¹ Burns, James MacGregor, The Deadlock of Democracy, rev. Spectrum ed. (Englewood Cliffs, N.J., 1963), p. 244 Google Scholar.

² Burns is probably the best-known contemporary advocate of this point of view.

³ While the insiders' theory is probably the prevailing one among Washington cognoscenti, it has not been given the literary circulation of the textbook theory. Hints and scraps of it may be found in the writings of White, William S.; see particularly Citadel (New York, 1956)Google Scholar.

⁴ For an extensive treatment of this subject at the state level see Wahlke, John et al. , The Legislative System (New York, 1962)Google Scholar. See also Wilson, James Q., “Two Negro Politicians: An Interpretation,” Midwest Journal of Political Science, Vol. 4 (11, 1960), pp. 346–69CrossRefGoogle Scholar.

⁵ Since minor changes in the composition of the House occur frequently, it is necessary to choose a date for any analysis of House membership. We chose January 7, 1964 because a House roster for that date was published in the Congressional Quarterly Weekly Report, January 10, 1964.

⁶ Throughout this article the South is defined as the eleven states of the Confederacy: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia.

We have included the five border states (Kentucky, Maryland, Missouri, Oklahoma, and West Virginia) in the North. Some observers of Congress classify some or all of these states as southern, or put them in a separate category. We saw no reason to do this, since the voting record of the 28 Democratic representatives from these states is similar to that of northern congressmen, and very different from the Southerners'. In 1963, for instance, the border congressmen had a Presidential Support score of 80, compared to 59 for the Southerners and 82 for congressmen from the northeastern states. The Presidential Opposition scores of the three groups were 9, 27, and 5, respectively. (These scores are based on individual voting records for 71 House roll calls during 1963 on proposals on which President Kennedy took a position. The support score is the percentage of the 71 roll calls on which the member supported the president. Since failure to vote lowers a member's support score, we also present the opposition scores. The individual scores are in Congressional Quarterly Weekly Report, March 13, 1964.)

Congressional Quarterly defines as “urban” any central city with at least 50,000 population and any suburban city with 100,000 or more people. “Suburban” areas are those “closely settled areas contiguous to central cities,” with the exception of cities over 100,000 in population. All other areas are classified as rural. See CQ Census Analysis: Congressional Districts of the United States (Washington, Congressional Quarterly, Inc., 1964), pp. 1786, 1792 Google Scholar. We have combined CQ's “urban” and “suburban” classifications for the purpose of testing the propositions stated in the text. We have classified districts as “rural” if they are 50% or more rural according to CQ, and “urban” if they are less than 50% rural.

⁷ The absence of southern overrepresentation on the top three committees is due partly to the recent appointment of four Northerners to the Appropriations Committee. During most of the period from 1947 to 1963 every southern state had a representative on that committee. (We are indebted to Richard F. Fenno, Jr. for this information.) If one splits the Appropriations Committee's 30 Democratic members on the basis of seniority, the former southern advantage becomes apparent. Eight of the top-ranking 15 members are Southerners, compared to only one of the 15 most recent appointees.

Appointments to the Ways and Means Committee are made on the basis of regional zones, but it should not be thought that this guarantees equitable representation on the committee to all parts of the country. The zones are not redrawn after every election to take account of changes in state delegations, and, perhaps for this reason, there is considerable variation in the number of representatives assigned to each zone. The range was from 14 to 24 in the 86th Congress. See Masters, Nicholas A., “Committee Assignments in the House of Representatives,” this Review, Vol. 55 (06, 1961), p. 347 Google Scholar.

⁸ There are 20 standing committees. Members generally consider three of them unimportant: the District of Columbia, House Administration, and Un-American Activities Committees. The first of these—important to permanent residents of the District—is in effect Washington's city council, the second is occupied with housekeeping, and the third reports virtually no legislation. Membership on one of these three committees does not preclude assignment to another standing committee.

⁹ This practice is not required by the rules, but is usually followed.

¹⁰ Two seats, one each in the North and South, were omitted because they were won by Republicans at the special election following the death of the incumbents.

¹¹ See, e.g., Burns, pp. 242–4. Lewis A. Froman, Jr., discusses several other propositions about relationships between competition and party loyalty at both state and national levels. See his Congressmen and their Constituencies (Chicago, Rand McNally, 1963), ch. 9Google Scholar; and the works cited there.

¹² Of course, failure to vote also lowers the Presidential Opposition score, but the rate of opposition in all northern cells is so low that 100% voting participation would not change the opposition scores significantly.

¹³ After examining voting records for the 87th Congress, Froman came to the same conclusion (ibid., p. 114). Other recent findings indicate that when region is controlled, length of service is not strongly related to the support that congressmen of either party give to their president. See Mitchell, Judson and Spink, George, “Presidential Support and Length of Service in the House of Representatives” (unpublished paper, Stanford University)Google Scholar.

¹⁴ See Key, V. O., Southern Politics (New York, 1949), ch. 18 and passim Google Scholar.

¹⁵ The figure for the Southerners would be 13% if we took account of George Mahon's accession to the chairmanship of the Appropriations Committee on the death of Clarence Cannon in May, 1964.

¹⁶ It might be thought that the larger number of by-passed Northerners results from the fact that there are more of them in the House. But Southerners comprise 44% of the membership of the highest-ranking half of the 17 major committees, and it is from this group that subcommittee chairmen are chosen.

¹⁷ For a full description of this study see Reiss, Albert J. Jr., Occupations and Social Status (New York, 1961.Google Scholar)

¹⁸ Ibid., pp. 6, 19, 200, 220 Data on white Southerners were obtained from the National Opinion Research Center through the Inter-University, Consortium for Political Research. We are grateful to Ralph L. Bisco of the Consortium and Patrick Bova of NORC for their kind help.

¹⁹ In addition to nominations to the service academies (which usually are heavily dependent on objective examination scores), the average congressman is lucky if he can name a Capitol policeman, elevator operator, or page. It takes a number of terms to acquire enough seniority to have much more appointment power.

²⁰ Snowiss, Leo M., “Chicago and Congress: A Study of Metropolitan Representation” (unpublished doctoral dissertation, University of Chicago, 1965), ch. 2Google Scholar.

²¹ Rovere, Richard H., “The Big Hello,” in The American Establishment, Harvest edition (New York, 1962), pp. 45–6Google Scholar.

²² The New York Times, March 16, 1964, p. 37.

²³ Congressional Quarterly Weekly Report, January 31, 1964.

²⁴ The classic description of the unstructured politics of most southern states is Key's book. See also Turner, Julius, “Primary Elections as the Alternative to Party Competition in ‘Safe’ Districts,” Journal of Politics, Vol. 15 (05, 1953), pp. 197–210 CrossRefGoogle Scholar.

²⁵ Jones, Charles O., “Inter-party Competition for Congressional Seats,” Western Political Quarterly, Vol. 17 (09, 1964), pp. 461–76CrossRefGoogle Scholar.

²⁶ There probably would be fewer southern non-competitive seats (not to mention fewer Democratic congressmen) if there had been Republican candidates in every southern district in the 1964 election. In several southern states Republican candidates either won or held their opponents to less than 65% of the vote in every district that they contested.

²⁷ Burns, p. 242.

²⁸ The wholesale reapportionment resulting from recent Supreme Court decisions is also likely to increase urban and suburban representation in both North and South.

²⁹ This is an example of how a rational strategy for Republican parties at the state level—conceding a minimum number of seats to the Democrats—is harmful to the party's interests on the national level because it enables these Democrats to build up seniority in Congress.

³⁰ This includes Representative Albert W. Watson of South Carolina, a Democrat who backed Senator Goldwater in 1964, was stripped of his seniority for this defection by the Democratic Caucus, and then switched parties.

³¹ Representatives Grant and Roberts of Alabama and Winstead of Mississippi, all defeated by Republican candidates, headed a total of four subcommittees. A fifth subcommittee was chaired by John Bell Williams of Mississippi, who shared Watson's fate (but did not switch parties). Northerners took over Williams' subcommittee and one of the two chaired by Grant.

³² Procedures were changed to reduce the Rules Committee's power to prevent floor consideration of legislation and avoid appointing conference committees. Most Southerners resisted these actions.

³³ There are, of course, many Southerners, including powerful committee chairmen like Carl Vinson, who are (or were) loyal to Democratic presidents all or most of the time. One cannot validly label all southern congressmen as invariably dissident. Nevertheless, there is also no doubt that the southern wing of the party is the major source of deviance.

³⁴ Some political scientists have argued that large legislative majorities are less responsive to the president, because they lead to factionalism. It is unclear whether this proposition refers simply to a majority for the president's party, or an effective majority for his policies. Even if it refers to the latter, there are some difficulties. Most important, it would seem that the smaller the majority, the easier it is for a potential defector to impede the president's program, since he needs to win fewer other dissidents to destroy the majority. As the majority grows, so does the size of the splinter group necessary to make an impact, and hence so does the magnitude of the defector's task. For discussions of this “law of economy” see Schattschneider, E. E., Party Government (New York, 1942), pp. 85–96 Google Scholar. William H. Riker, who is cited on this point, actually discusses the weakness of big majorities only when they are so big that they become “a coalition of the whole.” See his A Theory of Political Coalitions (New Haven, Yale University Press, 1962), p. 56 Google Scholar.

Statements of this proposition are usually accompanied by reminders that President Roosevelt, having won enormous congressional majorities in 1936, found his legislative program bogged down soon thereafter. This appears to be the major item of evidence to support the proposition. It does not appear that presidents have restrained their efforts to elect friendly congresses for fear of suffering Roosevelt's fate. A systematic examination of the historical record found that, if anything, big majorities are better than little ones. See Goodman, Jay, “Legislative Majorities and Presidential Success” (unpublished paper, Brown University Google Scholar.)