Preview Mode Links will not work in preview mode

Mar 21, 2022

Since 1801 the UK has undertaken the mammoth task of counting its entire population, transforming over the years to uncover a wealth of information about the people of our country. With all the data collected for Census 2021 our most comprehensive quality assurance programme ever is currently under way, drawing on the unique insight and expertise of local authorities across England and Wales to help us produce the best possible statistics for every local area. With results expected in early summer, Statistically Speaking meets the people who ran the first digital-by-default census during the pandemic to find out how it went and what the results will able tell us.

Transcript:

MILES FLETCHER

Welcome to ‘Statistically Speaking’ the podcast where numbers talk and we talk to the people behind them. This month we peek behind the scenes of one of the UK’s biggest mass participation events, the Census. 

Almost every ten years since 1801 the UK has undertaken the massive task of counting its entire population. But what began as a headcount to measure population growth has gradually transformed over the years to add a wealth of information about all of the people in our country. 

So now, with all the forms collected, a very large programme of quality assurance is going on and that means taking data from an array of alternative sources, information gathered by the government for other purposes, to see how far that corroborates the picture that's emerging from the census forms. And also for the first time this year we're drawing a unique insight and expertise of councils across England and Wales as to use their local knowledge to help us produce the best possible statistics for every area. 

Taking the further evidence gathered from those 250 organisations into account, we now aim to publish the first results in early summer 2022. However, some releases of early data have already helped to shed light on the effects of the pandemic, and more recently, the response to the Russian invasion of Ukraine. 

Today I'm joined by Pete Benton, one of the people chiefly responsible for running the first ‘digital by default’ census during the pandemic, and Sanjay Jagatia one of our vital Census Community Advisors. 

Pete Benton, the England and Wales census, it's the largest statistical operation certainly that any government has to undertake, and it's sometimes described as the largest regular peacetime operation that UK government does. But just give us a sense of the scale of the whole thing and what's involved in numbers.

PETE BENTON

Well, the simplest way to think about it is we're asking everybody, every household in the country, to do something in the same way on the same day. That's quite a challenge. You know, people compare running a census with standing up the London Olympics. That’s a big job but you don't have to get everybody to take part. And in simple numbers, there are about 25 million addresses in England and Wales and we need to get a response from everybody.

MILES FLETCHER

The ONS, it's a relatively small organisation for a Herculean task like this. It isn't a case of everybody in the ONS downing tools for a few weeks, is it? And this time around we're measuring COVID and the economic impact and all of that. So, who are the census people, where do they come from, and how long do they stay with us?

PETE BENTON

Well, it really does take us 10 years to plan every census. The 2021 census - I actually started with a team in June 2011, just after we'd done the 2011 census. Now, the first question wasn't how are we going to do the next one? The first question was, do we need another one or can we get the information that is so desperately needed about the size of the population from data sets that already exist? So, we spent three years looking at questions, specifically after the 2011 census, and then in about 2014 came fairly firmly to the conclusion that we do need a census in 2021. And there were seven years-worth of efforts and some people were still around from 2011. There were even some from 2001, one or two people who'd even been part of the 1981 census so you get a little bit of corporate memory, but most people are new when each census comes along. We have a small team to start with. And then it just grows. And in particular, in the last year or two, you fill up the office with the people that are going to run the operation.

MILES FLETCHER

As well as the people in the office, you've got to organise this huge army of people, this ‘fieldforce’, who actually, as you say, go out and knock on doors.

PETE BENTON

You're right. When I joined back in the 2001 census, just after that, that census took about 70,000 staff. And the census before that took over 100,000. By the time we got to 2011, with advances in technology able to support people in different ways, we brought that down to 35,000 staff, and for the 2021 census it was under 20,000. So actually, we’ve become better and more efficient as technology enables us to change the way we do things.

MILES FLETCHER

Now there was less activity on the doorstep perhaps this time of course because for the first time we were trying to get everybody, if they could, to do it online. What were the challenges of that particularly?

PETE BENTON

Well, interestingly, it’s always been a self-completion job, or at least certainly for recent decades. You get a paper questionnaire come through your door, you fill it in and you give it back or post it back. 

The big difference this time was that we wanted the vast majority of people to do it online. There was no online option for 2011. By the time we'd given everybody a paper questionnaire, they mostly did it on paper. 16% in 2011 did it online. But for 2021 we’re at 89% of households who completed the census having done it online, and 11% having done it on paper. Now that's quite a shift in a decade and so far, I think that's the highest online percentage of any census around the world. There are a few going on at the moment. Australia's has only recently happened and we're still waiting to hear how theirs went. But that shift to doing online actually genuinely for most of the population did make it quicker and easier. And for us, in ONS, because it's all electronic, we can see immediately when every household has responded and can see those that still need some further help.

MILES FLETCHER

The technology on the day. It was smooth, it was efficient?

PETE BENTON

It was fabulous. So of course we thought long and hard about how big might the peak actually be and we scaled the system to be able to cope with our predicted online demand. And we've actually worked very, very close with our estimates. 

At the peak we had 60,000 submissions in a 10 minute period at near lunchtime on census day. And that's the submissions, but when you think that people are looking at our website and they're going through every question, the load on the website was quite phenomenal. 

During that peak period we had 30,000 requests per second for information to our website, whether it was to serve the next question on the questionnaire, or to get a bit of extra information about the census. And because we built it in the cloud that just scaled naturally and easily. 

We worked with the cloud providers to make sure that we understood how to make things dynamically scale according to the numbers of users, and it worked absolutely seamlessly. You can imagine there were a few people biting their nails in the days running up to it, with something quite that big, but on the day it went without a blink.

MILES FLETCHER

Now there’s always a bit of a PR push to let everybody know that the Census is taking place. This time around we had some new events and happenings taking place. Do you want to just talk us through how we drew people's attention to it, and got it onto the media and into the news?

PETE BENTON

It was a mixture of things. I mean, every address got a postcard through the door two or three weeks before to say “the census is coming, watch out for it”. Every address got a letter through the door a bit later to say “census is coming, here's your access code”. So, every address had something that said it's coming. 

But sitting around that was a huge campaign. There were TV adverts, there were radio adverts. There was an awful lot of stuff just in the news because we were pushing it out there and people got behind it. But we also for the first time had 300 local staff, six months before the census, just working on local community engagement, getting the word out. We worked really closely with local authorities to use all of their local networks to promote the census to every local group so that everybody knew that it was coming, and also just how much it mattered for their community to be counted. Because if they weren't counted they wouldn't be in the numbers, and services couldn't be planned to meet their needs - services like the NHS and school places and transport - these things all depend on the number of people in a local area. That campaign was essential to get the response rates up.

MILES FLETCHER

And it's the campaign that has to reach the parts that other campaigns can't reach because, when you're trying to reach every single household in the country, there's always quite a proportion you cannot reach through the media. Who are those people, and how did we get them particularly?

PETE BENTON

So, we translated all of our materials into 50 different languages. And we had specific adverts in certain languages on the radio and on the telly, and we were tailoring the content and tailoring the networks we were using to get the broadest reach we possibly could. And we estimate that after census day, well over 95% of people were aware of the census and knew that it was coming. It was incredibly successful and it translated into over 97% occupied households actually filling in a census, because only one person for each address needs to do that. So you don't have to have every person knowing that it's coming but at least have one person from every household.

MILES FLETCHER

And that engagement of course comes to a head as Census Day approaches. We have buildings being lit up in the census colour of purple; we had mentions on soap operas; you're on the BBC One Show as well, banging the drum for the Census. And what was fascinating is that you could see in real time, because so many responses were coming in online, the completion rates were being boosted by those appearances as they took place.

PETE BENTON

Absolutely. We had real time dashboards telling us both how many people were on our website and how many were completing the questionnaire at any given point in time, and you genuinely could see spikes where the number of people on our website jumped fivefold in a particular minute when the census was mentioned on Gogglebox. And you saw peaks when the census was talked about in EastEnders and on The One Show and BBC News, and then when the TV advert went out on the Saturday night before census day, you could see it in the Ant and Dec broadcast. So through the spikes you could see the impact of the advertising that we did.

MILES FLETCHER

So, you could see some of these early returns going well, but what's it like to be in charge and responsible for that huge operation on the morning of census day itself? Does it feel like Christmas day or do you wake with a sense of trepidation about whether or not it's all gonna work?

PETE BENTON

Well, by the time you get to Census Day, interestingly, you kind of have a sense of how it's going because the letters land on every doorstep three weeks before Census day. So you've got a sense of how many people have been responding, and what times of day they've been responding, over previous days and previous weekends. 

And interestingly, we had the weekend profile from the two previous weekends before census weekend and we could see on the Saturday and on the Sunday what time of day the most responses came in. And it was typically about 11 or 12 o'clock.  So by 11 o'clock on the Sunday of census day we knew that was the peak hour probably, from what we've seen the week before. 

We knew by seven or eight o'clock how big that peak was likely to be, because on previous weekend days we'd seen that eight o'clock peak was about a quarter of what the peak was going to be at lunchtime. I can't remember the exact numbers, but I was looking at seven or eight in the morning to try and predict “how high is the peak?” and what's the total likely to be. And by the time we got to 10 o'clock I think we were comfortable that the peak looked like it would be manageable within our projections and it certainly was.

MILES FLETCHER

And that continued then according to your expectations during the day. Did you have a sense then, come the end of that day, just how many people had completed and whether you were on course for success?

PETE BENTON

Yeah, we could see that we were ahead of what we'd expected at that point. And that is a census-takers dream because you can imagine knowing that you've got this fixed window as a census day. It's a fixed day, there's a period of letters that are going to come out and remind those that haven't responded after the census date, and then a fixed period of field staff. And you know you're on tenterhooks waiting to see whether the response is going to come in quick enough and be big enough. And by the end of the Census Day, we could see actually things were going well and at that point, you can breathe and say, “Okay, we're going to be alright”, you know. When things go wrong it can be horrible. There are stories from around the world of websites breaking on census day and front page news for weeks as the public follow this kind of scenario unfolding, and you know that God willing, we won't end up there, but you know that it's possible and so as you get towards the census there's both nerves and excitement kicking in.

MILES FLETCHER

Okay, so by the end of census day you've got a pretty good picture of how well you're doing, but you know that not every household can possibly have completed it. What do you do next? How do you pull in those remaining households that haven't taken part?

PETE BENTON

So there's probably three big things that we do. We continue the publicity campaign, we look for the opportunities to get on the radio and on the telly and just talk about how well it's gone. We put press releases out so the conversation continues, and that was really successful. 

But then we can see of course, which addresses have responded and which haven't. There's a slight lag in the ones that have come back as paper questionnaires - it takes a while for those to be receipted. But nonetheless, we can see which addresses haven't responded, and at that point we send a reminder letter. The most effective thing we can do is send a letter - it would cost an awful lot more to send a person out on the street. 

Firstly, we try and get people to respond without needing a reminder through all that publicity campaign. Secondly, we send reminder letters and people can get three or four over a period of three or four weeks. And then after about a week, that's when in earnest we start sending out field support staff to go knock on the door and say “we see you haven't done the census yet, can we help?” And most people would say “oh, I've done it” or “I’m about to do it”, and they would genuinely get on and do it. But there were some that needed a few more reminders than that. But eventually we could see that those letters were working and that the field responses were working. 

And of course there was also the option for people to fill it in on telephone, if they couldn't get online weren't able to fill it in on paper, but could do it more easily over the phone. And so by making multiple options available and making sure they were well promoted, we got a fabulous response.

MILES FLETCHER

There was an issue wasn't there with some people who thought they completed it, but hadn’t pressed ‘Submit’, and were then getting chased up.

PETE BENTON

That's right, the Census is a legal document and you actually have to say “to the best of my knowledge and belief this is a true reflection of my household” and press the submit button. And there were a few people who had got to the end, thought they've done the job, but hadn't pressed ‘Submit’. And so actually, they ended up getting a letter that was very specifically worded that said “we can see you've started the census online, but you haven't yet pressed ‘Submit’ would you like to get back and finish it off?” And we saw a good response to that as well. 

I'll let you into a little secret. I actually got one of those letters. It wasn't because I had forgotten to press ‘submit’. It was because my kids actually weren't at home on census day. Two of them are out of the country. And I was debating with them. When are you coming home and are you to be counted in the census because there are some quite precise rules. Because if you're out of the country for more than 12 months, you shouldn't be counted, but if you’re out of the country for less than 12 months you should. So we were just discussing “when are you going to be home? Do you think you'd be home within 12 months?” They said yes, but we didn't quite get there by Census day. I’d filled in my bits, my wife had done hers, as had my stepdaughter, but I was just waiting for two bits from my other kids. And it was fabulous that they could using our access code from our household which I shared securely with them via a secure link. They could go in and fill in their details, pass it back to me and eventually I pressed ‘submit’ a few days after census day.

MILES FLETCHER

How difficult was the decision to go ahead with this census, despite all the years of planning? How difficult was it to press the button and go ahead despite the fact that there was a pandemic still in full swing and the country was still in effective lockdown?

PETE BENTON

Well, you can imagine that was one of the biggest decisions that we've made. And in principle there were essentially two questions that we asked. The first was “can we keep the public safe? And can we keep our field staff safe?” And the second was, “can we get good statistics?” And you can gather from the fact that we went ahead that our answer to both of those questions was “yes, we can keep people safe and we can produce good statistics” but we spent a lot of time looking into that and we actually changed a fair bit of the operation. 

All of our recruitment and training for our field staff moved to being online where sometimes we might have gone into somebody's home to give them support, we were very clearly not going to be going into anybody's home. We kept two metres distance, we kept masks on at any point where we did do a field visit, and even doing those field visits there’s a big decision to go ahead and continue. We had options just to use letters to remind people. But we went ahead and we got good statistics. 

We had to give some extra guidance here and there for certain questions. When we asked “Where do you normally work from, at home or in the office?” We always ask that because many people always have worked at home. But people we saw as we tested some of those things out were kind of thinking well, “what do you mean, do you mean now or do you mean prior to the pandemic?” So, we gave some guidance on a few questions to say answer now, as it is now. So that guidance was clear and straightforward. 

But there will be some interesting artefacts that we find from this census. So one of the questions the census asks is “Where do you live? Where do you work? And how do you travel to work by car, by train or by bike or by walking?” And of course in the middle of a pandemic, most people, or many people, weren’t travelling to work. Post-pandemic we're seeing more people travelling to work, albeit they will never get back to the levels before. So the census itself will show not many people travelling actually, and the census is used for planning transport. How wide do the roads need to be? How many trains do we need? That's all based on census information. And so we're going to need to find ways to update that information after the census for some of those things.

MILES FLETCHER

So the operation benefited hugely as it turned out, from having what was literally a captive audience – people were at home and ready to take part, as it turned out in large numbers. But just how big a problem is it that it was done in those unusual circumstances that you mentioned? Indeed, might it even be necessary to do one again after a few years, rather than wait another 10?

PETE BENTON

Well, actually, when you think about it, it was critical that we did it at that time. We just left the European Union and we were at the heart of the pandemic, and it gives us a fabulous baseline. We know that the population will continue to change as we adjust to a new normal, whatever that might be - no one knows what that's going to look like. 

So having that baseline to say where were we, and then every year to refresh those statistics and say: How has the population changed? How has the way we work changed? How has the kinds of jobs that we're now doing changed? And how do we travel to work - how has that changed? 

So what we're now planning is frequent updates to our statistics, both how many people are in a local area and what types of people are in a local area, using all kinds of new information sources to chart out the change post pandemic and post Brexit baseline.

MILES FLETCHER

So it's a successful operation. People were convinced about the need to take part. They took part in very large numbers online. Just how successful do you think it was then?

PETE BENTON

We estimate that we got a response from over 97% occupied households. Now that is quite something. We set ourselves a target of covering 94% of the whole population in the census responses, but that is just the first step. 

You know, knowing how many addresses have responded is the beginning. But different addresses have different sizes. Some have more people in than others. And it might be that some of the addresses that didn't respond, the few, were either larger or smaller than the average address. So we don't know yet what percentage of people we covered and there are some fairly clever processes that we go through to actually work out who we’ve missed.

MILES FLETCHER

In historic terms, how successful is a 97% household completion rate?

PETE BENTON

Well, in 2011 we estimate that we hit 94%, and something similar in 2001. So 97% is pretty good. But actually we've got a bit more information this time about those vacant addresses than we would have had in the past. In the past, all the record books that the collectors used were on paper. For the first time in 2021 all the field staff had a mobile phone with an app on it that gave them their workload and they could record the outcome of every address that they visited. And so, within a moment, we knew centrally every address that the field had said “this one is vacant”. 

We've never had during the operation that kind of information before, we've always waited till the end and we've always talked about what percentage of the population we've covered. So in a sense there isn't a comparable figure, the comparable figure will be ‘what percentage of people did we count’. 

But nonetheless we're pretty sure that 97% is good going. Of course, in 1801 we have no idea how many addresses were covered. We've got the results, we know something about how they did it, but they didn't have computers, they didn't have address lists. They didn't have extensive records. A lot of it was managed by people just keeping track locally of what was going on. 

So we're getting to a world where information becomes more and more finely detailed as we go through running these operations

MILES FLETCHER

And to get a better handle on what you don't know as well as what you do know.

PETE BENTON

Absolutely. And in the last three censuses, we've actually made quite a big step in how we assess that total coverage of the population. 

To plan public services you need to know about everybody, every different community group, every age group in every part of the country. As you can imagine, doing a census, there are some groups that are more willing to respond than others. Older people tend to be more diligent in completing forms, certainly more diligent than students and young men or busy young families.

So we know that sometimes we miss people because they're busy, and they just don't quite get the census done, and we make adjustments for that. But what we don't want to do is just add a percentage of all age groups. We want to assess who have we got a high response rate from, and who have we got a lower response rate from, and what adjustments do we need to make to our statistics so that the results do the very best at representing the whole population, even if we didn't count every single one of them. 

And there's a technique we use called ‘capture / recapture’, and it's used for wildlife if you want to count how many ladybirds there are on an island or how many fish there are in a pond or even how many taxis there are driving the streets of London. There's a technique you can use, and then we apply it to the census. So let me talk you through that step by step. And if you're really interested in this, there's a little article online. It's called Trapped Catfish and Roach, the beginner's guide to census population estimates. If you Google that it's only a few pages long and all you need to read is the first page but I'll explain it to you now. 

So, you get the idea. Just supposing you've got a pond in your garden and you'd like to know how many fish are in it, and you could drain it obviously, but that's a lot of work and it wouldn't do the fish much good. So instead you catch as many as you can, and supposing on the first day you catch 100 fish, then you very carefully put a little tag on them. Then you put them back in your pond. You let them all mix up and you go back the next day and you go fishing again, and this time you catch 50. And you find that 25 already have one of your tags on them. How many fish out there in your pond?

MILES FLETCHER

I don't know. I’ll blunder a guess and say 400.

PETE BENTON

So what you do know: you saw 100 Fish on day one. And on day two, you saw another 25 that didn't have a tag, so at least 125 are in the pond. You can be sure of that. But you can guess there's going to be more swimming around that you've never seen. 

But the interesting thing that you learn on day two is that half the fish in the pond probably have a tag because you caught 50, 25 had a tag - half of them. So you can kind of have a guess that maybe on day one you tagged half the fish. Given that you tagged 100 that means there are probably around about 200 in the pond in total. 

And you’d be right, give or take a few. And the question is give or take how many? And the interesting thing is the more you catch on day one and the more you catch on day two, the more confident you can be. 

So we catch as many as we can in the census - that 97% that I told you about is our first day and we don't put a red tag on people because everybody's got a name and an address, and a date of birth, and the sex, and we know that from the census returns. So then we go out again about six weeks after the census and rather than ask people to fill the form in we say will you just do a quick interview with me? And we asked them their subset of the census questions again. Then we do it just in 1% of postcodes, and we list all the addresses blind we don't give them an address list this time. We say here's a boundary on a map. Here's a postcode. Go and list the addresses and go and interview as many people as you can. And they've got names and addresses, and sex, and date of birth. And we literally match the two together, and we go to about 300,000 addresses when we do that. And in doing that we can do those sums. How many do we catch first time? How many do we catch second time? And we can estimate the total but because we know everybody's age and sex we can see how it varies for different age groups. And we also ask ethnicity and we see how it varies for different ethnic groups. And we also can see different housing types. So big houses, small houses, and we break it all up into those subgroups, do the estimating and then add it all up. And that tells us in total what percentage of people we've missed in those 1% of postcodes, where we did that second count. 

And then we take that information and we kind of generalise to the other 99% of postcodes where we didn't do the second count and with a whole bunch of fancy maths, we can estimate the total population.

MILES FLETCHER

And that is how you fill in the missing 3% of households.

PETE BENTON

Exactly. And we pride ourselves on not just doing that nationally, but getting it so that right down at the small areas of the country. All the totals add up to the total number of people living there as best we can estimate it. 

So, a lot of work. It takes us a while. And interestingly, we're the only country in the world that attempts to do that process before publishing the census results. Other countries will typically publish the raw count, accepting they've missed a few, and then they'll spend the next year or so looking through the data and adjusting their overall totals but they don't adjust the small area data in the same way that we do. 

MILES FLETCHER

So we're at the point now where all the data has been gathered. ONS has gone away to go analyse it all and we have the wait for the final published statistics, including of course the most important population estimates of how many people are in the country and where they live.

PETE BENTON

So, the first job is not just to collect the stuff that came back online, but also to scan every response that did come back on paper. Whilst it was only about 11% of addresses, that still amounts to somewhere near 3 million paper questionnaires. Each one of those is 30 pages long - it takes a while to scan all of that and to catch the data and we do it with computers using optical text recognition. They're all in there, very secure, confidential processing operation. 

So it took us till about August to get all of that data captured. Then we put it in the mix and we merge the whole lot together. And you know, the interesting thing is you find some houses that have responded twice. So they've sent you a paper return and somebody else has sent you an online return. And even sometimes, people are counted in more than one location different people in different addresses. Parents count kids. Like mine. Parents of divorced families may well both count their kids, so we do a check next of any duplicates and we remove the duplicates. And then sometimes people do give us some unusual answers, like, you know, children who are three years old and they're married or working in a factory and we do our very best to try and tidy up some of that so that at least what we're publishing is consistent and, you know, we'll kind of say for three you're probably not married and fix some of that. Sometimes people do skip a question and we'll do our best to estimate. Based on their near neighbours, what might the answer have been if they didn't fill in the question about what kind of heating they have in their house? If 99% of the houses around them have got gas central heating, we'll take a punt and say it is probably gas. So you get better statistics that way. 

So, there's a bunch of tidying up to do. And then there's the work to estimate how many people we've missed by linking it up with our Census coverage survey data. That in itself takes another while and then our estimates come out. And then the next job is to say okay, do those estimates make sense? 

And we compare it with other sources for how many people have got a National Insurance number, how many people are registered with a doctor, and do our results sit comfortably with those other sources. So, a lot of work to make sure the results are plausible and make sense to when they come out. They're as valuable as they can possibly be. 

MILES FLETCHER

So this huge sense checking operation takes part so that the statistics as you say are as valuable as can possibly be. What can we expect to see first then, is everything going to be released? 

PETE BENTON

So the first thing that we will do is publish that size of the population. By age group and sex for every local authority in the country, so that every local authority can see how many people are here and they can plan on the school places they need, and how many houses might need to be built. And then every couple of weeks after that, we will keep releasing more and more detail like how many people have got a job, what kinds of jobs people do, how's that vary by ethnic group.

And then we will publish something like 5 billion numbers, 5 billion statistics, and that takes us a while to get it all quality assured, to get it all published. But for the first time we will also make available a flexible table builder so people can take a look and say, well, could you give me a count of how many people are this age and working in this kind of profession, who are of this ethnic group, who are gay or whose gender is different from what was assigned at birth, and people can actually produce their own statistics.

Now we double check to make sure there's nothing that's being published that has any risk of identifying a person, because you've got small numbers. But that flexible table builder will mean people can answer some of the questions that might have taken longer in the past more quickly. 

MILES FLETCHER

Billions of data points. How do you make it so detailed to be so very useful, and yet protect anonymity?

PETE BENTON

Well, we spent a long time making sure the questions work in the first place so that when people see this question, they're not answering what they think we're asking. They're answering what we're actually asking. 

But then on top of that, there's a lot we do to make sure that when you see a table that has got small numbers in it, and you might think “oh, that's Joe that lives up the street, oh I’ve just found out that he works in the clinical profession” say, well, we just don't publish those statistics for one thing. We don't publish data that is that detailed on all topics so that you know people could find that out. But there's also some sophisticated methods that we call ‘statistical disclosure control’ – world leading methods, actually. That means that when you look at a given table, you can't be confident that a number that you see actually is absolutely correct. 

In the totals, the numbers are right. But if there are some fairly unique people with unusual characteristics that would stand out in a table, interestingly, we’ll often swap them with a household somewhere else. We will take one record and literally move it somewhere else, and swap somebody else back so that you don't know for sure who's there. 

And then even when we publish the numbers at the end, sometimes we’ll change a number too. If there's a count of two people we’ll change it and make it a bit bigger or a bit smaller and we can do that to any particular number in any table. And we've got some really world leading methods to help us do that. 

So when you look at the numbers, they're accurate, they give you a good picture in total of what's going on nationally and in a local area, but you can't identify any individual with any kind of confidence. 

MILES FLETCHER

So there is a very great deal more to anonymity than simply withholding people's names and addresses.

PETE BENTON

Absolutely. And in a world with more and more data, these methods get more and more important.

MILES FLETCHER

The big question, of course, is whether that was the last census of its kind that we're going to see in England and Wales.

PETE BENTON

Well, I wish I could answer that question today. It's certainly our intention. We are working as hard as we can to find new ways of counting the population, using data that exists in government. How many people are getting benefits? How many people are paying taxes? And can we join that information together to get a good clear, detailed picture of local communities. And we're making really good progress, but we're not going to make a decision until we've been able to compare the results of our alternative new methods with the results from the 2021 census. So, we'll do that and come 2023 we'll make a recommendation - do we have another census or don’t we?

MILES FLETCHER

Pete Benton – thanks very much.

SANJAY JAGATIA

My name is Sanjay Jagatia. I'm actually based in Coventry. As part of my voluntary work, I'm also the chair of the Hindu Think Tank UK, which is an umbrella organisation for the Hindu community across the country - almost 1.5 million Hindus across the country from the last census. And then also the Vice Chair of the organisation of people of Asian origin in Britain. And it was through that that I felt that you know, I wanted to have the Indian community being represented as part of census 2021. And when I saw the advert come through, I thought this was an excellent opportunity for me to personally do whatever I can to encourage as many from the Indian community to take part in census 2021.

MILES FLETCHER

You’ve told us already that you thought that working for the census was a life changing experience for you. Tell us about how it has changed things for you? 

SANJAY JAGATIA

I think for me it was life changing because I was one of the sceptics initially about how census data is used and what benefit, more than anything else, what benefit does it have on me as an individual and as me representing my community? What benefit has it had before? 

I hadn't noticed and I hadn't realised the enormity of the whole process and how important it is, not just for us as communities, but how local authorities and how everybody else will use that data. So, we had to be counted. And that for me was life changing. 

It was also life changing to see people from the age of 18 to 99 in some cases, wanting to be part of this. And the enthusiasm that has built for the first three, four months coming up to census - that for me was really, really important because I haven't seen that engagement ever before in my community for any issue that's happened. Even when you look at voting days, election days and all the other things that happen, that enthusiasm was not there as much as it was on this occasion to be part of census.

MILES FLETCHER

And what does an engagement manager do to make sure that happens, that people do get themselves covered in the census? How do you go about that work? 

SANJAY JAGATIA

When we first started, I did research in the area of all the various Indian community organisations, key individuals, people who had a large social media following as well, and started to engage with them initially because without them, I knew that my work would have been very, very hard to be able to reach their communities and their membership that they had. 

So that piece of work that we did right at the beginning of the census, which was a three week period of researching the entire area to find out where people were, and who the key individuals were, that we can actually engage with the census. And what I did went a little bit further and I identified individuals who we made census champions - people who were able to work with me right throughout the period of the six months of the census.

MILES FLETCHER

How did you persuade them that championing census was a worthwhile thing for them to do?

SANJAY JAGATIA

I think you know that the previous census 2011 and in Kenya prior to that, it was clearly evident that we saw that the Indian community’s engagement was very, very little compared to other groups and minority groups as well. So, this was a key area for me to be able to say you had your opportunity or we have our opportunity of having our voice heard, you know, whether it was roads, transport, education, and what may be available for our communities in our local authority areas. 

And I think having examples of how other communities have actually used the ONS Census data previously for their faiths and their communities was a really good way for me to be able to engage with my own community to encourage them to take part in the census. 

MILES FLETCHER

What do you think it was that was preventing your audience from previously taking part? Was it just a lack of awareness? 

SANJAY JAGATIA

I think the awareness this time round for Census 2021, in my opinion from my community, there was a lot more awareness a lot earlier. And I think that was because of the focus group meetings that were held with key individuals from each community in each faith. And I think back then it started to filter down to communities at least six, seven months in advance of Census happening. 

And I think from my point of view, I think it was a lack of awareness of census and how important census can be for each individual and I think that's why there was a low uptake on Census previously. 

But on top of that, as well, I think that there is an area where we have cultural differences, language differences as well. And I think that that was key, that a lot of information previously was very much in a mainstream English language, which was sometimes very difficult for those within the community who are elderly to actually understand that. 

MILES FLETCHER

And did you just target areas where there were known to be lots of people with Indian heritage? Or did you seek to look at some areas, or identify some areas, that were perhaps less well known? 

SANJAY JAGATIA

Yeah, I mean, to be honest with you, from my personal point of view, I wanted to have a look at those areas where there was a low uptake before, to try and see how I can actually engage them a lot earlier than those who I knew that there was larger, you know, quorums of people from the community. So my initial stages were actually looking at those areas where there was a low uptake previously. 

MILES FLETCHER

So you're signed up, you've got census champions, you've got a mission to go out and identify those areas where we could engage Indians, people of Indian heritage, a specific community with the census. Tell us how it went out there on the road, on the doorstep, on the airwaves. How did it go? 

SANJAY JAGATIA

Yeah, I mean, to be honest with you, right from the beginning, in October, November time, when I was recruited for the role, it was very daunting. You know, because we were in a major pandemic. We were in lockdown at the time and I was thinking to myself, as well as my other colleagues from other areas that how on earth are we going to engage with people because it's going to be virtual. 

So, it was daunting to begin with. But you know, when I look at it now , we had so much support from the ONS where we were able to have all of our questions answered, the regular meetings that we had, the updates that we had as well, which then allowed me to start to work with the key champions, also the key individuals that were going to be working with myself and started to put together items and articles that they could use on social media, through their E newsletters, and trying to capitalise on what they were doing already, because they were also in a COVID period and they were engaging with their communities on a regular basis. 

So, I then asked to be part of any communication that they were making to their communities and regularly feeding them information about the census that is coming. It's 10 days away, it is here now. You know, these are the information that you can actually go on and do online. 

I was very, very fortunate that throughout my period of the engagement, I conducted over 188 Zoom calls and these calls ranged from anything from 10 people on the Zoom call to some cases, over 800, 900 individuals. Along with that there was you know, where we got temples involved. So I had a zoom call with just temple heads, you know, the President and the Secretary. So that was something that I was able to do and it really, really proved worthwhile for me. 

MILES FLETCHER

So, because of lockdown and the pandemic, and all the uncertainty of that, you had to go electronic and run down the streets to get the right people in front of their laptops. 

SANJAY JAGATIA

Absolutely. And I think if I look back at it today, I really don't think - and I honestly believe this - I don't think that I would have had the success that I had in Coventry had it been just face to face engagement. What was daunting to start off with actually proved to be so beneficial to be able to have that opportunity of reaching more people at the same time.

You know, I was thinking of holding meetings in community centres and in places of worship and you know that you'd get 10 / 12 people coming to those any one time. It's difficult to drag people out of their homes to come for a meeting to talk to them about something like Census, but the fact that they were able to actually engage with myself, and me to engage with them in the comfort of their own homes was a fantastic opportunity. 

MILES FLETCHER

What sort of reservations did you encounter that people might have had about taking part in this Census? Perhaps of dealing with officialdom and handing over information to officialdom, perhaps? How did you how did you overcome those? 

SANJAY JAGATIA

I think that the biggest barriers that I had right from the beginning was that even though the ONS wasn't a governmental department as such, people felt that the information that they were going to be giving may have an impact on their benefits and their rent and all sorts of different issues that they may have personally, you know, whether they were illegally living in the UK, for example, and I think that was a reservation to start off with. The amount of information that was being asked initially, there was this fear about letting that information go out into the public domain. 

Having said that, once we started to get the dummy or the draft questionnaires that we were going to get people to fill in. I think the whole spectrum changed because it wasn't as in-depth information that they were thinking initially that they had to provide.

MILES FLETCHER

And of course it's an important message at every census time that all of the information you share, all the personal information that you share, of course, remains absolutely confidential for a very long time indeed. Do you think the people you were dealing with were perhaps more receptive to hearing that message from you perhaps as someone with status in the in the Indian community?

SANJAY JAGATIA

Yeah, I mean, I think one of the key things there is that it wasn't just myself but the idea of engaging community leaders who they have a very good link with. That was a key thing for me that they listened to the community leaders and the key influencers and then their reservations were sort of somewhat limited thereafter. 

MILES FLETCHER

Overall then share with us if you would some of the most rewarding bits of this experience. What would you chalk up as your biggest successes and most rewarding experiences? 

SANJAY JAGATIA

I think the biggest thing for me was actually census day, you know where I had zoom meetings right from 7:30 in the morning, right the way till 11 o'clock at night, and the urgency in people wanting to make sure that they take part in the census on that one particular day. Even though they had the opportunity a week before to start this, it was census day itself and the urgency of “yes, I want to have my vote. I want to have my say. I want to be able to say what I want to do going forward in the census”. 

I think that for me, that particular day was fantastic. Not just online but to be able to go out and help people fill out their census form as well going out into the city centre, going out into some of the areas where lockdown restrictions slightly had been lifted at that time. And that was a really, really a brilliant occasion for me. 

MILES FLETCHER

I'm sure Sanjay a lot of the credit for that then rests with you. The proof of the pudding will come where we see the results of course, but in the meantime, thank you very much for joining us today and thank you very much for everything you did for 2021 census.

Thanks to Pete Benton and Sanjay Jagatia. 

In the next episode of ‘statistically speaking’, we're joined by the National Statistician and the chief data scientist at the ONS data science campus. We hear from them how a data revolution and the fight against COVID changed UK stats forever, how we kept the numbers coming at a time of crisis, and how ONS statisticians are using faster forms of data to gain ever clearer new insights.

You can subscribe to new episodes of this podcast on Spotify, Apple podcasts and all the other major podcast platforms. You can also get more information by following the @ONSfocus Twitter feed.

The producers of ‘Statistically Speaking’ are Elliot Cassley and Julia Short. 

I'm Miles Fletcher. Goodbye

ENDS.