Caroline Selman
speaker
38 appearances
1 recordings
1 series
first heard Oct 2024
last heard Oct 2024
Caroline Selman’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsNo recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.
Appearances
Really? I've not come across that before.
Yeah, so like you say, we are currently working with Centring and Law Centre round about supporting people in relation to sanctions. So that's when people have money taken away if they haven't, for example, attended a work-focused interview with the DWP. And although they tend to be things that are put in place for not particularly serious things like being a bit late to an interview,
The consequences for people are really very serious. So you're looking at sort of about 100% of somebody's standard allowance if you're a single person, sometimes for really quite a long amount of time. And one of the key concerns we have about the sanction system is it's been shown not to work. in terms of supporting people into work or better paid work.
But it has been shown to be really harmful repeatedly across both sort of independent studies and also some of the government's own evidence as well. So there's a core concern as to why that regime should be being used at all.
But the piece of work that we're doing with Central Good Law Centre is about supporting people to challenge sanctions if they are challenged, because people can ask for an internal review and appeal them. And what we find is that if people do challenge them, they're very often successful.
I think most people, regardless of where you are on the political spectrum, if you have something that's been found not to work and which causes harm, that's probably not a policy that is a good one to be applying to people. More broadly, we have also got concerns about how fit DWP is to understand who is being impacted by things and how.
What we've seen recently in terms of that statistics is that some people are more likely to be impacted or harmed by the sanctions regime than others, which also raises a whole host of concerns about how decisions are taken by the DWP in terms of...
how policies impact people but another area that we do a lot of work on is how the DWP is increasingly using things like data analytics and machine learning that's been identified by the National Audit Office as having a real risk of bias within it but where the DWP is on record as saying they don't have the data or the way of knowing whether that bias is in place but still rolling it out at pace.
The thing that rang really true was that bit about the impact on trust as a result of sanctions. And I think, you know, that's not just in terms of our own experience in relation to it. It's also, you know, borne out by DWP's own findings. And then I think in terms of some of our own sort of experience in this area. So we often come at stuff from an access to justice perspective.
So how we're supporting people to exercise their rights or challenge incorrect or unfair decisions. But one of the things that really comes through so strongly in that is that wider pernicious impact of justice.
Everything that John's just been describing of people's experience of the system as whether it's one that you see as supportive and one that you can trust or whether it's one that you feel is set up to potentially assume the worst of you or to be punitive.
We see from casework that quite a common story is, like you say, where somebody has been awarded something by the DWP, relied on that as being something that they are entitled to, have often gone back and double checked and said, am I definitely entitled to that before then relying on that to spend it?
And then finding that they weren't entitled to it and having it recovered from them, causing the financial hardship that comes from that. So that's something that is very familiar to us from casework.
And again, in terms of DWP's own data, in terms of that context of official error, so where that mistake has been made by DWP, we know from statistics from a couple of years back, which have stopped publishing now, that three quarters of the overpayments they had on their debt management system were things that were caused by DWP.
And that's partly reflective of a change in policy and the legal framework which was introduced when universal credit was introduced, which basically gave DWP the power to recover overpayments even when it was their mistake. That said, they do have discretion as to whether to do it. And one of the concerns that we have is that they don't really apply that discretion before they apply deductions.
generally the default position is to recover that overpayment regardless of the context and regardless of the potential harm for somebody if it's recovered, even if it is to do with their own mistake.
In terms of what people can do is that people can, in those contexts, get in touch with DWP to ask for some relief in terms of the rate that it's recovered, or they can ask for it to be waived as well. So if particularly in that kind of circumstance where they've relied on something and checked about it, that is the kind of context where they can and should be able to ask for a waiver.
What we've found is that that is not well communicated to people and that people don't know that they have the ability to request that and ask for that and that in some circumstances DWP really should be granting that waiver, which is something we're quite keen for people to be more aware of and for DWP to do more as well so that people know about it as well.
Yeah, so what we know, and DWP has been very public about the fact that it's putting about £70 million into developing what it refers to as data analytics and machine learning, in particular in order to try and predict or identify whether cases should be investigated for fraud.
The concerns we have about that is that those things have been flagged up by, for example, the National Audit Office as having particular inherent risks of bias, which is also something that we know from other contexts in terms of risks associated particularly with predictive tools, where there's a risk that what happens is that they are baking in pre-existing biases that might exist in the system or flushing out some other forms of bias.
Showing 1–20 of 38 · page 1 of 2
Next →