Competing Elite-Selection Mechanisms

In this issue:

• Competing Elite-Selection Mechanisms—It's convenient to identify rising stars early in order to make sure they get some of the world's limited opportunities to show their stuff. But this can very easily lead to broken selection systems. It works best when there's a mix of formal and informal selection mechanisms, and when elite status is not on autopilot, but partly consists of having good judgment about which status ladder to climb.
• Left-Tail Outcomes—An update to the Diff theory that the future of terrorism and political assassinations is much weirder.
• Downgrades and Giffen Goods—A world where the US is less creditworthy is a world where everywhere else is a bit more risky.
• Bearer Instruments—A permissionless, decentralized asset is much easier to steal.
• Currency—US borrowers are increasingly turning to Europe.
• Shipping—Lower global shipping volumes can be great for shippers if they take the form of a start-stop market where there's rapid cycling between shortage and glut.

The Diff May 19th 2025 13:14 • Click to play audio

Competing Elite-Selection Mechanisms

A few weeks ago, The Diff briefly reviewed the book The Chosen, a history of elite college admissions in the twentieth century. Realistically, elite status is a continuum, or actually many continua based on different status ladders. But in practice, many societies find that it's helpful to have some kind of boolean value that designates some of us as being worthy of special responsibilities. This isn't a complaint by any means: it's actually a clever kind of information compression, because it's saving lots of collective effort to triage based on surface-level credentials. In fact, this practice of granting fairly young people long-term membership in a special class based entirely on their potential is so useful that other institutions do it, too.^[1]

Google and Meta have Associate Product Manager and Rotational Product Manager programs, both of which are designed to be a fast learning experience on both sides: unusually early responsibility for the candidates, and, for the employer, an early look at who's able to execute.^[2] Procter & Gamble's Assistant Brand Manager role performed a similar function decades earlier—at one point in the 70s, Jeff Immelt and Steve Ballmer shared an office while they had this title. They'd both end up being CEOs of companies that had the highest market cap in the world at some point during their tenure.

This is distinct from another more ad hoc approach at pressure-cooker professional services companies like investment banks and consulting firms, where someone can get the equivalent of a temporary promotion by being put in charge of some time-sensitive and important part of a project (e.g. building and owning the financial model for a multi-billion dollar sell-side M&A transaction). That's a status that can be revoked at any time, and can easily be a way to shift blame when things aren't going well. Those businesses are structured around high attrition in the first few years, so it doesn't make sense for them to invest a lot in human capital that they won't actually collect a return on (and it also makes them more willing to risk employee burnout during crunch times, because they'll be hiring new analysts soon anyway).

People running big organizations always face the problem that most of the people who will be in charge after they're gone will be hired at some point during their tenure, i.e. it's every CEO's job to at least consider whether their company's new hires are CEO material, and whether potential company lifers are actually getting identified early. If the business is designed around outcomes that can be measured in a short period of time—did we get the deal? Did we deliver the project?—they don't need as much formal structure. But at other companies, there are a few truly front-office people who directly contribute to revenue, and a huge number of people who indirectly contribute to them, and who are often pretty indispensable.

When CEOs are doing this, they're often looking at external markers of status and upward mobility, and there's a rich ecosystem for this. Y Combinator is the standout example, having nabbed some of the Ivies' franchise of finding the most promising young people and verifying that important people ought to trust them with responsibility or at least capital. There are companies whose imprimatur is valuable, though people leaving those companies also have to deal with the fraught question of why, exactly, they're leaving such a favorable environment.^[3]

Bringing in outside talent implicitly means treating the world outside the company as a better judge and incubator of talent than the company itself. Which could very well be true: the assessment and cultivation of talent can't be something that everyone's equally good at, and any time there's variance in skills, there's room for a market. So the buyside outsources the first cut of talent assessment to investment banks who hire analysts and teach them critical skills like Excel, PowerPoint, basic business knowledge, and strategies for dealing with chronic sleep deprivation. And those banks also outsource their talent triage to colleges, who in turn outsource some of it to the College Board, high schools, etc.

This actually turns out to be a pretty robust system, in part because it's an organic one: when a bank stops training good analysts, it stops being able to recruit them, and after a while the lack of a buyside alumni network means it's not really involved in major deals. Schools also rise and fall in prominence—CCNY produced nine Nobel Prize winners when it had strictly test-based admissions and the Ivies had discriminatory policies, but in 1970 the Harvard of the Proletariat switched to open admissions for high school graduates in New York City, which basically turned it into a brand-new organization that kept the building, brand name (but not brand equity), and staff (for a while) of its predecessor.

The part of this that's durable is the meta aspect: there's a constant shift in who's able to give a meaningful imprimatur to whom, and even a shift in how credentialed different career paths are, and it doesn't run in just one direction. It's a lot harder to become a doctor or lawyer than it was in the past—med school became a US requirement in the 1920s, and law school became a de facto requirement for lawyers over about half a century wrapping up in the 70s.^[4] Law and medicine are still very lucrative fields, with some tradeoffs, and they attract lots of people who are passionate about both. But they're not where people go when they're optimizing for either maximum impact (tech) or maximum profit (tech or finance, mostly depending on personal inclination, and increasingly these two fields bleed together in terms of exit opportunities). It's not a coincidence that the domains with fuzzier and more variable criteria for elite selection are also domains where there's a lot more excitement right now.

It's inevitable that we need some kind of rubric for deciding who's talented enough to be worth paying attention to and who isn't. But it's also inevitable that this kind of ranking will be gamed, since everyone who is up for inclusion wants to get it, and everyone who's been vetted already wants the set of new entrants to be as small as possible—doctors have a strong financial incentive for there not to be many more doctors, and even extending the amount of education required means that a given new doctor has fewer active doctor-years in which to dilute other doctors' economics. And, of course, public-spirited reasons to say that not everyone should be wielding scalpels and prescribing fun and/or dangerous substances to whoever they want.^[5] Fields can raise standards when they professionalize, and when it actually makes sense to tell a teenager what specific things they should do today in order to maximize their odds of having a particular job in their thirties. But that's also the point at which the variance is going down and the impact of any one participant is also diminishing. The curse of legibility in a career path or a status ladder is that you can have more confidence in the future if it's a lot like the present.

Disclosure: Long GOOGL, META.

Diff Jobs

Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:

• Ex-Citadel, D.E. Shaw team that’s building the intelligence layer for casualty risk transfer that turns claims and underwriting data into liquidity for the trillion-dollar insurance liability market is looking for a principal full-stack software engineer with 10+ years of end-to-end product development experience in a fast-moving environment. (Boston, NYC)
• YC-backed startup using AI to transform how companies quantify and optimize engineering productivity is hiring formidable full-stack and AI engineers. Experience with React + Typescript, Go, or Python on the ML side a plus. (SF)
• Well-funded, fast-moving team is looking for a full-stack engineer to help build the best AI powered video editor for marketers. Tackle advanced media pipelines, LLM applications, and more. TypeScript/React expertise required. (Austin, Remote)
• A Google Ventures-backed startup founded by SpaceX engineers that’s building data infrastructure and tooling for hardware companies is looking for a product manager with 3+ years experience building product at high-growth enterprise SaaS businesses. Technical background preferred. (LA, Hybrid)
• An OpenAI backed startup that’s applying advanced reasoning techniques to reinvent investment analysis from first principles and build the IDE for financial research is looking for software engineers and a fundamental analyst. Experience at a Tiger Cub a plus. (NYC)Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.

Elsewhere

Left-Tail Outcomes

The Diff recently had a grim piece about how future political assassins will increasingly be motivated by very niche ideologies, because it takes some level of planning and effort to carry out an attack, and there's a correlation between being functional enough to do this and being aware that it's not an especially effective way to advance one's agenda. But the weirder your belief system, the more likely you are to be both competent enough to act but weird enough to think one-off disorganized violence will actually help your agenda.

So, a suicide bomber attacked an IVF clinic on Saturday. Plenty of ways to pattern-match that one to extremist actions on behalf of fairly mainstream policies, but no: the perpetrator turns out to be a "promortalist," i.e. someone who believes that the suffering of existence is severe enough that its moral to kill. (His manifesto says that this entails painless killing, but that sounds like a question of the interest rate on utils and the actuarial math of suffering, not something core to the argument.)

This belief system is at least nominally descended from antinatalism, which is a bit less niche-y: The New Yorker recently recirculated their profile of antinatalist philosopher David Benatar. (For what it's worth, I read Better Never to Have Been years ago and proceeded to continue existing and bring a few more people into existence. It's a fun reductio of utilitarianism and/or the non-aggression principle, but more the kind of reductio that leads you to weaken or discard the original idea.) But it's a weird one, and one that most antinatalists would strongly reject. Since it's a fringe variant of an already non-mainstream idea, there was basically no chance that someone looking at the headline could have any idea what the bomber's deal was. Expect that to get more common.

Downgrades and Giffen Goods

The US has lost its last AAA bond rating. The US is a weird special case when it compares to creditworthiness, because the US's reserve currency status makes it similar to a Giffen Good. For a Giffen Good, when the price goes up, consumption increases, through the specific mechanism that people can't afford alternatives (so, in a country where rice is the staple crop and where people spend most of their money on food, an increase in the price of rice can mean that they can't afford pork, and eat more rice to compensate). For the US, the analogous feature is that when there's some macro factor that threatens the US government's ability to service its financial obligations, that's such catastrophically bad news for the rest of the world that risk assets get sold off, and treasuries are the default thing people buy (if you sell an asset, just mechanically, you're often buying dollars and, also often, automatically sweeping them into some kind of account that is ultimately buying either t-bills or low-risk assets benchmarked to them). As of this writing, long bond yields are up a bit, but shorter-term ones are mixed and close to flat, so it's not showing the Giffen pattern just yet.

A broader question is: how high should the US's credit rating be? The number of AAA-rated companies has declined over time—Microsoft and Johnson & Johnson are the two survivors today, compared to around 60 in 1980, 15 in 2000, and four as of the publication date of this article, which happens to be about the last big news cycle regarding US credit ratings. One stylized fact about the world is that an aging US population has demand for extremely safe dollar-denominated assets, and an aging global population has disproportionate demand for dollars even if their consumption is priced in some other currency. For the US government to produce enough bonds to satisfy that demand probably requires a fiscal policy that isn't compatible with rating agencies treating the US's credit rating as perfect with a straight face. But it can be perfect-enough to be a universal denominator even if it's not quite perfect—and, in fact, it gets more universal the more bonds the government issues.

Bearer Instruments

There was a brief period where a pretty easy way to tell a gripping fictional crime story was:

1. Find anything good from back when bearer bonds were a meaningful financial instrument.
2. Update the story to replace the bonds with a USB stick that has a crypto private key.

In this case, reality is a bit ahead of fiction, as the crypto world has dealt with a spate of kidnappings meant to get at people's crypto stashes ($, WSJ). This is in some ways very much in keeping with the ideology of crypto: if you don't rely on trusted third parties, you can't reverse transactions, even if those transactions were made under duress. It turns out that one thing this does is that it makes crypto a tool for converting duress into enough money never to work again, albeit with a looming legal liability. This situation gives crypto people an incentive to adopt things like multi-signature wallets that make it impossible for just one person to give up the funds. Which is also a triumph for crypto: if you must, for practical reasons, trust some sort of third party, you should ideally be able to choose exactly which one you trust, and with what.

Currency

More American companies are issuing euro-denominated bonds ($, FT). There are two ways to read this:

1. They're locking in funding that matches some of their other costs and revenues, which is a good way to mitigate currency risk. And they're doing it at a time when rates are lower in Europe than they are in the US. Or
2. It's an early shift in the global currency system's center of gravity, with more investors choosing to think in terms of currencies other than the dollar.

It's probably mostly the first point, but worth watching whether a higher supply of American-issued, euro-denominated debt doesn't push rates closer to parity, and whether this kind of issuance continues at a similar pace if the rate gap shrinks.

Shipping

You would think that the sector most exposed to tariff-driven downside would be the one that ships goods from place to place, but the big container shipping companies are doing just fine. Shipping is a sector where demand is more volatile than supply, so prices can swing between a) whatever low price is just high enough that it isn't worth scrapping part of the fleet, or b) high, but just low enough that factories and retailers aren't shutting down. So higher volatility in the volume of goods shipped can be net good for the shipping companies, even if it's accompanied by a lower absolute volume of shipping. The ships are not quite a call option, since they're usually in the money and pricing is typically somewhere between the two extremes outlined above. But when they're volatile, it's faster to cut costs and reduce capacity than it is to add it, so they tend to benefit.