Federal Frustration, State Satisfaction? Voters and Decentralized Governmental Power.
In the United States, the federal government and the state governments experience periods of increasing and decreasing influence over policymaking as well as times of waxing and waning citizen affection. In recent years, conditions seem to have been improving for state governments on both counts. Leaders of both political parties have treated citizen feelings about state government as a significant force in national politics.
For the Republicans, Newt Gingrich's "Contract with America," with its focus on decentralized governmental authority, helped the GOP to take over both the U.S. House and the U.S. Senate in the 1994 midterm elections.  In 1996, Bob Dole frequently cited the Tenth Amendment and promoted its "reserved" state powers clause during his presidential campaign.  Even Democratic President Bill Clinton, who once proposed that the federal government establish a comprehensive national health-care program, sought to seal his reelection in 1996 by accepting Republican plans to turn over to the states much of the control over the nation's welfare policy.  This renewed interest in federalism by leaders of both political parties appears to be a prudent political strategy because recent public opinion poils and in-depth interviews with citizens have shown considerable public enthusiasm for state governments and discontent with the federal government. 
This study will examine the extent to which citizen views about their own state government are relevant for national politics. It attempts to answer two questions: What factors explain public views concerning the federal government and one's own state government? How was vote choice in the 1996 presidential election affected by citizen opinions about these two orders of government?
PERSPECTIVES ON FEDERAL AND STATE GOVERNMENTS
A variety of recent research projects have tapped the considerable public dissatisfaction with the government in Washington, D.C.  Although this discontent does not represent a rejection of the legitimacy of the American political system,  sentiment may be building for some fine-tuning, perhaps by reducing the federal government's power in favor of the states. In fact, Washington policymakers already are making that adjustment. 
Academics who have studied state governments have focused on two areas of comparison between the federal government and the state governments: efficiency and responsiveness. Linda Bennett and Stephen Bennett wrote that increasing the power of state governments is a natural path for a federal government to follow when it is being criticized for poor performance.  Increasing the power of state government for some programs may be a more efficient use of taxpayer resources because states are closer to the points of service delivery.  Other scholars doubt that state governments, which they tend to view as more parochial, are more capable than federal authorities. 
Affinity for the different governments may take quite different forms and have different impacts in different states. Examinations of state voting patterns suggest a decoupling of federal and state political evaluations in recent elections. 
Of course, as V. O. Key once observed, one's feelings about one's own state government may depend on factors that vary from jurisdiction to jurisdiction.  Robert Erikson, Gerald Wright, and John McIver found that states tailor their public polices to match more precisely the wishes of their distinct electorates.  This pattern of state policy responsiveness can be found in a variety of issue areas, including abortion, civil rights, civic culture, economic policy, and welfare.  Edward Lascher, Michael Hagen, and Steven Rochlin found that public policy in states with citizen initiatives generally are not more responsive to public opinion than in states without them, largely because even the states without referenda are quite responsive to public opinion. 
Federalism and its consequences for public opinion and electoral politics are important areas for academic inquiry in large part because of the increasing emphasis politicians have placed on the idea of reducing the power of Washington. The increased citizen discontent of the 1990s and the rise to power of House and Senate Republicans in 1994 suggest the need to reexamine feelings about state power and their consequences for both public opinion and electoral politics. Finally, the restoration of key (though admittedly imperfect) questions about comparative governmental power in the 1996 American National Elections Study (ANES) allows for a renewed examination of this concept's impact.
How might these apparent public feelings of federal frustration and state satisfaction translate into citizen orientations and actions? Partisanship and ideology are often keys to public opinion formation, as are one's age, race, income, education, and views about government's perceived responsiveness and competence.  All are considered here. Once the possible determinants of these opinions have been examined, this study then turns to the impacts of federalism on national electoral politics.
Republicans have been more consistent and more aggressive in recent years in their expressions of fondness for state government and of hostility to the federal government. They seem to be in the best position to profit from this shift, and recent GOP gains in the federal, state, and local arenas suggest some electoral utility to this appeal.  This pro-state government trend, if relevant to federal politics, should have helped Dole, who made revived federalism a key part of his campaign, but not Clinton, whose views on the topic were inconsistent. 
Five hypotheses are presented here:
H1: Higher levels of political efficacy, perceived governmental competence, liberalism, and Democratic partisanship will lead to more positive evaluations of the federal government.
H2: Lower levels of perceived competence, efficacy, liberalism, and Democratic partisanship concerning the federal government will lead to more positive evaluations of one's state government.
These hypotheses propose that support for the federal government and one's own state government can be explained largely through two key avenues: individual background measures and evaluations of the federal government's performance.
H3: The more one's state government represents one's own partisan and ideological orientations, the more favorably it will be viewed.
H4: The less professionalized one's own state legislature and the poorer the state's economic performance, the more one will support the federal government.
These two hypotheses propose that people evaluate their state governments in a partisan context and through that government's objective performance.
H5: The more one dislikes the federal government and big government in general, the more likely one was to vote for Republican Bob Dole in the 1996 presidential election.
This hypothesis suggests that such feelings about federalism matter for presidential elections in ways consistent with party traditions and campaign rhetoric.
DATA AND MEASURES
At the center of this analysis are two questions asked in the 1996 ANES for the first time in two decades. The first of these questions asked the interviewees to identify the level of government in which they have the most faith and confidence; the second question asked the respondent to say which level inspires the least faith and confidence. Table 1 shows the results of ANES questions.
The federal government was the leading source of the least faith and confidence in 1996; 48 percent identified the federal government, as compared to 34 percent who selected local government and 19 percent who objected most to state government. The lack of enthusiasm for the federal government was roughly comparable to that of the mid-1970s, when trust fell greatly in the wake of the war in Vietnam and the Watergate scandal.
When citizens were asked about where they placed the greatest confidence, the pattern was reversed: the states finished first with 37 percent, as compared to 33 percent for local governments and 30 percent for the federal government. The idea of enhancing state governments apparently finds a ready audience of citizens, and one considerably enhanced from 1976, the last time these questions were asked in the ANES.
State governments were much more highly regarded in 1996 than they had been two decades earlier, especially as the controversial "states' rights" legacy of racial discrimination continued to fade from public consciousness with the passage of time. In addition, the increasing accountability and professionalization of state governments in recent years may have triggered changing, though not always positive, feelings about state government. 
This second question, which government is the most disliked, forms the basis of the public opinion part of this project (tests using the "most liked" question-results not shown-offered similar results in the opposite direction to those reported below). In the final part of this study, which relates to vote choice, the two measures found in Table 1 are combined into a single variable that measures one's relative fondness for state government over the federal government. This measure was created through a procedure similar to that used by Phillip Roeder, except that the variable used here was calculated without any explicit reference to feelings about local government, which is not a subject of this study.  Further information is found in the notes to Table 3.
This study also makes use of three general questions tapping public opinion about the ideal size of the federal government. These three ANES questions had relatively similar levels of support for the federal government, ranging from 45 percent to 54 percent, and are combined into a three-question index here.
This study also incorporates six measures not in the ANES that relate to the perceived responsiveness, performance, and culture of one's state government. 
The first measure is a State Partisan Affinity variable that measures the extent to which the partisan composition of the lower chamber of each state legislature differed from the respondent's partisan leanings.  Independents in states with a rough parity between the parties received low values, as did strong Republicans and Democrats in states where one's own party had large legislative majorities. Strong Democrats in states with large Republican-majority legislatures, and vice versa, had the highest values. Residents of Nebraska, which has a nonpartisan legislature, were dropped from this analysis.
A second state variable measures the absolute distance between one's ideological position and state taxation per capita.  Strong liberals in hightax states, conservatives in low-tax states, and moderates in mid-level tax states received low scores. Conservatives in high-tax states and liberals in low-tax states received the highest scores.
This study also uses a State Populism measure offered by James Burns, et al., which ranks states on the extent to which they penit citizen initiative, referendum, and recall of elected officials.  The fourth state-specific measure is one of Legislative Professionalism developed by Samuel Patterson.  The fifth measure is the state's average unemployment rate in 1996. 
The sixth measure provides for a test of the significance of state political cultures. Using Daniel J. Elazar's model, the study tests whether residents of what he termed the moralistic states are more pro-national in orientation, while citizens of traditionalistic and the individualistic states are more pro-state. 
Table 2 contains two logistic regressions testing the extent to which one's relative hostility to either the federal government or one's own state government can be predicted by demographic, partisan, and ideological measures, and by political attitudes. Ordinary Least Squares (OLS) regression is an improper statistical technique for dichotomous dependent variables (was the respondent the most hostile to the federal government or not, in one instance, or to their own state government, in the other). Although a straight-fitting OLS regression line does not fit logistic distributions, a related technique known as logistic regression provides regression coefficients like those in OLS regression and, therefore, relatively easily interpreted results. One key difference between the two techniques is that the effectiveness of the overall model can be measured both by an r-square statistic and by the percentage of the cases predicted correctly. Table 2 contains unstandardized coefficients, an r-square measure, and ca se-classification results.
With respect to the results relating to the federal government, one notices first the powerful influence played by party identification and, to a lesser extent, ideology. Liberals and Democrats are the most positively disposed toward the federal government, as hypothesized. Other influential measures in this model include an efficacy measure (No Care) and how positively one views the economy's performance over the past year, as hypothesized. The federal model, which has a Cox/Snell r-square of.159, correctly predicts 67.3 percent of the cases: 64.4 percent of the cases where the federal government was least liked and 70.2 percent of the cases where Washington was not the least liked.
Only one state-specific measure shows a statistically significant relationship with feelings about the federal government: whether the state is classified by Elazar as moralistic. In a departure from what was hypothesized, the more moralistic the state, the more likely a resident was critical of the federal government, a result perhaps of the shrinking policy role Washington sought for itself in 1995 and 1996.
The right column uses the same independent variables to predict whether an individual was most hostile to his or her own state government (Pearson's r = -.46 for the two dependent variables used in the two models is reported in Table 2). Although the state model has a higher overall prediction rate, 81.6 percent, this is a highly misleading statistic; the high percentage comes from the fact that the model does not effectively distinguish people relatively hostile to state governments from those more hostile to the federal government. The Cox/Snell r-square reading of .07 demonstrates the fundamental weakness of the state model, as does a closer look at the classification pattern.
Even so, many of the same independent variables are influential in both the federal and state models. Strong Republicans, conservatives, and those who evaluate the economy's recent performance more negatively are most likely to feel positively about state government. Strangely, not one of the six state-specific measures has a statistically significant relationship with individual feelings about state government, even though the measures were designed to elicit state government evaluations explicitly.
Why might the differences between predictions in the left and right sides of Table 2 be so dramatic? Democratic citizens seem far more committed to a positive evaluation of the federal government than Republicans to a positive evaluation of the state governments. After all, Republican ideas of reducing federal power can be, and sometimes are, coupled with proposals to reduce state government authority as well. 
The results here suggest that support for one's own state government does not spring largely from frustration with the federal government. People generally do not turn to their state capitals because of a sense that Washington doesn't care about them, though economic matters do have some relevance to one's feelings about federalism. One's feelings about one's state government have even less to do with the state-specific measures: the responsiveness, the professionalism, and the partisan or taxation affinity for one's own state. Broad orientations toward the political world like partisanship and ideology are much more important for predicting support for state government than any substantive evaluation of what one's own state government has or has not done.
At least part of the disparity between the two analyses in Table 2may stem from the fact that most of the questions available in the ANES are clearly tied to the federal government. The government in Washington is implicitly or explicitly behind the measures of efficacy and perceived governmental competence used here. Although this study added measures to elicit state-specific measures and feelings beyond what is available in the ANES, explicit questions of state government efficacy and performance may be more successful in predicting public opinion with respect to state government than is this model.
Table 3 also uses logistic regression, this time to test whether citizens' feelings about big government and state government helps predict whether a citizen voted for Bob Dole in 1996. The model includes three new independent variables: a system trust index, an index of support for the principle of big government, and a state preference measure derived from the two questions identified in Table 1. The first two measures were not used as independent variables in Table 2 because of their potential for measuring something quite similar to that sought by the dependent variable. More information about the new variables is found in the notes to Table 3. Given that the model examines only Dole and Clinton voters, one would simply have to reverse the signs on the coefficients to present the Clinton results.
The measures provide for a correct prediction of 92 percent of the voters who said they cast ballots for either Dole or Clinton; 94 percent of the Clinton voters were correctly classified, and 89.5 percent of the Dole voters were identified correctly. Partisanship, ideology, and race, the staples of voting research, all proved effective measures, as did perceived economic performance, another variable often found to affect electoral politics.
Three variables relating to federalism are also effective predictors. The Big Government index and the State Preference measure are both in the expected direction. The more one dislikes big government and the more one likes state government when compared to the federal government, the more likely one was to vote for Bob Dole. Taxation affinity, the third state-related variable, suggests that the less happy one was with the taxation orientation of one's state, the more likely one was to vote for Bob Dole. Perhaps low-tax conservatives in high-tax states would be particularly likely to vote, and to vote for Bob Dole, over tax matters because of frustration with their own state government.
This model was somewhat better at identifying Clinton voters than Dole voters. This suggests, as does Table 2, that Democrats have done a better job in inculcating the values of big government among their voters than the Republicans have done in convincing their supporters of the dangers of big government. The findings are consistent with the New Deal and Great Society traditions of the Democratic party, and with the pronouncements by the Republicans-especially by Gingrich and in the 1996 campaign by Dole himself-about the evils of an overly centralized government. It may be harder to criticize one government and sing the virtues of another than some Republicans thought.
In Table 3, many of the measures that relate to the responsiveness and performance of state government were not statistically significant predictors as hypothesized. What mattered here for predicting vote choice, beyond the traditional demographic and economic measures, were the broad beliefs about the advantages and disadvantages of big government in the abstract. This suggests that the Republican message of more power to state governments is still largely ideological, and has few linkages to the actual performance of state government. The federalism issue is not useless at an ideological level, however. It helped some people make the decision to vote for Bob Dole, even after partisan and ideological factors are taken into account.
The results indicate that citizen evaluations about state government are largely ideological and partisan in orientation. They are not primarily a result of performance shortcomings on the part of the federal government. Nor do feelings about state governments correspond to a given state's actual performance. Further, the findings here much more strongly supported the link between the attitudinal measures and federal government feelings than any supposed link between those attitudinal measures and feelings about state governments. The relative weakness of the state government model may be partially due to the fact that opinions may be more consistent regarding one federal government than 50 different state governments. Even so, most of the state-specific measures added to this study did not improve the predictive power of the vote choice or state preference models.
There are opportunities for further research on how feelings about the different orders of government are derived. Attempts to employ possible predictor measures for public views about state government have been hampered by the American National Elections Study's concentration on federal issues. Might questions relating to state government efficacy predict levels of public feelings about state government? This study cannot say.
One can also wonder whether citizens who dislike federal government power may also dislike state government power. The state questions used here, though a dramatic improvement on what has been available in past surveys, may force respondents to choose which government they like the most even if they dislike them all intensely. To deal with this potential problem, future surveys might use thermometer measures or at least a five-point like-dislike scale to tap more precisely citizen orientations toward the different governments.
Unfortunately for the Republicans, who in recent years have been concentrating on the virtues of the state governments, their supporters are somewhat less certain about how to apply evaluations of one's own state government to the presidential contest. Perhaps there can be substantial differences between what is said about the federal government and what is actually desired by citizens. Although reducing the totality of government functions may be popular rhetoric, one can wonder how desired that approach actually is by voters. Citizens continue to expect the federal government to provide a high level of public services, and elected officials rely on effective service delivery to help remain in the good graces of their constituents.
This study, with its generalized comparison of federal versus state government power and the impacts of these views on candidate choice, might profitably be tested further in surveys of different state electorates. Distinct state political cultures could affect the results in ways obscured in a nationwide analysis of this type. The findings of the Elazar measure suggest further consideration of state political culture. Another fruitful analysis could be the impact of public opinion about state governments on state elections.
This study is time-bound. The ANES returned to its comparative evaluations of governments' questions in 1996 after a 20-year hiatus. Further research into the changing nature of comparative evaluations of government and politics over time also should be an important part of future research. We do not yet know whether a revived states' rights doctrine will remain prominent in the minds of voters and candidates. If it does, future elections should be tested to see whether the basic orientations toward the state and federal governments, and toward the concept of big government generally, continue to predict candidate vote choice in presidential contests.
Stephen J. Farnsworth is Assistant Professor of Political Science at Mary Washington College in Fredericksburg, VA. His research interests include federalism, presidential elections, the news media and political trust, and his work has appeared in the Harvard International Journal of Press/Politics and the New York Times. He was a 1999 winner of the Best Chapter Advisor Award from Pi Sigma Alpha, the political science honor society, and was named by MWC's Class of 1999 as the college's most inspirational professor.
AUTHOR'S NOTE: I wish to acknowledge the assistance of Diana Owen, Jack Dennis, Jim Lengle, Ronald Weber, and the anonymous reviewers. I also thank the InterUniversity Consortium for Political and Social Research (ICSPR) for access to the data and Mary Washington College for its financial support. An earlier version of this paper was delivered at the 1997 Annual Meeting of the Southern Political Science Association. All errors remain the author's.
(1.) James E. Campbell, "The Presidential Pulse and 1994 Midterm Congressional Election," Journal of Politics 59 (August 1997): 830-57; Newt Gingrich, to Renew America (New York: HarperCollins, 1995); Gary C. Jacobson, The Politics of Congressional Elections, 4th ed. (New York: Longman, 1997).
(2.) Richard A. Harris, "The Era of Big Government Lives," Polity 30 (Fall 1997): 187-192.
(3.) Scott Keeter, "Public Opinion and the Election," The Elections of 1996: Reports and Interpretations (Chatham, NJ: Chatham House, 1997); Michael Nelson, "The Election: Turbulence and Tranquility in Contemporary American Politics," The Elections of 1996, ed. Michael Nelson (Washington, D.C.: CQ Press, 1997); Theda Skocpol, Boomerang Health Care Reform and the Turn Against Government (New York: Norton, 1997); Carol S. Weissert and Sanford F. Schram, "The State of American Federalism 1995-1996," Publius: The Journal of Federalism 26 (Summer 1996): 1-26.
(4.) Stephen C. Craig, The Malevolent Leaders (Boulder, CO: Westview, 1993); Samuel C. Patterson, Randall B. Ripley, and Stephen V. Quinlan, "Citizen Orientations Toward Legislatures: Congress and the State Legislatures," Western Political Quarterly 45 (June 1992): 315-338; Phillip W. Roeder, Puhlic Opinion and Policy Leadership in the American States (Tuscaloosa, AL: University of Alabama Press, 1994).
(5.) Craig, The Malevolent Leaders, Jack Citrin, "Who's the Boss? Direct Democracy and Popular Control of Government," Broken Contract? Changing Relationships Between Americans and Their Government, ed. Stephen C. Craig (Boulder, CO: Westview, 1996); John R. Hibbing and Elizabeth Theiss-Morse, Congress As Public Enemy: Public Attitudes Toward American Political Institutions (Cambridge, UK: Cambridge University Press, 1995); David Matthews, Polities for People: Finding a Responsible Public Voice (Chicago, IL: University of Illinois Press, 1994).
(6.) Stephen C. Craig, "The Angry Voter: Politics and Popular Discontent in the 1990s," Broken Contract? Changing Relationships Between Americans and Their Government, ed. Stephen C. Craig (Boulder, CO: Westview, 1996).
(7.) Nelson, "The Election"; Skocpol, Boomerang.
(8.) Linda L. M. Bennett and Stephen Earl Bennett, Living with Leviathan: Americans Coming to Terms with Big Government (Lawrence, KS: University of Kansas Press, 1990).
(9.) John Herbers, "The New Federalism: Unplanned, innovative and here to stay," Governing 1 (October 1987): 28-37; Alice Rivlin, Reviving the American Dream: The Economy, The States and the Federal Government (Washington, D.C.: Brookings Institution, 1992).
(10.) Max Kaase and Kenneth Newton, Beliefs in Government (Oxford: Oxford University Press, 1995); Bruce A. Wallin, "Federal Cutbacks and the Fiscal Conditions of the States," Publius: The journal of Federalism 26 (Summer 1996): 141-159.
(11.) Lonna Rae Atkeson and Randall W. Partin, "Economic and Referendum Voting: A Comparison of Gubernatorial and Senatorial Elections," American Political Science Review 89 (March 1995): 99-107; Richard G. Niemi, Harold W. Stanley, and Ronald J. Vogel, "State Economics and State Taxes: Do Voters Hold Governors Accountable?" American journal of Political Science 39 (November 1995): 936-57; Robert Stein, "Economic Voting for Governor and U.S. Senator: Electoral Consequences of Federalism," journal of Politics 52 (February 1990): 29-53. But see also John E. Chubb, "Institutions, the Economy and the Dynamics of State Elections," American Political Science Review 82 (March 1988): 133-154.
(12.) V. O. Key, Southern Politics in State and Nation (New York: Alfred A. Knopf, 1949).
(13.) Robert S. Erikson, Gerald C. Wright, and John P. McIver, Statehouse Democracy: Public Opinion and Policy in the American States (New York: Cambridge University Press, 1993).
(14.) Frances Stokes Berry and William D. Berry, "Tax Innovation in the States: Capitalizing on Political Opportunity," American Journal of Political Science 36 (August 1992): 715-742; Frances Stokes Berry and William D. Berry, "The Politics of Tax Increases in the States," American journal of Political Science 38 (August 1994): 855-839; Rim Quaile Hill, Democracy in the Fifty States (Lincoln, NE: University of Nebraska Press, 1994); Rim Quaile Hill, Jan E. Leighley, and Angela Hinton-Andersson, "Lower Class Mobilization and Policy Linkage in the U.S. States," American journal of Political Science 39 (February 1995): 75-86; Tom W. Rice and Alexander F. Sumberg, "Civic Culture and Government Performance in the American States," Publius: The journal of Federalism 27 (Winter 1997): 99-114; Matthew E. Wetstein and Robert B. Albritton, "Effects of Public Opinion on Abortion Policies and Use in the American States," Publius: The journal of Federalism 25 (Fall 1995): 91-105.
(15.) Edward L. Lascher, Jr., Michael Hagen, and Steven A. Rochlin, "Gun Behind the Door? Ballot Initiatives, State Politics and Public Opinion," Journal of Politics 58 (August 1996): 760-775.
(16.) Steven J. Rosenstone and John Mark Hansen, Mobilization, Participation and Democracy in America (NewYork: Macmillan, 1993); Sidney Verba, Kay Lehman Schlozman, and Henry E. Brady, Voice and Equality: Civic Voluntarism in American Politics (Cambridge, MA: Harvard University Press, 1995).
(17.) James E. Campbell, "The Presidential Pulse and the 1994 Midterm Congressional Election," Journal of Politics 59 (August 1997): 830-57; Gary C. Jacobson, The Politics of Congressional Elections, 4th ed. (New York: Longman, 1997).
(18.) Harris, "The Era of Big Government Lives,"; Nelson, "The Election."
(19.) Thad Beyle, "Being Governor," The State of the States, ed. Carl E. Van Horn, 2nd ed. (Washington, D.C.: CQ Press, 1993); Malcolm Jewell, Representation in State Legislatures (Lexington, KY: University of Kentucky Press, 1982); Peverill Squire, "Professionalism and Public Opinion of State Legislatures," Journal of Politics 55 (May 1993): 478-491.
(20.) Roeder, Public Opinion and Policy Leadership in the American States.
(21.) Erikson, Wright, and McIver, Statehouse Democracy; Alice Rivlin, Reviving the American Dream: The Economy, The States and the Federal Government. (Washington, D.C.: Brookings Institution, 1992).
(22.) Statistical Abstract of the United States, 117th ed. (Washington, D.C.: U.S. Bureau of the Census, Department of Commerce, 1997).
(24.) James MacGregor Burns, et al., State and Local Politics: Government by the People, 7th ed. (Englewood Cliffs, NJ: Prentice-Hall, 1993).
(25.) Samuel C. Patterson, "Legislative Politics in the States," Politics in the American States: A Comparative Analysis, eds. Virginia Gray and Herbert Jacobs 6th ed, (Washington, D.C.: CQ Press, 1996).
(26.) Statistical Abstract of the United States.
(27.) Daniel J. Elazar, American Federalism: A View from the States, 2nd ed. (NewYork: Crowell, 1972).
(28.) John Kincaid, "De Facto Devolution and Urban Defunding: The Priority of Persons Over Places," Journal of Urban Affairs 21(2:1999): 135-167.
Citizen Evaluations of Their Governments: Most and Least Faith and Confidence "Do you have more faith and confidence in the federal government, the government of this state or the local government around here?" Most Confidence 1996 1976 1974 1972 1968 Federal 30% 32% 32% 46% 50% State 37 28 30 24 20 Local 33 40 38 30 30 "Which level of government do you have the least faith and confidence in: the federal government, the government of this state or the local government around here?" Least Confidence Federal 48% 47% 47% 27% 32% State 19 20 16 25 24 Local 34 33 37 48 45 Source: American National Elections Studies (ANES). Note: Percentage may not all add up to 100 because of rounding. Faith and Confidence in Federal and State Governments: Logistic Regression Analyses Variable Name Federal (b) State (b) Education -.04 .05 Income -.13 -.01 Sex -.06 -.02 Race -.38 .31 Party ID -.27 [***] .16 [**] Ideology -.13 [*] .17 [*] No Care .16 [*] .003 No Say .10 -.10 Complex .10 -.04 Economy Past .24 [**] -.26 [*] Populism .13 .01 Professionalism .05 -.26 Unemployment .07 .04 Partisan Affinity .005 -.01 Taxation Affinity .000 .000 Moralistic -.26 [**] .09 N 979.0 979.0 -2 Log Likelihood 1187.226 863.663 Cox/Snell r-square .159 .070 Classification: Predicted versus Observed Preference Federal Predicted Least Not Least % Correct Observed Least 313 173 64.4 Not Least 147 346 70.2 Total Percent Correct 67.3% State Observed Least 0 180 0.0 Not Least 0 799 100.0 Total Percent Correct: 81.6% (*.)p [less than] .05 (**.)p [less than] .01 (***.)p [less than] .001
Notes to Table 2:
NO CARE: "Public officials don't care much about what people like me think."
NO SAY: "People like me don't have any say about what the government does."
COMPLEX: "Sometimes politics and government seem so complicated that a person like me can't really understand what's going on.
ECONOMY PAST: "Would you say that over the past year the nation's economy has gotten better, stayed about the same, or gotten worse? (If better) Would you say much better or somewhat better? (If worse) Would you say much worse or somewhat worse?"
POPULISM (0-3 scale): Does state offer citizen initiative, citizen referendum, and citizen recall?
PROFESSIONALISM: Citizen, hybrid, or professional state legislature?
UNEMPLOYMENT: The state unemployment rate (1996 average).
PARTISAN AFFINITY: The absolute partisan distance between an individual and the lower chamber of the state legislature before the 1996 elections. The seven-point party identification scale was spread along the range of level of GOP control.
TAXATION AFFINITY: The absolute ideological distance between an individual and his or her state government's tax collections per capita. The seven-point ideological scale was spread along the tax collection range at the following percentiles (5-20-35-50-65-80-95).
MORALISTIC: Pure moralistic states (n = 9) scored a 3; moralistic dominant states (n = 8) scored a 2; states with a strong moralistic strain (n = 10) scored a 1; states without a strong moralistic strain (n = 23) scored a zero.
(DV) FEDERAL (Dichotomy): If federal government was identified as the least favored government a zero was recorded; variable was scored a 1 if state or local government was identified by respondent as least favored. Cut value for logistic regression set at .48.
(DV) STATE (Dichotomy): If state government was identified as the least favored level of government a zero was recorded; variable was scored a 1 if federal or local government was identified by respondent as least favored. Cut value for logistic regression set at .19.
Voting for Dole and Feelings About Big Government: Logistic Regression Analysis Variable Name Coefficient (b) Standard Error of b Education .05 (.11) Income .27 (.17) Sex -.12 (.33) Race -3.43 [**] (1.19) Party ID .91 [***] (.11) Ideology .68 [***] (.15) Trust Index .09 (.09) No Care -.17 (.19) No Say .04 (.16) Complex .12 (.15) Economy Past -1.18 [***] (.21) Big Government Index -.53 [***] (.14) Populism -.18 (.19) Professionalism .16 (.26) Unemployment .07 (.18) Partisan Affinity -.001 (.01) Taxation Affinity -.002 [*] (.0008) Moralistic .11 (.22) State Preference .23 [*] (.11) N 675.0 -2 Log Likelihood 289.415 Cox/Snell r-square .613 Classification: Predicted versus Observed Vote Predicted Predicted Percent Total Percent Clinton Dole Correct Correct Observed Clinton 348 22 94.05% Observed Dole 32 273 89.51% 92.00% (*.)p [less than] .05 (**.)p [less than] .01 (***.)p [less than] .001
Notes to Table 3:
Cut value for logistic regression analysis was set at .58, because 58 percent of the ANES respondents who reported voting for Clinton or Dole said they voted for Clinton. All other candidates were excluded from this analysis.
See notes to Table 2 for descriptions of all measures except those that follow.
TRUST INDEX (five variable alpha .67): (1) "How much of the time do you think you can trust the government in Washington to do what is right-just about always, most of the time, or only some of the time?" (2) "Do you think that people in government waste a lot of the money we pay in taxes, waste some of it, or don't waste very much of it?" (3) "Would you say the government is pretty much run by a few big interests looking Out for themselves or that it is run for the benefit of all the people?" (4) "Do you think that quite a few of the people running the government are crooked, not very many are, or do you think hardly any of them are crooked?" (5) "Over the years, how much attention do you feel the government pays to what people think when it decides what to do-a good deal, some, or not much?'
BIG GOVERNMENT INDEX (three variable alpha = .75): "I am going to ask you to choose which of two statements I read comes closer to your own opinion. You might agree to some extent with both, but we want to know which one is closer to your views," (1) "ONE, the less government, the better; or TWO, there are more things that government should be doing? (2) "ONE, we need a strong government to handle today's complex economic problems; or TWO, the free market can handle these problems without government being involved. (3) "ONE, the main reason government has become bigger over the years is because it has gotten involved in things that people should do for themselves; or TWO, government has become bigger because the problems we face have become bigger."
STATE PREFERENCE: (1) State least favored, Federal most favored; (2) State least, Federal middle (Local most); (3) State middle, Federal most; (4) State middle, Federal least (Local most); (5) State most, Federal middle; (6) State most, Federal least.
|Printer friendly Cite/link Email Feedback|
|Author:||Farnsworth, Stephen J.|
|Date:||Jun 22, 1999|
|Previous Article:||Asymmetric Effects of Intergovernmental Grants: Analysis and Implications for U.S. Welfare Policy.|
|Next Article:||The United Kingdom: Federalism in Denial?|