Prof. Robert P. Merges, University of California at Berkeley School of Law
August 10, 2015

Fritz Machlup famously said that if we did not have a patent system, there is not enough evidence to show we should create one.  But he also said that because we do have one, there is not enough evidence to get rid of it.1

Machlup thus falls right in between a ringing endorsement and a harsh rebuke.  Yet this tepid assessment is still – 57 years after it was made – the starting point for most discussions of the field’s Big Question: Are patents justified on the economic evidence, or are they not?  This is such a conventional starting point for discussion that patent scholars have missed something important.  We know quite a bit more about how patents work now than we did 57 years ago.  The field has not stood still.  In fact there is now a much richer body of empirical data on patents than there was when Machlup laid down his lukewarm dictum.

Despite all this new data, we are in my view no closer to a convincing empirical answer than we were in 1958.  We have accumulated a large body of knowledge concerning how people and companies view and use the patent system.  We have detailed information on the behavior of many components of the patent system: courts, patent offices, litigating parties, licensors, and licensees.  But even with all this rich information we are no closer to answering the Big Question.  It is as if we were trying to decide whether to build a new highway, and we had extensive information about paving surfaces, on-ramps and exit ramps, optimal spacing of rest stops, and the proper number and size of road signs; but we still lacked the basic information of whether the cost of the highway was worth it, given the next best available transportation routes and the volume of car traffic we might expect.  We know a lot, in other words, but we don’t yet have the Big Answer.

An Illustration of the Problem: New Theories, Pro and Con

To show why uncertainty over the Big Question still exists, it may help to review two waves of recent IP scholarship.  One develops ideas about the benefits of patents that were largely unexplored in Machlup’s day.  The other examines carefully industries where IP has largely not been available – and argues that thriving innovation in these industries demonstrates that IP rights are not necessary in them, and possibly in other industries as well.

One new branch of literature emphasizes patent-related benefits in the form of enhanced incentives to disclose, exchange, and license information.  Briefly put, this research emphasizes the transactional role IP rights play in economic activity.  Traditional theory simplified research activity as taking place inside a large firm; the role of patents was to allow a firm to recoup expensive R&D costs.  This newer literature emphasizes patents as facilitators of exchange.  Thus patents are said to make small, independent “idea shops” more viable, because patents make possible a specialized market in pure ideas.  Patents affect not only the overall volume of innovation, but, critically, they affect where innovation takes place.  Because small companies have some recognizable benefits, patents may indirectly affect the overall volume and quality of innovative ideas.  But they do so indirectly: by making it easier for more innovative small firms to earn income through the licensing of innovative ideas to other firms that engage in actual production.

In some related research, two economists have shown that the availability of published patent applications (which came into U.S. law in 2000) led to (1) a higher volume of patent licensing, and (2) more rapid licensing of inventions than in earlier periods.2  These results show that patents tend to speed up transfer of ideas, and promote more rapid diffusion of innovations throughout the economy. Again, the effect of patents on innovation is positive but in a way different in kind from the classic treatment of patents as incentive to aggregate R&D spending.

The other post-Machlup branch of theory I want to mention looks at industries where IP rights are not available, but that seem to foster significant innovation nevertheless.  From French chefs3 to stand-up comics,4 and from fashion designers5 to tattoo artists,6 creative people working in various industries develop norms and practices that provide an adequate, and perhaps often superior, alternative to formal IP protection.  Some have taken to calling these areas, collectively, IP’s “negative spaces.”7  From these studies, a consensus theory has begun to emerge, which holds that IP law is far less necessary than many have traditionally supposed.  The absence of IP, they say, not only fails to impede creative contributions in these areas; in some cases at least there is more activity.  From a descriptive or positive beginning, in other words, negative space theory often moves to a more normative point: the essential wrongheadedness of the traditional story that IP rights are always and everywhere necessary to call forth creative works.

It would not be news to Machlup that there are other ways besides patents to encourage creative effort. What is new is the detailed ethnographic studies behind negative space theory.  This is not speculation; the deep-dive description of the industries being analyzed adds a level of specificity far beyond the high-level theorizing current in Machlup’s day.

Creeping Into Positive Territory?

On the surface the two trends in IP scholarship I have discussed might appear to cancel each other out. One more checkmark on the positive side of the ledger, one more on the negative side.  Is that all there is to it – a new stalemate on a bigger game board, replacing the old stalemate?

Maybe, but I think there may be some room for optimism on the “pro patent” side of things.  To begin, the study of patent-related transactional benefits is fairly new.  While it is difficult to estimate financial benefits, it is not hard to imagine that they may prove to be substantial.  Just the one study mentioned, on higher patent licensing volume and quicker deals in the aftermath of required patent application publications, might be thought to add very significantly to the case for patents.  If licensing increases, the incentives to invent new products and technologies increase; and the incentives to form specialist “idea shops” increase as well.  This could be an important benefit for the patent system.

On the other side, though the negative space literature looks impressive, the truth is that, taken together, the “negative space” industries studied to date (other than fashion) total roughly $12.75 billion in revenue per year (tattoos, French cooking, and stand-up comedy).  The fashion industry is much larger – but the problem is that many believe the “negative space” story (thriving innovation without IP rights) does not fit the industry at all.8  And if we remove fashion from the list of negative space industries, the fields studied in this research represent only a small percentage of economic activity that appears closely tied to IP protection. The International Intellectual Property Protection Alliance (IIPPA), for example, estimates that the “copyright industries” alone add $1 trillion to the U.S. economy each year.9  It is very difficult to arrive at similar estimates for the “patent industries” because companies in so many industries obtain large numbers of patents every year.  But we can say that the pharmaceutical industry is worth $ 340 billion per year alone,10 and that the chemical industry totals roughly $450 billion.11  Add in medical devices, not to mention information technology, and it is apparent that the accounts of negative space fields show we have a long way to go before deciding that many other industries would benefit from the elimination of IP rights.


The transactional benefits of patents are slowly being revealed, but the magnitude of the economic impact is still somewhat speculative.  Meanwhile, the negative space fields studied so far are small compared to the aggregate of all industries, but that could change as this literature grows and develops.

All of which leaves us perhaps a little closer to the Big Answer we seek – whether patents are a net positive for economic activity.  Closer, but as a fair observer would be forced to conclude: not quite there yet.