(covering polls and related articles from the week of February 7–13, 2004)
In this edition of Public Opinion Watch:
• It’s the White Working Class, Stupid
• Social Security: That Dog Won’t Hunt Department
• Will the Real Party Preferences of the American Electorate Please Stand Up?
It’s the White Working Class, Stupid
There are many theories about what drove the 2004 election results, and some of the more fanciful ideas (exurbs, fast-growing counties, evangelicals, Hispanics, values voters) have been critiqued in various editions of Public Opinion Watch. Now, with the release of the raw data from the 2004 National Election Pool (NEP) exit poll, it is possible to do some closer analysis of trends that really were of high salience. One such trend was the movement of white working class voters away from the Democratic ticket.
Here some findings from an initial pass through the NEP national data:
- In 2000, Gore lost white, working-class (defined as whites with less than a four-year college degree) voters by 17 points; this year, Kerry lost them by 23 points, a swing of 6 points against the Democrats. In contrast, Gore lost college-educated whites by 9 points and Kerry lost them by 10 points—not much change.
Therefore white, working-class voters were responsible for almost all of Bush’s increased margin among whites as a whole (which went from 12 to 17 points). And Bush’s increased margin among whites, of course, was primarily responsible for his re-election.
- Almost all of the white, working-class movement toward Bush was among white, working-class women, rather than white, working-class men. Bush won white, working-class men by almost identical margins in the two elections (29 points in 2000 and by 30 points in 2004). But he substantially widened his margin among white, working-class women, going from a 7-point lead in 2000 to an 18-point lead in 2004. That 11-point swing against the Democrats among white, working-class women is arguably the most important single fact about the 2004 election.
- Looking at married versus single white, working-class women, both groups appear to have swung substantially against the Democrats. Single, white, working-class women (38 percent of white, working-class women) went Democratic by 15 points in 2000, but only by 2 points in 2004. Married, white, working-class women (62 percent of white, working-class women) gave Bush a 15-point margin in 2000 and more than doubled that margin, to 31 points, in 2004. Since married, white, working-class women are the bulk of this group and had a slightly larger pro-Republican shift, they are responsible for most of the shift toward Bush among white, working-class women, but their single counterparts clearly made an important contribution as well.
- But why did these shifts against the Democrats among the white, working class occur? That’s a topic that deserves a lengthy discussion; here are some data to ponder from the NEP poll:
Among white, working-class voters, 66 percent said that they trusted Bush to handle terrorism, compared to just 35 percent who said the same about Kerry. That’s pretty bad, but check this out: 55 percent of these voters said that they trusted Bush to handle the economy and only 39 percent said the same about Kerry. Guess that Kerry message about the economy didn’t quite get through to the white working class!
It’s also interesting to note that there wasn’t much of a difference in these sentiments among men and women in the white working class: 55 percent of white, working-class women said that they trusted Bush to handle the economy and 40 percent said that they trusted Kerry, while 56 percent of white, working-class men said that they trusted Bush on the economy and 37 percent said that they trusted Kerry.
That’s something to ponder. Not only were white, working-class women alarmed about terrorism, but they were also, in contrast to previous elections, no more likely to find the Democratic economic message more compelling than did their male counterparts.
Social Security: That Dog Won’t Hunt Department
ICR surveys of 1,236 and 1,231 adults for Washington Post/Kaiser Family Foundation/Harvard University, released February 10, 2005 (conducted February 3-6 and February 4–6, 2005)
On February 6, Nick Confessore argued in the New York Times that “Going for Broke May Break Bush” and the next day Ron Brownstein commented that “Bush’s Social Security Equation Comes Up Short on Money, Trust.” And in last Friday’s New York Times, the bloody implications of indexing Social Security benefits to prices, instead of wages, were copiously detailed.
Things just aren’t coming up roses for the president as he continues to stump for his Social Security plan. It looks like it’s going to be harder than Bush anticipated to move public opinion in his direction. And that’s for a very good reason: the underlying structure of public opinion is hostile to Bush’s approach.
Consider these results from a just-released Washington Post/Kaiser Family Foundation/Harvard survey of public opinion on Social Security issues:
- Only about a quarter (27 percent) of respondents say that Social Security is in a crisis, 46 percent say that Social Security has major problems but is not in a crisis, and the rest say that the program has minor problems or no problems. Note that the “crisis” figure in this question is actually substantially lower than it was in the late 1990s, when it reached as high as 36 percent.
- The only ways of fixing Social Security’s future financial problems that garner majority support are two alternate wordings about benefit cuts for the wealthy: “reducing the rate of growth in benefits for wealthy retirees only” (60 percent) and “cutting guaranteed benefits for wealthy retirees only” (54 percent). But that’s not what Bush proposes to do—he proposes to cut everybody’s benefits, which gets quite a different reception: “reducing the rate of growth in benefits for future retirees” gets only 30 percent support and “cutting guaranteed benefits for future retirees” receives just 13 percent support.
- No matter whether the accounts in the president’s plan are referred to as “private” or “personal,” they get about the same middling level of support in the abstract (that is, without any tradeoffs or costs). But that 54 percent to 57 percent majority support drops to a dead-even 46 percent to 46 percent split once Bush’s name is associated with the plan and drops much further when some of the plan’s tradeoffs and costs are mentioned.
This can be seen in two ways. First, follow-ups to the general question of support for private/personal accounts show sharp drops in support for these accounts when costs/tradeoffs are mentioned. Specifically, support drops to 34 percent when it is pointed out that those who open accounts but make poor investment decisions would wind up with lower benefits than under the current system; to 29 percent if it is true that the plan including private/personal accounts would not by itself solve Social Security’s financial problems; and to 22 percent if the government would have to borrow $700 billion or more to set up these accounts.
Second, a question that mentions both the stock market option for Social Security contributions and changes in guaranteed benefits yields majority opposition to such an approach. People are opposed, by 52 percent to 43 percent, if the change in guaranteed benefits is referred to as “reducing the rate of growth in benefits,” and are opposed by an overwhelming 66 percent to 30 percent, if the change in benefits is simply referred to as “cutting guaranteed benefits.”
- Finally, just 9 percent believe that creating private/personal accounts would, by itself, solve Social Security’s financial problems. And slightly more people believe young people will wind up with less money under a personal accounts system (35 percent) than believe that they will wind up with more money if these accounts were available (33 percent). Another 24 percent believe that young people will do about the same under a personal accounts system as under the current system. That means there’s a 59 percent to 35 percent majority against the idea young people will gain with a personal accounts system.
As a certain ex-president might have put it: that dog just won’t hunt. We’ll see how long it will take for Bush to face this bitter truth.
Will the Real Party Preferences of the American Electorate Please Stand Up?
Gallup poll of 1,010 adults for CNN/USA Today, released February 8, 2005 (conducted February 4–6, 2005)
Gallup poll of 1,008 adults, released February 11, 2005 (conducted February 7–10, 2005)
Gallup’s February 4–6 approval rating for Bush (57 percent) seemed to many, including myself, as being distinctly on the high side and, teamed as it was with a lopsidedly Republican sample (a nine-point Republican advantage in party identification, as Steve Soto was able to establish), seemed distinctly lacking in credibility.
It’s noteworthy that no other recent poll seems to be able to come close to Gallup’s 57 percent rating in that poll. ABC News/Washington Post, for example, had Bush’s approval rating at 50 percent for January 26–31—but then, that was mostly before the Iraqi elections, so perhaps the Washington Post poll couldn’t capture that big post-Iraqi election surge toward Bush (Gallup’s own explanation for Bush’s high rating in their poll).
Newsweek, however, polled during February 3–4—after the Iraqi elections—and found only a 50 percent rating for Bush. But the Gallup poll was February 4–6 so perhaps this surge was late developing?
But, inconveniently for Gallup, Washington Post/Kaiser Family Foundation/Harvard University were in the field February 3–6 (see above) and they also found just a 50 percent approval rating for Bush.
And here’s the coup de grace: the new Fox News poll, not generally known for being unfriendly to the president and low-balling his approval rating, found his approval rating on February 8–9 to be only 51 percent. Moreover, Fox’s 51 percent rating is actually a point below their mid-January rating for Bush, while Gallup’s recent 57 percent rating is six points above their mid-January rating for the president.
Well, something’s happened here. But I don’t believe it’s among the general public= I think it’s more in Gallup’s polling and—as indicated above—we have some obvious suspects.
The role of sample party identification in driving these approval results is further suggested by considering the new Ipsos-Associated Press poll that has Bush’s approval rating at just 45 percent, in rather stark contrast to Gallup’s 57 percent rating in their poll with the nine-point Republican party identification advantage. The Ipsos poll also has a lop-sided party identification advantage—but this time for the Democrats (+12). The fact that the Ipsos party identification figure is for registered voters and the Gallup figure for all adults hardly seems adequate to account for this vast difference.
In truth, neither figure seems terribly credible and, therefore, both approval figures are probably outliers driven by the party identification composition of their samples. Certainly neither figure should be taken particularly seriously on its own, though you could average the two if you wish. In that case, you get a 51 percent approval rating for Bush, pretty much in line with other figures from recent polls.
What is to be done about these wacky partisan samples, which give such misleading pictures of current politics? Perhaps it’s time to revive “dynamic party identification weighting,” an idea whose time finally may have come. Arguably, this is the time to pursue such an innovation, away from all the passions induced by a political campaign. And, if pollsters did so, I think it would help smooth out poll results and avoid the fake surges that are starting to erode faith in the veracity of polling.
Of course, down at Gallup and many other polling headquarters as well, the view is probably that all is fine. I can assure them that all is not fine and it is time to trade in their stonewalling for a bit of listening and openness to change.
That may be awhile coming, judging from how Gallup has played their most recent poll, conducted February 7–10. In that poll, lo and behold, Bush’s 57 percent approval rating has nosedived to 49 percent, in the same neighborhood as other recent polls. My immediate guess, though Gallup did not publicly release these data, was that this lower rating for Bush was also accompanied by the disappearance of that big Republican Party identification advantage.
You’d think that their first poll’s heavily Republican sample and outlier status relative to other polls would have made the Gallup folks cautious in interpreting Bush’s approval rating drop in their second poll as representing a real political trend. Think again! Gallup, in fact, constructed a whole political story to fit their data, starting with their title, “From Public’s Perspective, Past Week Not Good One for Bush: Approval ratings drop,” and continuing thus:
“The new poll, conducted Feb. 7–10, shows his approval rating is back down to 49% as the news focus has shifted to his proposed federal budget and his plans for changing the Social Security system.”
Could be. But an alternative, more parsimonious hypothesis is this: Bush never had a popularity spike to begin with; little changed over the period covered by Gallup’s two polls and Gallup’s swing against Bush in their latest poll is an artifact of their changing sample composition, not the product of a real shift in public opinion.
That viewpoint receives strong support from data unearthed by Steve Soto on the party identification distribution of the later Gallup poll’s sample. It is +1 Democratic. So, from February 4–6 to February 7–10, the period covered by the two polls, Gallup claims there was a ten-point swing in party identification, from +9 Republican to +1 Democratic.
That strikes me—and most others outside of Gallup headquarters, I suspect—as pretty damn implausible. But then most of us don’t subscribe to the Gallup philosophy: reality’s job is to explain Gallup’s data, not the other way around.
Ruy Teixeira is a joint fellow at the Center for American Progress and The Century Foundation.