
Michele Boldrin on Intellectual Property 05/18/2009
What this covers
This is an EconTalk conversation in which Russ Roberts questions Michele Boldrin about Against Intellectual Monopoly, the book Boldrin wrote with David Levine. Much of the argument is theoretical. Boldrin says Marshallian competitive dynamics already let a first entrant recover its fixed cost through rents before imitators erode them. He rebuts the standard counterargument, a game in which a rational imitator never copies, by pointing to its hidden assumption of costless, unlimited copying. He also traces the "ideas are a public good" premise to Arrow's 1962 paper and to the growth theory that revived it. In his telling, Schumpeter has a good side (competitive creative destruction) and a bad side (the case for large oligopolies), and George Stigler's neglected 1956 paper on competition and productivity supports the first.
The second half is largely historical and sectoral. Boldrin describes early PC software and other 20th-century industries as periods of heavy copying with little patenting. He recounts how Germany's ban on chemical product patents coincided with its rise past Britain, and how Swiss chemicals and pharmaceuticals grew for a century without patents. On drugs, he cites the industry's own estimate that roughly 80% of a new drug's cost is clinical trials, argues that new compounds often come from startups built on publicly funded research, and suggests public funding or mandatory cost-sharing could replace patents. He also questions whether patenting accelerated innovation after Bayh–Dole.
The conversation then turns to music, where copyright arrived late in continental Europe, and to private responses to Napster and the VCR. Roberts adds a public-choice point: once patents exist, incumbents lobby to keep rivals out, whatever the economic merits. Boldrin concedes that patents or prizes could be justified where innovation is enormously costly and imitation fast, but he treats that as an empirical question. He also warns that patents on tiny steps in software can block thousands of other products.
Michele Boldrin argues that 'intellectual property' is really an intellectual monopoly, akin to tariffs. He holds that Marshallian competition through imitation already rewards innovators enough in most cases, so patents and copyrights mainly shift effort from producing to rent-seeking. Even pharmaceuticals do not need patents, because their main cost, clinical trials, could be handled through public funding or mandatory cost-sharing.
- Historically, new industries take off with heavy copying and little patenting; patents arrive later, when mature incumbents use them to block entrants.
- The growth-theory case for patents confuses abstract, non-rival ideas with productive ideas, which are embodied in people and are rival and excludable.
- Cases like Swiss, German and Indian chemicals and pharma, Travelpro, and the music industry show innovation thriving without patents, while IP rents discourage innovation.
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
George Stigler, in a little-known 1956 paper (written 1950) for the 25th anniversary of Chicago's Social Science Research Building, made the anti-Schumpeterian argument. He ran a regression on 1920s–1940s industry data, using competition indices to explain labor productivity growth, and found that more competitive industries had faster productivity growth. Labor productivity is the true measure of innovation, unlike the patent counts current research uses.
“there is a paper by stickler in a completely unknown volume published on the 25th anniversary of the social science research building at Chicago”
For a long time Germany prohibited patents on chemical products, on the ground that natural compounds cannot be privately appropriated, while allowing patents on processes. This forced firms to imitate products but compete hard on cheaper processes. German chemistry thereby went from behind Britain, which was dominant in 1840, to dominance by 1890.
“in Germany patents on chemical products were prohibited for a long time”
A game-theoretic reductio defeats the standard argument on its own terms. - An innovator (Russell) sinks $100M; an imitator (Michele) can copy for $10. - If competition is Bertrand with unbounded capacity, imitation drives price to marginal cost: the innovator loses $100M and the imitator loses $10. - A rational imitator therefore never imitates, and the only equilibrium is that the innovator gets a monopoly and is never imitated. This contradicts observed reality, which shows that the faulty assumption is unbounded capacity, i.e. costless copying.
“I claim that that is the only equilibria of this game”
The endogenous-growth argument is wrong at its premise because it confuses abstract ideas with productive ideas. Abstract ideas like '2+2=4' are non-rival. But productive ideas, the ones actually known and used in production that have social value, are embodied in someone's brain and are rival and excludable: you have to bribe or torture someone to get them. So the case for monopoly protection collapses.
“our argument confuses abstract ideas with our actual productive ideas and my claim is the productive ideas ... are completely rivals and completely excludable do they typically embodied typically in the brain of somebody”
Patents become central once an industry matures. After the shakeout few players remain, profit from innovating becomes harder, and the rate of return falls, so legal protection becomes the good way to make money. Rent-seeking takes money from others, whereas production and exchange create wealth.
“only when they industry mature were when that there are a few players left over after the shake”
Alfred Marshall's textbook account of a competitive industry already solves the fixed-cost problem through imitation: - A first entrant builds a plant (a fixed cost) and earns rents above marginal cost because capacity is low and demand is high. - Imitators then enter, capacity rises, and price falls along the demand curve. - Entry continues until rents just cover the fixed cost of the last plant built, which is the long-run competitive equilibrium. So innovators recover fixed costs without monopoly.
“their rant typically will cover the fixed cost of the factory and more and then there is somebody else looking at him”
The Boldrin–Levine model allows patents or prizes to be socially useful in special cases. These are cases where the cost of innovating is gigantic, imitation is very easy, and imitators can expand capacity very quickly. The question is quantitative and empirical: a trade-off between the fixed cost of innovation and the speed of imitation and capacity build-up.
“if the cost of innovating is gigantic and it's really easy to imitate and copy an ending particular if it is really easy to expand capacity very quickly then my argument fails”
The Swiss chemical and pharmaceutical industry evolved for over a century without patents, which the Swiss constitution prohibited for chemical products until 1978. Its origin lies in French dye makers who fled to French-speaking Switzerland in the second half of the 19th century when France introduced patents. They left so they could keep producing without paying royalties to the holder of the key patent on the fuchsine dye. Under Anglo-Saxon lobbying, Italy's constitutional court struck down its pharma patent ban in 1978 and Switzerland changed its rules about a year later.
“the Swiss pharmaceutical company evolved for more than a hundred years in complete absence of patents”
Alfred Marshall's textbook account of a competitive industry already handles fixed costs. The first entrant sets up capacity at a fixed cost and, because demand is high and capacity low, earns rents that cover the fixed cost and more. Imitators then enter, quantity rises and price falls along the demand curve until rents just cover the fixed cost of the last plant built: the long-run competitive equilibrium. Boldrin and Levine say this core theoretical point is not new.
“if you look at his description which i think is i would steal our economics 101 description of our competitive industry works”
The standard pharmaceutical argument is that drug research has extreme fixed costs and very uncertain returns, so without patents no one would seek new drugs. This is both true and false, and evidence is limited because pharma is secretive. By the industry's own estimate, about 80% of the average cost of a new drug is clinical trials for FDA approval, not invention.
“by their own pharmaceutical industry estimate about eighty percent of the body compute to be the average cost of a new drug is clinical trial”
There is an almost deterministic pattern in the big innovative industries of the 20th century. At their origin, many small firms compete and copy each other like crazy, with very little patenting even where patents are available. PC software is the classic example: Bill Gates himself said that if patents and copyright had been used in the 1970s there would be no software industry.
“at the origin of the industry that is a flurry a large number of people's small firms typically small innovators competing struggling copying each other like crazy”
The notion that ideas and innovation are public goods comes from Kenneth Arrow's famous 1962 paper (in the Nelson-edited volume) on information and technological advance. Arrow concluded that innovation should therefore be financed by the public purse. That is more theoretically coherent than granting monopolies, if ideas really were public goods. Arrow confused abstract and productive ideas, and he was pushing for NSF taxpayer funding of basic research. New growth theory in the 1980s simply revived this.
“the ideas and innovation are public goods and as such should be financed we are the public purse”
New effective active compounds come mostly from small labs and startups, typically spun out of publicly funded university research. A researcher funded by NIH hits something commercial, works with the university patent office, forms a one-product startup, patents it, and licenses it to big pharma, which handles legal and regulatory work and marketing. Big pharma has thus become essentially a commercializer and 'legalizer' of others' inventions.
“strange enough they come from small lab from small startup company typically the outcome all University lab research the research was was financed by public money”
People imagine patents as the seed of innovation. Historically, though, in many if not most cases patents come late in the game, used to keep out competitors after the innovation has already taken place.
“patents come in later in the game to keep out competitors after the innovations already taken place”
Even if some invention arguably deserves a patent, the political process will predictably favor entrenched incumbents over would-be competitors. This public-choice dynamic is disastrous.
“the way the political process is going to react to that is to favor the entrenched existing competitor at the expense of the would-be competitor”
Copying costs that are close to zero are not the same as zero, and the difference matters enormously. Only perfectly costless copying would produce unbounded capacity and price-at-marginal-cost destruction of innovator rents.
“the being closed is not the same thing as zero something is it makes a big difference there”
The bigger culprit for losing the dynamic view of competition is general equilibrium theory, more than game theory. It reduced competition to static price-taking at a common price and lost the dynamic process, as Stigler complained.
“I put it more or not the general equilibrium theories that took it to an abstract view of competition for the point that it was boiling down to you know essentially price taking”
There are two Schumpeters, a Dr. Jekyll and a Mr. Hyde. The good Schumpeter is the 1911 German book The Theory of Economic Development, which describes competitive creative destruction. The bad Schumpeter is Capitalism, Socialism and Democracy (c. 1942–43). Fearing that Soviet planning was working and capitalism was failing, that book wrongly concluded that innovation requires big oligopolies and multinationals able to plan long-term, a view that heavily influenced industrial organization.
“it's like dr. Jekyll and mr. Hyde ... the theory of economic development is the good Schumpeter that's a book publishing German in 1911”
Even if some things ought in principle to be patented, the political process will react by favoring the entrenched incumbent at the expense of would-be competitors. That public-choice dynamic is disastrous.
“the way the political process is going to react to that is to favor the entrenched existing competitor at the expense of the would-be competitor”
The limited available data do not show that pharmaceutical innovation exploded after the Bayh–Dole Act and the expansion of pharma patents. Meanwhile big pharma has found a stable modus vivendi of huge, stable profits by monopolizing markets. The industry's structure is driven mostly by the legal framework rather than technology, and it does not have to be this way.
“the little did I have does not show that since the bail dole Act and the expansion of patents in pharmaceutical arm innovations I've exploded”
When patents are available, firms turn to government to keep out competitors even when there is no good economic case. Economists' theoretical justifications give them cover to pursue self-interest in socially destructive ways.
“we as economists unfortunately then have given people to cover to pursue their self-interest in ways that are socially destructive”
Patenting every small step ('every small epsilon') is becoming extremely damaging in areas like software. Complex products use thousands of inputs, so patents on individual steps can block thousands of other new processes; emerging empirical research documents this negative impact.
“patenting of everything every small epsilon is sorry to become extremely damaging in areas like software”
The belief that musicians would stop making music without the tens of millions copyright brings them is a myth. Copyright for music is a latecomer, arriving in continental Europe in the second half of the 19th century.
“we're all living in a mythological world in which we think if you know the musician doesn't make tens of millions of dollars out of their song they won't sing anymore”
The Napster debate ignored private solutions that would emerge without strong copyright, such as earning money from concerts or bundling extras like lyric sheets and videos. The VHS and Betamax episode illustrates this: movie studios feared VCRs would destroy their industry, yet theaters improved to compete and studios made substantial money from video.
“it ignored the potential for private solutions to emerge”
Publicly funded research that yields a patentable innovation is then turned into a private monopoly, as in pharmaceuticals. The public finances the research for its claimed positive externalities, but most of the social benefit is captured via the monopoly, which makes the arrangement even more rentier.
“a lot of publicly funded research to the moment it produces a some ideas or innovation that is patentable turn into a private monopoly”
So-called intellectual property is not genuine property but belongs to the same class as tariffs and exclusive-right monopolies. As a general principle it is overall inefficient, with some exceptions or special cases, and there is neither a theoretical nor a practical reason to treat ideas as property.
“we just find that this particular one is of the same class as Terrace and exclusive right monopoly right”
The most negative reaction came from the new growth (endogenous growth) literature of Paul Romer and followers, who argue that fixed costs of creating ideas are central to growth. - In that model, ideas or 'blueprints' are non-rival public goods that everyone can use at zero cost once discovered, and they are non-excludable without legal protection. - With constant returns in standard inputs, this yields increasing returns from the technology externality. - Hence a temporary monopoly is needed to give people an incentive to discover ideas.
“the most negative reaction we have got ... was from this whole new growth ... or endogenous growth literature our old friend call it Paul Romer”
Alternatively, keep the current system but replace patents with mandatory licensing priced at the cost of trials. If a drug's trials cost $800 million, a second firm wanting to sell the same active compound pays the first $400 million to split the cost, and a third firm splits again. The FDA would deny approval until entrants have purchased their share, avoiding the social waste of redoing trials.
“all they have to do is give me 400 we'll split the cost I've already done the cost at the clinic”
The 'pharma needs patents because of high costs' argument uses clinical trials as a smokescreen. Clinical trials are a genuine public good imposed by regulation because of a principal-agent problem: firms have a temptation to shade safety results. Public money should therefore fund clinical trials on a competitive basis, for example by redirecting NIH funds, rather than basic research, which produces private goods embodied in labs.
“the clinical trial part that a smokescreen why because that should be if there is something on which we should put public money”
Because copyright lets the music industry 'milk the cow' without innovating, it failed to exploit the potential of digital distribution (MP3, the internet). Over more than a decade it made only small steps toward what Napster was already doing, which proves that badly allocated IP creates bad incentives for innovation.
“the incredible potential of digital distribution of music ... has not been exploited by this industry purely because ... they have no incentive they have a nice rant to extract”
Instead of the FDA and public funding, private testing agencies with reputations could emerge to certify drug safety. Boldrin agrees this market-based redesign is the more radical option, but takes political reality as given and prefers public funding of trials.
“you also could argue that private testing agencies with reputations would emerge instead of an FDA which is I think would happen”
India had a thriving pharmaceutical industry producing good drugs at low prices while not enforcing patents. Having adopted patents under the WTO, Boldrin predicts its firms will monopolize and 'legalize', becoming less efficient and less creative while making more profits. Ronald Jones's research suggests this is unfortunately ongoing, though some Indian firms are becoming big world players.
“the Indian had a very thriving pharmaceutical in this was producing a lot of good stuff a low prices they were not enforcing patents”
Historical data show that the Marshallian imitation process does work for most innovation. Patented innovations are only a subset of all innovations, although that subset has grown as patent law has been extended to more fields. This suggests patents may not be needed even where they are granted.
“wait a second it does work for most innovation in fact the innovations cover by patents are a subset of innovations”
The fact that students pay large tuition to hear professors explain ideas that are centuries old and freely available in books shows that ideas are not costlessly transferable once discovered. Productive knowledge must be reproduced in each mind at real cost.
“why do they waste their time listen to you me there is a free ideas being reproduced millions of time”
Piracy is so successful largely because of the gap between the price consumers pay, about $20 for a CD, and what music could legally sell for online, a couple of dollars. A gigantic gap 'blows off' moral restraint. If downloads cost about $5, many people would refrain from pirating, both because piracy is time-consuming and for ethical reasons.
“you can still have to pay 20 bucks for a CD for music could be sold legally on the internet at a couple bucks three bucks”
Economists have a terrible tendency to look for theoretical proofs, for example that externalities justify intervention or that high fixed costs justify patents. Whether fixed costs actually matter is an empirical question, and in many cases they are small relative to early opportunities and profits.
“in economics we have this terrible tendency to look for proofs so we prove that externalities say justify government intervention”
For a camera phone, the abstract idea of putting a camera in a phone becomes totally public once produced. The real value lies in implementation (quality, emailing, uploading), which is much more private. Granting a patent merely for thinking of it first is absurd.
“the whole value of it is the implementation and how it's put into the phone”
Imitation may lower returns, but it pushes competitors to compete on dimensions other than price, such as usefulness and customization. Innovators face a choice between rent-seeking in Washington and serving consumers, and patent justifications tilt that choice toward rent-seeking.
“it encourages the different competitors to compete on things other than just price”
Although virtually all music is available free via P2P, the CD industry suffered only a modest dent and profitability has not fallen much. 'Free' downloads are not really free: they are time-consuming, low quality, hard to find and often incomplete.
“yes there was a dent in the CD industry but not a big deal profitability hasn't gone down very much”
In a new industry, possibilities for cutting costs and improving the product are large and demand is huge, so mutual copying benefits all firms. The opportunity cost of spending time on legal fights over copying is too high compared with innovating and expanding capacity. This is why early industries feature little patent enforcement.
“the opportunity cost or wasting my time fighting you legally because you copy me it's just too high”
For everyday economic ideas, nobody thinks the first mover deserves exclusivity, for example the first jeans shop on a street or the first to put cameras in cell phones. Imitation (copying good ideas, improving quality, lowering cost) is what real-world competition is, and it drives prices toward equalization. So it is puzzling why exclusivity should apply only to some ideas.
“we would not like the idea the principle that the first guy that opens the jeans shop in that particular happening has to have the exclusive”
Cutting the copyright term to about ten years would be plenty under current market conditions. Some profits would fall, but the music industry would stop being a lottery. The big companies that gatekeep and push a few stars would lose their reason to exist, opening doors to many more musicians.
“if I cut copyright for what it is now to ten years say that it would be plenty”
In policy terms, IP should be dismantled gradually, like trade tariffs, because abolishing all restrictions overnight would cause 'insurrection'. Shortening patent and copyright terms is a good starting experiment, as opposed to the current trend of lengthening them. Co-author David Levine would abolish them overnight.
“I see them a bit like trade tariffs you know I'm a free trade guys but I also understand from a practical point of view that if I abolish every trade restriction overnight I'm gonna have insurrection”
Copyright terms have been extended repeatedly while technology has cut production and distribution costs and multiplied market size. As a result, today's superstar musicians earn far more than Frank Sinatra or interwar stars did. Their opportunity cost is not ten times larger, so the extra rent is not needed as an incentive.
“the superstar musicians now makes so much money compared to what even a Frank Sinatra or the famous singers say in between the two wars would have made”
When imitation is very easy and cheap, the fixed cost of innovating was usually also very low. Travelpro, the first to make the rolling carry-on suitcase, was easy to copy and did not patent the idea. It was not driven out of business: it had lead time, made abundant profits, and remains a leading company.
“when imitation is very easy very cheap because the fixed cost of innovating was also very low and the Travelpro example is the perfect one”
IP lawyers and artists are self-interested in the copyright debate. Claims by artists that copyright is crucial to their productivity should be taken with a grain of salt.
“when an artist tells you the copyrights crucial for their productivity you have to take it with a grain of salt”
The Anglo-Saxon countries caught up with German chemical technology essentially by seizing German industrial secrets after Germany lost the two World Wars, on a much more massive scale after the second.
“we caught up with them essentially because we took away all their industrial secret after that they lost two wars”
From informal sampling of his undergraduates, Boldrin estimates that only about 10–15% of the content on a typical iPod was purchased; people buy one tune and share with roughly nine friends. The effective price is therefore about one-tenth of the $0.99 list price, and Apple appears to tolerate this.
“their attitude is one out of ten meaning I pay for one tune and I share with nine buddies”
Game theory has made mixed contributions on this issue. One useful contribution shows that patenting can induce too much effort, because winner-take-all races create a lottery.
“you can have too much effort in the presence of patenting because of the effort of the winner takes all”
Copyright for music is historically a latecomer: in continental Europe it arrived in the second half of the 19th century and was initially short. It has since been extended repeatedly.
“copyright music is a late comer by the way in general writing continental Europe it comes on in that second half of the 19th century”
Big pharma has reached a modus vivendi with the legal and regulatory system that gives it huge, very stable profits and monopolized markets. The industry became what it is partly through technology but mostly through legal structure, so blaming firms misses that the structure is the problem, and it need not be this way.
“they have found a modus vivendi that they are making huge profits very stable their profitability they've created a way of monopolizing markets”
The most negative reaction in the economics profession came from new/endogenous growth theory, Paul Romer and his followers, whose framework depends on fixed costs of innovation and the resulting monopoly rents to drive growth.
“the most negative reaction we have got for a long time ... was from this whole new growth ... or endogenous growth literature our old friend call it Paul Romer”
The book tells a Schumpeterian, Marshallian story of competition in which new industries create and destroy and early winners are soon competed down. That view was doubted from roughly the 1980s–90s as game theory became the way to analyze innovation. Game theory emphasized first-mover advantages, path dependence and strategic deterrence.
“there was a strong argument from the the game theoretic approach which was being first is really important”
Innovators often choose not to patent because patents carry out-of-pocket costs. Disclosure also invites competitors to engineer around the patent, so some innovators prefer secrecy.
“the risk that by patenting the the device or the idea that you open the opportunity for competitors to try to work around the patent”
One theoretical justification for patents is that imitation creates an externality: the copier benefits from the innovator's work, which leads to underinvestment in innovation, and patents realign incentives. Boldrin accepts this as good theory but insists its practical relevance must be tested.
“one way of justifying part is from the basic that is an externality is being extradited from imitation which induces an inefficiency”
Digitization and the internet are technological progress that has made parts of the music industry's current structure and jobs socially useless, as has happened at GM and Chrysler. Those hurt will fight back, and losers may deserve some compensation. But a legal defense of their monopoly position is not justified.
“the digitalization and the internet makes the current structure there is some of those jobs just useless from a social point of view”
The social product of the millions or billions of dollars the NSF has spent on economics research is dubious: good for the economists who got checks, with limited and sometimes negative spillovers. Government-funded research is a very mixed bag.
“the social product from the millions maybe billions of dollars that the NSF has been on economics research is I shudder at the value of it”
Stigler's The Theory of Price, now rarely used but subtle and deep, has a chapter on 'the quicksilver nature of competition' telling the Marshallian story. In game-theoretic models, by contrast, the world appears barely competitive, full of strategic moves such as incumbents threatening to flood the market and price at zero to deter entrants. These are always theoretical and never real.
“called the Quicksilver nature of competition and it's really the marshallian story”
Boldrin and Levine's public defense of Napster ('Why Napster is right') was not a claim that it was legal under current law. It was a provocation showing what current technology makes possible and how the music industry's property-rights system blocks technological advance.
“it's showing what current technology can allow us to do and what the current property rights system in the music industry is preventing is blocking technological advances”
The standard argument for patents and copyright: an innovator who invests heavily in a new idea will see quick-entering competitors destroy its profitability without exclusive rights, so the incentive to innovate would disappear.
“if I don't have the exclusive rights to the concept behind it then new competitors can come in very quickly destroy any profitability”
It is puzzling, and telling, how much content people give away on blogs without direct profit, and how much students pay for teaching of ideas that are freely available.
“I think it's fascinating how much stuff we give away our only blog”
Pirate sites succeed largely because consumers still pay about $20 for a CD when music could be sold legally online for two or three dollars. Roberts adds that he is surprised iTunes still charges $9.99 per album.
“you can still have to pay 20 bucks for a CD for music could be sold legally on the internet at a couple bucks three bucks”
Boldrin and Levine resisted for years going public with the claim that patents are not useful, because the conclusion from their theoretical model 'sounded crazy'. Studying the history of actual innovation revealed a consistent pattern that convinced them to stake out the position empirically.
“I remember resisting for a long while saying okay let's go in public and tell people that actually we think patents are not useful”
Legal scholars gave the warmest reception to the Boldrin–Levine thesis. IP scholars such as Larry Lessig and Mark Lemley (Stanford) see the system as out of control and turned into pure rent-seeking, so they want it scaled back dramatically rather than abolished. Lemley has recently moved close to Boldrin's position.
“their basic point is that if the fingers gone so much out of control it has to be scaled back dramatically because it's become pure rent seeking”
Roberts expected bands to bypass distribution systems and expected iTunes to become more of a record label because of its pleasant interface. This has happened much less than he thought, suggesting distribution systems provide real value.
“I thought iTunes would become more of a record label because of its the interface is so a pleasant to use”
Roberts praises the book's implicit and explicit ideas about competition and especially its empirical nature.
“the empirical nature of the book is also a very good thing”