Do Jews deny being Jewish when called on the telephone? The goal of this test was to indicate whether there is a substantial incidence of Jewish "denial" among people known to be Jews. If the rate of Jewish denial among Jews in a phone interview was found to be quite high, this would suggest that the American Jewish population estimate in NJPS 2000-01 may be too low.

A total of 39,917 phone numbers from 5 different Jewish federation lists (Cleveland, Chicago, Kansas City, Palm Beach and San Antonio) were submitted to RoperASW, a survey research firm with headquarters in New York City, which conducted the telephone dialing. Phone calls were made on Saturday, October 5, 2002 and Sunday, October 6, 2002.

Neither interviewers nor respondents were informed of the study's sponsor. A screening interview nearly identical to the NJPS 2000-01 screener was administered, and respondents were classified as either Jewish, PJB or non-Jewish. For the purposes of this test the presence of a Jew or PJB in the household was sufficient to categorize the household as Jewish.

If a household was classified as non-Jewish after being asked the four screening questions about each adult in the household, the respondent (on behalf of the household) was asked a few additional questions concerning the presence of other Jewish relatives and whether anyone in the household had ever considered him/herself Jewish (and why).

Of the 997 valid responses to the survey, 66 (6.6%) of the households contacted were identified as non-Jewish. Assuming 100% of the people on the federation lists were Jewish, the maximum possible Jewish denial rate from this study would be 6.6%. However, the Kansas City sample inadvertently included a list of JCC members that contains known non-Jews. When removing the Kansas City respondents from the larger sample, the Jewish denial rate from the remaining sample becomes 3.5% (25/723).

Of the remaining sample (Chicago, Cleveland, Palm Beach and San Antonio), 25 out of 723 households initially qualify as potential Jewish "deniers." However, one of these 25 households previously considered itself to be Jewish because it once had a Jewish grandparent. In a separate question, 6 other households indicated that they once had a Jewish grandparent or other relative. In these 7 cases, Jewishness at some level is being affirmed, not denied. Subtracting these 7 households from the original 25, the adjusted estimated ceiling rate for Jewish denial becomes 2.6% (18/723).

This 2.6% ceiling estimate, however, is based on the initial assumption that the 4 submitted federation lists were 100% Jewish. Otherwise, a seeming Jewish "denial" may instead be a non-Jew affirming the fact that he/she is not Jewish.

Before the Jewish denial test was conducted in October 2002, UJC asked participating federations for lists of donor phone numbers that were purged of non-Jews and business numbers. Although there was not sufficient time before the test to determine the criteria by which the lists were assembled, the UJC Research Department has since verified with federation professionals in each participating community the composition of the lists that were sent. As a result of this verification, there is reason to believe that some instances of Jewish "denial" may instead be instances where non-Jews were contacted:

In 2 of the 4 federation lists that were used, "prospects" were included in lists submitted to UJC. Prospects are names of non-federation donors who are constantly added to the federation database, although there exists no systematic criteria for doing so. There is also no way to verify that every prospect added is, in fact, Jewish.

Moreover, federation personnel do not believe that prospects are ever purged from the database. For the purposes of this Jewish denial test, this may result in calling a phone number that once belonged to a Jewish prospect but now belongs to a non-Jew. In addition, officials in the remaining 2 communities indicated that, as the names of their donors were not coded by religion, they could not rule out the possibility of some non-Jews appearing on their lists.

Conclusion

In conclusion, the estimated ceiling rate for Jewish denial is 2.6%. However, this ceiling assumes a 100% Jewish sample. A closer examination of federation lists suggests that this may not be the case, however. The actual ceiling from this test may therefore be lower, potentially as low as 0. In addition, these specific lists cannot considered to be a random or representative sample of the larger American Jewish population.