I work in an information-heavy professional field. The world of data data data and more data, team-shared spreadsheets in cloud storage, monthly and quarterly metrics, predictive modeling, and all of that. And since just a year after the founding of DailyKos, I’ve been a participant here (Torta is my 2nd screen handle, started in 2004). I’ve followed “the polling” for each Presidential election with great interest since the Clark/Dean wars of 2003 gave way to Bush 2 defeating John Kerry in November 2004.
I sincerely appreciate all the great work the Good News Roundup folks do in here, and everybody else. And I hope and predict that the 2024 election season has more twists and turns to come, and some of them are likely to hurt Trump. Yet I can’t help feeling troubled on two fronts, at the moment.
(1) Trump and Biden are currently polling head to head at an even level, more or less. That is 4 points worse for us than Hillary Clinton ‘16 vs Trump, and 8 points worse for us than Joe Biden ‘20 vs Trump. And,
(2) Perhaps in response to this, a surprisingly large and loud chorus of voices on DailyKos is proclaiming “this polling is all corrupt and inaccurate”, “polls are worthless,” and “ignore the polls.”
Polls, collectively speaking, are not worthless. We have a depth of information here to work with.
And this is what it tells us.
Polling, 2016. Contrary to the narrative that “the polls were wildly inaccurate in 2016”… the final National polling average in early November 2016 had Hillary Clinton approximately 4 points up over Donald Trump. Hillary Clinton won the popular vote by 2.1 points. The polling was off by… only two points. That’s it. Two points. Does everyone understand this? And the result was just enough to play out electorally in just a certain way in Michigan, Wisconsin, Pennsylvania, giving Trump thin victories there and giving him the US Presidency.
2016 Takeaway: If our candidate is nationally “4 points up over Trump,” that is likely to actually mean only 2 points up, and it is likely to mean a razor thin fight across multiple battleground states. 4 points up was not enough for us, in 2016.
Polling, 2020. Things looked great for Joe Biden heading into November 2020. He was 8 points up in the final National polling average on the morning of Election Day. And so it landed: Joe Biden defeated Donald Trump decisively. It still took a day to call it, but it was a thorough victory. But here’s what got lost in the celebrating: Joe Biden won the National popular vote by 4.5 points. The polling was off by… 3 or 4 points. In Trump’s favor. I’m not sure enough of us really noticed that, at the time, and connected it to the 2016 result and understood it.
2020 Takeaway: Our candidate was polling nationally at “8 points up over Trump”, but that actually meant only 4 or 5 points up over Trump. And it turns out, that’s absolutely good enough for both an electoral victory and a popular vote victory.
So here we are. It’s only April. Things can change. But they pretty much didn’t, between April and November, during the last two cycles where Trump was our opponent.
And if the election were to happen tomorrow, then — based on the vast storehouse of polling information over the past decade — predictive modeling suggests very strongly that Trump would defeat Biden.
We need to see polling that has Joe Biden up at least 5 points over Donald Trump. And preferably 7 or 8 points up. And we have just six months to go, even a bit less in some states when you factor in Early Voting.
I hope and pray, fervently, that things change for the better. And we need to do what we can to make that change happen. Maybe it’s volunteering, maybe it’s donating, maybe it’s both. And I also heartily recommend a personal “each one reach one” outreach to the disaffected, disconnected and less-likely-to-vote people you and I might know, in our everyday lives.
But I tell you this much: sitting back and scoffing “polls are worthless”, “these polls are bullshit”, etc, is a grave mistake. Please, I beg you, please do not do that. One individual poll can certainly be junk. But over the long haul, it’s a body of information that has great value. We have information here, mountains of it. And we have brains to assess its accuracy/inaccuracy, and the ability to extrapolate and predict from there.
Joe Biden is doing a superb job as President and as our 2024 Nominee. And yet — all the data here is warning us that this is a “perilous fight,” to quote the national anthem — it will be anything but a confident cakewalk. Based on what voters this year are telling pollsters, over and over, Donald Trump’s pathway to the 2024 presidency is in better shape than his 2016 pathway was.
As for things out of our control… the Trump Trials will hopefully have an impact; it is unprecedented, to say the least, that a defeated ex-President tries to run for office again while facing multiple court trials as a criminal defendant. And Trump’s increasingly unhinged rhetoric and demented-seeming mental state will hopefully also have an impact. And there are always unexpected events.
But it’s April 2024 and here we are. With a lot of polling information, current and 2020 and 2016, information that — when connected together — tells us things we may not want to be hearing. But it’s real.