Myths about out-of-work benefit claims?

There’s been a lot of talk about ‘benefit myths’ over the last few years – the things that people believe about the benefits system that aren’t actually true. I’ve almost finished a paper on this – watch this space! – but in the meantime I wanted to write about one new finding in the paper: the public’s beliefs about out-of-work benefit claims in general. And it doesn’t show exactly what you might expect.

The new data comes from a really interesting new survey by Ipsos MORI in collaboration with the Royal Statistical Society and Kings College London, on misperceptions around personal finance (which the head of their Social Research Institute, Bobby Duffy, wrote about in the Guardian).  Buried away in this, though, is a question on perceptions of out-of-work benefit claims that Ipsos MORI kindly included, prompted by an email discussion – Bobby Duffy and the Ipsos MORI team have a longstanding interest in public perceptions, and their data forms part of the wider paper I’m writing. So this blog post owes a debt to them, and I encourage you to read Bobby’s various writings over the past decade on misperceptions – there are few people in the country who have thought as long and hard as him about the things we wrongly believe!

The new questions

Previous studies have asked people about the level of unemployment or sickness benefit claims – but no-one has previously asked the public how many out-of-work benefit claims they think there are in total.  So Ipsos MORI asked the following two questions:

“At present, out of every 100 people of working age, how many do you think are claiming out of work benefits?   By out of work benefits we mean benefits for people who are not in work because they are unemployed, disabled, carers or single parents.                  

“Fifteen years ago (the year 2000), out of every 100 people of working age, how many do you think were claiming out of work benefits then?”

This allows us to investigate three possible myths: what people think now, what people think claim levels were fifteen years ago, and what people think the trend has been (which we get from subtracting the answer to the second question from the first question).  The data comes from an Ipsos MORI i:Omnibus web panel, weighted to match the general population on age, gender, region, social grade, working status and main shopper status, and conducted in June 2015 (a few further details available here).  And I’ve previously blogged about the true figures here (though note these are for Aug 2014, as we don’t yet have the May 2015 figures – I’ll update this accordingly when we do).

So do the public believe ‘myths’?

Straight up, I’ll admit that I thought people would wildly overestimate the proportion of people who claim out-of-work benefits in Britain. But I was wrong.

The average person thought that 15% of the working-age population claimed out-of-work benefits in 2015 – and this is only slightly higher than the true figure of 11.8%.  Moreover, they also thought that 15% claimed out-of-work benefits back in 2000, when the true figure was 14.3%.  It just isn’t true that the average person is completely wrong about this; instead, I was wrong to expect it!

That’s not to say that most people got the answer right – the public may be right on average, but this doesn’t mean that individual people were right, just that people who were wrong in different directions cancelled each other out.  If we take ‘correct’ to mean a ten percentage point window around the true value (people tend to give round numbers), then 28% were correct about the current true figure, with 23% too low and 49% too high.  And because people overestimated by very high amounts – shown in the figure below – the public’s average estimate (the mean) is further away from the truth, at 23%.

Figure 1 - Perceived level of claims

This also means that most people were wrong about the trend in out-of-work benefit claims.  We know that 2.5% fewer of the working-age population claims out-of-work benefits now than they did fifteen years ago – which is a noticeable reduction (of 17% of the original level of claims), particularly given the recent recession.  Yet only 34% of people believed that claims had gone down, with 16% thinking they’d stayed the same, and half of people believing they had gone up. Even some of the people who believed that claims had gone down were wrong, with 12% of people thinking that claims had gone down by 10 percent of the working-age population or more.  The full range is shown in the Figure below.

Figure 2 - perceived trend in claims

(It’s worth noting that this is likely to be an upper bound on levels of knowledge, because the sort of person that takes part in web surveys tends to be slightly more knowledgeable than the average person (according to previous studies).  And this is a sample survey of 1,100 people, so there’s obviously sampling error around this that I haven’t discussed here).

I’d summarise this as: people often believe things about the benefits system that aren’t true, but that doesn’t mean they’re wildly wrong about everything.  Reviewing more than 40 other indicators of people’s beliefs about the benefits system, my wider paper comes to a broadly similar conclusion. I’ll put up more about the paper when it eventually comes out, but as always, I’m interested in your thoughts in the meantime, so please do comment below…

5 responses to “Myths about out-of-work benefit claims?”

  1. How does it look like broken down by income ? Also I am always uncomfortable with that kind of item because people vary in their anchor ( i.e. What percentage they see as a lot or not so much, as a lot less or a lot more …)

    • Thanks for comments Charlotte! I’ll break it down by income for a book chapter later in the year, which will also look at how it relates to attitudes, so watch this space (as it says above). However, I think it’s important to look separately at beliefs and attitudes (‘too much’ etc), even if they’re closely entwined within people’s worldviews – conceptually I think they’re quite different. For example, it’s an issue if people wildly overestimate the level of benefit fraud, even if relatively low levels of fraud (say the private insurance average of around 5-10%) are seen as ‘too much’.

      • My point was different: responses to these items cannot be compared across individuals because of anchoring effects (see our dear Kahneman), just a comment on descriptive inference 🙂 I would be fine saying that someone who says more than 50 % are claiming benefit is different from someone saying less than 50 % are (because 50 is a widely shared anchor) but comparing someone who says 20 and someone who says 40….less comfortable…are these individuals truly different in their priors?

    • Oh, I see – interesting! Are there any studies of anchoring effects as they relate to perceptions of the benefits system or welfare state? I’d instinctively think there’s some random noise here, but not anything that really changes the interpretation above. (Incidentally, in the longer paper that reviews belief questions, I separate out the continuous questions (which are more subject to innumeracy) from the categorical questions, and they basically show the same thing).

  2. A reader – being honest, my mum! – queried whether I’d taken into account the changing retirement age of women in the definition of ‘working age’. The ONS figures do acccount for this – these are consistently figures for people aged 16-64, irrespective of the women’s retirement age in that year. However, an increasing proportion of women aged 60-64 are no longer pensioners, and this is likely to INCREASE the out-of-work benefit claim rate (because they are more likely to claim out-of-work benefits than state pensions) – so if anything, the trend is underestimated. A really interesting comment on how this plays out for labour market figures in general is from Paul Bivand here.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: