¹ This four point summary is not a précis of any particular work on the problem. The main sources are: Miliband, Ralph, The State in Capitalist Society (London: Weidenfeld and Nicolson, 1969)Google Scholar; Poulantzas, Nicos, Political Power and Social Classes (London: New Left Books, 1973)Google Scholar; Holloway, John and Picciotto, S., eds, The State and Capital: A Marxist Debate (London: Edward Arnold, 1978)Google Scholar; and Draper, Hal, Karl Marx's Theory of Revolution (New York: Monthly Review Press, 1980).Google Scholar

² Block, Fred, ‘The Ruling Class Docs Not Rule: Notes on the Marxist Theory of the State’, Socialist Revolution, VII (1977), 6–28Google Scholar. We focus on Block because his is perhaps the clearest statement of the argument for the relative autonomy of the state. The criticism we make of Block can, however, be made of most formulations of the relative autonomy of the state. A very similar argument is made against Miliband and Poulantzas in Jessop, Bob, ‘Recent Theories of the Capitalist State’, Cambridge Journal of Economics, 1 (1977), 353–73Google Scholar. Jessop tends on this argument to dismiss the use of the concept of the relative autonomy of the state as largely rhetorical. We believe, however, that the concept can be given theoretical precision, and that Poulantzas gives important leads in that direction. These leads are easily ignored because, as Jessop properly indicates, Poulantzas has a contradictory tendency to either endow the state with complete independence from the economic base or to deny it any independence at all. (Jessop, , ‘Recent Theories of the Capitalist State’, p. 358.)Google Scholar

³ Block, , ‘The Ruling Class Does Not Rule’, pp. 8–9.Google Scholar

⁴ Block, , ‘The Ruling Class Does Not Rule’, p. 20.Google Scholar

⁵ Block, , ‘The Ruling Class Does Not Rule’, pp. 25–6.Google Scholar

⁶ Block, , ‘The Ruling Class Does Not Rule’, p. 26.Google Scholar

⁷ For Poulantzas, political struggle is the ‘motive power of history having as its object the state’, and ‘the state is not an “entity” with an intrinsic instrumental essence, but it is itself a relation, more precisely the condensation of the class struggle’. See Poulantzas, , Political Power and Social Classes, p. 27Google Scholar, and Classes in Contemporary Capitalism (London: New Left Books, 1975), p. 26.Google Scholar

⁸ See Poulantzas's discussion of the uneven representation of class fractions, and the contradictory unity of the state, in Classes in Contemporary Capitalism, p. 164Google Scholar; see also Mahon, Rianne, ‘Canadian Public Policy: the Unequal Structure of Representation’, in Panitch, Leo, ed., The Canadian State: Political Economy and Political Power (Toronto: University of Toronto Press, 1977), pp. 165–98.Google Scholar

⁹ See the criticism in Miliband, Ralph, ‘Poulantzas and the Capitalist State’, New Left Review, No. 82 (1973), 88–9.Google Scholar

¹⁰ See Mouzelis, Nicos, ‘Ideology and Class Politics: a Critique of Ernesto Laclau’, New Left Review, No. 112 (1978), 45–61Google Scholar, for the necessity of analysing ideology with reference to institutions.

¹¹ See Lindblom, Charles, The Intelligence of Democracy (New York: The Free Press, 1965).Google Scholar

¹² We say genuine because there are obviously other weaker senses of autonomous state action. Incremental decisions which favour particular clients of government over other contending interests involve a sense of autonomy. But this is a trivial sense of the autonomy of the state from the point of view of a theory of the class organization of the state, or from the point of view of any conception of politics as a determinant of social and economic change, whatever the underlying theory of social and economic organization involved. A less trivial sense of the autonomy of the state may be involved in ethnic politics, where these are not tied to a discourse of class relations, and where the state divides and rules by the invocation of nationalist ideology, and by manipulating rewards and incentives to ethnic constituencies. The autonomy of the state in this area is, however, equivocal, as in the history of the national question in Canada. See our ‘Elite and Public Opinion Before the Quebec Referendum: A Commentary on the State in Canada’, Canadian Journal of Political Science, XIV (1981), 745–74Google Scholar. Our restricted conceptualization of the autonomy of the state is quite different from the attempt to theorize a much more expansive autonomy of the state in Nordlinger, Eric, The Autonomy of the Democratic State (Cambridge, Mass.: Harvard University Press, 1982)Google Scholar. Nordlinger makes much of the autonomy of the state to pursue its own preferences when these do not diverge from societal preferences; he argues for the considerable ability of the state to influence societal preferences so that they shift from opposition to convergence with state preferences; and he argues for a greater incidence of state initiatives contrary to societal preferences than is conventionally assumed in the major schools of contemporary political theory. This argument is to our mind excessively abstract, and fails to give any theoretical content to the notion of ‘preferences’. It suffers, therefore, from the typical failure of abstracted Weberian theory to distinguish socially significant interests from the more ephemeral and heterogeneous goals and preferences of individuals and groups. In these terms, autonomy is the ability to enact preferences, and there are endless possibilities for constructing any state policy as the autonomous action of some set of state officials.

¹³ Positional samples necessarily involve some assumptions about the distribution of influence for which it is difficult to provide precise evidence. For example, we chose to interview the chief executives of large firms and not board chairpersons even though in some leading firms it is likely that the chairperson plays a more important role. Short of conducting a separate study of decision making in each of the sectors of interest it would not have been easy to improve on these samples which are based, essentially, on common-sense assumptions about the relative importance of different positions in large organizations. Further, our sectoral samples do not provide an estimate for any sensible ‘universe’ of decision makers. The number of cases in each category reflects our decision, based largely on our research objectives in the overall analysis, of how to distribute scarce resources. Just over two hundred of the interviews are with state managers and an equal number are with businessmen, while only about fifty are with labour leaders. This is not intended to reflect an evaluation that the labour movement contributes only one-fourth as much as the other two groups to the universe!

¹⁴ See, for example, the review by Schuman, Howard and Johnson, M. P., ‘Attitudes and Behaviour’, Annual Review of Sociology, II (1976), 161–207.CrossRefGoogle Scholar

¹⁵ This latter finding illustrates an important difference in the mode of identifying the structural differentiation and the relative autonomy of the state in Marxist as opposed to non-Marxist political analysis. For Marxist theory the problem is identified at the level of an institutional aggregate (government) tied by a functional connection (maintaining the cohesion and unity of the social formation) to the essential workings of the mode of production (the accumulation of capital). For non-Marxist political sociology, on the other hand, the problem is defined at the level of the specific functions of the separate institutions of government (representation for parliament, policy management for the bureaucracy, etc.), the differentiation of role orientation within institutions, and the autonomy of the goal-directed action of individual vole-incumbents. That these perspectives disclose quite different research interests and findings with regard to the difference between politicians and bureaucrats is evident by comparison of our work with Aberbach, Joel D., Putnam, Robert D. and Rockman, Bert A., Bureaucrats and Politicians in Western Democracies (Cambridge, Mass.: Harvard University Press, 1981).Google Scholar

¹⁶ For other evidence of our general argument contra Block that business leaders are involved in the ideological articulation of their common class interests, rather than simply confining themselves to short-term profit-making, one has simply to look at the organization and activity of numerous business sponsored ‘think tanks’ and associations in Canada and other countries. That business confidence and expectations are, therefore, moulded by ideological discourse, and that ideological confrontation between business leaders and state managers can be intense, is evident in the public pronouncements of businessmen who elevate the state to the position, formerly reserved for the working class, of public enemy. See Vogel, David, ‘Why Businessmen Distrust Their State: The Political Consciousness of American Corporate Executives’, British Journal of Political Science, VIII (1978), 45–78CrossRefGoogle Scholar. Block's attempt to account for business politics and ideology is, in the light of this kind of evidence, strained, to say the least: ‘Ruling class members who devote substantial energy to policy formation become atypical of their class, since they are forced to look at the world from the perspective of the state managers.’ (Block, , ‘The Ruling Class Does Not Rule’, p. 13.)Google Scholar

¹⁷ For an elaboration of the following paragraph, see our ‘Elite and Public Opinion Before the Quebec Referendum’.

¹⁸ This general conclusion is close to the theoretical argument in Jessop, , ‘Recent Theories of the Capitalist State’Google Scholar. Although we emphasize the concept of ideology, which has no place in his argument, we very much approve his general outline of the analysis of the role of the state in economic policy formation. ‘The economic state apparatuses and their means of intervention are not neutral, but are integrated into the movement of capital and constitute a field of conflict between different interests. This means that state intervention has inherent limitations in securing the conditions for capital accumulation and is always subject to the inevitable influence of various class and popular-democratic struggles.’ (Jessop, , ‘Recent Theories of the Capitalist State’, p. 371.)Google Scholar