Forgotten Rationale for Patents

Patents are legally defined rights, but the patent system is intended to have an economic payoff.  This works in two ways, often described as “social contracts”.  First, patents are granted to induce investment in innovation — both the research behind the invention and the resources needed to make inventions into products in the marketplace.  The basic idea is that what looks to be an inefficient and undesirable monopoly in the short run will promote innovation in the long run — especially if large investments are needed.  Yet studies have shown that patenting can actually be a substitute for, not a complement of, R&D spending in high-tech.

The second quid pro quo is that patents publicly disclose knowledge (rather than keeping it as a trade secret) in exchange for a limited period of exclusivity.

Both tradeoffs are problematic.  Patents are not the only way for innovators to capture returns on their investment.  Other means include secrecy, first-mover advantages, learning speed, complexity, reputation, complementary assets, complementary services and copyright in the case of software.  Extensive economic research [1,2,3] shows that patents are the most important means to secure return on innovation mainly for pharmaceuticals and chemicals.

As a general principle, in free market economies competition drives innovation.  Patents make sense where there is substantial market failure.  The strongest case for intervention is in pharmaceuticals, where the costs of development and testing are high and the cost of copying and reproduction is low.

The disclosure tradeoff also works relatively well in pharmaceuticals.  Patents are few, competitors know who claims what, and at least in the case of new molecules, the claims made are unambiguous.  In software, nobody reads patents — except lawyers when necessary.  Not only are there far too many software patents, their validity and value are uncertain and they are written by lawyers for lawyers.  Moreover, lawyers advise engineers and technologists not to read patents, because if infringement is found to be “willful,” defendants may be liable for treble damages.

Information may be free, but acquiring accurate, complete, and meaningful knowledge about patents is expensive and risky.