Paul Testa

Assistant Professor of Political Science at Brown. Tweets about politics of Race and Criminal Justice. Occasional cat photos.

[Paul Testa @ProfPaulTesta]

One last time. Tuxedo returns to form, khakis making their first and last appearance. The combination of crocs, jersey and bow tie remains undefeated.

Eternally grateful to have taught such wonderful students, even if their sartorial suggestions are a bit suspect.

[Paul Testa @ProfPaulTesta]

Tweed upsets tuxedo, but crocs, jersey and bowtie remain undefeated. What would @dieworkwear say?

Thanks as always to my fashion savy students in https://t.co/yhYg9mWIV0 #LetEmKnow

[Paul Testa @ProfPaulTesta]

Tweed upsets tuxedo, but crocs, jersey and bowtie remain undefeated. What would @dieworkwear say?

Thanks as always to my fashion savy students in https://t.co/yhYg9mWIV0 #LetEmKnow

[Paul Testa @ProfPaulTesta]

Apparently, khakis are in right now, but the pairing of tuxedo, crocs, bow tie and soccer jersey remains a timeless combination. Thanks https://t.co/BwjiEndcvH

[Rob Blair @robert_a_blair]

What do elites want from the foreign aid they receive? Conjoint experiment in 141 countries w/ @samanthajcuster and Phil Roessler, now online @AJPS_Editor https://t.co/xr5c1CvqVs

[Michael Rios @MikeBRios]

Election narratives, meet data analysis 🔍 Our team at @ucla_vrp is breaking down election night trends, and this is just the beginning. Stay tuned for more insights—follow us to stay informed. 🗳️📊 #Election2024

[Tim Bale @ProfTimBale]

Worth remembering before the blame-game gets too personal: "Democrats join 2024’s graveyard of incumbents", @jburnmurdoch https://t.co/2S3Y9JMpk6 via @ft

[Robert Ford @robfordmancs]

Decided to go through this systematically. Incumbent government performance in wealthy democracies since March 2022, when Ukraine invasion really spiked things upwards: 🧵

[John Sides @johnmsides]

My initial take on Trump's win:

https://t.co/OagyPxsFFg

His gains were widespread, so explanations should start with the broadest factors -- not with bespoke stories about states, cities, counties, and groups.

The simplest explanation: party of unpopular incumbent loses.

1/2

[Efrén Pérez @EfrenPoliPsy]

I don't have many strong and informed opinions, but when it comes to PoC politics, that's one exception because I study them. Lots of ill takes here on PoC & the election. Rather than criticize, let me provide some evidence-based input on the election and the preceding months🧵

[Taubman Center @TaubmanCtr]

— Dr. Paul Testa, Assistant Professor of Political Science at Brown University. Watch here https://t.co/dUDC5sLxD5

[Paul Testa @ProfPaulTesta]

@almostseamus 🐳🐳🐳

[Paul Testa @ProfPaulTesta]

Seeing lots of folks dunk on this tweet for various reasons. The analysis is probably not great, but not for the reasons people think.
🧵

[Nate Silver @NateSilver538]

There are some confounders here, but the inflation thing is actually statistically significant! Each additional $100 of inflation in a state since January 2021 predicts a further 1.6 swing against Harris in our polling average vs. the Biden-Trump margin in 2020.

[Paul Testa @ProfPaulTesta]

@tcarpenter216 I felt similarly. Here’s my attempt at a constructive critique:

https://t.co/VdBnYYufaI

[Paul Testa @ProfPaulTesta]

Seeing lots of folks dunk on this tweet for various reasons. The analysis is probably not great, but not for the reasons people think.
🧵

[Paul Testa @ProfPaulTesta]

@ekrembaser Thanks Ekrem! Hope you’re well 😀

[Ben Casselman @bencasselman]

Nate's data is coming from this tracker from the @JECRepublicans. They don't have a state-level inflation estimate either, though. They just use BLS's estimate of regional inflation and apply it to an estimate of household spending when Biden took office.
https://t.co/9krefDNnjZ

[Paul Testa @ProfPaulTesta]

@jonahbgelbach Yeah, this was my best attempt with polling averages he shares at Silver Bulletin. Not sure what's the added state in these models. Could also be a difference in calculating the 2020 margins (two party vs total votes) But seemed close enough for Twitter

[Paul Testa @ProfPaulTesta]

@marquesgz lol. I ❤️ Marques

[Paul Testa @ProfPaulTesta]

Peter's right (of course), but I still think the larger issue is more with the measure of inflation, than leaving out ",r"

[Peter Hull @instrumenthull]

@ProfPaulTesta The point is not that robust and homoskedastic SEs may be similar in practice. The point is there's basically no reason to use the latter when the former (or better HC's) is available. And Nate should know this

[Paul Testa @ProfPaulTesta]

@Brendan_Duke It was your post (and point) that put the bee in my bonnet. Thanks for explaining the JEC measure!

[Paul Testa @ProfPaulTesta]

@instrumenthull But have you have you done a qq-plot of your residuals and tested the normality assumption?

[Paul Testa @ProfPaulTesta]

And here's a working link to the gist:

https://t.co/STln1mKXQu

[Paul Testa @ProfPaulTesta]

Looking at inflation by Census division there's a lot of variation. Maybe you could say something like Harris is underperforming the 2020 benchmark in richer Blue states, and overperforming in poorer Red states but story for inflation isn't all that clear with these data.