The High Cost of Protectionism: AI Edition

Economists go on endless diatribes against tariffs and quotas as costly policies that raise prices to consumers and slow economic growth. There is considerable truth to this story, even if economists and politicians often exaggerate their case to push favored policies. While virtually all economists will go to their graves touting the evils of protectionism they almost all ignore the most costly forms of protectionism: government-granted patent and copyright monopolies.

Most tariffs raise the price of the protected items by somewhere in the range of 10-25 percent. By contrast, patent and copyright monopolies often raise the price of protected items by 1000 percent, or even 10,000 percent. Many high-priced drugs that enjoy patent monopolies or related protections can sell for tens of thousands of dollars. Their generic versions might sell for $30 or $40 a prescription.

There is a similar story for copyrights. Items that could be transferred at near zero cost over the Internet can instead sell for hundreds or even thousands of dollars. This is most evident with costly software, but also true for recorded music and video material, video games, and a variety of other material subject to copyright protections.

There is a clear rationale for patent and copyright monopolies, these monopolies provide an incentive for innovation and creative work. But every type of protectionism has a rationale, having a rationale doesn’t prevent a trade tariff or quota from being a protectionist policy.

The point here is that these government-granted monopolies are huge interventions in the market. They arguably are justified, but it is close to nuts to just assert they are the free market. (Alternative mechanisms are discussed here and in chapter 5 of Rigged [it’s free]).

It is understandable that people on the right, who generally support policies that redistribute income upward, would try to hide patent and copyright monopolies as just the natural working of the market. It is absolutely mindboggling that many on the left also perpetuate this blatant misrepresentation of reality.

Anyhow, let’s get the playing field set. Granting these monopolies is a choice by governments, they can set different policies. In the case of AI, the New York Times reports that it seems as though China is rapidly catching up, and possibly even taking the lead, by pursuing open-source policies rather than relying on patents, copyrights, and related protection.

The idea of open-source with reference to AI is that all the coding is freely available to anyone to review and build upon. It is understandable that this could be a more effective way to advance the technology since breakthroughs could quickly be built upon by others working in the same area. Also, researchers could learn from failures as well and avoid pursuing similar dead-ends. (This would be a great approach to drug or vaccine development.)

The other obvious development is that the finished product is very cheap. The developer may look to recover costs by charging servicing fees and/or relying on direct government support. Either way, end users will not be prevented from being able to take advantage of a useful product by its high price.

Anyhow, given all the hype in the business world around AI it would certainly be ironic if Chinese firms surged past their leading U.S. competitors because they relied on an open-source process whereas our firms relied on old-fashioned protectionism. Who knows, maybe even the “free trader” economists would notice one day.

This first appeared on Dean Baker’s Beat the Press blog.

The post The High Cost of Protectionism: AI Edition appeared first on CounterPunch.org.