Tipsheet

Something About Gallup's Latest National Survey Data Isn't Adding Up

Guy Benson
Guy Benson  |  September 30, 2024 10:25 AM

I'm trying to reconcile Gallup's latest national survey data, about which we've done some writing already, with modeling from 'experts' who continue to cast Kamala Harris as favored to win the election.  We noted last week that the Gallup numbers show Republicans leading for the first time in the third quarter of a presidential election year on the broad metric of party identification.  Including leaners, 48 percent of respondents told the firm that they identify more as Republicans, while 45 percent said they lean toward the Democrats.  As others have noted, since at least 2008, this particular Gallup outcome has presaged the subsequent national vote results -- to an uncanny degree.  If you missed it in our last post, take a look at this:


Around 2004, the party ID levels were tied, and George W. Bush was re-elected by roughly three points. But it's not just the top line party ID numbers that caught my attention.  Gallup asked Americans which party they trust more to handle the issue they care about the most -- whatever that issue might be.  The GOP holds a five-point lead on that question.  For the last 20 years, every time Democrats lead on this question, they've won the White House.  Every time they they've trailed Republicans on it, they've lost.  They trail on it this year. On which party is more trusted to make America prosperous, a generalized economy question, Republicans are up by six points, 50-44 percent.  On a question on keeping the country safe, the GOP's advantage is larger, 54-40 percent.  A majority of Americans say they want the federal government to do less (55 percent) versus more (41 percent), Joe Biden's job approval is just south of 40 percent, views of the economy are sour (-28 percentage point index scale), and Republicans have a tick higher favorability rating than Democrats overall.  On nine of ten issues polled, the GOP leads.  No wonder Gallup's own headline on their data is, '2024 Election Environment Favorable to GOP.'

And yet, Kamala Harris has a two-point national lead over Donald Trump, according to the RealClearPolitics average.  In aggregate, across the seven major battleground states, the two campaigns are almost exactly tied.  Now, if 2016 and 2020-level polling misses (or even close) are replicated this cycle (in the same direction), the current numbers would likely point to a Trump win.  But if pollsters have corrected some of their persistent issues, and if the polling is either basically accurate or the misses favor Democrats this time, Harris would be on track to prevail.  Regardless, if Gallup's major indicators, which they've been tracking for decades, are on point, why isn't Trump winning comfortably?  All of the data I relayed above should suggest he's the clear favorite.  If he ends up losing, there will be some very uncomfortable conversations and recriminations within the Republican Party, which should not have any business losing an election in which they're so well situated, per Gallup's polling.  If the party, and its leader, underperform such powerful historic trends, that will be the biggest red flag we've seen in a long time -- even including underperformances and losses in 2018, 2020, 2021 (Georgia), and 2022.  

On the other hand, is Harris really the frontrunner?  Over the last few days, we've seen fresh Fox News polls giving Trump a one-point edge in hyper-crucial Pennsylvania, as well as slim leads in North Carolina and Arizona (albeit with Harris slightly ahead in Georgia, bucking virtually all other polling).  The New York Times/Siena polling shows Wisconsin extremely tight, and Michigan exactly tied, showing notable erosion for Harris since their last surveys of those states.  For reference, just before the 2020 election, NYT/Sienna gave Biden an 11-point lead in Wisconsin (a worse than ten-point miss).  Meanwhile, Atlas Intel, a polling firm that bills itself as the most accurate of the 2020 cycle (with receipts) is out with a new slate of polls.  I take all of these results with grains of salt, as all polling should be viewed, but why should we disregard numbers like this when we endlessly discuss polls churned out by organizations that produced much less accurate results four years ago?


Atlas' national margin -- which was almost exactly correct in 2020 -- reportedly has Trump leading by 2-3 points.  Maybe that's totally wrong and they'll look far less impressive after November 5th.  However, consider this:

Recommended

Someone Should Tell That Bucks County Dem Where She Can Shove Her Shoddy Apology Matt Vespa
Why Again Do We Still Have a Special Relationship With the Tyrannical UK? Kurt Schlichter


Look at that trend line.  Across all four Hispanic demographics highlighted by NBC, Trump has gained on Biden's 2020 margins, including pulling into a tie among men, and within ten points among young Hispanics.  Again, maybe none of this actually plays out when the votes are counted, and maybe Democrats' ballot operations and ground game will cut into whatever progress Republicans seem to be making.  But do you remember this poll, commissioned by published a few weeks ago by the NAACP, on black voters?  It showed Harris under-shooting Biden's margins against Trump, with Trump's standing improving slightly, especially among black men under the age of 50 (among whom he garnered 26 percent support in the survey).  These are all imprecise projections, but if Kamala Harris is in fact failing to hit Biden's 2020 victory coalition benchmarks among multiple demographics, is she making up ground elsewhere?  Among whom?  

Maybe she'll do better with young people, single women, and suburban women than Biden did.  But Biden carried those groups strongly last time.  Maybe her turnout will be more robust and effective than Biden's in 2020.  Is there evidence supporting those theories?  There is evidence of Trump making progress, sometimes dramatically, among important constituencies and demographics (including the Teamsters, to take another recent example).  Those gains could prove to be mirages, or they may be offset, but dismissing them out of hand as we analyze this race seems foolish. I'll leave you with this observation from watching a college football game over the weekend:


"Interesting choice on the dig, considering the recent assassination attempt, in which Trump was targeted by one of her supporters at a golf course."