The Need for a Systematic Comparative Approach to National Population Surveys of Jews. Barry A. Kosmin

Transcription

1 KOSMN 33 The Need for a Systematic Comparative Approach to National Population Surveys of Jews Barry A. Kosmin National counts of Jewish populations have been fraught with unpleasant consequences for their sponsors since biblical times(see Samuel 24: 1-18). n those times they were unpopular because of the belief that unfair advantages and dangerous unnatural powers were conferred on the initiators of such projects. n the contemporary academic world, social scientists are taught that surveys have to be approached with caution because they famously are prone to error because of bad or fluctuating design, discrepancies in samples, and poor execution (Kosmin, 1979). Unfortunately, the National Jewish Population Surveys of and brought little naches (satisfaction) to their sponsors in part because they took place in an intellectual vacuum and ignored the historical and international comparative framework that should underpin all major social scientific studies of the Jews. Their sponsors also failed to recognize that the essence of science, especially as applied to national baseline data collection, is replicable data with standardized, and detailed classification rules applied consistently. Rabbinical and biblical literature, as well as gentile authorities. traditionally viewed the Jews both as a nation and as a religious community. After Emancipation during the 19th century, this fabric of unity began to unravel. n Western Europe, some Jews chose to define themselves solely as a religious group, eliminating the national aspect. n Eastern Europe, particularly in Russia, and also to some extent in the Austro-Hungarian Empire, Jewishness was expressed by modernizers, such as the Bundists and Zionists, as a secular national category comparable to other nationality groups dwelling in these multi-ethnic empires. nternationally, official Jewish population and social statistics reflect these historical developments. For instance, for more than a century the Jews of Switzerland and reland have been recorded as a religious or confessional group in their respective national decennial censuses. By contrast, the Jews of the USSR/CS were recorded as a nationality on the basis of descent in censuses that provide a similar rich time series. North American Jews are heirs to three traditions: the pre-modern religion-nation; the Western modernizers who defined themselves as a group with a distinct religion and who maintained the nationality of their host country; and the East European modernizers who defined themselves as a secular nationality on the basis of Yiddish or Hebrew ~ culture. Though largely of Eastern European stock, American Jews live. in a society similar to Western Europe, a society of unitary nationality but with multiple religious groups. Contemporary Jewish identification ~

2 34 CONTEMPORARY JEWRY is problematic because self-definition must take into account this historic memory of varying criteria as to what constitutes membership of the Jewish collectivity (Ritterband, et.al., 1988). Furthermore, Jews are considered one of America's three (now perhaps four) great faith groups. Yet nearly all studies of religiosity show that Jews are more secular in their observance than Protestants or Catholics (Kosmin & Lachman, 1993; Kosmin, 2002). However, America has provided for the voluntary construction of another variation of the national option suited to modern pluralistic societies. This comprises ethnic identities based upon national descent groups-hyphenated Americans with special cultural traits. The failure of the official cognitive system, the U.S. Census, with the tacit support of Jewish agencies, to allow the recording of Jewish ethnicity is a major failing (Lieberson & Walters, 1988). t works to depress national Jewish population counts because of the signal it sends discouraging positive Jewish identification among non-religious people of Jewish background. This is especially problematic given the findings of the 1990 NJPS. These findings showed: The high level of secularization among American Jews is reflected in the fact that fewer than half include religion in their definition of what it means to be a Jew, and only a tiny number believe its meaning is solely that of being a religiously based group. (Lerer, Keysar & Kosmin: 60) This is not a new phenomenon. t was recorded and recognized by the founders of Jewish social statistics, such as Arthur Ruppin in the bulletins of Zeitschriftfiir Demographie und Statistik der Juden ( & ) and in Yiddish by Jacob Lestchinsky in the Bieterfar Yiddishe Demografiye, Statistik un Ekonomik ( ). This recognition that Jews were essentially a sui generis, "ethnoreligious" population fortunately was recognized in North America in Canada's official census statistics. t is interesting to note that this took place within Canada's official policy of multiculturalism, with its echoes of Habsburg Europe. Jewish demography owes a debt of gratitude to the analytical work of Louis Rosenberg at the Canadian Jewish Congress and, more recently, to Jim Torczyner. Their lobbying efforts secured a Jewish category in both the religion and ethnic-origin questions of the Canadian Census and with it a long-term time series ( ) of Canadian Jewry. These two questions thus combined the two modem European Jewish identity traditions and so allowed the production of a more inclusive (and accurate) count of the Jewish population (see below) that allows secular and cultural Jews to identify themselves on their own terms., f ' KOSMN 35 The research design and analytical categories (Jewish identity constructs, e.g. Jews by religion, Jews by choice, Jews of no religion, etc.) in the Highlights report of the 1990 National Jewish Population Survey (Kosmin, et.al., 1991) illustrated an awareness of the intellectual and social science tradition within which we worked ( was a member of that team). t also reflected a felt political imperative to give a full and inclusive account of the Jewish population within the overall national context (i.e. the CUNY National Survey of Religious dentification). These same motivations lay behind the more recent decision to undertake its replicate, the CUNY American Jewish dentity Survey 2001 (Mayer, Kosmin & Keysar). My late colleague Egon Mayer (z'l) and our sponsor Felix Posen both shared these ideals and a deep knowledge of the European Jewish experience and of our intellectual roots. We believed in our research that we had a responsibility to record objectively all the types of Jewish identity and ties that exist in contemporary American society, as well as to create a time series with the potential for international and historical comparability. Furthermore, the study of contemporary American Jewish populations is not just a matter of demographic methods and statistical precision. Considerable sociological imagination is required to do an, adequate job. Many of the conceptual and practical problems that arise in an attempt to identify a sample of Jewish Americans are not unique to the Jewish group. They arise whenever one asks people in an individualistic, privatized, and pluralistic society to assign themselves to a specific group at the expense of alternative identity options. As Mayer, Ariela Keysar and pointed out: Particularly in an environment where individuals may hold multiple notions of self, and hold membership in multiple, noncontinuous communities and associations, establishing any fixed notions of identity is problematic. One of the hallmarks of contemporary American society in particular is that individuals can lay claim to a variety of identities, like so many "screen names" in cyberspace, with varying degrees of commitment to each. The relative salience of these diverse identities can fluctuate with the psychic economy of the individual as a result of evolving circumstances. n such an environment, it becomes difficult to speak of anyone's identity as a permanent fixture ofthe self. (Kosmin, Mayer & Keysar, 2001: 31) This type of sociological and societal awareness informed NJPS 1990 too. Sidney Goldstein and suggested with regard to that multi-stage study: \

3 36 CONTEMPORARY JEWRY The screening process used in sample surveys, though scientific in method, is basically subjective in nature. We use questions involving terms or groups as keys in order to unlock doors, but we cannot predict who will enter. There are no correct answers in a fluid and dynamic society. nstead, as in this case, we are able to rely on measuring the variation in responses across groups; and across time when we have the luxury of multiple screenings. (Goldstein & Kosmin, 1992:242) Jewishness is hard to define today because it is multifaceted: There are different Jewish populations for different purposes. How one views what makes a Jew-religious belief and practice or descent and biology-obviously affects one's opinion of how individuals enter or leave the group. Orthodox Judaism's rules and criteria on personal status are commonly cited as the normative system but they are challenged today from two powerful directions. n theory, a faith community is open to those who want to believe and practice, including converts. t expels and disavows heretics and both those who deny the existence of the God of srael and those who adopt another belief system. Reform Judaism, which now has a plurality among American Jewish denominational preferences, has adopted this approach as its standard for membership. ts approach holds that beingj;aised a Jew is more important than biological descent, so it is irrelevant whether the Jewish parent is the mother, and converts are welcome. The nationality principle is linked to biological descent. From this perspective, even the Jewish neshama (soul) is seen more as a product of nature than nurture. f Jews are a descent group then it is both difficult to join or leave the group. f Jews are a nationality, they can be like Frenchmen or Hungarians, of any or no religion. They still remain Jews. The Nuremberg Laws applied this model and there are echoes of it in the State of srael's Law of Return (particularly as it is now applied to the Ethiopian Falash Mura) and the programs of the Russian-oriented political parties in srael. This intellectual baggage is significant today for two reasons. First, there is no consensus across the Jewish world as to which membership criteria are paramount. Second, what was largely theoretical 40 years ago for NJPS 1970 and affected only a few cases, affects hundreds of thousands of people today due to vast increases in the rates and numbers of intermarriages over the last few decades. There is also the question of when the inclusion of peripheral populations or, as U.O. Schmelz and Sergio DellaPergola termed them, the "extended" and "enlarged" Jewish populations, make theoretical and practical sense (1996: 437). For political purposes such as in regard to ri'. t.1 ~ ~, f P KOSMN 37 anti-semitism, where most of the perpetrators use a wide definition of their target population, then the Jewish community too should be more inclusive. The same thinking should apply on the analytical level for social indicators. t would be inaccurate and unhelpful to exclude gentiles from analysis of household composition, since it would exaggerate the number of one-parent families. The same logic applies to economic and occupational data. Gentile sources of household income cannot be excluded without making the data meaningless. Other cases involve close judgment calls as to the relevance of including or excluding certain sections of the total household population. An assessment of what constitutes the "population at risk" may well involve ideological assumptions, but that is inevitable. The key responsibility of the data collectors is that they construct the research design and the categorization of identity constructs and sub-populations so that such decisions are possible both in theory and practice. To this academic mix can be added societal realities related to the practical questions of recognition in everyday life. They operate irrespective of personal self-identification or elite categorization. Who does the average person in the street or at the Hillel house regard as Jewish? s it Ruth Goldberg whose mother is Lutheran and father is Jewish or Sean McCarthy whose mother is Jewish and father Catholic? And what do we do with William Brown, who has four Jewish grandparents but holds to no religion, yet attends a Buddhist meditation group episodically? What do the grandparents of each person think about their descendant's Jewish status? Who is likely to be solicited by the campaign of the local Jewish federation? Since one person's Jew is, literally, another person's gentile, NJPS 1990 offered a range of identity constructs. These enabled different analysts and Jewish agencies to decide for themselves who is to be included or excluded from their "Jewish population" according to their agenda, mandate and criteria. We made sure that these identity types were carefully constructed and nested so they could fit or be slotted into the categories existing in previous national censuses and surveys, as will be demonstrated below for Canada and the United Kingdom. Unfortunately, NJPS failed to apply a similar approach with its Sample Allocation Codes (SAC), which were unique to that study. Moreover, the 1990 survey's identity constructs were created by analysts ex post facto-leo after interviewing was completed on the basis of recoverable data of household rosters and following two or three screenings of individuals using a detailed religion item This allows anyone to examine or reassign identity types. By contrast, the 2001 SACs were allocated by computer program during the fieldwork phase-leo the screener operationally defined who was a Jew in order to define intermarriage. n addition, whereas all households and individuals whatever their Jewish

4 38 CONTEMPORARY JEWRY identity type were administered the same questionnaire in 1990, in 2001 the SACs were used to allocate short or long surveys using a priori definitions. This means that analysts who disagree with the initial allocation cannot view a full response from peripheral households and individuals in order to make up their own minds on the validity of the identity ascribed. n contrast to the iconoclastic and mechanistic approach of NJPS , the American Jewish dentification Survey 2001 further developed the 1990 refinement of identity constructs: Jewish parent: Religion Judaism (BJR), No Jewish Parent: Religion Judaism (JBC), Jews by Religion (JBR), Jews of No Religion (JNR), Jews of Other Religions (JOR). This produced a whole array of carefully defined Jewish populations, all of which can be aggregated up or pared down to produce the "number of Jews" one requires: adult population; child population, core Jewish copulation (5.34M); halachic adult population (3.61M), Jewish and Jewish origins population (7.69M), total American Jewish & kindred population (9.85M). t is my contention that the NJPS was an unfortunate regression in Jewish population studies because it largely ignored many of the issues have touched upon above. t was predicated on trying to fit Jews into a faith model. This claim relates to a succession of research design, fieldwork methodology and analytical decisions that had a cumulative effect on the enterprise. Mark Shulman, a distinguished survey methodologist wljo undertook one of the two UJC-commissioned independent audits of the 2000 study, discovered a series of methodological flaws and statistical problems that led him toward the same conclusions. He is reported as stating that the survey inflated "the proportion who are most religiously identified," and there was a "skew toward Jews who are more religiousiy identified".1 Certainly, the number of secular and cultural Jews, or JNRs, reported by NJPS was several hundred thousandpersons fewer than reported by AJS n order for the casual reader to understand how this situation occurred and to fully comprehend the heated debate and controversy concerning NJPS 2000 it is necessary to recall its history. NJPS 1990 was criticized on two main counts. First, it was criticized for publishing an intermarriage rate for individuals of 52% for the period that related to definition of the population at risk-i.e. the validity of using the authors' criteria of anyone born or raised Jewish rather than those just currently Jewish. The second criticism was the claim by some Orthodox leaders that the 1990 study had somehow undercounted Orthodox Jews.2 n order to meet the first criticism, the Sample Allocation Code system was adopted. n order to meet the second criticism, the decision was taken to dispense with the multi-stage 1990 methodology, KOSMN 39 which relied on a using the EXCEL omnibus survey to screen households and accumulate samples over a 12-month period. t was claimed that this design led to unacceptably low response rates, especially among "traditional" Jews. nstead, a "boutique" (i.e. proprietary) custom-made survey was commissioned with a single-stage design, combining screening and full interviews. t was also hoped that this would speed up the fieldwork, which was an issue as the survey was geared to have almost twice as many respondents as the 1990 survey. The National Technical Advisory Committee of the CJF/NJPS planned to begin the fieldwork stage in January 2000, but unfortunately its start was delayed for eight months while negotiations went on to change the design. This hiatus was due to external pressure on the UJC from donors reacting to a lo-page letter from an alternative group of social scientists that wanted questions amended and the overall approach of the study made more practical and relevant, according to their perspectives. Eventually, the 2000 survey went into the field in August 2000 with an amended and lengthened questionnaire, but it ran into difficulties because it was hard to get enough respondents. After a year, in April 2001, there were only 3,100 interviews, rather than the 4,500 planned.3 This situation led to the fieldwork being extended, including an additional delay, while a further $600,000 was raised for fieldwork costs. The fieldwork finally was completed in August 2001 after 18 months in the field-hardly the speedy operation originally envisaged. However, this crisis necessitated a number of decisions that undermined the integrity of the survey process. A decision was made to pay a small inducement for participation partway through the survey, thus creating two classes of respondents. t was particularly difficult to meet the Orthodox "quota," so in this case a "snowball" or multiplicity technique was adopted to recruit this sub-sample. Respondents were asked to identify other Orthodox households, but the downside of this approach is that it is then very difficult to calculate accurately selection probabilities. These features, combined with the low cooperation rate and the stratified sampling design, mean that there are real statistical problems that cast doubt on the claim that this is a truly random national sample survey. Undoubtedly, intricate and complex weighting problems were created. As a result, there are grounds for claiming that NJPS was seriously flawed in both its research design and conceptual and analytical approach to the enumeration of American Jewry. This criticism has nothing to do with the saga of the "lost data" and other practical problems that dogged its progress, though they were undoubtedly linked. The elongated period of the fieldwork means that analysis of certain demographic events is complicated. There are also practical problems that tend to invalidate some of the data, such as the fact that attitudes

5 40 CONTEMPORARY JEWRY toward srael were collected from some respondents during the Oslo period and from others after the outbreak of the second intifada. believe a clear error of judgement by the National Technical Advisory Committee caused some of the above problems. This error was to ignore the lessons learned from the 1970 survey-for instance, the decision to adopt a geographically stratified sampling strategy based on assumptions about the national distribution of the Jewish population. n my lead article in the Journal of Jewish Communal Service special issue on the CJF 1990 National Jewish Population Survey (NJPS), made note of the 1990 team's determination to avoid repeating the mistakes of the 1970 NJPS: t is widely accepted that this first NJPS was not a wholehearted success. t was based upon face-to-face interviews and a very complex sampling methodology that relied on local federations' lists. t ran afoul of poor quality address lists, budget overrun, supervision problems, and eventually dissension among the academics in the research team. (Kosmin, 1992A: 293) With the full support of the professional leadership of at the Council of Jewish Federations, NJPS 1990 operated within the parameters of liberal Western academic social-scientific scholarship. in practice, this means that, in keeping with current polling standards, we tried to keep our questions unbiased or value neutral, and respondents were provided with the opportunity to give candid answers, rather than normative ones. Self-reports of Jewish identity were accepted, and we stressed to both our interviewers and our respondents our desire to seek the truth, rather than what they thought we wished to hear. t was logical that we had a moral imperative to continue on this track when we have reported the findings. The lay and professional leadership of CJF were committed to academic freedom and full reporting of the results. (Kosmin, 1992A: 296) The 1990 research team (of which was part), also felt that we were fully transparent when we made it clear to our audience from the outset: Definitions of units of analysis, such as whom you regard as Jewish and at what point in time, now or at birth, can all affect the findings. Moreover, important statistics, such as intermarriage rates, are not only affected by definitions about Jewish- KOSMN 41 ness but also about marriage. s analysis restricted only to legal marriage and then just to first, not subsequent marriages? f any of these variables is altered, even before one introduces controls, such as sex, and age, differing and possibly confusing but still accurate rates of intermarriage can be produced. (Kosmin, 1992A: 297) This should not have been news for anyone who knew the essential literature on Jewish demography since it was pointed out by Lestchinksy in his pioneering study of conversions in Prussia in Nevertheless, there was an orchestrated campaign to undermine the reputation of the CJF's 1990 study because of its reported intermarriage rate. This attack eventualz even led to ridiculous but defamatory claims of "falsification of data." This unacceptable level of criticism occurred despite the fact that at the outset of the analytical phase Sidney Goldstein and acknowledged that our approach was open to debate. We presented a paper at the annual conference of the 1991 Population Association of America (PAA) that demonstrated how different Jewish populations with very different social characteristics emerged using different identity points in the initial screening process (Goldstein & Kosmin, 1992). My skepticism with regard to a highly stratified sampling design for a national survey was a direct result of my assessment of the problems associated with the 1970 NJPS (Massarik & Chenkin, 1973). firmly believe we have to adopt simple and straightforward equal-probability sampling strategies when we deal with a largely unknown population. Making assumptions about where Jews reside by relying on Jewish communal organizations' administrative records is a formula for failure. Both the 1970 and 2000 studies ran out of money because they made ill-informed and thus costly assumptions about the geographical distribution of the population. A truly scientific approach makes no a priori assumptions and gives every unit in the universe equal probability of inclusion. That is the approach that was adopted for the 1990 NJPS and AJS 2001, where the universe sampled was all telephone households in the continental United States. Let me provide a contemporary comparison as an exemplar for the need to avoid making assumptions about contemporary Jewish populations. The 2001 U.K. Census included a voluntary religion question for the first time (Kosmin, 1998). The national total reporting Jewish was 268,000, which was very close to the Board of Deputies of British Jews' 1995 national estimate of 285,000 (Schmool & Cohen, 1998). Since one must assume an undercount for refusals and "no religion" responses, it immediately appears that the official community underestimated its constituency. Nevertheless, one might expect that the characteristics of the Jewish religious population should be well known to Jewish com-

6 42 CONTEMPORARY JEWRY munal authorities in a geographically bounded country like Britain with low rates of both immigration and regional migration and a relatively centralized synagogue system with a long history of synagogue membership counts. Yet the census showed that the Jewish population was much more widely spread than community leaders had thought and in fact Jews were found in all but one of the country's local government units. More central to my argument is the fact that 14% of the reported Jewish population did not live in previously known concentrations of Jews. Whereas the Board of Deputies estimated 8,350 Jews living in peripheral rural areas, the Census recorded 38,470. They were off by a factor of nearly five (Graham, 2003). My next criticism of NJPS is that it precludes the option of comparative analysis. One way to check the validity of our survey results and analytical models is to see how they stand up against more robust population counts. The most obvious control for the United States is the Canadian census. We were able to do this in 1990 because our identity constructs closely matched the categorization used for the analysis of the 1991 Canadian Census findings for the religion and ethnic-origin questions as shown in Table 1 (Torczyner & Brotman, 1995). Once allowance is made for the substantial differences in the size of the two populations, the patterns of Jewish self-identification are very similar. Among the "core" populations, the proportion of people who identify as Jewish both by.ethnic origin and by religion is almost identical (79% and 76%) in Canada and the United States. The data also suggests that Canada has proportionately fewer secular or cultural Jews than the United States. This is a finding of important sociological value. The proportion of Jews by choice (gentile ancestry) is exaggerated in the Canadian case by the categorization system, which in this regard is not compatible between the two national systems. Even though they may report more than one origin, some Canadian Jews prefer to identify ethnically with the original homeland of their ancestors (e.g. Hungary, England, Morocco) rather than as Jews. The result is that there probably is a greater proportion of respondents counted as Jewish solely in terms of religion than there are in reality. The lower half of the table brings into the statistics the peripheral population of Jews who self-identify as Jewish in terms of ethnic origin and ancestry but presently follow another religion, such as Buddhism or Christianity. This diverse population consists of apostates who recently have left Judaism as well as those never raised as Jews, who have had little personal contact with the Jewish community in any form but are the children and grandchildren of Jews. As one might expect, given the greater proportion of third- and fourth-generation Jews in the United States, this fringe is relatively larger among Americans. The difference, however, is not large. The statistical data suggest that the societal pro- KOSMN 43 Table 1 U.S. and Canadian Jewish Populations by dentity Constructs Population Group US Canada (Definition*) 1990 NJPS 1991 Census JEWS BY RELGON Jewish Ethnic Origin BJR(US) JRJE (Canada) Other Ethnic Origin JBC (US) JROE (Canada) JEWSH BY ETHNC ORGN 4,210, , ,395,000 No Religion JNR (US) 1,1200,000 NRJE (Canada) TOTALCOREJEWSH POPULATON (CJP)(US) 5,515, , ,070 JSD (Canada) 356, '"JEWSHETHNC ANDRELGOUSPREFERENCEPOPULATON Jewish Ethnic Origin/with Other Religion JCO/JOR/JCOR (US) 1.325, ORJE (Canada) Core/Standard Jewish Population CJP(US) 5,515, JSD (Canada) 356, , TOTAL JEWSH & JEWSH ORGNS POPULATON 6,840, , '" Abbreviations, in the order in which they appear, are as follows: BJR Born Jewish: Jewish by JSD (Canada) Jewish by religion. or Jewish by religion religion and Jewish by ethnic origin, or no JRJE Jewish by religion and Jewish religion and Jewish by ethnic origin by ethnic origin JCO Adult born/raised in Judaism. converted to JBC Jewish by choice another religion JROE Jewish by religion and other JOR Adult of Jewish ancestry, born/raised and than Jewish ethnic origin currently of another religion JBR Jewish by religion JCOR Child under 18 being raised in another JNR Secular Jews. no religion religion NRJE No religion. Jewish by ORJE Other religion and Jewish by ethnic origin ethnicity only (see note ) CJP Core Jewish Population

7 44 CONTEMPORARY JEWRY cesses in both nations are similar and that the resulting Jewish identity constructs run along similar lines. Canadian Jewry is indeed more traditional or normative than American Jewry at present, but the gap is not wide. Finally, would like to return to the United Kingdom in order to demonstrate why we must cast a wider net socially as much as geographically in our data collection than the study attempted. Social scientists have an obligation to record and understand the social processes that Jewish communities are undergoing in the contemporary world, particularly in periodic national baseline studies. The voluntary census question in the United Kingdom incorporated a variant in Scotland, which has its own independent government census operation. Thus, in Scotland two religion questions were offered, one relating to the past and one to the present: CUrrent religion: Jewish (6448) 1000/ct- " "-, '. KOSMN 45 Chart Census Results: Jews in Scotland (percentages are of current Jewish population) Total accessions (converts) Upbringing: Not Jewish or refused CUrrent: Jewish (787) 12% What religion, religious denomination or body were you brought up in? What religion, religious denomination or body do you belong to? The wording of the questions could be interpreted as biased toward eliciting membership rather than merely religious identification. Certainly, the proportion stating no religion was much higher in Scotland at 27% than that obtained by.a simpler What is your religion? version used in England and Wales, which produced 14% of No Religion responses. However, for Jewish purposes the Scottish findings (see chart) provide considerable insights into identificational processes occurring in a small but relatively prosperous English-speaking nation with well-documented synagogue membership statistics: 1,952 household membership units in 2001 in 11 congregations of which 10 were modern Orthodox (Schmool & Cohen, 2002). The rich data trove provided by the two-stage questioning also is immediately obvious. The most striking feature is the "churning" of the Jewish population. Whereas 6,448 persons are currently Jewish, 7,446 persons were raised Jewish. This means that secessions from Judaism far exceed accessions by 1,785 to 787 persons. Nevertheless, the number and proportion (12%) of entrants (converts) far exceeds the number the community had thought. The three-part Ever-Jewish population, which includes Jews by upbringing and converts, totals 8,133 (7, ) persons, whereas the permanently or Always Jewish population numbers only 5,661. This means the Ever-Jewish population exceeds the Always Jewish population by 30%. These findings show that the loss to other religions (lco) is 10% of the raised Jewish population and 8% of the Ever Jewish. The chart also Total secessions (losses) Upbringing: Jewish CUrrent: Not JeWish or refused (1785) 28% Source: Scotland's Census Upbringing: Jewish (7446) 115% Registrar General's 2001 Census Report to the Scottish Parliament, Table 9, p 31. shows that the Secular or Cultural Jewish population, which define as those who now report No Religion or Refuse the current question is about 18%. Equivalent statistics for Jews by Religion in the United

8 46 CONTEMPORARY JEWRY States are provided in the report of the CUNY American Religious dentification Survey n the United States, adult accessions or "inswitchers" are estimated at 171,000,or 6% of the current adult JBR population, whereas secessions from Judaism are estimated at 291,000- though in this case the majority went to the No Religion (JNR) category. (Kosmin, Mayer & Keysar, 2001: 23). These comparative statistics seem to suggest a lower rate of churning in the United States than in Scotland, though the ratios among the sub-categories and the overall trajectories are similar in direction. This type of information is essential for understanding the social dynamics of contemporary society, and it is the responsibility of Jewish social science today to strive to obtain and fully analyze such valuable and intriguing data. Unfortunately, as suggested earlier, as a result of weaknesses in design and Sampling Allocation Code (SAC) issues, NJPS cannot readily provide equivalent data. KOSMN 47 work that acknowledges that Jews are a world people with a distinguished social science literature. NOTES 1Jewish Telegraphic Agency, September 29, Jewish Telegraphic Agency, February 4, Jewish Telegraphic Agency, April 2, 2001 and The New York Times November 30, Forward, September and The New York Times September 12, t is the duty of social scientists studying Jews to be loyal to their disciplines and training. f necessary, they must challenge the prevailing assumptions of Jewish leaders and communal institutions and certainly they must not to be their ciphers. We have to recognize local "demographic studies" are commercial undertakings sponsored by fund-raising bodies: the local Jewish federations. The natural tendency of the researcher who is commissioned to undertake them is to do the bidding of their client and to do it as quickly and efficiently as possible so that costs are controlled. Sidney Goldstein and have explained why we believe such studies, though valuable for local institutional purposes, are not of good enough quality to serve as national baseline data even if aggregated (Kosmin & Goldstein, 1998). A decennial national survey must be analogous to the national cens\js, which is generally not seen as the plaything of the governing political party or any faction or interest group. Unfortunately, the American Jewish communal system does not seem to be mature enough to recognize that requirement. Hence, we have the rather recent calls by some of our colleagues to transfer this responsibility to the public or government sector and to lobby for the inclusion of a religion question in the U.S. Census. This would be welcome, though doubt it is a viable political possibility. However, the data have presented clearly show that a question on current religion would only supply a partial and therefore inaccurate picture unless it is combined with an ethnic origin question and, preferably, a religious upbringing question. Only by adopting this comprehensive approach to data collection could we properly measure population dynamics. t is imperative that we enumerate the large and growing proportion of "no religion" cultural Jews and calculate the losses to the core Jewish population through defections to other faiths and religions. We must also place contemporary American Jews within a valid comparative frame-! tn r J ~ ~ ~,! t,.,f&

9 48 CONTEMPORARY JEWRY REFERENCES Goldstein, Sidney and Barry Kosmin "Religious Self-dentification in the United States : A Case Study of the Jewish Population," Ethnic Groups, 9, Graham, David "The Geographical Distribution of the UK's Jewish Population in 2001," Unpublished Working Paper, JPR: London. Kosmin, Barry A "Demography and Sampling Problems." n D. Bensimon (Ed.) Communautes Juives ( ): Sources et methods de Recherche ( ). Paris: nstitut National des Langues et Civilasations Orientales A. "Counting for Something: The Why and Wherefore of the CJF 1990 National Jewish Population Survey," Journal of Jewish Communal Service, 68 (4), B, "The Permeable Boundaries of Being Jewish in America," Moment, (August) , "A Religion Question in the British Census?" Patterns of Prejudice, 32 (2)~ , "As Secular as They Come," Moment (June). Kosmin, Barry A. & Sidney Goldstein "Why Local Surveys are Overly Optimistic,"New YorkJewish Week, Kosmin, Barry A., Sidney Goldstein, Joseph Waksberg, Nava Lerer, Ariella Keysar and Jeffrey Scheckner Highlights of the CJF 1990 National Jewish Population Survey. New York: Council of Jewish Federations. KOSMN 49 Lieberson, Stanley & Mary Walters From Many Strands: Ethnic and Racial Groups in Contemporary America, New York: Russell Sage Foundation. Lerer Nava, Ariela Keysar & Barry Kosmin "Being Jewish in America: Religion or Culture?" n David M. Gordis & Dorit P. Gary, eds., American Jewry: Portraits and Prognosis (pp ), West Orange, NJ: Behrman House. Massarik, Fred & Alvin Chenkin "United States National Jewish Population Study: A First Report," American Jewish Year Book 1974, Philadelphia: Jewish Publication Society. Mayer, Egon, Barry A. Kosmin & Ariela Keysar American Jewish dentity Survey 2001, New York: Graduate Center CUNY. Ritterband, Paul, Barry Kosmin & Jeffrey Scheckner "Counting Jews: An Essay in Methods," American Jewish Year Book 1988, ( ), New York: American Jewish Committee. Schmelz, U.O. & S. DellaPergola "World Jewish Population, 1994," American Jewish Year Book 1996, ( ), New York: American Jewish Committee. Shmool, Marlena & Frances Cohen Profile of British Jewry, London, Board of Deputies of British Jews British Synagogue Membership 200, London, Board of Deputies of British Jews. Torczyner, Jim L. & Shari L. Brotman "The Jews of Canada: A Profile from the Census," American ljewish YearBook 1995, ( ), New York:American Jewish Committee. Kosmin, Barry A. & Seymour Lachman One Nation Under God: Religion in Contemporary American Society, New York: Harmony Books. Kosmin, Barry A., Egon Mayer & Ariela Keysar American Religious dentification Survey 2001, New York: Graduate Center CUNY.