¹ See MacRae, Duncan Jr., “Partisanship and Issues in Congressional Voting,” Paper delivered at the 1968 Annual Meeting of the American Political Science Association, September 2–7, Washington, D.C.Google Scholar In relation to the differentation of the impact of constituency and partisan factors see Froman, Lewis A. Jr., Congressmen and Their Constituencies (Chicago: Rand McNally and Company, 1963)Google Scholar.

² Senate-House comparisons of roll call voting have been made utilizing other methods, e.g., Truman, David B., The Congressional Party (New York: John Wiley and Sons, Inc., 1959)Google Scholar; Lewis A. Froman, Jr., op. cit., pp. 69–84; Westerfield, H. Bradford, Foreign Policy and Party Politics (New Haven: Yale University Press, 1955)Google Scholar.

³ Clausen, Aage R., “Measurement Identity in the Longitudinal Analysis of Legislative Voting,” this Review, 61 (December. 1967). 1020–1035Google Scholar.

⁴ MacRae, Duncan Jr. and Schwartz, Susan Borker, “Identifying Congressional Issues by Multidimensional Models,” in Midwest Journal of Political Science, 12 (May, 1968), 181–201CrossRefGoogle Scholar.

⁵ Clausen, op. cit.

⁶ The .70 correlation criterion is approximately equivalent to the standard .90 level of reproducibility used in Guttman scaling. However, it should be noted that the requirement of a minimum of .60 for individual roll call pairings has, in practice, yielded clusters dominated by inter-item correlations of .80 and above. The setting of a minimum correlation of .60 for individual correlations guards against the possibility of an item with an extreme division being included in a cluster despite wildly fluctuating correlations that average out to .70 or better. The exclusion of such items is justified on the grounds that highly variable correlations are poor indications of a common underlying dimension. Actually, this danger had already been minimized by excluding items more extreme in their division than 90–10. For a discussion of the use of Yule’s Q in roll call analysis, see MacRae, Duncan Jr., “A Method for Identifying Issues and Factions from Legislative Votes,” this Review, 62 (March, 1968), 909–926Google Scholar.

⁷ The scoring was performed using a program constructed by Keith Billingsley.

⁸ There are two aspects of this scoring procedure which require explanation in terms of the assumptions made: (1) the scoring of absences, and (2) the simple summation scoring. One alternative to our scoring of absences would be that of excluding absences in the scoring. This assumes that the roll calls on which no absences are observed, with respect to each legislator, are a representative sample of the set of roll calls in the scale. We prefer not to make this assumption, treating the absence instead as partially a function of indifference or abstention. In each case, a middle scoring for the legislator is a reasonable option although more elaborate procedures might be used such as scoring absences according to a mean computed for the entire body on the given roll call.

The second feature of the scoring procedure, the simple summation of values across roll calls, recognizes that individuals occupying the same position on a policy dimension may have somewhat different voting patterns. They are assumed to be alike only in the probabilities of voting Yea and Nay, probabilities which vary between 0 and 1 probabilities assumed by the Guttman scale model. Only across a number of roll calls will the order positions on the dimension become clear as individual idiosyncracies are absorbed into the more general pattern of voting.

Actually, the scoring technique does not produce results which deviate sharply from those obtained by the Guttman scoring method. The scoring procedure does relax the deterministic response criteria of the Guttman model. On the other hand, the requirement of high Yule Q correlations among roll calls places constraints on the variations in the voting patterns of individuals assigned the same score. For a discussion of probability models and the simple summation scoring of monotone items, see Torgerson, Warren S., Theory and Methods of Scaling (New York: John Wiley and Sons Inc., 1958), pp. 360–402Google Scholar.

⁹ Congress and the Nation: 1946–1964 (Washington, D.C.: Congressional Quarterly Service, 1965), pp. 1398–1399Google Scholar.

¹⁰ Ibid., pp. 1205, 1403–1404.

¹¹ The correlations reported here and in the remainder of the paper are Pearaon r's.

¹² Urbanization measure on Congressional districts is percentage urban minus percentage rural, based on data provided by “Congressional Quarterly Weekly Report, No. 5” (February 2, 1962), pp. 156–159Google Scholar. Urban areas contain a central city of 50,000 or larger, including suburban areas containing a city of 100,000 or larger; suburban areas consist of closely settled areas contiguous to central cities; rural areas contain cities smaller than 50,000. The urbanization index for states is a weighted composite of the district indices.

¹³ Percentage blue collar measure on Congressional districts is based on data provided by “Congressional Quarterly Weekly Report, No. 29” (July 20, 1956), p. 853. The blue collar definition includes craftsmen, foremen, machine operators, private household help, service employees, and all laborers except those who work on farms. The state measure of blue collar is a weighted composite of the district indicesGoogle Scholar.

¹⁴ This proposition is implied at different points in the literature but we have found no author who states it clearly enough to warrant citation.

¹⁵ Coastal: Alaska, Hawaii, Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont, Delaware, New Jersey, New York, Pennsylvania, California, Oregon, Washington; Interior: Illinois, Indiana, Michigan, Ohio, Wisconsin, Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, South Dakota, Kentucky, Maryland, Oklahoma, Tennessee, West Virginia, Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, Wyoming; South: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Texas, Virginia.

¹⁶ Mean scores computed on the individual Congresses in each set are sufficiently homogeneous to warrant the aggregation for each set.

¹⁷ Kesselman, Mark, “Presidential Leadership in Congress on Foreign Policy: A Replication of a Hypothesis,” Midwest Journal oj Political Science, 9 (November, 1965), 401–406CrossRefGoogle Scholar.

¹⁸ Previous studies concerned with the impact of personnel turnover include Lewis A. Froman, Jr., op. cit., especially Chapter 8; and Anderson, Lee, “Individuality in Congress: A Research Note,” Midwest Journal of Political Science, 8 (November, 1964), 425–429.Google Scholar Using a measure based on votes on reciprocal trade legislation, Froman concludes that districts represented by the same individual through several Congresses exhibit more stability than do districts characterized by a substantial turnover in personnel. He does not address the question of what occurs when the partisanship of the districts’ representation changes. Anderson tests and confirms Froman’s hypothesis, using both foreign involvement and domestic liberalism-conservatism scales. Also, see Stone, Clarence N., “Issue Cleavage Between Democrats and Republicans in the U.S. House of Representatives,” Journal of Public Law, 14 (1965), 343–358Google Scholar, for a discussion of the impact of change in the partisan representation of a district. Stone suggests that the impact varies substantially among different policy areas.

¹⁹ With one exception, a Democratic Senator replacing a Republican Senator in Ohio, partisan turnover produces a change in the expected direction: a Democratic legislator more liberal than a Republican.

²⁰ Clausen, op. cit., pp. 1031–1033.

²¹ Miller, Warren E. and Stokes, Donald E., “Constituency Influence in Congress,” this Review, 57 (March, 1963), pp. 45–46Google Scholar.

²² Schubert, Glendon, “The 1960–1961 Term of the Supreme Court: A Psychological Analysis,” this Review, 26 (March, 1962), 90–107Google Scholar; Spaeth, Harold J., “Warren Court Attitudes Toward Business: The ‘B’ Scale” in Judicial Decision Making, Schubert, Glendon (ed.), (London: The Free Press of Glencoe, 1963), pp. 73–110Google Scholar.