Voters and Values in the 2004 Election

A poorly devised exit poll question undermined meaningful analysisof voters’ concerns in the 2004 presidential election.Twenty-two percent of voters picked "moral values" from a listof "issues" describing what mattered most in their vote, morethan selected any other item. Various commentators have misinterpretedthis single data point to conclude that moral values are anascendant political issue and to credit conservative Christiangroups with turning George W. Bush’s popular vote defeatin 2000 into his three million–vote margin of victoryin 2004. We suggest, rather, that while morals and values arecritical in informing political judgments, they represent personalcharacteristics and ill-defined policy preferences far morethan any discrete political issue. First by conflating moralsand values and then by further conflating characteristics andissues, the exit poll’s "issues" list distorted our understandingof the 2004 election. In this article, we examine the flawsin the 2004 National Election Pool exit poll’s "most importantissue" question and explore the presumed rising electoral importanceof moral values and the conservative Christians who overwhelminglyselected this item. Using national exit poll data from 1980through 2004 and other national surveys, we find that the moralvalues item on the issues list cannot properly be viewed asa discrete issue or set of closely related issues; that itsimportance to voters has not grown over time; and that whencontrolled for other variables, it ranks low on the issues listin predicting 2004 vote choices. The aggregated exit poll dataalso show that the voting behavior of conservative Christiansis relatively stable over time, and these voters were not primarilyresponsible for Bush’s improvement in 2004 over 2000.     

Preelection Polls in Israel: Structural Constraints on Accuracy

This article examines the last two Israeli election campaignsof 1981 and 1984, and suggests that structural variables greatlyaffect the accuracy of reported preelection polls. Pollsters'quest for more detail and refinement driven by competition andmass media pressures for highly processed predictions is liableto result in less accurate reports. The kind of errors in polls'projections are situation-bound and greatly depend on the natureof the floating vote in the system. Interpollster accuracy,however, is mainly determined by pollster–sponsor dependencies.Pollsters working for political parties projected results morefavorable to their sponsors or less favorable to their sponsors'opponents. Moreover, the stronger the pollster–sponsorstructural dependencies, the greater the inaccuracy.     

RACE-OF-INTERVIEWER EFFECTS IN A PREELECTION POLL VIRGINIA 1989

All published preelection surveys of the 1989 Virginia gubernatorialcontest overestimated the vote share of the black candidateand eventual victor, Douglas Wilder. We offer a "social desirability"interpretation of the polls' inaccuracies and hypothesize thatclaiming support for Wilder was the socially desirable responsefor some whites, especially when the interviewer was black.We show a race-of-interviewer effect on the vote intention ofwhite respondents of 8–11 percentage points in a preelectionsurvey of Virginia voters. The effects were greatest among whiteDemocrats and among whites who were more uncertain of theirvote intention. We discuss the implications of these findingsfor race-of-interviewer research and for improving the accuracyof preelection forecasts in contests with black and white opposingcandidates.     

BUILDING A PROBABLE ELECTORATE FROM PREELECTION POLLS: A TWO-STAGE APPROACH

Prominent among sources of nonsampling error in preelectionpolls are different methods of screening likely voters. Buildingon past research, we use National Election Studies data to constructa simple, practical, two-stage model of voter turnout. We comparethe performance of this model to alternative estimators, usingdata from the 1988 and 1992 presidential elections. By allowingfor truncated interviews with up to 20 percent of respondents,our two-stage model provides turnout estimates that are slightlymore accurate and significantly less expensive than traditionalmethods for screening likely voters.     

THE OBSESSION TO FORECAST: PRE-ELECTION POLLS IN THE ISRAELI PRESS

The generous coverage of pre-election polls in the media hasstimulated a debate on the effects of publicized polls on voters.Data for this study come from a content analysis of all pressreports on polls during six pre-election campaigns in Israel(1969–88). The findings highlight the impressive increaseof poll reports both by space and by prominence in the press.Analysis of the reports' content reveals that the style of reportingpolls has become dominated by "horse-race" journalism, wherepredictions and popularity ratings are the leading themes. Inaddition, although the reports have become more detailed andinformative, they still do not adequately define or interpretmethodological deficiencies. Finally, the analysis examinesthe intercorrelations between attributes of coverage and predictionserrors. The fact that the independent variable, media coverageof polls, has changed significantly calls for the inclusionof this factor in any study of polls' effects.     

An Elite View of the Polls

This research note examines the views of election polls of asample of political activists: delegates elected to the 1980state conventions of the Oklahoma Democratic and Republicanparties. Findings show that although just l3 percent of thedelegates overall place high levels of trust in the resultsof major national polls, nearly a majority (45 percent) feelthat poll results affect voter behavior. Both trust in electionpolls and perception of their impact on voters are found toincrease with involvement or interest in electoral politics.     

Rating the Polls: The Views of Media Elites and the General Public

To measure the credibility of the polls in 1984, Gallup conductedtwo surveys, one of the press and one of the general public.Results of the press survey indicated a favorable reaction tothe accuracy of the preelection polling in 1984, although therewas a division of opinion about whether the polls were betterat explaining the vote than predicting it, and whether the pollsenhanced or interfered with the electoral process. The public'sperception of the polls was a generally positive one both asto their accuracy and their overall effect.     

The Poll as a News Event in the 1980 Presidential Campaign

Public opinion polls made up 15 percent of the news events coveredduring the 1980 presidential election campaign, but news storiesgenerated by polls were no more likely to receive better playin the 50 newspapers studied than other stories about the campaign.Horserace polls were popular with editors during the last twoweeks of the campaign.     

The Accuracy of the National Preelection Polls in the 2004 Presidential Election

The 2004 presidential election campaign provided a venue fora wide variety of polling, and it was not without its controversies.In the end, the final estimates of the preelection polls, thebread and butter of the polling industry, were very good atsuggesting it would be a close race, with Bush the likely winner.In historical perspective, the overall performance was aboveaverage for the period since 1956. Issues raised in the medialeading up to the end of the campaign and the final estimates,however, created some controversy, especially about the likelyvoter methodology used by different organizations. There werealso some anomalies at the end of the campaign as some firmsand collaborators ended up producing different estimates ofthe outcome depending on likely voter definitions or the modeof data collection.     

The Role of Corporations in the 1980 U.S. Congressional Elections*

The issues of whether and how corporate campaign donations affect the operation of American elections have recently become the focus of public and scholarly debate. Using Federal Election Commission campaign finance data, this analysis identifies a link betwren contributions from business PACs and candidate success in the 1980 U.S. congressional elections. Findings also confirm an association between corporate support and legislative voting for candidates re-elected in 1980. Candidates heavily supported by business PACs during the campaign did better at the polls and were more likely to vote conservatively if elected than those not so strongly endorsed by economic interests. Results are discussed in terms of various theories of the state as well as their relevance to potential election finance reforms.     

WHY THE DEMOCRAT-REPUBLICAN PARTISANSHIP GAP VARIES FROM POLL TO POLL

Polls taken in late 1980 and 1984 varied by up to 20 pointsin the spread they reported between Democratic and Republicanidentifiers. We found three systematic and sizable differencesacross polls: polls that sample only voters, polls that emphasize"today" or the present in their question wording, and pollsthat are taken close to election day (at least in circumstanceslike those of 1984) tend to favor the party currently advantagedin the presidential race.     

Sampling Telephone Numbers and Adults, Interview Length, and Weighting in The California Health Interview Survey Cell Phone Pilot Study

This article describes several features included in a CaliforniaHealth Interview Survey cell phone pilot study that differ fromearlier cell phone surveys conducted in the United States. Onedifference is that the study used a screening design and onlyadults living in cell-only households were interviewed. Mostof the previous studies used dual frame designs. Another differencewas the development and implementation of a within-householdadult sampling procedure to cover adults when cell phones wereshared in the household. The study was also intended to determineif conducting a cell phone interview of the same scope and lengthas a regular telephone interview was feasible. Most previoussurveys were focused on cell phone topics and were shorter thanthe comparable landline surveys. We then explore some interestingproblems that arise in weighting a cell phone survey using ascreening design. We conclude with a discussion of planninga new cell phone survey in 2007 that incorporates findings fromthis study.     

Likely (and Unlikely) Voters and the Assessment of Campaign Dynamics

Only in recent years has the "likely voter" technology beenextended to polls well in advance of an election. In the caseof the 2000 U.S. presidential election, CNN/USA Today/Galluptracking polls indicated considerable fluctuations in likelyvoter preferences, greater than among the larger pool of registeredvoters surveyed. This article explores how Gallup’s likelyvoter model exaggerates the reported volatility of voter preferencesduring the campaign. Much of the reported variation in candidatepreference reported by Gallup in that election is not due toactual voter shifts in preference but rather to changes in thecomposition of Gallup’s likely voter pool. The findingshighlight dangers of relying on samples of likely voters whenpolling well before Election Day.     

A Review and Proposal for a New Measure of Poll Accuracy

This article proposes a new measure of the predictive accuracy(A) of election polls that permits examination of both accuracyand bias, and it applies the new measure to summarize the resultsof a number of preelection polls. We first briefly review pastmeasures of accuracy, then introduce the new measure. Afterthe new measure is described, the general strategy is to applyit to three presidential elections (1948, 1996, and 2000) andto compare the results derived from it to the results obtainedwith the Mosteller measures. Then, the new measure is appliedto the results of 548 state polls from gubernatorial and senatorialraces in the 2002 elections to illustrate its application toa large body of preelection polls conducted in "off-year" raceswith different outcomes. We believe that this new measure willbe useful as a summary measure of accuracy in election forecasts.It is easily computed and summarized, and it can be used asa dependent variable in multivariate statistical analyses ofthe nature and extent of biases that affect election forecastsand to identify their potential sources. It is comparable acrosselections with different outcomes and among polls that varyin their treatment or numbers of undecided voters.     

"SECRET BALLOTS" AND SELF-REPORTS IN AN EXIT-POLL EXPERIMENT

A controlled exit-poll experiment on Election Day (November3, 1992) shows that refusals to answer questions and other evasiveforms of responding were significantly lower when respondentswere given a self-administered "secret-ballot" questionnairethan when they were interviewed face-to-face. While there weresome suggestive interactions of this mode-of-data collectioneffect with age and sex, and with the partisan atmosphere ofthe precinct in which the interviews were conducted, they wereborderline in significance, inconsistent in pattern, or difficuitto interpret. More important, comparisons with official electionreturns (the truth) indicated that the secret-ballot techniquewas more accurate than were face-to-face interviews in estimatingthe final outcome on the most socially sensitive of three self-reportedvotes: a vote for or against a tax levy for elderly services.Using a genuine secret ballot in the tradition of the olderGallup preelection polls can thus increase the validity of self-reportsin exit polls today.     

Picking the Winners: Politician vs. Voter Predictions of Two Controversial Ballot Measures

Abstract Both politicians and voters were asked to predict outcomesof two Oregon ballot measures in 1982. As expected, politicians'predictions always were closer to the mark than voters' were.Further, voters showed stronger signs of wishful thinking (the"Looking-Glass effect") in their predictions than did politicians.Using published preelection polls apparently improved politicians'accuracy in 1982, as well as voters' accuracy in a separate1984 survey. No other sources of data improved predictive accuracy.Findings have implications for theories of representative governmentand are consistent with a new theory of public opinion.     

What's Missing from National Landline RDD Surveys?: The Impact of the Growing Cell-Only Population

The number of cell phone only households has continued to grow– 12.8 percent of all households by the end of 2006, accordingto the National Health Interview Survey (NHIS). To assess theextent and nature of the potential bias in landline telephonesamples created by the absence of cell-only households, thePew Research Center conducted four independent dual frame studiesin 2006, three of which included at least 200 interviews withcell-only respondents. Despite the fact that there are significantdifferences between cell-only and landline respondents on manyimportant variables, across the four surveys with differentsubstantive content we find that including a cell-only samplewith a landline RDD sample produces general population estimatesthat are nearly identical to those from the landline samplealone. Yet, while the noncoverage problem is currently not damagingestimates for the entire population, we find evidence that itdoes create biased estimates on certain variables for youngadults, 25 percent of whom are cell-only according to the mostrecent government estimate.     

Preelection Polling in the 1982 Illinois Gubernatorial Contest

All the polls in advance of the 1982 Illinois gubernatorialelection proved incorrect in their prediction of a substantialvictory for James Thompson, the Republican. After examiningsome of the popular explanations for the "missed call," thispaper shows that while the voters "preferred" Thompson, therewas a very substantial increase in the number of straight partyballots cast for the Democrats, compared to 1978. This paperconcludes that in those states where straight party vote isan option, the poll takers probe for the possibility of a partyline vote. Information used in this paper to examine five hypothesesstems from a preelection poll of 1,200 Illinois voters.     

Pre-election polling and third party candidates

I analyze voters’ incentives in responding to pre-election polls with a third party candidate. Third party supporters normally have an incentive to vote strategically in the election by voting for one of the major candidates. But these voters would vote third party if the third party candidate is doing surprisingly well in the polls. Because voters are more likely to vote third party if the third party candidate is doing well in polls, voters who like the third party candidate best have an incentive to claim they will vote third party in the polls so that more voters will ultimately vote third party in the election. The differing incentives faced during polls and elections accounts for why third party candidates do better in polls than in elections.