Sign in

Jonathan Libgober

@jlibdoesecon.bsky.social
1.1K followers 842 following 97 posts

AP in Econ at USC. I do Information Economics & more, both pure and applied. Best friend of Spinoza (Nosik) Petrov-Libgober, husband of @lenapetrova.bsky.social, and brother of @blibgober.bsky.social.

PostsRepliesMedia
Jonathan Libgober @jlibdoesecon.bsky.social · 08/08/2026
With the new semester approaching, I wrote a short note on how I'm thinking about AI in education: jonlib.com/assets/paper... It's 60% to explain my views to students, 30% to help myself think through this, and 10% to contribute to the broader conversation. Reactions welcome!
jonlib.com
141
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
You can see the thread for the paper here: bsky.app/profile/jlib...
000
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
…and see how “jaggedness” influences team size. Not much to say other than I tried to make it as fun as I could. Comment welcome on this part too—especially if you’re not used to reading theory papers.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
This is *definitely* new territory for me, but once I had the idea for how this might work, I couldn’t help myself from trying to do it. The interactive shows how Gaussian processes are being built from independent components, lets you simulate the attention problem…
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
My doing so was in large part inspired by seeing this example: kevinbryanecon.com/AIandhumanch...
kevinbryanecon.com
Training AI for When Humans Will Use It
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
I just posted a new paper which I've been working on for a while now. I made an interactive that takes you through the main ideas in a more…interactive…way. jonlib.com/papers/organ...
jonlib.com
Organization Design for Complex Worlds — an interactive walkthrough
An interactive walkthrough of the core technical ideas in "Organization Design for Complex Worlds": the Karhunen–Loève decomposition, the water-filling attention problem, and optimal team size.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
There are several other extensions in the paper, plus a companion interactive that I'll post about soon. Paper: jonlib.com/assets/paper... Comments very welcome!!!! I’m about to present it for the first time and am very excited to have it out there.
jonlib.com
000
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
As a result, the complexity that remains becomes relatively more concentrated outside management’s attention—so teams grow to handle that. Proving that formally (even after having the intuition from the main model) is surprisingly involved, for reasons I'll leave to the paper.
110
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
The intuition is actually very simple. Management pays attention to the most significant components of complexity. Those are also the components it would most like to simplify.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
Probably my favorite part of the paper is an extension where the firm can adjust complexity endogenously. This has a natural interpretation in terms of AI.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
Crucially, proportionally scaling all variation up or down does *not* change team size. What matters is where the variation lies.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
...i.e., how sharply can the conditions facing nearby workers differ? The main comparative statics are more general, but a simple takeaway is that, within a natural family of environments, more jaggedness—holding total variance fixed—means larger teams.
110
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
A technical contribution is showing how to use this decomposition in the context of an org econ model. One reason the continuum formulation is nice is because calculus is nice. But another is that it lets me formalize a notion of local complexity in terms of jaggedness...
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
Solving the attention problem involves some beautiful mathematics from spectral theory. Essentially, a Gaussian process can be decomposed into a sum of independent modes—akin to principal-components analysis.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
The organization chooses team size. Larger teams make it harder for individual workers to adapt to their local conditions. But they also leave headquarters with fewer team-level points of contact to understand and coordinate.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
Headquarters, meanwhile, faces an attention problem. It can acquire any signal about the realized state process, subject to a constraint on mutual information.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
Some short highlights: The model formulates an organization as a continuum of workers. Each worker is trying to match a local state, with states correlated across workers according to a Gaussian process.
100
Jonathan Libgober @jlibdoesecon.bsky.social · 17/07/2026
New paper!!! How does complexity relate to organization design? And what does the answer say about how organizational hierarchy might change with new technologies like AI? Thread below. If you don't have time for that, the picture basically summarizes the main idea.
230
Jonathan Libgober @jlibdoesecon.bsky.social · 06/05/2026
Really excited to see this in print!! It's even more meaningful as my first publication with a student (at least, one who remained a student from start to finish). Also appreciative of everyone who helped make the paper better--can't even describe how happy I am with how the paper turned out.
020
Jonathan Libgober @jlibdoesecon.bsky.social · 17/01/2026
@ EconSky: If any of you have taught (or are teaching/planning to teach) any of my papers, could you please DM me? I would be very grateful. Any honest DM will receive a personalized picture of my Spinoza (Nosik) Petrov-Libgober (my dog) as a token of gratitude.
010
Jonathan Libgober @jlibdoesecon.bsky.social · 31/10/2025
Ungated version here: uploads.strikinglycdn.com/files/fbd829...
uploads.strikinglycdn.com
010
Jonathan Libgober @jlibdoesecon.bsky.social · 31/10/2025
My paper on research registries with Eliot Abrams and John List is now forthcoming at Economic Journal: academic.oup.com/ej/advance-a... Very happy to finally have finished up this project, although I learned a ton from working on it.
academic.oup.com
Research Registries and the Credibility Crisis: An Empirical and Theoretical Investigation
Abstract. We analyze one prominent policy solution to the credibility crisis in experimental research—research registries—with primary focus on the AEA RCT
160
Jonathan Libgober @jlibdoesecon.bsky.social · 28/10/2025
This was a fun one! Glad to see it in a great home.
020
Jonathan Libgober @jlibdoesecon.bsky.social · 28/07/2025
I've discovered that you have much more fun as a tourist if you don't know what the exchange rate is.
110
Jonathan Libgober @jlibdoesecon.bsky.social · 26/07/2025
I feel like I've heard critiques of economics as a discipline since as long as I've studied economics. Has anyone done a historiography of this? I guess it's a sign of the field's influence that people seem to care so much, so maybe it's a good thing. At the same time, I can't help but think...why?
8221
Jonathan Libgober @jlibdoesecon.bsky.social · 15/07/2025
Huge congrats to my amazing colleague and her coauthors on this fantastic paper.
020
Reposted by Jonathan Libgober
Rajiv Sethi @rajivsethi.bsky.social · 28/06/2025
New post on prediction markets, opinion polls, and the New York City mayor's race: open.substack.com/pub/rajivset...
open.substack.com
Markets, Models, and Mamdani
Political prediction markets have been around for decades, but rose to unprecedented visibility and prominence during the last election cycle in the United States.
022
Reposted by Jonathan Libgober
Rajiv Sethi @rajivsethi.bsky.social · 18/06/2025
New post on the realization after two decades of Poindexter's dream: rajivsethi.substack.com/p/betting-on...
rajivsethi.substack.com
Betting on Geopolitical Violence
Back in 2003, a Pentagon office led by Admiral John M.
064
Jonathan Libgober @jlibdoesecon.bsky.social · 17/05/2025
Meh, I think it'd only work about half as well as it should... (and thanks!)
000
Jonathan Libgober @jlibdoesecon.bsky.social · 17/05/2025
Seems that I jumped ahead of this a little bit...
270
Jonathan Libgober @jlibdoesecon.bsky.social · 17/05/2025
Less related to recent events than it may seem, but I wanted to post this at some point: My Retractions paper, with brilliant coauthors Duarte Goncalves and Jack Willis, was accepted at Restud earlier this year, and is now on the journal's website. www.restud.com/retractions-...
restud.com
Retractions: Updating from Complex Information - The Review of Economic Studies
We modify a canonical experimental design to identify the effectiveness of retractions. Comparing beliefs after retractions to beliefs (a) without the retracted information and (b) after equivalent ne...
073
Jonathan Libgober @jlibdoesecon.bsky.social · 09/05/2025
An important paper on the role of subjective beliefs on the business cycle, now published in Restud. academic.oup.com/restud/advan...
academic.oup.com
Survey Data and Subjective Beliefs in Business Cycle Models
Abstract. This paper develops a theory of subjective beliefs that departs from rational expectations, and shows that biases in household beliefs have quant
010
Jonathan Libgober @jlibdoesecon.bsky.social · 29/04/2025
My paper on algorithmic pricing, joint with In-Koo Cho, is now online at the Journal of Economic Theory. www.sciencedirect.com/science/arti...
sciencedirect.com
Learning underspecified models
This paper examines learning dynamics under non-parametric model uncertainty. We choose the monopolistic profit maximization problem (Myerson (1981)) …
0213
Jonathan Libgober @jlibdoesecon.bsky.social · 22/04/2025
My dog will literally climb on top of me while I'm working on my laptop but then start to growl in discomfort if I hug him back. Make it make sense.
110
Reposted by Jonathan Libgober
Guy Grossman @guygrossman.bsky.social · 14/04/2025
Attempted murder of a Jewish governor and his family on Passover is not a headline story for @nytimes.com and @washingtonpost.com. I’m relieved that Shapiro and his family survived the attack.
928769
Jonathan Libgober @jlibdoesecon.bsky.social · 08/04/2025
As an economist, here is my take on how to understand the justification for Trump's tariffs. en.wikipedia.org/wiki/Kettle_...
en.wikipedia.org
Kettle logic - Wikipedia
2323
Jonathan Libgober @jlibdoesecon.bsky.social · 07/04/2025
Will be waiting for them to add Russian...
000
Jonathan Libgober @jlibdoesecon.bsky.social · 13/03/2025
I've used a variant of this example as a one-player game, which makes it even weirder imo...
220
Jonathan Libgober @jlibdoesecon.bsky.social · 13/03/2025
I've taught game theory for several years now (maybe like, 10 times?) but somehow it only now just hit me that mixed strategies in Bayesian games are pretty damn weird.
020
Jonathan Libgober @jlibdoesecon.bsky.social · 02/03/2025
The only part I don't like about this is that now we may be forced to spell it "defence."
010
Reposted by Jonathan Libgober
Adam Kinzinger @adamkinzinger.substack.com · 28/02/2025
13757962117476
Jonathan Libgober @jlibdoesecon.bsky.social · 19/02/2025
I was motivated to post this after back-to-back seeing the clip of Roger Wicker with Manu Raju and the tweet from Don Bacon.
010
Jonathan Libgober @jlibdoesecon.bsky.social · 19/02/2025
If I shared the ITYSL meme with Tim Robinson in the hot dog suit every time a Republican said that Putin is out of control, I wouldn't have time to do anything else.
1294
Jonathan Libgober @jlibdoesecon.bsky.social · 17/02/2025
Rubio has like, what, 5 months left on the job?
010
Jonathan Libgober @jlibdoesecon.bsky.social · 16/02/2025
Good to know I am fighting the woke agenda by showing that some inequalities are actually binding equalities.
020
Jonathan Libgober @jlibdoesecon.bsky.social · 15/02/2025
Roses are red, A noise ball is blue, But when you find out it's noise, The info will not undo. uploads.strikinglycdn.com/files/8779d7...
uploads.strikinglycdn.com
010
Jonathan Libgober @jlibdoesecon.bsky.social · 10/02/2025
Chiefs can still win this if Mike Pence has the courage to do the right thing.
030
Jonathan Libgober @jlibdoesecon.bsky.social · 08/02/2025
If only I knew how to stop it...
000
Jonathan Libgober @jlibdoesecon.bsky.social · 08/02/2025
Unfortunately I rely on social media for all of my tips on how to best support my mental health...
110