Recent poll data, and particularly Professor Samuel A. Stouffer’s Communism, Conformity, and Civil Liberties,1 raise an almost infinite series of questions concerning what might be termed the ecology of tolerance for political and cultural dissent; they show, on a great variety of civil-liberties issues, that the educated are relatively more tolerant than the uneducated—this is the most salient difference, even when one sorts it out from age, income, and region; that the young are more tolerant than the old; men than women; people in the Far West and East than in the South and Midwest. “Tolerance,” of course, is a word of many meanings; I shall not seek to unravel them here, but rather to develop Stouffer’s own conclusion that tolerance as he defines it goes together with wide orbits of education, work, and travel; I shall hope to clarify his and similar data and to suggest some of their implications for the future of political leadership. And since I believe that the data may be somewhat clarified by a better understanding of the methods used in obtaining them, I want to dwell on some of the interviewing problems involved, and especially on Stouffer’s innovation in interviewing simultaneously a national cross-section and a sample of “community leaders”—mayors, bar association presidents, PTA heads, DAR regents, and so on. This research design is a contribution to our understanding of the institutionalization and rationalization of political attitudes, but one requiring development of a more systematic distinction between anthropological interviewing of informants and poll interviewing of respondents on such bifurcated political surveys.