¹ This study is a revision of a paper delivered at the APSA annual meeting of 1963. I was assisted in preparation by a grant from the Ford Foundation to Oberlin College for the study of public affairs. The assistance of Robert Murphy and Virginia Woodcock is also acknowledged with deep appreciation.

² Jewell, Malcolm E., “Party Voting in American State Legislatures,” this Review, Vol. 49 (Sept., 1955), pp. 773–791Google Scholar, and studies cited by Jewell, , The State Legislature (New York, 1962)Google Scholar, ch. 3.

³ A good discussion is Truman's, David, The Congressional Party (New York, 1959), pp. 10–14Google Scholar.

⁴ Lockard, W. Duane, “Legislative Politics in Connecticut,” this Review, Vol. 48 (March 1954), pp. 166–173Google Scholar, especially p. 167.

⁵ Ibid., p. 171.

⁶ Ibid., pp. 172–173.

⁷ Committee on Political Parties, APSA (Schattschneider, E. E., Chm.), “Toward a More Responsible Two-Party System,” this Review, Vol. 44 (September, 1950)Google Scholar, Supplement.

⁸ Bailey, Stephen K., The Condition of Our National Political Parties (Santa Barbara, Calif.: The Fund for the Republic, 1959)Google Scholar.

⁹ Data on Ohio may be found in my study, “The Outline of Ohio Politics,” Western Political Quarterly, Vol. 13 (September, 1960), pp. 702–721CrossRefGoogle Scholar.

¹⁰ Some supporting evidence for this observation may be found in my case study, Governor Freeman and the Minnesota Budget, ICP Study 60 (University, Alabama: University of Alabama Press, 1961)Google Scholar.

¹¹ Wahlke, et al. , The Legislative System (New York, 1962), p. 349Google Scholar.

¹² Dexter, Lewis, “The Representative and His District,” Human Organization, Vol. 16 (Spring, 1957), pp. 2–13CrossRefGoogle Scholar.

¹³ Eulau, Particularly et al. , “The Role of the Representative: Some Empirical Observations on the Theory of Edmund Burke,” this Review, Vol. 53 (September, 1959), pp. 742–756Google Scholar, and Jones, Charles O., “Representation in Congress: The Case of the House Agriculture Committee,” this Review, Vol. 55 (June, 1961), pp. 358–368Google Scholar.

¹⁴ Derge, David R., “Metropolitan and Outstate Alignments in Illinois and Missouri Legislative Delegations,” this Review, Vol. 52 (December, 1958), pp. 1051–1065Google Scholar.

¹⁵ Havens, Murray Clark, City Versus Farm? (University, Alabama: Bureau of Public Administration, University of Alabama, 1957)Google Scholar.

¹⁶ Flinn, “The Outline of Ohio Politics,” op. cit.

¹⁷ Rice, Stuart, “The Behavior of Legislative Groups: A Method of Measurement,” Political Science Quarterly, Vol. 40 (March, 1925), pp. 6072CrossRefGoogle Scholar.

¹⁸ This is in effect a time sample, and it is taken to be a random sample of party opposition votes in the 1959 session. Findings are spoken of as applying to the entire session, and probabilities are given. Another time sample taken from the 1963 session is treated in the same way. If the data analyzed do not constitute random samples, then the probabilities add nothing, of course; and certain statements must be taken to apply only to the data analyzed and not to the entire legislative sessions from which they were drawn.

¹⁹ But see Froman, Lewis A. Jr., “Inter-Party Constituency Differences and Congressional Voting Behavior,” this Review, Vol. 57 (March, 1963), pp. 57–62Google Scholar. As I understand him, he seems to argue that intra-party differences relate to differences in constituency and that Democrats and Republicans tend to represent different kinds of constituencies. Therefore, differences between the parties are due to constituency influences. This is not necessarily true since intra-party differences may have little or nothing to do with inter-party differences, as Table II in this study shows. Furthermore, Froman's phi coefficients which he uses to measure the relations of constituency to intra-party differences are so weak as to suggest negative rather than positive findings.

²⁰ MacRae, Duncan Jr., “The Relations Between Roll Call Votes and Constituencies in the Massachusetts House of Representatives,” this Review, Vol. 46 (December, 1952), pp. 1046–1055Google Scholar, reprinted in Eulau, , Eldersveld, , and Janowitz, (eds.), Political Behavior (Glencoe, Ill., 1956), pp. 317–324Google Scholar.

²¹ Dye, Thomas R., “A Comparison of Constituency Influences in the Upper and Lower Chambers of a State Legislature,” Western Political Quarterly, Vol. 14 (June, 1961), pp. 473–481CrossRefGoogle Scholar.

²² MacRae, op. cit., p. 323 in Eulau et al., Political Behavior. In order to understand clearly what MacRae is reporting, it should be recognized that his use of the term “typical” is somewhat unusual although convenient for lack of anything better. “Typical” may refer to central tendency, but here it means, in a sense, representative of the extreme. The finding actually summarized by reference to typical and atypical districts is that members who come from districts which have most of that characteristic which distinguishes districts represented by their party from districts represented by the opposing party are more loyal to their own party than are members who come from districts which have least of the characteristic which distinguishes districts represented by their party. Hence, use of the term “typical” in this context may be unconventional; but it is certainly convenient.

²³ Pesonen in a recent study [Pesonen, Pertti, “Close and Safe State Elections in Massachusetts,” Midwest Journal of Political Science, Vol. 7 (February, 1963), pp. 54–70]CrossRefGoogle Scholar replicates MacRae's study using a more recent session of the Bame legislature. He uses an improved measure of party loyalty, and reaches the same conclusion as MacRae concerning the relation of constituency and party voting. However, he works with only seven roll calls; and does not measure the strength of the observed relation.

²⁴ Dye, op. cit., pp. 476–477.

²⁵ Ibid., p. 477.

²⁶ Ibid., pp. 479–480.

²⁷ Members were also scored on 60 party agreement votes. (10 of the original 183 roll calls were dropped since one party or the other was evenly divided making it impossible to characterize the vote as a party agreement or a party opposition vote.) The scores of Democratic and Republican members on party opposition votes were each correlated with their scores on party agreement votes. The product moment coefficients were less than .1 in each case, indicating that a party opposition vote and a party agreement vote are distinctly different events in the life of the legislature.

²⁸ See Mood, A. M., Introduction to the Theory of Statistics (New York, 1950), pp. 318–326Google Scholar. The procedure is designed to permit comparison of variance around the over-all mean with the sum of the variance around class means. It is specially useful in the present situation since it allows flexibility in the creation of classes.

²⁹ Fifty party opposition roll calls were selected. They are every such vote which occurred between May 15 and the July 11 recess. Members who voted on less than 35 of the 50 selected votes were dropped: 3 Republicans and 3 Democrats. Others were given scores expressing the percentage of their votes cast in support of positions taken by a majority of Democrats.

³⁰ MacRae, op. cit., reprinted in Eulau et al., pp. 321–322.

³¹ Pesonen, op. cit., p. 62.

³² Dye, op. cit., p. 477.

³³ Patterson, Samuel C., “The Role of the Deviant in the State Legislative System: The Wisconsin Assembly,” The Western Political Quarterly, Vol. 14 (June, 1961), pp. 460–472, 467CrossRefGoogle Scholar.

³⁴ Becker, R. W. et al. , “Correlates of Legislative Voting: Michigan House of Representatives, 1954–1961,” Midwest Journal of Political Science, Vol. 6 (Nov., 1962), pp. 384–396CrossRefGoogle Scholar. Actually, the number of members from districts classified by the authors as close is so small as to make the finding doubtful. What the data do show clearly, however, is that so few members of the Michigan House come from close districts that electoral insecurity could not influence the behavior of many members.

³⁵ MacRae, Duncan Jr., and MacRae, Edith K., “Legislators' Social Status and Their Votes,” American Journal of Sociology, Vol. 66 (May 1961), pp. 699–603CrossRefGoogle Scholar.

³⁶ Fuller, Margaret G., Leadership in the Michigan Legislature (Master's thesis, Michigan State University: published by the author, 1957), pp. 27–28Google Scholar.

³⁷ MacRae, Duncan Jr., “Roll Call Votes and Leadership,” Public Opinion Quarterly, Vol. 20 (Fall, 1956), pp. 543–558CrossRefGoogle Scholar, at p. 552.

³⁸ Respondents did not differ significantly from from non-respondents in terms of constituency characteristics or in terms of party support scores; in fact, they were remarkably similar. Respondents were, therefore, treated as a proper sample from which inferences could be drawn. If there is objection, then it is only necessary to consider what follows as applying to about two-thirds of the members of the 1963 Ohio House and not to the whole House.

³⁹ McClosky, Herbert et al. , “Issue Conflict and Consensus Among Party Leaders and Followers,” this Review, Vol. 53 (June, 1960), pp. 406–427Google Scholar.