POLETICAL PARIISANSHEP IN FOUR STATE LEGISLATURES Thesis for fin Degree of DH. D. MICHIGAN STATE UNWERSITY Bernard W. Klein 1966 'i‘fSIl LIBRARY ~ Michigan State University This is to certify that the thesis entitled POLITICAL PARTISANSHIP IN FOUR STATE LEGISLATURES presented by Bernard w. Klein has been accepted towards fulfillment of the requirements for Ph. DI degree in Political Science Cgf/Mbigéc -, (72/4 L, Major professor Dateiebruary 21, 1966 0-169 A i Amy—own". ffimfi—'v4v“ . .i n-mm. - ABSTRACT POLITICAL PARTISANSHIP IN FOUR STATE LEGISLATURES by Bernard W. Klein This dissertation is an analysis of the relation— ship between partisan voting and seventeen other variables in the legislatures of California, New Jersey, Ohio, and Tennessee. The independent variables consist of partisan attitudes, constituency factors, perceptions of the polit- ical party in the legislature, career expectations, per- sonal factors (age, education, experience), sense of effi— cacy and liberal—conservative ideology. Attitudinal, constituency, and personal data were obtained from the material gathered by the Four State Legislative Study directed by Heinz Eulau, John Wahlke, LeRoy C. Ferguson and William Buchanan in 1957. The roll calls used to determine partisan voting behavior were gathered for the same session as those in which the interview data were gathered. A partisan roll call was defined as one in l Bernard W. Klein which the majority of one party differed with the majority of the other party. The relationships that were confirmed most conclu- sively were: 1. Holders of leadership positions in the legislature tended to be high party voters to a greater degree than non—office holders (62.5%). Where party was a factor in the legislator's de- cision to run for that office, the legislator would tend to be a high party voter (56.4%). Those legislators who intended to seek reelection tended to be high party voters (56.4%). High party voters were more likely to make a dis- tinction between the behaviors of a "party man“ as against an "independent" than would low party voters (56.4%). The hypothesis that was negated most conclusively was the relationship between sense of efficacy and high partisan voting. Low party voters tended to have a higher sense of efficacy than high party voters (62.5%). 2 Bernard W. Klein The agreements and disagreements on the different variables were then compared among the units of the four states and I found that the low and high party voters of California agreed more with their counterparts in Tennes- see than with those of New Jersey and Ohio. The low and high party voting legislators of New Jersey and Ohio agreed more frequently with each other than with those in California or Tennessee. This appears to suggest party situation in a particular state could be a crucial factor in the relationships of partisan voting to other variables. POLITICAL PARTISANSHIP IN FOUR STATE LEGISLATURES BY Bernard W. Klein A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Political Science 1966 ACKNOWLEDGMENTS I would like to express my appreciation for the patient guidance accorded me by my dissertation committee composed of Professors Charles Press, Chairman, and LeRoy C. Ferguson and Charles R. Adrian. I am especially grate- ful to Professor Ferguson Who guided this dissertation in its early stages and made available to me the data of the Four State Legislative Study. His organization of the un- published portions of that study was most helpful to this dissertation as well as to possible future studies. I owe a great debt of gratitude to my family, and the schools which I attended for having develOped and encouraged my desire for education generally and my interest in the field of political science Specifically. My colleagues at the University of Michigan Institute of Public Administration as well as my other personal friends all provided me with much needed encouragement for which I am most grateful. Bernard W. Klein ii TABLE OF CONTENTS Page ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . . . . . . . vi Chapter I PARTISANSHIP IN STATE LEGISLATURES . . . . . . . l The Research Literature on Voting in Legislatures. . . . . . . . . . . . . . . . 2 1. How much partisan influence exists in legislative bodies? . . . . . . . . . 8 2. Interest Groups . . . . . . . . . . . . 9 3. Legislative Leadership. . . . . . . . . lO 4. Social Background & Constituency. . . . ll 5. Political Culture . . . . . . . . . . . 12 6. Legislative Strategy. . . . . . . . . . l4 7. Role Theory . . . . . . . . . . . . . . 15 8. Career Expectations . . . . . . . . . . l6 9. Psychological Approach. . . . . . . . . l6 Scope of Study. . . . . . . . . . . . . . . . l7 Methodological Problems Concerning Partisan Votes . . . . . . . . . . . . . . . . . . . 19 II THE POLITICAL SETTING. . . . . . . . . . . . . . 3O Classification. . . . . . . . . . . . . . . . 32 California. . . . . . . . . . . . . . . . . . 37 New Jersey. . . . . . . . . . . . . . . . . . 41 iii Table of Contents—-cont. Chapter Ohio. . . . . . . . . . . . . . . . . . Tennessee . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . III LEGISLATIVE ATTITUDES TO PARTISANSHIP AND LEGISLATIVE VOTES . . . . . . . . . IV CONSTITUENCY AND PARTY VOTING BEHAVIOR . . Competitive Nature of the District. . . Urban—rural Nature of the District. . . Summary . . . . . . . . . . . . . . . . V PERSONAL CHARACTERISTICS AND PARTISANSHIP. Age . . . . . . . . . . . . . . . . . . Education . . . . . . . . . . . . . . . Occupation. . . . . . . . . . Summary . . . . . . . . . . . . . . . . VI CAREER FACTORS AND PARTISANSHIP. . . . . . Party as Source of Decision to Become a Legislator. . . . . . . . . . . . . . Legislative Experience. . . . . . . . . Offices in the Legislature. . . . . . . Running for Reelection. . . . . . . . . Summary . . . . . . . . . . . . . . . . iv Page 44 46 49 51 69 7O 78 87 89 89 94 99 103 107 107 112 117 121 125 Table of Contents--cont. Page Chapter VII PERCEPTIONS OF PARTY IN THE LEGISLATURE. . . . . 128 Influence on Individual Members . . . . . . . 130 Party Influence on Individual Members and Voting Behavior . . . . . . . . . . . . . . 135 Activities Influenced by Parties. . . . . . . 138 Summary . . . . . . . . . . . . . . . . . . . 143 VIII PARTY IDENTIFICATION BY INDIVIDUAL LEGISLATORS. I46 Justification for not voting with Party . . . 153 Advantages of Supporting Party Leaders. . . . 159 Summary . . . . . . . . . . . . . . . . . . . 164 IX EFFICACY, LIBERALISM AND VOTING . . . . . . . . 167 Liberalism—-Conservatism and Partisanship . . 173 Summary . . . . . . . . . . . . . . . . . . . 181 X SUMMARY AND CONCLUSIONS . . . . . . . . . . . . 183 Testing of Hypotheses . . . . . . . . . . . . 184 Situational Conditions. . . . . . . . . . . . 192 California. . . . . . . . . . . . . . . . . 200 New Jersey. . . . . . . . . . . . . . . . . 202 Ohio. . . . . . . . . . . . . . . . . . . . 203 Tennessee . . . . . . . . . . . . . . . . . 205 Conclusions . . . . . . . . . . . . . . . . . 207 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . 210 Table 10 11 12 13 LIST OF TABLES PARTISAN ROLL CALLS IN EACH STATE, 1957 SESSION. PARTY EVALUATION SCALE ITEMS . . . . . . . . . PARTY EVALUATION SCALE SCORES BY STATE . DISTRIBUTION OF PARTY EVALUATION SCALE TYPES IN FOUR STATES . . . . . . . . . . . . . . . . . RELATIONSHIP OF PARTISAN VOTING TO PARTISAN ATTITUDES IN CALIFORNIA . . . . . . . . . . . RELATIONSHIP OF PARTISAN VOTING TO PARTISAN ATTITUDES IN TENNESSEE. . . . . . . . . . . RELATIONSHIP OF PARTISAN VOTING TO PARTISAN ATTITUDES IN NEW JERSEY . . . . . . . . . RELATIONSHIP OF PARTISAN VOTING TO PARTISAN ATTITUDES IN OHIO . . . . . . . . . . . . . . RELATION OF COMPETITIVENESS OF DISTRICTS TO PARTISAN ATTITUDES. . . . . . . . . . . . . . COMPETITIVE NATURE OF DISTRICT AND PARTY VOTING IN CALIFORNIA . . . . . . . . . . . . . . . . COMPETITIVE NATURE OF DISTRICT AND PARTY VOTING IN NEW JERSEY . . . . . . . . . . . . . . . . COMPETITIVE NATURE OF DISTRICT AND PARTY VOTING IN OHIO . . . . . . . . . . . . . . . . . . . COMPETITIVE NATURE OF DISTRICT AND PARTY VOTING IN TENNESSEE. . . . . . . . . . . . . . . . . vi Page 23 54 55 56 65 66 67 68 73 74 75 76 78 List of Tables--cont. Table 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 RELATION OF RURAL-URBAN NATURE OF DISTRICT TO PARTISAN ATTITUDES. . . . . . . . . . . . . . URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN CALIFORNIA . . . . . . . . . . . . . . URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN NEW JERSEY . . . . . . . . . . . . . . . . URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN OHIO . . . . . . . . . . . . . . . . . . . URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN TENNESSEE. . . . . . . . . . . . . . . . . RELATION OF AGE TO PARTISAN ATTITUDES. . . . AGE AND PARTISAN VOTING. . . . . . . . . . . . RELATION OF EDUCATION TO PARTISAN ATTITUDES. . . EDUCATION AND PARTISAN VOTING. . . . . . . . EDUCATION BEFORE COLLEGE AND PARTISAN VOTING RELATION OF OCCUPATION TO PARTISAN ATTITUDES MEMBERSHIP IN LEGAL PROFESSION AND PARTISAN VOTING. . . . . . . . . . . . . . . . . . . RELATION OF MENTION OF PARTY AS SOURCE OF DECISION TO BECOME LEGISLATOR TO PARTISAN ATTITUDES . . . . . . . . . . . . . . . . MENTION OF PARTY AND PARTY VOTING- - - - - - - RELATION OF YEARS OF LEGISLATIVE EXPERIENCE TO PARTISAN ATTITUDES. . . . . . . . . . . . . . YEARS OF LEGISLATIVE EXPERIENCE AND PARTY VOTING vii Page 82 83 84 85 86 91 93 95 96 98 101 102 110 111 115 116 List of Tables-~cont. Table 30 31 32 33 34 35 36 37 38 39 4O 41 42 43 44 45 RELATION OF HOLDING OFFICE IN THE LEGISLATURE TO PARTISAN ATTITUDES . . . . . . . . . . OFFICE HOLDING IN LEGISLATURE AND PARTY VOTING . RELATION OF INTENTION TO RUN AGAIN FOR THE LEGISLATURE TO PARTISAN ATTITUDES . . . . . INTENTION TO SEEK REELECTION AND PARTY VOTING. RELATION OF PARTISAN ATTITUDES TO PERCEPTION OF PARTY IMPACT ON LEGISLATIVE DECISIONS . . PERCEPTION OF PARTY IMPACT ON LEGISLATIVE DECISIONS AND PARTY VOTING. . . . . . . . RELATION OF PARTISANSHIP TO PERCEPTION OF PARTY INFLUENCE ON PARTIES IN THE [STATE] LEGIS- LATURE. . . . . . . . . . . . . . . . . . . . PARTY INFLUENCE ON INDIVIDUAL MEMEBERS . . . ACTIVITIES PERCEIVED AS BEING INFLUENCED BY POLITICAL PARTIES . . . . . . . . . . . . PARTY ACTIVITIES AND PARTY VOTING. . . . . . BEHAVIOR OF PARTY MAN AND PARTY VOTING . . . BEHAVIOR OF PARTY MAN AS COMPARED TO PARTY VOTING JUSTIFICATION FOR NOT VOTING WITH PARTY. . . JUSTIFICATION OF NOT VOTING WITH PARTY . . . . ADVANTAGES OF SUPPORTING PARTY LEADERS . . . ADVANTAGES OF SUPPORTING PARTY LEADERS RELATED TO PARTY VOTING . . . . . . . . . . . . . . . viii Page 119 120 122 124 131 132 134 137 141 142 150 151 155 157 162 163 List of Tables--cont. Table 46 47 48 49 50 51 52 53 54 55 56 LEGISLATIVE EFFICACY AMONG THE FOUR STATES RELATION OF SENSE OF LEGISLATIVE EFFICACY TO PARTISAN ATTITUDES. . . . . . . . . . . SENSE OF EFFICACY AND PARTY VOTING . . . . RELATION OF LIBERAL-CONSERVATIVE IDEOLOGY TO PARTISAN ATTITUDES. . . . . . . . . . LIBERALISM--CONSERVATISM AND PARTISAN VOTING . RESULTS OF TESTING OF SEVENTEEN HYPOTHESES CONFIRMATIONS OF SEVENTEEN VARIABLES AMONG UNITS . . . . . . . . . . . . . . . . . CALIFORNIA--1957 . . . . . . . . . . . NEW JERSEY--l957 . . . . . . . . . . OHIO--1957 . . . . . . . . . . . . . . . . TENNESSEE--1957. . . . . . . . . . . . . . ix THE Page 168 171 172 176 179 185 193 196 197 198 199 POLITICAL PARTISANSHIP IN FOUR STATE LEGISLATURES CHAPTER I PARTISANSHIP IN STATE LEGISLATURES Much has been written over the years regarding the role of political parties in the American political system. Studies have compared political parties in many nations, as exemplified by the work of Sigmund Neuman. Others have examined functions performed by the parties, organization and control of parties, and party ideologies, party's role in government and a host of other variables. Much of the literature concerning American polit— ical parties has been reformist in nature or been descrip- tive only of legal and institutional factors. Only re— cently have political scientists undertaken behavioral studies which attempt to analyze party activity as it 1 . . . . Slgmund Neumann, ed., Modern Political Parties, Approaches to Comparative Politics (Chicago: University of Chicago Press, 1956). relates to governmental decision making. One aspect of these studies is the effect of partisanship on decision making in legislative bodies. Potentially the greatest impact of partisanship in a legislature is voting, inasmuch as shaping policy is supposed to be the culmination of all previous politi- cal activity (i.e., recruitment, campaign, election, or- ganization of legislature, etc.). Stated in another manner, voting behavior in a legislative body should provide a mirror of the effectiveness of the political party in the entire legislative process. Voting behavior is thus the focus of this study on the effect of parties on legislative decision making in four states. The Research Literature on Voting in Legislatures This study relies most heavily on behavioral studies of the legislative process both as to methodology as well as findings. However, it should not be assumed that earlier studies of formal organization of legisla— tures, committees and procedures though primarily reform oriented, are not of value. These earlier studies provide useful information as well as some underlying hypotheses that behavioral studies have been able to utilize. Inas- much as it would be too time consuming and of doubtful value to cite all previous works dealing with the legis- lative process, this writer will cite only those studies which have attempted to relate and assess the significance of political parties in the legislative process. Special emphasis is given to those which concentrate on roll call votes and the influence of political parties. Excellent bibliographic essays already exist that review the legis— lative process in general. One of these is one by Norman Meller entitled "Legislative Behavior Research”2 and an- other is that of John Wahlke entitled "Behavioral Analyses of Representative Bodies."3 An additional bibliography without annotations was prepared by the Michigan Senate Fellows and is titled "The Legislative Process--A Bibli— ography in Legislative Behavior."4 Norman Meller, "Legislative Behavior Research," Western Political Quarterly, XIII (March 1960), pp. 131- 153. 3John Wahlke, "Behavioral Analysis of Representa- tive Bodies," in Austin Ranney, ed., Essays On the Behav- ioral Study of Politics (Urbana: University of Illinois Press, 1962), pp. 173-190. 4Michigan State Senate Fellows of 1963—64, ”The Legislative Process--A Bibliography in Legislative Inasmuch as the legislative branch of government received much attention in the early days of the founding of our country, it is not surprising to find that much of the early literature of the legislative process is con— cerned with how legislatures can take their "rightful" place in the scheme of government. This "reform” liter— ature has consistently decried the failures of legislatures to function more "efficiently" and "reSponsibly" and has contained various suggestions for improving legislatures. Briefly stated, the tenets of the reformers contain such proposals as that there is a crying need for greater con— trol over lobbyists and lobbying activities, a need for more staff help for legislators, reform of the committee system, change in the size of legislatures, and use of machines for voting.5 Some of the recommendations of this reform literature as it pertained to Congress cul— minated in the Legislative Reorganization Act of 1946. .This literature will be briefly reviewed first. Behavior (East Lansing: Institute for Community Develop- ment and Services, Michigan State University, 1963. 5For example, George B. Galloway, Congress at the Crossroads (New York: Crowell, 1946); Ronald Young, This is Congress (New York: Knopf, 1943); Thomas K. Finletter. Can Representative Government Do the Job? (New York: Rey- nal & Hitchcock, 1945); Estes Kefauver and Jack Levin, A_ The literature of reform of the legislative process has been marked by ambivalence in its treatment of politi— cal parties. The earlier approach shied away from assign- ing political parties a legitimate place in the legislative process. The tendency in this earlier literature was to decry the parties presence in legislative bodies and postu— late the superiority of nonpartisanship, a system of elec- tion which was adopted in many local government units and in two state legislatures.6 In later years there was a change in the attitude of some reformers toward the posi- tion that the political parties should play a greater role in the legislative process but that reform within the party system was necessary in order that they be responsive and . . . . . . 7 reSpon31ble in carrying out their m1531on. The reform Twentieth Century Congress (New York: Duell, Sloan and Pearce, 1947); and Dean Acheson, A Citizen Looks at Congress (New York: Harper & Bros., 1956). 6For example, Donald C. Blaisdell, American Democ— racy Under Pressure (New York: The Ronald Press Co., 1957); Stuart Chase, Democracy Under Pressure, Special Interest vs. The Public Welfare (New York: The Twentieth Century Fund, 1945); Kenneth G. Crawford, The Pressure Boys: The Inside Story of Lobbying in America (New York: J. Messner, 1938); Karl Schriftgiesser, The Lobbyists, The Art and Business of Influencing Lawmakers (Boston, Little Brown & Co., 1957). 7Toward a More Responsible Two Party System, A Re- port of the Committee on Political Parties, American Polit- ical Science Association (New Ybrk: Rinehart & Company, Inc., 1950). literature that has evidenced greater concern over the role political parties should play in the legislative process is well illustrated by the recent book of James MacGregor Burns.8 Burns states that there is actually ”four party politics in America“ and is concerned that action must be taken for "Americans to regain control of 9 their national politics . . . . Another very significant body of reform literature on the legislative process which has a bearing on the role of political parties is the literature describing Specific interest groups. Its underlying hypothesis is that inter- est groups rather than parties are the key factor in in- fluencing roll call votes in legislatures. These ”pressure" groups are seen as working through both political parties. In the conceptual framework used in many of these studies, government policy is seen as the result of the interplay of the various interest groups rather than through party 8James M. Burns, Congress On Trial (New York: Harper & Bros., 1949). See also the case study of Stephen K. Bailey, Congress Makes a Law: The Story Behind the Em- ployment Act of 1946 (New York: Columbia University Press, 1950). James M. Burns, Deadlock of Democracy (Englewood Cliffs, New Jersey: Prentice-Hall, 1963). competition.lo Many writers of such studies conclude that a strengthening of the political parties is necessary if we are not to have government by pressure group. In recent years there has also been a blossoming forth of many studies that I will refer to as the behav— ioral approach to politics. This approach differs from earlier approaches in that behavior becomes the prime ob- ject of study rather than structure or institution or his- torical or legal analysis. This approach as applied to the legislative process would concern the student more with what "is“ that is the human behavior associated with legislative bodies rather than with what "ought to be.” Legislatures lend themselves to the behavioral approach because of the large numbers of people involved and the many types of behavior that lend themselves to systematic study. 10Peter H. Odegard, Pressure Politics: The Study of the Anti-Saloon League (New York: Columbia University Press, 1928); E. E. Schattschneider, Pressure and the Tar- ‘iff (New York: Prentice-Hall, 1935); Belle Zeller, Pres- sure Politics in New York (New YOrk: Prentice-Hall, 1937); Dayton McKlan, Pressures on the Legislature of New Jersey (New York: Columbia University Press, 1938); and Fred W. Riggs, Pressures on Congress (New York: Kings Crown Press, Columbia University, 1950). In presenting the findings of the behavioral lit- erature relevant to the present study, this writer shall present them in terms of how they tend to explain the re— sults of roll call votes in legislative bodies. Some studies of course can be categorized under several of the headings that follow and will be cited several times where— ever it is apprOpriate to do so. 1. How much partisan influence exists in legislative bodies? An early work at the turn of the century, dealing with the question of the role of the parties on legisla— tion was that of A. Lawrence Lowell.ll Not only did Lowell come to grips with certain methodological problems (i.e., defining party votes) but he tested hypotheses that led him to conclude that the political parties in American legislative bodies he studied played a very minor role in roll call votes. The findings of Lowell influenced much of the thinking of students of the legislative pro- cess. Stuart A. Rice continued Lowell's approach to the 11 A. Lawrence Lowell, ”The Influence of Party Upon Legislation," Annual Report of the American Histor- ical Association, Vol. 1 (Washington, D. C.: U. S. Gov- ernment Printing Office, 1901), pp. 321-543. study of legislative bodies and developed an index of co- hesion to measure party voting and his conclusions are not very different from those of Lowell.12 It is in only more recent years that other studies have emerged that suggest there are ideological differences between the two parties in present day legislatures. There has been a great out— pouring of roll call studies in which party cohesion is the crucial variable. A major finding is that certain issues are more likely to evoke party differences. 2. Interest Groups In addition to some of the descriptions of individ- ual interest groups mentioned earlier, there have been few studies using behavioral methods but some develOpment of a theoretical framework in which to study these groups has occurred. The major theoretical framework of many of these 2See Malcolm E. Jewell, ”Party Voting in American State Legislatures," American Political Science Review, Vol. XLIX (September, 1955), pp. 773-791; W. Duane Lockard, "Legislative Politics in Connecticut," American Political Science Review, Vol. XLVIII (March, 1954), pp. 166-173; Duncan MacRae, Jr., ”The Relation Between Roll Call Votes and Constituencies in the Massachusetts House of Represen— tatives,” American Political Science Review, Vol.-XLVI (December, 1952), pp. 1046—1055 and several others. 3 . . . . . . Stuart A. Rice, Quant1tat1ve Methods in Polit1cs (New YOrk: Knopf, 1928). 10 studies has been developed by Arthur F. Bentley and ex- . 14 panded upon by Dav1d Truman and Bertram Gross. These men viewed the legislative process as a field of combat among competing interests. Most of the studies of inter- est groups that have emerged have been in the category of individual case study or a study of a particular bloc of legislators as representative of a particular pressure group. 3. Legislative Leadership The behavioral approach to the study of legisla- tive leadership has focused mainly upon the power rela- tionships that exist in legislative bodies. Woodrow 14 Arthur F. Bentley, The Process of Government; A Studyiof Social Pressures (Bloomington, Indiana: The Principia Press, Inc., 1949); David B. Truman, The Gov- ernmental Process (New York: Knopf, 1951); Bertram M. Gross, The Legislative Struggle (New York: McGraw-Hill, 1953); Samuel J. Eldersveld, ”American Interest Groups: A Survey of Research and Some Implications for Theory and Method," in Henry W. Ehrmann, ed., Interest Groups on Four Continents (Pittsburgh: University of Pittsburgh Press, 1958), pp. 173-197; and Earl Latham, Group Basis of Politics—-A Study in Basing Point Legislation (Ithaca: Cornell University Press, 1952). 15 . For example, Odegard, Zeller, Schattschneider, 0p. cit.; and E. Pendleton Herring, Group Representation Before Congress (Baltimore, Maryland: The Johns Hopkins Press, 1929); David B. Truman, The Congressional Party; A Case Study (New York: John Wiley, 1959); and Donald C. Blaisdell, ed.,"Unofficial Government: Pressure Groups and Lobbies"in The Annals (Philadelphia, 1958). 11 Wilson pointed out in 1885, that power in Congress is ex- ercised by committees and their chairmen. Since then some writers have attempted to measure power in quantita- tive terms and have developed several interesting schemes. The role of parties does not figure very prominently in these studies with the exception of that of Truman. 4. Social Background and Constituency A common study has been of the charactersitics, age, group membership, etc., revealed in biographies. The constituency approach to the legislative process attempts to push beyond the studies of the characteris— tics of individual legislators to study the social and demographic characteristics of the constituency influence upon legislators goes beyond formal interest group influ— ence and seeks an explanation of legislative results by comparing constituencies and roll call votes.17 The 6Woodrow Wilson, Congressional Government (Bos- ton: Houghton Mifflin, 1885); Mary P. Follett, The Speaker of the House of Representatives (New York: Longmans, Green, 1896); Ralph K. Huitt, "The Congressional Committee: A Case Study," American Political Science Review, Vol. 48 (1954), pp. 340-365; Duncan MacRae, Jr., "Roll Call Votes and Leadership," Public Opinion Qparterly, Vol. 20 (Fall 1956), pp. 543-558. 7 . . Lew1s A. Froman, Jr., "Inter-Party Constituency Differences and Congressional Voting Behavior," American 12 combining of studies of parties and constituency has yielded fruitful results in interpreting legislative re- sults and has offered greater promise than the more sim- plified study of urban—rural conflict or other single conflict. When tied in with party voting, studies have shown that constituency influences have accounted for de- viations in party voting. More recently it has been hy- pothesized that party and constituency are much more syn- onymous than was previously thought.18 5. Political Culture Another approach toward explaining the results of the legislative votes is that of using the political cul- ture of the constituency as the independent variable and voting behavior and other legislative behavior as the de- pendent variable. Such is the basic approach used by Frank Sorauf in his study of the Pennsylvania Legislature.19 Political Science Review, Vol. 57 (March, 1963), pp. 57—62; also Congressmen and Their Constituencies (Chicago: Rand McNally & Co., 1963); Julius Turner, Party and Constituency: Pressures on Congress, "Studies in Historical and Political Science" (Baltimore: The Johns Hopkins Press, 1951). 18 ' . Frank J. Sorauf, Party and Representation (New York: Atherton Press, 1963). 19Ibid. 13 He argues that the nature of the constituency appears to have greater influence on determining the type of politi— cal organization than do prospects for party victory. A similar approach was used by Leon Epstein in his study of Wisconsin politics.21 He finds that legislators from urban areas tend to be more educated and do not have a record of long party activity.22 Briefly stated, these students have combined demographic, social, and economic factors to describe the politics of their respective states. Epstein compared his findings with hypotheses that have been presented by previous writers, particu- larly V. 0. Key.23 Lewis Froman also showed a relation— ship between Congressional voting and the political en- vironment of the Congressional constituencies.24 He ar— gues that partisanship and constituency overlap rather emerge as separate variables. Other studies which use ZOIbid., p. 49. Leon Epstein, Politics in Wisconsin (Madison: University of Wisconsin Press, 1958). 22Ibid., p. 109. 23Ibid., p. 99. 24 . Froman, Op. Cit. 14 political culture concepts to explain legislative behavior are those which analyze social backgrounds of legislators. 6. Legislative Strategy This approach explains legislative behavior in terms of strategies that can be developed on given issues. These strategies take the political party into considera— tion as one factor but do so in terms of coalitions that can be mustered on the issues. There is still very little of an empirical nature that has emerged using this game theory framework but it does present some new approaches to studying the role of parties in the legislative pro- cess . 25Donald R. Matthews, "United States Senators and Class Structure," Public Opinion Quarterly, Vol. 18(Spring, 1954), pp. 5-22; Joseph A. Schlesinger, "Lawyers and Ameri— ican Politics," Midwest Journal of Political Science, Vol. 1 (May, 1957), pp. 26-39; John C. Wahlke, William Buchanan, Heinz Eulau, and LeRoy Ferguson, "The Political Socializa- tion of American State Legislators," Midwest Journal of Political Science, Vol. 3 (1959), pp. 188-206; and Donald R. Matthews, U. S. Senators and Their World (Chapel Hill: University of North Carolina Press, 1960). 26 . . L. S. Shapley and Martin Subik, "A Method for Evaluating the Distribution of Power in a Committee Sys- tem," American Political Science Review, Vol. XLVIII (September 1954), p. 789; R. Donald Luce and Arnold A. Rogow, "A Game Theoretic Analysis of Congressional Power Distributions for a Two-Party System," Behavioral Science, Vol. I (April 1956), p. 83. 15 7. Role Theory In recent years there has been a greater acceptance of and use of role theory as a framework for the study of legislative behavior. The basic hypothesis of role theory is that there are certain roles in a legislative body that legislators play either consciously or otherwise and their behavior is conditioned by their own perceptions of their roles as well as the role expectations that develop in a collectivity such as a legislature. This approach relies most heavily upon systematic interviews and the setting up of categories of roles. Ralph Huitt does this in his study of the roles of Congressional committee members.27 The most far reaching use and presentation of role theory is the Four State Study of Wahlke, Eulau, Ferguson and Buch- anan. They have set up legislative types derived from the behavior of individual legislators. They show that there are certain norms and role expectations in a legis- lative body to which legislators tend to conform. 27Ralph Huitt, Op. cit. 28 . . . John C. Wahlke, Heinz Eulau,-William Buchanan and LeRoy C. Ferguson, The Legislative System (New York: John Wiley & Sons, 1962). 16 8. Career Expectations A variation of role theory would take into consid- eration that legislators in addition to acting according to their role perceptions of the present will also act according to what their future political career expecta- tions would require. Such variables as the desire to seek reelection, desire to run for higher office would tend to govern a legislator's behavior. Career expecta- tions as a form of behavior have been used by Sorauf, Epstein and the Four State Study, and by Schlesinger in respect to Governors and Krislov for southern Attorneys General.29 9. Psychological Approach This approach to the study of legislative behavior would tend to explain legislative and other types of po— litical behavior in terms of the psychological background of the individuals. The most noted proponent of this ap— proach has been Harold Lasswell. This approach has opened new facets for studying politics but still leaves us 9Ibid.; see also Sorauf, 0p. cit.; and Epstein op. cit.; Samuel Krislov, "Constituency vs Constituional- ism: The Desegregation Issue and Tensions and Aspirations of Southern Attorneys General,“ Midwest Journal of Polit- cal Science, Vol. III (Feb. 1959), pp. 75-92. 17 without a method of operationalizing the process by which . 30 the individual psyches bring about group results. Leg- islators are not apt to volunteer themselves for psycho— logical testing. Scope of Study The scope of this study is to examine the voting behavior of one session of four state legislatures and treat these data as the dependent variables. Factors thought to be closely associated with partisan activities are the independent variables. These include the cate— gories already reviewed in the discussion of behavioral literature on legislatures such as extent of partisanship, legislative leadership, constituency factors, role percep- tion, career expectations, and personal factors such as age, education, and experience of legislators. In addi— tion certain other concepts will be employed as independ- ent variables such as legislator's feelings of efficacy, his reported party influence on legislative behavior, and his political attitudes vis-a-vis conservatism and liberalism. 0 Harold Laswell, Power and Personality (New York: W. W. Norton & Co., 1948), and Psychopathology and Politics (Chicago: University of Chicago Press, 1930). 18 These data will be related to the roll call votes of the 1957 sessions that could be classed as partisan votes in the four state legislatures. The information about the independent variables was gathered in interviews during the 1957 session of the Legislatures in California, New Jersey, Ohio, and Tennes- see as part of the State Legislative Research Project con- ducted by Professors LeRoy Ferguson, Heinz Eulau, John Wahlke, and William Buchanan. Hereinafter, this study shall be referred to as the Four State Study. Several articles have appeared by some or all of these authors and two books containing findings of their study have appeared.31 The interview data pertaining to legislators' 31John B. McConoughy, "Some Personality Factors of State Legislators in South Carolina," American Political Science Review, 44 (Dec. 1950), pp. 897-903; Wahlke, Eulau, 0p. cit.; William Buchanan, Legislative Partisanship, The Deviant Case of California (Berkeley & Los Angeles: Uni- versity of California Press, 1963); John C. Wahlke, William Buchanan, Heinz Eulau, and LeRoy C. Ferguson, "The Poltical Socialization of American Legislators," Op. cit.; John C. Wahlke and others, "The Legislator as a Specialist," West- ern Political Quarterly, 13 (September 1960), pp. 636-651; Heinz Eulau, William Buchanan, LeRoy Ferguson and John Wahlke, "The Role of the Representative: Some Empirical Observations of the Theory of Edmund Burke,” American Po— litical Science Review, 53 (September 1959), pp. 742-756. See also John C. Wahlke and others, “American State Legis- lators' Role Orientations Toward Pressure Groups,“ Journal of Politics, 22 (1960), pp. 203-227; and John C. Wahlke and Heinz Eulau, Legislative Behavior: A Reader in Theory and Research (Glencoe: The Free Press, 1959). 19 perceptions of the role of political parties is but a small portion of all the data gathered in connection with the larger study, but it is that portion upon which the present study relies very heavily. Use was also made in Chapter III of this study of the Guttman scale of partisan attitudes that was prepared by the four authors of this study. The findings of their scale were compared with the partisan voting behavior of the legislators of these four states. Methodolpgical Problems Concerning Partisan Votes In setting out to do a study of voting behavior of legislators, certain methodological problems were ap- parent at the outset. One of the most important of these was defining a ”partisan roll call." This problem was faced by all students of legislative voting behavior who have used roll call analysis as a tool of research. The definition of a partisan roll call used in this study is any vote in which the majority of those voting from one party differed from the majority of those voting from the Opposite party. This definition of a partisan roll 20 call was also used by Robert Scigliano in his study of the Michigan Legislature.32 Other solutions to this problem have been setting a percentage of party divergence such as 75% of one party differing with 75%.of the Opposing party, which would de- fine a "partisan roll call." Stuart Rice developed an index of cohesion which briefly stated is the figure that results when the proportion voting against the party ma— jority is subtracted from the percentage that constitutes the party majority. In other words if a group of 100 legislators Of the same party divided 75-25 their cohesion index would be 50 (75%.minus 25%). If they were unanimous, their cohesion index would be 100 or absolute cohesion, if they were evenly split, the index number would be 0 (50%,minus 50%). Rice's measure has been used by Jewell and Lockard. There are certain advantages and disadvantages to each definition. It is the feeling of this writer that 32 . . . . . . Robert G. SCigliano, Michigan Legislative Report, 1954; Regular and Special Sessions (East Lansing, Michigan: Governmental Research Bureau, Michigan State College, 1955). 33 . . . Malcolm E. Jewell, The State Legislature-"Politics and Practice (New YOrk: Random House, 1962), p. 50; Duane Lockard, Op. cit., p. 68. 21 the simple majority difference in defining a partisan roll call is best for purposes of this study because it is less complex to arrive at because calculations are simpler for large numbers of roll calls and that one does not achieve any more exact definition by setting an arbitrary percent- age requirement above a simple majority. Another methodological problem which had to be faced was the treatment of unanimous or near unanimous roll calls. This presented no problem by use of this study's definition of a "partisan roll call" but does pose a problem in other studies. Jewell does not include votes in which 90% of both parties voted on the same side. He states that "such roll calls prove nothing about the role of the parties on issues which provoked disagreement in the legislature."34 Lockard in his study of Connecti- cut does include such votes though he notes that they are relatively infrequent. Keefe makes use Of unanimous roll calls and claims that a unanimous vote on a bill does not necessarily mean that the issue is an unimportant one. 34Jewell, Ibid., p. 48. 35 . . - William J. Keefe, ”Comparative Study of the Role of Political Parties in the State Legislatures," Western Political Quarterly, Vol. IX (September 1956), pp. 726-742, p. 741. 22 Another methodological problem which had to be faced was the treatment of abstentions. Jewell makes no record of those abstaining from voting on a given issue while Keefe counts abstentions as negative votes. This study makes no special provisions for abstentions. Wil— liam Buchanan in his study of uses a much refined system of certain bills.36 It should be repeated study a partisan roll call is majority of those voting from majority of those voting from tempt was made to analyze the the California Legislature inclusion and exclusion of that for purposes of this defined as one in which the one party differed from the the opposite party. No at- content of the bills being voted on or the number of roll calls on a particular bill. Each roll call was treated equally and all unanimous or near unanimous roll calls were excluded since they did not meet the initial definition of partisanship. The num- ber of roll calls during the 1957 session which met the definition of a partisan roll on the following page. call are shown in Table l 36Buchanan, Op. cit. 23 TABLE 1 PARTISAN ROLL CALLS IN EACH STATE, 1957 SESSION California Assembly — 108 Senate - 25 New Jersey Assembly - 55 Senate — 57 Ohio House of Representatives - 65 Senate - 30 Tennessee House of Representatives — 20 Senate - 66 It should be stated again that the reasons a leg— islator votes along with or in opposition to the majority of his own party can be varied and no attempt was made to ascertain these reasons. To attempt to get beneath the reasons for every vote would have been beyond the scope of this study. After determining which roll calls met the defini- tion of "partisan," each legislator's votes were counted to determine the number of times he voted with the major- ity of his party. The legislators of each of the eight chambers under study were then placed in rank order by the highest party regularity to the lowest and separated 24 at the median point of partisan support (or as close to it as possible) into low partisan Democrats or high parti- san Democrats, low partisan Republicans or high partisan Republicans. The results were as follows: California Assembly: low Democrats - 18 low Republicans 23 high Democrats - 19 high Republicans 19 California Senate: low Democrats - 10 low Republicans 9 high Democrats - 10 high Republicans 11 New Jersey Assembly: low Democrats — 13 low Republicans 17 high Democrats — 7 high Republicans 21 New Jersey Senate: low Democrats — 2 low Republicans 8 high Democrats — 6 high Republicans 6 Ohio House of Representatives: low Democrats — 21 low Republicans 47 high Democrats - 21 high Republicans 48 Ohio Senate: low Democrats — 6 low Republicans 11 high Democrats - 6 high Republicans 11 25 Tennessee House of Representatives: low Democrats — 39 low Republicans — 11 high Democrats — 39 high Republicans - 10 Tennessee Senate: low Democrats — 13 low Republicans — 3 high Democrats - 14 high Republicans - 3 The method followed in this study is a simple one. Low and.high partisan legislators of each chamber of each state are compared according to independent var- iables already described. Many legislators were found to show a contradic— tion between their partisan attitudes as expressed in the interview data presented in Chapter III and the ex— tent to Which they voted with their party on roll calls. Therefore it was decided that the findings for each in- dependent variable tested would be presented in brief summary form. A lengthier treatment summarizing the results of the various hypotheses and noting their in- terrelations is reserved for the final chapter where a theory of partisanship in state legislatures is presented. There have been several attacks on roll call studies in recent years, some of them justified. It is pointed out by critics that bills are not given weights but rather all roll calls are treated as if they were of 26 equal importance--which they are not. Other criticisms are that there can be many roll calls on a single piece of legislation and that agreement on certain bills by both parties on the final roll call vote for the record does not make these bills less partisan. The significant partisan votes, thOugh, may have occurred in unrecorded votes in the Committee of the Whole. It is also claimed that even though a vote is finally a partisan one accord- ing to the definition set forth here, there can be other factors that may be more crucial than the party consider— ation in explaining the outcome.37 While this writer recognizes all these limitations to the use of roll calls, there are still benefits to their use in providing another dimension to political behavior. Perhaps taken alone these roll calls might not be very revealing, but combining these data with other types of data enriches our knowledge as long as we recognize the limitations. 37Wilder Crane, "A Caveat on Roll Call Studies of Party Voting," Midwest Journal of Political Science. IV (1960), pp. 237-249; Fred I. Greenstein and Alton F. Jackson, "A Second Look at the Validity of Roll Call Analysis," Midwest Journal of Political Science, VII (May 1963). pp. 156-166. 27 Also, the dividing of members in each chamber Of each state into two categories (high party voters and low party voters) might have tended to squeeze too wide a spread of party voters into too few categories. To test whether this was happening, I tested the hypotheses by dividing the legislators in each chamber into three cate- gories of party voting to see if by creating a middle category 13 would sharpen the more extreme categories of high and low party voters. The results showed no sig- nificant differences from those of our original procedure and so this study presents only two categories of party voting. Also it should be noted that fashioning two cate- gories may be distorting where, as in the New Jersey As- sembly, 94% to 100% was the range of low to high party voters. I assume this explains why New Jersey differed on many variables from the other states where party co— hesiveness was not as prevalent. DeSpite these reservations, I felt enough find- ings of significance were made to have proven this study worthwhile. 28 Summary This chapter began with a review of the literature which stressed the central focus of this study: the ex- planation of voting in legislative bodies. Less attention was paid to the traditional literature of the legislative process than to behavioral studies since from these this study drew both data and concepts. This study makes use of the interview data gathered in 1957 in the Four State Legislature Study conducted by Professors Heinz Eulau, John C. Wahlke, William Buchanan, and LeRoy C. Ferguson. TO this interview data, partisan roll call analysis for the 1957 session of these four legislatures was compared. The definition of partisan voting used was any roll call vote in which the majority of one party differed from the majority of the other party. Partisan voting will be compared with the follow— ing categories Of political behavior: 1. Partisan identification. 2. Interest groups as a factor in party voting. 3. Legislative leadership. 4. Constituency factors in partisan voting. 5. Political culture of legislators' districts. 29 6. Legislative strategy. 7. Role theory. 8. Career expectations of legislators. 9. Personal factors. The chapter concludes with a discussion of the limitations of roll call analysis as applied in this study. CHAPTER II THE POLITICAL SETTING Since this study is concerned with political party systems in four different states, I have attempted to skirt some of the problems of defining political par- ties or of evaluating the contributions political parties make to society in general. I have briefly summarized what one text on the subject lists as the functions of political parties: 1. Parties provide the vehicle leading to the nomin- ation of candidates and mobilization of support for the nominees. 2. Parties perform a "brokerage“ or mediating func— tion by helping to work out compromises and adjustments of the diverse demands of groups in society. 3. Parties serve as organizers of public Opinion and translators of diverse Opinions into public policy. 4. Parties help define legitimate areas of political controversy and proper techniques of political influence. 3O 31 5. Parties are agencies for the management and Operation of the government. The carrying out of these functions of political parties at the level of state legislatures is most often impeded by institutional and structural arrangements such as malapportionment, staggered elections, separation of powers (which makes possible divided control of the exec- utive and legislative branches of government), and strong committee systems. Apart from these institutional—structural limita— tions, the functioning of the political party on the state legislative level is also impeded by demands of peOple in a legislator's district, demands of pressure groups, and by the development of traditions and norms within the legislative body itself which tend to limit the functioning of the political party in the legisla- tive process. Because of these limitations, it is necessary to begin with an understanding of the party systems of the four states dealt with in this study. I will begin 1 . . Austin Ranney and Willmore Kendall, Democracy and the American Party System (New York: Harcourt, Brace and Company, 1956), pp. 85-87. 32 by reviewing the classification of partisanship within the four states and then in turn describe political ac- tivity and traditions and relevant institutional charac- teristics in each state. Classification In each of the fifty American states there is a party system that performs some of the functions of political parties described above. There are great var- iations however in the patterns of the political prac- tices and organization. Most attempts at classifying state party systems have concentrated on attempting to measure the extent to which state party systems deviate from a competitive two party pattern. V. 0. Key has developed a rough measure of the extent of two party competition by arranging states according to the number of years of Democratic control of the governorship during the period of 1929-1956. Tennessee was included among the twelve states that had an unbroken series of Democratic governors while California was listed among the states where Democratic gubernatorial victories were infrequent. New Jersey 33 was grouped with six other states where Republicans held the governorship for slightly over half the period and Ohio was one of the twelve states where control of the governorship between the two parties was approximately equal during this period.2 Ranney and Kendall classified the forty—eight state systems according to degree of competitiveness by combining electoral results in each state for the offices of President, senator and governor for the period from 1914—1954. Under their scheme California, New Jersey, and Ohio were included among the twenty-six party systems that were classified as "two party" on the basis that the second party had won more than 25% of all elections for the three Offices during the period. Tennessee was typed among the twelve states with a "modified one-party" sys- tem. The criticism by Joseph A. Schlesinger of the above procedure for classifying party competition is that combining votes for the several offices results in an overestimation of the degree of competition in the 2 . . . V. 0. Key, Jr., Politics, Parties and Pressure Groups (New York: Thomas Y. Crowell Company, 1958), p. 313. 3 . Op. Cit., pp. 161-64. 34 states. He has suggested a second dimension for measur— ing competitiveness, namely the rapidity with which the parties alternate in their control of an Office. Schles- inger's scheme has combined a measure of party control of the Office of governor for the period 1870—1950 with a measure of the cyclical character of that competition. Under his classification system one party is strongly dominant in most states and that even in the more com— petitive states there is a low rate of alternation be- tween the parties. In dividing the states into categories according to this two dimensional scheme, Schlesinger includes Ohio among the nine "competitive" states in which the minority party won at least a third of the elections during the period, and there was a relatively high degree Of alternation between the parties. New Jersey is classified as "one party cyclical" (Democratic) among the eight states where one of the parties had an overall predominance in the control of the governorship but in which the minority party has been able to win short periods of control. California (Republican) and Tennessee (Democratic) were in the category Of "one party predominant" along with fourteen other states where 35 minority parties won occasional victories in the guber- natorial races, but failed to succeed themselves very often. The phenomenon of divided partisan control of state governments tends to limit the value of measuring party competitiveness by control of the governorship. V. 0. Key suggests that the lesser party in a state may expect to be more successful in electing governors than in winning legislative majorities. As an illustration of this, Key has grouped states according to average per— centage Of Democratic strength in the lower houses for the period 1924—1956. In this grouping, Tennessee was included among the states where the Democratic percentage of legislators averaged about 90% for the entire period. New Jersey was among twelve states grouped as "strong Republican," while California and Ohio were in an inter- mediate group which leaned toward Republican control, 4Joseph A. Schlesinger, "A Two—Dimensional Scheme for Classifying the States According to Degree of Inter- Party Competition," American Political Science Review, XLIX (1955), pp. 1120-28; also "The Structure of Compe- tition for Office in the American States," Behavioral Science, V (July 1960), pp. 197-210; and Robert T. Go- lembiewski, "A Taxonomic Approach to State Political Party Strength," Western Political Quarterly, XI (1958), No. 3, pp. 494-513. 36 although the Republicans might occasionally win control of the lower house of the legislature.5 Another source of information regarding the re- lationship between party systems and state legislatures is the report of the American Political Science Associa- tion Committee on State Legislatures. This report indi- cated that New Jersey was one of seventeen states with a strong degree of party spirit or cohesion in the legis— lature; Ohio was among the eleven states where party cohesion was moderately strong; and California and Ten— nessee were in the group where party cohesion was weak or nonexistent. Although there is some variation in the results achieved by these various classification schemes, they do give us some hint as to how the legislators perceive the importance of political parties in each state. We will now consider some of the political background of each Of the four states under study. 5Key, Op. cit., p. 314-15. 6 . . Belle Zeller, ed., American State Legislatures (New York: Thomas Y. Crowell Company, 1954), Chapter XII. 37 California California has a long tradition of nonpartisan- ship dating back to the Hiram Johnson era (1911) but has over a period of ten or fifteen years been develOping a competitive two party system. It selects a number of separately elected executive officials who like the leg- islators are elected on a party ballot. In recent years the major parties have split control of such Offices and the governorship. The tremendous influx of people from other states during and after World War II appears to have been a major factor in increasing the Democratic vote and making partisanship respectable within both parties. One im- portant result of this trend has been the elimination of cross—filing, California's unique institution which permitted candidates of one party to file in the other party's primary and take their seats unopposed if they win both. This system was somewhat weakened in its effect by a provision adopted in 1952 requiring a candidate's party affiliation to appear beside his name on the ballot. In previous years one—half to three—fourths of the legis- lators won their seats by cross-filing, but only a fourth 38 of the 1957 legislature had done so, and these were, for the most part, either very senior incumbents or represen— tatives of one—party districts. Cross-filing was finally abolished altogether in 1959 when the Democrats gained control of both the legislature and the governorship.7 The party function of narrowing the field of po- tential candidates and campaigning for their election is performed by local and state Democratic and Republican Assemblies. These are independent groups of partisans who have organized to control their respective parties. The apportionment of legislative seats appears to have a less inhibiting effect on party competition than it does in many states. The distribution of Senate seats gives preponderant power to northern, inland areas at the expense of southern California and the Bay cities, but the Assembly seats have been reapportioned decennially to keep them approximately representative of the rapidly growing and shifting population.8 Democrats hold a size- able number of seats in rural as well as urban areas. 7William Buchanan, California Legislature: Struc- ture and Practices, Unpublished working paper, 1958; also Legislative Partisanship: The Deviant Case of California (Berkley: University Of California Press, 1963). 8Ibid. 39 The tradition of nonpartisanship was reflected in the organization of both the Senate and the Assembly at the time of this study. There were no formal party Officers in the Senate. That body was run by the Presi- dent pro tempore and four other members of the Rules Committee who were selected by the entire body. By cus- tom the minority party is represented by two members on the Rules Committee regardless of its strength. In 1957, the Senate seats were divided equally between the parties, but two Republican members voted for a Democratic Presi- dent pro tempore, giving the Democrats a 3-2 majority. Committee posts were rather evenly divided between the two parties. Although the Republicans held a majority of the Assembly seats in 1957, the leadership system of the body was based on a bi-partisan "Speakers' Coalition.“ This group, which supported a moderate Republican for the Speakership in 1957, consisted of more Republicans than Democrats, more Southern Californians than north- erners, and most of the rural bloc. The Speaker appointed committees and chairmen, and generally dominated the As— sembly. The Majority Floor Leader was not considered a 4O partisan Officer, but the Speaker's "personal represen- tative on the floor." By contrast, the Minority Floor Leader was a partisan officer, although his choice was influenced by the Speaker and, in fact, he represented the Democrats within the coalition. Coalition members held all important chairmanships and dominated the com- mittees. A Democrat was chairman of the Rules Committee and Democrats held 9 of 24 other chairmanships. In the past, neither the formal nor informal state party organizations took much interest in the Cal— ifornia Legislature so far as program was concerned, but in 1957 the state Democratic organization provided an experienced, salaried secretary to coordinate the ef— forts of Democratic legislators. Democratic Assemblymen and 8 or 10 Senators held a weekly caucus to discuss strategy. A few members consistently failed to attend. Republicans held similar conferences, but there was no evidence that either Democratic or Republican caucuses were considered binding, even on those attending.1 9Ibid. lOIbid. 41 New Jersey Politics in New Jersey is intensely partisan and politicians have little use for independents. However, the differences between parties at the time of this study were more organization— than policy-inspired. Policy issues were often used to try to improve the position of the party organizations. By getting on the “right" side of an issue it was hoped that the party would get credit for supporting popular legislation.11 The main strength of both parties in New Jersey lies in the county organizations, and there is wide var— iation among the counties in the nature of party member- ship, organization, and policy orientation. South Jersey Republicans, whose center of power is the smooth-working Atlantic City machine, tend to be conservative. In the northern counties such as Bergen and Pasaaic, there are many "modern" Republicans, and even some affiliated with the AFL-CIO. While for many years the Democratic Party was little more than a branch of Frank Hague's Hudson County 11 . . LeRoy C. Ferguson, New Jersey Legislative Notes, Unpublished working paper, 1958; Murray Frost, Roll Call Cohesion in the New Jersey Legislature, 1956—1957, Un- published Masters Thesis, Michigan State University, 1959. 42 machine, Governor Meyner's overwhelming victory in 1957 with very little support from Hudson County demonstrated the emergence of a state-wide Democratic Party. Certain former Republican strongholds such as Camden and Mercer counties are now considered safely Democratic. At the time of this study, the Republicans controlled both houses of the legislature, but in the election of 1957 the Democrats gained control of the Assembly for the 1958 session. The apportionment Of seats in the Senate (one senator from each county) made it unlikely that rising Democratic strength would be reflected in that body. Neither party has worked out any satisfactory means of coordinating the activities of party members in the legislature with interests of the state party organization. This is particularly difficult for the party which does not control the governorship. At the time of this study, the Republican Party leadership had almost complete control of all legisla— tive activity in both houses of the legislature even though a Democrat, Robert Meyner, was governor. Poli— cies were worked out in the majority party caucus. 43 Conferences were held before each daily legislative ses— sion. This system gave the majority party members a good Opportunity to become informed about and discuss all pending legislation. Minority party members had little opportunity to contribute to the work of either house and were not well informed about pending legisla- tion. It was quite common for Democrats in the Assembly to request a recess so that they could discuss the bills that the majority conference was going to bring to the floor. This changed in 1958 when the Democrats gained control of the Assembly and exercised the same tight con— trol over legislation that the Republicans had previously. While party discipline in New Jersey appeared at times to be strict, there was no evidence that legis- lators considered this particularly burdensome. Many legislators were gratified to receive guidance from the party in policy matters and to receive from party disci— pline a degree of protection from interest and district pressures. 2Ferguson, Ibid. 44 Ohio At the time of this study the Republicans had strong well disciplined majorities in both houses of the Ohio Legislature. The Republican state organization was well—knit and influential.13 In the Legislature, the Republicans held regular caucuses on policy matters and the decisions of these caucuses were binding on individ- ual members unless their objections were accepted by party leaders. So strong was the majority that the party leaders could afford to allow individual members to deviate from the party position when they felt it was a matter of conscience or constituency Opinion. The Re- publican unity in the Ohio Legislature was also strong enough to bridge the gap between the rural and metropol- itan legislators, thus preventing rural—urban factionalism. The Democratic Party on the other hand was beset by factionalism and its state organization was weak. Dur- ing the period of this study there were three factions among the Democratic legislators, a reflection of the same division in the state. There were the pro-Lausche 3 . . . Heinz Eulau, Institutional Parameters for Leg- islative Behavior in Ohio, Unpublished working paper, 1958. 45 Democrats (Lausche had served five terms as Governor be- before being elected Senator in 1956, an office he held at the time of the study), conservative in attitudes and coming most often from rural, down-state counties, the anti-Lausche Cleveland-based group which was powerful enough to get one of its members elected minority leader in each house, and finally the loosely knit "New Deal” faction whose leader in the state was Michael V. DiSalle of Toledo. This last group resented the leadership given the Democratic members Of the Cleveland group.14 The Democrats appeared to be a very ineffective opposition in the 1957 session. Rather than playing the role of the militant minority, they were more likely to play down criticism of the powerful majority in order to gain approval of some of their own legislation. Democrats in each house caucused on policy matters, but few con- sidered themselves bound by caucus decision. In both houses of the Ohio Legislature the major— ity party leadership was able to exercise tight control over all activity. The Speaker of the House was consid- ered to be the formal leader of the Republican majority l4Ibid. 46 in addition to being the presiding Officer of the House. He was assisted by the Speaker pro tem and was the center of the Chamber's activity. He was chairman of the Rules Committee and controlled the calendar and all legislative business. He appointed all committees and designated their chairmen. The President pro tem of the Senate en— joyed the same type of control over the Senate. He, too, was chairman of the Rules Committee and controlled the flow Of business. Both the Speaker of the House and the President pro tem of the Senate worked closely with the Republican Governor on legislative business. This har- mony between both branches of government made public pol— icy in Ohio in 1957 an almost complete Republican affair.15 At the time of this study, Mr. William O'Neill, a Re- publican, was Governor of Ohio. Tennessee As had been noted earlier, Tennessee is usually designated as having a "predominantly“ one-party system. existence of a Republican minority which exercises con— siderable power in the eastern part of the state. 15Ibid. 47 distinguishes Tennessee from one—party states of the deep south where the Republican Party is almost non- existent. Also, in Tennessee, the Democratic Party is "bi-factional" in contrast to the multi-factionalism which exists in some of the other one-party states. V. 0. Key has suggested that the existence of a Repub— lican Opposition contributes to the creation Of one tightly organized Democratic faction.l6 At the same time, this cohesive faction generates within the Demo- cratic Party an opposition group, producing a bi-faction- alism within the dominant party. The parties, as such, function very little in the Tennessee Legislature. Just before the opening day of a session, the Democrats caucused in order to agree upon a slate of officers for the chambers and also for constitutional officers who must be elected by the leg- islature (in joint session) in the Opening days——Secre- tary of State, Treasurer, and Comptroller of the Treas- ury. However, the choice of the Democratic nominees 1 . . . 6V. 0. Key, Jr., Southern Politics in State and Nation (New YOrk: Knopf, 1949), p. 300. See also Ran— ney and Kendall, Op. cit., pp. 194-197. 17John C. Wahlke, A Note on the Habitat and Habits of Tennessee Legislators, Unpublished working paper, 1958. 48 for such posts was more an administration choice than a party choice. The governor is usually able to get the caucus to accept his nominees for the various positions, including those for the speakership and for floor leaders. Beyond this, the Democratic caucus usually is dormant. The Republican members also nominate candidates for cham- ber and constitutional Offices at their first caucus meeting but unlike the Democrats, they hold meetings a few times during the session and the party takes a stand on legislative measures. In practice there was little party solidarity among either party and even the Repub- licans did not feel bound to vote together after a caucus. The Governor at the time of this study was Frank Clement, a Democrat. The administration floor leader was picked by the governor and was not considered a party agent in any sense. He presented the administration side of measures supported by the governor, and was also looked upon by the members for many purposes as an agent for the whole house. He was looked to for motions to adjourn, to alter procedures in certain cases, and to shape the debate on each succeeding bill. The Republican floor leader was 49 more of a party agent in that he was designated by the party caucus. But he did very little "leading" on the floor, since most bills do not divide Democrats and Re- publicans. Opposition to administration measures was as likely to be led by any member, regardless of party who was interested in passing (or defeating) the bill in question. It can be said that most activity of the Ten- nessee Legislature took place without political parties being a relevant factor.18 §11_mna_rx The foregoing information reveals that there is considerable variation among the four states of this study in regard to political structures, traditions, practices, and situations. The four states represent different degrees of party discipline and competition. ~We would have reason to expect that perceptions of party role by the legislators in Tennessee would differ from that in New Jersey and Ohio and that there would be sim- ilarity between these two states. In California, party 18Ibid. 50 would be perceived as being more salient for legislators of that state than for those in Tennessee, but probably less so than in New Jersey and Ohio. Because of the long period prior to this study in which the Republicans dom- inated the legislatures of New Jersey and Ohio, we might expect that the political party would be more important for the majority than for the minority legislators. In California where there was a greater non-partisan approach to legislative matters, we might expect Democrats to be more partisan. In Tennessee we might expect more intra party differences than inter party differences. CHAPTER III LEGISLATIVE ATTITUDES TO PARTISANSHIP AND LEGISLATIVE VOTES One of the important pioneers in the technique of submitting roll call votes in legislatures to Guttman scaling is George M. Belknap. In his study of the United States Senate, he took the votes on key labor issues to determine the standings of Senators in their attitudes toward labor.1 The Guttman scale is designed to measure one single attitude at a time. This concept of Guttman scal— ing is known as unidimensionality. Members of a given population are placed in order with regard to some set of attitudes or behaviors; in our case the set of atti— tudes is that of political partisanship. In a perfect Guttman scale it would be possible to reproduce the exact 1George M. Belknap, "A Method of Analyzing Legis— lative Behavior," Midwest Journal of Political Science, II (1958), pp. 377—402. 51 52 pattern of responses to each item in the scale from a single score. A perfect 100% coefficient of reproduci— bility is seldom attained and Guttman set a 90% repro- ducibility for a satisfactory scale. Such a scale would have errors in only 10% of reSponses to individual scale items. (An error occurs when a respondent gives a posi— tive response to one or more different items on the scale and a negative response to an earlier item.) In addition to the criterion of 90% reproducibility for the entire scale, each individual item should have not more than 15% error if it is to be included in the scale. Legislators were asked a series of questions about their attitudes toward political parties. These items were then combined.ir1tflmafour states to form a partisanship scale in which those favorable or opposed to party action in legislature would be at different extremes. The party scale items and total scores are pre- sented in Table 2. These four scale items were combined into a five point scale that was used to score each individual leg- islator according to the evaluation of party in his re- sponses to the four items. The distribution of these 53 scale scores by state and by party is shown in Tables 3 and 4. In order to interpret these tables, it will be helpful to remember the following: 'Within the 10% mar- gin for error in the scale, we can say that the legis- lators with a scale score of “1" gave a negative response to all of the party evaluation items, even to the extent Of thinking it best to eliminate party labels in state legislative elections. Those who received scale score "2" would retain party labels, but felt that every indi— vidual should take an interest in government directly, not through a political party. In number "3" are people who felt that parties had some usefulness, but who would not vote with their parties if it would mean loss of sup- port in their districts. The legislators assigned to category "4" considered maintaining the party's record more important than losing support in their districts, but saw no need for the parties to take clear-cut Oppos- ing stands on more of the important state issues. Fin- ally, those assigned to scale score "5“ made a positive response to all four of the party evaluation items in the scale. 54 TABLE 2 PARTY EVALUATION SCALE ITEMSl Agree Undecided Disagree Total l. The best interests of the people would be better served if leg- islators were elected without party labels. 22% 1% 77% 100% 2. Under our form of N _ 468 government every in- dividual should take an interest in gov— ernment directly, not through a polit- ical party. 47% -- 53% 100% N = 454 3. If a bill is important for his party's record, a member should vote with his party even if it costs him some sup- port in his district. 46% 2% 52% 100% N = 452 4. The two parties should take clear-cut, Oppos— ing stands on more of the important state issues in order to en— courage party respon— sibility. 29% 1% 70% .100% N = 460 l . . LeRoy C. Ferguson, Partisan Roles of State Legisla— tors--A Working Paper for the State Legislative Research Project, Michigan State University, 1959, pa 32. 55 TABLE 3 PARTY EVALUATION SCALE SCORES BY STATE AND PARTY Low High 1 2 3 4 5 INDEPENDENTS PARTY MEN Weak partisans...Strong California Total (N=108)* 34% ll 27 8 20 100% Democrats (N= 50) 20% 14 32 6 28 100% Republicans (N= 58) 46%, 8 23 ll 12 100% New Jersey Total (N= 79) 5% 8 35 41 11 100% -Democrats (N= 27) 11% ll 44 30 4 100% Republicans (N= 52) 2% 6 31 46 15 100% Ohio Total (N=159) 6%1 12 19 44 19 100% Democrats (N= 49) 14% 16 23 29 18 100% Republicans (N=110) 2% 11 17 51 19 100% Tennessee Total (N=111) 18% 29 27 21 5 100% Democrats (N= 89) 19% 27 26 24 4 100% Republicans (N= 22) 14% 36 32 9 9 100% FOUR STATE TOTAL (N=457) 15% 15 26 29 15 100% *The "N" consists of the total number in each category for whom it was possible to assign scale scores. There were forty-seven legislators who either were not interested or did not respond to enough of the partisan scale items to be scored. The number assigned scales scores were Calif- ornia, 108 (90%); New Jersey, 79 (100%): Ohio, 159 (92%); and Tennessee, 111 (84%). Source: Ferguson Working Paper, p. 28. 56 TABLE 4 DISTRIBUTION OF PARTY EVALUATION SCALE TYPES IN FOUR STATES Calif. N. J. Ohio Tenn. Total (N=108) (N=79) (N=159) (N=lll) (N=457) Independent Democrats 16% 8% 9% 37% 17% Independent Republicans 29 5 9 10 13 Weak Partisan Democrats 15 15 7 21 14 Weak Partisan Republicans 12 20 12 6 12 Strong Partisan Democrats 16 ll 14 22 16 Strong Partisan Republicans 12 :1 1:. :1 2e TOTAL 100% 100% 100% 100% 100% Source: Ferguson Working Paper, p. 30. 57 It is not surprising to find that New Jersey and Ohio legislators gave much higher evaluation to party than those in California and Tennessee. At the same time the party evaluation scores for the latter two states were closer together than we might have expected from what we knew about the party systems of those states. Another fact to be noticed is the tendency seen in Table 4 toward bipolarization in California. There was a considerably larger portion of California legis— lators in the lowest scale category than in any other state, while at the same time that state had the largest percentage of legislators in the highest scale category, even exceeding Ohio by a slight margin. This bi—polar tendency, which we note was partly a reflection of the differences between Republicans and Democrats, might be a reflection of the somewhat confused and transitional nature Of the party situation in California at the time of our study. We would expect that the index of party evaluation for California legislators after the 1958 election might be considerably higher than it was at the time Of this study. 58 The distribution of scale scores by party shows that Republicans gave a significantly higher evaluation than Democrats to party in both Ohio and New Jersey, and the data suggest that Democrats consider party more im- portant than Republicans in California and Tennessee. It would appear to be worth noting that in New Jersey and Ohio (and possibly in Tennessee) legislators who be- long to the majority and traditionally dominant party were likely to give a higher evaluation to party than those who were accustomed to being in the minority. In order to use the party evaluation scale to interpret other data on party perceptions in the legis— latures, the five groups representing different degrees of party evaluation were collapsed into three scale types. Accordingly scale scores "1'I and ”2” were combined to form a scale type which we labeled "independents.“ Even though it might seem wrong to use this term inasmuch as all legislators were either Democrats or Republicans, the label is justified by the fact that all of them felt that it was better for an individual to deal with the government directly rather than through a political party and one-half of them had taken the extreme position that 59 it would be better if legislators were elected without party labels. The "independents" made up 30% of the total group to whom scale scores were assigned, and since the remain- ing 70% all felt that it was better to work through a political party, they were labeled "party men." These "party men" were further divided into two groups--weak partisans and strong partisans. The weak partisans (scale score "3") differed from the strong partisans (scale score "4" and "5") in that they had said that they would not vote with their parties if it would cost them support in their districts. In addition to placing party above district, about one-third of the regular party men felt that parties should take more clear-cut, opposing stands on state issues. These groupings are indicated in Table 4. Almost one-half of the legislators in both Calif- ornia and Tennessee were "independents” but it should be noted that a higher proportion of California independents took the extreme position of being willing to abolish party labels. A majority of the legislators in both Ohio and New Jersey were strong party men, and while 60 New Jersey had a significantly higher percentage of weak party men than Ohio, the latter had a slightly higher proportion Of independents. The availability of the data on party evaluation attitudes for each of the four states makes it possible for us to test the relationship between the scale types constructed on the basis of interview data and the cate- gories established on the basis of the legislators' vote ing behavior during the session. It would appear logi- cal to presume that legislators who ranked "high" in party loyalty on the basis of roll call voting records would be more likely than those who were "low“ in party voting loyalty to have been classified as party ”regulars” on the basis Of their responses to the items in the party evaluation scale. As a corollary of this we might assume, logically, that legislators who were “low” in party voting loyalty might be more likely than other legislators to have been classified as "independents“ on the basis of their reSponses to the items in the party evaluation scale. In making these assumptions we would need to re- mind ourselves, however, that many other factors in addi- tion to a legislator's abstract attitude toward party 61 might have affected his record of party loyalty as it was reflected in actual roll-call votes. The subject matter of the particular bill being voted on, special problems in a legislatorTsconstituency, and varying de- grees of pressure exerted by party leaders are only a few of the intervening variables that might have affected this relationship. Nonetheless the hypothesis that party voting loyalty would be positively related to favorable attitudes toward party seemed a reasonable assumption. The results of this testing are shown in Tables 5, 6, 7, and 8. The hypothesis is given rather strong support by the data on California Democrats in that legislators high in party voting loyalty were much more likely than those low in party voting loyalty to have been classified as party ”regulars" on the basis of their party evaluation attitudes. The California Democrats who were ”low“ in party voting loyalty were, on the other hand, more likely to have been classified as ”independents” on the basis of the party evaluation scale. The data for the California Republicans, however, tend, if anything, to negate the hypothesis. It will be 62 recalled that more than one—half of the California Repub- licans were classified as "independents" on the basis of their attitudinal responses. These "independents” made up about equal proportions of the "high" and "low" cate- gories established on the basis of party voting records. There were only thirteen "regular" party men among the California Republicans, and nine of these were categor- ized as "low" in party voting loyalty. The differences between the two parties with re- gard to the relation of party voting to party attitudes is partly to be understood in terms of what we know about the political situation in California at the time of the study. While Republican legislators continued to reflect the traditional California attitude of independence or blurred lines of partisanship, the Democrats were begin- ning to place much more emphasis on party loyalty and party organization. The hypothesis also is rather clearly negated in the case of the Tennessee Democrats (Table 6) in that these legislators who were "low" in party voting loyalty were much more likely than those who were "high" to have been classed as "independents" on the basis of their 63 attitudinal responses, while the "party men" made up relatively higher proportions of the "low" party voting category. There were too few Tennessee Republicans for an adequate analysis, but the data suggest that the re- lationship went in the same direction in that party. In New Jersey (Table 7) there was no discernible relationship between party voting and attitude toward party among the Democrats, but for the Republicans there was a marked positive correlation between ”high" party loyalty in voting and a favorable attitude toward party regularity expressed in the interview. The strongest support for the hypothesis is found in the data for the state of Ohio (Table 8) where we note that in both parties the "high" party loyalty voting cate- gOry included a much higher proportion of "regular“ party men than the "low" voting category. The "low" voting categories in both parties in Ohio contained relatively higher proportions of both "moderate" party men and in- dependents." In summary, it appears that in four of eight pos- sible cases (California Democrats, New Jersey Republicans, and both parties in Ohio) the hypothesis that there would 64 be a positive relation between party loyalty in voting and favorable attitudes toward party expressed in the interview was sustained. In one case (Tennessee Demo- crats) the hypothesis was rather clearly refuted, and in two cases (California and Tennessee Republicans) the data suggest refutation. The case of the New Jersey Democrats is inconclusive. 65 TABLE 5 RELATIONSHIP OF PARTISAN VOTING TO PARTISAN ATTITUDES IN CALIFORNIA CALIFORNIA Independents weak. Strogg Total Partisans Partisans House low Dems 5 5 5 Senate low Dems 5 3 1 TOTAL LOW DEMS 10 43% 8 35% 6 22% 100% House high Dems 6 5 8 Senate high Dems 1 3 4 TOTAL HIGH DEMS 7 26% 8 30% 12 44% 100% House low Reps 12 3 8 Senate low Reps 3 2 1 TOTAL LOW REPS 15 52% 5 17% 9 31% 100% House high Reps 9 6 4 Senate high Reps 8 2 0 TOTAL HIGH REPS 17 59% 8 27% 4 14%. 100% 66 TABLE 6 RELATIONSHIP OF PARTISAN VOTING TO PARTISAN ATTITUDES IN TENNESSEE _. w TENNESSEE Independents weak. Strogg Total Partisans Partisans House low Dems ll 12 11 Senate low Dems 5 3 4 TOTAL LOW DEMS 16 34% 15 33% 15 33% 100% House high Dems 18 4 9 Senate high Dems 7 4 1 TOTAL HIGH DEMS 25 58% 8 19% 10 23% 100% House low Reps 4 4 0 Senate low Reps l l 1 TOTAL Low REPS 5 45% 5 45% l 10% 100% House high Reps 6 2 1 Senate high Reps 0 0 2 TOTAL HIGH REPS 6 55% 2 18% 3 27% 100% RELATIONSHIP OF 67 TABLE 7 PARTISAN VOTING TO PARTISTAN ATTITUDES IN NEW JERSEY NEW JERSEY Independents Weak. Strong Total Partisans Partisans House low Dems 3 5 5 Senate low Dems 0 2 0 TOTAL LOW DEMS 3 20% 7 47% 5 33% 100% House high Dems 2 4 1 Senate high Dems 1 l 3 TOTAL HIGH DEMS 3 25% 5 42% 4 33% 100% House low Reps 3 9‘ 5 Senate low Reps 0 2 6 TOTAL LOW REPS 3 12% 11 44% 11 44% 100% House high Reps 1 4 16 Senate high Reps 0 1 5 TOTAL HIGH REPS 1 4% 5 18% 21 78% 100% RELATIONSHIP OF 68 TABLE 8 PARTISAN VOTING TO PARTISAN ATTITUDES IN OHIO === Weak Strong OHIO Independents Partisans Partisans Total House low Dems 6 6 7 Senate low Dems 2 l 3 TOTAL LOW DEMS 8 32% 7 28% 10 40% 100% House high Dems 4 4 10 Senate high Dems 3 0 3 TOTAL HIGH DEMS 7 29% 4 17% 13 54% 100% House low Reps 5 10 27 Senate low Reps 3 3 4 TOTAL LOW REPS 8 15% 13 25% 31 60% 100% House high Reps 6 4 37 Senate high Reps 0 2 8 TOTAL HIGH REPS 6 10% 6 10% 45 80% 100% CHAPTER IV CONSTITUENCY AND PARTISAN VOTING BEHAVIOR Recently much has been written Showing the rela- tionship of constituency to voting behavior. It has been generally hypothesized that both the party competi- tiveness, social composition, and the urban-rural nature of a district are crucial variables in explaining legis- lative voting behavior. Duncan MacRae was among the first researchers to relate constituency characteristics to party voting. In his study of the Massachusetts House of Representatives he found that legislators from "safe" districts tended to vote with their party more often than representatives from "close" districts,2 and that districts whose compo- sition was different from the party average tended to elect representatives more independent of party than others. Duncan MacRae, Jr., "The Relation Between Roll Call Votes and Constituencies in the Massachusetts House of Representatives," American Political Science Review, XLVI, NO. 4 (December, 1952), p. 1049. 69 70 David Truman in a study of constituency and vot- ing studied the voting cohesion within state delegations of 13 states in the U. S. House of Representatives. Briefly stated, he found that a high degree of vote co- hesion occurred among Representatives from the same state. He suggests that a Representative's colleagues from the same state are a source Of guidance and reassurance in determining a Representative's vote on controversial issues. Frank Sorauf in his Study of the Pennsylvania Legislature makes the point that party and constituency tend to be similar in that State. Unlike Wisconsin where Republicans tend to be the rural and small city party and Democrats the urban big city party, Pennsylvania has a number of small mill communities controlled by Democrats aS‘well as districts in the big cities of Philadelphia 4 and Pittsburgh. 3David B. Truman, "The State Delegations and the Structure Of Party Voting in the United States House of Representatives," American Political Science Review, L (1956), pp. 1023-1045. Frank J. Sorauf, Party and Representation: Leg- islative Politics in Pennsylvania (New YOrk: Atherton Press, 1963), pp. 43-61. 71 Competitive Nature of the District The Four State Legislative Study classified leg- islative districts in three of the four states as "one- party," "semi-competitive,” and "competitive." Tennessee was excluded because there were an insufficient number of legislators from semi-competitive (10%) and competitive (8%) districts. The results of correlating partisan at- titudes with the competitive nature of the district are found in Table 9. In interpreting the following table it should be pointed out that in both Ohio and California, Democratic legislators were more likely to come from competitive districts than Republican legislators. In New Jersey, Democratic legislators were more likely to come from one- party districts than Republican legislators. The table shows that competitiveness of district appears to be re- lated to degree of partisanship, but oppositely for the two parties. There are a higher proportion of strong partisans from competitive districts than from one party or semi-competitive districts among Democrats. But the reverse is true for Republicans. Stated in another way and considering the two parties separately, strong parti- sans made up a Significantly smaller proportion Of the 72 Democrats who came from one party districts, and a larger proportion of those who came from semi—competitive and competitive districts. Independent and moderate Democrats were significantly less likely to come from the competitive than from one-party districts. Among the Republicans, strong partisans were significantly more likely to come from one-party districts than from semi—competitive and competitive districts. A possible explanation of the difference between the two parties with respect to the relationship of attitud- inal partisanship to competitive nature of the district is that the Democrats have been traditionally in the legisla— tive minority in the three states involved. Perhaps Democrats from one-party districts assume weaker partisan attitudes be- cause they see no opportunity for their party to plan an im— portant role in the legislative process, while the Republi- cans from one—party districts assume stronger partisan atti- tudes because they are accustomed to being in control of the legislature. On the other hand, a Democrat in a competitive district has to assume a stronger partisan role even in run- ning for the legislature if he knows that his party will probably be in the minority. This would not necessarily be the case for the traditionally dominant Republicans. RELATION OF COMPETITIVENESS OF DISTRICT TO 73 TABLE 9 PARTISAN ATTITUDES Independent Weak Partisans Strong Dems Reps Dems Reps Dems Reps Total CALIFORNIA (N= 46) One- Party 17% 29 13 13 13 15 100% (N= 38) Semi- competitive 8% 44 8 16 8 16 100% (N= 25 ) Competitive 25% 8 29 4 34 -— 10Q% NEW JERSEY (N= 20) One- Party 15% —- 35 10 -- 40 100% (N= 20) Semi— competitive -- 5 5 25 15 50 100% (N= 39) Competitive 8% 8 10 23 15 36 100% OHIO (N= 64) One- Party 2%> 8 5 12 3 70 100% (N= 51) Semi— Competitive 10% 14 2 16 14 44 100% (N= 44) Competitive 20%» 5 l6 7 32 20 100% THREE STATE TOTAL (N=130) One- Party 9% 14 12 12 6 47 100% (N=109) Semi- competitive 7% 23 5 17 12 36 100% (N=107) Competitive 17% 7 17 12 26 21 100% Source: LeRoy C. Ferguson,WMmking Paper, op. cit., p. 32. 74 Let us now consider the relationship between partisan voting behavior and the competitive nature of the district. (Table 10) COMPETITIVE TABLE 10 IN CALIFORNIA NATURE OF DISTRICT AND PARTY VOTING *— V- One Semi Compet- Total Party Compet- itive itive House low Dems 7 4 7 18 Senate low Dems _2_ _9_ _§_ .19 Total 14/50% 4/13% 10/33% 28/100% House high Dems 9 4 6 19 Senate high Dems ._2 ._2 ._6 llg Total 11/38 6/21 12/41 29/100 House low Reps 10 10 3 23 Senate low Reps _§_ _3 ._Q ._2 Total 16/50 13/41 3/ 9 32/100 House high Reps 9 10 O 19 Senate high Reps _§_ _6_ ‘_9 .li Total 14/47 16/53 0 30/100 55/46% 39/33% 25/21% 119/100% In California we find the Republicans bearing out the findings of the Four State Study slightly more than the Democrats. As can be seen in Table 10, none of the high 75 party Republicans come from competitive districts while at least a few of the low party voting Republicans did. Among the Democrats there were slightly more low party voting Democrats from one—party districts and slightly more high voting Democrats from competitive districts. TABLE 11 COMPETITIVE NATURE OF DISTRICT AND PARTY VOTING IN NEW JERSEY fI—i 14L; One Semi Compet- Total Party Compet- itive itive House low Dems 4 3 6 13 Senate low Dems _1_ _Q_ ._1 ._2 Total 5/33% 3/20% 7/47% 15/100% House high Dems 5 2 O 7 Senate high Dems _Q_ _l_ ._4 _J§ Total 5/42 3/25 4/33 12/100 House low Reps 2 5 10 17 Senate low Reps _3_ _2 _§_ _8 Total 5/20 7/28 13/52 25/100 House high Reps 4 6 11 21 Senate high Reps ._l _43 _2_ _Q_ Total 5/19 9/33 13/48 27/100 20/25% 22/28% 37/47% 79/1oo% In New Jersey the low partisan voting Democrats are more likely to come from competitive districts than from one-party districts, while more high partisan voting 76 Democrats come from one—party districts. however, is too small to be significant. The difference, Among Republi— cans we find no significance at all in that both high party voting and low party voting Republicans tend to come mostly from competitive districts more than from one-party or semi-competitive districts. COMPETITIVE TABLE 12 NATURE OF DISTRICT AND PARTY VOTING IN OHIO One Semi Competi— Dist. Total Party Competi- tive Domin. by tive Opposite Party House low Dems 3 3 l3 2 21 Senate low Dems _Q_ _3. _3. g_ ‘_6 Total 3/11 6/22 16/60 2/7 27/100% House high Dems O 2 18 1 21 Senate high Dems _Q' _5 _1 .9 _6 Total 0/0 7/26 19/70 1/4 27/100% House low Reps 17 17 11 2 47 Senate low Reps _§_ _§_ _§_ .9 .11 Total 20/34 22/38 14/24 2/4 58/100% House high Reps 31 14 3 O 48 Senate high Reps _1_ _4_ _Q_ 9_ ll Total 38/64 18/31 3/ 5 0/0 59/100% 61/36% 53 /31% 52/30% 5 /3% 171/100% 77 Ohio Republicans bear out the findings of the Four State Study more than do the Democrats. The over— whelming number of high party voting Republicans come from one—party districts; almost twice as many as among the low voting Republicans who came mostly from semi- competitive and competitive districts. Among the Demo- crats most of the low and high party Democrats came from competitive districts. None of the high party voting Democrats came from one-party districts though a few of the low party voting ones did. As stated earlier Tennessee was not considered as to the relationship of competitive nature of the dis- trict and partisan attitudes because there were too few semi-competitive or competitive districts. The same de— letion is justified in comparing the relationship of par- tisan voting to competitive nature of the district. About the only possible significance is that slightly more high party voting Republicans came from competitive districts than low party voting Republicans. Among Democrats there is absolutely no relationship between the nature of district and party voting since an equal number of low party voting Democrats and high party voting Democrats come from one- party districts. 78 TABLE 13 COMPETITIVE NATURE OF DISTRICT AND PARTY VOTING IN TENNESSEE -_._ One Semi Compet- Party Compet— itive Total itive House low Dems 36 2 l 39 Senate low Dems 12_ _1' _Q_ .13 Total 48/92% 3/ 6% l/ 2% 52/100% House high Dems 34 3 2 39 Sen. high Dems 14 ._9 _J2 14' Total 48/91 3/ 6 2/ 3 53/100% House low Reps 6 3 2 11 Senate low Reps _2. _Q_ ‘_1 ._3 Total 8/57 3/21 3/21 14/100% House high Reps 4 2 4 10 Sen. high Reps _l ._2 _9_ _3 Total 5/38 4/21 4/31 13/100% 109 13 10 132 Urban—Rural Nature of the District Another variable frequently considered to be re- lated to political partisanship is the urban-rural nature of the population of the constituency. The literature of political science is replete with textbooks and articles 79 lamenting the usual under—representation of urban inter- ests and consequent over—representation of rural inter- ests in our state legislatures. In his study of urban—rural conflict in Illinois, David R. Derge compared the voting behavior on contested roll calls between the Cook County delegation and the ”downstate" legislators for the sessions 1949—1957. Both parties had representatives in all areas because of Illi— nois cumulative voting system. His findings showed that there was very little cohesion within Cook County or downstate delegations and that party conflict was far more prevalent than this type of conflict.5 The Four State Study divided the districts in the four states according to whether they were located in standard metropolitan statistical areas as defined by the United States Bureau of the Census. Districts out- side standard metropolitan statistical areas were then 5David R. Derge, "Urban—Rural Conflict: The Case in Illinois," in Legislative Behavior edited by John C. Wahlke & Heinz Eulau (Glencoe: Free Press, 1959), pp. 218—22; also "Metropolitan and Outstate Align- ments in Illinois and Missouri Legislative Delegations," American Political Science Review, LII (December 1958), pp. 1051-1065; and Robert S. Friedman, "The Urban-Rural Conflict Revisited," Western Political_guarterly, XIV (June 1961), pp. 481-495. 80 classified according to the percentage of rural—farm population residing in each. In California only, the central city districts of Los Angeles, San Diego, and San Francisco were separated from other metropolitan area districts. According to their method of classifi- cation, New Jersey was the most "urban" state with 81% of the legislators coming from standard metropolitan areas and no districts that contained 30% or over rural- farm population. In California 60% of the legislators came from metropolitan districts, and there were none from areas that were 40% or over rural—farm. Ohio was considerably more "rural" than California and New Jersey, but even there 45% of the legislators came from metro- politan areas, and there were only twelve who came from districts that were 50% or over-rural farm. Tennessee had the most "rural" legislature with only 20% of the members coming from metropolitan districts, while 42% came from districts that were 50% or over rural-farm. The distribution of partisan attitudes according to the urban—rural nature of the district is shown for each individual state, and for the combined totals in Table 14. In the combined totals for the four states, 81 legislators from districts outside the metropolitan areas were significantly more likely to express attitudes char- acteristic of independents when compared to those from metropolitan districts. MetrOpolitan legislators were significantly more likely to assume attitudes of strong partisans. The differences are accounted for largely by the responses of Democrats. The difference in proportion of independents and strong Republicans who come from met- ropolitan and non—metrOpolitan areas are more negligible though in the same direction as Democrats. Independent Democrats were significantly more likely to come from rural districts, but this is because Tennessee has a plentiful supply of both rural districts and independent Democrats. In summary, in reSpect to individual states, it was found that for both parties in California, for the Democrats in Ohio and Tennessee, and for the Republicans in New Jersey there was a relationship between strength of attitudinal partisanship and the proportion of metro- politan people in the legislative district. The strong Republicans in Ohio are an exception in that they are much more likely than any of the other partisan categories to come from districts outside metropolitan areas. RELATION OF RURAL-URBAN NATURE OF DISTRICT '10 PARTISAN ATTITUDES 82 TABLE 14 ==1Independent ‘Weak Strong Tbtal Deps Reps Dems Reps Dams Reps CALIFORNIA Central City districts of metropolitan areas (N= 20) 10$ 15 10 10 30 25 100$ Other metropolitan area districts (N- 45) 16% 33 11 13 11 16 100$ NOn-metro districts (N- 43) 19$ 32 21 12 14 2 100$ NEW JERSEY Metropolitan area (N= 64) 9$ ‘6 17 14 13 41 100$ Non-metro districts (N- 15) - - 7 46 7 4O 100$ OHIO Metropolitan area districts (N. 69) 13$ 6 9 27 36 100$ Non-metro districts (N- 90) 7$ 11 14 4 58 100$ W Metropolitan area districts (N- 22) 23$ 9 36 - 27 5 100$ Non-metro districts, under 50$ rural-farm. (N- 41) 31$ 15 22 12 15 5 100$ Non-metro districts, 50$ or over rural-farm (N- 48) 48$ 6 13 4 27 2 100$ FOUR STATE TOTAL Metropolitan area districts (N-220) 13$ 13 15 10 20 29 100$ Nonemetro districts under 50$ rural-farm (N-177) 15$ 16 13 16 8 32 100$ Nonrmetro districts, 50$ or over rural, farm (N- 60) 40$ 8 12 7 25 8 1001. Tbtal NOn-metro districts (N-237) 21$ 14 13 13 13 26 100$ Source: Ferguson working Paper, 0 . cit., p. 34. 83 Comparing the voting behavior with the urban rural nature of the district, we once again find a disparity in most instances with the findings of the Four State Study. TABLE 15 URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN CALIFORNIA Central Other Non A . C LIFORNIA City Metro Metro Total House low Dems 3 10 5 18 Senate low Dems 9_ _g_ _§. ‘19 Total 3 11% 12 43% 13 46% 28 100% House high Dems 9 5 5 19 Senate high Dems .9 _§_ _;1 19. Total 9 31 8 28 12 41 29 100 House low Reps 5 l4 4 23 Senate low Reps Q_ _3_ _§_ _2_ Total 5 l6 17 53 10 31 32 100 House high Reps 5 8 6 19 Senate high Reps Q, _4_ _1 311 Total 5 17% 12 40% 13 43% 30 100% In California we find most of the low Democrats coming from non-metropolitan districts with very few from the central city (Los Angeles). Among high partisan 84 voting Democrats a substantial number come from central city districts but also from non—metropolitan districts. Among Republicans there is no significant relationship between voting behavior and the urban-rural nature of the district. TABLE 16 URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN NEW JERSEY Central Other Non 1 City Metro Metro Tota House low Dems 0 12 l 13 Senate low Dems Q. _l. .1 _g Total 0 13 87% 2 13% 15 100% House high Dems 0 7 O 7 Senate high Dems Q_ _5 ‘Q __5 Total 3 O 12 100 0 12 100 House low Dems 0 l4 3 17 Senate low Reps Q __4 4 _8 Total 0 18 72 7 28 i 25 100 House high Reps O 17 4 21 Senate high Reps Q_ _4_ g_ _Q Total , O 21 78% 6 22% 27 100% New Jersey high and low party voting Democrats both came from metropolitan districts and the same is true of Republicans (Table 16). 85 TABLE 17 URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING IN OHIO Total Non—Metro Non-Metro OHIO Metro Non- _30% R.F +30% R F Total Metro House low Dems 9 12 9 3 33 Sen low Dems _4' ._2 _g_ _9_ _§_ Total 13 32% 14 34% 11 27% 3 7% 41 100% House high Dems 20 1 l O 22 Sen high Dems _§_ _9_ _Q_ _9_ _6_ Total 26 93% l 3 1 3 0 28 100 House low Reps 26 21 17 4 68 Sen low Reps _3. _§_ _§_ _9_ .19 Total 29 33 29 33 25 29 4 5 87 100 House high Reps' 7 41 31 10 89 Sen high Reps _g_ _2_ _2_ _Q_ 29_ Total 9 8% 50 46% 4O 37% 10 9% 109 100% In Ohio we find twice as many high party voting Dem- ocrats coming from metropolitan areas as low party voting Democrats while the reverse is true among Republicans (Table 17). The Republicans of high party vote come from non metro— politan areas to a much greater extent than from metropolitan districts. This would bear out the findings of the Four State Study as to Ohio. URBAN-RURAL NATURE OF DISTRICT AND PARTY VOTING 86 TABLE 18 OF TENNESSEE Metro Total Non—Metro Non—Metro Total Non—Metro—30% R F +30% R F House low Dems 15 24 15 9 63 Senate low Dems _6_ _Z_ _4_ _3_ 29 Total 21 25% 31 37% 19 23%. 12 15% 83 100% House high Dems 1 38 13 25 77 Senate high Dems _Q. 14_ _§_ _2' g§_ Total 1 l 52 50 l8 17 34 32 105 100% House low Reps 0 ll 5 6 22 Senate low Reps _Q. _3. _3. ._0 ‘_§ Total 0 14 50 8 29 6 21 28 100% House high Reps 4 6 4 2 16 Senate high Reps _Q_ _3_ _2. ._1 ‘_6 Total 4 18% 9 41% 6 27% 3 14% 22 100% Tennessee also bears out the findings of the Four State Study in that the overwhelming number of high party voting Democrats were from non-metropolitan districts (Table 18). Low party voting Democrats were also more from non- metropolitan districts but there were a substantial number of low party voting Democrats from metropolitan areas. Among the Republicans both the high party voting and the low party 87 voting members were mostly from non-metrOpolitan districts. However, there were some high voting Republicans from metro- politan districts but no low voting Republicans. 2259221 In Chapter 4 I have related constituency factors to partisan voting behavior. The two constituency factors that were considered were the competitive nature of the legisla- tor's district and the urban-rural nature of the district. I hypothesized that high party voters tended to come from non—competitive districts and low party voters from competitive districts. This hypothesis held true among California Republicans. The California Democrats behaved the Opposite, but not strongly so. New Jersey Democrats bore out the hypotheses but the Republicans showed no pat- tern at all. In Ohio, the Republicans confirmed the hypothe- sis and the Democrats showed no pattern at all. Tennessee had too few competitive districts to yield any results along this dimension. Thus out of sixteen units of study, the hypothesis was confirmed by seven units, negated by five units, with four units neither confirming nor negating the relationship 88 between party voting and the competitive nature of the district. In regard to the urban-rural nature of the district, my hypothesis was that high party voters tended to come from metropolitan districts while low party voters tended to come from non—metrOpolitan districts. This was confirmed among California Democrats but Republicans showed no clear rela— tionship between these variables. This hypothesis was fur- ther rendered difficult to prove by virtue of the fact that almost all units in New Jersey came from metropolitan dis— tricts and most Tennessee legislators came from non metro- politan districts. Ohio Democrats confirmed my hypothesis while the Republicans tended to refute it. Thus of sixteen units of study, the hypothesis was confirmed by only four units, negated by five units with seven units neither confirming nor negating any relation- ship between partisan voting and urban-rural nature of the district. CHAPTER V PERSONAL CHARACTERISTICS AND PARTISANSHIP We turn to a consideration of the personal char- acteristics as they might relate to partisan attitudes and partisan voting behavior. The personal characteris- tics which we shall consider are age, education, and occupation. Age The ages of the legislators were coded in five year intervals (31-35, 36-40, etc.) and the interval that was closest to the median age of legislators in each of the four states was used as a breaking point. In Calif— ornia the median age was 50, in New Jersey it was 46.2. in Ohio 47.9, and in Tennessee 45.2. Table 19 shows the findings of Four State Study on the relation of age to partisan attitudes. 89 90 TABLE 19 RELATION OF AGE TO PARTISAN ATTITUDES Independent Weak Strong Total State and Age Dems Reps Dems Reps Dems Reps California 50 and under 18% 19 l6 13 21 13 100% Over 50 13% 40 13 12 10 12 100% New Jersey 45 and under 11% 3 14 28 ll 33 100% Over 45 5% 7 16 14 12 46 100% Ohio 50 and under 7% 6 8 15 23 41 100% Over 50 13% 13 6 7 3 58 100% Tennessee 45 and under 29% 14 23 5 24 5 100% Over 45 47% 4 l8 8 21 2 100% Source: LeRoy C. Ferguson, Working Paper, op. cit., p. 36. We note that in California independent Republicans were more likely to be over the median age for Republicans of that state, and the data also show strong attitudinal Democrats to be younger than the other partisanship cate- gories. With regard to New Jersey legislators the data sug- gest that weak partisan Republicans were likely to be younger 91 and strong Republicans older than the other partisanship categories while in Ohio the strong Democrats were likely to be below, and strong Republicans above the median age bracket for the legislature of that state. The data also indicated that Tennessee independent Democrats were some- what more likely to be above the median age of all Tennes_ see legislators. In general there did not appear to be any strong significant relationship between age and parti- san attitudes. Let us now turn to see if there is any significant relationship between age and partisan voting behavior. We see in Table 20 that in California the low vot- ing Democrats are almost evenly divided age—wise but that the high partisan voting Democrats are twice as likely to be under the median age of legislators of that state as over that age. Among the Republicans once again the low partisan voting are almost evenly divided below and above the median age while high partisan voting Republicans were much more likely to be over the median age as under it. New Jersey Democrats were exactly the opposite of California Democrats in that the low party voting Demo— crats were more likely to be under the median age and the 92 high party voting Democrats tended to be above the median age of legislators for that state. The New Jersey Repub- licans followed the reverse pattern with low voting Repub- licans tending more to be over the median age and the high voting Republicans tending to be almost evenly divided above and below the median age. In Ohio the Democrats showed no relationship be— tween age and partisan voting in that both the low party voters as well as the high party voters were mostly under the median age for legislators of that state. The Repub— 1icans were somewhat similar to the Republicans of Calif- ornia in that low party voters tended to be found a bit more under the median age and high party voting Republicans were more prevalent above the median age. Tennessee low party voting Democrats tended more to be under the median age bracket while high party voting Democrats were significantly more prevalent over the median age bracket of legislators for that state. Although small in number, the Republicans tended to have high party voters substantially greater under the median age while low party voters were almost evenly divided above and below the median age bracket. 93 TABLE 20 AGE AND PARTISAN VOTING California 50 & under Over 50 N A Total Low Dems 15 48% 13 42% 3 10% 31 High Dems 19 68 9 32 28 100% Low Reps 15 45 17 52 l 3 33 High Reps 12 39 18 58 1 3 31 100% New Jersey 45 & under Over 45 N A Total Low Dems 9 60% 6 40% 15 High Dems 4 31 8 62 l 7% 13 100% Low Reps 9 35 16 62 1 3 26 100 High Reps 14 48 13 45 2 7 29 100 Ohio 50 & under Over 50 N A Total Low Dems 17 63% 9 33% .l 4% 27 100% High Dems 19 70 6 22 2 8 27 100 Low Reps 3O 52 23 4O 5 8 58 100 High Reps 26 44 31 53 2 3 59 100 Tennessee 45 & under Over 45 N A Total Low Dems 29 56% 22 42% 1 2% 52 100% High Dems 21 4O 32 6O 53 100 Low Reps 7 50 6 43 l 7 14 100 High Reps 9 69 4 31 13 100% 94 Education The legislators in the four states of this study were a highly educated group in comparison to the total population. Only 2% of them had not attended high school, 81% had attended college, and almost one-half of them had some form of graduate or professional education. The Four State Study did reveal a positive rela- tion between completion of college and strength of partisan attitudes assumed by the legislators. Legislators who had not completed college were much more likely to express the attitudes of independentsixlall states except Tennessee and even there the data suggested such a relationship. As can be seen in Table 21, the completion of college work either at the bachelor's degree or graduate level was used as the breaking point to determine the relationship between the amount of education and partisan attitudes. In Table 22 we find that for the most part there is a relationship between education and partisan voting behav- ior as there was between education and partisan attitudinal behavior. However, there are some notable exceptions. Note that California high party voting Republicans were as likely as not to be college graduates while low party voting Repub- licans were more likely to be college graduates. 95 TABLE 21 RELATION OF EDUCATION TO PARTISAN ATTITUDES Independent Weak Strong Total Dems Reps Dems Reps Dems Reps California Less than completed college (N=47) 19% 48 6 9 9 9 100% Completed college (N=61) 13% 15 21 15 21 15 100% New Jersey Less than completed college (N:25) 16% 12 8 24 4 36 100% Completed college (N=54) 4% 2 l9 19 15 41 100% Ohio Less than completed college (N=66) 20% 15 4 14 12 35 100% Completed college . (N=93) 2% 4 9 ll 16 58 100% Tennessee Less than completed college (N=57) 38% 16 12 4 26 4 100% Completed college (N=54) 34% 4 3O 9 l9 4 100% Four State Totals Less than completed college (N=195) 24% 23 8 ll 15 19 100% Completed college (N=262) 12% 6 18 13 18 33 100% Source: LeRoy C. Ferguson, Working Paper, op. cit., p. 38. “ct-g “__.... - 96 TABLE 22 EDUCATION AND PARTISAN VOTING Less than College College Completed N A Total Completed California Low Dems 12 40% 16 53% 2 7% 30 100% High Dems 6 19 23 74 2 7 31 100 Low Reps 12 35 20 59 2 6 34 100 High Reps 15 47 15 47 2 6 32 100 New Jersey Low Dems 4 23% ll 65% 2 12% 17 100% High Dems 2 14 10 72 2 l4 14 100 Low Reps 8 30 17 63 2 7 27 100 High Reps 6 21 21 72 2 7 29 100 Ohio Low Dems 12 41% 14 48%» 3 11% 29 100% High Dems 10 34 15 52 4 14 29 100 Low Reps 22 37 31 52 7 11 60 100 High Reps 16 26 41 67 4 7 61 100 Tennessee Low Dems 17 32% 33 61% ‘4 7% 54 100% High Dems 24 44 23 42 8 14 55 100 Low Reps 6 38 6 38 4 24 16 100 High Reps 5 33 6 4O 4 27 15 100 97 In New Jersey both low and high party voting Dem- ocrats were mostly college graduates. Republicans too were mostly college graduates in both voting categories but high party voters were in greater number than low party voters among those who were college graduates. In Ohio there were slightly more low party voting Democrats among the non-college graduate legislators than among the high party voters but not significantly differ- ent. The Republicans followed our expectations more closely in that low party voters were more predominant in the non-college graduate category than high party voters and more college graduates were high party voters than low party voters. Tennessee Democrats went the opposite of our ex- pectations in that college graduates were greater among the low party voters and high party voters were almost split evenly between college graduates and non—college graduates. The Republicans were evenly divided all around both as to voting and level of education. Since it has been hypothesized that more highly educated persons tend to identify more closely with a political party and inasmuch as dividing college graduates 98 failed to yield any close relationship to partisan voting behavior, a further attempt was made to see whether those legislators educated beyond college were any more partisan in their voting behavior than those with lesser education. The results are found in the following table. TABLE 23 EDUCATION BEYOND COLLEGE AND PARTISAN VOTING I Non- Beyond college College college N A Total California Low Dems 12 40% 7 24% 9 30% 2 6%1 30 100% High Dems 6 l9 8 26 15 48 2 7 31 100 Low Reps 12 35 10 29 10 29 2 7 34 100 High Reps 15 47 9 28 6 l9 2 6 32 100 New Jersey Low Dems 4 24% 5 29%. 6 35% 2 12% 17 100% High Dems 2 l4 3 21 7 50 2 15 14 100 Low Reps 8 3O 9 33 8 30 2 7 27 100 High Reps 6 21 7 24 14 48 2 7 29 100 Ohio Low Dems 12 41% 3 10% ll 38% 3 11%: 29 100% High Dems 10 35 7 24 8 28 4 13 29 100 Low Reps 22 37 9 15 22 37 7 11 60 100 High Reps 16 26 19 31 22 36 4 7 61 100 Tennessee Low Dems 17 32% 11 20% 22 41% 4 7% 54 100% High Dems 24 44 9 16 14 26 8 14 55 100 Low Reps 6 37 4 25 2 13 4 25 16 100 High Reps 5 33 3 20 3 20 4 27 15 100 99 We find very similar patterns as we did when we compared college graduates versus non college graduates and partisan voting behavior only a bit more pronounced. In those states where college graduates are more partisan in their voting behavior than non college graduates, those who were educated beyond the college level were partisan in a greater proportion than either the non college or college graduates. In those states and parties where college graduates tend to be less partisan in their vot- ing behavior than non college graduates, than those edu- cated beyond college, tend to vote less partisanly. Ex- ceptions would be Republicans in Ohio who split evenly between high and low party voting at the beyond college level and Tennessee where Republicans were almost evenly Split in all educational categories. Occupation The distribution of partisan attitudes within various occupational groups of legislators is shown in Table 24. We might have expected a smaller number of independents among lawyers than among other occupations since the legal profession is most often associated with 100 professional political careers. The data of the four states bears this out in that lawyers were significantly less likely to be independents than those in other occu— pations. The percentage of lawyers in the legislatures under study were as follows: California, 30%; New Jersey, 52%; Ohio, 36%; and Tennesse, 30%. The Democrats and Re- publicans had equal proportions of lawyers in Ohio, but in the other three states Democratic legislators were more likely than Republicans to be in the legal profession. In comparing occupations with partisan voting be- havior we have once again separated lawyers from the other occupations and hypothesize that in keeping with the atti- tudinal data, lawyers would tend to be high party voters in greater prOportion than members of other occupations. California Democrats bear out our hypothesis in that among lawyers the high party voters predominated while among non-lawyers more were low party voters than high party voters. The California Republicans behaved differently in that more Republican lawyers were low party voters than high party voters and the non-lawyers were split down the middle between the two voting categories. (See Table 25) 101 TABLE 24 RELATION OF OCCUPATION TO PARTISAN ATTITUDES Independent Weak Strong Four States Total Total Dems Reps Dems Reps Dems Reps Professions other than law (N= 35) 14% 3 6 17 31 29 100% Law (N=l66) 13% 5 20 13 19 30 100% Wholesale & retail trade (N= 46) 22% 13 15 9 15 26 100% Banking, real es- tate & insurance (N= 50) 20% 22 12 6 12 28 100% Other occupations (N=110) 23% 19 5 15 13 25 100% Agriculture(N= 50) 14% 28 18 8 8 24 100% California Lawyers (N= 31) 16% 10 16 13 29 16 100% Other occupations (N= 77) 16% 38 14 12 10 10 100% New Jersey Lawyers (N= 41) 5% 2 22 17 12 42 100% Other occupations (N= 38) 11% 8 8 23 ll 39 100% Ohio Layers (N= 58) 2% 5 10 16 19 48 100% Other occupations (N=101) 14% ll 5 10 12 48 100% Tennessee Lawyers (N= 36) 36% 3 36 6 l9 - 100% Other occupations (N= 75) 38% l3 l3 7 24 5 100% Source: LeRoy C. Ferguson, Working Paper, Op. cit., p. 40 102 TABLE 25 MEMBERSHIP IN LEGAL PROFESSION AND PARTISAN VOTING Lawyers Non—Lawyers N A Total California Low Dems 9 30%, 19 64% 2 6% 30 100% High Dems 12 4O 16 53 2 7 30 100 Low Reps 9 26 23 67 2 7 34 100 High Reps 6 19 24 75 2 6 32 100 New Jersey Low Dems 8 47%. 7 41% 2 12% 17 100% High Dems 8 57 4 29 2 l4 14 100 Low Reps 12 44 13 48 2 8 27 100 High Reps 13 45 14 48 2 7 29 100 Ohio Low Dems 9 32% 17 61% 2 7% 28 100% High Dems 10 35 17 59 2 6 29 100 Low Reps 22 37 36 6O 2 3 60 100 High Reps 22 36 37 61 2 3 61 100 Tennessee Low Dems 24 45%, 28 52% 2 3% 54 100% High Dems 13 24 4O 73 2 3 55 100 Low Reps l 6 13 81 2 13 16 100 High Reps 2 13 ll 74 2 13 15 100 103 In New Jersey there was no voting difference among lawyers of either party. Among the non—lawyers, the only significant difference was between the Democrats Where more non-lawyers were low party voters than high party voters. In Ohio there was almost no relationship at all in either party between party voting and the legal pro- fession. Tennessee Democrats behaved along opposite lines of our hypothesis. Most lawyers were low party voters and most non-lawyers were high party voters. Among Ten- nessee Republicans there was no significant difference except that slightly more non—lawyers were low party voters than high party voters which is in keeping with our expectations. Summary Chapter 5 considered the relationship of the leg- islators personal characteristics and partisan voting. The personal characteristics under consideration were age, education, and occupation. 104 As in the Four State Study the legislators of each state were divided into those falling above or below the median age bracket for that state. These were then compared with partisan voting, the hypothesis being that younger leg- islators tended to vote more partisanly while older legisla- tors were low partisan voters. The results of relating these two variables were most irregular. High voting Democrats in California confirmed the hypothesis but the low voting Demo- crats were evenly divided in age but high voting Republicans confirmed the hypothesis. New Jersey Democrats negated the hypothesis while the low voting Republicans were evenly div— ided in the age brackets. Ohio Democrats demonstrated no pat— tern at all in the relationship between age and partisan vot- ing and Ohio Republicans negated the hypothesis. Tennessee low voting Democrats tended to fall below the median age and high party voting Democrats were above the median age. High party voting Tennessee Republicans confirmed the hypothesis while low party voting Republicans were almost evenly divided in age. Thus, in the sixteen units of study the hypothesis was confirmed by seven units, negated by seven units with two showing no relationship between age and partisan voting behavior. 105 Education The legislators were divided into those who were college graduates and those who were not, with the hypothesis being that a higher level of education would lead to more partisan voting and vice versa. California high party voting Democrats were overwhelmingly college graduates but so were the majority of low party voting Democrats. California Re- publicans on the other hand showed low party voters to be mostly college graduates with the high party voters evenly divided. New Jersey low party voting Democrats and Republi- cans negated the hypotheses while high party voters of both parties confirmed it. . Ohio low party voting Democrats failed to show any relationship between education and party voting but high party voting Democrats and both types of Republicans con- firmed the hypothesis. Tennessee Democrats and Republicans were somewhat divided all around. Thus out of sixteen units, the hypothesis was con- firmed by five, negated by seven with four being neutral. I also set up a category of legislators who went be- yond college expecting them to be more partisan voters but the results showed no appreciable difference in this division of legislators as between college and non college graduates 106 Occupation I hypothesized that lawyers would tend to be more dedicated to politics as a profession and hence would be more partisan in their voting behavior. California Democrats confirmed the hypothesis. Low party voting California Repub- licans negated the hypothesis and high party voting Republi— cans were evenly divided between lawyers and non-lawyers. New Jersey showed no appreciable relationship between member- ship in the legal profession and partisan voting. Low party voting Democrats of New Jersey negated the hypothesis. Ohio legislators showed no relationship between party voting and occupation. Tennessee Democrats negated the hypotheses while Republicans showed no relationship between these two variables. Thus out of sixteen units, the hypothesis was confirmed in seven, negated in seven with two units being neutral. CHAPTER VI CAREER FACTORS AND PARTISANSHIP We now turn to career characteristics of legisla— tors and how these relate to partisanship. The career characteristics that we shall consider are the source of the decision to enter the legislature, years of legislative experience, whether or not legislator held party office in the legislature and the legislator's intention to seek re- election. It is assumed that these variables should have an effect upon the legislator's partisanship both in atti- tude and in voting. Party as Source of Decision to Become a Legislator The nomination of candidates for public office is generally considered to be one of the major functions of political parties. In view of this it was surprising to find that in all of the states except New Jersey, very few legislators mentioned the party as a source of their decision to become a legislator. The question asked was: 107 108 "Just how did it come about that you became a legislator?" The prOportion of those who mentioned party were: Calif- ornia, 22%; New Jersey, 61%; Ohio, 20%; and Tennessee, 13%. The proportion of Republicans and Democrats who mentioned party were approximately the same in California but in Ohio and Tennessee, Republicans were more likely than Democrats to give party as one of the sources of their decision to run. In New Jersey, however, a larger proportion of Demo- crats than Republicans mentioned party as a sponsor of their candidacy. To the extent that legislators did mention party as a source of their decision to enter the legislature it would be expected that this would be related to the strength of their partisan attitudes and voting behavior. The data presented in Table 26 gives some limited support to this expectation. In the combined totals for the four states, independents were significantly less likely than partisans to mention party as one of the sources of their decision to become a legislator. There appeared to be no such re— lationship in California, but in Ohio the strong Republicans differed from other members of their own party in regard to this variable. There were so few Tennessee legislators who gave party as one of the sources of their decision that we 109 cannot make any strong statements about their partisan atti— tudes, but there is some indication that independent Demo— crats were even less likely than other Democrats to give the party as one of their sources of decision. It should also be noted in Table 27 that weak Democrats in New Jersey were more likely to mention the party than any of the other legislators in that state. This is mainly because the majority of the weak Democrats in New Jersey came from Hudson County, where the strong Demo- cratic machine exercises almost complete control over legis- lative nomination. In regard to voting behavior as it relates to the party as a source of a legislative career we find almost no relationship in either party in California. We find no re- lationship among the Democrats of New Jersey but we do find a significant relationship between these two variables among the Republicans of New Jersey. Those who mentioned the party as a source of their decision to run for the legisla- ture were likely to vote with their party more frequently and those who didn't mention the party as a source of their decision were more likely to be low party voters. RELATION OF MENTION OF PARTY AS SOURCE OF DECISION TO 110 TABLE 26 BECOME LEGISLATOR TO PARTISAN ATTITUDES Independent Weak Strong Total Dems Reps Dems Reps Dems Reps California Party mentioned (N= 24) 12% 30 21 9 17 12 100% Party not mentioned (N= 84) 17 30 l3 13 15 12 100 New Jersey Party mentioned (N= 48) 10 4 23 19 8 36 100 Party not mentioned (N= 31) 3 6 3 23 16 49 100 Ohio Party mentioned (N= 31) 10 3 6 10 3 68 100 Party not mentioned (N=128) 9 10 7 13 17 44 100 Tennessee Party mentioned (N= 14) 14 14 22 22 28 -- 100 Party not mentioned (N= 97) 40 9 21 4 22 4 100 Four States Total Party mentioned (N=ll7) ll 10 18 15 11 35 100 Party not mentioned (N=340) 20 14 12 ll 18 25 100 Source: LeRoy C. Ferguson, Working Paper, Op. cit., p. 49a. 111 TABLE 27 MENTION OF PARTY AND PARTY VOTING Yes No Total California Low Dems 5 20% 20 80% 25 100% High Dems 7 25 21 75 28 100 Low Reps 7 24 22 76 29 100 High Reps 5 17 24 83 29 100 New Jersey Low Dems ll 73% 4 27% 15 100% High Dems 9 75 3 25 12 100 Low Reps ll 44 14 56 25 100 High Reps 17 63 10 37 27 100 Ohio Low Dems 3 12% 23 88% 26 100% High Dems 3 12 22 88 25 100 Low Reps 2 4 51 96 53 100 High Reps 13 24 44 76 57 100 Tennessee Low Dems 8 16% 43 84 51 100% High Dems 2 25 6 75 8 100 Low Reps 3 21 ll 79 14 100 High Reps 2 15 ll 85 13 100 112 Ohio Democrats had no relationship at all between party voting and party as a source of decision but the Republicans had a very significant relationship. Those Ohio Republicans who mentioned the party voted more fre- quently with the party and those who didn't mention the party as the source of their decision tended to vote less regularly with their party. In Tennessee neither the Democrats nor the Repub- licans showed any significant relationship between the two variables with the overwhelming majority of the Demo- crats not mentioning the party as the source of their decision to enter the legislature. Legislative Experience Another variable that we shall consider is that of legislative experience as it relates to partisan atti— tudes and voting. It would be expected that longer ex— perience in the legislature would mean a stronger pro- fessional political career pattern and would result in stronger partisanship. To test this hypothesis, the legislators in each state were divided into two groups 113 using as a breaking—point the median number of years of service in each legislature. The median in California was six years of service, but five years was used in order to have two groups of approximately equal size. In Tennessee the median was two years of service, but because of clustering at that point, the legislators of that state were divided into two groups--those with and those without any experience in the legislature. As shown in Table 28 this hypothesis is not borne outin regard to attitudinal partisanship. In the combined totals for the four states it can be seen that the independents were significantly more likely than the other partisanship categories to have had above the median number of years of legislative service. In Calif- ornia and Tennessee the same significant relationship is evident between years of service and likelihood of being an independent and the data suggest that such may also be the case in Ohio. California strong Democrats were sig- nificantly less likely than other Democrats to have had more than a median number of years of legislative service. It was only in the case of New Jersey Republicans that the strong partisans were significantly more likely than 114 weak partisans and independents to have had more than the median number of years of service in the legislature. Our findings regarding legislative experience and partisan voting does present some marked differences from the partisan attitudinal behavior. (See Table 29) In California those Democrats with less than five years ex- perience were more likely to be strong voting partisans while more Democrats with over five years experience were likely to be low partisan voters. Among Republicans there were no significant relationships between experience and partisan voting. New Jersey Democrats showed no relationship at all between these two variables since they Spread equally in all categories of voting and experience. The Republi— cans on the other hand showed a significant relationship in that those with more than the median number of years experience were more likely to be low party voters. This bears out the attitudinal findings. Ohio Democrats showed very little significant re- lationship between voting and experience while Republicans were the same as New Jersey Republicans only more pronounc- edly. Those with more experience were more likely to be low partisan voters while those with less experience were more likely to be high partisan voters. 115 TABLE 28 RELATION OF YEARS OF LEGISLATIVE EXPERIENCE TO PARTISAN ATTITUDES Independent Weak Strong Total Dems Reps Dems Reps Dems Reps California Five years or less (N= 57) 14% 21 16 12 25 12 100% Over five years (N= 51) 17 39 14 12 6 12 100 New Jersey Three years or less (N= 43) 9 7 16 26 14 28 100 Over three years (N= 36) 6 3 l4 l4 8 55 100 Ohio Four years or less (N= 83) 8 5 10 l6 19 42 100 Over four years (N= 76) ll 13 4 8 9 55 100 Tennessee No previous legis— lative experience (N= 51) 28 8 25 8 25 6 100 Two or more years (N= 60) ~45 12 16 5 20 2 100 Four States Total Median or less years of service (N=234) 14 10 16 15 21 24 100 Above median years of service(N=223) 21 17 ll 9 ll 31 100 Source: LeRoy C. Ferguson, WOrking Paper, Op. cit., p. 53. Ian. L‘s-1.. YEARS OF LEGISLATIVE EXPERIENCE AND PARTY VOTING 116 TABLE 29 California Five years . Over Total or less five years Low Dems 8 25% 24 75% 32 100% High Dems 14 42 19 58 33 100 Low Reps 22 _ 28 78 36 100 High Reps 9 26 25 74 34 100 Three years Over New Jersey or less three years Total Low Dems 9 53% 8 47% 17 100% High Dems 8 50 8 50 16 100 Low Reps 14 48 15 52 29 100 High Reps 10 31 22 69 32 100 Four years Over Ohio or less four years Total Low Dems 12 39% 19 61% 31 100% High Dems 14 45 17 55 31 100 Low Reps ll 18 51 82 62 100 High Reps 20 32 43 68 63 100 Tennessee No prev1ous legis- Two or Total lative experience more years Low Dems 7 13%, 49 87% 56 100% High Dems 6 11 51 89 57 100 Low Reps 2 l3 14 87 16 100 High Reps 2 13 13 87 15 100 117 Tennessee Democrats showed no significant rela- tionship between these two variables nor did the Republi- cans. Offices in the Legislature We would also expect that one who holds an office in the legislature would in turn be more of a career poli- tician and tend to be a stronger partisan. The one excep- tion we might expect to find would be in California which at that time had a lack of partisanship in its organiza- tion. Table 30 shows the distribution of partisan atti- tudes among office holders and non—office holders in each state. For purposes of this analysis, office holding was interpreted broadly, so as to include not only presiding officers and floor leaders, but also committee chairmen and members of important procedural committees such as rules. The attitudinal data tends to verify our expecta- tions. In California, in keeping with its nonpartisan tradition of the time independents were more likely than partisans to hold legislative offices. There appeared to be no relationship between strength of partisanship 118 and office holding in Tennessee but in New Jersey and Ohio there is a strong indication in the data that office holders were more likely than rank and file members to be strong attitudinal partisans. The voting behavior revealed some marked deviation from the attitudinal behavior. As seen in Table 31 California Republicans had a very significant relationship between office holding and partisan voting behavior. The majority of the office holders were high voting partisans and the majority of the non-office holders were low party voters. New Jersey Democrats showed a significant relationship between hold— ing office in the legislature and party voting while the Republicans did not. In Ohio the reverse was true with the Democrats equally divided as to party voting and office holding but the Republicans showed a clear relationship. Attitudinally they both showed a relationship. Neither Democrats nor Republicans in Tennessee had any signifi- cant relationship between legislative office holding and partisan voting but neither did they show any such rela- tionship in partisan attitudes. 119 TABLE 30 RELATION OF HOLDING OFFICE IN THE LEGISLATURE TO PARTISAN ATTITUDES Independent Moderate Strong Total Dems Reps Dems Reps Dems Reps California Office holders (N= 54) 15% 40 13 17 ll 4 100% Non—office holders (N= 54) l7 l9 l7 7 20 20 100 New Jersey Office holders (N= 39) 3 8 5 23 5 56 100 Non-office holders (N= 40) 12 2 25 18 18 25 100 Ohio Office holders (N= 58) 3 12 3 9 5 68 100 Non-office holders (N=101) 13 7 9 14 20 37 100 Tennessee Office holders (N= 29) 38 10 28 3 21 -— 100 Non-office holders (N= 82) 37 10 18 7 23 5 100 Source: LeRoy C. Ferguson, Working Paper, 0p. cit., p. 55. OFFICE HOLDING 120 IN LEGISLATURE AND PARTY VOTING Office Non-office Total Holders holders California Low Dems ll 39% 17 61% 28 100% High Dems 13 45 16 55 29 100 Low Reps 14 44 18 56 32 100 High Reps 22 73 8 27 30 100 New Jersey Low Dems l 7%. 14 93% 15 100% High Dems 4 33 8 67 12 100 Low Reps 16 64 9 36 25 100 High Reps 18 67 9 33 27 100 Ohio Low Dems 4 15% 23 85%. 27 100% High Dems 4 15 23 85 27 100 Low Reps 22 38 36 62 58 100 High Reps 31 53 28 47 59 100 Tennessee Low Dems 16 31% 36 69% 52 100% High Dems 15 28 38 72 53 100 Low Reps 4 29 10 71 14 100 High Reps 3 23 10 77 13 100 121 Running for Reelection The final career variable that we have considered is that of the legislator's expectations with regard to running for reelection to the legislature. We would tend to expect that those legislators intending to run again would be more likely to be professional politicians and likely to be more partisan in both attitude and voting behavior. Substantial majorities of the legislators ex- pected to run again in all of the states except Tennessee where only about a third of the legislators planned to seek reelection. The percentages were California, 70%; New Jersey, 76%; Ohio, 61%; and Tennessee, 34%. The data in Table 32 tend to confirm our expecta— tions in that among the attitudinal partisans in the four states strong partisans were significantly more likely to plan to run again. There were also differences in certain of the groups within the four states. Strong Republicans in New Jersey and Ohio tended slightly more not to seek reelection. The data for Tennessee conform most closely to the four states combined. Almost 2/3 of the strong partisans in Tennessee planned to run again, but among the independents less than 1/5 had such plans. 122 TABLE 32 RELATION OF INTENTION TO RUN AGAIN FOR THE LEGISLATURE TO PARTISAN ATTITUDES Independent Weak Strong Total Dems Reps Dems Reps Dems Reps California Will run again (N= 71) 17% 27 16 5 21 14 100% Will not run again or uncertain (N= 31) 13 36 16 23 6 6 100 New Jersey Will run again (N= 60) 7 7 15 18 13 40 100 Will not run again or uncertain (N= 19) 11 -- 16 26 5 42 100 Ohio Will run again (N= 96) 5 l3 7 13 15 47 100 Will not run again or uncertain (N= 63) 16 3 6 ll 13 51 100 Tennessee Will run again (N= 38) 21 5 24 3 39 8 100 Will not run again or uncertain (N= 73) 46 12 19 8 14 l 100 iEer States Total Will run again (N=265) ll l4 14 10 20 31 100 Will not run again or . un(Beertain (N=186) 27 12 14 13 ll 23 100 \ S . cytlnrce: LeRoy Ferguson, Working Paper, op. cit., p. 57. 123 There appears to be very little relationship be- tween party voting and intention to seek reelection to the legislature. (See Table 33) California Democrats did show some relationship in that more high party voting Democrats intended to seek reelection than not. Among Republicans the low party voters were more likely to seek reelection. Once again California might be an exception since its legislative politics at that time were very non partisan. New Jersey Democrats showed no relationship be- tween these two variables but Republicans did. High vot— ing Republicans were more likely to seek reelection while low voting ones were less likely to seek reelection. The same was true in Ohio where Democrats showed no signifi— cant relationship but Republicans did. However, among Ohio Republicans not intending to seek reelection more were high party voters than low party voters. Tennessee Democrats showed very little relationship between party voting and intention to seek reelection ex- cept that among those not seeking reelection, most tended to be low party voters. Though the numbers are small Tenne- ssee Republicans planning to seek reelection tended to be higher party voters. INTENTION TO SEEK REELECTION AND PARTY VOTING 124 TABLE 33 Run Unde- Will not . . . N A Total again c1ded run again California Low Dems 18 60% 6 20% 1 4% 5 16% 30 100% High Dems 22 71 2 7 3 10 4 12 31 100 Low Reps 18 53 6 18 3 9 7 20 34 100 High Reps 15 47 9 28 2 6 6 19 32 100 New Jersey Low Dems 10 59% 3 19% 2 12% 2 12% 17 100% High Dems ll 79 l 7 O 0 2 14 14 100 Low Reps 17 63 2 7 6 23 2 7 27 100 High Reps 22 76 4 l4 1 3 2 7 29 100 Ohio Low Dems 14 48% 8 28% 4 14% 3 10% 29 100% High Dems 15 52 5 l7 5 l7 4 14 29 100 Low Reps 38 63 13 22 2 3 7 12 60 100 High Reps 31 51 19 31 7 ll 4 7 61 100 Tennessee Low Dems 17 32% 21 39% 12 22% 4 7% 5 100% High Dems 18 33 20 36 9 l6 8 15 55 100 Low Reps 2 13 6 37 4 25 4 25 16 100 High Reps 4 27 3 20 4 27 4 26 15 100 125 W This chapter dealt with career factors as they related to partisan voting behavior. These career factors are party recruitment of candidates, legislative experi— ence, holding of party office in the legislature and in— tention to seek reelection. Party as a Source of Decision in candidacy The hypothesis was that those legislators who men- tioned their party as a source of decision to run for elec- tion to the legislature would tend to be high party voters and vice versa. No such relationship was found among either party in California and only among the Republicans of New Jersey. This relationship was shown among Ohio Republicans but not among the Ohio Democrats while in Tennessee this hypothesis was borne out only among the low party voters of both parties. Thus, of my sixteen units of study, the hypothesis was confirmed by nine units and negated by seven units. Legislative Experience The expectation that longer years of experience meant more dedication to politics as a profession and a 126 higher degree of partisan voting was not borne out in most units. This hypothesis was negated by the California Dem- ocrats, and there was no relationship between these two factors among the California Republicans. New Jersey Dem- ocrats exhibited no relationship and Republicans ran counter to expectations. The same was true in Ohio and in Tennessee where neither party showed any relationship between partisan voting and legislative experience. Of sixteen units, the hypothesis that there is a relationship between partisan voting and legislative ex- perience was confirmed in seven units, negated in six with three units neither confirming nor negating the hypothesis. Office Holding in Legislature The hypothesis that party office holders in the legislature would tend to be high party voters was con- firmed in more instances than any other of our hypotheses. A relationship between office holding and high party vot- ing was established among California Republicans, New Jersey Democrats and Ohio Republicans. Tennessee, where party organization in the legislature was virtually non- existent, there was no relationship between these two variables. 127 In my sixteen units, ten confirmed the hypothesis while six negated it. This hypothesis was confirmed more strongly than any other hypothesis in this study. Intention to Seek Reelection I hypothesized that those intending to seek reelec— tion would tend to be high party voters and vice versa. Most legislators intended to seek reelection and very little relationship was shown between this intention and the strength degree of partisan voting behavior. A relation- ship was shown only among New Jersey Republicans and an inverse relationship among Ohio Republicans. Tennessee did show a slight tendency toward confirming a relationship between intention to seek reelection and partisan voting. Of the sixteen units, nine confirmed while seven negated the relationship between intention to seek reelec- tion and partisan voting. CHAPTER VII PERCEPTIONS OF PARTY IN THE LEGISLATURE Having previously discussed some of the independ— ent variables related to partisan attitudes and voting behavior, we now turn to the question of whether the par— tisan attitude or voting behavior assumed by the legisla— tor is related to his perceptions of the operation of parties in the legislative process as indicated by his replies to the open-ended questions which covered this subject. In the first of this series of questions the legislators were asked, "How would you describe the part played by political parties in the legislature?" Replies to this question were coded in terms of whether the legis— lator perceived either the Republican party or Democratic party or parties generally, as having an impact on leg- islative decisions. Of the legislators who replied to this question, 42% indicated that parties had an impact on legislative decisions, and 39% said that parties had little or no impact. A few (8%, mostly minority party 128 129 members in New Jersey and Ohio) said that one of the par- ties had impact, while the other party had little or no impact, and 11% of those who answered the question did not mention party impact on legislative decisions in their reply. Table 34 shows the relation of partisan attitudes to these differing perceptions of the impact of parties on legislative decisions. In terms of role theory we would expect that independents would be less likely than strong partisans to perceive parties as having an impact on legislative decision making, and Table 34 demonstrated this in a striking manner. For the combined totals of the four states, independents in both parties were sig— nificantly less likely than weak or strong partisans to perceive party impact on the decision-making process in the legislature. In turning to the relation between perception of party impact on legislative decisions and party voting we note that once again we fail to find as significant a relationship as was found in between expressed attitudes and perception of the parties' impact. The total numbers in these categories are greater than the number of persons 130 responding since there was some overlap in response. In Table 35 we note that in California the Democrats did not perceive much impact and neither did the Republicans. The low party voting Republicans were quite emphatic in this. In New Jersey most Democrats and Republicans of both voting categories felt that the parties had an impact on legislative decisions. Ohio Democrats perceived an impact though low voters felt it more strongly than high voters. Republi- cans perceived an impact of party on legislative decisions and more low party voters than high party voters did. In Tennessee we find the peculiar phenomenon of low party voters in both parties perceiving party impact more than high voters but those who did not perceive party impact were equally divided among all voting categories. Influence on Individual Members The replies to the previous question were also coded in terms of the amount of influence the legislator perceived political parties as exerting on the behavior of individual members. 131 RELATION OF PARTISAN ATTITUDES TO PERCEPTION OF PARTY IMPACT ON LEGISLATIVE DECISIONS Democrats Republicans Four State Total ' Inde- Weak Strong Inde- Weak Strong pendent pendent (N=78) (N=6l) (N=73) (N=6l) (N=55) (N=125) One or both parties, or parties generally have impact on leg- islative decisions. 24% 38% 37% 31% 56% 58% One party has impact, other party little or no impact. 6 8 15 10 2 5 One or both parties, or parties generally have little or no impact. 57 48 34 49 35 22 No mention of party impact in reply to the question. 13 6 14 10 7 15 Total . . 100% 100% 100% 100% 100% 100% Source: LeRoy C. Ferguson, WorkingpPaper, op. cit., p. 63. 132 TABLE 35 PERCEPTION OF PARTY IMPACT ON LEGISLATIVE DECISIONS AND PARTY VOTING One or both Parties do parties have not have Total impact impact California Low Dems 10 42% 14 58% 24 100% High Dems 13 46 15 54 28 100 Low Reps 8 28 21 72 29 100 High Reps 15 46 18 54 33 100 New Jersey Low Dems 14 70%, 6 30% 20 100% High Dems 13 100 O 0 13 100 Low Reps 21 91 2 9 23 100 High Reps 21 95 l 5 22 100 Ohio Low Dems 16 59%, ll 41% 27 100% High Dems 15 54 13 46 28 100 Low Reps 34 6O 23 4O 57 100 High Reps 33 70 14 30 47 100 Tennessee Low Dems 14 29% 35 71% 49 100% High Dems 8 18 36 82 44 100 Low Reps 6 46 7 54 13 100 High Reps 4 36 7 64 11 100 133 Of those who were asked this question, 31% made comments which indicated that they perceived parties as having considerable influence on the behavior of individ- ual members. There were 17% who saw parties as having "some" or "increasing" influence on individual members. More than half of these were Californians who saw parties as having increasing influence. Twenty—seven percent of the legislators saw little or no influence by party on individual members, and the remaining 17%.made no mention of party influence on individual members in their replies to the question. The relationship of partisan attitudes to percep- tion of party influence on individual members is shown in Table 36. Here again in the combined totals the independ— ents in both parties were significantly less likely than moderate or strong partisans to receive party influence. The positive relationship between strength of partisan attitudes and perception of party influence is not as marked as it was in the case of perception of impact on decision making. This is possibly due in part to the fact that fewer of the legislators answered the Open ended ques- tion in terms of the dimension of influence on individual members. 134 TABLE 36 RELATION OF PARTISANSHIP TO PERCEPTION OF PARTY INFLUENCE ON INDIVIDUAL LEGISLATORS Question: "How would you describe the part played by political parties in the [state] legislature?" Four State Totals Democrats Republicans _4_ Inde- pendent (N=78) Weak (N=6l) Strong (N=73) Inde- pendent (N=6l) Weak (N=55) Strong (N=125) One or both par- ties, or parties generally have much influence. One or both par— ties, or parties generally have some influence. One party has influence, other party little or no influence. One or both par- ties, or parties generally have little or no in- fluence NO mention of in- fluence of party on individual leg- islators in reply to question. Total _— 17% 20 41 14 100% 31% 20 29 15 100% 25% 18 14 31 12 100% 16% 28 10 23 23 100% 42% 16 27 13 100% 46% 13 14 25 100% Source: LeRoy C. Ferguson, Working Paper, Op. cit., p.67. 135 Party Influence on Individual Members and Voting Behavior The relationship between perceived party influence on individual members and partisan voting behavior is quite irregular in most units. (See Table 37) Once again the total members in these categories are greater than the number of legislators because of the multiple punch coding of the responses. In California only one Democrat perceived the party as having much influence on individual members. Most Democrats perceived some party influence or thought this influence to be increasing, although almost as many high party voting Democrats saw little or no party influ- ence on the individual members. Among the Republicans, very few saw much party influence on individual members although more high voters saw this than low voters. How- ever, more low voters saw some or increasing party influ- ence than did high voters and more high voters saw little or no influence than low party voters. In New Jersey most Democrats perceived much party influence on individual members though more low voters saw this than high voters. Most Republicans also perceived 136 much party influence though slightly more high voters per— ceived this than low voters. Among the few Republicans who saw little or no influence more were also low party voters than high. Among Ohio Democrats, more high voters than low voters perceived a great deal of influence and there was no difference in the other categories along voting lines. Republicans in Ohio fit our model in that high party voters also perceive much party influence, low party voters mostly perceived little or no party influence on individual members. Tennessee Democrats were the Opposite of Ohio Democrats in that more low voters perceived much party influence than high voters though more low voters also perceived no influence than high party voters. Republicans though small in the number of reSponses were in similar proportions to the Democrats in that low voters perceived more influence than high voters and also more low voters perceived little or no influence at all. One group of high party voters, however, did see the parties influence on individual members as increasing. 137 PARTY INFLUENCE ON INDIVIDUAL MEMBERS Much or Some or Little considerable increasing or no Total influence influence influence California Low Dems O 0% 16 67% 8 33%, 24 100% High Dems l 4 13 50 12 46 26 100 Low Reps l 3 17 59 ll 37 29 100 High Reps 4 12 15 45 14 42 33 100 New Jersey Low Dems 19 90%. 0 0% 2 10% 21 100% High Dems 17 94 1 6 O O 18 100 Low Reps 26 87 0 0 4 13 30 100 High Reps 28 97 O 0 l 3 29 100 Ohio Low Dems ll 41% 6 22% 10 37% 27 100% High Dems 14 45 6 19 ll 35 31 100 Low Reps 17 45 7 18 14 36 38 100 High Reps 25 68 5 l4 7 18 37 100 Tennessee Low Dems 10 19% ll 21% 32 60% 53 100% High Dems 7 16 10 22 28 62 45 100 Low Reps 4 40 l 10 5 50 10 100 High Reps 2 20 5 50 3 30 10 100 138 Activities Influenced by Parties The same question of party influence was also coded in terms of activities mentioned as being influenced by party. The activities most frequently mentioned are shown in Table 38, and the activities are distributed by state, by party, and by partisan attitudes. The activities most frequently mentioned were voting with the party or against the opposition party and develOpment of party policy. The state by state totals shown in Table 38 show considerable differences among the states in number of party activities mentioned. New Jersey legislators men- tioned most of the different activities and 99% of them mentioned some party activity. It was surprising to find that California legislators were more likely than those in Ohio to mention some party activity, and even in Ten- nesse there was a higher percentage of members who men- tioned party activities than in Ohio. Because this study deals to a large extent with voting behavior, the reSponses on party activities were divided into those Who mentioned voting with the party 138 139 and those who did not mention voting with the party. These responses were then compared with the voting be- havior of the legislators. In Table 39 we find that the California Democrats once again show no consistent pattern in that high party voters both mentioned and failed to mention party voting in greater numbers than low party voters. The Republicans who mentioned voting as an important party activity were split evenly between low and high voters but among those who did not mention voting, the majority were low party voters. In New Jersey we note that twice as many Democrats in both low and high categories mentioned voting as did not mention voting as a party activity. Republicans were almost evenly divided among those mentioning and those failing to mention voting as a party activity. The overwhelming majority of Ohio Democrats did not mention voting and slightly more of these were low party voters than high party voters. Republicans sur- prise us in that among those not mentioning voting most were high voters and among those who do mention voting. slightly more were low party voters. 140 In Tennessee we see that most Democrats did not mention voting as a party activity and both those who did as well as those who did not are close to evenly divided between high and low party voters. The Repub- licans of Tennessee, though too small in number to en- able us to draw significant conclusions do follow closer to our expectations in that more high voters mentioned voting than low voters and more low voters failed to mention voting than high voters. .Hb .Q ..ufio .mo .uoaom wcwxuoa .coosmuOh .u homoa "coksom am am am m ea am as am 5mm .zv maoupm >2 mm em e e ma om -- Aom .zv xao: e m NH 4 a :H a: $N “an uzv condonedoosH : :3 NH 3 -u m cm u- “mm nzv mascaansaom 62 mm ea 2 a sa mm em Aom .zv uuouooson H mm ma 4 o ma am am AmHH.zv ooaooccoa ma om ma 2 m 3 mm “mm Aooa.zv mcouom -- in e -- -- -- mm mm Aom uzv goo: -- an e e m m am am Rom -zv apnoeaoaesH an m on n a m mm “mm Haa.zw asaoaapaaom m o oa m -- 4 mm am an . apaaoosoo m m: oa a m m Hm mma AmoHuzv oano as : 8 o : m on *8 8.1 .5 macho me -- a» e Ha -- 36 ea» Awm .zv soc: o» oa ow oa -- -- on mo» Aoa .zv apnoeaoooocH om m mm o -- m mm “mm Ago .zv assuaapaaom we -- o» Ha HH -- em mew ism .zv apauooaon Mu om a no m a H on use Awe .zv Newton :62 11 mm .9 -- Hm ea oa Hm mmm Amm uzv msoapm ma :3 e m3 an as am “ma Amm .zv no»: on mm m om mm o w mom Am: .zv opooocoooesH o2 mm m om Na ma ma mom Amn .zv acaoaapsaom as ea m ma mm 6 am mmm “on .zv apoaooaoo oa mm m om ma m ma umm AoHH.zv oasuooaaao sea ova oma A.opo .mmoupwEEOO A.opo Samoa nonuo cowuocsw cowumoso coflpmozw macaw mo ucmsucHOQ .pCostHpuoaaoou one mo Eonmoun .6633 on man Op Samoa Op Samoa -ooam Honocom -Qo .thowmmo mo nosey mofip 0v :OHpHoommo howaom o.hpuoa -ou ca mpw>wu oocofipcos cw oocoflo mo nose chap «0 cofioooaov upon co voommo ca no anemone nonuo on scapao .06 one soap moguw>wpoo uses weave -oamwuoa oowm ouspoamflmoa Hawoodo wca>og are ac whoa uoano co AOHHOQ ouos oocoHucoz owmaoOQo oz ->Hooo ponpo -uso ooHpH>Hpo< mo coHpouacowuo oaafin so wcauo> -eso ca mcfiuo> mo pgosdoao>on Ill 1' =«onzponHwOH ”weapog on» a“ oofipuoa an Ochoam when one onfiuoooo 30» canon 30m: mafimdm AHflommm mMHBHbHBo< mm mgm .9 ..eeo .mo .homom mgexhoz acooswuoh .o homoa "oopsom 38 m H em a m ewe Amaauzv ooaaoqqoe ma Ha ma ma m ma ea: “mo .zv mcouom we m o oa -- m em: AOm .zv saw; am e 4H m e Hm “mm flow .zv aucoecoaoeoe mm a me 33 0 ea am: AoHH.zv acaoaapaaom m m mm ma -- 6H um: Adm .zv opoaoosoo 03 o ma HH m 2H em: “Hoauzv ofiao -- mm ma om ma mm mom Ao: .zv msoupm e e me e 4 am eon “mm uzv saw: -- om om -- oa on mo: AoH .zv opaoesoooeqH : ma mm ea 6 am man flan uzv aqooaapsaom : om Hg P ma 3: “a: Rem .zv upoaooaoo m mm ma ma oa Hm um: Ame uzv Newton :02 an m em 6 m H mm Awo .zv oaauoeaaao onwa aceeoosg oe Samoa xenon poneo :oaaep hoeHoA: ecoocoeooce nos heeom :e cocoHecos mo sceeouemeqesoo mooeo> Hoeosuo asymoun one cos Segue Mo u0e>on nos heuom mooondo ho xenon co Noumea o coxoe henna oe oxoeeo cooaeon ooos -op one cone «0 u0e>ozop cos heuoa mo s30 mo aceeoue no: ososoo con: .heuom nee: cOeeoceeoeo 02 once ocoeecoz seweooeo oz qu>onop uoneo scenesoo oeuoedsm xenon nee: ooeo> ooeo> nos henna :ueoo muoneo has one one oeoo dos.mehoe n he: one noozeon oogouomweo one opeuoooo so» oazo: 30m « oeocom ho oosom oeoeo one ce ooos ace .oshoe nodeseo no .ocomeeuomwcoc .mxoeuo>oa .oecoocomoUQe mo coneuoooo one muoneo one .mmmflwmmmm mo poneuoooo ohm cos geoeuoo mosseoaoemoa oEOm GH: z<2 rfim¢k ho moH> mm mm *Hm How .zv psoeaoooeeH : NH om oH mm “a: HmoH.zv aquOHHnsoom oH mH :H mm mm mom Hon .zv aeouooson o mH mm Hm am an: HomH.zv oHso m om mH oH a: mo: Ho: "av weaken e Hm aH H mm am: How .zv x66: -- om om om o: mom HoH .zv paoeaoaoenH m H NH «H H: Hm mHn .z aeaOHHpsaom -- m mm P mm H: rm .2 neouooson 0 mm mH oH em up: He» .zv Nachos 56: om em m m :: mom Ham .zv oHauoeHHoo Apnoea o noxoe moooe no Heuoe one eon no: moosoov eoohoese oersn awkaHe .oHoHOa .soHeoe> nee: oeo> oe moocoemsdo Hoeeoa smoHHOQ: one GH e0: :.HHHQ hogoseeeosOo -aoo Hancouoa oe huomooooc Ho>oz -HHo Honeo o eon oH eH con: '1 non: o me ee dog: a me ouone con: :«Heuoe 6H: nee: oeo> oe Hopsoa o How metamoovu eon mH eH xcHne son 06 moocoeosdoheo eons goons: Whm Roz mom ZOHB3 ooooonnon. H H H m mm Hm noon :me n m 1H m NH H 0 H mm onom boa m H m H n n H S m S as: 6:. m m m m m m m 6‘8 >3 028 H m m H m H HH Rom SE n H H m m m 3 oaom 6H m m H H. m 2.3 63 m H H n H m 1H n mg :3 .mzolouoh. ion 5 H H H H m H 33H 63 mm H N N H m omom non H H H c. c. name 63 mH m m H H H H 2.89 30H oHHHHOMHHoO hehum oosooe cameo uuoeeoa onoHe He?» no monhe macs: noxoe non HHoHeou :HHHQ uOHBHoo oeo» 0e HHHoenoo 6H menu oHHoHHoo non: 12893 do can: aoHHHOHHH Hancouom Soooon no know .333 enooNo huoo emooxo have s 9H honose oe mhoue < z .385 .Ho>oz sooooq e0: :25 .638 eon noooon e0: 3 «H uHeosoo unoo son: Pugh EH3 GED» 902 he onathHBmDh m: gm. 158 Ohio Democrats divided almost equally along vot- ing lines in justifying not voting with the party when a bill is contrary to personal convictions. However, low voting Democrats in greater number than high party voters felt that constituency problems were a justifica- tion while high party voters felt in greater numbers that it was not necessary to vote for a ”bad bill." More high party voting Democrats believed that voting with the party was only necessary when caucus had taken a stand and also was not necessary when the leadership is wrong. Republi— cans in greater numbers felt that it was not necessary to vote with the party on matters of personal conviction and more high party voters gave constituency problems as a reason than did low party voters. More high party vot- ers felt that voting with the party was only necessary where the caucus has taken.a.stand..Four low party voters felt that it was never necessary to vote with the party while no high party voters felt this way. In Tennessee most legislators failed to distin- guish between behavior of a party man and an independent and therefore did not answer this question. Even with the few who did, more low Democratic party voters than i/_l j 159 high party voters gave personal convictions as a justifi- cation for not voting with the party. Tennessee Republi- cans showed no pattern at all on this question. Advantages of Supporting Party Leaders In the final question of the party series the legislators were asked to mention some of the advantages of going along with their party leaders when they sought support on a bill. Here again, since this question was not asked of those who did not distinguish between a party man and an independent, we have only a few cases in California and Tennessee, but most of the New Jersey and Ohio legislators replied to the question. In New Jersey the advantage most frequently men- tioned was helping the party to carry out its program. In Ohio the advantage most frequently mentioned was get- ting patronage and personal favors such as good committee assignments. Other advantages that were frequently men- tioned were getting favorable consideration for one's own bills, getting support in the election and helping to maintain the two party system by following the party label. 160 The distribution of the replies are shown by state and for New Jersey and Ohio by party and partisan atti- tudes in Table 44. In New Jersey the Republicans were more likely to mention the advantage of getting their own bills passed, but the Democrats were more likely to men- tion helping the party, maintaining the two party system and getting support in elections. Democrats in Ohio were also more likely than Republicans to mention getting elec— tion support, but Republicans were more likely to mention helping the party carry out its program. In the four state totals shown in Table 44 we note that the strong partisans were significantly more likely than independents to mention helping the party organiza- tion and getting patronage and other personal favors, while the independents were more likely to regard getting election support as an important advantage. In comparing voting behavior with the advantages of supporting party leaders we note in Table 45 that there are almost no significant results to this comparison. In California of the small number who responded to this question low voting Democrats and high voting Re- publicans felt that there was some advantage in getting 161 one's own bills passed by supporting party leaders. More low voting Democrats than high voting Democrats and more high voting Republicans than low voting ones felt that there were no advantages of supporting party leaders. In New Jersey more low party voting Republicans than high party voting ones felt that an advantage of supporting party leaders was in helping the organization. A few high voting Republicans felt that there were no ad- vantages in supporting party leaders while no other voting category felt that way. Ohio high party voting Democrats and high voting Republicans both felt that supporting party leaders helped get one's own bills passed. The high voting Democrats and low voting Republicans felt that getting support in elec~ tions was another advantage of supporting party leaders. Most of all the respondents of both parties in Ohio felt that getting personal favors was a decided advantage Of supporting party leaders. The few who felt that there were no advantages in voting with the party leaders were low voters in each party. .3 .HH :30 do anogom wfixnoa .eOnsmnoh .o nomoH «condom \ .owoeqo>oo ono gone ones cocoHeoos oEOo oosooon uOOH none onoa DHDepat n: m o: mH mm mH mom Han.zv msonam Hm oH om 6H om NH umm HHe .22 How: mm mH mm oH MH mm mom HHo .zv pnoeaoooesn aHapoa oeapm neon rm .. Pm u- n :H *:m ANN uzv oooooncoa mm H oo nH om om Hem Hem .zv meonpm m: HH on HH Hm PH *Hm Ham .zv Ham: m: HH on HH mH mm nHm Hem .zv apnoesoaoeeH e: m on nH om wH umm HmoH.zV usaoHHnsoom en mH Ho pH pH Hm gem Hm: .zv upanoosoo on 6 en nH mm mm *Hm HwnH.zv oHao 62 mm m mH oH n: mH umH Ho: .zv wsonem ,6 mm H HH wH o: HH mHm Hem .zv no»: 11 -- -- -- om om om *oH HoH .zv psoesoooeqn om 6 HH NH em HH mom HHn .zv usooHHnaaom mm .. HH mH mm mm uHH AFN nzv neonooson Hm H HH :H H: NH ueH Hm» .zv Hoanoe so: Hm am 6 mH om wH teem HHm .zv uHenonHHao Hence .oeo noon: ooeooHo one sonwonm hence ,I mcHenondso .oenosnweooo son HopoH aoHHon eso annoo no owoeco> xenon nee: ooeeHsEoo poom .soeoho henom Owe nHoz anoHeoN -oo one gene wcHeo> :H .ono>ou Hocoonoe :Hoeceoa aneHHHn -Hnowno “Hon :oHeooHo common ones ocoHecoz oomoeco>oo oz .owoconeom eoo -oeoeconn onsmnH .menom mHom sH enoneso eoo oHHHn cfio e00 ‘. :«HHHQ o co enomeso noon xooo none con: onooooH xenon n50» sen: w:OHo wceom no mowoeso>oo one no oEOm ono eons: mmMQn< :3 mqmOH oooooqnoa Rom sea umom 30H 38 68 mg 30H flu anon Hme omom 30H canon anm oaon 30H Niounoh 3oz aaom Hme onom 30H 38 6H 2.89 FOR OHHnnOHHHoo <2 nofio onooooH Lanna H2: m5»? 5 mowoenob -8266 02 comma: ocHsm doom a 3 nenom one»: Hence Lon poo omoaone uom eoo aoeohm henom N HHHoeHnH o: .. moono enobonm noHeon -qumno henna AHom 8366:. 5 29H .63... poo , 0E2 saw on. Ea mung H553 059mg; ho agaba 935a. coooom oHHHm 55 e00 164 Since so few reSponded in Tennessee, about all we can note is that only four low party voting Democrats and one high party voting Republican felt that patronage was an advantage in supporting party leaders. Almost no one in any other state considered patronage as a factor. w This chapter dealt with the identification of the symbols of the political party by the individual legislators. The variables under consideration were identification of the behavior of a "party man" as against an "independent" and the advantages given by the legislators for supporting party leaders and the justification for not supporting party leaders. One of the hypotheses that I put forth was that those legislators who mentioned voting as one of the be- haviors of a "party man" would tend to be high party voters while those who failed to mention voting would tend to be low party voters. Unfortunately, this hypoth— esis was difficult to prove or diSprove since the major— ity of legislators in both California and Tennessee made no distinction between the behavior of a "party man“ as 165 against an "independent." New Jersey members of both parties and of both voting categories mentioned voting more often than not. Ohio legislators of both parties were the only ones to confirm the hypothesis to any ex- tent. Justification for Not Voting With Parpy The responses to this question did not lend them- selves to any meaningful hypotheses. Inasmuch as this question was only asked of those who perceived behavior differences, California and Tennessee did not have enough respondents. Even in California, however, no high party voting Democrats felt that "it was never necessary to vote with the party" while most high voting Republicans did give this response. Most responding legislators of all voting categories gave reasons of personal conviction as justification for not voting with the party. Advantages of Supporting Party Leaders The responses to this question also failed to provide significant results. California and Tennessee once again had too few respondents. New Jersey low vot- ing Republicans felt that supporting party leaders helped 166 the organization. Ohio high voters of both parties felt that supporting the party leaders helped get one's own bills passed. Most of all the Ohio respondents felt that getting personal favors was another decided advantage in supporting party leaders. The hypothesis that these legislators who per- ceived behaviors of ”party men" as being different from behaviors of independents would also tend to be high party voters was confirmed by nine units and negated by seven. The hypothesis that among those legislators who perceived different behaviors those mentioning voting would also tend to be high party voters and vice versa was confirmed by six units, negated by four with six units being neutral. CHAPTER IX EFFICACY, LIBERALISM AND VOTING One of the final variables to be considered is the legislator's sense of legislative efficacy. Four items in the interview schedule were designed to try to measure the legislator's perception of whether or not he was doing a competent and effective job in his legislative work. On the basis of these items, which formed a Guttman-type scale, the legislators were divided into three groups which were labeled "high," "medium," and "low" in the question of legislative efficacy. The items used each called for an agree-disagree response and were as follows: 1. There is so little time during a session to study all the bills that sometimes I don't know what I'm voting for or against. 2. Many of the bills are so detailed and technical that I have trouble understanding them. 167 168 3. So many groups want so many different things that it is often difficult to know what stand to take. 4. My district includes so many different kinds of people that I often don't know just what the people there want me to do. The state distributions are shown in Table 46. TABLE 46 LEGISLATIVE EFFICACY AMONG THE FOUR STATES Sense of Legislative California New Jersey Ohio Tennessee Efficacy High 44% 59% 54% 48% Medium 22 28 23 26 Low 34 13 23 26 Total 100%, 100% 100% 100% Source: LeRoy C. Ferguson, Working Paper, Op. Cit p.59. We would expect that strong partisanship is linked with a professional political career and would therefore expect that strong partisans more likely than independents 169 would perceive themselves as efficacious legislators. While we might not expect this to be so in California or Tennessee since they were weak party states, we would ex- pect this hypothesis to hold true particularly among strong Republicans in Ohio and New Jersey. The data shown in Table 47 on the relation of sense of legislative efficacy to partisan attitudes do not give much encouragement to our hypothesis. In the four state totals the data suggest that strong Republicans may be slightly more likely than other Republicans to be in the high efficacy category, and it also appears that weak Democrats may be somewhat more likely than other Democrats to be in the high efficacy bracket. In California it ap— pears that the independent Republicans may be more likely than other Republicans to be high in perceived legislative efficacy, and the data for that state also suggest that strong Democrats are somewhat less likely than other Dem- ocrats to make high efficacy scale scores. There appears to be no relation between partisan attitudes and sense of legislative efficacy among the New Jersey legislators and about the only significant differ— ences are found in Ohio. As seen in Table 48 the strong 170 Republicans are significantly more likely than other Re- publicans to be in the high efficacy category. In Tennes- see independents appear to be lower in sense of efficacy than other legislators. There is a strong indication in the Tennessee data that independent Democrats are more likely to be in the low rather than the high efficacy bracket. There appears to be very little more relation be- tween voting behavior and sense of efficacy than there was between attitudinal behavior and sense of efficacy. In fact the findings are very similar. In California the Democratic low party voters have a high sense of efficacy while the high party voters have a low sense of efficacy. The same is true of low party voting Republicans who have a sense of high effi- cacy. However, high party voting Republicans are almost evenly split between low and high sense of efficacy. Once again this might be explained by the peculiar political setting of California at that time. New Jersey low voting Democrats seem to be split in their sense of efficacy while most of the high party voting Democrats appear to have a high sense Of efficacy. TABLE 47 171 RELATION OF SENSE OF LEGISLATIVE EFFICACY TO PARTISAN ATTITUDES Sense of Legislative Independent Weak Strong . Total Efficacy Dems Reps Dems Reps Dems Reps California Low (N= 37) 19% 19 l6 14 21 11 100% Medium (N= 24) 13 32 4 21 l7 13 100 High (N= 47) 15 36 19 6 ll 13 100 New Jersey Low (N= 10) 10%. -- 3O —— 3O 30 100% Medium (N= 22) 14 9 9 l4 9 45 100 High (N= 47) 4 4 15 28 9 40 100 Ohio Low (N= 35) 11% 9 6 20 20 34 100% Medium (N= 36) ll 14 3 ll 14 47 100 High (N= 86) 8 7 8 9 13 55 100 Tennessee Low (N= 28) 50% 7 l4 —— 18 11 100% Medium (N= 28) 36 17 ll -— 36 -- 100 High (N= 55) 31 7 29 13 18 2 100 Four State Totals Low (N=110) 23%. ll 14 11 21 20 100% Medium (N=110) 18 18 7 ll 19 27 100 High (N=235) 14 12 17 l3 13 31 100 Source: LeRoy C. Ferguson, Working Paper, Op. cit., p. 61. 172 TABLE 48 SENSE OF EFFICACY AND PARTY VOTING Low Medium High N A Total California Low Dems 6 20%, 3 10% 14 47% 7 23% 30 100% High Dems 15 48 5 l6 7 23 4 13 31 100 Low Reps 4 12 9 26 16 47 5 15 34 100 High Reps 12 38 7 22 ll 34 2 6 32 100 New Jersey Low Dems 4 24% 5 29% 6 35% '2 12% 17 100% High Dems 3 21 2 14 7 50 2 15 14 100 Low Reps 2 7 7 26 16 59 2 8 27 100 High Reps l 3 8 28 18 62 2 7 29 100 Ohio Low Dems 5 17% 8 28% 12 41% 4 14% 29 100% High Dems 7 26 3 11 13 48 4 15 27 100 Low Reps 14 23 12 20 27 45 7 12 60 100 High Reps 8 l3 14 23 34 56 5 8 61 100 Tennessee Low Dems 13 24% 13 24% 23 43% 5 9% 54 100% High Dems 12 22 12 22 22 4O 9 16 55 100 Low Reps 2 l3 3 l9 7 43 4 25 16 100% High Reps 4 27 2 13 5 33 4 27 15 100 Among the Republican majority, both the low party voters and the high party voters had a high sense of efficacy and almost 173 none of them had a low sense of efficacy regardless of their voting behavior. Ohio Democrats for the most part had a high sense of efficacy and this cut across voting lines. The Republicans on the other hand behaved truer to our hypothesis in that high party voters had a higher sense of efficacy than did low party voters. However, even the low party voters had the high sense of efficacy that seems to stem from being in the majority party. Tennessee Democrats behaved similarly to Ohio Repub- licans in that more Democrats of both voting categories had a high sense of efficacy. However, they showed no relation at all between voting and sense of efficacy. They are Split almost evenly in each efficacy category. The Republicans on the other hand though small in number do follow our hypothe- Sis for a minority party. The low party voters have a higher sense of efficacy than the higher party voters. Libergiism--Conservatism and Partisanship The final variable to be considered is the liberal- conservative ideology as it relates to partisan attitudes 174 and partisan voting. Four items in the interview schedule were designed to measure attitudes in the area of civil rights, government activity in economic matters, and central-versus- local government responsibility for solving local problems. On the basis of their replies to these items, each legislator was classified in the the Four State Study as either a "lib- eral" or a "conservative." The Specific items which called for an agree or dis- agree response were 2 l. A man whose loyalty has been questioned before a legislative committee, but who swears under oath that he has never been a communist, should be per— mitted to teach in our public schools. 2. The government has the reSponsibility to see to it that all peOple, rich or poor, have adequate housing, education, medical care, and protection against un- employment. 3. Business enterprise can give us our high standard of living only if it remains free from government regu- lation. 175 4. The most pressing problems which local governments face cannot be solved without new state taxes. These items were used to form a Guttman-type scale that was used to score each individual legislator according to the degree of liberalism or conservatism used in his pat- tern of replies. The distribution of partisan attitudes within the liberalism-conservatism categories is shown in Table 49. In order to interpret the table it will be helpful to know that there were considerable differences in the percentages of liberal and conservative legislators in the four states. Of the total group of legislators 61%.were classified as liberals, and in the individual states the proportion of liberals was as follows: California, 75%; New Jersey, 65%; Ohio, 52%; and Tennessee, 59%. Democrats were significantly more likely than Republicans to be liberals in California and Ohio, and a similar difference existed in New Jersey. Even in Tennessee, the Democrats made up a slightly higher percentage of the liberals than of the conservatives, al— though the difference was negligible. TABLE 49 176 RELATION OF LIBERAL-CONSERVATIVE IDEOLOGY TO PARTISAN ATTITUDES R H Democrats Republ1cans Total Inde- Inde- Strong Weak pendent Eendent Weak Strong California Conservatives (N= 26) - 4% 8 46 15 27 100% Liberals(N= 82) 21% 18 18 25 ll 7 100% New Jersey Conservatives (N= 28) 7%, l4 - 36 36 100% Liberals(N= 51) 14% 15 12 4 12 43 100% Ohio Conservatives (N= 76) 3% 3 4 10 18 62 100% Liberals(N= 82) 26% ll 15 7 6 35 100% Tennessee Conservatives (N= 46) 24% 13 41 ll 7 4 100% Liberals(N= 65) 22% 26 34 9 6 100% Total Conservatives (N=l76) 9% 7 14 15 18 37 100% Liberals(N=280) 21% 17 20 12 9 21 100% 177 Table 49-—continued. THREE STATE TOTAL (CALIFORNIA, OHIO, AND TENNESSEE) ‘r Democrats Republicans Strong Weak Inde- Inde- Weak Strong pendent pendent (N=65) (N=51) (N=72) (N=57) (N=39) (N=93) Conservatives 20% 20% 32% 44% 54% 60% Liberals 80 80 68 56 46 40 Total 100% 100% 100% 100% 100% 100% Source: LeRoy C. Ferguson, Working Paper, Op. cit., p. 43. The data in Table 49 suggests that there may be a pos- itive relationship between degree of partisan attitude and ideological position in three of the four states involved in this study. In California it appears that strong Republicans were more likely than other Republicans to be conservative and strong Democrats were more likely than other Democrats to be liberals. The data for legislators in the state of Ohio sug- gest that independent Republicans were likely to be more lib- eral than other Republicans, and that strong Democrats were more likely to be liberal than other Democrats. In Tennessee, it appears that independent Democrats were more likely than 178 weak Democrats (but not than strong Democrats) to be conservatives. In New Jersey, the relation between strength of partisan attitudes and ideology is reversed. Strong Re- publican legislators in New Jersey were Significantly more likely to be liberals than other Republicans, and independ- ent Democrats were more likely than other Democrats to be liberals. Because of this deviancy, New Jersey was left out of the second part of Table 49, where the percentages of liberals and conservatives within each partisan cate- gory are shown for the combined totals for the other three states. In this part of the table we can see clearly the tendency for both independent Democrats and independent Republicans to deviate from their parties' dominant ideo- logical position. Independent Democrats are likely to be more conservative than other Democrats, and independent Republicans are more likely than other Republicans to be liberal. In comparing the liberal—conservative ideology with voting behavior we have come up with similar findings to the relation of partisan attitude and ideology except much less clearly defined. (See Table 50) LIBERALISM--CONSERVATISM AND PARTISAN VOTING 179 TABLE 50 Conserv- Liberalism N A Total atlsm California Low Dems 3 10%, 20 67% 7 23% 30 100% High Dems O 0 27 87 4 13 31 100 Low Reps 10 30 18 55 5 15 33 100 High Reps 13 41 17 53 2 6 32 100 New Jersey Low Dems 3 20% 10 67% 2 13% 15 100% High Dems 3 20 10 67 2 13 15 100 Low Reps 13 48 12 45 2 7 27 100 High Reps 9 31 18 62 2 7 29 100 Ohio Low Dems 7 24% 19 66% 3 10% 29 100% High Dems l 3 24 83 4 14 29 100 Low Reps 28 47 25 42 7 ll 60 100 High Reps 41 68 15 25 5 7 60 100 Tennessee Low Dems 18 33% 31 57% 5 10% 54 100% High Dems 20 36 26 47 9 17 55 100 Low Reps 5 31 7 44 4 25 16 100 High Reps 5 33 6 40 4 27 15 100 180 In California for instance most Democrats of both voting categories were liberals though the high party voters were more pronouncedly so. Republicans were closer to evenly divided than they were attitudinally though more high party voters were conservatives than low party voters. New Jersey once again runs contrary to the other three states but not as Sharply as in its partisan atti- tudes and ideology. Most Democrats of both voting cate- gories were liberals and were in fact evenly divided be- tween liberals and conservatives. As in their partisan attitudes low party voting Republicans were more conserva- tive and high party voters contained more liberals. In Ohio high party voting Democrats were almost all liberals but then so were many low party voting Demo- crats. Republicans fit our hypothesis as they did in attitudinal partisanship in that high party voters were predominantly conservatives and low Republican party voters were predominantly liberal. Tennessee results differ from the attitudinal relation to ideology in that high party voting Democrats tended to be found slightly more among the conservatives while low party voters were slightly more among the 181 liberals. Republicans were about evenly divided with slightly more high and low party voters being among the liberals than among conservatives. Summary This chapter was concerned with efficacy and liberal—conservative ideology as they relate to partisan voting behavior. Efficacy It would be expected that those legislators who were high party voters would have a high sense of effi- cacy and vice versa. However, in ten of the Sixteen units the Opposite was true and low party voters were more likely to have a higher sense of efficacy than were the high party voters. This result was even more evident in the non par- tisan state of California and one party state of Tennessee than it was in the two party states of New Jersey and Ohio. Of the sixteen units, five confirmed the hypothesis but ten units negated the hypothesis and one unit did nei- ther. It would appear that low party voters have a higher sense of efficacy than do high party voters. 182 Liberalism—Conservatism A scale was constructed to measure the legislators along a liberal—conservative dimension. It would be ex- pected that Democratic liberals might be more party ori— ented and hence high party voters while Republican high party voters would tend to be conservative. In California most Democrats of both voting types were liberal and Re- publican high party voters were more conservative than low party voters. Most New Jersey Democrats of both vot— ing categories were liberals though the relationship be— tween liberalism and high party voting was negated among Republicans. Ohio Democrats were mostly liberals but the Re— publicans confirmed the hypothesis and their high voters were conservative and low party voters predominantly lib- eral. Tennessee Democrats behaved slightly more liberally. It would appear that majority and minority status of each party in each state might be an important factor account— ing for any relationship between ideology and voting be- havior. Of the Sixteen units seven confirmed the hypothesis while six negated it. Three units showed no relationship between liberal-conservative ideology and party voting. CHAPTER X SUMMARY AND CONCLUSIONS I have examined the relationship between low and high partisan voting in four state legislatures and sev- eral other factors which are of concern to students of the legislative process; attitude of legislators to par— tisanship, constituency variables, personal and career variables, perceptions of the political party by the legislators, legislative efficacy and liberal—conservative ideology. The roll call votes were those of the 1957 ses— sions of the legislatures of California, New Jersey, Ohio, and Tennessee in which more than half of one party differed from more than half of the other. The interview data were gathered during that same session in conjunction with the Four State Legislative Study. Most legislative studies, especially those involv- ing the use of roll call analysis have been limited to one state. My study suggests that there is less compara- bility among states than researchers commonly suppose. 183 184 I was unable, as will be explained below, to find a con— sistent relationship for all four states between partisan voting and any of the variables described above. Never— theless, it would appear that the data gathered in my study take on significance when they are combined with the situational data regarding the political party system prevalent in each of the four states at the time of this study. Thus I compared strong and weak voting partisans Of each party of each state with those in the other three states to isolate patterns of similarity in reSpect to the variables. I then examined the results in each state and related it to the political situation of that state and this proved to be the most rewarding feature of my study in reSpect to positive findings. Testing of Hypotheses I examined the relationship between partisan vot- ing and seventeen variables. In making tests within the four states, the strong and and weak voting partisans for each party and each legislative house were treated separ- ately. Thus each hypothesis was tested for Sixteen group- ings hereinafter referred to as the sixteen units. (Table TA” 51 m5 0? 351m: 0? 5mm amass-s 13 14 15 16 17 1112 78 10 5 6 Cllifomis . - I'.v.-I' Den High Dans Lav Reps High Reps 12 (\l 185 UK 13 lav Jersev Lev Dems High Dans Low Reps High Reps Ohio 12 d‘ 6 11 Dov Dans nigh Dans Low Reps High Rs pa 6 11 Mass-u Down-l High Dans Lav Reps High Bop. I O p a I} 0 C . h «.8 .H 0- .60 ex: “H fig:- 13 0 £0 . E85 '10-, hoe sauce 000 URI 0 “\HH \030 O\:~o beN FO‘O PEN O‘rso O 0 H0 mom O~t~o hem hr»: r-r-N Imp 186 The variables tested are listed here as seventeen hypotheses: Partisanship Scale-—Low party voters tend to be low attitudinal partisans and vice versa. Competitive Nature of District--A more party competi- tive district makes legislators lower party voters and a "safe" district makes for higher party voting. Urban-Rural Nature of District—-High party voters tend to come from urban metropolitan districts and low party voters from rural. Age—-Younger legislators tend to be high partisan voters and older ones less partisan. Education—-Higher education level makes for higher party voting and vice versa. Occupations--Lawyers tend to be more partisan in vot- ing behavior than non lawyers. Party as a Source of Decision to Enter Legislature—- Where party is mentioned as source of decision a leg- islator votes more partisanly and vice versa. 10. 11. 12. 13. 187 Legislative Experience-~Legislators with longer ex- perience tend to vote more partisanly and those of lesser experience less partisanly. Office Holding in Legislature—-Office holders are more likely to be high partisan voters than non office holders. Intention to Seek Reelection——Those seeking reelection tend to be more dedicated to politics and hence high voting partisans. Party Impact on Legislature-—Legislators who perceive the party as having an impact on the legislature would tend to vote with their party more than legislators who do not perceive this impact. Party Impact on Individual Members—-Those legislators who perceived great party influence on individual mem- bers would tend to vote more with their party than those who did not perceive this influence. Activities Influenced by Parties-—Those legislators who mentioned voting as an activity influenced by the 14. 15. 16. 17. 188 party would tend to vote with their party more fre- quently and vice versa. Behavior of Party Men--High party voters are more likely to make distinction between party men and independents. Behavior of Party Men (Continued)--High party voters more likely to mention voting as a behavior of party men . Efficacy--High party voters would tend to have a higher sense of legislative efficacy. Liberalism—-Conservatism-—Low party voters tend to be more conservative while high party voters tend to be more liberal. Surprisingly, I found that for my main hypothesis (Hypothesis #1) only six of the sixteen units showed a positive relationship between the attitudes legislators expressed towards legislative partisanship and partisan voting. Eight of the sixteen units Showed a negative re— lationship while two units neither confirmed nor negated the relationship between holding partisan attitudes and 189 and partisan voting. In only six units did a consistent relationship exist between expressed partisanship and par— tisan voting behavior. Very few of the hypotheses were confirmed to a greater degree than one would expect from chance. The variable that evoked the greatest uniformity among the four states was that dealing with office holding in the legislature. My hypothesis that holders of leadership positions in the legislature would be high party voters to a greater degree than non office holders held true in ten of the Sixteen units of study (62.5%). The only other variables that held true in more than half of the sixteen units were the effect of the party as a source of recruit- ment to the legislature, intention to seek reelection, and legislative perception of ”party men." I hypothesized that in those cases where the party was a factor in a legisla- tor's choice to run for the legislature, that legislator would tend to be a high party voter. This held true in nine of the Sixteen units (56.4%). I also hypothesized that intention to seek reelection was an indication of more than a casual interest in politics and hence those legislators intending to seek reelection would tend to be 190 high party voters. This too was confirmed in nine of the units (56.4%). The only other hypothesis that was borne out in nine units was that high party voters were more likely to distinguish between a "party man" and an "in— dependent" than were low party voters, (56.4%). The hypothesis that was negated in most instances was that dealing with legislative efficacy. I hypothe- sized that high party voters would tend to have a high sense of efficacy and low party voters a low sense of efficacy. This relationship was disProven in ten of the sixteen units. Thus it appears that low party voters were more likely to have a high sense of efficacy than high party voters (62.5%). The only other hypothesis that was disproven in more than half of our units of study was that of party impact on individual members. I hypothesized that high party voters are more likely than low party voters to perceive a party impact on indi- vidual members of the legislature. However, it is im- portant to note that the negation of this hypothesis occurred almost completely in the states of California and Tennessee, one a non partisan type state and the other a one—party dominated state. In the two party 191 states of New Jersey and Ohio this hypothesis was borne out almost completely. My other hypotheses were neither proven nor neg- ated sufficiently to enable us to reach any conclusions. These findings of weak relationships suggest several possibilities: l. The partisan voter measure was inappropriate for truly measuring differences in partisan voting. I discussed the adequacy of this measure chosen in Chapter I and concluded it was useful for my purposes after having taken the precaution of increasing the divisions of low and high to three categories (high, medium, and low) and of making the dividing line between high and low partisans at fifty percent. The results were un- changed by these procedures. I also discuss the ade- quacy of my definition of a partisan roll call. 2. The attitudinal data may be inaccurate or presented in misleading ways. The precautions taken by the in- terviewers as reported in the Four State Study suggest the former was not the case. However, the construc— tion of the partisanship scale might be criticized 192 but this alone would not account for the results. A number of variables were data concerned with specific characteristics easily ascertained (urban nature of district, age, education, and occupation of legisla- tor, etc.) and the results were the same for these as for the attitudinal variables. 3. The relationships may vary by state depending upon the political situation in that state at the time of the survey. Thus in California where a coalition of Republicans and Democrats controlled the organization of both houses, a higher sense of political efficacy might be felt by low party voters than by high party voters. In a tightly organized partisan Situation, the reverse might be true. Situational Conditions The possibility Number 3 seemed promising. First tested was the amount of agreement among the sixteen units. Table 52 shows the number of agreements among the sixteen units of this study. 193 TABLE 52 CONFIRMATIONS 0F SEVENTEEN VARIABLES AMONG THE UNITS California New Jersey Ohio Tennessee California ID ED LR HR ID ED LR HR HD LR ER ID ED LR HR LD -- 4 12 4 10 11 ED 4 -- 10 1+ LB 12 6 10 HR 4 6 New Jersey LD H538 Ohio 8 H98 Tennessee 8 BEE 194 It was a disappointment to find that while there was a possibility of agreement among my units to study in forty-eight cases, such agreement on a majority of vari- ables occurred in only nineteen cases, roughly forty per— cent. Though the results Of the study shed some light on the relationship of the political party to the legislative process, it fell short of being even close to definitive. It is Significant, however, that in the overwhelming ma- jority of cases of agreement, low party voters tended to agree with other low party voters regardless of party both in the same and other states while high party voters tended to agree more frequently with high party voters in their own and other states. Out of seventeen variables we find that California low Republicans agreed fourteen times (82.3%) with Tennessee low Democrats. This was the highest percent- age of agreement between any two of the Sixteen units. There was also substantial agreement (70.5%) between Cali- fornia high Democrats and Ohio high Democrats, California low Democrats and low Republicans, New Jersey high Repub— licans and Ohio high Democrats, Ohio low Democrats and low Republicans, and Tennessee low Democrats and Ohio low Dem- ocrats. 195 The only instances where low party voters agreed with high party voters or vice versa with any degree of regularity were the California high Republicans and the New Jersey low Democrats (50.3%), the New Jersey high Dem- ocrats and Ohio low Republicans (50.3%) and Tennessee low Republicans and high Republicans (53%). Tennessee was the only state where there was a clear cut difference along party lines as against high or low party voting lines in more than half the number of hypotheses under consideration. A further examination of the Situational factors was achieved by examining those hypotheses in respect to which a strong relationship was found with certain variables for Specific partisan groupings and relating these to the political situation within the state. This method shows some interesting relationships by providing an additional dimension to the descriptive materials. The significant relationships found within each state are marked with asterisks (*) in Tables 53, 54, 55, and 56. Following this is a brief description of the state political Situation, noting how this relates to a number of the findings. 196 TABLE 53 California--1957 Situationa1--Republican Governor Goodwin Knight Attorney General Pat Brown 1958 Knowland lost to Brown Republicans becoming more split Democrats on rise Last year of cross-filing--only fi'of legislators cross-filed Senate equally divided between parties House had Republican majority but bi—partisan leadership-- coalition control Low Dems *Attitudes correlated with voting Came from safe districts High in education *Lawyers Not recruited by party Long experience Few holders of office in legislature Tended to seek reelection See little party impact on legislature *See party impact on individual members Don't see party influencing voting Perceive party men Did not mention voting *High sense of efficacy Liberal High Dems *Attitude correlated with voting *Came from competitive districts N0 relationship to urban-rural dist's YOung in age High in education *Non-lawyers Not party recruited Long experience Few legislative office holders Tend to seek reelection See little party impact on legislature See no party impact on individual Don't see party as influencing voting Perceive party'men Did not mention voting *Low efficacy Liberal Low Reps Attitudes not correlated with voting No correlation to safensss or district *Came from urban districts No correlation to age *High in education Non-lawyers Not recruited by party Long experience *Fewer office holders Tended to seek reelection See little party impact on legislature See party influence on individuals DonFt see party influence voting no not perceive party men fifiigh sense of efficacy Liberal High Reps Attitude not correlated to voting *Came from safe districts *Rural 01d of age *NO correlation to education Non-lawyers Not recruited by party Long experience *Mbre office holders Tended to seek reelection See little party impact on legislature *See no party impact on individual Don't see party influencing voting Don't perceive party men *Low efficacy Liberal 197 TABLE 54 NEW JERSEY--l957 Situational--Republicans split between liberals and conservatives Beginnings of emergence of state-wide Democratic Party (Meyner election in 1953) Republicans controlled legislature but Democrats gained control of Assembly that year MOst control of legislative activity was worked out in Republican caucus Party discipline appeared strict at times ,Low Dems Low'Reps Neutral on correlation between atti- tudes and voting *Competitive district *Young in age High in education Lawyers Recruited by party No correlation to eXperience Few Office holders Intend to seek reelection Saw party impact on legislature Saw party impact on individual members Mentioned voting as party influence Perceived party men Mention voting of party men .High in efficacy Liberal Attitudes correlated with voting com- petitive district *old in age High in education *Not recruited by party No correlation to experience *New office holders Intend to seek reelection Saw party impact on legislature *Saw no party impact on individual members Neutral relationship to mention of voting Perceived party men Mentioned voting of party men High in efficacy *Neutral in ideology High Dems High Reps Neutral in relation between atti- tudes and voting ' *Safe district *Old in age High in education Lawyers Recruited by party No correlation to experience Few office holders Intend to seek reelection Saw party impact on legislature Saw party impact on individual members Mention voting as party influenced Perceived party men Mentioned voting of party men High efficacy __Liberal Attitudes correlated with voting Competitive district *Younger in age High in education *Recruited by party *High in experience *Many office holders Intend to seek reelection Saw party impact on legislature *Saw party impact on individual members Neutral relationship to mention of voting Perceived party men Mention voting of party men High in efficacy Liberal 198 TABLE 55 OHIO--l957 I m Situational--Republicans had strong disciplined majorities in both houses, strong enough to bridge the gap between urban and rural members Democrats split between pro-Lausche conservative, rural Demo- crats and Cleveland-anti-Lausche group Strong leadership control of legislative activities among Republicans but not among Democrats Low Dems Low Reps *No correlation between attitude and voting Competitive districts Neutral on urban-rural nature Young in age *Neutral in education Non lawyers Not recruited by party Long experience Few office holders Intend to seek reelection See party impact on legislature *See little party impact on individual members Didn't mention voting as party influenced Perceived party men High in efficacy *Liberal High Dems *Correlation between attitude & voting Competitive districts *Came from urban districts Young in age *High in education Non lawyers Not recruited by party Long exPerience Few office holders Intend to seek reelection See party impact on legislature *Ses party impact on individual members Didn't see voting as party influenced Perceived party'men *Saw voting as behavior of party man High in efficacy Liberal *No correlation between attitude and voting *Competitive districts *Neutral on urban-rural nature *Young in age High in education Non lawyers Not recruited by party Long Experience *Few office holders Intend to seek reelection *See little party impact on legislature *See little party impact on individual members Didn't see voting as party influenced Perceived party men Saw voting as behavior of party men High in efficacy *Conservative High Reps *Correlation between attitude & voting *Safe districts *Came from rural districts *Older in age High in education Non lawyers Not recruited by party *Meny office holders Intend to seek.reelection *See party impact on legislature *See party impact on individual members Didn‘t see voting as party influenced Perceived party men Saw voting as behavior of party men High in efficacy *Liberal 199 TABLE 56 TENNESSEE--l957 Situational--Predominantly Democratic but with some Republican power in eastern part of state Democratic Party was bi-factional rather than.multi-factiona1 Very little party functions in legislature, caucus was usually dormant in both parties Party officers in legislature chosen by Governor and parties are simply not a relevant factors Low Dems Low Reps No correlation between attitudes No correlation between attitudes and voting and voting Neutral on competitive nature of dis't Came from rural districts *YOung in age *High in education Non lawyers Not recruited by party High in experience Few office holders Intend to seek reelection *Saw party impact on legislature *Saw party impact on individual members Did not mention voting as party influenced Did not perceive party men High in efficacy Liberal High Dems *Came from.safe districts Came from.rural districts *Old in age Neutral on education Non lawyers Not recruited by party High in experience Few office holders *Do not intend to seek reelection *Saw party impact on individual members Did not mention voting as party influenced *Did not perceive party men *High in efficacy Neutral in ideology High Reps No correlation between attitudes and voting Neutral on competitive nature of dis't Came from.rural districts *Old in age *Low in education Non lawyers Not recruited by party High in experience Few office holders Intended to seek reelection *Didn't see party impact on legislature *Saw no party impact on individual members Did not mention voting as party influenced Did not perceive party men *Saw voting as behavior of party man High in efficacy Liberal- No correlation between attitudes and voting *Neutral on competitive nature of dis't Came from.rura1 districts ' *Young in age Neutral on education Non lawyers Not recruited by party High in experience Few office holders *Intend to seek reelection *Saw no party impact on individual members Did not mention voting as party influenced *Perceived party men *Neutral in efficacy Neutral in ideology 200 California The political situation in California at the time of this study was one of a growing Democratic ascendency with Republicans become less cohesive. This legislature was elected in the last year in which cross—filing took place but only one-fourth of the legislators were elected on both ballots. The<30vernor was a Republican, but a Democrat, Pat Brown, held the office of Attorney General. In the following election, 1958, Mr. Brown defeated Wil— liam Knowland in the race for Governor. The Senate was equally divided between the two parties. The House had a Republican majority though it was controlled by a bi- partisan coalition. Even though the House was more urban in composition and the Senate more rural, the coalition control enabled more rural control of the House than their numbers would have justified. Given this political setting, it was not surpris— ing to find that among the strong partisan Democrats there was a relationship between favorable attitude to- ward party and party voting. But such a relationship did not exist among the Republicans. It was not surpris- ing to find that the majority of legislators in all four 201 voting categories perceived very little party impact on the coalition controlled legislature. None of the groups saw the party as influencing voting. Because the Democrats were then in the process of becoming a more conscious force in the legislative process, both high and low voting Democrats perceived definite be- haviors of "party men" as differing from definite behaviors of ”independents" whereas neither category of Republicans exhibited this perception. The findings also are related to the factional fights in both parties. The high party voting Democrats tended to be non—lawyers who came from competitive dis- tricts, young in age with high partisan attitudes even though they perceived little party influence on the leg- islature at that time. Republicans on the other hand had older members from safe districts mostly rural in nature as their high party voters. They perceived no party influence at all and their attitudes on party bore no relationship to their party voting. 202 New Jersey At the time of our study the Republicans had a tightly knit legislative organization in New Jersey. A state—wide Democratic Party was fast emerging with the election of Robert Meyner as governor and the achievement of Democratic control of the Assembly later that year. In the legislature, control of activity was worked out in the Republican caucus and party discipline appeared to be strict. The behavior of each party was almost reversed from what it was in California. In New Jersey it was the Republicans rather than Democrats whose partisan attitudes bore a relationship to voting and who had more differences between high and low party voters. The only differences between low and high party voting Democrats were that the low party voters tended to be younger and from competitive districts while the high party voters tended to be older and from safe districts. Low voting Republicans on the other hand were older, while the high party voters tended to be younger. New Jersey being much more party oriented than California, and the organization and Operation of the legislature reflects this fact. Only low party voting 203 Republicans did not owe their recruitment to office to their party. All four voting categories saw a party im- pact on the legislature and all but low party voting Re- publicans saw a party impact on individual members. The high voting Republicans were those high in experience while the other three categories were neutral on this variable. Holders of party office in the legislature were mostly high voting Republicans. All four categories perceived behavior of "party men" as significant and the majority of all four voting categories mentioned party voting as one of the behaviors of "party men." All four categories of legislators were high in their sense of efficacy and all but the low party voting Republicans were liberal in ideology. Ohio Ohio Republicans had strong disciplined majorities in both houses of the legislature and as a result were able to bridge the gap between rural and urban members. Democrats were split between the pro-Lausche conservative rural members and the Cleveland anti—Lausche group. Re- publicans had strong leadership control of legislative activities but this was not the case with the Democrats. 204 The low party voters of both parties showed no re- lationship between party attitudes and partisan voting while the high party voters did. High party voting Demo- crats appeared to be pitted opposite to high voting Repub— licans on most variables. Ohio legislators came closest to fitting a model of two party competition. High partisan voting Democrats came from urban districts while low partisan voting Democrats showed no pattern on this variable. High partisan Democrats were higher in education than most of the low partisan Demo- crats and saw party impact on individual members while low partisan voting Democrats did not. High partisan voting Democrats saw party voting as one of the behaviors of "party men” while this was not the case among the low voters. The high partisan voting Republicans tended to come from safe, rural districts and were older in age than the low voting Republicans. The high partisan Republicans had many holders of office in the legislature and unlike their low voting colleagues, they perceived party impact both on the legislature and on the individual members. Ideologically, the high party voting Republicans were 205 liberal while the low voters were more conservative. Both types of Democrats fell mostly in the liberal cate— gory. In general Ohio behaved like a two party state in that there were significant differences along several dimensions both within each party and between each party. Tennessee Tennessee was almost a classic one party state, predominantly Democratic, but with some Republican power in the eastern part of the state. There was very little party function in the legislature and the caucus of each party met very rarely. Party officers in the legislature were chosen by the governor and parties were simply not a relevant factor in the legislative process. Considering this situation it is not surprising to find that none of the four categories showed any re- lationship between party attitude and party voting. The majority of the legislators of all categories came from rural districts with only the low party voting Republicans coming from safe districts. 206 The low voting Democrats tended to be in the younger age group while the high voters were older; ex- actly the opposite of the Republicans where the older members tended to be low party voters and the younger ones tended to be high party voters. Low party voting Democrats tended to be higher in education than high party voters while the Republicans showed no difference in this regard. Interestingly, only low voting Democrats saw party impact on the legislature; high Democrats saw no such im- pact and Republicans of both types were neutral on this. Low party voters of both parties saw party impact on the individual members of the legislature while high voters of each party saw no such impact. None of the four cate- gories saw voting as being party influenced, and only high voting Republicans perceived "party men" as against "inde- pendents." Only low party voting Republicans did not intend to seek reelection though all but high voting Republicans had a high sense of efficacy. Both types of Democrats were liberal in ideology while both types of Republicans were neutral in the relationship between party voting and ideology. ‘207 Conclusions My major overall finding was that partisan voting behavior is in part a situational variable, that is under certain political conditions legislators with certain at- titudinal or other characteristics are likely to be strong partisans whereas given a different political situation their voting behavior might be markedly different. The most consistent findings in respect to any variable, and these it should be noted occurred only slightly more than chance, were relationships between party voting and variables suggesting positive party ac— tivity (recruitment of legislators, holding party leader- ship offices, intention to seek reelection, and clear identification of party men in the legislature). That these conditions of positive party activity did not char— acterize all four states perhaps accounts for some of the lack of confirmation of hypotheses in respect to these variables. Among the most important factors in the political situation of a state is the tradition of non partisanship or one party dominance. Legislative behavior in these 208 states would be expected to differ substantially from states which have more of a two party tradition. In this study this was borne out by the fact that members of the four voting categories of California tended to agree on more items with their counterparts in Tennessee than with those in either New Jersey or Ohio. Likewise, the four voting categories in New Jersey tended to agree with their counterparts in Ohio more than with either those in Calif- ornia or Tennessee. There were no units in New Jersey that agreed on 50% of the hypotheses with any units in Tennessee. It would appear that some of the results point out that the need was for combining several approaches in order to derive a more complete, composite picture of po- litical parties in legislatures. Among these approaches would be greater "in depth" study of the political party situation prevalent in each state and the bearing that situational variables have on partisan voting in legis— latures. Certainly the expectations of a political party that is the probability of its winning or losing control of legislature or governorship, the degree to which he enunciates a program, its relationship to the present gov— ernor, and other factors in many cases account for differ— ences in behavior. 209 These variables and many others that appear prom- ising were not Operationalized in this study. Their sig— nificance was unanticipated when the study was undertaken both by the present writer and presumably by the researchers who conducted the four state study. In undertaking future research, a number of pre- cautions would seem to be in order. It is probable that the relationship between partisan voting behavior and other variables can result from Opposite motivations. Great care would have to be exerted to differentiate between situa- tional elements as against idiosyncratic elements as they affect legislative partisanship. More work is necessary in Operationalizing situational variables so as to be able to relate them to other aspects of legislative behavior. This study by pointing out several variables which were unrelated to the political party voting in the legislture hopefully does help delineate and suggest some of the fruitful areas for further research. BIBLIOGRAPHY 139219 Acheson, Dean. A Citizen Looks at Congress. New York: Harper and Brothers, 1956. Bailey, Stephen K. Congress Makes A Law: The Story Be- hind the Employment Act of 1946. New York: Col- umbia University Press, 1950. Bentley, Arthur F. The Process of Government; A Study of Social Pressures. Bloomington, Indiana: The Principia Press, Inc., 1949. Blaisdell, Donald C. American Democracy Under Pressure. New York: The Ronald Press Co., 1957. (ed.). "Unofficial Government: Pressure Groups and Lobbies," The Annals of The American Academy of Political and Social Science. Philadelphia, 1958. Buchanan, William. Legislative Partisanship! the Deviant Case of California. Berkeley: University of California Press, 1963. Burns, James M. Congress on Trial. New YOrk: Harper and Bros., 1949. . Deadlock of Democracy. Englewood Cliffs, New Jersey: Prentice-Hall, 1963. Chase, Stuart. Democracy Under Pressure, Special Interest vs.the Public Welfare. New YOrk: The Twentieth Century Fund, 1945. 210 211 Crawford, Kenneth G. The Pressure Boys: The Inside Story of Lobbying in America. New York: J. Messner, 1939. Derge, David R. "Urban-Rural Conflict: The Case in Illi- nois,“ in Legislative Behavior edited by John C. Wahlke & Heinz Eulau. Glencoe: Free Press, 1959. Eldersveld, Samuel J. "American Interest Groups: A Sur- vey of Research and Some Implications for Theory and Method,“ in Henry W. Ehrman (ed.) Interest Groups on Four Continents. Pittsburgh: Univer- sity of Pittsburgh Press, 1958. Epstein, Leon. Politics in Wisconsin. Madison: Univer- sity of Wisconsin Press, 1958. Finletter, Thomas K. Can Repgesentative Government Do the Job? New York: Reynal & Hitchcock, 1945. Follett, Mary P. The Speaker of the House of Represen- tatives. New York: Longmans, Green, 1896. Froman, Lewis A. Jr. Congressmen and Their Constituencies. Chicago: Rand McNally & Co., 1963. Frost, Murray. Roll Call Cohesion in the New Jersgy Leg- islature, 1956-57. Unpublished Master's thesis, Michigan State University, 1959. Galloway, George B. Congress at the Crossroads. New York: Crowell, 1946. Gross, Bertram M. The Legislative Struggle. New York: McGraw—Hill, 1953. Herring, E. Pendleton. Group Representation Before Con- gress. Baltimore: The Johns Hopkins Press, 1929. Jewell, Malcolm E. The State Legislature-Politics and Practice. New YOrk: Random House, 1962. Kefauver, Estes, and Levin, Jack. A Twentieth Century Congress. New YOrk: Duell, Sloan and Pearce, 1947. 212 Key, V. 0. Jr. Politics, Parties and Pressure Groups. New York: Thomas Y. Crowell Company, 1958. . Southern Politics in State and Nation. New York: Knopf, 1949. Laswell, Harold. Power and Personality. New York: W.'W. Norton & Co., 1948. . Psychopathology and Politics. Chicago: Uni- versity of Chicago Press, 1930. Latham, Earl. Group Basis of Politics-—A Study in Basing Point Legislation. Ithaca: Cornell University Press, 1952. Lowell, A. Lawrence. "The Influence of Party Upon Legis- lation," Annual Report of the American Historical Association. Vol. 1. Washington, D. C.: U. S. Government Printing Office, 1901. Matthews, Donald R. United States Senators and Their World. Chapel Hill: University of North Caro- lina Press, 1960. McKean, Dayton D. Pressures on the Legislature of New Jersey. New York: Columbia University Press, 1938. Neumann, Sigmund (ed.). Modern Political Parties, Approaches to Comparative Politics. Chicago: University of Chicago Press, 1956. Odegard, Peter H. Pressure Politics: The Study of the Anti Saloon League. New YOrk: Columbia Univer— sity Press, 1928. Ranney, Austin (ed.). Essays on the Behavioral Study of Politics. Urbana: University of Illinois Press, 1962. Ranney, Austin, and Kendall, Willmore. Democracy and the American Party System. New York: Harcourt, Brace and Company, 1956. 213 Rice, Stuart A. Quantitative Methods in Politics. New York: KnOpf, 1928. Riggs, Fred W. Pressures on Congress. New York: Kings Crown Press, Columbia University, 1950. Schriftgiesser, Karl. The Lobbyists, The Art and Busi- ness of Influencing Lawmakers. Boston: Little Brown and Co., 1951. Schattschneider, E. E. Pressure and the Tariff. New YOrk: Prentice-Hall, 1935. Scigliano, Robert G. Michigan Legislative Report, 1954; Regular and Special Sessions. East Lansing, Mich- igan: Governmental Research Bureau, Michigan State College, 1955. Senate Fellows of Michigan State, 1963—64. The Legisla— tive Process--A Bibliography in Legislative Be- havior. East Lansing, Michigan: Institute for Community Development and Services, Michigan State University, 1963. Sorauf, Frank J. Party and Representation. New York: Atherton Press, 1963. Toward a More Responsible Two Party System. A Report of the Committee on Political Parties, American Po— litical Science Association. New YOrk: Rinehart & Co., Inc., 1950. Truman, David B. The Congressional Party; A Case Study. New York: John Wiley, 1959. . The Governmental Process. New York: KnOpf, 1951. Turner, Julius. Party and Constituency: Pressures on Congress. "Studies in Historical and Political Science." Baltimore: The Johns Hopkins Press, 1951. 214 Wahlke, John C., and Eulau, Heinz. Legislative Behavior: A Reader in Theory and Research. Glencoe: The Free Press, 1959. Wahlke, John C.: Eulau, Heinz; Buchanan, William; and Ferguson, LeRoy C. The Legislative System. New YOrk: John Wiley & Sons, 1962. Wilson, Woodrow. Congressional Government. Boston: Houghton Mifflin, 1885. Young, Ronald. This is Congress. New YOrk: Knopf, l943. Zeller, Belle (ed.). American State Legislatures. New YOrk: Thomas Y. Crowell Company, 1954. Zeller, Belle. Pressure Politics in New York. New York: Prentice Hall, 1937. Articles and Papers Belknap, George M. "A Method of Analyzing Legislative Behavior," Midwest Journal of Political Science, II (1958), pp. 377-402. Buchanan, William. "California Legislature: Structure and Practices." Unpublished working paper, 1958. Crane, Wilder. "A Caveat on Roll Call Studies of Party Voting," Midwest Journal of Political Science, IV (1960), pp. 237-249. Derge, David R. "Metropolitan and Outstate Alignments in Illinois and Missouri Legislative Delegations," American Political Science Review, LII (December 1958), pp. 1051-1065. Dye, Thomas R. "A Comparison of Constituency Influence in the Upper and Lower Chambers of a State Leg- islature," Western Political Quarterly, XIV (June 1961), pp. 473-480. 215 Eulau, Heinz. "Institutional Parameters for Legislative Behavior in Ohio.” Unpublished working paper, 1958. Eulau, Heinz; Buchanan, William; Ferguson, LeRoy; and Wahlke, John. "The Role of the Representative: Some Empirical Observations on the Theory of Edmund Burke," American Political Science Review, LIII (September, 1959), pp. 742-756. Ferguson, LeRoy C. New Jersey Legislative Notes. Unpub- lished working paper, 1958. Friedman, Robert S. "The Urban-Rural Conflict Revisited," Western Political Quarterly, XIV (June 1961), pp. 481-495. Froman, Lewis A. Jr. "Inter Party Constituency Differences and Congressional Voting Behavior,” American Politi- cal Science Review, LVII (March 1963), pp. 57—62. Golembiewski, Robert T. "A Taxonomic Approach to State Political Party Strength," Western Political Quar- terly, XI (1958), No. 3, pp. 494—513. Greenstein, Fred 1., and Jackson, Alton F. "A Second Look at the Validity of Roll Call Analysis,” Mid— west Journal of Political Science, VII (May 1963), pp. 156-166. Huitt, Ralph K. "The Congressional Committee: A Case of Study," American Political Science Review, XLVIII (1954), pp. 340-365. Jewell, Malcolm E. "Party Voting in American State Legis- latures," American Political Science Review, XLIX (September 1955), pp. 773-791. Keefe, William J. "Comparative Study of the Role of P0- litical Parties in State Legislatures,” Western Political Quarterly, IX (September 1956), pp. 726-742. 216 Krislov, Samuel. "Constituency vs Constitutionalism: The Desegregation Issue and Tensions and Aspira- tions of Southern Attorneys General," Midwest Journal of Political Science, III (Feb. 1959), pp. 75-92. Lockard, W. Duane. "Legislative Politics in Connecti- cut," American Political Science Review, XLVIII (March 1954), pp. 166—173. Luce, R. Donald, and Rogow, Arnold A. "A Game Theoretic Analysis of Congressional Power Distributions for A Two Party System," Behavioral Science, I (April 1956), p. 83. MacRae, Duncan Jr. "Roll Call Votes and Leadership," Public OpinionyQuarterly, XX (Fall 1956), pp. 543-558. "The Relation Between Roll Call Votes and Con- stituencies in the Massachusetts House of Repre— sentatives," American Political Science Review, XLVI (December, 1952), pp. 1046-1055. Matthews, Donald R. "United States Senators and the Class Structure," Public Opinion Quarterly, XVIII (Spring 1954), pp. 5-22. McConoughy, John B. "Some Personality Factors of State Legislators in South Carolina," American Politi- cal Science Review, XL (Dec. 1950), pp. 897-903. Meller, Norman. "Legislative Behavior Research," Western Political Quarterly, XIII (March 1960), pp. 131— 153. Schlesinger, Joseph A. "A Two-Dimensional Scheme for Classifying the States According to Degree of Inter—Party Competition," American Political Science Review, XLIX (1955), pp. 1120-28. "Lawyers and American Politics," Midwest Journal of Political Science, I (May 1957), pp. 26-39. 217 Schlesinger, Joseph A. "The Structure of Competition for Sh apley I Wahlke I Wahlke, Office in the American States," Behavioral Science, V (July 1960), pp. 197-210. L. S., and Shubik, Martin. "A Method for Evalu- ating the Distribution of Power in a Committee System," American Political Science Review, XLVIII (September 1954), p. 789. John C. ”A Note on the Habitat and Habits of Tennessee Legislators." Unpublished working paper, 1958. John C.: Buchanan, William; Eulau, Heinz; and Ferguson, LeRoy C. "American State Legislators' Role Orientations Toward Pressure Groups," Jour— nal of Politics, XXII (1960), pp. 203-227. . "The Legislator As a Specialist," Western Po- litical Quarterly,-XIII (September 1960), pp- 636-651. "The Political Socialization of American State Legislators," Midwest Journal of Political Science, III (1959), pp. 188-206.