¹ There are exceptions. A recent study of Canadian citizens and decision makers reported that mass-elite differences were small or reversed on issues of public order or protecting society's moral values by banning certain types of publications and that the public was more tolerant than the political elites on issues of wire-tapping. See Fletcher, Joseph, ‘Mass and Elite Attitudes About Wiretapping in Canada: Implications for Democratic Theory and Politics’, Public Opinion Quarterly, 53 (1988), 225–45CrossRefGoogle Scholar; and Sniderman, Paul M., Fletcher, Joseph F., Russell, Peter H. and Tetlock, Philip E., ‘The Fallacy of Democratic Elitism: Elite Competition and Commitment to Civil Liberties’, British Journal of Political Science, 21 (1991), 349–70, p. 361CrossRefGoogle Scholar. A similar finding has been reported in a study of British MPs and citizens by Barnum, David G. and Sullivan, John L., ‘Attitudinal Tolerance and Political Freedom in Britain’, British Journal of Political Science, 19 (1989), 136–46, p. 145CrossRefGoogle Scholar. More generally, a study based on some of the data reported in this article found smaller differences in political tolerance between Knesset Members and the general public in Israel than in other countries and greater intolerance towards the right-wing group, Kach, among the MKs than the general public. See Shamir, Michal, ‘Political Intolerance Among Masses and Elites in Israel: A Re-evaluation of the Elitist Theory of Democracy’, Journal of Politics, 53 (1991), 1018–43, p. 1026CrossRefGoogle Scholar. Finally, Duch, Raymond and Gibson, James L., ‘“Putting Up With” Fascists in Western Europe: A Comparative, Cross-Level Analysis of Political Tolerance’, Western Political Quarterly, 45 (1992), 237–73CrossRefGoogle Scholar; and Gibson, James L. and Duch, Raymond, ‘Elitist Theory and Political Tolerance in Western Europe’, Political Behavior, 13 (1991), 191–212CrossRefGoogle Scholar, found that Western European political elites are sometimes less tolerant of fascists than ordinary citizens. Despite these somewhat circumscribed exceptions, most research shows greater elite tolerance to be the rule.

² Stouffer, Samuel, Communism, Conformity, and Civil Liberties (New York: Doubleday, 1955)Google Scholar. See also Dahl, Robert, Who Governs? (New Haven, Conn.: Yale University Press, 1961)Google Scholar; McClosky, Herbert, ‘Consensus and Ideology in American Politics’, American Political Science Review, 58 (1964), 361–82CrossRefGoogle Scholar; Prothro, James and Grigg, Charles, ‘Fundamental Principles of Democracy: Bases of Agreement and Disagreement’, Journal of Politics, 22 (1960), 276–94.CrossRefGoogle Scholar

³ But see McClosky, Herbert and Brill, Alida, Dimensions of Tolerance (New York: Russell Sage, 1983), pp. 236–43Google Scholar and Shamir, , ‘Political Intolerance Among Masses and Elites in Israel’.Google Scholar

⁴ Most notably, Berelson, Bernard, Lazarsfeld, Paul and McPhee, William, Voting (Chicago: University of Chicago Press, 1954), pp. 313–23.Google Scholar

⁵ See McClosky, and Brill, , Dimensions of Tolerance, p. 250Google Scholar; Nunn, Clyde A., Crockett, Harry J. and Williams, J. Allen, Tolerance for Nonconformity (San Francisco: Jossey-Bass, 1978), pp. 60, 97, 103Google Scholar; Prothro, and Grigg, , ‘Fundamental Principles of Democracy’Google Scholar; and Stouffer, , Communism, Conformity, and Civil Liberties, p. 127.Google Scholar

⁶ Dye, Thomas R. and Zeigler, Harmon, The Irony of Democracy (Pacific Grove, Calif.: Brooks/Cole Publishing Company, 1990), chap. 5.Google Scholar

⁷ Renshon, Stanley, ‘Psychological Perspectives on Theories of Adult Development and the Political Socialization of Leaders,’ in Sigel, Roberta S., ed., Political Learning in Adulthood (Chicago: University of Chicago Press, 1989), pp. 203–64.Google Scholar

⁸ Altemeyer, Bob, Enemies of Freedom (San Francisco, Calif.: Jossey-Bass, 1989), p. 95.Google Scholar

⁹ McClosky, and Brill, , Dimensions of Tolerance, pp. 232–4.Google Scholar

¹⁰ Of course, the normative consequences of these two explanations are enormously at odds with one another. Selective recruitment is especially congruent with democratic elitism and conservative variants of democratic theory. Adult political socialization, on the other hand, implies a more optimistic view and suggests that substantially greater political involvement of ordinary citizens could attenuate their intolerance.

¹¹ See, for example, Eulau, Heinz and Czudnowski, Moshe M., Elite Recruitment in Democratic Polities (Beverly Hills, Calif.: Sage Publications, 1976)Google Scholar; and Verba, Sidney, Nie, Norman H. and Kim, Jae-On, Participation and Political Equality (Chicago: University of Chicago Press, 1978).Google Scholar

¹² We are assuming that most of the selective recruitment effects are from childhood and early adult socialization processes and that adult political experience is the heart of the socialization hypothesis. This distinction, of course, is probably not as tidy as our discussion implies.

¹³ Stouffer, , Communism, Conformity, and Civil LibertiesGoogle Scholar, defined community leaders as school board presidents, mayors, chairs of local political party organizations and so on.

¹⁴ Jackman, Robert, ‘Political Elites, Mass Publics, and Support for Democratic Principles’, Journal of Politics, 34 (1972), 753–73.CrossRefGoogle Scholar

¹⁵ Nunn, , Crockett, and Williams, , Tolerance For Nonconformity, p. 152.Google Scholar

¹⁶ McClosky, and Brill, , Dimensions of Political Tolerance, pp. 252–3.Google Scholar

¹⁷ Sniderman, Paul M., Personality and Democratic Politics (Berkeley: University of California Press, 1975), pp. 204–22Google Scholar; Sullivan, John L., James Piereson and Marcus, George E., Political Tolerance and American Democracy (Chicago: University of Chicago Press, 1982), pp. 145–56Google Scholar; and Sullivan, John L., Shamir, Michal, Walsh, Patrick and Roberts, Nigel, Political Tolerance in Context: Support for Unpopular Minorities in Israel, New Zealand, and the United States (Boulder, Colo.: Westview Press, 1985), pp. 186–90.Google Scholar

¹⁸ Sniderman, , Personality and Democratic Politics, pp. 262–72Google Scholar, found that ‘the critical dimension of self-regard which most sharply separates leaders and followers is interpersonal competence’. His analysis showed that, among samples of the general public, individuals with high self-esteem participated more than people with low self-esteem: among those high on ‘interpersonal competence’, 45 per cent were also high on his participation scale, while among those low in interpersonal competence, only 26 per cent were high on the participation scale. He also found that leaders (national convention delegates) had higher interpersonal competence than followers (the general population). Political elites may be more tolerant than the public in part because they are higher in self-esteem and lower in dogmatism even before entering politics. A corollary of this thesis is that individuals who are excessively dogmatic might be weeded out during the democratic process of selecting political leaders. Highly dogmatic candidates will not be sufficiently flexible to adjust to changing circumstances and will more often lose elections.

¹⁹ Of course, longitudinal data would be the best suited to testing hypotheses about selective recruitment and adult political socialization, but we lacked resources to collect such data. See Appendix A for a discussion of our data collection procedures.

²⁰ Verba, , Nie, and Kim, , Participation and Political Equality, pp. 290–303.Google Scholar

²¹ See, among others, McClosky, and Brill, , Dimensions of Political Tolerance, p. 293Google Scholar; Sullivan, , Piereson, and Marcus, , Political Tolerance and American Democracy, pp. 175–86Google Scholar; and Sullivan, et al. , Political Tolerance in Context, pp. 195–9.Google Scholar

²² Although this might seem to be tautological, it is not. Early empirical researchers discovered that while citizens of the United States demonstrated a consensus on principles such as freedom of speech and minority rights, they were sharply divided on whether to apply these principles to specific unpopular political groups (Prothro, and Grigg, , ‘Fundamental Principles of Democracy’Google Scholar; McClosky, , ‘Consensus and Ideology in American Politics’)Google Scholar. This research demonstrated a sharp distinction between support for democratic norms or principles and applied political tolerance. Many citizens who endorsed freedom of speech as a principle were prepared to deny it to specific groups they disliked and/or feared. Later research showed that, although the earlier findings were true, people who most strongly and consistently endorsed democratic principles tended to be more tolerant towards specific disliked groups (Sullivan, , Piereson, and Marcus, , Political Tolerance and American Democracy, pp. 202–7)Google Scholar. Endorsement of specific legal guarantees for these principles – such as ensuring the same legal protections even for those accused of treason – also enhances political tolerance (Sullivan, et al. , Political Tolerance in Context, pp. 217–22).Google Scholar

²³ Sniderman, , Personality and Democratic Politics, p. 195Google Scholar; Sullivan, , Piereson, and Marcus, , Political Tolerance and American Democracy, pp. 159–61, 188–90Google Scholar; and McClosky, and Brill, , Dimensions of Tolerance, pp. 260, 368.Google Scholar

²⁴ Browning, Rufus P. and Jacob, Herbert, ‘Power Motivation and the Political Personality’, Public Opinion Quarterly, 28 (1964), 75–90CrossRefGoogle Scholar; and Eulau, and Czudnowski, , Elite Recruitment in Democratic Societies.Google Scholar

²⁵ See, for example, Stouffer, , Communism, Conformity, and Civil LibertiesGoogle Scholar; and Nunn, , Crockett, and Williams, , Tolerance for NonconformityGoogle Scholar. McClosky, and Brill, , Dimensions of Tolerance, pp. 481–8Google Scholar, compared their mass and elite samples using a 52-item civil liberties scale. That scale contained a mixture of items from various sub-scales, including free speech, free press, symbolic speech, academic freedom and so on. Only a few of the questions referenced specific groups. Most of the items were quite general and did not mention specific action by specific target groups, but instead specified generic ‘rapists’, ‘students’, ‘arrested persons’, ‘a professor’ and so on.

²⁶ See Sullivan, et al. , Political Tolerance in Context, chap. 6, in particular Table 6.1, p. 151.Google Scholar

²⁷ The fact that political elites in our surveys are more likely to dislike right-wing groups, however, does not necessarily translate into less tolerance towards the right wing, even given our attempt to provide equally difficult tests for the public and their elected representatives. Political elites may select right-wing targets more often, but tolerate them more than most citizens tolerate their own targets.

²⁸ Because of data limitations, personality variables are not available for Britain and the United States.

²⁹ See Appendix B for a discussion of how these variables were measured.

³⁰ See Appendix B for a discussion of how these variables were measured.

³¹ In particular, we have loaded the dice against the political socialization hypothesis by including controls for the norms of democracy. Support among political elites for democratic norms is often viewed as likely to increase as a result of their level of participation (McClosky, and Brill, , Dimensions of Political Tolerance, p. 234).Google Scholar

³² Jackman, , ‘Political Elites, Mass Publics, and Support for Democratic Principles’Google Scholar; McClosky, and Brill, , Dimensions of Political Tolerance, pp. 474–5Google Scholar; and Nunn, , Crockett, and Williams, , Tolerance for Nonconformity, pp. 151–2.Google Scholar

³³ Education had no direct impact on tolerance in an earlier data set of American citizens when personality variables were controlled. See Sullivan, , Piereson, and Marcus, , Political Tolerance and American Democracy, pp. 221–5.Google Scholar

³⁴ This result is consistent with other depictions suggesting that Israeli political culture puts less emphasis on minority rights and freedom of speech and more on majority rule, than the Anglo-Saxon democracies. See Shapiro, Y., Israeli Democracy (Ramat Gan: Massada, 1977)Google Scholar; and Shamir, Michal and Sullivan, John L., ‘The Political Context of Tolerance: The United States and Israel’, American Political Science Review, 77 (1983), 911–28, pp. 914, 922, 923.CrossRefGoogle Scholar

³⁵ In the 1985 survey, only 70 per cent of the Israeli public and 69 per cent of MKs endorsed ‘freedom of speech’ compared to figures around 90 per cent in the other countries.

Among elites in the other three countries, an interesting pattern emerges. Procedural norms have an impact only in the United States, which is the only one of the four with a written constitution. Political elites are generally in an advantageous position to comprehend the major form in which democratic norms are ‘packaged’ in a society, and the elites of New Zealand and Britain pay particular attention to the abstract, general norms of democracy while those in the United States respond to how these norms are embodied in legal protections and procedures.

³⁶ Shamir, Michal, ‘Kach and the Limits to Political Tolerance in Israel’, in Elazar, D., Penniman, H. and Sandier, S., eds, Israel's Odd Couple: The 1984 Elections and the National Unity Government (Detroit, Michigan: Wayne State University Press, 1990), pp. 159–68.Google Scholar

³⁷ In all of the other countries, support for democratic norms and tolerance are strongly and positively related. The discovery of a negative relationship allows us to apply this explanation – the paradox of tolerance – to Israel.

³⁸ As discussed later, political elites are not less threatened than the public by their least-liked groups. They simply are more tolerant even in the face of similar threat perceptions. Among samples of the public, in all four countries, perceptions of threat and both procedural and general democratic norms had direct and significant effects on tolerance. Dogmatism also had a strong impact on citizens' tolerance in all three countries in which it was measured.

³⁹ There was of course a very large difference in tolerance between these two groups of MPs. The mean among MPs who selected Sinn Fein was 16.1 while among those selecting all other groups it was 22.4 (18 is the scale midpoint).

⁴⁰ Our findings are consistent with Searing's conclusion, using different methods, that while both socialization and selective recruitment were at work among British elites, socialization factors were more important than selective recruitment. In parallel with our research on tolerance, one of the variables he examined was support for the ‘rules of the game’. See Searing, Donald, ‘A Theory of Political Socialization: Institutional Support and Deradicalization in Britain’, British Journal of Political Science, 16 (1986), 341–76.CrossRefGoogle Scholar

⁴¹ McClosky, , ‘Consensus and Ideology in American Polities’, p. 375Google Scholar; McClosky, and Brill, , Dimensions of Tolerance, p. 233.Google Scholar

⁴² Searing, Donald, ‘Political Involvement and Socialization in Great Britain’, in Clarke, Harold D. and Czudnowski, Moshe M., Political Elites in Anglo-American Democracies (DeKalb, Ill.: Northern Illinois University Press, 1987), pp. 109–46, at p. 109Google Scholar. Searing earlier had warned, however, that among elites, support for democratic norms is founded upon political partisan bias, and that elites are fully capable of abrogating such norms when it is in their best interests to do so. See Searing, Donald, ‘Rules of the Game in Britain: Can the Politicians Be Trusted?’ American Political Science Review, 76 (1982), 239–58CrossRefGoogle Scholar. Also see Shamir, , ‘Political Intolerance Among Masses and Elites in Israel’, pp. 1032–6Google Scholar, for an explanation of Israeli elites' tendency to abrogate these norms for political gain.

⁴³ As noted earlier, some of our tests for selective recruitment actually control for differences that could be due to adult political socialization experiences.

⁴⁴ Sullivan, , Piereson, and Marcus, , Political Tolerance and American Democracy, pp. 186–94Google Scholar; Sullivan, et al. , Political Tolerance in Context, pp. 199–203Google Scholar; Stouffer, , Communism, Conformity, and Civil Liberties, chap. 8Google Scholar; and Peffley, Mark and Sigelman, Lee, ‘Intolerance of Communists During the McCarthy Era: A General Model’, Western Political Quarterly, 43 (1990), 93–111.CrossRefGoogle Scholar

⁴⁵ The threat questions used in the sample of the American public were not identical to those used in the survey of MCs so a direct comparison would be misleading. The results for the Israeli public are from the 1980 survey.

⁴⁶ Stouffer, , Communism, Conformity, and Civil Liberties, p. 195.Google Scholar

⁴⁷ Avery, Patricia, Bird, Karen, Johnstone, Sandra, Sullivan, John L. and Thalhammer, Kristina, ‘Exploring Political Tolerance with Adolescents’, Theory and Research in Social Education (forthcoming, 1992).Google Scholar

⁴⁸ Marcus, George E., Sullivan, John L., Theiss-Morse, Elizabeth, and Johnstone, Sandra, ‘Political Tolerance and Threat: Affective and Cognitive Influences’ (Williamstown, Mass.: Williams College, unpublished manuscript, 1992).Google Scholar

⁴⁹ Although both public and elites are strongly committed to democratic norms and this commitment in turn is highly correlated to tolerance for both groups, perhaps elite political experience fosters a qualitatively more profound commitment to democratic norms, which serves to uncouple threat perceptions from levels of tolerance. Gibson, James L. and Bingham, Richard D., ‘Elite Tolerance of Nazi Rights’, American Politics Quarterly, 11 (1983), 403–28CrossRefGoogle Scholar, found that among members of the American Civil Liberties Union and Common Cause support for democratic norms played a very strong role in determining their responses to the 1978 conflict over the Nazis' right to march in Skokie.

⁵⁰ In addition to these considerations, the dual acts of electoral campaigning and engaging in policy making and governance carry with them unique and ubiquitous educational experiences that can enhance tolerance. Both acts increase contact with ideological diversity, require the ability to compromise, and carry the responsibility to take a ‘sober second thought’ before acting, particularly when restricting the basic rights of citizens. The responsibility to govern also means that when elites are asked whether a particular group ought to be outlawed or prevented from exercising free speech, their answers are not entirely hypothetical. The issue positions of legislators have consequences – they are often translated into policy. The general public's views on issues of political tolerance do not generally carry with them such profound responsibility or weight.

⁵¹ Fletcher, , ‘Mass and Elite Attitudes About Wiretapping in Canada’Google Scholar; McClosky, , ‘Consensus and Ideology in American Polities’Google Scholar; and Milbrath, Lester and Goel, M. L., Political Participation, 2nd edn (Chicago, Ill.: Rand McNally, 1977).Google Scholar

⁵² Fried, Richard M., Nightmare in Red (New York: Oxford University Press, 1990), p. 36.Google Scholar

⁵³ Gibson, James, ‘Political Intolerance and Political Repression During the McCarthy Red Scare’, American Political Science Review, 82 (1988), 511–29, p. 518.CrossRefGoogle Scholar

⁵⁴ Shamir, , ‘Political Intolerance Among Masses and Elites in Israel’.Google Scholar

⁵⁵ Searing, Donald, ‘Rules of the Game in Britain’Google Scholar, and ‘Political Involvement and Socialization in Great Britain’. Thus although there is something about the adult political socialization of political elites that may enhance their political tolerance, the structure of partisan politics can also cause them to set aside their tolerance for political gain. McClosky, and Brill, , Dimensions of Tolerance, pp. 88–92, 117–29 and 255Google Scholar, found that many non-political elites (such as judges, lawyers, clergy) were also more tolerant than the general public, just as community leaders were. Some of them were even more tolerant than local politicians. Shamir, , ‘Political Intolerance among Masses and Elites in Israel’, p. 1028Google Scholar, also found that journalists and intellectuals were more tolerant than MKs.

We ran the regression equation reported for Israel in Table 4, including dummy variables for both intellectuals and journalists, and found that journalists were significantly more tolerant than the intellectuals, MKs and ordinary citizens, even controlling for selective recruitment. There were no significant differences among the latter three groups. Clearly, then, other occupations can provide experiences that encourage political tolerance, above and beyond selective recruitment. Unlike politicians, these other occupations may lack the partisan political advantages that can attenuate tolerance, thus leading to even greater levels of tolerance than among political elites.

⁵⁶ Barnum, David G. and Sullivan, John L., ‘The Elusive Foundations of Political Freedom in Britain and the United States’, Journal of Politics, 52 (1990), 719–39CrossRefGoogle Scholar; Barnum, David G., Sullivan, John L. and Sunkin, Maurice, ‘Constitutional and Cultural Underpinnings of Political Freedom in Britain and the United States’, Oxford Journal of Legal Studies (Autumn, 1992, forthcoming)CrossRefGoogle Scholar; Gibson, , ‘Political Intolerance and Political Repression During the McCarthy Red Scare’Google Scholar; and Gibson, James L., ‘The Structure of Attitudinal Tolerance in the United States’, British Journal of Political Science, 19 (1989), 562–70.CrossRefGoogle Scholar

⁵⁷ Sullivan, et al. , Political Tolerance in ContextGoogle Scholar; Barnum, and Sullivan, , ‘Attitudinal Tolerance and Political Freedom in Britain’, p. 137Google Scholar; and Gibson, , ‘The Structure of Attitudinal Tolerance in the United States’, p. 564.Google Scholar

⁵⁸ Shamir, , ‘Political Intolerance Among Masses and Elites in Israel’, pp. 1024, 1038–40.Google Scholar

⁵⁹ Sprinzak, Ehud, ‘Kach and Kahane: The Emergence of Jewish Quasi-Fascism’, in Arian, Asher and Shamir, Michal, eds, The Elections in Israel 1984 (New Brunswick, NJ: Transaction Books, 1986), pp. 169–87.Google Scholar

⁶⁰ Shamir, , ‘Kach and the Limits to Political Tolerance in Israel’.Google Scholar

⁶¹ Barnum, and Sullivan, , ‘Attitudinal Tolerance and Political Freedom in Britain’, p. 137.Google Scholar

⁶² See Sullivan, , Piereson, and Marcus, , Political Tolerance and American Democracy, for the exact items.Google Scholar