Flawed Marginal Cost Theory Should Not Be Applied to Intellectual Property

Recent calls for the government to lower prescription drug prices by overriding patent rights include proposals for the establishment of a marginal cost pricing system in the pharmaceutical industry (and in patent-based markets in general). As a previous article in this series details, some academics and advocates are now suggesting that the government use a federal law (known as § 1498) to force companies to sell patented drugs at the amount it costs to produce one unit of the drug—also known as the marginal cost. But while providing cheap prescription drugs for all sounds like a noble endeavor, the proposal ignores economists’ near complete rejection of the marginal cost theory and threatens to disrupt the development of potentially life-saving medicines.

Marginal Cost Theory Fails Economic Scrutiny

Proposed 80 years ago by American economist Harold Hotelling, traditional marginal cost pricing theory maintains that maximum societal benefit is realized when products are sold at a price that corresponds to the cost to produce one additional unit of a good. Not included in marginal cost are any “fixed” costs incurred prior to the production of that additional unit, such as rents, property taxes, salaries, or any other overhead costs not dependent on the amount of goods produced. To arrive at a balance that would maximize general welfare, Hotelling proposed that government revenues be used to cover fixed costs in industries where these costs were significant, such as power plants, railroads, and other public utilities. With the fixed costs of these industries subsidized by the government, companies could then offer consumers goods and services at the reduced marginal cost of production.

Despite some initial acclaim, it wasn’t long before Hotelling’s proposal was systematically challenged and repudiated by a number of preeminent economists. As detailed in a 1946 article by Ronald Coase, the key problem with Hotelling’s marginal cost proposal was that it assumed the government has very good information on the marketplace and consumer demand. Without good information and the will to use it, Coase explained that a system of government subsidies would result in a “maldistribution” of resources and income that would ultimately result in losses similar to those the system was aimed at alleviating. This critical problem of a less-than-omniscient government was echoed by others over the years, and now, as Professor John Duffy explains, “[t]he rejection of the Hotelling thesis is so complete that reputable economics encompasses the very opposite of Hotelling’s view.”

New Proposals Disregard Parallel Shortcomings 

Notwithstanding the widespread dismissal of Hotelling’s marginal cost theory, proponents of expansive and unprecedented use of § 1498 insist that that because intellectual property and innovative products are different from the public utilities referenced in Hotelling’s work, they will benefit from a government-run marginal cost pricing scheme. In a 2016 article calling for § 1498 to be used to lower drug prices, proponents argue that allowing the government to impose prices that reflect the marginal or “generic” cost of production will maximize social benefit. Their argument is based on the theory that intellectual property and information are non-rivalrous goods with a marginal cost of zero and that any “supra-marginal” costs result in inefficient deadweight loss.

But proposals for using significant government subsidies to address pricing in innovative industries ignore the fact that the critical problems with marginal cost pricing in public utilities are just as, if not more, apparent in intellectual property. Responding to arguments in favor of a system of public subsidies to cover fixed costs in a broad range of innovative and creative industries, Professor Duffy’s 2008 article on marginal cost controversy identifies four basic problems that, while originally raised to refute Hotelling’s proposal, directly apply to intellectual property.

The first and most obvious problem Professor Duffy confronts is the distortionary effects of taxes the government would need to impose on the general public to fund the subsidies marginal cost pricing requires. Proposals to implement government subsidies often assume, with little-to-no evidence, that tax distortions would be smaller than any distortions flowing from IP rights and non-marginal pricing. Professor Duffy contends that because there is no evidence that IP rights result in a price divergence any greater than other areas of the economy, they may actually be small in comparison to the taxes that would be required in a subsidy scheme.

The second problem is Coase’s aforementioned issue of government ignorance and the misallocation of resources that would result from a lack of information. Essential to government implementation of a successful marginal cost scheme is sufficient information about the demand for intellectual property—specifically whether consumers would have demanded the good if they had to pay the total cost. As Professor Duffy explains, gathering data in an attempt to answer this question would be impossible “because consumers are not paying full costs and never expected to do so.” Inaccurate assumptions that attempt to fill gaps in the data will inevitably lead to inefficiencies and the misallocation of resources.

Professor Duffy then challenges the notion that a marginal cost scheme would redistribute income in favor of those consuming the goods targeted by the price controls while using taxes to spread the cost of the subsidies to the general public. After questioning the fairness of this type of distribution, Professor Duffy moves on to an in-depth critique of the efficiency of such a system and a warning of the rent-seeking behavior likely to follow. He explains that this type of wealth transfer will result in inefficiencies when individuals are likely to expend resources to bring about the transfer. Professor Duffy describes a pricing scheme that promises far cheaper costs to consumers:

That regime will appeal most strongly to those who currently consume those goods, for they will obtain the largest benefit. But the economic efficiency of a reward system comes exclusively from increasing consumption among those consumers who value the good (or some units of the good) below the price that would be charged by the right holder under the current IP system.

A system that targets certain areas for subsidies invites consumers of those products to expend resources in attempt to secure the subsidies, thereby opening the door for rent-seeking behavior and increased inefficiencies.

Finally, Professor Duffy emphasizes Coase’s criticism that marginal cost proposals incorrectly compare marginal cost pricing to single-pricing schemes. Coase explained that the industries identified by Hotelling are free to rely on “multi-part” pricing, or price discrimination in which no deadweight loss occurs. Professor Duffy applies Coase’s argument to intellectual property and stresses that rights owners have a true incentive to minimize deadweight loss in an attempt to capture the resulting gains. Because rights owners have this incentive, have the power to charge different prices to different categories of consumers, and have potential constraints of competition from other innovative technology, Professor Duffy concludes that “it is by no means clear that the IP right holder will cause greater distortions than the government’s revenue agents.”

Incentives to Innovate Must be Preserved

While it may be true that costs associated with providing intellectual property to consumers are low or close to zero, the fixed costs of innovation are often very high. Creators and innovators rely on intellectual property rights to commercialize their products and recoup the massive fixed costs required to develop innovative products and creative works. Government mandated marginal cost pricing schemes implemented through § 1498 would threaten innovators’ ability to both recoup past losses and to fund the development of new technology. In a 2005 article on economics and IP law, Richard Posner warns of this dilemma:

Marginal-cost pricing would maximize access to existing intellectual property and deter or expel inefficient entrants, but it would reduce, indeed often eliminate, the incentive to create the property in the first place.

Recent calls for the government to impose marginal cost pricing in the innovative industries gloss over the serious flaws that have led to the widespread dismissal of such proposals in the past. While those advocating for § 1498 do suggest a type of risk-adjusted compensation for companies subject to the marginal cost scheme, they couch the proposal in highly abstract terms and fail to consider that appropriate compensation would be impossible to accurately calculate. Additionally, as Professor Duffy notes, the literature upon which these proposals rely “treats the marginal cost pricing problem of intellectual property as a unique phenomenon” without thoroughly examining whether IP is distinguishable enough from Hotelling’s public utilities that it would benefit from government price control.

The delivery of affordable drugs to people in need should be something that innovative companies and the government work together to ensure, but flawed mechanisms and unsound economic theories should not be the foundation of such an important endeavor. Before rushing to implement a marginal cost pricing system, proponents would be wise to consider the drastic effects it could have on incentives to innovate—and ultimately the development of the next ground-breaking drugs.