On white flight (from the comments)

by  Tyler Cowen in

Are whites fleeing from Asian-heavy California public schools?  One recent paper suggested maybe so, but abc raises some doubts:

I don’t want to dismiss the paper out of hand, as I have seen time and again the challenges communities face both in and outside of the school setting in accommodating demographic change.

However, I don’t think the headline result in this paper is particularly credible. First, there isn’t a well-articulated research question to guide the choice of regression. Second, the authors implicitly rely on the “an instrument is always better” fallacy rather than explaining why their instrument yields more reliable estimates than naive OLS for the (unstated) question of interest. Taken together, the paper is undergrad-thesis level material elevated only by a click bait topic and result. If we want to make bold claims about White animosity towards Asians (a claim that also constructive of such animosity and counter-animosity from Asians towards Whites) we should demand substantive evidence. This paper does not present such evidence.

Some key takeaways:

(1) The authors note that a mechanical housing market replacement would suggest a one-for-one effect, but say that their -1.47 effect is above that threshold. However, if we check the confidence interval using a conservative 1.96 critical value and the estimated standard error of the coefficient estimate, we have -1.47 + 1.96*0.268 = -0.96 so that we are not statistically significantly different from -1 by this measure.

(2) The naive OLS estimate in high-SES regions is -0.6, well below the fixed enrollment effect of -1. The authors speculate that OLS may be biased downward because the error term include unmeasured district quality changes that draw in both Asians and Whites. (Note such a correlation only operates if enrollment is not capped, so inconsistent with that model.) The authors don’t document any of these omitted variable issues, however, and just assert that their instrument will be better.

(3) Authors do not substantively engage issues with their IV. First, the IV doesn’t account for changes in composition of immigrants over time (increasing wealth and education of Asian arrivals relative to earlier waves) nor does it account for movement of second-generation Asian families. If there is no omitted variable bias but the instrumented entry is lower than the actual entry, then mechanically the coefficient will have to be higher to offset this effect and restore least-squares minimization.

(4) The instrumented Asian inflows coefficient could pick up effects from Asian-agglomeration effects. A one unit increase in Asian enrollment from pure fixed-pattern immigration flows made lead to shifts of previously settled Asians or shift the direction of subsequent immigration. For example, a settled Korean in Riverside who sees large increases in Korean population in Orange County may see OC as being more attractive than before and move into the area. This induced shift may be only partially captured by the first-stage prediction, leaving the 2nd stage coefficient of interest to increase in magnitude.

(5) Various sensitivities lead to surprising results. First, the instrument behaves poorly in some subsamples, e.g. the bottom-half of the SES scale. Why should we believe an instrument in one data subset when it plainly fails in the complement? Second, the instrument is insignificant in the Bottom Tercile of the above-median SES group (appendix table 2). Third, the IV estimate is only -0.841 in the top tercile of the above-median SES group, again below the key -1 threshold if enrollment caps are binding. Taken together, are we to think that we can identify white flight using this instrument only for the 66.6th to 83.3th percentile bucket?

(6) There’s just a big background trend issue that one has to worry about here. The theory of white flight begs the question of “flight to where?” However if we just look at Appendix Figure 2 during this time period there is a big drop in total White enrollment (and a small decline in Black enrollment) while Asian and Hispanic enrollment see big increases. To what extent are we just finding that aging out of whites in high-SES regions is being replaced disproportionately by Asians?

(7) A couple other wrinkles: how are mixed-race students handled? how would demographic shifts in total enrollment by district affect the 1-to-1 threshold? If child population is shrinking over time (e.g. because families are leaving CA, children per family is declining) then normal churn would predict more than 1-to-1 replacement of new-cohort race versus previous-cohort race.

So perhaps the right answer is “no”?

Against human-AI collaboration

by  Tyler Cowen in

From a new NBER working paper by Nikhil Agarwal, Alex Moehring, Pranav Rajpurkar, and Tobias Salz:

Radiologists do not fully capitalize on the potential gains from AI assistance because of large deviations from the benchmark Bayesian model with correct belief updating. The observed errors in belief updating can be explained by radiologists’ partially underweighting the AI’s information relative to their own and not accounting for the correlation between their own information and AI predictions. In light of these biases, we design a collaborative system between radiologists and AI. Our results demonstrate that, unless the documented mistakes can be corrected, the optimal solution involves assigning cases either to humans or to AI, but rarely to a human assisted by AI.

I am more optimistic in my views, noting there may well be contexts such as radiology where the collaborations fail.  I collaborate with Google’s AI all the time, and I am pretty sure that joint effort does better than either myself or “Google with no human” unaided.  Still, this is a cautionary note of some import, as many humans are not good enough to work well with AIs.

The Brazil Tax

by  Alex Tabarrok in

The thread with many Brazilians sharing similar stories is worth perusing.

Intergenerational transmission of mental health problems

by  Tyler Cowen in

We estimate health associations across generations and dynasties using information on healthcare visits from administrative data for the entire Norwegian population. A parental mental health diagnosis is associated with a 9.3 percentage point (40%) higher probability of a mental health diagnosis of their adolescent child. Intensive margin physical and mental health associations are similar, and dynastic estimates account for about 40% of the intergenerational persistence. We also show that a policy targeting additional health resources for the young children of adults diagnosed with mental health conditions reduced the parent-child mental health association by about 40%.

That is from a new NBER working paper from Aline Bütikofer, Krzysztof Karbownik, and Fanny Landaud.

Why Singaporean democracy is like a social media graph

by  Tyler Cowen in

The Singaporean polity is far more democratic than most underinformed outsiders realize.  Nonetheless it is frequently observed that the same party — PAP — keeps on winning elections.  Furthermore, there is an extreme method of gerrymandering, so PAP might win sixty percent of the vote and end up with ninety percent or so of the seats in the legislature.

Whatever you think of that arrangement, I expect it will prove difficult to undo.  In Singapore, civil service jobs are extremely important.  They are well paid and attract a very high quality of elite, and they are a major means of networking and advancing your own reputation, if only because so many other elites are in government as well.  (On top of that, the start-up scene is not so dynamic there, so the opportunity cost of public service is lower than in say CA or NYC.)

All that contributes to Singapore having an extremely high quality civil service.

But these networks have elements of natural monopoly to them.  If you are a talented Singaporean, of course you will view the PAP as the natural vehicle for your efforts, even if you disagree with a lot of PAP policies.  Working from within the PAP would be the most logical attempt to change the system, as there is “no other game in town.”  That is better for you, and it is better for Singapore as well.

So, if only for careerist reasons (put aside performance and reelection issues, though of course they are important), PAP is a self-replicating network that maintains a very high degree of influence over the Singaporean polity.  It is the network you have to join.

You could imagine the PAP someday suffering a shocking electoral loss, just as the election of Trump shocked many American elites.  But alternate parties would not have the talent infrastructure to staff their own regime with their preferred points of view, not with anything remotely resembling the current level of competence.  For better or worse, PAP affiliates still would be running most of the government.

And that is one reason — by no means the only reason — why it is difficult for Singaporean democracy to become truly contestable.

You will note of course that many American cities — some with roughly the population of Singapore — also keep on electing the same party repeatedly.  If you want to change Chicago city politics at the electoral level, working through the Democratic Party is probably your primary option.  And so this problem of natural monopoly political networks extends well beyond Singapore.

Singapore has less policy accumulation

by  Tyler Cowen in

…we find that Singapore (1) has only produced about one-fourth of the environmental policy measures of an “average” democracy and (2) is constantly the country with the lowest level of policy accumulation in our sample. These findings hold even when controlling for alternative explanations, such as the effectiveness of the administration and the government’s ability to opt for stricter and more hierarchical forms of intervention.

Here is more from Christian Aschenbrenner, Christoph Knill, and Yves Steinebach.  Via the excellent Kevin Lewis.

Peter Isztin on the incidence of AI

by  Tyler Cowen in

This short paper considers the effects of artificial intelligence (AI) tools on the division of labor across tasks. Following Becker and Murphy (1992) I posit a “team” with each team member being assigned to a task. The division of labor (that is, the number of specialized tasks) is here limited not only by the extent of the market, but by coordination costs. Coordination costs stem from the need for knowledge in multiple tasks, as well as from monitoring and punishing shirking and other malfeasance. The introduction of AI in this model helps the coordination of the team and fully or partially substitute for human “generalist” knowledge. This in turn can make specialization wider, resulting in a greater number of specialized fields. The introduction of AI technologies also increases the return to fully general knowledge (i.e.education).

Not a certainty, but definitely worth a ponder.  Here is the draft. 

My excellent Conversation with David Bentley Hart

by  Tyler Cowen in

Here is the audio, video, and transcript.  Here is the episode summary:

David Bentley Hart is an American writer, philosopher, religious scholar, critic, and theologian who has authored over 1,000 essays and 19 books, including a very well-known translation of the New Testament and several volumes of fiction.

In this conversation, Tyler and David discuss ways in which Orthodox Christianity is not so millenarian, how theological patience shapes the polities of Orthodox Christian nations, how Heidegger deepened his understanding of Christian Orthodoxy, who played left field for the Baltimore Orioles in 1970, the simplest way to explain how Orthodoxy diverges from Catholicism, the future of the American Orthodox Church, what he thinks of the Book of Mormon, whether theological arguments are ultimately based on reason or faith, what he makes of reincarnation and near-death experiences, gnosticism in movies and TV, why he dislikes Sarah Ruden’s translation of the New Testament, the most difficult word to translate, a tally of the 15+ languages he knows, what he’ll work on next, and more.

Hart is probably the best-read CWT guest of all time, with possible competition from Dana Gioia?  Excerpt:

COWEN: If you could explain to me, as simply as possible, in which ways is Orthodox Christianity not so very millenarian?

HART: Well, it depends on what you mean by millenarian. I’d have to ask you to be a bit more —

COWEN: Say the Protestant 17th-century sense that the world is on the verge of a very radical transformation that will herald in some completely new age, and we all should be prepared for it.

HART: Well, in one sense, it’s been the case of Christianity from the first century that it’s always existed in a time between times. There’s always this sense of being in history but always expecting an imminent interruption of history.

But Orthodoxy has been around for a while. It’s part of an underrated culture, grounded originally in the Eastern Greco-Roman world, and has a huge apparatus of philosophy and theology and, I think, over the centuries has learned to be patient.

The Protestant millenarianism you speak of always seems to have been born out of historical crisis in a sense. The rise of the nation-state, the fragmentation of the Western Church — it’s always as much an effective history as a flight from history.

Whereas, I think it’s fair to say that Orthodoxy has created for itself a parallel world just outside the flow of history. It puts much more of an emphasis on the spiritual life, mysticism, that sort of thing. And as such, whereas it still uses the recognizable language of the imminent return of Christ, it’s not at the center of the spiritual life.

COWEN: How does that theological patience shape the polities of Orthodox Christian nations and regions? How does that matter?

HART: Well, it’s been both good and bad, to be honest. At its best, Orthodoxy has cultivated a spiritual life that nourished millions and that puts an emphasis upon moral obligation to others and the life of charity and the ascetical virtues of Christianity, the self-denial. At its worst, however, it’s often been an accommodation with historical forces that are antithetical to the gospel, too.

It’s often been the case that Orthodoxy has been so, let’s say, disenchanted with the millenarian expectation that it’s become a prop of the state, and you can see it today in Russia, in which you have a church institution. Now, this isn’t to speak of the faithful themselves, but the institutional authority of the state — of the institution, rather, of the church more or less being nothing but a propaganda wing of an authoritarian and terrorist government.

So, it’s had both its good and its bad consequences over the centuries. At its best, as I say, it encourages a true spiritual life that can teach one to be detached from ambitions and expectations and the violent projects of the ego. But at its worst, it can become a passive participant in precisely those sorts of projects and those sorts of evils.

Recommended, interesting throughout, and yes I do ask him about the Baltimore Orioles.

Toward a theory of military loyalties

by  Tyler Cowen in

A critical element in civil wars is military fragmentation. Yet, we have a limited understanding of why military elites fight in civil wars and on what side. In this article I develop a theory of the economic and professional motivations of military elites. I test this theory using the case of West Point graduates in the American Civil War. I argue that in addition to home state, economic and professional interests were a major influence on West Pointers. Graduates with connections to Southern cash crops were less likely to fight for the Union and more likely to fight for the Confederacy. Higher ranking graduates were more likely to fight for both sides, as they were better positioned to compete for promotion. I test this argument using a new dataset of more than 1000 West Point graduates’ wartime allegiances and antebellum careers and find strong evidence in support of my expectations.

That is from a newly published paper by Peter B. White.

Wednesday assorted links

by  Tyler Cowen

1. More Okie-Dokie.  And some RFK Okie-Dokie.

2. Paul Robinson podcast on Russian liberalism.  And Classical Liberalism by Country, Volume III, is now out.

3. Nabeel’s notes on puzzles.

4. Do “multiple selves” adapt more quickly?

5. Markets in everything: People are paying to break printers with sledgehammers in smash rooms.

6. Claude 2 is now out from Anthropic.  And NYT coverage of Anthropic.

Can the School Choice Movement Liberate Childhood?

by  Alex Tabarrok in

Richard Hanania has a very good post on the rapidly expanding school choice movement and his hopes for a radical rethinking of education.

The first thing to point out about public education is that it involves an extreme restriction of liberty beyond anything we usually accept. How common is it for government to force you to be in a certain place at a certain time? What I call “time-place” mandates are rare. Sometimes you have to go to the DMV, but even then you spend a short amount of time there, and can generally choose when to go. Sometimes people have to respond to subpoenas or jury duty, but those are uncommon events in most people’s lives. Government says to do your taxes, though you only have a deadline and can fill out the paperwork whenever and under whatever conditions you want.

The only substantial populations of individuals who have their lives structured according to time-place mandates in a free society like ours are prisoners, members of the military, and children. The mandates for children have gotten less strict over the years now that all states allow homeschooling, but opponents of school choice for all practical purposes want to do what they can to shape the incentive structures of parents so that they all use public schools (liberal reformers tend to like vouchers that can be used at charter schools, but not ESAs, which give parents complete control). Of course, children don’t have the freedom of adults, and so others are by default in control of how they spend most of their time. But it’s usually parents, not the government, that we trust in this role. Given the unusual degree to which public education infringes on individual liberty and family autonomy, the burden of proof has to be on those in favor of maintaining such an extreme institution.

…To me, the true promise of the school choice movement isn’t that it might simply save a bit of money or avoid the worst excesses of public education. Rather, it presents an opportunity to rethink childhood…On what basis did we as a society decide that the ideal way to spend a childhood was to attend government institutions 5 days a week, 7 hours a day, 9 months a year, for 12 years? That most of that time should be spent sitting at a desk, with say one hour for lunch and one for recess?

My hope is that states with universal ESAs will see radical experimentation. Maybe some parents would send their kids to a traditional school for six months of the year, and then have them apprenticing or interning in the workforce the rest of the time. Imagine having a few months experience working at a law firm during eighth grade, grabbing coffee for corporate executives in ninth grade, following around a pipe fitter in tenth grade, and helping around a gym in eleventh grade.

I too would like to see radical experimentation in education but I’m struck by how conservative and homogeneous schools are, regardless of their public or private status. Private schools, despite having the autonomy, have not pioneered novel teaching methods. Montessori was innovative but that was a hundred years ago. A few private schools have adopted Direct Instruction, but how many offer lessons in memory palaces, mental arithmetic or increasing creativity?

I am enthusiastic about developments coming out of Elon Musk’s school and Minerva but it’s still remarkable how similar almost all private schools are to almost all public schools. The global adoption of a nearly identical education model is also disturbing, as I harbor significant skepticism that we’ve reached an optimum. I see this as more of an outcome of world-elite consensus, similar to what we saw with COVID policy, with basically only Sweden bucking the trend and coming under intense pressure for doing so.

Online education and AI ought to greatly expand the potential range of experimentation but the demand for experimentation appears to be low.

Hanania has more of interest to say. Read the whole thing.