Okay, I’m baffled.
Joe Biden’s estimated approval over at 538 dropped to a new day-end low this week (via the invaluable Taegan Goddard’s Political Wire) at a fairly miserable 37.4%, although it’s since recovered a bit and sits at 38.0% as I write this.
The thing is that for a while this year it appeared that Biden was on a slow uphill path. He spent much of April around 39.5%….not great of course, but it was easy to build a sensible and (for him) optimistic story around it. The continuing pandemic and then supply chain issues and high inflation had destroyed his popularity, but with inflation fizzling and unemployment still low, good economic news was finally beginning to help Biden recover. Granted, it wasn’t yet clear that the upward trajectory was real, but if it was it would make sense, and it should have continued as long as the news remained basically the same.
Well, as far as I can tell the economic news is still pretty good and high inflation is receding, but Biden has apparently slumped. Why? I have no idea. It just doesn’t really make sense to me that June 2024 should be his low point.
Look, the big lesson here is probably that I should stop checking approval ratings day-to-day, but I know I’m not going to do that and it’s likely that at least some readers are just as foolish as me. So here’s what I’m thinking about what might be happening, even as I’m trying to remind myself not to start analyzing short-term polling fluctuations.
First: Maybe it’s not real. The 538 polling average is good, and it attempts to correct for polling firm biases, but it’s still vulnerable to a run of surveys that come in particularly high or low. Either because several pollsters with similar biases happen to release results on top of each other — or because several polls randomly happen to show up off in the same direction. I don’t really see signs of that here, but I suppose it’s possible.1
Second: Maybe it’s not real for a non-random reason. The leading suspect for changing reactions to public opinion surveys this month would be Donald Trump’s conviction in New York. Perhaps that has Trumpy voters so upset that instead of trying to avoid getting polled, they’re eager to register their feelings, thus (theoretically) pushing Biden’s approval numbers down.
That’s plausible, but there’s a problem: Over at 538’s November trial heat average, Trump if anything has lost a little ground against Biden head-to-head since the May 31 conviction. So if it’s about temporary shifts in response rates, we’ll need a fairly complicated story: Trump supporters are showing up in unusual numbers thus pushing Biden down a bit, and also hiding a more significant anti-Trump shift in head-to-head polling. Is it possible? Sure. Likely? I wouldn’t think so.
Third: Biden’s dip in approval is real, and some event in the news triggered it…but frankly I just don’t see it. Gasoline prices are down over the last month. The news from Gaza and Ukraine has been grim, but not particularly more grim since April than before. The roll-out of his asylum policy isn’t the type of thing that typically would change approval ratings. Are there people who moved from approving of Biden to not approving because Trump was convicted of state charges in New York? I’m skeptical. And I’m more than skeptical that the Hunter Biden trial has had even a mild short-term effect of any kind.2
Fourth: The dip is real, and some event we don’t know about yet triggered it. This would be an economic downturn that people are feeling, but hasn’t yet appeared in the main statistical reports. If this is the case, and it persists, Biden is in very big trouble in November. But as far as I can tell, it isn’t true. For example, the inflation number out Tuesday didn’t show any sudden spike; indeed, the contrary.
It’s still possible that the economy is worse than the numbers would indicate, which might account more broadly for why Biden is stuck at or below 40% approval, but there doesn’t seem to be any further negative change.
So what’s going on? If I had to guess, my money would be on the first one — it’s just random variation, and we’ll see that soon enough. But no question that it’s harder to believe now than it was six months ago that Biden can count on delayed reactions to an improving economy to make him significantly more popular before the election.
At the same time, Biden is close to tied in head-to-head polling despite his lousy approval numbers.3 Which just gets me back to what I said a few weeks ago: There are so many unusual things about this presidential election cycle that we should all be very cautious about everything. Or as Jamelle Bouie said on Wednesday: “I can’t read any analysis of this presidential election without concluding that literally no one knows anything about what will happen in November other than one of two people will be the winner.”
Of course, it’s also possible that Biden is actually doing worse than it seems, with recent polls actually coming in too high. That’s the thing about this kind of speculation; it’s important to remember that random error can go either way.
Remember, we’re looking here for something that would have dragged him down from where he was, not for what made him unpopular in the first place. So it’s not, say, housing — unless something recently got (even) much worse about housing.
Yes, Biden probably has to win the national vote by 2 or 3 percentage points to win the elecctoral college. Even on that, however, there’s more uncertainty than people think. And yes, in both directions.