THE UNIVERSITY OF VIRGINIA VOL. 57, NO.7

THE 1980 ELECTION: UNDERSTANDING THE REAGAN VICTORY IN VIRGINIA By David B. Magleby

Mr. Magleby is a research associate on the staffofthe the popular vote in the nation as a whole. fairly stable, in the 50 to 60 percent range. In Institute of Government. After having cast its twelve electoral votes in Virginia, in contrast, turnout has grown 1976 for Gerald Ford, Virginia in 1980 from only 21 percent in 1948 to 49 percent in On the eve of the 1980 election, most returned to its more typical role ofvoting for 1980-with a dramatic growth of 18 per­ pollsters had declared the presidential race the national winner. (In only one other centage points occurring between the 1960 to be too close to call. While some public election since 1932-that of 1960-has and 1968 presidential elections alone. opinion surveys during the final week of Virginia failed to cast its electoral votes for This significant increase in the proportion the campaign had detected a growing the winning candidate.) of the voting age population casting ballots trend towards former California Governor This news letter will assess the reasons for in the state largely reflects growth in the Ronald Reagan, the fact that President the Reagan victory in Virginia by examining overall levels of registration, with the Carter had narrowed Reagan's substantial survey data gathered in public opinion polls greatest proportional growth occurring early lead made it difficult to assess which conducted both before and after the No­ among blacks. Between 1960 and 1971 the candidate might enjoy a last-minute surge. vember election. The discussion begins with proportion of the black voting age popula­ Yet when the votes were counted, Reagan a brief review, over time, of the state's voter tion registered to vote more than doubled had won the election by a surprisingly turnout and vote for president. Then, after rising from 23 percent to 52 percent. Growth decisive popular vote margin of 10 percent; explaining the methodology of the polls, the in the proportion of whites registered to vote and in carrying forty-four states, Reagan analysis will focus on our findings of why was less dramatic, rising from 46 percent in had garnered 489 electoral votes, compared Virginians voted as they did for president, 1960 to 60 percent in 1971. 1 Several factors with only 49 for President Carter. with particular attention to the relative account for this increase in registration. Virginia's voters chose Reagan over importance ofcampaign issues and ideology, Among the most important were passage in Carter by a margin almost identical to the candidate appeal, and party identification. 1964 of the 24th Amendment, which abol­ nationwide breakdown. In Virginia, 53 Finally, some attention will be devoted to the ished poll taxes as a condition for voting in percent voted for Reagan, compared to 51 question of whether the results of the 1980 federal elections; the 1966 U.S. Supreme percent nationwide; 40 percent chose Carter, election constitute a party realignment in Court decision extending the abolition of compared to 41 percent nationwide; and 5 Virginia. poll taxes to state elections;2 and the percent voted for John Anderson, while the suspension of literacy tests as mandated by independent candidate received 7 percent of TURNOUT the 1965 Voting Rights Act. Thus, an Slightly less than half of Virginia's voting essential backdrop for the 1980 presidential age population cast ballots in the 1980 election is the recognition that the pool of general election, only slightly below the Author's Note: The Virginia data used in this news letter have participating voters in Virginia elections state's record high turnout of 51 percent of been combined from two separately designed surveys. One, since 1960 has been expanded substantially. conducted under the direction of the author, was carried out by age-eligible voters, achieved in 1968. Nation-' sixty University of Virginia students as part of a course on public wide, Virginia ranked 43rd in turnout. THE VOTE FOR PRESIDENT opinion and elections. The students had completed several weeks of Several factors account for Virginia's rela­ instruction on question wording, sampling, and interviewing tively low turnout in 1980, including the In presidential elections since 1932 the technique. Funding was provided by the Dean of the College of absence of contests for statewide office breakdown of the vote by party in both Arts and Sciences, the Student Council, and the Woodrow Wilson which made the election less stimulating, and ':ir~inia and the nation has been quite Department of Government and Foreign Affairs. This poll used a the widespread belief that Governor Reagan Similar. In the four elections from 1932 random-digit dialing technique in which every telephone number in through 1944, when Fr.anklin D. Roosevelt Virginia had an equal chance of being sampled. was going to carry Virginia by a substantial margin. was the Democratic party nominee, Virginia The second survey, also by telephone, was conducted between Nationally, voter turnout has declined in was substantially less Republican in voting October 16 and 19 under the auspices of Mr. Larry Sabato, every presidential election since 1960. Part of choice than the nation as a whole. In assistant professor of government and foreign affairs, for a elections since 1948, however, the fortunes of syndicate of newspapers and television stations, using a statewide this decline is due to the extension of the sample provided by Survey Sampling, Inc., of Waterbury, voting franchise to the 18-21 year olds in Republicans have been virtually the same in Connecticut. The author would like to thank Mr. Sabato for his 1972, since this group is among the least Virginia and nationwide. In 1952, 1956, kind permission to use the data from that survey. likely to turn out. Figure 1 examines turnout I See Charles V. Hamilton, The Bench and the Ballot ( ew The combined sample of the two pre-election surveys included in presidential elections since 1948 for both York: Oxford University Press, 1973), pp. 238-9.

slightly over 1,000 respondents. Of that number, 686 were Virginia and the nation as a whole. As this 2 Harper v. Virginia State Board of Elections, 383 U.S. 663 reinterviewed in the week following the election. figure shows, turnout nationally has been (1966).

INSTITUTE OF GOVERNMENT / UNIVERSITY OF VIRGINIA / CHARLOTTESVILLE / MARCH 1981 26

Figure 1 acquired in childhood as the result of parental socialization and tends to remain Turnout of Voting Age Population in Presidential constant throughout adult life. In our poll, Elections in Virginia, 1948-1980 an individual's party identification was determined by coding the responses to the following questions: Generally speaking in politics, 60% do you think of yourself as a u.s. Democrat, a Republican, an inde­ pendent, or what? 50% ...,...... •...... •...... • (If Republican or Democrat) ..•• Would you call yourself a strong ••• Virginia Republican (Democrat) or a not 40% .... very strong Republican (Demo­ •• crat)? E ...•.. 0 ...... (If independent) Do you consid­ t: 30% •...... ~ •• er yourself as leaning toward the r- •••• Democratic or Republican party? •• This series of questions produces these seven 20%; categories of partisanship: strong Republi­ can, weak Republican, independent Repub­ 10% lican, independent, independent Democrat, weak Democrat, and strong Democrat. Our survey of Virginia .voters_ io_1980 0% revealed that the two parties are divided almost evenly, the Democrats having 45 1948 1952 1956 1960 1964 1968 1972 1976 1980 percent and the Republicans, 42 percent.4 Presidential Elections Because Democrats are less likely to vote, if nonvoters had been part of our sample, the SOURCES: Estimates of the voting age population were provided by the Tayloe Murphy Institute, University of Virginia, and the U.S. Bureau of the Census, Projections o/the Population 0/ Voting Age/or States: November Democrats would enjoy a wider margin over 1980. The State Board of Elections provided figures for total votes cast in Virginia, and the national average data the Republicans. Party identification consti­ were derived from U.S. Bureau of the Census, Statistical Abstract ofthe U.S. 1979, p. 513, and Congressional tutes a standing decision about partisan Quarterly Weekly Report, November 8, 1980 for the 1980 election, p. 3297. politics; although individuals may change party preference, that happens quite rarely. 1972, and 1980 Republicans won by substan­ change by interviewing the same respond­ Rather, most people use their partisan tial margins in both Virginia and the nation. ents at two different times. More than 1,000 predisposition to simplify political choices Similarly, elections that were close nation­ respondents were interviewed initially be­ and to filter out conflicting information. All wide, such as those in 1960 and 1976, were tween October 16 and 19, and 686 were other things being equal, a person will vote also close in Virginia. The only exception to reinterviewed in November, in the week for the party with which he identifies. Things this pattern is the 1968 election, when the following the election. At least three efforts are not always equal, of course, and national result was close but Richard Nixon were made to reach each of the pre-election sometimes either a candidate oran issue may won Virginia by II percent. It is important to respondents during the post-election survey. prompt a voter temporarily to defect from remember, however, that independent Unlike aggregate voting data, which his preferred party. But this is precisely the George Wallace obtained a substantial 24 report the vote by voting precincts, cities, feature of party identification that is most percent of the Virginia popular vote in 1968, and counties, survey data gathered from distinctive; party preferences persist beyond a proportion that almost certainly affected public opinion polls permit us to compare the short-term forces ofa particular election. the votes for Nixon and for Humphrey, the individual voters, thereby advancing our In 1980, how important was partisanship Republican and Democratic nominees. 3 understanding of why people voted as they in the voting choice of Virginians? Table 1 did. Surveys also permit us to learn ifcertain summarizes the party identification of EXPLAINING THE REAGAN APPEAL individual characteristics such as age, sex, Virginia voters in our sample, and the vote IN VIRGINIA and religion are correlated with the political for president for each of the categories of lust as Ronald Reagan produced a acts being studied. Our polls revealed, for party identification. Voters who considered convincing national victory in 1980 by instance, that Reagan voters in Virginia were themselves to be Republicans (strong, weak, winning forty-four states and capturing a of both sexes, and of diverse ages and and independent Republicans) as well as majority in all regions, so too in Virginia the religious backgrounds. Men and women those identifying themselves as strong Reagan campaign demonstrated strong voted for Reagan in almost equal propor­ Democrats were very loyal to their party's appeal in all sections of the state, winning tions (men, 58 percent for Reagan; women, nominee. Former Governor Reagan's suc­ sixty-eight of Virginia's ninety-five counties 56 percent for Reagan). Reagan voters were cess in Virginia is in large part accounted for and twenty-five of forty-one cities. Reagan both young (56 percent of the 18-24 year not only by the loyalty ofall three categories won by substantial margins in the suburbs olds) and old (55 percent of those over age of Republicans, but also by the defection of (Reagan-59 percent, Carter-33 percent), the 65); and Catholics, Protestants, and Jews in roughly one-third ofthe weak and independ­ rural areas (Reagan-53 percent, Carter-42 Virginia all voted for Reagan (58 percent of ent Democrats. Reagan also carried the pure percent), and the urban corridor (Reagan-53 the Catholics and Protestants, and 55 independents, by a margin in excess oftwo to percent, Carter-39 percent). In 1976 Carter percent of the Jews). one. not only had carried Virginia's central cities Previous studies of voting behavior in Issues and Ideology. The economy and the by substantial margins but also had beaten other states and nationally have identified related issues of inflation and unemploy­ Gerald Ford in Virginia's rural areas as well. three primar, explanatory variables for ment were perceived by the vast majority of Thus, the 1980 vote was an across-the-board presidential voting: the individual's party Virginia voters in our survey (78 percent) to victory for Reagan, a victory that public identification, the issues of the campaign, be the most important problem facing the opinion surveys done at the University of and the presence or absence of the candi­ country in 1980. Coupled with this was the Virginia help to explain. date's personal appeal. What role did each of widespread sentiment that Carter had done These polls were designed as a panel study, these factors play in the 1980 presidental poorly in managing the economy; this which permits the researcher to assess vote in Virginia? 4 Those independents who leaned towards one ofthe parties were

3 George Wallace carried five states and received forty-five Party Identification. The most important classified as partisans. Past research by the author and others has electoral votes in 1968. Of the southern states, Wallace had his influence on voting is that ofparty identifica­ demonstrated that partisan independents are in effect more like lowest percentages in Virginia and Texas. tion, a psychological loyalty that typically is partisans than independents. 27 perception was a central determinant in the said that they were either middle-of-the road a way as possible. Similarly, they will election outcome both in Virginia and ideologically, didn't know their ideology, or attempt to call attention to the opponent's nationwide. On the one hand, Reagan's had not thought about liberalism and shortcomings-which, they argue, will affect campaign was successful in reminding voters conservatism. In a national poll, George the capacity to govern. Unlike 1976 when, on of their economic woes, while on the other Gallup found 55 percent of adult Americans balance, both Ford and Carter were per­ hand Carter was unsuccessful in diverting to be similarly non-ideological. This large ceived positively, the candidates in 1980 had attention away from the economy to other body of non-ideological voters has long been predominantly negative images, at least in issues like foreign and defense policy, thought to be responsible for a campaign Virginia. Nearly two thirds (65 percent) of energy, or Reagan. Thus, Carter had to face strategy of appealing to the moderate or Virginia voters in our poll, when asked, a referendum of sorts on his stewardship of centrist voter, and the success of Governor offered reasons to vote against Carter, while the economy. Each candidate attempted to Reagan in not allowing himself to be painted 58 percent offered reasons to vote against make his opponent the issue; Carterdepicted as an extremist explains part of his drawing Reagan. In contrast, 38 percent offered Reagan as "dangerous" and "trigger-happy," power among Virginia voters. reasons to vote for Carter, and 43 percent while Reagan revised the "misery index"­ Despite the widespread attention given by offered reasons to vote for Reagan. Carter the sum of the inflation and the unemploy­ the media to the religious right generally, was perceived as being unable to handle the ment rates, first employed by candidate and the Moral Majority specifically, Virgin­ job, as well as being a weak leader. Reagan Carter against Gerald Ford. In an NBC ia voters surveyed overwhelmingly indicated was thought to lack experience and to have News/ Associated Press nationwide pre­ that these activities did not affect theirvoting traits of personality that might not be suited election survey, potential voters saw Reagan choices. Only 3 percent of Virginia voters to the presidency. On balance, candidate as better able to handle the country's indicated that they wanted to vote in ways appeal, like issues, worked to Reagan's economic problems by a margin of almost encouraged by the Moral Majority, while 9 advantage. As will be demonstrated, the two to one (Reagan-41 percent, Carter-21 percent volunteered that they wanted to vote campaign-especially the second debate­ percent, Anderson-II percent, no difference­ against the Moral Majority. The remainder and the last-minute hostage announcement 26 percent).s And in the end, most voters in said either that they did not know of the tended to reinforce the voting intentions of Virginia and elsewhere perceived Reagan as Moral Majority or that it had no effect on Reagan supporters, to sway a majority ofthe the candidate most likely to deal effectively their vote. A slightly higher percentage of undecideds, and even to convert some of the with the economic problems. Virginia voters (11 percent) indicated that Carter and Anderson voters to the Reagan How important to Virginia voters were the they would follow the political advice of camp. ideological predispositions of the candi­ their religious leaders; but for most Virgin­ ians in 1980, religion and politics were CHANGES IN VOTING INTENTIONS dates? While the liberalism-conservatism DURING THE CAMPAIGN dimension may be a part of the political separate spheres. In the three-week period between our first system in the minds of some voters, 40 Candidate Appeal. The third factor that and second polls, Republican Ronald percent of the Virginia voters in our survey often influences presidential voting, in Reagan's support in Virginia jumped 15 perceived themselves as neither conservative addition to party identification and issues, is percentage points. Table 2 presents the cross nor liberal. Of those who did perceive the presence or absence of a candidate's tabulation of the pre-election voting inten­ themselves as ideological, the majority were personal appeal. Often, as discussed above, tions of our sample and the vote actually conservative, by a ratio of three to one. issues are part of the public's general reported in the post-election survey. As that Republican identifiers on the whole tended reaction to a candidate. However, aspects of table indicates, Reagan's substantial in­ to be more conservative, but even strong personality, background, and character also crease in the last three weeks is largely Democrats were almost equally divided can playa distinct role in how candidates are accounted for by his having gained a two-to­ between conservative and liberal. Thus, even perceived and which candidate wins. In 1952 one advantage among the voters who were though the Democrats enjoyed a slight edge and 1956, for instance, Dwight Eisenhower undecided in October. Reagan also did in the proportion of party identifiers, the was perceived as an excellent and likable better in wooing Carterand Anderson voters ideological perspective of voters who pro­ candidate, this positive reaction stemming to join his bandwagon. The finding that so fessed to have an ideology was conservative. from his experience as the Allied Command­ many Virginia voters decided so late for An important finding ofour poll, however, is er in World War II, his distance from whom they would vote is consistent with the that 40 percent of those who voted in 1980 "politics," and his substantial personal national survey conducted on election day charm. 5 Survey conducted between October 22 and 24, 1980 for NBC by CBS News and the New York Times, News/The Associated Press, as reported in "Opinion Roundup," 3 Candidates and their managers work hard which found 25 percent of those who voted Public Opinion 20 (December/ January 1981). at portraying the candidates in as appealing reporting that they decided how to vote during the last week of the campaign. Our poll provides useful insight into why Table 1 so many voters were undecided at the end of Party Identification and Voting for President in Virginia, 1980 the campaign, and why Reagan did much better in winning their votes. Because the Percent Voting for campaign had focused on the negative aspects of both candidates, clearly many Party Identificationa Reagan(R) Carter (D) Anderson (I) Other voters had a difficult time finding reasons to Strong Republican (13%) 93% 4% 3% vote for any of the candidates. Among these Weak Republican (8%) 79 17 4 undecided voters in Virginia, the events of Independent Republican (21 %) 85 7 5 3 the last week-particularly the debate and the announcement that the Iranian Parlia­ Independent (12%) 65 24 5 6 ment would vote on the possible release of Independent DeIl}.ocrat (14%) 33 56 7 4 the hostages-worked to Reagan's benefit. Weak Democrat (16%) 36 50 9 6 In addition, Carter proved unable to pacify Stong Democrat (15%) 10 85 4 1 the widespread concern of Virginia voters about the economy. On October 28, exactly one week before SOURCE: Surveys described in author's note. the election, candidates Reagan and Carter NO!E: Some responses, reflecting missing information, "don't know," etc., have been excluded from this table, met to debate in Cleveland, Ohio. Over 80 leavmg an N = 581. Also, percentages in this table may not always equal 100 percent, due to rounding. percent of Virginia voters surveyed reported that they had watched part or all of the aThe figure in parentheses represents, in each case, the percentage oftotal respondents that identified themselves as, for ~xample, Strong Republican. For all categories of the Republican vote, the total is 42 percent; for all debate. Of that number, 53 percent thought categones of the Democratic vote, the total is 45 percent. Reagan had won and 12 percent thought Carter had won. Perhaps more importantly, 28

Table 2 Democratic. What turned the tide for Reagan in Virginia was the combined effect Change in Vote Intentions among Virginians, October to November, 1980 of the central issue, the economy, and the widespread disenchantment of weak and Voters' Pre-election Percent Reporting a Vote for independent Democrats, as well as pure Vote Intentions Carter (D) Reagan (R) Anderson (I) Other independents, with Carter. Did the 1980 election constitute a realign­ Carter 88% 10% 1% 1% ment in the nation as a whole? Before Reagan 3 96 1 1 examining this question, it is important to underscore the meaning of a realignment. In Anderson 19 22 58 a realigning election, there would be sub­ Undecided 30 60 7 2 stantial evidence of a lasting change or shift in political preferences, usually party prefer­ ences. In other words, a party or a candidate SOURCE: Surveys described in author's note. could do well in a particular election and yet OTE: Some responses, reflecting missing information, "don't know," etc., have been excluded from this table, voters could still remain committed to their leaving an N =580. Also, percentages in this table may not always equal 100 percent, due to rounding. preexisting party preferences. All available evidence indicates that voters were doing precisely this in 1980. This does not mean over 60 percent perceived the news media to voters, the failure to secure the release of the that a successful Reagan presidency could have declared Reagan the winner, compared hostages made 8 percent more likely to vote not shift the Republican party into majority to only 6 percent who felt Carter had been for Reagan, compared to the 4 percent who status in the electorate. It simply means that declared the victor. This widespread percep­ stated that the Sunday announcement made in the minds of many voters, a vote' for tion that Reagan had won the debate them more likely to vote for Jimmy Carter. Reagan or Republican senatorial candidates solidified the Reagan vote, reassuring these in 1980 did not mean that they stopped voters that their preference was correct. The perceiving themselves as Democrats. CONCLUSIONS: 1980 NOT YET A Does the fact that roughly a third of the debate seemed to change the voting inten­ REALIGNMENT tions of only 13 percent of Virginia voters, weak and independent Democrats in Virgin­ ia voted for Republicans Reagan and Bush with Reagan being the primary beneficiary The act of voting represents the interac­ of the changes. constitute a realignment in the Common­ tion of a variety of forces. For some voters, wealth? Probably not. As discussed above, their lasting party attachments preclude Similarly, most Virginia voters reported Republicans carried Virginia by wider voting for the opposing party's candidate, that the announcement of a possible resolu­ margins in 1952, 1956, and 1972, only to see a while for others the issues and candidates tion of the year-old hostage crisis did not Democratic resurgence in some subsequent simply reinforce their standing party prefer­ affect their voting choice. The fact that the elections. As the evidence in this news letter ences. For still others, the particular circum­ hostages were not released on Sunday, demonstrates, Virginia, at the presidential stances of a campaign may motivate them to November 2, 1980, despite the widespread level at least, has been a competitive two­ abandon party and to vote primarily on the reports of such a possibility and the party state since 1948. basis of issues or a candidate's personal announcement by the Iranian parliament But whatever the future of the party that an end to the hostage crisis might be at appeal. system in Virginia and the nation, the hand, again tended to reinforce the choice of In the 1980 election in Virginia, Republi­ Republican victory in 1980 was substantial. those who had already decided to vote for cans remained loyal to their party's standard In terms of presidential voting, the voters Reagan. Apparently, most voters had had bearer. Strong Democrats, despite the chose the candidate that they perceived as their hopes dashed too many times to trust unpopularity of their nominee and his most attractive, most likely to change the the media speculation about the imminent apparent inability to deal with the economy, status quo, and most likely to succeed in release of the hostages. Among Virginia also were loyal and voted overwhelmingly dealing with the economy. Erratum: The February 1981 issue was incorrectly identified as Vol. No.6; it properly should be Vol. 57, No.6.

THE UNIVERSITY OF VIRGINIA

Entered as second-class matter w Charlottesville, Virginia LET E R

(ISSN 0042-0271) Editor I Clifton McCleskey Assistant Editor I Sandra H. Wiley Published each month from September through August by the Institute ofGovernment, University of Virginia, Charlottesville, Virginia 22903. The views and opinions expressed herein are those of the author, and are not to be interpreted as representing the official position ofthe Institute or the University. Entered as second-class matter January 2, 1925, at the post office at Charlottesville, Virginia, under the act of August 24, 1912. c 1981 by The Rector and Visitors of the University of Virginia. Printed by the University Printing Office.