May 16, 2010 - May 22, 2010


Circular Charts Suck 'Outliers'

Topics: Outliers Feature

Tom Jensen rethinks the enthusiasm gap.

Anthony Wells asks what went wrong in the British pre-election polls.

The Pew Research Center re-assesses the cell phone challenge to surveys.

Alex Lundry shares a Stephen Few paper on why circular charts suck.

Reliance on Cable News: More Than We Thought?

Topics: Cable news ratings , Ezra Klein , Health Care Reform , Kaiser Family Foundation

Ezra Klein shares my fascination with the questions in the monthly tracking poll by the Kaiser Family Foundation (KFF) that ask Americans how they get their information about the health care reform law.

For the last two months, KFF's pollsters have asked a unique set of questions that probes a list of possible information sources and tells respondents to include both the "channels" and "websites" for various broadcast and cable networks. The somewhat startling finding is that Americans say they have gotten information more often from "cable TV news channels such as CNN, FOX, or MSNBC, or their websites" than from "national broadcast network news channels such as ABC, NBC, or CBS, or their websites."


Klein concludes:

This is, to say the least, weird. For one thing, many, many fewer people watch cable news than watch network news or listen to the radio. Yet this poll shows that cable news outranked both those sources in the "how did you learn about the law" question and in the "how important was this source in teaching you about the law" question.

I had a similar reaction when I saw the similar KFF findings from last month, having recently compiled the following ratings statistics (provided to me by the Nielsen Company) for a chapter in a forthcoming academic text. In mid-February 2010:

  • The average combined audience for the three broadcast network evening news programs (ABC World News Tonight, CBS Evening News, and NBC Nightly News) was just over 24 million.

  • The average combined audience for the three broadcast network morning shows (the Today Show, Good Morning America, and the Early Show) was a little over 13 million.

  • The average combined audience for the three major cable news networks (CNN, Fox News and MSNBC) over the course of their broadcast day added to roughly 2 million.

The last bullet requires some explanation: Nielsen reports a rating based on the average audience size for any given minute of the broadcast day. Thus, we know from the statistics above that the combined cable news audience at any given moment of the day is smaller than the combined audience of the evening news broadcasts on any given evening.

What these statistics cannot tell us is how the cumulative cable audience compares. In other words, I'd like to know what percentage of Americans watches a half-hour evening news broadcast at least once a week versus how many watch at least a half-hour of cable news programming at least once a week? The standard ratings don't tell us that. They also tell us nothing about how many more Americans "tune in" to each network online. (The category totals above are based on adding together Nielsen's ratings for individual networks using a method described by Markus Prior in his 2009 article in Public Opinion Quarterly, "The Immensely Inflated News Audience: Assessing Bias in Self-Reported News Exposure;" similar Nielsen data is updated regularly on the indispensable web site, TV by the Numbers).

The Nielsen Company could produce such an estimate from their ratings data and release it into the public domain, but as far as I know, has not done so. I wish they would.

This month's KFF poll includes a follow-up that asks those who say get information about health reform from a cable news networks to identify the network they rely on most. Among other things, they identify 12% of Americans who say they rely mainly on Fox News and the Fox News website for information about the health reform law. Only 15% of these Americans rate the new health reform law favorably, 78% rate is unfavorably (see pp. 4-5 of the findings).

CA: 2010 Sen, Gov (Kos 5/17-19)

Topics: California , poll

DailyKos.com / Research 2000
5/17-19/10; 600 likely voters, 4% margin of error
400 likely Republican primary voters, 5% margin of error
Mode: Live telephone interviews
(Kos release)


2010 Governor
46% Brown (D), 42% Whitman (R) (chart)
47% Brown (D), 37% Poizner (R) (chart)

2010 Senate
47% Boxer (D), 40% Campbell (R) (chart)
48% Boxer (D), 39% Fiorina (R) (chart)
47% Boxer (D), 38% DeVore (R) (chart)

2010 Governor: Republican Primary
46% Whitman, 36% Poizner (chart)

2010 Senate: Republican Primary
37% Campbell, 22% Fiorina, 14% DeVore (chart)

Favorable / Unfavorable
Jerry Brown: 48 / 43
Meg Whitman: 47 / 38
Steve Poizner: 39 / 43
Barbara Boxer: 49 / 42 (chart)
Tom Campbell: 43 / 42
Carly Fiorina: 40 / 43
Chuck DeVore: 38 / 44
Arnold Schwarzenegger: 27 / 69 (chart)
Dianne Feinstein: 46 / 45 (chart)
Barack Obama: 62 / 33 (chart)

OK: 2010 Gov (CHS&A 4/19-22)

Topics: Oklahoma , poll

Cole Hargrave Snodgrass & Associates (R)
4/19-22/10; 500 registered voters, 4.3% margin of error
Mode: Live telephone interviews
(Mary Fallin campaign post)


2010 Governor
52% Fallin (R), 30% Edmondson (D)
52% Fallin (R), 30% Askins (D)

US: Elena Kagan (Fox 5/18-19)

Topics: national , poll

Fox News / Opinion Dynamics
5/18-19/10; 900 registered voters, 3% margin of error
Mode: Live telephone interviews
(Fox release)


President Obama has nominated Elena Kagan to serve on the U.S. Supreme
Court. If you were voting on Kagan's nomination, would you vote to confirm her
or not?

39% Yes, confirm 29% Not confirm

Overall, do you think Elena Kagan is qualified to serve as a justice on the
Supreme Court or not?

45% Qualified, 26% Not qualified

US: National Survey (Economist 5/15-18)

Topics: National , poll

Economist / YouGov
5/15-18/10; 1,000 adults, 3.6% margin of error
Mode: Internet
(Economist release)


Obama Job Approval
46% Approve, 47% Disapprove (chart)
Dems: 82 / 12 (chart)
Reps: 7 / 90 (chart)
Inds: 42 / 53 (chart)
Economy: 40 / 52 (chart)
Health care: 42 / 50 (chart)

Congressional Job Approval
14% Approve, 63% Disapprove (chart)

2010 Congress: Generic Ballot
Registered voters: 51% Democrat, 44% Republican (N=747) (chart)
Adults: 49% Democrat, 41% Republican

State of the Country
33% Right Direction, 54% Wrong Track (chart)

AR: 2010 Sen (Rasmussen 5/19)

Topics: Arkansas Senate

5/19/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Senate
66% Boozman (R), 28% Lincoln (D) (chart)
60% Boozman (R), 33% Halter (D) (chart)

Favorable / Unfavorable
Blanche Lincoln: 28 / 61
John Boozman: 64 / 28
Bill Halter: 41 / 50

Job Approve / Disapprove

Pres. Obama: 32 / 67
Gov. Beebe: 62 / 28

US: National Survey (Kos 5/17-20)

Topics: national , poll

DailyKos.com / Research 2000
5/17-20/10; 1,200 registered voters, 2.8% margin of error
Mode: Live telephone interviews
(Kos release)


Favorable / Unfavorable
Barack Obama: 54 / 42 (chart)
Nancy Pelosi: 38 / 52
Harry Reid: 30 / 58
Mitch McConnell: 25 / 60
John Boehner: 23 / 58
Democratic Party: 40 / 53
Republican Party: 31 / 62

State of the Country
41% Right Direction, 55% Wrong Track (chart)

US: National Survey (ARG 5/17-20)

Topics: National , poll

American Research Group
5/17-20/10; 1,100 adults, 2.6% margin of error
Mode: Live telephone interviews
(ARG release)


Obama Job Approval
49% Approve, 48% Disapprove (chart)
Dems: 85 / 11 (chart)
Reps: 8 / 89 (chart)
Inds: 44 / 53 (chart)
Economy: 44 / 52 (chart)

AZ: 2010 Gov (Rasmussen 5/17)

Topics: Arizona , Govenor

5/17/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Governor
Brewer (R) 52%, Goddard (D) 39%
Martin (R) 41%, Goddard (D) 40%
Goddard (D) 42%, Munger (R) 41%
Mills (R) 45%, Goddard (D) 38%

Favorable / Unfavorable
Terry Goddard: 48 / 45
Jan Brewer: 60 / 36
Dean Martin: 50 / 29
John Munger: 32 / 34
Buz Mills: 48 / 29

Job Approval / Disapproval
Pres. Obama: 39 / 60
Gov. Brewer: 64 / 35

MN: 2010 Gov (MPR 5/13-16)

Topics: Minnesota , poll

Minnesota Public Radio / University of Minnesota
5/13-16/10; 701 adults, 5.8% margin of error
Mode: Live telephone interviews
(MPR story)


2010 Governor: Democratic Primary
38% Dayton, 28% Kelliher, 6% Entenza

2010 Governor: General Election
35% Dayton (D), 31% Emmer (R)
315 Emmer (R), 29% Kelliher (D)
32% Emmer (R), 28% Entenza (D)

Job Approval / Disapproval
Sen. Klobuchar: 64 / 36 (chart)
Pres. Obama: 51 / 49 (chart)
Sen. Franken: 48 / 52 (chart)
Gov. Pawlenty: 43 / 57 (chart)

SC: 2010 Dem Gov Primary (Rasmussen 5/17)

Topics: Governor , Primary , South Carolina

5/17/10; 404 likely Democratic primary voters, 3% margin of error
Mode: Automated phone
(Rasmussen release)

South Carolina

2010 Governor: Democratic Primary
30% Sheheen, 22% Rex, 4% Ford

US: Health Care (Kaiser 5/11-16)

Topics: National , poll

Kaiser Family Foundation
5/11-16/10; 1,210 adults, 3% margin of error
Mode: Live telephone interviews
(Kaiser release)


As you may know, a new health reform bill was signed into law earlier this year. Given what you know about the new health reform law, do you have a generally favorable or generally unfavorable opinion of it?
41% Favorable, 44% Unfavorable (chart)

Do you think _______ will be better off or worse off under the new health reform law, or don't you think it will make much difference?
You and your family: 29% Better, 30% Worse, 32% No difference
The country as a whole: 43% Better, 35% Worse, 13% No difference

Party ID
35% Democratic, 21% Republican, 32% independent (chart)

US: National Survey (Fox 5/18-19)

Topics: Immigration , National , Oil

Fox News / Opinion Dynamics
5/18-19/10; 900 registered voters, 3% margin of error
348 Democrats, 5% margin of error
319 Republicans, 5.5% margin of error
178 independents, 7% margin of error
Mode: Live telephone interviews
(Fox News: Story Approvals, Toplines Approval, Story Immigration, Toplines Immigration, Story Oil Spill, Toplines Oil Spill)


Obama Job Approval
45% Approve, 46% Disapprove (chart)
Dems: 79 / 12 (chart)
Reps: 12 / 84 (chart)
Inds: 39 / 48 (chart)
"Attempted Bombing of Times Square": 52 / 18
"Oil Spill in the Gulf Coast": 44 / 36

Approve / Disapprove
Congress: 22 / 65 (chart)
Hillary Clinton: 59 / 19
Joe Biden: 43 / 34
Janet Napolitano: 40 / 31
Eric Holder: 36 / 27

41% Help Make United States a Better Place to Live, 31% Hurt and Make United States a Worse Place to Live

"Allowing local and state law enforcement officials to question anyone who they think may be in the country illegally"
65% favor, 30% oppose

"Your state passing an immigration law like Arizona's new law"
52% favor, 31% oppose

Boycotts of Arizona
23% support, 66% oppose

Increasing Offshore Drilling in Coastal Areas
54% favor, 38% oppose

Cause More Damage
36% Hurricane Katrina, 50% Gulf Coast Oil Spill

Party ID
39% Democrat, 35% Republican, 20% independent (chart)

NV: 2010 Sen Primary (PPP 5/12-16)

Topics: Nevada , Senate

Public Policy Polling (D) / Patriot Majority
5/12-16/10; 651 likely Republcian primary voters, 3.8% margin of error
Mode: Automated phone
(Jon Ralston Tweet)

Patriot Majority is PAC that normally advertises against Republican candidates.


2010 Senate: Republican Primary
29% Angle, 26% Lowden, 24% Tarkanian, 5% Chachas, 5% Christensen (trend)

Year of the Primary 'Outliers'

Topics: Outliers Feature

Ezra Klein wonders whether 2010 is "the year of the primary challenge."

Chris Bowers says the June 8th primaries are shaping up to be just as important as Tuesday's.

Frank Newport says jobs and the economy are still Americans' top priorities.

National Journal insiders name incumbents big losers of the week.

Democracy Corps argues that Obama's economic message is "demonstrably ineffective" and offers alternatives.

John Sides considers whether the Tea Party will succeed.

Gary Langer asks how Rand Paul's fumbling of a civil rights question could influence public perceptions of the Tea Party movement.

Nate Silver takes a look at Rasmussen's latest offering in Kentucky.

Reid Wilson reviews divergent California GOP Governor polls.

Tom Jensen compares Tim Burns to Scott Brown.

The Pew Internet & American Life Project segments Americans based on their usage of mobile devices.

CO: 2010 Gov (PPP 5/14-16)

Topics: Colorado , poll

Public Policy Polling (D)
5/14-16/10; 1,060 likely voters, 3% margin of error
Mode: Automated phone
(PPP release)


2010 Governor
44% Hickenlooper, 44% McInnis (chart)

Favorable / Unfavorable
John Hickenlooper: 47 / 33
Scott McInnis: 31 / 30

Job Approval / Disapproval
Gov. Ritter: 34 / 52 (chart)

PA: 2010 Sen, Gov (Rasmussen 5/19)

Topics: Governor , Pennsylvania , Poll , Senate

5/19/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)
Update: Governor


2010 Senate
46% Sestak, 42% Toomey (chart)

2010 Governor
49% Corbett, 36% Onorato (chart)

Favorable / Unfavorable
Pat Toomey: 56 / 25
Joe Sestak: 52 / 33
Tom Corbett: 57 / 26
Dan Onorato: 46 / 32

Job Approval / Disapproval
Pres. Obama: 47 / 52 (chart)
Gov. Rendell: 42 / 55 (chart)

DFA Survey: Priming the Vote in Arkansas?

Topics: Arkansas , Democracy for America , Measurement , Priming , Questionnaire Design , Research2000

This morning, we posted results from a new survey of Arkansas conducted by the firm Research 2000 on behalf of Democracy for America. The survey was conducted on Tuesday night among 500 "Democratic voters" who said the had voted in the Democratic primary that day. It shows Bill Halter with a not-statistically-significant two point lead over Senator Blanche Lincoln (48% to 46%).

While the DFA release is less than transparent about several aspects of their methodology (including how voters were sampled and whether and how the results were weighted), they do disclose the two most important things you need to know about this particular poll: The order in which questions were asked and the fact that Democracy for America is a political action committee that has formally endorsed Bill Halter.

Why does question order matter? Take a look at the questions they asked in the order they presented in the DFA Release (emphasis added):

QUESTION: Who did you vote for in the Democratic Primary election today, Blanche Lincoln, Bill Halter, or D.C. Morrison?

QUESTION: What do you think is the most important problem facing this country today?

QUESTION: Generally speaking do you think Democrats in Washington, DC are fighting hard enough against corporate special interests like Wall Street and Insurance Companies?

QUESTION: Generally speaking, do you think Democrats in Washington, DC are fighting hard enough to challenge the Republican policies of the Bush years, aren't fighting hard enough to change those policies, or are fighting about right?

QUESTION: Is the issue of national health care reform very important, somewhat important, or not important when deciding how or if to vote?

QUESTION: Do you favor or oppose the health care reform bill passed by Congress this year?

QUESTION: If oppose, do you think it goes too far or doesn't go far enough?

QUESTION: Do you favor or oppose the national government offering everyone the opportunity to voluntarily buy into a system like Medicare that would compete with private health insurance plans?

QUESTION: Is the issue of the economy very important, somewhat important, or not important when deciding how or if to vote?

QUESTION: Who do you blame more for our nation's current economic problems: The government or Wall Street?

QUESTION: What would do more to improve our nation's economic conditions: Decreasing government spending OR tightening government regulation of Wall Street and corporate executives?

I'm going to read you several pairs of statements, please tell us which comes closest to your views:

QUESTION: Democrats' economic policy is more focused on helping Wall Street than helping Main Street, OR Democrats' economic policy is more focused on helping Main Street than helping Wall Street

QUESTION: Blanche Lincoln is more on my side than on the side of the lobbyists and special interests, OR Blanche Lincoln is more on the side of the lobbyists and special interests than on the side of people like me.

QUESTION: Bill Halter is more on my side than on the side of the lobbyists and special interests, OR Bill Halter is more on the side of the lobbyists and special interests than on the side of people like me.

QUESTION: If the Democratic primary for Senate results in a run-off election between Blanche Lincoln and Bill Halter on June 8, do you plan to vote?

QUESTION: If yes, do you plan to vote for Blanche Lincoln or Bill Halter?

I probably oversimplify when I say that pollsters fall into roughly three schools of thought in terms of when they ask the vote preference question in a pre-election survey interview. Most media pollsters typically ask about vote preference as the very first question, or at least only after any questions to screen for likely voters, on the theory that asking any other questions might bias vote preference. Most campaign pollsters typically ask a few very neutral, general items before the vote (such as candidate favorable ratings or the so-called "right direction-wrong track" question), on the theory that they only serve to get voters thinking about the candidates (something they will inevitably do before voting) but do not prime specific issues or themes. And then there's the third category: Pollsters who either fell asleep in questionnaire design class or (dialing back the snark) are willing or eager to risk creating bias in the vote preference question.

Now I exaggerate a little, and I am certainly going to pick on Research 2000 today, but I have seen a disturbing number of independent media polls ask vote preference questions deep into surveys following questions with the potential to prime preferences.

The question order used in the DFA Arkansas survey, however, is especially problematic. Consider that if you visit Bill Halter's web site, you will learn that he "supported a bill that would have allowed the public to buy into a system that would have also provided more competition and choice." And more important, if you have been watching any commercial television in Arkansas the last month or so, you have seen advertisements from the Halter campaign claiming that Halter believes that "Washington is broken" and "cares more about special interests than you," that Senator Lincoln has "gone Washington," that she voted "aye" on "bailing out Wall Street" and has taken "millions in campaign cash from Wall Street."

In other words, the questions that immediately precede vote preference get voters thinking at considerable length about precisely the themes emphasized by the Halter campaign in its advertising. And we know that Halter has a perceived advantage on these themes, at least among the 500 respondents to this survey, because of the results of the two questions that immediately precede the vote preference question: By a margin of 46% to 42%, the voters sampled say Lincoln is "more on the side of the lobbyists and special interests than on the side of people like me." They say Halter is on the side of "people like me" by a margin of 35% to 14% (with 51% unsure).

Academics have a word for what this sort of question order can do to vote choices. it's called "priming." In 1987, political scientists Shanto Iyengar and Donald Kinder published the book, News That Matters, that details a set of experiments that proved that "by priming certain aspects of national life while ignoring others," news programs could "[set] the terms by which political judgements are rendered and political choices made" (p.4). The same principles extend to survey questions.

I asked Del Ali, the president of Research 2000, to confirm and explain the order of the questions on this survey. His response:

It was asked in the order that DFA had asked us too. I think you have a good point, however, let me a few points that I think one has to consider:

The poll was pretty close to the actual primary vote, therefore, one could argue that primary voters who are committed to their candidate and have something at stake by the fact they showed up to vote are not going to jump ship on any exit polling questions that may be deemed as negative towards one candidate or the other. In fact, I felt the questions were fair for all the candidates in that we asked the same question about each of them when it came to lobbyists, Wall street, etc... I don't think there is anything that we asked that is close to Lincoln beats her husband and Halter is the next Father Flannagan

I know the criticism is that we should have asked the run-off question right after we asked whom did they voted for on Tuesday, and that is fair. However, I do not feel we moved any of the voters preference and am confident that this race is a true toss up as of today.

And to be fair, it is certainly possible that vote preferences in Arkansas are now so firmly rooted that the order of questions on this survey had no priming effect. More straightforward surveys by other organizations may yield similar results. We will see. For now, however, we have good reason to be skeptical about the results of this survey. If it turns out to be something of an outlier, you will certainly know why.

SC: 2010 Gov Primary (Rasmussen 5/17)

Topics: poll , South Carolina

5/17/10; 931 likely Republican primary voters, 3% margin of error
Mode: Automated phone
(Rasmussen release)

South Carolina

2010 Governor: Republican Primary
30% Haley, 19% McMaster, 17% Barrett, 12% Bauer

AZ: 2010 Gov Primary (Rasmussen 5/17)

Topics: Arizona , poll

5/17/10; 541 likely Republican primary voters, 4% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Governor: Republican Primary
45% Brewer, 18% Martin, 18% Mills, 3% Munger

KY: 2010 Sen (Rasmussen 5/19)

Topics: Kentucky , Senate

5/19/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Senate
59% Paul (R), 34% Conway (D) (chart)

Favorable / Unfavorable
Rand Paul: 69 / 28
Jack Conway: 44 / 45

Job Approval / Disapproval
Pres. Obama: 36 / 63
Gov. Beshear: 46 / 52

GA: 2010 Gov Primary (InsiderAdvantage 5/18)

Topics: Georgia , poll

5/18/10; 423 likely voters, 5% margin of error
Mode: Automated phone
(InsiderAdvantage release)


2010 Governor: Republican Primary
23% Oxendine, 15% Handel, 14% Deal, 5% Johnson, 2% Chapman, 2% McBerry

CA: 2010 Sen, Gov (PPIC 5/9-16)

Topics: California , poll

Public Policy Institute of California
5/9-16/10; 2,003 adults, 2% margin of error
1,168 likely voters, 3% margin of error
411 likely Republican primary voters, 5% margin of error
Mode: Live telephone interviews
(PPIC release)


2010 Governor: Republican Primary
38% Whiman, 29% Poizner (chart)

2010 Governor: General Election
45% Brown (D), 32% Poizner (R) (chart)
42% Brown (D), 37% Whitman (R) (chart)

2010 Senate: Republican Primary
25% Fiorina, 23% Campbell, 16% DeVore (chart)

2010 Senate: General Election
46% Boxer (D), 40% Campbell (R) (chart)
50% Boxer (D), 39% DeVore (R) (chart)
48% Boxer (D), 39% Fiorina (R) (chart)

Job Approval / Disapproval
Gov. Schwarzenegger: 23 / 65 (chart)
Pres. Obama: 59 / 37 (chart)
Sen. Feinstein: 50 / 35 (chart)
Sen. Boxer: 50 / 38 (chart)

AR: 2010 Sen Primary Runoff (DFA 5/18)

Topics: Arkansas , poll

Democracy for America* (D) / Research 2000
5/18/10; 500 Democratic Primary Voters, 4.5% margin of error
Mode: Live telephone interviews
(DFA release)

*Democracy for America is a political action committee that has endorsed Halter and solicits contributions to his campaign.


2010 Senate: Democratic Primary Runoff
48% Halter, 46% Lincoln (chart)

Tales of Turnout 'Outliers'

Topics: Outliers Feature

Gary Langer sees a turn in consumer confidence as good news for incumbents.

Jennifer Agiesta tells a tale of turnout in AR & PA, and adds more on Kentucky.

The Analyst Institute reports on how to increase turnout (via Lundry).

Kos calls out Rasmussen for setting narratives, but not polling late.

Nate Silver asks what yesterday's primaries really meant.

Jay Cost questions PA-12's bellwether status; Tom Jensen and Tom Schaller add more on PA-12.

Mark Mellman reviews the four "anti" strains of public opinion.

Glen Bolger laments the death of independence.

David Hill approves of AAPOR's revised polling standards.

The British Election Study posts initial results (via Sides).

AZ: 2010 Sen (Rasmussen 5/17)

Topics: Arizona , poll

5/17/10; 1,000 likely voters, 3% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Senate
57% McCain (R), 28% Glassman (D)
49% Hayworth (R), 33% Glassman (D)

Favorable / Unfavorable
John McCain: 53 / 45
Rodney Glassman: 28 / 41
JD Hayworth: 45 / 44

Job Approval / Disapproval
Pres. Obama: 39 / 60
Gov. Brewer: 64 / 35

CA: 2010 Sen, Gov Primaries (M4 5/12-16)

Topics: California , poll

M4 Strategies / Small Business Action Committee
5/12-13,16/10; 600 likely Republican primary voters, 4% margin of error
Mode: Live telephone interviews
(M4: Senate, Governor)


2010 Senate: Republican Primary
33% Campbell, 28% Fiorina, 15% DeVore (chart)

2010 Governor: Republican Primary
49% Whitman, 32% Poizner (chart)

Incumbent Armageddon Effect: Epilogue

Topics: Arkansas , Incumbent Rule , Pennyslvania , Primary elections

Yesterday I speculated about whether we might see the return of largely abandoned "incumbent rule" which some of us used to interpret pre-election polling in the 1980s and 1990s but that largely disappeared over the last decade. And then last night, in Arkansas and Pennsylvania, we saw exactly the pattern that the old "rule" would have forecast: The challengers in both states performed much better than the final pre-election polls would have suggested.

Having re-opened this can of worms, you might expect me to crow about a newly revived "incumbent rule," but having thought on it further, I want to offer some cautions. We ought not make too much of the results of two somewhat unique primary elections. The current political environment creates huge opportunities for little-known political outsiders, to be sure, but "rules" of interpretation require far more empirical support than the examples from last night.

Let's review how the final polls compared to the outcome, starting with Pennsylvania:

2010-05-19-PA results.png

Our final trend estimate based on all the polls (not just the four from the final week listed above) shows the two candidates essentially tied (43.1% for Sestak to 42.5% for Specter), but Sestak won by 8 points (54.0% to 46.0%). Some will read these results as representing a roughly three-to-one "break" of the undecided vote to Sestak and thus fitting the theory behind the incumbent rule: Undecideds voters are really undecided about the incumbent and thus end up deciding to support the challenger at the last minute. Some of that may have been true, but I can think of a number of reasons to be cautious about such a conclusion:

First and foremost, the sum of all polling in this race showed a huge trend to Sestak that kicked into high gear after he started running television advertising on April 20. That trend ends on our chart with a virtual tie, but we should remember that only two polls continued calling through the weekend, and those two both wrapped up on Saturday. All four in the final week had field periods of 3-5 days, so the final polling snapshot had an epicenter of roughly Thursday.

Some looked at the results of the daily rolling-average tracking conducted by Muhlenberg University and saw a leveling off of the trend, but as I argued last week, it is all too easy to make too much of the sampling noise inherent in this sort of tracking. Muhlenberg tracked for 18 days. If we compare their first eight days (April 20-May 6) to their last eight (May 9-May 16) by averaging non-overlapping releases, they show Sestak closing a six point deficit (41% to 47%) to a dead heat (45% to 45%). At the same time, they show Sestak's personal rating jumping nine points, from 44% favorable-12% unfavorable to 53%-14%, but with nearly a third of likely voters (34%) still unable to rate him during the final week.

Second, we have good reason to assume that both trends -- Sestak gaining recognition and vote along with it -- continued and even accelerated over the last few days of the campaign. Local television stations and newspapers that had covered the race only sporadically all did Senate race stories over the last few days. Our trend line ends on Saturday, May 16.

Third, this race is one instance where our "more sensitive" trend estimate (available through the "smoothing" tool on the chart) provided a more appropriate summary of the data. You can make your own judgement by using the embedded chart below, which I have set up to show just Sestak's vote using the more sensitive setting. Try clicking through to the fully interactive version and toggling both the sensitivity setting (Tools/Smoothing) for both Sestak and Specter (Tools/Choices). The more sensitive setting showed Sestak with a 3.3 percentage point lead (44.7% to 44.1%) as of Saturday. That estimate is closer to the actual result.

Fourth, the data I obtained yesterday about the final week undecided voters from both the Quinnipiac and Muhlenberg polls did not suggest a hidden anti-Specter vote. Fewer than one voter in five of those who were still "undecided" had formed a negative opinion of Specter, although roughly half were neutral or expressed no opinion of Specter. If Sestak continued to gain support it probably came both from totally undecided voters and others who would have expressed a weak preference for Specter until they learned more about Sestak over the final weekend

Fifth, and probably most important, the problems of sampling and identifying likely voters confounds our ability to determine how undecideds "break," and both tasks are much, much harder in primary elections. Consider that only 12% of Pennsylvania's registered voters, and only 24% of its registered Democrats, cast a ballot in the Senate primary. So pollsters have a huge challenge identifying voters even when sampling from registered voter lists (never mind the related obstacles of covering those with missing phone numbers or that live in cell-phone only households).

Thus, it would not surprise me at all if many polls sampled the likely electorate too broadly -- including too many non-primary voters who might find a Republican-turned-Democrat like Specter more appealing. If so, the lessons learned yesterday are not likely to apply to general election polling in the fall.

Now let's turn to Arkansas:


At first blush, the Arkansas results seem like an even better example of the incumbent rule in action. Sen. Blanche Lincoln won virtually the same percentage of the vote has she received on our final trend estimate (both round to 45%), while challengers Bill Halter and D.C. Morrison picked up support as compared to the final polling.

The Arkansas polling yielded no clear trends comparable to Pennsylvania, but these results present other challenges. Only two pollsters (Daily Kos/Research 2000 and Mason Dixon) released results for the last two weeks, only one of those (Research 2000) was conducted in the final week and even finished interviewing last Wednesday, a full six days before the election.

All the difficulties of sampling and modeling primary turnout are also present in Arkansas. The turnout for yesterday's Democratic primary represented just 20% of registered voters in Arkansas. It's possible that undecided voters "broke" to Halter and Morrison. It is also possible that sampling or the definition of likely voters led to an understatement of support for Halter and Morrison .

Beyond that, one aspect of the Arkansas contest is unique to primaries: Also-ran D.C. Morrison garnered 13% of the votes cast, and some of that vote may have come from voters wishing to express a preference for "none-of-the-above." If so, that is the sort of phenomenon that is more likely in a primary election where less rooted in the party affiliation of the candidates.

The bottom line is that primary election polling is much more difficult -- and more prone to error -- than polling in general elections. Either way, we are probably better off avoiding "rules" for poll interpretation and looking more closely at the evidence available for each contest.

CO: 2010 Sen (PPP 5/14-16)

Topics: Colorado , poll

Public Policy Polling
5/14-16/10; 1,060 likely voters, 3% margin of error
Mode: Automated phone
(PPP release)


2010 Senate
44% Bennet (D), 41% Norton (R) (chart)
45% Bennet (D), 39% Buck (R) (chart)
44% Bennet (D), 36% Wiens (R) (chart)
43% Romanoff (D), 41% Norton (R) (chart)
41% Romanoff (D), 38% Buck (R) (chart)
41% Romanoff (D), 37% Wiens (R) (chart)

Job approval / Disapproval
Sen. Bennet: 34 / 44 (chart)
Pres. Obama: 45 / 50 (chart)
Sen. Udall: 38 / 41 (chart)

Favorable / Unfavorable
Jane Norton: 20 / 32
Ken Buck: 19 / 24
Tom Wiens: 8 / 16
Andrew Romanoff: 31 / 26

AZ: McCain 52 Hayworth 40 (Rasmussen 5/17)

Topics: Arizona , poll

5/17/10; 541 likely Republican primary voters, 4% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Senate: Republican Primary
52% McCain, 40% Hayworth (chart)

NH: 2012 Pres (UNH 4/18-28)

Topics: new hampshire , poll

University of New Hampshire
4/18-28/10; 228 likely Republican primary voters, 6.5% margin of error
198 likely Democratic primary voters, 7% margin of error
Mode: Live telephone interviews
(UNH release)

New Hampshire
2012 President: Republican Primary
41% Romney, 12% Palin, 11% Giuliani, 9% Huckabee, 6% Paul, 5% Gingrich, 3% Pawlenty, 2% Santorum

< em>Democrats: Do you plan on voting for Barack Obama in the 2012 New Hampshire Democratic primary or do you plan to vote for
another Democratic candidate?
68% Obama, 6% Another candidate

CT: 2010 Sen (Rasmussen 5/18)

Topics: Connecticut , poll

5/18/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Senate
50% Blumenthal (D), 39% Simmons (R) (chart)
48% Blumenthal (D), 45% McMahon (R) (chart)
53% Blumenthal (D), 37% Schiff (R) (chart)

Favorable / Unfavorable
Richard Blumenthal: 66 / 32
Rob Simmons: 54 / 37
Peter Schiff: 44 / 29
Linda McMahon: 54 / 42

Job Approval / Disapproval
Pres. Obama: 59 / 39 (chart)
Gov. Rell: 56 / 53 (chart)

US: National Survey (Pew/NatJournal)

Topics: National , poll

Pew Research Center / National Jounal / Society for Human Resource Management
5/13-16/10; 1,002 adults, 4% margin of error
Mode: Live telephone interviews
(Pew: report, toplines; National Journal: post)


State of the Country
28% Satisfied, 64% Dissatisfied (chart)

Job Rating
Congress: 13% Excellent/Good, 82% Only Fair/Poor (chart)
Pres. Obama: 39% Excellent/Good, 58% Only Fair/poor (chart)

From what you've seen and heard so far, do you think the Senate should or should not confirm Elena Kagan to the Supreme Court?
33% Should, 21% Should not, 46% Don't know

Party ID
32% Democrat, 23% Republican, 35% independent (chart)

Super Senate Tuesday Live Blog

Topics: Arkansas , Election Night , kentucky , Liveblog , Pennsylvania

We're live blogging!

Remember: (1) No exit polls tonight, at least not the ones conducted by the network consortium and (2) we'll have at least one special guest.

Primary Preview Outliers

Topics: Outliers Feature

Sean Trende asks if 2010 is anti-incumbent, anti-liberal, or anti-Democrat; Jay Cost says no to "anti-incumbent."

Glen Bolger says 2010 is not 1994.

Chris Good, Evan McMorris-Santoro and James L. preview today's primaries.

Ed Kilgore and Quinn MCord and Stephanie Palla discuss Kentucky.

Tom Jensen and Jamie Schufflebarger weigh in on Pennsylvania.

Dan Roem looks at Blanche Lincoln's Senate battle.

Jamie Schufflebarger sets the stage in Oregon.

Nate Silver takes a look at PA-12.

Ana Marie Cox live-blogs a Gallup event on predicting midterm outcomes (hit replay).

We will be live blogging the election returns starting at around 6:00 p.m. Eastern Time. Be sure to tune in.

CO: 2010 Sen, Gov, and 2012 Pres Primaries (PPP 5/14-16)

Public Policy Polling (D)
5/14-16/10; 448 likely Republican primary voters, 4.6% margin of error
442 likely Democratic primary voters, 4.7% margin of error
Mode: Automated phone
(PPP release)


2010 Senate: Republican Primary
31% Norton, 26% Buck, 5% Greenheck, 5% Wiens

2010 Senate: Democratic Primary
46% Bennet, 31% Romanoff (trend)

2010 Governor: Republican Primary
50% McInnis, 15% Maes (trend)

2012 President: Republican Primary
29% Palin, 25% Romney, 18% Huckabee, 16% Gingrich, 9% Paul

FL: 2010 Gov (Rasmussen 5/16)

Topics: Florida , poll

5/16/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Governor
43% McCollum (R), 35% Sink (D) (chart)
41% Scott (R), 40% Sink (D)

Favorable / Unfavorable
Bill McCollum: 48 / 33
Alex Sink: 38 / 32
Rick Scott: 39 / 28

Job Approval / Disapproval
Pres. Obama: 50 / 49 (chart)
Gov. Crist: 57 / 42 (chart)

AAPOR 2010: Thoughts from a First-Time Attendee

Topics: AAPOR 2010

Last week, I had the opportunity to attend the annual conference of the American Association for Public Opinion Research. The conference featured presentations from private sector, government, and academic researchers about their methods and findings, in addition to the release of reports by two AAPOR task forces - one on online panels, one on cell-phone surveying.

On the whole, I had a phenomenal experience. I truly enjoyed the spirit of collaboration as attendees and presenters shared best practices and supported each other's research.  I had an opportunity to meet an impressive group of established public opinion researchers, and also got to meet many young students and professionals who are doing fascinating work. (I now believe that pollsters are excellent conversationalists precisely because they're so good at asking questions.)

I think that one of the great benefits of attending AAPOR came in seeing how research is conducted by those in other industries.  For example, political pollsters deal with a variety of pressures that are lessened in academic research: the speed of data production, the need to insert your findings into the conversation quickly, as well as client demands and cost pressures.  While a major academic study may consume years of a doctoral candidate's life, a campaign poll typically needs rapid turnaround and subsequent immediate release in order to remain "fresh."  A campaign rarely if ever has time to improve its coverage and conduct in-person face-to-face interviews of populations missed by land-line and cell-phone surveys, for instance. Weeks or months of post-stratification are a luxury not afforded to those in the world of campaign polling.  

At AAPOR, you get exposure to "the ideal" - projects refined by the most advanced and rigorous techniques, exploring the toughest challenges of sampling, processing and analysis that the survey research field faces.  It highlights ways to improve your methods, regardless of field, and helps a researcher facing time and cost pressures make informed decisions about what is critical to producing useful data.  And for a political pollster, AAPOR is a great time to focus on these issues exclusively, away from discussion about whose clients won more races or who got which race predictions closest.

There was one thing that surprised me a bit about the AAPOR conference, and I'd love to hear comments on this from those who have been to the conference before or who have been involved in the organization more deeply.  Essentially, if AAPOR is the "American Association for Public Opinion Research," one might logically assume the conference would devote a substantial portion of time to the findings of public opinion research in addition to the methods of collecting data.  

A great example of a panel that balanced these two was the Gary Langer/Matthew Warshaw presentation about ABC News' "Where Things Stand" research in Afghanistan.  I walked away with a greater understanding of how to conduct research in the most incredibly challenging circumstances, but I also learned what the people of Afghanistan think about the future of their nation.

However, the vast majority of content from the conference was about the process of social science research. In some cases, it was not necessarily even about opinion research in the strictest sense of the word "opinion", but rather the collection of demographics.  This is understandable, given that at a professional conference, everyone is trying to figure out how to do what they do better, but I felt there was a very narrow focus on the methods of research and less attention paid to what we're finding.  

Why do we conduct opinion research in the first place? We do it to learn about certain groups of people and audiences.  Developing a research methodology that perfectly captures cell-only populations is as useful as the research findings it generates.  So what are we finding? Opinion research conducted by another organization about, say, shifting attitudes in America about the media, have a great deal of application to my work as a political pollster, even if that research presentation does not impact the methods of how I do my work in the future.

With the wealth of knowledge possessed by the various professional and academic organizations in AAPOR, it would be great to see more panels highlighting the findings of public opinion professionals.

In the end, I think it is critical that more political pollsters take the opportunity to focus on their methods in order to create the highest quality data. We often measure political pollsters by the accuracy of their results and how often their numbers are "on the money" when final ballots counts are in. A conference like AAPOR gives researchers the tools to make sure they are right rather than lucky. There is a great deal that political polling professionals can learn from their counterparts in other industries and I feel very thankful that I had the opportunity to attend this conference and learn from their experiences.

US: Elena Kagan (Rasmussen 5/16-17)

Topics: National , poll

5/16-17/10; 1,000 likely voters, 35 margin of error
Mode: Automated phone
(Rasmussen release)


Favorable / Unfavorable
Elena Kagan: 43 / 44

The U.S. Senate has the constitutional authority to confirm all Supreme Court nominees. Based upon what you know at this time, should the U.S. Senate confirm Elena Kagan as a Supreme Court Justice?
39% yes, 39% No

A Midterm Super Tuesday

The phrase "angry voters" is a redundant one in 2010, and we will likely see this on full display today in three states. Our sense is that after today there will be a large number of Congressional incumbents who will be wishing that they had chosen to retire in 2010.

Since every newspaper in America has told us that these races are "bellwether predictors," we'll examine several of them. But before we do so, here are some observations on the big picture political environment:

  • As much as anything else, voters are anti-Washington. Yes, voters are "anti-incumbent," but that's because incumbents are associated with Washington. You will also see non-incumbent candidates lose, especially if they're more "connected" to Washington than their opponents. For evidence, look no further than the latest WSJ/NBC poll, which shows that only 25% of voters trust the government in Washington "to do what is right most or all of the time."
  • Mid-term elections are low turnout events and angry voters tend to turn out. Polling is difficult in mid-terms because we are not necessarily sure who is going to vote. The composition of the turnout is the problem for incumbents (or those who are perceived to be Washington insiders) today. That is why Specter and Grayson will lose and Lincoln will be held under 50%.
  • It would be a mistake to read too much into these races. Yes, they are a measure of voter anger with Washington and incumbents, but they will tell us little in terms of projecting how badly Democrats will lose in the fall. Each race has its own unique characters and characteristics that overlay the national mood.

Pennsylvania's Democratic Senate Primary - Sen. Arlen Specter v. Rep. Joe Sestak

Arlen Specter is going to lose today because this is the absolute worst environment for him to be running in. He was either going to lose now or in the fall, so it was only a matter of time. He is the embodiment of Washington, and to top it off he switched parties to keep his seat. His opponent, Representative Joe Sestak, fired off one of the best ads by any candidate this month: it showed Specter talking about the switch to "keep" his seat while picturing him with George W. Bush. Specter is simply the wrong candidate at the wrong time.

Now, Specter has been underestimated before and has been polling almost dead-even with Sestak, so he is far from out of it. However, the trend in polling has clearly not been moving in his favor--he's fallen steadily from a 62%-24% advantage a year ago to his current tie with Sestak (they're deadlocked at 44% each). For an incumbent to be mired in the low forties--and apparently moving backward in polls with just a few days remaining--is very dangerous territory.

At this point, we believe that all signs are pointing to a Sestak victory. Vice President Biden has been raising money and even cut a radio spot for Specter, but it appears that the Obama political operation has thrown Specter under the bus; there have been no Presidential appearances for the Senator in the last several weeks. Of course, after the President was unable to save Corzine, Deeds or Coakley, maybe that's a good thing for Specter.

Kentucky's Republican Senate Primary - Rand Paul v. Trey Grayson (formerly Sen. Jim Bunning's (R) seat)

Now this race is the polar opposite. Rand Paul is the perfect candidate for these times. And his opponent fell into the trap of exhibiting his connections to Washington at a time when doing so is a huge liability.

Sixteen months ago, this looked to be a ho-hum contest to replace the retiring Jim Bunning in a relatively straightforward Republican nomination and victory. Secretary of State Trey Grayson has done and said all the right things and has even gotten what should be the right endorsements, including that of Kentucky's senior senator and the current minority leader, Mitch McConnell. So why is he down about ten points to ophthalmologist and first-time candidate Rand Paul? It's because Paul has successfully positioned himself as an outsider and as someone who takes the concerns of the anti-Washington Tea Party seriously; his endorsements from the Tea Party's "renegade" heroes like Sarah Palin and Sen. Jim DeMint have helped, too.

This does represent a chance for Tea Partiers to demonstrate whether they are truly able to help get their own candidates across the finish line. (Sen. Bob Bennett's defeat in the Utah caucus was also a surprising upset by a Tea Party insurgency, but the caucus structure greatly favors underdogs with enthusiastic supports--as the Clinton campaign discovered in 2008.) Grayson has run a competent campaign, but Rand Paul has in one way acted like his father's son, expertly stoking the current anti-Washington, anti-incumbent sentiment. Like the Democratic contest in Pennsylvania, that's what this race is ultimately about. Of course, that's also the Tea Party movement's defining feature--even more so than low taxes or smaller government--so in this way it is very much a chance for the Tea Party to flex its muscles. We're comfortable predicting that he will be the first true "Tea Party" candidate for statewide office in a general election.

Arkansas's Democratic Senate Primary - Sen. Blanche Lincoln v. Lt. Gov. Bill Halter

In the end, poor Blanche Lincoln may end up wishing she had lost the primary outright, because the next several months are probably going to be very painful. She is an establishment candidate who is also in trouble within her own party, making her almost unelectable in the fall.

She is still leading Lt. Gov. Bill Halter in most polls, but as another incumbent stuck in the mid-40s, she's also vulnerable, and the increasingly aggressive tone of her campaign shows that she's clearly worried. Of course, there's also a very good chance we won't have a winner on Tuesday night: per Arkansas election law, she will need to break 50% of the total vote to avoid an early June run-off election.

While she may have been somewhat vulnerable before, Lincoln's turn in the spotlight during the health care reform battle was disastrous. She was criticized from both sides; on the left, for pledging to join a filibuster against the public option, and on the right (in a conservative Southern state) for negotiating with the Administration on the final bill. Ultimately, she did vote for the bill: a major issue in a state where 65% of voters would like to see it repealed.

While all of the candidates we've covered also face tough general election contests, Lincoln (or Halter) is in an even more difficult position than the rest. Currently, Rep. John Boozman leads a very crowded field in the Republican primary. But in a hypothetical general election matchup, he currently leads either Democrat by more than 25 points--no, that's not a typo.

We predict a narrow victory for Lincoln tomorrow (with a runoff to come) but, ultimately, we think this seat is almost certain to swing to the GOP in November.

Pennsylvania's 12th District - Special election to replace the Late Rep. John Murtha (D)

This election is important to both parties for deeply symbolic reasons beyond the chance for Republicans to pick up another House seat. Both parties are clearly aware of the significance: the DCCC has spent $800,000 on advertising, and political rock stars like Bill Clinton, Joe Biden, Newt Gingrich and Scott Brown are all campaigning in the district. And, as Politico reports, labor and interest groups are pouring money into the race as well.

The reason this race has garnered national attention is that it is the sort of tough-but-winnable race that Republicans will need to carry if they hope to even think about a majority in the next Congress. The Democratic candidate, Mark Critz, is a former aide to the late John Murtha. He's positioned himself as an heir to Murtha's legacy, which might seem an odd choice given the country's immense anti-incumbent sentiment. However, this is a district that was tailor-made for Murtha through gerrymandering, and the popular Murtha has been coasting through elections and bringing home federal dollars for decades. On the other hand, PA-12 does have a Cook PVI of R+1, indicating that it's a swing district in federal elections, and the demographic lean is strongly toward Republicans, with an electorate that is 94% white, 13% military veterans and has a median age of 41.8--all well above the national and statewide averages.

A deeper look at polling in the district points to other potential landmines for any candidate with a "D" next to his or her name in swing districts like this. As PPP observed back in April, only 28% of voters in the district express support for health care reform (59% are opposed). Even Democrats in the district support it by just a 43/39 margin. While Critz has spoken out against the passage of health care reform on the campaign trail, this issue is a key anger point for conservative voters and all Democrats are wedded to the bill, regardless of their stated positions. And another PPP poll over the weekend found that the President and his party are somewhat unpopular, with 55% of voters disapproving of Obama and 63% holding an unfavorable opinion of Nancy Pelosi

But there are some technical factors in this election that seriously help Critz. Because the election will be held on the same day as the Democratic primaries for Senate and Governor, there should be a boost to their turnout, given the interest in the contest between Sestak and Specter we described above. Also, the party primaries are actually being held on Tuesday as well, so voters will be asked to vote for Burns twice (Critz has fewer serious opponents in his primary).

At this point it looks like Critz will gain a very narrow victory today. The Senate race will drive up Democratic turnout.

However, if the interest in the Specter-Sestak battle for the soul of the Democratic Party is overstated and Democratic turnout sags, it's very possible that Burns could pull off an upset. If he does, the sky may in fact be falling for Democrats and anything will be possible for Republicans on Election Day. Either way, expect the media to seize upon the result as a key piece in the narrative leading to November.

Special thanks to John Zirinsky and Peter Ventimiglia for their thoughts and insights.

NY: 2010 Sen (Rasmussen 5/12)

Topics: New York , poll

5/12/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)

New York

2010 Senate
51% Gillibrand (D), 28% DioGuardi (R)
51% Gillibrand (D), 31% Blakeman (R) (chart)
46% Gillibrand (D), 27% Malpass (R)

Favorable / Unfavorable
Kirsten Gillibrand: 57 / 37 (chart)
Bruce Blakeman: 29 / 33
David Malpass: 24 / 31
Joe DioGuardi: 29 / 32

Job Approval / Disapproval
Pres. Obama: 61 / 38 (chart)
Gov. Paterson: 28 / 69 (chart)

Incumbent Armageddon Effect?

Topics: Arkansas , Incumbent Rule , Pennyslvania , Primary elections

As you almost certainly know, hotly contested primary and special elections are being held today in what NBC's Chuck Todd has dubbed Super Senate Tuesday because of closely watched Senate primaries in Arkansas, Kentucky and Pennsylvania. I'm most most intrigued with how the polls perform in the two Democratic primaries involving incumbent senators (Blanche Lincoln in Arkansas and Arlen Specter in Pennsylvania) because of the potential lessons they may teach about what to look for in polling on incumbents for the rest of the year (Chris Cillizza's readers apparently share my interest, as they prefer the moniker "Incumbent Armagedon" for today's contests).

Our final trend estimates show Senator Blanche Lincoln leading challenger Bill Halter by nine percentage points (45.4% to 35.9%) in Arkansas and virtual dead-heat in Pennsylvania between Senator Arlen Specter (42.5%) and Joe Sestak (43.1%). If you use our chart's smoothing tool to pick the more sensitive estimate, the chart gives Sestak a roughly three point lead (44.7% to 41.4%), mostly because it gives greater weight to the one "outlier poll" from Suffolk University that has Sestak leading by 9 points. Three others all have the Pennsylvania race well within their respective margins of error.

In the last three general elections, polls have collectively been mostly unbiased, and as such, polling averages and trend estimates that minimize sampling error have been very accurate predictors of the final margins. Primary elections, however, have often been a very different story. Individual polls tend to be more variable and there are many examples of contests where the final polls have collectively missed the final margins by a mile -- most recently in the Democratic presidential primaries in 2008, especially New Hampshire, South Carolina and the Virginia's Democratic primary for Senate in 2009.

There are many reasons why primary election polling is more error prone -- see AAPOR's investigation of the 2008 primaries for a complete catalogue -- but I'll highlight three: Smaller electorates are harder to sample and model, final snapshots sometimes miss late trends, and the lack of party preferences leave more voters feeling truly uncertain about their choice until confronted with a ballot.

All of that is a long way of saying: I wouldn't be at all surprised if the actual results in Arkansas, Pennsylvania and Kentucky turn out differently than final polls suggest.

And that brings me to my interest in the two Democratic primaries featuring incumbent Senators. Polls in both contests feature relatively large percentages of the vote that are still undecided -- 11% to 15% in Arkansas and 11% to 16% in Pennsylvania. The safest interpretation is that many of these uncertain voters are unlikely to vote, but some see the potential for what some pollsters used to call the "incumbent rule," the idea that incumbents "get what they get" on the final poll an undecided voters "break" decisively for challengers. Such a pattern would put Sestak comfortably ahead in Pennsylvania and would have Lincoln falling short of the 50% necessary to avoid a run-off in Arkansas.

The most concise answer to that comes in the form of a "tweet" last night from our intern, Harry Enten:

Can we PLEASE STOP with undecideds break for challenger! THEY HAVEN'T IN 10 yrs. THEY DIDN'T IN TEXAS GOP OR ILLINOIS DEM GOV PRIM EITHER!

He is right that the pattern we saw in polling in the 1980s and 1990s largely disappeared over the last ten years. However, we did see something of an incumbent effect in the New Jersey Governor's race last year, especially in polls conducted with live interviewers. The fact that Republican Chris Christie did consistently better on automated survey, and that the automated surveys came much closer to the final margins, suggests that the undecided respondents on live interviewer surveys "broke" decisively in Christie's favor.


As such, note that all of the final week polls in Pennsylvania and Arkansas used live interviewers, and the anti-incumbent sentiment evident last fall has only intensified over the last six months.

Of course, all of this just speculation, though there are clues in the final surveys themselves. Feeling especially curious about Pennsylvania, I asked the pollsters at Quinnipiac and Muhlenberg Universities to run run cross-tabulations of the favorable ratings of the two candidates among the voters who were undecided on their final polls. These tabs yield roughly 150 undecided respondents in the Quinnipiac poll conducted last week and exactly 111 undecided respondents interviewed by Muhlenberg University from May 6 to May 16.


The results, as shown in the table below, are more or less consistent: Both show most of these voters to be disengaged. Roughly two-thirds don't know or can't rate Sestak and nearly half are unable to rate Specter. Only 18% of the undecideds on the Muhlenberg poll and 19% on the Quinnipiac poll rate Specter unfavorably. Put simply: it's hard to see a big break to Sestak coming from the undecided vote.

Of course, undecided voters are just one potential source of error. As always, we'll just have to wait for the actual results later tonight.

KY: 2010 Sen Primary (Magellan 5/16)

Magellan Strategies (R)
5/16/10; 809 likely Republican primary voters, 3.5% margin of error
Mode: Automated phone
(Magellan release)


2010 Senate: Republican Primary
55% Paul, 30% Grayson (chart)

US: Generic Ballot (Gallup, Rasmussen 5/10-16)

Topics: Generic House Vote , National


5/10-16/10; 1,600 registered voters, 3% margin of error
Mode: Live telephone interviews
(Gallup release)

2010 Congress: Generic Ballot
46% Republican, 45% Democrat (chart)

5/10-16/10; 3,500 likely voters, 2% margin of error
Mode: Automated phone
(Rasmussen release)

2010 Congress: Generic Ballot
43% Republican, 38% Democrat (chart)

Column: AAPOR's Transparency Initiative

Topics: AAPOR , AAPOR Transparency Initiative , Disclosure , Peter Miller

My new column reviews AAPOR's Transparency Initiative as described in detail this past weekend by outgoing AAPOR President Peter Miller. I hope you'll click through and read it all. It may not be quite as consequential as health care reform, but in the polling world it has the potential to be a very big deal (to paraphrase our Vice President).

This column follows up on two items posted last week: The first reviews the rationale for the initiative, and the second features a video interview with Miller and includes the full list of participants "so far." Regular readers will know that AAPOR's initiative jives neatly with my own ongoing interest in improving disclosure of polling methodology (discussed most completely here).

US: National Survey (Zogby 5/14-17)

Topics: National , poll

5/14-17/10; 2,063 likely voters, 2.2% margin of error
Mode: Internet
(Zogby release)


2010 House: Generic Ballot
43% Democrat, 42% Republican (chart)

Obama Job Approval
50% Approve, 50% Disapprove (chart)

State of the Country
40% Right Direction, 52% Wrong Track (chart)

No To Chartjunk 'Outliers'

Topics: Outliers Feature

Chris Bowers argues that GOTV efforts are already baked into polling and reviews the final Pennsylvania polls.

Andrew Romano argues that tomorrow's primaries won't be a big deal.

Gallup will live tweet their Election 2010 event tomorrow at #gallup2010.

Jennifer Agiesta notes an ethnic divide over Arizona's immigration law.

Frank Newport reflects on the AAPOR Conference.

ABC News' Afghanistan polling is honored by AAPOR.

Resurgent Republic examines subgroup shifts over the last year.

Stan Greenberg analyzes the UK elections.

Infosthetics blogs a study claiming that "chart junk" aids recall; Andrew Gelman responds.

AAPOR2010 Interviews Wrap-Up

Topics: AAPOR , AAPOR 2010

I'm back in DC and just filed a column that should appear on NationalJournal.com later this afternoon about the Transparency Initiative announced over the weekend at the conference of the American Association for Public Opinion Research (AAPOR). Kristen also tells me she's preparing a post on her experiences as a first-time conference attendee. So while this post will not truly "wrap-up" our coverage of the AAPOR conference, I do want to provide a collection of links to all of the interviews we conducted this past weekend at this year's conference (with some final thoughts below):

A few things to remember about AAPOR's annual conference (largely cribbed from my comments last year): First, our interviews barely scratch the surface of the breadth and depth of subjects covered at the conference and the findings presented. We tend to focus on topics related to pre-election polling, but the conference covers a much wider array of methodological issues that are often highly technical and not easily distilled into a quick video interview.

Second, remember also that many of the findings presented at the conference -- including some that made their way into our interviews -- are very preliminary. The quality of the "papers" presented at the AAPOR varies widely (and I put that word in quotations because most are just Powerpoint presentations). Only a handful will eventually appear in academic journals, and those that do are at the beginning of a long peer review process that may ultimately lead their authors to different conclusions.

Finally, some words of thanks: First, thanks to all of those we interviewed for making themselves available. Second, to all of the folks at AAPOR for their logistical support. Third, to Lisa Mathias at the Winston Group for creating the animation that appears at the beginning of each video. Finally, last but definitely not leas, a huge personal thank you to the Winston Group's Kristen Soltis for sharing the interviewing duties. I got considerably more sleep this weekend than I would have otherwise, and our interviews are made that much better by having a newcomer's perspective. Thank you Kristen and thank you all!

IL: 2010 Sen (Giannoulias 5/11-13)

Topics: Illinois , poll

Greenberg Quinlan Rosner (D) for Alexi Giannoulias
5/11-13/10; 600 likely voters, 4% margin of error
Mode: Live telephone interviews
(Sun-Times story)


2010 Senate
44% Giannoulias (D), 44% Kirk (R) (chart)

US: National Survey (Ipsos 5/6-9)

Topics: National , poll

Ipsos / McClatchy
5/6-9/10; 1,016 adults, 3.1% margin of error
905 registered voters (for generic ballot question)
Mode: Live telephone interviews
(Ipsos release)


State of the Country
34% Right Direction, 60% Wrong Track (chart)

Obama Job Approval
52% Approve, 43% Disapprove (chart)
Dems: 84 / 14 (chart)
Inds: 40 / 47 (chart)
Reps: 18 / 80 (chart)

2010 Congress: Generic Ballot (chart)
All respondents: 46% Democrat, 39% Republcan
Registered voters: 45% Democrat, 42% Republican

Which of these comes closer to your view?
60% Offshore drilling is necessary so that America can produce its own energy and not depend on other countries for oil
35% Offshore drilling is a bad idea because of the risks to the environment

As you may know, the state of Arizona recently passed a new law that once police stop a person, requires local law enforcement officials to verify the immigration status of people they reasonably suspect of being in the country illegally. Do you strongly favor, somewhat favor, somewhat oppose or strongly oppose this new Arizona State law.
61% Favor, 36% Oppose

Party ID
31% Democrat, 23% Republican, 46% independent (chart)

FL: 2010 Sen (Rasmussen 5/16)

Topics: Florida , Senate

5/16/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Senate
39% Rubio (R), 31% Crist (i), 18% Meek (D) (chart)

Favorable / Unfavorable
Charlie Crist: 57 / 41 (chart)
Kendrick Meek: 33 / 36
Marco Rubio: 46 / 43

Job Approval / Disapproval
Pres. Obama: 50 / 49 (chart)
Gov. Crist: 57 / 42 (chart)

TX: 2010 Gov (Rasmussen 5/13)

Topics: poll , Texas

5/13/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)


2010 Governor
51% Perry (R), 38% White (D)

Favorable / Unfavorable
Rick Perry: 55 / 43
Bill White: 47 / 41

Job Approval / Disapproval
Pres. Obama: 42 / 57
Gov. Perry: 55 / 42

PA-12: 2010 House Special (PPP 5/15-16)

Topics: poll

Public Policy Polling (D)
5/15-16/10; 831 likely voters, 3.4% margin of error
Mode: Automated phone

(PPP release)

Pennsylvania 12th Congressional District

2010 House Special Election
48% Burns (R), 47% Critz (D)

KY: 2010 Sen Primary (PPP 5/15-16)

Topics: Kentucky , poll

Public Policy Polling (D)
5/15-16/10; 1,065 likely Republican primary voters, 3% margin of error
Mode: Automated phone
(PPP release)


2010 Senate: Republican Primary
52% Paul. 34% Grayson, 4% Martin, 2% Stephenson, 1% Scribner (chart)

PA: 2010 Sen, Gov (Quinnipiac 5/12-16)

Topics: Pennsylvania , poll

5/12-16/10; 951 likely Democratic primary voters, 3.2% margin of error
Mode: Live telephone interviews
(Quinnipiac release)


2010 Senate
42% Sestak, 41% Specter (chart)

2010 Governor
39% Onorato, 11% Williams, 10% Wagner, 9% Hoeffel (chart)

AAPOR 2010: Courtney Kennedy

Some of the items I was most interested in learning about at AAPOR this year were the findings of the task force on cell phone survey research. Given my particular interest in understanding young voters, I am particularly concerned about the lack of coverage among that group with landline-only research. The task force presented a number of findings that acknowledged the increased costs and challenges of cell-phone sampling (cognitive shortcutting, potential risks to respondent safety, response rates, etc.) Courtney Kennedy's award-winning student paper on whether or not cell phone respondents employ cognitive shortcuts when responding to surveys. Essentially, are respondents paying attention and giving us good data? As the recipient of the Seymour Sudman student paper award, Kennedy tackled a critical question in understanding how to conduct better research using this sample frame. I was lucky enough to be able to ask her about her work for a moment at the AAPOR conference this weekend in Chicago.

AAPOR 2010: Pew's Jocelyn Kiley

Topics: AAPOR 2010 , Cell Phones , Hispanics , Pew Research Center

One of the things I loved most about the AAPOR conference was the opportunity to learn from pollsters of different disciplines. The lessons one organization learns about how to reach a unique population are often useful to researchers of all varieties. In this case, Pew presented its findings about how best to reach Hispanics in general public opinion surveys. From issues in language and translation to interviewer hand-offs to the prevalence of cell phone use, Pew's findings highlighted the challenges in ensuring Hispanics are properly represented in survey research. For campaign pollsters, particularly those operating in states with a high proportion of Hispanic voters, knowing how to get a representative snapshot is becoming more and more critical to monitoring political attitudes. I had a chance to chat with Jocelyn Kiley about the research and its importance to political polling.

AAPOR 2010: Chris Wilson and Bryon Allen

Topics: AAPOR 2010 , ANES , independents , Turnout

While AAPOR's panels are predominantly comprised of academics and professional non-partisan researchers, it was nice to run into a handful of political pollsters who had presentations as well. Chris Wilson and Bryon Allen from Wilson Research Strategies dug into the ANES data to answer a basic question: what matters more, persuasion or turnout? Is it more critical to move the middle or to energize your base? Their research points to persuasion as key. I pulled them aside for a moment before their panel to find out about their research.

AAPOR 2010: Washington Post's Jennifer Agiesta

Topics: AAPOR 2010 , election results , Jennifer Agiesta , third parties

On the second day of the AAPOR Conference in Chicago, I had a chance to catch up with Jennifer Agiesta of the Washington Post who chaired a panel session on candidate preferences and election outcomes. The panel featured presentations on a number of topics that impact how public opinion measures elections, including third parties and whether or not having your name at the top of the ballot gives you a distinct advantage. I caught up with Jennifer after the panel to chat with her about the presentations and to find out what challenges she's anticipating as a media pollster heading into the 2010 elections.

AAPOR2010: George Bishop

Topics: AAPOR , AAPOR 2010 , George Bishop , Measurement

One last AAPOR interview (from me) for tonight: On Saturday, I talked to University of Cincinnati Political Science Professor George Bishop, who presented a paper on interpreting and misinterpreting public opinion trends, about recent polling on Arizona's immigration law.

I wrote about Bishop's well known work on survey questions on fictitious issues in a National Journal column this past November.

We are conducting interviews this week with participants at the annual conference of the American Association for Public Opinion Research (AAPOR). The following links take you to all the videos and occasional Twitter updates from me, Kristen Soltis and others at the conference.

AAPOR2010: Robert Erikson

Topics: AAPOR , AAPOR2010 , Robert Erikson

This latest interview from the AAPOR Conference in Chicago features Columbia University Political Science Professor Robert Erikson discussing his paper on individual level stability and change in presidential elections from 1952 to 2008.

Long-time Pollster readers may remember the guest contributions from Erickson includes an article he co-authored in 2006 with Joseph Bafumi and Chris Wlezien on forecasting House seats from Generic and Congressional polls and another in 2008, co-authored with Karl Sigman, on SurveyUSA's 50 State Polls and the Electoral College.

We are conducting interviews this week with participants at the annual conference of the American Association for Public Opinion Research (AAPOR). The following links take you to all the videos and occasional Twitter updates from me, Kristen Soltis and others at the conference.

AAPOR2010: David Rothschild

Topics: AAPOR , AAPOR2010 , David Rothschild

This latest interview from the AAPOR Conference in Chicago features David Rothschild, a PhD candidate at the Wharton School of the University of Pennsylvania, discussing his paper on using voter expectations to forecast election outcomes.

We are conducting interviews this week with participants at the annual conference of the American Association for Public Opinion Research (AAPOR). The following links take you to all the videos and occasional Twitter updates from me, Kristen Soltis and others at the conference.