Was it the demography?: A reassessment of U.S. Jewish population estimates, 1945–2001

Abstract

National Jewish population surveys in the United States provide comprehensive data and insights about the Jewish community’s
development. Controversy accompanied the 2000–2001 National Jewish Population Survey (NJPS) since inception and through release
of the final report. This paper focuses on the plausibility of Jewish population size—namely, a decline since 1990—as estimated
from the 2001 NJPS. While the size of the U.S. Jewish population is difficult to ascertain, growth momentum reached its peak
around 1990, followed by moderate decline. Population aging and a decline in the wish to identify as a Jew underlie a decrease
in the Jewish population. In retrospect, the 1957 Current Population Survey and 1990 NJPS probably better represented the
total Jewish population, while the 1970 NJPS and 2001 NJPS better covered the more Jewishly identified sections. NJPS 2001,
in any case, is an important and usable source.

Full text

DELLAPERGOLA 85
Was It the Demography?: A Reassessment of
U.S. Jewish Population Estimates, 1945-2001
Sergio DellaPergola
Abstract
National Jewish population surveys in the United States pro-
vide comprehensive data and insights about the Jewish com-
munity’s development. Controversy accompanied the 2000-
2001 National Jewish Population Survey (NJPS) since incep-
tion and through release of the final report. This paper focuses
on the plausibility of Jewish population size—namely, a
decline since 1990—as estimated from the 2001 NJPS. While
the size of the U.S. Jewish population is difficult to ascertain,
growth momentum reached its peak around 1990, followed by
moderate decline. Population aging and a decline in the wish
to identify as a Jew underlie a decrease in the Jewish popula-
tion. In retrospect, the 1957 Current Population Survey and
1990 NJPS probably better represented the total Jewish popu-
lation, while the 1970 NJPS and 2001 NJPS better covered the
more Jewishly identified sections. NJPS 2001, in any case, is
an important and usable source.
Large-scale national Jewish population surveys conducted periodically
in the United States provide a comprehensive inventory of data and
insights about the community’s main development paths and patterns.
Beyond the specific themes investigated, research plays a leading role in
portraying “the state of the nation” and in providing, if not foundations,
at least an anchor to collective identity and imaginary inside and outside
of the Jewish public. Given these broad implications in the context of an
ongoing spirited debate about continuity and change among American
Jewry (Rosenberg Farber and Waxman, 1999; Cohen and Eisen, 2000;
DellaPergola, 2000; Liebman, 2001; Goldscheider, 2002; Verbit, 2002),
it could have been expected that the 2000-01 National Jewish Popula-
tion Survey (NJPS) would spark controversy. Indeed, controversy
accompanied this major study from its inception, increasing after publi-
cation of the preliminary results in October 2002 and—following the
high-profile postponement of its release on the eve of the 2002 UJC
General Assembly—the issuing of the final report in September 2003.
Fundamental disagreements emerged in the organized Jewish com-
munity and among research specialists regarding the management of
NJPS and the quality of its data. Actually, the questions at stake
involved far broader facets of the delicate relationship between the Jew-

86 CONTEMPORARY JEWRY
ish institutional system and scientific research. Since the launching of
NJPS 2001, much has been written on these themes, mostly in the form
of editorials and Op-Eds in the Jewish and general press (e.g., DellaPer-
gola, 2003a; Goldberg, 2003; Saxe and Kadushin, 2003; Sheskin, 2003).
However, a systematic discussion of the major issues revolving around
the NJPS data description (Kotler-Berkowitz et al., 2003) and analysis
in historical perspective is still to come.
This paper focuses on one basic issue: the plausibility of Jewish
population size—namely a population decline since 1990—as estimated
from the 2001 NJPS, in the context of the chain of factors that may or
may not have determined the apparent demographic changes. Such eval-
uation cannot be limited to reviewing the NJPS findings only, but calls
for an assessment of the major determinants of Jewish population
change over the last several decades. Not less significantly, we need to
review NJPS in comparison with the main sources of data on U.S. Jew-
ish population since the end of World War II, trying to evaluate what
part of the results may reflect actual population changes and what part
may stem from variations in the quality of data. In this paper, we shall
discuss whether and to what extent the results of various major national
data sets can be reconciled over a span of several decades. While trying
to ascertain whether or not we should address NJPS 2001 as another
contribution to the mainstream of American Jewish social research, the
main question we ask is: Was it the demography that determined the
apparent Jewish population decline between 1990 and 2001, or was it
data quality?
Conceptual and Operational Determinants of Jewish Population
Size and Change
How many Jews live in the United States may not be the most important
statistic to emerge from national population studies such as the NJPS
(Kotler-Berkowitz et al., 2003) or the American Jewish Identification
Survey (AJIS) (Mayer et al., 2002), both undertaken in 2001, but the
statistic’s importance cannot be underestimated. At any given time, pop-
ulation size is a synthetic indicator of an array of different demographic
and sociocultural determinants and trends. In this respect, quantitative
measures do not provide more than a broad descriptive container whose
contents need to be enriched by additional qualitative evidence for both
analytical and planning purposes. Nonetheless, numbers are important
because they may unveil the basic dynamics of growth, resilience, or
decline among the group investigated in historical perspective, and the
nature of mutual relationships between the Jewish group and society’s
majority or other minorities in comparative perspective.
U.S. Jews constitute a particular case of the general class of sub-
populations whose total size is intensely affected by both biological-

DELLAPERGOLA 87
demographic and cultural-identificational variables and who respond to
three operational mechanisms or components of change: (a) how many
newborn babies are identified with the given group by their parents over
a certain period of time and how many die from within the same group
over the same period; (b) how many persons identified with the group
immigrate into the country and how many emigrate over the same
period; and (c) how many join the given collective, no matter by what
procedure, and how many sever their links with the given collective.
The preceding argument assumes a significant amount of continuity
over time in terms of both composition of the investigated population
and rules for its definition. Indeed, while the definitional criteria of Jew-
ish populations can be demonstrated not to have been constant in long-
term historical experience (Cohen, 1999; Corinaldi, 2001), a basic
assumption widely shared by past historians and social scientists—first
among them demographers—was a relatively coherent frame of refer-
ence transcending space and time (Herman, 1977). Population charac-
teristics in a given locale at a certain point of time could generally be
explained in the light of demographic and identificational trends occur-
ring over an earlier time span locally or elsewhere, in the case of inter-
national migrations.
A more recent orientation in the scholarly debate has attempted to
emphasize less such spatial-temporal continuity—inherently related to
genealogy—giving more attention to changes in the local environmental
circumstances. In general terms, relativist or post-modernist arguments
have questioned the real or imaginary roots of corporate ethnic identities
(Anderson, 1991), and have related them to myths that, while some-
times weakly tied to historiographic evidence, may nonetheless gather
significant momentum and relevance. This perspective stresses a
group’s ability to shape, invent or reinvent its own sense of personal and
collective belonging and solidarity regardless of fixed constraints pro-
vided by the observation of past behaviors or attitudes. A similar argu-
ment applied to the study of Jewish community identities calls for the
need to constantly redefine basic terms of reference, such as the bound-
aries and contents of Jewishness. Perceptions of the ultimate gist of
Jewish identification tend to shift from a more conventional past notion
of Jewishness as a mode of being intimately related to a given set of
beliefs and behaviors, to Jewish identification as expressed through con-
necting, traveling, and journeying, or even surfing and zapping—mean-
ing eclectic, selective, subjective reconstruction of the subject matter
and its meaning (Horowitz, 2002; Dencik, 2002; Cohen and Eisen,
2000). The notion of historical continuity tends to lose relevance facing
the pressure of new needs and new norms inherent in contemporary cul-
ture and society.

88 CONTEMPORARY JEWRY
An important definitional predicament is associated with these
changes of context and contents. The paradigmatic “Who is a Jew?”
question constitutes an ever-elusive issue in the field of Jewish popula-
tion studies, in the long term and, particularly, in recent years. A major
problem in Jewish population estimates available in the literature,
whether by individual scholars or by Jewish organizations, is a lack of
coherence and uniformity in the definitional criteria followed (DellaPer-
gola, 1992).
The study of Jewish populations may rely on normative or on oper-
ational definitions of the target population. Halachah (Jewish rabbinical
law) provides a clear and authoritative normative definition of who is a
Jew. Even alternative definitions stemming from the adoption of patri-
linearity in the attribution of a Jewish identity, while admittedly incon-
sistent with traditional halachah, remain in the domain of normative
definitions. However, for empirical research purposes, it is not possible
to undertake the stringent controls involved in ascertaining each individ-
ual’s Jewish identity according to such technical criteria. Therefore,
Jewish populations usually are identified in censuses or surveys through
operational criteria, such as the more or less accurate proxies offered by
simply pre-coded variables like religion or ethnic origin, or based on
indirect and rougher information on countries of origin (Rosenthal,
1975) and the like.
One important complicating factor in contemporary definitions is
the increasing frequency of intermarriage. Intermarriage provides the
major context for the growing number of individuals whose Jewish
identification is one among several possible ancestries. Moreover, such
personal Jewish identities may become the object of controversy
between different religious or legal authorities. Consequently, many
individuals do not know whether or when to identify as Jewish or prefer
not to, regardless of their eligibility according to some objective crite-
rion. Others who know about their own Jewish identity do not deem it
mutually exclusive with other religious or ethnic identities. This may
conflict with the normative assumption that Jewish identity is incompat-
ible with other religious identities.
To appropriately appraise Jewish population trends one needs to
address the broadest possible definition of the collective. Yet, to make
the study of any finite population meaningful and worthwhile, we need
working definitions. Definitions imply certain standards, the alternative
being an amorphous approach unable to generate analytic conclusions
of any sort. It is therefore essential that data collectors allow for wide
and flexible analytic opportunities to data users who within the broadest
possible initial definition may later decide on specific definitional typol-
ogies according to their own assumptions and research goals.

DELLAPERGOLA 89
Jewish population definitions for the purpose of empirical research
long have followed operational rather than normative paths. The con-
cepts of “core” and “enlarged” Jewish population were developed for
analytical purposes in order to disentangle the cluster of Jews and non-
Jews who share their daily lives in the same households, and to trace
virtual boundary lines where in reality such group boundaries have
become increasingly flexible, porous and interchangeable (DellaPer-
gola, 2003b). The core Jewish population concept includes all those
who, when asked, identify themselves as Jews or, if the respondent is a
different person in the same household, are identified by him/her as
Jews, and those of Jewish parentage who are identificationally indiffer-
ent or agnostic but do not formally identify with another religious
group. The enlarged Jewish population concept also includes all other
persons of Jewish parentage who are not Jews currently (or at the time
of investigation) and all the additional non-Jewish members (spouses,
children, etc.) in mixed religious households. In addition, the Law of
Return, Israel’s distinctive legal framework for the acceptance and
absorption of new immigrants, awards immediate citizenship and other
civil rights to all current Jewish new immigrants and to their Jewish or
non-Jewish spouses, children, and grandchildren, as well as to the
spouses of such children and grandchildren. Unfortunately, much recent
research has confused these various definitional concepts. It should be
noted that an enlarged Jewish population may be growing while the
respective core Jewish population is declining.
A great amount of latitude hence characterizes the definitional solu-
tions adopted in sociodemographic research, with obvious conse-
quences for the ensuing population counts. Today, normative
formulations persist at one extreme of the Jewish definitional contin-
uum, such as in Israel’s Law of Return, where “a Jew is who was born of
a Jewish mother or was converted (in Hebrew: nitgaiyer) and does not
belong to another religion” (Corinaldi, 2001). Such definition carries
important practical implications in the daily life of Israel society and its
relationship to world Jewry. It does not only follow normative guide-
lines but also assumes Jewish identity to be mutually exclusive with
other identities. Within the operational limits of social scientific
research in the United States, definitions have tended to evolve from the
straightforward question “What is his religion?”, and its “Jewish”
modality (U.S. Census Bureau, 1958), through the several multivariate
ideal constructs devised for the 1990 NJPS (Kosmin et al., 1991), reach-
ing increasingly more nuanced solutions in the NJPS 2001. In the latter
survey, at least in the version initially processed and circulated by UJC,
“a Jew is defined as a person whose religion is Judaism, OR whose reli-
gion is Jewish and something else, OR who has no religion and has at
least one Jewish parent or a Jewish upbringing, OR who has a non-

90 CONTEMPORARY JEWRY
monotheistic religion and has at least one Jewish parent or a Jewish
upbringing” (Kotler-Berkowitz et al., 2003). It is important to note that
such definition is neither normative nor mutually exclusive with being
identified with other groups. These criteria were strongly supported by
the sponsoring institution (UJC) as better responding to cognitive and
policy planning needs vis-à-vis their perceived constituency. NJPS 2001
therefore operated based on a broadly conceived Jewish population def-
inition, relying on non-dichotomic classification criteria and therefore
allowing for a large number of intermediate categories of uncertain res-
olution. Such definition not only was operational in nature but also sup-
posedly reflected particular servicing considerations. At the extreme
end of this ongoing definitional transformation, no consistent criterion
may be left besides the purely subjective and willingness of people to
identify themselves with the given group at a given point in time.
In such a fluid conceptual context, a derived question transcending
the strictly technical issues concerns the relevance of demography as a
useful tool in the appraisal of U.S. Jewish population, and of Jewish and
other sub-populations generally. If a population can be born, die, resur-
rect itself, or be targeted essentially as the expression of social propensi-
ties or organizational needs at a given time and place regardless of
historical identification, comparing populations and population change
over time and across space lose significance. Jewish population continu-
ity and discontinuity may become decreasingly connected with demo-
graphically traceable patterns, and increasingly a function of cultural,
political or budget-related tides, fashions or propensities. The conse-
quences for Jewish social scientific research and for social research in
general are complex. Some would argue that the highly voluntaristic
patterns of identification of contemporary Jewish populations make
unfeasible any attempt to quantify the number of Jews locally, or glo-
bally (Schnapper, 1987). Even if one rejects this position, Jewish popu-
lation figures and estimates should always be taken as orders of
magnitude, surrounded by variable and sometimes significant margins
of error. While reflecting improvements and corrections, Jewish popula-
tion estimates highlight the increasing complexity of the sociodemo-
graphic and identificational processes underlying the definition of
Jewish populations; hence, estimates of their sizes. This is made even
more complicated by growing international and internal migration,
which often lead to double counts of people on the move. Consequently,
the analyst has to come to terms with the paradox of the permanently
provisional character of Jewish population estimates (DellaPergola,
2002).
A related question is whether American Jewish population and
identification trends should be interpreted primarily in the framework of
American society, or as part of a global configuration of Jewish commu-

DELLAPERGOLA 91
nities. While the obvious answer is both, conclusions drawn about the
same sociodemographic and sociocultural patterns according to either
perspective can be quite different. Significant assimilation and erosion
from an internal perspective can at the same time appear as significant
distinctiveness and resilience from an external perspective. Another
question is whether the criteria used to define the U.S. Jewish popula-
tion should be the same or different as criteria used to define Jews in
other countries. Some exceptionalism provided by the general societal
context—in this case American—normally should be expected to be
found in the analysis of Jewish social patterns at the substantive level.
However, were we to make the assumption that the definitional criteria
themselves should take into account American exceptionalism, the
implication would be that Jewishness has become a subsidiary attribute
of “Americanness,” and therefore cannot really be compared across the
gamut of Jewish populations and communities worldwide (see the con-
ceptual discussion in Horowitz, 2002).
In the light of these concerns—actually worthy of far deeper elabo-
ration than feasible here—the central theme of this paper needs to be
reformulated: Under what circumstances can we reconcile Jewish popu-
lation sizes and characteristics emerging from various national data
sources, in light of definitional or other cultural intervening factors?
Should the 2001 NJPS data be read as part of an unfolding process of
sociodemographic change among American Jewry, or as an unrelated
one-time shot in the history of Jewish population research? And perhaps
more fundamentally, what can demography contribute in the ongoing
debate about NJPS 2001?
Local vs. National Sources for U.S. Jewish Population Estimates
As appropriate to a country where there is separation of Church and
State, U.S. population censuses never have addressed the question of
religion (Good, 1959). For the most part, the Jewish community has
opposed the inclusion of such a question in U.S. censuses. Nonetheless,
a significant research tradition has developed concerning the size and
characteristics of religious groups. Jews in particular have been scruti-
nized through a variety of direct and indirect research techniques (Robi-
son, 1943; Seligman, 1950; Goldscheider, 1982; Cohen, Woocher,
Phillips, 1984; Winter and Levin, 1984; Schmelz and DellaPergola,
1984; Ritterband, Kosmin, Scheckner, 1988; Goldstein, 1989; Schwartz,
Scheckner, Kotler-Berkowitz, 2002).
Over time, if we disregard indirect estimation methods, national
Jewish population estimates in the United States were obtained through
two basic approaches. One was the systematic compilation and summa-
tion of local population estimates periodically collected by local Jewish
communities and federations. The other was independent data-collec-

92 CONTEMPORARY JEWRY
tion efforts at the national level at one time or another. Both approaches
have significant precedent and comparability, reflecting the research ini-
tiatives of Jewish organizations and, to some extent, general public
agencies.
There clearly are advantages and disadvantages to both national
and local studies. First, in term of relevance, both approaches are signif-
icant as each captures relevant data though different dimensions of the
corporate community and its relationship with the broader social con-
text. Local studies can claim greater proximity with the subject matter,
while national studies are conducted at a distance by people who may
not understand significant local peculiarities. Local studies are con-
ducted by a variety of investigators and organizations, with different
methods, and one may assume (or hope) that any mistakes will not
become cumulative. National studies are conducted with one common
methodology, and in case of bias, mistakes cumulate nationally. On the
other hand, national studies have the advantage of standardized defini-
tions and research methods, which local studies lack, and are clearly
superior in capturing one central feature of American society: a high
volume of internal migration and geographical mobility (Goldstein and
Goldstein, 1996). Frequent permanent or temporary passages of individ-
uals from one locale to another do not only affect geographical distribu-
tions but also Jewish identificational patterns; hence, nationwide counts
of Jews.
In the following we shall review some of the major findings in the
assessment of U.S. Jewish population size, while at the same time not-
ing some of the strengths and weaknesses of the various sources avail-
able.
Local Jewish Population Estimates
The tradition to determine the number of Jews in the U.S. materialized
primarily through the yearly publication of detailed population esti-
mates in the American Jewish Year Book (AJYB) issued by the Jewish
Publication Society and the American Jewish Committee (in recent
years the AJCommittee has became the sole publisher of the AJYB).
The AJCommittee indeed attributed great prominence to U.S. Jewish
demography in its yearly publications, which comprised some of the
best overviews of the subject matter and its implications (e.g. Seligman,
1950; Goldstein, 1981). Since the end of World War II (our period of
concern here) yearly estimates were based on compilations of data gath-
ered through a wide network of local sources. Some of these local data,
mostly for larger Jewish communities, reflected the results of local Jew-
ish population surveys. A serious problem with national estimates
derived from a conglomerate of hundreds of local estimates is that the

DELLAPERGOLA 93
national estimates lack a common basis of definitional and methodolog-
ical premises.
To provide a recent example, if we refer to the total U.S. Jewish
population estimate of 6,155,000 reported in the 2003 AJYB (The Edi-
tors, 2003), about 80% of this total figure reflected people covered by
local population surveys, but the remaining 20% depended on local non-
survey sources (often in small areas) whose reliability cannot be evalu-
ated (Table 1). A useful synopsis prepared by Sheskin (2001) and subse-
quent updates indicate that by 2002 the baselines for local population
estimates were spread over more than 20 years. A total pool of 48 com-
munity surveys, excluding repeated surveys of the same locality, could
be divided into three time periods. Of the total 80% of current U.S. Jew-
ish population thus covered, 34% derived from community surveys exe-
cuted about at the time of the 2001 NJPS, and therefore were quite
updated. Another 32% reflected studies undertaken in 1993-97, and a
further 14% were based on studies of 1982-92. Demographic and identi-
ficational changes that occurred between the time of data collection and
the 2001 NJPS were not incorporated in estimates that covered the
majority of the national total.
Other significant problems with drawing national totals from compila-
tions of local population estimates, besides the time lag mentioned
above, include:
(a) Definitions of target population in local surveys. Most, but
not all, surveys refer to a “core” Jewish population concept, but
sometimes they adopt an “enlarged” population concept. A few
Table 1
U.S. Jewish Population Estimates
by Type and Year of Source, 2002
Year of
last survey
Number
of
Surveys Total
Households Total
Persons Core
Jews % Core
Jews
Grand total 6,155,000 100.0
No survey 1,203,400 19.5
Total surveys 48 2,349,342 5,756,750 4,951,600 80.5
1998-2002 19 986,273 2,486,800 2,115,300 34.4
1993-1997 20 959,267 2,262,500 1,958,500 31.8
1982-1992 9 403,802 1,074,450 877,800 14.3
Source: adapted from Sheskin (2001), The Editors (2003).

94 CONTEMPORARY JEWRY
of the AJYB estimates for U.S. areas covered by surveys origi-
nally reported “enlarged” data including the non-Jewish mem-
bers of households.
(b) Coverage of target population definitions. Different posi-
tioning and contents of screeners and subsequent questionnaire
modules tend to produce variation in Jewish population defini-
tions and size.
(c) Bases for population estimates. The respective qualities
range between excellent to very poor. The minimum request
should be that the local estimates published in the AJYB are
accompanied by a quality rating (e.g.: A to D) based on year of
the baseline figure, quality of the baseline, and intervening (if
any) adjustments.
(d) Sampling techniques. Sampling frameworks include Ran-
dom Digit Dialing (RDD), Distinctive Jewish Name (DJN)
lists, Jewish community lists, and snowball lists—gauged
expanding from an initial roster of informants. The complete-
ness, accuracy and updating status of these lists is open to
questioning.
(e) Data collection techniques. These include telephone or
face-to-face interviews. Choice of respondent within the
household—selective or all-inclusive; based on rigorous pre-
determination or not—may affect the results.
(f) Response rates. These tend to be quite different, reflecting
variability in research designs and techniques, or even the
capacity of fieldwork organizations.
(g) Questionnaire contents. Different surveys use different
questionnaires and unavoidably do not systematically cover the
same topics or ask identical questions for the same topic. This
affects the possibilities of data evaluation and comparisons.
(h) Residential eligibility definitions. Reflecting frequent sea-
sonal mobility within the U.S., this likely generates double
counts of individuals with more than one residence (e.g. snow-
birds).
(i) Handling of mobile people. According to the 2001 NJPS, as
many as one-fifth of the households changed their city of resi-

DELLAPERGOLA 95
dence in the five years prior to survey, and one-half of them
across states. Some of these mobiles have a greater chance of
being included in more than one frame for sampling.
(j) Handling of bi-local populations. Typically, college stu-
dents can be reported at both home and study residences.
(k) Geographical boundaries of localities and regions. Possi-
ble overlaps exist in tangential suburban areas in surveys con-
ducted in neighboring metropolitan areas.
(l) Geographical catchment areas. Inconsistencies may occur
between federation areas, Standard Metropolitan Areas, or
local authority boundaries.
(m) Criteria for estimating missing parameters. In the absence
of detailed individual data, estimates drawn from household
numbers mostly for smaller communities may rely on errone-
ous assumptions. One example is the now clearly obsolete
multiplier of 3 Jewish members per household still often
assumed by local sources.
(n) Criteria for assessing local change. In the absence of hard
data, local pride may cause sources in growing communities to
report on growth, and in dwindling communities not to report
decline.
(o) Authorship of survey reports. Finally, besides variable pro-
fessional skills, ranging between top scholars to unknown sur-
veyors, a possible “author effect” may reflect different
disciplinary training, policy perspectives and agendas, possibly
turning up with non-comparable conclusions even when based
on similar findings.
The national Jewish population totals derived from local estimates and
reported in Table 1 indeed appear quite discombobulated. The figure
initially reported by the AJYB for the number of U.S. Jews in 1945 was
4,770,000, relying on a 1936 Census of Religious Bodies that had evalu-
ated the total Jewish population at 4,770,647 (Engelman, 1947). In sub-
sequent years, the AJYB reported estimates alternating periods of
stability and years of moderate or more sudden change. From 1946 to
1953, the reported estimate was 5 million, then it started to grow and in
1960 passed the 5.5 million mark. Total U.S. Jewish population figures
above 6 million first were reported during the early 1970s, and then

96 CONTEMPORARY JEWRY
after downward corrections were made following the 1970 and 1990
NJPS studies (see AJYB issues relating to 1997 and subsequent years).
Since the creation of the North American Jewish Data Bank (NAJDB)
in 1986 as a joint venture of CJF-UJC (Council of Jewish Federations-
United Jewish Communities) and the Center for Jewish Studies at the
Graduate School and University Center of the City University of New
York (currently at Brandeis University), local Jewish population survey
compilations and estimates were centrally handled at NAJDB. Follow-
ing release of the 2001 NJPS results, publication of an alternative
national Jewish population estimate based on the compilation of local
Jewish population estimates was temporarily discontinued (The Editors,
2003).
National Jewish Population Studies
Major national Jewish population studies periodically have been under-
taken in the United States. We briefly review here the main technical
characteristics of major surveys based on significant numbers of Jewish
cases (see Table 2 for a synopsis of the main parameters). We do not
refer to the many instances of general national surveys that include Jews
as part of a relatively small total sample, such as NORC-GSS or similar
surveys. The latter studies, while important in a general comparative
research perspective, do not allow for detailed insights on the Jewish
group because of the limited possibility of detailed processing based on
few cases and high sampling variability. All of the following databases
rely on sample techniques and their results therefore are subject to sam-
pling variability and errors.
(a) The 1957 CPS. A sample survey of the U.S. civilian popu-
lation was undertaken in March 1957 as part of the Census
Bureau’s Current Population Survey. This was the only
instance in which a question on religion was included this peri-
odical survey. The sample was spread over 330 sample areas
comprising 638 counties and independent cities. A total of
about 40,000 dwelling units and other living quarters were des-
ignated for the sample, and completed interviews were
obtained from about 35,000 households (a response rate of
over 87%). Of the remainder, 1,500 were households for which
information could not be obtained, and the rest were vacant
dwellings or others not to be enumerated. In the survey, sub-
jects voluntarily answered the question “What is your reli-
gion?” Of all persons 14 and older in the survey, 96% reported
a religion. The estimating procedure involved the adjustment
of weighted sample results to independent estimates of the
civilian non-institutional population in the United States (U.S.

DELLAPERGOLA 97
Census Bureau, 1958). Jews represented 3.2% of total adults.
Assuming a Jewish household size 10% smaller than the
national average, Jewish households would comprise about
3.5% of total households, or 1,225 respondents out of a
national sample of 35,000. Assuming an average Jewish house-
hold size of 2.95, the survey would cover 3,614 individuals in
Jewish households. The total Jewish population aged 14 and
over was estimated at 3,868,000, and after allocation of
1,107,000 children below 14 in households where both head-
of-household and wife were reported to be Jewish, and 38,000
Jewish children out of a total of 64,000 children in households
where only one of the spouses was reported to be Jewish, a
total estimate of 5,013,000 obtained (Glick, 1960).

98 CONTEMPORARY JEWRY
Table 2
Synopsis of Main National Sources of Data on U.S. Jewish Population
1957-2001
Parameter CPS NJPS NJPS NJPS AJIS
Full name of study Current Population
Survey National Jewish
Population Study National Jewish
Population Survey National Jewish
Population Survey American Jewish
Identity Survey
Year 1957 1970-71 1989-90 2000-01 2001
Sponsor U.S. Census
Bureau Council of Jewish
Federations (CJF) CJF CJF – United Jewish
Communities (UJC) The Graduate
Center of the City
University of New
Yo r k
Sampling
stratification Counties Jewish density areas Census divisions Jewish density areas Census divisions
Identification of
respondents National sample Partly Jewish lists,
partly random area
sampling
Random Digit
Dialing (RDD) RDD RDD
Data collection
technique Face to face Face to face Telephone Telephone Telephone

DELLAPERGOLA 99
Jewish population
definition Based on
respondent’s
religion
Based on four-
question screener Based on four-
question screener Based on four-
question screener Based on four-
question screener
Jewishness resolution Dichotomic Dichotomic Dichotomic Non-dichotomic Dichotomic
Total interviews 35,000 5,750 125,813 >175,000 50,282
Net sample size:
Jewish households 1,225 5,750 2,441 4,523 1,668
Total individuals in
Jewish households 3,614 17,135 6,514 10,403 4,337
Total survey response
rate 87% 79% 24% 28% 18%
Table 2
Synopsis of Main National Sources of Data on U.S. Jewish Population
1957-2001 (cont.)

100 CONTEMPORARY JEWRY
(b) The 1970-1971 NJPS (for brevity: 1970). The first full-
scale national Jewish population study was sponsored by CJF
and was directed by Fred Massarik and Bernard Lazerwitz.
Respondents were located through a combination of Jewish
community lists and area probability sampling. For eligibility,
the respondent had to provide an affirmative reply, for himself
and/or for one or more household members, to one or more of
the following questions: (1) Was person born Jewish? (2) Is
person Jewish now? (3) Was person’s father born Jewish? (4)
Was person’s mother born Jewish? Answers to these questions
could be exclusively yes or no. A total of 5,750 face-to-face
interviews were completed, 2,950 of them from lists and 2,800
from area sampling (Massarik, 1974). The response rate was
79%. At an average household size of 2.98, 17,135 individuals
in Jewish households were covered by the survey. The Jewish
population was estimated at 5,370,000, not inclusive of at least
50,000 persons in institutions. Including non-Jewish household
members, an enlarged total population of 5.8 million obtained.
(c) The 1989-1990 NJPS (for brevity: 1990). This major survey
also was sponsored by CJF and was advised by a National
Technical Advisory Committee (NTAC) directed by Sidney
Goldstein, with Barry Kosmin Research serving as director on
behalf of CJF. The survey was based on a national RDD sam-
ple. Fieldwork was handled by ICR, a Pennsylvania company.
Over the course of a year, 125,813 Americans were screened to
determine the Jewish qualification of their households in 1990.
This resulted from the answers provided to any of four screen-
ing questions: (1) What is your religion? (2) Do you or anyone
else in the household consider themselves Jewish? (3) Were
you or anyone else in the household raised Jewish? (4) Do you
or anyone else in the household have a Jewish parent? Answers
to these questions could be exclusively yes or no. Following
this first stage, 5,139 households qualified. The response rate
to the initial screener interview was about 50%. Of the house-
holds potentially identified as Jewish, 2,441 households were
traced and contacted again and completed interviews at a total
survey response rate of 24%. The total of individuals in the
households finally covered was 6,514. Due to this multi-stage
procedure, there were some inconsistencies in the identifica-
tion of eligible households and in their willingness to partici-
pate at each stage. The estimating procedure involved
weighting the total sample of 125,813 screened responding
households to match the most current U.S. Census Bureau esti-

DELLAPERGOLA 101
mates of basic demographic characteristics (Kosmin et al.,
1991; Goldstein, 1992; Goldstein and Kosmin, 1992; Waks-
berg, 1996). The core Jewish population, inclusive of Jews
with no religion and Jews by choice, as well as 100,000 Jews in
institutions, was estimated at 5,515,000. Of these, 4,210,000
were born Jews whose religion was Judaism, 185,000 were
converts to Judaism/Jews by choice, and 1,120,000 were born
Jews with no religion. The total number of Jews and other per-
sons of Jewish origin was estimated at 6,840,000. The total
number of individuals in 3.2 million qualified Jewish house-
holds (including gentiles) was estimated at 8,200,000, includ-
ing 500,000 households without any current “core” Jew.
(d) The 2000-2001 NJPS (for brevity: 2001). This survey was
again sponsored by CJF, later UJC, and advised by a National
Technical Advisory Committee chaired by Frank Mott and
Vivian Klaff, including senior researchers and Jewish commu-
nity planners, among them a former president of the Population
Association of America and a former deputy director of the
U.S. Census Bureau. On behalf of UJC, Jim Schwartz and later
Lawrence Kotler-Berkowitz served as research director, with
Lorraine Blass project manager. The survey was based on a
national stratified RDD sample. The United States was divided
into seven strata based on pre-survey estimates of Jewish popu-
lation density, with sampling probabilities proportional to Jew-
ish density in each stratum. Fieldwork and initial statistical
evaluation was handled by Roper-ASW, a New York company.
Over 175,000 households were screened for possible inclusion
based on four questions: (1) What is your or other adult’s reli-
gion, if any? (2) Do you or other adult have a Jewish mother or
a Jewish father? (3) Were you or other adult raised Jewish? (4)
Do you or other adult consider your/him/herself Jewish for any
reasons? Answers to these questions included options other
than yes or no—or, in other words, allowed a non-dichotomic
resolution to Jewish population definition. Such screening cri-
teria were known from the start to produce results not strictly
comparable with the 1990 NJPS. This modified research
design reflected an explicit mandate by UJC management to
the NTAC following a redefinition of service policies by the
sponsoring organization and a rejection of the “core-enlarged”
Jewish population conceptual definition. The final unweighted
sample included 4,220 Jewish respondents and 303 people of
Jewish background (PJB), for a total of 4,523 Jewish house-
holds; 625 non-Jews of Jewish background; and 4,027 non-

102 CONTEMPORARY JEWRY
Jews, for a total of 9,175 respondent households. The 4,027
non-Jewish households were interviewed for a National Survey
of Religion and Ethnicity (NSRE) to collect data necessary for
weighting and thus estimating the size of the Jewish popula-
tion, and to provide comparative data to Jews and PJBs on
socio-demographic topics. The total survey response rate was
28%. At an average household size of 2.3, the total number of
individuals in Jewish households covered was 10,403. Weights
were directly or indirectly estimated and applied to the original
data to adjust for the number of telephone lines in the house-
hold, and to bring sample household and respondent data to
U.S. Census totals for sampling strata, age, gender and region
(Kotler-Berkowitz et al., 2003). Following claims of exces-
sively low respondent rates, selective population undercounts,
and other inappropriate procedures during and following the
fieldwork, the NJPS was subjected to independent professional
scrutiny. It was concluded that while handicapped by several
methodological shortcomings such as low response rates,
inconsistent survey coverage of relevant sub-populations, and
loss of documentation, the study stood within the range of pro-
fessionally acceptable research standards and biases (Schul-
man, 2003). The total Jewish population was estimated at 5.2
million, including over 100,000 persons in institutions; of the
total, 4.3 million were identified as more Jewishly connected
and 800,000 as more weakly identified Americans of Jewish
background. Respondents from the former group were admin-
istered a long-form questionnaire; respondents from the latter
were administered a short-form questionnaire. The total num-
ber of Jews and non-Jews of Jewish background was estimated
at 6.7 million. The total of individuals in 2.9 million house-
holds with at least one Jewish member was estimated at 8.7
million.
(e) The 2001 AJIS. The fact that while a major national survey
was being undertaken (the NJPS) an alternative source of data
should be developed is an emblematic indicator of the charged
atmosphere that prevailed at the time of the NJPS. There were
three stated or implicit rationales for AJIS and the related
American Religious Identity Survey (ARIS): (1) a quest for a
study conducted at least nominally with a methodology identi-
cal to that of the 1990 NJPS, unlike the 2001 NJPS; (2) an ana-
lytic focus on secular Jewish identification, which some
maintained was not adequately covered in the NJPS design and
questionnaire; (3) a desire by the AJIS private sponsors to take

DELLAPERGOLA 103
major social research out of the responsibility and control of a
major Jewish community organization such as UJC. AJIS was
based on a national RDD sample. Fieldwork was handled by
ICR, the same company that did the 1990 NJPS. Out of all suc-
cessful contacts, a total of 50,238 respondents agreed to be
interviewed. After a series of screening questions quite similar
to those of NJPS 1990, 1,668 respondents qualified to be
included in a survey of American Jewish households. The
response rate was 18%. With an average household size of 2.6,
the total number of individuals in households covered was
4,337. The whole screening and actual survey interviewing
was conducted in one stage (Mayer, Kosmin, Keysar, 2002).
The estimated core Jewish population, inclusive of Jews with
no religion, Jews by choice, and Jews in institutions, was
5,340,000. Of these, 3,460,000 were born Jews whose religion
was Judaism, 170,000 were converts to Judaism/Jews by
choice, and 1,710,000 were born Jews with no religion. The
total of Jews and others of Jewish origin was 7,690,000. The
total of individuals in all households surveyed, including those
without any current “core” Jew was 9,740,000, excluding the
institutionalized population. These latter data (and not those of
the 2001 NJPS) conceptually match the figures reported above
from the 1990 NJPS.
One problematic feature emerging from Table 2 is the decreasing rate of
response achieved in major Jewish population studies, which does not
augur well for the future of survey research.
Independent Assessments of Jewish Population Size
Besides the major databases and summations just reviewed, there have
been several attempts to review critically the available evidence in the
light of a more systematic approach to population research. Some of
these efforts deserve attention, and the results are summarized in
Figure 1.
Dr. Ira Rosenwaike, a specialist on the demography of aging at the
University of Pennsylvania, produced a synthetic attempt to reconstruct
the unfolding of U.S. Jewish population in the light of available data and
some hypotheses about each determinant of population change (Rosen-
waike, 1980). His critique of the national data, obtained from the con-
glomerate of local estimates, brought him to suggest somewhat lower
data, beginning with 4,359,000 in 1945, and growing to 5,581,000 in
1975 through a gradual slowing of the pace of growth. These estimates
did not take into account the 1970 NJPS.

104 CONTEMPORARY JEWRY
Figure 1: U.S. Jewish Population Various Estimates (Thousands), 1945-2005
2500
3000
3500
4000
4500
5000
5500
6000
6500
1945 1950 1955 1960 1965 1970 1975 1980 1985 19 90 1995 2000 2005
AJYB Rosenwaike ICJ CPS NJPS
AJIS Proj1957 Proj1970 Proj1990
CPS
1957 NJPS
1970 NJPS
1990 NJPS
AJIS
2001
Note: AJYB = American Jewish
Year Book; ICJ = Institute of
Contemporary Jewry; CPS =
Current Population Survey;
NJPS = National Jewish Popula
-
tion Survey; AJIS = American
Jewish Identity Survey;
Proj1957 = Projection based on
CPS 1957; Proj1970 = Projec-
tion based on NJPS 1970;
Proj1990 = Projection based on
NJPS 1990.

DELLAPERGOLA 105
Beginning with 1980, under the general direction of Roberto Bachi,
demographers at the Hebrew University’s Institute of Contemporary
Jewry (ICJ) took responsibility for the elaboration of the annual Jewish
population estimates for each country in the world published by the
AJYB (for the background and initial articles of the series still under
way, see Schmelz, 1969; Schmelz, 1981a; Schmelz and DellaPergola,
1982). Yearly estimates were thus suggested for the United States, inde-
pendent of other sources—specifically, the national estimates also pub-
lished in the AJYB by the NAJDB. The initial figure of 5,690,000
reflected an assessment of 1970 NJPS results (Massarik, 1974; Lazer-
witz, 1978; Schmelz, 1981b) and subsequent population trends. In the
following years an effort was made to evaluate each component of pos-
sible population change, resulting in cautious increases tending to zero
population growth. These estimates were periodically reduced, reflect-
ing the results of 1990 NJPS and the 2001 NJPS and 2001 AJIS, respec-
tively, and the consequent determination of new population baselines.
The ICJ estimated a Jewish population increase of 185,000 between
1990 and 2000, as against 621,000 according to NAJDB. On the eve of
NJPS 2001, the ICJ U.S. estimate stood at 5.7 million.
Population Projections
In addition to these year-by-year accountancy exercises, a further way to
arrive at Jewish population estimates at various dates is to take a known
baseline and undertake a population projection by detailed age and sex
groups. Some of these projections are reported here, initially ignoring
the impact of international migration and only focusing on the effects of
age-sex specific birth and death rates. Figure 1 displays results of a pro-
jection of the 1957 CPS results to 1970; a projection of 1970 NJPS data
up to 2005; a similar projection of 1990 NJPS data; and the original total
estimates from the 2001 NJPS and 2001 AJIS. The NJPS data discussed
here include both the more strongly connected (Jews) and the more
weakly connected (PJB) individuals among the total sample, and rely on
the data weighting procedures originally suggested by the UJC Research
Department (Kotler-Berkowitz, 2003).
Demographic projections, besides detailed initial population distri-
butions, require a set of assumptions concerning the expected level of
fertility and mortality (and migration) in subsequent years. The impact
of international migration will be discussed below. In the present set of
projections, we assume a life expectancy at birth gradually increasing
from 67.5 for men and 71.7 for women in 1957-61, to 76 and 80, respec-
tively, in 2000-04. These estimates reflect known patterns from avail-
able Jewish and general research on mortality levels in the U.S.
(Goldstein, 1986; Rosenwaike, 1994).

106 CONTEMPORARY JEWRY
“Effectively Jewish” total fertility rates—a measure of the average
number of Jewishly identified children expected to be born to a woman
assuming indefinite continuation of age-specific fertility levels observed
in each period—were 2.80 in 1957-62, 2.49 in 1962-67, 1.92 in 1967-72
(DellaPergola, 1980), and were posited at 1.5 thereafter. The latter fig-
ure tends to be significantly and downwardly affected by the growing
frequency of out-marriages and by the tendency of a large proportion of
the outmarried households not to choose a Jewish identification for their
children. According to NJPS 2001, the total average number of children
born to Jewish women aged 40-44 was 1.86 (Kotler-Berkowitz et al.,
2003). However, among Jewish respondents less than 50 years old, 3%
of the children of the in-married and 53% of the children of the out-mar-
ried were not being raised Jewish resulting in a total of 27% of total
children (Barack Fishman, 2004). Factoring the Jewishness of children
into a completed family size of 1.86, an estimate of 1.36 obtains. The
already mentioned 1.5 total fertility estimate is thus likely to represent
an overestimate, but will nonetheless be kept here as the basis for our
projections in the light of the consistent evidence about low U.S. Jewish
fertility over the last decades (Mott and Abma, 1992). One consider-
ation is that a higher proportion of the children of Jewish women are
raised as Jews, compared with the children of Jewish men who out-
marry.
As shown in Figure 1, projected Jewish population estimates (with-
out migration) run from an earlier to the next baseline. Quite impor-
tantly, the estimates attained for the later date generally seem to match
sufficiently well the results obtained through fresh data collection. In
1970, the projected estimate from a 1957 baseline of 5,013,000 would
be 5,226,000, versus an actual NJPS 1970 estimate of 5,370,000 (with-
out persons in institutions). The difference between the actual and pro-
jected estimates is 2.8%. It should be mentioned that we estimated the
1957 five-year group age-sex Jewish composition by splitting a number
of broader age groups within each sex out of the somewhat less detailed
data originally published (U.S. Census Bureau, 1958 and 1968; Gold-
stein, 1969). In 1990, the projected estimate from a 1970 baseline of
5,370,000 would be 5,491,000, versus an actual NJPS 1990 estimate of
5,515,000—a difference of 0.4%. In 2001, the projected estimate from a
1990 baseline of 5,515,000 would be 5,367,000, versus an actual NJPS
2001 estimate of 5,035,000 (without persons in institutions)—the actual
data being 6.2% lower than the projection. Interestingly, if we project
the 1970 NJPS to 2001, a better fit is obtained with the actual 2001 data
than is obtained by projecting the 1990 NJPS data—5,282,000, or a dif-
ference of -4.7% between the actual and projected results.

DELLAPERGOLA 107
Comparative Evaluation of Jewish Population Size
When comparing these various data sets and results, a gap typically
emerges between national Jewish population estimates primarily
derived at the local and at the national level. Four times in the last 50
years nationally derived total Jewish population figures turned out to be
lower than the available sum of local Jewish population estimates: on
the occasion of the 1957 CPS, followed by the three NJPSs of 1970,
1990 and 2001, as well as the 2001 AJIS. Four times out of four looks
like a pattern indeed. This can be construed by the supporters of the
local estimates approach as evidence that national studies provide
underestimates of the real demographic situation. The symmetric argu-
ment can be provided by supporters of national data-collection efforts
that the sum of local Jewish population data significantly overestimates
the real population size. Yearly independent assessments of U.S. Jewish
population size, specifically the ongoing ICJ series, consistently fall
between the high of local data compilations and the low of periodical
national surveys and projections based on those surveys (see Figure 1).
Stretching these projections for another short period to 2005 enhances
the impression that the Jewish population in the United States reached a
peak around 1990 and subsequently began to diminish.
In the light of this evidence, the purported decline in the number of
American Jews between the 1990 and 2001 NJPSs clearly may not be an
absolute truth due to the margins of error that are inherent in any statisti-
cal survey and to NJPS particularly. But population decline definitely
looks like a possibility in the light of longer-term Jewish demographic
trends. Due to immigration during the 1990s, U.S. Jewry actually should
have increased by about 250,000, to over 5.7 million, even assuming
zero population growth. The fact that the reported NJPS total Jewish
population (with people in institutions) was 5.2 million points to an
even larger gap between the expected and the actual. The suspicion of
significant undercounting in the 2001 NJPS thus would seem plausible.
A useful independent check obtains through another nationwide survey:
The AJIS found 5.35 million Jews in the United States in 2001, a figure
also considerably lower than the 1990 NJPS estimate. As already noted,
there were significant conceptual and methodological differences
between NJPS and AJIS, but both provided independent, large-scale,
representative efforts to portray U.S. Jewry. The consistency of findings
among these two surveys cannot be attributed to consistent technical
biases. One may conclude that at least part of the apparent recent reduc-
tion in U.S. Jewish population size seems to reflect the actual interplay
of demographic and identificational variables and is not only an artifact
of survey inadequacies.

108 CONTEMPORARY JEWRY
Reconciling Age Composition across Subsequent Surveys
A review of the changing age composition of U.S. Jewry offers an
opportunity for a more in-depth study of patterns of demographic conti-
nuity and change reflecting the birth rate, international migration, and
retention within a Jewish definitional framework. Age-specific compar-
isons also provide important evidence to the search for inter-survey con-
sistency or inconsistency. Figure 2 compares the age distributions of
U.S. Jews in 1957, 1970, 1990, and 2001, according to the four major
sources of data (see the more detailed discussion below).
The 1957 CPS offers a snapshot of American Jewry at the height of
the baby boom era. The rapid growth of the child cohorts reflects prima-
rily the relatively high and growing Jewish Total Fertility Rates, which
reached a peak of 2.80 around 1955, versus a peak of 3.50 among total
U.S. whites around 1960 (DellaPergola, 1980). We also note that the
baby boom to some extent constituted an echo effect of the large cohorts
born in the years just before and after World War I, before the Great
Depression. Interestingly, that earlier period of demographic expansion
occurred at a time of very rapid contraction in Jewish fertility levels.
The average number of children born to foreign-born Jewish women
aged 45 and over declined from 7.2 in 1910 (vs. 5.7 among total U.S.
immigrant women), to 3.8 in 1940 (the same as among total women)
(Grabill, Kiser, Whelpton, 1958). The post-World War II baby boom
probably started a little earlier among Jews than among total whites, and
peaked earlier at a level equivalent to 80% of the total white maximum.
The subsequent baby bust years are well-documented in the 1970
and 1990 NJPS data. The “fat” baby-boom cohorts move to the right in
the Figure 2 displays and are substituted by “leaner” cohorts. In 1990, a
baby boom echo effect appears, but the 2001 data show it to have been
short lived and of modest import. The size of relevant birth cohorts is
not at all comparable with that of the parental cohorts, because unlike
during the late 1940s through the early 1960s, the mechanical effect of
large parental cohorts was not accompanied by an actual fertility
increase during the 1980s and 1990s.
Another feature emerging in the 2001 data is a significant leveling-
off of the original baby boom cohorts. This “disappearance” can be
related to a declining willingness to identify as Jewish even by the loose
definitional criteria of the 2001 NJPS, or to special problems of survey
coverage among those specific age cohorts—or, most likely, a combina-
tion of both causes. The fading away of large older cohorts and their
substitution by smaller younger cohorts, possibly accompanied by iden-
tificational attrition among younger adults, in any case constitutes the
underlying mechanism of Jewish population stagnation or decline.

DELLAPERGOLA 109
Figure 2: U.S. Jewish Population by Age, 1957-2001
0
100000
200000
300000
400000
500000
0-4 5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
1957
0
100000
200000
300000
400000
500000
0-4 5-910-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
1970
0
100000
200000
300000
400000
500000
0-4 5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
1990
0
100000
200000
300000
400000
500000
0-4 5-910-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
2001

110 CONTEMPORARY JEWRY
Table 3
U.S. Jewish Population, by Age, Actual and Projected (Thousands)a1957-2001
Age 1957
Actual 1970
Projected
(1957)
1970
Actual % Diff.
Actual-
Projected
1970
1990
Projected
(1970)
1990
Actual % Diff.
Actual-
Projected
1990
2001
Projected
(1990)
2001
Actual % Diff.
Actual.-
Projected
2001(90)
2001
Projected
(1970)
% Diff.
Actual-
Projected
2001(70)
Total 5,013 5,226 5,370 2.8 5,491 5,515 0.4 5,367 5,035 -6.2 5,282 -4.7
0-4 468 310 246 -20.7 322 398 23.8 250 264 5.9 224 18.0
5-9 436 360 355 -1.4 352 366 4.1 264 251 -5.2 254 -1.4
10-14 299 431 532 23.6 329 272 -17.5 383 280 -27.0 311 -9.9
15-19 292 466 520 11.6 272 263 -3.0 380 332 -12.7 349 -4.8
20-24 233 377 473 25.3 244 329 35.0 285 373 30.8 336 11.0
25-29 310 284 276 -2.8 352 431 22.5 256 307 19.7 281 9.1
30-34 350 262 250 -4.7 525 448 -14.7 309 316 2.5 238 32.8
35-39 370 249 306 23.1 512 479 -6.3 412 297 -27.9 317 -6.3
40-44 359 324 344 6.3 463 470 1.6 439 335 -23.7 496 -32.4
45-49 470 349 389 11.5 268 344 28.5 465 390 -16.0 509 -23.4
50-54 388 341 391 14.7 239 229 -4.0 467 439 -6.0 478 -6.7
55-59 319 370 357 -3.5 285 250 -12.3 355 278 -21.7 284 -2.1
60-64 216 399 286 -28.2 308 237 -23.0 225 217 -3.5 217 -0.3
65-69 166 282 240 -14.9 325 283 -13.2 212 234 10.4 244 -4.0
70-74 138 211 196 -7.0 291 249 -14.5 208 253 21.4 255 0.2
75+ 200 210 207 -1.5 405 466 14.9 456 469 2.8 497 -5.6
Median 36.6 37.4 35.5 38.4 37.6 41.6 41.5 43.3
a Minor discrepancies due to rounding. Projection data without international migration.

DELLAPERGOLA 111
The same trends are examined in greater detail in Table 3 and in Figures
3, 4, and 5, with an eye both to the substantive changes in population
composition and the succession over time of different data sources of
unequal quality and each with its own idiosyncrasies. Each of the three
figures compares detailed age compositions at an initial date and then at
a later date both for population projections from that initial date and data
obtained by new research. Comparisons are shown for 1970 vs. 1957;
1990 vs. 1970; and 2001 vs. 1990. One general conclusion that applies
to the three comparisons is that whereas the actual age compositions at
each period’s beginning and end are quite different, the projected and
actual data at period’s end appear relatively more germane. If we take a
difference of more than 20% between the actual versus projected cohort
sizes as a standard of clear inconsistency, then out of a total of 16 five-
year cohorts, five fall out of range in 1970 (three on the high side, mean-
ing the actual numbers are bigger than the projected ones, and two on
the low side); five cohorts in 1990 (four on the high side and one on the
low side); and six cohorts in 2001 (two on the high side and four on the
low side). Interestingly, if we address a projection from NJPS 1970 to
NJPS 2001, the correspondence with NJPS 2001 is much better than in
the case of the NJPS 1990 projection: Only three age cohorts fall out of
a range of error of plus-minus 20% (one on the high side and two on the
low side).
The overall conclusion is that allowing for gradual improvements
in health standards and life expectancy, and recalling that these compar-
isons do not take into account immigration to the United States, the
actual survey data seem generally in line with an assumption of persis-
tently low Jewish fertility and its consequences for Jewish population
size and structure. Moreover, since the data in Figures 3 to 5 do not
make allowance for net international migration, which should produce
visible population increases, this suggests that a counteracting determi-
nant of population decrease is at work. For our analytical purposes, this
can be termed identificational assimilation or erosion.
More detailed inspection of Figures 3 to 5 and Table 3 reveals some
inconsistencies. As compared to the 1957 CPS projected data, the 1970
NJPS obtains a particularly large excess of people aged 10-14 (born
1955-60), 20-24 (born 1945-50), and 35-39 (born 1930-35). Twenty
years later, the 1990 NJPS displays an underestimate of the expected
number of people 30-34 (born 1955-60), and 55-59 (born 1930-35). This
suggests that in the 1970 NJPS those cohorts may have been overesti-
mated in the first place. In the 2001 NJPS, the same 1955-60 cohort (now
aged 40-44) emerges again as one of the most underestimated compared
to the projection of NJPS 1990. That points either to the survey’s inabil-
ity to reach members of this specific age cohort, or to a continuously
declining propensity of its members to identify Jewishly, or both.

112 CONTEMPORARY JEWRY
Figure 3: U.S. Jewish Population by Age 1957 Actual, 1970 Projected (1957) and 1970 Actual
0 100000 200000 300000 400000 500000
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
1970 Actual
1970 Projected
1957 Actual

DELLAPERGOLA 113
Figure 4: U.S. Jewish Population by Age 1970 Actual, 1990 Projected (1970) and 1990 Actual
0 100000 200000 300000 400000 500000
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
1990 Actual
1990 Projected
1970 Actual

114 CONTEMPORARY JEWRY
Figure 5: U.S. Jewish Population by Age 1990 Actual, 2001 Projected (1990) and 2001 Actual
0 100000 200000 300000 400000 500000
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75+
2001 Actual
2001 Projected
1990 Actual

DELLAPERGOLA 115
The slightly younger cohort born 1960-65 and aged 35-39 in 2001 also
would appear to have been significantly undercounted, while it appar-
ently had been overcounted in NJPS 1990 vs. the projected figures from
1970 NJPS. Other undercounts in 2001 possibly matched by overcounts
in 1990 involve the 10-14 and 55-59 age groups (born 1986-91 and
1941-46, respectively). In contrast, apparent overcounts of a cohort at a
later date matched by apparent undercounts of the same cohort at an ear-
lier date involve those aged 20-24 in 1990 (born 1965-70), and those
aged 20-24 and 70-74 in 2001 (born 1976-81 and 1926-31, respec-
tively).
These examples illustrate a certain imperfection in the correct
matching of detailed age data when moving from one data source to
another. The apparent alternation of overcounts and undercounts cannot
be too surprising in view of sampling errors, but also in view of the dif-
ferent methods and response rates of different surveys. A further cause
of inaccuracy is the lack of an exact time reference point for most of the
surveys, most of which are conducted over a period of one year or more.
In spite of these limitations, the coherence of the overall picture allows
the reader to draw conclusions about the major thrust of demographic
trends with sufficient degree of confidence. Based on 16 age cohorts,
the average absolute magnitude of inaccuracy when matching actual and
projected cohort data (Table 3) results as follows: 1970 vs. 1957, 12.5%;
1990 vs. 1970, 14.9%; 2001 vs. 1990, 14.8%; 2001 vs. 1970, 10.5%.
Overall, when matching actual vs. projected age data the NJPS 2001
degree of accuracy does not look exceptional in comparison with the
previous two national surveys of 1970 and 1990.
Figure 6 summarizes the main gist of the preceding discussion by
comparing the total percent share of the 0-14 and 65+ age groups
between 1957 and 2001. Data reflect actual survey results and projec-
tions. The clear outcome of an aging population translates into a rever-
sal of the relative predominance of the percent of children versus the
percent of the elderly. A relatively young Jewish population in the late
1950s and early 1960s turned into a relatively elderly one by 1990, and
significantly so in 2001. These findings again confirm the fairly good
consistency of different survey data spread over 44 years in portraying
the demographic development of U.S. Jewry. The 2001 age configura-
tion for the first time displays a higher percent of persons aged 65+ than
below 15. The median age of U.S Jews rose from 36.6 in 1957 and 35.5
in 1970 (vs. 27.9 among the total U.S. population in 1970) (Glick,
1974), to 41.5 in 2001 (vs. 35 for the total U.S. population). Such an age
configuration foreshadows further aging and population decline over the
next several years, unless very significant changes occur in levels of
Jewish fertility, international migration, or Jewish identificational reten-
tion.

116 CONTEMPORARY JEWRY
Figure 6: U.S. Jewish Population at Ages 0-14 and 65+ (Percent) 1957-2001
5
10
15
20
25
1957 1960 1970 1980 1990 2001
0-14(1957) 0-14(1970) 0-14(1990) 0-14(2001)
65+(1957) 65+(1970) 65+(1990) 65+(2001)

DELLAPERGOLA 117
International Migration
The population data and estimates reviewed so far only partly account
for international migration. Figure 7 and Table 4 indicate levels of Jew-
ish migration to the U.S. as reflected by various data sources. These
comprise data and estimates concocted from a variety of institutional
sources, such as HIAS (Hebrew Immigrant Aid Society) for Jews from
the former Soviet Union or the Immigration and Naturalization Service
of the Department of Justice (INS) for Israelis (HIAS; Campbell and
Lennon, 1999), as well as retrospective analyses from the 1990 and
2001 NJPSs. Data on migration to Israel (aliya) also are presented for
reference (Israel CBS). The 2001 NJPS estimated numbers of immi-
grants are based on the actual figures for male and female Jewish
respondents, and on an assumption of a similar number of immigrants
among non-respondent Jewish household members.
There has been significant variation in the volume of Jewish immi-
gration to the U.S. since 1945, and especially between 1957 and 2001.
Following relatively low initial levels of variation of only a few thou-
sand per year, immigration increased during the 1970s, especially from
the Soviet Union, decreased again during the 1980s, and significantly
grew during the first half of the 1990s. These fluctuations primarily
reflect variations in Jewish emigration opportunities and circumstances
in the countries of origin, but also variable quotas of admission into the
United States. A constant inflow of Jewish immigrants came from the
(former) Soviet republics, but also from Israel, Iran, South Africa, Latin
America, and other countries. With regard to the volume of Jewish emi-
gration from the United States, it is apparent that it corresponded only to
a small share of the volume of immigration. While Israel probably was
the main country of destination, there is some indirect evidence that
emigration to other countries followed patterns of growth and contrac-
tion similar to those of aliya to Israel (DellaPergola, Rebhun and
Raicher, 2000).
The NJPS 2001 international migration data compare quite well
with other data sources. The somewhat lower NJPS data can be
explained by their relating to the more restricted “core” definitional con-
cept (in its 2001 version), as against the “enlarged” population concept
of Jews and their non-Jewish family members reflected in immigration
data such as those from HIAS. NJPS 1990 data could not yet catch the
emerging migration exodus from the FSU, which actually began at the
end of 1989.
Once we add estimates of the immigration input to the population
projections already presented in Figure 1, we obtain the upgraded esti-
mates in Figure 8. The alternative preferred here was to allocate the
cumulative net impact of immigration as assessed through the 2001
NJPS. One interesting pattern emerges out of the generally good fit

118 CONTEMPORARY JEWRY
Figure 7: Estimates of U.S. Jewish Immigration and Aliyah Various Sources, 1957-2000
0
25
50
75
100
125
150
175
200
(Thousands)
1957-60 1961-65 1966-70 1971-75 1976-80 1981-85 1986-90 1991-95 1996-00
Current (enlarged
)
1990 (core)
2001 (core)
Aliyah

DELLAPERGOLA 119
between the various estimates and projections of Jewish population size
adjusted for migration. The 1970 NJPS original figures and subsequent
projections with net migration look somewhat higher than an ideal line
that would connect the 1957 and 1990-based series of estimates. One
explanation might be that the 1970 NJPS slightly overestimated the U.S.
Jewish population. This is not implausible if we recall that NJPS 1970
partly relied on random sampling and partly on Jewish lists, whereas the
1957 CPS used a random national sample, and the 1990 NJPS was the
Jewish survey that more than others stressed a purely random sampling
procedure.
One additional piece of information appended to Figure 8 is derived
from an earlier set of Jewish population projections by ICJ researchers,
based on a 1970 NJPS baseline extended to 1975 (Schmelz and Della-
Pergola, 1983). Of the eight different scenarios that originally were
obtained according to different assumptions of fertility, international
migration, and assimilation, the two suggesting intermediate population
levels are displayed in Figure 8. The highest and lowest population esti-
mates expected by these eight scenarios for the year 2000 were
5,571,000 and 4,639,000, respectively. The two intermediate scenarios
were “A: low fertility, moderate assimilation, moderately positive
Table 4
Estimates of U.S. Jewish International Migration
1957-2000
Year Current
estimatesa1990 NJPS
core Jewish
population
2001 NJPS
Jewish
respondents
2001 NJPS
To tal
estimateb
Aliyah
to
Israelc
1957-60 21,900 21,200 16,300 32,600 1,400
1961-65 44,300 35,000 21,500 43,000 3,700
1966-70 38,900 23,700 10,700 21,400 14,500
1971-75 38,800 21,200 17,500 35,000 23,200
1976-80 101,400 116,100 37,100 74,200 13,500
1981-85 60,000 38,700 16,500 33,000 13,000
1986-90 141,700 25,000 37,600 75,300 8,100
1991-95 175,200 80,400 160,900 9,800
1996-00 70,000 48,300 96,600 8,200
a Sources: Diamond (1977); HIAS.
b Assumes total immigrants to be twice the number of immigrant
respondents.
c Israel Central Bureau of Statistics.

120 CONTEMPORARY JEWRY
Figure 8: U.S Jewish Population Various Estimates (Thousands) 1957-2001
Note: 1957+netmigr = Projection
based on 1957 CPS plus migration
estimate; 1970+netmigr = Projection
based on 1970 NJPS plus migration
estimate; 1990+netmigr = Projection
based on 1990 NJPS plus migration
estimate; NJPS = National Jewish
Population Survey; AJIS = American
Jewish Identity Survey; SchmelzDPa
= Schmelz and DellaPergola (1983),
model a; SchmelzDPd = Schmelz
and DellaPergola (1983), model d.
4500
4750
5000
5250
5500
5750
6000
1957 1960 1965 1970 1975 1980 1985 1990 1995 2000
1957+netmigr 1970+netmigr 1990+netmigr NJPS 2001
AJIS 2001 SchmelzDPa SchmelzDPd

DELLAPERGOLA 121
migration balance,” with 5,321,000 people (SchmelzDPa model in
Figure 8), and “D: stronger assimilation, moderately positive migration
balance, somewhat higher fertility” with 5,178,000 people
(SchmelzDPd model in Figure 8). In reality, international migration
happened to be stronger than had been predicted, but Jewish fertility
turned to be weaker. Assimilation, as reflected by rising rates of out-
marriage, actually increased as predicted. It is intriguing to find that
despite these forecasting imperfections, these two intermediate scenar-
ios ended by falling exactly on the target determined by the 2001 higher
AJIS and lower NJPS estimates.
Concluding Remarks and Further Reflections
When the author of a New York Times Op-Ed piece announced that “the
[NJPS 2001] false population decline must be corrected before it further
sours communal discourse” (Goldberg, 2003), was the intention to casti-
gate an incompetent and unusable survey or to cast doubt on the possi-
bility that Jewish population in the United States might ever decline?
Beyond the skirmishes in the American Jewish and general media about
the quality or even legitimacy of NJPS, our preceding analysis suggests
a number of conclusions for sociodemographic research and Jewish
community policies.
First, inasmuch as American Jews still constitute not only a cultural
construct but also a population, it is appropriate to apply demographic
concepts and tools to assess them. The evidence indicates that while the
absolute Jewish population size is difficult to ascertain with exactitude
and probably falls within a range of variable sizes depending on the
quality of available data, the momentum of growth reached its peak
around 1990. What has followed since and is bound to follow in the
foreseeable future is moderate population decline (see also the indepen-
dent assessments in Klaff, 1998; DellaPergola, Rebhun, Tolts, 2000).
The continuous stretching of the definitional criteria for inclusion not-
withstanding, the reduction in the number of Americans who are willing
to identify as Jews or persons of Jewish background seems to be predict-
able and real. This conclusion is not another version of the “lachrymose
theory of Jewish history” (Rawidowicz, 1986; Goldscheider, 2002), but
a straightforward consequence of demographic dynamics cumulated
over several decades. Following the continuous and accelerating rise in
the median age of American Jews, there is hardly any doubt that the
number of Jewish deaths in the U.S. is now higher than the number of
newborn children whose parents impart them with a Jewish identity (of
any sort). Consequently, the effectively low Jewish birth rate reflects
more the consequences of marriage patterns and frequent religious het-
erogamy than a fertility level significantly different from the peer non-
Jewish population (Phillips, 1997; Waite, 2002; Kotler-Berkowitz et al.,

122 CONTEMPORARY JEWRY
2003). It is also true, on the other hand, that U.S. Jewry is not one homo-
geneous block. Among its different constituencies one discerns patterns
of growth among the relatively small Orthodox sector of the population
and patterns of contraction among other population sectors (DellaPer-
gola and Rebhun, 1998-1999).
Besides population aging, a second factor in Jewish population
decline is a blurring of the wish to identify as Jews. In the 2001 NJPS,
the surveyors assessed 4.3 million Jews and another 800,000 persons of
Jewish background. It surely was a mistake not to ask more questions
about the Jewish identity of the representative sample of these 800,000
more weakly identified respondents. Answers might have revealed
some interesting links between some of the marginally identified at the
Jewish community periphery and persistent patterns of Jewish identity,
as indeed was the case with the 1990 NJPS (DellaPergola, 1991). It may
be assumed, however, that in most instances a lackluster response would
have been received in 2001. It therefore can be argued that the survey’s
partial disregard of the identificationally weaker cases actually created
an overly Jewish communal profile based on the respondents.
Even if part of the apparent Jewish population decline reflects
undercoverage, it seems plausible that those less eager to participate in a
survey sponsored by a clearly identifiable Jewish organization include a
disproportionate share of the more weakly identified. Jewish survey
response is, among other things, an indicator of Jewish identification.
Our discovery of the unique patterns of declining identification and/or
undercounting among the 1955-60 cohort calls for the formulation of
analytic hypotheses about the underlying causal determinants. The spe-
cific combination of period and lifecycle circumstances experienced by
this particular cohort may conceivably have affected both its identifica-
tional patterns and its availability for in-depth interviewing. A follow-
up study particularly designed to investigate this specific population
group might provide interesting clues that could help evaluate NJPS
2001 as well as previous surveys.
Cross-comparisons should be undertaken between regional sub-
totals obtained from the sum of local studies and from NJPS 2001. They
should not be limited to simple population totals but extended to charac-
teristics such as age, major indicators of Jewish identification and, most
significantly, measures of migration within the United States. After per-
forming these tasks, we will be in a better position to evaluate the
respective merits of NJPS 2001 in comparison with other local and
national Jewish population studies.
The present overview clearly has stressed the unequal quality of the
available databases. NJPS 2001 cannot escape its own share of critiques,
and probably also the need for some statistical adjustments as against
the originally published data. However, it would be a serious mistake to

DELLAPERGOLA 123
focus exclusively on one source. A broader historical prospective on the
data is needed to better understand not only the substantive issues, but
also the respective strengths and weaknesses of various data sources. At
the end of our overview, it may be possible to infer that the 1957 CPS
and the 1990 NJPS were more broadly representative of the characteris-
tics of the total U.S. Jewish population, whereas the 1970 NJPS and the
2001 NJPS (at least as assessed through the initial procedures for data-
weighting) tended to be somewhat more representative of the more Jew-
ishly identified sections. NJPS 2001, in any case, belongs in the family
of important, legitimate, and usable sources.
Of course, an intriguing question is what our findings imply for the
future of Jewish population research in the United States. The substan-
tive answer cannot be separated from a few additional reflections stem-
ming from the experience with NJPS 2001. These reflections involve
several sensitive aspects in the research-oriented and public domains.
Topics calling for evaluation extend well beyond the technical quality
attained in the 2001 study, its validity on methodological grounds, its
comparability with other studies, and the credibility of its main conclu-
sions—namely, an apparent reduction in Jewish population size and
assessments of its underlying causes. Frank discussion, sincere intro-
spection, and clear directives for the future are necessary concerning:
(a) the need for substantial investments in research as a basis for Jewish
community planning and management; (b) the significant amount of
resources invested in the 2001 NJPS by an American Jewish community
facing other competing needs; (c) the ability of the organized Jewish
community to withstand objective scrutiny of its own trends and charac-
teristics and to learn from scientific research; (d) the role of the United
Jewish Communities as the central pillar of the American Jewish com-
munity, and hence the leading body responsible for major research and
policy investments; (e) the analytic goals and contents to be preferred
for systematic community studies, namely the respective merits and
complementarity of qualitative vs. quantitative research orientations,
and of sociodemographic vs. policy-oriented research; (f) the profes-
sional competences and decisional autonomy of the panel of investiga-
tors designed to design, supervise, and direct the study; (g) the policy
conclusions to be drawn following research, and the new investments
needed to translate understanding of the facts into community programs;
(h) the mechanisms and tools of policy decision-making facing possible
gaps between the results of systematic research and existing pre-concep-
tions among public opinion makers or lay leaders; (i) the issue of the
relationship. and sometimes conflict of interests, between research-
sponsoring organizations and the community of professional investiga-
tors.

124 CONTEMPORARY JEWRY
The last of these topics deserves special attention, as it seldom has
been discussed in the framework of Jewish studies. Researchers, or at
least some among them, depend in their work on resources provided by
sponsoring organizations. The mode of operating of the sponsor-
researcher relationship and the amount of independence available to
investigators may critically affect the results of research; indeed, it cru-
cially shaped the unfolding of the 2001 NJPS. This and each of the other
abovementioned items are worth detailed consideration well beyond the
purposes of this article.
One important lesson to be drawn is that a keen and systematic
reconsideration of what went right and what went wrong with NJPS
2001 is necessary, along with a serious and careful reading of its results.
In particular, all those concerned should be ready to consider the two
following propositions: (a) the emerging long-term demographic trends
challenge the assumption of continuing and level population growth or
stability among U.S. Jewry; (b) the complications in defining, identify-
ing, and reaching the Jewish constituency for research purposes now
have approached the limits of the possible. Short of acknowledging
these two facts, NJPS 2001 may have offered the last opportunity for the
scholarly profession, for the U.S. Jewish community leadership, and for
the public at large to have at their disposal such a large-scale and com-
plex tool for factual assessment and policy planning.
Acknowledgments
The author was a member of the National Technical Advisory Commit-
tee of the 1990 NJPS and 2001 NJPS. Research for this paper was con-
ducted at the Division of Jewish Demography and Statistics of the A.
Harman Institute of Contemporary Jewry, at the Hebrew University of
Jerusalem. The author thanks his colleagues Uzi Rebhun, Mark Tolts,
Dalia Sagi, Beny Anderman, and Judith Even for invaluable help and
advice. David Marker, Vivian Klaff, and two anonymous reviewers pro-
vided useful suggestions to this paper.

DELLAPERGOLA 125
REFERENCES
Anderson, B. 1991. Imagined Communities. London and New York:
Verso.
Barack Fishman, S. 2004. Double or Nothing? Jewish Families and
Mixed Marriage. Hanover and London: Brandeis University Press.
Campbell, G. and E. Lennon. 1999. Historical Census Statistics in the
Foreign-Born Population of the United States: 1850 to 1990. Wash-
ington, D.C.: U.S. Bureau of the Census.
Cohen, S.J.D. 1999. The Beginnings of Jewishness: Boundaries, Variet-
ies, Uncertainties. Berkeley: University of California Press.
Cohen, S.M., J.S. Woocher and B.A. Phillips, (eds.). 1984. Perspectives
in Jewish Population Research. Boulder/London: Westview.
Cohen, S.M. and A.M. Eisen. 2000. The Jew Within: Self, Family and
Community in America. Bloomington and Indianapolis: Indiana
University Press.
Corinaldi, M. 2001. The Enigma of Jewish Identity: The Law of Return,
Theory and Practice. Serigim-Leeon: Nevo (Hebrew).
DellaPergola, S. 1980. “Patterns of American Jewish Fertility.” Demog-
raphy, 17, 3, 261-273.
_______. “New Data on Demography and Identification among Jews in
the U.S.: Trends, Inconsistencies and Disagreements.” Contempo-
rary Jewry, 12, 67-97.
_______. 2000. “Thoughts of a Jewish Demographer in the Year 2000.”
Contemporary Jewry, 21, 98-116.
_______. 2002. “Demography.” In M. Goodman with J. Cohen and D.
Sorkin (eds.) The Oxford Handbook of Jewish Studies. Oxford:
Oxford University Press, 797-823.
_______. 2003a. “Good Data, Troubling Numbers.” Jerusalem Post,
September 11, 2003, 9.
_______. 2003b. “World Jewish Population, 2003.” American Jewish
Year Book, 103, 588-612.

126 CONTEMPORARY JEWRY
DellaPergola, S. and U. Rebhun. 1998-1999. “American Orthodox
Jews: Demographic Trends and Scenarios.” Jewish Action, Fall
5759, 30-33; Spring 5759, 82-83.
DellaPergola, S., U. Rebhun and R.P. Raicher. 2000. “The Six Day War
and Israel-Diaspora Relations: An Analysis of Quantitative Indica-
tors.” In E. Lederhendler (ed.) The Six-Day War and World Jewry.
Bethesda: University Press of Maryland, 11-50.
DellaPergola, S., U. Rebhun and M. Tolts. 2000. “Prospecting the Jew-
ish Future: Population Projections, 2000-2080.” American Jewish
Year Book, 100, 103-146.
Dencik, L. 2002. “Homo Zappiens” – A European-Jewish Way of Life
in the Era of Globalisation.” Paideia, 1-34.
Diamond, J.J. 1977. “A Reader in the Demography of American Jews.”
American Jewish Year Book, 77, 251-319.
Engelman, U.Z. 1947. “Jewish Statistics in the U.S. Census of Religious
Bodies (1850-1936).” Jewish Social Studies, 9, 127-174.
Glick, P.C. 1960. “Intermarriage and Fertility Patterns among Persons in
Major Religious Groups.” Eugenics Quarterly, 7.
Glick, P.C. (ed.) 1974. The Population of the United States. Washington,
D.C.: U.S. Government Printing Office.
Goldberg, J.J. 2003. “A Jewish Recount.” The New York Times, Septem-
ber 17, 2003, 27.
Goldscheider, C. 1982. “Demography of Jewish Americans: Research
Findings, Issues, and Challenges.” In M. Sklare (ed.) Understand-
ing American Jewry. New Brunswick-London: Transaction Books,
1-55.
_______. 2002. “Social Science and the Jews: A Research Agenda for
the Next Generation.” Contemporary Jewry, 23, 196-219.
Goldstein, A. 1986. “Patterns of Mortality and Causes of Death among
Rhode Island Jews.” Social Biology, 33, 87-101.

DELLAPERGOLA 127
Goldstein, S. 1969. “Socioeconomic Differentials among Religious
Groups in the United States.” The American Journal of Sociology,
74, 6, 612-631.
_______. 1981. “Jews in the United States: Perspectives from Demogra-
phy.” American Jewish Year Book, 81, 1981, 3-59.
_______. 1989. “American Jewish Demography: Inconsistencies that
Challenge.” In U.O. Schmelz and S. DellaPergola (eds.) Papers in
Jewish Demography 1985. Jerusalem: The Hebrew University, 23-
42.
_______. 1992. “Profile of American Jewry: Insights from the 1990
National Jewish Population Survey.” American Jewish Year Book,
92, 77-173.
Goldstein, S. and A. Goldstein. 1996. Jews on the Move: Implications
for Jewish Identity. Albany: State of New York University Press.
Goldstein, S. and B. Kosmin. 1992. “Religious and Ethnic Self-Identifi-
cation in the United States 1989-90: A Case Study of the Jewish
Population.” Ethnic Groups, 9, 219-245.
Good, D. 1959. “Questions on Religion in the United States Census.”
Population Index, 25, 3-16.
Grabill, W.H., C.V. Kiser and P.K. Whelpton. 1958. The Fertility of
American Women. New York: J. Wiley.
Herman, S. 1977. Jewish Identity: A Social Psychological Perspective.
Beverly Hills and London: Sage Publications.
HIAS. Statistical Report. New York: The Hebrew Immigrant Aid Soci-
ety (annual).
Horowitz, B. 2002. “Reframing the Study of Contemporary American
Jewish Identity.” Contemporary Jewry, 23, 14-34.
Israel Central Bureau of Statistics. Statistical Abstract of Israel. Jerusa-
lem (annual).
Klaff, V. 1998. “Broken Down by Sex and Age: Projecting the Ameri-
can Jewish Population.” Contemporary Jewry, 19, 1-37.

128 CONTEMPORARY JEWRY
Kosmin, B.A., S. Goldstein, J. Waksberg, N. Lerer, A. Keysar and J.
Scheckner. 1991. Highlights of the CJF 1990 National Jewish Pop-
ulation Survey. New York: Council of Jewish Federations.
Kotler-Berkowitz, L., S.M. Cohen, J. Ament, V. Klaff, F. Mott, and D.
Peckermen-Neuman. 2003. The National Jewish Population Survey
2000-01: Strength, Challenge and Diversity in the American Jewish
Community. New York: United Jewish Communities.
Lazerwitz, B. 1978. “An Estimate of a Rare Population Group – The
U.S. Jewish Population.” Demography, 15, 3, 1978, 389-394.
Liebman, C.S. 2001. “Some Research Proposals for the Study of Amer-
ican Jews.” Contemporary Jewry, 22, 99-114.
Massarik F. 1974. “National Jewish Population Study: A New United
States Estimate.” American Jewish Year Book, 75, 296-304.
_______. 1977. “The Boundary of Jewishness: Some Measures of Jew-
ish Identity in the United States.” In U.O. Schmelz and S. DellaPer-
gola (eds.) Papers in Jewish Demography 1973. Jerusalem: The
Hebrew University, 117-139.
Mayer, E., B. Kosmin, and A. Keysar. 2002. American Jewish Identity
Survey: An Exploration in the Demography and Outlook of a Peo-
ple. New York: The Graduate Center of the City University of New
York.
Mott, F. and J. Abma. 1992. “Contemporary Jewish Fertility: Does Reli-
gion Make A Difference?” Contemporary Jewry, 13, 74-94.
Phillips, B.A. 1997. Re-examining Intermarriage: Trends, Textures,
Strategies. New York, The Susan and David Wilstein Institute of
Jewish Policy Studies, and the American Jewish Committee, The
William Petschek National Family Center.
Rawidowicz, S. 1986. Israel the Ever-Dying People and Other Essays.
Rutherford: Farleigh Dickinson University Press.
Rebhun, U. 1997. “Similarities and Dissimilarities in National and
Community Surveys: The Case of American Jews.” In S. DellaPer-
gola and J. Even (eds.) Papers in Jewish Demography 1993 in
Memory of U.O. Schmelz. Jerusalem: The Hebrew University, 55-
78.

DELLAPERGOLA 129
Rebhun, U., S. DellaPergola and M. Tolts. 1999. “American Jewry: A
Population Projection, 1990-2020.” In R. Rosenberg Farber and
C.I. Waxman, Jews in America: A Contemporary Reader. Hanover
and London: Brandeis University Press, 33-50.
Ritterband, P., B.A. Kosmin and J. Scheckner. 1988. “Counting Jewish
Populations: Method and Problems.” American Jewish Year Book,
88, 204-221.
Robison, S. 1943. Jewish Population Studies. New York: Conference on
Jewish Relations, Jewish Social Studies, 3.
Rosenberg Farber, R. and C.I. Waxman. 1999. Jews in America: A Con-
temporary Reader. Hanover and London: Brandeis University
Press.
Rosenthal, E. 1975. “The Equivalence of United States Census Data for
Persons of Russian Stock or Descent with American Jews: An Eval-
uation.” Demography, 12, 275-290.
Rosenwaike, I. 1980. “A Synthetic Estimate of American Jewish Popu-
lation Movement over the Last Three Decades.” In U.O. Schmelz
and S. DellaPergola (eds.) Papers in Jewish Demography 1977.
Jerusalem: The Hebrew University, 83-102.
_______. 1994. “Causes of Death among Elderly Jews in New York
City, 1979-1981.” International Journal of Epidemiology, 23, 2,
327-332.
Saxe, L. and C. Kadushin. 2003. “Questioning the Validity.” Jewish
Week, September 19, 2003.
Schmelz, U.O. 1969. “Evaluation of Jewish Population Estimates.”
American Jewish Year Book, 70, 273-288.
_______. 1981a. “Jewish Survival: The Demographic Factors.” Ameri-
can Jewish Year Book, 81, 61-117.
_______. 1981b. World Jewish Population – Regional Estimates and
Projections. Jerusalem: The Hebrew University.
Schmelz, U.O. and S. DellaPergola. 1982. “World Jewish Population.”
American Jewish Year Book, 82, 277-290.

130 CONTEMPORARY JEWRY
_______. 1983. “The Demographic Consequences of U.S. Jewish Popu-
lation Trends.” American Jewish Year Book, 83, 141-187.
_______. 1988. Basic Trends in American Jewish Demography. New
York: The American Jewish Committee, Jewish Sociology Papers.
Schnapper, D. 1987. “Les limites de la démographie des Juifs de la
Diaspora.” Revue française de sociologie, 28, 319-332.
Schulman, M. 2003. “National Jewish Population Survey 2000-2001:
Study Review Memo.” New York: United Jewish Communities.
Schwartz, J., J. Scheckner and L. Kotler-Berkowitz. 2002. “Census of
U.S. Synagogues, 2001.” American Jewish Year Book, 102, 112-
150.
Seligman, B.B. 1950. “The American Jew: Some Demographic Fea-
tures.” American Jewish Year Book, 51, 3-52.
Sheskin, I. 2001. How Jewish Communities Differ: Variations in the
Findings of Local Jewish Population Studies. New York: American
Jewish Data Bank.
_______. 2003. “NJPS 2000-01 by the Numbers.” Forward, September
26, 2003.
The Editors 2003. “Jewish Population in the United States, 2002.”
American Jewish Year Book, 103, 155-179.
U.S. Bureau of the Census 1958. “Religion Reported by the Civilian
Population in the United States, March 1957”, Current Population
Reports, Population Characteristics, Series P-20, No. 79. Washing-
ton, D.C.
U.S. Bureau of the Census 1968. “Tabulations of Data on the Social and
Economic Characteristics of Major Religious Groups, March
1957.” Washington, D.C..
Verbit, M.F. 2002. “A Research Agenda for American Jewry.” Contem-
porary Jewry, 23, 1-13.
Waite, L.J. 2002. “The American Jewish Family: What We Know, What
We Need to Know.” Contemporary Jewry, 23, 35-63.

DELLAPERGOLA 131
Waksberg, J. 1996. “The Methodology of the National Jewish Popula-
tion Survey.” In S. Goldstein, A. Goldstein, Jews on the Move:
Implications for Jewish Identity. New York: SUNY Press, 333-359.
Winter, J.A. and L.I. Levin (eds.). 1984. Advancing the State of the Art:
Colloquium on Jewish Population Studies. New York: Council of
Jewish Federations, 2 vols.