What a Charlotte Observer Columnist Gets Wrong About Polling

We knew the criticism of our polling model was inevitable, so it was no surprise when a columnist for the Charlotte Observer attacked the wording of our HB2 question in our recent North Carolina poll. Truth is, we knew the way we worded the question was going to be an issue to some, and it turns out we were right.

On Monday, May 2nd, RABA Research, a bipartisan polling firm that I co-founded with fellow Democrats and Republicans, released a statewide North Carolina poll showing approval for its new, anti-transgender HB2 law underwater with 39% of voters approving of the new law and 50% disapproving. There was another poll released around the same time from the conservative-leaning firm Civitas showing 47% of North Carolina voters approving of the law and 43% are opposed to HB-2.

Both polls had Republican Governor Pat McCrory behind in his re-election campaign.

Here’s what columnist Peter St. Onge wrote for the Observer soon after the polls were released:

“Civitas, which calls itself ‘North Carolina’s conservative voice,’ asks respondents if they agree that the state legislature and governor were right to ‘make sure North Carolina had a uniform law across the state and stop the Charlotte Bathroom Ordinance from taking effect.’ Forty-seven percent said yes, with 43 percent saying the ‘state legislature and governor should not have gotten involved.’ But in no question did Civitas state that HB 2 also removed anti-discrimination protections for gays and lesbians. So we don’t know how respondents see the law with that in mind. RABA, a self-proclaimed bipartisan pollster based in Iowa, D.C. and New York, also fails Polling 101. Its poll asked simply: ‘Do you approve or disapprove of HB 2 — the state’s new anti-transgender law?’ We agree that the law is anti-transgender, but poll questions shouldn’t state opinion (at least without offering the choice of an opposing opinion). That steers respondents toward an answer, and RABA got the answer that the question pointed toward, with 50 percent disapproving to 39 percent approving. So yes, questions count.”

We agree “questions count,” which is precisely why we worded it the way we did. It all boils down to whether or not you believe the law is “anti-transgender.” Our view is whether or not you support or oppose the law, your basic understanding of it is that it is, in fact, “anti-transgender” and saying so is not partisan, but rather plain-spoken fact. Growing, mainstream, non-partisan media outlets such as Vox and Buzzfeed understand this and characterize the HB-2 bill simply as “anti-transgender” or “anti-LGBT” in their reporting and even their headlines.

We live in a Buzzfeed word full of clickbait and distractions, which is why we need to start polling in the current world we live in.

There is a debate currently raging about landlines, cell phones, Internet and the overall state of polling. The debate needs to happen because the polling industry has been off the mark lately. The questions about phones and methodology are important, but my biggest concern is not only how do we get people to agree to take our polls, but also how do we get them to stay on the line and finish them? Ask any credible pollster and they will tell you completion rates for polls are at an all-time low, which means fewer people are finishing the poll and the ones who do are many times a small network of professional poll takers who love answering poll questions (this is especially true in battleground states where campaigns never stop). Of course this then leads to poor results because the professional poll takers who took the poll are not representative of the overall electorate who show up and actually vote.

So, how do you get people to stay on the line and complete surveys? You can start by using clear, crisp language that is reflective of conversations real voters are having at the office water-cooler and headlines they are reading in their Facebook feed. Common sense tells us that whether you support or oppose HB-2, you know it is “anti-transgender,” which is exactly why you either support or oppose the bill. You know Bruce Springsteen and NASCAR are speaking out against HB-2 because they believe the bill is anti-transgender, and you either support or oppose their decision to do so.

The RABA wording on HB-2 is in stark contrast to the Civitas poll asking voters about the “Charlotte Bathroom Ordinance,” a bureaucratic-sounding thing that likely confuses respondents (now, whether confusion was the goal is up to you to decide).

Buzzfeed, Vox and other growing online media outlets understand the appeal of simplicity while many editorial writers and legacy polling firms in today’s climate do not.

So, since I am a generous soul who appreciates a healthy debate, I am going to give Peter St. Onge and the Charlotte Observer a C- in “Polling 101.” He understood the importance of wording poll questions and recognized that HB-2 is politically toxic for Governor Pat McCrory. However, by criticizing the RABA Research characterization of HB-2 as “anti-transgender” and dismissing the results he fails the common-sense test. Yes, Pat McCrory is losing. Yes, Pat McCrory is losing because most North Carolinians oppose HB2. Yes North Carolina voters are blaming Pat McCrory for signing HB2 into law. This is what our RABA poll shows, and we stand by our numbers and the framing of our questions 100%.

P.S. RABA is actually bipartisan and made up of real, live Republicans and Democrats, not merely a “self-proclaimed bipartisan pollster” as described in the Observer column. I’ll let that slide.