BI podcast transcription: Barking and Dagenham

Rhian Gladman [00:00:12] Welcome to the first episode of The Nudges for Social Good podcast from the Local Government Association. My name is Rhian Gladman  and I've been managing the behavioural Insights programme here at the LGA for the past four years. And really what we're trying to achieve with this podcast is to demystify behavioural insights and help busy local government officers and councillors to understand more about how they can actually use these tools and techniques to help deal with their biggest service challenges. So on each episode, we'll be talking to a local government officer, a councillor who's actually undertaking behavioural insights projects within their own council so that you can learn from them and hopefully start to implement some of these projects in your own council. So I'm here today with Tim Pearse from the London Borough of Barking and Dagenham. Thanks for coming in, Tim. Tim is an old friend of the programme, has worked with us from the very beginning. So it's great to have you in here today. I'll let you introduce yourself and your career background. 


Tim Pearse [00:01:15] So I started life many moons ago as an economist in central government, but increasingly became interested in, I guess, what you call behaviour economics. So that's the intersection of economics with psychology. And as a result of that, joined the behavioural Insights team, which was a part of government but left to become a social purpose business, where I headed up the local government team, I did four years there, working across all areas of local government services, doing consulting projects. And six months ago, I left left the behavioural science team to join the London Borough of Barking and Dagenham as their behavioural science lead. So doing kind of similar stuff, but from the inside instead of from the outside.


Rhian Gladman [00:02:06] So welcome to local government, great to have you.


Rhian Gladman [00:02:11] So I guess a key question to kick things off is if I was completely new to behavioural insights, I'm working in a busy council, I might have heard someone in the office talk about behaviour insights or they might have read, you know, the nudge book or something. What exactly does it mean? What  is behavioural insights?


Tim Pearse [00:02:30] Yeah, that is a really good question. I think there's confusion about it. And different people have different definitions, but I'll give you mine. And that is basically that a lot of traditional government work is based on getting people to do stuff, but that is usually done with essentially carrots and sticks. Right. So for stuff that we don't want people to do, we use sticks like taxes or fines or prison and for stuff that we do want people do like eating healthier, whatever, we offer carrots. And so those things change our behaviour because they change our incentives to do something. Behavioural insights is different to that. It tries to change behaviour without changing people's incentives. It draws on psychology and behavioural economics to do that. And a lot of it is focused on the environment in which a decision is taken. So as a simple example, when you're asked a question will make a difference to your answer and also who asks you a question could make a difference. So thinking quite carefully about how decisions are made and in what context could also shape behaviour.


Tim Pearse [00:03:41] And this is a sort of can supplement the traditional kind of carrots and sticks approach.


Rhian Gladman [00:03:48] Healthy eating and carrots.


Rhian Gladman [00:03:51] Right, worked really well.


Rhian Gladman [00:03:54] Okay. So could you give us an example of a specific behavioural insights project that you've undertaken in your career so far? What was the behaviour you wanted to change and what did you do? And what was the impact? If you could talk us through that, that would be really helpful for out for our council audience.


Tim Pearse [00:04:10]  Yeah, definitely. So there's been, you know, a lot of work on the sort of quick wins around sort of how to change letters on council tax. So I'll try and take an example that's a bit more complex than that to show you how it could be applied, so in this case, criminal justice and it's a project I did while I was still out the behaviour insights team and we did it with Kent County Council and it was part funded by the LGA as part of the Parent Insights programme. And it was about domestic abuse. And so the thing that the behaviour we wanted change was to increase the uptake of domestic abuse support services, which was at that time just under 40 per cent in Kent. And the focus is really on kind of standard risk victims, though. Those are the lower end where they could be higher risk. But, you know, police were not aware at that time. And so trying to trying to get them to take up that offer could improve the improve their safety. Now, our research suggested that take-up was low, partly because victims did not understand the process. So the process is that after an incident occurs, victim support gives you a call two or three days afterwards. But unfortunately that comes from an unknown number. And the first thing that they say when you pick up the phone is they ask you to confirm your your name. That's for security. But obviously, it puts a kind of barrier in place. And although officers were kind of describing this at the scene, a lot of people didn't realise this is exactly how it happened. And they also didn't realise the full services available from victim support. So they didn't understand the breadth of the offer, including stuff like financial advice or the offer of kind of like physical locks on your windows as a broad support that people didn't realise was there.


Rhian Gladman [00:06:02] I guess there's a really important other dimension there as well. You know, if you put yourself in the shoes of somebody who has been a victim of domestic violence and then 48 hours later, you're going to receive a call from a withheld number, how likely are you to pick up that call?


Rhian Gladman [00:06:19] I guess it's about understanding that what would drive that behaviour as well, isn't it?


Tim Pearse [00:06:25] That's exactly right. And I think this is the kind of interesting thing about behaviour science, because as I mentioned, you know, police officers would sort of explain this process to victims. But if you have just been assaulted or something, that is not the best time to intake information. And so even though that looks like a good process on paper, thinking carefully about how people actually take in information, make decisions gives you a kind of a different answer to what good might look like for that process. And so hence we kind of focused on the environment in which that decision is taken. So thinking, well, is a sort of verbal briefing at the time good enough or could we be missing people with that? And so what we decided to do is we kind of decided to design a small business card type thing that could be left with the victim at the time, which would show the sort of benefits of the service. So give them the breadth of the victim support service. It would make the process clear, but it would also show that people were already on that process. So it wasn't something they had to opt into. They'd already reported someone to police.


Tim Pearse [00:07:36] They would be receiving a call and showing this sort of, you know, first step on this journey, we thought help people to sort of go with it rather than think of it, thinking of it as a kind of like something to opt into or get additionally.


Rhian Gladman [00:07:50] Can I just ask there as well about how did you manage that kind of potential issue around the perpetrator potentially finding the business card?


Rhian Gladman [00:07:57] And obviously, it's a sensitive issue just for other councils who'd be keen to, you know, take this project and roll it out.


Tim Pearse [00:08:04] Yeah. Yeah. So, you know, this was a, you know, an interesting project because cause of that sort of safety concern and that has to be the first priority. So our thinking on that and obviously our discussion with the police and the council was was very much around, this would be kind of it would be clear that this was a kind of standard piece of information that is given out to everyone. So if you call the police and it's a domestic incident, you would get this and it would be clear if you'd looked at that that that was the case. It wasn't the case that someone had kind of reached out and asked for it, you know, on a sort of personal basis. But also, it was you know, it was small, could be concealed, it could be thrown away the instant they got it. So we kind of wanted to give as much flexibility to the victims as possible. But yeah, safety was a big, big concern.


Tim Pearse [00:09:05] Shall I talk about the impact?


Rhian Gladman [00:09:08] That would be great, that kind of impact evaluation and the results. 


Tim Pearse [00:09:15] So again, measuring this was tricky because again, in an ideal world, we wanted to do this as a randomised controlled trial. And what that means is that you basically take all that all the victims of an incident over time period and give half of them the cards and half of them, not the cards. But we thought they'll get too fiddly, asking officers to try to remember to, you know, give half the victims a card and half not. So we basically went for a kind of a staggered roll out approach where we every few weeks a new station in Kent would start giving out the cards as standard.


Tim Pearse [00:09:55] So there are 12 stations and over a three month period, each one of those kind of - all the officers that responded to a domestic incident would hand out these cards. And then the way we measured it was basically by looking at the response rates to the victim support call in those areas where the cards were being handed out versus those that weren't. And so unfortunately, there were a couple of technical problems with the data set thing, you know, and these are common in kind of, you know, big analytical pieces of work. In this case, we found that unfortunately in quite a few cases, the officer that had attended the scene was not listed. And so we couldn't actually track back as to whether they'd given the card or not. So we did lose a fair few of our incidents, which was a shame. And that meant we were slightly under powered. So we didn't have the statistical power to say that the results were, you know, 95 per cent confidence right. So we would need to do that at a bigger scale. But what we did see was kind of a directionally positive result. So it you know, the numbers were positive. So the people that received the cards were 6.3 per cent more likely to take the call and 3.6 per cent more likely to accept the service. So we think that's positive. It didn't hit statistical significance. But, you know, we made the case to do it larger to to test that. And those seem like small numbers. And that is sometimes a kind of criticism lodged at behavioural insights. But if those are the case and if they were rolled out, that would lead to hundreds of lower risk victims taking up services that could protect them and potentially sort of save their life. So given that it's a relatively cheap and straightforward kind of way of doing things, we recommended that Kent did roll this out as standard.


Rhian Gladman [00:12:05] Can I ask to just bring you back to the directionally positive result? What do we mean by, you know, if I just keep this really simple, what do we kind of mean by that?


Tim Pearse [00:12:14] Yeah. Now, that's a really good question. And, you know, we want to be clear out there. So what I mean by that is we measured it and basically what came back is that the people that got the cards were 6.3 per cent more likely to take the call from victim support, i.e. pick it up. And then also 3.6 per cent more likely to actually take up a service once they'd spoken to victim support. However, there is a chance that was down to pure chance that through, you know, kind of randomness, that that was higher. Because one of them had to be high and that one was higher. Now, we don't know that's a a high chance, but we can't kind of rule that out at what you'd call kind of conventional levels of statistical significance, which is kind of like a one in 20 chance. So we think, as you know, it would be a greater than 1 in 20 chance. And so we have to sort of be a bit more nuanced about those findings.


Rhian Gladman [00:13:13] Okay. But in terms of, you know, the cost of intervention, the potential impacts. Yeah, it's a good one to roll out and to encourage other areas to to look into it?


Tim Pearse [00:13:24] Yes. Basically, our thinking on this was: is there any evidence of a backfire or negative impacts? And that was, no. Was the high costs involved? No. Do we have reason to be optimistic about it? 


Tim Pearse [00:13:40] Yes. And and so in discussion, you know, with the police and the LGA, we felt it was a good enough basis to say keep doing it and if possible, kind of, you know, measure and let's see what comes out of that.


Rhian Gladman [00:13:53] Okay. Excellent. So obviously, now that was when you working with the behaviour Insights team. Now, you've recently arrived in Barking and Dagenham. And you know, what of the behavioural insights projects that you are looking to work on in your current role at the council and and if you're can say about which of them have the greatest potential.


Tim Pearse [00:14:15] So, I mean, it's great to join a council where there are so many things that are going on that I could be involved in. And so that's always that's always a real pleasure, actually. And there's lots of people that want to do this and are interested. And so all that's fantastic.


Tim Pearse [00:14:34] Unfortunately, only one of me, which means at the moment.


Tim Pearse [00:14:40] But so so it is kind of limited is what I can do. But I guess there's a few things I'm working on right now, which are things like increasing health checks for the over 45s in the borough, improving school attendance, which again, is kind of interesting because quite, quite nuanced. It's not just about kids going off, there's all sorts of things around loneliness and social isolation and sort of mental health stuff. So really interesting issues around that. But also working with our social workers and thinking about how our processes around things like assessments currently work and can we kind of make them more behaviourally informed to support our professionals and also improve our services for the families that we work with. So that's a kind of a bit of a dip into it. So in terms of where I think the impact could be greatest, I'm not sure is the honest answer. I think one of the areas I'm really excited about and doing quite a bit on at the moment is around debt. So it's quite a significant issue in Barking and Dagenham. So, you know, people in high cost credit or arrears or whatever it is, and that basically is driving quite a bit of the kind of wider, poorer outcomes you might see. You know, people come into our housing service or people come in or require social care. And so it's an area of focus for the council. And so we're taking a kind of a behavioural approach to this. I'm particularly interested in how behavioural science can do the prevention side of things. And that's obviously the key to demand management. And I think behavioural insights is well-placed to do that. So what we're currently doing is taking data about people that are falling into council tax arrears or falling behind with their rent for their social housing and using that as a kind of an indicator of of people getting into financial difficulties. And then for those that are deemed most at risk, we basically send them a text. And the idea here is not to sound too much like a council, but just to say very simply, you know, hello, we noticed you fell into rent arrears, we have a service called the Homes and Money hub, which offers free financial advice. Just text us back with the number one, if you would like a callback. So making it really easy for people to engage, you know, the people maybe. I mean, you know, in sort of complex situations, we just want to get to them as quickly as possible before these kinds of things spiral out of control.


Rhian Gladman [00:17:26] And so is it literally that kind of high level wording, that's the type of wording in the text?


Tim Pearse [00:17:34] Yeah. So we want to be crystal clear, plain English sound kind of basically quite friendly and like an offer of support. And it also comes from our homes and money hub, which is quite trusted and seen as separate to the council.


Rhian Gladman [00:17:50] Okay. Is it that brand recognition of that service in the community? 


Tim Pearse [00:17:54] So it's not hi, we're the council and we need you to pay your rent or pay your council tax, it's hi we're the homes and money hub and we offer free financial advice.


Tim Pearse [00:18:05] And so, you know, obviously we can't text lots and lots of people that and we probably wouldn't want to because some people dip in and out of council tax debt. And that's part of life. But for  the for the small amount of people we have texted, we've got about a 50 per cent response rate, which is really positive. We have been trying this with letters before. The letters got very, very few responses. And what's nice as well is that the the issues that people have come in with have been things that we could help them with. So people falling into arrears, who weren't aware they were eligible for housing benefit.


Tim Pearse [00:18:43] You know, really nice, quick things you can get people straight, straight, sorted out, get them out the door. You know, you can get on with their lives, basically. So I feel like that and you know, that offers great potential. And I'd like to start thinking about where, you know, how we could build on that. What is a what is a Rev's and Ben's service look like that, the sort of, you know, screens people out who are vulnerable and gives them a different pathway to others. You know, that's still very much early, early thinking. But I think it could have quite big impact for people who are financially vulnerable.


Rhian Gladman [00:19:17] So a lot of councils across the country are looking at using behavioural insights techniques to manage demand into their services. So I guess with this one, there is a concern that, you know, if you text a lot of people, you're going to get huge demand back into your service. And that's an issue. So could you expand a bit more around around that manage demand.


Tim Pearse [00:19:42] That's absolutely right. And I think it's very much a proposition to do to kind of like, you know, move demand upstream. So there is there is a little bit of reaching out to people and ask them to come in, which may sound like the opposite of what you want to do but it is actually essential basically for it to tackle some of these more acute issues. So we know that Barking has a really good insight hub. And so we use the, you know, the data they hold to work out who we think is, as I say, high risk. And that can be based on a range of data so we're only texting those people who we think are most likely to require our help. Now, we could text more and we'd probably be able to help more people, but it sort of we'd also be texting people that we couldn't help. So there's that sort of real cut off. And I think what we were quite keen to do is like, let's just start doing something. And I am literally talking about five to 10 texts a week. Let's just start doing this, see what we're thinking, and then work out, you know, see what how effective this is and then see if we want to expand it. But it was very much like, let's start this small and let's start this on the bit we most likely think need help. And this will end up in a reduction in demand. And then let's evaluate it and see if we if we want to expand it. I think that will either reduce demand even more if we'll get kind of diminishing returns.


Rhian Gladman [00:21:05] Yeah, I really like that idea - picking up on what you said about just starting something and actually try something and let the insight lead you to where actually our resources at the council can be best used and best focused. Is that something you would suggest to other councils - you know, start small?


Tim Pearse [00:21:25] And so I think one of the things that is great about behavioural insights is that you can just start something small, you know, be a change in one sentence, in one letter. It's as simple as that. Just do it. You know, obviously go through the right channels, don't go too outrageous without checking it. But do something, get it out there, do what you can to measure it. And that can be, you know, ideally quantitative. But again, a lot of this data is collected already. So you just need to take it, even if it's a before and after or even if it's sort of some qualitative work to understand how services or staff are feeling about it get it measured and then you have something to go back to your organisation with to say, we did this, it showed this, we want to do more of it or we think we should do it over there.


Rhian Gladman [00:22:12] Really interesting. And great advice to councils. And, you know, just to crack on and try something. So that's the kind of work programme you've got coming up at Barking and Dagenham, which sounds really exciting and lots to be done. So you've been at behavoural insights team and now you're looking at things from Barking and Dagenham. Where do you see behavoural insights going for local government in the future? What kind of trends can you see on the horizon? Where could we be going with this?


Tim Pearse [00:22:45] So I think there is really big potential actually. And I think the kind of thing we might need to kind of develop is too many projects are focused on one very small part of the system. But, you know, BI can have big impacts, but some of the impacts can also be more marginal. And where I think the biggest opportunities are is to look across a whole pathway and put in several behavioural insights type changes and seeing what the overall impact of that is.


Tim Pearse [00:23:25] Say, for example, if we take debt, you know, thinking about how are we identifying people in the first place, how are we reaching out to them? What are the other touchpoints we could be using? You know, when they do come in, how do we how do we work with them? How do we help them stay out of debt and putting small nudges, as they're called, all along that pathway, I think even if each one of them only has a 5 per cent impact, could lead to quite a significant impact overall. And I think that's one of the real delights of doing this in local government. There's so many whole pathways that are, you know, totally council run or run with very close partners, that it really offers a chance to go kind of canvass or behavioural insights kind of across a system which could be really impactful and you can do one at a time and so it still has that kind of incremental and easy to do approach. But just just look, you know, it's worth looking beyond a single point and going a bit broader than that.


Rhian Gladman [00:24:25] I love that. I think that's fantastic. Like, look through the whole end to end service, the whole customer journey through our services. So you'd almost put several of these interventions in place and then would you sort of measure at the start and then measure at the end - do it that way? Or you could still measure those interventions or do it either way?


Tim Pearse [00:24:45] Yeah, I think it is a good question. It kind of depends on the system you're working on.


Tim Pearse [00:24:50] You know, measurement is is important. But but it's also mustn't sort of dominate everything else. And that's that's the thing that it's been difficult. You know, for me personally, I'm always keen to measure. But you must be careful that it's sort of proportional and that it doesn't sort of the tail doesn't start worrying the dog. So I'd always say definitely do measure but then look at a system you're in and look at where data is already collected and then be kind of pragmatic about what that means about where it is possible to measure. And so, yeah, I mean it'll be fantastic if at each different point you could you could measure and in many systems that will be possible because you just this is just stuff that is collected. So it's a great place to be doing this. In other areas, it'll be harder. And then it's just a case of, you know, a beginning to end type thing and seeing if it looks like there's impact. But even kind of small amounts of, you know, measurement, as long as it's reasonably robust to make a strong case for doing more of it or can tell you where things are not working. So when things need to change and that's an important, you know, part of this to be iterative and ready to to adapt.


Rhian Gladman [00:26:00] Yeah, I think that's very important to look at. You know, if we do a small nudge in one part of a very complex system, the rest of the system may kind of segment our nudge and negate the nudge, really. So I think it's looking at it in the whole system. And moving on because we have done these interventions in specific parts, haven't we? But actually, in the future can we as a like government sector start to look throughout our whole service, understand what's driving the behaviour across that whole service? It's not going to be easy, but I really think it's a good ambition to have to ensure that these interventions are actually complementing each other rather than maybe working against the system that they operate within.


Tim Pearse [00:26:43] And that's that's the opportunity, because if you can go upstream, then even it doesn't have to be huge holistic interventions that change behaviour. If you go upstream, smaller interventions like signing someone up for housing benefit, make an appointment with them at the citizens advice bureau, can actually set them on a different path. And again, that's one of the things I found really exciting about working in local government. And, you know, there's so much knowledge in the services about where demand is coming from. You know, that's already helping you understand where to start looking for potentially thinking about behavioural insights.


Rhian Gladman [00:27:19] I mean, there are some key points that have come out of our conversation for listeners already around, you know, start something, start small, and then also moving on to looking at interventions across the whole service, not to let measurement take over everything.


Rhian Gladman [00:27:36] They're really interesting points to take away from you there. But it would be useful to get your kind of your top tips really for a council listening to this thinking gosh, where do I start. I just need like four or five key top tips to leave listeners with so that they can actually start something, as you say.


Tim Pearse [00:27:55] Yeah, definitely say so. You know, as I mentioned, one of the nice things about local government is that so much of it is applicable and so many services and they all join up. So the first thing is sort of, you know, be picky. I'd say there are some areas that is going to be much easier to do this stuff in than others. And so think carefully about where where you start  and do that without alienating key people, of course.


Tim Pearse [00:28:23] So I'd say, you know, one of those could be, for example, like to what extent is it a priority as a council? Because you want to kind of the wind behind you. You want it to be in an area that people are going to be interested in. And if you do get a good result, that they'll be something that people pay attention to. So priority for the council, a whole would be would be great.


Tim Pearse [00:28:43] Think carefully about if it really is a kind of behavioural type problem. So think carefully. Is there an identifiable group of people and you know, there's a clear behaviour that it would be realistic to change. Those are some key things to consider. Sometimes it can be hard to reach people or hard to know which people are doing which things. But, you know, if you've got a clear target group and you know what you want to change about them, that would be fantastic. We've already talked about it but start testing quickly for a start. There's lots of evidence out there about what works, including on the LGA website. There's several things that have been done before and have worked. But you know, but beyond that as well. So already like changes to letters, you know, BIT and the LGA did a top tips on how to to write letters about about income. So just  start somewhere, test it quickly. Get going, show people the agileness of the approach in that it can get results. And finally. Yeah, do try and do some measurement. Do not let it rule everything, but showing that the impact of this is an important part of showing the value. And so and as I say, a lot of this data is collected already. So have that in the back of your mind as you're thinking. Where should I do this? How should I do this?


Rhian Gladman [00:30:07] Tim, that's fantastic. Thank you very much for your time today. Tim Pearse from the London Borough of Barking and Dagenham. Appreciate you coming in to share your thoughts and wish you all the best with your future projects.


Rhian Gladman [00:30:17] And we'll have to get you back along in six month's time or so to report back on how those projects have fared so that we can continue to share that with local government. Thank you very much. So as we mentioned there is further resources on the projects that Tim has mentioned today, particularly the Kent one - we've got the report there from that project. We've got the business card that you can download for free and use. So if you go to and search for behavioural insights on the search function of the home page there. All of our resources are there and they're all free to download and use. And if you enjoyed this podcast, please do pass it on to your colleagues and friends and spread the word. And again, if you would like to feature the work that your council is undertaking, we'd love to hear from you. So thanks for listening. And until next time, bye now.