CLASH OF POLITICS OR CIVILIZATIONS? SECTARIANISM AMONG YOUTH IN LEBANON [*].

THE SECOND LEBANESE REPUBLIC, created by the 1989 Ta'if Agreement, is rife with paradoxes. Recognizing that sectarianism was one of the principal causes of the conflict that had racked the country since 1975, those Lebanese politicians who gathered in Ta'if implicitly and, in some cases, explicitly, pledged to reform the political system to overcome this defect. Yet what emerged from Ta'if--more in its implementation than in its actual content -- was not a "de-confessionalized" Republic, but one in which the balance of power between confessions had been tilted in favor of Sunni and Shiite Muslims, away from Maronite Christians. Thus Ta'if reinforced confessionalism by reforming it, correcting the historical imbalance in favor of Maronites that had contributed to the destabilization of the political system.

That the Second Republic did not overturn but in fact reinvigorated confessionalism gave rise to another paradox, which is that the former critics of that system of sectarian balance have become its protagonists, while its former advocates are now its opponents. Chief among the latter are Shiite Muslims, whose new found political strength, combined with their plurality of the population, compels many of them to view confessionalism as a system which denies Shiites the power they would exercise under one based on majority rule. Secularized, middle class Maronites, who formerly constituted a major subcomponent of the anti-confessional forces, appear now to be fearful of the rising power and numbers of Shiite and more willing to accept the confessionalized status quo.

Predominant among Ta'if's arrangements for the domestic political system was the conversion of the dominant presidency, the sole preserve of Maronites, into a power-sharing troika comprised of a weakened Maronite President, a somewhat strengthened Sunni Prime Minister, and a Shiite Speaker of Parliament whose role, formerly far less important than the other two, was significantly upgraded. In the event this tripartite division shifted power dramatically toward the Prime Minister, and hence toward the Sunni community, reducing the role of the President--and therefore of the Maronite community -- apparently more than was envisioned at Ta'if. The term "troika" is, moreover, misleading, for the three do not pull in tandem as is implied in that term.

Related to the reallocation of power between these three key roles was what was thought at the time of Ta'if to be a redistribution of power between the executive and legislative branches to the benefit of the latter, as well as the devolution of power from the national to the local level. Upgrading the role of the Speaker, increasing the size of the parliament, and reducing the power of the President, were all thought to have paved the way for an invigorated Chamber of Deputies. But post-Ta'if parliaments have for the most part been more acquiescent to the executive than their predecessors, and probably less representative as well. No power has yet been devolved to any unit of local government.

Ta'if was to have been a new political beginning, yet it perpetuated the political power of the "warlords" who had divided the running of the country between themselves for the better part of fourteen years of conflict. It failed to provide for the emergence of a new political class to formulate policy for rebuilding of the country on new foundations. Ta'if was also to have signaled a thaw in inter-confessional hostilities, but instead it may have simply frozen animosities while facilitating the restoration of a central government, leaving unresolved the vital issue of inter-confessional relations, especially at the level of the "street."

The Second Lebanese Republic, in sum, is fraught with political paradoxes that intertwine sectarian relations, constitutional arrangements, economic reconstruction and foreign intervention. While these paradoxes have been much described and analyzed, what has received scant attention is their impact on the attitudes and behavior of Lebanese of different sects. Are these paradoxes exacerbating inter-confessional hostilities or reducing them? Are they stimulating a heightened sense of Lebanese nationalism or undermining it? Are they bringing about a greater commitment to Lebanese political institutions or alienating the population, or some sections of it, from them?

The behavioral evidence with which these questions might be answered is mixed. Occasional outbursts of public protest punctuate what appear to be long periods of apathetic indifference by the public. Elections seem to have stimulated considerable political activity, yet they have been condemned as contests with pre-determined outcomes for positions of little importance. The lack of overtly expressed inter-confessional hostility coexists with considerable covert grumbling and expressions of fear and resentment, especially on the part of Maronite Christians. Acquiescence to Syrian will, coupled with public pronouncements of fealty to Damascus, coexist with privately expressed resentment. In sum, just as the post-Ta'if political system appears to be paradoxical in many regards, so too does political behavior and presumably, therefore, the attitudes underlying that behavior.

The purpose of this investigation is to attempt to assess the impact of the paradoxes of Ta'if on the attitudes of a key sector of the Lebanese population. Since the Second Lebanese Republic, like the First, is a "precarious republic"[1] built upon an elaborate division of roles and power between religious sects, the most vital issue is whether or not the political preferences, attitudes, and behavior of the members of the different sects are converging or diverging. Presumably if Maronites, Sunni, Shiites, Druze, Orthodox and others are becoming more alike in their personal experiences, general beliefs, and political attitudes and the means by which they act upon them, then the political load to be borne by the relatively fragile institutions of the precarious republic will be reduced. Conversely, if those beliefs, attitudes and behaviors are diverging, greater stress will be placed upon governing institutions.

A hypothesis suggested by the paradoxes described above is that they are exacerbating tension and hostility between sects, thereby causing the members of those sects to diverge in their attitudes and ultimately, presumably, their behavior. The counter hypothesis is that the "unfinished business" of Ta'if, including the presence of foreign forces (Israeli as well as Syrian); the continuation and in some senses the intensification of confessionalism; the political predominance of warlords after the war has ended; and the lack of resolution of questions involving the roles and responsibilities of the key political institutions - are all creating a shared frustration among those of all sects and possibly even agreement on what needs to be done to improve the Second Lebanese Republic.

The method by which these alternative hypotheses were investigated was an attitudinal survey of students enrolled at five Lebanese universities. The rationale for drawing the sample exclusively from these young adults enrolled in universities lay in the fact that, more than their parents, they are the products of Ta'if, having arrived at adulthood under its provisions. Their beliefs and attitudes can thus be expected to reflect the paradoxes of Ta'if more than do those of their parents. Moreover, youth tend to hold attitudes comparatively intensely, thus facilitating the task of identifying trends in their early phases or when they are fairly weakly held across the population as a whole. The apparent intensity with which some educated youth are attracted toward Islamist or primarily Christian movements is a case in point. It is also true, of course, that these youths will have an increasing impact on the political system as they enter full adulthood, since they may reasonably be expected to hold leadership p ositions in the elite. Hence their attitudes will be more germane to future outcomes than those of their parents. A final, pragmatic consideration was that, in the Middle East, it is very much easier to administer questionnaires to students than to any other types of respondents.[2]

THE SAMPLE

Surveyed in mid-1996, the sample of 917 university students was drawn from a range of faculties across university campuses in Lebanon. Of the institutions involved, three were in Beirut: the American University of Beirut (AUB); the Lebanese American University (LAU) Beirut campus; and Hagazian University. Located outside Beirut were the LAU Byblos campus, and Balamand University. The largest contingent of students (57.5%) came from AUB. The greatest concentration of students in the sample overall was in the Arts and Sciences Faculty (44.9%), followed by Engineering (15.7%) and Business (13.6%). In each university Arts and Sciences is the largest faculty (comprising, for example, 51% of AUB), so that the sample was reasonably representative of faculty distributions in the student body as a whole. The gender distribution in the sample was almost symmetrical (50.5% female, 49.5% male). Over 70% of respondents were in the 20-24 age group, spread fairly evenly over first to third years of enrolment. While 89.2% w ere Lebanese citizens, just over one-fifth of the sample declared some other citizenship or nationality (including Jordanian, Palestinian, North or South American and Australian). It is sometimes advantageous for students in Lebanon to enroll as "foreign" students rather than Lebanese citizens, since the admission requirements for the former are less rigorous. [3] In cases of dual nationality - which is allowable in Lebanon - there may thus have been some under-reporting of Lebanese citizenship, especially since non-Lebanese were instructed not to answer the questionnaire.

This group of students could be seen as enjoying elite status in educational terms. Over half the sample (51.4%) rated their own academic performance as above average, while 58% of their fathers and 45% of their mothers had some kind of university degree. Yet it would be erroneous to perceive the students as a homogeneous, "born to rule" social stratum. Only a third of the sample rated their family's living standard as above average, and their fathers' occupations, as described by the offspring, were very diverse and rather difficult to categorize; for while twenty-eight per cent of cases could be put into the "business/entrepreneur" category, and a further 23% into various categories of "professional," a host of other high- and low-status occupations were named as well. The occupation of the mothers was described as "home duties"--or its equivalent--in well over half the responses to this question (although one quarter of respondents answered only with respect to their fathers). Given the diversity of stude nts' responses to items relating to their socioeconomic status, the sample appeared to be representative of a relatively broad segment of the population at large. It should also be recalled that scholarships, loans and grants make tertiary education in Lebanon accessible to considerable numbers of youth of lower socio-economic status. [4]

When it came to the important issue of sect, the sample was largely made up of non-Christians. The sample, as noted earlier, was dominated by students at AUB. Although this university was founded by Christian missionaries, Sunni Muslims formed the majority of the students from as early as 1890, and it is this sect which has continued to play a predominate role in the institution. Constituting almost one-third of the overall sample (32.2%), the Sunni combined with the Shiite respondents (12%) to form a Muslim sub-group totaling 44.1% of the entire sample, with Druze students amounting to a further 6.5%. The 13.5% of Maronite Christians (the largest Christian sect in the country) were only slightly more numerous than the Shiites in this sample, and were outnumbered by the Orthodox who formed almost a quarter of the sample (24.1%). A scattering of other affiliations and sects (including a few Armenians) made up the remainder. Statistical information relating to students' sectarian affiliations is not released b y the universities, so that there is no way of knowing whether this sample was representative of affiliations in the university (or across universities) as a whole. What is known, however, is that the Shiites who are now generally regarded as the most numerous sect in Lebanon appeared underrepresented in this particular student body, and that the Maronites were likewise fewer than their broader distribution would suggest. In the absence of census data on sect affiliations within the Lebanese population, the sample cannot be proven to be either representative or not representative in this regard. However, the large sample and the presence within it of sizeable sub-groups representing all the major sects meant that the inter-sect differences could be analyzed in some detail, as will be seen below.

QUESTIONNAIRE DESIGN AND ADMINISTRATION

The questionnaire was drafted by the Australia-based research team, piloted and modified before being taken to Lebanon and administered in situ by one team member. The questionnaires were administered to students in class by their professors--a method guaranteed to ensure a very high return rate (if not necessarily great spontaneity). Utilizing the good offices of the team's faculty and staff colleagues at the five universities overcame the potential problems faced by "foreign" researchers in establishing trust vis-a-vis a conflict-weary and justifiably suspicious local population. Items were mainly close-ended questions with a 5-point scale of possible responses (typically, ranging from "Strongly agree" to "Strongly disagree"), although respondents were also given the option of writing in answers to some open-ended questions as well. Apart from a number of questions eliciting basic bio-data about the respondents and their families, the students were also asked to answer questions relating to their experienc es, attitudes and behavior. [5] The completed questionnaires were collected and taken to Macquarie University in Sydney, Australia, where the analysis of the data took place. The fact that the survey was brought to a successful conclusion despite the concurrent turmoil caused by Israel's "Grapes of Wrath" operation was a tribute to the diligence and cooperativeness of all those involved in facilitating the completion of the questionnaires on all the campuses surveyed.

METHOD OF ANALYSIS

Investigations of inter-sectarian relations via attitudinal surveys are problematical. Asking direct questions is virtually impossible, risks alienating respondents and is unlikely to result in candid responses. The method chosen in this study to investigate whether or not sects are diverging or converging in their personal experiences, general beliefs, political attitudes and behavior was an inferential one. It was assumed that if attitudes toward general issues and--more specifically--toward political issues, as well as reports of behavior vary greatly by sect, it suggests growing divergence between those sects. Conversely, convergence is suggested if inter-sect attitudinal differences are relatively small. The investigation, in sum, was directed at finding out not what Lebanese university students think about members of their own and other sects, but whether their attitudes and behavior vary as a result of their sect, and to suggest possible reasons why.

The sample, reflecting the distribution of sect membership in the population as a whole, included small numbers of the more marginal sects. [6] Some of these were discarded from the analysis (such as the two Jewish respondents), while other sub-groups were collapsed into single categories. This merging of sect categories was done only for groups with strong theological and social affinities and after the initial statistical analysis had revealed high correlation on key variables. On this basis the 34 Greek Catholics were merged with the Maronites, and students identifying themselves as Greek Orthodox were combined with all those who described themselves simply as "Orthodox." Druze, as members of a schismatic sect of Islam, were combined with Sunni and Shiite only when the analysis required an overall "Muslim" category. Most of the tabulated findings below relate to the four major sect groups: Sunni, Shiites, Maronites and Orthodox, with the occasional inclusion of Druze as a separate category.

The three independent variables whose interaction and impact the team sought to investigate were sect, gender and class or socio-economic status (SES). Preliminary analysis revealed that sect and gender explained more of the variance with respect to selected dependent variables than did SES. Accordingly, sect and sex were both treated as the variables of interest, with SES assigned the role of covariate. Thus, at each stage of the analysis SES was examined to see if it was exerting an effect independent of sect or gender. If it did, then SES was regarded as a "confounding variable" and its precise effect was investigated through its status as a covariate in that analysis; if not, it was excluded. In effect, then, SES was found to be a much less important predictor of behavioral and attitudinal differences than the other two independent variables. Where the impact of gender on the findings reported here was both significant and independent of sect, it is reported as such. [7] To measure complex variables, suc h as socio-economic status, extent of political participation, and levels of satisfaction with the political system, scales were created by combining responses to two or more questionnaire items where responses were highly significantly interrelated. The composition of the various scales is explained in the text below.

FINDINGS: PERSONAL BACKGROUND AND EXPERIENCES

Divergence/convergence can be measured according to the degree of variance in relevant personal experiences, in attitudes as reflected in responses, and in self-reported political behavior. Since war and reconstruction were expected to be the source of the most profound interventions into the lives of these respondents, two questionnaire items attempted to assess the differential impacts of these events. The first asked whether or not the respondent's immediate family had suffered loss of life, injury, displacement from home, and/or property loss or damage as a result of civil war. Responses, awarded one point each, were added, so the maximum possible total was four and minimum was zero. The second question asked the extent to which the respondent's family had benefited in economic terms from reconstruction, with the five closed-ended options ranging from "very much" to "not at all."

These two items produced starkly varied responses. With regard to the first, Shiites reported significantly more suffering from the war than either the Sunni or the Orthodox. [8] The finding with regard to the greater suffering of the Shiites is not accounted for by socio-economic differences, which were not significant between the major sects. It might be the case, therefore, that Shiite attitudes on a range of issues reflect their disproportionate wartime suffering. [9] At the same time, and despite the youthfulness of the respondents, few members of the sample remained altogether unaffected by the war. Only 30% reported no suffering at all; one tenth reported a war-related family death; and displacement and loss of/damage to property had been experienced by the families of 45% and 52% respectively.

Although fewer Shiites reported having benefited from reconstruction than members of any other sect, the difference was not significant. Indeed, responses to this item suggest convergence rather than divergence, with just under half of all valid responses (49.3%) being "not at all." Across all sects, less than ten percent of respondents reported that their families had benefited "very much" or "quite a lot" from reconstruction. If this sample of Lebanese did not suffer equally as a result of the civil war, they have certainly failed in more or less equal measure to benefit from reconstruction. The overall implication may well be that their experiences are converging rather than diverging, for the impact of reconstruction is presumably of greater salience for these young Lebanese than was wartime suffering, for that occurred at least seven years ago and probably considerably more.

Socio-economic status is another determinant of personal experiences. If family backgrounds--as measured by parents' education, occupation and income--differed markedly between sects, that would result in different personal experiences, and hence in diverging attitudes and behavior. In order to measure socio-economic status, a scale was constructed from responses to items that asked mother's and father's occupations and levels of education, and a question that asked respondents to indicate whether their family's standard of living was above average, average, or below average. The analysis revealed systematic and significant differences between males and females, with females having higher socio-economic status. [10] This may reflect a propensity on the part of lower SES families to educate sons rather than daughters. Within sects there were no significant male-female differences, and - more importantly - between the sects being examined in this study there were no significant socio-economic differences. The rank order of the sects is shown in Table 1. The maximum mean score possible was 23, and the sample's overall mean score was 15.84.

                Rank Order of Socio-economic Status by Sect
Sect     Mean
Sunni    16.25
Shiite   15.72
Maronite 15.61
Orthodox 15.56
Druze    14.92

Note: The only significant difference in the rankings (t=4.28,df=863,p[less than]0.005) was between Druze and 'Other Christian'; the latter had the highest ranking at 17.65. But since the latter category was a composite one, containing members of a number of sects, it has been omitted from most of the analysis.

These data indicate marked convergence of the major sects as measured by the socio-economic status of this sample of their youth. These sect groupings in turn comprised 92% of the total sample.

ATTITUDES TOWARD POLITICAL ISSUES AND INSTITUTIONS

Attitudes were elicited in response to a wide range of questions about political issues and institutions, some of which were assumed to be of special concern to university students. Issues that were dealt with included perceived degree of democracy and access to decision making; protection of civil liberties; and prospects for settling the Arab-Israeli conflict together with the possible impact of such a settlement. [11] Respondents were asked to evaluate the informal institution of confessionalism and were also asked which political tendency or movement best represented their views. On a more personal level, an item that sought to measure a sense of helplessness (ihbat-- a widely reported and much discussed feeling, especially among Maronites) was included.

Confidence In The Political System

From responses to six questionnaire items relating to a range of political issues, a 30-point scale labeled "Confidence in the Political System" was constructed. [12] Responses to the politically relevant items generally substantiated a hypothesis of divergence of attitudes, and this was especially the case with regard to degree of confidence in the political system. The Druze, Shiites and Sunni were all significantly more confident than the Maronites; [13] and Shiites as well as Sunni were also significantly more confident in the system than the Orthodox. [14] Overall mean scores on the "Confidence" scale are set out in Table 2.

       Levels Of Confidence In The Political System, By Sect (N=878)
Sect     Mean
Sunni    15.2
Shiite   15.0
Druze    14.2
Orthodox 13.2
Maronite 12.2

As Table 2 indicates, the Muslim sects (including Druze) ranked as being significantly more confident in the political system than the most numerous Christian sect in Lebanon (the Maronites), and the two major Muslim sects were also significantly more confident than the other major Christian group, the Orthodox. Expressed another way, Maronites were more likely to hold negative views than Sunni, Shiites and Druze, while the Orthodox were more likely to be negative about the system than either Sunni or Shiiite. The comparatively low level of confidence in the political system by Christians signals disaffection from it, while the much higher level of confidence manifested by Muslims of all sects (including Druze) indicates a correspondingly greater sense of political trust and efficacy.

It is also interesting to note from Table 2 that Shiite respondents, revealed in this investigation to have suffered more from the civil war than several other sects (see above), had a mean level of confidence in the political system almost as high as Sunni. This confidence may be based at least partially on the enhanced political role played by Shiites since Ta'if, a role that no doubt has done much to reduce the level of bitterness that might otherwise have persisted as a result of disproportionate Shiite suffering, not to mention their previous history of political marginalization. Shiites suffered greatly in the war, but they did not lose it. Maronites in this sample, on the other hand, may perceive themselves as having lost both the war and the peace, in that the wartime suffering they endured has not been compensated for by political gains, so that they are doubly bitter. Maronite estrangement from the body politic is also indicated by their pronounced sense of helplessness.

The word ihbat, generally translated as "helplessness," has come into common parlance in Lebanon to describe a widespread feeling of political impotence, especially among Christians. As a further measure of political confidence (or rather, the lack of it), respondents were asked to indicate on a 5point scale how much they shared that feeling of being unable to control their political destiny. On the basis of the data presented in Table 3, both Maronites and Orthodox were significantly more likely to feel helpless than Sunni. [15] It will be seen from Table 3 (where scores range from a possible low of 1 to a possible high of 5) that the Muslim sects (including Druze) had lower mean "helpless" scores than both Christian sects.

Thus, on both the 6-item scale measuring confidence in the political system, and on the single "helplessness" item, Maronites and Sunni are at opposite ends. Upon further collapsing the sects to give a broader Muslim/Christian dichotomy, the results shown in Table 4 unequivocally support the assertion that, at least among university students, Christians in Lebanon today feel significantly more helpless than Muslims.

The evidence in Tables 3 and 4 suggests strongly that Muslims and especially Sunni are considerably more satisfied with the political status quo than are Christians, especially Maronites, who appear to feel both less satisfied with and less able to change it.

             Level Of Shared Feeling Of Helplessness (Ihbat),
                              By Sect (N=815)
Sect     Mean
Maronite 3.42
Orthodox 3.19
Shiite   3.13
Druze    2.98
Sunni    2.80

Note: Tables 3 and 4 do not include either Armenians or "Other Christian" (because of the small numbers in those categories) although both groups were included when calculating the means for the sample as a whole. F=5.34, df=6,860, p[less than]0.0005

             Level Of Shared Feeling Of Helplessness (Ihbat),
                      By Major Sect Grouping (N=815)
Sect Group          Mean
Maronite/Orthodox   3.28
Shiite/Sunni/Druze  2.89
F=22.77, df=1,771,p[less than]0.005

Views On Confessionalism

Because confessionalism is so central to the Lebanese political system; because the civil war was commonly attributed to it; and because Ta'if--at least in the hopes of some--was to have commenced the process of dismantling confessionalism rather than reinforcing it, diverging attitudes toward it could point to a threat to future political stability. Accordingly, respondents were asked to indicate on a 5-point scale the extent to which they agreed with two propositions; namely, that confessionalism was necessary for political stability in Lebanon, and that confessionalism was holding back Lebanon's development. [16] The results were cross-tabulated against sect groupings using the Chi square test ([X.sup.2]) for significance. [17]

To identify overall differences between the "Muslim" view and the "Christian" view on this crucial issue, Sunni, Shiite and Druze were combined into one category while another category was formed from Maronites and Orthodox. Strong and significant differences between the sect groupings emerged: Christians were significantly more likely to agree that confessionalism had a stabilizing effect than Muslims (including Druze), who were more likely to take the opposite, negative stance. These results can be seen in Table 5.

        Respondents' Views On Whether Confessionalism Is Necessary
                    For Stability, By Major Sect Groups
         Muslim/Druze Maronite/Orthodox
         % (N)        % (N)
Agree    29 (128)     37 (132)
Neutral  20 (90)      24 (85)
Disagree 51 (230)     40 (144)
Note: Not all % columns add up to 100, due to
rounding up [X.sup.2]=10.75,df=2,p=0.005

In order to isolate differences between particular religious groups, the results were also analyzed by sect, as shown in Table 6.

        Respondents' Views On Whether Confessionalism Is Necessary
                          For Stability, By Sect
          Sunni   Shiite  Maronite Orthodox  Druze
          % (N)    % (N)   % (N)    % (N)    % (N)
Agree    25 (71)  34 (37) 33 (51)  39 (81)  35 (20)
Neutral  25 (72)  13 (14) 30 (46)  19 (39)   7 (4)
Disagree 50 (140) 53 (57) 37 (58)  42 (86)  58 (33)
Note: Not all % columns add up to 100, due to rounding
up [X.sup.2]=31.45, df=8, p=0.0001

Sunni were significantly less likely to agree with the proposition that confessionalism is necessary for stability (25%), with another quarter of them feeling non-committal about it and fully 50% disagreeing with the proposition. Fifty-three per cent of the Shiites were similarly negative, perhaps because they have traditionally been marginalized by this system, and/or because they are now the largest sect in Lebanon and presumably could expect to exercise greater political influence were confessionalism to be abolished. The most critical stance of all was adopted by Druze (58% against). As traditional Christian support for the institution would lead one to expect, Orthodox (39%) were significantly much more likely to agree on the stabilizing effect of confessionalism than any other sect (although 42% were in disagreement). The picture with the Maronites, however, was somewhat less clear. Maronites supporting the proposition (33%) were only as numerous as their representation in the sample would predict.

The evidence of this survey suggests that Christians are more likely to believe that confessionalism is necessary for political stability and presumably, therefore, many of them would seek vigorously to defend it. Yet, the evidence also indicates that members of all sects are divided in their beliefs about confessionalism, thereby mitigating the degree of divergence in Muslim-Christian attitudes. A majority of Muslims (51%) and 40% of Christians in the sample did not believe that confessionalism was necessary for political stability. But very substantial minorities of both sect groupings (29% and 37%, respectively) held the opposite view, suggesting the possibility that cross-sect coalitions could form around the issue of deconfessionalization. In the sample as a whole 46% disagreed that confessionalism is necessary for stability, while only 32% agreed with that suggestion.

Interestingly, the divergence manifested on the stability issue was not replicated when respondents gave their views on whether confessionalism is holding back Lebanon's development. On this issue there were no significant inter-sect differences. [18] Indeed, 61.5% of all respondents agreed (or strongly agreed) that Lebanon was being held back by confessionalism; 23% took a noncommittal stance, and only 15% disagreed. In other words, Lebanese students in this sample largely agreed that confessionalism impedes development, regardless of their sect. The respondents' discriminating approach to the two questions on confessionalism is evidence both of the political sophistication of this group of young Lebanese and of the centrality of this particular issue in Lebanon's political system. Support for its reform, therefore, although obviously substantial, is far from universal and opposition to it would be considerable, when and if such reforms are placed on the political agenda.

Views On The Arab-Israeli Conflict

The Arab-Israeli conflict and Lebanon's role in it is another central political issue on which a diversity of views could be expected. As with the previous two questions, respondents were asked to express their agreement or disagreement (on a 5-point scale) with statements that: "There will soon be a settlement of the Arab-Israeli conflict" and "If there is a settlement of the Arab-Israeli conflict, Lebanon will prosper." The only significant sect-related finding in the analysis of responses to the first proposition was that Maronites were more likely than Shiites to agree that there would be a settlement. [19] Almost half the Maronites (49%) felt this way, compared to only 35% of Shiites. The rank order of the sects' levels of agreement with the proposition is shown in Table 7. The maximum mean score possible was 5.

        Rank Order Of Levels Of Agreement As To Whether There Will
              Be An Arab-Israeli Settlement, By Sect (N=878)
Sect     Mean
Maronite 3.29
Sunni    2.98
Orthodox 2.97
Druze    2.97
Shiite   2.72

It can be seen that, in terms of percentage distributions, the optimism of the Maronites was in strongest contrast to the relatively pessimistic outlook of the Shiites, with the Sunni, Orthodox and Druze somewhere in between.

A greater number of significant differences between members of the various sects were found in their responses to the second proposition on the Israel issue, namely, that Lebanon will prosper if there is a settlement. Maronites were found to be significantly more hopeful in their expectation of national prosperity than Shiites, and Maronites were also more likely to be positive regarding this issue than either Sunni or Druze. [20] The rank order of the sects' levels of agreement with the proposition (with a possible maximum score of 5) is shown in Table 8.

Maronites were the only respondents among whom a majority agreed or strongly agreed with the proposition regarding prosperity, with some 55% of them so doing, compared to minorities in the Muslim groups (including Druze) ranging from 34% to 37%.

          Rank Order Of Levels Of Agreement As To Whether Lebanon
   Will Prosper If The Arab-Israeli Conflict Is Settled, By Sect (N=878)
Sect     Mean
Maronite 3.47
Orthodox 3.10
Shiite   2.91
Sunni    2.88
Druze    2.83

Because this conflict is of such importance to Lebanon, significant differences in attitudes about it, even on these relatively indirect measures, could indicate divergences by sect that would become very important at a political level when and if the prospects for a settlement improve and Lebanon is forced to formulate a clear policy. These data imply that the largest Christian sect in Lebanon, the Maronites, is significantly more likely to be favorably disposed to a settlement than are the Muslims. That divergence could conceivably exacerbate inter-sectarian tensions associated with a more active peace process involving Lebanon.

Political Tendencies

When it came to respondents' identification with a political tendency or movement, a majority of those answering the question (56%) identified Lebanese nationalism as the tendency which best represented their views. [21] "Arab nationalist" was second at 15 per cent, followed by "Islamist" at 7 per cent and "Syrian nationalist" at 4 per cent. A number of other choices were expressed in write-in responses, but they comprised a very small proportion of the total; for example, Aounist and Lebanese Forces together rated only 1.5 per cent, and Hizbollah (included in Table 9 under "Other") had only one declared supporter. [22] General disillusionment with the political system was manifested by the 11% of those responding who stated in writing that no existing political tendency represented their views, while 7% of the sample as a whole chose not to answer the question at all.

In every sect Lebanese nationalism was most favored, but the percentage distributions set out in Table 9 show how the members of different sects varied in their support for other tendencies. [23]

         Respondents' First Preferred Political Tendency, By Sect
              Sunni   Shiite  Maronite Orthodox Druze
              (N=269) (N=l08) (N=150)  (N=201)  (N=54)
              %       %       %        %          %
Lebanese Nat. 38      55      71       67         57
Arab Nat.     28      14       4        7         15
Syrian Nat.    2       1       3        7          7
Islamist      13      19       1        -          -
Aoun/Leb.F.    -       -       3        3          -
Other          5       6       3        8          7
No tendency   13       7      14        9         13

Notes: Not all % columns add up to 100, due to rounding up. 'No tendency' denotes those who commented that they felt no existing tendency represented them, while those not answering have been omitted from the table altogether. [X.sup.2]=166.45,df=24,p[less than]0.0001

A plurality of Sunni (38%) supported Lebanese nationalism, but a larger proportion of them supported Arab nationalism (28%) than did members of other sects such as Shiite (14%), Orthodox (7%) and especially Maronites (4%). Thirteen per cent of Sunni joined 19% of Shiites in expressing support for Islamism, in which few Christians and no Druze concurred. The preferences of Druze and Shiites were distributed very similarly (except with regard to Islamism) and were also in line with the preferences of the sample as a whole, with 55% and 57% respectively supporting the Lebanese nationalist tendency.

The overall patterns of Maronite and Orthodox preferences were very similar to each other. Both sects strongly favored Lebanese nationalism above all other tendencies (71% and 67% respectively). Their broad rejection of Arab nationalism has already been noted. Interestingly, both groups gave the Aounists/Lebanese Forces a very low rating (3% in both groups). Overall, the two Christian sects mirrored each other's views, while there was much more disparity between the three non-Christian sects. Table 10 presents the preferred political tendencies for Muslims and Christians as a whole.

               Respondents' Preferred Political Tendencies,
                          By Major Sect Grouping
                     SECT GROUP
Preferred tendency   Muslim/Druze       Maronite/Orthodox
                     %            (N)         %            (N)
Lebanese nationalist 55.5         (193)       92.0        (242)
Arab nationalist     28.4         (99)         7.6         (20)
Islamist             16.1         (56)         0.4          (1)
[X.sup.2]=101.17,df 2,p[less than]0.005.

The clear and significant differences between the two major sect groupings immediately suggest that the political outlooks of Muslim and Christian students are quite different. The Christians are overwhelmingly more likely to be Lebanese nationalists, with no other political tendency accounting for more than 7 per cent of that sub-group. Muslims, on the other hand, are significantly more likely to be attracted to both Arab nationalist and Islamist viewpoints, with these tendencies together accounting for more than 38 per cent of Sunni and almost one third of Shiite respondents. It may well be the case that previous generations of Lebanese Muslim students, especially Sunni, identified more strongly with Arab nationalism, but the present level of support for that political tendency, especially if combined with that for Islamism, still sharply differentiates Muslims from Christians (although the Druze are somewhat an exception with regard to Islamism). It should also be noted that Christians are much more homog eneous in their support for political tendencies than are Muslims. The proportion of Sunni supporting Arab nationalism, for example, is more than twice the proportion of Shiites who do so; similarly, a much larger percentage of Shiites favor Lebanese nationalism than do Sunni.

In sum, responses to questions concerning respondents' attitudes toward political issues and institutions suggest considerable divergence among sects, an observation confirmed by significant differences between sects in their political outlooks. Sunni, although more likely to be skeptical about the benefits of confessionalism, were also significantly less critical of the political system than the Christians, having the highest mean level of confidence in the system. They not only suffered significantly less in the civil war than, for example, the Shiites, but were also significantly less likely than Christians to report suffering a sense of helplessness, having the lowest mean level of such ihbat overall. Sunni were generally not hopeful about the benefits of a possible peace with Israel, being significantly more pessimistic than the Maronites. Finally, Sunni were more likely to be committed to Arab nationalism than members of other sects and were second only to Shiites in their support for Islamism.

Maronites tended to be at the opposite end of the continuum from Sunni on all of these items and scales. They were significantly less satisfied with the political system than any Muslim sect, and suffered significantly more than Sunni from a sense of political helplessness (having the highest mean ihbat score). More of them expected peace with Israel than did the proportion in other sects, and they were significantly more likely to pin their hopes on the consequences of such a peace than members of any of the Muslim sects. Although they acquiesced in the stabilizing effect of confessionalism more than members of other sects, large sub-groups were also critical of it. Maronites were relatively more supportive of Lebanese nationalism, and less supportive of Arab nationalism, than were the adherents of other sects.

Shiites and Orthodox tended to fall between Sunni and Maronites on all of these dimensions, although Orthodox were closer to Maronites than Shiites were to Sunni. Compared to Sunni, for example, the Shiites in the sample were more committed to supporting Lebanese nationalism (and Islamism) than Sunni, while favoring Arab nationalism less. Shiites were significantly less optimistic than Maronites about prosperity in case of a settlement with Israel; were significantly more confident in the political system than the Christian sects; and were in the middle ranking on the scale of "helplessness". Orthodox sympathies, in contrast, paralleled those of the Maronites in that they were overwhelmingly likely to favor Lebanese nationalism and to reject Arab nationalism (and Islamism). Orthodox respondents also had similar views to those of Maronites on the Arab-Israeli issue, and had a similar sense both of lacking political confidence and of helplessness, but were significantly more likely to support confessionalism t han their fellow-Christians.

These data suggest that the attitudes of Sunni are consistent with their status as the politically dominant sect, while those of Maronites are consistent with attitudes of a sect that has been politically marginalized. Shiite and Orthodox attitudes are somewhat more difficult to characterize, although the former appear to support an interpretation of a sect that has recently "arrived" politically. The Orthodox seem to be in an ambivalent position between Maronites on the one hand and the Muslim sects on the other, although much closer to the former. Druze, consistent with their status as a schismatic sect of Islam, manifest somewhat idiosyncratic attitudes, but on most major political issues are to be found at the "Muslim" end of the attitudinal spectrum.

If the analysis was to end at this point the conclusion would be that the political attitudes of Lebanese students diverge according to sect and that those attitudes tend to reflect the current political standing of the sect to which the individual students belong. In other words, sectarian politics is mirrored by sectarian attitudes in the sense that those attitudes differ in predictable fashion by sect. According to this scenario, Ta'if clearly has not eroded sectarian differences even though it has enabled a political system to operate within a context of divergent attitudes. It remains an open question, however, as to whether the system would be sufficiently strong to confront a major crisis that triggered sharply varied sectarian reactions.

SELF-REPORTED POLITICAL BEHAVIOR

In addition to asking respondents about their attitudes on various issues, the questionnaire also sought information about their political behavior as another potential measure of convergence/divergence.

Political Participation

To measure respondents' involvement in the political system, a 14-point political participation scale was constructed by combining the variables for joining a party, demonstrating, contacting one's Deputy, actual and intended voting, contributing to campaigns and discussing politics. Using this scale it was found that students of lower socio-economic status (SES) were slightly more likely to participate in politics than those of higher SES. [24] Generally, however, the level of political participation was low, with the mean being 6.3 (out of a possible 14) and only 14 of the 917 respondents scoring twelve or above. But the scale also revealed differences in political participation, by sect, which were statistically significant. Comparing sects, Shiites were significantly more likely to participate than either Orthodox or Sunni, while Maronites were more likely to participate than Orthodox. [25] The results for all sects are shown in Table 11.

           Rank Order Of Mean Levels Of Political Participation,
                                  By Sect
Sect     Mean
Shiite   6.77
Maronite 6.62
Druze    6.44
Sunni    6.06
Orthodox 5.95

Although variations in the propensity of different sects to participate in politics are statistically significant, the question remains open as to whether those variations signify highly divergent approaches to politics, or whether the overall low level of participation, which points to convergence, is the more important finding. One way of evaluating these alternative explanations is to compare propensities to engage in types of political participation that imply a high degree of commitment and radicalism. These two types of participation are, respectively, having joined a political party or group, and having participated in a political demonstration.

Party Membership

Only one tenth of the sample had ever joined a political party, and even fewer (7%) were actually a member of one when surveyed. The apparent lack of interest in party membership is consistent with the students' generally low opinion of the parties and may also have been connected to the fact that 44% were too young to vote, when surveyed. Given the uniformity within the sample on this issue, it was not surprising that there were no significant inter-sect differences.

Participation In Demonstrations

In this age group, a more reliable indicator of political activism is willingness to take part in political demonstrations, and 17% of the sample reported having done so. There were significant differences between the sect groupings; Christians (21.9%) were more likely to have demonstrated than Muslims and Druze (13.2%). [26] The percentage distributions within the various sects are set out in Table 12.

                    Percentage Of Respondents Who Have
                        Ever Demonstrated, By Sect
Sect     %    (N)
Maronite 23.3 (35)
Orthodox 20.9 (42)
Shiite   16.7 (18)
Druze    14.8  (8)
Sunni    11.5 (31)

The data suggest that the overall measure of political participation conceals a significant difference in level of intensity between Christians and Muslims. The former's greater propensity to demonstrate--particularly on the part of the Maronites--may suggest dissatisfaction with the status quo; a belief that other forms of participation are ineffective; or high levels of intra-sect organization. Orthodox respondents are, overall, not particularly active in terms of their political behavior, but join the Maronites in their willingness to demonstrate. Conversely, the low level of participation by Sunni, whether measured broadly or only on the item that suggests willingness to resort to comparatively extreme political behavior, may indicate their relative satisfaction with the status quo and/or the relatively low level of organization among Sunni. Shiites, who have the highest level of participation measured broadly, are less likely to have demonstrated than respondents of the Christian sects. Belying their im age, Shiite respondents seem disproportionately to be participating within the system".

Preferred Means Of Bringing Change

In order to contrast reported past political behavior with their intended actions in the future, and in order to investigate the possible consequences for Lebanon of their apparent disenchantment with the current system, respondents were asked to indicate whether or not they supported various methods of bringing about political change. The options ranged from voting, through party activism and public rallies, to armed struggle. The results shown in Table 13 broadly confirm earlier findings regarding the divergence between sect groupings in terms of their political inclinations.

        Respondents' Views On The Best Way To Bring About Political
          Change, By Percentage Within Major Sect Groups (N=782)
                       Muslim/Druze Maronite/Orthodox
Favored method                    %                 %
Voting in elections            38.7              31.3
Working within a party         30.4              23.6
Public rallies                 25.1              35.0
Armed struggle                 12.5              13.1
Note: In this multiple response item respondents
could choose as many options as they liked. The percentages
therefore do not add up to 100.

Muslims (including Druze) were significantly more likely than Christians to believe that political change would be affected by means of voting in elections and by working with parties or groups [27] and were much less inclined to favor joining public rallies against the government. [28] Significant differences also emerged with respect to expected distributions within sects: Sunni and Druze were less likely to believe in the efficacy of public rallies, while Shiites, Maronites and Orthodox were all rather more likely to do so [29] --results which are consistent with the findings regarding those sects' respective levels of involvement in past demonstrations. Table 13 also shows, however, that armed struggle was rejected by 87.2% of all respondents, with neither sect nor religious grouping exerting any significant effect on the results.

Different types and levels of political participation between respondents seem to point to divergence rather than convergence. Rough characterizations of respondents from the major sects indicate that Sunni tend to political quietude, Shiites to broad engagement with the political system, and Christians of all sects to a comparatively disproportionate amount of political participation that suggests high commitment and possibly dissatisfaction with the existing, formal channels of political participation. These differences in political behavior appear to substantiate observed differences in attitudes, with Muslims being much more accommodated to the provisions of Ta'if than Christians. The exception lies in attitudes to the confessional system upon which Ta'if is based, about which Muslims have greater reservations than Christians.

LIFE STYLES AND ECONOMIC OPPORTUNITIES

The divergence of political attitudes and behavior according to sect would be particularly worrisome were it to be underpinned by sharp differences in degree of religiosity, perceptions of economic opportunity, and by patterns of attitudes held by males and females - a set of indicators that might be thought of (for want of a better term) as defining a "life style" dimension. This, however, appears not to be the case, as the students in the sample demonstrated considerable homogeneity in these areas.

Religiosity

Respondents' religiosity was evaluated by an item that asked them how they would describe themselves in religious terms, with the five close-ended responses ranging from "very religious" (scoring 5) to "not religious at all" (scoring 1). No significant differences were found between sects. Indeed, as the results in Table 14 show, the mean responses for each sect were very similar, and the sample as a whole could be described as quite religious.

           Levels Of Self-Described Religiosity, By Sect (N=775)
          Mean
Druze    3.316
Orthodox 2.817
Shiite   2.789
Maronite 2.754
Sunni    2.751

While a standard assumption of many casual observers is that Shiites are more religious than those of other sects in Lebanon, the responses of these students from Lebanon's major sects suggest very similar levels of religiosity.

Economic opportunities

Primary economic challenges facing youth include preparing for a career and obtaining suitable employment, challenges that have been particularly difficult for Lebanese university students who have had to endure a civil war and its devastating economic consequences. In order to gauge the level of optimism/pessimism regarding these key economic challenges, respondents were asked how easy they thought it would be to find a job for which university study qualified them; the extent to which they agreed that things are economically more difficult for the current generation than for the previous one; and the extent to which they agreed that the only way to succeed economically is to find work in another country. They were also asked to indicate what they thought most likely to help them get a job, including family connections in and out of Lebanon; political connections; their own qualifications; and/or their own initiative. [30]

Responses to these questions did not reveal any significant differences between sects. Well over half the sample (55%) were quite uncertain about their prospects for finding a job, with a further one-fifth of respondents expecting to find it very or quite difficult. An overwhelming majority (84 per cent) agreed or strongly agreed that things were more difficult for this generation. As Table 15 shows, in the rank order of importance regarding what would most help in finding a job, qualifications ranked first but were closely followed by family connections, with political connections and one's own initiative or merit placed rather further down the list. Just over one third of responses related to the need for connections abroad - a reflection of the fact that 37.5% of the sample had agreed, in responding to another item, that success was linked to emigration.

        Rank Order Of Sample's Responses Regarding Factors Thought
                   Most Likely To Help Get A Job (N=917)
                                   % of responses
Own qualifications                      66.3
Family connections in Lebanon           65.9
Political connections                   41.0
Own initiative/merit                    40.7
Family connections outside Lebanon      34.0
Note: In this multiple response item respondents
could choose as many options as they liked.
The percentages therefore do not add up to 100.

These data suggest that regardless of sect, Lebanese university students, even from elite institutions, anticipate difficulty in finding employment and think that the challenges they face are more difficult than those faced by their parents. Both one's own qualifications and motivations, on the one hand, and connections on the other, are widely considered important for economic success. In sum, regardless of sect, they see the challenges before them and the ways to meet those challenges in very similar ways. Perceived access into the economic system does not appear to vary appreciably by sect.

Gender

Gender-determined differences in attitude and behavior also lend support to an interpretation of convergence between sects, for the differences that did exist tended to do so across all sects. This suggests that patterns of gender relations are quite similar regardless of sect, which in turn indicates that sub-cultural differences between sects are relatively small. Males, for example, were found to be significantly more likely to have suffered in the civil war, [31] and much more likely to participate in politics than females, [32] whatever their respective sects. Interestingly, Muslim females (Sunni, Shiite and Druze) reported exactly the same mean level of political participation (5.8 compared to the overall mean of 6.3 in the sample) as did Christian females (Maronite and Orthodox). The findings regarding political participation are discussed in more detail below.

Males across the sample were somewhat more likely to expect that they would find a job than females. [33] The perceived necessity of emigrating appeared to be reflected in the fact that females were only half as likely as males to think that connections outside Lebanon would be useful [34] --a finding presumably related to the fact that more males of all sects leave Lebanon to find work. As mentioned earlier, females in the sample were of significantly higher socioeconomic status than males [35], with no significant inter-sect differences. This suggests among other things that parents pursue similar strategies when making choices concerning the education of their sons and daughters, and that they do this regardless of sect, despite the currency of certain stereotypes regarding, for example, the cloistering of Muslim women. Females were 1.7 times more likely than males to think that their own qualifications would help them find a job, [36] and they rated their own academic performance significantly higher tha n did males. [37] Women also had some views which set them apart from men; females were much less likely to be optimistic about either the chances for a settlement of the Arab-Israeli conflict, or of the consequences of such a settlement than males [38], while their declared levels of personal religiosity and rejection of armed struggle were both significantly higher than those of men in the sample. [39] These gender-based differences obtained in all sects.

Only in relation to attitudes toward the political system did some sect-based differences emerge between the responses of males and females in the same sect (see Table 16). The wide gap between male and female Shiite respondents in their levels of confidence is unique, as is the fact that Maronite females in the sample expressed more confidence in the political system than did male Maronites. [40]

       Ranking Of Mean Levels Of Confidence In The Political System,
                              By Sect And Sex
         MEANS
Sect     Male  Female
Shiite   15.90 14.17
Sunni    15.60 14.86
Druze    13.64 14.96
Orthodox 13.21 13.16
Maronite 11.77 13.00

Overall, and despite some observers' tendency to stereotype Muslim women's opportunities, attitudes and life styles as somehow more inhibited and restricted than those of Middle Eastern Christian women, the data strongly support an interpretation of convergence of sects in gender-related differences of attitude and behavior. It thus appears to be political issues, rather than those based on gender relations, that separate the sects.

The findings reported above indicate that, while both attitudes and behavior in relation to political issues and institutions tend to suggest divergence between sects, the students' responses to what might broadly be called "life style" and economic opportunity issues indicate convergence. It will be recalled that there was also more convergence than divergence within the sample on some related issues discussed earlier - issues such as their failure to benefit from reconstruction and their view that economic development is being impeded by confessionalism. Lebanese youths appear not to have strong sub-cultural differences as a result of being a member of a particular sect, but they do have significantly different political attitudes. The good news is that shared cultural orientations and life experiences probably make communications and negotiations easier. The bad news is that the markedly different political attitudes of Christians and Muslims will presumably make such communications and negotiations essen tial to avoid deterioration of the political system or to enable it successfully to confront a major crisis.

CONCLUSION

The findings of this survey of Lebanese students point to a paradox. On the one hand what might be thought of as the deeper, underlying attitudes of these students tend to be converging. There are very few significant differences between Muslims and Christians on how they perceive the economic challenges of life, their levels of religiosity, or in the patterns of attitudinal differences of young men and women. This may in turn suggest that the homogenizing forces of education, class and youth have a stronger impact on these non-political aspects of life than does being a member of a particular sect. Certainly, when asked to comment on how they saw the overall political and economic future of Lebanon, the students who responded were equally divided between optimists (43.2%) and pessimists (43.8%) with the remainder undecided or apathetic. Sect was immaterial to these results.

On the other hand, political attitudes and behavior do appear to be diverging on the basis of sect. Maronites and Sunni tend to be most strongly polarized, with the former being most critical of the political system and alienated from it, while the latter are most content with the political status quo. Christian and Muslim students generally, if not quite as polarized as the findings for Maronites and Sunni might suggest, nevertheless think about politics and behave politically in distinctive ways. The patterns that emerge from these differences suggest that Muslims tend to be much more supportive of the second Lebanese Republic, in terms of both attitudes and behavior, than do Christians, who seem to be more alienated from it and more willing to engage in extra-systemic political behavior.

A convergence of attitudes about life style and economic opportunity appears to be one characteristic of the young educational elite of Lebanon. Whether such convergence provides a basis upon which differences in political attitudes and behavior can be overcome probably depends on the intensity of feelings regarding a particular issue to be negotiated. If the issue were to be an extremely divisive one, such as changing the confessional system or making peace with Israel, the divergence of attitudes suggests that the task would place a heavy burden on inter-sect relationships. But the political fault lines that divide these students by sect do not divide them in terms of their life styles or their relationship to the economy. This final paradox tempers somewhat the conclusion that the Second Lebanese Republic is built upon the shaky foundations of sectarian divergence.

Christine Asmar teaches at the University of Sydney, Australia; Maroun Kisirwani is a professor at the American University of Beirut; and Robert Springborg is a professor at Macquarie University, Australia.

NOTES

(*.)This research was funded by a grant from the Australian Research Council (ARC). Support and facilities for the research were provided by the American University of Beirut and by Macquarie University in Sydney. The authors wish to express their gratitude to Cathy Rytmeister and especially to Mary Lam for their work on the statistical analysis, and to the students at Macquarie University who assisted with the coding. The authors are also grateful to all those in Lebanon who facilitated the on-campus administration of the surveys, namely: Randa Antoun, Tarif Bazzi, George Bitar, Mary Deeb, Itimar Diab, Farid El-Khazin, Margarit Fabri-Helou, Nizar Hamzeh, Wadi' Jureidine, Albert Kur'an, Michel Nehme, Musa Nimah, Elie Salem and Edward Vitale. Above all, the authors wish to thank the Lebanese students whose cooperation made the whole study possible.

(1.) This term was coined by Michael Hudson in his 1968 book Precarious Republic: Political Modernization in Lebanon. (New York: Random House).

(2.) In Sell's study of Egyptian students, for example, he noted that sampling students helps to overcome the difficulty of surveying populations unaccustomed to strangers soliciting their opinions. He also commented that elite students are political "bellwethers," since they are likely to become tomorrow's leaders. Ralph R. Sell. "International Affinities in Modem Egypt: Results from a Social Distance Survey of Elite Students." International Journal of Middle East Studies 12 (1990):60.

(3.) Lebanese students (but not foreigners) must obtain the Lebanese Baccalaureate--a difficult task.

(4.) According to official university figures for the year of the survey (1996/97), 23% of undergraduates at LAU received financial aid, while 50% at Balamand and 30% at AUB held some kind of scholarship.

(5.) Questions covered their religion and religiosity; their political preferences and behaviors; their experiences both of civil war and of post-war reconstruction; their view of institutions in Lebanon such as democracy, confessionalism, parliament, parties and groups (including Islamic); the Arab-Israeli conflict; their own and their country's economic prospects; the political future of Lebanon; and their personal sense of efficacy in the face of all these issues and developments.

(6.) Two percent of the students refused to answer the question on sect, writing some comment to the effect that it was sectarian differences which had caused Lebanon's problems, and a further 0.7% simply left it blank.

(7.) Analysis of variance (ANOVA) was selected as the most powerful statistical test available for the analysis of responses involving both numeric variables, and ordinal variables that could be upgraded to numeric. In cases where the ANOVA test was not appropriate, the Chi-square ([X.sup.2]) test was used instead. An advantage of the Chi-square test is that it enables percentage distributions of responses from various groups to be included in tables and in the text, which clarifies some issues for the reader.

(8.) Generally, only statistically significant findings (p[less than]0.05) will be reported in the text, although tables setting out percentage distributions will be provided where further detail is necessary. Here the two significant findings were, respectively: t=3.92,df=863,p[less than]0.0005; t=3. 18, df=863, p[less than]0.001.

(9.) The mean responses of the major sect groups regarding civil war suffering were: Shiite 1.50, Druze 1.37, Maronites 1.32, Orthodox 1.11, Sunni 1.04.

(10.) t=5.01, df=864, p=0.025

(11.) The political context in which the students lived meant that they were unlikely to be frank about their attitudes to the Syrian presence and no such question was put to them, despite the undoubted pertinence of the issue for most Lebanese.

(12.) Respondents were asked to indicate on a 5-point scale the extent to which they agreed with statements that the political system is democratic; that civil liberties are guaranteed; and that a) political parties, and b) Islamic groups, effectively represent people's interests. They were also asked to indicate (again, on a 5-point scale) how satisfied they were with the current parliament's performance, and how much citizens can influence government policies. The Alpha reliability coefficient for the six items of the scale was .6798, indicating a high inter-item correlation, and therefore suggesting that the scale was tapping a single dimension.

(13.) Respectively: t=3.39, t=5.76 and t=7.58 while for all three df=863, p[less than]0.0005

(14.) Respectively: t=3.94, t=5.52 while for both df=863, p[less than]0.0005

(15.) Respectively: t=5.15, t=3.95 while for both df=860, p[less than]0.0005

(16.) The categories used in both items ranged from "Strongly agree" through "Agree" and "Neither agree nor disagree" to "Disagree" and "Strongly disagree". The small number (2-2.5%) who chose not to respond have been omitted from the tables.

(17.) The Chi square statistic ([X.sup.2]) which was used in this part of the analysis shows up statistically significant differences between actual results and the results which would be expected on the basis of how various groups are distributed within the sample.

(18.) F=1.95, df=6.832, p=0.071

(19.) t=3.9, df=863, p[less than]0.0005

(20.) Respectively: t=3.7, t=4.99 and t=3.51; in all 3 cases, df=863, p[less than]0.0005

(21.) The term "Lebanese nationalist" does not necessarily denote a particular political tendency, but rather may indicate a more general feeling of patriotism and loyalty to one's country, quite possibly heightened by the presence of foreign troops within the homeland.

(22.) There exist a number of possible explanations for the somewhat unexpected finding regarding Hizbollah. The Shiite practice of taqiya or dissimulation, which amounts to concealing the truth in order to protect one's life, may well be operating in a context where there is considerable - and probably justifiable - fear of Israeli agents. On a more prosaic level, there are simply not very many Shiites at the universities to begin with; in this sample, for example, Shiites constituted a mere 12% of the total students surveyed.

(23.) In Table 9 the large number of categories containing fewer than 5 cases means that the assumptions on which Chi-square analysis is based, are in fact violated. The percentage distributions are therefore given only for their intrinsic interest, and the high level of statistical significance should not be given too much weight.

(24.) r=-0.0829,df=881,p=0.014

(25.) Respectively: t=3.48,df=863,p[less than]0.0005; t=3.14,df=863,p[less than]0.001; t=3.2,df=863,p[less than]0.001

(26.) [X.sup.2]=l0.34,df=1,p=0.001

(27.) [X.sup.2]=4.64,df=1,p=0.031; [X.sup.2]=4.43,df=l,p=0.035

(28.) [X.sup.2]=9.27,df=l,p=0.002

(29.) [X.sup.2]=17.16,df=4,p=0.002

(30.) All these questions were closed-ended, with 5 options (ranging from "Strongly agree" to "Strongly disagree" offered in the first four questions, and six Yes/No options in the last question.

(31.) t=2.21,df=878,p=0.028

(32.) t=8.36,df=912,p[less than]0.0005

(33.) t=1.98,df=912,p=0.048

(34.) [X.sup.2]=33.15,df=2,p=0.005

(35.) t=2.67,df=912,p=0.008

(36.) [X.sup.2]=l5.79,df=2,p=0.0004

(37.) t=2.49,df=912,p=0.013

(38.) Respectively: t=5.44,df=912,p[less than]0.0005; t=3.33,df=9l2,p=0.00l

(39.) Respectively: t=2.94,df=9l2,p=0.003; [X.sup.2]9.39 df=1,p=0.002

(40.) One might be tempted to explain these findings in terms of stereotypes regarding the ways Muslim women differ from Christians in public life - in other words, that Muslim women feel less confident than their men because they participate less, and that Christian women are more confident because they participate more. But it has already been noted above that, in this sample, female levels of political participation--while lower than those of men--are identical across sects.