The reported situation

At 9.30 AM on the 29 November 2021, the latest results of the 2021 Census of England and Wales describing the religious and ethnic composition of the nation made it into the public domain. The Office for National Statistics (the ONS) told that in 2021 the Census of England and Wales registered 271,327 people who self-identified as Jews in response to the religion question. To be precise, these are people who ticked the ‘Jewish’ box in response to a question ‘What is your religion?’. A decade earlier, in 2011, there were 265,073 such people. That means an addition of 6,000, or growth of 2.4%, in a decade. For comparison, the population of England and Wales as a whole grew by about 6.3% during this period.

Just how credible is it?

This picture is rather incredible. Anyone with sound knowledge of the recent demographic realities of British Jews would say so.  Konstantin Stanislavsky, a notable Russian theatre producer and actors’ trainer, was rumored to react to the less-than-authentic performance of his trainees with ‘I do not believe it!’ snap. A demographer of British Jewry is expected to have a Stanislavsky moment in relation to the newly released 2021 Census results on Jews by religion in England and Wales. Why?

What makes populations grow and decline, largely but not exclusively, is the relationship between people born to this population and those who die. It is called the balance of births and deaths, or the natural balance. When the number of births is higher than the number of deaths, populations are expected to grow. The truth of the matter is that the natural balance of British Jews was strongly positive in during the decade in-between Censuses. Previous research tells us that – British Jewish communal statistics is among the best in the world. With 4,500-5,000 births and only 2,500 deaths occurring every year, the balance is about 2,000-2,500 per year. 2,000-2,500 persons are added to the British Jewish population every year, on average! Now think about a decade-long period: 2,000 multiplied by 10. Despite all not-so-recent cultural developments… a la ‘truth is a narrative’ and all that, 2,000 multiplied by 10 has never been less than 20,000. And 2,500 multiplied by 10 is never less than 25,000. So, without any other forces interfering, the expected addition to the British Jewish population between the 2011 and 2021 Censuses is in the range of 20,000-25,000 then, not 6,000.

Migration matters, of course. Jews leave Britain and they come to it, driven by million different motivations, with Zionism being not the last one. Migration balance, the difference between the number of immigrants into a country and the number of emigrants leaving the country, can be positive or negative just like the natural balance. And it can add or deduct from the population. The scope of Jewish emigration from Britain is reasonably well-documented. On average, about 1,000 British born Jews left Britain annually between 2011 and 2021, so about 10,000 accumulatively. About 70% of them went to Israel, but that is a side point. All these numbers are well grounded because Central Bureau of Statistics in Israel keeps records of the immigrants and surveys of British Jews allow an insight into where else they tend to emigrate, if not Israel. Now, trickiest part to figure out is the immigration of Jews into Britain. It exists. Britain is a developed Western economy, and it attracts a lot of migrants. The proportion of foreign-born in England and Wales today (about 17%) is above the level observed in the United States. Jews come among others. We know that for fact. Israelis come quite a bit, also for various reasons: to work, study, play, relax, escape, you name it. The ONS, the British equivalent of the CBS-Israel, does not register Jewish migrants specifically. Until it does, we have to live with uncertainty as to their number. However, that should not stop us. The situation is clear enough: with the natural balance at 20,000-25,000 per decade and just the outgoing stream of migrants at 10,000 per decade, not accounting for any immigrants into the country whatsoever (knowing that they exist), we can say that the anticipated addition to Jewish population of England and Wales is about 10,000-15,000 and not 6,000. And, given that the immigration stream has not been taken into account at all, it is highly likely to be more than that. An addition of 10,000-15,000 between 2011 and 2021 is a minimum we should expect to see. So, no matter how I turn my crystal ball, some Jews are missing from the religion data in the 2021 Census data release. Where are they? Some are in the data but not explicitly, some are not and will have to be estimated indirectly. Read on and they will be shown to you.

The ‘ethnic Jews’

First, and it would be hard not to notice, identity politics have been quite impactful in the past decade. People self-analyse and search for different, perhaps more genuine, ways to understand and present themselves. British statistical establishment, the ONS, took notice and put a significant effort into finding out how minorities prefer to be named and labelled in their preparatory work in the years leading up to the 2021 Census. Extensive consultations and data gathering exercised were carried out by the ONS among Jews and Sikhs, for example. Jews have been enjoying the identity games as much as anyone else, really. In fact, Jews are somewhat ahead of many others. For as long as one can remember Jews understood themselves as a religious group and an ancestral, ethnic, tribal group, seeing no contradiction between the two terms. This has been well documented and even popularised. In censuses of England and Wales, there are two ways in which a Jew can self-identify as such. One way would be by ticking the ‘Jewish’ box in the religion question. The other way would be by writing-in ‘Jewish’ in response to the question ‘What is your ethnicity?’. Obviously, a Jewish individual can do both: tick a ‘Jewish’ box in the religion question and write-in the ‘Jewish’ answer in the ethnicity question. Since 2001, a certain proportion of Jews did exactly that while others, a small group, avoided self-identifying as Jewish by religion and just identified in purely ethnic terms. 12,000 Jews in 2001, 34,000 Jews in 2011 and 67,000 Jews in 2021 made sure to self-identify as both religious and ethnic Jews. There is a very unambiguous trend of increase there. The provenance of this development is obscure at present.

The extent to which the political debates in the months leading to the 2021 Census influenced the increase in the size of this group will remain a matter of speculation for a while. Never before the question of how Jews should be captured in the Census, as a religious or an ethnic group, attracted so much attention from the Jewish community, both the members and the leadership. Debates on this issue took place at the level of the Board of Deputies of British Jews, no less. Desirability of relating to Jews as an ethnic group rather than a religious groups was highlighted by David Baddiel in a political pamphlet ’Jews Don’t Count’ , published just before the Census. Statistical methodology was not Baddiel’s concern; the perceived exclusion of Jews and obfuscation of their minority status through non-application of an ethnic label to them was. It is possible that Baddiel was not too aware of the preparatory work conducted by the ONS. Authors of political pamphlets are rarely interested in the mundane minutae of the work of statistical establishments. Shame, really, it could sharpen their pamphlets…But this is not my main concern here, to say the very least. Have these debates affected the Jewish responses? Can they be understood as causing a surge in the number of ‘ethnic Jews’? It is genuinely unclear now. The trend of increase was well defined beforehand; without any such debates the number of ethnic Jews nearly tripled between 2001 and 2011. Thus, the more important question is how one benefits from information on ‘ethnic Jews’ . Can incorporation of this information can make the Census counts of British Jews more comprehensible?

It would be a mistake to simply add ethnic Jews to Jews by religion. A lot of Jews identify as both so such an addition is nonsensical and would lead to double counting. In the past, the ONS released data allowing an insight into the number of ethnic-only Jews, i.e. those who self-identified as Jews only in response to the ethnicity question but not in response to the religion question. This group can be unproblematically added to Jews by religion to obtain a fuller count of Jews. In 2001-2011 the ethnic-only Jews were a minority of 23%-25% of all ethnic Jews (3,000 in 2001 and 8,500 in 2011). With respect to the 2021 Census such insight is not yet available but, given the stability of the situation with ethnic-only Jews in the past, we are allowed some assumptions. So: if the past realities hold in 2021, then there are about 17,000 ethnic-only Jews in 2021 (25% of 67,000). These ethnic-only Jews (the suspected 17,000 in 2021, as well as the verified 3,000 in 2001 and 8,500 in 2011) could and should be added to Jews by religion. I will do so in a moment. Yet, this is not the end of story, although the end is near.

The haredi

The haredi….they have been undercounted in the British Census. That first happened in 2001 when the whole idea of reporting religion in the Census was utterly novel (hence the voluntary status of the religion question in the Census in England and Wales). Policy uses of this information were not as clear, let alone established, as they are today. In general, the ethos of evidence-driven policy was not as developed as it is today. Some haredi at the time could have avoided responding to the religion question – though that does not mean that they avoided responding at all. If they indeed avoided answering the religion question, they would have become invisible as Jews in the 2001 Census data. It is a matter of scholarly consensus today that in 2001 haredi Jews were undercounted by about 30%. A decade later, in the 2011 Census, haredi Jews were undercounted by 25%.

With respect to the 2011 Census, research suggested that the undercount was not due the avoidance of the religion question but due to inadequacy of the Census paper questionnaire for registering especially large families. The questionnaire only allowed space for six people to be reported. Larger families had to request the ‘Continuation Questionnaire’ from the Census team specifically, and who would do that…? Population of Britain does not consist of the Census enthusiasts, haredi or not ….Filling in the Census form online, where there were no limits on the number of family members who could be listed, was an option already in 2011 but not many haredi were willing or able to use the Internet then. Given that close to 40% of haredi families have six or more people in them, it is easy to see how a lot of haredi could have ‘fallen though the cracks’ of the Census. To be sure, families as families were reported well in the 2011 Census. We know that much because the comparison of haredi communal telephone directories (the so-called ‘Shomer Shabbos’ directory), listing haredi families in Hackney and Haringey, with the 2011 Census records in these areas confirms similar number of families in both sources. It is people within families, and especially young children, who got underreported and undercounted.

In 2021 Census, haredi got undercounted again. Although it is premature to report on the exact scope of undercounting, and the reasons for it, it is not too early to flag up the very fact that it may exist and to provide some initial assessment. One does not require bionic eyes to see the undercount in 2021. Just consider the case of Hackney, an area of homogeneous haredi settlement in London. In 2021, haredi population of Hackney can be estimated in the region of 32,000.  This estimate is based on three different sources – all external to the Census. The first is the telephone and address directories of haredi households (the so-called Shomer Shabbos directory), used extensively by haredi communities both for internal communication and for outreach and advertising by haredi organisations. The second is statistics of haredi school children, which is a by-product of the national School Census conducted annually in England and Wales by the British educational authorities. The third is the announcements of births celebrations in the haredi community which are routinely published in communal circulars.

 The exact details of the method have been made public in previously. Guess what: 2021 Census records indicate the presence of 17,400 Jews by religion in Hackney, quite far from the anticipated 32,000. If the existence of ethnic Jews is accounted for, then the 2021 Census figure for haredi Jews in Hackney is somewhere in the range of 19,000-24,000. The top boundary of this range (24,000) is obtained if it is assumed that none of those Jews self-identifying as ethnic Jews have marked themselves as Jewish by religion and so all of them should be added to Jews by religion to obtain the full count of Jews in Hackney, a highly improbable situation in itself. This is a demographer’s second Stanislavsky moment with respect to the recent release. For the whole universe of haredi communities in Britain (including haredi residing in Hackney, Haringey, Bury, Salford and Gateshead, but excluding those in Barnet) the estimate for 2021 is about 64,100. In the 2021 Census data, these areas contain 35,600 haredi identified via the religion question and 38,100-45,600, if ethnic Jews are added. These figures suggest a minimal level of 25% undercount of haredi.

The situation, troubling as it is, is not an utterly unprecedented and unheard of, strictly speaking. Undercount of minority populations, especially those with large families and many children, is a documented issue and have been observed in the USA, Canada and Australia, just to name a few. To the extent possible, it should be ‘repaired’ for the data on minorities to become useful. This process has been long under way in Britain: much of the time of scholars of Jewish demography here has spent inventing the alternative population estimation system for haredi. The issue, however, ought to be acknowledged more confidently and widely. The alternative system should really replace the official system which is dysfunctional. Policy makers relying on the official data on haredi communities, i.e. data coming from the Census, ought to switch to the alternative system.

The workings and results of that alternative system have been reported already, and are also reported here in part, and more on that will be published in the following months. It is important to understand that not just that the number of haredi may be compromised but several other indices that are critical to get right for policy planning purposes may be compromised too. If children are undercounted specifically, for example, then fertility figures will be underestimated. This will have a knock-on effect on service provision involving haredi children and women. According to the official estimates, there would be fewer children, and fewer children per woman, than the situation is. Further, the undercount may significantly affect estimates of family size, making them appear smaller than they are in reality, with a knock-on effect on all policies taking into account family size. The consequences of not acknowledging the limitations of the official statistics and unawareness of the provenance and strengths of the alternative estimation systems are high for all involved, haredi policy makers and their counterparts at a national and local level. Last but not the least: the processes that led to undercount of the haredi are expected produce similar effects in other populations with large families, e.g. British Muslims with the same projected knock-on effects in policy.

The correction and the ‘take home’ lessons

So, what is the number of Jews in England and Wales, and what is the likely trajectory in numbers – if one adjusts for the two groups of missing Jews: the ethnic Jews and the haredi Jews? The results of that exercise are captured in a single exhibit.

The real number of Jews in England and Wales is in the region of 314,000. And so, the first ‘take home’ lesson is: there are more British Jews than meets the eye. So far, the story of the 21st century is the story of growth for British Jews. In the last decade, population of Jews in England and Wales grew by about 10%, a rate higher than in general population (6.3%). That is both remarkable and to be expected: British Jewish population possesses a numerically significant and vigorously expanding component that the general British population lacks, e.g. the haredi component. Haredi Jews, a population with total fertility rate of 6-7 children per average woman and high longevity, growing at 3.5%-4.0% per annum, increased their share in Jewish population from 13% in 2011 to about 26% in 2021. Haredi growth critically shapes the trajectory of population growth of British Jews; metaphorically speaking, haredi Jews ‘take British Jewry with them’ on a growth journey. It is equally clear, incidentally, that the original Census figures of Jews by religion paint an illogical picture of growth. Growth of 2.0%-2.4% per decade appears both low and strangely static in view of the existence of the haredi.

This leads us to the second take home lesson: British Jewish population has been growing. In the last decade, probably about 29,000 Jews were added to it. Haredi Jews account for the large part of this growth, probably as much as 90%. This conclusion aligns well with the independently produced British Jewish communal statistics on births and deaths. With zero migration balance, i.e. with the numbers of Jewish migrants leaving Britain and arriving to it being equal, the anticipated addition to the British Jewish population is in the range of 20,000-25,000. As explained in the opening sections of this document, the balance of births and deaths of British Jews is strongly positive and have been that for a while now due to haredi influence. (Parenthetically, the balance of births and deaths of mainstream Jewish population is probably negative). Given that (1) the actual estimated addition to British Jewish population is in the region of 29,000 and (2) the anticipated addition is in the region of 20,000-25,000 as a result of the positive natural balance only, one is led to suspect that British Jewry experiences positive migration balance, i.e. the number of Jewish immigrants to Britain may be higher than the number of Jewish emigrants from Britain. This is simply because more Jews are added than what the natural balance suggests. For anyone who knows anything about the dynamics of the British Jewish community, the obvious candidate countries of origin of Jewish immigrants are Israel and France. Further releases of the Census results relating to the religious and ethnic composition of foreign-born population should clarify this issue.

The finding of growth makes Jews in England and Wales a special case in the European Jewish arena; that would be the third take home lesson. Many Jewish populations in European countries possess delicate natural balances. These are ageing and not very fertile populations. In addition, some are ‘drained’ by emigration. There are some European Jewish populations that are growing, for sure. These are Austria and Belgium, highlighted by the recent research. British Jewish population, at the very least Jews in England and Wales, is part of this family of growers. The defining feature of this family is the presence of the haredi. They generate population growth, to a very large extent, if not exclusively. The position of British Jews as a ‘grower’ population is truly unique in the context of European Jewish ‘population giants’, i.e the largest Jewish populations of Europe: France (above 400,000 Jews), Germany and Russia (each with above 100,000 Jews). In combination, about 70% of Jews in Europe live in Jewish ‘population giants’ mentioned above. The fate of these Jewish populations is very different from England and Wales. French Jewish population is declining due to migration, German Jewish population is declining due to negative natural balance (preponderance of deaths over births). Russian Jewish population is in decline due to both these forces operating in tandem.

It is important to be open about the uncertainty in figures. It exists, hence the fourth take home lesson: these estimates contained in this paper are not final and may be revised somewhat. The weakest element in the re-estimation of the number of Jews in 2021 is the assumption about the ethnic-only Jews. It is assumed that the scope of this phenomenon, in relative terms, is stable and comparable to 2001-2011. It is a reasonably well-supported assumption: in 2001-2011 the proportion of ethnic-only Jews out of all Jews who used ethnic label was very stable. Still, it is an assumption not certainty. Greater appetite for ethnic self-identification among British Jews is evident and it may express itself in many forms. It is possible that the new fashion results in an increase of popularity of the ‘ethnic-only’ option, for example. It is also possible that, in contrast, the ethnic-only option does not change, or even decreases, in popularity but just the size of the population body using the ethnic label alongside the religious label increases. All that may or may not affect the estimations. The effects, if any, will be clarified only when the ONS releases more data from the 2021 Census, specifically on Jews by ethnicity. Dramatic changes are not anticipated. Much of the population growth of British Jews is due to the developments in the haredi sector. That part of the estimation is secure. The changes, when and if they are made, will affect the number of ethnic-only Jews added and the conclusions about the nature of population growth outside of haredi community, and the role migration plays in it.

The era of Relative Simplicity when the number of Jews could be established mostly through counting the ‘Jewish’ responses to the religion question in the census is over. We are now moving into the era of Reconstructive Surgery. Such is the fifth take home lesson. Now, crafting the number of British Jews would require careful work to identify subgroups (e.g. purely ethnic Jews, haredi) that are not visible via the lens of the religion variable, taking care to include all relevant people and avoid double-counting. In truth, this has been in the making for some time. Haredi Jews could not be simply counted but had to be variously estimated using sources other than the Census already in 2001-2011. Today, with the added issue of ethnic Jews, estimation gets trickier. The whole regime of work with the Census data goes a notch up in complexity. It is no longer possible, for example, for the policy makers or experts in Jewish demography, to simply download data containing Jews by religion from the ONS 2021 Census data depository and base their conclusions just on that. Working in this way may mislead considerably with respect to the number of Jews at a given point in time and how that number changed over time, both nationally and locally. The new regime of work would need to be developed. This is not an unsurmountable obstacle. For a long time, Jews in Canada were registered partly as a religious group and partly as a ancestral group in the Canadian Census and experts in Jewish demography had to engage in ‘reconstructive surgery’ as described above. That experience can be built upon, and/but it will take time for full orientation in the new realities to emerge.