Working papers // The Levy Economics Institute 507
The old ways in which surveys of Jews handled marginal cases no longer make sense, and the number of cases involved is no longer small. I examine in detail the public-use samples of the two recent national surveys of Americans of recent Jewish originthe National Jewish Population Survey (NJPS) and the American Jewish Identity Survey (AJIS)and also explore the implications for the American Jewish Committee annual surveys of Jewish political opinion. When Jews are defined by the question What is your religion, if any? the effect is not primarily to eliminate secular or culturally oriented Jews. However, large majorities of the children of intermarriage will fail to reply Jewish. Accordingly, the paper turns to two competing procedures for treating respondents of recent Jewish origin who do not report themselves to be Jewish by religion. The core Jewish population includes respondents who answer that they have no religion. I find this procedure problematic because the meaning of the no religion response has also changed: it no longer captures people with close connections to the Jewish world who deny the religious connection out of principle. Yet two out of three are the products of intermarriage. I tentatively suggest an alternative principle: self-identity. Americans of recent Jewish origin who are not Jews by religion should be asked (as they were in the 200001 the NJPS) whether they consider themselves Jewish for any reason. Those that reply in the affirmative should be counted as Jews. The paper examines the proportions of people affected by limiting surveys of American Jews to Jews by religion, and the results of using one or another procedure for deciding who else is a Jew. As an example, some demographic outcomes are tabulated using different definitions, as are responses to the question How close do you feel to Israel?