¹ The past two years have produced others sympathetic to this position: Lockard, Duane, “State Party Systems and Policy Outputs,” in Political Research and Political Theory, ed. Garceau, Oliver (Cambridge: Harvard University Press, 1968), pp. 190–220Google Scholar; Sharkansky, Ira and Hofferbert, Richard I., “Dimensions of State Politics, Economics and Public Policy,” American Political Science Review, 63 (09, 1969), 867–879.CrossRefGoogle Scholar

² Dye, Thomas R., Politics, Economics, and the Public: Policy Outcomes in the American States, American Politics Research Series (Chicago: Rand McNally & Co., 1966).Google Scholar Dye also considers highway, educational and tax policies and morality regulation and concludes that party competition does not have a measurable impact on these policy outcomes once socioeconomic variables are controlled. In a later work Dye finds an association between income inequalities and degree of centralization of governmental spending and services at the state level: “Income Inequality and American State Politics,” American Political Science Review, 63 (March, 1969), 161. See also Dawson, Richard E. and Robinson, James A., “Inter-Party Competition, Economic Variables, and Welfare Policies in the American States,” Journal of Politics, 25 (05, 1963), 281–289CrossRefGoogle Scholar; Hofferbert, Richard I., “The Relation Between Public Policy and Some Structural and Environmental Variables in the American States,” American Political Science Review, 60 (03, 1966), 73–82.CrossRefGoogle Scholar

³ Casstevens, T. W. and Press, C., “The Context of Democratic Competition in American State Politics,” American Journal of Sociology, 68 (03, 1963), 542–543.CrossRefGoogle Scholar See also Crittenden, John, “Dimensions of Modernization in the American States,” American Political Science Review, 61 (12, 1967), 989–1001.CrossRefGoogle Scholar Crittenden suggests that differences in taxing, spending and employing between the states may connote levels of effort in the attempt to provide a uniform level of services. Fenton, John H., People and Parties in Politics, American Government Series (Glenview, Illinois: Scott. Foresman and Company, 1966), p. 44.Google Scholar Multiple correlations showed that 55 per cent of the variations in aid to dependent children from state to state were related to two-party competition plus urbanism plus the per-capita income of the state, and the partial regression coefficients indicated that 75 per cent of the total relationship was attributable to the two-party competition variable. Per-pupil expenditures and total welfare expenditures also correlated significantly with two-party competition.

⁴ Sharkansky, Ira. Spending in the American States, American Politic Research Series (Chicago: Rand McNally, 1968).Google Scholar Sharkansky does not use the combined expenditures of state and local governments used by the previous writers, but considers state expenditures alone. He finds that a combination of three variables: previous expenditures, the state percentage of state-local expenditures, and total local government expenditures per capita can account for 95 per cent of the interstate differences in total state spending per capita for 1962. He finds, however, that the most powerful correlates of the percentage of change in total expenditures per capita from 1962–1965 account for only 50 per cent of interstate variation. I believe that many of the variables Sharkansky calls independent are in effect dependent variables. In this category are his “government measures”: federal aid, taxes, state-local financial relationships, state employees, as well as his measure of previous expenditure. None of these can adequately explain expenditure change which, I maintain, is where the activity of political parties comes in. In his Regionalism in American Politics (New York: Bobbs-Merrill Co., Inc., 1970), he comes to the same conclusions with respect to current expenditures but does not consider change. In another article, Sharkansky finds that a favorable recommendation from the governor seems essential for agency budget success in the legislature: “Agency Requests, Gubernatorial Support and Budget Success in State Legislatures,” American Political Science Review, 62 (December, 1968), 1224–1231. It is assumed by all the above writers that socioeconomic variables indicate policy preferences of the electorate. A study by Leon D. Epstein indicates the difficulty of making any assumptions about the policy preferences of the statewide electorate even in an election in which the major issue was tax policy. See Epstein, Leon D., “Electoral Decision and Policy Mandate: An Empirical Example,” Public Opinion Quarterly, 28 (Winter, 1964), 564–572.CrossRefGoogle Scholar

⁵ Key, V. O., Politics, Parties, and Pressure Groups, 5th edition (New York: Thomas Y. Crowell Co., 1964), pp. 295–304Google Scholar; Lockard, Duane, New England State Politics, 2nd edition (Princeton: Princeton University Press, 1959), pp. 320–340Google Scholar; The Politics of State and Local Government (New York: The Macmillan Company, 1969), p. 179; and Jewell, Malcolm E., “Party Voting in American State Legislatures,” American Political Science Review, 49 (09, 1955), 773–791.CrossRefGoogle Scholar See also Jewell's, discussion of this same finding in his book The State Legislature: Politics and Practice (New York: Random House, 1959), pp. 112–117Google Scholar, and Fenton, John H., Midwest Politics (New York: Holt, Rinehart and Winston, 1966).Google Scholar Fenton uses a measure of “effort” in order to prove that issue-oriented state parties are more responsible than job-oriented state parties. “Effort” is measured by the percentage of per-capita income collected as per-capita state and local tax revenue from own sources. This is based solely on personal income and does not reflect other indices of a state's financial resources (i.e., corporate income and wealth, property valuations), which, if available in addition to personal income data, would be a more accurate measure of a state's financial ability. See Barber, James D., “Leadership Strategies for Legislative Party Cohesion,” Journal of Politics, 28 (05, 1966), 347–367CrossRefGoogle Scholar; Patterson, Samuel C., “Legislative Leadership and Political Ideology,” Public Opinion Quarterly, 27 (Fall, 1963), 407CrossRefGoogle Scholar; McCally, Sarah P., “The Governor and his Legislative Party,” American Political Science Review, 60 (12, 1966), 923–942.CrossRefGoogle Scholar

⁶ Lockard, , “State Party Systems and Policy Outputs,” p. 207.Google Scholar See the review of the literature on apportionment in Dye, Thomas R., “State Legislative Politics,” in Politics in the American States: A Comparative Analysis, ed. Jacob, Herbert and Vines, Kenneth N. (Boston: Little, Brown and Co., 1965), pp. 157–165.Google Scholar See also Pulsipher, Allan C. and Weatherby, James L. Jr., “Malapportionment, Party Competition and the Functional Distribution of Governmental Expenditures,” American Political Science Review, 62 (12, 1968), 1207–1219CrossRefGoogle Scholar, and the “Communications to the Editor” on the subject, American Political Science Review, 63 (June, 1969), 528–530.

⁷ Jacob, Herbert, “The Consequences of Malapportionment: A Note of Caution,” Social Forces, 43 (12, 1964), 256–261CrossRefGoogle Scholar; Hofferbert, Richard I., “Relation Between Public Policy and Structural and Environmental Variables,” pp. 73–82.Google Scholar and Dye, , Politics, Economics, and The Public, pp. 270–281.Google Scholar

⁸ Froman, Lewis A. Jr., Congressmen and Their Constituencies (Chicago: Rand McNally and Co., 1963), pp. 85–97.Google Scholar

⁹ Derge, David R., “Metropolitan and Outstate Alignments in Illinois and Missouri Legislative Delegations,” American Political Science Review, 52 (12, 1958), 1051–1065.CrossRefGoogle Scholar Since this study, it has been acknowledged that party is a more important determinant than rural-urban conflict, and subsequent studies have dealt with the presumed relationship between similarity of district and degree of ideological cohesion within the party contingent. Thomas Flinn also tested this theory of party unity based on similarity of district in the Ohio Assembly of 1949–1959. He concluded that differences in constituency do relate to intraparty differences with members from districts typical of the party supporting the party position more often than other members. However, constituency characteristics will not account for all the variance or even most of it. Flinn, Thomas A., “Party Responsibility in the States: Some Casual Factors,” American Political Science Review, 58 (03, 1964), 61–66.CrossRefGoogle Scholar

¹⁰ Dye, Thomas R., “A Comparison of Constituency Influences in the Upper and Lower Chambers of a State Legislature,” Western Political Quarterly, 14 (06, 1961), 473–481CrossRefGoogle Scholar; Sorauf, Frank J., Party and Representation: Legislative Politics in Pennsylvania (New York: Prentice-Hall, 1963), pp. 141–144.Google Scholar Both writers used all party opposition votes, whether or not the issues were economic. See also MacRae, Duncan, “The Relationship Between Roll Call Votes and Constituencies in the Massachusetts House of Representatives,” American Political Science Review, 46 (12, 1952), 1046–1055.CrossRefGoogle Scholar We do not know whether the roll calls selected by MacRae to reflect class differences were administration supported measures or whether deviations were sizable enough to prevent passage.

¹¹ MacRae, pp. 1046–1055; Dye, , “A Comparison of Constituency Influences,” p. 477Google Scholar; Flinn, , “Party Responsibility in the States,” p. 67Google Scholar; Patterson, Samuel C., “The Role of the Deviant in the State Legislative System: The Wisconsin Assembly,” The Western Political Quarterly, 14 (06, 1961), 460–472.CrossRefGoogle Scholar

¹² Becker, Robert W., Foote, Frieda L., Lubega, Mathias and Mosma, Stephen V., “Correlates of Legislative Voting: Michigan House of Representatives, 1954–1961,” Midwest Journal of Political Science, 6 (11, 1962), 384–396.CrossRefGoogle Scholar The number of members from districts classified as close was so small as to make the finding doubtful. With respect to primary competition, the hypothesis that a member would support the state party position if he won in a close primary is open to question. He would probably support the ideological position of the wing of the party from which he received help. This might not be the state-party position.

¹³ Sorauf, p. 141. Sorauf discovered a correlation between the results of the gubernatorial race in a legislative district and party loyalty. This may indicate a degree of state party organization and support in the district which does influence the behavior of a legislator.

¹⁴ Froman, pp. 110–121.

¹⁵ This research was performed on 140 legislative sessions in thirty-four northern states during 1946–1960. (Nebraska and Minnesota were excluded because their legislatures were elected on a nonpartisan basis. New Mexico did not publish a record of the legislative proceedings during this period). See the discussion justifying the use of veto votes to measure gubernatorial policy influence. In the present study two indices were used to measure the degree to which the governor's party supported him on the veto vote: the Index of Administration Support, which measures the degree to which the governor's party members support him and is obtained by dividing the number of votes cast by party members who voted for the governor by the total number of party members who voted; and the Index of Administration Success, which indicates the degree to which the governor was successful in obtaining enough votes from his party to uphold his veto, and is obtained by dividing the percentage of those who voted in favor of the governor by the percentage of his party votes he needed in order to sustain his veto (McCally, , “The Governor and his Legislative Party,” pp. 926–930).Google Scholar

¹⁶ Key, V. O., American State Politics: An Introduction (New York: Alfred A. Knopf, 1956), pp. 109–118Google Scholar; and Schlesinger, Joseph A., How They Became Governor (East Lansing: Governmental Research Bureau, Michigan State University, 1957), p. 27.Google Scholar For the measures for each factor, see McCally, p. 931.

¹⁷ This finding is confirmed by Francis, Wayne L., Legislative Issues in the Fifty States: A Comparative Analysis, (Chicago: Rand McNally and Co., 1967), p. 35.Google Scholar Francis also discovered that factional conflict must be viewed in relation to the size of the majority and minority parties. Within majority parties, the degree of factional conflict is lowest when the party has between 60 and 69.9 per cent of the seats and highest when the party has between 80 and 89.9 per cent of the seats.

¹⁸ An R of .55 was obtained between the success of a governor and the above independent variables (McCally, pp. 935–940).

¹⁹ Sorauf, , Party and Representation, pp. 43–62.Google Scholar

²⁰ Francis, , Legislative Issues in the Fifty States: A Comparative Analysis, p. 23.Google Scholar

²¹ Four states were eliminated because there were no economic issues among the vetoes: Illinois, 1957; Ohio, 1957; Michigan, 1957; Oklahoma, 1953. Typical of the vetoes used were: rent control, sales tax, income tax, property tax, labor union control, public welfare, old age assistance. The four high support sessions were also eliminated because the support for the governor was so high that divisions among the legislators could not be measured. The sessions eliminated in this way were: Kansas, 1959-D; Massachusetts, 1949-D and 1957-D and Massachusetts 1955-R. In all of them the governor was over 100 per cent successful in receiving the required support of his party and garnered over 82 per cent of the members. Not enough of Utah's thirty-three legislators ran again to make an analysis meaningful.

²² MacRae, , “Relation Between Roll Call Votes and Constituencies,” p. 1051.Google Scholar In all of th sessions there was a negative correlation between per cent urban and owner-occupancy of over .45 except for Oregon and Iowa where it was negative but lower. The generalization may be made that within both parties, a high per cent owner-occupancy represents the rural end of a scale and a high degree of a “city” variable such as population or per cent urban represents the urban extreme. In this research, two economic variables were used for each legislator, one an urban variable and the other the per cent owner-occupancy. I tried not to use any independent variable which was intercorrelated with another independent variable at higher than .50 because highly inter-correlated independent variables lead to coefficient estimates (in a sampling sense) biased away from their true value (if an underlying population could be assumed) toward zero. The following exception to this rule had to be made: In Arizona, 1947 and 1949, per cent urban and per cent owner-occupancy correlate at —.58 and —.57 respectively; for Colorado, population correlates —.61 with owner-occupancy, —.58 with legislator's election and —.72 with legislator's primary; also for Colorado, owner-occupancy correlates .76 with legislator's primary.

²³ A good explanation of the technique appears in Blalock, Hubert M., Social Statistics (New York: McGraw-Hill Book Company, Inc., 1960), chaps. 1719.Google Scholar For two applications see Stokes, Donald E., Campbell, Angus and Miller, Warren E., “Components of Electoral Decision,” American Political Science Review, 52 (06, 1958), 367–387CrossRefGoogle Scholar; and Alker, Hayward R. Jr., “Dimensions of Conflict in the General Assembly,” American Political Science Review, 58 (09, 1964), 642–657.CrossRefGoogle Scholar

²⁴ The regression equation for explaining a legislator's index of support (Sup.) is:

In this equation the i subscript may indicate any legislative district being analyzed. K is an appropriate constant; Pop., is the population in thousands represented by the legislator; Urb., is per cent urban in the legislator's district; Own₁ is the per cent owner-occupancy in the legislator's district; L.E., is the legislator's per cent of the postsession election vote; L.P., is the legislator's per cent of the postsession primary vote: G.P., is the Governor's per cent of the postsession gubernatorial primary vote in the legislator's district.

²⁵ This makes the comparison between the variables more equitable. Changes in the dependent variable are measured in terms of standard deviation units for each of the other variables, a method which assures the same variability for each of these variables. Beta weights are obtained by multiplying a concrete b-coefficient times the standard deviation of its independent variable and dividing by the standard deviation of the dependent varable. (K equals zero when the independent variables are standardized.)

²⁶ Schlesinger, Joseph A., Ambition and Politics: Political Careers in the United States, American Politics Research Series (Chicago: Rand McNally & Co., 1966), pp. 86–88.Google Scholar

²⁷ Salisbury, Robert H., “The Analysis of Public Policy: A Search for Theories and Roles,” in Political Science and Public Policy, ed. Ranney, Austin (Chicago: Markham Publishing Co., 1968), pp. 163–174.Google Scholar

²⁸ Sharkansky, and Hofferbert, , “Dimensions of State Politics, Economics, and Public Policy,” pp. 876–879.Google Scholar See also Cnudde, Charles F. and McCrone, Donald J., “Party Competition and Welfare Policies in the American States,” American Political Science Review, 63 (09, 1969), 858–866.CrossRefGoogle Scholar