2022 Governor's Race

How Accurate Were New Mexico’s Midterm Polls?

Polls leading up to election day showed MLG up an average of 6.6 points

Election Day post-mortems always reek of copeful conspiracy theories blaming fraud for their candidate’s loss when qualifications, campaign messaging, and simple demographics offer the more logical explanation of disappointing Election Day outcomes.

In New Mexico, there are 1.42 Democrats for every Republican voter, a difference of 178,518 voters, according to Secretary of State voter registration data as of October 2022. For a Democrat incumbent to win re-election by only 45,0142 votes–less than half her margin of victory in 2018–should be seen as progress for Republicans in the state.

Alas, it is not.

Other than claiming, without evidence, that voting machines are fixed to favor Democrats, the most common theory of how a Democrat could win in a Blue state (shocker) is: voter suppression in the polls.

Polling companies, the theory goes, skew polling data in order to make it look like Democrats are more popular than they actually are. They inflate Democrat support in their polls in order to discourage Republicans from casting a ballot. If the polls show your candidate has no chance of winning, why bother voting?

There are instances where this appears to have occurred.

In 2020, Trump’s campaign accused ABC News/Washington Post of voter suppression after a poll showed Biden up 17 points in Wisconsin. Biden won Wisconsin by 20,000 votes, less than 1%. For potential Republican voters seeing this in the news, surely it could have the effect of “voter suppression.”

There are obvious holes in this narrative. For one, ABC/WaPo wasn’t the only poll that election (obviously). Fox News and Marquette had Biden up only 5; Susquehanna had it at 3, Trafalgar only 1. Two, the ABC/WaPo poll came out two weeks before Election Day, with dozens of polls before and after that made clear the race was significantly closer. Three, when “trust in the media” has never been lower, voters shouldn’t be taking every poll as Gospel.

If your constituency still doesn’t know that polls are an inexact science, that voter sentiment on a Monday and Tuesday two weeks before an election may be drastically different than on the Saturday and Sunday before Election Day, or that the voting is a civic duty regardless of whether or not your guy’s in the lead, then maybe the problem isn’t the media.

The gaping hole in the “voter suppression” narrative is that it only seems to affect Republicans.

Why do Democrats still show up even when polls show their candidate trailing? In the Georgia gubernatorial race, polls showed Republican incumbent Brian Kemp up by as many as 14 points, yet enough Democrats showed up to cut that number in half. Kemp won, but only by 7.6 points.

And if wide margins discourage Republican turnout, then wouldn’t close margins spur GOP voter turnout? Not according to the data.

In the New Hampshire Senate race, polls had the Democrat up only 1.4 points, with two polls right before Election Day putting the Republican up one. The race was a statistical tie, and yet the Democrat won by 9. The same thing happened in Michigan, where the RCP poll average was 1 point–another statistical tie. Democrat incumbent Gretchen Whitmer won that race by 10.6 points, meaning even in a close race Republicans failed to show up.

In New Mexico, two polls showed the Democrat incumbent up 12 and 16 points–in September and early October, more than a month before the election. Literally every other poll taken showed Michelle Lujan Grisham up 3 to 8 points. Two weeks before the election, Trafalgar showed Republican Mark Ronchetti up 1. The last poll before Election Day, from The Hill/Emerson, had Grisham up only 2.

Are Republicans only reading the bad news? Do they see the one poll a month before Election Day then stop watching the news entirely until after the election? Do they only believe polls certain polls?

Perhaps the better question is, if “voter suppression” is real, why does it only affect Republicans?

Democrats will show up to vote for candidates who aren’t even alive. In Allegheny County, Pennsylvania, the Democrat candidate LITERALLY DIED–a month before Election Day. He won with 86% of the vote.

Libertarians seem equally dedicated to their candidates, even when they know they have zero chance of winning.

In Bernalillo County, a Libertarian candidate for sheriff whose only qualification for office was flying a dildo-dangling drone into a mayoral campaign in 2021–nonetheless secured 4% of the vote!

So while Democrats will show up to vote for a dead guy and Libertarians will vote for a “sheriff dongcopter,” Republicans are so discouraged from three minutes of election news a month before the election that they don’t bother to vote?

Not likely.

Eleven polls were conducted over the course of the New Mexico gubernatorial general election campaign cycle. The average of these polls, according to RealClearPolitics.com, gave Grisham a 6.6-point advantage.

She won by 6.2 points.

Were there outlier polls? Yes. Did they discourage some voters? Possibly. But a snapshot of voter sentiment at a given time is not the predetermined outcome, and if voters don’t understand that, the problem isn’t polling.

3 replies »

  1. Let’s do some simple math questions. 1. How accurate are are the rolls? How often are they purged, and are the 7.5k that we lost to COVID in the last two years still active? Next – polls. They can be statistically manipulated and skewed. There’s a trend being circulated called ballot harvesting vs getting out the vote. SOS has ceased publishing a lot of data which voter voter voted for who. Did we actually have any Democrats vote for the Republican candidates. As for mail in ballots, Will we ever see if those ballots that are counted equal those who request one.

    Too many questions, and without transparency and open audits – THAT contributes to voter suppression.

Leave a Reply