The Economics of Information II: Production

The Nature of the Resource: market relevant characteristics of information

As a resource, information has unique characteristics. Many of the things we commonly think about as resources are competitive in use: if I chop down a tree to build some furniture, that tree is not available to you to build your house; if I drink a beer, that same beer is not available for you to drink. Economists refer to such resources as “rival” — my use leaves less for you to use. However, if I use information, it has no impact on the amount of information left for you to use — certainly the fact that you’re reading this blog doesn’t leave any fewer ideas for the next person to read! Economists refer to such resources as non-rival. But information is unlike other non-rival resources: the more people use information, the better it gets. The reason I am writing this blog is so that I can get feedback on the ideas here and improve them. I coin the term ‘anti-rival’ to describe information, but would be surprised if I were the first to do so.

In Guns, Germs and Steel, Jared Diamond suggests that this anti-rival nature of information is what led to the rapid development of technologies and civilizations. If we look back over time, the rate of technological progress was exceptionally slow for the first 200,000 years or so of human existence — small bands of hunter-gatherers roamed the countryside, looking for food, and technological advances were separated by millennia. Then along came agriculture. Agriculture allowed denser populations. Ideas could circulate faster in these denser populations, and improve as they did so. Written language emerged, so that ideas could be stored and transmitted easier than ever. As the rate of flow of knowledge increased, so too did the rate of change of technology. The generation of new ideas and new technologies exploded. Mercantilism and industrialization led to more rapid communication of ideas between cities and across cultures, contributing to an even more rapid rate of increase in knowledge.

A recent biography of Genghis Khan provides another example. Genghis conquered most of Asia, the Middle East and even Eastern Europe. Everywhere he went, he adopted new technologies and spread them from place to place. Equally important, he opened up and protected trade routes, allowing people and ideas to continue to spread. As ideas spread, new users found ways to improve them. The spread of information through Genghis’ conquest ultimately paved the way for the European Renaissance and the industrial revolution to which it led. Genghis Khan could be considered the father of the modern age. As many people in the commons movement point out, information is like grass that grows longer and more nutritious the more it is grazed, so everyone should be free to graze on it as much as possible.

Most conventional economists respond that without the incentives created by private property rights (patents and copyrights) and the resulting profits they allow, there is no incentive to create information in the first place. The assumption here is that markets are the only way to allocate resources. An alternative would be publicly funded knowledge freely available to all. To decide which approach is correct, we must answer two questions: first, how do we allocate scarce resources (scientists, etc.) towards the production of information, and second, once the information exists, how should it be allocated among its consumers? In both cases, our goal should be to achieve the desirable ends for society (see economics of information part one.)

The first issue we have to look at is incentives. The old market story was the inventor in his garage, e.g. Thomas Edison, working long hours driven by the pursuit of profit. While even 100 years ago that story was incomplete (Tesla and Salk for example refused to patent their inventions, as their goal was apparently to contribute to the public good), today the situation is entirely different, as most research scientists work for a salary regardless of whether they are in the public or private sector. It’s hard to imagine that a scientist would work harder for a salary paid by Bristol Meyers Squibb than by the National Institute of Health so the incentive argument on the level of the corporate scientist would appear moot. There is also the evidence from mathematics (you can’t patent mathematical theorems, but that doesn’t seem to slow the rate at which they are produced) and open source software that the role of monetary incentives may be exaggerated in the first place. Unfortunately, there is an increasing trend for university professors to patent their inventions, and the large rewards they reap may provide added incentives. Joseph Stiglitz has recently suggested rewarding worthwhile inventions with a prize, but then making the information public property. Whether or not monetary incentives do drive worthwhile inventions would be an interesting but difficult hypothesis to test. In the meantime, there is no reason we could not develop a publicly funded system that provides incentives equal to the existing system.

The second issue is productivity. Is the private sector simply better organized and more productive than the public sector and publicly funded university scientists? In the private sector, the goal is to be the first to develop a patentable product. Firms therefore jealously guard their information, and are unlikely to share it with competing firms. We might have dozens of pharmaceutical firms striving to develop a new treatment for obesity (a disease of the rich), each of which has its own team of research scientists strictly forbidden from sharing information with others. This of course is not strictly true — one of the reasons that Silicon Valley generated so many new ideas was that the dense numbers of scientists and programmers were able to learn from each other. But this just serves to illustrate the greater point that information is anti-rival, and the more it can be shared, the faster it is likely to improve. When research is publicly funded, researchers freely share their information (while it’s true that even academic researchers guard their data and ideas prior to publication, the driving goal is to publish, i.e. share, information, not to keep it to oneself — though unfortunately this is changing as academics increasingly rush to patent their ideas). What is a more effective research model? 100 teams of scientists not allowed to talk to each other, or 10 teams allowed to communicate freely? As an academic, I recognize that my productivity would plunge towards zero if I were not allowed to build on the research of others. Even if private sector scientists are smarter or work harder than public sector and university scientists, which I personally find implausible, the fact that they cannot share their knowledge must seriously reduce their productivity.

Other productivity problems emerge from patents as well. In the case of patent trolling, private firms buy up as many patents as possible, not with the intention of using them, but with the intention of demanding royalties if anyone else ever does. Any resources dedicated to such activities are purely unproductive (known as rent-seeking behavior in economics), and actually decrease incentives for other inventors. Other firms apply for as many patents as possible simply to prevent competitors from using the ideas in the patents — basically, they are trying to slow the advance of knowledge! With the resulting proliferation of patents, one study found that every new medicine potentially conflicted with an average of 50 existing patents. Resources spent resolving the resulting legal conflicts are simply transaction costs, which virtually all economists agree should be minimized. As an example, one team of scientists wanted to developed rice genetically modified to produce vitamin A to improve the health of the world’s poor. After they developed this ‘golden rice’, they found that it infringed on 70 existing patents held by 32 separate companies. With licensing arrangements too complex to negotiate, the team had to end up ceding commercial rights to AstraZeneca but retained non-commercial rights for research and poor farmers.

The third issue is resources. Would the public sector allocate adequate resources towards R&D? Many governments already invest a lot of money in basic research, which they recognize as too expensive for individual firms to carry out. They also provide huge subsidies to private sector research, even when the results are patented by the private sector. While government sponsored research might require an increase in taxes, it could also reduce other demands on both government and private resources. Health care provides one of the best examples. Most people are not aware that even in the US, half of every dollar spent on health care is provided by the government. The skyrocketing cost of pharmaceuticals is rapidly increasing both private sector and government expenditures. If government sponsored research on pharmaceuticals were freely shared by all, pharmaceutical costs would likely plunge, freeing up government resources to spend on research and private sector resources that could be used to pay additional taxes. The private sector can fund its research through profits on patents, but those profits ultimately come from the taxpayers pocket. Should it matter to the taxpayer whether they pay for R&D through monopoly profits or through higher taxes?

In addition, an enormous amount of research in the pharmaceutical industry is on ‘me too’ drugs — if one company patents a highly profitable drug, other companies will look for something similar that they can patent, even if their products offer no improvements on existing drugs. This is an enormous drain on research resources. The same is true in other industries — do we really need thousands of models of digital cameras, for example? Another tactic is to slightly modify a product when an existing patent expires, again offering no significant value. Without the profit motive, research funds would not be wasted on these activities. Even if the government proves unwilling to dedicate as much money to R&D as the private sector, if knowledge were better directed and freely shared, presumably much less money would be required.

Next topic: Economics of Information III: consumption