Karsten on Free Software

Glyn Moody pointed me to a recent draft paper (.pdf) by economists  Michele Boldrin and David K. Levine. It’s an interesting read. Here are the notes I made while reading. I’m posting them in order to make the arguments in the paper accessible to more people. I or FSFE don’t necessarily share these views.

There is no empirical evidence that patents serve to increase innovation and productivity, unless we use patents themselves to measure innovation and productivity. [p1] This is the “patent puzzle”: there is no increase in rate of technological progress despite ever-stronger patent protection [p1].
Innovation is hardly ever born out of patent protection. Instead, it is the fruit of highly competitive-cooperative environments.  Innovators profit mainly from first-mover advantages, not patents. [p1]

Political demand for strong patent protection comes from old and stagnant firms, not new ones. Mature industries turn to patents only after an initial “hot” phase of innovation and rampant growth, as their growth potential shrinks and the industry becomes concentrated. [p1] Case in point: While Apple released its first iPhone in 2007, it did not really start to use patents to fight competition from Android until 2010. [p4]

Patents block competition and innovation

The basic problem of patents is that existing monopolies block future innovation. This is made worse because modern products consist of many thousands of potentially patented ideas. Anyone who holds a patent on any of these ideas is potentially in the position to block the product or levy a tax on it.

In mature industries, patents are widely used to inhibit innovation, prevent entry and encourage exit of competitors: [p5]

“The dead hand of dying institutions ­ Texas Instruments was famous for this and now we have the example of Microsoft ­ gets hold of the industry as they attempt to tax consumers, new entrants and any potential competitor.” [p5]

There is no statistically significant correlation between measures of productivity and patenting activity. [p17] But there is a significant correlation between competition and productivity growth. [p18]

In an example of the inefficiencies created by the patent system, Research in Motion paid 612.5 million USD to NTP for a patent license. Patent was later invalidated, but RIM didn’t get its money back. “In this setting it is no surprise that patent trolls hope to get rich quickly.” [p6]

Patents don’t actually serve to spread knowledge through the publication of ideas that would otherwise have been kept secret. Disclosures in modern patents don’t provide enough information to replicate the technology. Also, innovators will patent exactly when the amount of time for which the workings of an invention can be kept secret is smaller than the duration of the patents. This may merely shift the focus of innovation away from things where the workings of the invention can more easily be kept secret. [p2f.]

Patents in the pharmaceutical industry

Pharmaceutical companies may value patents more highly not because the initial fixed costs of developing a new product, but because disclosure is more meaningful when it comes to drugs. The chemical formula is available to competitors essentially for free. The clinical trials to prove that a new drug works account for 80% of the development costs.

On the other hand, the high monopoly prices of life-saving drugs imposed by the pharmaceutical industry have a great social cost. Instead of patents, a prize system might work better, fostering innovation while minimising the social costs. [p4f] Pharmaceutical products often sell for hundreds of times their marginal cost of production. [p6]

Abolition is the only solution

It may be theoretically possible to design a patent system that actually promotes innovation. But the political economy of patents makes it impossible to build such a system. So Boldrin and Levine argue that the only solution is to advocate outright abolition of patents: [p10]

“Being not a “property” right but rather a “monopoly” right, patent possessors will automatically leverage whatever initial rents their monopoly provides them with in order to increase their monopoly power until all potential rents are extracted and, probably, dissipated by the associated lobbying and transaction costs.” [p11]

Over the last two centuries, governments have progressively enlarged and strengthened the monopoly powers granted by patents. “At each stage of this process of enlargement the main driving force were the rent-seeking efforts of large, cash rich companies unable to keep up with new and creative competitors.” [p14]

Policy recommendations

Patents are akin to trade restrictions, as they prevent the entry of foreign competitors into national markets. Trade restrictions were vastly reduced over the past decades. It is time that the same reduction is applied to the patent system.

As the abolition of the patent system will require a transitional period, Boldrin and Levine make a number of policy proposals for this interim phase, such as [p21f.]

• Stop the expansion of the patent system, and stop favouring patent holders in the judicial process.
• Anti-trust and competition policies are key to fostering innovation.
• Stop exporting US / European policy on patents to the rest of the world through WIPO, WTO and TRIPS – the balance of trade in patents might easily reverse within a couple of decades.
• Tailor patents’ length and breadth to the needs of different industries. Move away from one-size-fits-all approach.
• Reverse the burden of proof for patent applications. Patents should only be granted when strictly needed on economic grounds.
• Use prizes to foster innovation, and ensure that the results of publicly funded innovation becomes available to the public.