Aeimit Lakdawala

Even holds for professional economists. bsky.app/profile/aeim...

(5/5) Paper available here: aeimit.weebly.com/uploads/2/5/...

Co-authored with:

Benjamin Kay (Fed Board) sites.google.com/site/benjami...
&
Jane Ryngaert (Notre Dame) janeryngaert.github.io#!/home

(4/5) Bias is specific to GDP growth forecasts (not inflation/unemployment/rates). Why only GDP? It's harder to predict. When data is clear, politics doesn't matter. When there is uncertainty, partisan beliefs creep in.

Different beliefs about how well tax cuts work is important driver of bias

(3/5) The partisan bias is asymmetric: appearing primarily under Republican presidents.

Bias also comes at the cost of forecast accuracy. Republican affiliated forecasters make relatively larger forecast errors when there is a Republican president.

(2/5) We matched Wall Street Journal Economic Forecasting Survey participants to their political affiliations using publicly available data (voter rolls, donations, partisan employment). The consensus forecasts from WSJ survey are similar to commonly used (but anonymous) SPF and Blue Chip.

NEW PAPER: We know household expectations show a strong partisan bias, but do professional economists let politics cloud their forecasts too? Surprisingly, we find yes!

Republican forecasters predict ~0.4 pp higher GDP growth when Republicans hold the presidency vs Democratic forecasters. #EconSky

Interested!

Impressive!

Not completely disagreeing with you here. Just think it’s going to be too hard for a moderate like Shapiro to get folks excited.

Personally I would prefer someone like Ezra Klein but we know that’s not happening. Given the constraints of the political system I would much rather see AOC than Shapiro, Whitmer etc…

There will likely be anti-elite backlash from a lot of the Trump policies and she is well positioned to take advantage of it. My opinion is that there is a lot of appeal to populist candidate that runs on “make the wealthy pay their fair share” on both sides of the aisle.

I think she is the most “ authentic “ candidate the dems can put forward really, And I think that seems to matter a lot.

You know things are real bad when Jason Furman starts getting spicy!

An emergency rate cut from the Fed? A fixed income freakout

FT piece on this: www.ft.com/content/e8b5...

The rise in 10y appears to be due to market microstructure" or plumbing issues related to the basis trade. Will be interesting to see if 10y yields move lower once these are resolved (assuming fed funds expectations stay low).

10y Treasury

CME fed watch

10 year Treasury at 4.35% (up from 3.9% just a couple of days ago). At the same time, expectations for fed funds rate at the end of 2025 have gone down from ~3.6% a week ago to ~3.3%

10 year Treasury back below 4%

down to almost -3% now , but again seems to be over-reaction to early data.

Yes, was just being facetious. Although won’t be surprised if the “doge in dsge” title is actually used.

Who's working on "DOGE in a DSGE model" paper right now? #econsky

Seems to be related to trade data and surge in imports.

seasonality issues likely to be important here. Some evidence of slowdown in recent data but nothing of this magnitude.

Biggest decline in GDPNow in a short time that I can recall (excl. covid). Estimate now is for a 1.5% fall(!) in GDP in Q1-2025

You should let us know here which parts you agree/disagree with.

In Widows, stataRun seems to conflict with Copilot somehow. Can't figure it out.

Has anyone successfully figured out how to execute Stata from VS Code in Windows? #EconSky

Congrats! Super cool paper.

Intro slides from when I used to teach 3rd year PhD students. ec831.weebly.com/uploads/2/5/... Not sure this is what you’re looking for but has a couple of simple examples that may help.

100% agree with this. Empirical papers especially are typically messy. It’s ok if all the robustness tests do not work out perfectly. As a referee I often request authors to include “negative” robustness results. But as an author when I have tried this myself it has not been easy to publish.

Why even bother with the *