Hostname: page-component-586b7cd67f-g8jcs Total loading time: 0 Render date: 2024-11-28T13:51:41.807Z Has data issue: false hasContentIssue false

Partisanship Reinstated? A Comparison of the 1972 and 1976 U.S. Presidential Elections

Published online by Cambridge University Press:  27 January 2009

Extract

Recent presidential elections in the United States have been marked by widely divergent landslide victories. In 1964 Lyndon Johnson led the Democrats to a sweeping victory over Barry Goldwater. In 1968 Richard Nixon captured the White House for the Republicans in a contest that was close only because George Wallace ran as a third-party candidate and siphoned off a large share of the conservative vote. With Wallace eliminated from the 1972 race, Nixon easily won a lopsided re-election victory over George McGovern. Early in the campaign the 1976 election also appeared to be heading towards an overwhelming victory, this time for the Democrats. Yet, Jimmy Carter won by only two percentage points.

Type
Articles
Copyright
Copyright © Cambridge University Press 1978

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 For a discussion of those factors affecting a particular election result see Converse, Philip E., Clausen, Aage R. and Miller, Warren E., ‘Electoral Myth and Reality: The 1964 Election’, American Political Science Review, LIX (1965), 321–36CrossRefGoogle Scholar; Converse, Philip E., Miller, Warren E., Rusk, Jerrold G. and Wolfe, Arthur G., ‘Continuities and Change in American Politics: Parties and Issues in the 1968 Election’, American Political Science Review, LXIII (1969), 10831105CrossRefGoogle Scholar; Miller, Arthur H., Miller, Warren E., Raine, Alden S. and Brown, Thad A., ‘A Majority Party in Disarray: Policy Polarization in the 1972 Election’, American Political Science Review, LXX (1976), 753–78CrossRefGoogle Scholar; and Miller, Arthur H. and Miller, Warren E., ‘Issues, Candidates and Partisan Divisions in the 1972 American Presidential Election’, British Journal of Political Science, v (1975). 422–3.Google Scholar For a good general discussion of the relevant elections, see Asher, Herbert, Presidential Elections and American Politics (Homewood, Ill.: The Dorsey Press, 1976), pp. 159–90.Google Scholar

2 The data used in this report were gathered by the Institute for Social Research at the University of Michigan as part of the Center for Political Studies' 1976 national election survey, made possible by a grant (no. 76–13562) from the National Science Foundation. In the total election survey of 2,875 eligible voters, 50·5 per cent said they had voted for Carter and 48·5 per cent had chosen Ford; the 1 per cent who had voted for either McCarthy or Maddox were excluded from any analysis in this report.

3 The data over time suggest that while the growth in non-identifiers (i.e. ‘pure’ independents combined with ‘apoliticals’) reached a peak in 1974 (18 per cent) and declined during the following two years (to 15 per cent), the shifts between 1972 and 1976 were clearly not statistically significant. Perhaps it would thus be more reasonable to conclude that popular support for the political parties had really not changed during that four-year period.

4 The ‘feeling thermometer’ is a scale used to rate either social groups or political leaders; it ranges from zero degrees to 100 degrees with 50 as the neutral category. The high end of the scale represents the most positive feelings toward a group or candidate; the low end, the most negative feelings.

5 See Miller, et al. , ‘A Majority Party in Disarray’.Google Scholar

6 A principal hypothesis suggested by Scammon, Richard M. and Wattenberg, Ben J. in The Real Majority (New York: Coward McCann and Geoghegan, 1970)Google Scholar is that in order for Democrats to win presidential elections, they must de-emphasize the so-called ‘social’ issues (race relations, law and order, welfare reform) and return to traditionally Democratic economic issues.

7 In 1972 economic difficulties had not yet become very widespread, nor were they as salient as social concerns. They thus played a very minor role in that election. See Miller, et al. , ‘A Majority Party in Disarray’.Google Scholar

8 The recent recession has renewed interest in the study of economic effects on voting behaviour. For some valuable insights on the topic, see the symposium in American Political Science Review, LXIX (1975)Google Scholar which is comprised of the following articles: Arcelus, Francisco and Meltzer, Allan H., ‘The Effect of Aggregate Economic Variables on Congressional Elections’, 1232–9Google Scholar; Bloom, Howard S. and Price, H. Douglas, ‘Voter Response to Short-Run Economic Conditions: The Asymmetric Effect of Prosperity and Recession’, 1240–54Google Scholar; Goodman, Saul and Kramer, Gerald H., ‘Comment on Arcelus and Meltzer’, 1255–65Google Scholar; and Arcelus, Francisco and Meltzer, Allan H., ‘Aggregate Economic Variables and Votes for Congress: A Rejoinder’, 1266–9.Google Scholar

9 A multivariate analysis predicting the vote decision from evaluations of how well the government was dealing with both inflation and unemployment revealed an equal impact by each type of policy performance assessment. The standardized regression coefficients for both variables were identical (·14). The multiple correlation (·24) was only slightly larger than the zero-order correlation between the economic policy evaluation variables and the vote (·20). Thus concern about unemployment had an impact on the vote similar to but clearly not independent of that found for inflation.

10 For a report on trends in consumer sentiment and economic outlook during the period preceding the election, see Curtin, Richard T., ‘Recovery in Consumer Sentiment Resumes’, Economic Outlook USA, III (1976), 60–1.Google Scholar

11 Citrin, Jack, ‘Comment: The Political Relevance of Trust in Government’, American Political Science Review, LXVIII (1974), 975–6Google Scholar, argues that the trust-in-government items simply tap support for the incumbent authorities and not any deeper attitudes toward the political regime.

12 The five standard trust-in-government items asked in SRC/CPS surveys since 1958 are worded as follows: (a) Do you think that people in the government waste a lot of money we pay in taxes, waste some of it, or don't waste very much of it? (b) 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? (c) 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? (d) Do you feel that almost all of the people running the government are smart people, or do you think that quite a few of them don't seem to know what they are doing? (e) 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?

13 For an explanation of contemporary political distrust, see Miller, Arthur H., ‘Political Issues and Trust in Government: 1964–1970’, American Political Science Review, LXVIII (1974), 951–72.CrossRefGoogle Scholar

14 CBS/New York Times Poll, New York Times, 19 10 1976, p. 1.Google Scholar The table which accompanied Robert Reinhold's article showed a significant growth (from 20 per cent to nearly 40 per cent) in the proportion of the public who perceived Carter as liberal between June and October. Unfortunately, the accompanying article, entitled ‘Survey Shows Carter Holds Lead Because He is Seen as Moderate’, missed the main point expressed by the data in the table.

15 For a presentation of the data documenting McGovern's relative unpopularity, see Miller, and Miller, , ‘Issues, Candidates and Partisan Divisions’, pp. 393434.Google Scholar

16 The discussion of candidate effects has focused on the presidential candidates only because comparable longitudinal data is not available for the vice-presidential candidates. A later report will, however, investigate the impact of the vice-presidential candidates on the outcome of the 1976 election.

17 The correlation reported here indicates the relationship between a measure based on the difference in the thermometer ratings of the two candidates (Ford-Carter) and the vote. The individual Ford thermometer correlated with the vote at 61; the Carter thermometer correlation was 71.

18 Earlier studies that have considered reactions to candidates as explanations of the vote include Stokes, Donald E., ‘Some Dynamic Elements of Contests for the Presidency’, American Political Science Review, LX (1966), 1928CrossRefGoogle Scholar; Kelley, Stanley Jr., and Mirer, Thad W., ‘The Simple Act of Voting’, American Political Science Review, LXVIII (1974), 572–91CrossRefGoogle Scholar; and Kagay, Michael R. and Caldeira, Greg, ‘I Like the Looks of His Face: Elements of Electoral Choice, 1952–1972’ (paper presented at the annual meeting of the American Political Science Association, San Francisco, 1975).Google Scholar

19 For a more extensive discussion of change in attitudes towards Carter's performance during the general election campaign period, as well as further analysis of the impact that Ford's incumbency had on the election outcome, see Miller, Arthur H. and Miller, Warren E., ‘Partisanship and Performance: Rational Choice in the 1976 Presidential Election’ (paper delivered at the annual meeting of the American Political Science Association, Washington, 1977).Google Scholar

20 Downs, Anthony, An Economic Theory of Democracy (New York: Harper and Row, 1957), Chap. 3.Google Scholar

21 A complete statement of the normal vote concept and computational methods can be found in Converse, Philip E., ‘The Concept of a Normal Vote’, in Campbell, Angus, Converse, Philip E., Miller, Warren E. and Stokes, Donald E., Elections and the Political Order (New York: Wiley, 1966), pp. 939.Google Scholar

22 Pomper, Gerald M., Voters Choice (New York: Dodd, Mead, 1975), Chap. 3.Google Scholar

23 Miller, et al. , ‘A Majority Party in Disarray’.Google Scholar

24 While turnout has not been discussed in this report, it should be noted that if a larger proportion of the eligible electorate had participated, Carter's margin should have increased because he was the choice of 60 per cent of non-voters. The 1976 turnout was the second lowest in modern history with only slightly more than 53 per cent of eligible voters participating; it is surpassed only by the 1948 election when 52 per cent participated. The 1976 turnout figure is derived from the official presidential election totals as reported in the New York Times (13 12 1976, p. 65)Google Scholar with an estimated 148 million eligible voters, not adjusted for institutionalized individuals, as the base.

25 A four-variable, multiple equation path analysis including party identification, a single composite issue proximity measure, a composite candidate thermometer, and the vote, was used to compute the relative total effects that parties, candidates and issues had on the vote in 1976. The model underlying the analysis is as follows:

The model assumes that party identification is a prior predisposition to attitudes towards issues and candidates. It also assumes that attitudes towards issues are not determined by candidate ratings.

26 The total effects for the 1972 and 1976 elections are presented below. The 1972 results are from Miller, et al. , ‘A Majority Party in Disarray’, p. 769.Google Scholar

The resulting total effects clearly show that in 1976 party identification had a greater impact than issues or candidates. For a concise statement of path analysis and the technique of computing total effects, see Stokes, Donald E., ‘Compound Paths in Political Analysis’, in Herndon, James F. and Bernd, Joseph L., eds., Mathematical Applications in Political Science V (Charlottesville: University Press of Virginia, 1971), pp. 7092Google Scholar