 Welcome everyone to the Family Finance User Conference. I'm Gem Buckley from the UK Data Service. So we'll make a start. So welcome everyone to the 2023 Family Finance User Conference, introducing our first session. So in our session today, we have presentations from the ONS. We're going to have all of the presentations together and then go for questions at the end. But please be free to add your questions at any point into the chat as we go along and we'll pick them all up as we go at the time for questions. And we'll start with a presentation on household finance statistics at the ONS. Statistics for Public Good by John Morris. So John Morris is a divisional director at the Office for National Statistics, leading the production and analysis of household finance statistics, along with other topics such as crime, scales and time use. Prior to joining the Office for National Statistics in 2019, John was the chief statistician at the Ministry of Justice and previously held analytical and operational roles in the Department for Work and Pensions and the Home Office. And with a passion for delivering statistics for the public good, John has increased the prominence and scope of household finance statistics at the Office for National Statistics. So if I invite John to start his presentation. Morning and thanks, Jen, for the introduction. You make me sound a lot more relevant than I really am to this topic, but it's great to see so many here today on a subject that is very close to my heart, but also a very welcome point in our history with the cost of living crisis to be here to discuss a topic that really matters for everybody. And if I could ask Daniel if he could move on one slide, that would be great. As Jen said, we'll go through all of the ONS presentations and then questions at the end. But if you do want to drop a question into the comment or the chat bar, please go ahead with that. I'll try not to look at it as people who know me know I will talk forever and I've only got 10 minutes to get through everything, everything that we need to say. So the first slide here is very much taking you through what we're trying to achieve as an office. So about the relevance of the information that we provide to users, whether it be policymakers, whether it be the general public or analysts like yourself. So we do a hell of a lot here and we try and be as responsive as possible. And earlier this year, we held an event on measuring the cost of living and how we can add to the public debate. So this is just a quick slide showing you some of the things that we have managed to produce just in the last year and how it's been taken by press and other people, commentators in the media. So it just shows you the range of things we get up to. Daniel, if you could just move on another one, that'd be great. So what have we been doing? Many of you know that O&S do a lot in this space and we produce statistics on spending, wealth and income and have done for a number of years. I mean, this is that provides a vital source of information for policymakers with our wealth and assets survey being, as an example, the only measurement of wealth in the UK. And it's something that we're very proud of and something that we really want to make sure users get to know and make best use of. But we also cover a lot of other bits around the survey of living conditions and the living cost and food survey. And we work very closely with other people on today's session, but DWP, the Joseph Fran Tree Foundation, making sure that we can answer the questions that everybody has in this space. The statistics we produce tell people about how wealth, income and spending are distributed across the UK and provides a vital set of data and information that is out there and updated regularly. Ideally, we would like to be updating these even more regularly, but we will come on to that in a couple of slides time. So what else have we been doing? So if we move on again, Daniel, on to the next slide. So you're going to hear from three different teams with the ONS looking at some of the work that we've already done to date. So Lee will be talking to you about the work that we've done on incomes and also on family spending, but we'll also be talking to some of the developments that we've been doing on understanding how administrative data could be used in providing more timely data on incomes. So there's a lot going on in this space. If you go on again, Daniel, I'm desperate to keep the time where I can. Oh, no, we've gone one too far. So if you go back one, sorry, Daniel, that's me just going ahead on myself. So if we look ahead to what we're doing this year, we're hoping to publish even more information on some of the survey changes that we've recently made, particularly with regard to pensions. We're looking to improve the pipelines on how we produce the data and the methods that we are employing. So at the moment, we are looking at some of the methods that we use to look at defined benefit pension schemes and how to make sure that they are still relevant and are the best way of doing things. So we've employed the government's actually department to help us on this to make sure that we get the right expertise in this, but we're also willing to hear and not just willing, but really want to hear from users and how they think we can make best use of the methods we have or any changes that they think are important. But we never stop. So we have a set of regular publications on the redistributed role of taxes and benefits and our normal effects of taxes and benefits release will be coming out later this year, actually later this month. And we've also got a lot of analysis around the cost of living crisis that we're currently living in. We're also looking at, as the slide says, on updating our measures of system poverty through the EU Silk Measure, which is a commitment we made earlier in the year to get that out and as up to date as we could, which will take us to 2019, but we're also looking at how we can improve that and how we can improve the poverty measures across the piece with our colleagues in DWP and beyond. We're also looking to update our small area income estimate. So again, this is something that we know users are very keen on and something that we are very keen on updating and make sure that it's relevant to people and can be used more regularly. We're also looking at how we can be more responsive to some of the key topics of the day. And as the slide says, we're trying to update our financial resilience measures. So I mean, this is part of the wellbeing side of things and how people can cope with sudden shocks to their lifestyles or also how they can carry on affording to live as they really need to. And as part of our regular releases of things, our wealth and Great Britain release will be going out in early next year and that will take us up to date to 2022, which again is something I think is really impressive in that we've managed at a time when survey responses aren't as good as we would like them to be, that we can still get out that important information to our users in time for them to make use of it. And part of the other bits that we're trying to do is make sure that we are relevant to what stakeholders and users need and we're setting up a new set of expert groups. So please get in touch with us if you would like to be part of this and you can help shape the work that we take forward. And finally, but not least, is that we really like to thank everybody who many of you on this call who responded to our household financial statistics consultation and provide us a lot of feedback on how we can make improvements and how we can take things forward. We're still working through those very detailed comments. So please bear with us, they will be coming out in the coming weeks. So, and we can take it from there as questions once they're out, but please bear with us while we get through all of these. As I said, there's been a lot of feedback from that and a lot to get through. I will leave it there for now, but please stay on the call throughout the day and I really hope you enjoy today's conference. I'm really excited a lot of the topics and I will stay as much as I can through the day, but I do have a few other sessions I have to go to, but I will definitely be here throughout this one. So please put your questions into the chat bar or raise your hand when it comes to that point at the end of the presentations. So I will leave it there for now and I will hand on to the next of our presenters. So if we go on the next slide, Daniel, that'd be great. So Beth and Danielle, over to you, I think. So yes, next we've got Danielle who's going to be presenting with Beth Jones on financial mental models research. So Beth is a senior social research methodologist at the ONS specialising in qualitative research, cognitive testing and survey design. And she works in the qualitative data collection, methods expert group and has led research on a variety of topics and surveys, including the census, admin data, health data, COVID, finance data and the crime survey. And she's going to be presenting with Danielle Watson who's an assistant methodologist within the methodology and quality division. And she joined the ONS in 2021 and is a member of the qualitative and data collection methods expert group within the social statistics hub at the ONS. So over to you. Wonderful, thanks very much. Yeah, so Danielle and I are going to take you through a recent qualitative project that we worked on. So you can move to the next slide. Thank you. So this project looked to explore how household finances are understood, conceptualised and information about it reported by members of the public. So we looked specifically at the topics of income, wealth and expenditure with the aim of using this information to inform how we might improve the quality of financial data collected by ONS. And our research used what's called a mental model approach. Now a mental model is a concept, a framework, a worldview that a person holds in their mind about a particular topic. They're internal representations of kind of an external reality and people use mental models to interact with the world, to filter and store new information. And more importantly for us to influence their choices or behaviour. And the construction of mental representations is really based on kind of unique life experiences and perceptions. And in the context of our research, a mental model is an explanation of a participant's thought and answer in process. And mental models are really important for us in expert understanding, where usually the basis for development of policies and strategies may contain, you know, kind of inaccurate assumptions about what the public knows or expects about different topics. And especially when it comes to issues of trust. So you can move on to the next slide then I thank you. So to recruit for this research, we initially explored the use of a recruitment company when we use regularly internally within ONS. But it wasn't really feasible with the deadline or budget for this particular project. So instead we advertised the research internally, asking staff and colleagues to circulate the research with their friends and family and they thought would be applicable for the research. And this internal call out was open to anyone over the age of 18 that was living in England and Wales. And we provided a 50 pound voucher as a thank you for helping with the research. The only restriction for the research was that we weren't able to accept ONS staff as volunteers due to the risk of bias. But we had lots and lots of ONS staff recommending friends and family, passing the message on, which was really great. We had such a huge response from the call out. We ended up with in excess of kind of 200 suggested volunteer emails in our inbox and we contacted these potential volunteers and purposefully sampled from those that replied expressing interest and those that completed, excuse me, our screening questions. And we were looking to make sure we had a diverse range of characteristics from age, sex, ethnicity, employment status, qualifications, household income and household composition. We ultimately recruited 31 participants with 27 attended interview, a small number of participants scheduled for interview. Unfortunately, we're no longer able to attend for a variety of reasons. And due to a couple of these reasons we were unable to recruit anybody classified as unemployed. And the reason for this was a mix of non-response and like I said, not being able to attend interview. If you could move on again, thank you. Okay, wonderful. So to meet our deadline, we opted for a two-part rapid qualitative analysis approach instead of a usual full end-to-anthematic analysis. Our interviews with members of the public were attended by a lead interviewer and an observer with the observer recording quite detailed notes about the participant's behavior, their thought processes, their circumstances, what they were indicating about their mental models and how that then translated into their answers to wealth income and expenditure questions. All the data was anonymized and all the identifiable information was removed from the observation sheets and the interviewers were then transcribed by a third-party organization called Language Empire. So for the first part of the analysis we conducted a high-level thematic analysis on these observation sheets. This was combined with listening to the audio recordings where necessary to provide us with context of some of the comments and this was done while we waited for the transcriptions to be returned. And once we had received the transcripts from the third-party, we conducted a second part of the analysis using our own more in-depth analysis process to ensure we hadn't missed any kind of important themes or lost any of that context, excuse me, and also allowing us to extract supporting quotes and any evidence that we might want. And then a final report was delivered outlining these detail findings and considerations. And just before I pass over to Danielle, just a note that qualitative analysis and the goal of that is to provide insights by presenting a range of responses, perspectives, experiences, so Danielle's going to take you through some of the key findings, but without quantifying anything. So I'll pass over to Danielle. Great, thanks Beth. So if we could go to the next slide then please. So I'm going to be walking you through our findings now. I'm going to start with income. So participants overwhelmingly associated income with money received from an employer or for work that they've done. Other less common examples included pensions, benefits, dividends, selling items online, mainly on an ad hoc basis, investments, or buy telepropathy. These weren't always offered spontaneously though. They were often confirmed when the interviewer probed sources of income other than wages or salary, or they were sometimes later recorded in the interviews because they were second sources rather than a mainstream of income. Understanding of grace and net income varied. Some were confident that grace is before tax and next is after, but others were less confident mixing up the two or just didn't know at all. Checking bank statements online via banking apps were a popular method of monitoring income. The level of monitoring varied though, some daily, some once a week, some once a month to no monitoring at all. Other methods of monitoring included checking pay slips, keeping notes on paper or on their phone, as well as relying on someone else, such as a spouse or partner to monitor. Where income was the same each month, participants thought that it would be fairly easy to provide this information in a survey. They suggested that they could give reasonably accurate and estimated figures and more accurate sources and figures could be sourced from bank statements or pay slips if needed. Participants whose monthly income varied felt that it would be more difficult to provide the information because they often work different hours, so naturally that income fluctuates with this. And if we could have the next slide, please. Thank you. So participants were generally willing to provide their own income and information in a government survey run by ONS because they trust the data would be handled appropriately and securely and that it'd be used to benefit society. Probably worth noting here though that this level of willingness could be a byproduct of our sampling method because our participants had some level of association with ONS already because we advertised internally, it would be worth exploring this further in a wider population. Willingness was often associated with how easy or difficult it would be, so participants were willing if they thought it would be easy and less willing if they thought it would be more difficult. Having nothing to hide was also mentioned here as a reason for being willing to participate. Method of contact appeared to affect willingness too with participants stating that they'd be more likely to participate if they'd been invited by a letter rather than by a telephone. This is because it was easier to judge the authenticity of a letter and they could contact ONS directly themselves rather than having someone cold calling them. Assurances around data protection were also an important factor here too. When asked about providing information for others in the household, those in relationships were willing to answer for their spouse or partner as long as they had permission. Where findings were kept mostly separate, it was suggested that we should be asking members of the households individually rather than assuming that there is a head of a household as this is quite outdated. In unrelated households, so those living with housemates or friends, willingness to answer for others was low because it would be more difficult for participants to answer on behalf of someone else. They also felt that finances are a sensitive topic. Many would consider income private information and would be uncomfortable asking housemates how much they earn or how much money they have. Many in house shares felt that although they live with these people, they didn't really have a relationship with them and wouldn't call them friends. We heard things like, I don't know them, I don't really speak to them, I just live with them. So that means a lot of these questions would be inappropriate. Okay, next slide please. Thank you. So looking at findings for spending now then, participants found spending very easy to define. We heard phrases such as using money to buy things, paying for goods and services, outgoings, expenses, and the amount of income that you use. Participants recalled things they spend money on by thinking about their regular direct debits or by thinking about life events such as birthdays or Christmas. Feelings associated with spending such as guilt about certain purchases, like having their nails done, was also mentioned as a reason for remembering spending too, particularly for those with co-dependence and children. Again, monitoring varied here from not really at all to regularly checking bank apps and checking how much money had been spent and how much was left. App notifications were mentioned as being really useful here for monitoring spending because every time money leaves their account, they receive a notification on their phone. Other participants kept more detailed spending notes though, so this was on their phone or in an Excel spreadsheet that they update regularly. Some felt that categories of spending would help prompt their money, but others felt that it would require a lot of work on their part to calculate these separate figures for each category. Cash and ad hoc spending, such as pop into the shop for a snack, were mentioned as being more difficult to track and recall because people don't generally take their receipts and these purchases are often small and not as regular. Spending bands wouldn't necessarily help here. We'd be effectively just putting in an extra step for the participant to answering the question and it wouldn't necessarily make the answers any more accurate. Okay, next slide, please. So proxy responding would be quite easy for joint expenses, but more difficult for personal spending because participants were not always aware of everything their partner spends their money on. Those in a relationship were generally willing to ask their partners about their spending and provide the information on their behalf, again if they have permission, but separate surveys were suggested for more accuracy. It would be very difficult for participants in unrelated household to answer by proxy because they'd have very little idea about how their housemates spend their money. Some said they could guess on what they could see, such as deliveries to the house, but they wouldn't be confident. Some wouldn't be willing to ask for this information, again, because they just didn't have much communication with these people on a day-to-day basis anyway. And they felt that spending and money in general was a private subject and quite a taboo subject. We also found that some were reluctant to share spending information in a survey due to distrust of government. Okay, next slide, please. Thank you. So looking at findings for wealth now then, we found that wealth was a lot harder for participants to define compared to income or spending. It was very much associated with having money and possessions or being in good health, as well as relationships and life purpose. There were also participants who struggled to define wealth at all, though. The term assets was largely related to possessions, so some of the examples people gave included property, cars, musical instruments, collections, valuable items, and even a private jet. We also had examples of financial assets, such as pensions, investments, and savings. We found that wealth was generally monitored less closely than income and spending. Participants might occasionally check local house prices more out of interest and sometimes check their savings and investments, but they didn't tend to do this as frequently as checking their income or spending. Ability to provide wealth information varied depending on the type of wealth. So for example, the value of property could be estimated by checking current house prices online and the value of savings could be sourced from bank statements relatively easily. But for a total figure of wealth, this would be difficult. So it'd probably just be rough estimates. And for some, even this would be difficult. When we suggested banded response options instead, opinions varied between thinking that this would help and make it easier to not change in the difficulty at all because they wouldn't make the estimates any more accurate. It was suggested that different types of wealth would be easier to value rather than providing one total figure. So willingness to provide wealth information in a survey range from unwilling to very willing. Reasons for being willing to participate, including trust in OAS, having nothing to hide, already providing such information to HMRC and wanting to help inform economic policy. Some suggested their willingness would increase if given advanced explanation of why the government needed this wealth information and what the survey would be actually useful. Some were willing but weren't confident in the accuracy of the information they could provide. So felt that their participation in a survey about wealth wouldn't really be that worthwhile or insightful. Those who were more unwilling felt that it's private information that they'd rather just not share. Selecting a band to indicate wealth could increase willingness in some instances, but the bands would need to be fairly broad in order to keep the information private. For participants who have joint wealth with their spouse or partner, their ability to provide proxy information would be similar to providing their own information because these figures would be roughly the same. Where spouses or partners kept wealth and assets separately, participants were willing to provide the information, again, as long as they had permission in the first place to do so. They would involve the other person in this process of responding or suggested that partners should be surveyed separately to improve quality. Participants in unrelated households would really struggle with this and wouldn't be willing to ask information for similar reasons we've already discussed. Finances are private. Next slide, please. Great, so just to finish up with some considerations for the future. So the next steps in this area would be to develop an instrument that took these mental models on board and then test that approach to ensure that it translated well into something that adequately collected all types of income, spending, and wealth. There was also evidence that a language barrier caused some confusion for participants whose first language was not English. Although a translator was present at interview, there was no way for our researchers to know how the topic guide and the questions were being translated. And this is something that we think would benefit from further research as well. And with household composition being a strong factor, contributing to the differences in ability and willingness to provide financial information, especially on behalf of other household members, we would recommend considering how separate surveys for household members would affect response rate and data quality compared to that proxy response. And lastly, further research with people who would be classified as economically active and employed. So that's looking and available for work or waiting to start a new job would be beneficial. As like we mentioned at the beginning, we were unable to recruit this group for this research and the mental models of this population may be different to those sampled. And that's us all done, thanks very much. Thank you, that was brilliant. Just a reminder to put any questions you have in the chat, but then we'll now move on to our next presentation. So we now have a presentation on developing admin-based income statistics from Michael Cole. So Michael Cole is head of social statistics admin first at ONS. His area leads on the development of admin-based population characteristics statistics. And Michael is a government statistician that has worked in a range of different departments and has been at the ONS since 2021. So Michael, over to you. Thank you, Jennifer. So yes, thank you everybody. I'm going to take us a little bit away from, so to be as for a second, I'm going to think about what we're doing in this piece. Using admin-based income statistics as an alternative traditional CVS measures. So we're going to jump to the next slide, please. And we're going to start a little bit of context first before we start, that's all right. Next slide, please, Daniel. Cool, thank you. So there's never been an income question on census because we know it would affect negatively how individuals respond, but you weirdly know that the need for user, if you know the occasion needs while I'm going in particular, I think it's small area income statistics. And we know too that the current CVS measures that are produced by ourselves at ONS do struggle to meet that user need. The closest being income estimates for small areas which are published every two years at MSLB level only. So we've been doing a lot of work over the last more seven years to explore the feasibility of using admin-based measures of income. And seeing if that can better meet that user need going forward too. Next slide, please, Daniel. Thank you. And it's really important to be able to put this into the wider context. So we here at ONS are committed and we have an ambition to embed and reuse admin data across our products and statistics going forward. And this is in the wider context of our ambition to put admin data at the core of our future population and migration system. And just a plug here too, to flag that the consecration on the future system was launched on the 29th of June and that will run until the 26th of October too. So great. Next slide, please, Daniel. So what we've been doing in this space is a question. And I'm going to talk a little bit about through our research date and talk with you how we've made that and how it's being used to produce both admin and statistics but also some related products that are built from it as well. Next slide, please, Daniel. Cool, thank you. But in developing admin-based income statistics, it's really important that we do so to a national and aligned standard. So we've been working really hard to make sure that our admin-based income statistics that we've been producing are built from and use the camera hand book as our guide book to make this. It can actually comparable, but also capturing those components that we recognize need to be captured as part of this process as well. And we've been developing our statistics aligned to those standards and we recognize that there are some missing components currently, but I'll come to that a bit later on but we are working really hard to try to make sure that we create a complete possible income picture from data that's reflective of the standard. Just going forward too. Jump forward, please, Daniel. So what's all we've been doing in this space? So over the last seven, almost seven years now, we've been iteratively developing our admin-based income statistics. So going back to 2016, we saw our first release go out and that at the time only used some information from the PYE system and benefits information. And we're able to then demonstrate how we were able to reduce individual gross annual income distributions at local authority and by age and by sex. And over the last, almost seven years, we have been iteratively developing that with updates ad hoc, but incrementally intensive, adding additional information in there allowing us to do a lower level geography, but also to doing adding additional components of income too. So you can see a really quick timeline of that development over the last couple of years and the latest release of our business income statistics went out in December, 2022. And for the first time, allowed us to include the benefit information from use of credit, personal independence payment. And again, was allowed us to demonstrate how we can reduce this admin-based income statistics at local authority, yes, but also down to LSEA level by age and by sex as well. And we are obviously very keen to continue that development going forward too, but I thought I would just highlight a little bit of development I've been doing and how I've been progressing that this work over the last couple of years. Next slide, please Daniel. So how do we make our admin-based income statistics? And again, I'm just gonna just quickly talk through our process. It is our methodology, we begin at individual and household level. So we start on the left-hand side, bringing in a number of data sets from across government. So information from the department of weather pensions and also HMRC, so capturing employment income, self-settlement income, income people receive from benefits, tax credits, housing benefits, shell benefit, use of credit, and personal independence payment. And we also do some invitation to capture some of the elements as well. We take all that information or that data at an individual level and we create a combined income-linked data set which we then use as our source to take through this process. We link that onto our statistical population data set which again is an individual level data set that's been created and that's built to produce a stock of the resident population of England and Wales. And then from there, moving from left to right, we work from that point to remove individuals that haven't been necessarily linked. We add in disimputation for some other components leading us through to a measure of gross income that is then produced from which we're then able to calculate proxy tax national insurance amounts to derive a net level individual income and also a household level of income using equalization which then led to our outputs which again, we've been released over a number of years but was recently in December which allow us to produce individual equalized occupied address levels of income at the sales by age, sex, regional authority and LSRA crucially. If you'll go to the next slide please. Thank you. So how do our admissing home stats compare in contrast and what's the coverage of those in terms of how will the capture income? So on the left hand side, just some really high level summary stats taken from a recent December. So using our population base of the statistical population data set, we were able to capture and demonstrate that we were able to show that we were able to find some level of income for 92.5% of individuals in there and 98.6% of occupied addresses required some information. Crucial question is, of course, are we capturing all the aspects of income? Are we capturing that income correctly? And yes, we recognize that there was some more to do there but the right hand side just shows really quick appraisal against that Canberra in National Handbook definition of gross household income. And we are capturing quite a lot of chunks from there in terms of income from employment. And we have work underway and planned to further improve that in terms of capturing income that is received from property and income from household production services from consumption as well in transfers as well. So it's just a really, really quick highlight of our coverage date, which shows real promise and to an appraisal of our assessment against that versus the national standard which we're trying to work this to meet. The next slide, please, Daniel. So just give you a flavor. You may be familiar with our income base admin so I'll just give you a flavor for what we are able to produce. So here, just a quick snap shot just to show you the types of geocatholic breakdowns you can receive from that environment from our statistics. So here, just highlighting some really high level interactive maps that are produced alongside the statistical summary and also the tabs I've got alongside the release. So here, just short showcase and that way we have a lot of ability to drill down to LSA level and provide that median net in digital income level information. So here, I'm showing both the coverage on the left-hand side and also the right at that median net individual income value as well. While our income sets have been developed over the last couple of years, we also know the value and the importance of users to produce multivariate statistics build on this. So over the last couple of years too and most recently in February this year we released a number of multivariate case studies one, which I'll come to in a second for the first time allow us to produce income by ethnicity you use it up in data. Can we do the next slide please? Thank you Daniel. So here, just short case and that recent work from February this year to develop a multivariate income by ethnicity statistics. So here we're using our database income sets but also combining that with our work on ambassive ethnicity as well. And for the first time as I say, we're able to identify both in income but also ambassive ethnic group for those individuals as well and produce that for the first time in a way that wouldn't be possible to same granularity in time. It's what our current series just can do. We'll hope to be able to iterate this in time going forward as well. Equally two, back to that original point I made at the start in terms of we know that there's an ongoing need for income information. We know too that it will be detrimental to put it on census. We're also doing a lot of work in delivery of a income and census-based product as well. Can we do the next slide please Neil? So we're also working towards producing a census type income release which will combine wrapping-based income stats and also census 21 data as well. And we are working at a pace on that and we hope to be able to have that delivered by April next year. But we have recently put out a progress update that provides an overview of what we have been doing in the space in our progress to develop a census type income data release in time too. So I'll pause there which has been a whistle-stop tour of admiss income stats. And yeah, obviously I'll pass across to our next speaker today. Okay, so in our last session from ONS, we have a presentation on household income and expenditure analysis from Lee Colvin. Lee Colvin joined the ONS as a government senior statistician in 2019, currently working in the household income and expenditure team and leading on the ad hoc production of income and expenditure statistics. His area of interest is measuring the impact of cost of living rises and the development of household finance statistics. And prior to working at the ONS, he was a secondary biology and science teacher for 14 years. So Lee over to you. Thank you, Jen. This morning I'll take you through the household income and expenditure analysis for financial year-ending 2022. Next slide please. So today I'll be looking at two of the current ONS annual releases that have come out. First one being the household disposable income and inequality publication for financial year-ending 2022, in which I'll look at the changes to the mean and median household income, how household income changes across the distribution and what has contributed to the changes of household income. The second release that we'll be looking at will be the family spending publication for financial year-ending 2022. We'll be looking at what households are spending the most on, how expenditure is changing across the distribution and how expenditure compares to pre-pandemic levels. All income and expenditure values in this presentation are at 2022 prices. Thank you. Next slide please. So what surveys feed into these two publications? Well, we have the living cost of food survey which feeds into the family spending publication in a sample size of 5,000 households in which households fill out an expenditure diary which covers two weeks. It has coverage of the UK and it's an annual survey. The household finance survey feeds into the household disposable income and inequality publication. Again, it has some size of 17,000 households in UK coverage and it's an annual survey. So what has happened to the household income across the UK? Next slide please. So we can use mean or median to describe household income. The problem with the mean is that it can be influenced by individuals or by individual households. So this is why we're gonna focus on the median today as we're refocusing what's happened to the overall UK population. So the annual household median fell, income fell during the financial year-ending 2022. It fell by 0.6% which is 180 pounds to 32,349 pounds annually compared with an average increase of 1.7% per year for the last 10 years since the financial year-ending 2013. And if we look at the mean for the same time period, it increased by 0.9% which is 344 pounds to 39,328 pounds annually compared with an average 1.4% increase per year for the last 10 years from the financial year-ending 2013. So it's clear to see from the mean that there are households within the UK where their income is growing but for the overall population in the UK, it's actually decreasing. Next slide please. So if we look at it over a timeframe, we can see that the fact of the average yearly increase in median disposable income has leveled off in the last three years. So between the financial year's ending 2020 and 2022, the median income actually increased an average by 0.7% a year where over the last 10 years from the financial year-ending 2013, it had an average of 1.7% increase per year. And even in the most recent year, the median has actually fallen. Next slide please. How has this changed across the population? Well, disposable income from the poorest households is decreasing. So the median disposable income decreased by 3.8% which is 572 pounds to 14,500 pounds annually in the financial year ending 2022. And this was mainly driven by reductions in the original income and cash benefits. If we look at the figure on the right-hand side here we can see quite clearly for the bottom quintile that it is now lower than the financial year ending 2020 and equally the same for the second quintile. So the bottom 40% of all UK households median disposable income now is now lower than it was in financial year ending 2020. So how does this commit to the richest fifth of households? Well, their actual disposable income increased for financial year ending 2022 by 1.6% which is 1,034 pounds to 66,000 pounds which brought them more in line with the financial year ending 2020. And this was mainly driven by an increase in the original income that was not fully offset by an increase in direct taxes. Next slide please. So how does this compare to households expenditure? In this section I will be using the mean income due to the fact that we don't have a median expenditure as of yet and that's something we'll be looking into in future works. So the weekly household expenditure has increased by 6% in the financial year ending 2022 but which was an increase of 28 pounds to 528 pounds weekly where this is compared to mean disposable income which increased by 0.9% in the same time period. Could we go to the next slide please? So what are households spending their money on? So in this section we'll be looking at a category called housing net and energy. In this category this does not include anything to do with mortgage payments. It only includes rent, maintenance of dwellings and home improvements and energy bills. So 52% of UK households expenditure is spent on what you could consider essential categories. They have a total household spend of 528 pounds 80 per week and 52% of that is spent on mainly rent and energy bills. Transport in the forms of buying and maintaining and running personal transport, food and non-alcoholic drinks, mortgage interest payments, council tax and et cetera. Can we move on to the next slide please? How is this changing across the distribution? Well it's quite clear the poorer households are making different expenditure decisions. They are spending proportionally more of their expenditure on housing, fuel and power. So they have a total expenditure of 329 pounds 80 a week and their highest expenditure is going on mainly rent and energy bills, food and non-alcoholic drinks, transport in the form of the maintenance and running of their own personal transport. Now when we compare this to the richest fifth of households who have a total expenditure of 811 pounds 20 a week, their highest expenditure is going on mortgage payments because they're more likely to be homeowners. Transport in the forms of buying and purchasing second hand cars in particular and in the housing net and fuel and power, it's mainly on home improvements and energy bills. Now while the richest fifth have an expenditure which is more than double that of the poorest fifth of households, the richest fifth disposable income, means disposable income is six times that of the poorest fifth of households within the UK. So how does expenditure compare to the financial year ending in 2020? Next slide please. So we could quite clearly see the expenditure has not returned to the pre-pandemic levels, so financial year ending 2020. Now during the pandemic certain groups where card cop categories were hit, particularly spending on restaurants, hotels and recreational cultural services. So if we focus on these two groups to start with, we can see therefore restaurants and hotels, though the spending increased in financial year ending 2022 by 82%, it still remains 32% below the financial year ending 2020. And this is similar for the recreational cultural services. They increased in the financial year ending 2022 by 37%, but still 24% below the financial year ending 2020. And if we look on the figure on the right hand side here, you can see that this is the case for the majority of card cop categories, the spending has increased in the financial year 2022, but still nowhere near the expenditure of the year financial year ending 2020. To say we're moving into the time for questions. That's fine. Next slide please. So our coming work is gonna be, we're working on with the OPN, the Opinions and Lifestyles Survey Team cost-living analysis in which living costs and foods are very dangerous playing a part. There's the effects of tax and benefits and income statistics, which looks at the distribution effects of individuals on households and indirect taxations and benefits for financial year ending 2020. And long-term methods development, we're looking at an improved combination of income and expenditure data to allow better comparisons and to improve coherency of income and expenditure statistics. Next slide please. Thank you for your time. If you want to contact us, please contact us on the two emails below for income. Please contact us on the HIE inbox after expenditure. Please contact us on the family spending. Thank you.