The Lemon Problem

Why the best endometriosis surgeons are the hardest to find — and why you are part of the reason

At two o’clock in the morning, in a country you do not practice in, a woman is deciding who will perform her next surgery.

She has done her homework. She has read the patient forums until her eyes hurt. She has watched the surgical videos of a surgeon in one city and the case demonstrations of a surgeon in another, and compared the before-and-after posts on two Instagram accounts. Google reviews have been read — three hundred for one surgeon, twelve for the other. Fellowship pedigrees, hospital rankings, book chapters, podcast appearances, testimonial videos — the research is thorough.

Six weeks from now she will fly to a city she has never been to, and hand her body to a surgeon she has chosen through this process.

Now, one of two things will happen.

She will be lucky. Or she will not.

The point of this essay is not that she is being foolish. She is being rational. She is doing exactly what a well-informed patient in a market with asymmetric information can do: she is optimising over the signals available to her.

The problem is that the signals available to her do not track surgical quality.

---

Akerlof

In 1970, a young economist named George Akerlof published a paper that almost did not get published. Three journals rejected it and an editor said the argument, if correct, would imply that most markets could not function. And since most markets did function (apparently), the argument had to be wrong.

The paper became one of the most cited economics papers of the 20th century. Akerlof shared the Nobel for it in 2001.

The argument fits in one paragraph.

Imagine a market for used cars. Some are good, let’s call them peaches; some are bad, let’s call them lemons.

The seller knows which is which. The buyer does not.

Because the buyer cannot tell, she is only willing to pay an average price — halfway between what a good car is worth and what a lemon is worth. At that price, the sellers of good cars cannot recoup their value; so they exit the market. And because the sellers of lemons get more than what their cars are worth, they want to sell. So all good cars leave the market. The average car becomes worse. The buyer’s expected value drops.

She is willing to pay less. More good sellers exit. In the limit, the market collapses into lemons only.

Akerlof’s point was not really about used cars. It was about any market in which the buyer cannot verify quality and the seller can.

Health insurance. Hiring. Anything sold on trust.

Endometriosis surgery.

---

The Market

Endometriosis surgery is one of the purest lemon markets in modern medicine.

The buyer — the patient — has almost no way to directly evaluate surgical quality. She cannot watch the operation. She cannot verify that the disease was completely excised. She cannot know whether her post-operative pain relief is because of what the surgeon did, because of what her body did, or because of what her endocrinologist did after. She cannot know, a year later, whether the disease she has today was left behind or newly grown. Even her second-opinion surgeon, looking at what remains, often cannot tell.

She has, at most, a handful of proxy signals:

Fellowship credentials. Hospital marketing and centre-of-excellence designations. Surgical videos on YouTube. Case demonstrations at conferences she cannot attend. Patient testimonials on the surgeon’s website. Reviews on Google and health-rating sites. The referral of a gynaecologist who is himself downstream in the same lemons market.

Of these signals, how many correlate meaningfully with the thing she is actually trying to measure — whether this surgeon will excise her disease completely, without collateral damage, with a durable outcome?

The honest answer, if we are willing to give it, is: a few of them, weakly.

The rest are noise she cannot distinguish from signal.

---

The Mirror

You may be reading this thinking about the surgeons at the loud end of the market — the ones with the ring lights and the branded content and the DMs open — and you may feel something between amusement and contempt for what they do. You may believe you have opted out. That the quality of your work speaks for itself. That your patients find you through referrals, not marketing. The market’s noise is somebody else’s problem to fix.

This is a misreading of your position.

You are not outside the market. You are inside it, occupying a specific corner of it, sending a specific set of signals whether you intend to or not. Your quiet is a signal. Your absence from social media is a signal. The volume of your practice — set by referral patterns you did not choose — is a signal. The city you happen to work in is a signal. The five patients this month who chose the loud surgeon over you did so on the basis of the signals you were, in effect, sending or refusing to send.

Here is what Akerlof’s model says about the good seller who refuses to participate. It does not say she preserves the market’s integrity. It says she exits. And her exit makes the average worse.

The refusal of excellent surgeons to make their excellence legible is not neutral. It has victims. The victims are the patients who cannot find them — and who therefore choose someone else. Someone easier to find.

---

The Kind of Signals that Work

In 1973, Michael Spence (who would share the 2001 Nobel with Akerlof) published the paper that answered the obvious follow-up question. If markets with hidden quality collapse, how do the ones that survive, survive?

His answer was signalling.

Sellers and buyers develop conventions, costly, hard-to-fake signals, that allow quality to become visible. A university degree signals capability because obtaining one is costly enough that low-capability people cannot easily fake it. No one (including you) respects the content. Most of it isn’t even directly useful.

Warranties on products that you buy signal reliability because a manufacturer of unreliable goods cannot afford to offer them. Peer review signals scientific credibility because reviewers who approve bad work damage their own reputations.

The Spence test for a signal is simple: is this signal costly enough that a low-quality operator cannot afford to send it?

Apply that test to the ten proxies on the patient’s list.

A curated YouTube channel of beautiful cases fails the test. Every surgeon has beautiful cases; the channel shows those and no others. Curated testimonials fail the test. So does a polished Instagram feed. So do most media appearances. A fellowship pedigree passes weakly — the cost is real, but the correlation with skill is loose. Hospital centre-of-excellence designations depend entirely on who is issuing them; most fail the test.

But some signals pass, and pass cleanly.

Published outcomes with meaningful follow-up pass. Bad outcomes cannot be edited. Peer-refereed audit passes. Willingness to publish and openly discuss one’s own complications passes — no low-quality operator can afford this signal, because a full accounting of their complications would end them. Long-term follow-up data of a defined cohort passes. The referral pattern of other surgeons in the same subspecialty, the operators who cannot be fooled, passes.

And one more, which is perhaps the strongest quality signal a surgeon can send, and which is almost never faked:

Sending patients away when you are not the right person for their case, publicly and by name.

---

What next?

None of this is easy. The infrastructure for costly signals — shared registries, third-party outcome verification, transparent complication reporting — is underbuilt in gynaecological surgery, and the individual surgeon cannot fix that alone. But the individual surgeon can decide which ones to invest in and which ones to refuse.

If you are a surgeon who has been telling yourself that quality speaks for itself and you do not need to signal, the argument of this essay is that you have misread the market you operate in.

You do not need to become the loud surgeon. You do need to become a legible one.

The practical version of this is unglamorous.

Publish your outcomes, including the ones you would rather not. Track your patients past the six-week mark; publish what you find at one year, at three, at five. Present your complications at conferences, not just your saves. Write things that other surgeons will read; the referral pattern of your peers is the signal patients cannot see but that shapes what they can. Send patients to better surgeons when the case demands it, and let both patients and colleagues watch you do it. Refuse to compete on the signals that don’t track quality, and invest, instead, in the ones that do.

The point of these actions is not to build a personal brand.

It is to make excellence visible in a market where, as things currently stand, it is not.

---

At two o’clock in the morning, in a country you do not practice in, the woman deciding who will perform her next surgery has now made her choice. She is booking a consultation with the surgeon whose signals she could see.

Somewhere else — perhaps in the same city, perhaps on the same social media platforms she has been scrolling — is the surgeon who would have given her the outcome she is hoping for.

She will never find them. Because they are not legible.

That is the market for lemons.

And it is the market we all work in, whether we like it or not.

---

That’s it for this week. See you in the next one.