Big Pharma Is Paid to Treat the People Who Would Have Recovered Anyway

The argument I want to lay out I first met in Nassim Nicholas Taleb, and I have not been able to shake it. It does not require believing that anyone in the pharmaceutical industry is wicked. It requires only looking at the shape of the market and asking what a rational company does in it.

The Shape of the Market

Illness is not evenly distributed. At any moment a very large number of people are mildly unwell and a very small number are seriously ill. The mildly unwell have colds, aches, low moods, indigestion, trouble sleeping, minor infections. The seriously ill have the diseases that kill.

Now consider what it takes to treat each group.

The mildly unwell will, mostly, get better on their own. Their condition is self-limiting. This means that almost anything you give them will appear to work, because recovery was coming anyway and the pill was in the room when it arrived. Treating them is easy, the market is enormous, and every customer is a repeat customer, since they will be mildly unwell again next month.

The seriously ill are the opposite case. To cure them you have to find out what is actually wrong and fix it. That takes research that fails most of the time, at enormous cost, for a small number of patients, many of whom will not survive long enough to be repeat customers.

What a Rational Company Does

It goes where the money is. The money is in the large, easy, self-limiting group.

So the industry's effort tilts toward symptomatic treatment of the mildly ill: something for the ache, the mood, the sleep, the reflux. Products that are taken for years by people who are not very sick. That is not a failure of the industry. It is the industry doing exactly what its incentives instruct.

The seriously ill get what is left over, plus whatever a public research budget and a few determined people manage to produce.

Why This Is Worse Than Neglect

If the problem were merely that the sick are underserved, it would be a problem of allocation. It is worse than that, for a reason that follows from a fact nobody disputes: every drug has side effects.

Give a drug to a small number of seriously ill people and the side effects are a price worth paying; the alternative is worse. Give a drug to a very large number of mildly unwell people who would have recovered untreated, and the side effects are a cost imposed for no benefit. At the scale of a whole population, that is not a rounding error. It is a large quantity of harm, spread thin enough that nobody can see it and no single case can be blamed on anyone.

So the industry is not just failing to heal the very ill. It is doing a measurable amount of damage to the mostly well, and being paid for it, and the payment is what keeps the pattern in place.

Not a Conspiracy

I want to be careful here because this argument attracts people who believe things I do not.

Nobody in a pharmaceutical company needs to intend any of this. The scientists want to cure disease. The executives want to hit targets. The regulators want to approve things that pass the tests. Each acts sensibly inside their part of the machine. I have argued elsewhere that organisations are steered by interests, not by their members, and this is a clean instance: the incentive matrix produces the outcome regardless of who sits in the chairs.

That is what makes it hard to fix. Replacing the people changes nothing.

What Might Change It

The note I wrote to myself ended with the word "How?" and then "TBD", and I am not going to pretend I have moved much past that. But the shape of an answer is visible, because the problem is the shape of the incentive.

Pay for outcomes rather than prescriptions. If revenue followed recovery rather than consumption, the self-limiting market would stop being a gold mine, because those patients recover whether or not you sell them anything.

Separate research from marketing. The same organisation currently decides what to investigate and what to sell, so it investigates what will sell. Institutions that only research, funded to attack the diseases that kill, would not face that pull.

Run the trials publicly. A company testing its own product against a placebo, on a population that was going to improve anyway, reports a success. An independent trial, designed by people with no revenue riding on it, would find far fewer of them.

Each of these runs into the same obstacle, which is that the industry is a concentrated interest and the mildly unwell are a diffuse one, and I have written enough about how that contest goes not to be optimistic. But at least the diagnosis is clear. The industry is not broken. It is working perfectly, at the wrong job.