Wars and rumours of wars: The contexts of cultural conflict in American political behaviour

British Journal of Political Science

Volumn 36, Issue 1, 2006, Pages 61-89

  Layman, Geoffrey C ^a   Green, John C ^b  

^a UNIVERSITY OF MARYLAND   (United States)

^b UNIVERSITY OF AKRON   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 29144458957     PISSN: 00071234     EISSN: 14692112     Source Type: Journal    
DOI: 10.1017/S0007123406000044     Document Type: Article

References (148)

1. Lincoln Caplan, 'Dissent is fine, but spare the bite', Washington Post Weekly Edition, 14-20 July 2003, pp. 22-3.
2. Byron E. Shafer, The State of American Politics (Lanham, Md.: Rowman & Littlefield, 2002);
3. Pew Research Center for the People and the Press, 'The 2004 Political Landscape Evenly Divided and Increasingly Polarized', Press Release, 5 November 2003;
4. John C. Green and Mark Silk, 'The New Religion Gap', Religion in the News, 6:3 (2003), Special Supplement, 1-3, 15;
5. Pew Research Center for the People and the Press, 'Voters Liked Campaign 2004, But Too Much 'Mud-Slinging'; Moral Values: How Important?', Press Release, 11 November 2004;
6. Kevin Eckstrom, 'In 2004, Christians stood at front lines', Washington Post, 1 January 2005, p. B7.
7. Morris P. Fiorina, with Samuel J. Abrams and Jeremy C. Pope, Culture War? The Myth of a Polarized America (New York: Pearson Longman, 2005).
8. James Davison Hunter, Culture Wars: The Struggle to Define America (New York: Basic Books, 1991);
9. see also James Davison Hunter, Before the Shooting Begins: Searching for Democracy in America's Culture War (New York: Macmillan, 1994).
10. Hunter, Culture Wars, pp. 43-4. We use Hunter's terms for ease of presentation, but they are potentially misleading because of their political connotations. Religious traditionalists (rather than the orthodox) and modernists (rather than progressives) are probably more useful terms.
11. Hunter, Culture Wars, p. 44.
12. Hunter, Culture Wars, p. 43.
13. What constitutes a 'moral' issue, is, of course, debatable. That is evidenced by the dispute among American political commentators over what the plurality of voters who cited 'moral values' as their most salient concern in 2004 election-day exit polls had in mind. However, in recent political discourse, the language of morality has most commonly been used in reference to sexual, gender and reproductive issues.
14. Hunter, Culture Wars, p. 281.
15. Hunter, Culture Wars, p. 128.
16. Fiorina, Abrams and Pope, Culture War?
17. Rhys H. Williams, ed., Cultural Wars in American Politics: Critical Reviews of a Popular Myth (New York: De Gruyter, 1997), pp. 291-2;
18. James L. Guth, John C. Green, Corwin E. Smidt and Lyman A. Kellstedt, The Bully Pulpit: The Politics of Protestant Clergy (Lawrence: University Press of Kansas, 1997);
19. John C. Green, James L. Guth, Corwin E. Smidt and Lyman A. Kellstedt, Religion and the Culture Wars (Lanham, Md.: Rowman and Littlefield, 1996);
20. Clyde Wilcox, Onward Christian Soldiers: The Religious Right in American Politics (Boulder, Colo.: Westview, 2000);
21. Geoffrey C. Layman, The Great Divide: Religious and Cultural Conflict in American Party Politics (New York: Columbia University Press, 2001).
22. Robert Wuthnow, The Restructuring of American Religion (Princeton, N.J.: Princeton University Press, 1988);
23. Robert Wuthnow, The Struggle for America's Soul: Evangelicals, Liberals, and Secularism (Grand Rapids, Mich.: Eerdmans, 1989);
24. Gertrude Himmelfarb, One Nation, Two Cultures (New York: Knopf, 1999);
25. John Kenneth White, The Values Divide (New York: Chatham House, 2003).
26. The two most recent efforts to directly test the culture wars thesis find rather limited support for it: Fiorina, Abrams and Pope, Culture War?
27. Wayne Baker, America's Crisis of Values: Reality and Perception (Princeton, N.J.: Princeton University Press, 2005).
28. Fiorina, Abrams and Pope, Culture War?
29. Nancy J. Davis and Robert V. Robinson, 'Are the Rumors of War Exaggerated? Religious Orthodoxy and Moral Progressivism in America', American Journal of Sociology, 102 (1996), 756-87;
30. Nancy J. Davis and Robert V. Robinson, 'Religious Orthodoxy in American Society: The Myth of a Monolithic Camp', Journal for the Scientific Study of Religion, 35 (1996), 229-45;
31. Nancy J. Davis and Robert V. Robinson, 'A War for America's Soul: The American Religious Landscape' in Williams, ed., Cultural Wars in American Politics, pp. 39-62;
32. N. J. Demerath III and Yonghe Yang, 'What American Culture War? A View from the Trenches as Opposed to the Command Posts and the Press Corps', in Williams, ed., Cultural Wars in American Politics, pp. 17-38;
33. Paul DiMaggio, John Evans and Bethany Bryson, 'Have Americans' Social Attitudes Become More Polarized?' American Journal of Sociology, 102 (1996), 690-755;
34. Christian Smith, Michael Emerson, Sally Gallagher, Paul Kennedy and David Sikkink, 'The Myth of Culture Wars: The Case of American Protestantism', in Williams, ed., Cultural Wars in American Politics, pp. 175-95;
35. Alan Wolfe, One Nation, After All: What Middle-Class Americans Really Think About: God, Country, Family, Racism, Welfare, Immigration, Homosexuality, Work, the Right, the Left, and Each Other (New York: Viking, 1998).
36. Demerath and Yang, 'What American Culture War?'
37. Davis and Robinson, 'Religious Orthodoxy in American Society';
38. Daniel V. A. Olson, 'Dimensions of Cultural Tension among the American Public', in Williams, ed., Cultural Wars in American Politics, pp. 237-58.
39. Philip E. Converse, 'The Nature of Belief Systems in Mass Publics', in David E. Apter, ed., Ideology and Discontent (New York: Free Press, 1964), pp. 206-61;
40. Angus Campbell, Philip E. Converse, Warren E. Miller and Donald E. Stokes, The American Voter (Chicago: The University of Chicago Press, 1964);
41. Philip E. Converse and Gregory B. Markus, 'Plus ça Change ... : The New CPS Election Study Panel', American Political Science Review, 73 (1979), 32-49.
42. Campbell et al., The American Voter
43. Richard A. Brody and Benjamin I. Page, 'Comment: The Assessment of Policy Voting', American Political Science Review, 66 (1972), 450-8;
44. John Zaller, The Nature and Origins of Mass Opinion (New York: Cambridge University Press, 1992);
45. Warren E. Miller and J. Merrill Shanks, The New American Voter (Cambridge, Mass.: Harvard University Press, 1996).
46. Converse, 'The Nature of Belief Systems in Mass Publics'.
47. See, for example, Edward G. Carmines and James A. Stimson, Issue Evolution: Race and the Transformation of American Politics (Princeton, N.J.: Princeton University Press, 1989);
48. Norman H. Nie, Sidney Verba and John Petrocik, The Changing American Voter (Cambridge: Harvard University Press, 1976).
49. Converse, 'The Nature of Belief Systems in Mass Publics', p. 209.
50. Converse, 'The Nature of Belief Systems in Mass Publics', pp. 209-10;
51. see also Michael X. Delli Carpini and Scott Keeter, What Americans Know About Politics and Why It Matters (New Haven, Conn.: Yale University Press, 1996).
52. Hunter, Culture Wars, p. 127.
53. Converse, 'The Nature of Belief Systems in Mass Publics', p. 209.
54. Converse, 'The Nature of Belief Systems in Mass Publics', pp. 210-12.
55. Bernard R. Berelson, Paul R. Lazarsfeld and William N. McPhee, Voting: A Study of Opinion Formation in a Presidential Campaign (Chicago: University of Chicago Press, 1954);
56. Robert Huckfeldt and John Sprague, Citizens, Politics, and Social Communication: Information and Influence in an Election Campaign (Cambridge: Cambridge University Press, 1995).
57. The meaning of terms such as 'individualism' and 'communalism' may vary across contexts. For instance, white evangelical Protestants are highly individualistic in their theology, demanding a personal conversion for admission to the faith, emphasizing individual prayer and Scripture reading, and placing great importance on the personal moral behaviour of believers. In the context of economic and social welfare policy, that theological individualism tends to translate into economic individualism and support for conservative issue positions (see, for example, David C. Barker and Christopher Jan Carman, 'The Spirit of Capitalism? Religious Doctrine, Values, and Economic Attitude Constructs', Political Behavior, 22 (2000), 1-27;
58. Guth et al., The Bully Pulpit). However, in the context of moral policy, theological individualism can have a strong communitarian character, generating demands for the legal enforcement of moral codes. In another context, evangelical theology among black Protestants gave rise to the Civil Rights movement, which was a highly communitarian enterprise, but also one focused on the individual rights of African-Americans. Here, we use these terms to refer to the character of religious beliefs and values. Their application to policy positions is shaped by religious, policy and political contexts.
59. Guth et al., The Bully Pulpit;
60. Kenneth D. Wald, Dennis E. Owen and Samuel S. Hill Jr, 'Churches as Political Communities', American Political Science Review, 82 (1988), 531-48;
61. Kenneth D. Wald, Dennis E. Owen and Samuel S. Hill Jr, 'Political Cohesion in Churches', Journal of Politics, 52 (1990), 197-215.
62. Barker and Carman, 'The Spirit of Capitalism?'
63. Guth et al., The Bully Pulpit;
64. Timothy A. Byrnes, Catholic Bishops in American Politics (Princeton, N.J.: Princeton University Press, 1991).
65. Carmines and Stimson, Issue Evolution;
66. V. O. Key, The Responsible Electorate (Cambridge, Mass.: Harvard University Press, 1966);
67. Nie, Verba and Petrocik, The Changing American Voter;
68. Paul M. Sniderman, Richard A. Brody and Philip E. Tetlock, Reasoning and Choice: Explorations in Political Psychology (New York: Cambridge University Press, 1991);
69. Thomas M. Carsey, Campaign Dynamics: The Race for Governor (Ann Arbor: University of Michigan Press, 2000);
70. Layman, The Great Divide.
71. David E. RePass, 'Issue Salience and Party Choies', American Political Science Review, 65 (1971), 389-400;
72. Brody and Page, 'Comment';
73. Jon A. Krosnick, 'Government Policy and Citizen Passion: A Study of Issue Publics in Contemporary American Polities', Political Behavior, 12 (1990), 59-92;
74. Geoffrey C. Layman and Thomas M. Carsey, 'Party Polarization and "Conflict Extension" in the American Electorate', American Journal of Political Science, 46 (2002), 786-802.
75. In fact, even though they are arguing against the idea of a mass cultural conflict with increasing political relevance, Fiorina and his colleagues contend that the impact of existing religious cleavages on electoral behaviour and outcomes might be made greater by parties and candidates who are culturally polarized: Fiorina, Abrams and Pope, Culture War?
76. The 2000 NES interviewed roughly half (1,006) of its respondents in person and roughly half (801) over the phone. Since the 1972 study, the NES has measured respondents' attitudes and their perceptions of party and candidate positions on most political issues by asking them to place themselves, the two parties and candidates on seven-point scales. The face-to-face interviews in 2000 used the traditional seven-point scale format, but the phone interviews mainly used a branching format resulting in only five categories. To maintain continuity with the 1992 and 1996 surveys, our analysis includes only the face-to-face respondents in 2000.
77. We combine these three NES surveys for our analysis so that our results will not be greatly affected by the idiosyncrasies of a particular election campaign and will reflect the general connection between religion and contemporary political behaviour. This does have the potential to mask ongoing changes in that connection, but if the political impact of the orthodox-progressive religious divide grew over this period, then pooling the three surveys would deflate, rather than inflate, its current importance. Moreover, we conducted our analyses separately for each of the three years and did not find systematic changes in the political influence of religion.
78. David C. Leege and Lyman A. Kellstedt, eds, Rediscovering the Religious Factor in American Politics (Armonk, N.Y.: M. E. Sharpe, 1993).
79. Wuthnow, The Restructuring of American Religion;
80. Hunter, Culture Wars;
81. Lyman A. Kellstedt, John C. Green, James L. Guth and Corwin E. Smidt, 'Is There a Culture War? Religion and the 1996 Election' (presented at the Annual Meeting of the American Political Science Association, Washington, D.C., 1997).
82. Hunter, Culture Wars, p. 44.
83. The question on view of the Bible is the only one in the NES that is specifically about religious beliefs. Its response options are the Bible 'is a book written by men and is not the Word of God', 'is the Word of God but not everything in it should be taken literally, word for word', and 'is the actual Word of God and is to be taken literally, word for word'. We have recoded the responses to range from - 1 (least orthodox) to 1 (most orthodox).
84. Although literal and inerrant views of Scripture are more central to theological orthodoxy in Protestantism, especially among evangelicals, than in other traditions, authoritative views of the Bible are a component of religious orthodoxy within most major religious traditions: Lyman A. Kellstedt and Corwin E. Smidt, 'Doctrinal Beliefs and Political Behavior: Views of the Bible', in Leege and Kellstedt, eds, Rediscovering the Religious Factor in American Politics, pp. 177-98;
85. Kellstedt et al., 'Is There a Culture War?' For example, we used the Third National Survey of Religion and Politics, conducted at the University of Akron in 2000, to create an index of Catholic traditionalism consisting of responses to questions about praying the Rosary, confessing to a priest, and believing that the Pope is infallible. The correlation between that measure and a view of the Bible item very similar to that in the NES was relatively strong (0.30) and highly significant (p < 0.001).
86. Another NES item that is often used as an indicator of doctrinal orthodoxy is identification as a born-again Christian: see Layman, The Great Divide. We did not use born-again identification as a measure of belief orthodoxy here because it is much more central to evangelical identifications and beliefs than it is to those in other traditions. However, including the born-again indicator produces results nearly identical to those presented here.
87. Worship attendance and prayer are normative in all major religious traditions, and religious salience should tap into commitment in all religious contexts. Protestants may attach more importance to Bible reading than do Catholics and members of other traditions. However, past research has shown that Bible reading does distinguish between more and less orthodox Catholics: see James D. Davidson, Andrea S. Williams, Richard A. Lamanna, Jan Stenftenagel, Kathleen Mass Weigert, William J. Whalen and Patricia Wittenberg, S.C., The Search for Common Ground: What Unites and Divides Catholic Americans (Huntington, Ind.: Our Sunday Visitor, Inc., 1997). And, in the Third National Survey of Religion and Politics, the correlation between Bible reading and Catholic traditionalism is relatively strong (0.38) and quite significant (p < 0.001).
88. This analysis produced only one factor with an eigenvalue greater than 1. That factor had an eigenvalue of 2.69 and explained 67 per cent of the total variance in the five indicators. The factor loadings of the four indicators ranged from 0.79 to 0.85.
89. Religious commitment ranges from - 1 to +1, and has a reliability coefficient (a) of 0.84. We use this measure for the analyses in Tables 1 and 3. In Table 2, commitment is measured through a confirmatory factor analysis of observed indicators, treated as having measurement error.
90. Not surprisingly, our measures of doctrinal orthodoxy and religious commitment are highly correlated with each other r = 0.52, p < 0.0001). However, because beliefs and behaviours are conceptually distinct aspects of religion (see Leege and Kellstedt, Rediscovering the Religious Factor in American Politics;
91. Green et al., Religion and the Culture Wars), we leave them as separate variables here. We did conduct all of our analyses with a single measure of religious orthodoxy, combining beliefs and behaviours, and the political impact of that variable was nearly always stronger than that of the separate commitment and doctrine variables. Thus, conducting our analysis with separate measures of doctrinal orthodoxy and religious commitment may create a more rigorous test of the political impact of the orthodox-progressive divide.
92. Because all of our analyses examine the political impact of the orthodox-progressive divide within particular religious traditions (i.e., take into account religious context), the most straightforward analyses are those focusing on religious contexts. Thus, although we introduced policy contexts first in developing our argument, we begin the analysis with religious contexts.
93. Lyman A. Kellstedt and John C. Green, 'Knowing God's Many People: Denominational Preference and Political Behavior', in Leege and Kellstedt, eds, Rediscovering the Religious Factor in American Politics, pp. 53-71;
94. Lyman A. Kellstedt, John C. Green, James L. Guth and Corwin E. Smidt, 'Grasping the Essentials: The Social Embodiment of Religion and Political Behavior', in Green et al., eds, Religion and the Culture Wars, pp. 174-92;
95. Layman, The Great Divide.
96. We discuss the assignment of religious affiliations to religious traditions in Appendix A.
97. The vote is coded 0 for Democratic and 1 for Republican. We coded the other four variables to range from 0 (most liberal/Democratic) to 1 (most conservative/Republican). We use logit to estimate the vote model and regression for the other models. The political alignment index is the factor score from a principal components factor analysis of the four indicators of political behaviour. The analysis produced only one factor with an eigenvalue greater than 1. That factor had an eigenvalue of 2.38 and explained 74 per cent of the total variance in the four indicators. The factor loadings were 0.89 for party identification, 0.73 for ideological identification, 0.90 for comparative candidate evaluations, and 0.90 for the presidential vote. We use this measure for the analyses in Tables 1 and 3. In Table 2, political alignment is measured through a confirmatory factor analysis of observed indicators.
98. The control variables are income, education, region of residence (South vs. non-South), gender, age and union membership. We do not control for race because our models include a dummy variable for black Protestants and only non-blacks are included in our evangelical and mainline Protestant categories.
99. In order to account for the possibility that the intercepts in our models vary across the three survey years in our pooled data set, we include dummy variables for 1996 and 2000 respondents, with 1992 respondents serving as the comparison group. Even with dummy variables for year, there remains the possibility of non-constant error variance across years and that errors are correlated across observations within years. Although regression and logit coefficient estimates remain consistent in the face of such problems, their standard errors may be inaccurate: see Nathaniel Beck and Johnathan N. Katz, 'What to Do (and Not to Do) with Time-Series-Cross-Section Data in Comparative Politics', American Political Science Review, 89 (1995), 634-41;
100. Nathaniel Beck and Richard Tucker, 'Conflict in Space and Time: Time-Series-Cross-Section Analysis with a Binary Dependent Variable' (presented at the Annual Meeting of the American Political Science Association, San Francisco, 1996).
101. To account for this possibility, we compute robust standard errors using Huber's and White's formula for heteroscedasticity-consistent standard errors and relaxing the assumption of independent errors within years (by clustering the data on survey year in STATA): Peter J. Huber, 'The Behavior of Maximum Likelihood Estimates Under Non-Standard Conditions', Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, 1 (1967), 221-33;
102. Halbert White, 'A Heteroskedasticity-Consistent Covariance Matrix and a Direct Test for Heteroskedasticity', Econometrica, 48 (1980), 817-38.
103. Appendix B presents the full set of regression and logit coefficients.
104. The estimated slope coefficients for each tradition are taken directly from the regression and logit estimates. The slopes for evangelicals, the comparison group, are the coefficients on religious commitment and doctrinal orthodoxy themselves, and the slopes for each other tradition are the sum of the coefficient on commitment/orthodoxy and the coefficient on the interaction between commitment/orthodoxy and the dummy variable for the particular tradition. The standard errors for evangelicals are simply the standard errors around the coefficients on commitment and orthodoxy. For all of the other traditions, standard errors can be computed based on the variance of the coefficient on commitment/orthodoxy, the variance of the coefficient on the interaction term for that tradition, and the covariance of the two coefficients: see James Jaccard, Interaction Effects in Logistic Regression (Thousand Oaks, Calif.: Sage, 2001);
105. James Jaccard, Robert Turrisi and Choi K. Wan, Interaction Effects in Multiple Regression (Thousand Oaks, Calif.: Sage, 1990). Of course, the slopes and standard errors for any tradition besides evangelicals can also be obtained by re-estimating the model with that tradition as the comparison group.
106. To determine whether the differences between traditions are statistically significant when commitment, orthodoxy and all of the control variables are at their mean values, we recoded commitment, orthodoxy and the control variables so that their mean values were all equal to 0. Tests on the dummy variable coefficients then provided the information on statistical significance reported in Table 1.
107. In other words, the interactions between religious commitment and the mainline Protestant, Catholic and black Protestant dummy variables (with evangelical Protestants serving as the comparison group) are negative and statistically significant in all five models. The interactions for Catholics and black Protestants are all significant at p < 0.05. The interactions for mainline Protestants are significant at p < 0.05 in the ideology, candidate evaluation and vote models, and at p < 0.10 in the party identification and political alignment models.
108. In other words, when we make mainline Protestants the comparison group, the interaction between religious commitment and the Catholic dummy variable is not statistically significant for any dependent variable.
109. For each dependent variable, we estimated separate models with evangelical Protestants, mainline Protestants and Catholics as the comparison groups. In every instance, the interaction between religious commitment and the black Protestant dummy variable was negative and statistically significant (p < 0.05).
110. There are only three instances in which the difference between the effect of doctrinal orthodoxy for two traditions is statistically significant: the effect on candidate evaluations is more positive for black Protestants than for mainline Protestants, the effect on candidate evaluations is more positive for black Protestants than for Catholics, and the effect on ideology is more positive for mainline Protestants than for evangelicals. Only the first instance is significant at p < 0.05. The other two are significant at p < 0.10.
111. Eric C. Lincoln and Lawrence H. Mamiya, The Black Church in the African-American Experience (Durham, Ca.: Duke University Press, 1990);
112. Layman, The Great Divide;
113. Clyde Wilcox, 'Race, Religion, Region, and Abortion Attitudes', Sociological Analysis, 53 (1992), 97-105.
114. In the bottom portion of Table 1, the predicted values or probabilities for each tradition (when doctrinal orthodoxy, religious commitment and all of the control variables are held constant at their mean values) are followed by a series of superscript letters which indicate statistically-significant (p < 0.05) differences with other traditions. For example, an 'a' following a predicted value or probability indicates that the prediction for a particular tradition is significantly different from that for evangelical Protestants, while a 'b' indicates that the prediction is significantly different from that for mainline Protestants. That the value for black Protestants is followed by an 'a' and a 'b' for each dependent variable indicates that, when orthodoxy, commitment and the control variables are held at their means, black Protestants are significantly different from both evangelical and mainline Protestants on each variable.
115. Wuthnow, The Struggle for America's Soul;
116. Layman, The Great Divide.
117. See, for example, Miller and Shanks, The New American Voter.
118. Hunter, Culture Wars.
119. We use the political alignment index here for two reasons. First, party identification, comparative candidate evaluations and the presidential vote all reflect individuals' feelings of identification with or support for the two major political parties; and liberal and conservative ideologies are closely identified with the Democratic and Republican parties. Thus the four indicators all tap into one's degree of alignment with the major combatants in American national politics. Secondly, we estimated our structural model for each of the four indicators separately and the results were very similar to those presented here.
120. See, for example, Herbert McClosky and John R. Zaller, The American Ethos: Public Attitudes Toward Capitalism and Democracy (Cambridge, Mass.: Harvard University Press, 1984);
121. Stanley Feldman, 'Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values', American Journal of Political Science, 32 (1988), 416-40;
122. Stanley Feldman and Marco R. Steenbergen, 'The Humanitarian Foundation of Public Support for Social Welfare', American Journal of Political Science, 45 (2001), 658-77.
123. Although orientations towards social welfare and racial issues typically fall into two distinct attitudinal dimensions (see Layman and Carsey, 'Party Polarization and "Conflict Extension" in the American Electorate'), we do not separate the indirect effects of orthodoxy through the two sets of attitudes because we assume that they are both influenced by egalitarianism and support for limited government.
124. Our model also includes attitude towards the death penalty as an indicator of racial attitudes. Race plays a major role in the debate over the death penalty, and in our exploratory factor analysis of domestic policy attitudes, death penalty attitude loaded much more strongly with racial issues than with other types of issues.
125. These spending questions are the only indicators of environmental and defence and foreign policy attitudes that are included in each of the 1992, 1996 and 2000 NES surveys. Also, there are no indicators of core values associated with environmental or defence and foreign policy attitudes that are included in all three surveys.
126. To provide a scale for the latent variables, we constrain the factor loading for one observed indicator of each latent variable to be equal to 1. All of the observed indicators are coded to range from 0 for the least orthodox, most liberal or most Democratic orientation to 1 for the most orthodox, most conservative or most Republican orientations. Thus, the latent variables take on the same range of values.
127. We include the standard demographic controls and dummy variables for 1996 and 2000 in the equation for every endogenous political variable. Because the literature on voting behaviour widely identifies evaluations of the incumbent president, evaluations of prevailing economic conditions, and assessments of presidential candidates' personal traits as shaping vote choice, candidate evaluations, and perhaps even party identification (see Campbell et al., The American Voter;
128. Morris P. Fiorina, Retrospective Voting in American National Elections (New Haven, Conn.: Yale University Press, 1981), the political alignment equation also includes these three variables. Economic evaluations are measured by assessments of trends in the national economy over the past year. We did not include assessments of personal finances in the model because the question was asked to only half of the sample (and half of the face-to-face respondents) in 2000 and because the literature suggests that evaluations of the national economy are more closely related than personal financial conditions to electoral choice. We coded both economic evaluations and presidential approval ratings so that higher scores in all three survey years should be associated with stronger support for the Republican party and its candidates. In 1992, with a Republican president in office, approval ratings range from disapprove strongly to approve strongly, and economic evaluations range from the national economy got much worse over the past year to the economy got much better. In 1996 and 2000, with a Democratic incumbent in place, the variables are coded in just the opposite direction.
129. Assessments of candidates' personal traits are measured by questions asking how well certain characteristics describe both parties' presidential candidates. Questions about five of these characteristics - moral, really cares about people like you, knowledgeable, provides strong leadership, and honest (asked as 'dishonest' in 2000) - were included in all three survey years. We took the difference in each respondent's assessment of how well each of these five traits described the Republican and Democratic candidates. A principal components factor analysis of these five difference scores yielded only one factor with an eigenvalue greater than 1 (which explained 67 per cent of the total variance) and all five difference scores loaded strongly (0.75 or higher) on that single factor. Our measure of assessments of candidate traits is the factor score from that analysis.
130. This approach does not allow us to examine the effect of religious tradition, independent of religious orthodoxy, on political attitudes and behaviour. However, we have established that independent effect in Table 1.
131. Appendix Table A2 presents all of the measurement and structural (path) coefficients in the models. The coefficients on the control variables are available from the authors upon request.
132. We do not exclude missing values from our analyses. We estimate our model using Amos 4.0, which computes full information maximum likelihood (FIML) estimates even in the presence of missing data: see T.W. Andersen, 'Maximum Likelihood Estimates for a Multivariate Normal Distribution When Some Observations are Missing', Journal of the American Statistical Association, 52 (1957), 200-3.
133. Wothke and Arbuckle describe this FIML procedure and show that the estimates produced by it are more consistent and efficient than those produced by pairwise or listwise deletion of missing observations (Werner Wothke and James L. Arbuckle, 'Full-Information Missing Data Analysis with Amos' (Chicago: SPSS White Paper, 1996), see http://www.spss.com.
134. All non-zero direct effects in Table 2 are statistically significant at p < 0.10. Direct effects that are not statistically significant are given a value of 0 in the table, and are treated as 0 in computing indirect and total effects.
135. Green et al., Religion and the Culture Wars.
136. Andrew Kohut, John C. Green, Scott Keeter and Robert C. Toth, The Diminishing Divide: Religion's Changing Role in American Politics (Washington, D.C.: The Brookings Institution, 2000).
137. Layman, The Great Divide.
138. We use only the 1996 NES for this analysis because it was the only one of the three presidential-year NES surveys from 1992 to 2000 to ask respondents both how important the abortion issue was to them and to place the two parties (and not just their candidates) on the abortion scale. Abortion was the only moral issue on which questions about salience and party positions were asked in any of the three surveys.
139. We use the same measures of religious commitment, doctrinal orthodoxy and political alignment as in the analysis in Table 1, but here for only 1996 NES respondents. We return here to simple models including only demographic controls and not controls for issue attitudes and core values. We do so for two reasons. First, we already have shown that the impact of religious commitment and doctrinal orthodoxy on political behaviour is largely indirect, exerted through issue attitudes and policy predispositions. The purpose of this analysis is not to again uncover the nature of commitment and doctrine's indirect effects, but to assess the political contexts in which their impact on political behaviour is felt. Secondly, the sample sizes for some of the groups for which we estimate our model are quite small. That makes the estimation of models that include a wide range of policy attitudes and political values very difficult.
140. The number of Jewish respondents in some of these subgroups is too small for us to estimate meaningful coefficients on the Jewish variable and its interaction with orthodoxy. So, we exclude Jews from these analyses.
141. We do not show the predicted values of political alignment for various traditions when religious commitment and doctrinal orthodoxy are at their sample means (as in Table 1) because the differences between traditions are not affected much by the salience of and awareness of party differences on abortion. We performed the same analyses with party identification, ideology, candidate evaluations and the vote as separate dependent variables. The results were all very similar to those for the political alignment index.
142. Kohut et al., The Diminishing Divide.
143. See, for example, Mark J. Rozell and Clyde Wilcox, eds, God at the Grass Roots: The Christian Right in the 1994 Elections (Lanham, Md.: Rowman and Littlefield, 1995);
144. Mark J. Rozell and Clyde Wilcox, eds, God at the Grass Roots 1996 (Lanham, Md.: Rowman and Littlefield, 1997);
145. John C. Green, Mark J. Rozell and Clyde Wilcox, eds, Prayers in the Precincts: The Christian Right in the 1998 Elections (Washington, D.C.: Georgetown University Press, 2000).
146. Kohut et al., The Diminishing Divide.
147. Layman, The Great Divide;
148. Fiorina, Abrams and Pope, Culture War?