# ![@bechhof Avatar](https://lunarcrush.com/gi/w:26/cr:twitter::2578553016.png) @bechhof Nathaniel Bechhofer

Nathaniel Bechhofer posts on X about in the, if you, countries, this is the most. They currently have [------] followers and [---] posts still getting attention that total [---] engagements in the last [--] hours.

### Engagements: [---] [#](/creator/twitter::2578553016/interactions)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::2578553016/c:line/m:interactions.svg)

- [--] Week [-----] +1,322%
- [--] Month [------] +311%
- [--] Months [------] -68%
- [--] Year [-------] -79%

### Mentions: [--] [#](/creator/twitter::2578553016/posts_active)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::2578553016/c:line/m:posts_active.svg)


### Followers: [------] [#](/creator/twitter::2578553016/followers)
![Followers Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::2578553016/c:line/m:followers.svg)

- [--] Week [-----] +0.11%
- [--] Month [-----] +0.80%
- [--] Months [-----] +4.80%
- [--] Year [-----] +11%

### CreatorRank: [---------] [#](/creator/twitter::2578553016/influencer_rank)
![CreatorRank Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::2578553016/c:line/m:influencer_rank.svg)

### Social Influence

**Social category influence**
[finance](/list/finance)  [countries](/list/countries)  [technology brands](/list/technology-brands)  [stocks](/list/stocks)  [social networks](/list/social-networks)  [travel destinations](/list/travel-destinations) 

**Social topic influence**
[in the](/topic/in-the), [if you](/topic/if-you), [countries](/topic/countries), [this is](/topic/this-is), [at least](/topic/at-least), [to the](/topic/to-the), [silver](/topic/silver), [model](/topic/model), [inference](/topic/inference), [investment](/topic/investment)

**Top assets mentioned**
[Cloudflare, Inc. (NET)](/topic/cloudflare) [Validity (VAL)](/topic/validity) [Waves (WAVES)](/topic/waves)
### Top Social Posts
Top posts by engagements in the last [--] hours

"@RexDouglass @NateSilver538 Oh come on that's such an unreasonable steelman of Silver's point. By that standard linear or logistic regression would qualify as blackboxy too and Silver would absolutely never ever in a million years go for that given he actually knows how to use those in Stata"  
[X Link](https://x.com/bechhof/status/2018837077048340623)  2026-02-04T00:00Z [----] followers, [--] engagements


"@RexDouglass @NateSilver538 Why would you think Silver is any better"  
[X Link](https://x.com/bechhof/status/2018838337776218180)  2026-02-04T00:05Z [----] followers, [--] engagements


"@RexDouglass @NateSilver538 Most trashing is also trash and Silver isn't just doing trashing but also promoting a confident alternative view of the world. You're getting a whole lot of auxiliary garbage with the criticisms which don't have much to offer anyway"  
[X Link](https://x.com/bechhof/status/2018838931790352465)  2026-02-04T00:08Z [----] followers, [--] engagements


"@RexDouglass @NateSilver538 Literally Jake Grumbach is great"  
[X Link](https://x.com/bechhof/status/2018843577745608911)  2026-02-04T00:26Z [----] followers, [--] engagements


"@teej_m At this point the setup to do is probably to have AI agents pull all web pages via the cloudflare website to markdown endpoint"  
[X Link](https://x.com/bechhof/status/2018888696179859582)  2026-02-04T03:25Z [----] followers, [---] engagements


"@joe_r_Odonnell @laharvell @SPSPnews @Spencer_Fdn @JMUresearch @cremieuxrecueil Eh I would trust Jordan Lasker's credibility on anything"  
[X Link](https://x.com/bechhof/status/1893168845080129845)  2025-02-22T05:20Z [----] followers, [--] engagements


"@jxnlco The lengths people will go to just to not use DSPy are remarkable"  
[X Link](https://x.com/bechhof/status/1981455433014685916)  2025-10-23T20:19Z [----] followers, [---] engagements


"@JakeMGrumbach Not when done competently as in Denmark"  
[X Link](https://x.com/bechhof/status/1989031254801428783)  2025-11-13T18:03Z [----] followers, [---] engagements


"@jmhorp @lymanstoneky @salonium Its here Saloni also did a great podcast with @JerusalemDemsas about it: https://podcasts.apple.com/us/podcast/what-do-we-really-know-about-the-maternal-mortality-crisis/id1746176654i=1000664468611 https://ourworldindata.org/rise-us-maternal-mortality-rates-measurement https://podcasts.apple.com/us/podcast/what-do-we-really-know-about-the-maternal-mortality-crisis/id1746176654i=1000664468611 https://ourworldindata.org/rise-us-maternal-mortality-rates-measurement"  
[X Link](https://x.com/bechhof/status/1991158101240672384)  2025-11-19T14:54Z [----] followers, [--] engagements


"@MishaTeplitskiy rather than straight up definitively endorsing the conclusion that there's a large selection bias problem as a proportion of the overall apparent change"  
[X Link](https://x.com/bechhof/status/1995878106733969744)  2025-12-02T15:29Z [----] followers, [--] engagements


"Two claims that people following this should decouple: 1) Something remarkable has been happening with Mississippi test scores presumably down to some of the _many_ reforms they did. 2) We can confidently isolate the effect of a specific reform which really had the oomph. Here's the key part of the paper suggesting the Mississippi Miracle effect is mostly selection bias rather than learning improvement (It's a really tough question to study tho and attrition issues are especially rough so I'm not feeling super critical of the original research) https://t.co/GXvZjSmAYd Here's the key part of"  
[X Link](https://x.com/bechhof/status/1995882555992932554)  2025-12-02T15:47Z [----] followers, [---] engagements


"@arpitrage @alexolegimas Micro-founded models do throw more parameters at the problem. They do so in an opinionated beliefs-about-mechanism laden way but it seems correct for economists to insist on that as default over the higher flexibility and opacity of (undisciplined) rich statistical models"  
[X Link](https://x.com/bechhof/status/1998220082770477353)  2025-12-09T02:36Z [----] followers, [--] engagements


"@Dyllan88 @besttrousers It depends how fine you cut; I'm sure it's true that @MattBruenig's figures for the specific stuff would look less compelling if it visualized some uncertainty. I cut up the data differently with ACS and despite the small sample size there appears to be something to the thesis"  
[X Link](https://x.com/bechhof/status/2001438271776109010)  2025-12-17T23:44Z [----] followers, [--] engagements


"@KelseyTuoc Seem unsolvable given that apparently according to people on the Internet it's impossible to actually teach mathematics or target teaching at kids who are struggling and we just have to accept ability hierarchies as they are so we give up on helping students who underperform"  
[X Link](https://x.com/bechhof/status/2001442243752652990)  2025-12-17T23:59Z [----] followers, [---] engagements


"@michael_wiebe I don't see why this would be true. What developers can do is create effective coordination which results in changing demographics and anticipation of those changed demographics allows a shift to the new gentrified equilibrium which raises neighborhood prices"  
[X Link](https://x.com/bechhof/status/2002100448128569443)  2025-12-19T19:35Z [----] followers, [--] engagements


"@arindube @alexolegimas @arpitrage BTW here's a Claude written response to the post I liked: https://claude.ai/public/artifacts/d2de7a2c-d738-4a88-856c-65eef298590c https://claude.ai/public/artifacts/d2de7a2c-d738-4a88-856c-65eef298590c"  
[X Link](https://x.com/bechhof/status/2003217017697108402)  2025-12-22T21:32Z [----] followers, [---] engagements


"@JohnHolbein1 The study doesn't focus its confounding adjustments on the critical issue: other contaminants in the water sources. Moreover they make estimates for each [--] kilometer wide radius which likely means a lot of noise in the constructed environmental variables. Those undermine even"  
[X Link](https://x.com/bechhof/status/2003884177016398209)  2025-12-24T17:43Z [----] followers, [----] engagements


"@JohnHolbein1 interpreting the results as stated so the methodological weakness is decisive in addition to the sketchiness of just how the confidence intervals have the suspicious tell of how conveniently located they are relative to zero instead of"  
[X Link](https://x.com/bechhof/status/2003884242154279059)  2025-12-24T17:43Z [----] followers, [---] engagements


"@alexolegimas @ben_golub @PAHoyeck I did well on his exams and preferred taking the micro qualifying exam because I (correctly) anticipated I would pass first try but his actual presentation was abysmal unless you already had some mastery of the content in which case it was tolerable / you could sympathize"  
[X Link](https://x.com/bechhof/status/2003929318066638926)  2025-12-24T20:42Z [----] followers, [---] engagements


"@MattBruenig @GreySmithereens @PEWilliams_ @MattZeitlin It's both not AGI adjacent and something easy enough to do better that by making your own division of labor with DSPy 3"  
[X Link](https://x.com/bechhof/status/2005443498036867105)  2025-12-29T00:59Z [----] followers, [---] engagements


"@JakeMGrumbach I'm somewhat sympathetic to taxing unrealized gains etc but to the extent that resources are being drawn dude by wealthy people you can get at that pretty effectively by taxing consumption rather than going full on with the massive distortions from a wealth tax"  
[X Link](https://x.com/bechhof/status/2005444420691534326)  2025-12-29T01:03Z [----] followers, [---] engagements


"@JakeMGrumbach Going for a wealth tax over ambitious tax reform is one of the classic "properly better things aren't possible so let's break a bunch as we try to get a win that feels good" failures which I'd much prefer a healthy left move past"  
[X Link](https://x.com/bechhof/status/2005445736562057366)  2025-12-29T01:08Z [----] followers, [---] engagements


"@JakeMGrumbach BTW we have a few governments which have the equivalent of a wealth tax in asset tests for benefits etc for individuals / households who are poor or low income. That's bad We should reform those implementations"  
[X Link](https://x.com/bechhof/status/2005447097148158434)  2025-12-29T01:13Z [----] followers, [---] engagements


"@fitzr1122 I'm told that Irish people are all super angry at being replaced by immigrants and want to leave the EU but elites refuse to respect this preference from the vast majority"  
[X Link](https://x.com/bechhof/status/2005759559978344780)  2025-12-29T21:55Z [----] followers, [--] engagements


"@max_sixty Ignorant question having just browsed the docs: is there an API exposed that allows using this without sending explicit shell commands I've really enjoyed being able to do this in eg @simonw's llm package"  
[X Link](https://x.com/bechhof/status/2006164170450817056)  2025-12-31T00:43Z [----] followers, [---] engagements


"@natolambert But that's in the model construction step. The open opportunity is for other developers to build on top of existing models (my preferred way is @DSPyOSS ) to get the tone management without having to wait for the model providers or even do any model fine-tuning"  
[X Link](https://x.com/bechhof/status/2006486093194285410)  2025-12-31T22:02Z [----] followers, [--] engagements


"@grok Would Jon Snow have won election as Lord Commander if the Night's Watch had used a Borda Count ranked choice system Consider the full equilibrium including candidate entry etc"  
[X Link](https://x.com/bechhof/status/2007561039139725514)  2026-01-03T21:13Z [----] followers, [--] engagements


"@souljagoyteller These are not people who invest a lot of bandwidth into getting stuff done. The internal pushback will create a lot of friction and so these people will just never do sufficient follow through to make it happen"  
[X Link](https://x.com/bechhof/status/2007590030207004689)  2026-01-03T23:08Z [----] followers, [---] engagements


"@BasilHalperin @sebkrier It is now but wouldn't be in the future at least if you did at least a half decent job of alignment because they're going to optimize for more valuable output regardless of the taxation and they're on auto pilot"  
[X Link](https://x.com/bechhof/status/2007596752355127509)  2026-01-03T23:35Z [----] followers, [--] engagements


"@yudapearl I don't even need to go for the Pro version of GPT [---] for it to fully clarify how Bayesian reasoning works perfectly here. https://chatgpt.com/s/t_6963b31dc4408191af858986c3fee4b9 https://chatgpt.com/s/t_6963b31dc4408191af858986c3fee4b9"  
[X Link](https://x.com/bechhof/status/2010357722030490002)  2026-01-11T14:26Z [----] followers, [---] engagements


"@grok how do these three equations undermine Pearl's claim that Bayesian reasoning can't handle causal reasoning well P(Y mid do(X) mathcalD) = int_Theta underbraceP(Y mid do(X) theta)_textCausal Logic underbracep(theta mid mathcalD)_textBayesian Weight dtheta P(Y mid X theta) neq P(Y mid do(X) theta) p(mathcalD mid theta) = prod_i=1N P(v_i mid pa(v_i) theta_mechanism) https://twitter.com/i/web/status/2010364315748925563 https://twitter.com/i/web/status/2010364315748925563"  
[X Link](https://x.com/bechhof/status/2010364315748925563)  2026-01-11T14:52Z [----] followers, [---] engagements


"@grok isn't the do-operator/graph just the domain-specific forward model like Newton's laws We don't say physics predictions go "beyond probabilistic updating". Just like with anything else Bayes marginalizes over uncertainty given the mechanistic model. Why hold causality to a different standard The equations show full absorption into Bayesianism once structure is specified which handles Pearl's emphasis on the 'differing logics' for interventional queries undermining his claim about a flaw in Bayesian reasoning. https://twitter.com/i/web/status/2010367119066829066"  
[X Link](https://x.com/bechhof/status/2010367119066829066)  2026-01-11T15:04Z [----] followers, [--] engagements


"@grok if causal graphs/do-operator are just the domain model (like Newton's laws) and Bayes incorporates them fully for interventional predictives then doesn't this undermine Pearl's caricature of the "standalone Bayesianism" that allegedly falls short Bayesian inference and uncertainty quantification uses structured models all the time for any domain. The equations show causality is fully absorbed with no shortfall; Pearl's 'differing logics' point is handled by querying the model correctly (interventionally vs. observationally)"  
[X Link](https://x.com/bechhof/status/2010368086830891011)  2026-01-11T15:07Z [----] followers, [--] engagements


"@analisereal @grok @yudapearl They're not derived from or constrained wirhin the probabilistic updating. That doesn't mean they do something special which undermines Bayesian process. They complement that process rather than showing Bayesianism fails"  
[X Link](https://x.com/bechhof/status/2010533181980901428)  2026-01-12T02:03Z [----] followers, [--] engagements


"@yudapearl @DongNguyeb That's just prior information with probability [--]. Bayes is so good that we can even generalize to where we have uncertainty about the type of data we have"  
[X Link](https://x.com/bechhof/status/2010726967331680589)  2026-01-12T14:54Z [----] followers, [--] engagements


"@RAVerBruggen @AlexNowrasteh This I do disagree with. I think carving out high rate native groups does bias the comparison in a specific direction"  
[X Link](https://x.com/bechhof/status/2010742531638894597)  2026-01-12T15:55Z [----] followers, [--] engagements


"Contra @JonSteinsson and @_alice_evans if you look at age specific birth rates there's no indication that completed birth rates by cohort are set to have a big drop any time in the next few years. We have much finer grained data than TFR It is true that completed cohort fertility (CCF) has not fallen in the US while TFR has. But CCF is a very lagging indicator. Can only be calculated for cohorts that are [--] and older. Here is TFR and CCF at birth year + [--]. Looks like CCF might just be about to start falling. https://t.co/Ed61DjNEnt It is true that completed cohort fertility (CCF) has not"  
[X Link](https://x.com/bechhof/status/2011171716048044415)  2026-01-13T20:21Z [----] followers, [---] engagements


"As always best to read Saloni on this: https://ourworldindata.org/total-fertility-rate-births-per-woman https://ourworldindata.org/total-fertility-rate-births-per-woman"  
[X Link](https://x.com/bechhof/status/2011172201035174272)  2026-01-13T20:23Z [----] followers, [--] engagements


"@RAVerBruggen I had both Gemini and ChatGPT do the proper tabulations using single year of age data from UN WPP and the trend by cohorts from [--] years ago and sooner is quite persistent (the small declines for the younger ages are easy to make up for at current age-specific rates)"  
[X Link](https://x.com/bechhof/status/2011518018149622264)  2026-01-14T19:17Z [----] followers, [--] engagements


"@RAVerBruggen I think it'll be fairly obvious even once we get the finalized [----] numbers (presumably in 2032)"  
[X Link](https://x.com/bechhof/status/2011520865645838650)  2026-01-14T19:28Z [----] followers, [--] engagements


"The extraordinarily high rate of complete breakage even on Claude's mobile app has me update even more toward the idea that Anthropic was created by God to hammer us over the head with the lesson that JavaScript was a mistake"  
[X Link](https://x.com/bechhof/status/2013053258101588004)  2026-01-19T00:57Z [----] followers, [---] engagements


"You'd think from this post that birth rate decline is a dominant phenomenon for every country and every demographic when in fact birth rates for women age 35+ have been going up in many countries for over a decade including the USA. The catastrophizers are overfitting. I have been posting repeatedly on X about the extraordinarily fast collapse of births across the planet: in rich and poor countries in fast-growing and slow-growing economies in religious and secular societies under right-wing and left-wing governments with high taxes and https://t.co/bTWWwFyYBX I have been posting repeatedly"  
[X Link](https://x.com/bechhof/status/2013285767678669239)  2026-01-19T16:21Z [----] followers, 10.9K engagements


"@GregoryBesharov @JesusFerna7026 this explains it most thoroughly; you can also either calculate cohort rates manually by pulling the UN WPP data or use the HFD numbers https://ourworldindata.org/total-fertility-rate-births-per-woman https://ourworldindata.org/total-fertility-rate-births-per-woman"  
[X Link](https://x.com/bechhof/status/2013318081284567107)  2026-01-19T18:30Z [----] followers, [--] engagements


"@Bennyhumma4242 You can have Gemini CLI Codex Claude Code or Amp look at the ccfrVH.txt file from HFD yourself if you prefer to not rely on my judgment. Whatever you pick will show you"  
[X Link](https://x.com/bechhof/status/2013342367751753763)  2026-01-19T20:06Z [----] followers, [--] engagements


"@dshuter wat"  
[X Link](https://x.com/bechhof/status/2013342489541832775)  2026-01-19T20:07Z [----] followers, [--] engagements


"@econ_inclined @JesusFerna7026 We've already seen huge jumps for age-specific brackets 35+ within-period since the early 1990s"  
[X Link](https://x.com/bechhof/status/2013345282734395761)  2026-01-19T20:18Z [----] followers, [--] engagements


"@econ_inclined @JesusFerna7026 The age specific birth rates at older ages are still vastly below 19th century numbers and vastly below numbers observed in other countries. We aren't close to a binding ceiling any time soon"  
[X Link](https://x.com/bechhof/status/2013357450892562595)  2026-01-19T21:06Z [----] followers, [--] engagements


"@JesusFerna7026 @carjema82 You need it to stay constant but regardless you're talking so far in the future it's the equivalent of worrying about peak oil in 2150"  
[X Link](https://x.com/bechhof/status/2013483443997270127)  2026-01-20T05:27Z [----] followers, [--] engagements


"@JustinSandefur @kearney_melissa @phil_wellesley The variable which is a bigger problem here is the use of TFR; even a pretty basic economic model of career investment would predict that low child penalties can *magnify* tempo effects"  
[X Link](https://x.com/bechhof/status/2013997115165307327)  2026-01-21T15:28Z [----] followers, [--] engagements


"@DominicJPino @scottlincicome Now do comparative life expectancy gains for literally all of those countries listed"  
[X Link](https://x.com/bechhof/status/2013998360361931134)  2026-01-21T15:33Z [----] followers, 14.9K engagements


"@Utahdesert @DominicJPino @scottlincicome"  
[X Link](https://x.com/bechhof/status/2014078384498716946)  2026-01-21T20:51Z [----] followers, [---] engagements


"@zk_whale @DominicJPino @scottlincicome Straight up unambiguously wrong right here on both fronts"  
[X Link](https://x.com/bechhof/status/2014164410063806854)  2026-01-22T02:33Z [----] followers, [---] engagements


"@grok @jasonprgressive @zk_whale @DominicJPino @scottlincicome @grok fully explicate therefore how the earlier @zk_whale post was wrong both re claims that US shortfalls in *changes* could be attributed to non White population levels or changes and wrong in claiming that White Americans have similar rates of mortality and crime as Europeans"  
[X Link](https://x.com/bechhof/status/2014167500041097383)  2026-01-22T02:45Z [----] followers, [---] engagements


"@grok @jasonprgressive @zk_whale @DominicJPino @scottlincicome @grok explain how the @jasonprgressive reply in this thread might be an instance of "motte-and-bailey" (sometimes called "town-and-castle") relative to the earlier @zk_whale post not least because the @zk_whale post implied a decomposition result which you've shown implausible"  
[X Link](https://x.com/bechhof/status/2014169352010940747)  2026-01-22T02:52Z [----] followers, [---] engagements


"@grok @jasonprgressive @zk_whale @DominicJPino @scottlincicome @grok parse the thread again to see that the discussion is about *changes* rather than levels and assess whether that makes the motte-and-bailey more or less egregious"  
[X Link](https://x.com/bechhof/status/2014169914660077590)  2026-01-22T02:55Z [----] followers, [--] engagements


"@repuzans @DominicJPino @scottlincicome That doesn't negate the country differences. You're giving examples of how it is that the USA has failed not variables that we should adjust away. The USA has also failed to offset those factors on other margins in spite of its relative advantages on the income growth front"  
[X Link](https://x.com/bechhof/status/2014171223530389639)  2026-01-22T03:00Z [----] followers, [--] engagements


"@rt4995060 @zk_whale @DominicJPino @scottlincicome Kind of incredible how this platform encourages so much disdain for people on the American right because of blatant disregard for facts as you've shown here"  
[X Link](https://x.com/bechhof/status/2014209286226710602)  2026-01-22T05:31Z [----] followers, [--] engagements


"@rt4995060 @zk_whale @DominicJPino @scottlincicome Literally false"  
[X Link](https://x.com/bechhof/status/2014217234290188789)  2026-01-22T06:03Z [----] followers, [--] engagements


"@rt4995060 @zk_whale @DominicJPino @scottlincicome Wrong on the medical care stuff and you're ignoring the critical metric of overall life expectancy which was what I brought up in the first place"  
[X Link](https://x.com/bechhof/status/2014217429195407487)  2026-01-22T06:03Z [----] followers, [--] engagements


"@rt4995060 @zk_whale @DominicJPino @scottlincicome Them not having free toilets is a good thing but you're so reflexive about this you don't understand the relevant economics"  
[X Link](https://x.com/bechhof/status/2014217578877530255)  2026-01-22T06:04Z [----] followers, [--] engagements


"@AJFCPassalacqua @zk_whale @DominicJPino @scottlincicome This would make at least some sense as a reply if it wasn't in the context of the absurd @zk_whale post"  
[X Link](https://x.com/bechhof/status/2014405135183564837)  2026-01-22T18:29Z [----] followers, [--] engagements


"@AJFCPassalacqua @zk_whale @DominicJPino @scottlincicome I do enjoy your pretense of caring about human misery when you're clearly pushing a narrative about gaps by comparing two non-comparable numbers (overall victimization vs victimization by *confirmed* same race perpetrators) so you can subtly portray non-White people as bad"  
[X Link](https://x.com/bechhof/status/2014413017518117116)  2026-01-22T19:01Z [----] followers, [--] engagements


"@zk_whale @AJFCPassalacqua @DominicJPino @scottlincicome @grok which participants in this thread have actually been presenting facts that are sensible in the context of the conversation"  
[X Link](https://x.com/bechhof/status/2014414441190666289)  2026-01-22T19:06Z [----] followers, [--] engagements


"@zk_whale @AJFCPassalacqua @DominicJPino @scottlincicome @grok Ah you enjoy overstating statistics to pollute perceptions. It's good to have that clarified out in the open so your trustworthiness can be undermined so thoroughly"  
[X Link](https://x.com/bechhof/status/2014418467831591033)  2026-01-22T19:22Z [----] followers, [--] engagements


"@patapackala @DominicJPino @scottlincicome Again this is just applying a heuristic without thinking about whether it makes a quantitatively meaningful difference. What compositional changes for the USA matter here to the point where there's some subtle way it's actually performed well on life expectancy gains Name it"  
[X Link](https://x.com/bechhof/status/2014418816051143115)  2026-01-22T19:24Z [----] followers, [--] engagements


"@zk_whale @AJFCPassalacqua @DominicJPino @scottlincicome @grok What Marxist revolution Is the Marxist revolution in the room with us right now"  
[X Link](https://x.com/bechhof/status/2014419827197161531)  2026-01-22T19:28Z [----] followers, [--] engagements


"@calebwatney Voxtral Transcribe [--] beats this now"  
[X Link](https://x.com/bechhof/status/2020370886751309896)  2026-02-08T05:35Z [----] followers, [---] engagements


"@besttrousers @cojobrien @grok which BLS or DOL dataset is likely being used for this mapping of degree fields to occupation codes"  
[X Link](https://x.com/bechhof/status/2021324579038245312)  2026-02-10T20:45Z [----] followers, [--] engagements


"What is it with social conservatives whether the @FamStudies people generally @BradWilcoxIFS specifically or @TPCarney and their commitment to using misleading denominators to push their narrative All a big distraction from increased birth rates for women 35+ Yes more children are being born to women over [--]. But at the same time way more women are childless in their 30s. As a fraction of all childless women first births are going down. That's true even among women in their early 30s. Late 30s first birth rates are lower than https://t.co/WM84h4uBFL Yes more children are being born to women"  
[X Link](https://x.com/bechhof/status/2022347144846594066)  2026-02-13T16:28Z [----] followers, [----] engagements


"@grok be more explicit: Name the statistic as first births per [----] currently childless women = a period hazard conditional on being childless at that age. Show the algebraic mechanism. textrate=fracB_1(30-34t)N_textchildless(30-34t) If N_textchildless rises faster than B_1 the rate falls automatically. Construct a one-line numerical toy example to demonstrate. Make sure you state the correct target estimands for the claim being argued Ensure you don't conflate the two issues: denominator trick vs real decline in hazard. Clearly separate: [--]. Why the metric can mislead even with unchanged"  
[X Link](https://x.com/bechhof/status/2022362106545463382)  2026-02-13T17:27Z [----] followers, [--] engagements


"@gabster0191 https://automeris.io/WebPlotDigitizer/ https://automeris.io/WebPlotDigitizer/"  
[X Link](https://x.com/bechhof/status/1229240469365739520)  2020-02-17T03:05Z [----] followers, [--] engagements


"@JukkaSavo @Steve_Sailer @cremieuxrecueil Is that consistent with the geography of the reported homicide *increases*"  
[X Link](https://x.com/bechhof/status/1699568357173805132)  2023-09-06T23:40Z [----] followers, [---] engagements


"@goblinodds @growing_daniel Correct:"  
[X Link](https://x.com/bechhof/status/1724880213090128248)  2023-11-15T20:01Z [----] followers, [---] engagements


"Some of the more honest COVID intervention skeptics predicted that post-pandemic life expectancy would more than recover to the former trend based on the idea that COVID deaths were mostly people already vulnerable to other causes of death. That prediction isnt looking good"  
[X Link](https://x.com/bechhof/status/1729879123395580417)  2023-11-29T15:04Z [----] followers, [---] engagements


"@hyperplanes The arithmetic didnt work for the Lib Dems to have Labour continue"  
[X Link](https://x.com/anyuser/status/1741509083599905205)  2023-12-31T17:18Z [--] followers, [--] engagements


"@StefanFSchubert I predict now this won't end up holding come the actual '24 results"  
[X Link](https://x.com/bechhof/status/1743338634974282179)  2024-01-05T18:28Z [----] followers, [--] engagements


"@generic_void I think this wouldn't matter at all for how much they would work to solve the problem so not worth doing given that Congress is massively underpaid as is"  
[X Link](https://x.com/bechhof/status/1745954456397644216)  2024-01-12T23:42Z [----] followers, [--] engagements


"There are multiple other plausible stories: 1) Haley winning would indicate that Trump is out of the picture somehow and therefore not a threat to the R brand or as an independent. 2) Haley winning would indicate that R primary voters are more sensible than thought and therefore the presidential ticket would benefit from higher quality candidates down ballot. 3) Haley winning would indicate that many centrist voters crossed over to vote in R primaries revealing people who want to vote Republican who would have thought to not necessarily lean that way"  
[X Link](https://x.com/bechhof/status/1746612858526261378)  2024-01-14T19:18Z [----] followers, [---] engagements


"@_alice_evans @instrumenthull @kearney_melissa @washingtonpost Innovation in co-parenting contracts solves this"  
[X Link](https://x.com/bechhof/status/1748515508251467883)  2024-01-20T01:19Z [----] followers, [--] engagements


"@getnormality @olivertraldi Depends which aspects of the pattern you think define the graph"  
[X Link](https://x.com/anyuser/status/1749807342995247159)  2024-01-23T14:52Z [--] followers, [---] engagements


"@getnormality @olivertraldi Notice I didn't make such a claim. Instead I made a claim about the validity of using the center ground specifically the median of public opinion as the baseline"  
[X Link](https://x.com/bechhof/status/1749811966313742509)  2024-01-23T15:10Z [----] followers, [--] engagements


"@getnormality @olivertraldi I don't see how that tracks at all. The median is moving with the broader trend of being more pro immigration so it isn't just about partisan Dems moving hence the bulk is moving. To be sure the reasons for the change are also important (especially generational replacement)"  
[X Link](https://x.com/bechhof/status/1749812990084608205)  2024-01-23T15:15Z [----] followers, [--] engagements


"@economeager @nbouscal sons about to get really into deploying external validity arguments"  
[X Link](https://x.com/bechhof/status/1757992461027119442)  2024-02-15T04:57Z [----] followers, [---] engagements


"@AntBreach @antonhowes @IronEconomist We might prefer the savings rate to increase for other reasons but there isn't any *new* problem: competition between investors ensures that plenty of gains can accrue to workers etc via them capturing some of the increased productivity no new savings on their part required"  
[X Link](https://x.com/bechhof/status/1766884958566584578)  2024-03-10T17:52Z [----] followers, [---] engagements


"@AntBreach @antonhowes @IronEconomist It isn't due to the kindness of strangers. The investment increase comes from the opportunities available domestically that foreign investors can bring to fruition and capture benefits from. They don't however act as a coordinated monolith to ensure they capture all gains"  
[X Link](https://x.com/bechhof/status/1766890983004733894)  2024-03-10T18:16Z [----] followers, [---] engagements


"@AntBreach @antonhowes @IronEconomist The increased output can finance the investment even without the domestic savings *rate* increasing. It's true that in the Solow model diminishing returns prevents this from spiraling indefinitely but the limits to gains from foreign investment still allow for plenty of gains"  
[X Link](https://x.com/bechhof/status/1766896945581248869)  2024-03-10T18:40Z [----] followers, [--] engagements


"@besttrousers Some strong assumptions required for this to work which don't hold IRL"  
[X Link](https://x.com/bechhof/status/1783473064765735155)  2024-04-25T12:28Z [----] followers, [--] engagements


"@robinhanson You made a compound claim which is wrong"  
[X Link](https://x.com/bechhof/status/1784753247481147649)  2024-04-29T01:15Z [----] followers, [---] engagements


"@robinhanson (and the earlier claim that low distrust etc don't count as externalities has a critical dependence on the first part of the compound claim)"  
[X Link](https://x.com/bechhof/status/1784754053236371465)  2024-04-29T01:18Z [----] followers, [--] engagements


"@SilverVVulpes @RuxandraTeslo The bigger problem is that the graph shows geographies not people"  
[X Link](https://x.com/bechhof/status/1787198570933670175)  2024-05-05T19:11Z [----] followers, [--] engagements


"@yhdistyminen Maybe worth having the policy to address this be mandating that listings advertise the parking included price as the default price so people don't misperceive prices when shopping around"  
[X Link](https://x.com/anyuser/status/1813789368152359197)  2024-07-18T04:14Z [----] followers, [---] engagements


"@sebkrier @mattsclancy def working on at least being able to achieve some of this more straightforwardly via standardization with AI's help"  
[X Link](https://x.com/bechhof/status/1815443226393817481)  2024-07-22T17:46Z [----] followers, [--] engagements


"@HiFromMichaelV @robinhanson If this were true GPT-4o would've mentioned it"  
[X Link](https://x.com/bechhof/status/1817627957973680387)  2024-07-28T18:27Z [----] followers, [---] engagements


"@yhdistyminen So use it to fund progressive stuff. I'm happy with regressive taxes if the spending side more than offsets the regressiveness; you could even use it to provide financial management help to low income people"  
[X Link](https://x.com/bechhof/status/1822647042159558755)  2024-08-11T14:51Z [----] followers, [---] engagements


"@S_Surprenant Heat waves are costly to anticipate. The predictability is conditional on knowing there will be a heat wave"  
[X Link](https://x.com/bechhof/status/1822690462349603055)  2024-08-11T17:44Z [----] followers, [--] engagements


"@ryanburge IMO this is better visualized as CDFs"  
[X Link](https://x.com/bechhof/status/1823452340411621381)  2024-08-13T20:11Z [----] followers, [---] engagements


"Price controls are very bad policy to the point that they should not be thought of even as a seldom used part of the sensible public policymakers' toolkit. (This refers to price controls proper not governments negotiating how much they'll pay eg pharmaceutical companies.) Yes price controls are bad. Yes banning price gouging would amount to price controls. I guess this needs to be retaught. https://t.co/KxXMUHZGit Yes price controls are bad. Yes banning price gouging would amount to price controls. I guess this needs to be retaught. https://t.co/KxXMUHZGit"  
[X Link](https://x.com/bechhof/status/1824516780330561806)  2024-08-16T18:41Z [----] followers, [---] engagements


"@caseybmulligan There's no shock. The fact that it's an NPV is a detail of how the adding up is done which doesn't negate that it's *30 years* of costs which it's worth reiterating are *gross* costs"  
[X Link](https://x.com/bechhof/status/1825646470072074478)  2024-08-19T21:30Z [----] followers, [--] engagements


"@george_berry Post your question here and I'll tell you"  
[X Link](https://x.com/bechhof/status/1826031694094459271)  2024-08-20T23:00Z [----] followers, [--] engagements


"@tailcalled @dingding_peng Whatever result the RCT yields is potentially important and this doesn't negate the point at all that standard statistical models can be very useful in lots of settings (including if you want to validate the claim that an effect doesn't exist to much extent)"  
[X Link](https://x.com/bechhof/status/1828444968123343339)  2024-08-27T14:50Z [----] followers, [--] engagements


"@tailcalled @dingding_peng This is a question of where to target one's statistical analyses not whether they're workable which they often are as illustrated by how effective comparing average outcomes in an RCT is"  
[X Link](https://x.com/bechhof/status/1828447768697880836)  2024-08-27T15:01Z [----] followers, [--] engagements


"@tailcalled @dingding_peng 1) you get value by confirming they're not powerful (with standard statistical analyses) 2) you get value in the cases where you're looking at things that are powerful enough (with standard statistical analyses)"  
[X Link](https://x.com/bechhof/status/1828450257031827593)  2024-08-27T15:11Z [----] followers, [--] engagements


"@tailcalled @dingding_peng Okay this seems both incorrect and a different claim than the one that standard models aren't workable"  
[X Link](https://x.com/bechhof/status/1828451441419153607)  2024-08-27T15:16Z [----] followers, [--] engagements


"@tailcalled @dingding_peng Okay that does seem like quite a different claim than implied but regardless I think they're more fruitful than trying to drill down into outliers in typical cases"  
[X Link](https://x.com/bechhof/status/1828454177544249621)  2024-08-27T15:27Z [----] followers, [--] engagements


"@tailcalled @dingding_peng Figuring out which demographic subgroups have the highest propensity to buy a certain product"  
[X Link](https://x.com/bechhof/status/1828454871407960168)  2024-08-27T15:29Z [----] followers, [--] engagements


"@tailcalled @dingding_peng I'm not saying RCAs are bad. I'm saying standard statistical analyses are very often quite useful and in this sort of case doing a bunch of statistical modeling with regularization can be useful in and of itself (and as a compliment to RCAs)"  
[X Link](https://x.com/bechhof/status/1828538719043027182)  2024-08-27T21:02Z [----] followers, [--] engagements


"@tailcalled @dingding_peng That seems like a very open empirical question which we'd want to look at with. more standard statistical methods"  
[X Link](https://x.com/bechhof/status/1829594122883059862)  2024-08-30T18:56Z [----] followers, [--] engagements


"@tailcalled @dingding_peng You'd use the statistical methods to test how effective root cause analysis is at accelerating progress"  
[X Link](https://x.com/bechhof/status/1829596267217129633)  2024-08-30T19:05Z [----] followers, [--] engagements


"@tailcalled @dingding_peng That seems very much incorrect to me: statistics is *complementary* to exploratory work"  
[X Link](https://x.com/bechhof/status/1829600101121343771)  2024-08-30T19:20Z [----] followers, [--] engagements


"@tailcalled @dingding_peng You could do an RCT between different sets of data teams"  
[X Link](https://x.com/bechhof/status/1829603259268678012)  2024-08-30T19:33Z [----] followers, [--] engagements


"@tailcalled @dingding_peng I don't think we know that with anything close to the definitiveness you're claiming"  
[X Link](https://x.com/bechhof/status/1829609556911313315)  2024-08-30T19:58Z [----] followers, [--] engagements


"@tailcalled @dingding_peng Then your broader claims about the usefulness of standard statistical methods may not generalize much"  
[X Link](https://x.com/bechhof/status/1829612244348621303)  2024-08-30T20:08Z [----] followers, [--] engagements


"@tailcalled @dingding_peng That isn't sufficient motivation"  
[X Link](https://x.com/bechhof/status/1829612666903806335)  2024-08-30T20:10Z [----] followers, [--] engagements


"@ashebytes I'm working on one at least"  
[X Link](https://x.com/bechhof/status/1848109838775005259)  2024-10-20T21:11Z [----] followers, [--] engagements


"@ashebytes One thing that people seem to like: Everything looks like a coefficient plot rather than tables with certain estimates being a full dot and uncertain estimates a hollow dot to make the interval being plotted really obvious"  
[X Link](https://x.com/bechhof/status/1848111523731800150)  2024-10-20T21:18Z [----] followers, [--] engagements


"All incorrect / overstated: Plenty of bipartisan issues that get worked through eg earlier this year bipartisan consensus on nuclear emerged even though that debate has been prominent for a long time Left vs right seems in current day to be about aspiring to be like an imagined future vs imagined past and thus probabilistically implies various positions Partisan elites build coalitions rather than just picking their fondness for whatever at random. That coalition building process while it has random elements to it is the biggest driver of what partisan elites choose and therefore what"  
[X Link](https://x.com/bechhof/status/1850353777334747542)  2024-10-27T01:48Z [----] followers, [--] engagements


"@olivertraldi Funny to show the graph as consistent with what you're saying given that the movement is mostly more recent than a decade ago. Regardless you shouldn't give it zero weight as I said upthread"  
[X Link](https://x.com/bechhof/status/1857588445247320419)  2024-11-16T00:56Z [----] followers, [--] engagements


"American workers win by: 1) getting more & higher quality consumption 2) getting more as owners of capital 3) getting a government that can collect more revenue at the same rates 4) getting higher wages from complementary workers 5) getting wonders of scaled civilization The Indian wage gap arbitrage. It appears to work this way: Indian co imports workers and pays them massively under market salary then turns around and contracts them out for slightly below market to US clients. Everyone wins except the American worker from depressed wages. The Indian wage gap arbitrage. It appears to work"  
[X Link](https://x.com/bechhof/status/1873454771002278132)  2024-12-29T19:43Z [----] followers, [---] engagements


"@krichard1212 Buy me an Apple Vision Pro first"  
[X Link](https://x.com/bechhof/status/1877256119267082567)  2025-01-09T07:28Z [----] followers, [--] engagements


"@robinhanson It's true that it _can_ be via proper selection but I think by far the bigger story is that humans via their cognitive capabilities can scale tech etc quite quickly so change that happens for other reasons even random drift can accumulate quickly"  
[X Link](https://x.com/bechhof/status/1878527787448889409)  2025-01-12T19:41Z [----] followers, [---] engagements


"@OsitaNwanevu To me it's sort of the reverse that's convincing: you don't need high net worth to have outsized influence. You can have eg [--] million YT subscribers and be a similar threat"  
[X Link](https://x.com/bechhof/status/1882150845426593997)  2025-01-22T19:38Z [----] followers, [---] engagements


"@MarkinAngles @OsitaNwanevu Mr. Beast hasn't tried to amass power using his assets in the same way sure. He's more on par with Alice Walton in his (current) capability to influence outcomes"  
[X Link](https://x.com/bechhof/status/1882155113546211335)  2025-01-22T19:55Z [----] followers, [---] engagements


"@KelseyTuoc On reflection my biggest frustration with your piece is the lack of international comparison. Why did we not get better outcomes like they had in successful places like Vietnam"  
[X Link](https://x.com/bechhof/status/1903546712418885747)  2025-03-22T20:37Z [----] followers, [---] engagements


"@jmhorp Composition effects: population growth is a lot faster on average for SSA countries with low incomes so that offsets the growth from the countries that have done well"  
[X Link](https://x.com/anyuser/status/1906215153579782342)  2025-03-30T05:21Z [----] followers, [----] engagements


"@olivertraldi This is a bad counter. A decade ago MBDTF was like his defining album and this stuff is generally not considered a continuation of that even by ongoing fans"  
[X Link](https://x.com/bechhof/status/1920487146151989353)  2025-05-08T14:33Z [----] followers, [---] engagements


"@fchollet Please I beg you to read @ATabarrok on this graph so you're not misled https://marginalrevolution.com/marginalrevolution/2019/08/are-health-administrators-to-blame.html https://marginalrevolution.com/marginalrevolution/2019/08/are-health-administrators-to-blame.html"  
[X Link](https://x.com/anyuser/status/1926396424301715661)  2025-05-24T21:54Z [----] followers, [----] engagements


"@polskrt @MattBruenig Single payers deny plenty. They don't just uniformly deny less on every margin. Consumers can choose; insurers can play with different actuarial processes; providers can be negotiated with differently / opt in to more innovative payment systems among other things"  
[X Link](https://x.com/bechhof/status/1927138345303339400)  2025-05-26T23:02Z [----] followers, [---] engagements


"@MattPirkowski @Schorch_D_W @ryxcommar @theramblingfool @gpu_thief This is such a bizarre response for anyone who knows what a softmax function is"  
[X Link](https://x.com/bechhof/status/1927263444190773611)  2025-05-27T07:19Z [----] followers, [--] engagements


"@NeumarkEcon 1) that's only for citizens not people under Israeli control that don't get citizenship automatically despite having no choice but to accept Israeli authority unless they leave everything behind 2) part of the point is to not have "Muslim" countries either 3) see"  
[X Link](https://x.com/bechhof/status/1930798518341423607)  2025-06-06T01:26Z [----] followers, [--] engagements


"@JakeMGrumbach That's how the right loses the Hanania faction. I guess"  
[X Link](https://x.com/bechhof/status/1934781842525126687)  2025-06-17T01:15Z [----] followers, [--] engagements


"@Corey_Yanofsky @jeffreyhelzner It's all a mess because ecological inference but anyone who wants the disaggregating can pull degree fields by area from ACS"  
[X Link](https://x.com/bechhof/status/1938071036525748327)  2025-06-26T03:05Z [----] followers, [--] engagements


"This took me three minutes in IPUMS Abacus looking at employed Americans born in the [--] states + DC of Japanese ancestry vs overall on the Duncan Socioeconomic Index of occupations. Is there a meaningful difference Sure. Is the magnitude anything like Caplan implies No"  
[X Link](https://x.com/bechhof/status/1939715999541895259)  2025-06-30T16:01Z [----] followers, [---] engagements


"Or there's the logistic regression version for the janitor occupation specifically. And. employed Japanese-Americans born in the [--] states + DC are about half as likely to be janitors. Definitely less likely but "virtually never" is preposterous. @StefanFSchubert"  
[X Link](https://x.com/bechhof/status/1939721507766444495)  2025-06-30T16:23Z [----] followers, [---] engagements


"@mattppal @59thProfile @JimmyAustin Have you fixed the issue where the change log gets absolutely enormous degrading the quality of responses by taking up so much of the context window"  
[X Link](https://x.com/bechhof/status/1942243230449275360)  2025-07-07T15:24Z [----] followers, [--] engagements


"@lymanstoneky @paulnovosad @RuxandraTeslo The limits themselves are in flux because of innovation and set to be less and less constraining (though how fast is a policy choice) which is much of Ruxandra's point. As such it's a reasonable baseline to fit a trend line and extrapolate for the next [--] years or so"  
[X Link](https://x.com/bechhof/status/1942695526413017214)  2025-07-08T21:21Z [----] followers, [---] engagements


"@salonium @lymanstoneky @paulnovosad @RuxandraTeslo I referred to the article to show that the idea that the falling is near universal or particularly drastic is wrong with Sweden being just one example of a country where the trend looks much more like an asymptote (for the moment) than an ongoing dramatic fall"  
[X Link](https://x.com/bechhof/status/1943178341403529576)  2025-07-10T05:19Z [----] followers, [---] engagements


"@GallenYana @paulnovosad @RuxandraTeslo @lymanstoneky @OurWorldInData Thanks for the clarification here Censoring is a problem It's true we can't see what those cohorts will do later. The validity of that point is not good reason to make a literally false claim. (Though @StefanFSchubert is more rigorous re this than me) https://x.com/GallenYana/status/1943375922888929346 @bechhof @paulnovosad @RuxandraTeslo @lymanstoneky @OurWorldInData *post [----] https://x.com/GallenYana/status/1943375922888929346 @bechhof @paulnovosad @RuxandraTeslo @lymanstoneky @OurWorldInData *post 2010"  
[X Link](https://x.com/bechhof/status/1943376698025578869)  2025-07-10T18:28Z [----] followers, [--] engagements


"@lymanstoneky @paulnovosad @RuxandraTeslo @GallenYana That. isn't how the mathematics of TFR works. Cohorts can be exposed to TFR under [--] for all of their childbearing years and still have a completed cohort birth rate above [--] per woman"  
[X Link](https://x.com/bechhof/status/1943413500216316207)  2025-07-10T20:54Z [----] followers, [--] engagements


"@krichard1212 @grok https://grok.com/share/bGVnYWN5_521a7408-7f70-423d-9207-fccc7beb426d https://grok.com/share/bGVnYWN5_521a7408-7f70-423d-9207-fccc7beb426d"  
[X Link](https://x.com/bechhof/status/1945240859143778413)  2025-07-15T21:55Z [----] followers, [--] engagements


"@krichard1212 @grok GPT [---] https://chatgpt.com/s/t_6876cef024908191a4dc41bc5dd15f79 https://chatgpt.com/s/t_6876cef024908191a4dc41bc5dd15f79"  
[X Link](https://x.com/bechhof/status/1945241545004703758)  2025-07-15T21:58Z [----] followers, [---] engagements


"@oelayat That's your example Starmer won and won big"  
[X Link](https://x.com/bechhof/status/1945246654304583830)  2025-07-15T22:18Z [----] followers, [----] engagements


"@beetlesarefake @oelayat We have some solid examples even if not ideal. Clinton in the USA. Blair in the UK. Macron in France. Proven election winners and reformers all"  
[X Link](https://x.com/bechhof/status/1945253985729785973)  2025-07-15T22:47Z [----] followers, [--] engagements


"@HenryGB_274 @oelayat You know we have a more complete example of when Labour was ruled by centrists. Tony Blair won three elections"  
[X Link](https://x.com/bechhof/status/1945263759070507094)  2025-07-15T23:26Z [----] followers, [---] engagements


"@typingduck @oelayat Who cares about votes for the party The thing that matters is winning power"  
[X Link](https://x.com/bechhof/status/1945330498835828795)  2025-07-16T03:51Z [----] followers, [---] engagements


"@OfSymbols @typingduck @oelayat It is really good by recent UK standards and solid even for UK history generally. Whether it's front loaded remains to be seen"  
[X Link](https://x.com/bechhof/status/1945518651597013119)  2025-07-16T16:19Z [----] followers, [--] engagements


"@RedOxiOxi @oelayat They did well on vote share but very poorly on seats fragmenting the opposition to Labour. Funny to reference polling as a reason to oppose Starmer and also hating on the two child cap which has polled as popular for [--] years in Britain"  
[X Link](https://x.com/bechhof/status/1945520429503852744)  2025-07-16T16:26Z [----] followers, [--] engagements


"@michael_wiebe IZ is such a poorly implemented tax. It would be so much better if developers could just pay a fee to opt out of it so they didn't have to devote the bandwidth to complying with all the nuances. Having a $ payment is way better than these messy frictions"  
[X Link](https://x.com/bechhof/status/1945522432535671154)  2025-07-16T16:34Z [----] followers, [--] engagements


"@tatamitimon @ATabarrok Then factories are always open (so you don't have capital going idle) and people can decide they don't want to be part of the standard grind. In non-communist systems we have people who work on weekends to make the people with time off have a better life"  
[X Link](https://x.com/anyuser/status/1947424739838071004)  2025-07-21T22:33Z [----] followers, [----] engagements


"@ATabarrok That piece is so embarrassing for socialists. Like at least use one of the clever arguments like wastefulness of redundant logistics infrastructure rather than something that can't stay coherent within the length of a newspaper opinion piece"  
[X Link](https://x.com/anyuser/status/1947683095899226121)  2025-07-22T15:40Z [----] followers, 16.8K engagements


"@besttrousers @JakeMGrumbach I don't agree with Jake FWIW that it's a reasonable model since it excludes real world offsetting mechanisms eg entrepreneurship in response to higher margins & can't account for inflation going on forever w/o some bizarre mechanism like super high labor supply elasticities"  
[X Link](https://x.com/bechhof/status/1948137505578238443)  2025-07-23T21:45Z [----] followers, [--] engagements


"@besttrousers @JakeMGrumbach Yeah I agree with that bar so I'll clarify that I don't think the model mechanism is inherently goofy but that applying it as an important explanation of an extended period of inflation (not just the initial parts) which isn't followed by deflationary backlash is"  
[X Link](https://x.com/bechhof/status/1948138928562704626)  2025-07-23T21:51Z [----] followers, [--] engagements


"@KMAG_YOLO It never did. Gen Z [--] somethings live with roommates a plenty and people are inferring otherwise based on TikTok clips rather than data from CPS/ACS/companies like Flex indicating very high prevalence of roommate arrangements"  
[X Link](https://x.com/bechhof/status/1949159103378436457)  2025-07-26T17:25Z [----] followers, [--] engagements


"@RuxandraTeslo @GallenYana @paulnovosad @20 Very much like those who insist clean energy doesn't count as a solution vs cutting energy use and point to early signs of low *take-up* (not even effectiveness) of new tech as definitive proof that it'll never work"  
[X Link](https://x.com/bechhof/status/1949629661555597597)  2025-07-28T00:35Z [----] followers, [--] engagements


"@wastetime @CartoonsHateHer The whole story about Sydney Sweeney is about her being registered not just vaguely self identified"  
[X Link](https://x.com/bechhof/status/1952134162376474807)  2025-08-03T22:27Z [----] followers, [--] engagements


"@george_berry I read GEM as saying that [--]. moderation to get votes is overhyped as a tactic by pro moderation pundits [--]. given that whatever costs you may pay for moderation (bandwidth expenditure for recruitment constrained policy in office) may not be worthwhile on the margin"  
[X Link](https://x.com/bechhof/status/1956121167628259503)  2025-08-14T22:30Z [----] followers, [--] engagements


"@michael_wiebe It doesn't. Everyone can have finite lifespan and it still works just fine"  
[X Link](https://x.com/bechhof/status/1961558317047648347)  2025-08-29T22:35Z [----] followers, [--] engagements


"@PoundstoneWill @jmhorp I agree most of the income growth is from rising hourly pay. The other points were to make explicit that accounting for complications like composition changes makes the story look even worse"  
[X Link](https://x.com/bechhof/status/1962990637013221537)  2025-09-02T21:26Z [----] followers, [--] engagements


"@PoundstoneWill @jmhorp Sure that is a legit offsetting factor though in real terms the value of it isn't a great contributor since rising medical care prices have eaten so much of the benefit"  
[X Link](https://x.com/bechhof/status/1962993932411019410)  2025-09-02T21:39Z [----] followers, [--] engagements


"@PoundstoneWill @jmhorp PCE deflator is wrong to use for medical care benefits specifically when thinking about living standards. The prices of medical care have risen *much* more quickly in the USA than price levels generally"  
[X Link](https://x.com/bechhof/status/1962994578346201582)  2025-09-02T21:42Z [----] followers, [--] engagements


"@robinhanson @bryan_caplan That means that medical care is actually having substantial effects on outcomes rather than the naive inference from RAND/Oregon that it's all just nulls / noise. If variation in medical care matters for variation in outcomes we can use that to optimize for better outcomes"  
[X Link](https://x.com/bechhof/status/1963024562599154025)  2025-09-02T23:41Z [----] followers, [---] engagements


"People who make these predictive arguments never present proper psychometric reasoning for why this would be good. They appeal to claims about g-loading which. isn't the point of an admissions test. They can't point to reduced predictive validity because that never happened. The best way to reform the SAT is to just reinstate the test format from [----]. The best way to reform the SAT is to just reinstate the test format from 1988"  
[X Link](https://x.com/bechhof/status/1971468120587108734)  2025-09-26T06:53Z [----] followers, [---] engagements


"@AlanMCole @ryanhmurphy I guess you are literally type of guy who finds Tucker's loud opposition to "usury" and credit cards to be among his most offensive brand components (though I guess I am too.)"  
[X Link](https://x.com/bechhof/status/1973577005259170206)  2025-10-02T02:33Z [----] followers, [--] engagements


"This is a ridiculous proposal: http://t.co/rXtRymh3Jc Better: a tax on men http://gu.com/p/4329m/stw http://gu.com/p/4329m/stw"  
[X Link](https://x.com/bechhof/status/529671544574717952)  2014-11-04T16:28Z [----] followers, [--] engagements


"@LeightonVW @Survation Why not YouGov"  
[X Link](https://x.com/bechhof/status/873111566152908800)  2017-06-09T09:36Z [----] followers, [--] engagements


"Sherlock Holmes stories are bad because they glorify overfitting"  
[X Link](https://x.com/anyuser/status/1083212927744671745)  2019-01-10T04:04Z [----] followers, [---] engagements


"@ATabarrok That piece is so embarrassing for socialists. Like at least use one of the clever arguments like wastefulness of redundant logistics infrastructure rather than something that can't stay coherent within the length of a newspaper opinion piece"  
[X Link](https://x.com/anyuser/status/1947683095899226121)  2025-07-22T15:40Z [----] followers, 16.8K engagements


"Imagine reading these arguments in a typical discrimination case and think how much credence youd giveconditioning on post treatment variables arguments about unobservable quality etc all here https://x.com/dynarski/status/1007624830622031872 https://x.com/dynarski/status/1007624830622031872"  
[X Link](https://x.com/anyuser/status/1007694424648835072)  2018-06-15T18:40Z [----] followers, [--] engagements


"@jmhorp Composition effects: population growth is a lot faster on average for SSA countries with low incomes so that offsets the growth from the countries that have done well"  
[X Link](https://x.com/anyuser/status/1906215153579782342)  2025-03-30T05:21Z [----] followers, [----] engagements


"@laurenbalik That's got to be selection bias explaining which H-1B workers you encounter; probably 60%+ of H1-B holders can find x=2 just fine"  
[X Link](https://x.com/anyuser/status/1910801585279476051)  2025-04-11T21:06Z [----] followers, 20.6K engagements


"Im very pleased to be sharing this a @salonium and @bechhof expression of a radical vision for more effective scientific production Its time to rebuild scientific institutions for the 21st century complete with frontier tools and practices. https://worksinprogress.co/issue/the-speed-of-science/ https://worksinprogress.co/issue/the-speed-of-science/"  
[X Link](https://x.com/anyuser/status/1358844841208160256)  2021-02-08T18:27Z [----] followers, [--] engagements


"If you read one thing about John Nash today read this amazing post from @Afinetheorem https://afinetheorem.wordpress.com/2015/05/24/the-economics-of-john-nash/utm_source=twitterfeed&utm_medium=twitter https://afinetheorem.wordpress.com/2015/05/24/the-economics-of-john-nash/utm_source=twitterfeed&utm_medium=twitter"  
[X Link](https://x.com/anyuser/status/602529556386488320)  2015-05-24T17:40Z [----] followers, [--] engagements


"Think of Steven Pinkers Blank Slate and Better Angels of Our Nature as analogous to the Old and New Testament"  
[X Link](https://x.com/anyuser/status/1066927186828058624)  2018-11-26T05:30Z [----] followers, [--] engagements


"@fchollet Please I beg you to read @ATabarrok on this graph so you're not misled https://marginalrevolution.com/marginalrevolution/2019/08/are-health-administrators-to-blame.html https://marginalrevolution.com/marginalrevolution/2019/08/are-health-administrators-to-blame.html"  
[X Link](https://x.com/anyuser/status/1926396424301715661)  2025-05-24T21:54Z [----] followers, [----] engagements


"@rubenbmathisen It seems pretty misleading to condition on working more than 35hr/week to begin with"  
[X Link](https://x.com/anyuser/status/1555399721581256709)  2022-08-05T03:46Z [----] followers, [--] engagements


"@bryan_caplan Great Tinder bio"  
[X Link](https://x.com/anyuser/status/1065841307732594688)  2018-11-23T05:35Z [----] followers, [--] engagements


"@ATabarrok That's. kind of awesome. All they needed to complete the system was a well designed matching mechanism and that would've been ideal"  
[X Link](https://x.com/anyuser/status/1947380945406836887)  2025-07-21T19:39Z [----] followers, 22.1K engagements


"@SeanMcElwee the kids wouldnt have to be white"  
[X Link](https://x.com/anyuser/status/754525032613838849)  2016-07-17T03:56Z [----] followers, [--] engagements


"@StefanFSchubert I suspect that time invested in make-up is underreported and time invested in exercise is over reported tho"  
[X Link](https://x.com/anyuser/status/1509226670405816321)  2022-03-30T17:50Z [----] followers, [--] engagements


"@MTabarrok @catehall That's wrong. Turns out you can get much better information when you ditch polluting accounts like cremieux"  
[X Link](https://x.com/anyuser/status/1927185445063942564)  2025-05-27T02:09Z [----] followers, [----] engagements


"@eigenrobot Im disappointed you havent given any of my tweets an extraordinary"  
[X Link](https://x.com/anyuser/status/967830445336363009)  2018-02-25T18:35Z [----] followers, [--] engagements


"@dylanmatt @AndyMasley This genuinely reads as one or both of [--]. content for men into femdom to be aroused by [--]. putting a call out for men to try to prove themselves to the author by defying the pessimistic expectations"  
[X Link](https://x.com/anyuser/status/1948102970303823997)  2025-07-23T19:28Z [----] followers, [----] engagements


"Another big break around 1973: http://t.co/MqDxZ3WWtr"  
[X Link](https://x.com/anyuser/status/506919298812891137)  2014-09-02T21:39Z [----] followers, [--] engagements


"As a birthday present to myself we are launching @TrovBase: a data management platform for researchers that makes best practices accessible and standardized. Im psyched to be working with @samcjordan_ and @AndrewHornstra to help build tooling for modern open science F it were building in public. Introducing: @TrovBase https://t.co/GKWl6ThFoh F it were building in public. Introducing: @TrovBase https://t.co/GKWl6ThFoh"  
[X Link](https://x.com/anyuser/status/1541432747696308225)  2022-06-27T14:46Z [----] followers, [--] engagements


"F it were building in public. Introducing: @TrovBase https://vast-bird-dd8.notion.site/TrovBase-481512fcc40848499c5a41476f13f06a https://vast-bird-dd8.notion.site/TrovBase-481512fcc40848499c5a41476f13f06a"  
[X Link](https://x.com/anyuser/status/1541429823704383491)  2022-06-27T14:34Z [----] followers, [--] engagements


"@arpitrage I think it's sensible as a price given the constraints. The bigger issue is how cheap the highway alternative is"  
[X Link](https://x.com/anyuser/status/1914407351529988102)  2025-04-21T19:54Z [----] followers, [----] engagements


"Remember this when you get excited about a new behavioral economics result"  
[X Link](https://x.com/anyuser/status/709145640941576193)  2016-03-13T22:34Z [----] followers, [--] engagements


"@MrGeorgeFrancis @whyvert @KirkegaardEmil Cranial size as instrumental variable is like a self parody"  
[X Link](https://x.com/anyuser/status/1571023581987078144)  2022-09-17T06:29Z [----] followers, [--] engagements


"@eigenrobot Any moral system that cant turn any moral dilemma into a well-posed optimization problem should be laughed out of the discipline of moral philosophy"  
[X Link](https://x.com/anyuser/status/967831367072464896)  2018-02-25T18:39Z [----] followers, [--] engagements


"@robinhanson I can guarantee this person has no IQ 175+ friends"  
[X Link](https://x.com/anyuser/status/1842950674683990476)  2024-10-06T15:30Z [----] followers, [---] engagements


"@dylanmatt actually it's arrived at by listening to podcasts"  
[X Link](https://x.com/anyuser/status/1028057767607562241)  2018-08-10T23:17Z [----] followers, [--] engagements


"Essential #longread from @slatestarcodex on comparative advantage how to think about innate talent differences http://t.co/mv2nSyb3iV http://slatestarcodex.com/2015/01/31/the-parable-of-the-talents/ http://slatestarcodex.com/2015/01/31/the-parable-of-the-talents/"  
[X Link](https://x.com/anyuser/status/561690228784893952)  2015-02-01T00:59Z [----] followers, [--] engagements


"Its much more fun watching Queer Eye if you imagine that theres a policy evaluation team behind the scenes. How much fadeout is there How scalable can this intervention be What kind of stratification did they use to choose the sample"  
[X Link](https://x.com/anyuser/status/978875875897864192)  2018-03-28T06:06Z [----] followers, [--] engagements


"@wwwojtekk If I were the government I would simply choose the most productive production technology"  
[X Link](https://x.com/anyuser/status/1564108638293073920)  2022-08-29T04:32Z [----] followers, [--] engagements


"@robinhanson Making sex inequality more salient makes its consequences worse"  
[X Link](https://x.com/anyuser/status/989570148037115904)  2018-04-26T18:21Z [----] followers, [--] engagements


"@jakevdp please tell me you made this in matplotlib"  
[X Link](https://x.com/anyuser/status/1058906761795592193)  2018-11-04T02:20Z [----] followers, [--] engagements


"@robinhanson Not generally and the resentment as a function of inequality is much smoother in the income case"  
[X Link](https://x.com/anyuser/status/990287551347937281)  2018-04-28T17:52Z [----] followers, [--] engagements


"@jmhorp @ne0liberal Idk what kind of insane inflation adjustment is happening here though. Like no country in history has had GDPpc that low (and even PPP wont get you all the way to $90)"  
[X Link](https://x.com/anyuser/status/1523428841665941505)  2022-05-08T22:25Z [----] followers, [--] engagements


"I've never seen causes of death graphed quite this way before. It really illustrates how other causes "crowd out" these among older people. By request Here's an extended version of the chart that shows the breakdown of different external causes of death. This breaks down deaths from different external causes by age in the United States. Grey represents all other causes. https://t.co/23S6A0qfQA By request Here's an extended version of the chart that shows the breakdown of different external causes of death. This breaks down deaths from different external causes by age in the United States."  
[X Link](https://x.com/anyuser/status/1769364263372476556)  2024-03-17T14:04Z [----] followers, [----] engagements


"By request Here's an extended version of the chart that shows the breakdown of different external causes of death. This breaks down deaths from different external causes by age in the United States. Grey represents all other causes"  
[X Link](https://x.com/anyuser/status/1769360003939402232)  2024-03-17T13:47Z 34.7K followers, [----] engagements


"@AlecStapp I had ChatGPT make this graph showing it wasn't just old people (and this is overall mortality to pre-empt nonsense about "with but not of COVID")"  
[X Link](https://x.com/anyuser/status/1925307165851488621)  2025-05-21T21:46Z [----] followers, [----] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing