HUFFPOLLSTER: Here's What We Can Learn From The Polling Miss In Kentucky

A combination of factors led pollsters to miss the mark in the Kentucky governor's race. Marco Rubio, while still the underdog, is gaining significantly in the polls. And several low-polling candidates won't make it to the next Fox debate. This is HuffPollster for Friday, November 6, 2015.
Republican gubernatorial candidate Matt Bevin defies poll expectations and wins in Kentucky.
Republican gubernatorial candidate Matt Bevin defies poll expectations and wins in Kentucky.
Tom Williams via Getty Images

RECAPPING ELECTION DAY 2015 - Republicans scored some significant victories on Tuesday, including an unexpected win in the Kentucky gubernatorial race. Although polls indicated that Democratic candidate Jack Conway had a slight lead, Republican Matt Bevin won, taking 53 percent of the vote.

Why did Kentucky polls miss the mark? Although it’s difficult to know for sure, evidence points to the timing of the polls, the partisan composition of likely voters in the polls and the presence of an independent candidate. None of these alone explains the wide discrepancy between the polls and the result, but combined, they contributed to the miss. HuffPollster: “There were only six publicly released polls between August and Election Day. As HuffPost pointed out, that’s pretty sparse polling. Any last-minute campaigning or changes in opinion were completely missed -- simply because there were no polls.” [HuffPost]

The role of the independent candidate - More from HuffPollster: “Drew Curtis, the independent candidate, was another problem for Kentucky pollsters. He pulled an average of 7 percent support in polls, but only got 3.7 percent of the vote. If Curtis’ extra supporters in the polls turned out to be Bevin voters, this could explain part of the error in Bevin’s poll estimates.” Polls regularly overestimate support for third candidates when they are named in polls, according to research Edison Research’s Joe Lenski presented at the American Association for Public Opinion Research conference last May. However, Lenski also noted that not naming the third candidate leads to underestimates of the non-major party vote in polls. Either way, a third candidate has the potential to add to polling error.

The role of likely voter selection - About 31 percent of Kentucky registered voters showed up to vote on Tuesday. That is likely a smaller fraction of the state's potential electorate than the possible "likely voters" contacted by pollsters. It’s impossible to know whether tighter likely voter screens would have resulted in different poll results, but there is some evidence that the polls missed the mark in the distribution of voters by party identification. A series of tweets from Charles Franklin, director of the Marquette Law School Poll, compares poll registration and identification numbers to statewide party registration and past exit poll party identification. There’s a big difference between registration and identification -- as a result of changing partisanship and the decline of Southern Democrats, there are many more Kentucky voters who selected Democrat as their party choice when they initially registered to vote than say they are Democrats when a pollster asks if they are Democrats or Republicans in a poll. While the polls seemed to get fairly close on registration, their proportions of Democratic and independent identifiers were generally too high. Polls also underestimated the Republican vote among independent identifiers. Since lower turnout often favors Republicans, a likely voter screen that projects higher turnout would contribute to the pattern of overestimating those who identify as Democrats and Independents.

Higher turnout wouldn’t necessarily have changed election result, though - Franklin notes that there wasn't a gap in turnout rates by party in Kentucky in the last few elections.

The Kentucky polling miss in perspective - Harry Enten: “I would be careful of making too much of the Kentucky results. Only three polls not sponsored by a candidate came out during the final three weeks of the campaign. That’s far less polling than was conducted in other recent polling mishaps, such as in Israel and the United Kingdom over the past year. The Kentucky results match most of the bigger misses in the U.S. during the 2014 midterm elections, such as in the Maryland gubernatorial race and Virginia Senate election, when few polls were released during the final weeks of the campaign. That’s a good thing for 2016, when the most highly anticipated races will have lots of polls in the field.” [538]

Turnout did matter in the Virginia state senate races - University of Florida political science professor Michael McDonald: “The key race for control of the nearly-evenly divided Virginia Senate was the Tenth District, a district that stretches from portions of downtown Richmond City, west through portions of Chesterfield County to encompass the entirety of Powhatan County…. If all the voters in each precinct within the district turned out to vote at the same rate relative to the 2012 election, and these voters cast their ballots the same way those who actually voted did, [the Democrat] would have won by roughly 1,300 votes.” [HuffPost]

MARCO RUBIO GAINING IN THE POLLS -HuffPollster: "Republican presidential candidate Sen. Marco Rubio (Fla.) is experiencing a post-debate boost, three new surveys find, with two New Hampshire polls and a national poll this week all showing him rising to third place. A WBUR poll conducted by MassINC of likely Republican voters in New Hampshire shows Rubio with 11 percent of the vote, up 9 percentage points since a September WBUR poll. Rubio received a greater boost than any other candidate tested in the survey…..[F]avorable rating is also up by 10 points...making him the second-most favorably rated candidate in the field, just after former neurosurgeon Ben Carson….A Monmouth University New Hampshire poll released Monday also shows Rubio taking the third place spot in the race with 13 percent of the vote. His share of the vote has tripled since a September Monmouth poll….On the national stage, a Quinnipiac University poll released Wednesday also shows Rubio jumping to third place, with 14 percent. His favorable rating also holds strong in the poll at 71 percent. " [HuffPost]

Not solely a result of a strong debate performance - Philip Bump: "The last Monmouth poll also coincided with Rubio's worst position in the polls, as The Post's polling guru Scott Clement pointed out. He'd already gained six points since mid-September in the Real Clear Politics polling average in the state….That's additional good news for the Rubio camp, since it suggests that the senator was building momentum even before his good performance. What's more, Rubio and Carson are much more likely to be the second choices of voters -- an important consideration assuming the field of 15 ever gets any smaller….Combining first and second choices, Donald Trump, Rubio and Carson form a top tier of candidates in the state, each getting more than 25 percent combined support." [WashPost]

He's winning the endorsement game too - Jonathan Bernstein: "Marco Rubio is the most likely candidate to win the Republican 2016 presidential nomination. I said early on that Rubio was in a first tier of contenders with Jeb Bush and, before he dropped out, Scott Walker. There was a solid case for and against each of them. Well, the case against Walker turned out to be correct, while the one for Rubio has looked stronger and stronger. Ross Douthat of the New York Times recently described Rubio’s oddly intangible front-runner status. After good reviews for his debate performance (and terrible ones for Bush), that has changed. Rubio has picked up his first two endorsements from his fellow U.S. senators -- Colorado’s Cory Gardner on Monday and Montana’s Steve Daines on Tuesday. After getting off to a slow start in high-visibility endorsements, Rubio has been on a roll for a while now. He has nailed down seven members of the House since Sept. 21. Over the same period, the other 14 Republican candidates had 10 new House endorsements combined." [Bloomberg]

A shifting moment? - Nate Cohn: “In the fourth quarter of 1991 and of 2011, Democratic and Republican officials broke toward candidates who would ultimately become their party’s nominee: Bill Clinton and Mitt Romney….It’s still too early to say whether the Republican primary has reached a similar moment, but when analysts look back on the 2016 election, they may conclude that the G.O.P. reached a similar inflection point this week. Last week’s debate might have been a clarifying moment for party officials and donors, moving many toward deciding in favor of Marco Rubio and ultimately sending him on a path to the nomination." [NYT]

DIFFERENT RULES AND CANDIDATES FOR NEXT GOP DEBATE “Fox Business Network on Thursday announced the candidate line-up for the Nov. 10 Republican presidential debates.The candidates qualifying for the prime-time, 9 p.m. ET debate are: Billionaire businessman Donald Trump; retired neurosurgeon Ben Carson; Florida Sen. Marco Rubio; Texas Sen. Ted Cruz; former Florida Gov. Jeb Bush; former HP CEO Carly Fiorina; Ohio Gov. John Kasich; and Kentucky Sen. Rand Paul. The candidates qualifying for the earlier, 7 p.m. ET debate are: New Jersey Gov. Chris Christie; former Arkansas Gov. Mike Huckabee; Louisiana Gov. Bobby Jindal; and former Pennsylvania Sen. Rick Santorum. The criteria were different than for past debates. In a change, Christie and Huckabee ‎did not qualify for the prime-time event, while former New York Gov. George Pataki and South Carolina Sen. Lindsey Graham did not qualify for either; neither did ex-Virginia Gov. Jim Gilmore. To qualify for the prime-time debate, a candidate had to score 2.5 percent or higher in an average of the four most recent national polls. Candidates scoring under that had to receive at least 1 percent support in at least one of the four most recent national polls to qualify for the 7 p.m. debate.The four polls used were conducted by: Fox News; Investor’s Business Daily/TIPP; Quinnipiac University; and The Wall Street Journal/NBC News.” [Fox News]

CONFLICTING POLL HEADLINES ON FOREIGN POLICY Fox News and AP-GfK both released polls measuring public sentiment about Obama’s planned actions in Afghanistan and the Middle East, but with contrasting findings. Fox News: “A 54-percent majority of American voters approves of President Obama’s decision to send a small number of U.S. troops to Syria to help in the fight against the Islamic extremist group ISIS. … In addition, by a 49-38 percent margin, voters approve of Obama’s decision to delay withdrawal of troops from Afghanistan.” The AP: “More than 6 in 10 now reject Obama's handling of the threat posed by the Islamic State in Iraq and Syria, where Obama has been escalating the U.S. military's involvement in a bid to break a vexing stalemate. ... Concerns about Obama's strategy overseas resonate deeply when it comes to Afghanistan, where Obama abruptly dropped his plans last month to pull nearly all U.S. forces by end of 2016. Roughly a third of Americans said they approve of that revamped plan, with one-third opposed and another third neither in favor nor against.” [Fox News, AP]

Why the difference? - For one thing, the polls were conducted almost two weeks apart As the AP article noted, “The AP-GfK poll was conducted before Obama's announcement last week that up to 50 U.S. special operations troops will head to northern Syria.” Fox News asked whether respondents approve or disapprove of the specific policy Obama announced, while AP-GfK asked more generally if respondents approve of how Obama is handling ISIS. In addition, the questions about Afghanistan were structured differently. Fox News asked whether respondents “approve or disapprove” of delaying the troop withdrawal from Afghanistan, and AP-GfK asked whether respondents “favor, oppose, or neither favor nor oppose” the plan. The explicit “neither” option in the AP-GfK poll attracted nearly a third of respondents, whereas in the Fox News poll respondents didn’t have a middle option -- and only 12 percent volunteered a “don’t know” response.

“CONTROL” MATTERS IN ASKING ABOUT GUN LAWS In a test of how language influences attitudes, half of the respondents were asked one question and the other respondents were asked a question with different wording: Support for 'stricter gun control laws,' asked of the first group, is a negative 46 - 51 percent; Support for 'stricter gun laws,' without the word 'control,' asked of the second group, is 52 - 45 percent. Republican voters oppose either option by wide margins, while Democrats support either option by similarly wide margins. The most pronounced difference is among independent voters who oppose 'stricter gun control laws' 57 - 39 percent, but support 'stricter gun laws' 50 - 45 percent.” [Quinnipiac]


-Marco Rubio rises to third place in two New Hampshire polls, with Chris Christie also getting boost. [Monmouth, WBUR]

-A national Quinnipiac survey gives Ben Carson the lead, with Jeb Bush falling toward the bottom of the pack. [Quinnipiac]

-Ben Carson makes a move for the top spot in the GOP race. [HuffPost, NBC/WSJ]

-Hillary Clinton regains momentum in New Hampshire. [HuffPost, Monmouth]

-Most Americans are supportive of new regulations on health care costs. [Harris]

-Three new polls in the Louisiana governor’s race show Democrat John Bel Edwards leading Republican David Vitter by substantial margins. [MRI, JMC Analytics (R), Triumph (R)]

-Americans think the Spring Valley police officer who violently arrested a high school student was in the wrong. [HuffPost]

HUFFPOLLSTER VIA EMAIL! - You can receive this weekly update every Friday morning via email! Just click here. Enter your email address, and click "sign up." That's all there is to it (and you can unsubscribe anytime).

THIS WEEK'S 'OUTLIERS' - Links to the best of news at the intersection of polling, politics and political data:

-Nate Silver reminds us that we have no idea what the economy will look like in the fall of 2016. [538]

-Nate Cohn finds a path to the nomination for Ted Cruz. [NYT]

-Jonathan Bernstein argues that Ben Carson is not Barry Goldwater. [Bloomberg]

-Lynn Vavreck looks at how Donald Trump is capitalizing on “fault lines” in American society. [NYT]

-Swing voters are getting less common in American politics. [Monkey Cage]

-Amy Walter explains why Carly Fiorina was unable to make her popularity bump last. [Cook Political]

-Half of black millennials know a victim of police violence, according to a report from the University of Chicago. [AP]

-A new study shows how parents who both work full-time share household responsibilities. [Pew]

-Sam Stein offers a solution to the GOP’s debate mess: pickup basketball instead of polls. [HuffPost]

Before You Go

Popular in the Community