¹ Hacker, Andrew, Congressional Districting (Washington: The Brookings Institution, 1963), p. 45Google Scholar; the most thorough examination of the nationwide relationship between the popular vote and the partisan seat division is found in March, James G., “Party Legislative Representation as a Function of Election Results,” Public Opinion Quarterly, 21 (1957–1958), 521–542CrossRefGoogle Scholar.

² For the measure of election results for Table 1, and in the subsequent analysis of this study, slight adjustments are made for uncontested seats and atlarge seats. These adjustments are discussed in the Appendix. Throughout this study, the “North” is defined so as to exclude from analysis those districts in the former Confederate States and in the Border States of Kentucky, Maryland, Missouri, Oklahoma, and West Virginia.

³ Given the close fit of the regression line to the 1952–1964 observations (r = .97, σ_y/x = 2.6 per cent), departures from the regression line as great as those for 1966 and 1968 should occur less than one time in 100.

⁴ This result corresponds with Hacker's earlier finding that northern districts categorized as “safe Democratic” tended to have slightly smaller populations than “safe Republican” districts. Hacker also found districts with “changing” partisan loyalties to be the most overpopulated. See Hacker, Andrew, Congressional Districting, p. 79Google Scholar. The explanation for the absence of a Republican bias in congressional malapportionment is the familiar thesis that overrepresentation of rural Republican areas was offset by an underrepresentation of the suburbs rather than by underrepresentation of the central cities.

⁵ For a discussion of the tendency for Democrats to vote less frequently than Republicans do, see Converse, Philip E., “The Concept of a Normal Vote,” in Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., Elections and the Political Order (New York: John Wiley and Sons, 1966), pp. 9–39Google Scholar.

⁶ The mean district vote division is what the vote division would be if every vote is assigned a weight in proportion to the “value” of the vote in the particular district, with the “value” of the vote determined by the ratio of the average number of voters per district to the number of voters in the particular district. For example, in a district with half the average number of voters, the “value” of the vote would be two, and each vote would have twice the average weight in the determination of the mean district vote division. For an analysis of the variability of the “value” of the vote in state legislative contests (but based on population rather than turnout), see David, Paul T. and Eisenberg, Ralph, Devaluation of the Urban and Suburban Vote (Charlottesville: Bureau of Public Administration, University of Virginia, 1962)Google Scholar.

⁷ Although the manner in which the two-party vote is distributed among congressional districts has received little scholarly attention, a discussion of how the distribution of the vote affects British elections can be found in Butler, D. E., The Electoral System in Britain Since 1918, 2nd ed. (London: Oxford University Press, 1963), pp. 198–199Google Scholar. Butler concluded that the British Labour Party has been electorally disadvantaged by the concentration of Labour voters in safe constituencies. For a rigorous mathematical presentation of how the distribution of the vote affects a legislature's party composition, see Kendall, M. G. and Stuart, A., “The Law of Cubic Proportions in Election Results,” British Journal of Sociology, 1 (09 1950), 183–197CrossRefGoogle Scholar.

⁸ Elsewhere I have argued that during the 'fifties, incumbency status was only worth about an added two per cent of the two-party vote to the congressional candidate. See Erikson, Robert S., “The Advantage of Incumbency in Congressional Elections,” Polity, 3 (Spring, 1971), 395–405CrossRefGoogle Scholar.

⁹ Burnham, Walter Dean, Critical Elections and the Mainsprings of American Politics (New York: W. W. Norton and Company, 1970)Google Scholar.

¹⁰ Because redistricting plans are often suspected of providing safe seats for incumbents, one might also suspect that the extensive shuffling of district lines in the late ‘sixties was another source of insulation to incumbents, in addition to the electorate's increased propensity to vote for incumbents. But this did not appear to happen, as the gains of each party's veteran incumbents were about the same amount whether their districts’ boundaries shifted or not.

¹¹ The necessary information for categorizing the partisan nature of redistricting plans was obtained from Congressional Quarterly sources. Excluded are instances of “redistricting” involving the addition of one at-large seat, plans drawn by Federal Courts or nonpartisan legislatures, and the post-1962 “redistricting” in Rhode Island which amounted to the transfer of one small township when it was moved from one county to another.

¹² “Eastern States Roundup,” Congressional Quarterly (10 4, 1968), 147–148Google Scholar.

¹³ The four instances are Pennsylvania, 1960–1962 (Rep. to Bipartisan); California, 1966–1968 (Dem. to Bipartisan); New York, 1966–1968 (Rep. to Bipartisan); and New York, 1968–1970 (Bipartisan to Rep.). Not included are instances of partisan change in the districting plan in which the initial plan was over ten years old at the time of change.

¹⁴ The four states with bipartisan districting plans enacted for the 1962 election are Illinois, Massachusetts, New Jersey, and Pennsylvania. The eight states with bipartisan plans at the time of the 1968 election are California, Connecticut, Illinois, Massachusetts, New York, Oregon, Pennsylvania, Wisconsin.