Contemporary Jewry

, Volume 30, Issue 2, pp 149–153

On Sampling, Evidence and Theory: Concluding Remarks on the Distancing Debate


    • Brandeis University
  • Charles Kadushin
    • Brandeis University
  • Leonard Saxe
    • Brandeis University

DOI: 10.1007/s12397-010-9040-9

Cite this article as:
Sasson, T., Kadushin, C. & Saxe, L. Cont Jewry (2010) 30: 149. doi:10.1007/s12397-010-9040-9


This paper is a summary and discussion of articles contributed to Contemporary Jewry's special issue (vol 30, issues 2/3) on trends in American Jewish attachment to Israel.

We are gratified that our exchange with Steven M. Cohen and Ari Y. Kelman has generated such a substantial response. Our original contribution examined trends over a 20-year period in an attempt to assess what can be gleaned from empirical data about the attitudes of American Jews to Israel. In this brief final comment, we address concerns raised by several contributors regarding the limitations of our survey samples and the relationship between evidence and theory. But before we begin, we wish to identify an area in which there appears to be emerging agreement.

Many contributors (e.g., Cohen and Kelman, DellaPergola, Fellman, Keysar, Phillips, Miller and Dashefsky, Mintz Geffen, Wertheimer) appear to concur with our central finding that there is no evidence of distancing among people who identify as Jewish when asked about their religion.1 Importantly, this population includes many Jews who do not identify with a Jewish denomination and consider themselves to be secular or cultural Jews. We know this, for example, because 28% of respondents to the American Jewish Identity Survey (AJIS) who described their religion as Jewish also defined themselves as secular; an additional 16% identified as somewhat secular (see Perlmann 2009). In the American Jewish Committee (AJC) surveys examined in our initial paper, roughly one quarter of respondents identified as “just Jewish” rather than with one of the religious denominations. The General Social Survey (Davis and Smith 2009), the gold-standard of national attitude surveys, confirms that Jews tend to answer questions about religion differently from non-Jews. According to that survey, 5% of Protestants, 6% of Catholics, and 21% of Jews said they were not religious.2 In short, the category “Jews-by-religion” includes many respondents who do not consider themselves religious.

The emerging consensus that emotional attachment to Israel among American Jews-by-religion remained stable in recent decades represents a modest breakthrough. For many years, social scientists have cited surveys of Jews-by-religion and claimed to find evidence of distancing (e.g. Cohen and Eisen 2000; Cohen and Kelman 2007; Cohen and Wertheimer 2006; Rosenthal 2001; Seliktar 2002). Such distancing has been attributed to well-known sources, including generational distance from the Holocaust, the establishment of the State of Israel, and the Six Day War. More recently, distancing has been attributed to alienation from Middle East violence; opposition to Israel’s policies regarding the Palestinians; disdain for the dominance of Israel’s Orthodox rabbinate and other contemporary causes. Judging from the contributions to this volume, however, to the extent that distancing has occurred, Israel has had little to do with it. Rather, a new narrative has taken shape that attributes distancing to the changing demography of American Jewry, especially growth in the share of the population that identifies as Jewish but not by religion. “Jewish attachment to Israel is not decreasing per se,” states Phillips (this volume), “but the kinds of Jews least attached to Israel are growing in number.” Cohen and Kelman (this volume) estimate this latter group—Jews who “identify as Jewish, but see their religion as ‘none’”—as 20% and growing. But how large is this population, really?

Several studies, most prominently the AJIS, have documented an increase in the proportion of survey respondents who report a Jewish parent but profess no religion. This category is what we meant (this volume) by “Jews-by-ancestry,” and their increase makes sense given the rising incidence of intermarriage since the 1980s. But a key question is what proportion of the “no religion” respondents actually identify as Jewish. The question is important because: (a) the problem at hand—identifying trends in how American Jews feel about Israel—justifies a focus on people who regard themselves as Jewish by some criteria; and (b) those who critique our conclusions describe the population of “Jews, not-by-religion” as comprised of individuals who self-identify as Jewish.

NJPS 2000–2001 (United Jewish Communities 2003) is noteworthy in that it asked respondents who professed no religion but indicated Jewish origins (either a Jewish parent or having been raised Jewish) whether they consider themselves to be Jewish for “any reason.” The study estimates that only a little over one-third (36%) answered in the affirmative. Thus, Jews not-by-religion who self-identify as Jewish for “any reason” comprise about 6% of the total Jewish population, as defined by Cohen and Kelman (i.e., including Jews-by-religion plus individuals of Jewish origin who profess no religion but self-identify as Jewish).3 This population may be increasing due to intermarriage, but not necessarily. According to NJPS 2000–2001, just 20% of mixed origin respondents professed no religion (54% were Christian or some other religion, and 26% were Jewish), and just 24% of those who professed no religion also self-identified as Jewish for any reason (Perlmann 2009). Thus, based on NJPS 2000–2001, just 5% of children of intermarried parents qualify as self-identifying Jews of no religion. Critics of our analyses who believe that the population of self-identifying Jews of no religion is burgeoning due to intermarriage neglect the fact that many children of mixed marriages do not self-identify as Jewish for any reason (ibid). On the basis of available evidence, albeit limited (see Saxe et al. 2007), there is little reason to believe that the population of self-identifying Jews of no religion is rapidly increasing relative to the whole.

We also want to correct another misunderstanding: Several contributors to this issue (Ament and Kotler-Berkowitz, Cohen and Kelman, DellaPergola, Schick) claim that Saxe, Kadushin and colleagues argued that NJPS undercounted the most marginally engaged Jews, especially not-by-religion Jews. In fact, Saxe et al. (2007) argued that NJPS undercounted all non-Orthodox Jews. Moreover, Saxe et al. (2007) also clearly indicated that most not-by-religion Jews counted by NJPS did not consider themselves to be Jewish (p. 11).

More generally, we agree with critics who described the AJC samples as suboptimal, which is why we took pains to describe their limitations and how we arrived at our conclusions. Like Cohen and Kelman, we also believe in the value of theory (and even “side knowledge”). But good science requires that one distinguish between theory and empirical data and, also, that one’s assumptions be identified in analyses and conclusions. An increase in the not-by-religion population is apparently central to Cohen and Kelman’s argument, but it is not discussed in their original “Beyond Distancing” (2007) report and is not supported by specific evidence in their contribution to this volume. We accept that social scientists may wish to advance arguments derived from theory or, even, anecdotes. Theory and conjecture have a place in scientific discourse, but they must be labeled as such and not confused with systematic evidence.

We also take issue with the depiction of our approach as “strict constructionist” in contrast to our colleagues’ “sociological imagination.” For C. Wright Mills (1959), the point of sociological research is to escape the limited vantage point defined by personal networks, daily routines, and conventional wisdom. To argue, as do Cohen and Kelman (this volume) that “[f]or Jewish social thinkers and policy intellectuals … it’s not about the data … but the people whose lives are impacted by them” is to encourage arguments based on personal experience or professional expertise rather than systematic evidence. In contrast, we attempt to limit data-based conclusions to the evidence at hand, and make clear to readers the distinction between inferences derived from data and arguments based on anecdote or theory.

Finally, and most importantly, we agree with Cohen and Kelman, along with a number of the contributors, that we need to look beyond survey data to better understand the changing relationship of American Jews to Israel. Elsewhere, we and our Cohen Center colleagues have investigated trends in philanthropy, Israel travel, political advocacy, Israel education, political culture and the new media (see, e.g., Koren and Einhorn 2010; Sasson 2009, 2010; Sasson and Tabory 2010; Sasson et al. 2010; Saxe and Chazan 2008). This body of work cannot be summarized here beyond stating that the evidence across a number of domains points less to emotional distancing than to new patterns of engagement. In our view, the discourse on distancing has become a distraction. Setting it aside, unless and until evidence suggests otherwise, can open up a new terrain of inquiry into how the Diaspora-homeland relationship is changing.


The AJC has completed two new surveys during the gestation period for this special issue, and the trend for the “close to Israel” question has been upward. In 2009, 69% of respondents reported feeling close to Israel, up 2% from the previous year. In 2010, 74% reported feeling close to Israel. The downward slide noted in our original paper therefore reversed, and the 2010 level is near the 20-year high.


Protestants, Catholics and Jews were identified with a question about religious preference. Two survey years were combined and samples were weighted. The difference between Jews, Protestants and Catholics is statistically significant at the 0.05 level. Analysis by Kadushin.


Sixty-five percent of NJPS respondents were Jewish-by-religion; 12% were of Jewish origin but professed no religion (Perlmann 2009, Table 3). As noted, 36% of the no religion group—or 4% of all respondents—self identified as Jewish “for any reason.” Therefore, self-identifying Jews of no religion comprise 6% of the total Jewish population [4/(65+4)], when the latter is defined to include Jews-by- religion plus individuals of Jewish origin who profess no religion but self-identify as Jewish.


Copyright information

© Springer Science + Business Media B.V. 2010