Traditionally, pollsters trying to accurately assess voter intentions have struggled with a basic problem — figuring out who actually is going to show up to vote. In the 2000 election campaign, sharp fluctuations in the Gallup Organization’s daily tracking poll were blamed by some on difficulties in nailing down likely voters. Similar complaints arose during the 1998 congressional elections, when some critics of President Clinton charged that likely voter samples included too many Democrats sympathetic to Clinton. In 1999, the Pew Research Center undertook an experiment to study the accuracy of the likely voter models employed for decades by leading survey organizations, including Gallup and the Pew Research Center.

In this comprehensive experiment, we set out to discover how many of the voters we classified as “likely” actually voted. The Center used as a test case the closely-contested 1999 mayoral race in Philadelphia. It was one of that city’s closest ever — just 9,447 votes separated the victor, Democrat John Street, from his Republican rival Sam Katz. Polling was conducted in two waves among 2,415 registered voters: one survey was taken two weeks before the election, the second was conducted in the last week before voters went to the polls. Aside from the usual battery of questions assessing voting preferences and intentions of voting, participants were asked to provide their name and address — information which was used to match pre-election survey responses with actual voting records. Overall, we were able to match 70% of registered voters polled with Philadelphia voting records to determine whether they actually voted.

What we discovered was that the traditional methods originally developed by Gallup in the 1950s to sort voters from non-voters still work reasonably well, particularly when compared to the alternatives. This method uses an eight-item likely voter index designed to assess not only a voter’s preferences, but their past history of voting, interest in the campaign and knowledge of where to vote on Election Day. Using this index, the Center correctly predicted the voting behavior of 73% of registered voters.

Clearly, the index does not forecast the behavior of all respondents. The 73% accuracy rate means that 27% of respondents were wrongly classified — those who were determined as unlikely to vote but cast ballots (17%), or non-voters who were misclassified as likely to vote (10%). But the likely voter index successfully identified the preferences of respondents who actually went to the polls, according to the validation study. In the second wave of polling, conducted the week before Election Day, Street held a three-point lead among registered voters. But the dead heat among likely voters more accurately reflected the split among those respondents who actually voted.

More important, the result virtually mirrors the findings of a similar voter validation study conducted by Gallup during the 1984 presidential election, which correctly classified 69% of registered voters. While the polling business has undergone massive changes since then, the likely voter index remains a model of consistency.

Here is a summary of our principal findings. A more comprehensive analysis by Pew Center survey director Michael Dimock is the subject of a paper presented May 19 at the annual conference of the American Association for Public Opinion Research. (Complete paper; You can also contact Michael Dimock via email at mdimock@pewresearch.org).

One of the main successes of the likely-voter index is in identifying a pool of respondents which, if not a perfect replica of the electorate, shares a similar demographic profile with the voting public. It gives a better approximation of the electorate than using registered voters.

The survey of likely voters conducted two weeks before the election was nearly as accurate in assessing voters’ preferences as the one taken just prior to the election. Still, it is possible that the effectiveness of the likely voter screen would decline if used too early in an election campaign. In 1998, the Democratic composition of the likely voter poll changed significantly between September and Election Day.

Using a single measure to determine likely voters increases the risk of bias and subjects estimates to the vagaries of individual elections. Some individual questions, such as those relating to actual voting intention (“Do you plan to vote?”) result in too many respondents being classified as likely voters, thus increasing the chance of a Democratic bias. Other rifle-shot questions — focusing on a respondent’s level of attention to the campaign, for example, — may classify too few people as likely voters, resulting in a Republican bias.

Interestingly, the Pew experiment showed that, in particular, an individual question related to a candidate’s strength of support (how strongly the respondent supported the candidate) is not closely associated with voter turnout. Registered voters who expressed no preference in pre-election surveys were nearly as likely to vote as those who voiced strong support for candidates (and those with no preference were more likely to vote than those expressing moderate candidate support).

Clearly, Gallup’s index is not the only way of identifying likely voters — other indices, employing as few as four or as many as 15 questions, also are effective. These indices are superior to relying solely on samples of registered voters, and also are more effective than using regression analysis to determine a probability of voting for each respondent.

While the long and contentious election aftermath drew massive media coverage, the rising price of gasoline attracted the most public interest of any news story of 2000. In June, more than six-in-ten Americans (61%) said they paid very close attention to this story, which far outpaced the election and other news events of the year. Another quarter of the public followed the hikes in gas prices fairly closely, according to the Pew Research Center’s surveys of public attentiveness to domestic and international news stories.

Gas prices attracted considerably less press coverage than the election and other developments, but drivers are confronted with the price of gas every time they fill up their tanks. Consequently, interest in gas price hikes remained at a high level throughout the year. While overall attentiveness peaked in June, it reached 58% in March and 56% in October. Interest in gas prices has not been this high since the Persian Gulf crisis in 1990.

The October terrorist attack on the USS Cole ranked as the second most-closely followed news story, with 44% of Americans paying very close attention. Another automobile-related story, the recall of defective Firestone tires, ranked third, followed by the Michigan shooting of a 6-year-old girl by a classmate and the controversy over whether Elian Gonzalez should be returned to Cuba.

The post-election political and legal drama ranked only sixth, but it drew a relatively large and sustained audience for a political story. Nearly four-in-ten Americans (38%) tracked post-election developments very closely in the period just after the Nov. 7 election (Nov. 10-17), while another 38% followed this story fairly closely (76% total). That is comparable to the audience for major political stories of recent years, including the Monica Lewinsky scandal; 36% paid very close attention to that story during its peak period in September 1998 and another 36% followed it fairly closely.

While this year’s unprecedented political events found a sizable audience, many other important developments, in fields such as science and business, did not. Just 16% of the public paid very close attention to the breakthrough in mapping the human genome, and 17% very closely followed the merger between media giants AOL and Time-Warner. The federal ruling ordering the breakup of Microsoft drew more interest, with 28% paying very close attention.

Similarly, overseas stories failed to draw much interest, with the exception of the ongoing violence between Israelis and Palestinians. The Middle East conflict attracted very close attention from three-in-ten Americans, making it the year’s top international story (aside from the attack on the Cole, which occurred in Yemen). Fewer than one-in-five closely followed the political turmoil in Yugoslavia that led to Slobodan Milosveic’s ouster, and just 10% paid very close attention to the floods that ravaged Mozambique.

Not surprisingly, more college graduates tended to follow major stories in the areas of science, business and international affairs than those with high school degrees. In July, for example, one-quarter of college graduates paid very close attention to the genome announcement, compared to just 12% of high school graduates. By contrast, the highest-rated story among high school graduates that month (23% paying very close attention) was the video showing Philadelphia police beating a carjacking suspect; 16% of college grads followed that story very closely.

This year’s top stories mark a departure from the 1990s, in that death and disaster did not lead the list. In the previous two years, school shootings — in Littleton, Colo. and Jonesboro, Ark. — attracted most attention. Before that, the death of Princess Diana, the crash of TWA flight 800 and the Oklahoma City bombing were the top-ranked news stories.

Election’s Last Act Popular

Interest in this year’s presidential campaign climbed through the early party primaries, sagged in the long post-primary season and spiked as the election went into its drawn-out endgame.

And the audience for the long Florida recount saga declined, but only modestly, after the initial post-election period, according to Pew’s daily tracking poll of 5,719 adults conducted from Nov. 10 to Dec. 14. The percentage of those following the recount very closely decreased from 38% in the initial post-election phase (Nov. 10-17), to 32% for the period between Nov. 18-Dec. 7. The audience then increased slightly (to 34% very interested) for the series of climactic developments between Dec. 8-14 — the Supreme Court ruling stopping the recounts, Al Gore’s concession speech and President-elect George W. Bush’s victory address.

There were big differences in the attentiveness of self-described voters and those who said they did not vote. At least 40% of voters said they paid very close attention to the news from Florida and elsewhere during each phase of the recount saga, and at least eight-in-ten said they followed the story very or fairly closely.

Non-voters were far less interested in the election story to begin with, and their attentiveness dropped sharply as the recount fight dragged on. Fewer than one-fifth (22%) paid very close attention in the week of Nov.10-17; that number dropped to 12% the following week before rising slightly in the final two tracking periods. While 61% of non-voters said they were very or fairly interested in the story in the first week after the election, barely half remained at least somewhat interested in the succeeding weeks.

In the immediate aftermath of the election, Bush and Gore voters followed the recount battle in roughly equal numbers. But more supporters of the Texas governor stayed focused on the story. During the week of Nov. 10-17, 49% of Bush voters said they paid very close attention to the story, compared to 47% of Gore supporters. For the Nov. 27-Dec. 7 period, interest among Bush voters remained virtually unchanged, with 47% paying very close attention. But by then, 37% of Gore voters were intensely interested, a 10-point decline from the initial period. As the saga concluded, interest among supporters of both men rose slightly, but more Bush voters followed the story very closely (51% compared to 41% of Gore supporters).

The election news interest survey, conducted as part of the daily tracking poll of the Pew Internet & American Life Project, has a margin of error of plus or minus 1.5 percentage points. The list of the year’s top news stories is derived from monthly Pew Research Center surveys, which ask people how closely they follow selected news stories.

Overlooked amid controversies over hanging chads and divided courts were some important lessons from the presidential campaign. Here are previously unreleased findings from the Pew Research Center’s post-election survey, conducted Nov. 10-12:

There was more positive voting and less voting against candidates this time than in the three previous presidential elections. George W. Bush received about the same grades from his voters as his father did in 1988, while Al Gore earned the same kind of ratings from his backers that Bill Clinton received in 1996.

Abortion proved to be the sleeper issue of Campaign 2000 — it was volunteered as a decisive issue most often by Bush voters who said issues mattered most. It rated less important, but still highly, among Gore voters. For the vice president’s supporters, abortion was ranked nearly as important as the environment, and not much below Social Security and education.

An expected backlash from the Columbine massacre and other school shootings in favor of gun control never materialized. In fact, Bush voters who were motivated by issues cited his stance on gun control more often than issue-oriented Gore voters (14% to 4%).

Dick Cheney was a bigger plus for Bush voters than Joe Lieberman was for Gore supporters. More than one-third (35%) of those who backed the GOP ticket cited Cheney as an important factor in their choice, compared to 15% of Gore voters who said that about Lieberman. By contrast, neither Cheney nor Lieberman was much of a drag on their tickets.

Party was cited more often as a motivating factor by Gore voters (57%) than Bush voters (44%). This may be a reflection of greater GOP satisfaction with their candidate, which also surfaced in other ways throughout the campaign.

Still, Gore did not suffer much from the enthusiasm gap that was evident in many pre-election polls. The vice president’s support actually grew in the campaign’s final days and turnout was high among core Democratic constituencies.

Approximately equal percentages of Bush and Gore voters said they were urged to vote by the campaigns, but respondents in battleground states were contacted more often than those in less competitive places.

Fully 53% of voters in Ohio, Michigan, Wisconsin, West Virginia, Minnesota, Oregon, Washington, New Mexico, Pennsylvania, Missouri, and Florida said they were contacted over the phone by candidates, campaigns or other groups urging them to vote in a particular way, compared to only 38% of voters in other parts of the country.¹

Voters in the battleground states changed their minds more (or were persuaded to do so by the campaigns) than did voters in less competitive places. The horse race fluctuated dramatically in battleground states, but moved less in strong Bush and Gore states.

However, Gore may owe his popular vote victory win to strong support in Democratic states. Bush’s margin, on the other hand, fell in safe GOP states in the final weeks of the campaign.

Campaign 2000 firmly established the Internet as a major source of election news and information. But as the audience for online campaign news has expanded — increasing fourfold over the past four years — it has gone more mainstream in its preferences and pursuits. A majority now cites convenience, not a desire to tap new or different information sources, as the main reason they go online for election news. Many more election news consumers gravitate to the online addresses of major news organizations and web portals than seek out specialized political sites or the candidates’ own sites.

Nearly one-in five Americans (18%) say they went online for election news during this year’s campaign, up from 4% who did so in the 1996 campaign. Fully one-third of the online population, which itself has grown dramatically over the past four years, got election news from the Internet. Veteran online users were far more likely to get election news than Internet “newbies”: 45% of those who have been online for at least three years used the Internet to access election information, compared to 17% of those who began going online in the past six months.

The Pew Research Center’s latest nationwide survey — conducted during October and November among 4,186 online users, in association with the Pew Internet & American Life Project — finds that election information draws more election news consumers than participation in online political activities. Nearly seven-in-ten of those who went online for election news sought out information on the candidates’ positions.

This information clearly had an impact: 43% of election news consumers say it affected their voting decisions, up from 31% in 1996. The effect of online campaign news has been particularly pronounced among young people. Fully half of online election news consumers under age 30 say the information they received made them want to vote for or against a particular candidate. Still, there has been no indication that the Internet is actually drawing more young people — or for that matter, more people of any age — into the political process. Controlling for other factors related to participation, Internet users are no more likely to be engaged in the political process, and show no greater propensity to vote than do non-users.¹

Convenience is the Internet’s main appeal as a campaign news source. More than half of those who went online for election news (56%) cited convenience as their main reason for doing so, up from 45% in 1996. During that campaign, when the Internet had yet to fully emerge as a news medium, a majority of election news consumers (53%) said they went online because they weren’t getting all the news they wanted from traditional media; just 29% cited that factor in the current survey.

The online audience for election news has shown less interest in engaging in other political activities than it has in accessing political information. But the Internet is beginning to affect the way candidates and voters interact. A sizable minority of those who went online for election news (35%) registered their views in Internet political polls, while 22% used email to contact candidates and 5% made campaign contributions over the Internet.

Online Use Spiked at Campaign’s End

On a day-to-day basis, interest in online campaign news peaked around Election Day. Fully 12% of Americans went online for political news on November 7, and 18% went online the day after the election — a more than fourfold increase in the normal traffic for political news. The continuing drama of the post-election story and its fast-moving pace kept people online for election news in the week following November 7. Between 11% and 15% of all Americans followed the story online on any given day during this period.

The current poll was conducted as part of a larger effort to track Internet use over time. The Pew Internet & American Life Project has been conducting a daily tracking poll almost continually since March 1, 2000, measuring online usage. The level of campaign news consumption remained remarkably stable from March through October; during this period, roughly 3% to 5% of Americans got campaign news online on a typical day. This pattern was not influenced dramatically by important primary contests in March, nor by the conventions or the fall debates.

No matter who wins the presidency, George W. Bush or Al Gore, many Americans think that the victor will come to office because of the way the voting was conducted or counted rather than because he legitimately won the election in Florida. Ironically, this view has become more prevalent regarding a potential Bush victory since the most recent certification of the Florida vote.

The second wave of the latest Pew Research Center’s survey, conducted Nov. 27-30, finds the public thinking Bush will have legitimately won, by a 48% to 33% margin. But the first part of this survey, conducted Nov. 20-26, showed a wider 54% to 28% margin thinking Bush will be a legitimate winner. A plurality of voters, both before and after the Nov. 26 certification, say a victory by Gore would be a matter of way the votes were counted rather than a legitimate win.

Overall, there is mostly good news in this survey for the Bush camp — the percentage of Americans thinking he will be the next president rose from 58% last week to 66% in the polling conducted since Sunday. At the same time, the number of Americans who believe that Bush won and Gore should concede increased from 36% to 45%; and the number who said that Gore has no chance of winning rose from 31% to 41%.

Not surprisingly views on what might happen and what should happen are colored by partisanship — Republicans, almost to a person, think Bush will be president and Gore should concede. Among Democrats, however, fewer than one-in-four (23%) think the vice president should throw in the towel. However, over half of Democrats (56%) now think that Bush will be the next president.

It should be noted that even though many think that either candidate will have won the vote because of the way it was counted, most Americans still will accept either candidate as a “legitimate” president, according to other national polling conducted this week. This is because the vast majority think it impossible to get a completely accurate count of the Florida vote, according to a Gallup poll last week.¹

Even with an uncertain conclusion, Campaign 2000 gets better grades from the public than most recent presidential contests. Fully 83% of voters say they learned enough from the campaign to make an informed choice — a larger percentage than expressed that view in surveys conducted in the days following the three previous presidential campaigns. There was less mudslinging, more discussion of issues, and the debates were more useful than four years ago, said a sample of voters in the Pew Research Center’s quadrennial “weekend after” poll.

The poll, conducted Nov. 10-12 among 1,113 voters, also finds a greater number of respondents satisfied with the choice of candidates this year than in 1988, 1992 and 1996 — although most were only fairly satisfied with this year’s slate. Voters also are more generous in grading themselves this year, as 60% award themselves a grade of A or B for their performance, up from 43% in Pew’s survey following the 1996 campaign. But the public’s negative view of the press is unchanged from previous elections. Just 29% give the press top grades, while nearly four-in-ten (39%) graded the media D or F.

Media miscalls of the outcome of the presidential race on Tuesday have only intensified voters’ long-standing criticisms of press performance. Seven-in-ten voters (69%) voice anger or disappointment with the networks’ premature calls that George W. Bush had won the presidency. More than half of voters (52%) believe the networks’ earlier mistake of calling Florida for Gore may have had an effect on how people in other parts of the country voted (with as many as 58% of those in the West supporting this view). Little wonder that the perception that the media had too much of an influence on the outcome climbed to 53% in the current survey from 47% in 1996 and 46% in 1992.

Roughly nine-in-ten respondents (87%) want the networks to wait to announce winners until nearly all the votes are counted on election night rather than predicting a winner when they feel confident based on early returns. Almost as many (81%) think news organizations make projections in order to be the first to call the winner, rather than because they think it’s important that their audience know the outcome as soon as possible.

More GOP Satisfaction

In keeping with the enthusiasm advantage Bush enjoyed during the campaign, Republicans were more satisfied than Democrats with the choice of presidential candidates this year: 81% of Republicans vs. 66% of Democrats say they were very or fairly satisfied. This represents a dramatic increase in satisfaction on the part of Republicans, only 39% of whom were content with their choices in 1996. Independents were less satisfied this year than partisans — only 58% liked their choices.

By overwhelming margins, voters say they learned enough about the candidates and the issues to make an informed choice. Again, this represents a significant improvement from recent presidential elections. Men and women, young and old, Republicans and Democrats agree on this point. No major demographic group subscribes to the view that it was difficult to choose among the candidates because they didn’t learn enough from the campaign. The widely publicized indecision on the part of voters this year was not due to a lack of information about the candidates or the issues.

The presidential debates were quite influential in helping voters decide which candidate to support. Nearly two-thirds of voters (62%) found the debates helpful — up from 41% in 1996. More women than men say the debates helped them make up their minds (66% vs. 58%). Young voters were more influenced by the debates than older voters — 74% of those under age 30 vs. 57% of those age 65 and older say they found the debates instructive. Bush and Gore voters evaluate the debates about equally. Not only were the debates helpful, but a sizable minority of voters (20%) say they made up their minds definitively during the debate period, up from only 3% in 1996.

On balance, voters say there was more discussion of the issues in this campaign compared to past presidential elections — 46% vs. 36% who say there was less discussion of issues. Again, that is a significant improvement from 1996 when 65% of voters said there had been less discussion of issues.

At the same time, voters saw less mudslinging in this campaign compared to past presidential elections. By a margin of 46%-34%, voters say there was less, not more, negative campaigning this time around; another 16% said there was about the same amount this time. In both 1996 and 1992, voters said negative campaigning was on the increase. Bush and Gore voters have similar evaluations of the campaign in this regard — 49% of Gore voters saw less mudslinging this year. Among Bush voters, 43% said less.

As was the case in both 1992 and 1996, the candidates’ commercials were not well received by the voters. Two-thirds say the commercials were not helpful to them in deciding which candidate to vote for, down only moderately from 73% in 1996. Young voters are among the most critical of the candidates’ commercials — 72% say they were not helpful. Democrats judge the commercials as slightly more helpful than do Republicans (37% vs. 27%). Still most deemed them to be of little importance.

By the weekend, George W. Bush was further ahead of Al Gore in the battle for public opinion than he was in the vote recount in Florida. A strong majority of voters, including many Gore supporters, think he will have legitimately won the presidency if he prevails in the Florida recount. And a comparable majority rejects the idea of conducting a new election in Palm Beach County Fla., even though many voters there may have been confused by ballots.

A nationwide survey of 1,113 voters by the Pew Research Center gave Bush and the Republicans slightly higher grades than Gore and the Democrats for their handling of the uncertain outcome of the election. But the poll, conducted Friday through Sunday (Nov. 10-12) as the two sides wrangled over vote-count procedures, finds a wide political gulf on these and other questions relating to the election’s aftermath.

Overall, 51% of voters say Bush has done an excellent or good job of handling the electoral outcome, compared to 46% who say the same about Gore. Not surprisingly, supporters of each man strongly endorse their candidate’s handling of the confusing situation; 76% of Bush supporters say the Texas governor has done an excellent or good job, while seven-in-ten Gore backers (72%) give the vice president top grades.

But a substantial minority of Gore voters (41%) say Bush will be the legitimate victor of the presidential contest if he wins in Florida and thus gains a narrow majority in the Electoral College; 51% of Gore supporters disagree. There is more unanimity on this point among Bush voters. Fully 95% of those who support the Texas governor say he will be the legitimate victor in the presidential contest if he wins in Florida.

Generally, voters who question the legitimacy of Bush’s victory, should he win Florida, do so because of concerns over the accuracy of the vote count rather than because of Gore’s lead in the popular vote. While 15% have doubts about the accuracy of the vote count, just 6% cite Gore’s popular vote advantage (which remains unofficial) as a reason to question the legitimacy of a Bush victory in the Electoral College.

Most voters (67%) believe that if Bush does win in Florida, Gore should concede the election. Democrats are divided — 42% think the vice president should concede while 47% say he should fight the results in court because of the voting problems in Florida. Again, there is a consensus among Republicans, with 93% in favor of Gore conceding.

Silver Lining?

If there is an upside to these events, it may be the positive effect they have had on perceptions of the importance of voting. Nearly three-in-four voters (74%) say the election’s razor-thin outcome has increased their belief that every vote counts, while only 23% say it has raised doubts in their minds about the importance of voting. Moreover, there is less of a partisan divide on this question — solid majorities of both Gore and Bush supporters feel more strongly about the relevance of their vote as a result of this process.

At the same time, many voters harbor doubts that the final tallies in Florida and other closely contested states will indeed be accurate. A 49% plurality is confident that the results will be accurate, while 42% disagree. Gore supporters are far more skeptical than Bush voters. Just 35% of the vice president’s backers believe the final count will be accurate, compared to 64% of Bush supporters.

Whatever doubts voters have, however, only a minority (36%) supports a new election in Palm Beach County because of alleged ballot irregularities. Gore voters favor a new election by better than a two-to-one margin (64%-30%), while Bush supporters reject the idea, 90%-7%.

The Pew Research Center’s final pre-election poll of 1,301 likely voters, conducted Nov. 2-5, finds 45% backing George W. Bush, 43% for Al Gore, 4% for Ralph Nader, less than 1% supporting Pat Buchanan, and 8% undecided. The Bush margin is well within sampling error, and narrowed slightly over the course of the field period.

An analysis of undecided voters suggest that those in that category who do vote will divide fairly evenly between the two major party candidates. When undecided voters are allocated, the Center’s best estimate of the probable outcome of the popular vote is Bush 49%, Gore 47%, Nader 4% and less than 1% Buchanan. But the contest is so close that we cannot confidently predict the winner of the popular vote.

George W. Bush continues to hold a slim edge over Al Gore in the final days of Campaign 2000. A Pew Research Center poll of 1,307 likely voters conducted November 1-4 finds 46% favoring Bush, 43% Gore, with 3% for Ralph Nader and 1% for Pat Buchanan. These results are almost identical to those of a Center survey conducted October 25-29, which gave the Texas governor his first lead in Pew’s polls since the summer.

The current survey continues to find a significant potential for last-minute changes in opinion. As many now are undecided or say they might change their minds as was the case a week ago — 8% are undecided, 7% say they could still vote for Bush, and 8% say that about Gore among registered voters. This is comparable to findings of a Center survey on the final weekend of the 1996 race between Bill Clinton and Bob Dole. That contest was not nearly as close, but as many as 11% of voters made up their minds in the final days of the 1996 contest, according to Voter News Service exit polls. Late deciders broke for Dole, giving Clinton a narrower victory than anticipated.

There is no indication that revelations of Bush’s 1976 arrest for driving under the influence have had a material impact on voter attitudes. Although 79% of those interviewed Nov. 3-4 say they followed news stories about this to at least some extent, the vast majority (88%) say this does not raise serious doubts in their minds about voting for Bush. Just 11% say the story has led them to question voting for Bush. However as many, 12%, say the same about Gore, based on charges that the Democrats were behind the release of the story. Further, the poll found no meaningful difference in levels of candidate support before and after these disclosures.

As in previous polls Bush’s supporters are more enthusiastic than those backing Gore. In fact, Bush registers the highest percentage of strong supporters in the final weekend of the campaign as any candidate since Ronald Reagan in 1984. While this may augur better for Republican than Democratic turnout, the survey finds the “ground war” is deadlocked. One-in-four voters say they have been urged to vote by a campaign — 13% reported they were contacted by Bush or the Republicans, and 12% say Gore or the Democrats had been in touch.

Independents Dead Even

The new survey underscores several patterns that have been evident throughout the final month of the campaign, including a large Bush edge among men — especially men under 50 — and divisions among independent voters.

Among likely voters, Bush enjoys a 12-point lead among men (50%-38%), twice the lead Gore holds among women (48%-42%). (See table on page 4.) Bush holds an especially strong 16-point lead among men under 50 (51%-35%), and a narrower seven-point lead among men 50 or older (48%-41%). In contrast, Gore has only a three-point lead among women under 50 (47%-44%) and an eight-point lead among women 50 and older (48%-40%).

Meanwhile, independents remain split, with 42% favoring Gore and 41% supporting Bush. Bush holds a slight edge among the party faithful, with 89% of rank-and-file Republicans supporting the GOP nominee, compared to 83% of Democrats who support Gore.

Congressional Race Unchanged

Democrats continue to lead in the generic congressional ballot, 47%-43%. The trend in the congressional race has remained stable all year. In early October and July, Democrats led by the same four-point margin. In February, Democrats held a 47%-44% advantage.

Overall, satisfaction with congressional incumbents remains high. Nearly six-in-ten voters (59%) say they would like to see their own representative reelected, while 16% would not. In general, national issues and the tight battle for control of Congress have not weighed heavily on voters. In the Pew survey in early October, 42% said state and local issues would make the biggest difference in their congressional vote, against 21% who cited national issues.

With less than a week to go before the presidential election, George W. Bush’s advantage with the voters on personal qualities is now trumping Al Gore’s edge on the issues. A steadily growing plurality has come to see the GOP candidate as more likable, more honest, more able to get things done, and fewer worry about his qualifications for the presidency. At the same time, voters increasingly see the vice president as a typical politician, and his personality is cited as a reason for opposing his candidacy by a greater percentage than did so before the presidential debates. Gore is confronting these adverse trends despite the fact that voters have more confidence in him than his opponent to deal with health care, Social Security and the economy.

Consequently, while the horse race remains close, for the first time since July Bush has a slight edge over Gore among likely voters, 47% to 43% in the latest Pew Research Center survey. The Pew poll, the third survey since early October, was conducted among 1,963 adults, including 1,508 registered voters (1,062 likely voters), from Oct. 25-29.

While there is nothing conclusive about Bush’s small margin given the fluidity of voter attitudes in this cycle, it does suggest that voters may be drifting in a Republican direction as Election Day approaches. The favorable trend for Bush is most apparent over the past month among seniors, white Catholics, middle- to high-income voters and those in union households.

Voters mostly cite issue positions as reasons for backing their choice for president, but personal qualities — including leadership and experience — are the predominant basis for not supporting a candidate. Personal questions are also offered up frequently by swing voters as reasons for their uncertainty. Those who have not yet firmly decided how they will vote most often voice questions about Bush’s qualifications for the presidency. Doubts about Gore are broader: swing voters worry about his personality, truthfulness, and ties with President Clinton. In that regard, when asked directly, a 55% majority of swing voters say they prefer that Gore be less, not more, like Clinton.

There are signs that voters are slowly coming to terms with the decision they will make on Nov. 7. The poll finds fewer voters on the fence: The percentage of voters who still may switch their vote has inched down from 17% in early October and 16% in mid-October to 13% in the current poll. The percentage of undecideds, the other component of the swing vote category, remains stable at 7% — about what it was four years ago at this time.

Although a majority of voters say they are very interested in voting in the upcoming election because of the tightness of the race, prospects are dim that voter turnout will significantly exceed the level of four years ago. Measures of voter engagement in the current campaign suggest that electoral participation will probably mirror 1996 (when 49% of age-eligible citizens voted) rather than 1992 (when 55% voted). For now, the likely voter pool continues to look only somewhat more Republican than all registered voters. This is similar to 1988 when likely voters backed Bush over Dukakis by only a few points more than did all potential voters. In 1996 likely voters skewed even more Republican than did all registered voters.

The prospect of an even larger GOP turnout advantage, however, is seen in the fact that Gore voters are somewhat less enthusiastic about going out to vote for their candidate than are Bush backers. Underscoring this enthusiasm gap, fully 37% of Democrats say they don’t see a big difference between Gore and Bush on the issues, compared to just 26% of Republicans. Some Democrats are demoralized. The percentage of Democrats expecting a Bush win has grown from 13% in early October to 25% in the current survey. But Clinton still has the potential to energize the Democratic base — unlike swing voters, a plurality of Democrats say they wish Gore were more like the president.

Abortion never emerged as a major topic in the campaign, but it still is very much on the minds of some voters. Surprisingly, those who say issues are motivating them to support Gore and Bush rate abortion almost as high as more widely-debated issues like Social Security and health care. Bush supporters who set issues as a priority cite abortion as often as taxes when giving reasons for backing the Texas governor. And while it ranks below education and Social Security for Gore’s issue-oriented supporters, it is on par with health care as a reason to support Gore.

Bush’s Intensity Edge

While the race is still extremely tight, Bush continues to enjoy stronger backing from his supporters than does Gore from his. Nearly two-thirds (64%) of Bush voters say they strongly support the Texas governor. Gore gets strong support from 55% of those who say they will vote for him. In addition, Bush voters are somewhat less likely than Gore voters to say they might change their minds and vote for the other candidate. Only 9% of Bush supporters say there’s a chance they might vote for Gore, compared to 14% of Gore voters who say there’s a chance they might for Bush.

Overall, the pool of swing voters has diminished slightly in recent weeks. One-in-five voters (20%) now fall into this group, down from 23% in mid-October and 25% earlier in this month.

Among registered voters who currently support the Texas governor, fully 75% fall into the likely voter category; 68% of Gore’s voters are classified as likely to vote. This accounts for the fact that Bush’s lead over Gore widens when the pool of voters is narrowed from all registered voters to the 50% of the electorate who are most likely to turn out.

In recent weeks, Bush has solidified his lead with men. The gender gap is firmly in place as the candidates head into the final week of campaigning. Men favor Bush by a margin of 49%-39%, while women favor Gore 48%-40%. And Bush has made steady gains among older voters over the past month. Seniors now divide evenly between Gore and Bush (44%-45%). In early October, they favored Gore by a wide margin (50%-38%). Most of the erosion in Gore’s support has come from older men, who narrowly favored the vice president earlier this month and now lean toward Bush. Young voters continue to vacillate between the two major party candidates — favoring Gore by a wide margin in mid-October and now dividing evenly between the two, with 8% supporting Ralph Nader.

Gore has lost some ground among union voters in recent weeks. Though he still leads Bush among voters in union households, his advantage has narrowed from 31 points in mid-October to 12 points now. Voters from non-union households continue to narrowly favor Bush. Independents remain split — 40% preferring Bush and 36% choosing Gore. Swing voters prefer Gore over Bush (30%-20%).