Reconceptualizing religious change: ethno-apostasy and change in religion among American Jews

Sociology of Religion, Winter, 2006 by Benjamin T. Phillips, Shaul Kelner

H6: Ethno-religiously heterogeneous friendship networks will be positively associated with ethno-apostasy among American Jewish adults.

Among the American Jews most involved in non-Jewish religious communities are those raised in households where Judaism is not the only religion practiced. We hypothesize that, by integrating Jews into non-Jewish religious environments, early exposure to multiple religions will increase the likelihood that American Jewish adults will adopt a non-Jewish religion. It would lower the entry costs associated with acquiring cultural competence, constitute a set of cross-cutting relations that may offset the social pressures and incentives to remain within the Jewish fold, and increase the chances that the religious preferences will tend in a non-Jewish direction. We also hypothesize that it will increase the prevalence of ethno-apostasy by weakening the exclusivity of the hold that either form of religious identification, Jewish or non-Jewish, has on the individual.

H7: Mixed Jewish and non-Jewish religious upbringing in childhood will be positively associated with the adoption by American Jewish adults of a non-Jewish religion.

H8: Mixed Jewish and non-Jewish religious upbringing in childhood will be positively associated with ethno-apostasy among American Jewish adults.

METHODS AND DATA

National Jewish Population Survey 2000-01

This paper uses data from the National Jewish Population Survey (NJPS) 2000-01 and its partner study of households without any tie to Jewish identity, the National Survey of Religion and Ethnicity (NSRE) 2000-01 (n=4,027) (cf. Kadushin, Phillips, and Saxe 2005; Kotler-Berkowitz in this issue; Kotler-Berkowitz et al. 2004; Schulman 2003; United Jewish Communities 2003b; United Jewish Communities 2003c).

As the use of the NJPS is limited by various methodological constraints (cf., Kadushin et al. 2005 and Kotler-Berkowitz in this issue), it is important to review those that relate to this paper: lost screening data and the interaction of the survey instrument and definitions of Jewishness.

Lost screening data. The NJPS first screened for Jews and "People of Jewish Background" (PJB--those with some connection to Jewish identity but who did not consider themselves to be currently Jewish) before switching to a randomly selected respondent in order to complete the main interview. Data are missing, however, from about two-thirds of the cases where people were offered the main interview and refused. Respondents who rejected longer interviews were significantly more likely to come from ethno-religiously mixed households than those who agreed to participate in the main interview (Phillips, Kadushin, and Saxe 2004). As switchers and ethno-apostates are more likely to be found in such households, failing to adjust for this bias may result in invalid estimates. We use Phillips et al.'s (2004) revised weights, which take account of these biases.

Survey instrument. People of Jewish Background (PJB) received a different instrument than the Jewish subsample, omitting questions concerning current Jewish practice and adding questions specifically designed to learn more about people who switched to other religions. As the definition of a person as a Jew or PJB was based, in part, on the question "Do you consider yourself Jewish for any reason?" respondents who were raised Jewish but currently identify as Christian were interviewed as Jews if they answered affirmatively, and did not receive questions about switching. Data are systematically missing from a nonrandom 46 percent of cases; thus, questions like age at time of switching cannot be examined. Other variables excluded from analysis include travel to Israel and attending Jewish summer camp during childhood.