Sign in

Cyrus Samii

@cdsamii.bsky.social
4.4K followers 1.7K following 592 posts

NYU Politics prof. Methods to inform policy. Experiments on governance and conflict management innovations. cyrussamii.com

PostsRepliesMedia
Cyrus Samii @cdsamii.bsky.social · 07/10/2026
Time for you to write up your take!
020
Cyrus Samii @cdsamii.bsky.social · 07/10/2026
Re the comparison to calculators etc., I think with LLMs it’s a qualitative leap in capabilities. Then with regard to disclosing researcher vs LLM, the issue is with regard to claiming credit given the moment we are in.
100
Cyrus Samii @cdsamii.bsky.social · 07/10/2026
If you look at P’s note, the disclosure itself contains a principled withholding of model details particularly because it wasn’t open source.
110
Cyrus Samii @cdsamii.bsky.social · 06/10/2026
A thoughtful provocation: not disclosing about LLM use or nonuse is either dishonest (if you actually did use it) or reflects carelessness (if you actually didn’t use it). pmaronow.github.io/llm-disclosu...
pmaronow.github.io
292
Cyrus Samii @cdsamii.bsky.social · 23/09/2026
In grad school a fellow student once observed: things like this (utility maximization) are not theories—they are not falsifiable—but rather methodologies, ways of representing things. They may then be used to arrive at something falsifiable and at that point we have a theory.
020
Reposted by Cyrus Samii
Aaron Sojourner @aaronsojourner.org · 12/09/2026
Program's up! 30+ teams share evidence abt impacts on health, educ, & econ + community response. Great folks @ziibiing.com @ottergoose.net @aaronrosenthal.bsky.social @chrisuggen.bsky.social @casemcnich.bsky.social @ngraetz.bsky.social @craigstetter.bsky.social docs.google.com/document/u/1...
Documenting the Impact of US DHS Activity in MN

Conference Date: Sept 22, 2026

Hubert H. Humphrey School of Public Affairs

Purpose

In December 2025, the U.S. Department of Homeland Security (DHS) began a significant, sustained agent surge in Minnesota that affected the daily lives of many residents, but had severe impacts on immigrant communities and communities of color.

This one-day, in-person convening provides researchers from across disciplines and organizations with an opportunity to share and discuss research related to the effects of Operation Metro Surge and Operation PARRIS.
722162
Cyrus Samii @cdsamii.bsky.social · 22/09/2026
Great thread with follow up readings to the viral NHJ article.
030
Reposted by Cyrus Samii
Gwyneth McClendon @gmcclendon.bsky.social · 20/09/2026
NYU is hiring in American Politics! Deadline October 6. See more details here: apply.interfolio.com/193724 @rozenas.bsky.social @cdsamii.bsky.social @sanfordgordon.bsky.social @bryantjmoy.bsky.social @awilke.bsky.social @stasavage.bsky.social @jatucker.bsky.social @peterrosendorff.bsky.social
apply.interfolio.com
Apply - Interfolio {{$ctrl.$state.data.pageTitle}} - Apply - Interfolio
02024
Cyrus Samii @cdsamii.bsky.social · 16/09/2026
❤️ Seen in Boerum Hill this afternoon.
Poster with Bert and Ernie in sunglasses riding in car with subtitle The Horrors Persist But so Do We.
050
Cyrus Samii @cdsamii.bsky.social · 16/09/2026
Half of our seminars should be pre analysis plans and the other half results. Imagine if people had an up front stake in your research, and imagine getting great input before you finalize your analysis strategy. It should be the norm.
1122
Cyrus Samii @cdsamii.bsky.social · 16/09/2026
Looking forward to presenting this work on Thursday (not Friday) at UoT. It’s a pre analysis plan actually, with the theoretical model motivating the intervention and details on our measurement, methods, and targeting strategy.
131
Cyrus Samii @cdsamii.bsky.social · 13/09/2026
Horrific.
041
Cyrus Samii @cdsamii.bsky.social · 10/09/2026
Sounds amazing. I hear people these days use “rave” to refer to a night at a licensed, standing, commercial club. Very different from what I knew to be a rave back when I was a young lad!
000
Cyrus Samii @cdsamii.bsky.social · 09/09/2026
Alcaraz v Shelton - wild match. Alcaraz served poorly and made lots of uncharacteristic errors. But there’s no doubt that he was confounded by Shelton’s serve and next level athleticism.
000
Cyrus Samii @cdsamii.bsky.social · 06/09/2026
Jovic v Eala was elite! Wow! And for Jovic to have such a complete game at 18 is breathtaking.
020
Cyrus Samii @cdsamii.bsky.social · 03/09/2026
Tagger v Anisimova shaping up to be as good as hoped. Two smart and crafty players.
000
Cyrus Samii @cdsamii.bsky.social · 28/08/2026
An utterly reasonable take by MIT’s committee on AI and education. The argument about resisting temptation to replace RAs with AI was thoughtful: RAs are an investment in scientific community. I also liked the call to think about spaces and tools for offline learning and evaluation.
061
Reposted by Cyrus Samii
Federica Genovese @fgenovese.bsky.social · 24/08/2026
@acalacino.bsky.social myself and others have crunched the numbers (work in progress, see sources and deets here: osf.io/preprints/so...) and have some hard evidence of that: decline of climate change in news reporting since 2023 is esp. dramatic in the Global North ⬇️
Patterns of climate reporting in Global North vs Global South HQ news matched to GDELT climate tag articles. For notes see "When the Tail Wags the Dog: Climate Exposureand News Media Behavior in the Global South", Anthony Calacino, Mats Ahrenshop, Federica Genovese, Hayley Pring (work in progress)See Calacino, Ahrenshop, Genovese and Pring. https://osf.io/preprints/socarxiv/37juf_v1
28238
Cyrus Samii @cdsamii.bsky.social · 24/08/2026
Or, if a treatment depends on skill to be effective short term clumsiness may mean small effects but long term skill may be larger. Understanding what mechanism will unfold can help predict for the long term.
000
Cyrus Samii @cdsamii.bsky.social · 24/08/2026
Eg, if a treatment is exploitative, short term inelastic demand may mean more revenue but long run elasticity may mean fewer users and less revenue.
110
Cyrus Samii @cdsamii.bsky.social · 24/08/2026
Especially interesting are the examples of when short run doesn’t equal long run, hinting at mechanisms that might be predictable with a richer representation of choice behavior.
100
Cyrus Samii @cdsamii.bsky.social · 24/08/2026
Terrific round up of issues in using short term experimental results to infer long run effects. The issues and references could be a whole course.
150
Cyrus Samii @cdsamii.bsky.social · 17/08/2026
I’d suggest introducing a parameter that allows for for effect heterogeneity. Preserving the null standard error but then just shifting the effect (shift hypothesis) will overstate power when effect heterogeneity is present (which it always is). But nice work getting this together, helpful.
110
Reposted by Cyrus Samii
Jonathan S Doucette @jdoucette.bsky.social · 17/08/2026
What effect sizes can a design reliably detect, and how does that compare to the smallest effect size of interest? These questions are rarely pondered outside of experiments. I (with assistance from Claude) wrote an R package, "simpower", that helps with this: github.com/faranordk/si....
199
Reposted by Cyrus Samii
Paper Skygest Team @paper-feed.bsky.social · 19/08/2025
**Please repost** If you're enjoying Paper Skygest -- our personalized feed of academic content on Bluesky -- we'd appreciate you reposting this! We’ve found that the most effective way for us to reach new users and communities is through users sharing it with their network
2114143
Cyrus Samii @cdsamii.bsky.social · 14/08/2026
They use continuous DID methods to estimate the effects of exposure to the campaign. Of course there are issues with DID identification, but what I find most interesting about the study, and worthy of further consideration, is the attempt to link the micro to the macro.
040
Cyrus Samii @cdsamii.bsky.social · 14/08/2026
Methodologically, what is interesting is how they argue for the relevance of the micro-level outcomes: essentially, it comes from testimonials and private correspondence by key actors at the time, access to which was apparently hard won by Alsan and Neberai.
100
Cyrus Samii @cdsamii.bsky.social · 14/08/2026
A key driver of the dynamics was a campaign commissioned by the American Medical Association to counter the push toward national health insurance and cultivate support for private health insurance.
100
Cyrus Samii @cdsamii.bsky.social · 14/08/2026
They identify a critical juncture during the second Truman admin when a path toward national health insurance was open and when micro-dynamics in public opinion, politician behavior, and rates of subscription to private insurance shaped whether that path would be taken. (It wasn't.)
100
Cyrus Samii @cdsamii.bsky.social · 14/08/2026
The new Alsan and Neberai paper on why the US doesn't have national health insurance offers an interesting template for research on the origins of institutions academic.oup.com/qje/article/...
academic.oup.com
Why Doesn’t the United States Have National Health Insurance? The Political Role of the American Medical Association
Abstract. This article examines how the American Medical Association (AMA) helped shape the development of the U.S. health insurance system in the critical
1133
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
Just thinking about that Beraja and Yuchtman paper (see previous post): politics of regulation hinge on who has more stake in the innovators’ profits vs in recouping value from the appropriated resource.
000
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
Recent examples include "digital exhaust" that was the key input to targeted online advertising systems powering Google, Facebook, etc. or online intellectual property "stolen" to train LLMs by the AI giants.
000
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
This is a "theory of everything" paper that captures so much of the current economic and political moment.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
Empirics include qualitative case studies, trend analysis of property rights disputes in law reviews (showing they follow innovation trends), and simple counterfactual analyses of appropriation rents' effects on profits and growth. The case studies do most to establish the relevance of the theory.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
Crucially, appropriation rents depend on other potential claimants' awareness of the resource's value and the disposition of courts and the state in ownership disputes.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
The theoretical innovation is in defining innovators' expected returns as a sum of (1) conventional monopoly profits from the differentiated input variety created by the innovative use of the resource and (2) appropriation rents from controlling this resource.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
During such episodes, legal and political conflicts unfold as people realize they have rightful claims on the appropriated, and newly valuable, resources. Property rights and, as a consequence, profit trajectories are determined by how such conflicts are settled.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
Historically, innovators pioneering uses of these resources simply took them and realized extraordinary "appropriation rents". Today's tech giants are a recent episode of profits coming from innovation tied to appropriation rents.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
Recent examples are "digital exhaust" as an input to targeted advertising powering Google, Facebook, etc. or online intellectual property to train LLMs by the AI giants. Historical examples include electromagnetic spectrum for radio and TV broadcast, oil reserves for petroleum products.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
They present a model of economic growth based on producers deciding whether to appropriate (ie. just take without asking) or try to contract for access to resources with newly discovered uses. A defining feature of such resources is that they do not have property rights clearly assigned.
100
Cyrus Samii @cdsamii.bsky.social · 11/08/2026
What a *great* paper by Beraja and Yuchmtan from the NBER SI -- "Growth at the Resource Frontier". conference.nber.org/conf_papers/...
conference.nber.org
110
Reposted by Cyrus Samii
Carl Quintanilla @carlquintanilla.bsky.social · 10/08/2026
“.. nothing matters more to markets at present than the AI buildout. It’s such a sudden and massive stimulus .., Columbia Business School’s Stijn Van Nieuwerburgh argues that without it, the US would be in recession.” @bloomberg.com www.bloomberg.com/opinion/news...
45337123
Reposted by Cyrus Samii
ᴅᴀɴɪᴇʟ ᴍɪʟʟɪᴍᴇᴛ ✡️☮️❤️ @dlmillimet.bsky.social · 03/08/2026
The current issues of @aeajournals.bsky.social J Eco Perspectives has a nice primer on weak instruments and IV more generally. One of the "Don'ts" for applied researchers is copied below: don't search over a pool of instruments for a "strong" one and then go with that IV estimate.
1227
Reposted by Cyrus Samii
Melody Huang @melodyyhuang.bsky.social · 03/08/2026
New working paper with @corymccartan.com on the Gaussian Process limit of BART! (Bayesian Additive Regression Trees... not the Bay Area Rapid Transit sadly) arxiv.org/abs/2607.28844
arxiv.org
Seeing the Forest for the Trees: The Gaussian Process Limit of BART
Bayesian Additive Regression Trees (BART) have shown state-of-the-art performance in both prediction and causal inference problems. Previous theoretical work has attempted to explain BART's superior p...
1173
Cyrus Samii @cdsamii.bsky.social · 02/08/2026
Reading the essay, my take is that the issue is poor methodological and analytical standards. Economics and increasingly political science have pursued social goals with rigor. Psych has gotten much better over the past decade.
050
Reposted by Cyrus Samii
Adam L @adam-lg.bsky.social · 29/07/2026
"Over time, I have reluctantly concluded that my optimism was naïve. I now see evidence-based medicine as a deeply ill enterprise." Manski throwing bombs 💣💣💣 arxiv.org/abs/2607.23660 #statssky 📈📉
 I critique a set of entrenched methodological conventions that collectively create systemic dysfunction in statistical research for clinical decisions. These include: (1) the prevalent use of hypothesis tests to compare treatments, (2) remoteness from patient care of the methods used to evaluate the accuracy of predictions of patient outcomes, (3) poor practice of meta-analysis to combine findings across studies, and (4) widespread research with incredible certitude. It appears that the dysfunction is held in place by three factors: (i) rudimentary instruction in statistical methodology received by medical students and residents, (ii) reliance of clinical researchers on consulting biostatisticians, wo act as statistical gatekeepers in evaluation of grant proposals and paper submissions, and (iii) institutional practices of research funding agencies, medical journals, and governmental bodies that regulate medical treatment. I conjecture that systemic changes are necessary to break the existing impasse, moving statistical research to a better equilibrium.
2267
Cyrus Samii @cdsamii.bsky.social · 23/07/2026
The counterexample appears to be a fairly simple numerical example. I also see that the chatbot arrived there by iterating to discover necessary conditions. Quite something.
020
Reposted by Cyrus Samii
Martin Huber @causalhuber.bsky.social · 22/07/2026
📖 Nearly 3 years after the publication of Causal Analysis (MIT Press), I'd like to highlight the lecture slides accompanying the book. Available here as PDF and editable TeX files, together with datasets & code in R, Python, & Stata: dataverse.harvard.edu/dataset.xhtm... @mitpress.bsky.social
0317
Reposted by Cyrus Samii
Dean Yang @deanyang.bsky.social · 20/07/2026
If you're a development economist, join @ideadevecon.bsky.social! In particular, students and early-career scholars. Share this with anyone you know who should be part of the community!👇
011
Cyrus Samii @cdsamii.bsky.social · 12/07/2026
Sinner’s win today was so impressive. Hard to imagine how anyone could have done more than Zverev: check out the stats. Rallying consistently, with 135mph serves on the lines, 115mph second serves when needed, and 85mph forehands. Not enough.
050