That’s close to what it looks like to me - incumbency is an advantage when things are good and a disadvantage when they’re bad. I don’t agree it’s purely the economy - I think that things like the Iran hostage situation of Carter or Vietnam for Johnson played a significant role in their losses. (I certainly count running in the primaries, then running out when you don’t do as well as you hoped as a loss). In the opposite direction, W Bush had a middling economy (economic growth but not jobs) and some major bad points that could have been significant against him (Iraq War and Patriot act), but the Democrats ran Kerry against him, who’s ability to capitalize on those points was compromised by the fact that he had voted for both.
In the absence of good data, the null hypothesis is that the presidential election is just like every other election, which we have lots of good data for, not that it is a special unique snowflake that’s not subject to the same effects that result in an incumbency advantage everywhere else.
If you define a question to be specific enough that there’s little data for it, you can certainly convince yourself that it’s a special case that all the other data doesn’t apply to. And you might be right. But it’s not particularly good science or statistics.
Not sure I follow this. If your theory that the presidential election is somehow special and unique is not valid, then the presidential election is like other elections, and I’ve already provided tons of data for the claim that incumbency is an advantage in elections.
A useful question if you are interested in the answer, is to think about what sort of data would provide meaningful insight. There are a lot of data about presidential elections other than the binary answer of who the winner was. You could elaborate a theory of which factors are relevant to an election and develop a model, etc.
Nope. The null hypothesis is that it’s neither an advantage or disadvantage. If your position is that the null hypothesis is the thing you want to be true but have no evidence of, I don’t see how we can have any kind of productive discussion.
I agree that it’s a mistake to draw firm conclusions about the Presidency from data about other elected positions. However, I also disagree with the OP’s interpretation of the presidential data, because I don’t think that Johnson and Ford should be considered “incumbents”.
People often think of “incumbency advantage” as “whoever’s currently in the office has the advantage”. But that’s mostly not how it works. Mostly, people vote for the candidates that they like, and the guy in office is there in the first place because people liked him last election, and so this election they’ll probably continue to like him. In other words, if they elected him once, for whatever reasons, they’ll probably elect him again, for the same reasons. But it only works for people who were elected. When Johnson was elected Vice President, it mostly wasn’t because people particularly liked him: It was because people liked Kennedy. And Ford famously was never even elected to the Vice Presidency at all.
However he did proceed to win a term in his own right in 1964. The '64 election vs. Goldwater was the one with carryover from the JFK term. Election '68 featured incumbent LBJ in his own right, and very early in the primaries he realized he did not have the momentum to come out of them on top.
Nope right back. There’s no way to read iamthewalrus_3’s statement as not being exactly the same as your statement. You’re both saying the same thing.
So, we agree that incumbency is an advantage for elections that aren’t the presidential election, right? I found a paper, we both think it shows that?
So, the question then is: Is the presidential election a normal election subject to incumbency bias, or is it special in some way that makes incumbency bias irrelevant (or an actual detractor).
Again, you can always define a question narrowly enough that you think the data that exists doesn’t apply to it. But we have lots of data about elections.
Like, let’s consider the following series of questions:
Is incumbency an advantage for Representatives? Yes.
Is incumbency an advantage for Senators? Yes.
Is incumbency an advantage for Mayors? Yes.
Is incumbency an advantage for Governors? Yes.
Is incumbency an advantage for State Comptrollers? Yes.
Is incumbency an advantage for left-handed Episcopalian former-trucker City Council Members named Steve? Gosh, no way to tell. We just have too little data.
I’m starting to agree with you.
I call both 9 and 0 data points garbage for predictive purposes, so I guess we’re just arguing about how bad it smells.