One poll's wrap-up of pessimism vs. optimism among the electorate

The two latest national polls reflect distorted mirror images, while the second result shown demonstrates, if nothing else, the volatility of polling within the industry.   

The chief distortion between the two polls is that Quinnipiac surveyed likely voters; The Journal, registered.

The phrase "statistically insignificant" has been popular in 2024's presidential polling, and again it's applicable to Quinnipiac's +2 (with all candidates included) when compared with Suffolk University's recent finding of +5 for Harris. Nevertheless, insignificant becomes somewhat jarring for readers who glance at RealClearPolitics' poll listings — to see +5 suddenly morph into +1 on RCP's page.

The collective admixture of stats morphs also into a blur, thus one can be forgiven for asking, What does it all mean? There's nothing complicated involved, yet the flurry of +s careening from 1 to 5 of late reminds one of Shakespeare's "much ado," but perhaps about something? No more than the industry's internal volatility, it would seem.

As for the two polling results above, they look identical but they're not. Aside from their conflation of registered vs. likely voters, Quinnipiac is considered a reliable polling operation (2.8/3.0) by 538; The Journal's was done by Michael Bocian, GBAO (1.2/3.0) and David Lee of Fabrizio, Lee & Associates, (1.7). Take from that what you will. My take is that the WSJ cut its costs at the expense of readers, which is one way Rupert wound up with billions of dollars.

In Quinnipiac's internals, 98% of Democrats support Harris, 94% Republicans support Trump, and Independents are split at 45%, which isn't good. What is good is that Harris possesses a 21-point lead among women, 58%–37%. Trump's male support explains why I'm embarrassed by my own kind. Among us three-legged primates he has an 18-point advantage, 57%–39%. As I'm said for years, America's men exemplify brawn over brains.  

Also in the (sort of)  Good News Department is that RFK Jr.'s favorability rating as gathered by Quinnipiac is 32% — by itself, mind-bindingly bad, but an encouraging 10-point drop from what Suffolk gathered. Harris' 47% favorability matches her unfavorability, and Trump's favorability is underwater by 6 points, 47% to 52%.

Oh, and Quinnipiac discovered — it really did — that 1% of its respondents said they "haven’t heard enough about" Trump to form an opinion. Quinnipiac polled 1,611 respondents, nearly all them by conventional means. But somehow the pollster managed to reach 16 people who've been without newspapers, television news, internet service and even telephonic communication for lo these many years, as they hunkered deep in faraway caves. 

So my hat goes off to Tim Malloy, a top man (among the brains-over-brawn minority) with Quinnipiac University's polling operation. And I'll close with this observation from Malloy, one that may be the poll's most significant finding: "Nearly 7 in 10 voters see brighter days on the horizon. While Democrats are the most hopeful, followed by independents then Republicans; in the aggregate, Americans are choosing optimism." Kamala Harris' choice. 

Comments

[Anne J]

I still find polls so confusing. Is there usually a difference between polls taken leading up to an election and the results after voters go the actual polls to cast their ballots?

[PM]

Depends on the election, Anne. The main problem with polling since 2016 has been in undercounting Trump voters. But I still say 2016's were pretty accurate. The problem with them was in interpretation: too many commentators — like me — failed to appreciate the power of the 5% undecided right up to Election Day. Undecided voters tend to go for the challenger.

[Anne J]

In this election, I don't see Trump as the challenger because we've already seen him as president. But it's the amnesia of the electorate that gives me pause.

[VoiceOfReason]

The presence of pro-choice constitutional amendments on state ballots this November is a metric that shouldn't be ignored. See https://www.nbcnews.com/politics/2024-election/abortion-rights-ballot-measures-states-rcna168621 - where NBC noted "Voters in the swing states (Arizona and Nevada), blue-leaning states (Colorado, Maryland and New York) and red-leaning states (Florida, Missouri, Montana, Nebraska and South Dakota) will have the ability to directly decide the future of abortion access this fall." Post-Dobbs history has shown that whenever these initiatives are on the ballot, the GOP loses, which is why the GOP so desperately tries to keep them off the ballot (see Ohio, for example). I can't find much justification for thinking that voters would cast ballots in favor of such, and then vote for Trump. So I'm thinking that at least five states that went for Trump in 2020 will not this time around, because women will come out in large numbers to vote pro-choice. Heck, I even think Ohio is more in play than pollsters seem to.

[PM]

In Quinnipiac, VoR, 73% of women said the abortion issue is extremely or very important in their presidential choice. (Of course some of them are "pro-life," a wretched description.) Only 53% of men said the same, even though they're the ones who've been pushing harder to strip women of their rights.

[Mary]

Anne, of course Trump is the challenger. There is no-one else. Hopefully, he'll bomb but the polls are not entirely favourable, especially taking account of your country's eccentric electoral system. That's the way things work though and everybody knows that. You work within the confines of the rules, except when you're Donald Trump.

In reference to a previous post, Kamala Harris can maybe get away with an occasional dissemble, but greater transparency would be the better option. Everyone knows that there is always some degree of pragmatism in political campaigning. Still, (and this is mean and superficial, I know) Josh Shapiro would have been the better running mate. He looks the part and he comes from Pennsylvania. Opportunity lost.