In previous posts I asserted that the Centers for Medicare and Medicaid Services (CMS) pays much less than private insurers for the same healthcare services because price discrimination by healthcare providers is easy to achieve (here and here). In general, price discrimination gets a bad rap in the economic literature and is even illegal in many instances. But in the case of healthcare, it serves a useful purpose. This is an underappreciated strength of the US healthcare system.
Such is not the case for the pharmaceutical drug market, however. Pharmaceutical producers’ inability to prevent arbitrage severely limits their ability to price discriminate. If they could perfectly price discriminate, the prices charged to poor, uninsured people would be much lower than they are currently, perhaps only a few dollars for a round of treatments.
This is a bold claim, so I should support them with some evidence.
Price Discrimination in the Pharmaceutical Drug Market
Perfect or “first degree” price discrimination would allow the producer to sell each unit to each customer at a different price. Each customer would have to pay the absolute maximum that they individually could and would pay. A millionaire would pay perhaps a million dollars for a single pill. A homeless vagabond, maybe only ten.
As long as the price the customer pays is greater than the marginal cost of producing it, the transaction adds to the producer’s profit. Even if the cost of the research to bring the drug to market was several billion dollars, the marginal cost of producing an extra pill is only a few dollars at most. Refusing to sell a pill to a customer for anything less than $1,000, for example, would lower the producer’s profit if the most the customer can pay is a few dollars. Consequently, even a profit-maximizing pharmaceutical producer has an incentive to sell its drug to poor customers at a very low price. This is not just theoretical conjecture. Real pharmaceutical companies do this in a limited way. If you go to almost any pharmaceutical company website, you will see a statement that says something like, “If you are unable to afford ____, we can help”. The picture below is from the website for Harvoni, a very expensive hepatitis C treatment.
Of course, anybody can claim to be uninsured, poor and suffer from Hepatitis C. Pharmaceutical companies must contend with the same information asymmetry that private insurers face. It is difficult for them to be sure their product is not being resold to customers who could otherwise pay more.
Pharmaceutical Companies Set Their Prices Globally
The “resale” problem also applies to sales to healthcare systems in other countries. It would make perfect sense for pharmaceutical companies to negotiate a different price for each country. Countries with single-payer systems could negotiate a much lower price than fractious systems like the one in the U.S. But the incentive to purchase the drug in the low-price country and resell it to the high-price country at a small markup would be very great. Consequently, pharmaceutical companies normally hold the line when negotiating with single-payer countries and refuse to lower their price.
The following is from a description of the Democratic proposal to lower drug prices:
The legislation would create a maximum price to aid negotiations called the Average International Market price. Drawing on the idea of an international pricing index — which Trump has said he supports but that many Republicans dislike — the so-called AIM would be the average price of a drug in six countries (Australia, Canada, France, Germany, Japan and the United Kingdom) weighted on the basis of sales volume.
The underlying assumption here is that these other countries pay significantly less than the US for the same drugs. But as I have asserted before, this doesn’t happen to any significant degree. Pharmaceutical companies don’t price discriminate in the way this legislation assumes they do.
Here is a passage from an article in the Guardian about the UK’s decision to ration Harvoni, the treatment for hepatitis C mentioned above:
An estimated 215,000 people in the UK have chronic hepatitis C infection (160,000 in England), which new but costly drugs can cure. Addaction, a charity that helps people overcome drug and alcohol abuse, says the decision to treat [only] 10,000 people a year is “manifestly unfair”.
The article also reveals that the cost of Harvoni ranges from £26,000 to £78,000 in the UK depending on the length of treatment. These prices are consistent with those paid in the U.S.
The marginal cost to the producer of a round of treatments for one patient are only a few dollars. The producers of Harvoni forego profits by not selling the drug to the UK at a much lower price for the nearly 200,000 people who will not receive any treatments. The reason why is because they fear it would lower revenues from the patients who can pay the high price, especially from the U.S.
A policy to simply pay the average paid by other wealthy countries will achieve nothing. This Democratic proposal to lower drug prices is very similar to many other proposals. It sounds good. Its proponents can crow about how they are sticking it to the big, bad pharmaceutical companies. But it is unlikely to have any real effect. And if it did have an effect, it would probably do more harm than good.
Harvoni and all the other hepatitis C drugs that are now on the market only exist because their producers had a realistic expectation that they would recoup their investments selling at very high prices. If we arbitrarily decrease these prices, we decrease the incentive to develop new drugs.
The AIM component of the Democratic proposal would, by itself, be ineffective, but it is a step in the right direction. What’s missing is a coordinated effort by the wealthy countries to find a price that each can afford to pay. A price that adequately compensates the pharmaceutical companies for their costs of investment in research and development.
The solution to this kind of international problem was outlined long ago by an Italian genius named Vilfredo Pareto.
More about that next time.