Ruy Teixeira
Ruy Teixeira

(covering polls and related articles from the weeks of Oct. 25—Nov. 7)

In this edition of Public Opinion Watch:

• A Tour of the 2004 Exit Poll: What It Says and What It Doesn’t
• Lessons of the 2004 Election

A Tour of the 2004 Exit Poll: What It Says and What It Doesn’t

Edison Media Research/Mitofsky International exit poll of 13,360 voters for National Election Pool, released Nov. 2 (conducted Nov. 2)

Greenberg Quinlan Rosner poll of 2,000 voters for Democracy Corps/Institute for America’s Future, released Nov. 5 (conducted Nov. 2–3)

Here are some observations on the 2004 exit poll data, based on the latest version of the data available. There is much to be explained and understood about these data and certainly legitimate questions can be raised about some of the findings. But the first task is simply to clarify what the poll actually says and does not say, because there is considerable confusion about this.

The figures used here are not the final figures, but, based on my experience in previous election cycles, they probably vary only slightly from the final numbers available when the exit poll authorities (in this case, the National Election Poll [NEP]) release a cleaned-up dataset in a couple of months with final weights.

All 2004 figures discussed here refer to the NEP exit poll, not the Los Angeles Times exit poll, since the NEP poll is both substantially larger and far and away the most widely used and cited. All 2000 figures refer to the 2000 exit poll by Voter News Service (VNS), the NEP’s predecessor.

1. Gender. According to the NEP poll, Bush carried men by 11 points (55 percent to 44 percent), exactly the same margin he had in 2000 when he carried men by 53 percent to 42 percent. Among women, however, Kerry’s margin was only three points (51 percent to 48 percent), down from the 11-point margin Gore had in 2000 (54 percent to 43 percent). No matter how you measure the gender gap (add the margins and divide by two or simply subtract Democratic male support from Democratic female support), this means a substantial compression of the gender gap (from 11 or 12 points, down to seven) and it is entirely due to the Democrats’ reduced strength among women.

2. Race. According to the NEP poll, 23 percent of voters this year were minorities, up from 19 percent in 2000, indicating the continued rapid expansion of the minority electorate.

The NEP poll says, however, that Bush widened his margin among white voters—still 77 percent of voters—to seventeen points (58 percent to 41 percent), up from a 12-point margin (54 percent to 42 percent) in 2000. And among Hispanics, now 8 percent of voters, the poll indicates a Kerry margin of only nine points (53 percent to 44 percent), a dramatic compression from Gore’s 27 point margin (62 percent to 35 percent) among the same group in 2000.

However, there is some dispute about whether the compression of the Democratic margin was as severe as indicated by this poll. An exit poll of Hispanics only by the William c= Velásquez Institute of San Antonio, which sampled 54 counties in the 14 states with the largest number of Latino registered voters, had 68 percent voting for Kerry and only 31 percent voting for Bush.

To further sow confusion, the NEP data on Hispanics are now being reported in two different ways—as above, at 8 percent of voters and 53 percent to 44 percent Democratic support, and at 6 percent of voters and 56 percent to 43 percent Democratic support in Sunday’s New York Times. How did Hispanics suddenly get demoted to 6 percent of voters? The answer is complicated, but here it is: the Times, for purposes of its historical chart, uses a single race question to capture Hispanics, as opposed to a race question plus another question on whether the respondent is of Hispanic descent or not, which were included on both the 2000 and 2004 exit polls and are now used by CNN and practically everyone else. The Times’ reasoning for not using this new (and better) two-question measure of Hispanic respondents is that the historical Hispanic data in their chart will at least all be measured in the same way.

I certainly see the point in apples-to-apples comparisons. On the other hand, since the new measure is undoubtedly a better one and we don’t really believe the Hispanic proportion of voters in 2000 was only 4 percent (as the single question Hispanic series also indicates) and just 6 percent this year, it would be better, I think—as well as less confusing—for the New York Times to go with the data that is the best and simply acknowledge a discontinuity in the exit poll on Hispanics between 1996 and 2000.

People will, after all, pay the closest attention to the figures—both support rates and proportion of voters—from this year and, secondarily, from 2000. Given that the figures from the two-question Hispanic measure for those two years are (a) better and (b) comparable with one another, it strikes me as a good idea to feature those data rather than the misleading single-question Hispanic data. Again, the discontinuity can then be footnoted for those that get into the data that far, but, with this approach, the average Times reader would be provided with the most accurate measure of Hispanic turnout and presidential support.

The data on blacks are much more straight-forward. Among blacks, Kerry had an 88 percent to 11 percent margin, down only slightly from the 90 percent to 9 percent margin for Gore in 2000. In fact, except for 2000 and Mondale’s 1984 campaign, Kerry’s margin among blacks is the highest obtained by a Democratic candidate since the exit polls started in 1976.

3. Whites by gender. Democrats’ falloff among whites appears to have been concentrated almost entirely among white women, rather than white men. This year, Bush carried white men by 25 points (62 percent to 37 percent), only a point more than his 24 point margin in 2000 (60 percent to 36 percent). In contrast, he carried white women by 11 points (55 percent to 44 percent), a big improvement over the single point (49 percent to 48 percent) by which he carried this group in 2000.

4. Education. Democrats’ slippage by education group was concentrated entirely among the non-college-educated. Kerry split the college-educated as a whole evenly with Bush, just as Gore did in 2000, and actually carried those with a postgraduate education by 11 points (55 percent to 44 percent).

But while Gore lost the non-college educated as a whole by just two points (49 percent to 47 percent), Kerry lost them by six points (53 percent to 47 percent), including an eight-point deficit among those with some college (up from a six-point disadvantage in 2000) and a five-point deficit among those with just a high school diploma (up from just a single point disadvantage in 2000). Most startlingly, Kerry only carried high school dropouts by one point (50 percent to 49 percent), while Gore had carried the same group by 20 points.

Given that Bush’s increased margin came entirely from the non-college educated and given the increase in Bush’s margin among white voters, we would expect that Bush’s performance among white working class voters must have improved substantially. This cannot be estimated directly from the NEP poll because they haven’t yet released that level of detail on their data. However, the Institute for America’s Future and Democracy Corps conducted an extensive (2000 interviews) post-election survey and these data indicate that Bush won white working class voters by about 24 points. That compares to a 19-point margin in Democracy Corps’ 2000 post-election survey and a 17-point margin in the 2000 VNS exit poll.

Arguably, that’s the story of the election right there. An additional wrinkle on the white working class vote is that this falloff was likely concentrated among white working class women, not men, judging from the figures cited above on Bush’s big gains among white women, but no change among white men (however, this is an inference from the pattern of the data; no direct evidence on white working class women vs. men is available from the NEP or Democracy Corps surveys).

5. Income. It is fascinating to note that Kerry actually improved over Gore among income groups under $30,000: 63 percent to 36 percent versus 57 percent to 37 percent among those with less than $15,000 and 57 percent to 42 percent versus 54 percent to 41 percent among those between $15,000 and $30,000. He did about the same as Gore among the $30,000 to $50,000 group (50 percent to 49 percent versus 49 percent to 48 percent). But he lost considerable ground among those over $50,000, losing 56 percent to 43 percent versus 51 percent to 46 percent among the $50,000 to $75,000 group; 55 percent to 45 percent versus 52 percent to 45 percent among the $75,000 to $100,000 group and 58 percent to 41 percent versus 54 percent to 43 percent among those over $100,000.

6. Marriage. The “marriage gap” grew slightly in 2004. This was because, while Kerry’s margin among unmarried voters stayed about the same as Gore’s in 2000 (58 percent to 40 percent versus 57 percent to 38 percent), Bush’s margin among married voters expanded from nine to 15 points (57 percent to 42 percent versus 53 percent to 44 percent). This increased the marriage gap, depending on how you measure it, from 13 to 14 points in 2000 to about 17 points this year.

But Bush’s margin among those who are married and have children expanded more modestly, from 56 percent to 41 percent in 2000 to 59 percent to 40 percent this year.

Data available from Democracy Corps’ post-election survey make it possible to compare white married voters by gender with their counterparts in 2000. This comparison shows Bush’s margin among white married men staying about the same across elections and actually shrinking a bit among white unmarried men. But among white married women, his margin increases from nine to 18 points and, among white unmarried women, he actually achieves a tie, compared to a 15-point deficit in 2000.

7. Age. Kerry did very well with young voters this year, winning them 54 percent to 45 percent, compared to a narrow 48 percent to 46 percent margin for Gore in 2000. On the other hand, Kerry lost seniors by 52 percent to 47 percent, while Gore won them by 50 percent to 47 percent.

A few more words on the youth vote. This marks the fourth straight presidential election where Democrats have won the youth vote. It is also, of those four elections, the one where youths’ Democratic support was most out-of-line with the rest of population. In 2000, youth were only two points more Democratic than all voters; in 1996, they were 11 points more Democratic than all voters; and in 1992, they were four points more Democratic than all voters. But in this election, youth were 12 points more Democratic than all voters (+9 Democratic among youth versus -3 among all voters).

In this election, youth were about 17 percent of voters. That’s the same as the exit poll figure for 2000. Does this mean youth turnout didn’t go up? Not at all. Even assuming the exit poll figures are correct (and personally I prefer the Census voter supplement data for looking at the demographic composition of the voting pool and assessing turnout trends), they merely mean youth turnout didn’t go up any more than other groups in the electorate. In other words, youth turnout went up, but probably only three to four points, about the national average.

8. Religion and religious observance. Perhaps no feature of the 2004 election has received more attention than the allegedly central role of evangelical Christians and their high turnout in Bush’s victory.

But the evidence that evangelicals were so very, very important (as opposed to merely important, which seems reasonable) is shockingly thin. Perhaps the main piece of evidence for this claim is that 23 percent of voters in the NEP exit poll were white “born-again or evangelical” Christians, who supported George Bush, 78 percent to 21 percent.

That is indeed impressive. The trouble is, we have no idea how that compares to 2000, since the exit polls didn’t ask the same question last time. Instead they asked a very different question about being part of the “religious right,” which categorized 14 percent of voters as part of the white religious right. Clearly, to conclude from these two different questions that white evangelical turnout increased from 14 percent to 23 percent from 2000 to 2004 is inappropriate.

Thomas Edsall in the Washington Post on Monday nicely summarized the correct way to look at these data:

Exit polls do not permit a direct comparison of how many evangelical and born-again Americans voted in 2000 and 2004 because the way pollsters identified these voters changed. Four years ago voters leaving polls were asked: “Do you consider yourself part of the conservative Christian political movement, also known as the religious right?” In 2004, the question was changed to: “Would you describe yourself as a born-again or evangelical Christian?”

Fourteen percent answered “yes” in 2000 and 23 percent did so in 2004, but polling specialists said the 2004 wording virtually assures more affirmative answers.

Admirably clear. Okay, on to the next piece of evidence. This is the finding that 22 percent of voters—more than for any other issue—said “moral values” were the most important to their vote and these voters supported Bush 80 percent to 18 percent.

Again, pretty impressive. But again, we have no idea how this compares to 2000, when voters were not given a “moral values” or any other “values” choice but instead a list of actual issues (taxes, world affairs, Medicare/prescription drugs, health care, economy/jobs, education and social security). As Gary Langer, ABC News Polling director points out:

[T]he exit poll . . . asked voters what was the most important issue in their decision: taxes, education, Iraq, terrorism, economy/jobs, moral values or health care. Six of these are concrete, specific issues. The seventh, moral values, is not, and its presence on the list produced a misleading result.

How do we know? Pre-election polls consistently found that voters were most concerned about three issues: Iraq, the economy and terrorism. When telephone surveys asked an open-ended issues question (impossible on an exit poll), answers that could sensibly be categorized as moral values were in the low single digits. In the exit poll, they drew 22 percent.

Okay, next. The exit poll asks a question on the frequency of religious service attendance. And this question does show those who say they attend services more than weekly increasing slightly from 14 percent to 16 percent. On the other hand, the poll also shows those who say they attend weekly decreasing slightly from 28 percent to 26 percent, so the most observant segment of voters, those who attend services weekly or more, remained steady at 42 percent of voters. This hardly seems consistent with a wave of evangelical turnout.

Moreover, as Alan Abramowitz points out:

[B]etween 2000 and 2004, President Bush’s largest gains occurred among less religious voters, not among more religious voters. Among those attending services more than weekly and those attending every week, support for Bush rose by 1 percent, from 63 percent in 2000 to 64 percent in 2004. However, among those attending services a few times a month, support for Bush rose by 4 points, from 46 percent to 50 percent, among those attending only a few times a year, support for Bush rose by 3 points, from 42 percent to 45 percent, and among those never attending services, support for Bush rose by 4 points, from 32 percent to 36 percent.

Bottom line: the President made gains across the board among voters, regardless of their degree of religious commitment but he made his largest gains among less religious voters.

None of this seems consistent with the idea that evangelical turnout and intense support from the most religious Americans put Bush over the top in 2004.

On to religion itself, independent of level of observance. The NEP exit poll shows that Protestants supported Bush by 19 points (59 percent to 40 percent), compared to 14 points in 2000 (56 percent to 42 percent). Among Catholics, there was an even larger swing in Bush’s favor, going from a 50 percent to 47 percent Democratic advantage in 2000 to a 52 percent to 47 percent Republican advantage this year.

Unfortunately, we don’t know how white Catholics’ preferences changed this year, because NEP has not released the data. In 2000, Bush carried white Catholics by seven points (52 percent to 45 percent). It seems reasonable to assume that Bush carried this group by a significantly wider margin this year.

In addition, while Jews are a small proportion of voters (3 percent of voters this year) their margin for the Democrats shrank from 60 points in 2000 (79 percent to 19 percent) to 49 points (74 percent to 25 percent) this year.

On the other hand, among those who profess some other religion besides Christianity or Judaism, Democrats’ margin of support rose from 34 points in 2000 to 51 points this year. And among those who say they have no religion, Democrats’ margin of support rose from 31 points to 36 points across the two elections.

That’s all for now. But there’s much more to be covered and I hope to continue our tour soon. In the meantime, what we’ve covered so far should provide some food for thought.

Lessons of the 2004 Election

Well, a second term for George W. Bush it is. Not a smashing victory for him: he took the popular vote by a 51 percent to 48 percent margin and gained two new states (Iowa and New Mexico) by 50 percent to 49 percent margins, while losing one old state (New Hampshire) by a 50 percent to 49 percent margin.

What are the lessons Democrats can draw from Bush’s victory? How was Bush able to hang onto power despite the poor economy, Iraq, the health care crisis, and so on?

1. The limits of mobilization. Democrats put great stock in mobilization and the ground game. And Kerry did do better in many areas where there was intensive mobilization. For example, in Ohio, Kerry carried Franklin County (Columbus) by 41,000 votes, compared to Gore’s margin of just 4,000 last election, and carried Cuyahoga county (Cleveland) by 218,000 voters, compared to Gore’s margin of 166,000 in 2000. But these gains were mostly canceled out by Republican mobilization in conservative rural and exurban areas, so Ohio, in the end, was only slightly closer (2.5 percentage points) than it was in 2000 (3.5 points).

As another example, the exit polls indicate that 23 percent of voters this year were minorities, up from 19 percent in 2000. So Democrats were reasonably successful in getting minorities to the polls. But these data indicate that Hispanics only supported Kerry 53 percent to 44 percent, a dramatic compression from Gore’s 62 percent to 35 percent margin among the same group in 2000. (Note, however, that there is some controversy about whether Democratic support among Hispanics was really as low as these data indicate—see the discussion above.) And—much more consequential for the election—the exit polls say that Bush widened his margin among white voters to 17 points (58 percent to 41 percent), up from a 12-point margin (54 percent to 42 percent) in 2000. Weakened support among Hispanics and, especially, a bigger deficit among whites (still 77 percent of voters) was more than enough to cancel out the effect of more minority voters going to the polls.

2. The limits of anti-Bushism. Kerry had much to say that was very critical of Bush and certainly there was much to criticize in the areas of the economy, tax cuts, Iraq, health care, energy policy and so on. These criticisms were directed at genuine weak points in Bush’s record and there is good evidence that most voters shared at least some of these criticisms. Bush was not, and is not, a particularly popular incumbent, so attacking his record was an inevitable and important part of Kerry’s campaign.

The problem, however, was that Kerry never managed to convince many of the same voters who shared his criticisms of the Bush administration that he could and would do a better job in the areas he criticized. To cite just one example from the exit poll, voters were asked if they trusted Bush to handle the economy: 51 percent said no and 49 percent said yes. Not so good for an incumbent. But voters rated Kerry even worse: 53 percent said they didn’t trust him to handle the economy, compared to just 45 percent who said they did.

And all through the campaign, up to the very end, there was abundant evidence that voters did not think Kerry had a clear plan for Iraq or, for that matter, for the country in general. His campaign was notable for lacking signature themes and proposals that typical voters could easily grasp and identify with. Does anyone seriously believe that many voters knew or understood Kerry’s plan for Iraq? For health care? For the economy? How many voters knew the one or two thematic phrases (if they existed) that summarized what John Kerry stood for?

Let’s face it: not many. I believed this was a serious problem for Kerry all through the campaign, but thought, toward the end, that voters might be interested enough in getting rid of Bush that they would cut him slack on these specifics. That did not turn out to be the case.

3. The need for white working class support. The last three elections (2000, 2002, and 2004) have all had strong “culture war” components that have severely depressed white working class support for Democrats. Recall that Bill Clinton actually carried the white working class (whites without a four-year college degree) by a point in both his election bids. But in 2000, Al Gore lost these voters by 17 points; in 2002, Democratic congressional candidates lost this group by 18 points, and this year, the situation appears to have worsened further. That is implied by the exit poll finding, cited above, that Democrats lost whites as a whole by five points more than in 2000 and another exit poll finding that Democrats’ slippage by education group was concentrated entirely among the non-college educated. In addition, Democracy Corps’ post-election survey directly indicates serious Democratic slippage among white working class voters (see discussion above).

The fact of the matter is that Democrats cannot win when they do so badly among this very large constituency. John Judis and I always believed that the trends we described in our book The Emerging Democratic Majority could underpin a majority coalition given reasonable (not majoritarian, but competitive) performance among white working class voters. But this does not qualify as reasonable performance.

Democrats’ difficulties with this group surely have a great deal to do with these voters’ sense of cultural alienation from the national Democratic party and its relatively cosmopolitan values concerning religion, family, guns and other social institutions/practices. Even the war on terror has increasingly become more of a cultural issue linked to patriotism than a true foreign policy issue for many of these voters.

Given this sense of cultural alienation, it must be questioned whether candidates like Gore or Kerry can ever really be viable with these voters. Democrats may have to choose candidates in the future who do not so easily evoke this sense of cultural alienation and who can connect in a genuine fashion with these voters. I come to this conclusion reluctantly because I had thought that an effective campaign could overcome this obstacle by, in effect, using wedge Democratic issues such as health care and jobs to build support among this group. But the messenger appears to matter a great deal, just as having a message does (see point number two, above). The Democrats in the future will have to pay attention to both, I think.

Ruy Teixeira is a joint fellow at the Center for American Progress and The Century Foundation.

The positions of American Progress, and our policy experts, are independent, and the findings and conclusions presented are those of American Progress alone. A full list of supporters is available here. American Progress would like to acknowledge the many generous supporters who make our work possible.


Ruy Teixeira

Former Senior Fellow