Just how common is benefits conditionality?

There are newspaper claims that conditionality is rare, but wider claims conditionality is ubiquitous – this post explains how both can be true.

This is the 1^st of a series of upcoming rapid posts that relate to the Green Paper – including some new evidence on mental distress among claimants, how I think Labour can square the circle on welfare, and what a progressive agenda on ‘medicalisation’ might look like.

Last week, The Times ran a headline saying, “Three million claim benefits with no obligation to seek work”. At the same time, academic research sometimes says conditionality (that is, the threat of sanctions) is ‘ubiquitous’ in the UK, felt by ‘virtually all’ claimants.¹ Which is a puzzle – how can benefits conditionality be both rare and widespread?

In this blog post, I argue that we need a better understanding of conditionality. This doesn’t just help us figure out this puzzle - it is also necessary to help make policy choices about conditionality, including in the impending Green Paper.

The study behind this post

This blog post is based on a just-published open access paper with the amazing WASD team (co-authors are Lisa Scullion, Daniel Edmiston, Robert de Vries, Kate Summers, Jo Ingold, and David Young). It comes from a survey of 3,800 UC/ESA/JSA claimants in May-June 2022 – so it refers to the situation 2½ years ago, though this was after the point that conditionality had been reintroduced after Covid-19, so is likely to be pretty similar to the position today.

The paper has more details of the methods, and you can access the raw data from the UK Data Service. One detail to flag - UC is a household-level benefit where conditionality depends on both partners. In our results, we therefore often divide between ‘working households’ and ‘non-working households’ (those working 16+hrs/week or with a working partner are classified as ‘working’, as they are likely to be above the threshold which exempted people from conditionality) - see the paper for details.

Evidence that conditionality isn’t so common

The source for the Times piece above is the official administrative data on conditionality groups of people receiving benefits. We started by comparing this admin data to how people described their conditionality group in the survey, shown below:

This gap between admin data and survey data is predictable, and probably reflects three things: (1) some people are formally subject to conditionality but are exempted by their work coach (‘easements’) - which we don’t have any admin data on, unhelpfully; (2) people might be unaware of conditionality where it is applied ‘lightly’ (e.g. where they’re already doing what the work coach would have asked them); and (3) we know from a little DWP research that people don’t always accurately report things in surveys. We’ll come back to methods issues below.

What the survey lets us do, though, is to see what conditionality people actually experienced, irrespective of the conditionality group they said they were in. This is shown in the figure below, and for the time being, let’s focus on the people in the intensive conditionality group² in red.

Firstly, even among people that said they were in the intensive conditionality group, only 57% said they were explicitly asked to search for work/increase their hours/attend a training course in the last year, and only 26% said they received a ‘warning that [their] benefits will be stopped/reduced’ if they didn’t do these. While the vast majority of these claimants reported some experiences of conditionality, these were split roughly equally between those that were warned about a sanction / those that were asked to do something / those that just had to sign a claimant commitment. Nearly 20% of people didn’t remember any of these things happening in the last year.

But this shouldn’t be taken as evidence that conditionality is even rarer than it looks - because you can easily misunderstand what’s going on here. Let’s look a bit closer.

Evidence that conditionality is felt widely

To understand the full picture, you need to know two more things: that conditionality is reported by some people who aren’t meant to be subject to it, and that people who aren’t subject to conditionality nevertheless say that it harms their mental health.

Firstly, let’s go back to the last chart, but looking at the orange and green bars. These show that even for those that say that they’re not currently required to take steps towards work, some people still say that they’ve been asked to search for jobs/increase their hours/take training in the last year, and a small number say they’ve been threatened with sanctions if they don’t do this. This isn’t the majority experience - but conditionality does seem to leach across boundaries (which other research also shows).

Secondly, people often feel conditionality harms their mental health - particularly among people in the intensive conditionality group, but also among other claimants too. This is shown in the figure below:

It was particularly people with disabling mental ill-health that said that conditionality made their mental health worse (they were 12 percentage points more likely to say this, against a baseline of 22% saying it - so it’s a pretty big difference). But let’s put that to one side for the moment, because mental health is the focus of next week’s post…

A better understanding of conditionality

Putting this together, our argument in the paper is that there’s four things we need to understand:

1. Methods matter – surveys will contain some errors in how people report their conditionality group. But equally, it’s likely that qualitative research may over-represent people who have experienced conditionality (indeed, the brilliant Welfare Conditionality project only selected people that they knew were subject to conditionality). There’s nothing out there that simply gives you ‘the truth’.

2. Conditionality in practice may be different to conditionality on paper - we found some people reporting conditionality in practice when they said they shouldn’t actually be subject to it. It’s tricky to figure out what’s going on, but there’s wider evidence from DWP research that sometimes things that are meant to be ‘voluntary’ are sometimes perceived by people as ‘mandatory’.

3. The past casts a shadow over the present – some of the inconsistencies between questions are probably because they ask about different timescales (the paper explains this in more detail). But this reminds us that people move between different conditionality groups - so the intensification of conditionality for one group may influence people who are later in the ‘no requirements’ group. From talking to people receiving benefits, it’s striking how strongly people’s current feelings are influenced by past experiences, particularly really bad ones.

4. We need to think about implicit conditionality – while sometimes conditionality is about explicit requests or threats, it can also be about an implicit sense of threat and insecurity. As Finn (2021) puts it in, despite a surprising ‘lightness’ of explicit conditionality in Ireland at that time, the threat of sanctions “hangs over the entirety of the population and serves to ensure engagement even it if amounts to ‘tactical mimicry’ of a good jobseeker’”. So policymakers need to consider both explicit changes to conditionality practices, and the implicit sense of conditionality that people will take from past & proxy experiences, media reports and the like.

Or in short, as the paper concludes, “our paper helps to understand that [‘ubiquitous’ conditionality] does not mean that explicit conditionality is ubiquitous in the present moment, but rather that conditionality is widely-felt by claimants, either due to past experiences or because of an implicit sense of threat in the present.”

Final things for the Green Paper

There’s a key message here for the Green Paper: there’s a clear risk that headlines about intensifying conditionality will contribute to a wider climate of fear (‘implicit’), irrespective of the detailed changes in practice (‘explicit’). And given the wider evidence that conditionality has much more negative effects on people with mental health problems, this may end up pushing some people away from the labour market, as well as harming their health.

Finally, I haven’t talked about another finding of the paper: many people receiving benefits thought that conditionality was unreasonable. In particular, 52.2% of those with care/health-related barriers said their work coaches didn’t take their circumstances and barriers into account when asking them to do things, and 25.0% disagreed outright that what was asked of them by their work coach was reasonable. This draws attention to the need to create a proper system of administrative justice surrounding welfare conditionality, so that the exercise of discretionary power by work coaches is accountable and demonstrably fair (as I explained further here). Again, something to continue pushing for in the Green Paper and beyond…

1. See the full paper for citations - I haven’t put them here because the papers are often (but not always) more nuanced than there’s space to put in a blog post. The description in the paper is fairer to the authors concerned, but basically makes the same point.

2. The paper refers to people being in the ‘intensive conditionality group’, but this isn’t the language the DWP uses - which confusingly varies between ‘intensive work search’, ‘all work-related requirements’, and ‘searching for work’. But we’ve preferred the term ‘intensive conditionality’, partly because it’s a bit clearer, and partly because we’re sometimes (like in the 2nd figure) combining JSA claimants with UC ‘intensive work search’ claimants into a single group.