Friday, November 16, 2012

Solving the Patent Problem—and Much Else

I am at this instant attending a conference on what to do about problems with software patents. The purpose of this post is to argue that the real problem is more fundamental, and to propose a simple solution.

Under current law, an ordinary civil suit, whether for patent infringement or anything else, is a "heads I win, tails I break even" bet. If you persuade the court that it is more likely than not that your claim is right, you collect damages. If you don't, you don't.

That means that if you believe you are probably wrong, have one chance in five or ten of winning, it may still be worth suing. You could get lucky, and the other side could blink—agree to settle out of court—especially if your claim is for a lot of money. If you lose, you pay for your own lawyer and possibly, if the court thinks your claims are sufficiently unjustified, for the other side's legal costs. You do not pay for inflicting the risk of large damages on an innocent party.

This problem was faced more than two thousand years ago by the legal system of Periclean Athens, and they came up with a simple solution. Their equivalent of criminal law was privately prosecuted; the prosecutor usually got a large share of the fine paid by the convicted defendant. That created an obvious risk—accuse someone innocent, unpopular, and rich of a crime he did not commit, in the hope of either being paid off to drop the charge or persuading enough members of a large jury to get a majority vote for conviction.

Their solution was to make the prosecutor liable to a large fine, the equivalent of about three years of an ordinary workman's wages, if he failed to get at least 20% of the jury to vote for conviction.

How might we apply that approach to our legal system? A civil case, unlike a criminal case, is supposed to be decided by a "preponderance of the evidence." That means that if the plaintiff loses, it is more likely than not that the defendant was in the right, which means that it is more likely than not that the plaintiff was in the wrong, was trying to extort a payment to which he was not entitled. When the plaintiff wins, the legal system treats the result as if it was certain that he was in the right, so when he loses ...  .

Which suggests a simple solution to the problem both of software patents and of tort litigation more generally. Make the losing plaintiff liable. Base the amount he owes the defendant on the amount he claims,  giving plaintiffs, legitimate as well as illegitimate, an incentive not overclaim. 

One of the speakers at the conference brought up the example of a Pigouvian tax, such as a tax on air pollution produced by steel mills, and argued that the equivalent in the context of software patents would be an increase in patent maintenance fees. Charging for patents when the problem is patent litigation is like charging steel mills for steel instead of for pollution. The correct implication of his example is that there ought to be a cost for suing people, ideally for suing innocent people. That is what I have just proposed.

P.S. A recent online comment on a different blog post of mine demonstrated the risk of discussing IP issues without proclaiming one's allegiance to one side or the other of the current IP wars. Let me make myself perfectly clear. I hold no position on whether copyright law or patent law ought to exist. My views on the arguments for and against can be discovered, by those sufficiently patient, from the recordings of the course on IP Theory I am currently teaching.

P.P.S. It turns out that Athenian law came even closer than I thought to my proposal; although the details are unclear, there was, in at least some cases, a penalty owed by a losing plaintiff to the winning defendant in their equivalent of a tort suit:
"In some private cases in which the prosecutor had claimed a sum of money from the defendant, on losing the case he had to pay to the defendant one-sixth of the amount which he had claimed (epobelia, one obol per drachma). Demosthenes faced this risk when he prosecuted his guardian Aphobos in 364/3 for making away with his inheritance, but Kallimakhos did not when he claimed 10,000 drachmas in 400/399, even though epobelia was payable in a diamartyria or a paragraphe at that period. Either the law was changed between those dates or epobelia was payable only on some financial claims and not all. It is also uncertain whether it was payable on every acquittal or only when the prosecutor failed to get one-fifth of the votes."
(Douglas M. MacDowell, The Law in Classical Athens, Cornell university Press, Ithaca, N.Y., 1978, pp. 252-253.) 


Anonymous said...

The solution to patents litigation is pretty straightforward. End the patent laws. After all they are basically a government granted monopoly enforced by government courts. The reality of this is seen in the so called "patent troll" where some private entity brings a little efficiency to the system, and this is considered a terrible thing. But the whole thing is a gigantic patent troll. The only reason the patent system doesn't bring the economy to a grinding halt is that the courts are so unbelievably inefficient, and so the cost of enforcing the nightmare is high enough as to make it impractical to full enforce.

Would no patents introduce some problems? Of course, but surely you, as a libertarian, recognized that when the government gets involved, everything gets gummed up.

One need only look at the smart phone business where a brilliantly innovative company like Apple is now beginning to give up innovation and turning to patent litigation instead, with the consequential cost that we consumers loose the benefit of competition and innovation.

It is a lot easier to sue than it is to innovate and stay ahead of the curve, and a lot worse for society too.

MikeP said...

Indeed, this does not solve the patent problem.

The problem is that you need to patent software just to keep someone else from patenting it and suing you. This is only exacerbated by the recent change from first-to-invent to first-to-file.

A start up can focus on software and innovation and hope they don't get sued in the chance that their venture actually succeeds. Or they can put their precious resources into writing and filing patents long before it looks like they will get any return whatsoever.

Patents, at least in software, are incredibly counterproductive. Making losing plaintiffs of patent lawsuits pay more is a good step, but the damage of patent law has been done long before any lawsuit emerges.

Steve Winkler said...

I like this solution, but I think it falls short of the underlying problem. It is not just that wrongful lawsuits are pursued, but that patents themselves are by their nature doing more to restrain trade, investment, and innovation than they are doing to nurture these ends.

wtanksley said...

How about making the patents in question automatically lapse into the public domain?

rif said...

Would this actually work at all? I would think that under this regime, a big company that wanted to sue would spin off a subsidiary with no assets except the rights to the patent in question and enough money to fight the lawsuit. So you'd at least need some mechanism to go after the parent company, and so on?

David Friedman said...

to rif:

An alternative solution is that, in order to sue for $X, you have to post a bond of (say) $X/3. That would correspond very roughly to a rule in Roman law, although in that case the bond of the losing party forfeited to the court, not the other party.

Patrick said...

I listened to and enjoyed the first of your lectures (by the way, I think your link is broken--I used the link from your comment on the other blog.) In your discussion of patent and copyright terms, you talk about how the length of each should be how long it would take for someone else to come up with them (ignoring certain costs that should lower the terms further).
This idea dovetails well with the rule that independent creation is a defense to copyright infringement, since your independent creation of my copyrighted work is evidence that we have reached the upper limit of the appropriate term for that work. Have you considered the possibility that independent invention should likewise serve as a defense to patent infringement? It seems to me that the basic argument in favor of this rule is the same, and it would get rid of a lot of problems with, e.g., software patents (which virtually always involve independent reinvention.)

wtanksley said...

rif, courts regularly settle questions like that; the general term is that one has to ask whether the spinoff produces "arm's length" deals. If it seems that the parent company is getting special preference, the court may rule that they are in effect the same company.

Of course, my "forfeit the patent" idea is similar in that the value of what's given up is similar to the cost of the risk falsely put on the accused. But perhaps I should make it a bit weaker; perhaps I should have them forfeit all right to enforce the patent for the duration they falsely claimed the accused had been violating it, and it will be considered to be RAND-licensed thereafter -- so it'll still possibly be a future revenue stream, but not a big one.

(On the other hand, this scheme definitely has some odd possible abuses -- one could file a suit for a TINY violation period, perform discovery, and then expand the violation period depending on how bad the problem looks. But perhaps that's actually a GOOD thing, since innocents will be less abusable.)

The idea of the 1/3 damage bond would make patents almost completely useless for individual inventors without ruining their abuse for megacorps. If we MUST have patents, surely that's not the direction we want them to tilt.

David Friedman said...

With regard to my bonding idea ... .

If the plaintiff doesn't expect to win, we don't want him to sue. If there is good reason to think he will win, he ought to be able to find an insurance company to post the bond--for a price.

wtanksley said...

David, great point -- it was silly of me to forget that. If an inventor might be able to convince the court he should also be able to convince an insurance company. The devil's in the details, of course, but I take back my criticism.

Anonymous said...

WONDERFUL Post.thanks for share..more wait .. …

[url=]NFL Authentic Jerseys[/url]

[url=]jordans for cheap[/url]