« So, Did Those Ads in the Battleground States Really Work? | Main | More on Yesterday's Polls »

A Bush Bump?

Two polls released today--Gallup and ABC News/Washington Post--give small leads to Bush over Kerry in presidential trial heat questions. The Gallup poll (using RVs and the Kerry-Bush not Kerry-Bush-Nader trial heat) shows Bush with a 4 point lead (50-46), while the ABC News poll gives Bush a 1 point lead (49-48). (Note that this latter result is not from a standard Kerry-Bush trial heat question, which ABC News chose not to ask, but rather from combining a Kerry-Bush-Nader trial heat question with a followup to Nader supporters/undecideds on who they would support if Nader doesn't run or isn't on the ballot. Guess they just wanted to be different.)

So: two polls, two RV leads, one taken April 16-18 (Gallup), the other taken April 15-18 (ABC News).

Here are other RV Kerry-Bush results for April:

Newsweek, April 8-9....................Kerry, 50-43
ARG, April 6-9.............................Kerry, 50-44
Gallup, April 5-8..........................Kerry, 48-45
Fox, April 6-7..............................Kerry, 44-43
CBS News, March 30-April 1.........Kerry, 48-43

Note that each of these polls has Kerry ahead. And note that there were no relevant polls conducted in the period from April 10 to April 15, the start date of the new ABC News poll. But that poll and the new Gallup poll do indicate that Bush has edged ahead, so a plausible theory is that Bush received some sort of a bump up in that period.

What might that have been? Given the timing, Bush's advertisements do not seem like the logical candidate for such a bump. A more plausible possibility is his televised speech/press conference on April 13, where he presented no clear plans on how to deal with the problems in Iraq, but did urge Americans to stay the course, be tough and so on. I thought at the time it was possible he would get some sort of small, short-lived rally effect from these posturings and that may have come to pass. Such an effect was likely aided and abetted by the very low profile of the Kerry campaign which provided Bush with a relatively clear field to push the electorate in his direction.

I also thought that "the really significant political development in the recent period is the undercutting of support for Bush's war in Iraq and for his handling of the war on terror", not the ups and downs of the horse race, and I continue to think that.

Indeed, there are plenty of findings in these new polls that indicate Bush's troubles in these areas are here to stay. The Gallup poll shows the public about split (52-46) about whether it was or was not worth going to war in Iraq. And the ABC News poll has a similar split (51-47) on whether the war with Iraq was worth fighting. That poll also shows a close split on a related question: whether the US did the right thing in going to war with Iraq or whether it was a mistake (52-46). Last April, this same question was 81-16.

People also feel, by 51-34, that the middle east is less stable, rather than more stable, as a result of the war. They think, by 57-37, that the US will not be able to establish a stable democracy in Iraq. They also believe, by 35-29, that the war in Iraq has left the US in a weaker not stronger, position in the world (last April, the public thought the war will make us stronger, by 52-12). And about two-thirds (65 percent) now say that, given the goals vs. costs of the war, the number of casualties we are sustaining is unacceptable.

The ABC News poll also indicates that Bush's approval ratings in a wide range of areas have improved only marginally in some areas, while continuing to slip in others. His ratings continue to be net negative on the economy, Iraq, social security, health insurance, taxes, creating jobs, the budget deficit, prescription drugs and even same-sex marriage. Nor are people more convinced the country is moving in the right direction; these numbers continue to be dismal with wrong track (57 percent) far outnumbering right direction (42 percent).

The really positive changes for Bush are in a series of questions asking people who they trust more, Bush or Kerry, on a range of issues. In every area, Kerry's advantages are smaller or disadvantages greater than than they were in ABC News' March 7 poll. For example, Kerry was preferred on the economy by 12 in March, now he and Bush are tied. Or Kerry was ahead by 20 on health insurance in March, now he is ahead by just 6. And so on. These are big changes on these preferences, not just from early March but from the end of March Battleground poll I discussed yesterday and even from the (notoriously pro-Bush) Fox poll 10 days ago.

That suggests, again, not the effect of Bush campaign advertisements, but rather a rally effect compression of Kerry's advantages over Bush (rally effects usually benefit presidents across unrelated areas). Therefore, Bush's improved showing in these areas is unlikely to stick around, given an adequate push-back by Kerry's campaign.

This is probably as good a time as any for the Kerry campaign to start that push-back, including especially defining Kerry positively for voters. Bush, as the data clearly show, has been massively undermined in his core area of strength, and, despite his much-vaunted advertisements and (probably more important) having the field to himself for six weeks, has Kerry breathing down his neck.

If the Kerry campaign can kick their game up a notch, they should really start to make the Bushies sweat.

Comments

The peculiar recent bump in Bush's numbers (as well as the strange stability before that, despite the bad national security news, could be explained not just by a rallying effect, but by another factor, as I argued in the thread below. When bad national security news arrives, national security as an issue simultaneously becomes more important to the voter, and Bush owns an edge on that issue, even if that edge is also diminishing.

Depending on the rate at which the edge declines and the importance of the issue increases, you may get the paradoxical effect of stability or even of an increase in the President's popularity.

If this is indeed what's going on, the point at which to expect the REAL decline in Bush's numbers is, again somewhat paradoxically, when national security recedes more into the background, and the economy, and Kerry's continued edge there, becomes once again the dominant issue in the public's mind (and Bush's edge on national security has been reduced to new lows).

Ruy, I hope you are emailing the Kerry campaign with your analysis. They need to hear from you (and anyone else they might listen to) that it is time to make themselves heard.

The Kerry campaign have proven themselves to be complacent, ineffectual idiots unless they have their backs to the freaking wall.

I knew I supported Howard Dean for a reason, GOD I am going to go bitch-slap everyone in Iowa if Kerry loses.

In recent polling showing Bush ahead, Gallup had 40% of its sample Republican voters as opposed to 34% Democrats. I would like to know what is the actual voter identification with a specific party because other earlier polls I've seen indicate that the party identification is either equal or a slight edge to the Democrats. If true that would mean that Gallup-CNN's results are skewered by their sample.

The issue with the likely overrepresentation of Republicans brings up a more general question about polling. Namely, when, if ever, is it appropriate to adjust the sample AFTER the poll is taken, because it can be demonstrated to be skewed in relevant ways?

Suppose, for example that a highly disproportionate number of men turned up in a poll sample, say, 55%, when we know that, say, exactly 50% of voters are in fact male. And suppose that, as we know is true, males disproportionately favor Bush. Wouldn't it be correct scientifically to adjust the poll results so that this overrepresentation of males is taken into account?

Now of course this would be a case in which we could know for hard fact what the underlying numbers of male and female voters are. The number of Republicans vs. the number of Democrats is a softer number, though over the course of many polls might be well enough established that we could detect obvious discrepancies in a given sample -- and a 6% edge to Republicans has got to fall into this category.

In general are adjustments for such obvious discrepancies simply not allowed in polling? How about for harder facts such as male vs. female? What are the permitted kinds of adjustments, if any? Given that the polling technique itself obviously introduces its own skew (such as selecting only people who have phones, don't filter out unknown callers via caller ID, are willing to answer a long series of questions, etc.), isn't it artificial to treat weightings after the fact as somehow less scientific?

This race is over. Kerry is toast because Ralph Nader is going to do it again. He's going to hand the victory to Bush because Dems are split on the Iraq war.

Kerry's position is responsible but too many Dems want the US out of Iraq and they want it out now. Nader is going to get enough votes from disaffected Dems and anti-war independents to drag down Kerry's numbers.

On the economy, we're starting to see a turn around that will also blunt Kerry's advantage there. The turn around may not result in much job growth but the voters will buy it.

Bush's numbers are stabilized and growing because deep down the American people seem to be supportive of the idea of pre-emptive war. Bush's stiff resolve is admired and the thought of dumping him in favor of someone as conflicted about the war as Kerry is not too appealing.

The future is grim.

Is the glass 1/2 empty or 1/2 full?

I am 1/2 empty guy.

To begin with Ruy's analysis is a non-sequitur. After proclaiming he predicted a slight bump due to the press conference, nothing he says afterwards supports that contention.

In fact, he contradicts himself and basically admits that the bush ads plus a somnambulent Kerry campaign account for the recent poll findings.

For weeks Ruy and others have urged Kerry to strick hard on foreign policy and defense. I agree


I read the same advice now.

Is Kerry listening?

My fear is that Kerry is rather lazy, laid back and doesn't get going until the gong gets tough.

Then it will be too late.

On the one hand there's a third poll (Investors Business Daily) taken last week with Bush ahead 46-42, and a 46-42 Bush lead in PA as well. On the other hand, maybe that's all the rally effect. And on the third hand, Kerry's about to do exactly what Ruy thinks he ought to do-- two new ads called "Risk" and "Commitment." (Note that these are not the fundraising ads running on Lifetime and Bravo, but new broadcast ads for swing states.) Nobody's ever shown Kerry lazy-- tactically misguided, perhaps, but not lazy. My fear is that the rally effect will never go away-- that the worse the Iraq/ terror news gets, the bigger the Bush bump will be, no matter how clear it seems (to us) that Bush stands to blame for some of the bad news; consider the career of Ariel Sharon, who (at least until recently) profited politically from Israeli casualties, since he was the guy who got tough on Arabs.

"If the Kerry campaign can kick their game up a notch," you say. I' beginning to doubt they can. Kerry still speaks in that annoying "public meeting oration" style. He comes off to me as pompous, bureaucratic, confused and arrogant. And I'm a deep-blue Dem. Why the hell isn't he working on his style?

Why no reference to the Zogby poll, taken over the time period the ABC and Gallup polls were taken, which showed Kerry with a three point lead in the head-to-head and tied in the three-way?

The Republican overrepresentation in the Gallup is also critical, as Alan noted.

My take on it gibes with the Investors' Daily poll. Kerry is losing support, mostly to undecided, while Bush is not gaining much. Could this be a replay of the 1980 race, albeit on a lesser scale?

The problem with Kerry lately is that he's been flip flopping on unimportant issues but nevertheless playing into Republican hands by doing so. First, there was the refusal to release all of his war records--which he's now doing. (This business about Purple Hearts is particularly infuriating given Bush's "record" in the National Guard).

Then there's Teresa Heinz Kerry's tax returns. The on-going refusal to release them--however justified--still gives Republicans an issue. And my guess is Kerry will back track on this one too.

I do agree with frankly0. But there is absolutely no reason for Bush's poll numbers to be better than Kerry's as Bush has all but dropped the ball in Iraq (remember, he is the War President). I beleive he's handing things over to the UN because he can then blame them for all that will go wrong after June 30, no matter that our troops are there. While Kerry may have lost a little ground, it should not have been enough to give Bush any type of lead.
We have to remember one thing: Bush's administration is riddled with scandal. A lot has been said about the Plame outing coming to a boil this summer. If so, Bush should be toast (to quote Cheney about Saddam). I can't believe the polls showing Bush with a lead. There;s just too much wrong with this country domestically for me to believe that over 50% of the public loves this guy. It can't be so.

Sadly, 50% of the country does love this guy, and the glass is half-empty.

The greater the violence in Iraq, the more the masses flock to Bush.

The CV had been that getting troops coming home in droves just in time for the election would clinch it for Bush, while a quagmire would put him in trouble. Actually, it is really the opposite -- on going violence all the way through the election means a Bush win, and only US victory -- soon -- thus allowing domestic issues back into the campaign, gives Kerry a shot.

On another note, some of these state by state polls are horrible. The Emerging Democratic Majority thesis always rested on the hypothesis that the growth of young, Democratic knowledge workers and the like would offset the the ongoing loss of uneducated, socially conservative Democratic workers. But look at those PA, WV, OH numbers -- we've completely lost Joe Sixpack -- he's a core Republican now -- nothing short of a Depression would drive him back to the Dems -- and there are still too many of him.

And another thing -- we're still underestimating the danger of Nader -- if he turns just one of Oregon, Minnesota, or Michigan -- and he most certainly can -- it's over.

Out of nowhere, across the blogosphere the DeanNuts have returned.

Without any understanding of the election calendar, the nature of continuous poll shifts in a race (it ALWAYS happens), they make the same statements over and over again:

Kerry stinks. He is a ______ .
His campaign stinks.
I can't stand him.
I told you so.

My guess is that school is out.

I generally think even looking at head to head match-ups in April is an irredeemable waste of time. Reagan and Clinton both trailed by wide margins at this point in their challenger years, which obviously meant nothing in the grand scheme.

Nonetheless, since everyone is so hepped up about these minor results, a few observations:

1) I can accept the idea that bad news overseas causes temporary rallying effects for incumbents, but I think recent develiopments in Iraq are clearly a negative for Bush in the long run. Carter got a huge boost in his numbers when the hostages were first taken -- and even a sympathy bounce after the disaster of Desert One -- but no one doubted on Election Day that the crisis had played a big role in his defeat. There's no way a fiasco like this past month in Iraq has been could enhance any incumbent's prospects.

2) And actually...I'm not convinced these numbers even represent a bump. As I recall, Gallup two weeks ago had Bush leading 51-47, so 51-46 doesn't spell progress to me. Gallup and ABC -- the two polls to which all are alluding -- have been consistent Bush-favorers from Inauguration Day, for the precise reason cited here: over-representation of Republicans. It's amazing how little attention Zogby's poll of the same period has received, seemingly because it doesn't reinforce the "amazing Bush rebound" theme the press likes.

3) The Investors Daily poll doesn't strike me as remotely bad news, simply because Bush's number is so low. Why don't more people understand that Bush, as incumbent, needs to hit 50 in these polls? Any poll where I see Bush in the 40s (ABC's included) I view as positive for Kerry.

4) In the long run, poll internals are far more important than match-ups. And here the news is very bad for Bush. He only hits 50% apporoval in Gallup (likely due to its skewed sample), and right track/wrong track continues to go heavily against him. The only possible hope I see for Bush in these areas is a noticeable (if ephemeral) improvement on the job front between now and September. That could bump both those negative numbers toward more neutral territory.

5) But, considering the general vibe of the election...I think Woodward's book is just the latest piece of evidence that Official Washington has turned somewhat against the administration. This is very ominous news for a group that for the most part was treated with deference for 3 years. I also think reports that people in the military and CIA are actively opposed to Bush's re-election should be viewed as huge harbingers. How in the world does a GOP administration get re-elected without those two groups solidly behind them?

6) As far as Kerry in all this...I was for Dean in the primaries, and I think the obvious voter decision to play "safe" this year may have been misguided (GOP voters in 1980 went the opposite way, and were richly rewarded). But I don't think in the end the opposition candidate makes that much difference. The Bush administration will be re-elected or not based on facts on the ground. Right now, they don't lean his way, and, given that, how often or well Kerry speaks is almost irrelevant.

It's a very long way till November. Pacing is important and over exposure can be a bad thing.
Also events can change things dramatically and the right wing knows how to cause events. We need to pray and hang on tight. I wish there were no polls.

Couldn't agree with you more demtom. Especially on points 2 and 3. I think the main reason the Zogby poll doesn't get the same play as the other two is that he's not tied into a network. CNN and ABC will trumpet their polls all over the place. And the key number is Bush's. As a challenger, Kerry can be expected to get 2/3 of the undecided vote.

demtom, thanks for having the energy to say all that and calm the Net masses, many of whom are freaking out. I just always overreact when the same-old intra-party candidate-obsessed nonsense comes out.

I think you're objectively correct about everything in your post.

Thanks, guys. It's nice to know I'm not just shouting into the void.

I'm with you demtom. I mean seriously, people won't really begin to look closely at the candidates till at least after the conventions. Polls at this stage are really not much more than a test of parth loyalty. If the narrow margin is any indication, I think all we can really say for sure is that Democrat and Republican partisans won't be doing much defecting this year.

I think these polls show that there is not and probably never will be a clear front runner in this election. It is going to come down to the wire.

The "rally effect" helps Bush only today; tomorrow likely is different. Voters tend to support the president when a crisis hits. Jimmy Carter's approval ratings went up in November 1979 when the hostage crisis began; they went up again when the rescue mission failed. In time, these failures destroyed his support.

Voters will rally around the president, but they will cotinue to support the president only if his policies succeed. They do not reward stalemates or failure. Carter could not resolve the hostage crisis in 1980 and went down to a landslide defeat. The polls remained close until five days before the election. More and more, the 2004 election is resembling the 1980 election.