In recent years, election polling has taken a bad rap.
Donald Trump’s electoral college win over Hillary Clinton in the 2016 presidential election contradicted polls across the country, cratering credibility in the polling system. And 2020 didn’t help when polls showed now-President Joe Biden with larger lead in swing states than he ended up with.
As we navigate the 2022 election season, skepticism continues to swirl around this science of polling and exactly how exact it is.
Brian Sanderoff, president of Research and Polling Inc., has been conducting the Journal Poll since he established his polling company in 1986 and can speak to that. National statistician Nate Silver’s FiveThirtyEight website gives the New Mexico company an “A” rating and ranks it among the top 10 most credible pollsters out of more than 400 it rates.
And for the record, Sanderoff got New Mexico’s numbers right in the Trump-Clinton matchup.
To get that A rating he very much treats each poll as an exact science — taking into account myriad factors such as demographics and participation rates.
For example, the first step in any poll is crafting the questions so there are no biases. This is fairly straightforward when doing candidate “horse race” polls. It gets more complicated when trying to gauge the public’s opinion about specific issues such as abortion, gun control, crime and homelessness.
The process for our Journal Polls is this: Editors and political reporters at the Journal suggest a question, Sanderoff’s folks come back with their recommended wording, then we often haggle back and forth and finally agree to a question that passes Sanderoff’s high standards. (A few topics have ended up on the cutting-room floor because the wording just got too complicated.)
Sanderoff’s group then generates a random sample among the population he believes will be voting in that specific election based on actual voting histories. For example, different groups often vote in a primary compared to a general election or during a presidential election year vs. a nonpresidential year. It’s a far more accurate process than simply asking whether someone plans to vote. Then some recently registered voters are included in the mix.
And obviously the right number of Democrats, Republicans and independents are polled to match New Mexico’s demographics and population.
A total of 27 pollsters conducted the Journal’s most recent poll. It’s a time-consuming process. For every successful phone call there are multiple unsuccessful calls — no answer, an answer and a hang-up, or an answer but then a hang-up halfway through the call.
And it’s not cheap. The Journal could almost hire a reporter for the cost of the polling we will have done this year. Our media partner, KOAT-TV, has chipped in the past two years. This year, that will allow us to include congressional races in our next and final poll.
Even though Research and Polling Inc. is partially owned by the Albuquerque Publishing Company, that doesn’t give us special rates. The Journal newsroom is treated like any other client, which suits us fine. We contract with Research and Polling because of its reputation and proven record over 35 years.
Sanderoff laughingly acknowledges he gets plenty of calls from people upset when results don’t look good for their candidate, and he doesn’t hear much from those folks whose candidates do well in the polls.
So why do we continue to poll amidst this criticism and skepticism? We consider it a public service our community has relied on for nearly 40 years.
In an interview two years ago with senior editor Kent Walz, Sanderoff said, “Polls are valuable in showing Americans how other people think. We tend to surround ourselves with people who have similar value systems; polls show there is a different point of view.”
And the public remains interested and curious in what the latest poll shows.
We are asked months in advance of an election if and when we are conducting our Journal Poll. In fact, the August poll on the governor’s race had the highest reader engagement score of the week. The rest of the poll stories scored high as well.
So when is the next poll coming out? You will see it the week of Oct. 30, and it will include a few races not in our August poll.
Regarding that next poll, here is a personal request: If you get a call from a Research and Polling pollster, please stay on the line and participate. We need your help to get this right — again.
Until next time,