 Hello everyone. Today I'm going to talk about understanding the English indexes of multiple deprivation and then making use of this indexes, which I would demonstrate with the help of my ongoing research on poverty deprivation and dementia. I plan my talk like this. I would begin with a brief overview of the concepts of poverty and deprivation, measures of deprivation and then I will move on to discussing the English indexes of deprivation in detail and then my ongoing research. So coming to the concept of poverty and deprivation different economists have tried to define poverty or the state of deprivation in using different terms. Well, I'm not reading out the slide, but it is very clear if you read this slide that poverty is complex. It is not really easy to define poverty. There have been differences in explaining it both conceptually as well as from the measurement perspective. Poverty line or consumption norms that we would see soon actually distinguish poor from non-poor and also give the proportion of poor to non-poor or to be more precise proportion of poor in the total population. Well, Nobel has tried to define deprivation as not enough financial resources to meet the needs and Amartasen says that absence of entitlement is deprivation and so very quick round of understanding on what he means by entitlement. So Amartasen explains that entitlement could be of five types. One could be through trade that is you go to the market and buy the thing, maybe a loaf of bread or something. Maybe you produce it on your own or you work in a factory which produces loaf of bread, where you give your labor and get some wages or loaf of bread in return. It could be through inheritance or transfer or it could end disruption in either of this could cause deprivation. Again, deprivation could be for a short period of time or it could be a prolonged one. Now coming to the measurements of poverty and deprivation. The two most commonly used measures and the traditional measures of poverty or deprivation is the headcount measure. Headcount measure is based on income and the shortfall from what is defined as the poverty line is the number of persons below the poverty line is the headcount of the number of poorest. Somehow this measure suffered from the limitation that it is not sensitive to the distribution of income and even those within living below the poverty line what is the extent of deprivation that cannot be gauged from this headcount measure. The second measure minimum nutritional or food requirement also is somewhat similar to defining a poverty line that if somebody does not get specified intake of calories per day, then they are classified as poor or deprived. But again, this measure also has limitations that choices of commodities and habits, everything varies significantly across regions and cultures. And if you try to link this calorie norms or minimum nutritional or food requirement with income, then again, it would be based on the income based poverty line. And the nutritional requirements would in turn determine the income requirements. And therefore, the third measure of deprivation is the relative deprivation. And that is what we are going to discuss in detail in terms of English indices of multiple deprivation. I mean, how much a particular group of people or how much a particular geographical area or an individual is deprived with income in comparison with the comparable groups. That is what the concept of relative deprivation is. Now if you go to the background and history of English indices of multiple deprivation. It is a successor of the index of local deprivation so prior to the indices of multiple deprivation there used to be a local deprivation index. And the index of multiple deprivation was developed and commissioned in 1998 by the then department of environment transport and regions and it was published in the year 2000. Since then index of multiple deprivation has been published in 2004 2007 2010 2015 and the latest one is 2019. But one thing it is referred to as indices of multiple deprivation. It is not an index it is indices because it has seven sub components it has seven components or seven domains and each domain has its own index. Now, it's important to talk something about LSO a which is the lower layer super output area, because indices are estimated for lower level lower layer super output area. And this are basically neighborhoods, which are of similar or uniform population size, ranging somewhere between 1000 to 1500 or approximately 650 households. So indices are basically estimated for LSO a and then they are aggregated and made available for the higher administrative levels like local authority districts and so on. Now, as I mentioned, there are seven domains of the indices of multiple deprivation. The three domains are seen on this slide. One is income domain second is employment domain and third is the education domain education skills and training domain. Other four domains are health deprivation and disability crime barriers to housings and services and living environment. Each of the domains have been assigned weights. So income income domain and employment domain has been assigned 25.5% each education skills and training and has deprivation and disability is assigned 13.5% each. And crime barriers to housing and services and living environment is assigned 9.3% each. Now, this individual domains income domain comprises seven components, employment comprises six components, education skills and training comprises seven components. Health, health deprivation and disability for components crime for components barriers to housing and services for components and living environment to components. Now, each of this indices are estimated using different methods. This methods are described in the on the slide, and I'm not reading out considering the time constraints. We can always have a discussion in the question and answer session. So I move on to then the measures of the indices of deprivation. So in what different forms are these numbers estimated and made available to us. So one is the average rank and another is average score. So these two measures basically summary provide the summary of the average level of deprivation. Second is the proportion of LSOS in the top most deprived decil across the country is the degree to which higher level area that is, for example, local authority district is highly deprived. And third one is the local concentration summary, which identifies again the higher level areas, which have extreme levels of deprivation. Now for my research I have made largely made use of the average score and the proportion of LSOS in the top most deprived decil. Okay, so aggregation as aggregation is done at this level. So as you can see on the slide, and they are made available to us. And this data are available in public domain. So, as I mentioned earlier that this indices are estimated at the LSOL level, there are 332,844 LSOS across England. There are 38 local enterprise partnerships, 317 local authority districts, 191 clinical commissioning groups. So the indices are the scores calculated for LSOS that then aggregated to local enterprise partnerships, local authority districts and clinical commissioning groups. And the ones which are computed for local authority districts are further aggregated to get the estimates for the upper tier local authorities. So how is the aggregation done? So we have the average score for the LSOA. It is converted into a weighted average score by multiplying it with the LSOA population. And they all are then summed up and divided by the total population for the higher level administrative units. That is, could be local enterprise partnership or local authority or clinical commissioning group. So these are two levels. Now, if we look into the characteristics shrinkage method, I did not read out on the slide, but shrinkage method is used to derive the shrinkage method basically enhances the reliability of the measurement of the index of multiple deprivation score. Even the individual indices of multiple deprivation. And most of the domains also employ factor analysis to identify one single common factor and against the possibility of a more meaningful factor. Exponential transformation is also undertaken so that a lack of deprivation in one domain could compensate for the deprivation in the other domain. And this transformation, the expansion transformation is scale independent and therefore it is not affected by the size of the lower layer super output areas population. So these are basically the characteristics of the indexes of multiple deprivation to ensure that they give us a robust estimate of deprivation. However, just like any other measure for anything, including deprivation or poverty, even indices of multiple deprivation has some limitations. And factor analysis, which is used to identify the single most important factor also suffers from the limitation of replicability that factor analysis is based on correlation and correlation change over time and therefore it has the issue of replicability over time. Another issue or the limitation is that there is no robust method to validate deprivation measure for small areas. Again, the ways that are assigned to each of the domains. The reasons for assigning particular weights is not clearly explained. And there could be possible issues of double counting something that came to my mind was the universal credit claimants, which is introduced in the index index of multiple deprivation for 2019 universal credit claimants in no work requirements is included both in income and employment so it might result in double counting. Now I move on to my ongoing research on dementia and deprivation. So dementia is measured as the diagnosis rate of dementia or call it dementia prevalence and that is particularly for the age group of 65 years and above. Now if we look at the statistics of the diagnosis rate of dementia across England. The data are for the month ending January 2023. So the diagnosis rate ranges somewhere between 20 to slightly more than 80. The average score for index of multiple deprivation the combined index of multiple deprivation ranges from slightly less than 10 to slightly more than 40. The diagnosis rate is the percentage and therefore it has a fixed range from 0 to 100. Whereas average score of index of multiple deprivation is a weighted average of composite score of subcomponent. So they do not have any fixed range. But if you look at the maps on the left hand side of the light in the upper panel you will see that a large proportion of the. The boundaries in the map represent the local authority district. They are not at LSO level. The reason is the diagnosis rate of dementia the data for that is available at local authority district level. And therefore I have made use of the average score of index of multiple deprivation also at the local authority district level aggregated at local authority district level. If you look at the diagnosis rate of dementia for people aged 65 and above the rate is really very high and it's very high across the country. You will see very, very less regions which are actually green or slightly darker than green. Most of the regions you will find it in the brown and red area, brown and moving gradually to on the color scale towards red. Whereas if you look at the average score of index of multiple deprivation overall. Most of the areas appear to be in green. And of course you might notice that there is a small gray region in the map and that indicates unavailability of data for those regions. But again, a large proportion of the local authority districts are green in color, which means that the average score for index of multiple deprivation is very, very low, which means most of the areas across England are relatively less deprived. Now, if we compare specifically the health deprivation and disability domain. There, the deprivation seems to be very high. The score range from slightly less than minus two to slightly more than one. That is the scale at scale on which the health deprivation and disability index is measured. Also some kind of weighted average, but as you see shrinkage method is applied and then factor analysis is applied and then exponential estimation is applied. And finally, you come at the values which may even range from may even have a negative number. So higher the number more is the deprivation. Here you will see that again, a large proportion of the local authority districts show high levels of health deprivation and disability across England. And if you see the, there is a pattern which matches more or less in the panel in the upper panel and the lower panel, that is the map for dementia diagnosis rate and the map for the average score for health deprivation and disability. However, we will see that the correlations correlation is not very strong between the two. Okay, so now if we compare the average score of overall IMD and health deprivation, you will see that while overall deprivation seems to be low, health deprivation seems to be high from the color that you can see in the maps that health deprivation index is more brown and red, whereas overall IMD is relatively greener coming to the correlations. So the first column shows the correlation of different domains of deprivation overall index of deprivation with that of the dementia diagnosis rate. And we can see that there is a weak correlation or a mild correlation or medium correlation across different domains of deprivation. Well, as I mentioned that dementia diagnosis rate is for the people who are 65 years and above, and therefore, another index which is specially estimated for old age people who are also 65 years and above, income deprivation for old age people, this index is also calculated, estimated and made available by the government of UK. So that also I have taken into consideration for my research and the correlation is again very weak. But if you see the correlation between the dementia diagnosis rate, leaving environment and barriers to housing services, there is a negative correlation, which means that higher the barriers to housing and services, lower is the dementia diagnosis rate and higher or better the living environment, lower is the dementia diagnosis rate. Now, if we take into consideration the proportion of LSOs which are highly deprived in the top 10% or the top decibel, then you will see that you see the map, which is absolutely green in the bottom panel of the right of the left side of the slide. So, across England, there are very few, I mean, very few local authority districts which has slightly higher levels of higher proportion of LSOs which fall into the top 10 decide on deprivation. This is the case when we come to health deprivation and disability domain. When we talk about the proportion of LSOs which are highly deprived, the proportion is really very less. So this is in contrast with the average score where we saw that the average score shows that there is a high level of health deprivation across the local authority districts. It shows that the proportion of LSOs which are highly deprived in the local authority districts for the health deprivation and disability is again very, very low. Therefore, this requires further research further exploration and unlike the average score. So this is all I had to discuss today and the major takeaways I can think of from today's session is there are many methods of measuring poverty. However, the most recent method of estimating IMD has the least limitations and it covers a wide range of domains of deprivation. There are some limitations which have already discussed in detail. I'm not repeating it over here, but the indices of multiple deprivation are available in the form of average scores for LSOs and they are made available at higher administrative levels also. And the research that I mentioned which I am currently, which is an ongoing research that examines the patents and associations in dimension deprivation at local authority district level. The research will be association that is a negative correlation as I mentioned between the dimension diagnosis rate and deprivation for housing and services and living environment. And there is a very high level of correlation across among the domains of deprivation with the IMD. Now the way forward is we are planning to examine the pattern in correlation over time and we also plan to undertake the impact to examine the impact of dementia while holding for demographic characteristics. Thank you very much for patient listening and you in question and suggestion. Thank you.