Juan Dodyk

@itaisher It's funny and tragic how much he's still obsessed with writing and publishing blog posts

This blog post dismisses the thing that we actually have right now in favor of five minutes of speculation about what we might have in the future.

@haddadliv @Harvard Congratulations!! Awesome

I’m thrilled to share that I’ll be joining @Harvard this fall as a PhD student in Government! I’m deeply grateful to the many people who supported me along the way. Here’s a thread of thanks:

This is unironically good music youtube.com/watch?v=e-jUqW…

@ahall_research @ben_golub @rootsofprogress It’s 10x-ing garbage research for sure

AI is already 10x-ing academic research in the social sciences. In a guest post for @rootsofprogress, I explore how we can get to 100x. Some of my ideas: build more prototypes, define open problems with objective benchmarks to compete on, and keep pressing on dynamic, replicable, agentic research. Check out the post here: newsletter.rootsofprogress.org/p/ai-is-alread…

@brhkim @arthur_spirling sure

@jdodyk @arthur_spirling I think I agree, but it's been a heavy constraint people have to optimize around and changes a lot of what we do research. So not a bottleneck to producing work, but reducing this constraint is a very very good thing (and any reallocation of time towards Q's is also beneficial)

Just beginning to sense that we may be confusing coding for doing research. Ofc, coding may be necessary, but its not sufficient, and it's definitely not the same thing as research. And I say that as a very pro AI-in-the-workflow person.

@HunterWieman @RefineInk 99% of political scientists wouldn't notice either. not because it's subtle

I have found a fatal flaw in a paper. It is an RDD paper, and the authors fundamentally misunderstand and misapply RDD. It's a very basic error, albeit a bit subtle. Every AI model I have tried (including @RefineInk) cannot identify it, even when I tell them exactly where it is.

@cur7i7e te falto schoenberg en viena, el goat

@cachitorla nooooooooo

every tech guy you know working on their @openclaw "productivity" system right now

The AEA has posted eight "Recent Developments" lectures exploring highly topical issues in economics, presented by the best scholars in the field: aeaweb.org/conference/web… Well worth a watch!

This semester, we'll organize a PhD reading group on Industrial policy at @columbia_econ, with CBS, SIPA, and CPE, with @luchi_casal and Filip Milos (both postdocs at CBS) We'll cover recent advances in IP at the intersection of macro, growth, firm dynamics, trade-spatial, public finance, IO, etc. Find the full syllabus here: thomasbourany.github.io/files/Reading_…

New Working Paper! Why do policies that reinforce unequal opportunities persist? Some blame elites. I argue less secure families defend these "boundaries" against reform, and call this "boundary defense". I test this on a 2010 referendum against school reform in Hamburg, Germany.

We’ve written a review on *New Industrial Policy* for the Oxford Research Encyclopedia. We offer a unified framework for analyzing IP effects and survey evidence from both ex post event studies and ex ante model-based evaluations of IP. 🧵thread on key takeaways (link below)

@erazoca @cachitorla Jaja todos deben morir pero no matar

@erazoca @cachitorla Jaja no esperaba que fuera tan polémico el tuit. Obviamente no estoy a favor del suicidio!

[Warning: this may offend some people!] I think economists have done society a disservice by elevating applied-micro studies that focus exclusively on the relative effects of trade across different groups of workers but are silent about the aggregate effects. Differences-in-differences is a useful hammer, but not everything's a nail. The paper Arpit cites says that regions that were exposed to NAFTA had worse labor-market outcomes than less-exposed regions. This doesn't mean that NAFTA was bad for US workers! A story that is entirely consistent with their results is that NAFTA increased US manufacturing employment overall, just less so in more-exposed areas. Autor et al's "China Shock" narrative is the best example. It has colored the debate on trade to such an extent that many economists (not to mention the vast majority of Americans) are convinced that trade with China was a net negative for US workers. But all it says is that regions that were more exposed to Chinese import competition had worse labor-market outcomes than less-exposed regions. It doesn't say anything about the aggregate effects! An important paper by Wang et al. (nber.org/papers/w24886) shows that after accounting for the supply-chain effects of Chinese imports, overall employment and wages actually increased. What's even more striking is that they find positive labor-market outcomes overall even in regions that experienced large manufacturing employment declines via the Autor et al. channel. I think we also do society a disservice by putting so much focus on labor-market effects. Especially in the case of China, the biggest effects come from lower prices (i.e. higher real wages). Here, there are no tensions between aggregate and distributional effects. Everyone gained, but lower-income households gained the most! For example, this paper by an old friend of mine (ericksager.com/uploads/3/8/0/…) finds a "large fall in domestic prices" and that "[p]roduct categories catering to low-income consumers experienced larger price declines." And this paper by Mike Waugh (waugheconomics.com/uploads/2/2/5/…) which develops a cool model where trade interacts with heterogeneity in the price elasticity of demand across the income distribution. He finds that "gains from trade are pro-poor and that the average gains from trade [is] substantially larger than representative agent benchmarks." Don't get me wrong: trade isn't a pareto improvement, and the tradeoff between widely-dispersed aggregate gains and concentrated losses is real. But somehow I think society is interpreting many of our findings as saying that trade is a net negative, not a net positive, and that's not good.

🚨Excited for incredible lineup at @nberpubs conference on **Energy Markets, Decarbonization, and Trade** 🗓️This Thurs/Fri 3/20-21: Organized w Natalia Ramondo, supported by @SloanFoundation Full agenda and streaming link in next post below👇

@colin_fraser @slimer48484 Oh, sorry, haha I'm worse than an LLM. But it's just a consequence of Green-Tao, as someone noted: take f(x) = ax + b such that f(1), ..., f(k^2) are all prime. In particular f(1^2), ..., f(k^2) are all prime, so P(x) = ax^2 + b is a quadratic such that P(1), ..., P(k) are prime

@jdodyk @slimer48484 Is it? I’m trying to work out exactly how to put my question in terms of this theorem. What would be the k polynomials?

Answer smells like bullshit but I actually don’t know! Very cool that they invented a computer program that undetectably lies to you about facts.