¹ See, for example, Ripley, Randall B., Party Leaders in The House of Representatives (Washington: Brookings Institution, 1967)Google Scholar; Jones, Charles O., The Minority Party in Congress (Boston: Little, Brown, 1970)Google Scholar; Ripley, Randall B., Majority Party Leadership in Congress (Boston: Little, Brown, 1969)Google Scholar; and Froman, Lewis A. Jr., and Ripley, Randall B., “Conditions for Party Leadership: The Case of the House Democrats,” American Political Science Review, 59 (March, 1965), 52–63CrossRefGoogle Scholar.

² Gawthrop is one of the few scholars to point out this phenomenon. However, his primary concern was to describe the changing patterns of multiple committee assignments in the House. See Gawthrop, Louis C., “Changing Membership Patterns in House Committees,” American Political Science Review, 60 (June, 1966), 366–373CrossRefGoogle Scholar.

³ For the purposes of this paper, “majority party leaders” will mean the Speaker, the Majority Leader, and the Majority Whip.

⁴ Their decisions are voted on by the Democratic Caucus. This final approval, however, is most often a pro forma action. See Rohde, David W. and Shepsle, Kenneth A., “The Committee Assignment Process: A Case Study of Social Choice” (paper presented at the 67th annual meeting of the American Political Science Association, Chicago, 1971)Google Scholar.

⁵ Masters, Nicholas A., “Committee Assignments in the House of Representatives,” American Political Science Review, 55 (June, 1961), 345–357CrossRefGoogle Scholar; Ripley, , Party Leaders in the House of Representatives, pp. 57–61Google Scholar; Clapp, Charles L., The Congressman: His Work As He Sees It (Washington: The Brookings Institution, 1963), pp. 183–212Google Scholar; and Manley, John F., The Politics of Finance: The House Committee on Ways and Means (Boston: Little, Brown, 1970), p. 24Google Scholar.

⁶ On the temporal sequence of the assignment process, see Shepsle, Kenneth A., “A Model of the Congressional Committee Assignment Process: Constrained Maximization in an Institutional Setting” (paper presented at the 69th annual meeting of the American Political Science Association, New Orleans, 1973)Google Scholar.

⁷ Democratic vacancies on Ways and Means are filled by the Democratic Caucus. Ways and Means has been included as a committee to be observed because we assume the pre-eminent position of the leaders also applies to this committee.

⁸ The intent is not to raise the issue of the “realism” of the assumptions. Two classic statements by Friedman and Koopmans take opposite positions on this issue. See Friedman, Milton, Essays in Positive Economics (Chicago: University of Chicago Press, 1953), pp. 3–43Google Scholar; and Koopmans, Tjalling C., Three Essays on the State of Economic Science (New York: Mcgraw-Hill, 1957), pp. 129–166Google Scholar.

⁹ Masters, p. 346. On scarcity of positions also see Fenno, Richard F. Jr., The Power of the Purse: Appropriations Politics in Congress (Boston: Little, Brown, 1966), Table 2.8 at p. 64Google Scholar; and Bullock, Charles S. III,“Committee Transfers in the United States House of Representatives,” Journal of Politics, 35 (February, 1973), footnote 39 at p. 101CrossRefGoogle Scholar.

¹⁰ Masters, p. 357.

¹¹ Ripley, , Party Leaders in the House of Representatives, p. 60Google Scholar. Also, see Manley, , The Politics of Finance, p. 24Google Scholar.

¹² All the data come from the available Congressional Directory, closest to the opening of each Congress, which lists committee assignments. In some cases it was necessary to use the Congressional Directory for the second session of a Congress. Since the committee system was reorganized in 1946, beginning with 1927 also provides two distinct time periods of approximately the same length. A list of the standing committees and their respective sizes may be found in Cannon's Precedents of the House of Representatives, 3 (Washington: Government Printing Office, 1935), 8–9Google Scholar. The Congresses listed there include the 60th-73rd Congresses. The Committee on Memorials (3 seats) was created in the 70th Congress. The Committee on Un-American Activities has been excluded from the period preceding the Legislative Reorganization Act of 1946. Prior to 1927, the number of positions grew even in the midst of periodic adjustments in the committee system.

¹³ It is worth noting that alteration of committee size occurred prior to the time period examined in this paper. “In making committee assignments for Republicans in the Fifty-ninth Congress, Cannon was faced with a problem of too few choice seats for too many Republicans. He finally reached an agreement with the Democratic Minority Leader, John Sharpe Williams of Mississippi, that would increase the size of the committees and leave the Democratic representation undisturbed” (Ripley, , Majority Party Leadership in Congress, p. 28Google Scholar).

¹⁴ One potential source of measurement error can be disregarded. The sizes of the committees may be altered during a Congress. However, the committee rosters for the second session of each Congress indicate that generally the size of committees is not altered within a Congress. And where an increase or decrease does occur, the committee is not changed to its former size at the beginning of the next Congress.

¹⁵ There is an interesting interpretation to be made from the data of Figure 3. Note the rate of change in the interval 1931–1933. A party, long out of power and suddenly gaining majority status with an overwhelming numerical advantage may have strong incentives to expand rapidly the number of rewards early in the game. Were the Republicans suddenly to achieve majority status by a large margin the same phenomenon may occur. Note, however, that this interpretation does not fall out of the argument. The argument presupposes only a weak condition (an increasing function of time) on the time path of the total number of committee positions.

¹⁶ Bullock, Charles and Sprague, John, “A Research Note on the Committee Assignments of Southern Democratic Congressmen,” Journal of Politics, 31 (May, 1969), 493CrossRefGoogle Scholar. For one example of the use of committee transfers, see Goodwin, George Jr.,, The Little Legislatures: Committees of Congress (Amherst: University of Massachusetts Press, 1970), pp. 114–115Google Scholar. And the positions on desirable committees are scarce. See footnote 9 above.

¹⁷ Bullock and Sprague, pp. 496–501.

¹⁸ Bullock, Charles, “The Committee Assignments of Freshmen in the House of Representatives, 1947–1967” (dissertation, Washington University, St. Louis, 1968), p. 35Google Scholar.

¹⁹ As we have noted, the Democrats controlled most of the Congresses during this time period. Hence, for our purposes, the majority party empirically will mean the Democrats. Furthermore, the prestige rankings of committees for Democrats and Republicans are empirically distinct. See Bullock and Sprague, pp. 496–501. Although theoretical reasons dictate the use of majority party rankings, use of rankings for Republicans or House members as a whole, both distinct from Democrats alone, will not dramatically alter the results reported below. This occurs because there is a correspondence between the rankings for the more desirable committees, the ones I shall argue take up most of the overall expansion in seats.

²⁰ For the modern period see Goodwin, pp. 114–116. For the earlier period see Galloway, George B., The Legislative Process in Congress (New York: Thomas Y. Crowell, 1955), p. 128Google Scholar; Riddick, Floyd M., Congressional Procedure (Boston: Chapman and Grimes, 1941), pp. 115–116Google Scholar; Chiu, Chang-Wei, The Speaker of the House of Representatives Since 1896 (New York: Columbia University Press, 1928), p. 69Google Scholar; and Galloway, George B., Congress at the Crossroads (New York: Thomas Y. Crowell, 1946), p. 90Google Scholar.

²¹ See Bullock and Sprague, pp. 496–501.

²² It is impossible to reconstruct the particular circumstances of the creation of each new seat. For example, a new minority position might be created to reflect a new party ratio and to prevent a reduction in majority positions. Such an action is consistent with the argument.

²³ One might suppose there is interdependence here. That is, adding positions to a committee may increase the number of transfers to that committee and hence affect the prestige measure. But prestige is also influenced by departures from a committee. Therefore, in principle it is possible to observe low prestige committees increasing in size.

²⁴ The crucial notion of this argument is the relative time rate of change of supply and demand. We might consider the demand for positions to be fixed and constant over time. Alternatively, we may assume that interests proliferate as social, political, and economic systems become more complex thereby increasing the demand for positions. Or demand for positions may decline through time. As long as the time rate of change of supply is greater than the time rate of change of demand, the basis of exchange between the party leaders and followers is impaired.

²⁵ Clearly this proposition, the long-run consequence of a leadership management strategy, cannot be evaluated empirically in the same manner as the earlier portions of the argument. It takes time for the process to work itself out. Nonetheless, this argument can easily be formalized. Specific rate of change assumptions for D_t (demand) and S_t (supply) would make it possible to estimate at what point the committee system would be under stress, i.e., at what point conditions favorable to reorganization are likely to occur.

²⁶ The leaders may also increase the size of a committee in order to reshape its composition for policy reasons. A good example would be the expansion of the Rules Committee. See Cummings, Milton C. and Peabody, Robert L., “The Decision to Enlarge the Committee on Rules: An Analysis of the 1961 Vote,” in New Perspectives on the House of Representatives, ed. Peabody, Robert L. and Polsby, Nelson W. (Chicago: Rand McNally, 1963), pp. 167–194Google Scholar. In such cases the leaders' desire for compliance is directed at a work group rather than a specific individual. Although this theory consists of assertions at the level of individuals, leadership decisions directed at work groups are not at odds with the argument. As long as scarcity of positions is important in the relations between leaders and followers the argument holds.

²⁷ See Polsby, Nelson W., “The Institutionalization of the U.S. House of Representatives,” American Political Science Review, 62 (March, 1968), 158–160CrossRefGoogle Scholar.

²⁸ As we have seen new positions on some committees are not matched by a reduction in size of other committees. In the period examined prior to 1946 almost half the members held multiple assignments. And the number of multiple memberships has steadily increased since 1946. See Gawthrop, , “Changing Membership Patterns in House Committees,” pp. 366–367Google Scholar.

²⁹ There are other data which suggest the process discussed in this paper may be a general leadership management strategy, i.e., not confined solely to the House of Representatives. From 1927–1945 the time path of the total number of Senate committee positions is an increasing function of time. And this growth took place in the midst of an abundance of positions; the ratio of the number of seats to the number of senators was more than four to one. After the Reorganization Act of 1946, the number of positions once more steadily increased. After 1969 the time path begins to oscillate slightly. But some departure from the House pattern is not surprising. Inflation of seats has been much more pronounced in the Senate; even after 1946 the ratio of seats to members exceeded two to one.

³⁰ See Polsby, pp. 148–168; Polsby, Nelson W., Gallaher, Miriam, and Rundquist, Barry Spencer, “The Growth of the Seniority System in the U.S. House of Representatives,” American Political Science Review, 63 (September, 1969), 787–807CrossRefGoogle Scholar; and Abram, Michael and Cooper, Joseph, “The Rise of Seniority in the House of Representatives,” Polity, 1 (Fall, 1968), 52–85CrossRefGoogle Scholar.