Episode 412: How To Fix The Patent Mess

Oct 23, 2012
Originally published on October 24, 2012 8:47 am

Two big patent cases this summer in the smartphone industry:

1. A jury finds that Samsung violated Apple's patents, and orders Samsung to pay Apple $1 billion.

2. A judge throws out a case between Apple and Motorola (now owned by Google). The judge goes on to write an article in the Atlantic arguing that there are too many patents in America, and lots of industries could probably get along fine with no patents at all.

These radically different rulings were just the latest reminder that the world of software patents is a mess. Big companies that should be focused on inventing the next great thing are instead spending billions of dollars buying up patents and suing each other. Small companies have to worry that someone with some random patent is going to sue them and shut them down.

On today's show, we talk with Mark Lemley, who has some ideas for fixing the patent mess. Lemley is a professor at Stanford law school and an expert on software patents. Lemley also works for clients in the private sector, including Google.

For More: See When Patents Attack, our big patent story from last year, and The Case Against Patents, a paper by the St. Louis Fed.

Download the Planet Money iPhone App. Music: Fillagar's "Guilty Good Intentions." Find us: Twitter/ Facebook/ Spotify/ Tumblr.

Copyright 2018 NPR. To see more, visit http://www.npr.org/.

CAITLIN KENNEY, BYLINE: Siri, are you ready to record?



COMPUTER-GENERATED VOICE: Hello and welcome to PLANET MONEY. This is Siri. I don't know the answer to everything, but I do know this. The best way to listen to PLANET MONEY is to download their new iPhone app. You can get it in the App Store. It's awesome. On with the show.


The world of software patents - by all accounts, it is a mess. Even computer programmers think so.


STEPHAN BRUNNER: I have to say I actually worked on a whole bunch of patents in my career over the years. And I have to say that every single patent is nothing but crap.


This is a guy named Stephan Brunner (ph). Our colleague, Laura Sydell, found him at a park in San Francisco where tech people hang out at lunchtime. Stephan, he says his patents don't even make sense to him.


BRUNNER: I can't tell you for the hell of it what they're actually supposed to do and what they did because I did not - the company said we have to do a patent on this. And then they send in a lawyer. I, personally, when I look at them, I'm not proud at all because most of them, again, it's just, like, mumbo jumbo which nobody understands and which makes no sense from an engineering standpoint whatsoever.

KESTENBAUM: Brunner was part of a great story that Laura and our colleague Alex Blumberg did last year about patents. And since then, some huge things have happened. This summer, there was a big lawsuit between Apple and Samsung over patents. And the jury decided that Samsung had violated some of Apple's patents, and it awarded Apple $1 billion, a billion dollars.

GOLDSTEIN: One billion dollars. But there was also another case this summer, one that we didn't hear so much about. And it ended up very differently. It was also a patent dispute, this one between Apple and Motorola. And in this case, the judge threw the whole thing out. And this judge, a very well-known judge named Richard Posner, he felt so strongly about the whole issue that he actually wrote a piece in The Atlantic. And the piece was titled, "Why There Are Too Many Patents In America."

KESTENBAUM: So you had two similar cases, radically different rulings. It's one more reminder that the world of software patents is a mess. Big companies that we want to be making, say, the next awesome smartphone, they are spending a lot of money - billions of dollars - just buying up patents and basically hitting each other over the head with them, suing each other.

GOLDSTEIN: And small companies - the companies that we count on to come up with the next big breakthrough, those companies have to worry that someone with some random patent is going to come and sue them out of existence.

KESTENBAUM: Today we talk with someone who thinks he knows how to fix everything.


KESTENBAUM: Hello and welcome to PLANET MONEY. I'm David Kestenbaum.

GOLDSTEIN: And I'm Jacob Goldstein. David, the guy who thinks he knows how to fix all this, his name is Mark Lemley. You and I spoke with him recently. Lemley is a law professor at Stanford. He wrote a book about how he thinks the patent problem can be fixed. And we should note here he has done some private work for Google, which is obviously involved in a lot of this patent stuff. But he hasn't worked for their smartphone business.

KESTENBAUM: We began by asking him to run through how we got into this mess in the first place because a few decades ago, it turns out, patents on software were not allowed. Basically, a patent had to be for something physical - something, as someone said, that you could put in a bag and shake it, and it would make noise. So we asked Lemley what we thought was a simple question - when was the first software patent?

MARK LEMLEY: Well, because the court said you can't patent software, that actually turns out to be a hard question to answer. What patent owners did is they patented software by pretending it wasn't software. If I had a computer program that controlled the temperature in a pizza oven, I would write a patent claim to a pizza oven. And sort of buried in that patent claim is the fact that the only new thing about my pizza oven is the computer program I use to control it. So we saw software patents, I think, in that form in the '70s and '80s, but they tended to be disguised.

KESTENBAUM: Lemley says this went on for a while. And then in 1998, finally, the courts threw up their hands and said, you know what? This doesn't make any sense. You can patent software.

GOLDSTEIN: This decision from the courts, it basically opened the floodgates. All of a sudden, people were going to the patent office with all kinds of software ideas. And the patent office basically said, OK, you can have your patents. We have seen in the last several years an explosion in the number of patents being granted that dwarfs any previous period in our history.

LEMLEY: From the founding of the republic, it took us over a hundred years to get to the first million patents. And that's a hundred years that included some pretty fundamental innovations from the railroad to the electric light to the telephone and the telegraph. By contrast, now we're granting a million patents every four years. And maybe it's the case that we've got so many more great innovations that we need to be granting patents at more than 25 times the rate we were then. But I'm skeptical.

KESTENBAUM: We should point out not all of these new patents are software, but a huge chunk of them. He says a third to a half, he estimates.

GOLDSTEIN: And to understand why this is such a big deal, it's useful to remember just how powerful patents are. A patent grants the owner a monopoly over something for 20 years. Anyone else who wants to use the idea that's in the patent has to get the owner's permission. And this is basically the opposite of the way a lot of our laws work, right? A lot of our laws are about preventing monopolies, about breaking up a monopoly. Patents - they are the very rare example where the government is saying, here, we're going to give you a monopoly.

KESTENBAUM: And what are people getting monopoly power over? Here, I give you patent number 5132992, quote, "audio and video transmission and receiving system." Here I'll read you part of it. (Reading) The information is sent over a standard telephone cable or satellite broadcast channels to a receiver specified by a subscriber...

GOLDSTEIN: David, let me just jump in here and say, basically, what we need to know about this patent is it covers on-demand streaming of basically anything - video, movies, TV, music. It's a very broad patent. And Lemley says there are lots of broad patents like this one.

LEMLEY: Patent on the concept of electronic commerce, a patent on the concept of voice over Internet that's not limited to a particular invention. It's - it covers any implementation of those ideas.

GOLDSTEIN: So just to be clear, it's not a patent on some particular way to say talk over the internet, voice over IP. It's a patent that covers the very idea of talking over the Internet so that nobody else could compete against you.

LEMLEY: Right. And I think that's the problem. We've granted way too many patents that seem to assert that I own the area itself.

KESTENBAUM: So why were these patents granted? Well, the logic has been, look, you know, the big idea is sending video over the Internet. The details of how you do that - the argument is - you know, that's trivial. That doesn't matter. Like, some monkey could write the code.

LEMLEY: There's language from a federal circuit opinion that says it would be a mere clerical function to a skilled programmer to sit down and actually write the code. And I just think that's wrong.

KESTENBAUM: OK. So say you agree with Mark Lemley that this is a problem. What do you do to fix it? Do we hire a bunch of new people at the Patent and Trade Office? Do we need to write a new law? Lemley says, no, actually, don't need to do any of that.

LEMLEY: So we could write a statute to try to narrow it. But I think the answer is we've already got a statute that does that or is supposed to do that.

KESTENBAUM: What is that statute?

LEMLEY: It's Section 112 of the Patent Act and what it says is...

KESTENBAUM: Oh yeah, 112, yeah.


KESTENBAUM: What is 112?

LEMLEY: What it says is that I am entitled to a patent only if I give you a written description of my invention and enable people to make and use it without undue experimentation. In other words, you cannot just go to the patent office and say, hey, I've got this idea; I'm going to stream movies over the Internet. You have to tell the patent office exactly how you're going to do that, and what you get a patent on is the way of doing that, not on the very idea of streaming movies of the Internet. That's how the system is supposed to work.

When we gave Samuel Morse a patent on the Telegraph, he actually asked for - in addition to his narrower patents on the actual Telegraph he had built - a patent on the concept of using electromagnetism to communicate information. A patent which, if it had been allowed, would have incorporated, later, the telephone, the fax machine, the Internet itself. The Supreme Court said no. They said, for all we know - this was 1854 - later on, somebody's going to come up with a different way of using electromagnetism to communicate at a distance, and sure enough they did. You're not entitled to own all ways of doing it. You're entitled to own the way you actually did it.

KESTENBAUM: All right, so Mark Lemley rule number one on how to fix things. They should actually enforce - what is it? What part of the Patent Code is it?

LEMLEY: Section 112.

GOLDSTEIN: David, I just have to add a little caveat here - to be clear, a patent is good for 20 years so even if Morse had got the patent he wanted, it wouldn't have covered the Internet or the fax machine or whatever.

KESTENBAUM: Though, his point is that, you know, if - even if you had the idea to do work on, say, the telephone, like, you wouldn't have be able to do it if he'd been granted that patent.

GOLDSTEIN: Fair enough. No, it is a good point. And so let's just nail down Mark Lemley's rule number one - no overly broad patents. OK, so now let's get onto what we're going to call, logically enough, Mark Lemley's rule number two - no patents on things that are obvious. And this, too, is actually written into the rules of patents. You are not supposed to get a patent on anything that would be obvious to somebody who works in the field.

KESTENBAUM: But he says this is a case where the patent office has taken a very lenient view for a long time and granted a lot of patents on things that sometimes look pretty obvious. Now, Lemley says there have been some encouraging developments. There was a case in the Supreme Court that weirdly had to do with gas pedals. Some inventor had gotten a patent for combining these two things that already existed. There were adjustable gas pedals. And there were gas pedals that used sensors. And this guy made an adjustable gas pedal with a sensor. And the Supreme Court said, basically, that is obvious. We've had these two things separately. Anybody in the business trying to solve whatever problem you were trying to solve would have thought to combine them. Big deal, it's obvious.

LEMLEY: It's not worth granting a patent to somebody because everybody would come to it anyway. It's had some impact on software patents. We've seen more invalidations for obviousness. I don't think it has fundamentally changed the problem, in part, because there were just so many software patents out there. We've got upwards of a million software patents out there.

KESTENBAUM: And this gets to, possibly, the biggest problem. You've got this huge number of software patents. I mean, imagine a list of a million patents. And some number of those are too broad. OK, so let's say we go through and cross those out. Some are obvious. Let me cross those out, get rid of those. After you've done that, he says, you still have an enormous, enormous number of software patents that are arguably valid.

GOLDSTEIN: And this itself is a huge problem if you want to, say, make a smartphone. Just think of how complicated and amazing a single smartphone is. It's a camera. It's a movie camera. It's a computer...

KESTENBAUM: It's a phone.

GOLDSTEIN: It's a phone. It's not a great phone. But it knows up from down, and it knows which way is north. And it knows where you are on the globe. I mean, it is filled with clever little ideas. It's filled with all these tiny inventions. A single smartphone will draw on an enormous number of legitimate, valid patents.

LEMLEY: David Drummond, the chief legal officer at Google, says smartphone, all in, including the apps, might incorporate 250,000 patents.

KESTENBAUM: Which is a lot if I want to just - say, I want to make a smartphone. To start with, I'm going to need to get permission from 250,000 different patent holders potentially.

LEMLEY: Exactly. And that's - and that creates a set of problems that weeding out bad patents isn't going to solve entirely and that even sort of narrowing the scope of patents isn't going to solve entirely. You've still got the stacking problem. You've got to stack up a bunch of different royalty payments to a bunch of different people, each of whom have claims to a small piece of your product.

KESTENBAUM: And so one guy could just hold out and be like, I know everyone else has given you rights to use their patent, but if you want mine's, it's going to cost you $1 billion.

LEMLEY: Exactly. Cristina Mulligan at Yale has estimated that if you actually got every software company to evaluate every potentially relevant patent and get their lawyer to advise them, even if the lawyers did it quickly, you would need 2 million patent lawyers in the United States doing nothing other than reading and evaluating software patents.

GOLDSTEIN: We do not, for the record, have 2 million patent lawyers in this country. And what has happened instead is this patent arms race among smartphone companies. Lemley estimates that in the last two years, big smartphone companies have spent $15 billion just buying up patents to use in lawsuits against each other. This is bad, Lemley says. It makes our cell phones a little more expensive, and it discourages people from starting up new companies because they're afraid they'll get sued.

KESTENBAUM: So what do you do about this problem? The answer to this, Lemley says - let's call it rule number three - courts need to put patents in perspective. If someone is holding a single patent and sues saying, you didn't get my permission in your cellphone, the court should keep in mind that that is just 1 of 250,000 patents in the cellphone. Probably, it isn't that critical. I mean, that patent owner should not be able to say - get an injunction and shut down everyone's cellphone that uses the patent.

LEMLEY: We are, I think, making some progress towards this. We used to say if you win your patent case, you get an injunction automatically. We changed that rule in 2006, and I think properly so. So I can't just shut down your whole product no matter how small a piece my patent covers. We've got to consider how much of a hardship that would be, whether I really need an injunction and the like. We are in the process of doing the same thing with patent damages, trying to take greater account of the fact that if there are 250,000 patents out there that I might have to pay for, your one patent can't be entitled to 5 percent of my total revenues.

KESTENBAUM: Jacob, talking to Mark Lemley made me wonder. You know, is there is just a fundamental problem with software and patents? I mean, it seems like there's broad agreement you need patents in something like the drug industry, right? Otherwise, who's going to invest hundreds of millions of dollars trying to cure some disease if the moment you start selling your drug, some other company instantly can just make the pill and sell it for five cents each?

GOLDSTEIN: Software, though, it's not like that, right? I mean, software companies don't typically put hundreds of millions of dollars into a single piece of software. Judge Posner - the judge who threw out that patent lawsuit between Apple and Motorola - in his article for The Atlantic, he wrote, quote, "most industries could get along fine without patent protection."

KESTENBAUM: It's not just Judge Posner. A paper just came out by some economist at the St. Louis Fed pointing out that Apple, just by being first with the iPhone, it captured a huge section of the market. So there's still a strong incentive to innovate even if you don't have the monopoly power of a patent. I mean, like, imagine software patents were not allowed, right? You really think we wouldn't have the iPhone? Would Steve Jobs just said, no, I don't think I'm going to do it? So we asked Lemley, could we just get rid of patents for software?

LEMLEY: I think it's a hard question. I think there's pretty good evidence that patents do harm in software than they do in other industries. And the evidence that they benefit us is dubious.


KESTENBAUM: That seems like, yeah, let's get rid of it.

LEMLEY: That's a lawyer's answer to the question. The reason I hesitate is we just don't know exactly what the world would look like without software patents. So I'm a little hesitant to say let's just get rid of them. In any event, I don't think we can. I think that ship has long since sailed.


FILLIGAR: (Singing) It's guilty good intentions. My good intentions guilty. All this jealousy has got the best of me. It's big bad jealousy. Jealousy has been eating at me.

GOLDSTEIN: We will put links on the blog to Mark Lemley's book, to the St. Louis Fed paper we mentioned and to Alex and Laura's great story from last year about patents. That'll all be at npr.org/money.

KESTENBAUM: You can send us email. We're planetmoney@npr.org. I'm David Kestenbaum.

GOLDSTEIN: And I'm Jacob Goldstein. Thanks for listening.


FILLIGAR: (Singing) The boy who cried wolf is innocent. Would you believe me if I said the boy who cried wolf is innocent? Well, then won't you pin it on good intentions? Transcript provided by NPR, Copyright NPR.