Price Controls on Medicines Today Mean Fewer New Medicines Tomorrow

The most expensive drug in the world is the one that does not exist. It has no price tag, no press conference and no patient advocate holding up its photograph. Nobody has ever been interviewed on the evening news about the medicine that was never developed, because nobody knows to miss it.

That is the whole problem with the politics of drug prices, in one paragraph. The saving from a price control is visible: a number on a chart, a headline, a smiling official. The cost is invisible: a treatment that a company decided, years earlier, not to chase. I want to argue that this asymmetry is the most important fact in the debate, and that it should make conservatives, and anyone who wants new cures, deeply wary of letting Washington set the price of medicines.

I will make the argument in numbered premises, state the strongest objection fairly, and answer it. Let me begin with terms.

Terms and the claim

A price control is a legal limit on what a seller may charge or what a buyer may pay. It can be a ceiling, a negotiated figure backed by a penalty, or a formula. What matters is that the number is set by the government and not by the parties to the transaction.

Research and development, or R&D, for a drug is the process of discovering a candidate compound, testing it in the laboratory and in animals, running human trials in three phases and seeking approval from the Food and Drug Administration. I describe that path step by step in How a New Drug Gets Through the FDA and Why the Trip Takes So Long.

My claim: government price-setting for drugs reduces the number of new medicines that will exist, because investors fund research only when success can pay for the many failures that come with it. That is a claim about incentives, and it can be tested against how the system works, not against anyone's motives.

Premise one: the pipeline is long, expensive and mostly fails

A new drug starts as an idea in a lab. Thousands of compounds are screened to find a few that show promise. Those that survive go through preclinical testing, and a small fraction go to human trials, which run in three phases: a small first phase to check safety, a larger second phase to look for evidence that the drug works and to find the dose, and a large third phase to confirm the effect and watch for side effects across many patients. Then there is the application to the FDA.

At every step, most candidates fail. A compound that looked brilliant in cells may do nothing in people. A drug that works in a small trial may fail in a large one. A drug that works may have side effects that sink it. The honest summary, which every pharmaceutical economist accepts, is that the large majority of drug candidates that enter human trials never reach the market. Estimates of the total cost to bring one new drug to approval vary widely, from hundreds of millions of dollars to several billion, depending on whether one counts the cost of the failures and the cost of capital, and those estimates are disputed. I will not pick a number. The structure is what matters.

Here is the structure, shown with a hypothetical. Suppose a company funds ten drug candidates, and each costs $200 million to carry through the full process. That is $2 billion in total. Suppose nine fail and one succeeds. The one success has to earn back the entire $2 billion, plus a return for the investors who took the risk of losing everything, or the company will not fund the ten. The single successful drug does not just pay for itself. It pays for the nine that did not work. The arithmetic is not a theory about greed; it is what any portfolio of long-shot bets looks like.

Now introduce a price limit that cuts the successful drug's lifetime revenue by, say, a quarter. That is a hypothetical, and I label it so. The company's decision is not "do we still make a profit on the one success?" It is "do we still fund ten candidates?" If the revenue on a success falls enough, the answer for some candidates is no. And the candidates that go first are the riskiest ones, the ones with the highest chance of failure and the highest potential benefit, such as a first treatment for a disease with no current therapy. Nobody will ever see them. That is the fewer-medicines part.

Premise two: the patent is a bargain, and it has a clock

The Constitution anticipated this. Article I, Section 8, Clause 8 gives Congress the power "To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." That is the patent power, and it has a feature that the Framers wrote into the text: the right is for "limited Times." For a deeper look at this clause, see The Patent and Copyright Clause Is the Only Power With a Stated Purpose.

In the United States a patent lasts 20 years from the date of filing. That sounds long, but a drug company files its patent early, usually when the compound is first discovered, years before human trials. By the time the drug is approved, a substantial portion of the 20 years has gone. That is why the law has provisions to restore some of the lost time, as I will describe in a moment.

The patent bargain is this: in exchange for disclosing the invention and for a limited monopoly, the inventor gets a period to recover costs, after which anyone may copy the invention. That is not a loophole for the drug industry. It is a deal society struck, and it has a built-in expiry that benefits the public. When the patent ends, competition arrives and prices fall, often dramatically.

Premise three: Hatch-Waxman built the generic market

That expiry is no accident, and it is the best example I know of Congress getting the balance right.

The Drug Price Competition and Patent Term Restoration Act of 1984, known as the Hatch-Waxman Act after Senator Orrin Hatch and Representative Henry Waxman, did two things at once. First, it created a streamlined pathway for generic drugs. A generic maker could file an abbreviated new drug application, relying on the brand's safety and effectiveness data and proving only that its product is bioequivalent, without repeating the expensive clinical trials. Second, it gave brand-name makers something in return: a restoration of part of the patent term lost during FDA review, up to five years, with limits on the total, and periods of marketing exclusivity for certain new products.

The law also set up a procedure for challenging patents, and a reward of 180 days of exclusivity for the first generic company to successfully challenge a patent. The result was a flowering of generic competition. Today the great majority of prescriptions filled in the United States are filled with generic drugs, and they cost a small fraction of the brand price. That is what an incentive-based approach produced: a Congress that said, in effect, the inventor gets a limited period, then the public gets the competition.

The lesson is that there is a way to lower drug prices that works with the logic of the market rather than against it. You set a clear, predictable term of exclusivity, you make it easy for competitors to enter on the day it ends, and you let price fall by competition. The price falls anyway. Nobody had to hold a hearing about it.

Premise four: the Inflation Reduction Act changes the clock

The Inflation Reduction Act of 2022 took the other approach. It gave the Secretary of Health and Human Services authority to negotiate prices for certain drugs in Medicare that have been on the market for a number of years and that account for large spending. The law is long and has many provisions, but the ones that matter for this argument are the clocks.

A small-molecule drug, the kind taken as a pill and made through chemical synthesis, becomes eligible for selection seven years after its FDA approval, and the negotiated price takes effect after nine years. A biologic, a complex product made from living cells such as many cancer and autoimmune treatments, becomes eligible eleven years after licensure, and the negotiated price takes effect after thirteen. The law also has exceptions, including for drugs for rare diseases with a single designation, and for drugs for which generic or biosimilar competition is expected.

Why do the clocks matter? Because they tell an investor how long a drug can earn before the government sets its price. A pill and a biologic are treated differently, and critics noticed it immediately. A small-molecule drug gets two fewer years before it becomes eligible for negotiation than a biologic. The difference has been called the "pill penalty," on the theory that it steers investment toward biologics and away from small molecules, which tend to be cheaper to make and easier for patients to take. I am not claiming that this effect has been proved. The law is new enough, and the effects distant enough, that the data cannot settle it. I am claiming that an investor deciding where to put money, five or ten years before a drug even reaches the market, will read the clock and respond to it. Economists call this a response to expected returns. Anyone who has ever decided whether to open a restaurant has done the same thing.

Here is the logic in a sentence: price controls do not change the cost of making a drug, they change the return on the success, and the return is what pays for the failures.

The strongest objection

I promised to state the best case against my position, so here it is as its defenders would state it.

"Taxpayers fund the basic research. The National Institutes of Health spends tens of billions of dollars a year on research, and the basic science that leads to new drugs often begins in publicly funded labs. A drug company that takes a discovery from an academic laboratory, runs it through trials, and charges a high price is capturing a return on an investment the public made. Americans pay for the science and then pay again for the product. Price negotiation merely recovers part of what the public already paid for. Besides, the profits of the industry are high, and the claim that every dollar of revenue is needed to fund innovation is self-serving."

That is a serious argument, and part of it is true. The NIH does fund a great deal of basic research, and many important drugs trace their origins to academic discoveries. The Bayh-Dole Act of 1980 let universities patent inventions made with federal funds and license them to companies, precisely to move those discoveries into products. I do not dismiss the point.

Here is my answer, in four parts.

1. Basic research is not a drug. The NIH funds the discovery of a biological target or a mechanism. It does not fund the years of chemistry, toxicology, manufacturing, three phases of human trials and regulatory filing that turn a discovery into a medicine. That later part is where most of the money is spent and where most of the failures occur. A finding that a certain protein plays a role in a disease is not a treatment, and the distance between the two is exactly where private capital is at risk.
2. The public already gets a return through the patent term. Bayh-Dole and the patent bargain mean that the product enters the market, then, after a limited time, becomes a generic. The taxpayer's return is a cure that exists, followed by a cheap copy. The alternative, a discovery that sits in a journal because nobody will fund the trials, returns nothing.
3. "Profits are high" is an argument about the average, not the margin. An industry can show healthy profits on its past successes and still decide, at the margin, not to fund the next risky candidate if the expected return has fallen. Investors do not fund what happened; they fund what they expect. The decision to start a program depends on expected returns in the future.
4. If the concern is the public's return on NIH-funded work, the targeted fix is to address that directly, with licensing terms or transparency about government-funded patents, not a general price-setting power over every drug that a company develops with its own risk capital.

Measuring it against a principle

There is a deeper reason to be careful, and it is a matter of principle as well as economics. A price is information. It is a signal that tells everyone in the system how much a good is worth to the people who use it. When government replaces that signal with a decree, it does not abolish scarcity. It hides it. For the best historical illustration, see the story of how a German economist ended price controls on a Sunday and launched a recovery, in Ludwig Erhard ended price controls on a Sunday and launched a miracle. For a lighter treatment of what happens when a committee sets every price in town, see a thought experiment where every price in town is set by a committee.

The same structure shows up in the history of the FDA, where the visible harms of a bad drug drove the rules and the invisible harm of a delayed drug was left uncounted; see Frances Kelsey, Thalidomide and the 1962 Law That Slowed Every New Drug. The recurring mistake is always the same: counting what you can see and ignoring what you cannot.

What to do instead

I am not arguing that Americans should be content with high prices. They should not, and neither should patients who cannot afford a prescription. The question is how to get lower prices without killing the pipeline. A shortlist of approaches that work with incentives rather than against them:

• Speed generic and biosimilar entry. Faster approval of copies and an end to tactics that delay competition lower prices without a price decree. This is the Hatch-Waxman principle applied to the present.
• Make prices visible. A patient cannot shop for what she cannot see. Transparency in the middle of the system, discussed in Pharmacy Benefit Managers, the Middlemen You Never Hired, Explained, matters as much as the manufacturer's price.
• Shorten the path to approval. Every year of review is a year of lost patent life and delayed treatment. The reasons, and the cost of delay, are in the FDA post linked above.
• Let the terminally ill try. The Right to Try law of 2018 let terminally ill patients who have run out of approved options seek access to experimental drugs; see The Right to Try Law of 2018 and the Families Who Pushed It Through.

None of these requires the government to decide what a medicine is worth.

The case in brief

1. The visible savings from a drug price control are paid for by invisible losses: the treatments that are never developed because the expected return no longer covers the failures.
2. Drug development is a long-shot portfolio. Most candidates fail, so one success must pay for many failures, and a lower return on success means fewer candidates funded.
3. The Constitution's patent power grants exclusive rights for "limited Times," and the Hatch-Waxman Act of 1984 paired that limit with a fast path for generics, which lowered prices through competition, not decree.
4. The Inflation Reduction Act of 2022 sets different clocks for small molecules (selection at seven years, price at nine) and biologics (eleven and thirteen), and investors will respond to them. The fair objection, that the NIH funds basic research, answers a question about discovery, not about the costly steps that follow.

What you can do: read Article I, Section 8, Clause 8 in the Constitution, then ask your representative which bills before Congress would speed generic competition, and whether those bills are paired with any price-setting power.