The presidential election is now officially one year away. So for this week’s 2016 Slack chat, we chewed over each party’s general election prospects. What do we know a year out? As always, the transcript below has been lightly edited.
Check out our live coverage of the GOP debate.
micah (Micah Cohen, politics editor): All right, we’re one year out from the 2016 general election; what’s the outlook? Does either party have an advantage?
natesilver (Nate Silver, editor in chief): No. Good chat, guys!
harry (Harry Enten, senior political writer): Here’s what I would say: The generic presidential ballot shows a slight GOP advantage. YouGov has tested a generic Republican presidential candidate vs. a generic Democrat four times since the beginning of September. GOP holds a 2.25 point average lead. That suggests to me, at the least, that the Democrats do not have an advantage heading into next year.
natesilver: I’m open to explanations about why the probabilities are slightly different from 50-50. But I’m not sure why I should care about the generic ballot.
harry: Well, I think it gives us a general idea of the political environment overall. And it reflects the president’s approval rating as well. The rough line of when a president’s approval rating helps or hurts a candidate from his own party is about 48 percent.
Obama’s approval rating right now is 45 or 46 percent. Both of those numbers indicate to me that the environment is probably a little more favorable to the GOP than the Dems.
Not greatly so, but a little bit so.
natesilver: I’m not sure I’d say that 45 or 46 percent is meaningfully different from 48 percent. Not a year out, anyway.
If you run the numbers with Obama’s favorability ratings instead, for instance, you get a different answer.
harry: Let me ask you this: Is the difference between a 55 percent chance of the GOP winning meaningfully different than a 50 percent chance?
natesilver: Do you want me to get existential here?
micah: YES!
natesilver: Sure, it’s meaningful if it really were a difference between 55 percent and 50 percent. Something that made a 5 percentage point difference in the likelihood of Democrats or Republicans winning would be way more meaningful than 99 percent of the stuff that pundits call “game-changers.”
However.
These fundamentals-based presidential models kind of suck. They’re not nearly as precise or as accurate as they claim to be.
harry: Most of those fundamental models are based off economic measures. I don’t think I put any economic numbers in this so far.
natesilver: I guess I look at it more like this. My prior is that elections with a term-limited incumbent are 50-50. I’m looking for evidence that persuasively overcomes that prior. An extremely popular or extremely unpopular incumbent would clearly matter. But Obama’s popularity is about average.
micah: But this was one of my questions: Obama isn’t running; how much of an effect will he have on the race? Positive or negative — is Obama’s popularity really a big factor?
natesilver: He’ll have a fairly neutral effect, given his current popularity level.
harry: We only have approval rating data at this point in a campaign (September/October the year before) for six instances when an incumbent president didn’t run for re-election. Now, I took those and plugged them into a simple little regression. With Obama’s approval at 46 percent, the GOP is expected to win by about 2 percentage points. Again, there’s a huge margin of error, but signs point to a slight GOP edge.
natesilver: Dude. It’s not even six examples really. It’s four.
harry: Who are your four?
natesilver: Dwight Eisenhower, Ronald Reagan, Bill Clinton and George W. Bush are the only presidents in American history to be term-limited. Obama will be the fifth.
And I don’t care if you get the same regression results with four.
harry: And did you know that Obama’s approval rating is below the average approval rating for those four? And it’s not particularly close either.
natesilver: The problem is that running a regression model based on an n of four is inherently kind of ridiculous.
NERD FIGHT!
harry: The average approval rating of those four is 59 percent. Obama’s is 45 or 46 percent. I’m not saying this is anything close to a be-all end-all predictor.
But the evidence we do have suggests something slightly on the GOP side of 50-50.
natesilver: OK, but there’s other evidence that possibly points toward Democrats having a slight edge.
harry: Such as what? The blue wall? Your favorite blue wall?
micah: PERMANENT DEMOCRATIC MAJORITY!
natesilver: No, the blue wall is bullshit. Or, at least, mostly bullshit. But if you’re talking about minutia on the order of a 46 percent vs. 48 percent approval rating, then maybe there’s a very very small Democratic Electoral College advantage.
harry: Why do you think that advantage exists?
natesilver: Because they did have an advantage in 2012 and 2008, if you look at where the tipping point state was relative to the national popular vote.
natesilver: Those advantages are NOT very sticky from election to election. But, again, if we’re talking about shit that moves the probabilities by 5 percentage points one way or the other, then maybe?
harry: Now, let me state a point of agreement here: I concur that those advantages are not very sticky from election to election.
natesilver: It’s all minutia, and I don’t think we should be concerned about a minutia a year out.
harry: This site should be about arguing over the small stuff sometimes!
micah: All right, somewhat related question: We’ve already heard pundits and politicians say, “It’s very hard for a party to win the White House three elections in row.” I guarantee that will be said millions of times from now to November 2016. Is it true?
natesilver: The White House is not a metronome.
micah: So no?
harry: Small sample size on that one. I don’t agree with the concept that winning a third term is inherently more difficult. And, moreover, looking at the generic presidential ballot and Obama’s approval, this looks more like a close election than one that is clearly one-sided from start to end, a la 1952 or 2008.
natesilver: You have four examples of term-limited presidents. If you look for examples before the 22nd amendment was adopted (which I guess you have to do when you have a sample size of four): Elections with retiring incumbents seem to be about 50-50.
micah: Of course, who the parties nominate could change those numbers. It seems like this could be an election where the candidates make a huge difference, right? Let’s say Hillary Clinton wins, whether she faces Donald Trump or Ted Cruz or Marco Rubio will have a big effect on the odds.
natesilver: Yes. That’s the proverbial and maybe literal elephant in the room.
micah: What order of magnitude are we talking about?
harry: Yes, this is the question. This is one where I could find myself in agreement with Mr. Silver on whether it’s a 50-50. Cruz, for example, would be a historically conservative candidate. If he’s the GOP nominee, that could be worth a few percentage points and harm the Republicans.
natesilver: Which, in an election that otherwise looks about 50-50, could make a lot of difference.
If Clinton has a 75 percent chance of facing a 50-50 election, and a 25 percent chance of facing a 75-25 election (e.g., against Cruz, Carson, Trump, or a GOP electorate that gets all screwed up because one of those guys runs as a third party), then her overall chances of winning are 56 percent.
natesilver: Now, I think you can argue that Clinton would be a slight underdog against Rubio, for instance.
micah: What about vs. Jeb! Bush or John Kasich or Chris Christie?
natesilver: Sure, Kasich, in particular. I’m less sure about Jeb or Christie, just because their personal ratings have been pretty bad for a long time.
But Clinton’s not very popular either, obviously.
micah: OK, let me see if I have this right …
One year out, the election is probably about 50-50 (maybe 55-45 Republican, according to Harry), but that could be tipped toward the Democrats if the Republicans nominate Trump/Carson/Cruz or toward the Republicans if they nominate Rubio or Kasich. Moreover! Obama, with middling-but-not-horrible approval ratings, won’t have a huge effect on the race (also, the “it’s hard to win the White House three times in a row” maxim is bullshit).
harry: I think that’s mostly right, though I will say to the degree that Obama has an effect, it likely hurts, not helps, the Democratic candidate at this point.
natesilver: That’s basically right. Almost everything we’ve been talking about wouldn’t shift the probabilities much from 50-50. Maybe you can argue that slightly more of the tiebreakers line up on the GOP side instead of the Democratic side; Harry’s more convinced about that than I am, I guess.
But there’s some probability of something that would make a huge difference, which would be the GOP nominating Cruz/Carson/Trump. Or nominating an establishment candidate after a huge, messy nomination fight, that might include a brokered convention or something close to it. In those cases, Clinton — for all her flaws — is a clear favorite.
harry: Clear favorite undersells it, IMHO. (Of course, now I’m just being difficult.)
Check out our live coverage of the Republican debate.