Higher Education and Political Attitude Change:
Is Israel the Exception that Proves the Rule?
Kenneth D. Wald
Danielle Feinstein
University of Florida
Contact Information for Corresponding Author (Wald):
Department of Political Science
University of Florida
POB 117325
Gainesville, FL 32611-7324
Email: [email protected]
Fax: 352-392-8127
ABSTRACT
Israel presents a critical test of the well-established finding from other nations that higher
education strongly influences the political attitudes of young people. Most Israelis attend
university only after completing military service and are thus exposed to a powerful socializing
institution before beginning higher education. This sequence may pre-empt the influence of
university experience by socializing IDF members to durable political attitudes while they are
serving. By means of a survey of Israeli university students who entered higher education after
military service, we test whether exposure to higher education changes attitudes to Israeli Arabs
and the Israeli-Arab conflict.
KEYWORDS
tolerance, socialization, university, Arab-Israeli conflict, military
July, 2010
1
Despite calls to approach Israel as just another case in comparative political analysis, political
scientists often focus on the unique and arguably incomparable qualities of the Jewish state. Yet
as Yehezkel Dror (1996, 254) emphasizes, “recognizing Israel‟s uniqueness does not negate the
possibility of using Israel as a laboratory for employing . . . theories from which some
generalizations might be derived.” In fact, as comparativists have noted, deviant cases
sometimes provide the exception that proves the rule. Following that assumption, we draw on
one “deviant” aspect of Israel—its system of universal military service--to test general theories
of political education. Specifically, this study examines whether the prior experience of military
service by young Israelis diminishes or obviates the role of universities in political learning.
Studies conducted in the United States and other western societies have routinely emphasized the
importance of institutions of higher learning in shaping the political perspectives of young adults.
Because they are important venues frequented by late adolescents and young adults when they
are cognitively open to external influences, universities often produce notable changes in
political and social attitudes among their students. The association between individual-level
education and political attitudes has aptly been described as "one of the most stable and
consistent findings in empirical social research of contemporary American society" (cited in Dey
1997, 400). Israel may be different. Because it maintains a large military based on conscription,
most young Israelis postpone higher education until the completion of a lengthy period of
military service. At an age when most of their peers outside Israel are immersed in universities
and confronting new ideas and attitudes, young Israelis are typically exposed to another
institution with considerable socializing potential--the military. Given the intensity of military
service, one might well assume that young Israelis enter universities well-formed politically from
that experience and, hence, relatively impervious to the socializing influence of higher education.
We test that research question with survey data gathered at an Israeli university.
The first section of the paper summarizes the literature on higher education as a factor in political
socialization, focusing on the general impact of university education on political attitudes and the
mechanisms hypothesized to produce these effects. Most of this research comes from North
America. The next section concentrates specifically on Israel, summarizing the literature on
political socialization in the military and higher education. We then describe the research design,
process of data collection, and major variables included in this study. The paper reports the
results of our analysis and concludes with the implications of the Israeli findings for broader
theories of political education.
Political Socialization and Higher Education
Driven by an assumption that “child is father to the man,” the pioneering political socialization
studies emphasized the role of family and early childhood experiences in the development of
adult political orientations (Hess and Torney 1967; Easton and Hess 1962). These cross-sectional
studies reported strong correspondence between the political attitudes of pre-adult children and
their parents (Dalton 1980). When longitudinal research designs were employed, however, the
findings undercut the determinism of these early approaches, demonstrating relatively little
carryover between childhood and adult political orientations (Jennings and Niemi 1968; Searing,
Wright & Rabinowitz 1976). Under what became known as the “lifelong openness” model of
political education (Sigel 1989), scholars began to push the crucial period of socialization from
2
the elementary school years to late adolescence and early adulthood, a time when people tended
to develop a strong sense of personal and political identity. Of course, the very concept of
lifelong openness emphasizes a plasticity in political attitudes that makes it problematic to refer
to any particular period of life as crucial. Nonetheless, the nature of late adolescence/early
adulthood as a time associated with identity development and exploration—the impressionable
years, in short—makes that phase of life particularly conducive to political attitude change (Sears
and Levy 2003).
Newcomb et al‟s Bennington studies (Newcomb 1958; Alwin, Cohen and Newcomb 1992)
represent the classic expression of this perspective. Beginning in the 1930s, Newcomb began
tracking the political and social attitudes of several cohorts of women who entered an all-female
liberal arts college in rural Vermont. Following these students over time and well into adulthood,
he found two major patterns. First, the students‟ political attitudes changed substantially over the
course of their matriculation, moving from generally conservative to more liberal political
orientations. Second, the political worldviews developed in college showed remarkable staying
power over the lifespan. While there was some reversion to earlier perspectives after leaving
university, students who interacted with politically likeminded people as adults (principally
spouses) tended to retain the political views they acquired in college.
Because of the unique aspects of Bennington, a relatively isolated environment where largely
middle class students first encountered ideological diversity, one may wonder about the
generalizability of these findings. However, subsequent research with more diverse and
representative samples has tended to reinforce both the critical importance of early adulthood
and the college years as crucibles of sociopolitical learning and the long-term retention of
political outlooks developed during that stage of life (Guimond 1999, Jennings 1987, Lottes and
Kuriloff 1994, Pascarella and Terenzini 1991, Sears and Funk 1999, Sears and Levy 2003,
Jennings and Stoker 2008, Stubager 2009).
What mechanisms account for the recurrent finding of political attitude change among students
during higher education? For some scholars, universities are environments where young people
encounter new ideas, meet people with different backgrounds, are exposed to perspectives rare or
previously unavailable to them in their prior environments, and internalize them. In effect, this
“zeitgeist” model emphasizes that students are open to the political cues in the university
environment because the very nature of higher education emphasizes cognitive freedom,
intellectual experimentation, and exploration. During their years at university, then, students are
exposed to persuasive arguments and perspectives that were previously unfamiliar to them.
According to the literature, the major sources of new ideas and perspectives that students
encounter in university are peers and professors (Milem 1998; Guimond 1999). Ripe for social
influence because of their youth and cognitive openness, university students may well respond to
both mechanisms and thus experience attitudinal change about politics.1
The Israeli Context
Because of its constant security concerns and reliance on conscription, Israel has a much higher
proportion of citizens who serve in the armed forces than most societies. As a rule, women serve
two years and men from three to four depending upon their rank and unit. Such service following
3
the completion of secondary education has long been considered a central part of the citizenship
process and a venue where Israelis are socialized to the dominant codes of their society. From
statehood onward, service in the Israel Defense Forces (IDF) have been conceived as an
important agent of national integration and instrumental in forging Israeli identity.
Might such military service affect political attitudes and outlooks, serving as an agent of political
socialization in Israel? Through a variety of mechanisms, the literature on military sociology
suggests, “military service leads individuals to reconsider their identity, their attachments, and
the definition of their political community, bringing these into accord with their personal
experiences and hence with military policy” (Krebs 2004, 87). That makes it plausible to assume
that students who have been exposed to concentrated military service before higher education
might well be relatively well-formed politically by the time they first encounter the university
and thus develop resistance to whatever conflicting political cues are transmitted within
academe.2
Although there is no shortage of social scientific research on the impact of military service on
Israelis, most of that work examines psychological factors such as maturation or potential
influences on unit cohesion (e.g., Dar and Kimhi 2001). There is surprisingly little research about
the socializing influence of military service on the political and social attitudes of Israelis.3 In
fact, as Krebs (2004, 98) reports after reviewing this sparse literature, much of it classified and/or
unpublished, “no systematic empirical evidence supports" the widespread claim about the
political socializing influence of service in the Israeli military. This judgment applies even to
Arian‟s study which, utilizing a representative national sample of Israelis, found significant
sociopolitical attitude differences between Israelis who had and hadn‟t served in the IDF. Given
that IDF service is strongly related to gender, religiosity and nationality, it is difficult to
disentangle the causal influence of military service from the factors that predict it and thus
whether IDF service or the background traits correlated with it account for the political and
social differences observed between military veterans and those who did not serve.4
There is a similar shortage of research on the political impact of university experience in Israel.
While scholars have examined the impact of secondary education on a variety of political
orientations (Ichilov 1991), Israeli socialization research has not generally extended to the realm
of higher education. In one of the few such studies, Shapira et al, (1986) compared the
ideological identification and political system affect of a multi-university sample of Israeli
students with their counterparts in the West. While the authors found that Israeli students were
less countercultural than their peers elsewhere, the study did not attempt to discern whether
university experience itself accounted for the political dispositions of the Israeli students. Kedem
and Bar-Lev (1989) compared the political orientations of Israeli university students who had
participated in different youth movements during childhood. They found strong carryover
between the movement orientations encountered by their respondents during childhood and the
political values they later expressed as university students. This study hints at the relative
imperviousness of Israeli university students to political change during their period of exposure
to higher education. Nonetheless, neither of these studies enables us to reach strong conclusions
about whether military service renders Israeli university students resistant to political change
while they are studying for a degree.
4
Research Design
Existing data about the effects of higher education on political attitudes among Israelis are not
sufficient to yield definitive conclusions. Simple cross-sectional comparison of Israelis who did
and did not experience higher education suffers from indeterminacy because of selection effects.
Whether a person goes to university or not is strongly determined by ethnicity, national
identification, and religiosity, and other social background factors that influence sociopolitical
attitudes. Thus it is difficult to disentangle the independent effects of university experience from
these selection effects.5 The alternative design, a longitudinal analysis of birth cohorts across the
lifespan with periodic surveys of political attitudes, could in principle differentiate respondents
by whether they served in the IDF and whether they attended university. Because of the
prevalence of military service, however, the number of cases would have to be considerable to
isolate enough respondents who did not serve in the IDF prior to beginning higher education and
who resemble those who did in terms of background traits. In any case, such a survey has never
been conducted in Israel and the sheer costs of tracking respondents across the lifespan makes it
extremely unlikely that it will.
Faced with this reality, we opted instead for a cross-sectional design that tested the sociopolitical
attitudes of Israeli university students who had served in the IDF but varied in the amount of
time they had been exposed to higher education. This approach controls for selection effects in
military service and attendance at university by excluding respondents who did not serve in the
military before attending university. Because members of our student sample differ in terms of
such important political influences as gender, ethnicity, religiosity and other factors, we are able
to discern the net impact of additional exposure to higher education after controlling for these
social forces. This design does not allow us to test directly for the effects of military service on
political attitudes because we lack data on university students who did not serve in the IDF prior
to matriculation. Thus the results only allow inferences about what mechanisms account for the
presence or absence of any observed attitude change among Israeli university students who
entered higher education after discharging their military obligations.
The study was conducted during May and June of 2009 at Haifa University. Unable to obtain a
list of university students that would have permitted random selection of respondents and
unwilling to sample from a haphazard set of specific classes, the sample was instead recruited via
venue sampling, a variant of time-space sampling (TSS) typically utilized by researchers to reach
a “hidden” or “hard-to-reach population” (Muhib et. al 2001, 216). This technique of locating
respondents in geographic areas where they are known to cluster was adapted by identifying five
major locations where students typically gathered outside of class.6 Whether these sites were
inside or outside, they were all popular locations close to or part of an on-campus coffee house or
food venue.
Following Stueve and colleagues‟ (2001) model, the five designated venues were randomly
selected to be the surveying location of the day by using a random number generator. Surveying
took place throughout the school week, from Sunday through Thursday (the normal school
week). Three two-hour time slots were used as designated surveying times and were randomly
assigned to each day of interviewing. To randomize at the level of individuals, the researcher
approached every third student, asking the student (in Hebrew) to participate in the study.
5
Screening for the target population, Israeli Jewish students who had served in the military, was
conducted before and after approaching a potential respondent.7 Students who agreed to
participate were then asked to sign a consent form and to fill out a paper and pencil questionnaire
in Hebrew. Under the circumstances, it was not possible to measure response rate with precision
but that is estimated to be in the 20-30% range. After excluding ineligible cases, the procedure
yielded 276 completed interviews.8
Recognizing that Haifa University differs from other Israeli universities in its catchment area
(the north of Israel), high proportion of Arab students (estimated at 20%), and other respects, we
do not intend to generalize the marginal frequencies (the percentage of responses to each
question option) to all Israeli university students. As we are interested in the relationships
between variables rather than the distribution of the dependent variables per se, this sample bias
is not consequential (Schuman 1986). Within the University itself, the sample underrepresented
the proportions of females, undergraduate students, and those studying a liberal major (a concept
defined below), with disparities ranging from 12-19 percent among the three variables for which
we obtained aggregate information from the Registrar. Rather than employ weighting techniques,
we control for sample bias by including gender, major and student status as statistical controls.
This strategy is consistent with recent scholarship on survey research that demonstrates the
capacity of apparently biased samples to yield representative attitudes when the bias is unrelated
to the questions asked in the survey and the potential sources of bias are included in the
statistical model (Groves 2006, Groves & Peytcheva 2008). As the survey is not principally
about gender, major, or student status, these differences are unlikely to bias the results.
Variables
To make manifest what has been implied, we hypothesize that increasing levels of higher
education among Israeli university students who served in the IDF will predict higher levels of
tolerance toward Arab Israelis and more “dovish” attitudes to the Arab-Israeli conflict. These
two dependent variables were selected for strong theoretical reasons. Tolerance assesses the
degree to which respondents supported civil equality for the Israeli Arab minority.9 We selected
tolerance because numerous studies around the world have reported strong correlations between
this concept and the experience of higher education (Jennings & Stoker 2008, Stubager 2009,
Marquart-Pyatt & Paxton 2007). If any variable is likely to respond to higher education,
tolerance is thus the prime candidate. The measure was specified for Israeli Arabs because their
situation most strongly tests Jewish Israelis‟ commitment to civil liberties for unpopular
minorities (Soen 2003). Although the Israeli Proclamation of Independence promises basic rights
and freedoms to all residents, Israeli citizens of Palestinian descent have posed a challenge to this
dictum because they are widely regarded by Jewish Israelis as a potential security threat. Indeed,
a 1998 survey found such massive social distance between Jewish and Arab Israelis that the
authors used “hatred” to characterize the relationship (Pedahzur and Yishai 1999). In the 1980s,
Seligman and Caspi (1983) reported that a “strong minority” of Israeli Jews, around one-third of
a national sample, would deny basic political rights to their fellow citizens who were Arabs.
Almost thirty years later, a survey of Jewish high school students found that fully half rejected
the idea that Israeli Arabs should enjoy the same rights as Jews and an even larger proportion
would deny Arabs the right to run for the Knesset (Kashti 2010).
6
The tolerance scale drew questions from Seligson and Caspi (1983) about the right to vote, right
to public assembly, freedom of speech, and the right to hold public office. The fifth question
inquired about the right of Israel‟s Arab citizens to attend university. Scores on the five-item
tolerance scale ranged from 0-20 with higher values on the scale indicated more tolerant attitudes
and greater support for the civil equality of Israel‟s Arab citizens. The Cronbach‟s alpha was .91,
indicating the same high level of reliability observed in Seligson and Caspi‟s national sample,
and the mean tolerance score was 12.05.
The second dependent variable was a composite scale of attitudes to the Arab-Israeli conflict. By
all accounts, the ongoing conflict has been for decades the “central, critical, and engulfing issue”
in Israeli political life (Shamir and Arian 1999, 266). The conflict has such potency, Shamir and
Arian suggest, because it serves as a touchstone for ongoing dilemmas of Israeli identity related
to religion, culture, citizenship and other existential concerns. Precisely because it is so
foundational, we suspect that attitudes to the components of the scale are likely to be relatively
fixed well before Israelis begin university—particularly following military service—and much
less responsive than tolerance to the forces that promote attitude change among young adults
attending university. If tolerance is an “easy” case prone to support the research hypothesis,
attitude toward the Arab-Israeli conflict is a “hard” case that resembles a critical test.
This measure incorporates seven questions from several of Arian‟s studies (Arian 1989, 1992,
2002, 2003; Arian, Shamir and Ventura 1992) and adds another. These questions tapped attitudes
about core issues involving Israeli policy toward Palestinians and the surrounding Arab states—
questions about support for the two-state solution, land for peace, responsibility for failures of
peace initiatives, rights to residents of the Occupied Territories, whether Israel should encourage
its Arab citizens to leave, military strength, and relative preference of the rule of law versus
security interests. The items scaled with an alpha of .78, indicating a strong relationship among
them. Scores on the 8-item dove scale ranged from 0-24. The items were coded so that an
increase in value denotes a more dovish political opinion. The mean score of the sample was
13.4.
The core independent variable is years of higher education, coded from 1-4 with 1 for first-year
undergraduates and 4 representing graduate students.10 (There were, respectively, 116, 35, 76,
and 42 cases in the first year, second year, third year and graduate categories.) Two other key
variables corresponded to the major theories about how the university experience influences
student attitudes. Major indicated whether a student was studying in a traditionally “liberal” or
“conservative” field, a factor shown to affect political attitudes in other studies (Stubager 2009).
Since the overall purpose of this investigation focused on the impact of university experience, it
was logical to determine if a student‟s specific major, as opposed to the general college
experience, was the key determinant to a change in political attitudes.11 Majors in the humanities,
liberal arts, and social sciences were coded as liberal (and assigned the higher value) while life
sciences, economics, mathematics and business were classified as “conservative.” Another
explanation for university impact emphasizes a student‟s peer group. We attempted to get at this
difficult to measure concept by assessing the degree to which a student‟s friendship network was
dominated by other students at Haifa University. While providing less information than a fullblown network analysis, this variable nonetheless identifies students who are most socially
integrated on campus and who, therefore, should be open to attitude change.
7
Beyond these central variables, the models also included a number of controls that previous
research suggested as influences on political tolerance and attitudes to the Arab-Israeli conflict-gender, religiosity, ethnicity and kibbutz background.12 As Seligson and Caspi noted in their
pioneering study of tolerance among Israeli Jews, “the younger, less religious, better educated,
and more affluent Israeli Jews express higher levels of general tolerance” (1983, 62). Precisely
these same groups typically adduce more dovish attitudes to the Israeli-Arab conflict.
Additionally, men, Sephardic and/or religious Jews are more likely to be relatively more
politically intolerant and hawkish than their counterparts (Moore and Aweiss 2002). Sephardic
Jews who trace their heritage toward Arab lands are known for negative attitudes toward Arabs
and hawkishness (Seliktar 1984). By contrast, kibbutzim have historically been known for their
strong leftist values and their young are likely to develop high levels of both tolerance and a
more dovish orientation to the conflict (Abramitzky 2008). Accordingly, the models included a
dummy variable to indicate whether the student had ever lived on a kibbutz.
We included one variable among the predictors in the tolerance equation but not the equation
predicting attitudes to the Arab-Israeli conflict—ideological self-definition (left or right) based
on security issues. (This variable was also coded in a liberal direction.) When measuring
respondents‟ willingness to grant civil liberties to groups, Sullivan et al (1982) have argued the
importance of controlling for the perceived threat posed by the group to society. Precisely
because Israeli Arabs may be seen as a fifth column, Israeli Jews are quite likely to condition
support for their civil liberties on sensitivity to security issues.13 By including a left-right scale
based explicitly on security issues, we will control for the underlying threat perception. Because
ideological differences in Israel are so heavily conflated with attitudes toward the Arab-Israeli
conflict, including self-identification in the second equation would amount to predicting the
dependent variable by itself. Hence, the security variable is not entered to predict attitudes
toward the Arab-Israeli conflict.
Analysis
In the following analysis, we use the basic OLS statistical model to predict both dependent
variables (tolerance and the Arab-Israeli conflict scale). Although the dependent variables are not
measured at the ratio level, they resemble such variables with a wide range and normal or nearnormal distributions. Given the robustness of OLS to violations of assumptions and the intuitive
interpretation of OLS regression coefficients, it is the preferred statistical technique for this
multivariate analysis. For ease of interpretation, we remind readers that the dependent variables
are coded in a right to left direction so higher scores indicate a greater propensity to support civil
liberties for Israeli Arabs and a more dovish view on the conflict. The unstandardized
coefficients can be read straightforwardly as the impact of a unit change in the independent
variable on the dependent variables.
Table 1 (below) reports the findings for tolerance. Focusing on the critical independent variables,
we discovered that greater exposure to higher education (measured by year in school) does
indeed promote support for extending civil liberties to Palestinian Israelis. All other things being
equal, each additional year of matriculation increases tolerance by about one point on the scale.
The two factors thought to promote this tendency, a liberal major and an extensive friendship
network, do indeed carry positive coefficients but neither reaches conventional levels of
8
statistical significance. Of the control variables, only the security orientation exerted a strong
effect on tolerance. The more the student inclined to a leftist self-definition on security issues,
the stronger the commitment to Arab civil liberties. Each step from strongly right to strongly left
increased the scale value by roughly one and two-thirds points. Although the coefficients for
secularity, kibbutz residence, and ethnicity (Sephardi identification) pointed in the expected
direction, they fell far short of statistical significance. In all likelihood, the variable representing
security orientation soaked up the variance that would otherwise have been attributed to these
controls.
A similar analysis was conducted for the dovishness scale, yielding the results in Table 2. As it
did for tolerance, years in school had a significant, positive influence on students‟ attitudes
toward the Arab-Israeli conflict. The magnitude of the effect, an increase of one scale point for
each year in university, was similar across the two dependent variables. Neither of the two
proposed mechanisms, friendship network and major, had any appreciable influence on
dovishness. Among the controls, we found two with significant effects. By approximately three
points and one and a half points, respectively, students who had lived on a kibbutz and those who
self-identified as secular rather than religious were on the whole more dovish than others. That
finding was consistent with our expectations. Although the coefficient did not attain statistical
significance, students who considered themselves Sephardi were, as expected, more hawkish
than their peers.
Discussion
The statistical analysis in the previous section indicated that exposure to higher education among
Israeli Jewish students who had prior military experience was associated with greater tolerance
to Arab Israelis and a more dovish orientation to the Arab-Israeli conflict. This finding was
consistent with our expectation that tolerance would be quite responsive to higher education, a
pattern widely detected in other countries. We were a bit surprised that the university influence
was potent enough to change attitudes toward the Arab-Israeli conflict, an orientation that largely
defines Israeli political identity.
The finding of a statistical association between tolerance and years of higher education emerged
despite a series of control variables that are associated with both variables. Hence the
relationship is not spurious. However, there are potential rival hypotheses that need to be
considered before we accept this finding.
Perhaps, as our first alternative explanation, the impact of additional higher education on
tolerance and dovishness can be explained by differential drop-out patterns. This assumes that
students who are intolerant or uncomfortable among Israeli Arabs do not enjoy attending a
university with a large population of Palestinians or with a political ethos that promotes mutual
understanding. If these students are more likely to drop out, then the increase in tolerance and
dovishness across the four categories from first year to graduate students could simply be an
artifact of compositional differences. Although we lack data to test this argument, the Registrar
of the University indicated to us that there is a very low dropout rate from year to year and that
most students who leave do so because of personal problems or financial constraints. That makes
it very unlikely that the tolerance effect is a consequence of differential dropout.
9
As a second possibility, the results may be due largely to the graduate student component of the
sample. Post-graduate students are likely to be self-selected on the basis of attributes that make
them quite different from undergraduates and perhaps they are largely responsible for the
apparent growth in tolerance as years at university increases. We can test this hypothesis two
ways, by including age as a predictor (since graduate students are older on average than
undergraduates) and by rerunning the analysis with only undergraduates. The results from
supplementary analysis (results not shown) enable us to reject this alternative interpretation
unreservedly for tolerance. By controlling for age and (separately) removing graduate students
from the analysis, the coefficient for tolerance on years in school actually increased in
magnitude (from 0.86 in Table 1 to .95 and 1.06 respectively) and, despite the loss of cases,
retained the same statistical significance. None of the other variables changed sign or
significance by virtue of the changes in the model. For attitudes to the Arab-Israeli conflict, there
was some slight movement consistent with the alternative interpretation but it was not terribly
consequential. When the dataset was reduced by excluding graduate students (results not shown),
the impact of years in school increased without any loss of statistical significance. However, with
the addition of age as a predictor in the full dataset, there was some attenuation as the years in
school variable diminished slightly to .65 (from .83 in Table 2) and the significance level was
.059. Because age and years at school were each significant and positive influences on
dovishness when included in equations without the other, the drop in significance for year in
school seems likely due to multicollinearity. This suspicion is reinforced by the correlation of .50
(Spearman‟s rho) between year in school and age and the corresponding drop in tolerance (using
that term in a statistical sense) for both variables when the other was included in the equation.
Under the totality of circumstances, it does not appear that older students and/or graduate
students are responsible for the growth in dovishness associated with additional exposure to
higher education.
Finally, as Jennings and Stoker note, the apparent growth in tolerance over the university years
could be a function of social desirability effects. As they write, “the presumed effects of higher
education in the realm of political tolerance, in particular, have been challenged on
methodological grounds . . . that better educated people have simply learned to express more
socially desirable opinions in response to survey questions” (2008, 4). If tolerance of Palestinians
is “politically correct” in an academic setting, the increase in scale values associated with
additional years at university could simply index the greater awareness of such views as students
become more familiar with the culture and norms prevalent in the university. We doubt this is
actually the case. First, it is not so clear that political tolerance is de rigueur in the university
setting where the research was conducted. The frequencies for the items making up the tolerance
scale indicate that a large proportion of Israeli Jewish students do not exhibit particularly tolerant
perspectives. In fact, only a minority of the student sample endorsed Israeli Arab rights to vote,
demonstrate or hold public office, hardly evidence of a strong civil libertarian norm that could
induce yea-saying.14 Second, social desirability effects are much more likely to emerge in
surveys that match respondents with an interviewer—face to face or via telephone. When survey
questions are administered by secret ballot, notwithstanding the presence of an interviewer,
respondents are much more willing to report socially undesirable or non-normative views
telephone (Aquilino 1994). In this study, students completed an anonymous self-administered
questionnaire out of direct gaze of the person who recruited them to participate—a situation
10
much more like a secret ballot than a formal interview. Finally—and this is entirely subjective—
anybody who has spent time in Israel will attest that Israelis are highly unlikely to disguise their
views in conversation solely to avoid discomfiture.15
Thus, while we cannot rule them out definitively, we think it unlikely that the observed protolerance effect associated with attendance at universities is a consequence of unmeasured
factors. Rather, we believe it attests to the power of higher education to promote more empathy
for minorities. This effect largely overrides the other social factors that have traditionally been
associated with tolerance in cross-sectional studies and even persists with a control for security
concerns that might well diminish tolerance. Once Israelis attain university status, it appears, the
differences between ethnic groups, religious identification, and other traits largely fall away. The
effect of higher education is not massive but, considering the large number of factors that
influence tolerance, it is noteworthy. We find it especially important because, as noted, Israelis
typically enter university only after a period of military service that one might well imagine as
fixing attitudes for the entire lifespan.
At the same point, the findings remind us of the limits of universities. The impact of university
experience on attitudes toward the Arab-Israeli conflict was not quite so substantial in that
students from a secular and/or kibbutz background were distinctive in their dovishness even after
taking account of years of schooling. No such pattern was apparent on tolerance. However,
students apparently enter universities with orientations to the Arab-Israeli conflict that predate
matriculation and that do not lose potency even as the university environment poses challenges
to pre-existing views. That finding is consistent with the discovery that the residues of
participation in youth groups are apparent for students in higher education.
Indirectly, the results also undermine the model of the military as a powerful socializing agent.
The results are indirect because we could not compare the attitude change registered among
veterans of the IDF with people who did not serve prior to university. With a large number of
university students who did not serve in the IDF but who resembled those who did—a relatively
rare population—we might have found much larger changes in attitude over the course of a
university career.16 Even with this proviso, the results tend to reinforce Kreb‟s (2004) strong
skepticism that military service leaves powerful or lasting political residues among those who
have experienced it. For all the centrality of the IDF in Israeli experience, it does not seem to fix
political attitudes rigidly or inure veterans to other influences that shape political learning.
Thus the Israeli case, for all its uniqueness, offers valuable insights about the conditions that both
enhance and diminish the political learning of young adults. The basic finding of prior research
conducted elsewhere, that attendance at university is a significant socializing agent in political
development, was sustained even in this distinctive and unpromising environment. That lends
credence to Dror‟s notion, cited at the outset of the paper, that studies conducted in Israel may
reinforce generalizations developed in other contexts as well as undermine them.
11
Table 1
Determinants of Tolerance
Model
Unstandardized Coefficients
B
(Constant)
Std. Error
t
Sig.
4.217
2.809
1.501
.135
Gender
-0.148
.700
-.212
.832
Secular Jew
-0.858
.785
-1.093
.276
Sephardi
-0.571
.734
-.777
.438
Kibbutz background
0.674
1.101
.612
.541
Liberal major
0.874
.711
1.229
.220
Year in School
0.864
.294
2.935
.004
Left on security
1.689
.270
6.248
.000
Haifa friends
0.431
.316
1.364
.174
(N=219)
12
Table 2
Determinants of Dovishness
Unstandardized Coefficients
Model
B
Std. Error
t
Sig.
(Constant)
7.771
2.869
2.709
.007
Gender
-0.281
0.735
-.381
.703
Secular Jew
1.649
0.823
2.004
.047
Sephardi
-1.224
0.756
-1.620
.107
Kibbutz background
3.172
1.081
2.934
.004
Liberal major
-0.034
0.741
-.045
.964
Year in School
0.831
0.302
2.754
.006
Haifa friends
-0.030
0.334
-.089
.930
(N=195)
13
References
Abramitzky, Ran. 2008. “The Limits of Equality: Insights from the Israeli Kibbutz.” Quarterly
Journal of Economics 123: 1111-1159.
Alwin, Duane F., Ronald L Cohen, and Theodore M. Newcomb. 1992. Political Attitudes over
the Life Span: The Bennington Women after Fifty Years. Madison, WI: University of
Wisconsin Press.
Aquilino, William S. 1994. “Interview Mode Effects in Surveys of Drug and Alcohol Use: A
Field Experiment.” Public Opinion Quarterly 58: 210-240.
Arian, Asher. 2003. Israeli Public Opinion on National Security 2003. Tel Aviv: Kedem.
-----. 2002. Israeli Public Opinion on National Security 2002. Tel Aviv: Kedem.
-----. 1989. “A People Apart.” Journal of Conflict Resolution 33: 605-631.
-----, Michal Shamir, and Raphael Ventura. 1992. “Public Opinion and Political Change: Israel
and the Intifada.” Comparative Politics 24: 317-334.
Bachman, Jerald G., Peter Freedman-Doan, David R. Segal, and Patrick M. O'Malley. 2000.
“Distinctive Military Attitudes among U.S. Enlistees, 1976-1997: Self-Selection versus
Socialization.” Armed Forces & Society 26: 561-585.
Bachman, Jerald G., Lee Sigelman, and Greg Diamond. 1987. “Self-Selection, Socialization, and
Distinctive Military Values: Attitudes of High School Seniors.” Armed Forces & Society
13: 169-187.
Dalton, Russell J. “Reassessing Parental Socialization: Indicator Unreliability versus
Generational Transfer.” American Political Science Review 74: 421-431.
Dar, Yechezkel and Shaul Kimhi. 2001. “Military Service and Self-Perceived Maturation among
Israeli Youth.” Journal of Youth & Adolescence 30: 427-448.
Dey, Eric L. 1997. “Undergraduate Political Attitudes: Peer Influence in Changing Social
Contexts.” Journal of Higher Education 68: 398-413.
Dror, Yehezkel. 1996. “On the Uniqueness of Israel: Multiple Readings.” In Israel in
Comparative perspective: Challenging the Conventional Wisdom, ed. Michael Barnett,
245-263. Albany, NY: SUNY Press.
Easton, David and Robert D. Hess. 1962. “The Child's Political World.” Midwest Journal of
Political Science 6: 229-246.
14
Emler, Nicholas and Elizabeth Frazer. 1999. “Politics: The Education Effect.” Oxford Review of
Education 25: 251-273.
Goldberg, Giora and Meir Kasirer. 1985. “Voting Patterns among Israeli Soldiers.” Jewish Social
Studies 47: 77-88.
Groves, Robert M. 2006. “Nonresponse Rates and Nonresponse Bias in Household Surveys.”
Public Opinion Quarterly 70: 646-675.
Groves, Robert M. and Emilia Peytcheva. 2008. “The Impact of Nonresponse Rates on
Nonresponse Bias: A Meta-Analysis.” Public Opinion Quarterly 72: 167-189.
Guimond, Serge. 1999. “Attitude Change during College: Normative or Informational Social
Influence?” Social Psychology of Education 2: 237-261.
Hess, Robert D. and Judith V. Torney. 1967. The Development of Political Attitudes in Children.
Chicago: Aldine.
Ichilov, Orit. 1991. “Political Socialization and Schooling Effects among Israeli Adolescents.”
Comparative Education Review 35: 430-446.
Jennings, M. Kent. 1987. “Residues of a Movement: The Aging of the American Protest
Generation.” American Political Science Review 81: 367-382.
Jennings, M. Kent and Gregory B. Markus. 1977. “The Effect of Military Service on Political
Attitudes: A Panel Study.” American Political Science Review 71: 131-147.
Jennings, M. Kent and Richard G. Niemi. 1968. “The Transmission of Political Values from
Parent to Child.” American Political Science Review 62: 169-184.
Jennings, M. Kent and Laura Stoker. 2008. “Another and Longer Look at the Impact of Higher
Education on Political Involvement and Attitudes.” Paper presented to the annual meeting
of the Midwestern Political Science Association, Chicago, IL.
Kashti, Or. 2010 (March 11). “Poll: Half of Israeli High Schoolers Oppose Equal Rights For
Arabs.” Haaretz. Accessed on 18 July 2010 at http://www.haaretz.com/printedition/news/poll-half-of-israeli-high-schoolers-oppose-equal-rights-for-arabs-1.264564.
Kedem, Peri and Mordechai Bar-Lev. 1989. “Does Political Socialization in Adolescence Have a
Lasting Influence? The Enduring Effect of Israeli Youth Movements on the Political
Ideology and Behavior of Their Graduates.” Political Psychology 10: 391-416.
Krebs, Ronald R. 2004. “A School for the Nation? How Military Service Does Not Build
Nations, and How It Might.” International Security 28: 85-124.
Lottes, Ilsa L. and Peter J. Kuriloff. 1994. “The Impact of College Experience on Political and
Social Attitudes.” Sex Roles 31: 31-54.
15
Marquart-Pyatt, Sandra and Pamela Paxton. 2007. “In Principle and in Practice: Learning
Political Tolerance in Eastern and Western Europe.” Political Behavior 29: 89-113.
Milem, Jeffrey F. 1998. “Attitude Change in College Students: Examining the Effect of College
Peer Groups and Faculty Normative Groups.” Journal of Higher Education 69: 117-140.
Moore, Dahlia and Salem Aweiss. 2002. “Hatred of „Others‟ Among Jewish, Arab, and
Palestinian Students in Israel.” Analyses of Social Issues & Public Policy 2: 151-172.
Muhib, Farzana B., Lillian S. Lin, Ann Stueve, Robin L. Miller, Wesley L. Ford, Wayne D.
Johnson, and Philip J. Smith. 2001. “A Venue-Based Method for Sampling Hard-toReach Populations.” Public Health Reports 116: 216-222.
Newcomb, Theodore M. 1958. “Attitude Development as a Function of Reference Groups: The
Bennington Study.” In Readings in Social Psychology, eds. Eleanor E. Maccoby,
Theodore M. Newcomb, and Eugene L Harley, 380-391. New York: Holt, Rinehart and
Winston.
Pascarella, Ernest T. and Terenzini, Patrick T., eds. 1991. How College Affects Students:
Findings and Insights from Twenty Years of Research. San Francisco, CA: Jossey-Bass.
Pedahzur, Ami and Yael Yishai. 1999. “Hatred by Hated People: Xenophobia in Israel.” Studies
in Conflict and Terrorism 22: 101-117.
Schuman, Howard. 1986. “Ordinary Questions, Survey Questions, and Policy Questions.” Public
Opinion Quarterly 50: 432-441.
Searing, Donald, Gerald Wright, and George Rabinowitz. 1976. “The Primacy Principle:
Attitude Change and Political Socialization.” British Journal of Political Science 6: 83113.
Sears, David O. and Carolyn L. Funk. 1999. “Evidence of the Long-Term Persistence of Adults'
Political Predispositions.” Journal of Politics 61: 1-28.
Sears, David O. and Levy, Sheri. 2003. “Childhood and Adult Political Development.” In
Oxford Handbook of Political Psychology, eds. David O Sears, Leonie Huddy, and
Robert Jervis, 60-109. New York: Oxford University Press.
Seligson, Mitchell A. and Dan Caspi. 1983. “Arabs in Israel: Political Tolerance and Ethnic
Conflict.” Journal of Applied Behavioral Science 19: 55-66.
Seliktar, Ofira. 1984. “Ethnic Stratification and Foreign Policy in Israel: The Attitudes of
Oriental Jews towards the Arabs and the Arab-Israeli Conflict.” Middle East Journal 38:
34-50.
Shamir, Michal and Asher Arian. 1999. “Collective Identity and Electoral Competition in Israel.”
American Political Science Review 93: 265-278.
16
Shamir, Michal and John Sullivan. 1984. “The Political Context of Tolerance: The United States
and Israel.” American Political Science Review 77: 911-928.
Shapira, Rina, Eva Etzioni-Halevy, and Anat Barak. 1986. “Political Attitudes of Israeli
Students: A Comparative Perspective.” Higher Education 15: 231-246.
Sigel, Roberta S. 1989. “Adult Political Learning: A Lifelong Process.” In Political Learning in
Adulthood, ed. Roberta S. Sigel, 458-471. Chicago: University of Chicago Press.
Soen, Dan. 2002. “A Binational Society: The Jewish-Arab Cleavage and Tolerance Education in
the State of Israel.” Israel Affairs 9: 97-119.
Stubager, R. 2008. “Education Effects on Authoritarian-Libertarian Values: A Question of
Socialization.” British Journal of Sociology 59: 327-350.
Stueve, Ann, Lydia N. O'Donnell, Richard Duran, Alexi San Doval, and Juliana Blome. 2001.
“Time-Space Sampling in Minority Communities: Results With Young Latino Men Who
Have Sex With Men.” American Journal of Public Health 91: 922-926.
Sullivan, John L., James Piereson, and George E. Marcus. 1982. Political Tolerance and
American Democracy. Chicago, IL: University of Chicago Press.
17
Notes
1
For a good theoretical discussion of the mechanisms linking education and political attitudes/behavior generally,
see Emler and Frazer (1999).
2
Using a two-wave panel design to assess the political impact of military service in the US, Jennings and Markus
(1977) found on the whole rather modest differences in the attitudes of young men who had and had not served. The
differences in civic tolerance, a variable of special interest in this study, were quite weak. Among the predictors of
civic tolerance, the authors reported that college experience was much more powerful than military experience.
3
Early empirical studies of Druze college students (cited in Krebs 2004, 98) and members of the officer corps
(Goldberg and Kasirer 1985) revealed very little evidence of the Israeli military‟s socializing effect.
4
On the significance of self-selection in studying military service and attitude change, see Bachman et al (1987,
2000).
5
The technique known as propensity matching does offer a way to sort out these influences by creating subsamples
of respondents generally identical in social background and thus holding these influences constant (Jennings &
Stoker 2008),. Unfortunately, such matching often reduces sample size and greatly increases sampling error,
offsetting some of the advantages from the technique. To our knowledge, this technique has not been applied to
Israeli studies of support for civil liberties.
6
Only 1,200 of the estimated 18,000 students of the University of Haifa live in the dormitories; therefore, many
students remain on campus throughout the day and in between classes (http://uhaifa.org).
7
Because the study examines the attitudes of Israelis who served in the military and focuses on questions related to
Israeli Arabs (Palestinian Israelis) and the Arab-Israeli conflict, the sample was restricted to Jewish Israelis with
military experience. Generally, Israeli Arabs could be identified by language use and/or clothing. If an Israeli Arab
or otherwise ineligible respondent agreed to participate, s/he was told the purpose of the study and the target
population and thanked for willingness to participate. As an additional precaution, the survey asked whether the
student had served in the military and those who indicated they had not done so were dropped from analysis.
8
As an experiment, students in one interview period were offered a free slice of pizza to enhance participation.
Unlike their American counterparts, who can usually be bribed with free food, this incentive did not encourage
cooperation and about half of the participants politely declined a slice. We encourage follow-up research on this
critical difference in cross-national reactions.
9
The terminology to denote this group is contested. Jewish Israelis, roughly eighty percent of the population,
traditionally used the term “Arab Israelis” for Arabs holding Israeli citizenship. Under growing nationalism and
identification with Arabs in the Territories (another contested term!), many members of this group prefer to label
themselves as “Palestinian Israelis” or just Palestinians. We will use the terms interchangeably although we suspect
that most Jewish Israelis still think of the minority in question as Arabs.
10
Israeli higher education follows the British model with undergraduate degrees typically awarded after three years
of study and students taking courses solely in their major or cognate fields.
11
We acknowledge that it is unclear whether students pick majors that are congenial with their attitudes or adjust
their attitudes to match the climate of opinion prevalent among students in the major.
12
Respondents were coded as secular and Sephardi, if they selected those alternatives, respectively, from the
following questions:
What type of Judaism do you identify yourself with?
1
Secular
2
Conservative
3
Traditional
18
4
5
6
Religious
Strictly Orthodox
Other
Which of the following groups do you consider yourself to be a member of?
(Check as many as apply.)
1
2
3
4
5
6
Ashkenazi Jews
Sephardi Jews
Russian Jews
Ethiopian Jews
Other
None
13
Shamir and Sullivan (1983) found that fully two-thirds of Israelis selected what they called pro-Arab organizations
as the group they liked least in society and were, accordingly, least deserving of civil rights. Given the primacy of
security as a value predicting tolerance in Israel, that finding reinforces the validity of the measure in this context.
14
We summed the respondents who would approve of such restrictions and those who described themselves as
neutral. In fairness, the percentage that rejected these three restrictions outright was larger than the proportion that
supported them but it never reached a majority.
15
In Thomas Friedman‟s apt observation, a conversation between two Israelis sounds like an argument between any
other people.
16
Most Israelis who do not serve in the IDF before (or after) university are Orthodox women and Palestinian Israelis.
Those qualities, rather than military service itself, are likely to account for attitudinal differences between those with
and without IDF experience.