Family Finance Surveys User Conference 2026 - Session 2

So, I'd like to just thank everybody for braving the heat wave and getting here. I have to say this is the coolest room I've been in for weeks. So, very um tempted to never leave it. Um I'm delighted to see you all here today. I I've lost one badge, but I've substituted another. I'm Debbie Price. I'm one of the deputy directors of the UK Data Service. Uh and I'm also a professor at the University of Manchester. I'm absolutely delighted to be invited by Jen to chair this session um which is on pension wealth in our national surveys and also about the people behind uh the statistics. I have a vested interest currently using WAS um in our own research on gender pension gaps have used many of these have a lot of feeling for the people who are trying to still work on national surveys. We have a lot of challenges at the moment but today's session uh very close to my heart is about issues of pension wealth and uh wealth and pension wealth and pension measurement. Um uh increasingly important of course to policy makers, researchers and analysts how we measure wealth, how we value pension entitlements and what difference those methodological choices that we make make to the conclusions we draw about inequality and retirement resources. Um we're very fortunate to have speakers today who have been very closely involved in those debates and I welcome them to the conference. So each presentation in this session is going to be 15 minutes. I have a five minute and one minute marker which I will use for the speakers. Um we have 20 minutes for each session. We're going to take questions immediately after each presentation. So your questions for that presentation, let's have them then when the 20 minutes is up uh we will move on. So, with all of that, it's my very great pleasure to introduce our first speaker, um, Pete Dangerfield from the ONS, who's also going to be joined by Joe Freestone and Tanzy Arthur. They're all going to come and sit up front in a second. Pete heads the wealth and pensions team at the ONS. He oversees the publication of statistical releases, user engagement, and the development of the wealth and pensions microlevel output. So, we're very uh fortunate to have him here today. Prior to ONS, he worked at Natsen. He worked at Ipsos Mori and he's worked on a wide range of social surveys including our favorite in the gerontolo gerontology sphere Elsa um the British social attitude survey and the millennium cohort study. So without further ado, Pete and colleagues, if you want to come up and um I'll sit there with my Hi everyone. Um so I'm going to be giving an update on the wealth and pensions um survey and then Joe and Tanzy um will be uh so they're our development leads on the survey on the survey side and the analysis side and they'll be giving an update on some of the recent development work. Okay. Um I'm going to we haven't got much time so I'm going to go through this quickly. Um but just a reminder what wealth and asset survey is um bianial longitudinal survey um collects information on financial wealth, pension wealth, property wealth, household debt and also physical wealth um and the attitudes and behaviors around that. Very importantly, why do we want to collect measures of wealth? Um just income and consumption are crucial for assessing living standards, but they can only be gauged um in conjunction with information on wealth. Um so what I'm going to do is give a bit of an update on the last year. Um so yeah we have to have to face up. It's been a challenging period behind household finance statistics. Um that's amid operational pressures. Um and also within was itself the suspension of accreditation of core wealth outputs. So that was something that was fairly fresh uh last year when we presented. Um but what we want to do now is show some of the um quality recovery and some of the stuff that we've done to maybe stabilize these outputs. Um and that's in line with our household finance um within our plan for household finance stats. Um yeah, as I say, we've made initial progress on um stabilizing fieldwork for the W. Um also progress against those OSR recommendations that we kind of linked to in this presentation. Um and then but it is longer transformation is needed. We recognize that. Um so that's when we've stabilized things. We know that there's much more that we can do with this data. It's really useful important important data. Um yeah and just to say we are we are going to be publishing more um very very shortly on the ons economic statistics plan and that will feature household finance statistics too. Um so yeah just a bit about the status of WAS at the moment. So where we are so we are processing round nine at the moment um that has been delayed um so that but work is ongoing and work has been go a lot of work has been going into stabilizing it. Round 10 has closed so that covers the period 24 to March 26. Um what we hope to do is work on round 10 in parallel some of the earlier stages once we get to the later stages of round nine. Um and then round 11 has gone into field um which is you know is a great achievement in itself and Tamzi is going to take you through some of the development work that's gone into that. Um very quickly on this I don't want to labor it. Um round nine um sample is about um 11,700. Um we've been working on um reproducing an analytical pipelines to push that data through and just thinking about that as a sustainable development for the future. So it will help round 10 and later rounds too. Um there are two work streams. So there's the al editing and validation work that's going on at the moment and then we're coming to a new stage which is um the preparation work where we pass that data through imputation and then the calculation analytical derived variables. Um we're doing a staggered approach. So what we're doing is we're taking test data and we're running through that first and then when we get to the 24 month data we're hoping that that we're hoping that we've covered most of the issues with the data when we see it and it it is it is still a long way off. There's still much more work to do. So we are looking at delivery winter 27 to 28. Um but as I say we are working in parallel on the improvements. We hope that will help the future. Um yeah, just to show so this is round 10 as well we're showing here. Um so we did achieve similar responding households for round 10 slightly more. Um now what's interesting about this slide I suppose is so yeah responding households have gone down but if you look at the amount of addresses we're also issuing then clearly we're not we're it's taking more issued addresses. It's taking more is there's more cost per interview. So that's some of the problems we're facing and I'd expect it's some of the problems that are being felt across the board with these stat with these surveys. Um just round 11 quickly we are hoping we are aiming to achieve the 18,000 divided by two rounds. Um we're hoping to reach a target much closer to those prepandemic levels. Um but it will require a higher issued sample size because of that low response rate. Um so yeah just on to the OSR review. So this is something that I showed last year. Um just some of the some of the um items that the OSR have asked us um have recommendations for the survey and the last year's basically been working towards these and um and ensuring that our outputs are are are meeting this this standard. Um so first a very uh release um an ask from the OSR was to publish comparability with other data sources. [gasps] Um so yes that that is published on the UKDS. Now um what it does is it explains some of the external data sources that can be compared to the WAS and how comparable they are. Um it looks across lots of the data sources that can be compared. So, administrative survey, economic sources, um some familiar names there. So, the FRS, Ash, LCF, um and then some administrative sources, too. Um we've also updated our quality information. So, this is something that is ONSwide as well. We're updating the way we put out our quality work and making it maybe clearer for the users. Um so, what we're doing is we're producing data sources guide. So that's the wealth and asset survey user guide um which was put out recently. Um focus there is the data source itself um but also changes to the survey um survey design etc. Um and then we also have a focus piece on the estimates that we put out as well. So that's called the quality and methods guide and we have one for total household wealth in Great Britain which is our flagship publication for the survey. So again the f that focus is the publication what the statistics cover and um the data sources feeding in. Um also um we want to improve our statistical the information on statistical quality too. So we have always put out quality indicators around our metrics but what we've done is given a much greater level of granularity with those. Now um I'd recommend picking up the slides and taking a look at them. There's some work that um Joe and his team have done which um is really excellent but it shows standard errors across the breakdowns and um then then confidence intervals around those estimates um by the um demographics you see there. Um and we are looking to integrate those much more into our work and much more into our outputs into our ad hocs and into our publications too. Um we've widened user engagement um including um the release of a user engagement plan. Um but we also have an expert group um which meets every six months and um runs through progress and data issu progress updates data issues but we also explore kind of methodologies and kind of anything that we need to and we need to look at for the upcoming rounds. Um we do have extraordinary meetings and technical sessions too. Um and just to flag the stakeholder engagement plan which was released last December um as as just part of part of meeting this requirement. Um now I've missed four and that's because Joe is going to be covering that in our latest release. Um but we are also reviewing that big one at the end reviewing the end toend resources um for the production of stats. Um so what we are doing is building out the team at the moment. Um we have these parallel development teams now um which you'll hear you'll hear from both of those teams. Um and we're also on boarding a delivery manager who can um bring together all the various people that have to be involved in the production of these stats. Um and yeah and as I say building out the production teams but I've mentioned this before we're thinking about the future and just making these this production much more efficient through pipelines etc. Um, okay. I'm gonna pass to Joe who can go through our latest release. Thank you, Pete. Um, so as you'll have seen on Pete's slides, one of the uh key OSR recommendations was to review our decision not to produce back series. So, uh, since last year, work has been uh started and is ongoing on that. Um so firstly just a bit of a uh kind of fundamental overview as to why we're producing this. So um for round eight um ONS updated the way that we calculate defined benefit pension wealth um this was applied to the figures for round eight um and uh put in short terms the change was going from market-based discount rates to scapebased discount rates um and I'll give a bit more detail on that in a moment um and basically this has a direct impact on our pension wealth figures and therefore a knock-on effect on our uh total wealth figures. So because round eight was using this new methodology, rounds prior to eight uh were not using this methodology, uh it meant that the roundto- round comparisons were basically no longer fully possible. Yeah. Um and uh so therefore um a uh back series solves this by applying a consistent methodology across the rounds. Uh so a little bit more on the methodology specifically. Um so why did we update this pension methodology uh sorry pension wealth is the largest component of total wealth. Um it's a very impactful uh part of that. Um so uh basically uh we get kind of a lot of feedback on that from our users and stakeholders. Um so basically the other thing so defined benefit pension value it's uh kind of incredibly uh complex to calculate um it's uh uh kind of more so than other assets that we measure in the wealth and asset survey and also other pension uh wealth measures as well. um because basically it's an estimate of the current value of promised income in retirement. Um so basically following stakeholder feedback on this uh ONS commissioned GAD to look into ways that we can go about um calculating uh defined benefit pension wealth. Um and we got that independent advice from the government acturies department. um we kind of guided how this uh consultation went um and ONS were kind of prioritizing stability um in the way that we approached the calculations for defined benefit pension wealth. Um so the reason we did this was because uh basically um the market base rates that we were using previously were volatile. We found it was harder to unpick uh like wealth change from uh market fluctuations especially where there were large swings and also from the pension holder perspective um we felt that uh kind of the uh more volatile marketbased rates didn't fully reflect the stable nature of defined benefit pensions and the way that they were designed. Um so basically off the back of this uh we went with the GAD recommendation to go with a scape rate um as our kind of discounting factor um which is the superanuation contributions adjusted for past experience um and uh that's a kind of longer term or stable measure of economic growth and we applied this to the um round eight figures first and have just released uh round seven figures. is um on the 7th of this month, a few days ago, and we're planning to release the remainder of the back series back to wave 1 um by autumn this year. So, conscious of time, so I'm not going to go through this in too much depth. You can access this. It's all released and out there, but that's just a slice of some of the results um from the round seven back series release. Um, and yeah, I've I've covered most of this anyway, but basically shows uh the process so far and what we're planning on doing. So, you see there the autumn 2026 date for uh publishing the remainder of the back series. So, I will pass on now to Tanzy who's going to talk about round 11 development work going on. >> Uh, hi everybody. Um, I think I've got about one minute, so I'm going to I've only got I think two maybe three slides. Um, and so this is the fun bit, I think, in terms of all the the the difficult trying to get the wealth and asset survey uh uh d quality data and stuff um back back into a good space. And so I was looking at um the uh development of the uh from from round 10. I've got now I got one minute. Uh from round 10 uh which finished in in the end of March this year and we had to from Septemberish onwards we were developing and improving the the survey for round 11 and uh I want to give a bit of a shout out to some of the people in the household resilience team. So Emily Farra who's here and Maria Quattry as well who were great at getting it all going actually before we um collaborated with them and then others in in my team. So it was a real combined effort and the people who are the survey manager, lots and lots of people involved in it. Um plus the stakeholders. So the stakeholders gave their views. Um we themed those we talked to the field staff to find out well what it's like it to collect the data. Are there ways that we can improve how we collect the data. Um so we fed that in as well. We uh put forward some suggested uh quest changes to questions and responses. We um then said yes, we have got time to do cog testing even though it's incredibly tight. Uh but we'll we'll do it for the quality benefits. We doubled the cognitive testing because there were so many changes that were being suggested. Um but we managed to get it all all done in the time. We then um from that we then pulled all those uh um the information together to make some recommendations and um say these are the findings and then we moved from from January this year we were into the build and test of um a lot of changes. Some of them were quite small but uh things like an interviewer prompt to help the the interviewer explain about uh pensions. Oh, I need to wrap up. Okay. So um but anyway some of them were much more much bigger and so show cards for example were all updated to help again in terms of the bandings for for costings and then quality assurance is happening now. Um so I've finished the time so I'll just leave this slide up for people but this is some of the topic areas that were updated and the other slide that I was going to cover which I won't now is uh we've started round 12 which is going to be from 28 onwards and uh so we're doing our stakeholder consultation on method and survey design and we're also reviewing the materials to try and raise response. Um so thank you very much. [applause] >> So next I'm very pleased uh indeed to welcome to the podium Isaac Celestra from the Institute of Fiscal Studies of course one of our most important institutions. Um Isaac's been an IFS economist since 2019 and his current work focuses on taxation, wealth, investment and high-income individuals. And he's going to be talking about one of our most complex methodological issues uh about which as you have just heard reasonable experts disagree. So this is the IFS view. Thanks very much. I >> Yeah, thank you very much. So good afternoon everyone. Um I'm going to be talking about how we measure wealth in W and more specifically about how we measure pension wealth. Um or as I'm going to be arguing how we mismeasure uh pension wealth in W. So let me just start with a slide to sort of try and convince you that what I'm telling you is important and that it matters. So this is just total wealth broken down by different components as currently measured uh by the ONS. You've got property wealth, financial wealth, um physical wealth, that's you know your sofas, your cars, your Picassos, whatever. And then finally, you've got pensions wealth at the top. And you can see that pretty much every year pensions wealth is currently measured is the single biggest source of wealth for households in the UK. So this really matters how you measure it. It's going to make first order differences to who you think is wealthy and how wealthy you think uh households are. Okay, so that's the big picture. What am I going to talk about today? Well, the vast majority of this is going to be me talking about how we currently measure pension wealth in Waz um and what I see as the fundamental flaws in how we're currently doing that in the ONS and why I think it's leading to uh statistics on the distribution of wealth that are just fundamentally misleading. Then I'm going to talk to you about uh a new meth a new method for valuing pensions wealth that I've developed along with a colleague uh at the IFS. I'll be showing you some of the estimates that that new method produces. Um, and then if we have time, I'll talk about some of the unresolved issues that are still outstanding in terms of pension wealth measurement that I don't think our new methodology does solve. Okay. So, I think the place to start is to ask the question, what is W fundamentally trying to measure? Um, and I think the pretty simple answer to that is that W is trying to measure wealth at its current market price. So this is the question from the uh WS questionnaire on housing. People are asked about how much do you expect to get for your current home if you sold it today. So pretty straightforward. Um and there's a pretty big advantage to doing things this way, right? It means you can compare the values of different assets um over different asset classes and then you can compare individuals who may have different portfolios of wealth. The question is are pensions in some way fundamentally different? And in one sense they are they don't have a market price. There's a legal prohibition on buying and selling pensions. Uh and so there's no price for us to observe. Um and I think that is a meaningful difference. I don't want to underplay that. And I think it essentially leaves you as a statistician with two options. So the first option is to say look these things are just too um non-comparable to other sorts of wealth. you just can't compare them to uh you know Ferraris and houses and whatever else and so we're just going to exclude them from the statistics altogether or measure them separately. Now some statist statistical agencies do do this internationally. The big downside obviously is that as I showed you on that first chart these are a really big part of household wealth in the UK. So if you emit them that's going to lead to a very different distribution of wealth. It's going to lead to very different answers as to who you think is wealthy in the UK. So the second option that's left open to you and this is sort of broadly in the spirit of what the ONS currently does is to say okay well what would these things be worth if you could buy and sell them or I think the equivalent question you can be asking in your mind is what can I buy in the market that replicates the exact benefits that a pension has and you know how much does that cost. Okay, so before I go into exactly how you might apply this to a pension, let me just step back and think about assets in general. How how do we determine asset values? How are asset values determined in the market? There are basically two factors that are going to be important. The first is the future stream of benefits that asset is going to provide. So that might be the future rent that a house is going to provide. It might be the future dividends a share is going to pay out. Um it might be the future benefits that a pension will pay out. And then the second thing that's going to matter is the market interest rate. And that's going to tell you how much is a pound in the future worth in today's terms. And I think the way to think about this, the question you're trying to answer with the interest rate is you're saying, okay, I'm going to get X in a number of years time. How much would I need to save today in order to get that amount in the future with certainty? And you can see that when interest rates are really low, I have to save more now because my savings are just going to grow more slowly over time. So a given amount in the future is just way more valuable. Whereas if interest rates are really high, getting that same amount in the future is going to be less costly to me today. I don't have to save as much because my savings are going to compound over time. They're going to increase in value. It's going to be less costly in today's terms for me to get a given amount. And it's really worth emphasizing that this process of interest rates determining asset values is something that is happening all the time in the marketplace. Right? Not only do banks explicitly do this when they value uh assets of different kinds, this is something that implicitly happens when you think about house prices, share prices, bond prices, constantly interest rates are shaping the price of assets um and their market values. So this is something that happens automatically with all the other kinds of assets that was or was as I know everyone else calls it uh is measuring. Okay. So pensions, how do we deal with pensions? Well, there are basically two main kinds of pension in the UK as I'm sure everyone knows. You've got DC pensions, defined contribution pensions. Those are the sort of pensions where it's just a pot of money, a pot of investments. Um, now a simple starting point to value to value those sorts of pensions would seem to be to say, okay, well, how much is in the pot? We'll value it at that amount. That's what the ONS does. Seems like a reasonable starting point, although if we have time, I actually think it's a lot more complicated than that. And I think there are some serious questions to be asked about whether that actually is the best way to do this. DB pensions, define benefit pensions, is where it gets even more complicated. So, as I'm sure everyone in this roomh knows only too well, these are the sorts of pensions that are related to your average or final salary. So, basically, you have two steps to valuing these things. The first is you got to work out how much they're going to pay out at retirement. And if you have enough information about the individual and about the scheme, this is basically just a matter of arithmetic. You can do this. The second is you then need to say, okay, well, it's going to pay out this benefit in the future. How much is that worth in today's terms? And that's the process of discounting, applying an interest rate to convert that amount um into into present value terms. Um now the ONS has changed the way it does this over time as we just heard. Um and it's changed in a way that means increasingly we rely um in these statistics on the scape rate to do this process of discounting. Um and this is really going to be sort of the villain of today's tale. Okay, the scape rate. So, what is the scape rate? Uh, it stands for uh superanuated contributions adjusted for past experience. If you have no idea what that means, uh, which I won't blame you. Um, the scape rate actually is just a measure of forecast um, UK GDP growth. That's what the scape rate is. It's literally taken from the OBR's forecast of how much GDP is going to grow by. Now remember the question we are trying to answer here is how much do I need to save today in order to get a given amount in the future. I think it is pretty much the only reasonable answer to that is a market interest rate. That is the only sensible answer to the question how much do I need to save today in order to get a given amount in the future. A GDP forecast is just fundamentally unrelated to this question. It's it's not just I think I really want to emphasize I don't think this is just like a matter of opinion. This is like a category error in my view. This is a completely wrong sort of object to be using >> in doing this process. [laughter] [gasps] Um and just to illustrate why it matters the yellow line here is the scape rate. The green line is the 15-year guilt yield. This is just one measure of market interest rates. And you can see that over this period between 2008 and 2022 when we had a huge decline in market interest rates, you don't get any of that in the scape rate. So over this period where declining interest rates are, as was very well covered at the time, are pumping up asset values all across the economy. They're driving up the price of houses, they're driving up the price of shares, they're driving up the price of bonds. the way in which we measure pension wealth is just going to be sort of you know blissfully going along without taking that into account at all. And so it means that you're going to have a huge skew in how you me in who you think is wealthy and how you measure wealth, how you measure the distribution of wealth. You're going to have people whose wealth is concentrated in pensions completely unaffected by this big change in interest rates that's affecting the wealth of all other kinds of assets. Now hopefully I'll have convinced you by now that market interest rates are uh the way forward and that that's the correct object that you should be using in order to discount uh future pension wealth or future pension income rather. There is then a subsidiary question as to which market interest rate should we be using. Obviously there are many rates of return uh that are available in the market. there are you know rates the rate of return to stocks is not the same as that as corporate bonds or you know bitcoin or whatever what should we be using um basically the reason or the main reason why rates of return differ in the marketplace is risk so higher risk uh propositions have to command a higher rate of return because people don't like to take risk um and so what you want to do is you want to match the discount rate you're using to the risk profile of the asset you're discounting. Now, pensions are about as low risk as it gets, or at least DB pensions are. The amount that you've been promised is the amount you're getting absolutely no matter what. Even if the if the scheme defaults, you're 90% protected. And they don't default that often. Um, so in the methodology that I'm proposing, we use the UK guilt rate um as a proxy for essentially a risk-free interest rate. Yeah, I've seen it. [laughter] Um, so just to give you an outline of the of the of the method that we've developed, we published a paper on this fairly recently, myself and a colleague at at IFS. Um, the big difference to how the ONS measures these things is that we're using this guilt rate to do the scouting, the market guilt rate. We do some fancy stuff in matching maturities, but I won't I won't bore you with that now. We can talk about it later if you're interested. Um, we also apply our method consistently over time as as the ONS is now planning to do uh with their method and we also do some other stuff with joint life annuities which I won't bore you with. Let me just show you a few pictures as to how much of a difference it makes. So this is scape rate discounting the purple line and then the green line is guilt rate discounting our methodology. In the most recent wave the mean wealth differs by almost a quarter of a million pounds mean household wealth. And not only is the level very different, the trend is very different. Now, it matters not only for the aggregate amount of wealth, it also matters a lot for the distribution of wealth, for who we think is wealthy. So, there's a lot going on in this chart. You don't need to worry about most of it. What I'm showing you here is how our method differs from the ONS method in the most recent uh wave based on age. And the purple line is giving you the percentage change between the two methods. So what you can see is that our method increases wealth most for people who are youngest. Why is that? Well, if you're a long long way from retirement, the discount rate matters a lot because it's going to compound over many many years. And so when you measure the intergenerational wealth gap using our method, it looks a lot smaller than it does using the ONS method. So again, this is a pretty important statistic, a pretty important insight you might want to take from this data. And it's fundamentally different um when you use market interest rates. Again, maybe not the clearest chart in the world, but what we also find is that conversely, if you look at the wealth gap by education, it looks a lot bigger. So the green line at the top, our method gives uh a median uh wealth for the highest educated households about 50% higher than the ONS method. for mid uh mid-education households, it's only about 30% and it's almost zero for the lowest education households because DB pensions are much more concentrated amongst people who have high levels of education. So again, another pretty fundamental fact about how we think wealth is distributed that looks very very different. Um, so these are pretty first order changes to what we might think about uh the distribution of wealth in the UK just from essentially from this one incredibly important judgment that's being made uh in developing these statistics. Now um there are some additional this is actually only a limited list but there are some additional questions about uh how you might want to measure pensions that I probably don't have time to get into but have to take questions on them. I don't think our method is, you know, the final word on this, but I think it is a substantial improvement on probably the most important uh judgment to be made when uh undertaking the value uh of these pensions. I'll stop there and and leave time for some questions. [applause] Brilliant. And I'm early, so you get all the prizes. Um yes. Okay. All right. Is this on? >> Yeah. >> Hello. Uh in your graph showing uh real return on guilts, what deflator were you using to convert from >> uh so we actually we actually used uh index linked guilts and we we backed out the the real return using those. >> Well index linked to what to RPI or to CPI >> to to RPI and then we applied the wedge between RPI and CPI and then we made also made the adjustment the adjustment at the point where RPI and CPI become the same thing. So you have to do it differently there. But but yeah, it's it's it's it's CPI adjusted but using the RPI link guilt price. >> And I imagine that details in the p in your paper. >> Yeah. Yeah. All all there's a long appendex on all this on all this stuff. >> Yes. Yeah. This may be a naive question, but surely um the uh the the the effective value for an individual of a defined benefit pension depends a great deal on their career path and whether they get promoted to professor or remain you know or whether they actually bail out of academia halfway through their career etc. So really their risk premium is way over zero. >> Yeah. So that's an interesting point. So the way in which DB pensions are currently measured in W is they take your currently acred rights. They're not doing anything in terms of saying, "Oh, we're going to forecast your future earnings and work out how much you're going to occur in the future." I think that probably is right if you think about how it's consistent with other kinds of assets, right? We're not going to take into account uh the expected growth in your share portfolio or you know things that are going to expected things that are going to happen in the future with other sorts of assets. So I think actually probably the right judgment's been made there in terms of saying we're going to take the rights you've acred up to this point. If you left your job now this is what you would have basically. [clears throat] >> Yeah. I mean lots of parallels actually. Another little world I sometimes in well often inhabit is valuing pensions on divorce >> and on divorce. Absolutely. Yeah. It is it is a similar question of parallels but actually it's appro I mean in the divorce world they're much more concerned with getting a consistent estimate than revaluing pensions in the January March May and December of a particular year when you might get these fluctuations. So it's it's just a question about that payoff between something that looks more stable over time and something that could really depend on the actual date of the interviewer visit. I think there is yeah I think there's a question on so the stability question yes that's true but I doesn't seem to be that people are as concerned about stability for stock portfolios for house prices for many of these things are more volatile than uh than might be shown in the in the guilt rate right um there's also you could still use a market interest rate but smooth it in some way if you're really really worried about um if you're really really worried about day-to-day volatility so I don't think those are uh objections that hold a lot of Any questions online? Please proceed. I [laughter] I'll ask you a final question. >> No, absolutely. >> Obviously the the government acturies department has given different advice. >> Absolutely. >> Um and I know this debate has been going on for quite a long time between IFS and and the various government. I mean do you do you understand their position at all? Can you see where they come from at all or um >> truly not Isaac? Well, so I well I think there's a few points on that. I think it's worth noting that as Pete said the task that was given to the government actuarial department clearly stated that stability was a key criteria >> and I think that actually is sort of in some ways the root of this issue right that fundamentally the value of these things isn't stable and if you try and impose price stability it's like saying well we're going to just say that your house is kind of worth what it was in 2008 and we're going to upgrade it by 2% each year and we're going to we're just going to ignore the fact that house prices have changed a lot. Um, so I think wanting to impose stability on something that just fundamentally isn't stable is going to lead you to wrong answers. I think >> interesting because I think that's how our council tax works, right? [laughter] >> That's a that's a whole another topic. >> We have one we'll take one question from the um Q&A and then we'll wrap up. Thanks. a question um on mimeter that says sorry really a dumb question are the 30 39 year old group 40% wealthier now or at their retirement >> sorry I didn't fully take that in I apologize >> are the 30 to 39 year old group >> 40% wealthier now or at their retirement meant. >> Oh, uh, you mean that chart I showed where I was showing the differences? No, it's now >> now. >> Yeah. >> Thank you very much. Really fascinating talk, fascinating debate. It'll rumble on, I'm sure. Um, can we all thank Isaac in the >> Thank you so much. [applause and cheering] >> Um, and uh I'm I'm very pleased to introduce um Ariana. Just say uh Ariana, we will give you the full 15 minutes and five minutes for questions. Don't worry about it. We'll sneak five minutes from the lunch. Um she's going to take us to the real world behind the statistics. Ariana is a senior lecturer in economics at Goldsmith's University, University of London. Um her fascinating research uh which I've seen presented a few times and can highly recommend her book rethinking financial behavior um looks at how people respond to the growing pressure to manage financial risk. Um she's an associate member of Aston's Center for Personal Financial Wealth and she also is a steering member of the deecon network. Thank you very much Ariana. [applause] Thank you very much for the very nice introduction and for having the possibility to show a little bit how I use the bus for framing interviews. So my main focus is qualitative research but the bus was quite helpful to use with the interviewees but also to uh position the interviews overall. So since the time is quite limited I'm just going to tap in and out different aspects of the key insights from the book to show how I approached it. In the beginning just very briefly I know I don't need to talk much about probably here in this environment about the kind of pension system we have in the UK. So basically that we have a three- tiered pension system where the state pension basically just provides a very basic provision is not sufficient for you to uh sustain yourself in the future. Whereas then the main focus for everyone nowadays is to have workplace pensions actively engage with these workplace pensions and also have private pension investments. So from a theoretical perspective that's basically has been termed as the everyday investor which has been constructed in the US and UK particularly in this case where we expect the everyday person you and me to actively engage with financial asset portfolios and also use um financial strategies such as diversification hatching etc. So just very briefly from um from a theoretical viewpoint where this research is coming from which also intersects though with the policy environment. So the everyday investor subject I just very briefly introduced before that would expect with workplace pension that most people don't just sign up and use the default uh pension option but actually look at okay what is the percentage I would need for my future retirement income and also use the investments expected. So now it was increasingly realized people don't necessarily behave as expected. So here come in different literature strands which have been trying to explain this phenomenon but also try to provide policy solutions. So you have for instance behavior economics which was one of the key factors why we have now automatic enrollment based on the concept of nut that we are less likely to opt out than to opt in. In this case, financial literacy, as I'm sure you are aware, financial education is a huge factor now where people are caught up on to educate themselves, get to know different um financial concepts. And then the last part, more recently, the financial capability discussion. That's not just about receiving the education, but we also need to apply it, practice it, and have access to it. The the other element is that it was more and more broken down onto different contexts. So recognizing that different regions might need different financial education, different situations you are in uh require different financial education. So we just had the example with uh getting divorced, you need different financial knowledge in this case than for instance if you're just starting out to work. All of those elements though however focus on individual behaviors. So changing the behavior the individual needs to make without looking necessarily as the system itself. So that's where my research basically comes in to show how systemic constraints which are inherent in the system impact choices people make in their pensions decisions overall and might actually explain some of the elements which are often seen as financially irresponsible and actually show that it's in the context they experience it's a responsible financial decision. So that's where the research comes in and I know it's very just very brief with the literature but so I conducted um 50 uh 63 interviews with 58 households. So sometimes I interviewed uh members of the household together. Here there was a diverse range of different interviewed participants. The original focus was medium to high income households. So those who could actually to a degree conform to those norms of accumulating asset but that also meant individual based that here uh the interview participants sometimes earned way less than median income or high income in this case and it was where uh conducted in two waves. So basically I conducted the first half of the interviews looked at the percentage compared to the UK population overall and then the second uh wave I made sure to have it aligned with the UK population concerning self-employment ethnic background etc. So to have also correlation to the the UK environment what I've done now let's uh what I'm going to do go into now what I've done with the interviews and it kind of nicely connects now to the discussion what we had I asked them to tell me what kind of asset and liabilities they have. So it was not just interview questions but I talked through them through the different asset categories explained different elements and then asked them I showed them let me just try to go there I showed them in a very simplified form not that detailed but I showed them a very simplified form what a household balance sheets could look like what kind of assets can be included what kind of liabilities and then what I've done is either talked through with them the balance sheet overall or they send it to me afterwards. So 45 of the interviewees followed the request to provide their balance sheet or to talk through with me and I use that then actually to position this in correlation as well with the wealth and asset survey. So I used the wealth and asset survey over from the time period when I conducted the interviews and established an overall overview of household distribution, the kind of assets based on how many households own an asset and based on the values and I set up a similar overview then with the aggregate data from the interviews. So here this enabled me to actually show that the interviewees I had how they correlate to the worth and asset survey how they connect with the overall UK environment and I dis I disentangle them that as well with the income category. So it was really helpful not only for the interview side to see how interviewees engage with different terms in this case. So sometimes they talked about elements but they were not quite sure but when you actually explain the asset categories explain different elements they dive deeper into the different concepts. So it helps in the explanation but also to position the different balance sheets to the wider development in the UK. So just to show you so that's one of the examples of the house of balance sheets which was provided by one of the interview participants. So the basic elements what I've done then as well is I looked at the development overall in the UK since uh the 1980s how assets have developed overall and then since the first wave of the bus how the assets might have changed and have developed and the composition of assets. So it's clear in this case that the asset composition and how they can conform to the investor subject or not depends a lot on income. So the high income desides in the figure on the top you can see they're the most likely to be able to conform. So they have a more diversified portfolio. They have more financial wealth and don't just predominantly rely on property and pension wealth. Whereas the median income households rely predominantly on property and pension wealth and the lower income is then uh reliant on physical wealth. So in combination then with the interview data what I was able to do is identify why people make decisions concerning financial investment concerning their pensions in this case and one category which I termed here was the everyday asset manager and they basically reject some elements from the investor subject based on their income position but by also based on the recognition that financial institutions can be very uncertain regulations can change and the trust issue in that that case but also financial investment uncertainty overall. And you can see here there nearly all of them recognize okay the welfare state will not provide for you anymore. You need to provide for yourself. There's rising job insecurity which prevents you actually in some cases to provide and invest as expected but they feel forced that they have to do it. So it was an interesting theme which came out here in this case that it was not about most of them that they enjoyed investing besides the high income interviews but actually they felt the need to do it because there's no security there anymore. They rejected predominantly stocks and shares investments. So the investor subject itself and the discussion a lot of times which you can see now also in the policy environment introducing uh incentives or nudges for people to invest more in stocks and shares. This was seen as highly risky and as something okay for them they cannot risk so much of their livelihood of the financial security to invest in stocks and shares because they go up and down. So they were very skeptical of stocks and shares investment and for this and the other element was that they recognized that changes regulatory changes are too many. So what basically then happened is they emphasize three different levels. So they emphasize okay first I have savings if I saved enough for the deposit I'm going to step onto the property letter and then only then I will think about more actively about the pension investments. So how to deal with pensions if I need something at some of them even uh the ones younger ones who were on shared home ownership in this case already thought about or started thinking about as well what other private pension investment they can do and here it was clear that those are the three elements they mainly focus on this so that was the everyday asset manager that was the main subject position they they took on the majority of the interview participants But because I interviewed also different income levels and different backgrounds etc. What I've done then is actually looked at what other assets are integrated and how they they engage with it, what discourses they're employing and here I identified five different position and that had a lot to do what kind of constraints they experience in the system. So where are they self-employed where they're foremployed? So here self-employment you can see for instance that's by a person who was self-employed and she said okay she cannot even in a personal pension contribute money regularly and then not be able to access this until they retire because of the kind of condition they have with self-employment. They earn money in lump sums. They have insecurities in in between if the job is not uh or they don't get as many jobs as expected. So therefore pensions are not an option and they chose deliberately here cash is which were more accessible for them in this case. So it depended a lot on their background and there are different levels to it. As I mentioned already before just very briefly the calculative investor is the one which is closest to the everyday investor identity. So those are the ones really high income. They have a relatively good level of wealth. They enjoy investing. They enjoy speculating. They have a very competitive understanding and the interesting part with those none of these highincome households wanted to provide the exact numbers. It was mainly based on as you can see in the balance sheet there. They provided for instance the kind of investment they had but they didn't share the asset and liabilities. The nonasset manager were those who rejected the whole idea of engaging with assets. But interesting enough there were often those which still had access to DB pensions. So they had a form of security still in place where implicitly you could argue they don't need to engage as much as others with the the concept. The last part because of time issues um what I wanted to discuss is that what happens if you actually experience constraints. So here it goes it goes into the case what happens if you have carrying duties which are not sufficiently recognized in the current pension system. So here workplace pension are for instance paid for someone in one time of employment and earning at least £10,000. They are if you're taking a break you don't have access to them. If you're coming back and have multiple employment you don't have access to them. So that often affects actually mothers and carers because of the kind of environment we have in the UK where caring costs are still relatively high compared to other countries. A lot of women still take break once they have children and as well then come back part-time or multiple employments etc. And you can see that clearly in the interviews they mentioned so many times that carrying duties prevented them to be able to save and invest as expected by these asset norms in this case. And some even mention okay the children are getting uh bigger now so now they can see the end uh of the the light at the end of the tunnel. And this is just one example of a balance sheet there. They also then the interesting part is because often in financial education literature and also policy discussion it's focused on individual measures that we need to educate them more that they need to do more. But what you can see here they do know that they need to plan and invest for the future. they want to have something security for themselves as well and then they adjust those measures in line with what they're able to do with their work trajectory basically. So here they try to find flexible accessible savings and investment means in this case something which is more tangible which they can connect to overall. So here for instance cash ISIS was a main means for them to actually save for the future because if something happened they could still access this overall premium bonds felt more safe again accessibility flexibility property was seen especially in the case that it was connected it's your security it's your home and there were many cases where participants for instance had were able to buy the council house had shared home ownership etc. and they saw this as a means for them to prepare for the future. So within the constraints they experienced they actually tried to manage as best as possible and the younger ones interesting enough they recognized that in the future it will happen that they will have less possibility to save and invest and then try to save as much as possible before they have children or increase uh contributions overall. So the other ones uh if you're interested in because of time I can discuss in the questions or uh later on but the main aspect what I wanted to showcase here is that a lot with the wealth and assets first of all was I was able to connect it to the balance sheets with the households and then also see how they engage with assets and then in connection with the context what it shows that less that it should be less about just individual measures But we need to rethink how the system is constructed if it is constructed in a way which recognizes different life trajectories. It clearly is not for carers and someone who has caring responsibilities and that these people do know that they need to save and invest and try to find ways within the context they're finding themselves in. So it's time to rethink as well how the system is constructed. So yeah, I think that's my time up for now. Thank you very much. >> [applause]