When discussing unemployment and social security benefits, should those of us who believe in a more generous system try to avoid talking about ‘welfare’?
Many researchers and campaigners believe that that the term ‘welfare’ activates ideas about ‘handouts’ and dependency which reduce public support benefits. For example, at a recent event hosted by the Commission on Social Security, the charity communications expert Nicky Hawkins argued that “talking about the social security system is probably the best way to go, moving away from talking about welfare or benefits” [at 46:20].
She (and others) make a highly plausible case. Where the terms ‘welfare’ and ‘benefits’ have come to imply ‘handouts’ to ‘dependents’, ‘social security’ draws on an older, more universal idea of protection from uncertainty.
However, no-one has tested whether these terms actually make any practical difference to how the public think about the benefits system. This is something we therefore decided to do as a small part of a much larger project on the British welfare system during the pandemic.
Our study
To investigate the effect of different terminology on people’s welfare attitudes, we used a survey experiment. Respondents answered otherwise identical questions in which we randomly substituted the terms ‘welfare’, ‘benefit’, and ‘social security’. If, on average, people give more generous responses to questions which refer to ‘social security’ than to those which reference ‘welfare’, this is evidence that the terminology really does matter.
In our survey, everyone was asked a longstanding YouGov question about ‘benefit payments’, but the term in the brackets was randomly varied:
Thinking about the level of [Benefit/Welfare/Social Security] payments, do you think they are too high, too low, or is the balance about right?
Respondents could choose from one of four response options, which repeated the variable wording:
[Benefit/Welfare/Social Security] payments are too high
[Benefit/Welfare/Social Security] payments are too low
The balance is about right
Don’t know
(A quick note: our YouGov survey was conducted 15-17 June 2022 with 3,499 people; all data are weighted using the YouGov-supplied weights. Data are already available via OSF and will soon also be available via the UK Data Archive).
The results are given in the chart below:
Figure 1: Whether people thought payments are too high/low, split by whether asked about ‘social security’, ‘welfare’, or ‘benefits’
The most immediately striking finding is that attitudes were pretty similar whichever wording we used. But to the extent that there were statistically significant differences, attitudes were actually more negative if the question referred to ‘social security’: only 29% of respondents thought payments were too low when we asked about ‘social security’, compared to 35% when we asked about either ‘welfare’ or ‘benefits’).
Why doesn’t ‘social security’ produce more positive attitudes?
We think there are three things going on here (but please do comment below if you have a different take!). First, the term ‘social security’ hasn’t been used much in recent years in the UK, and so isn’t as well-recognised as ‘welfare’ or ‘benefits’. You can see in the figure above that more people said ‘don’t know’ when asked about social security than when asked about welfare or benefits. Indeed, the proportion of ‘don’t knows’ rises to almost 50% among younger respondents (aged 18-34) asked about ‘social security’.
Second, the argument in Britain about the potentially stigmatising nature of ‘welfare’ versus ‘social security’ has been heavily influenced by the debate in the US. For example, Liam Stanley makes the case against saying ‘welfare’ by citing a series of survey experiments conducted entirely in the US. However, the social meaning of these terms is very different in the two countries. The term welfare is much more commonly used in the US than it is here, where we are more likely to talk about ‘benefits’. ‘Welfare’ in America also has a clear, and heavily racialized history (the Black ‘welfare queen’ stereotype is an American invention) as a stigmatising term, in a way that is not matched in the UK. There is also the fact that ‘welfare’ and ‘social security’ actually refer to different payments in the US: where the social security system covers contributory benefits (including pensions), while ‘welfare’ refers specifically to means-tested benefits. In the UK, these terms all basically refer to the same thing. It is therefore not particularly surprising that there would be a strong effect of the ‘welfare’ terminology in the US, but little to no effect here.
It is possible that people are more likely to think of pensions as part of ‘social security’ than part of ‘benefits’ or ‘welfare’. (In a separate question in our survey, we found that only 21% thought that people claiming the state pension were ‘benefit claimants’). Given that support for more spending on retired people has gone down (at least to 2018), this may explain why people had slightly more negative attitudes towards raising social security vs. raising welfare. But even if this explanation is true, it doesn’t suggest that talking about ‘social security’ is automatically better than talking about ‘welfare’.
Finally, there is overwhelming evidence that framing matters – the way we talk about something obviously affects how we think about it. But on their own, using the term ‘social security’ rather than ‘welfare’ doesn’t change our frames – at least not in the UK. Indeed, there is nothing inherently negative about the term ‘welfare’ (we talk unreservedly about the ‘welfare state’). The talk by Nicky Hawkins is full of insights about how to tell compelling stories about the benefits systems – we strongly recommend you give it a listen, and we’ll return to it on the blog soon – but these stories are about more than the terms you use.
Or put simply: frames matter, but – in the UK at least – the words you use don’t.
This post was written by Ben Geiger and Robert de Vries, based on their work on public attitudes within the Welfare at a (Social) Distance project. We’ll have more reflections on both public attitudes and the Commission on Social Security on the Inequalities blog in the coming weeks.
Inequalities 
6 responses to “Should we talk about ‘social security’ instead of ‘welfare’?”
As an aside, our survey experiment approach is basically the same as the classic US studies cited by Liam Stanley as showing that ‘welfare’ is a terrible term to use. The main difference is that the studies that Liam Stanley cites often compare ‘welfare’ to assistance for ‘the poor’ or ‘unemployed’, and these don’t necessarily refer to the same thing (for example ‘assistance for the poor’ might include e.g. Medicaid as well as cash transfers)
Many anti-poverty organisations have had this discussion and moved on from it years ago. The Joseph Rowntree Foundation created a ‘Talking About Poverty’ initiative, which was based on extensive research and provided a range of frames – fully worked out metaphors and ideas, rather than just individual bits of language – that had been proven to shift negative attitudes significantly. It’s widely used in the sector, and it’s surprising not to see it mentioned in a piece about benefits and framing.
Yes, I know about the JRF work (and indeed loads of other work on framing, from both within academia and outside it). But clearly the discussion about which term to use hasn’t been settled, because it kept coming up at the Commission on Social Security events (and indeed, wider conversations that I’ve been part of). I’ll be returning to other things about framing on the blog over the year – but rather than doing comprehensive reviews, the point of each blog post will be to look at one particular thing!
Echoing what Ben said (sorry this is an edit of a comment I made before I saw Ben’s reply!) – we just wanted to do something quick as part of our broader WASD project to see if this idea that ‘welfare’ was a bad word to use (which, as Ben says, keeps cropping up) was supported in the UK. And as you can seem we found that it wasn’t. Now I’ll happily admit that this is a minor point relative to e.g. the JRF’s work on what actually helps to shift attitudes. But it’s something we hoped would be useful the next time the debate about ‘should we say welfare’ crops up again.
Thanks both. Just a bit depressing to see the same discussions going round again, and reaching similar conclusions. One of the nice things about the JRF initiative was that it got different organisations talking together about framing so that we could present a more coherent message. But if, as you say, people are still raising the question…
I found your finding interesting. In Australia, where I live, we have similarly been debating whether to use ‘social security’ or ‘welfare’. I think you are right that the national context makes a big difference. I’d like to see a similar survey done with the Australian public. I suspect the results would be similar to your results because our system is more similar to the UK than the US.