Many policies work for individuals but fail when scaled to entire populations. The core insight: what works for anyone does not necessarily work for everyone.

---

Government deposit insurance “ensures” that our bank deposits are safe. Anyone’s deposits are insured, but not everyone’s.

If your bank (and only your bank) goes bust, you will get your money back. This may be true even if your bank is quite large. However, if all the banks go down, your deposits, too, will be gone.

Why? No insurance fund in the world can cover all bank deposits simultaneously. Correlated risks are not covered. There is no “you will not get your money if every bank collapses” in the fine print. And yet it is plainly the case.

---

Index funds allow small investors to invest in a stock index such as the S&P 500. Beautiful, isn’t it? Well, it sure is beautiful if you are the only person on the planet who owns a portion of the S&P 500.

However, we simply do not know the effects of large portions of the market being in ETFs. Financial professionals have assured me that there is still enough price discovery. And the incentives are still correct: you can go against the grain. But there is much evidence that inclusion of a stock in the S&P 500 causally uplifts its price. ETFs are not necessarily price takers.

If investors realize they cannot beat the market by active investment, they will turn to ETFs and similar instruments. It simply cannot be good for price discovery if hundreds of millions of people mindlessly yeet hundreds of dollars a pop a month into index funds. Returns must inevitably decline. The stock market can make anyone a winner. Just not everyone.

---

During the 20th century, the American government recognized that college education makes people better off, spiritually and financially. This recognition was based on evidence from a time when few people went to college.

And thus it was decided that “more people have to go to college.” Student loans were heavily underwritten by the government, veterans received subsidized university education, and many jobs now explicitly or implicitly required a college degree.

What policymakers did not understand was that anyone can become better off from college, just not everyone. Empirical research agrees that on the individual level, a college education (still?) pays for the individual student. But on a social level, the evidence may well go to the contrary. As soon as education was “scaled up,” the benefits to society from an individual student’s journey tapered off. Credential inflation and rising costs were not foreseen.

The methodological individualism of economics leads us to obtain marginal, causal effects for some “average” individual. But from that you cannot conclude that the effect stays constant once the policy is scaled up.

---

Antibiotics are another modern miracle. If you get a bacterial infection, antibiotics can save your life (as they have saved mine at least two times). What’s not to like?

But if everyone takes antibiotics for every minor sniffle, bacteria evolve resistance. The drugs stop working. Doctors have known this for decades, yet overprescription continues. The cost, accelerating resistance, is diffuse and delayed.

Anyone can overuse antibiotics. Just not everyone.

---

What, then, does scale? Markets. Policymakers are unable or unwilling to understand the distinction between “anyone” and “everyone.” Fortunately, we can focus on policies that do scale.

Markets have an unparalleled track record of alleviating poverty and other awesome stuff. Is it not curious: on an individual level, anyone is made better off by mutually beneficial exchange. And everyone is made better off, too, if they join along!

As economists have recently pointed out, scaling up even from the most rigorous experimental evidence is wicked hard. Somehow, interventions appear to stop working. The issues come in various shapes and sizes, but what underlies them all is this distinction. We simply cannot know the general equilibrium effects of our policy proposals before we try them, and partial equilibrium evidence is not sufficient.

So here is a heuristic: be skeptical of policies that attempt to give everyone what currently works for anyone. And be more optimistic about expanding the scope of voluntary exchange, the one scaling success story we have.