Mark Blumenthal

POLLSTER UPDATE: Rick Scott's Lukewarm Improvement

2013-06-18T16:14:02-04:00

Rick Scott gets some good news in Florida, if good news means trailing by 10 points and a "re-elect" of just 35 percent. The growth of cell-phone-only households show no signs of slowing, and we are shocked, shocked, shocked that the number of “big data" political consultants is growing even faster. This is the HuffPost Pollster update for Tuesday, June 18, 2013.

IN FLORIDA, SCOTT'S POSITION WEAK BUT IMPROVING - A new Quinnipiac University survey of registered voters in Florida has mixed results for Republican Governor Rick Scott. One the one hand, Scott has narrowed the lead of his strongest potential opponent and has increased his job approval rating to his "best ever" as measured by Quinnipiac. On the other hand, Scott's approval rating stands at just 43 percent, and he trails both "Republican turned independent-turned-Democrat" Charlie Crist (37 to 47 percent) and Democrat Bill Nelson (38 to 48 percent) in hypothetical contests.[Quinnipiac]

A 'killer' reelect? On Twitter, Guardian polling blogger Harry Enten noted a "key number" in the Quinnipiac Poll: "50% think Scott doesn't deserve re-election. Only 35% think he does. That's killer." [@ForecasterEnten]

Several Republican pollsters took issue with Enten's characterization:

-Logan Dobson: "Lots of people win re-election with 50+% thinking they don't deserve re-election. Because that's a theoretical concept, not a ballot choice." [@Logan Dobson]

-Jim Hobart: "Yup, and the question is even less predictive than it once was, due to current dissatisfaction with government (state and fed)." [@HobartPOS]

-Matt Dabrowski: "[Quinnipiac] reads closer to a hard re-elect. 35% the traditional line. Scott not the favorite but I wouldn't count him out yet." [@MattDabrowski]

Background on the "re-elect" - Campaign pollsters have long relied on the so-called "re-elect question" to gauge support for the incumbent, especially at an early stage of the campaign when potential challengers are limited by low name recognition. For individual pollsters who have asked the same question the same way in hundreds of private surveys over the years, these measures may have some predictive value. But results from re-elect questions can be harder to interpret on public polls, partly because news media polls rarely ask them and partly because the question has many variations, particularly in the way pollsters pose the alternatives to support for the incumbent. Some, like Quinnipiac, offer "does not deserve to be reelected" as the alternative choice. Others ask "if it's time to give someone else a chance" or if the incumbent should be "replaced" or even "fired." Some break the alternative into two categories ("will you consider voting for someone else, or do you think you will vote to elect someone else?"). Each of these variants can produce higher or lower scores.

Predictive value? It is hard to gauge without a systematic evaluation of the meager public data, but as Logan Dobson points out, the Gallup poll has asked a similar question about national politics over the years and concluded that their version of the question "is not highly predictive" of presidential elections 18 months or more in advance. For example, in October 1994, only 38 percent of adults said President Bill Clinton deserved reelection. Two years later he defeated Republican challenger Bob Dole, 49 to 41 percent. [Gallup]

CELL-PHONE ONLY POPULATION TICKS UP AGAIN - A report released on Tuesday by the Centers for Disease Control (CDC) estimates that 36.5 percent of U.S. adults live in a household that has only mobile telephone service and no landline. CDC has tracked questions about telephone service usage since 2003 on a high quality, in-person survey known as the National Health Interview Survey (NHIS), because the agency tracks health behaviors with telephone surveys. Their tracking has shown a steady, relentless increase that now extends to more than half of adults under age 35 and 62.1 percent of those between the ages of 25 and 29. It also includes 50.5 percent of Hispanic or Latino adults, 59.7 percent of those who rent their homes and 54.3 percent of those living below the Census Bureau's poverty threshold. [CDC]

Undermines landline-only polling - National Journal Hotline polling editor Steven Shepard explains the import for polling: "Polls that rely too much on landline participation -- or call only landlines -- are going to reach far too many older and whiter voters, and not nearly enough younger or Hispanic voters. 'The potential for bias due to undercoverage remains a real threat to surveys conducted only on landline telephones,' the CDC report states...The increases among middle-aged Americans are particularly troubling for public polling. While only 23.9 percent of those aged 35--44 were cell-only at the end of 2009, that percentage has skyrocketed since. In 2012, a whopping 43.5 percent of Americans aged 35--44 were cell-only, and as those cord-cutting Americans move into their late 40s and 50s, it's expected that landline-substitution percentages among those cohorts will accelerate as well. In 2012, 28.4 percent of Americans aged 45--64 were cell-only, up from 14.9 percent in late 2009. [National Journal]

The move to interviews over cell phones - As Shepard explains, "It's illegal for automatic dialers to call cell phones, so those numbers must all be dialed by a human being, which is more costly. Some pollsters have made significant commitments to contacting cell-only Americans, despite the costs." These include most of the national live-interviewer media polls, including Gallup, Pew Research, ABC/Washington Post, CBS/New York Times, NBC/Wall Street Journal, CNN/ORC, Fox News, AP/GfK and the statewide polls conducted by Quinnipiac University and Marist University. However, "the FCC regulations prohibiting automated dialers from calling cell phones also means that automated polls, like the ones produced by Public Policy Polling on the Democratic side, and the newly-founded Harper Polling on the Republican side, exclude [cell-phone only] respondents entirely." Other automated polls including Rasmussen Reports and Survey USA "[use] the Internet to reach cell-only voters at a lower cost than live interviewer." [ibid]

COOK POLITICAL LISTS POLITICAL 'BIG DATA' CONSULTANTS - On Tuesday, the Cook Political Report unveiled a "Big Data At-A-Glance" feature, a extensive "directory of who's who and who's where in the fast-reorganizing, quickly capitalizing world of political Big Data." An introduction by Elizabeth Wilner includes first reports of two new firms: "The latest Democratic firm on the scene is BlueLabs. Formed by Chris Wegrzyn, Elan Kriegel and other senior Obama analytics alumni, it boasts the heads of predictive modeling, analytics technology, polling and battleground states analytics from the campaign's storied 'Cave.'...The newly formed Deep Root Analytics brings together some of the [Republican] party's savviest thinkers about media targeting, including Romney 2012 analytics chief Alex Lundry, the leadership of Lundry's firm TargetPoint Consulting, and Bush 2004 top targeter Sara Taylor Fagen. [Cook Political]

AMERICANS AGAINST ARMING SYRIAN REBELS - HuffPost's Emily Swanson: “Most Americans continue to oppose sending weapons to Syrian rebels after the Obama administration announced it would do so last week, according to several new polls. And a HuffPost/YouGov poll finds that even among those who believe that the Syrian government has used chemical weapons against rebels, more people oppose sending weapons than support it. According to the new HuffPost/YouGov poll, 53 percent of Americans say that the U.S. military should not provide weapons to rebels in Syria, compared to 19 percent that agree with the arms support. Respondents to the poll also opposed U.S. air strikes in Syria (48 percent to 19 percent) and sending U.S. troops to Syria (68 percent to 8 percent).“ [HuffPost]

LESSONS FROM PAST POLLING ON PRIVACY - Mike Mokrzycki: “Pollsters jumped into action after bombshell disclosures early this month about vast National Security Agency collection of telephone and Internet records. Polling observers quickly noted widely varying results, some indicating Americans were sanguine about the surveillance and others signaling greater alarm....All these commentators noted that question wording can make a big difference. That's always true in polling but especially so on a topic like privacy vs. surveillance, when the broader context includes fear of terrorism, and when many people may not be paying very close attention to the issue." [HuffPost]

POLLSTER UPDATE VIA EMAIL! - You can receive this daily update every weekday via email! Just enter your email address in the box on the upper right corner of this page, and click "sign up." That's all there is to it (and you can unsubscribe anytime). The Pollster Updates will continue to publish online, but if you'd prefer the convenience of email, it's just one click away.

TUESDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Almost as many Americans believe "White House officials ordered the IRS to target" Tea Party Groups (47 percent) as believe :those IRS officials acted on their own without direct orders from the White House" (49 percent). [CNN]

-The Public Mind poll finds big leads in New Jersey for both Christie and Booker. [Fairleigh Dickinson University]

-Stu Rothenberg explains why he never considered the Massachusetts Senate special election a "toss-up." [Roll Call]

-Gabriel Gomez's campaign releases an internal poll showing him 7 points behind. [Gabriel Gomez]

-Ruy Teixeira agrees that Georgia could be a 2016 battleground. [Think Progress]

-Harry Enten says Rubio's immigration push is the first step of the invisible primary for "establishment" support. [Guardian]

-Germans still love Barack Obama. [Pew Research]

-Nathan Gonzales finds few competitive contests for open U.S. House seats. [Rothenberg Political Report]

-Rich Morin reviews the Bullock et. al. study showing that monetary rewards can reduce "wrong poll answers" driven by partisanship. [Pew Research]

-Mark Hugo Lopez and James Hawkins share verbatim responses from a focus group of LGBT adults that helped shape their national survey. [Pew Research]

-Nate Cohen thinks we're likely to forget about the NSA scandal, allowing for a rebound of Obama's approval. [TNR]

-Four scholars find no evidence of a likely backlash to a Supreme Court decision advancing gay rights. [Monkey Cage]

-Nate Silver disputes Politico's take on the empirical nature of truth, “prov[ing] stuff," and the Beltway bubble. [Talking Points Memo]

POLLSTER UPDATE: CNN Poll Exaggerates Barack Obama Approval Decline

2013-06-17T18:28:40-04:00

Obama's approval rating has declined since May but not nearly as much as implied by a new CNN poll. Americans are just as opposed to military aid to the Syrian rebels now as they were before Friday. And pie charts still suck. This is the HuffPost Pollster update for Monday, June 17, 2013.

OBAMA APPROVAL FALLS ON CNN POLL - A new national poll from CNN/ORC shows "President Barack Obama's approval rating dropped eight percentage points over the past month...The president's approval rating stands at 45%, down from 53% in mid-May. And 54% say they disapprove of how Obama's handling his job, up nine points from last month. It's the first time in CNN polling since November 2011 that a majority of Americans have had a negative view of the president." [CNN]

Obama's approval has been declining... - Obama's job approval ratings have been drifting steadily downward since January, eliminating the gains that occurred during the fall reelection campaign and just before the inauguration. The Pollster chart's estimate of Obama's approval rating, based on all public polls, stands at 46.7 percent as of this writing, just slightly higher than the 46.4 percent recorded in early July 2012. [Pollster]

...But other polls do not show as steep a decline as CNN - The issue is mostly that CNN's prior survey, showing Obama's approval at 53 percent, was something of an outlier. Of 36 other polls logged in May, none except CNN had Obama's approval higher than 51 percent. [Pollster]

Obama's approval is down since May - Four other pollsters -- Gallup, Rasmussen Reports, Fox News and Economist/YouGov -- have tracked Obama's approval since the NSA revelations were first published in June 6. When compared to their prior surveys in May, the other organizations all showed declines in approval of 1 to 2 percentage points (averaging -1.7 points), and two of four showed slight increases in disapproval (averaging +0.8). None show anywhere near the dramatic pattern of the two CNN surveys.

Reactions from Twitter:

-The Guardian's Harry Enten: "Dear all media organizations: really isn't a bad thing to cite an average or other company's polls. Really it's not." [@ForecasterEnten]

-Political scientist Brendan Nyhan: [Obama's approval] might be drifting down a bit - but no evidence to support a swing of eight points. [@BrendanNyhan]

-Republican pollster Glen Bolger: "Like to see big data moves in two polls before proclaiming it real, but given Obama's recent probs, the @CNNPolitics suggests a big problem." [@POSGlen]

-The Washington Post's Greg Sargent: "Odd thing about CNN poll: Obama approval down 8 points, but majorities approve of the NSA programs" [@ThePlumLineGS]

-Republican pollster Steve Lombardo: "It is interesting that the headlines a few weeks ago after polls showed Obama approval stable - saying POTUS 'unaffected' by scandals ...Now of course Obama poll rating in 'free fall'. Neither true...The reality is that #Obama has had a rough 60 days and as long as this narrative continues, we will see a rising disapproval rating." [@SteveLombardo in three tweets]

TWO POLLS SHOW OPPOSITION TO SYRIAN WEAPONS - The Pew Research Center: "Last Thursday’s announcement that the U.S. would aid the rebels has not increased public support for action, and majorities of all partisan groups are opposed. Overall, 70% oppose the U.S. and its allies sending arms and military supplies to anti-government groups in Syria; just 20% favor this. Opinion is little changed from December of last year (24% favor) and support is down slightly from March, 2012 (29% favor)." [Pew Research]

Gallup finds similar result - "The slight majority of Americans -- 54% -- disapprove of the Obama administration's decision to send direct military aid to Syrian rebels fighting against the Syrian government, while 37% approve. Those who are following the situation in Syria closely -- about half of the public -- are significantly more likely to approve of the decision than are those who are not following the situation closely, although a majority of both groups disapprove." [Gallup]

AMERICANS DIFFER WIDELY ON WHAT NSA WAS DOING - HuffPost's Emily Swanson on finding from a new HuffPost/YouGov poll: "The poll shows that many Americans think the NSA's surveillance programs go beyond what President Barack Obama and other defenders of the programs have said that they entail. Thirty-eight percent of respondents said they believe the NSA programs include the ability to listen to any American's phone calls without a warrant -- almost as many as the combined 41 percent who said either that the NSA only listens to phone calls between Americans and foreigners (15 percent) or that the NSA doesn't listen to any American's phone calls without a warrant (26 percent). Forty-two percent said that they believe the NSA programs include reading any American's emails without a warrant, compared to only a combined 33 percent who said the NSA reads emails between Americans and foreigners (15 percent) or that it can't read any American's emails without a warrant (18 percent)." [HuffPost]

USA Today/Pew Research poll finds similar results - "A majority of Americans (54%) think the U.S. government has collected data about their phone or online communications; 39% say it has probably not. When asked how they would feel if they knew that the federal government had collected data about their phone and internet activity, 63% say they would feel that their personal privacy has been violated." [Pew Research]

POLLSTER UPDATE VIA EMAIL! - You can receive this daily update every weekday via email! Just enter your email address in the box on the upper right corner of this page, and click "sign up." That's all there is to it (and you can unsubscribe anytime). The Pollster Updates will continue to publish online, but if you'd prefer the convenience of email, it's just one click away.

MONDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-In Massachusetts, Boston Globe/UNH shows Markey leading Gomez, 54 to 41 percent. [Boston Globe]

-Nate Silver rates the odds of a Gomez upset as slim. [NYTimes]

-In New York, Andrew Cuomo's favorable rating falls...to 58 percent. [Siena]

-Greg Dworkin reviews what polling does and does not tell us about public reactions to the NSA revelations. [DailyKos]

-Nate Silver now believes the NSA story may hurt Obama...with Democrats. [NYTimes]

-Dan Hopkins finds new evidence that Latino voters grew less likely to support Republicans between 2008 and 2012. [WaPost]

-Carl Bialik reports that some statisticians and security experts say the NSA data collection program is "highly inefficient and has some serious downsides." [WSJ]

-In almost all sectors of the news media, stories supportive of same-sex marriage outnumbered those in opposition over the last three months. [Pew Journalism Project]

-Dictionary of Numbers puts numbers into context with a Chrome browser plug-in. [xkcd]

-PPP is polling on animals and pets. [PPP]

-Walter Hickey explains why pie charts are useless. [Business Insider]

-Kevin Collins reminds us of one perfectly acceptable type of pie chart [@KevinCollins]

POLLSTER UPDATE: Polls Show Opposition To U.S. Military Involvement In Syria

2013-06-14T16:31:38-04:00

"Military aid to the Syrian rebels bucks public opinion. Frank Newport reminds us that opinions on complex policy issues are rarely fixed or 'ironclad.' And if you know a dad who prefers polling data to power tools, you've come to the right place. This is the HuffPost Pollster update for Friday, June 14, 2013.

PRIOR POLLS SHOW OPPOSITION TO MILITARY INTERVENTION IN SYRIA - Polling conducted in recent months suggests that President Barack Obama is bucking public opinion in opting to authorize military aid to the Syrian rebels. Polls have typically shown that as many as two thirds of Americans oppose military action in Syria, even when questions that specify the aid come in the form of "weapons" rather than more direct military intervention. [Polls by CBS/NYT, Gallup, NBC/WSJ, Fox News & HuffPost/YouGov]

Chemical weapons mention draws more support, but also skepticism - Support for intervention rises when poll questions raise the possibility of chemical weapons use. For example, an April Pew Research survey found plurality support (45 to 31 percent) for military action, "if it is confirmed that the Syrian government has used chemical weapons against anti-government groups." Similarly, a May CNN/ORC survey found 68 percent of Americans thought military action justified "if the United States were able to present evidence that convinced you that the Syrian government has chemical weapons and has used them to kill civilians in that country." However, only 16 percent on the same survey felt "certain" that chemical weapons had in fact been used (another 67 percent thought it "likely but not certain"). A HuffPost/YouGov survey in April find that while 40 percent believed the Syrian government had used chemical weapons and only 6 percent had not, a majority (54 percent) were uncertain. [Pew Research, CNN, HuffPost]

Why Americans resist involvement - Paul Donaldson, senior project director for the Republican polling firm Public Opinion Strategies, blogs an explanation for why Americans have been reluctant to support military action in Syria: "The hesitation for U.S. involvement in Syria closely matches twenty-five years of past data showing the reluctance Americans have about intervening between two conflicting factions within a foreign country, especially when there is no perceived threat to U.S. national security. Looking at other civil-wars/foreign conflicts, we see similar opposition to sending in our military." Donaldson piece reviews poll data since the 1980s showing opposition to American involvement in the conflicts in Libya, Liberia, Kosovo, Rwanda, Iraq, Nicaragua and El Salvador, but initial support for involvement in conflicts in Afghanistan and Iraq. "As atrocious as the use of chemical weapons on Syrians is," Donaldson writes, "what drives Americans to support military action is the sense that our national defense is at risk." [POS]

GALLUP'S NEWPORT ON NSA POLL CONFLICTS - Gallup Poll editor-in-chief Frank Newport compares differences between their question on the NSA surveillance programs (showing 37 percent who "approved" of the NSA's activities) with a similar question asked on a Washington Post/Pew Research survey (showing 56 percent who judged the NSA activities to be "acceptable"). He reconciles the two by focusing on a conflicted middle group: "21% of Americans initially disapproved of the program in the Gallup survey, but in response to a follow-up question said that there could be circumstances in which it would be right for the government to created a database of telephone logs and Internet communications. That yields the aforementioned 58% who might find the activities acceptable, which is very close to the 56% in the Pew/Post survey who find the activities acceptable as 'a way for the federal government to investigate terrorism.'" Newport then makes a point that bears repeating: "The key in all of this is that people don't usually have fixed, ironclad attitudes toward many issues stored in some mental filing cabinet ready to be accessed by those who inquire. This is particularly true for something that they don't think a lot about, something new, and something that has ambiguities and strengths and weaknesses. That's why we find that random samples of the public [can] react differently to a concept, depending on how they are asked about it. Again, this is not a bad thing, nor should it cast negative aspersions on the American population. As is true with the analysis of relationships and differences between variables across all fields of science, it provides us better insights and understanding." [Gallup]

FRIDAY AFTERNOON FATHERS' DAY DATA DUMP - What would Father's Day be without data? Here's a summary of some new numbers released or featured this past week:

-"Being a father in this era of changing family structures and converging gender roles means more than bringing home a paycheck or delivering punishment to a misbehaving child...Americans expect dad to be more of a moral teacher and emotional comforter than a breadwinner or disciplinarian." [Pew Research]

-"A plurality of Americans believe kids derive too little support from fathers (45%). Still, an overwhelming majority (93%) also believe it's possible to be good fathers and have successful careers at the same time." [Fairleigh Dickenson University]

-"Fathers have nearly tripled the amount of time they spend with their children, from 2.5 hours in 1965 to 7.3 hours per week in 2011. [Pew Research]

-"Dads hope for quality time with family (dinner, grilling, outing) for Father's Day (40%), followed distantly by a gift card (13%) or a tech gadget/electronics (8%)." [Ipsos]

-"The DIY-enthusiast dad is more than an American cliché -- anyone stuck for a gift on Father's Day would do well with a toolkit or some power tools. US fathers rate home improvement stores Craftsman (#1), Lowe's (#4) and Home Depot (#12) among their top twenty best-rated brands." [YouGov]

-"One in two Americans think Father's Day is shortchanged compared to Mother's Day (51%)" [YouGov]

-In the U.K., "Dads are more likely to lie about their children's school grades and achievements than mums." [Parent Dish]

POLLSTER UPDATE VIA EMAIL! - You can receive this daily update every weekday via email! Just enter your email address in the box on the upper right corner of this page, and click "sign up." That's all there is to it (and you can unsubscribe anytime). The Pollster Updates will continue to publish online, but if you'd prefer the convenience of email, it's just one click away.

FRIDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

--76 percent of U.S. voters want Congress to pass an immigration reform bill this year, and 74 percent support the main provisions in the legislation being considered. [Fox News]

-Republicans like Paul Ryan more than other potential 2016 nominees. [Gallup]

-Public Policy Polling (D) finds 71 percent of U.S. voters favor background checks for all gun sales. [PPP]

-Scott Rasmussen says distrust of government is the theme driving all recent controversies in Washington. [Rasmussen Reports]

-Harry Enten believes the NSA issue will give Republicans an "in" with younger voters. [Guardian]

-John Sides is skeptical that political parties can "rebrand." [Monkey Cage]

-Chris Cillizza recalls the "best congressional approval chart. ever." [WaPost]

-Ellen DeGeneres is the most visible gay or lesbian public figure in America. [Pew Research]

-Andrew Sullivan worries that bisexuals, at 40 percent of the Pew Research LGBT sample, "skew the data dramatically." [AndrewSullivan.com]

-Just 40 percent of Iranians think religious figures should play a large role in politics. [Pew Forum]

-IPOS (not Ipsos) tracked the Iranian election using dual samples of both landline and mobile telephones. [Ipsos]

-Political Analysis publishes an issue on big data in political science. [Oxford Journals]

--14 percent of Americans believe there is at least a small chance of a zombie apocalypse. [YouGov]

-Burr Settles uses a Twitter content analysis to plot the differences between a "geek" and a "nerd." [Slackprop via Flowing Data]

Have something to add to the Pollster Update? Send tips/quotes/stories/charts/new data/juicy miscellanea to Mark Blumenthal (mark@huffingtonpost.com) or Ariel Edwards-Levy (Ariel.Edwards-Levy@huffingtonpost.com).

CORRECTION:The original version of this article incorrectly identified Ipsos -- rather than IPOS --as the company fielding tracking surveys in Iran.

POLLSTER UPDATE: Polls Find Americans Divided On Edward Snowden's Actions

2013-06-13T17:18:10-04:00

Americans worry about threats to their privacy, yet polls find division and uncertainty about Edward Snowden's actions (unless nudged to an answer). Most Americans view sexual assault in the military as an important issue. And Haley Barbour has a polling simile you've never heard before. This is the HuffPost Pollster update for Thursday, June 13, 2013.

REACTIONS TO SNOWDEN - The HuffPost/YouGov online poll was among several recent surveys to probe opinions on Edward Snowden, the contractor who leaked information on NSA surveillance programs: "Americans with an opinion [are] divided over whether Edward Snowden did the right or wrong thing...According to the new poll, 38 percent of Americans think that Snowden releasing top-secret information about government surveillance programs to the media was the right thing to do, while 35 percent said it was the wrong thing. Twenty-eight percent said they weren't sure...But [those] paying the most attention to the leak story were more likely to say they thought Snowden had done the right thing. Among those who said they had "heard a lot" about the story, 50 percent said that they thought Snowden had done the right thing, and 34 percent said he had done the wrong thing." [HuffPost]

Other polls show division or modest support - Three other surveys show generally similar findings. A Gallup poll found Americans divided on whether Snowden did the right thing (44 percent) or the wrong thing (42 percent). A Reuters/Ipsos online poll found 31 percent willing to label Snowden as a "patriot," 23 percent as a "traitor" and 46 percent who could apply neither label. A Time poll found significantly more support for Snowden's actions -- 54 percent described the leaking as a "good thing," 30 percent described it as a "bad thing." [Reuters, Gallup, Time]

Watch wording and context - Given that Snowden and the NSA controversy remain unfamiliar to many Americans, the text of these questions can have great influence over the results. Note, for example, that Time's question asked if Snowden "did a good thing in informing the American public or a bad thing" (emphasis added), while others offered more classically balanced answer options ("right or wrong," "right thing to do or the wrong thing to do," "patriot or traitor"). Also, the Snowdon question was typically the last or nearly last question asked in each poll about the NSA surveillance issue, so it could have been influenced by questions that came earlier. Before asking about Snowden's actions, for example, the Time poll read a series of not-exactly-neutral statements: They asked respondents if they believed "the courts will overturn the program and put a stop to it," if "the government will be forced to reduce or eliminate this surveillance program," and if "we'll find out that the program is even bigger and more widespread than we know even now."

CONCERNS ABOUT PRIVACY - The National Journal's Ron Brownstein, reporting on a new Allstate/National Journal Heartland Monitor Poll on privacy, data collection and surveillance fielded before last week's disclosures about the NSA: "Most Americans fear their privacy is eroding...In an era of spreading surveillance cameras, cascading revelations about government’s collection of domestic and international communications, and online advertising triggered by users’ purchasing history, the survey said that an overwhelming majority of Americans believe that business, government, social-media sites, and other groups are accessing their most personal information without their consent. Although most Americans see many benefits in the ongoing communication revolution, the poll found a consistent current of concern about the costs to privacy that come with these advances...That ambivalence is crystallized in the survey’s finding that a majority of adults believe the explosive increase in data available to business, law enforcement, and government is a more negative than positive development...On many of these questions, young people take a more benign view of the changes than their elders. On many questions, people with a college education are also more likely to see a net benefit from the new communications opportunities than are those without degrees. And in some cases, such as the expanded use of surveillance cameras, the survey said that most Americans were willing to sacrifice privacy to advance other goals, such as enhancing security. [National Journal, topline results]

POST-PEW POLL ON SEXUAL ASSAULT IN THE MILITARY: Pew Research: "A substantial majority of Americans (81%) view sexual assault in the military as an extremely or very important issue. But the public does not believe the problem of sexual assault is a bigger problem in the military than outside it: Just 11% say it is more of a problem inside the military, 23% say it is more of a problem outside the military, while 63% say it is about the same. And while 40% say the reports of sexual assault in the military represent underlying problems with military culture, a majority (54%) says they represent individual acts of misconduct." [Pew Research]

Division on whether Congress should step in - The Washington Post's Craig Whitlock and Jon Cohen: "Mirroring a sharp split among lawmakers, 45 percent of the public says Congress should step in and change military law as a primary means to tackle the issue. About the same number, 44 percent, says the problem should be handled by military leaders within the chain of command. Few people, however, have a great deal of confidence in either the Congress or military leaders to handle the situation, but Congress fares worse. Overall, twice as many Americans say they have 'no confidence at all' in Congress."
[WaPost]

Gallup shows more confidence in military than Congress - A new Gallup survey that did not ask about sexual assault: "Americans' confidence in Congress as an institution is down to 10%, ranking the legislative body last on a list of 16 societal institutions for the fourth straight year. This is the lowest level of confidence Gallup has found, not only for Congress, but for any institution on record. Americans remain most confident in the military, at 76%." [Gallup]

PEW LGBT SURVEY - On Thursday, the Pew Research Center released a report on a new nationally representative, online survey of America’s lesbian, gay, bisexual and transgender adults: "An overwhelming share...(92%) say society has become more accepting of them in the past decade and an equal number expect it to grow even more accepting in the decade ahead. They attribute the changes to a variety of factors, from people knowing and interacting with someone who is LGBT, to advocacy on their behalf by high-profile public figures, to LGBT adults raising families." [Pew Research]

Sampled from a pre-recruited panel - From a separate "Explainer" on how Pew Research conducted the survey: "The survey’s respondents are members of a large, randomly selected and nationally representative sample of households recruited and maintained by the GfK Group, a research company we’ve worked with in the past." They used the GfK "KnowledgePanel," formerly operated by the company Knowledge Networks. Why use an online panel? "Part of the decision was driven by the fact that attempting to locate a random sample of the LGBT population with a telephone or in-person survey is very expensive, given that such a small percentage of the adult population identifies as LGBT...In addition, the online, self-administered method provides certain advantages when studying sensitive topics. One of the most important is that a respondent is able to answer our questions privately and need not worry that other people will overhear answers to the questions." [Pew Research]

Have something to add to the Pollster Update? Send tips/quotes/stories/charts/new data/juicy miscellanea to Mark Blumenthal (mark@huffingtonpost.com) or Ariel Edwards-Levy (Ariel.Edwards-Levy@huffingtonpost.com).

EARLY VOTING's 'MODEST' RISE IN 2012 - GMU political scientist and turnout guru Michael McDonald weighs in on an uptick in early voting in 2012: "The percentage of voters who cast their ballot before Election Day modestly increased from 29.7 percent to 31.6 percent from 2008 to 2012, according to a Census Bureau survey. The Current Population Survey, November Voting and Registration Supplement asks respondents if they voted on Election Day or before Election Day. Thus, the Census Bureau counts persons who voted by any means of "early voting" as voting prior to Election Day, be it by a mail ballot or in-person at an election office or special early voting polling place. The increase of 1.9 percentage points in early voting rates in the past two presidential elections is in stark contrast to the sharp rise of 9.7 percentage points from 2004 to 2008, from 20.0% to 29.7%. [HuffPost]

BARBOUR: 'POLLS ARE LIKE STREETCARS' - At a panel on immigration reform sponsored by the Bipartisan Policy Center, Gov. Haley Barbour (R-Miss.) dismissed polling saying Republicans don't want reform: "We have polling all over the lot...The polling is kind of like street cars, if you miss one there will be another one in about 10 minutes." Well, ok then. [@danmericaCNN, h/t @LoganDobson, BPC]

POLLSTER UPDATE VIA EMAIL! - You can receive this daily update every weekday via email! Just enter your email address in the box on the upper right corner of this page, and click "sign up." That's all there is to it (and you can unsubscribe anytime). The Pollster Updates will continue to publish online, but if you'd prefer the convenience of email, it's just one click away.

THURSDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Automated survey of by Republican-affiliated Harper Polling finds Markey leading Gomez, 49 to 37 percent, in Massachusetts. [Conservative Intel]

-Cook Political Report moves the Massachusetts Senate race to "lean D." [@JenniferEDuffy]

-MassInc Polling Group explains why Gabriel Gomez rarely mentions the Tea Party in Massachusetts. [MassInc]

-Cory Booker begins with huge early lead among New Jersey Democratic primary voters. [Monmouth University]

-Democratic Governor John Hickenlooper running roughly even against Republican Tom Tancredo in Colorado. [Quinnipiac]

-Three pro- reform groups release automated polls in 29 states, conducted by PPP (D) and Harper Polling (R), showing support for immigration reform. [Politico]

-Harry Enten says the NSA leaks are eroding trust in government and holding down Obama's ratings as economic confidence rises. [Guardian]

-Chris Bowers argues for ignoring NSA leak polls. [DailyKos]

-Non-Hispanic whites "recorded more deaths than births last year for the first time ever" last year. [USA Today]

-Cable television embraces big data. [LightReading, via @AlexLundry]

-U.S. Rep. Ted Deutsch (D-Fla.) conducts a "survey" on PRISM and the NSA; requires that respondents provide full name, email address and "grant Congressman Ted Deutch permission to contact you." [Ted Deutch]

-An Onion slide show delivers what it promises: 40 numbers under 40. [Onion]

POLLSTER UPDATE: Do Americans Support Surveillance? It Depends On How You Ask

2013-06-12T16:01:17-04:00

Something about today's NSA poll results seems familiar. Steve Shepard asks a question about questions. And is Rand rising? This is the HuffPost Pollster update for Tuesday, June 12, 2013.

CBS FINDS OPPOSITION TO MASS PHONE RECORDS SURVEILLANCE - A new CBS News poll finds support for NSA tracking of terror suspects but opposition to the collection of phone records of ordinary Americans: "Seventy-five percent of Americans approve of federal agencies collecting the phone records of people the government suspects of terrorist activity, but a 58 percent majority disapproves of this type of data collection in the case of ordinary Americans...When asked if the government's collection of Americans' phone call records is a necessary tool to help find terrorists, a slight majority of 53 percent say it is. There is broad support for the government monitoring the internet activities of those living in foreign countries. Seventy-two percent of Americans approve of this practice and there is agreement on this across party lines." [CBS]

Different story than Post/Pew poll. Why? - The new CBS results shed further light on why initial poll findings on the NSA phone records surveillance have varied. They show that Americans are more willing to accept surveillance targeted at terror suspects but not for "ordinary" Americans, and many may both "disapprove" of the collection of phone records while also granting that the practice is a "necessary tool." The Washington Post/Pew Research poll released on Monday, which showed 56 percent support for the NSA phone record surveillance, described the program as affecting "MILLIONS of Americans" (not all) and asked whether the program was "an acceptable or unacceptable way for the federal government to investigate terrorism." [Pew Research]

Gallup finds results similar to CBS - "More Americans disapprove (53%) than approve (37%) of the federal government agency program that as part of its efforts to investigate terrorism obtained records from U.S. telephone and Internet companies to 'compile telephone call logs and Internet communications.'...These results are from a June 10--11 Gallup poll. Although the current survey context was different, these results are similar to those obtained in a May 2006 Gallup poll measuring support for a government program that 'obtained records from three of the largest U.S. telephone companies in order to create a database of billions of telephone numbers dialed by Americans.' In that survey, 43% approved and 51% disapproved." [Gallup]

Reuters/Ipsos: Many in the middle - A Reuters/Ipsos online poll also released on Wednesday found that while "nearly half of all Americans [53 percent) say the U.S. government's broad surveillance tactics are acceptable within limits," only 19 percent said their were either "completely acceptable" or "acceptable under most circumstances." Far more (35 percent) said the phone records surveillance was "acceptable under only a few circumstances," and 37 percent said they were "completely unacceptable." [Reuters]

Deja vu - Something eerily similar happened seven years ago, when an earlier story broke about the NSA gathering domestic telephone records. Media pollsters tried to gauge reactions, and seemingly minor differences in questions produced widely varying results. Back in 2006, we offered this take: "If we read a bit beyond the face value of each poll result, we can see evidence of a large number of Americans that feel strongly cross-pressured on issues of protecting privacy and investigating terrorism...The most clear-cut evidence of the underlying cross-pressure comes from two questions with nearly identical wording... asked on last week's CBS News poll: 'To reduce the threat of terrorism,' 69% said they would be willing to allow the government to monitor the communications of 'Americans the government is suspicious of.' At the same time, on a question with nearly identical wording, only 30% would be willing to allow the government to monitor the communications of 'ordinary Americans on a regular basis' in order to combat terrorism. The lesson we could all stand to learn here is that on issues of public policy no single question provides a precise, 'scientific' measure of the truth. The most accurate read of public opinion usually comes from comparing the sometimes conflicting results of many different questions on many different polls and understanding the reasons for those differences." [CBS.com, h/t @DemFromCT]

WHAT QUESTION IS BEST? - Via Twitter, Steve Shepard, polling editor at National Journal's Hotline, asks a perennial question prompted by the NSA polls: "What's more useful: Asking short questions that only those following issue will answer, or...asking more explanatory questions to probe public opinion, even if question design adds subjectivity or complexities?" [@SteveShepard compiled with Storify]

A sampling of pollster responses:

@hotlinesteve not either or...both have a place— Mark Mellman (@MarkMellman) June 11, 2013

@hotlinesteve test for level of awareness, BRIEF support/oppose, then opposing viewpoint q.— B.J. Martino (@bjmartino) June 11, 2013

@hotlinesteve @mysterypollster Doing both, short then detailed, can also be a good option in the right situation (and if we have the time)— Brian Stryker (@BrianStryker) June 12, 2013

@mysterypollster @markmellman @hotlinesteve Both have place, depends on purpose: understand attitudes/causes vs estimate split in pop (1/2)— Natalie Jackson (@NatalieMJ84) June 12, 2013

@mysterypollster @markmellman @hotlinesteve (2/2) Often seems academics tend toward former, media tends toward latter--need more crossover.— Natalie Jackson (@NatalieMJ84) June 12, 2013

2016 CHECK-IN: IS RAND RISING?

SCOTT BROWN, JOSEPH KENNEDY II WOULD TIE IN MASS. GOVERNOR RACE...IF EITHER WERE RUNNING - HuffPost: “Former Sen. Scott Brown (R-Mass.) and former Rep. Joseph P. Kennedy II (D-Mass.) would nearly tie in the 2014 Massachusetts governor's race, according to polling released Tuesday. An online University of Massachusetts Amherst/YouGov poll found that Brown would barely edge out Kennedy, 42 percent to 41 percent. Brown carried wider leads over two other Democrats, state treasurer Steve Grossman (55 percent to 26 percent) and Rep. Mike Capuano (45 percent to 33 percent). The poll also included matchups with two other Republicans: former state Sen. Richard Tisei and 2010 gubernatorial candidate Charlie Baker. Kennedy led Tisei by 40 percent to 33 percent, and Baker by 48 percent to 25 percent.”
[HuffPost]

WEDNESDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Most people think the NSA is lying about the scope of its phone tracking. [YouGov]

--45 percent of Americans give Barack Obama a "poor" rating for protecting civil liberties. [HuffPost/YouGov]

-Mark Mellman reviews the complexity of public opinion on the NSA surveillance. [The Hill]

-Seth Masket sees no hypocrisy in partisans shifting their reactions, since "it's hard for poll respondents to evaluate NSA surveillance...without considering who's in charge of it." [MischiefsOfFaction]

-Voteview reviews Nate Silver's take on the intra-party congressional divide on domestic surveillance. [Voteview]

-76 percent of Americans oppose race-based college admissions. [WaPost]

-Gridlock is the reason Americans offer for why they're down on Congress. [Gallup]

-Chris Christie’s job approval is at 70 percent among New Jerseyans. [Rutgers-Eagleton]

-David Hill sees a parallel between George Wallace and Republicans standing in the path of an immigration deal. [The Hill]

-Republican pollster Matthew Jason reviews recent polling on whether undocumented immigrants should be allowed to stay in the U.S. [POS]

-Quinnipiac is now conducting polls in Iowa and Colorado. [Quinnipiac]

-Nathan Yau plots data on what men and women want from speed dating. [Flowing Data]

-Private data being collected by someone other than advertisers? Outrageous. [The Onion]

POLLSTER UPDATE: What The NSA Polling Doesn't Tell Us

2013-06-11T17:50:00-04:00

Today we look at why two initial polls on the NSA surveillance controversy appear to tell different stories about where the public stands and ask, what if privacy 'hawks' don't trust pollsters to protect their privacy? And you want more Simon Jackman? You got him! This is the HuffPost Pollster update for Tuesday, June 11, 2013.

TWO NSA POLLS, TWO STORIES - Yesterday's update included results from a new Pew Research survey showing that a majority of Americans (56 percent) consider the NSA's tracking of phone records "acceptable." Late Monday afternoon, we published results from a new HuffPost/YouGov online poll showing a different result: "According to a new HuffPost/YouGov poll, 55 percent of Americans think collecting and analyzing Americans' phone records is 'an unnecessary intrusion into Americans' lives,' while 22 percent say that it is 'justified to combat terrorism.' By a 41 percent to 25 percent margin, a plurality of poll respondents said that collecting and analyzing Americans' phone records is an ineffective way of combating terrorism." But the two surveys were consistent in finding that most Americans are so far not closely following the story: Only 36 percent of respondents to the HuffPost/YouGov poll said that they had heard a lot about the NSA obtaining phone records of Verizon customers, while another 40 percent said they had heard a little and 24 percent said that they had heard nothing at all. Similarly, in the Post/Pew survey, just 27 percent of respondents said that they were following the story "very closely." [HuffPost]

Why the difference? - As the Pew Research Center's Michael Dimock recently argued, public inattention to a complex issue creates an environment in which relatively small differences in question wording can produce very different reactions. And in this case, the wording was different: "The Post/Pew question pointed specifically to a court order (albeit a secret one) and said that the program tracked 'millions of Americans.' The HuffPost/YouGov poll, on the other hand, noted in a previous question that the NSA was obtaining the records of all customers of Verizon, before asking if the program was 'justified to combat terrorism' or 'an unnecessary intrusion.'...The HuffPost/YouGov question asking whether the program was justified or an unnecessary intrusion also followed a question on whether tracking Americans' phone records was an effective or ineffective way of combatting terrorism, perhaps calling into question for respondents whether the program was was worthwhile, given its potential effectiveness. [Ibid]

Wording differences in Post/Pew poll - Columbia University Prof. Andrew Gelman: "At first glance, support for the surveillance seems slightly higher than before, with 51% supporting it in 2006, and 56% supporting it now. But look carefully at the questions: In 2006: 'secretly listening ... without court approval.' In 2013: 'getting secret court orders ...' So, more people support wiretapping now—-but the survey stipulates that the NSA got court orders. Sure, they're 'secret' court orders, but it means that a judge is somewhere in the loop. In contrast, the 2006 poll asked about extrajudicial wiretapping. On the other direction, the 2013 question refers to 'millions of Americans,' whereas the 2006 question asks about a more restricted class: 'people suspected of terrorist involvement.'" [The Monkey Cage]

PRISM polling may be different - Nate Cohn: "The Pew/Post poll did hint, though, that the public might not support PRISM—which includes email snooping—as much as the NSA's collection of phone records, which does not involve eavesdropping. Pew asked whether Americans thought the 'the U.S. government should be able to monitor everyone's email and other online activities if officials say this might prevent future terrorist attacks,' and found that 52 percent were opposed, compared to 45 percent in support. That's a net--22 points less support than the NSA's phone program, which was opposed by 41 percent of respondents." [TNR]

Twitter reactions to NSA polling:

I would like to scream at everyone comparing the 2006 & 2013 poll Qs on surveillance: THE QUESTIONS WERE WORDED DIFFERENTLY!!! It matters.— Natalie Jackson (@NatalieMJ84) June 11, 2013

The fact that officials won't really tell the public what they are doing makes me skeptical of polls saying the public is okay with it.— Nick Confessore (@nickconfessore) June 11, 2013

Non-tweeting/polling friend has good Q: Might polling on privacy/surveill. be espy prone to diff. non-response? Do privacy hawks take polls?— Mike Mokrzycki (@mikemokr) June 11, 2013

DO PRIVACY 'HAWKS' PARTICIPATE IN SURVEYS? That last question by pollster Mike Mokrzycki is worth considering. All surveys require participation from respondents who share private attitudes and demographic details on the usually implicit promise that their confidentiality will be protected. Is it possible that the Americans most concerned about invasions of their privacy are less likely to participate in surveys? And if so, how would we know? This issue has been considered in depth by methodologists investigating whether such concerns might help explain a long term decline in survey response rates. Reviewing available experimental evidence for a chapter published in 2006, scholars Eleanor Singer and Stanley Presser found that concerns about privacy and confidentiality helped to reduce cooperation with the Census and other government surveys, but concluded that "the net effects of such attitudes, although statistically significant, are relatively small." [Advances in Telephone Survey Methodology]

Evidence of privacy 'bias'? - We put the question to Scott Keeter, director of survey research at the Pew Research Center, who has led a series of studies on the problem of declining response rates in surveys. "It's a reasonable theory," he replied via email, "but I am aware of only modest evidence in support of it." Keeter flagged the Singer/Presser review above, then summarized findings from Pew Research: "In our 2003 nonresponse study, we included a question that read, 'How much do you worry that computers and technology are being used to invade your privacy?' We didn't find any difference between the lower and higher response rate versions in answers to the question. (Question 13 here). Our 2012 nonresponse study found big differences between in estimates of sociability and volunteer activity from our survey and identical questions asked on the Current Population Survey (we found much higher levels on both variables), but it's not clear to me that those measures are capturing concerns about privacy, at least not directly. [Pew Non-Response studies from 2003 and 2012]

WHY DEMOGRAPHICS ALONE MIGHT NOT GIVE THE GOP A 2014 BUMP - Harry Enten: “Did Democrats do so poorly in 2010 because of poor turnout among their core constituencies: minorities and young voters? After all, the percentage of eligible voters that turns out drops by about 20pt between midterm and presidential years. Are most of those minority and young voters? If so, Republicans would have fared far worse in 2010 with presidential year turnout and will do well in 2014....Combining the impact of race and age, you'd be looking at a 3pt Republican gain between the 2008 and 2010 electorate. That still would have been enough for the Republicans to win more than 45 seats and win back the House. Keep in mind, though, the largest portion of the electorate that is 'minority' are youth. Putting together the age and race effects, therefore, may make the presidential and midterm electorates seem more different than they are...Pew data suggest that Republicans are actually not gaining an advantage; or if there is one, it's minimal. My own estimate is that it's probably about 2pt when looking at all the data. That's consistent with the idea that the current party affiliation differences between age groups has disadvantaged Democrats in midterm elections, while the differences in turnout by different racial groups likely don't have much of an effect." [Guardian]

GEORGE W. BUSH'S RATINGS IMPROVE - HuffPost: “For the first time since 2005, more Americans now view former President George W. Bush favorably than unfavorably, according to a Gallup poll released Tuesday. Forty-nine percent have a favorable view of Bush, while 46 percent view him unfavorably, the poll found. His ratings have risen by more than 10 points among both parties since he left office, with 84 percent of Republicans and 24 percent of Democrats now rating him favorably. At Bush's lowest point, in 2008, just 32 percent of Americans rated him positively, according to Gallup. Like most recent presidents, however, he saw his numbers rise after leaving the White House.“ [HuffPost]

Not a surprise:

Re-upping: There is nothing surprising about GW Bush's rising approval rating. The classic ex-prez pattern. fivethirtyeight.blogs.nytimes.com/2013/04/25/pre…
— David Leonhardt (@DLeonhardt) June 11, 2013

NEWS ABOUT US - Some good news to report from the Pollster home office: Stanford political scientist Simon Jackman, who created the poll averaging model for HuffPost Pollster that correctly forecast the outcome of the presidential election in 51 out of 51 contests (all 50 states plus D.C.) is officially on-board to continue in that role through 2016. Look forward to seeing more of his work on the The Huffington Post! [51 out of 51]

TUESDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-MassInc/WBUR poll finds Markey leading Gomez, 46 to 39 percent. [WBUR]

-Support for affirmative action hits a historic low. [NBC/WSJ]

-Nate Silver sees the domestic surveillance controversy dividing both parties in the 2016 primaries. [NYTimes]

-The American National Election Studies releases its time series data for 2012. [ANES]

-Seth Masket speculates about how the NSA uses call metadata. [Pacific Standard]

-A graphic designer redoes the PRISM slides. [SlideShare]

-Finding Paul Revere with metadata. [Kieran Heily via @AlexLundry]

POLLSTER UPDATE: Booker Leads In First New Jersey Special Election Polls

2013-06-10T16:57:17-04:00

Booker leads by a mile in two new New Jersey polls. Markey leads by smaller margins on five June polls in Massachusetts. And Pew Research finds a majority view the NSA phone tracking as acceptable, but your pollster's mileage may vary. This is the HuffPost Pollster update for Monday, June 10, 2013.

FIRST NEW JERSEY SENATE POLLS PUT BOOKER AHEAD - HuffPost: “Newark Mayor Cory Booker has an overwhelming lead in the Democratic primary for New Jersey's open senate seat, according to the first polls released since a special election was set for later this year. In a Rutgers-Eagleton poll, Booker took 55 percent of the primary vote, followed by Rep. Frank Pallone (D-N.J.) at 9 percent and Rep. Rush Holt (D-N.J.) at 8 percent. A Quinnipiac poll had similar results, with Booker leading at 53 percent of the primary vote, and Holt and Pallone trailing at 10 and 9 percent, respectively. Neither poll included state Assembly Speaker Sheila Oliver, who announced her candidacy after the surveys were taken." [HuffPost]

Pollster's chart gives Booker an average of more than 50 percent among Democrats [Pollster]

PEW: MAJORITY SEES NSA PHONE TRACKING AS 'ACCEPTABLE' - A new Pew Research/Washington Post poll released Monday afternoon and conducted since Thursday is one of the first with reactions to the NSA revelations: "A majority of Americans – 56% – say the National Security Agency's (NSA) program tracking the telephone records of millions of Americans is an acceptable way for the government to investigate terrorism, though a substantial minority – 41% – say it is unacceptable. And while the public is more evenly divided over the government's monitoring of email and other online activities to prevent possible terrorism, these views are largely unchanged since 2002, shortly after the 9/11 terrorist attacks." But, perhaps more important, many Americans are not yet following the story: "Roughly a quarter (27%) of Americans say they are following news about the government collecting Verizon phone records very closely. This is a relatively modest level of public interest. Only another 21% say they are following this fairly closely, while about half say they are following not too (17%) or not at all (35%) closely." Remember, when pollsters have to describe the facts of a story to ask about it, their results measure reactions to the text of the questions as much or more than existing opinions. So watch the wording carefully. [Pew Research]

Pollsters doubt 'public revolt' brewing - Politico's James Hohmann talks to pollsters who foresee a "meh" response from the public coming over recent disclosures of a top secret U.S. surveillance program: “The outrage is coming from the people who write, but not the people who vote," said Democratic pollster Jefrey Pollock, president of Global Strategy Group, adding that the type of surveillance revealed this week is seen as “a necessary evil."
“People are willing to kind of bite the bullet a little bit if it helps stop terrorist attacks," said Republican pollster Ed Goeas of the Tarrance Group...“I wouldn't want to minimize the concern over privacy at all because it's definitely there. But at the same time, especially in the wake of Boston and the constant threat people are feeling...protection is foremost," said Carroll Doherty, associate director of the Pew Research Center for the People and the Press. “In this general tradeoff, when push comes to shove...more people consistently since 9/11 said protecting the country is a greater concern than restricting civil liberties." [Politico]

let's just do a/b testing for "hero" and "traitor" in headline descriptors— Shani O. Hilton (@shani_o) June 10, 2013

On the Gallup watch, Obama stands at +5 net approval... which is basically the same as the day before that, and the day before that, etc.— Harry Enten (@ForecasterEnten) June 10, 2013

MASSACHUSETTS: MARKEY LEADS - With a special election to fill the U.S. Senate seat vacated by John Kerry just two weeks away, most polls are showing Democratic Rep. Ed Markey leading Republican nominee Gabriel Gomez. On a new Suffolk University telephone poll released on Monday, Markey has a seven percentage point lead (48 to 41 percent), while a new UMass Amherst/YouGov online poll shows Markey leading by 11 (51 to 40 percent). All together, Markey has a nominal lead on all five surveys conduct in June. On the Pollster chart, which combines results from all of the public poll, Markey leads by nearly seven percentage points (48.6 to 41.7 percent), a margin that is essentially unchanged over the past month. [Suffolk, UMass Amherst, Pollster chart]

NOT POLLING LIKE 2010 - Boston journalist David Bernstein finds "not much evidence" that "the Massachusetts Senate race is as close as (or closer than) the January 2010 special election in which Scott Brown defeated Martha Coakley." Why not? "The problem with looking at the 2010 polling—and a big reason why Scott Brown's surge snuck up on so many of us—is that nobody was bothering to poll the race. That was compounded by some really, really bad polling." Bernstein notes that the two most optimistic results in the two weeks of the 2010 campaign, from Coakley's perspective, were produced by Research 2000, "which would later be disgraced, accused of fabricating results....Either way, if the idea is that Gomez is following a Brown-like trajectory, this is the week he needs to prove it. If he isn't showing up even or ahead in some polls by then—polls other than McLaughlin—he'll have to pull off a much bigger come-from-behind than Brown did." [Boston Magazine]

Suffolk not the only word on Mass. Senate we're getting in the next 24 hours -- New @wbur/@massincpolling poll tomorrow.— Steven Shepard (@HotlineSteve) June 10, 2013

MONDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Hillary Clinton's favorable rating slips since April, but stands at 94 percent with Democrats. [Gallup, ForecasterEnten]

--18 percent of Americans mention jobs as the most important issue, down from 29 percent in 2010. [Gallup]

-For the first time, a third of American adults own tablet computers. [Pew Internet]

-Rasmussen gives McAuliffe a three point edge over Cuccinelli in Virginia (44 to 41 percent). [Rasmussen Reports]

-Iowa Gov. Terry Branstad (R) leads Democratic challenger Jack Hatch by 28 points. [Des Moines Register (PDF)]

-In NYC mayoral races, Micah Cohen finds, the early leader in the polls usually wins. [NYTimes]

-Drew Linzer compares his 2012 battleground state forecasts from October 9 to those from the Obama Analytics "cave." [@DrewLinzer]

-A survey of hill staffers finds more who trust corporations than trust organized labor. [Politico]

-Carl Bialik profiles Census director nominee John Thompson and the challenges he faces. [WSJ]

-The racial demographic “tipping point,' in several charts. [AP]

-PPP polls "LOLGOP" in Michigan. [PPP]

-How Abraham Lincoln microtargeted the Appalachians. [NYT, via Mischiefs of Faction]

POLLSTER UPDATE: Americans Have Little Expectation Of Online Privacy

2013-06-07T15:57:36-04:00

Polls show that while Americans have little expectation of privacy about their phone records and online data, they're wary of growing government surveillance. Polls did "herd" in 2012, but why? And hell hath no fury like Ed Tufte reviewing ugly Powerpoint slides. This is the HuffPost Pollster update for Friday, June 7, 2013.

POLLS SHOW LOW EXPECTATIONS OF PRIVACY - Following revelations that the National Security Agency is collecting domestic phone records and electronic communications, the National Journal shared new survey results showing that Americans assume that their "communications history" are more prone than "other forms of personal information" to access without their consent: "In the poll, 85 percent of adults surveyed said it was likely that their 'communications history, like phone calls, e-mails, and Internet use,' was 'available for businesses, government, individuals, and other groups to access without your consent.' That was a higher percentage than believed that any other kind of private information, such as medical and financial records, is being obtained without their approval." [National Journal]

Most not anxious about government monitoring - Citing similar data, GWU political scientist John Sides is skeptical that a backlash against domestic surveillance is brewing: "For one, most Americans do not express much anxiety about domestic surveillance. In a recent article (gated), political scientists Samuel J. Best, Brian S. Krueger, and Shanna Pearson-Merkowitz reported the results of a 2007 survey in which they explicitly asked whether Americans were anxious, worried, or scared about “the government monitoring the activities of people like you.” Only about 30% of Americans said that they were “somewhat” or “very” anxious, worried, or scared...What would create more of a public anxiety would be a concerted pushback from Congress against the NSA, and especially a bipartisan pushback. As I wrote regarding drone attacks, real public concern about civil liberties is most likely to arise when elected leaders express concern. But instead of a bipartisan pushback, I am seeing more evidence of a bipartisan shrug." [The Monkey Cage]

Washington Post polling director Jon Cohen: "65% not worried about others knowing what websites they visit." [@jcpolls]

Yet also little enthusiasm for expanded monitoring, concerns about going too far - The new National Journal poll asked respondents about potential steps to "to improve national security." Only 10 percent supported "expanded government monitoring of cell-phone and e-mail activities. Similarly, a CNN/Time/ORC poll conducted in April found far more opposition (59 percent) than support (38 percent) for "expanded government monitoring of cell phones and email, to intercept communications." The same survey found Americans more concerned that "the government will enact new anti-terrorism policies which excessively restrict the average person's civil liberties" (61 percent) than " that the government will fail to enact strong, new anti-terrorism policies" (31 percent). It also found, by a 49 to 40 percent margin, that Americans are not willing "to give up some civil liberties if that were necessary to curb terrorism in this country." [CNN]

WATCH OUT NSA: ED TUFTE HATES YOUR POWERPOINT SLIDES:

Dreadful spy-PRISM deck sets new record for most header logos per slide: 13 washingtonpost.com/wp-srv/special… #powerpoint #ppt twitter.com/EdwardTufte/st…— Edward Tufte (@EdwardTufte) June 7, 2013

List of spy-PRISM collected information includes nearly everything, except PPT decks. No useful information at all? twitter.com/EdwardTufte/st…— Edward Tufte (@EdwardTufte) June 7, 2013

PRISM "providers": classic PPT statistical graphic: 13 logos, 10 numbers, 9 bubbles, 1 giant green arrow. #powerpoint twitter.com/EdwardTufte/st…— Edward Tufte (@EdwardTufte) June 7, 2013

POLLSTER 'CONVERGENCE' IN 2012 - On Wednesday, we reported on some recent speculation and debate about pollster "herding," or the theory that some pollsters weight their results to fall into line with polling averages, particularly late in the campaign. Steve Koczela, who leads the MassINC polling group, emailed with a few charts showing that pollsters did indeed herd together toward the end of campaign 2012. Using data downloaded from Pollster, Koczela calculated the standard deviation in the Obama-Romney margin for polls conducted during each month of 2012. The number shows the degree of variation in the results. Sure enough, the range of variation declined throughout the summer and fall and then plunged in October.

Since there were far more polls in October, Koczela also calculated the same statistic by week for the last seven weeks of the campaign. The first week in November has a lower standard deviation (1.4) than any of the last 12 weeks.

But what does it mean? As Koczela points out via email, "it is hard to separate potential herding from the greater precision that becomes possible as election day approaches." The pattern of poll results tending to converge late in the 2012 campaign is similar to what David Moore found in looking at national polling conducted in October 2008, a phenomenon that he labeled "the Convergence Mystery." Moore's post drew replies from survey "experts" published later in the online journal Survey Practice, all of whom voiced skepticism that the pattern signaled intentional modification of results to come into line with other polls. Perhaps the most compelling explanation echoes Koczela's point. Rutgers University political scientist Cliff Zukin: "Absent some major occurrence in the campaign, one would expect citizens to become more certain about their behaviors the closer it gets to election day-both in whether and for whom to vote." [Pollster.com, Survey Practice]

Have something to add to the Pollster Update? Send tips/quotes/stories/charts/new data/juicy miscellanea to Mark Blumenthal (mark@huffingtonpost.com) or Ariel Edwards-Levy (Ariel.Edwards-Levy@huffingtonpost.com).

ATTITUDES TOWARD FEMALE BREADWINNERS DON’T REFLECT NEW REALITY, DEM POLLSTER ARGUES Margie Omero: “Pew notes about two-thirds (65 percent) of moms of children under six work outside the home. Gallup routinely shows a majority of women would prefer to work outside the home rather than stay at home to raise children, "if they were free to do either." Yet Pew shows a slim majority (51 percent) feel children are "better off" with a mom at home full-time, and about a third (34 percent) say children are "just as well off." However it is men who are largely driving this result. By two to one, men say children are better off with a mom at home (57 percent better off, 29 percent just as well off). The margin among women is in single digits (45 percent better off, 38 percent just as well off). So it's not just Republican leaders who have more traditional views toward gender equality -- it may be men as a whole. As families and the workforce change, many men's views are not changing in kind.“ [HuffPost]

CLINTON’S FAVORABILITY DROPS HuffPost: “Hillary Clinton's favorability rating dropped 12 points since last year, according to a Bloomberg poll released Thursday night. Fifty-eight percent of those polled by Bloomberg now view her positively, down from a high of 70 percent in December. The former secretary of state's rating is still relatively high. HuffPost Pollster's average, which combines all publicly available polling, shows that a 52 percent majority views her favorably. But that's down from the nearly 60 percent she enjoyed in November, and lower than the rating was during her tenure as secretary of state....her numbers are now closer to where they were at some points during 2008. Clinton continues to dominate early 2016 polling. HuffPost Pollster shows her taking an average 61 percent of the Democratic primary field, with Biden, her closest rival, more than 50 points behind.” [HuffPost]

FRIDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

--58 percent of Americans disapprove of President Obama's handling of the IRS situation. [Gallup]

-More Americans see China as friendly than unfriendly to the U.S. [Gallup]

-NYC mayoral candidate Bill Thompson doesn't believe the polls, thinks someone could win without a runoff. [Capital New York]

-Nate Cohn sees little hope for Gabriel Gomez in Massachusetts. [TNR]

-Chris Cillizza and Sean Sullivan consider the end of the political fight and continuing cultural divide on gay marriage. [WaPost]

-First Read compiles their polling on favorability for 2016 Republican hopefuls. [NBC]

-Ron Brownstein explains why Republicans could win in 2014 "without solving any of the underlying demographic challenges that threaten them" in 2016. [National Journal]

-Republican pollster Robert Blizzard argues that winning over young voters to the GOP means "fixing the party’s problems with Hispanics." [POS]

-Brookings posts the video of its panel on the Politics of Marijuana Legalization, featuring Sean Trende and Anna Greenberg. [Brookings]

-Taren Stinebrickner-Kauffman says, "you know nothing, campaigners." [YouTube]

-Anthony Wells provides a British perspective on Gallup's methodology review. [UK Polling Report]

-Bob Blaemire explains how voter files have become the "big data" of modern campaigns. [Campaigns & Elections]

-RTI evaluates video interviewing via Skype. [SurveyPost]

-HUD is redesigning the American Housing Survey. [HUD User]

-Justin Wolfers uses Twitter to graph payroll growth over the past year. [@JustinWolfers]

-Mapping the rise of craft beer with interactive geospatial data. [New Yorker]

-Hans Rosling uses legos to explain population growth and climate change. [Flowing Data]

POLLSTER UPDATE: The 'Stunning' Shift On Same-Sex Marriage

2013-06-06T16:47:11-04:00

A new Pew Research poll confirms a strong continuing trend in favor of same sex marriage. Obama's internal polling analytics beat the polling averages. And the CBS/Times poll shuts down its in-house call center. This is the HuffPost Pollster update for Thursday, June 6, 2013.

MAJORITY OF AMERICANS THINK GAY MARRIAGE IS ‘INEVITABLE’ - HuffPost: “Gay marriage remains a divisive issue, but the majority of both supporters and opponents now agree it's here to stay. Polling released Thursday by Pew Research finds that 72 percent of Americans say that recognition of same-sex marriage is "inevitable," up from 59 percent in 2004. Today, 59 percent of those who oppose legalizing gay marriage believe it nonetheless will happen, and 85 percent who support legalization think it will happen.” [HuffPost]

A continuing trend - Pew Research: "Support for same-sex marriage has edged above 50% for the first time in a Pew Research Center survey. Currently, 51% favor gays and lesbians marrying legally, while 42% are opposed...As support for same-sex marriage has increased, so too has the percentage favoring legal agreements for gays and lesbians that would give them many of the same rights as married couples. Currently, 67% favor gays and lesbians being allowed to enter legal agreements that would give them many of the rights of married couples, while just 28% are opposed." [Pew Research]

Republican pollster Alex Lundry: "These Pew numbers are absolutely STUNNING. This is the most profound public opinion shift in modern political history." [@AlexLundry]

Democratic pollster Mark Mellman: "fastest changing view in history of polling MT @pewresearch: For 1st time in our polling, over half (51%) support legalizing [#GayMarriage]" [@MarkMellman]

OBAMA'S ANALYTICS OUTPERFORMED...US - Bloomberg Businessweek's Josh Green persuaded the Obama campaign to share a bit more information about the performance of their internal tracking, including an assessment of how their internal horse-race tracking stacked up against the polling aggregators: "The Obama folks make two big claims: that they were more accurate—and more accurate earlier in the race—than their competitors. To back this up, they agreed to share, for the first time, the results of their own nightly forecasting model, code-named 'Golden,' that was based on 62,000 simulations of the November election and distributed each day at 2 p.m. to David Axelrod, Jim Messina, and the rest of the campaign’s brain trust." Green also obtained an analysis showing how the Obama "Analytics forecast" outperformed the polling averages produced by Pollster, FiveThirtyEight, RealClearPolitics and TalkingPointsMemo.” [Businessweek]

CBS/TIMES: POLITICS DROVE IRS TARGETING - CBS News: "Most Americans regardless of party believe political reasons drove the Internal Revenue Service to single out for burdensome and unnecessary scrutiny some conservative groups applying for tax-exempt status, according to a CBS News/New York Times poll out Thursday. The public splits across party lines, though, about whether President Obama and his administration were involved. Sixty-eight percent of respondents- 80 percent of Republicans, 60 percent of Democrats and 66 percent of independents - said they think the IRS targeting was motivated by politics, rather than adherence to the tax code policy. But while forty-four percent think the Obama administration had a hand in the targeting, 40 percent said they believe the agency acted on its own." [CBS]

CBS shuts down call center - The methodology blurb at the end of the CBS report included an easily overlooked change that marks the end of an era in news media polling: "Data collection was conducted on behalf of CBS News/The New York Times by Social Science Research Solutions of Media, PA." Previously, interviewing for CBS News polls, including those conducted in recent years in partnership with the Times, had been fielded at a Manhattan based call center owned and operated by the television network. But industry sources confirm that the CBS has opted to shutter its in-house call center. Both the CBS and Times/CBS polls will continue as before, but with telephone interviewing conducted by outside firm SSRS. Most news media polls now subcontract interviewing to outside firms. The CBS/Times partnership was the last major media polling operation to use an in-house call center.

Marjorie Connelly remembers - Connelly, a polling editor at the New York Times, via email: "I got started in this business as an interviewer for the Times/CBS News polls when I was in college. At that time, the interviewing was still being done at The New York Times building on West 43rd Street in the advertising department – a large room with desks, telephones and a staff that cleared out by 5:00 pm and didn’t work weekends. Desks for daytime interviewing were located throughout the newsroom – we used the desks of reporters on vacation or out on assignment. It worked that way while it was still being done by pencil and paper – from 1975 to 1991. The introduction of CATI interviewing necessitated the relocation of the interviewing operation to CBS offices on West 57th Street. For many years, there has been a dedicated survey room in the Broadcast Center."

CAN GALLUP BOUNCE BACK? - The New Republic's Nate Cohn: "The most important Gallup news might not have been any innovative finding from the report, but a continued commitment to transparency. Gallup deserves considerable credit for releasing their raw data through the Roper Center for Public Opinion archives, which allowed Blumenthal to diagnose Gallup’s failings early on. Blumenthal and Ariel Edwards-Levy reported Tuesday that Gallup would make the raw data from their experiments and future surveys available through the Roper Center, which will allow analysts to confirm that their changes have yielded improvements...The survey is extremely well-funded and has huge samples, especially over a multi-week period. They’ve brought in a highly regarded team of survey methodologists, and their commitment to transparency means that analysts should ultimately be able to confirm that their efforts have paid dividends. It just might be enough to restore confidence in America’s oldest polling firm." [TNR]

Or time for alternative methods? - The Guardian's Harry Enten: "I'd look to other sources in preference to a survey that interviews some number of respondents in one survey, a different set of respondents in the next survey, and so on. One of the biggest takeaways from the American Association for Public Opinion Research (AAPOR) conference is on the usefulness of panel research. That is, you have a set number of respondents, weighted to the correct population parameters, who get interviewed over and over again. This leads to less volatility, and you can actually see how different respondents are reacting to the campaign." [Guardian]

GOP FEMALE OFFICE-SEEKERS FACE A TOUGH ROAD - Christine Matthews and Kristen Folkerts: “A majority (58%) of Americans would like to see men and women represented equally in the U.S. Congress, according to a new HuffPost/YouGov poll. Despite efforts like Rutgers University CAWP 2012 Project "20% in 2012" campaign, women still account for only 18% of Congress. While many would like to see women members of Congress at parity with men, many challenges remain to getting them even close. For Republican women, the challenges are even steeper....Republican men were far more likely than anyone else to think Congress should be mostly men. They also thought that women's interests were already fairly represented in Congress...so why bother electing women?” [HuffPost]

POLLSTER UPDATE VIA EMAIL! - With a few technical glitches now out of the way, you can receive this daily update via email. Just enter your email address in the box on the upper right corner of this page, and click "sign up." That's all there is to it (and you can unsubscribe anytime). The Pollster Updates will continue to publish online, but if you'd prefer the convenience of email, it's just one click away.

THURSDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Pat Toomey would lead Joe Sestak by 5 points in a Pennsylvania Senate rematch. [Quinnipiac]

-Americans are feeling more optimistic about the economy, but most see no job growth. [YouGov]

-Brendan Nyhan warns journalists about making too much out of one poll. [CJR]

-Andrew Levison and Ruy Teixeira say Democrats still need working class whites. [TNR]

-Ed Markey leads Gabriel Gomez by 8 points in a PPP/League of Conservation Voters poll. [Politico]

-Joe Twyman, the director of YouGov’s Baghdad office, tells the story of conducting the first survey in Iraq after the fall of Saddam. [BBC]

-Google unveils model to project a film's opening weekend revenue "with 94 percent accuracy." [Hollywood Reporter]

-Dan Wagner's Civis Analytics is hiring. [Civis]

-A dialect survey maps "pop" vs "soda" and "coke" and "y'all" vs "you guys." [Flowing Data]

POLLSTER UPDATE: New Polls A Mixed Bag For Obama

2013-06-05T16:33:05-04:00

New national polls find good and bad news for Obama. Stu Rothenberg takes Lake Research to the woodshed. And where else can you find a debate about poll aggregators? This is the HuffPost Pollster update for Wednesday, June 5, 2013.

NBC/WSJ POLL FINDS GOOD AND BAD NEWS FOR OBAMA - The good news? His overall job and favorable ratings are stable. NBC'S First Read: "According to our new NBC/WSJ poll, President Obama has absorbed the political punch from the trio of controversies (IRS/Benghazi/leak investigations) that have hit his administration in the last few weeks. His overall job-approval rating stands at 48%, up 1 point since April, and his fav/unfav rating is at 47%--40%, which is essentially unchanged since that last poll." [First Read]

The bad news? - NBC's Mark Murray: "But just 46 percent give him high marks for having strong leadership qualities (down from 53 percent in January); only 42 percent give him high marks for being honest and straightforward (down from 47 percent); and just 21 percent give him high marks for changing business as usual in Washington (down from 28 percent)." [NBC]

Also mixed view of Obama's role - The Wall Street Journal's Patrick O'Connor and Rebecca Ballhaus: "A majority of poll respondents, some 55%, said IRS scrutiny of conservative groups raised some level of doubt about the administration's 'overall honesty and integrity.' Similar shares said the same about the Justice Department's subpoena of reporter phone records and about the administration's handling of the deadly terrorist attacks on a U.S. diplomatic outpost in Benghazi, Libya...Americans at this point don't hold the president personally responsible for any of the three major controversies that have dominated Washington since early May, the Journal/NBC poll found. In each case, the share saying Mr. Obama isn't responsible or only slightly responsible was larger than those who say he is mainly or totally responsible.” [WSJ]

BLOOMBERG POLL FINDS JOB APPROVAL DROP, IRS SKEPTICISM - Bloomberg's Mike Dorning: “Almost half of Americans say President Barack Obama isn’t telling the truth when he says he didn’t know the Internal Revenue Service was giving extra scrutiny to the applications of small government groups seeking tax-exempt status. Forty-seven percent of Americans say they don’t believe Obama compared with 40 percent who say he is being truthful, according to a Bloomberg National Poll of 1,002 adults conducted May 31 through June 3...Obama’s job approval rating declined six percentage points since the last Bloomberg poll in February, to 49 percent from 55 percent, returning to its lowest level since last September....The president’s personal popularity has been more resilient. Fifty-three percent of Americans say they have a favorable impression of him, down from 56 percent who said so in February.” [Bloomberg]

Do the two polls conflict? Keep in mind that the Bloomberg had not polled since mid-February, while NBC/WSJ tracks presidential approval on a monthly basis. Since February, Bloomberg shows a six percentage point decline while NBC/WSJ shows a two-point drop (from 50 to 48 percent). Not surprisingly, the Pollster chart, which incorporates all national polls measuring Obama's approval rating, shows a trend that falls somewhere in between, falling almost three points since February 12 (from 50.4 to 47.7 percent as of this writing), and roughly one percentage point (from 48.7 percent) since May 4. [Pollster]

ROTHENBERG SAYS POLLSTER CROSSED A LINE - Election handicapper Stu Rothenberg questions whether the Democratic polling firm Lake Research Partners crossed an ethical line in seeking to "discredit" a May USC Price/Los Angeles Times survey that they claimed understated the standing of their client,Wendy Greuel, in the L.A. mayor's race she lost to Eric Garcetti. The poll proved to be accurate. "Clearly, the five Lake Research members were implying that the Times poll was little more than a thinly veiled effort to help the candidate preferred by the newspaper’s editorial page. Accusing the Times of being involved in a conspiracy to use a bogus poll to help Garcetti was nothing short of a rash, unsubstantiated accusation of bias and fraud...The charge — and the entire memo — was all the more head-scratching because Lake Research Partners never referred to its own survey. It cited other polls, but never its own data, which seemed like an obvious omission. If it had the data to contradict the Times survey, why didn’t Lake present it? That’s normally what campaigns and pollsters do. In fact, the firm didn’t have recent polling." [Roll Call, see also our review of the L.A. polling]

Rothenberg talks to Lake Research - More: "[Lake Research pollster Bob] Meadow pointed out that the memo I’ve cited was not distributed to the media but was meant primarily for Greuel’s supporters and contributors, both of whom might be deflated by the news of the LA Times poll. It sought to remind supporters that 'polls come and go' and reassure them that 'they shouldn’t panic.'" According to Rothenberg, Meadow also questioned the timing of the poll and argued that in addition to providing data to a campaign, a pollster "should also support its campaigns 'where possible.'" Rothenberg: "While I agree with Meadow that consultants can try to help their campaigns 'where possible,' I would argue that pollsters have an obligation to make certain that campaigns don’t misuse their survey data and that they, themselves, don’t become little more than spin doctors for their campaigns. Professional pollsters ought to protect their reputations first, even ahead of promoting their candidates." [ibid]

WEIGHTING TO AVERAGES? - Elizabeth Wilner's meditation on "polling addiction" that we linked to on Monday included worries about "poll aggregators," such as Pollster, RealClearPolitics and FiveThirtyEight: "[Poll Aggregators] argue that averaging together individual trial-heat results allows sound surveys to compensate for sketchier ones. Some even take pains to try and address the issue of tracking-poll volume. Their case is backed up by the relative accuracy of their averages. But their work diminishes the value of the individual sound surveys....What if some pollsters start weighting their results so they are more in line with the aggregators’ averages? As a rule, pollsters don’t want to be outliers—look at what Gallup has gone through. Over the years, we’ve seen certain pollsters produce trial heats that are outliers at first, then magically fall in line with the majority of other polls shortly before a vote. [Cook Political]

Diminishes value? - Political scientist and statistician Drew Linzer, who aggregated polls during the 2012 election at the site Votamatic.org, took to Twitter to respond: "Fair worries here, but poll aggregators 'diminish the value of individual sound surveys'? No...Also inaccurate: aggregators 'argue that averaging together trial-heat results allows sound surveys to compensate for sketchier ones.' Poll aggregation is mainly about canceling out sampling error, pure and simple. If we...can erase some house effects, that's a bonus. Even the 'high quality' polls have sampling error. Obviously, would prefer only high quality polls, but this is the world we live in. And if you think I'm not concerned about herding or other bad practices, read me fretting here. High quality polling will always have value. If anything, aggregation helps judge which pollsters are good/bad." [Tweets via Storify by @DrewLinzer]

WEDNESDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Gallup's U.S. Job creation index increased in May to the highest score since 2008. [Gallup]

--72 percent of Americans recognize Chris Christie, and both Democrats and Republicans rate him favorably. [http://www.gallup.com/poll/162911/new-jersey-gov-christie-broad-cross-party-appeal.aspx]

-Andrew Cuomo's job approval now at 53 percent, down from 74 percent in December. [Quinnipiac]

-The 2016 Republican presidential race is wide open. [YouGov]

-Google has a higher favorable rating than Apple or Facebook. [WaPost]

-Sean Trende explains why Chris Christie's handling of a Senate vacancy might be a sign of "political genius." [RCP]

-David Hill says Republicans should target "The Resentfuls" in 2014: recent college graduates who can't find jobs and are underwater on student loans. [The Hill]

-Mark Mellman says partisanship is the real advantage of incumbency. [The Hill]

-Seth Masket maps the states with the most ideologically extreme legislators. [Mischiefs of Factions]

-Ross Schulman responds to Quentin Hardy: "Nobody said Big Data is truth." [DisCo]

-Ninety-one percent of American adults own cellphones, and for the first time, a majority are smartphones. [Pew]

POLLSTER UPDATE: Reviewing The Gallup Review

2013-06-04T17:02:03-04:00

Gallup explains what happened in 2012, and some of it may sound familiar. Everyone has an opinion about Chris Christie's decision to call a special election in NJ. Meanwhile, voters think the House and Senate need more women. This is the HuffPost Pollster update for Tuesday, June 4, 2013.

GALLUP: HOW FOUR SMALL PROBLEMS CREATED ONE BIG MISS - "Gallup, the polling firm that took a major hit last year for finding better results for GOP presidential nominee Mitt Romney than other pollsters found, offered a detailed mea culpa on Tuesday, citing a confluence of small issues that tipped its results rightward...On Tuesday, at a press briefing and in a 17-page report, Gallup named four factors that lessened the accuracy of its polling. As expected, no one problem alone accounted for its difficulties in 2012. "None of these factors are large, in and of themselves," said Gallup Editor-in-Chief Frank Newport, "in the sense that they are not moving the numbers 10 points, 5 points." However, each of the four nudged Gallup's numbers slightly in Romney's favor, making them collectively "significant enough," according to Newport, "that we think they made a significant difference in our overall estimate of who was going to win the presidential election last fall." [HuffPost]

Four problems - The four problems were (1) their likely voter model generally, which according to Newport "was at least 1 point more towards Romney than the average of other polls that were using some time of likely voter model," (2) under-representation, within census regions, of the Eastern versus the Central times zones and the Pacific versus the Mountain Time zones, (3) the use of a series of yes/no questions to measure race, which distorted the weighting process and led to an over-representation of some white voters and (4) the use of non-standard "listed landline" sampling instead of random digit dial samples led to older and more heavily Republican pool of respondents. [Gallup]

Sound familiar? Not to say we told you so, but we reported extensively on the last two problems in June 2012 and March 2013.

Reactions -

-David Axelrod: "All of these were issues that were raised during the campaign, but met with stubborn resistance from the Gallup organization." [@DavidAxelrod]

-Stephen Shepard: “Gallup itself was a part of the reason that national poll averages were inaccurate. Gallup's polls exhibited a consistent Republican bias in 2012; meanwhile, Gallup and some other firms, like the automated pollster Rasmussen Reports, are overrepresented in averages because they conduct daily tracking polls in the months prior to the election, rather than more infrequent media pollsters that didn't skew as heavily toward the Republican candidate. There is little precedent for review Gallup's process, and the implications for the firm's future -- and survey research at large -- could be far-reaching.” [National Journal]

-Josh Marshall: "Appropriate for Prez Romney to show up at Gallup poll roll-out?" [@JoshTPM]

Transparency update - From our report: "Newport confirmed that Gallup will make the 'raw' respondent-level data [from 2012] available to scholars generally through the Roper Center Public Opinion Archives."

How Frank Newport's day started - Joe Scarborough, Introducing Newport on Morning Joe: "Frank, thank you for being with us. I've been a long time consumer of Gallup polls, and quoted them a good bit. Obviously, that changed after 2012. You guys, you said you've had an extensive look at what went wrong. Tell us, what went wrong at one of the oldest and most established polling organizations in America in 2012?" [MSNBC]

A MID-OCTOBER SPECIAL ELECTION IN NEW JERSEY - The experiments Gallup is planning for New Jersey just got more interesting. HuffPost's John Celock: "Saying that an elected senator is needed quickly, New Jersey Gov. Chris Christie (R) on Tuesday called for a special election in October to fill the seat of Sen. Frank Lautenberg (D), who died on Monday.
Christie announced that a special primary election will be held on Aug. 13 with an Oct. 16 general election to fill the seat, ending a debate over conflicting state laws and the possibility that the Senate election would be the same day as the November 2013 election." [HuffPost]

Polling commentariat reacts:

-Monmouth University pollster Patrick Murray explains Christie's decision. [Monmouth]

-Harry Enten calls Cory Booker the big winner. [Guardian]

-Nate Silver says the decision speaks to a weak Republican field. [NYTimes]

-Nate Cohn says the October primary date is "a clear sign that [Christie] doesn't think a GOP victory is plausible." [Electionate]

-Sean Trende "can't decide whether Christie is a genius re 2016 or a complete mullethead." [@SeanTrende]

-Nathan Gonzales says the NJ Senate seat remains "safe" for Democrats. [Rothenberg Report]

-Chuck Todd says "Christie's 1st 2nd and 3rd choice for senate seat is Tom Kean Sr., Tom Kean Sr. and Tom Kean Sr. [@ChuckTodd]

-Tom Kean Sr. was the most popular NJ governor in a 2010 Monmouth University poll. [AP, h/t @ForecasterEnten]

-Nate Silver makes the case for Christie to nominate a Senator with previous electoral experience (but doesn't mention Kean). [NYTimes]

MOST PREFER A 50/50 GENDER SPLIT IN CONGRESS - HuffPollster's Emily Swanson reports on a new HuffPost/YouGov poll: "According to the new poll, 58 percent of Americans say that, if they could choose, they would prefer a Congress made up of about the same number of men and women. Sixteen percent said they prefer a mostly male Congress, while 9 percent said they would prefer one composed of mostly women...By a 45 percent to 27 percent margin, poll respondents were more likely to say that the interests of women are not fairly represented in Congress. Women themselves were especially likely to say that women's interests are not fairly represented (by a 53 percent to 17 percent margin)...Generally speaking, poll respondents were aware that the current Congress has few women. Fifty-three percent of respondents said that less than a quarter of current members of Congress are women, while another 25 percent said that it's less than half. Women make up 18.3 percent of the current Congress, with 20 percent of the Senate and 17.9 percent of the House of Representatives. [HuffPost]

DEM POLLSTERS SEE REASON FOR HOPE IN GEORGIA - Anzalone Liszt Grove’s Zac McCrary and Brian Stryker put on their peach-colored glasses: “Georgia's 16 electoral votes give it more clout than swing states like Virginia, North Carolina, Colorado, and Wisconsin....If Georgia had gone Democratic in 2012, for example, Republicans could have won all of their Rust Belt/Great Lakes targets of Pennsylvania, Ohio, Wisconsin, Michigan, and Minnesota and still would have lost 274--264. Given the demographic shifts underway nationally, there are several opportunities for Democrats and several headaches for Republicans. Population trends can change and parties can re-position themselves, but after a decade of states like Virginia, North Carolina, and Colorado becoming legitimate swing states, Georgia could certainly be the next state to move from red to purple -- and potentially blue.” [HuffPost]

TUESDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-U.S. Consumer confidence continues to rise. [Gallup]

-New England College poll of Massachusetts finds Markey leading Gomez, 52 to 40 percent. [NEC]

-Pew Research finds a global divide on homosexuality. [Pew Global]

-Kristen Soltis Anderson joins HuffPost Live to discuss the CNRC report on how millennials rate the GOP. [HuffPost]

-Charlie Cook suggests five polling "signs" that will signal the direction of the 2014 midterms. [National Journal ]

-Partisans are more likely to tell the truth when paid an incentive. [WaPost]

-Derek Willis shares lessons from the 'Data Crunched Democracy' conference. [The Scoop]

-Oxford Journals puts the first issue of the Journal of Survey Statistics and Methodology online for free. [Oxford]

-AAPOR posts winners from its conference video contest (we liked the 2nd place video best). [AAPOR]

-Almost a third of British elementary school children think cheese is made from plants and a quarter think "fish fingers" come from chicken or pigs. [BBC, via @LoganDobson]

Gallup Poll Reveals 4 Reasons It Got The 2012 Election Wrong

2013-06-04T15:38:31-04:00

WASHINGTON -- Gallup, the polling firm that took a major hit last year for finding better results for GOP presidential nominee Mitt Romney than other pollsters found, offered a detailed mea culpa on Tuesday, citing a confluence of small issues that tipped its results rightward.

Over the final weeks of the 2012 campaign, Gallup's daily tracking poll showed President Barack Obama consistently trailing Romney, including a final survey that gave Romney a 1-point edge. Instead, Obama won by nearly 4 percentage points. While Gallup was far from the only polling firm to call the election wrong, its visibility and reputation, as well as the size of its error, made it one of the most notable.

On Tuesday, at a press briefing and in a 17-page report, Gallup named four factors that lessened the accuracy of its polling. As expected, no one problem alone accounted for its difficulties in 2012.

"None of these factors are large, in and of themselves," said Gallup Editor-in-Chief Frank Newport, "in the sense that they are not moving the numbers 10 points, 5 points." However, each of the four nudged Gallup's numbers slightly in Romney's favor, making them collectively "significant enough," according to Newport, "that we think they made a significant difference in our overall estimate of who was going to win the presidential election last fall."

Here’s a brief summary of those issues, several of which The Huffington Post addressed earlier in more detail:

Misidentification of Likely Voters. Like most other media pollsters, Gallup attempts to determine which respondents are likely to actually vote in November, using a procedure developed in the 1950s that involves a battery of questions, such as how often they've voted in the past and how much attention they're paying to the election. Last year, this likely voter model moved Gallup's estimate of the margin separating Obama and Romney 4 points in Romney's direction.

"That, in and of itself," Newport said, "was at least 1 point more towards Romney than the average of other polls that were using some time of likely voter model." Thus, "on that sense alone," he concluded, "they moved things too far in moving the sample towards Mitt Romney."

While Gallup is continuing to investigate its likely voter procedures with a series of experiments in the fall 2013 state elections, it pointed out one big issue in 2012: that many Obama voters said they hadn't given much thought to the election, removing them from the likely voter pool even though they intended to cast a ballot. Gallup is researching whether it needs to majorly overhaul or even replace the way it identifies likely voters.

Under-Representation of Regions. Gallup also weights its data by a variety of factors, including broad geographical regions such as the Midwest and the South. But each region contains several time zones. Due to differing response rates, Gallup didn't interview enough people in certain time zones within some regions, effectively undersampling states that vote more Democratic.

Faulty Representation of Race and Ethnicity. As HuffPost first reported in June 2012, Gallup in recent years has used an unusual method to ask about race that distorted the racial composition of its samples when the data were weighted. Unlike most other pollsters, who ask respondents to select from a list of racial and ethnic categories, Gallup asked respondents whether they identified with each racial and ethnic group one by one. This led to a disproportionate number of people who said they were multiracial, and that in turn distorted the weighting procedure, effectively giving too much weight to some white voters.

Earlier this year, Gallup eliminated the yes/no racial questions and made some additional technical changes to its weighting procedures. As a result, Newport said, "we [now] do a better job moving our sample to the parameters we have as our gold standard ... for overall race representation."

Nonstandard Sampling Method. Before 2011, Gallup had selected phone numbers using random digit dialing, or RDD, which calls randomly generated numbers. This is the procedure that most national media polls have used for decades. As reported by HuffPost earlier this year, Gallup made a significant change in 2011, when it dropped the RDD methodology for its landline sample, using instead numbers randomly selected from those listed in residential telephone directories.

Gallup was alone among national pollsters in making this move, which promised to cut costs by eliminating unproductive calling to business lines and nonworking numbers. But the change came with a downside: Not everyone who has a landline has a listed number. Although Gallup's initial research indicated that cell phone calls would cover the difference, they didn't: The listed sample turned out to be older and more heavily Republican than the RDD sample.

Gallup is now in the process of switching back to RDD sampling for both landline and mobile phones, Newport said.

"Although in theory the use of the landline listed sample made sense ... it probably was a contributing factor to a skew of the sample that we felt was inappropriate, and therefore, we're making that change," he said.

In addition, Gallup also identified a number of factors it said didn't influence its results. Among them: the tracking poll format, the use of the name "Gallup," the race and gender of interviewers, the handling of third-party candidates, the order in which candidates' names were read, and the way undecided voters were nudged into picking a candidate.

But the firm said it's continuing to research these and other areas, using two 2013 elections -- the New Jersey and Virginia gubernatorial races -- as a sort of polling test lab. The results, which won’t be released to media, are intended to shed light on some of the questions plaguing Gallup individually and the polling industry as a whole.

"We think this is a perfect environment in the real world to see what's going on," Newport said. "It's very difficult to do retrospectively without an actual election to use."

Gallup's research later this year will include experimental surveys that attempt to validate the actual turnout of sampled voters in order to assess the accuracy of the likely voter selection process. These surveys will be designed to test a theory on which HuffPost previously reported, that many nonlikely voters were, in effect, screening themselves out -- by declining to participate in the survey.

"Does the fact that [respondents] have gone through the gantlet to the point where they are in our survey," Newport asked, "could that make a difference in terms of saying that they're already more likely to vote than the people who don't enter the sample?"

The ongoing research is being led by Michael Traugott, a renowned political scientist and survey methodologist at the University of Michigan, and several other academic experts in survey methods and statistics. Traugott will organize a fall seminar involving students in the Joint Program in Survey Methodology at Michigan and the University of Maryland, who will participate in the experiments' design. Newport confirmed that Gallup will make the "raw" respondent-level data available to scholars generally through the Roper Center Public Opinion Archives.

POLLSTER UPDATE: New Jersey's Senate Battle Complicated By Frank Lautenberg's Passing

2013-06-03T16:49:25-04:00

While the political world mourns Sen. Frank Lautenberg, the inevitable succession speculation begins. Barack Obama's job approval has declined since January, but it's still hard to see a significant shift in the last month. And we preview what to expect from Gallup's 2012 review. This is the HuffPost Pollster update for Monday, June 3, 2013.

RIP: FRANK LAUTENBERG - "Senator Frank Lautenberg (D-N.J.) has died at the age of 89 of viral pneumonia, The Bergen Record reported Monday, citing the senator's office...The senator's office confirmed Lautenberg's death in a Monday morning statement: 'United States Senator Frank R. Lautenberg, the last remaining World War II veteran serving in the Senate, passed away due to complications from viral pneumonia at 4:02 a.m. today at NewYork-Presbyterian Hospital/Weill Cornell.'" [HuffPost]

Here's the list of U.S. Senators -- there were 115 of them -- who served in World War II. Lautenberg is the last. ow.ly/lELSv— The Fix (@TheFix) June 3, 2013

Lautenberg death highlights dwindling # of veterans on Hill. Among public, wider civilian-military divide as well.bit.ly/13gziVh— Carroll Doherty (@CarrollDoherty) June 3, 2013

NJ 6/3/13 View Senator Lautenberg’s job approval ratings over the last 17 years bit.ly/16EneBX— Quinnipiac Poll (@QuinnipiacPoll) June 3, 2013

THE SPECULATION BEGINS - Cook Political's Jennifer Duffy reports that New Jersey law is unclear on the timing of any special election to fill Lautenberg's seat: "Under state law, the timing of Lautenberg’s death complicates when an election might be held to succeed him. The only thing that is clear is that when a U.S. Senate vacancy occurs, New Jersey’s Governor does appoint someone to fill the seat until a successor can be elected. What state law makes complicated is the timing of the election to choose that successor...There are conflicting provisions in New Jersey state law about when an election to fill a vacancy will occur. According to one provision, the law says that if a vacancy occurs more than 70 days before the next regularly scheduled statewide general election, then it will be filled during that general election....[That provision would mean] a special election to be held on November 5 of this year....However, another provision in state law says that a special election will be held in the next regularly scheduled statewide general election (e.g., this November) only if the vacancy occurs more than 70 days before the state’s primary election...[that provision would mean] the vacancy would be filled at the next general election – or November 4, 2014. It should be noted that the Governor does have the discretion to call a special election anytime between now and November 4, 2014, including holding one on November 5 of this year. [Cook Political]

Nate Cohn - “Barring a Booker snub by the state's Democratic Party, the date of the election is unlikely to change the outcome. A later election would give a Republican appointee more time to capitalize on the benefits of incumbency, but it's unlikely that a year and a half in the U.S. Senate will do wonders for a Republican’s popularity in New Jersey. If the election is held in November 2013, Christie’s coattails might help the GOP’s Senate candidate. On the other hand, Democrats tend to benefit from higher turnout, and a general election with two big races might draw more voters than a Senate election would on its own.” [TNR]

Sean Trende - "Oddly enough, there aren’t any clear winners here. Christie likely comes out of this a loser however it plays out. Corey Booker and other Democratic candidates just had their path to the Senate complicated. And whoever gets appointed to the seat will have to face an unfriendly electorate twice in the next 18 months." [RCP]

Harry Enten - "Methodological note: Be careful using the 2012 exits from NJ. Small sample size for an exit combined with post-Sandy temporary migration." [@ForecasterEnten]

Kevin Collins - "Prediction: Christie will call for an early special for US Sen and will appoint a GOP-er with good name id and appeal to Dems, like Whitman." [@KWCollins]

OBAMA APPROVAL TREND REVISITED - In a rare moment of seemingly precise agreement, the Gallup and Rasmussen tracking daily tracking polls of President Barack Obama's job approval rating have both shown a slight decline from roughly 50 percent approve to 47 percent approve over the last week. However, though near the low end, both remain within the low end of the typically day-to-day variation these polls have exhibited over the last few months. That said, the Pollster approval chart, which tracks all available public polls, tracks a long term decline in Obama's approval rating since a post-election/inaugural bump that peaked in January. [Pollster: Obama Job Approval]

Franklin finds no significant shift - Pollster co-founder and political scientist Charles Franklin has been watching the approval polls closely for "a change in trend or mean since the Benghazi hearing on May 8 and the subsequent IRS and AP revelations," and finds none. From a June 3 update: "As more data are available it is undoubtedly true that there will eventually be some movement from the current trend, either up or down...[yet] A summary of the statistical models finds no statistically significant shift since May 8, either overall or for any partisan group (Dem, Rep or Ind)." [Polls And Votes]

PREVIEWING GALLUP'S MEA CULPA - In case you missed it, we previewed five things you can expect from the "extensive review" of 2012 election polling methods that Gallup will be presenting reporters in Washington on Tuesday. A quick summary: (1) A deep dive into survey methodology. The list of issues examined covers virtually every aspect of Gallup's process. (2) This is just part one. In particular, investigation into Gallup's likely voter model is continuing and conclusions on that subject will await experiments Gallup will conduct in conjunction with gubernatorial elections in November. (3) Don't expect one big thing. The past history of polling snafus suggests a series of small errors all in Romney's favor. (4) Several "small things," uncovered by our own investigation last year, are already know. Finally, (5) Gallup will be raising their bar on transparency, perhaps even releasing respondent level data on their October tracking to scholars. [HuffPost]

The threat of our 'polling addiction?' - Elizabeth Wilner: "Gallup’s mea culpa this week and yet another release of 2016 trial heats reminds us that the biggest threat to the health of public opinion polling may not be shrinking response rates or the rising cost of dialing cellphones, but our growing addiction to its results...Some news organizations are considering using Survey Monkey. These organizations in the past would have scoffed at the idea of doing online or unscientific opinion research. Now, because of their shark-like need to constantly poll or die, they’re talking with a company whose core business is online employee satisfaction surveys." [Cook Political]

MONDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-Sam Wang defends averages. [NYTimes]

-Nate Cohn says Lincoln Chafee's still in hot water. [TNR]

-Nate Silver's not surprised by the decline in Hillary Clinton's favorable rating. [NYTimes]

-Anna_Greenberg and Sean Trende discuss public opinion on marijuana legalization. [@BrookingsGS]

-University of Arizona researchers survey those crossing the Mexican border illegally. [WaPost]

-Quentin Hardy shares six myths about big data. [NYTimes]

-Stephen Few explains what being a "data scientist" really means. [Perceptual Edge]

-Mike Mokrzycki summarizes how he and others used Twitter at the 2013 AAPOR Conference. [Storify]

-Portlandia explains what it really means to be a nerd. [YouTube, via @JoshdelaRosa1]

Gallup's 2012 Mea Culpa: What To Expect

2013-06-03T14:31:25-04:00

WASHINGTON -- On Tuesday, Gallup will unveil new details of an "extensive review" of its 2012 polling at a press briefing in Washington. While the investigation is not yet complete, the explanation of why the pollster consistently understated President Barack Obama's support during the 2012 campaign is likely to be complex.

After the election, Gallup's editor-in-chief, Frank Newport, announced that the company would conduct an internal review of its election polling methods, led by Michael Traugott, a renowned political scientist and survey methodologist at the University of Michigan. Gallup has provided few hints of its specific findings so far, but given the scope of the investigation and the nature of similar polling misfires in the past, here's what to expect:

A deep dive into survey methodology: At a briefing in May attended by The Huffington Post and a handful of other pollsters and academic researchers, Newport and Traugott shared a long list of issues being investigating (reproduced at the end of this article). It covers virtually every aspect of the telephone survey process, from drawing samples and interviewing voters to weighting data and selecting the likely electorate.

The company has done more than simply scrutinize the procedures used and data collected during 2012. Gallup has also conducted what Traugott described as "a series of experiments going forward ... involving various aspects of data collection," with the goal of determining if alternative procedures would have produced different results.

The outline of topics is evocative of a similarly extensive investigation conducted by the American Association for Public Opinion Research (AAPOR) of polling failures during the 2008 primary elections. The resemblance is not a coincidence; Traugott also led that investigation and was the primary author of its final report.

This is just part one: At the May briefing, Traugott and Newport cautioned that many of their experiments, particularly those relating to the selection and modeling of likely voters, remain ongoing. So while this week's briefing will likely present in-depth findings about the way Gallup selects and interviews its adult samples, those hoping for an in-depth dissection of Gallup's likely voter model on Tuesday may be disappointed.

Traugott explained in May that because experiments on likely voter procedures only make sense "in the context of a campaign," Gallup decided to undertake "a major experiment in conjunction with one or both of the 2013 gubernatorial elections. Virginia almost certainly, and possibly New Jersey." Analysis of these efforts will released after the November elections.

Don't expect one big thing: If the past is a guide, the news from Tuesday's briefing may be tough for reporters to summarize. Investigations of some of the most infamous polling failures of the last two decades -- including exit poll problems in Florida in 2000 and nationwide in 2004, and errors in surveys conducted before the New Hampshire primary in 2008 -- found not one big culprit, but a series of small errors all creating statistical bias in the same direction.

"Every survey has errors," said exit pollster Joe Lenski, referring to the many choices that pollsters make about how they draw samples, select respondents, ask questions and identify likely voters. Any of these choices can create small, typically random errors in one direction or another.

"It's just a matter of, are they small and do they cancel each other out," Lenski explained. "When they're small and they're all in the wrong direction, they make you look bad."

Several "small things" are already known: Gallup's problems in 2012 did not begin with the fall campaign. A Huffington Post investigation found three factors that appeared to contribute to a Gallup "house effect" that lowered President Obama's job approval rating. These involved the questions Gallup asked to ascertain the race of its respondents, the targets used to weight the racial composition of its adult samples and a consistent underweighting of non-white adults. In October, Gallup announced methodological changes that appeared to address the weighting issues. The firm also changed its race questions early this year, dropping the format criticized by HuffPost's investigation.

A return to transparency? Gallup has long been a leader in pushing pollsters to be transparent about their methods. The company founder, George Gallup, was the first to propose a "national standards group for polling" that became the National Council on Public Polls, and he played a leading role in establishing the Roper Center Public Opinion Archives, where Gallup and other public pollsters have long deposited their raw data for use by scholars.

Though Gallup has been on the defensive over its 2012 misfire, it appears poised to reset the curve for transparency with its ongoing investigation. The topics that Newport and Traugott are reviewing include aspects of the survey process that pollsters are often hesitant to discuss. And, as Traugott explained in May, Gallup "has agreed to make all of this information publicly available." He stressed that both he and Newport, as former presidents of AAPOR, "are firmly committed to transparency."

One unresolved issue is whether Tuesday's event will include a release of raw respondent-level data that would allow other scholars and researchers to test other theories. Newport said in May that although "final decisions" had not been made, "we think we'll try to make the data available as well."

So while Gallup's Tuesday review may not answer all questions, it promises to give polling aficionados much to consider.

Outline of Topics Covered by Gallup's Ongoing Investigation

A. Survey Design Factors
1. Tracking design
2. Call design, 3 call versus 5 call
3. Listed landline versus RDD [random digit dialing]
4. Respondent selection procedure
5. Cell and landline quotas
6. Gender quotas
7. Spanish language interviews

B. Survey Field Process
8. Gallup name
9. Gender composition of interviewers
10. Race composition of interviewers
11. Interviewer effects
12. Distribution of interviews by geography and density within regional quotas
13. Distribution of interviews by local time of interview
14. Interviewer Probing of DKs [don't knows] and Refs [refusals]

C. Post Field Handling of Data
15. Weighting
16. Ballot wording and placement within survey
17. Candidate order on the ballot
18. Handling of third party candidates
19. Process of screening for registered voters
20. Process of screening, adjusting for estimated likely electorate

D. Continuing research, including November 2013 VA and NJ Gubernatorial Elections

POLLSTER UPDATE: Hillary Clinton's Favorability Drops

2013-05-31T15:47:45-04:00

Hillary Clinton is looking a little more partisan to Republicans again. The Obama campaign shares their trendless trendlines from the 2012 campaign. And Twitter data-crunches democracy. Or something. This is the HuffPost Pollster update for Friday, May 31, 2013.

CLINTON DROPS IN FAVORABILITY, BUT STILL LEADS POTENTIAL 2016 RIVALS - HuffPost: “If the 2016 presidential election were today, Clinton would beat both former Florida Gov. Jeb Bush (R) and Sen. Rand Paul (R-Ky.) by 8 points, according to Quinnipiac University, leading Bush 48 percent to 40 percent and Paul 49 percent to 41 percent....Fifty-two percent of voters held a positive impression of her, down from 61 percent in a February Quinnipiac poll.” HuffPost]

Benghazi effect? - In Quinnipiac's news release, assistant polling director Peter Brown linked the drop in Clinton's favorable rating to the Benghazi controversy: "The drop in her favorability is substantial among men, Republicans and independent voters. One reason for her drop may be that 48 percent of voters blame her either a little or a lot for the death of the American ambassador in Benghazi." [Quinnipiac]

Question order effect? - TNR's Nate Cohn tweets an alternative hypothesis, "that Clinton's lower numbers...might be due to question sequence." He notes that a previous Quinnipiac survey "didn't ask any horse race questions," but the current survey led with questions matching Clinton against both Bush and Paul before before asking the Clinton rating that "could shift context" for Republican voters. [@ElectionNate]

Seen in other polls - While Cohn's point is fair -- and important given the speculation about the impact of the Benghazi story -- the change in context on the Quinnipiac poll mirrors the ongoing and inevitable shift in the nature of news coverage of Clinton as speculation mounts about a potential 2016 candidacy. Other polls as tracked by the Pollster chart, which combines all publicly available surveys that ask a favorable rating, finds that her average rating is now under 53 percent, down from nearly 60 percent in November. [Pollster]

CHART OF THE DAY - Bloomberg/BusinessWeek's Josh Green posts an epilogue to his Thursday story on the new business venture of Dan Wagner and much of the rest of the Obama campaign's data analytics staff. Green got the Obama campaign to release "a chart based on internal data that shows how the Obama campaign’s swing state model performed against the much maligned Gallup poll over the last several months of the race. This was the campaign’s daily 'horserace projection of the outcome,' based in part on 10,000 analytics interviews conducted each night to support Wagner's modeling. "Gallup shows a huge drop for Obama—really, an outright collapse—after the debacle of the first debate. At the time, Obama’s staffers were claiming to the press that, yes, their internal numbers showed the president’s weak showing had hurt his support, but that the fall was brief and quickly stabilized right about where his level of support had been all along....Based on this data, though, the Obama campaign looks to have been telling the truth." [Businessweek, see also Pollster: "How The Internal Data Got it Right"]

Not Just Gallup - While Gallup consistently understated Obama's position compared to other national polls, it was far from the only pollster to show a dramatic decline for Obama following the first debate, followed by a recovery in the aftermath of Hurricane Sandy. As we reported in November, most national polls, including those conducted by Pew Research and ABC News/Washington Post, showed a trend roughly parallel to Gallup's following the first debate. [Pollster]

WHAT’S THE FUTURE FOR DATA AND DEMOCRACY? The University of Pennsylvania held a conference on Friday on “Data-Crunched Democracy.” Some tweets from the event:

Obama data/analytics team spent significant amount of time educating staff/field on things like "What is probability?" #datapolitics— Alex Lundry (@alexlundry) May 31, 2013

The real scandal abt voter file #datapolitics is that data aren't transparent *enough*Some states restrict access to parties & candidates— Kevin Collins (@kwcollins) May 31, 2013

RT @prisonrodeo: Apparently all the people tweeting #datapolitics are discovering that academics & others have different goals.— The Victory Lab (@victorylab) May 31, 2013

If Facebook and cats aren't the secret to winning campaigns, then what's the point? #datapolitics— Joshua Darr (@joshuadarr) May 31, 2013

WHERE’S MY NEWSLETTER? Alas, the email version of this update will be delayed slightly, due to some technical issues. If you’ve already signed up, you’ll receive them as soon as we officially launch. In the meantime, as always, you can find our updates on the Pollster page.

FRIDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-68 percent oppose U.S. military action in Syria. [Gallup]

-37 percent of Republicans say voting for immigration reform with a path to citizenship is a "deal-breaker" for them. [WaPost]

-Younger Americans are less concerned with "family values issues." [POS]

-Nate Silver sees risk for the GOP in Bachmann's retirement. [NYTimes]

-Dave Wasserman says the GOP has a built-in turnout advantage for the midterm elections. [Cook Political Report]

-Harry Enten says Chafee's party switch culminates New England's shift from Republican to Democrat. [Guardian]

-Alex Lundry finds dissimilarities in public opinion on gay marriage and abortion. [Daily Caller]

-Jaime Settle describes her research using Facebook data to assess how political mobilization messages sent through social media affected behavior. [GNIP]

-Adam Sage analyzes Twitter conversation networks at the 2013 AAPOR Conference. [RTI]

-The Market Research Association defends the American Community Survey (ACS). [Research]

-Nicholas Kulish and Chris Cottrell argue that restrictions on the German census in the name of privacy led to a big miss on population estimates. [NYTimes]

-Auto industry trade group survey finds 81 percent of Americans concerned that "computer hackers could take control of an automated vehicle." [NYTimes]

-For the Nerd who has (almost) everything: R-Stats playing cards. [Ebay]

-YouGovUS finds pollster joy in the Spelling Bee. [@YouGovUS]