¹ Aronson, Sidney H., Status and Kinship in the Higher Civil Service (Cambridge: Harvard University Press, 1964)CrossRefGoogle Scholar has demonstrated that Jackson's efforts in this direction were at best only moderately successful; Jackson had to rely on the aristocracy of landed gentry for many of his higher-level appointments.

² Representative bureaucracy for the purposes of this paper shall mean a national bureaucracy in which the bureaucrats resemble the general public in terms of social origins and general values. At no time will active representation, i.e., representation of the specific clientele interests of the agency, however aggregated, be considered a representative bureaucracy. This paper does not test whether actions of the bureaucracy are representative, whether a representative bureaucracy is a responsive bureaucracy, or whether the bureaucracy represents organized interests.

³ The definition of responsive, responsibility, accountability, and other similar terms will not be considered in this paper because of space constraints. These important definitions are discussed in Gilbert, Charles E., “The Framework of Administrative Responsibility” Journal of Politics, 21 (08, 1959), 373–407CrossRefGoogle Scholar; and Spiro, Herbert J., Responsibility in Government (New York: Van Nostrand Reinhold, 1969)Google Scholar, among other works.

⁴ Representative Bureaucracy: An Interpretation of the British Civil Service (Yellow Springs: Antioch Press, 1944)Google Scholar.

⁵ Ibid., p. 215. The idea of bureaucracy being a tool of the dominant classes of society did not die with Kingsley. Sjoberg, Gideon, Brymer, Richard A., and Farris, Buford, “Bureaucracy and the Lower Class,” Sociology and Social Research, 50 (04, 1966), 325–337Google Scholar, claim bureaucracy performs a dominating function for whites vis-á-vis Mexican-Americans and other minorities. With some distortion much of the clientele-orientated new public administration also contains this class analysis of bureaucracy. Cf. Marini, Frank, ed., Toward a New Public Administration (Scranton, Pennsylvania: Chandler, 1971)Google Scholar.

⁶ Kingsley, p. 282.

⁷ “The Responsibility of Administrative Officials in a Democratic Society.” Political Science Quarterly, 61 (12, 1946), 562–598CrossRefGoogle Scholar.

⁸ “Bureaucracy and Constitutionalism,” American Political Science Review, 46 (09, 1952), 808–818CrossRefGoogle Scholar, quotation at p. 811.

⁹ Riper, Paul P. Van, History of the United States Civil Service (White Plains, New York: Row Peterson, 1958), p. 552Google Scholar.

¹⁰ have omitted from this discussion both of the empirical discussions of representative bureaucracy because the present analysis covers and extends them. See Subramaniam, V., “Representative Bureaucracy: A Reassessment,” American Political Science Review, 61 (12, 1967), 1010–1019CrossRefGoogle Scholar; and Warner, W. Lloyd, Riper, Paul P. Van, Martin, Norman H., and Collins, Orvis F., The American Federal Executive (New Haven: Yale University Press, 1963)Google Scholar.

¹¹ Mosher, Frederick C., Democracy and the Public Service (New York: Oxford University Press, 1968)Google Scholar.

¹² Ibid., pp. 12–13.

¹³ The reasons that all group conflicts are not resolved in Congress or the upper levels of the executive are many: the ambiguity of legislative policy, the discretionary decisions which arise during the execution of policy, the occasional political abdication to experts, or simply the desire to avoid making a decision in the political arena. The reasons that conflict is forced down into the bureaucracy are not important here; only the fact that it is forced down is important.

¹⁴ Actually, few of the theorists discussed below incorporate the values of bureaucrats into their theory (which I feel is the key to control), although the addition of values strengthens the theory and could be derived from Mosher (1968) and Van Riper (1958).

¹⁵ Cf. Hitch, Charles, Decision-Making for Defense (Berkeley: University of California Press, 1965)Google Scholar; Hitch, Charles and McKean, Roland, The Economics of Defense in a Nuclear Age (Cambridge: Harvard University Press, 1960)CrossRefGoogle Scholar; Schultze, Charles L., The Politics and Economics of Public Spending (Washington: The Brookings Institution, 1968)Google Scholar.

¹⁶ Some of the available cases studies of traditional control indicate it may be effective. See Stanley, David T., Changing Administrations (Washington, D.C.: The Brookings Institution, 1965)Google Scholar; and Scher, Seymour, “Regulatory Agency Control Through Appointment.” Journal of Politics, 23 (11, 1961), 667–688CrossRefGoogle Scholar. For a contrary view see Kerr, James, “Congress and Space: Overview or Oversight,” Public Administration Review, 25 (09, 1965), 185–192CrossRefGoogle Scholar.

¹⁷ Exceptions to the general trend are not usually considered advocates of representative bureaucracy. Ostrom, Vincent, The Intellectual Crisis in American Public Administration (University, Ala.: University of Alabama Press, 1973)Google Scholar, supports the concept of active representation; Fenno, Richard, The Power of the Purse (Boston: Little, Brown, 1966)Google Scholar demonstrates how much of the political system is based on the need for every actor to advocate his position; and Redford, Emmette S., Democracy in the Administrative State (New York: Oxford University Press, 1969)Google Scholar emphasizes the need for bureaucracy to provide channels of access for all concerned groups.

¹⁸ Brim, Orville G. Jr., “Socialization Through the Life Cycle,” in Socialization After Childhood, ed. Brim, Orville G. and Wheeler, Stanton (New York: John Wiley, 1966), p. 3Google Scholar.

¹⁹ Ibid., pp. 18–19.

²⁰ Ibid., pp. 25–32.

²¹ The organization's role as a socializer has been demonstrated many times. Kaufman, Herbert, The Forest Ranger (Baltimore: Johns Hopkins Press, 1960)Google Scholar describes the socialization process of the forest service; and Baldwin, Sidney, Politics and Poverty (Chapel Hill: University of North Carolina Press, 1968)Google Scholar describes the attempts of the Farm Security Agency to socialize its field workers. The socialization process goes on in all agencies to a greater or lesser degree.

²² Barber, James Alden Jr., Social Mobility and Voting Behavior (Chicago: Rand and McNally, 1970)Google Scholar.

²³ The conclusions drawn and the diagram presented here are solely my interpretation of the results presented by Barber and may not reflect his own conclusions. The presentation, however, in my opinion, does not misrepresent the arguments presented in his excellent study.

²⁴ Warner, W. Lloyd, Riper, Paul P. Van, Martin, Norman H., and Collins, Orvis F., The American Federal Executive (New Haven: Yale University Press, 1963)Google Scholar.

²⁵ Kilpatrick, Franklin P., Cummings, Milton C., and Jennings, M. Kent, The Image of the Federal Service (Washington: The Brookings Institution, 1964)Google Scholar; M. Kent Jennings, “Study of Federal Executives,” data provided through ICPR.

²⁶ Subramaniam, “Representative Bureaucracy.” In his article Mr. Subramaniam lists the original sources of his information. Population figures were taken from the U.N. Statistical Yearbook.

²⁷ The accepted argument on this position is Centers, Richard, The Psychology of Social Classes (Princeton: Princeton University Press, 1949)Google Scholar. Objective social status is measured in terms of occupation, income, etc. Subjective social status is the social class to which the respondent feels he belongs.

²⁸ Yntema, Dwight, “Measures of Inequality in the Personal Distribution of Wealth or Income,” American Statistical Association Journal, 28 (12, 1933), 423–433CrossRefGoogle Scholar; Alker, Hayward R. and Russett, Bruce M., “Indices for Comparing Inequality,” in Merrit, Richard L. and Rokkan, Stein, eds., Comparing Nations (New Haven: Yale University Press, 1966), 349–372Google Scholar. Other discussions of the issues include Lorenz, M. O., “Methods of Measuring the Concentration of Wealth,” Quarterly Publications of the American Statistical Association, 9 (06, 1905), 209–219CrossRefGoogle Scholar; and Bowman, Mary Jean, “A Graphical Analysis of Personal Income Distribution in the United States,” American Economic Review, 35 (09, 1945), 607–628Google Scholar.

²⁹ The Gini index is calculated by the following equation:

where X by definition is the equation of the 45-degree line and f(X) is the equation of the line of inequality. Actually this method of calculating R assumes the equation of the line of inequality is known. Since occupation, the major variable, is discrete rather than continuous, R will be calculated by summing areas of triangles. This procedure while avoiding the assumption of a regular curve of inequality will slightly underestimate the Gini index because the individuals are probably not randomly distributed throughout the interval but clustered toward the top of the interval.

³⁰ The ordering for occupations (in ascending order of status) is as follows: unskilled labor, farm labor, skilled labor, clerical and sales workers, miscellaneous, farm owners, small businessmen, large businessmen, and professionals. The nonurban occupations are difficult to position and were placed somewhat arbitrarily. Again, a slight reordering does not significantly affect the size of Gini index.

³¹ That the bureaucracy as a whole is representative of the people should not be surprising because the category of “civil servants” includes all employees of the federal government encompassing not only all GS positions but wage board employees, postal carriers, etc. Since these occupations are relatively open without regard to education and perhaps do not constitute a significant upward movement on the part of the employee, it is likely that the total service will be very representative.

³² The reason for the quantum jump between the GS 11–13 classifications and those in the higher civil service (GS 14+) may be partially explained in terms of the two groups' composition. The higher civil service includes the bureau chiefs, division heads, and section chiefs, the day to day leaders of the civil service. The 11 through 13 classifications, on the other hand, contain two distinct groups of people. First, the management and professional personnel marked for promotion to higher levels, they entered the civil service at GS 5, 7, or 9 and probably reached the 11–13 group in two to eight years after entry. These personnel should be similar to the higher civil service in composition. The other major category is those civil servants who have been in the civil service for significant portions of their lives, have been passed over for promotions, and have little likelihood of ever entering the higher civil service. Since the latter group is more numerous than the former, the index of concentration is closer to the lower categories than it is to the higher civil service.

³³ The distribution of the respondents' subjective social class is as follows:

³⁴ All figures presented here are from the Fulton report and as such do not reflect the changes made as a result of the report, specifically the combination of the administrative and executive classes. Nigro, Felix A., “What Has Happened to Fulton?” Public Administration Review, 33 (03/04, 1973), 185–187CrossRefGoogle Scholar has argued, however, that the reforms did not affect the distinction between the elite grades of the former administrative class and the more common grades of the executive class. If this is true, then the present distribution would not significantly differ from the findings presented here.

³⁵ Any implications drawn from the data presented in Figure 7 must be qualified. The statistics on the civil service were taken from many diverse surveys and the population statistics were taken from the U.N. Statistical Yearbook for the appropriate year (in most cases the earliest year reported). In addition to problems related to the accuracy and comparability of these sources, occupational status is not consistent across national boundaries. Despite these complications and limitations, the indices do conform to nonempirical impressions of the elite nature of given bureaucracies and are probably accurate to the nearest tenth.

³⁶ Recent discussions of representative bureaucracy in terms of race and sex include Nachmias, David and Rosenbloom, David H., “Measuring Bureaucratic Representation and Integration,” Public Administration Review, 33 (11/12, 1973), 590–597CrossRefGoogle Scholar; McGregor, Eugene B. Jr., “Social Equity and the Public Service,” Public Administration Review, 34 (01/02, 1974), 18–28CrossRefGoogle Scholar.

³⁷ The data collected by Warner, Van Riper, Martin, and Collins (1963) would be ideal for such analysis, but all efforts to locate existing copies of the data failed. The Jennings data used in this section is limited by a sample size of less than one thousand. While thoroughly adequate for generalizations about the entire civil service, the numbers shrink precipitously when parcelled out by bureaus, a fact which accounts for which departments were reported.

³⁸ Fishbein, Martin and Coombs, Fred, “Basis for Decision: An Attitudinal Approach Toward an Understanding of Voting Behavior,” paper presented at the 67th Annual Meeting of the American Political Science Association, Chicago, 09 7–11, 1971Google Scholar, contend issue positions can be interpreted as evaluative components of attitudes and, thus, as values.

³⁹ The subsample of public employee households has the limitation of including all public employees, not just those holding leadership positions. All employees rather than just federal employees were included because the number of federal employees was too small to permit analysis. A better test, if data were available, would be to compare upper-level civil servants with both the public and a subset of the public with similar status achievement.

⁴⁰ From Capital, Vol. Ill, pp. 649 cited in Bottomore, T. B., Karl Marx: Selected Writings in Sociology and Social Philosophy (New York: McGraw-Hill, 1964), p. 190Google Scholar.