¹ Key, V. O. Jr., Southern Politics in State and Nation (New York, 1949)Google Scholar.

² American State Legislatures, ed. Zeller, Belle (New York, 1954), p. 189Google Scholar.

³ Lowell, A. Lawrence, “The Influence of Party upon Legislation in England and America”, Annual Report of the American Historical Society for the Year 1901, Vol. 1, pp. 319–542 (1902)Google Scholar.

⁴ See Lockard, W. Duane, “Legislative Politics in Connecticut”, this Review, Vol. 48, pp. 166–73 (March, 1954)Google Scholar; Keefe, William J., “Party Government and Lawmaking in the Illinois General Assembly”, Northwestern University Law Review, Vol. 47, pp. 55–71 (March, 1952)Google Scholar; and Keefe, William J., “Parties, Partisanship, and Public Policy in the Pennsylvania Legislature”, this Review, Vol. 48, pp. 450–64 (June, 1954)Google Scholar. Comparisons with these studies are made later in this paper.

⁵ Eliminating these two categories facilitates comparison between the states requiring a roll call on all bills and those (e.g., Massachusetts) that do not. Inclusion of these categories in the studies by other writers has been a major factor leading them to apparently contradictory conclusions about party influence in legislatures of various states.

⁶ The omitted roll calls were 30 on two issues in the 1945 session of the Washington Senate and 51 on one issue in the 1947 session of the Colorado House.

⁷ Rice, Stuart, Quantitative Methods in Politics (New York, 1928), p. 209Google Scholar.

⁸ In a study of party voting in a similar New England state legislature, Connecticut, for eleven sessions (1931–1951), W. Duane Lockard concluded that “parties, far from being relatively insignificant, play a dominating role.” Connecticut, like Massachusetts, does not require roll calls on final passage of all bills and usually has an even lower number of total roll calls per session than does Massachusetts. This figure for Connecticut therefore presumably covers only the most controversial bills. Excluding the small number of unanimous votes (seven per cent in the Senate and three per cent in the House) from Lockard's statistics, we find that the parties took opposite sides in 90 per cent of the Senate votes and 83 per cent of the House votes, slightly less than in Massachusetts.

The Connecticut totals show a high level of unity for both parties, although unfortunately in this category the roll calls on which the parties agreed and disagreed are not separated. The percentage of all roll calls (excluding unanimous ones) on which each party had a cohesion index of 80 or more in the Senate was 59 per cent for the Democrats and 76 per cent for the Republicans, and in the House 83 per cent for the Democrats and 72 per cent for the Republicans. The Republicans controlled the House in all eleven sessions, and the Democrats Controlled the Senate in most of the sessions. These figures are similar to the high levels in other urban, industrial states. (Presumably the number of roll calls with 90 per cent of both parties agreed, included by Mr. Lockard, was too small to distort the comparison in a state not requiring roll calls.) See Lockard, “Legislative Politics in Connecticut.”

⁹ William J. Keefe reached a different conclusion in a study of the 1951 Pennsylvania legislature, that “the policy decisions which confronted the Pennsylvania legislature were mainly nonpartisan.” This is largely because 82 per cent of the Senate roll calls and 70 per cent of those in the House were unanimous. Keefe recognizes that the requirement for a roll call on all bills leads to this high figure. It would seem that the test of party voting should be concentrated on controversial issues that lead to some dispute.

If unanimous votes are excluded, Keefe's figures for 1951 appear to show somewhat less party voting than do those for 1945. In 1951 the percentage of roll calls with the parties on opposite sides was 34 in the Senate and 43 in the House, compared to 64 and 81 in 1945. A large part of the difference is probably accounted for by the inclusion of roll calls with over 90 per cent of both parties in agreement, which in many states are very numerous. The party votes (both parties with an index of cohesion of at least 80) formed a high proportion of all votes on which the parties were opposed in both studies. The House figure for this was 69 per cent in both years, but the Senate figure dropped from 81 per cent in 1945 to 64 per cent in 1951. See Keefe, “Parties, Partisanship and Public Policy in the Pennsylvania Legislature.” The party vote figures are based on a letter to the writer, January 10, 1955.

¹⁰ William J. Keefe, in a study of the Illinois sessions of 1949 and 1951, concluded that parties played a small role in lawmaking. As in his study of Pennsylvania, this resulted largely from the inclusion of unanimous votes, which averaged two-thirds of the total roll calls in the two Illinois sessions. It also resulted from the inclusion of roll calls on which at least 90 per cent of both parties voted on the same side. His figures show that the total of these roll calls exceeded all other non-unanimous roll calls in 1949. Keefe does not list the total number of roll calls on which the parties were opposed and does not compare the unity level of the two parties. These differences in standards make a comparison of the two sessions difficult, but Keefe did find that roll calls in which the parties were opposed and in which both had a high level of unity dropped from 1949 to 1951, particularly in the House. See Keefe, “Party, Government and Lawmaking in the Illinois General Assembly.” The comparison is also baaed on the previously-mentioned letter to the writer.

¹¹ By comparison, Professor Lowell's study in 1901, which used virtually the same criteria as this study, showed high party unity in New York, less in Ohio and Illinois, and very little in Massachusetts and Pennsylvania. He noted that only these latter two states had large and consistent Republican majorities.

The parties took opposite sides in about one-half of the roll calls in three states, and in somewhat more than half in Massachusetts and New York. The percentage of party votes (the parties opposed and both with a cohesion index over 80, by his standard) was 23 and 45 in the New York Senate and House, 15 and 10 in the Ohio Senate and House, 13 in the Illinois House, and from one to six in the other legislative bodies. The Democrats had much more unity than the Republicans in Massachusetts, somewhat more in New York, Pennsylvania, and Ohio, and less in Illinois. See Lowell, “The Influence of Party upon Legislation in England and America.”

¹² For the sake of convenience and brevity this “metropolitan and big city versus small city and rural” alignment is described in this paper as urban-rural.

¹³ See Keefe, “Parties, Partisanship and Public Policy in the Pennsylvania Legislature,” and “Party Government and Lawmaking in the Illinois General Assembly.”

¹⁴ See American State Legislatures, pp. 189–92.