Brendan Nyhan

[Update (6/30/10): Serious questions have been raised about the validity of Research 2000’s polls. The results discussed below should thus be viewed as potentially suspect until the matter is resolved.]

Josh Tucker (a political scientist at NYU) emails to ask if there are significant regional differences in the data on the state of the GOP brand that I blogged about yesterday.

Tucker, like many other bloggers, was struck by a chart created by Steve Benen highlighting differences a September Daily Kos poll found in views of the Republican Party by region:

As a point of comparison, I checked both the May 1993 Pew poll featured yesterday and a CBS/New York Times poll from late 1994 and there weren’t huge regional differences in Republican favorability between the South and the rest of the country. The same applied in a 2006 CBS/NYT poll.

The Pew poll I blogged about yesterday (the Religion & Public Life Survey) isn’t available online, but I checked several recent survey questions about the GOP image for which raw data is available in the Roper Center database. The 2009 survey that most closely replicates the Kos question about views of “Republicans in Congress” (a CBS poll) shows a smaller difference between the South and other regions, though it was conducted in March:

In addition, two other surveys asking closely related questions about approval of Republicans in Congress and views of the Republican Party show no obvious divergence between the South and the rest of the country:

One objection is that the CBS and USA Today/Gallup polls took place before the anti-Obama backlash had gotten underway. However, the CNN poll above was conducted July 31-August 3 and shows relatively similar views of the Republican Party by region.

There are important cultural and political differences between the South and the rest of the country, but those differences may be less dramatic than the Kos question suggests. It would be useful if other polls could break out their results by region to see if the Kos finding holds more generally.

(Cross-posted at Pollster.com)

How weak is the Republican brand right now? This issue came up yesterday when a Media Matters criticized The Hill for failing to mention the GOP’s poor polling numbers in a story on the 2010 elections. Similarly, I recently suggested that that the damaged Republican brand might limit the number of seats that the party picks up. But is the party really worse off than previous opposition parties at this point in the election cycle?

As a first cut at the question, I pulled all the relevant polling on approval of the party in Congresss and party favorability from the Roper iPoll database for the periods leading up to the four most recent midterms (1994, 1998, 2002, and 2006). In both cases, the results are consistent, but I’ll focus on the favorability questions since Pew and CBS asked comparable questions about party favorables in each cycle.*

The overall finding is simple — the GOP’s standing relative to the Democrats on both measures is worse than any opposition party in the sample. For instance, the Pew data show that the Republicans are currently viewed more negatively than any minority party in the previous four midterms in terms of both net favorables and the difference in net favorables between parties:**

The CBS results (not shown) are even more dramatic. In June, when the question was most recently asked, Republican net favorables were -30% and Democratic net favorables were 25%, which swamps the comparable results from the previous cycles.

In short, there’s no question that the GOP party brand is in worse shape than any opposition party in recent memory. The question, however, is whether this difference in party valence will (a) persist through next November and (b) translate into fewer GOP House seats at the polls, especially once we account for the generic Congressional ballot, which should (in principle) take much of this difference into account (see Alan Abramowitz’s model, for instance). Those questions remain to be addressed.

* Also, the approval question seems to be less closely related to electoral outcomes — for instance, disapproval of Republicans in Congress was high in September 1994.

** I chose the survey closest to the current point in the electoral cycle, though the exact date varied. Net favorables are defined as the percentage of Americans who have a favorable view of the party minus the percentage who have an unfavorable view.

(Cross-posted to Pollster.com)

The grim news from the New York Times yesterday is that the swine flu vaccine is likely to cause an outbreak of misperceptions about its consequences for people’s health:

As soon as swine flu vaccinations start next month, some people getting them will drop dead of heart attacks or strokes, some children will have seizures and some pregnant women will miscarry.

But those events will not necessarily have anything to do with the vaccine. That poses a public relations challenge for federal officials, who remember how sensational reports of deaths and illnesses derailed the large-scale flu vaccine drive of 1976.

This time they are making plans to respond rapidly to such events and to try to reassure a nervous public — and headline-hunting journalists — that the vaccine is not responsible.

Every year, there are 1.1 million heart attacks in the United States, 795,000 strokes and 876,000 miscarriages, and 200,000 Americans have their first seizure. Inevitably, officials say, some of these will happen within hours or days of a flu shot…

Officials are particularly worried about spontaneous miscarriages, because they are urging pregnant women to be among the first to be vaccinated. Pregnant women are usually advised to get flu shots, because they and their fetuses are at high risk of flu complications, but this year the pressure is greater…

“There are about 2,400 miscarriages a day in the U.S.,” said Dr. Jay C. Butler, chief of the swine flu vaccine task force at the federal Centers for Disease Control and Prevention. “You’ll see things that would have happened anyway. But the vaccine doesn’t cause miscarriages. It also doesn’t cause auto accidents, but they happen.”

Unfortunately, human beings are terrible at counterfactual thinking — in this case, assessing what would have happened without the shot. That hasn’t changed since 1976, but the problem is likely to be much worse this time around for several reasons:

-The scale of the problem is much worse (the swine flu is already a global pandemic).
-Many people have been influenced by the false claim that childhood vaccines are responsible for the increase in autism diagnoses (among other things).
-Every crazy story blaming the vaccine for heart attacks, miscarriages, etc. will be spread around the world via cable news and the Internet within hours.

But what I’m really worried about is the possibility that an anti-vaccine backlash will become an ideological issue. Glenn Beck asked his audience Monday night “Am I going to allow the government to give my child a flu shot? … I have already made my decision, but everyone has to make their own decision … Who will allow their children to be given a flu shot by this government?” Given the frenzied opposition to President Obama on the right, it’s likely that the vaccination campaign will be portrayed as some sort of socialist plot.

Via an NPR story on Houston, one of the great news ledes of all time:

The constitutional right to have a giant inflatable gorilla in a bathing suit and sunglasses grabbing consumer attention from a Houston business rooftop is the key issue in a trial that began in federal court on Wednesday.

Eternal vigilance is the price of freedom… to display things like this:

God bless America!

Update 9/29 4:52 PM — Sarah Cook has an awesome tweet linking to this post:

First they came for the giant inflatable gorillas, and I didn’t speak up because I wasn’t a giant inflatable gorilla: http://tr.im/A8Ke

In a comment on my post about the 2010 midterms yesterday, Charlie Cook protests the “lack of focus on actual Congressional districts”:

What I find interesting about this whole conversation is the lack of focus on actual Congressional districts. When you look at the 84 CD’s currently held by Democrats, that went for either Bush 2004 or McCain 2008, the 48 Democratic seats that went for Bush and McCain, the 54 seats that were in Republican hands four years go, it is very clear that the party’s vulnerability exceeds their margin of 40 seats.

In particular, Cook says, the remaining Southern Democrats who hold competitive seats are vulnerable:

I was interested in your comment, “There’s no comparable regional partisan shift working against the Democrats right now.”

Have you been in the South lately? The level of anti-Obama, anti-Democratic and anti-Congress venom is extraordinary, and with 59 Democrat-held seats in the region, 22 in or potentially in competitive districts, this is a very serious situation for Democrats. I have had several Democratic members from the region say the atmosphere is as bad or worse than it was in 1994.

This is not just about President Obama. It is anti-Congress and anti-Democratic Congress.

While the election is obviously 13 months away and much can change, that means it could get better, or snowball and get worse. To the extent that Democratic performance in 2008 was elevated by unusually high African-American turnout, that exposure to decline is even greater.

At this point, Democratic members in the South, Border South, Mountain states, in districts with heavy rural and small town populations as opposed to urban and suburban, particularly those with few transplants from other parts of the country, and fewer college graduates, are at particular exposure. Some of these members have either never had a tough race or haven’t in many years, with campaign organizations that are hardly sharped to a fine edge.

So while the Democratic performance in the generic Congressional, which is substantially lower than it was during the periods leading up to the 2006 and 2008 elections, when these majorities were built, that is only part of the case for why this may be an extremely challenging election for Democrats.

How seriously should we take these objections? On the first point, Cook’s job is to focus on the details of individual races, so it’s not surprising that he thinks we should do so. But it’s easy to be drawn into highly idiosyncratic narratives and end up losing sight of the big picture. In particular, individual House races are a noisy, lagging indicator of national trends (see, for instance, the House races that suddenly became competitive very late in 1994 and 2006). Political scientists try to abstract away from these details and analyze the underlying process that generates House election outcomes. Cook argues that many House seats held by Democrats are potentially vulnerable, but majority parties always hold marginal seats. The question is whether the number of potentially vulnerable Democratic members is significantly greater than, say, the number of vulnerable Republicans in 2006. (In technical terms, what does the seats-votes curve look like for 2010 relative to previous elections?)

In terms of Cook’s second point about the South, I’m open to the idea that the regional shift against Democrats is not complete, making some members there particularly vulnerable. But with only 22 in competitive or potentially competitive races, it’s not clear that enough Southern Democrats will lose to create a 1994-style landslide.

Update 10/5 9:20 AM: See my post comparing the GOP image across the last five midterm election cycles for more on the 1994 comparison. I also considered regional differences in a second post, which did not find large differences in GOP favorability between the South and other regions in 1993-1994 or 2009.

(Cross-posted to Pollster.com)

During his interview with President Clinton yesterday on Meet the Press, David Gregory asked a question that is increasingly occupying the minds of prominent Democrats — “do you worry about a repeat of ’94 politically?”

Vice President Joe Biden raised a similar concern last week, telling attendees at a Democratic fundraiser in Delaware that “If [Republicans] take them back [35 Democratic House seats in traditionally Republican districts], this [is] the end of the road for what Barack and I are trying to do.”

So is the House really in play? Analysis by several political scientists suggests that the answer is yes. Democrats could lose the House, which would take a 40 seat swing, but a 1994-style landslide seems unlikely.

A Sept. 3 column by Alan Abramowitz, a political scientist at Emory University, analyzes a statistical model of midterm elections since 1946 and concludes that Democrats should expect significant losses:

Democrats are likely to lose at least 15 seats in the House of Representatives in 2010 and their losses could go as high as 30-40 seats. The Senate looks more promising for Democrats because there are as many Republican as Democratic seats up for election next year but a loss of 3-4 seats is entirely possible.

Under what Abramowitz calls “what might be considered a worst case scenario for Democrats” in which “President Obama’s approval rating sinks into the low 40s next year” and “Republicans take a 5 point lead on the generic ballot,” he projects a GOP gain of four seats in the Senate and 41 seats in the House — just enough to take back control of the lower chamber.

Tom Holbrook at UW-Milwaukee cautions, however, against Charlie Cook’s warning that “wave elections, more often than not, start just like this: The president’s ratings plummet; his party loses its advantage on the generic congressional ballot test; the intensity of opposition-party voters skyrockets; his own party’s voters become complacent or even depressed; and independent voters move lopsidedly away”:

Let’s look first at presidential approval… [I]t appears that early reports of presidential approval are fair predictors of midterm losses (r=.60, p=.02), though there are still a number of data points that are significantly off the regression line… Still, there is a clear relationship here, one that might foreshadow the outcomes of next year’s elections. Based on the Obama’s current level of approval (52%, Gallup polls in September), the trend in these data predicts a seat loss of 28 seats for the Democrats in 2010.

Should this be cause for Republican jubilation and Democratic hand-wringing? Not quite. The problem with forecasts like this one is that the sample size is small enough (n=15) and the forecasting error is large enough (standard error of forecast=17.9) that a 95% confidence interval around the prediction ranges from a loss of 67 seats to a gain of 11 seats. In other words, the prediction from these data encompasses everything from a complete Democratic collapse to historic gains for the Democrats. The best guess is still substantial Democratic losses, but with plenty of hedging.

What about Obama’s standing among independents? Is it particularly important to the Democratic fortunes in the 2010 elections? It turns out that presidential support among independents is no more or less important than the overall level of presidential approval…

Finally, we turn to the generic congressional ballot. It is clear from other research that the generic ballot predicts well in the fall of election years (see Abramowitz), but is it really of much use 14 months out? In a word, no…

The third political scientist to weigh in, Andrew Gelman at Columbia, suggests that the current generic Congressional ballot numbers for Democrats are roughly consistent with a Republican vote swing that would be large enough to take back the House (though he admits he’s extrapolating — see Holbrook’s finding above).

Finally, low-volume trading on the Intrade prediction market puts the probability of a Republican takeover of the House at 37%.

In the end, Democrats seem likely to suffer significant losses (especially if the economy hasn’t started to turn around), but Holbrook is right to emphasize the level of uncertainty, which is still relatively high. As to the 1994 question, here’s what Clinton said to Gregory yesterday:

PRES. CLINTON: It, it–there’s no way they can make it that bad, for several reasons. Number one, the country is more diverse and more interested in positive action. Number two, they’ve seen this movie before, because they had eight years under President Bush when the Republicans finally had the whole government, and they know the results were bad. And number three, the Democrats haven’t taken on the gun lobby like I did, and they took 15 out of our members out. So I don’t think it’ll be–whatever happens, it’ll be manageable for the president.

I’d put a slightly different spin on the first and third points. From a political science perspective, 1994 was the culmination of the long decline of Democratic dominance among whites in the South — many incumbents were vulnerable on issues like guns, gays in the military, etc. because their districts had changed. There’s no comparable regional partisan shift working against the Democrats right now. Clinton’s second point can be similarly reinterpreted — the damage done to Republican brand under President Bush may restrict Republican gains in this election relative to 1994.

Update 9/28 11:52 AM: I missed a more recent Abramowitz analysis, which argues that a repeat of 1994 is unlikely due to the growing proportion of non-white voters in the electorate — a development that is likely to damage the GOP’s prospects due to its weakness with those groups (see Clinton’s first point above).

Update 9/29 7:19 AM: The Hill points out another reason that a 1994-style wave election is unlikely – the lack of retirements by incumbents:

In the last three “wave elections,” the party that lost a large number of seats has been hampered by incumbents not running for reelection. But so far in the 2010 cycle, not a single House member has announced his or her retirement, though 18 — seven Democrats and 11 Republicans — have said they will run for higher office.

(Cross-posted to Pollster.com)

Last week I criticized those conservatives who rushed to describe President Obama’s decision to withdraw a missile defense system from Eastern Europe as “appeasement.” The comparison to the appeasement of Hitler was simply absurd on its face. And, as I pointed out in a comment, the rationale for the move was unknown — how could conservatives even know if it was intended as appeasement?

Today, however, the picture has started to come into focus. The New York Times reports that Obama’s decision was part of a quid pro quo in which Russia agreed to support tougher sanctions against Iran:

President Obama, in his first visit to the opening of the United Nations General Assembly, made progress Wednesday on two key issues, wringing a concession from Russia to consider tough new sanctions against Iran and securing support from Moscow and Beijing for a Security Council resolution to curb nuclear weapons.

The successes came as Mr. Obama told leaders that the United States intended to begin a new era of engagement with the world, in a sweeping address to the General Assembly in which he sought to clearly delineate differences between himself and the administration of President George W. Bush.

One of the fruits of those differences — although White House officials were loath to acknowledge any quid pro quo publicly — emerged during Mr. Obama’s meeting on Wednesday afternoon with President Dmitri A. Medvedev of Russia, the first between the two since Mr. Obama decided to replace Mr. Bush’s missile defense program in Eastern Europe with a version less threatening to Moscow.

With a beaming Mr. Obama standing next to him, Mr. Medvedev signaled for the first time that Russia would be amenable to longstanding American requests to toughen sanctions against Iran significantly if, as expected, nuclear talks scheduled for next month failed to make progress.

“I told His Excellency Mr. President that we believe we need to help Iran to take a right decision,” Mr. Medvedev said, adding that “sanctions rarely lead to productive results, but in some cases, sanctions are inevitable.”

White House officials could barely hide their glee. “I couldn’t have said it any better myself,” a delighted Michael McFaul, Mr. Obama’s senior adviser for democracy and Russia, told reporters after the meeting. He insisted nonetheless that the administration had not tried to buy Russia’s cooperation with its decision to scrap the missile shield in Europe in favor of a reconfigured system.

Privately, several administration officials did acknowledge that missile defense might have had something to do with Moscow’s newfound verbal cooperation on the Iran sanctions issue.

I hope we can all agree that there’s a categorical difference between this sort of routine great powers diplomacy and giving land to Hitler so that he won’t invade other countries.

[Update (6/30/10): Serious questions have been raised about the validity of Research 2000’s polls. The results discussed below should thus be viewed as potentially suspect until the matter is resolved.]

Tom Jensen of Public Policy Polling continues his interesting polling on political misperceptions in a new national poll (PDF) that was conducted Sept. 18-21 and released today.

As in the New Jersey poll released last week, PPP’s national poll includes contrasting “truther” and “birther” questions. However, Jensen consulted with me on the wording of the “truther” question, which I had criticized, and ended up selecting new wording I adapted from a Scripps Howard/Ohio University poll. The new question removes the ambiguity associated with the phrase “advance knowledge,” which could be interpreted to refer to the August 2001 memo Bush received warning of the threat from Al Qaeda:

Old question: “Do you think George W. Bush had advance knowledge of the 9/11 attacks?”

New question: “Do you think President Bush intentionally allowed the 9/11 attacks to take place because he wanted the United States to go to war in the Middle East?”

Here are the poll results for the two misperceptions by party:

The primary finding is that the Obama birth certificate misperception has become far more prevalent among Republicans (42% no, 22% not sure) than the 9/11 misperception for Democrats (25% yes, 12% not sure). The percentage of Republicans who directly endorse the myth has increased substantially since the Daily Kos poll released in late July (which found 28% of Republicans said Obama wasn’t born in this country and 30% weren’t sure).

In terms of the 9/11 myth, the PPP results are generally consistent with what Scripps found in 2006, though reported levels of 9/11 misperceptions are lower among Democrats* (note: the Scripps results are juxtaposed below with the Daily Kos poll results referenced above):

The difference in 9/11 misperceptions between the two polls could be the result of variations in question wording (among other things, the original Scripps question refers to “people in the federal government” rather than President Bush) or differences in the response options (yes/no versus a scale of likelihood). 9/11 misperceptions may also have declined over time, particularly since Bush is no longer in office.

(Cross-posted to Pollster.com)

Matthew Yglesias calls it “striking” that Mitt Romney and so many other conservatives are engaging in opportunistic position-taking against TARP:

Dave Weigel points out that Mitt Romney is now slamming the TARP bill that he once favored. Shocking to see that guy change his position on something.

But Romney aside, it’s striking to see the number of conservatives who’ve decided that an initiative proposed by George W. Bush and Hank Paulson and endorsed by the GOP congressional leadership was and is secretly some socialist plot. Similarly with the idea that Ben Bernanke, former Bush administration official, is running some sort of rogue left-wing operation at the Fed. Obviously the economy’s still in a bad position today, but the evidence really does indicate that the whole suite of recovery measures (stimulus, TARP, Fed programs) are having the desired effect and things are turning around. I think this strategy of opportunistically pretending to have opposed this stuff could really come around to bite the right in the butt if things are looking better in 12 months and I think it’s very likely that things will be looking better in 12 months.

While it’s true that “opportunistically pretending to have opposed this stuff could really come around to bite the right in the butt if things are looking better in 12 months,” that’s a rational calculation that politicians often make. If TARP and the stimulus work and the economy is in great shape in 2012, it’s extremely likely that President Obama will be re-elected regardless of what Romney does. So he has every incentive to put his chips down on the side of predicting TARP’s failure, which could help position him to defeat Obama in 2012 if the economy is doing poorly. The same reasoning applies for the GOP more generally in 2010 and 2012.

In an online column last week, Eric Alterman and co-author Mickey Ehrlich denied any parallel in levels of partisan support for the “birther” and “truther” misperceptions:

David Paul Kuhn at Time.com attempted evenhandedness by introducing the results of supposed “truther” polls conducted in 2007. His claim is that the same number of Democrats believed that Bush knew about the 9/11 attacks in advance as Republicans believe Obama is not a citizen. However, this comparison doesn’t work. While Bush was handed a memo explaining that Osama bin Laden was determined to attack in the continental United States and decided to go fishing that day, no evidence exists anywhere to dispute Obama’s citizenship.

However, the parallel isn’t based simply on one potentially ambiguous poll question (a Rasmussen poll asking “Did Bush know about the 9/11 attacks in advance?”). As I showed back in August, Democratic support for the proposition that “People in the federal government either assisted in the 9/11 attacks or took no action to stop the attacks because they wanted to United States to go to war in the Middle East” in 2006 was quite comparable to birther misperceptions among Republicans in recent polls. What’s bizarre is that the Kuhn post linked by Alterman and Ehrlich cites me making precisely this point. Did they even read it?