agreed they were produced without any political interference.
This doesn’t stop some people on Twitter from expressing the view that our labour market figures are misleading or manipulated to support the government of the day.
Here are some the most common misapprehensions I’ve spent a lot of time rebutting and, in the hope we can persuade anyone who remains undecided on the veracity of our statistics to join the majority who trust our data.
“The jobs figures are rigged by the Government”
The main numbers do not come from government data on jobs or benefits. Instead they all come directly from the Labour Force Survey (LFS), a very large household survey carried out by the ONS right across the UK. That’s us, talking directly to real people in their tens of thousands, and logging information about their own experiences. I should also take this opportunity to say thank you to the thousands of people who take time to respond to our survey.
We ask, among many other things, whether they are working, seeking work or not active in the labour market at all – in other words, we’re measuring their actual labour market behaviour.
Ministers and their advisers have no influence on the ONS employment figures and no longer see the figures before they become public. We are independent of ministerial control and accountable, through the UK Statistics Authority, to Parliament rather than the Government.
The claimant count is derived from benefits data provided by the Department for Work and Pensions (DWP). But it it’s compiled by ONS statisticians operating under a strict code of professional impartiality. It also ceased long ago to be part of the headline employment indicators – those are all drawn from the ONS survey.
“Employment growth is driven by people working an hour a week or those on zero hours contracts”
Only a very small proportion of people in employment work very low hours – in January-March 2019 only 1.4 per cent of workers usually worked six hours a week or less. For many years this proportion has been falling, not rising, so is not a new thing. Likewise, only about 2.6 per cent of workers are in a zero-hours contract in their main job, and this proportion appears now to have peaked or to be falling.
Yes, someone who only works for an hour a week is technically defined as employed because that’s the standard international definition of employment (more on that below). But the numbers of such people would seem to be vanishingly small.
“The definition of employment and unemployment has been changed by the government to massage the figures so our figures aren’t comparable with those for other countries”
The definitions used in the figures – what counts as employment and what as unemployment – are agreed internationally at the forum of the International Labour Organisation (ILO). The reason for this is, of course, precisely to ensure that different countries’ figures are comparable. These definitions are long-standing and haven’t been altered. Because they are set internationally, they are not subject to alteration by the UK government and there have been no changes at all to these definitions since the change of government in 2010.
“The latest figures are not on the same basis as figures for the 1970s”
Our employment and unemployment statistics go back to 1971 and we use the same ILO definitions throughout, so it is not true to say that our latest figures are not comparable with those for the 1970s. Our figures for the 1970s and 1980s are not the same as when those figures were originally published.
That is because unemployment statistics used to be based on the number of people claiming unemployment-related benefits. However we have used ILO definitions as our main indicator for over 20 years now because the previous method did not follow international standards and bred scepticism about the independence of the figures. Our unemployment figures have been based solely on the LFS since the spring of 1992 when we started doing that survey on a continuous basis. The figures from 1971 up until the spring of 1992 are based on other surveys and indicators but use the same ILO definitions as the figures from spring 1992 onwards.
“ONS includes unpaid family carers in the employment figures”
This is an understandable mistake, but a mistake all the same. There is a small sub-category of employed people called “unpaid family workers” (though it only amounts to 131,000 people compared with 27.6 million employees and 4.9 million self-employed). That’s less than half of one per cent of all people in employment. However, it does not cover unpaid carers, who would not count in our employment figures if their caring responsibilities prevented them from being available for work. ‘Unpaid family workers’ are actually people who work in a family business without receiving a formal wage or salary but benefiting from its profits instead.
“Unemployment is a count of benefit claimants, and so ‘sanctioned’ claimants and those not claiming benefits are excluded from the figures”
That’s not the case at all – as I said, the unemployment figure comes from the ONS survey, not the DWP which manages the benefits system. In this, all people aged 16 or over who are not currently working but are both looking and available for work count as unemployed, no matter what benefits they may or may not be receiving. So having one’s benefit claim sanctioned does not directly affect this at all.
“The school leaving age has been raised to 18, so 16- and 17-year-olds have been removed from the unemployment figures”
It is not in fact the case that the ONS doesn’t measure unemployment among this age group. Our published figures show that in January-March 2019, there were 81,000 unemployed 16- and 17-year-olds, of whom 61,000 were still in full-time education. Of course, it is perfectly possible still to be in full-time education and count as unemployed – a student who was actively looking for a part-time job would meet the ILO definition and so would count in our numbers. Nor is it correct to say that the school leaving age has been raised to 18: while 17-18 year-olds are required to undertake some form of education or training, there is no requirement for them to remain at school. People in that age group can enter the labour market and some choose to do so.
These, then, are the main misconceptions I’ve had to deal with. I suppose it’s understandable and only human to doubt figures when they don’t match your own experience. In addition, our statistics cannot capture every aspect of the labour market that is important to people. We’ve also been reporting record high employment levels at a time when according to other ONS figures the economy has not exactly been booming for everyone. Wage growth also has by historical standards been quite slow, and adjusting for inflation, wages remain below their 2008 levels.
The ONS survey is probably the biggest regular, long-running household survey in the UK. Its findings aren’t precise – there’s a degree of uncertainty in any sample survey, of course. But the ONS operates transparently and in the public interest. If anyone wants to know more about how the figures are compiled or what they say, we’re always happy to explain.
Richard Clegg is an employment statistician at the Office for National Statistics. After 15 years as senior statistician presenting the monthly UK’s employment figures, he is retiring from the ONS.