Nate Cohn

Nate Cohn @Nate_Cohn

Would you support or oppose the following electoral system? (described in following tweets; poll in this tweet) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus as i recall, you did not think it was worth expanding the dataset before 1948 or so over the summer--which is a defensible position and one in keeping with your position here. but it also means we probably don't ever get a ton of macrolevel data — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus if you think polarization has fundamentally changed the model--and maybe it has--then we'll probably never collect enough elections in a single political era — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus i don't think we can — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus i think it is entirely possible, even likely, that it has constrained the effect of the economy on elections. how much? dk. and of course we'd want to know what the effect of the economy was before polarization, as well — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus the polarization point--whether our fundamentals 'model' of midterm elections changed over the last 12 years or so--is an important and serious one. it's also just a really hard question, and one that also makes it harder to answer other questions — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus (all of this separate from your polarization point, of course) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus (you may recall that was my remedy in the presidential election context, interestingly, as you wind up adding many of those seeming 'outlier' cases like 9/11, Vietnam, etc. which would probably really effect a presidential election but didn't happen in years divisible by four) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus i think people in this thread already offered one: the plain evidence that the economy affects presidential approval, which is much easier to see when you expand beyond n=17 to the dataset to the whole of a presidency — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus if it is a driving factor--and i'm not sure it is, but very well could be--i think it would be very difficult to identify with such a small number of data points and our inability to specify a full model of presidential approval — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus if it is a driving factor--and i'm sure it is, but i think it very well could be--i think it would be very difficult to identify with such a small number of data points and our inability to specify a full model of presidential approval — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus and well that's the disagreement. you're convinced there's basically no effect, though you wouldn't rule it out at the margins in a close race. i think it's just as likely that it could be 'a' driving factor with the potential to be decisive in many contexts — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus i think most of the data that you've introduced suggest it's not 'the' driving factor, but you keep making arguments that dismiss the idea that it's even 'a' driving factor, and i just don't think you have the goods on that — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus i don't think it's the 'driving factor' (i don't even think that's obviously true of presidential elections, as you know). but i am very open to the idea that it's 'a' driving factor, and wouldn't say someone was wrong for believing Ds could do well in '22 on that basis — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus i think we disagree on whether it's not 'a driving' factor v. 'the driving' factor — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus and we talked about this a lot over the summer, so i won't recapitulate my views about how to think about this, but you can start by crossing off the years with military deployments/operations to see how it applies in this case — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus one possibility is that the economy doesn't matter in even number years that aren't divisible by four; the other possibility is that the economy always matters quite a bit, but it's one of many factors, and so when you reduce to n=17 you don't have a clear relationship — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus and that's leaving aside that the reasonable argument advanced by others in this thread--that there's plenty of reason to think that the economy affects evaluations of the president, and everyone agrees presidential approval is predictive — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus but in this thread you've made a bolder argument that shifts the burden of proof. you're not simply arguing that the economy isn't the dominant factor; you're arguing it's basically irrelevant if it doesn't clearly show up in a regression of n=17. and that doesn't follow! — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus there are two things separate things in this thread. you have noted, correctly, that there is, at best, a fairly limited relationship between the economy and midterm outcomes. that's been known true for decades — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus (there are 17 on the scatterplot) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus (and that's leaving aside that different model specifications may yield some pretty different results!) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus well of course I wouldn't predict midterm results based on GDP and GDP alone! but i also wouldn't conclude that material economic conditions have no effect on the president's party based on 15 data points — PolitiTweet.org

Nate Cohn @Nate_Cohn

@gelliottmorris @julia_azari @smotus not to rehash our debate about the fundamentals from the summer, but here again there just isn't anywhere close to enough data to reach the conclusion your advancing in this thread — PolitiTweet.org

Nate Cohn @Nate_Cohn

This is the only response I respect (sry to david rose memes) — PolitiTweet.org

Nate Cohn @Nate_Cohn

A merited question. I am saying it to Canada — PolitiTweet.org

Nate Cohn @Nate_Cohn

Canada: we’re always here, when you’re ready — PolitiTweet.org

Nate Cohn @Nate_Cohn

@richardmskinner @DouthatNYT well i understand why someone could still be upset at the decision, but ross's column asked why free speech liberals aren't more upset about it, not cultural conservatives — PolitiTweet.org

Nate Cohn @Nate_Cohn

@DouthatNYT and isn't the seuss estate equivalent to the author in this process, which would presumably be even less scary? this position would make more sense to me if it was proceeded by big anti-seuss protests, but if that happened i missed them entirely — PolitiTweet.org

Nate Cohn @Nate_Cohn

@DouthatNYT here's what i don't understand: isn't it fundamentally different--and fundamentally less scary--if a book is 'disappeared' by a publisher choosing not to publish, which is not only well within their right but a kind of decision they're expected to make all the time? — PolitiTweet.org