A blazing row erupted earlier this week with the publication of a charity research report on food banks – the latest in a series of blazing rows on food bank use in the UK. At stake was the claim that food bank use is related to issues with the benefits system, a claim that has been made repeatedly but which has been steadfastly rejected by the Government. The charity Sense about Science got in touch with me to ask about the charity report, and after sending them a briefing that underpinned their own blog post, I decided to write a fuller explanation here.
Probing the latest report
So what exactly is the row about? The initial claim, as reported by the Guardian, is that “at least half of all food bank users are referred because they are waiting for benefits to be paid, because they have had benefits stopped for alleged breaches of jobcentre rules or because they have been hit by the bedroom tax or the removal of working tax credits.” The study, says the article, is “the most extensive research of its kind yet carried out in the UK.” Yet in response, the DWP reportedly dismissed the claim, saying “the report itself concludes it can’t prove anything.”
Which inevitably leads us back to chasing down the numbers, and examining them with a careful critical eye. The report containing the numbers is called Emergency Use Only: Understanding and reducing the use of food banks in the UK, and was published by the Child Poverty Action Group, the Church of England, Oxfam UK and The Trussell Trust (the largest food bank provider in the UK). It was written by a respected team of researchers led by Jane Perry, and including Tom Sefton, who I know from our respective stints at CASE, and have always been hugely impressed by. The key table from the report is this one:
A sign of the pedigree of these researchers is that they themselves are very open about the limitations of these numbers – and they are right to be. The Guardian claim above comes from a new survey of 903 food bank clients, with the questionnaire filled in by food bank staff in just three Trussell Trust food banks (see p99-100 of the report for details). Almost no information is given on the response rates to the survey, other than to note that the food bank in Durham – which accounted for over 600 of the responses – ‘was completing the form with every client’.
This survey therefore can’t be taken to be statistically representative of the three Trussell Trust food banks in question (with the possible exception of Durham), let alone food banks in the country as a whole. And this is admitted by the researchers themselves, who (p19) lay out exceptionally clearly that this data can’t ‘statistically represent all food bank users – either at those three food banks or nationally.’ The Guardian does have a short caveat about the limitations of the study further down the page, but I still think it gives a misleading impression about how confidently we can say that at least half of food bank referrals in general are for benefit-related reasons.
Why the DWP’s view still flies in the face of the evidence
Yet despite these limitations, the DWP’s outright denial of any link between the benefits system and food bank use seems to fly in the face of the evidence. This might appear slightly counterintuitive after I just went into the limitations of the Parry et al study, so let me explain the three reasons behind this.
Firstly, the Parry et al data is not statistically representative – that is, we can’t reasonably assume (with a known margin of error) that the true benefit-related referral rate across the country is the same as among the people they spoke to. But is still has a good claim to be representative in the broader sense of being typical. This is partly because the reasons that people are referred to these particular food banks are pretty similar to the reasons that people are referred to food banks across the country (see p102). But it’s also because these three food banks are a bit different from one another (one is in Durham, one in London, one in the suburban South-East), but the rate of benefits-related referrals is pretty similarly across all of them (50-55% due to delays, sanctions, or being found fit-for-work). So while the DWP are quoted as saying in the Guardian article that “the report itself concludes it can’t prove anything – it uses self-selecting data,” it actually seems reasonable to suppose that this is likely to be about right across the country. It’s not as good as having a random sample across the country; but unless someone spends tens of thousands of pounds on this, the rate in this study seems like our best guess.
Secondly, the Trussell Trust have data for everyone that is referred to them – a much bigger and more nationally comprehensive data source. And this similarly shows that benefit delays or benefit changes are cited as the single reason for referral in 45% of cases. These data aren’t perfect, as they’re filled by all sorts of different people (including some people in Jobcentres…) who refer their clients onto the food bank. And in fact, one the most interesting findings of the Parry et al report (p27) is that the referral data may underestimate the role of food banks, because nearly half of the people whose reason for referral was ‘low income’ also had “also reported one or more ‘acute crises’ attributable to problems with benefits.” So we have good reason to think that the national data might be a bit of an underestimate, but at least it’s national.
Finally, detailed qualitative research – such as in the Parry et al report itself (whose biggest is a really valuable analysis of why people turn to food banks), or by Hannah Lambie-Mumford (or this), or a team at Herriot-Watt in Scotland, or my friend Kayleigh Garthwaite’s ongoing ethnography in Stockton-on-Tees – consistently show how delays and sanctions in the benefits system lead to food bank use. And frontline charities such as Citizen’s Advice come to the same conclusion, including when they log the reasons for food bank referrals systematically. And this should be no surprise when we know that there are massive backlogs for some benefits (e.g. PIP), and that benefit sanctions have been rising – if some people in need aren’t getting benefits and are struggling to survive, then we would expect them to turn to food banks. This is to say nothing about the ‘genuineness’ of claims, or whether people are to blame for sanctions, or whatever – it’s just that there’s an obvious link between benefit delays/sanctions and food bank use.
Smoke and mirrors
In summary, we’re not quite sure about what proportion of food bank users are there because of benefits-related issues – and the Guardian made out that the numbers were slightly more solid that they actually are. But the best guess is that the numbers in the Guardian article are about right, and that benefits play a part in around half of food bank referrals. And whatever the precise figure, we can be absolutely sure that benefits are playing a part in some food bank referrals – partly because it would be slightly astonishing if none of the people left without benefits went to food banks, and partly because of overwhelming evidence from several different sources. The DWP’s denial is probably more because a not-quite-existing world is easier for the Government to deal with politically, than because there’s really any doubt that there is a link.
A related issue here is whether food bank use is rising because of increased food insecurity, or simply because there are more food banks in existence (as the Government have claimed). One of my PhD students has been doing some work on this over the summer, so when the work is ready to release to the world, we’ll get back to you on the blog.