Social Science and Consensus in Estimates of the US Jewish Population: Response to Sasson and DellaPergola

In response to Isaac Sasson and Sergio DellaPergola’s commentaries on our assessment of the validity of the Pew Research Center's 2020 estimate of 7.5 million US Jewish adults and children (Tighe et al. 2022), we address key points of agreement and contention in the validity of the estimate; in particular, how the Jewish population is identified and defined. We argue that Pew’s definition of the Jewish population is consistent with major studies of American Jewry, from NJPS 1990 to recent local Jewish community studies. Applying a consistent definition that includes the growing group of “Jews of no religion” with one Jewish parent, as Pew Research Center does, allows for a faithful comparison across national and local studies and a more accurate understanding of levels of Jewish engagement and expressions of Jewish identity.

We appreciate the responses by Sergio DellaPergola (this issue) and Isaac Sasson (this issue) to our assessment of the validity of the Pew Research Center's 2020 US Jewish population estimate (Tighe et al. 2022).At a time when there is intense politically freighted discussion of Jewish identity in both the United States and Israelthe homes of the two largest Jewish communities in the world-academic debate about the demography of US Jewry takes on added significance.The discourse has implications not only for how we understand Jewish life, but also for the policy that engages Jews with their identity.

Sasson
As Sasson noted, estimating the Jewish population of US Jewry through surveys is "messy," but it is also necessary because of the lack of census data and the utility of being able to assess changes in both the size and characteristics of US Jewry.Sasson concurs with our view that new methods, including those employed by the Pew Research Center in developing their 2020 population estimates and their description of the population, helped to overcome sampling and response rate problems.To be sure, identifying US Jews-a relatively small population-is challenging.But the difficulty of the task should not deter us from trying to develop better methods.
Sasson also shares our perspective on how to define Jewish identity.There are multiple possible ways to conceptually think about Jewish identity, and operationally, to measure it.To enable comparisons over time, one needs a consistent definition, but that definition has to address the ways in which Jewish identity and its expressions have evolved over time.We believe Pew's operationalization of Jewish identity serves both of these goals.As Sasson notes, there is empirical support for our position, and it is consistent with contemporary sociological views of racial and ethnic identity.Identity is not a single thing, fixed at birth, but involves multiple strands that may change over time.
Finally, we are pleased that Sasson agrees with our view that to understand Jewish identity, one needs to cast a wide net.This approach allows one, then, to examine, compare, and contrast different ways and intensities of expressing one's Jewishness.To be sure, boundaries are needed.We drew ours at the consensus position developed by Pew that includes secular and non-religious Jews, regardless of whether or not one or both parents are Jewish, and that excludes those who have another religion.It is imperfect, but enables the meaningful understanding of contemporary Jewish identity.

DellaPergola
Although we value the collegial tone of DellaPergola's response, not surprisingly, we take strong exception to his rejection of the growing consensus among scholars of contemporary Jewish life about the size of the US Jewish population and its implications.Our objections to the comments of our interlocutor are rooted in epistemology and norms regarding how to draw inferences from empirical evidence.
DellaPergola claims that our focus is on a narrative of US Jewish life rather than "the social scientific study of [the] Jewish population and society."For us, population dynamics and their use to tell a story of a group of individuals are inextricably intertwined.We employed well-accepted psychometric and research design principles to assess the validity of the Pew Research Center's population estimates.Following Donald Campbell's dictum that "validity is a property of inferences" (Shadish et al. 2002), Pew's use of the estimates as the basis for assessing the intensity of attitudes of US Jews toward Jewish life and their levels of participation in communal life is highly relevant to the evaluation of those estimates.

3
Social Science and Consensus in Estimates of the US Jewish… As such, we systematically assessed both Pew's methodological approach to developing population estimates and their validity in establishing a denominator for comparing the attitudes and behavior of US Jews.We considered several dimensions of construct and concurrent validity and were transparent with respect to the bases for our judgments and how we understand the differences across other available data.In documenting the near-consensus of US social scientists about the current size of the Jewish population, we examined the various approaches to including or excluding particular subgroups of individuals who identify as Jewish, along with the implications of doing so.
DellaPergola's complaint is that by documenting the growing consensus among studies of US Jewry and growing agreement about the boundaries of Jewish identity, we are seeking to silence dissenting voices.The very fact of the present debate suggests that no one is being silenced.To the contrary, our position is offered as a counterpoint to the "official" statistics about the size of the US Jewish population, as promulgated by Israel's Central Bureau of Statistics (CBS).For many years, DellaPergola's analyses have formed the basis for the CBS's world Jewish population report (see, e.g., Central Bureau of Statistics 2022).These counts have not only been promulgated by the Israeli government, but have also been the basis for annual reviews in the American Jewish Year Book (e.g., DellaPergola 2004, 2015, 2017, 2022).
DellaPergola argues that counting US Jews is neither the "most important" nor "most exciting" issue associated with understanding world Jewry.Undoubtedly, numbers inadequately represent the meanings associated with Jewish identity and the factors that shape how individuals engage with their Jewish identities.Estimates of the size of the population, nevertheless, serve an essential function: They provide the denominator that enables assessments of the vibrancy of Jewish identification, the effectiveness of communal policy, and the depth and breadth of Jewish engagement with world Jewry.A denominator helps us understand what is normative and what is unique in the population, as well the areas where Jewish identity is stable and those where it is fragile.Especially important in the contemporary context, estimates provide a basis for seeing the diversity in the Jewish population and the variety of ways in which individuals express their Jewishness.Quantitative indicators of population size also make comparisons across time and groups possible.
There is a long history of Jewish discourse and debate about population counts that began well before the modern era.The Hebrew Bible describes multiple censuses of the Jewish people following their escape from Egypt and 40 years of wandering in the desert.However, textual tradition in Judaism expresses concern about the conduct of censuses and the fear that some members of the community may not be counted or that results may be misused.Contemporary debates continue the focus on the proper method for counting populations.Given that we cannot take a census of US Jews, researchers make statistical projections by various sampling techniques.
Responses to questions about Jewish identity are used to make generalizations about the total population (cf.Saxe et al. 2014).It is inherently complicated, and in the US case, it is made difficult because the government does not collect religious identification information in its census.
The crux of our differences with DellaPergola is about where to draw the line in differentiating Jews from non-Jews.Contrary to his assertion that we advocate only a single way to define membership in the Jewish community, we appreciate the diverse ways in which Jewish Americans relate to their Jewish identities.At the same time, just as he, along with the CBS, reports a single number to represent the Jewish population (the "core" population in DellaPergola's terms), we do likewise.The difference is that we do not selectively exclude people on the basis of their status as offspring of intermarried parents.DellaPergola excludes "Jews of no religion" if they have one Jewish parent, but includes them if they have two Jewish parents.
We believe that our approach-consistent with that of other researchers-is more faithful to long-standing principles about the conduct of Jewish sociodemographic studies.The approach we advocate is based on decades of prior research on US Jewry, in particular national studies carried out since NJPS 1990 (Kosmin et al. 1991), the work of the Pew Research Center (2013, 2021), and local Jewish community studies sponsored by Jewish Federations (see discussions by Aronson et al. 2016;Kotler-Berkowitz 2016;Sheskin 2013).Characteristic of a definition that represents a consensus position, it is a centrist approach and is neither the broadest, nor most narrow, technique for defining the boundaries of who should be considered Jewish.Applying this definition provides a referent that allows for comparison across local and national studies regarding the expression of Jewish identity, the penetration of Jewish education, and the attitudes of a diverse range of Jewish individuals.Absent agreement on a common definition, it would be impossible to compare findings and detect trends.
We concur with DellaPergola that, as social scientists, application of Halacha (Jewish law) cannot be used as the sole basis to determine Jewish identity.Instead, as the basis for determining "who counts," we rely on individuals' reports of their Jewish family backgrounds and their understandings of their Jewish identity.Rather than presuming that we are the "custodians of objective truth," we empower survey respondents to provide definitional responses.These responses are categorized by criteria widely accepted by social scientists of the Jewish community, and they can be reinterpreted by various US Jewish communities' views of membership and belonging.
The disciplinary predicament, as DellaPergola describes it, is perhaps the core epistemological problem.His demographic equation of births minus death, plus immigration minus emigration, plus accessions minus secessions, seems straightforward.It might be if there were official statistics in the United States on Jewish births, deaths, migration, and conversions.There are not, just as there are no official statistics on any religious group.As a result, and as we discussed in Tighe et al. (2022), the methodology used for surveys on which US Jewish population estimates are based have had to use new techniques to improve estimation.Many of these changes have been related to addressing the known bias of sample surveys to underestimate the size of rare groups, particularly the unengaged/unaffiliated.Given the voluntary nature of participation in surveys, it is not surprising that it is easier to identify those who are the most engaged and interested in their Jewish identity.

3
Social Science and Consensus in Estimates of the US Jewish… Further complicating the application of the standard demographic formula is the reality that, rather than a fixed attribute ascribed at birth, being Jewish is a personal, social, and cultural identity, the boundaries of which can change over time.This is especially so for Jews in the United States, who live as a minority in a culture that, despite the recent rise in antisemitism, has been increasingly accepting and positive toward Jews (Pew Research Center 2017, 2019).
Intermarriage of Jews and non-Jews has been especially important to this discussion and has reshaped the landscape of how US Jews identify and engage with the community.Estimates of "accessions and secessions" are particularly difficult, not only because of conversions, but also because of other changes in self-identification.Intermarriage has become more normative, increasing from the 50% or higher rate reported by NJPS 1990 (Kosmin et al. 1991) to 70% among recent non-Orthodox marriages (Pew Research Center 2021).The implications of intermarriage have also changed (Sasson 2013;Sasson et al. 2017).Those born to intermarried parents prior to 1981 were unlikely to identify as Jews when they became adults (less than 30%).However, among millennials (born 1981-1996) born to one Jewish parent, nearly two-thirds identify as Jews as adults.This shift has dramatic demographic implications.
Our methodological approach to those who are born to one Jewish parent forms the basis of DellaPergola's primary critique.He claims that we have upended decades of sociodemographic research by accepting Pew's operational definition of who is Jewish (see Pew 2021, p. 18), but our definitional approach is consistent with earlier studies of the US Jewish population, including DellaPergola's earlier work.
DellaPergola (DellaPergola 2003;Schmelz and DellaPergola 1992) defines someone as Jewish if that individual self-identifies as a Jew and does not have another religious identification.In his early formulation (e.g., Schmelz and DellaPergola 1992), he included "those who converted to Judaism or joined the Jewish group informally" (Schmelz and DellaPergola 1992, p. 486).Later (see, e.g., DellaPergola 2003), he excluded those who do not formally convert.Nevertheless, he maintained Sklare's (1971) position that being part of the Jewish population "does not depend on any measure of that person's Jewish commitment or behavior-in terms of religiosity, beliefs, knowledge, communal affiliation or otherwise" (Schmelz and Della-Pergola 1992, p. 486).In chapters for the 2004, 2014, and 2016 American Jewish Year Book, DellaPergola added criteria to satisfy a mutually exclusive framework based on religion as well as ethnicity (DellaPergola 2004, pp. 493-494;2015, pp. 307-309;2017, p. 258).
In reporting on the Pew 2013 survey, DellaPergola (2015) defined exclusive ethnic Jewish identity based on a broadly worded screener question that included an option for "partly" Jewish.If an individual indicated that they were "partly" Jewish, they were excluded-even through some individuals, when asked, explained that they identified as "partly" not because of mixed ethnicity, but because they were not practicing.The 2020 Pew survey did not include a "partly" option, and Della-Pergola again revised his framework (DellaPergola 2022, p. 356)."Jews of no religion" who were the offspring of one Jewish parent, regardless of whether it was the mother or father, were excluded from his estimation of the core population.This decision seems an incongruous revision to his framework and refutes decades of prior research that DellaPergola claims to be upholding. 1long with rejecting consensus methodological frameworks, DellaPergola also discounts any approaches other than single cross-sectional national surveys as a basis for population estimates.He views summaries of local Jewish population studies (e.g., Sheskin and Dashefsky 2022) as unusable for estimating the size of the population and rejects the use of sophisticated data synthesis techniques to develop estimates.In fact, all three approaches can contribute to developing reliable population estimates.As described in Tighe et al. (2022; see also Saxe et al. 2014), each has advantages and limitations.
With respect to data synthesis, our work is a part of growing efforts across social science fields to use "big data" to provide answers to complex problems (Dutwin 2020).The data synthesis approach undertaken by the American Jewish Population Project (AJPP) (Saxe et al. 2021;Tighe et al. 2019) has been validated by a number of studies.In particular, a study using Canadian census data examined three population estimation methods (our modeling, meta-analysis methods, and pooled designbased cross-survey analysis) for their accuracy in evaluating Canada's Jewish population size (Magidin de Kramer et al. 2018).The accuracy and precision of these approaches were assessed by comparing the estimates of the proportion of the adult Jewish population in the three major Canadian cities generated by each method with benchmarks drawn from the Canadian census.The data synthesis method was also validated by researchers who compared a data synthesis of national surveys to the UK census (Claassen and Traunmüller 2018).These results support the view that single national surveys are not the only vehicle for developing estimates of a rare population such as US Jews.
Our conclusion that the US Jewish population is significantly larger than the 1990 NJPS estimate (by 2 million individuals) seems implausible to DellaPergola, but along with our findings being in line with other approaches to population estimates, it is also the view of other social scientists who have reviewed the evidence (e.g., Rebhun in press).There are multiple reasons, in addition to those elaborated in Tighe et al. (2022), to accept the plausibility of the Jewish population increase.
One possible explanation is that earlier studies underestimated the size of the Jewish population.We are not the first to suggest this theory.Referring to the 1970 Jewish population estimate of 5.4 million, Lazerwitz (1978) suggested that the correct estimate could have been as high as 6 million individuals.Similarly, Kosmin et al. (1991), writing about NJPS 1990's estimate of 5.5 million, was concerned 1 3 Social Science and Consensus in Estimates of the US Jewish… that Jewish heritage was underreported by respondents, and in addition, that recent immigrants had been undercounted.In their report on the 2002 survey of Heritage and Religious Identity, Tobin and Groenman (2003) estimated the Jewish population could be just over 6 million individuals.
Another explanation for the Jewish population increase may relate to the surge in immigration after 1990, in particular from countries of the Former Soviet Union (FSU), but also from Israel and South America (Gold 2015;Sheskin and Dashefsky 2022;Tolts 2020).As DellaPergola notes, the numbers of immigrants claimed to be served by advocacy organizations may be inflated.However, aid organizations, such as the Hebrew Immigrant Aid Society, report only the number of immigrants they aid.As Sarna (2013) has noted specifically with the case of Russian Jews, "the community of Russian-speaking Jews cannot be calculated just on the basis of those who arrived in the United States as refugees or accepted aid from Jewish relief organizations prior to 2001."In addition, many of these immigrants came to the United States after living in Israel or another Western country and may not be included in the counts by aid organizations.
As mentioned earlier, and unacknowledged by DellaPergola in relation to the population increase, is the changing dynamic of intermarriage.As noted above, the impact and ramifications of intermarriage in the Jewish community constitute the crux of our difference in perspective with DellaPergola.That the majority of the children of one Jewish parent now identify as Jewish has far-reaching implications for demographic projections (Saxe et al. 2014(Saxe et al. , 2015)).Contrary to earlier views that intermarriage would result in loss of the next generation of Jews (e.g., Bayme 1991; Phillips 2005), intermarriage may not only lead to more Jewish children, but may also result in the retention of adult Jews who marry non-Jews.It changes the fundamental premises of the demographic formula.Simply put, "demography is not destiny."DellaPergola excludes 1.5 million individuals from the Jewish population who are the offspring of one Jewish parent families but do not identify as Jewish religiously, including some who are halachic Jews (i.e., children of Jewish mothers who claim Jewish identity by long-standing and well-accepted sociological criteria).Excluding these individuals who have traditionally been included in population estimates not only yields a different narrative of US Jewish life than widely understood by most in the Jewish community, but also impedes a comparison of research findings between studies and over time.As illustrated by efforts to understand political behavior of US Jews, the inclusion or exclusion of "Jews of no religion" in an analysis also yields different portraits of US Jewry (Wald and Martinez 2022).
Excluding some Jews creates confusion with broader implications.Thus, for example, the ANU Museum ("we," in Hebrew), the re-envisioned Museum of the Jewish People located on the campus of Tel Aviv University, introduces visitors to the museum with an exhibit on Jewish diversity and the contributions of world Jewry.It includes a video display that depicts the population of the present-day Jewish people by country.As of early 2023, 2 the exhibit led with the Jewish population of Israel (shown as 6.8 million), followed by the Jewish population of the United States (shown as 5.7 million).Apparently based on (now out-of-date) information provided by the Israel's Central Bureau of Statistics, the key problem is not one of underestimation, but rather that the population it encompasses conflicts with the educational message of the exhibit about the ethnic, cultural, and religious diversity of Jews worldwide and their extraordinary contribution to modern life.
The ANU presentation is based on DellaPergola's estimates of what he calls the "core Jewish population" and does not reflect the complexity of DellaPergola's distinctions among various subgroups among those who claim Jewish identity (in particular, those who are part of the "core Jewish population" and those whom he calls "partly Jewish."The idea that some Jews are "full Jews" and others are "part Jews" is counter to long-standing and widely held views of Jewish identity.It is especially problematic for a museum that wants to tell the story of the Jewish people, not just those who are "core."It is particularly problematic when applied to the Pew 2020 data because it selectively excludes secular individuals ("Jews of no religion") who have only one Jewish parent.A different criterion for Jewish belonging is applied to those with two Jewish parents.
To be sure, boundaries with respect to religious and ethnic identification are needed.Our dispute with DellaPergola is about how narrowly these boundaries should be drawn and how we understand the implications of these boundaries.To use the data to understand contemporary Jewry necessitates agreement about who should be included among the Jewish people.Complex analyses of subgroups are essential, but an overall number is needed that encompasses the Jewish population and enables comparisons over time and across communities.Our efforts to describe a consensus in who should be included in the population is designed to promote understanding of Jewish life.
Our focus is on US Jewry, but we appreciate the importance of being able to compare US Jews with those in Israel, the Jewish homeland.We are particularly pleased that the present discussion has engaged us with two Israeli demographers.We would also like to acknowledge the work of Hebrew University sociologist and demographer Uzi Rebhun.His position vis-à-vis the boundaries of who should be included in the US Jewish population parallels ours.In his chapter for the American Jewish Year Book 2022 (in press) marking the 50th anniversary of the first NJPS study, he reviews efforts to estimate the US Jewish population.Although Rebhun focuses solely on cross-sectional national studies, his conclusions are very similar to our own, both about the growth of the US Jewish population and our definition of who should be included.With respect to the Pew 2020 estimate, Rebhun would include most of those whom DellaPergola excludes.His one modification is to exclude the 200,000 children who are being raised Jewish along with another religion.Rebhun's estimate of the total US Jewish population is 7.3 million adults and children.This figure is within the confidence interval of Pew's 7.5 million estimate.

3
Social Science and Consensus in Estimates of the US Jewish…

Conclusion
Among those of us who study and work with US Jewish communities, the boundaries as defined by Pew enable analyses that can inform scholarship as well as practice.Excluding individuals who were Jewish by prior definitions not only makes the portrait of US Jewry less accurate, but undermines our ability to understand developments in US Jewish life.Thus, for example, to view the increasing involvement of families with only one Jewish parent in Jewish education for their children, one would need an estimate of the total number of these families (e.g., for Israel education programs: see, Saxe et al. 2022).Using an inclusive approach for establishing a denominator for understanding Jewish identity and behavior-consistent with past practice-helps further an understanding of many of the issues critical to Jewish communities, both in the United States and around the world.
As social scientists, we seek to understand contemporary Jewish identity and make our work accessible and useful to those concerned with Jewish life.We are pleased that our assessment has provoked commentary by Sasson and DellaPergola that is both supportive and critical.This exchange has reinforced our view that trying to reach scientific consensus is not only an important part of the pursuit of truth, but essential to the understanding of contemporary Jewish life.Indeed, the argument that building consensus silences dissent is fallacious.It functions to sharpen our focus and pushes us to be transparent with the facts and logic that we use to support our arguments.In a disputatious political and social environment, we hope that this discussion models the kind of conversations that are needed.