Menu

XRepublic

Unfiltered Political News

Is Roy Moore winning? Don't ask the pollsters

Roy Moore appears to have inched back in front of Democrat Doug Jones in the latest Alabama Senate election polls, according to the oft-cited RealClearPolitics average — a change in fortune from mid-November, when sexual misconduct allegations against Moore first surfaced.

The reality? No one really has a clue about where things stand with Alabama voters in the Dec. 12 special election.

For all the national attention and the millions of dollars spent to win the seat, there’s relatively little public polling in the contest. Only three public surveys in the average have been conducted since the Thanksgiving holiday, and odds are you’ve never heard of two of the three pollsters.

And that’s precisely the problem. The most important and closely watched election in the nation is taking place in the equivalent of a polling black box. There are no established, in-state polling institutions or dominant regional media outlets to fill that void. Since it’s not typically a politically competitive state, outside pollsters don’t have much experience in Alabama either. Outside of Fox News and a Washington Post poll released Saturday, national media outlets or major pollsters haven’t yet stepped forward to survey the race.

On top of all that, the Alabama Senate race is a special, off-year election being held just before Christmas — layering on more elements of uncertainty and mystery.

“This is a state where no one has a real track record,” said Patrick Murray, director of the Monmouth University Polling Institute, “because who bothers to poll Alabama statewide races?”

That’s left the field occupied by a number of little-known polling firms, most of whom use experimental or less-accepted methodologies. In general, those surveys show some movement toward Moore in recent weeks, after the publication of the first news stories in which women said Moore assaulted or pursued them when they were teenagers and the Republican was in his 30s.

The most recent poll, conducted by The Washington Post and George Mason University, showed Jones leading Moore, 50 percent to 47 percent — well within the margin of error. Only 3 percent of voters support a write-in candidate or are undecided, the poll shows.

Interestingly, because of the Post’s role in the race — the newspaper’s editorial board issued an unusual, out-of-market endorsement of Jones early in the campaign and later published the first accusations against Moore — interviewers did not identify the poll’s sponsors until the end of the interview. Moore and his allies have attacked the Post for interfering in the campaign, with Moore strategist Brett Doster calling the paper “a worthless piece of crap that has gone out of its way to railroad Roy Moore.”

But despite the Post poll, Moore remains narrowly ahead in the average, posting small leads in the other three polls that currently comprise it. That includes a survey from JMC Analytics and Polling, a Louisiana-based consulting firm, that shows Moore leading Jones, 49 percent to 44 percent among likely voters.

The JMC poll was conducted solely by contacting voters on landline phones, a controversial methodology at a time when nearly half of Alabama adults live in households without a landline. Those in cellphone-only households skew younger; a stunning 52 percent of voters in the JMC poll are age 65 or older. Fewer than half of voters in the Republican primary this summer were seniors; primary voters would generally be older than the general electorate.

John Couvillon, the JMC pollster, said he was far more worried about achieving the proper racial balance in his poll — three-quarters of respondents were whites, while blacks accounted for 23 percent — than the age breakdown.

“You’re talking about a special election, during the off year, being held during the holiday season. You’re going to have disproportionally an older electorate,” said Couvillon. “We can quibble about the percentage of voters who are older, but when we’re talking about a Deep South state, the more salient factor is the white-black breakdown.”

The third post-Thanksgiving poll is from the student-run Emerson College Polling Society. That poll included a segment surveyed online to account for voters without landline phones and was weighted by age. The poll, which excludes undecided voters or those who might write in another candidate, shows Moore leading by 6 percentage points. That’s closer than the previous Emerson poll, conducted shortly after the first allegations against Moore.

There were two other polls not included in the RealClearPolitics average. One is a one-day, internet survey conducted by Change Research, which was founded this summer, showing Moore ahead by 5 points — statistically unchanged from the firm’s last poll, which had Moore leading by 4 points.

Meanwhile, a conservative nonprofit, America First Policies, released a poll last week showing Moore and Jones neck-and-neck, with the Republican ahead by a single point. After the scandal first broke, Jones was ahead by 8 points, the group said — though it didn’t release that survey at the time. America First Policies is closely aligned with President Donald Trump and was founded by former advisers to Trump’s presidential campaign.

That’s five polls in the past week — only three of which meet the qualifications to be included in the RealClearPolitics average. Compare that with the last special Senate race conducted outside of the usual election cycle to be decided by only a single-digit margin: the 2010 special election in Massachusetts. In that race, there were 11 polls in the final week — though two were conducted by Research 2000, the firm about which questions were later raised regarding the legitimacy of its data.

The limited polling from lesser-known outlets underscores the difficulties in surveying the race. There’s no playbook for polling voters in a special election — in which historical turnout patterns aren’t always predictive — let alone one under these unique circumstances. That leaves pollsters — particularly those who are seeking to include those voters who will turn out and exclude those who won’t — mostly guessing about what turnout might look like.

“It’s a special election, [and] there’s a particularly unusual situation with one of the candidates who should be the front-runner,” Murray said.

Murray also pointed out that in Alabama, Republican voters uneasy with Moore can select a straight Republican ticket instead of marking the box next to Moore’s name — even though the Senate race is the only contest on the ballot. That could lead to more votes for him than poll respondents who are willing to admit that they intend to cast a vote for him later this month.

For now, the limited polling suggests Moore may have reclaimed some of the momentum following Jones’ surge immediately after the scandal broke.

“Having polled the general election twice and the [primary] runoff two times, I can kind of appreciate now the ebbs and flows that have occurred in this race,” Couvillon said. “You can make the argument that Republican voters who supported Trump are coming home to Roy Moore now.”

Powered by WPeMatico