“The case against patents can be summarized briefly: there is no empirical evidence that they serve to increase innovation and productivity, unless productivity is identified with the number of patents awarded – which, as evidence shows, has no correlation with measured productivity. Both theory and evidence suggest that while patents can have a partial equilibrium effect of improving incentives to invent, the general equilibrium effect on innovation can be negative. A properly designed patent system might serve to increase innovation at a certain time and place. Unfortunately, the political economy of government-operated patent systems indicates that such systems are susceptible to pressures that cause the ill effects of patents to grow over time. Our preferred policy solution is to abolish patents entirely and to find other legislative instruments, less open to lobbying and rent seeking, to foster innovation when there is clear evidence that laissez-faire undersupplies it. However, if that policy change seems too large to swallow, we discuss in the conclusion a set of partial reforms that could be implemented.” Written by economics professors Michelle Boldrin and David K. Levine, published in the winter issue of the Journal of Economic Perspectives. Via John Siracusa.
One easy rule for limiting patents: if you can’t provide a technical drawing then it’s not patentable.
That won’t solve anything because the primary problem with patents today is that they are almost exclusively written by and for lawyers. Ever seen an engineer try to read a patent? They’re told they need to consult a lawyer to understand what it truly says… and that’s the problem.
Patents are supposed to be written so that anyone with average skills in the subject can reproduce the patented item. That hasn’t been the case ever since lawyers worked their way into the equation. So the first REAL change MUST be to make it illegal to have a lawyer work on a patent. If the average person in the field cannot read the patent, it’s invalid. No more “legalese” in patents.
That will cure 95% of the problems.
Patents must be written in working Python.
Using 7zip on the highest settings, compression should not reduce the size below 70% of the original.
Ironically, that would make them far easier to read, even for non-programmers.
Yes. And most sane people see the obvious problem with that – but none-the-less it is an intentional and essential ingredient in the US patent system. “Fixing” this makes no sense, because the parties with any actual say in reforming the system want it this way…
Its like the argument against fighting in hockey. Does fighting have anything to do with the actual game? No. Does the league try to stop it? No. Why? Because the fans like it, and the fans are the whole point.
Likewise, in the US Patent System, the lawyers are the whole point. We can all argue about the roots of patent law and what was intended by the framers, but its really too late – the inmates are running the prison.
The only way to fix it is to scrap it and start over. I am utterly convinced of that. I have read hundreds of good ideas regarding reform – yet the only ones that ever see the light of day are half-ass measures that often cause more harm than good. Its obvious at this point that reform has become a corrupted issue – those with any power to actually effect change only do so in half-hearted attempts to appease the rest of us.
Who will evaluate whether the drawing adequately represents the invention clammed by the patent? Why would this be more reliable than the current system of evaluating the text?
The English legal system already uses the concept of a hypothetical ‘reasonable person’. That ‘person’ is not expected to be real or present.
The law is used to having to deal with vagueries, so the idea that a reasonable person with relevant education or experience must be able to recognise that a patent properly and clearly represents the idea is entirely sensible.
No doubt the American system has a similar concept it could apply.
Patents are not granted by “a reasonable person”, they are granted by patent office clerks. The problem with bad patents is not that the legal system can’t cope with them. You can already invalidate them in court. The problem is that the litigation is expensive, often enough to put small players out of business.
Also you wouldn’t want to depend on “reasonable person’s” (i.e. layman’s) opinion on highly technical subjects, that are what technology patents are about.
You don’t seem to understand what is legally meant by “reasonable person”.
Feel free to enlighten me with links to credible resources.
http://en.wikipedia.org/wiki/Reasonable_person
If your from the US, our patent law has a similar concept…
http://en.wikipedia.org/wiki/Person_having_ordinary_skill_in_the_ar…
I think the overall point your missing is that the patent clerks job is to judge the patent based on these legal fictions… Their personal expertise in technical matters has nothing to do with it.
In other words, the patent clerk is supposed to put themselves into the shoes of someone having “ordinary skill in the art” when judging a patent – and I strongly stress supposed to. The reality is they pretty much just approve anything and let the courts work it out, which is most of the problem in itself.
Edited 2013-02-06 03:09 UTC
So how does this contradict anything I’ve said?
Decisions to grant patents are made by patent clerks. Why do you think they will be any better at evaluating drawings than at evaluating text, which they currently do? If a clerk grants a bad patent that affects you, your best bet at invalidating it is typically going to court.
Next, from your link
So this is a standard that jury can use to make a decision. In the context of this thread it’s not even clear who you apply this standard to. The patent clerk? He/she has no responsibility for granted patents. In any case, the decision is made by the jury (laymen), based on their understanding of what’s “reasonable.” The only way people with relevant skill and knowledge are allowed is as expert witnesses, and both sides are sure to bring their own to argue their way.
You’re contradicting yourself. There’s no way one can put themselves into the shoes of a person having relevant skills without actually having them. You cannot answer the question “Would this be obvious to me if I had ordinary skill in the field?” without the skill.
But this is not the way patent clerks work or are supposed to work. Patent clerks are experts in their subject patents. They basically treat them as text. For every new patent they find claims in previous patents that read similar to new claims, and ensure new claims are worded sufficiently differently. They do some sanity checking, like rejecting claims that are obviously too broad, but given they are not experts or even practitioners in the field, there’s not much they can do. I posted about this before: http://www.osnews.com/permalink?525537
Mostly because you said this:
This may or may not be correct, but you don’t seem to understand what is meant in law.
You don’t seem to understand my comment. The point is that it’s not clear who the “reasonable person” standard is supposed to be applied to. But in any case it’s too late: most of the harm from a bad patent is done before it goes into trail.
It doesn’t. You ask for relevant links… I posted a few.
If the system was actually designed to work, then yes – the patent clerk. But keep reading…
If I was unclear… I don’t mean the clerk should not use their own technical expertise. What I mean is they should judge the patent through the scope of one who is “ordinarily skilled in the art”. Exceptional talent at deciphering legalese is not generally relevant nor should it be – yet that is nearly the sole area of experience with the average patent clerk. If it is a software patent one ordinarily skilled in the art should be able to read it – i.e. someone skilled in software should be reviewing a software patent…
Yes, that is true. It is also part of the problem. If the standard of the law is that a patent should be useful to someone ordinarily skilled in the art, and if it is judged in such a way in court (that is what the law says), then it is reasonable to expect that it must meet this requirement to be granted in the first place. But it absolutely is not… That is why I strongly stressed that it was how I thought it was supposed to work.
It obviously doesn’t work this way and frankly I don’t know how it could without putting undue burden on the patent clerks. I really don’t have a solution to the problem – just pointing it out.
Exactly. It’s not reasonable to expect PTO to have people skilled in every field they grant patents, though. So you need switch from clerk reviews to peer reviews.
Edited 2013-02-06 13:43 UTC
The link didn’t work for me. This did:
http://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.27.1.3
Patents are useful, but companies abuse them because they are relatively cheap to obtain and maintain. So maybe we should make them expensive to obtain and maintain.
I propose a patent tax system. This is how it would work.
The first 10 patent applications made by an individual or a company would attract a nominal “tax” of, say, $100 per patent. So the system allows an individual to have 10 cheap patents per year.
The next 90 patents are a bit more expensive. These would cost $1,000 each. This would allow smaller companies to protect their invention without making it too expensive.
Any other patent costs $100,000 to file.
These fees would also apply when acquiring patents, so if either filed or bought 10 patents in a year, I would pay the government $100 per patent. The next 90 patents bought or filed would incur taxes of $1,000 and the rest would attract the $100,000 tax.
Suddenly, companies would really think twice about submitting 5000 patent applications a year (IBM, I am looking at you!).
A similar tax system could be developed for patents you own. So each patent generates a licensing fee to government. The first 100 patents are license free ($0). The next 900 cost the company (or individual) $2,000 per year (each), and the rest cost you $10,000 per year.
Anything that resembles ownership would be treated as ownership, therefore a company couldn’t get away by creating a subsidiary that only owns 10 patents. The owner of the company would be deemed to own those patents, and therefore would be liable for all patents owned directly or indirectly. Exclusive licenses to a patent would be treated in the same way as ownership, with a an initial fee for the exclusive license, and further annual fees for each year in which the patent remains exclusively licensed.
Probably flawed, but this might solve the patent mess.
I’ve actually heard variations of this one before… The problem is how do you incorporate such an idea into a system where there are already a few million active patents in existence?
IBM has something like 40,000 active patents right now. Do you make them pay license for them? At the rates you stated that would come up to 390 million and some change… a year. Of course it would taper off over 20 years, but it doesn’t seem fair to financially punish a company for a bunch of patents they filed for legally in the past.
So you exempt all existing patents and only do it going forward. Now IBM, Microsoft, and others who hold large patent portfolios have an artificial advantage that lasts for 20 years – because smaller players just starting out have to pay huge amounts of money for the privilege of having more than 100 patents – the big guys got there huge portfolios for free – and they will probably use them aggressively. Doesn’t seem fair…
Maybe a balance could be struck… Really though, my main issue with it is it is a form of “sin tax”. It is a tax like those applied to tobacco – the point is not to generate revenue, but to limit a behavior. I just find it ironic that the patent system is so broken taxing it to disuade it’s use sounds like a good idea…
Im still in the “gut the whole thing and start over” camp…
The patents system is there to promote progress, and I think it does meet some of its goals. I know patents get a really bad rep in the software circles because the dominant image of them is of them being trivial – the sort of things that someone could knock out lots of in no time.
The idea is to try and encourage, or force, companies to only patent things they genuinely believe are worth patenting, and want to put in production. By graduating the cost, companies will have to decide which of their inventions they want protection for. If a company has a single patent on a very popular and good drug, $50k per year is chump change. What it ought to prevent is a company submitting 10,000 patents just so every rival infringes on some obscure patent that the company doesn’t think is really worth much, but allows them to extract economic rents from other players.
I wouldn’t exempt existing patents. I might impose lower costs for existing patents – basically grandfather them in at lower rates. However the objective should be to prevent hoarding of patents whose only use is to prevent competitors from trying to compete at all.
One tweak that could be added would be to allow companies to turn over patents to the public domain and thereby relieve themselves of the cost of keeping patents they have no use for. So a company could either sell off patents that it has no use for, or could turn them over to the public domain.
Millions of patents is a bad thing!
It was to wrestle power away from the guilds. Unfortunately, the guilds are back in the guise of corporations.
“Don’t put your trust in revolutions. They always come around again. That’s why they’re called revolutions. People die, and nothing changes.”
— (Terry Pratchett, Night Watch)