This article marks an attempt to clarify the teachings of Robert Michels. It suggests that inPolitical Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (1915), Michels presented a favorable account of the compatibility of organization and democracy.

Other treatments attribute to Michels a thesis of the following kind: (1) Large, organizationally complex associations, compared with small, simple associations, are likely to be governed by cliques whose powers (disposable resources, freedom of action, security of tenure) are abundant and whose policies (use of official status and resources) deviate from the policy preferences of their constituents. (2) Increments of Organization (of scale, or members, and of complexity, or procedural formality, functional differentiation, stratification, specialization, hierarchy, and bureaucracy) augment the powers and the policy-deviating propensities of leadersvis-à-vis followers.

¹ See S. M. Lipset's introduction to the Collier Books edition (1962) ofPolitical Parties, and the commentaries cited by Lipset. For additional statements or approximations of this version of Michels's thesis, and some challenges, see the following:Blau,P. M. andScott,W. R.,Formal Organizations (Chandler,1962), pp. 48,228Google Scholar;Brooks,R. C.,Political Parties and Electoral Problems (Harper,1933), p.30Google Scholar;Coker,F. W.,Recent Political Thought (Appleton-Century-Crofts,1934), p.328Google Scholar;Dahl,R. A. andLindblom,C. E.,Politics, Economics and Welfare (Harper,1953), pp.279–85Google Scholar;Gouldner,A. W., ed.,Studies in Leadership, pp.418–35Google Scholar (T. W. Adorno) and 477–504 (B. Barber);Hughes,H. S.,Consciousness and Society (Vintage,1951)Google Scholar, ch. 7;Keller,Suzanne,Beyond the Ruling Class (Random House,1963) pp. 72–3, 80, 263,273–74Google Scholar;Kornhauser,Arthur and others, eds.,Industrial Conflict (McGraw-Hill,1954)Google Scholar, ch. 9 (L. H. Fisher and G. McConnell);Lasswell,H. D. andKaplan,A.,Power and Society (Yale,1950)Google Scholar;MacIver,R. M.,The Web of Government (Macmillan,1959), pp. 122, 140,434Google Scholar;McKenzie,R. T.,British Political Parties (Praeger,1964 ed.), pp.15–17Google Scholar, ch. 11;Merriam,C. E. andBarnes,H. E., eds.,A History of Political Theories: Recent Times (Macmillan,1924), pp. 56–67,383Google Scholar;Nomad,Max,Aspects of Revolt (Farrar, Straus & Cudahy,1959)Google Scholar, ch. 1;Presthus,Robert,The Organizalional Society (Knopf,1962), pp. 4,41–52Google Scholar;Pfiffner,J. M. andSherwood,F. P.,Administrative Organization (Prentice-Hall,1960), p.338Google Scholar;Sartori,Giovanni,Democratic Theory (Wayne State University Press,1962), pp. 42, 82, 100, 120–28,134Google Scholar;Spitz,David,Patterns of Anti-Democratic Thought (Macmillan,1949), esp. p.27Google Scholar and the treatment in Part II of James Burnham'sThe Machiavellians;Truman,D. B.,The Governmental Process (Knopf,1955), pp.137–55Google Scholar; andWaldo,Dwight, “Development of Theory of Democratic Administration,” this Review, Vol.46 (031952), pp.100–01Google Scholar.

² Michels does not use the terms “democracy,” “oligarchy,” and “organization” in a consistent or coherent manner. The terminological difficulties have been probed byCassinelli,C. W., in “The Law of Oligarchy,” this Review, Vol.47 (Sept.1953), p.3 ffGoogle Scholar. However, Michels persistently associates democracy with equality, with conditions suggesting the notion of popular sovereignty, and with the “system in which delegates represent the mass and carry out its will.” On the other hand, he speaks of “The notion of the representation of popular interests, a notion to which the great majority of democrats … cleave ….”Political Parlies, trans.Eden, andPaul,Cedar (Dover Publications,1959), esp. pp. 1–2, 27,401Google Scholar. References hereafter will be to this edition unless designated otherwise.

³ Pp. 33, 40.

⁴ P. 402.

⁵ Pp. 329, 406.

⁶ Pp. 3, 5, 5n, 11, 63, 190, 194, 201.

⁷ Pp. 11, 32, 390, 400, 402.

⁸ We are excluding here Michels's arguments for the indispensability of Organization and his suggestions that the process of Organization tends to be self-accelerating. Attention is confined to the question of what arrangements can be compatible with the presence of Organization.

⁹ For a sophisticated discussion of these processes and of some implications, see Langer, above, note 1, esp. ch. 3.

¹⁰ Pp. viii, 22, 168, 402, 408.

¹¹ For example,Easton,David,The Political System (Knopf,1959), pp.56–7Google Scholar.

¹² P. 32.

¹³ P. 401.

¹⁴ P. 376; emphasis added. He also says, “A party is neither a social unity nor an economic unity.” (p. 387) His characterization ofchange in parties, however, presupposes initial unity. See section III below, underSocial Pluralism.

¹⁵ Pp. 27–8. Omitted from this quotation is a contradictory remark, illuminating Michels's chronic confusion about the difference between hypothesis and history. He remarks that the “equality of like men” is “manifested” in some cases (i.e., Socialist labor groups) “by the mutual use of the familiar ‘thou,’ which is employed by the most poorly paid wage-laborer in addressing the most distinguished intellectuals.” (p. 27) If poor laborers and intellectuals are associated, then “equality of like men” is absent. The group is not a social democracy, although it may employ equalitarian rituals and it may be pledged to the attainment of social democracy.

¹⁶ P. 33.

¹⁷ P. 401.

¹⁸ P. 36.

¹⁹ Pp. 31–2, 36, 206, 400.

²⁰ Pp. 21–2.

²¹ P. 365.

²² P. 365n.

²³ For example, see Sartori, above, note 1, pp. 121–26.

²⁴ See pp. 176, 185–87, 272, 392.

²⁵ P. 365.

²⁶ Pp. 187–8.

²⁷ Pp. 187, 189, 371, 307, 401 (his capitals), 376. Some writers have suggested that Michels's Law of Oligarchy deals with the general subject of goal-reorientation as determined by internal group processes, rather than with the particular Subject of democracy. For example, seeEldersveld,Samuel, “American Interest Groups,” inInterest Groups on Four Continents, ed.Ehrmann,H. W. (University of Pittsburgh Press,1958), p.184Google Scholar. Philip Selznick relies heavily on Michels's contributions to understanding the “unanticipated consequences” of Organization and of circumstantial adaptation. See esp.Selznick's, “An Approach to the Theory of Bureaucracy,”American Sociological Review, Vol.8 (1943), p.1CrossRefGoogle Scholar.

²⁸ SeeGouldner,A. W., “Metaphysical Pathos and the Theory of Bureaucracy,” this Review, Vol.49 (1955), p.3Google Scholar. Gouldner argues that Michels, Selznick (inTV A and the Grass Roots), and other modern theorists of group organization quite arbitrarily assume that the “unanticipated consequences” wrought by Organization will be deplorable.

²⁹ Part I, ch. 3, esp. pp. 41–3.

³⁰ P. 238.

³¹ Pp. 22, 237, 281.

³² P. 238; emphasis his.

³³ P. 11; emphasis added. Elsewhere (esp. ch. 2) Michels voices doubt that such a commitment has ever truly animated a particular social group.

³⁴ P. 40; also pp. 289–90.

³⁵ P. 295.

³⁶ Pp. 283–8.

³⁷ Pp. 271–82.

³⁸ Pp. 265–70.

³⁹ Pp. 185–87.

⁴⁰ P. 387.

⁴¹ P. 268.

⁴² P. 275f.

⁴³ Pp. 185–7.

⁴⁴ P. 278.

⁴⁵Ibid.

⁴⁶ Part IV, chs. 2, 3, 4. The downward movement involved in the “social exchange” is ideological and affiliational rather than socio-economic. Idealistic or opportunistic bourgeois join Socialist groups, usually as leaders.

⁴⁷ P. 279.

⁴⁸ P. 89.

⁴⁹ P. 1.

⁵⁰ P. 189.

⁵¹ P. 389. A very similar argument (omitting the “teleology” of “class” fidelity), is advanced byCarr,E. H. inThe New Society (1951)Google Scholar and is dissected by Langer, above, note 1, p. 263.

⁵² P. 280.

⁵³ The early Socialist platform was “the best one from which to advocate the interests of the workers … so that the renunciation of this platform almost always involves the loss of opportunity for defending working-class interests” (p. 116).

⁵⁴ P. 367.

⁵⁵ Pp. 398–9.

⁵⁶ P. 367.

⁵⁷ P. 6.

⁵⁸ P. 367.

⁵⁹ P. 374.

⁶⁰ P. 394; also pp. 367–74. Michels amended this assumption in the light of the Bolshevik and Fascist triumphs. He acknowledged the prowess of elitist-insurrectionary (non-“mass”) parties during crisis periods. He also suggested that these parties alone can maintain a moral and social integrity, since they do not need to emasculate their doctrines for the sake of pluralistic electoral support.Michels,R., “Some Reflections on the Sociological Character of Political Parties,” this Review, Vol.21 (111927), p.3Google Scholar.

⁶¹ P. 371.

⁶² P. 374.

⁶³ P. 367.

⁶⁴ P. 393–5.

⁶⁵ P. 297. The second-generation Socialist leaders, in addition to being psychologically and sociologically more representative than their predecessors, also were ‘ethnically’ more representative. The early leaders (and many followers) were bourgeois, militant, and Jewish, (pp. 258–63, 28, 324, 342)

⁶⁶ Pp. 289, 319, 171.

⁶⁷ P. 306; emphasis added.

⁶⁸ Pp. 89, 165.

⁶⁹ Pp.389, 114.

⁷⁰ P. 80.

⁷¹ Pp. 71–2.

⁷² P. 67.

⁷³ P. 301.

⁷⁴ Pp. 400, 50–2, 79, 98.

⁷⁵ P. 400.

⁷⁶ “The massper se is amorphous, and therefore needs division of labor, specialization, and guidance.” This “incompetence” is “incurable.” (p. 404). Sartori observes that what is rendered as “leadership” in the English translation ofPolitical Parties appears asFührerstum and assisterma di capi in the German and Italian editions. The latter terms allegedly connote “rulership,” or “headship” or some sort of arrangement more sinister than what is conveyed by “leadership.” Sartori, above, note 1, p. 110.

⁷⁷ Pp. 83–83, 86.

⁷⁸ P. 83n.

⁷⁹ P. 76.