January 30, 2004
Patents and Research
Earlier, I mentioned Alex Tabarrok's idea of rewarding intellectual discovery in proportion to effor. Now, Brother Lowe has picked up on the idea, and pointed out some flaws.
I think that passage of such a law would turn out to be the Cost Accountant and Patent Lawyer Full Employment Act of 200x. As I said, I can see the theoretical appeal. But in practice, I think that this attempt to reward the costs of innovation would only create more expensive makework, none of which would have anything to do with research itself.
Some of the issues that Lowe raises were pointed out by commenters on my earlier post. My main response is to say that all methods of reward in the intellectual property sphere are flawed. In some cases, such as business process patents, I think that patents are more flawed than the alternatives. In other cases, patents are not the most flawed alternative. As noted in another recent post, Lawrence Lessig agrees that in the case of drugs, patents have some clear benefits.
Having said that, I think we should encourage Tabarrok and others continue to think "outside the box." One thought I have is to use prizes rather than patents as rewards for research. Some prizes could be offered ex ante. For instance, the government might offer a prize for developing an AIDS vaccine.
Other prizes could be offered ex post. Suppose somebody develops what they believe is unique and valuable distance learning software. They could apply to the "prize office" (replaces the patent office) with a description of what their software achieves and ask the prize office to determine a fair prize for such an achievement.
I know, I know, there are all sorts of issues. But conceivably, after enough prizes have been given out and enough precedents have been set, the system could be workable.
Maybe we should offer a prize for the best alternative to the patent system.
Posted by Arnold at 11:56 AM | Email this entry
| Category: intellectual property
What is the problem we're trying to solve here? It seems to me we're reacting to things like a $500 million judgement for Eolas. We want to reward innovation, but we also understand that execution makes or breaks the success of products.
What if we put a cap on patent infringement awards? It could be a multiple of dollars invested in development of the investment. Let's make it 100. When applying for a patent, the applicant would have to supply a reasonable accounting of development expenses. If a company incidentally infringes on a patent -- as Microsoft probably did in Eolas, rather than scouring the website of the USPTO and finding an interesting way to embed obejcts in web pages -- its liability would be bounded and there would be a clear incentive for both parties to negotiate a fair license.
The prize idea has been championed extensively by the Tofflers and the Newt Gingrich wing of the Republican Party, although as a stimulus for invention and exploration rather than a solution to the patent quagmire. A very positive idea!
-Brad Posted by Brad Hutchings on January 30, 2004 12:50 PM | Permalink to Comment
The prize idea could have merit as something that operated in addition to the patent system, as in the example of a prize for curing a given disease, but not as a replacement for the patent system.
The problem is who decides what the invention is worth, and on what basis? MY invention is clearly highly valuable, yours...not so much. You think the converse. So we both lobby the bureaucrat who decides the value, and who probably has no clue what an invention in that field is actually worth, so the whole thing becomes a beauty contest. The flaw is the implicit assumption that the value is not only knowable, but KNOWN to someone, if only he'll tell us.
Our present system admits that no one person knows the value of an invention, and so lets the market decide what it's willing to pay. That places the decision in the hands of those who know the needs that the invention purports to fulfill, and the alternative solutions, and therefore is in a position to decide what the invention is worth to them (as opposed to a valuation assigned by a bored bureaucrat who probably doesn't need the invention, doesn't know the field, and generally couldn't care less). Put another way, would you want a "fair" salary for you to be determined by someone at the IRS?
Incidentally, the "prize" idea has already been tried – by the Soviet Union. Bureaucrats awarded inventors an "inventor's certificate" and in some cases perhaps threw in a few rubles. Not coincidentally, the USSR was something less than a powerhouse of innovation.
Similar considerations apply to capping patent infringement awards. Artificially setting a cap would penalize holders of highly valuable patents (i.e., those who had made the greatest contributions), while protecting those of less valuable ones – which is hardly fair. In many instances, such a cap would not provide an incentive to negotiate a license, but quite the contrary, an incentive to infringe and take one's chances of being caught. A cap – unless it's so high as not effectively to constitute a cap – destroys the deterrence effect.
Suppose the cap were $100 million, and the market was $1 billion/yr. A rational competitor would then infringe, perhaps taking the market away from the original innovator (think of Microsoft vs. two guys in a garage), and view the risk of a $100 million infringement judgment as a cost of doing business. The only protection for a little guy is the possibility that the big guy MIGHT be socked with a huge judgment that would make the entire foray uneconomic. In a real life analogy, would you obey speed limits if speeding tickets were limited to $10?
Setting the cap as function of cash expended in development does not solve these problems either on a practical or theoretical basis. First, on a practical basis, a "reasonable accounting" is a quaint notion in view of the recent accounting scandals. Ethical, knowledgeable accountants often disagree about what is and is not reasonable. (Bear in mind that Hollywood deals are invariably defined in terms of the gross rather than the net, because once movie studio accountants finish most blockbusters end up "losing" money.)
Second, if a company has only one revenue-generating product, ALL of its costs – right up to the Christmas party – up to that point HAVE to be charged against development of that product, by accounting rules. On the other hand, a company that has two or more revenue-generating products now has to apportion costs among them, again inviting abuse. So the treatment a company receives would depend on its status, which is unfair.
On a theoretical basis, the cash expended in developing an invention in no way reflects the value of the invention. This is a rewarmed version of Marx's labor theory of value, and it doesn't wash. (I don't care how much time and money somebody spent developing this keyboard I'm typing on – all I care about is what it's worth to me.) A further theoretical difficulty with the cost accounting approach is that it rewards inefficiency and punishes efficiency. Why bother to strive for efficiency if you'll be made whole on the back end?
The bottom line is that there is no "one size fits all" solution. Issues need to be decided on a case by case basis, which is what our system does. Sometimes it does things that some people consider mistakes, but the point is that what constitutes a mistake is not always clear, even in retrospect. Let the people who are closest to the situation – those who want the invention, i.e., the market for it – decide what the invention is worth. Posted by Occam's Beard on February 2, 2004 01:29 PM | Permalink to Comment
THE Director of an Observatory, who, with a thirty-six-inch refractor, had discovered the moon, hastened to an Editor, with a four-column account of the event.
"How much?" said the Editor, sententiously, without looking up from his essay on the circularity of the political horizon.
"One hundred and sixty dollars," replied the man who had discovered the moon.
"Not half enough," was the Editor's comment.
"Generous man!" cried the Astronomer, glowing with warm and elevated sentiments, "pay me, then, what you will."
"Great and good friend," said the Editor, blandly, looking up from his work, "we are far asunder, it seems. The paying is to be done by you."
The Director of the Observatory gathered up the manuscript and went away, explaining that it needed correction; he had neglected to dot an m. Posted by The Literary Astronomer on February 2, 2004 11:51 PM | Permalink to Comment