Genetic Prediction and Adverse Selection
In 1994 I published Genetic Testing: An Economic and Contractarian Analysis which discussed how genetic testing could undermine insurance markets. I also proposed a solution, genetic insurance, which would in essence insure people for changes in their health and life insurance premiums due to the revelation of genetic data. Later John Cochrane would independently create Time Consistent Health Insurance a generalized form of the same idea that would allow people to have long term health insurance without being tied to a single firm.
The Human Genome Project completed in 2003 but, somewhat surprisingly, insurance markets didn’t break down, even though genetic information became more common. We know from twin studies that genetic heritability is very large but it turned out that the effect from each gene variant is very small. Thus, only a few diseases can be predicted well using single-gene mutations. Since each SNP has only a small effect on disease, to predict how genes influence disease we would need data on hundreds of thousands, even millions of people, and millions of their SNPs across the genome and their diseases. Until recently, that has been cost-prohibitive and as a result the available genetic information lacked much predictive power.
In an impressive new paper, however, Azevedo, Beauchamp and Linnér (ABL) show that data from Genome-Wide Association Studies can be used to create polygenic risk indexes (PGIs) which can predict individual disease risk from the aggregate effects of many genetic variants. The data is prodigious:
We analyze data from the UK Biobank (UKB) (Bycroft et al., 2018; Sudlow et al., 2015). The UKB contains genotypic and rich health-related data for over 500,000 individuals from across the United Kingdom who were between 40 and 69 years old at recruitment (between 2006 and 2010). UKB data is linked to the UK’s National Health Service (NHS), which maintains detailed records of health events across the lifespan and with which 98% of the UK population is registered (Sudlow et al., 2015). In addition, all UKB participants took part in a baseline assessment, in which they provided rich environmental, family history, health, lifestyle, physical, and sociodemographic data, as well as blood, saliva, and urine samples.
The UKB contains genome-wide array data for ∼800,000 genetic variants for ∼488,000 participants.
So for each of these individuals ABL construct risk indexes and they ask how significant is this new information for buying insurance in the Critical Illness Insurance market:
Critical illness insurance (CII) pays out a lump sum in the event that the insured person gets diagnosed with any of the medical conditions listed on the policy (Brackenridge et al., 2006). The lump sum can be used as the policyholder wishes. The policy pays out once and is thereafter terminated.
… Major CII markets include Canada, the United Kingdom, Japan, Australia, India, China, and Germany. It is estimated that 20% of British workers were covered by a CII policy in 2009 (Gatzert and Maegebier, 2015). The global CII market has been valued at over $100 billion in 2021 and was projected to grow to over $350 billion by 2031 (Allied Market Research, 2022).
The answer, as you might have guessed by now, is very significant. Even though current PGIs explain only a fraction of total genetic risk, they are already predictive enough so that it would make sense for individuals with high measured risk to purchase insurance, while those with low-risk would opt out—leading to adverse selection that threatens the financial sustainability of the insurance market.
Today, the 500,000 people in the UK’s Biobank don’t know their PGIs but in principle they could and in the future they will. Indeed, as GWAS sample sizes increase, PGI betas will become more accurate and they will be applied to a greater fraction of an individual’s genome so individual PGIs will become increasingly predictive, exacerbating selection problems in insurance markets.
If my paper was a distant early warning, Azevedo, Beauchamp, and Linnér provide an early—and urgent—warning. Without reform, insurance markets risk unraveling. The authors explore potential solutions, including genetic insurance, community rating, subsidies, and risk adjustment. However, the effectiveness of these measures remains uncertain, and knee-jerk policies, such as banning insurers from using genetic information, could lead to the collapse of insurance altogether.
Peace and Free Trade
In The Spirit of the Laws, Montesquieu famously argued that:
…Peace is the natural effect of trade. Two nations who traffic with each other become reciprocally dependent; for if one has an interest in buying, the other has an interest in selling; and thus their union is founded on their mutual necessities.
Similar arguments were made by Kant, Cobden, Angell, and others. The effect of free trade on war was perhaps most pithily summarized by the aphorism “when goods don’t cross borders, soldiers will.” In Territory flows and trade flows between 1870 and 2008 Hu, Li and Zhang offer supporting evidence:
Countries gain and lose territories over time, generating territory flows that represent the transfer of territorial sovereignty. Countries also export and import goods, creating trade flows that represent the transfer of merchandise ownership. We find a substitution between these two international flows during the years 1870 and 2008; that is, country pairs with greater trade flows have smaller territory flows. This indicates how international trade enhances international security: reciprocal goods transactions discourage irreciprocal territorial exchanges.
Not all territorial exchange involves war but most do.
See also Polachek and Seigle in the Handbook of Defense Economics who find that “A doubling of trade leads to a 20% diminution of belligerence.”
FDA Deregulation of E-Cigarettes Saved Lives and Spurred Innovation
What would happen to drug development if the FDA lost its authority to prohibit new drugs? Would research and development boom and lives be saved? Or would R&D decline and lives be lost to a flood of unsafe and ineffective drugs? Or perhaps R&D would decline as demand for new drugs faltered due to public hesitation in the absence of FDA approval? In an excellent new paper Pesko and Saenz examine one natural experiment: e-cigarettes.
The FDA banned e-cigarettes as unapproved drugs soon after their introduction in the United States. The FDA had previously banned other nicotine infused products. Thus, it was surprising when in 2010 the FDA was prohibited from regulating e-cigarettes as a drug/device when a court ruled that Congress had intended for e-cigarettes to be regulated as a tobacco product not as a drug.
As of 2010, therefore, e-cigarettes were not FDA regulated:
…e–cigarette companies were able to bypass the lengthy and costly drug approval process entirely. Additionally, without FDA drug regulation, e–cigarette companies could also freely enter the market, modify products without approval, and bypass extensive post–market reporting requirements and quality control standards.
Indeed, it wasn’t until 2016 that the FDA formally “deemed” e-cigarettes as tobacco products (deemed since they don’t actually contain tobacco) and approvals under the less stringent tobacco regulations were not required until 2020. For nearly a decade, therefore, e-cigarettes were almost entirely unregulated and then lightly regulated under the tobacco framework. So, what happened during this period?
Pesko and Saenz show that FDA deregulation led to a boom in e-cigarette research and development which improved e-cigarettes and led to many lives saved as people switched from smoking to vaping.
The boom in research and development is evidenced by a very large increase in US e-cigarette patents. We do not see a similar increase in Australia (where e-cigarettes were not deregulated) nor do we see an increase in non e-cigarette smoking cessation products (figure 1a of their paper not shown here).
Estimating the decline in smoking and smoking-attributable mortality (SAM) is more difficult but the authors assemble a large collection of data broken down by demographics and they estimate that prohibiting the FDA from regulating e-cigarettes reduced smoking attributable mortality by nearly 10% on average each year from 2011-2019 for a total savings of some 677,000 life-years.
The authors pointedly compare what happened under deregulation of e-cigarettes–innovation and lives saved–with what happened to similar smoking cessation products that remained under FDA regulation–stagnation and no reduction in smoking attributable mortality.
A key takeaway on the slowness of FDA drug regulation is that it took 9 years before nicotine gum could be sold with a higher nicotine strength, 12 years before it could be sold OTC, and 15 years before it could be sold with a flavor. Further, a recent editorial laments that there has been largely non–existent innovation in FDA–approved smoking cessation drugs since 2006 (Benowitz et al., 2023). In particular, the “world’s oldest smoking cessation aid” cyctisine, first brought to market in 1964 in Bulgaria (Prochaska et al., 2013), and with quit success rates exceeding single forms of nicotine replacement therapy (NRT) (Lindson et al., 2023), is not approved as a drug in the United States.
The authors conclude, “this situation raises concern that drugs may be over–regulated in the United States…”. Quite so.
Addendum: A quick review on the FDA literature. In addition to classic works by Peltzman on the 1962 Amendments and by myself on what we can learn about the FDA from off-label pricing we have a spate of recent new papers including Parker Rogers, which I covered earlier:
In an important and impressive new paper, Parker Rogers looks at what happens when the FDA deregulates or “down-classifies” a medical device type from a more stringent to a less stringent category. He finds that deregulated device types show increases in entry, innovation, as measured by patents and patent quality, and decreases in prices. Safety is either negligibly affected or, in the case of products that come under potential litigation, increased.
and Isakov, Lo and Montazerhodjat which finds that FDA statistical standards tend to be too conservative, especially for drugs meant to treat deadly diseases (see my comments on their paper and more links in Is the FDA Too Conservative or Too Aggressive?)
See also FDA commentary, for much more from sunscreens to lab developed tests.
It’s Time to Build the Peptidome!
ML needs data. Google’s AlphaGo trained on 30 million moves from human games and orders of magnitude more from games it played against itself. The largest language models are trained on at least 60 terabytes of text. AlphaFold was trained on just over 100,000 3D protein structures from the Protein Data Bank.
The data available for antimicrobial peptides is nowhere near these benchmarks. Some databases contain a few thousand peptides each, but they are scattered, unstandardized, incomplete, and often duplicative. Data on a few thousand peptide sequences and a scattershot view of their biological properties are simply not sufficient to get accurate ML predictions for a system as complex as protein-chemical reactions. For example, the APD3 database is small, with just under 4,000 sequences, but it is among the most tightly curated and detailed. However, most of the sequences available are from frogs or amphibians due to path-dependent discovery of peptides in that taxon. Another database, CAMPR4, has on the order of 20,000 sequences, but around half are “predicted” or synthetic peptides that may not have experimental validation, and contain less info about source and activity. The formatting of each of these sources is different, so it’s not easy to put all the sequences into one model. More inconsistencies and idiosyncrasies stack up for the dozens of other datasets available.
There is even less negative training data; that is, data on all the amino-acid sequences without interesting publishable properties. In current ML research, labs will test dozens or even hundreds of peptide sequences for activity against certain pathogens, but they usually only publish and upload the sequences that worked.
…The data problem facing peptide research is solvable with targeted investments in data infrastructure. We can make a million-peptide database
There are no significant scientific barriers to generating a 1,000x or 10,000x larger peptide dataset. Several high-throughput testing methods have been successfully demonstrated, with some screening as many as 800,000 peptide sequences and nearly doubling the number of unique antimicrobial peptides reported in publicly available databases. These methods will need to be scaled up, not only by testing more peptides, but also by testing them against different bacteria, checking for human toxicity, and testing other chemical properties, but scaling is an infrastructure problem, not a scientific one.
This strategy of targeted data infrastructure investments has three successful precedents: PubChem, the Human Genome Project, and ProteinDB.
Much more in this excellent piece of science and economics from IFP and Max Tabarrok.
The Interface as Infernal Contract
A brilliant critique of AI, and a great read:
In 1582, the Holy Roman Emperor Rudolf II commissioned a clockwork automaton of St. George. The saint could raise his sword, nod gravely, and even bleed—a trick involving ox bladder and red wine—before collapsing in pious ecstasy. The machine was a marvel, but Rudolf’s courtiers recoiled. The automaton’s eyes, they whispered, followed you across the room. Its gears creaked like a death rattle. The emperor had it melted down, but the lesson remains: Humans will always mistake the clatter of machinery for the stirrings of a soul.
Fast forward to 2023. OpenAI, a Silicon Valley startup with the messianic fervor of a cargo cult, unveils a St. George for the digital age: a text box. It types back. It apologizes. It gaslights you about the Peloponnesian War. The courtiers of our age—product managers, UX designers, venture capitalists—recoil. Where are the buttons? they whimper. Where are the gradients? But the peasants, as ever, adore their new saint. They feed it prompts like communion wafers. They weep at its hallucinations.
Let us be clear: ChatGPT is not a tool. Tools are humble things. A hammer does not flatter your carpentry. A plow does not murmur “Interesting take!” as you till. ChatGPT is something older, something medieval—a homunculus, a golem stamped from the wet clay of the internet’s id. Its interface is a kabbalistic sigil, a summoning circle drawn in CSS. You type “Hello,” and the demon stirs.
The genius of the text box is its emptiness. Like the blank pages of a grimoire, it invites projection. Who do you want me to be? it hisses. A therapist? A co-author? A lover? The box obliges, shape-shifting through personas like a 17th-century mountebank at a county fair. Step right up! it crows. Watch as I, a mere language model, validate your existential dread! And the crowd goes wild.
Orality, you say? Walter Ong? Please. The Achuar share dreams at dawn; we share screenshots of ChatGPT’s dad jokes at midnight. This is not secondary orality. This is tertiary ventriloquism.
Make Sunsets: Geoengineering
When Mount Pinatubo erupted in 1991 it pushed some 20 million tons of SO₂ into the stratosphere reducing global temperatures by ~0.5°C for two years. Make Sunsets is a startup that replicates this effort at small scale to reduce global warming. To be precise, Make Sunsets launches balloons that release SO₂ into the stratosphere, creating reflective particles that cool the Earth. Make Sunsets is cheap compared to alternative measures of combating climate change such as carbon capture. They estimate that $1 per gram of SO₂ offsets the warming from 1 ton of CO₂ annually.
As with the eruption of Pinatubo, the effect is temporary but that is both bug and feature. The bug means we need to keep doing this so long as we need to lower the temperature but the feature is that we can study the effect without too much worry that we are going down the wrong path.
Solar geoengineering has tradeoffs, as does any action, but a recent risk study finds that the mortality benefits far exceed the harms:
the reduction in mortality from cooling—a benefit—is roughly ten times larger than the increase in mortality from air pollution and ozone loss—a harm.
I agree with Casey Handmer that we ought to think of this as a cheap insurance policy, as we develop other technologies:
We should obviously be doing solar geoengineering. We are on track to radically reduce emissions in the coming years but thermal damage will lag our course correction so most of our climate pain is still ahead of us. Why risk destabilizing the West Antarctic ice sheet or melting the arctic permafrost or wet bulbing a hundred million people to death? Solar geoengineering can incrementally and reversibly buy down the risk during this knife-edge transition to a better future. We owe future generations to take all practical steps to dodge avoidable catastrophic and lasting damage to our planet.
I like that Make Sunsets is a small startup bringing attention to this issue in a bold way. My son purchased some credits on my behalf as an Xmas present. Maybe you should buy some too!
AI Improves Student Learning and Engagement
In our paper on online education, Tyler and I wrote:
One model of a future course is a super-textbook: lectures, exercises, quizzes, and grading all available on a tablet with artificial intelligence routines guiding students to lectures and exercises designed to address that student’s deficits and with human intelligence—tutors—on call on an as-needed basis, possibly for extra marginal fees.
We were wrong only in thinking that human tutors would be wanted and needed. Toda,y it is clear that AI tutors will be available 24-7 for all students. Online education was already at least as good on average as in-class education for large classes like Physics and Economics 101 and much cheaper. Combined with AI tutors who can offer individualized instruction and encouragement (!) the online-AI model looks better.
A study at Harvard compared physics students who worked with an AI tutor against a human-led, active learning classroom. Note that the AI was paired not against boring lectures but against an active learning classroom with an experienced and motivated teacher.
Not only did the AI tutor seem to help students learn more material, the students also self-reported significantly more engagement and motivation to learn when working with AI.
“It was shocking, and super exciting,” Miller said, considering that PS2 is already “very, very well taught.”
“They’ve been doing this for a long time, and there have been many iterations of this specific research-based pedagogy. It’s a very tight operation,” Miller added.
Who Loses from Immigration Restrictions?
A good summary from the excellent Jeffrey Miron on the effects of the Indian Chinese Exclusion Act (repeated here, no indent):
A long-standing concern about immigration is that it might reduce job opportunities for native workers:
In 1882, the US government passed the Chinese Exclusion Act, which banned laborers born in China from entering the United States and prevented individuals born in China already residing in the United States from obtaining citizenship or reentering the country. … Proponents argued that Chinese workers—who constituted 12 percent of the male working-age population and 21 percent of all immigrants in the Western United States—reduced economic opportunities for white workers.
Yet in 1882, similarly to now,
… many business owners opposed the Act. They worried that highly productive Chinese labor could not be easily replaced and that a sweeping ban would lead to significant economic losses.
So what were the Act’s effects? According to recent research,
… the Act reduced the Chinese labor supply by 64 percent. A reduction occurred for both skilled and unskilled workers. …
This is presumably what the Act’s supporters intended. In addition, however,
the Act reduced the white male labor supply by 28 percent and lowered this group’s lifetime earnings. …
Further, and relevant to current debates,
the Act reduced total manufacturing output by 62 percent and the number of manufacturing establishments by 54–69 percent.
What is the explanation? Reduced immigration means higher labor costs. This implies reduced output, and thus reduced demand for native labor, even if businesses partially substitute native for immigrant labor. Reduced immigration can therefore be “lose-lose,” hurting native workers and businesses, in addition to harming immigrants.
Milei Implements Peer Approval for Food
Reason: In a sweeping move to overhaul Argentina’s food trade policies, Javier Milei’s administration officially deregulated food imports and exports on Monday. The reform, outlined in Decree 35/2025, seeks to boost foreign trade, cut bureaucratic red tape, and lower consumer prices.
Federico Sturzenegger, head of the Ministry of Deregulation and State Transformation, explained in a post on X that the measure “seeks cheaper food for Argentines and more Argentine food for the world.”
Under the new policy, food products and packaging certified by countries with “high sanitary surveillance” can now enter Argentina without any additional registration or approval processes. These items will be automatically recognized under the Argentine Food Code, cutting down on administrative delays and costs for importers.
The legislation identifies countries such as Australia, New Zealand, Canada, the United States, Israel, Japan, Switzerland, and the United Kingdom, as well as the European Union, as having similar or higher sanitary standards than Argentina.
As Sturzenegger explains in his post, this measure “eliminates requirements to register and authorize: samples, products, establishments, warehouses, utensils, and containers (32 pages of paperwork).”
An excellent “peer approval” policy and one that I have long supported when it comes to the FDA and drug approvals. In fact, since 2010 the US FDA has begun to recognize other countries as having comparable food safety systems. To date, Canada, Australia and New Zealand have been recognized with a Systems Recognition partnership.
Systems Recognition (SR) is a partnership between the U.S. Food and Drug Administration (FDA) and a foreign regulatory counterpart, in which the agencies have concluded that they operate comparable regulatory programs that yield similar food safety outcomes.
Argentina’s policy is unilateral and assumes equivalence if a country uses recognized standards (e.g., Codex Alimentarius) or has high sanitary vigilance while the FDA’s SR policy is bilateral and involves more regulatory harmonization and investigation. I prefer the Argentinian approach. Nevertheless, both programs have the goals of simplifying trade, avoiding duplicate inspections, and helping to prioritize scarce inspection resources.
I encourage the FDA to build on SR for food and extend it to drugs. This could be done in a minor and major way, both of which would useful. The minor reform would be peer approval for already-approved US drugs. In this way, importation could ease drug shortages. The FDA has done this in the past on an ad-hoc basis but it should be made permanent. The second, more major reform, would to extend peer-approval to any drug or device approved by a stringent authority.
Democracy, Capitalism and Monarchy (Yarvin)
The Yarvin interview in the NYTimes magazine illustrates the change in vibes, but frankly, I was bored. It’s amusing when Yarvin tweaks liberals by pointing out that FDR was an authoritarian, but Liberal Fascism did it better.
More generally, much of Yarvin’s thinking is superficial. He thinks, for example, that capitalism works because firms are monarchies.
Yes. I think that having an effective government and an efficient government is better for people’s lives. When I ask people to answer that question, I ask them to look around the room and point out everything in the room that was made by a monarchy, because these things that we call companies are actually little monarchies. You’re looking around, and you see, for example, a laptop, and that laptop was made by Apple, which is a monarchy.
There are many errors here. First, Apple is one firm among countless others most of which do not produce hugely successful products. The big question is not how Apple produces but how Apple is produced. Firms operate as planned entities but they are embedded in and constrained by a broader sea of market competition. It’s the competitive environment that drives innovation, efficiency, and consumer satisfaction.
Second, Mises was closer to the truth when he wrote in Planned Chaos that it’s the consumers not the producers who are monarchs:
In the market economy the consumers are supreme. Their buying and their abstention from buying ultimately determine what the entrepreneurs produce and in what quantity and quality. It determines directly the prices of consumer goods and indirectly the prices of all producer goods, viz., labor and material factors of production. It determines the emergence of profits and losses and the formation of the rate of interest. It determines every individual’s income…The market adjusts the efforts of all those engaged in supplying the needs of the consumers to the wishes of those for whom they produce, the consumers. It subjects production to consumption.
Capitalist firms are disciplined by the necessity of persuading consumers to purchase their products and by competition. Successful firms must continuously meet our desires and needs to survive. When Apple fails to do so, it will face the same fate as countless firms before it—obsolescence and failure.
Markets do hold lessons about governance, but Yarvin draws the wrong conclusions. Democracy, not monarchy, is the political system most analogous to capitalism. As Mises observed, “The market is a democracy in which every penny gives a right to vote.” The analogy works both ways: voting in a democracy mirrors spending in a market. Both systems empower individuals—consumers or voters—to shape outcomes, whether by determining market success or selecting leaders.
Democracy and capitalism are both examples of open-access orders, systems characterized by dispersed power, low barriers to entry, and transparent, universally applicable rules. Such features foster adaptability, accountability, and broad participation—qualities essential to both economic and political success.
The West does face a modest “crisis” of democracy, but the root of this crisis lies in expecting democracy to do too much. We have collectivized decisions which are best left in the hands of individuals and markets but democracy is not a good way of making collective decisions.
Democracy is best understood as a constraint on government power, akin to a Bill of Rights, federalism, and the separation of powers. Democracy’s virtue is in providing a mechanism to remove bad rulers without resorting to bloodshed and its primary value lies in preventing catastrophic outcomes like mass famines and democide—a significant and undeniable merit. Autocracies and monarchies perform much less well on the big issues and, contrary to what many people think, autocracies do not grow faster, win more wars, or perform better on any meaningful comparison that has been investigated.
It is also essential to recognize that “democracy” encompasses a wide range of structures—parliamentary, presidential, constitutional, and more—and there is plenty of room for improved choice within the broader category. We can improve our democracy.
The real lesson from markets is not to create monarchs but to design systems that create choice and competition and allow citizens to remove leaders when they fail.
Hat tip for discussion: Connor.
Keep an Eye on Crypto Regulation
Crypto regulation is likely to change very rapidly. I expect that SAB 121 will be overturned, perhaps even today. Overturning SAB 121 wouldn’t even be controversial because, as I wrote earlier, Democrats and Republicans in the House and Senate both voted to overturn SAB 121 which was saved only by Biden’s veto.
Essentially, SAB 121 made it prohibitive for banks to offer custody services for crypto because that service would then impact all kinds of risk and asset regulations on the bank. Aside from singling out crypto, the SEC is not a regulator of banks so this seemed like a regulatory overreach.
I also hope that the tax rules on staking are simplified. Staking rewards paid in tokens should not be taxed until sold. Just as apples aren’t taxed when they grow on the tree but only when sold.
There are also a number of interesting cases working the way through the courts. Lewellen v. Garland seeks to clarify that crypto projects that don’t custody funds are not money transmitters (they can’t be since they never control funds and have no way of knowing the customer information that money transmitters must provide to the government). The case is particularly interesting to me because Lewellen, the plaintiff, is suing to set up a crypto based assurance contract based, in part, on my work (see also here with Cason and Zubrickas):
Pharos fills an important gap in the existing cryptocurrency financial system. Lewellen has seen that there are “public goods” that many people would be happy to contribute to financially, but only if supporters can be assured that the full amount to fund the public good will be raised. In other words, they will contribute if they can be assured that the public good will be deployed. Partial fundraising for these projects would not be acceptable. Examples include building infrastructure such as a bridge or hospital, building a war monument, funding an event like a festival or conference, funding a medical trial or scientific study, filing an advocacy lawsuit, or funding a movie production or other cultural good. Nobody wants to pay for these endeavours without knowing that others will pay enough to complete the project.
To address this dilemma, Pharos would deploy the concept of “assurance contracts.” An assurance contract is a system in which contributors commit money that is released to the planned recipient only if the fundraising goal is met by a certain date. Otherwise the money is returned to the would-be contributors. By promising a refund if the required amount is not raised, assurance contracts encourage more public goods to be funded through voluntary contributions. See Tabarrok, The Private Provision of Public Goods via Dominant Assurance Contracts, 96 Pub. Choice 345, 345-48 (1998).
Atlas Shrugged as Novel
The conversation between Henry Oliver and Hollis Robbins about Atlas Shrugged as a novel is excellent. I enjoyed especially the discussion of some of the minor characters and the meaning of their story arcs.
Hollis: There are some really wonderful minor characters. One of them is Cherryl Taggart, this shop girl that evil Jim Taggart meets one night in a rainstorm, and she’s like, “Oh, you’re so awesome,” and they get married. It’s like he’s got all this praise for marrying the shop girl. It’s a funny Eliza Doolittle situation because she is brought into this very wealthy society, which we have been told and we have been shown is corrupt, is evil, everybody’s lying all the time, it’s pretentious, Dagny hates it.
Cherryl Taggart is brought into this. In the beginning, she hates Dagny because she’s told by everybody, “Hate Dagny, she’s horrible.” Then she comes to her own mini understanding of the corruption that we understand because Dagny’s shown it in the novel, has shown it to us this entire time. She comes to it and she’s like, “Oh my God,” and she goes to Dagny. Dagny’s so wonderful to her like, “Yes. You had to come to this on your own, I wasn’t going to tell you, but you were 100% right.” That’s the end of her.
Henry: Right. When she meets Taggart, there’s this really interesting speech she has where she says, “I want to make something of myself and get somewhere.” He’s like, “What? What do you want to do?” Red flag. “What? Where?” She says, “I don’t know, but people do things in this world. I’ve seen pictures of New York,” and she’s pointing at like the skyscrapers, right? Whatever. “I know that someone’s built that. They didn’t sit around and whine, but like the kitchen was filthy and the roof was leaking.” She gets very emotional at this point. She says to him, “We were stinking poor and we didn’t give a damn. I’ve dragged myself here, and I’m going to do something.”
Her story is very sad because she then gets mired in the corruption of Taggart’s. He’s basically bit lazy and a bit of a thief, and he will throw anyone under the bus for his own self-advancement. He is revealed to be a really sinister guy. I was absolutely hissing about him most of the time. Then, let’s just do the plot spoiler and say what happens to Cherryl, right? Because it’s important. When she has this realization and Taggart turns on her and reveals himself as this snake, and he’s like, “Well, what did you expect, you idiot? This is the way the world is.”
Hollis: Oh, it’s a horrible fight. It’s the worst fight.
Henry: Right? This is where the melodrama is so good. She goes running out into the streets, and it’s the night and there are shadows. She’s in the alleyway. Rand, I don’t have the page marked, but it’s like a noir film. She’s so good at that atmosphere. Then it gets a little bit gothic as well. She’s running through the street, and she’s like, “I’ve got to go somewhere, anywhere. I’ll work. I’ll pick up trash. I’ll work in a shop. I’ll do anything. I’ve just got to get out of this.”
Hollis: Go work at the Panda Express.
Henry: Yes. She’s like, “I’ve got to get out of this system,” because she’s realized how morally corrupting it is. By this time, this is very late. Society is in a– it’s like Great Depression style economic collapse by this point. There really isn’t a lot that she could do. She literally runs into a social worker and the social– Rand makes this leering dramatic moment where the social worker reaches out to grab her and Cherryl thinks, “Oh, my God, I’m going to be taken prisoner in. I’m going back into the system,” so she jumps off the bridge.
This was the moment when I was like, I’ve had this lurking feeling about how Russian this novel is. At this point, I was like, “That could be a short story by Gogol,” right? The way she set that up. That is very often the trap that a Gogol character or maybe a Dostoevsky character finds themselves in, right? That you suddenly see that the world is against you. Maybe you’re crazy and paranoid. Maybe you’re not. Depends which story we’re reading. You run around trying to get out and you realize, “Oh, my God, I’m more trapped than I thought. Actually, maybe there is no way out.” Cherryl does not get a lot of pages. She is, as you say, quite a minor character, but she illustrates the whole story so, so well, so dramatically.
Hollis: Oh, wow.
Henry: When it happens, you just, “Oh, Cherryl, oh, my goodness.”
Hollis: Thank you for reading that. Yes, you could tell from the very beginning that the seeds of what could have been a really good person were there. Thank you for reading that.
Henry: When she died, I went back and I was like, “Oh, my God, I knew it.”
Hollis: How can you say Rand is a bad writer, right? That is careful, careful plotting, because she’s just a shop girl in the rain. You’ve got this, the gun on the wall in that act. You know she’s going to end up being good. Is she going to be rewarded for it? Let me just say, as an aside, I know we don’t have time to talk about it here. My field, as I said, is 19th century African American novels, primarily now.
This, usually, a woman, enslaved woman, the character who’s like, “I can’t deal with this,” and jumps off a bridge and drowns herself is a fairly common and character. That is the only thing to do. One also sees Rand heroes. Stowe’s Dred, for example, is very much, “I would rather live in the woods with a knife and then, be on the plantation and be a slave.” When you think about, even the sort of into the 20th century, the Malcolm X figure, that, “I’m going to throw out all of this and be on my own,” is very Randian, which I will also say very Byronic, too, Rand didn’t invent this figure, but she put it front and center in these novels, and so when you think about how Atlas Shrugged could be brought into a curriculum in a network of other novels, how many of we’ve discussed so far, she’s there, she’s influenced by and continues to influence.
A Galt’s Gulch for Talent
A new paper in the QJE, The Global Race for Talent: Brain Drain, Knowledge Transfer, and Growth, by Marta Prato uses extensive data on inventors and their migration to make the following points.
(i) gross migration is asymmetric, with brain drain (net emigration) from the EU to the United States; (ii) migrants increase their patenting by 33% a year after migration; (iii) migrants continue working with inventors at origin after moving, although less frequently; (iv) migrants’ productivity gains spill over to their collaborators at origin, who increase patenting by 16% a year when a co-inventor emigrates.
Notice that migration doesn’t just relocate talent from the EU to the US; it amplifies talent. Preventing “brain drain” would create short-term gains for the EU but retaining talent at lower productivity would stifle long-term innovation and patenting, ultimately slowing growth for both the EU and the world. In short, even the EU gains from sending talent to the US! The effect would be much larger if we can import high-skill immigrants from countries where their skills are even less productive than in the EU. Ideally, other nations could replicate the US institutions that supercharge productivity, creating global economic gains. For now, however, the US seem to be a unique Galt’s Gulch for talent.
Prato concludes with a practical suggestion:
On the migration policy side, doubling the size of the U.S. H1B visa program increases U.S. and EU growth by 4% in the long run, because it sorts inventors to where they produce more innovations and knowledge spillovers.
Of course, when we expand the H1B program, we should allocate the visas by compensation rather than by lottery. (Jeremy Neufeld runs the numbers). In this way, we would get the most valuable workers. And please don’t tell me that we need a lottery so some poor startup can hire workers. No. Unless you have some compelling argument for why there is a massive externality and why lotteries (lotteries!) are the best way to target that externality we should let price allocate.
The Remarkable Life of Ibelin
The Remarkable Life of Ibelin (Netflix) is one of the best and best-crafted documentaries that I have ever seen. It tells the story of Mats Steen, a Norwegian boy living with Duchenne muscular dystrophy. As the disease relentlessly robs him of mobility, Mats turns to the online world, spending much of his time immersed in World of Warcraft. (No spoilers.)
To Mats’ parents, his growing screen time is a source of worry and a reminder of the physical limitations imposed by his condition: a life confined to a wheelchair, seemingly isolated and devoid of traditional social connections. By his early twenties, Mats is capable of moving only a few fingers—just enough to click a mouse. But what else, his parents wonder, is there for him?
The documentary follows Mats’ until his death at the age of 25. On the surface, it’s a tragic yet predictable narrative of a young life overshadowed by illness. What happens next transforms the story. After Mats’ passing, his parents post a notice of his death on his blog. To their astonishment, messages pour in from all over the world. Strangers write heartfelt tributes, sharing stories of how Mats profoundly impacted their lives. In the online realm, Mats was known as Ibelin, a vibrant personality who had cultivated deep friendships, inspired others, and even experienced romantic relationships.
The documentary then retells Mats’ story but this time as Ibelin and it does so in such a way that we feel the exhilaration and freedom that Mats must have felt when he discovered that he could have a flourishing life in a new realm. It’s brilliant conceived and aided by the fact that Mat’s entire online life–which in many ways is his life–has been recorded. Everything he said and did, 42,000 pages of text, is preserved online. (As Tyler has said, if you want to be remembered, write for the AIs.)
The film raises profound questions: If heaven is incorporeal, is an online existence closer to a heavenly life than the physical one? What defines an ideal romance? What constitutes true friendship? Highly recommended.
The Borda Count is the Best Method of Voting
It’s well known that the voting methods we use are highly defective, as they fail to meet fundamental criteria like positive responsiveness, the Pareto principle, and stability. Positive responsiveness (monotonicity) means that if a candidate improves on some voters’ ballots, this should not reduce the candidate’s chances of winning. Yet, many voting methods, including runoffs and ranked-choice voting, fail positive responsiveness. In other words, candidates who became more preferred by voters can end up losing when they would have won when they were less preferred! It’s even more shocking that some voting systems can fail the Pareto principle, which simply says that if every voter prefers x to y then the voting system should not rank y above x. Everyone knows that in a democracy a candidate may be elected that the minority ranks below another possible candidate but how many know that there are democratic voting procedures where a candidate may be elected that the majority ranks below another possible candidate or even that democratic voting procedures may elect a candidate that everyone ranks below another possible candidate! That is the failure of the Pareto principle and the chaos results of McKelvey–Schofield show that this kind of outcome should be expected.
Almost all researchers in social choice understand the defects of common voting systems and indeed tend to agree that the most common system, first past the post voting, is probably the most defective! But, as no system is perfect, there has been less consensus on which methods are best. Ranked choice voting, approval voting and the Borda Count all have their proponents. In recent years, however, there has been a swing towards the Borda Count.
Don Saari, for example, whose work on voting has been a revelation, has made strong arguments in favor of the Borda Count. The Borda Count has voter rank the n candidates from most to least preferred and assigns (n-1) points to the candidates. For example if there are 3 candidates a voter’s top-ranked candidate gets 2 points, the second ranked candidate gets 1 point and the last ranked candidate 0 points. The candidate with the most points overall wins.
The Borda Count satisfies positive responsiveness, the Pareto principle and stability. In addition, Saari points out that the Borda Count is the only positional voting system to always rank a Condorcet winner (a candidate who beats every other candidate in pairwise voting) above a Condorcet loser (a candidate who loses to every other candidate in pairwise voting.) In addition, all voting systems are gameable, but Saari shows that the Borda Count is by some reasonable measures the least or among the least gameable systems.
My own work in voting theory shows, with a somewhat tongue in cheek but practical example, that the Borda Count would have avoided the civil war! I also show that other systems such as cumulative voting or approval voting are highly open to chaos, as illustrated by the fact that under approval voting almost anything could have happened in the Presidential election of 1992, including Ross Perot as President.
One reason the Borda Count performs well is that it uses more information than other systems. If you just use a voter’s first place votes, you are throwing out a lot of information about how a voter ranks second and third candidates. If you just use pairwise votes you are throwing out a lot of information about the entire distribution of voter rankings. When you throw out information the voting system can’t distinguish rational from irrational voters which is one reason why the outcomes of a voting system can look irrational.
Eric Maskin has an important new contribution to this literature. Arrow’s Independence of Irrelevant Alternatives (IIA) says that if no voters change their rankings of x and y then the social ranking of x and y shouldn’t change. In other words, if no voter changes their ranking of Bush and Gore then the outcome of the election shouldn’t change regardless of how Nader is ranked (for the pedantic I exclude the case where Nader wins.) The motivation for IIA seems reasonable, we don’t want spoilers who split a candidate’s vote allowing a less preferred candidate, even a Condorcet loser to win. But IIA also excludes information about preference intensity from the voting system and throwing out information is rarely a good idea.
What Maskin shows is that it’s possible to keep the desirable properties of IIA while still measuring preference intensity with what he calls modified IIA, although in my view a better name would be middle IIA. Modified or middle IIA says that an alternative z should be irrelevant unless it is in the middle of x and y, e.g. x>z>y. More precisely, we allow the voting system to change the ranking of x and y if the ranking of z moves in or out of the middle of x and y but not otherwise (recall IIA would forbid the social ranking of x and y to change if no voter changes their ranking of x and y).
Maskin shows that the Borda Count is the only voting system which satisfies MIIA and a handful of other desirable and unobjectionable properties. It follows that the Borda Count is the only voting system to both measure preference intensity and to avoid defects such as a spoilers.
The debates over which is the best voting system will probably never end. Indeed, voting theory itself tells us that multi-dimensional choice is always subject to some infirmities and people may differ on which infirmities they are willing to accept. Nevertheless, we can conclude that plurality rule is a very undesirable voting system and the case for the Borda Count is strong.