In the Wake of Grokster...
...the Justice Department is conducting criminal investigations of file-sharing networks. This development illustrates a point I made in a previous posting (a Lessig point) about the relationship of substitution between law and technology. The Grokster decision last week, if it holds up, will facilitate circumvention of copyright law by file sharing, by placing the sellers of the software for such sharing beyond reach of the copyright law. The liability of the sharers themselves is not affected; and already as we know hundreds of them have been sued by the recording industry. But copyright law also authorizes criminal sanctions. The Justice Department has not yet indicated an interest in prosecuting individuals who download just an occasional copyrighted song. But it is sometimes possible to deter unlawful behavior by a very slight threat of prosecution. Economists have the useful concept of an "expected cost." If there is a 1 percent probability that you will incur a $100 cost, the expected cost is $1 ($100 x .01), and if you're risk averse, you will spend up to $1 to avoid it. If the expected cost is an expected cost of punishment, it may be very great even if the probability of punishment is slight: many kids will stop downloading copyrighted songs if they think there is a 1 percent probability that they will be sent to prison for 6 months. So we can think of what the DoJ is doing (whether you like it or not) as the law pushing back against technology--trying to defeat technological circumvention of law by jacking up legal sanctions, in this instance possibly to a higher level than the RIAA can achieve with its civil suits.
| Permalink | technorati
Comments (14)
I have two questions. One of them, I'll ask here, and the other later. Why did the DoJ make no arrests and only siezed equipment? How does one determine proper punishement for infringement of this type?
..except the risk isn't 1%. its unbounded. probably lower. for anybody young enough to ride a motorbike, or smoke, or consume soft drugs in a college dorm, its within the norms for a risky lifestyle.
so it won't have this effect. it might, if they lynch somebody or fry them in the chair, or throw eggs at them on the steps of congress.
its been illegal to purchase alcohol as a minor for a very long time but it didn't stop somebody very famous using false ID recently did it? and I'd say their risks of 'getting caught' were well in excess of 1%...
One important aspect of the raids that the NY Times article doesn't go into, but others have reported, is that the raids concentrated on the "hubs" of the network. Hubs, superpeers, etc. are critical components of many filesharing networks. Take out these crucial nodes and the network is disproportionately degraded when compared to the effects of removing a node from the edge of the network.
By focusing limited enforcement resources on hubs, law enforcement magnifies their effects. The expected costs for hubs are significantly increased, and the smaller number of hubs means the networks become increasingly unusuable much quicker.
As noted by Prof. Ed Felten some time ago, and as came up at a recent conference on P2P law at Seaton Hall Law School, the economic risk aversion model doesn't yet work for P2P.
The following is a serious oversimplification, but I think it makes an important point nevertheless:
Assume there are 25,000,000 file sharers in the US right now. (The number is probably higher).
Assume that the music industry sues 5000 people a year (which is more than they are suing each year right now), with an average settlement award of $5000 (which is higher than the publicly reported settlement amounts). The average file sharer has a 1/5000 risk of having to pay $5000.
Then, the per file-sharer risk is $1 a year. (As the always-insightful Mr. Miller points out above, for quasicentralized networks with hubs, supernodes, or central servers, this isn't accurate. However, given the popularity of new gen P2P technologies such as BitTorrent, and the continuing value of gnutella-type "flat" P2P systems, such quasicentralization doesn't entirely resolve the economic issue). This is not an incentive for an economically rational file sharer to stop.
Raising the fiscal penalty won't help since most file sharers simply won't be able to pay. Confescating their equipment similarly won't act as a penalty of more than $5000. Forcing file sharers (who substantially overlap with the media industry's best customers) into bankruptcy is not a useful long-term business model. And suing substantially more than 5000 individuals for copyright infringement in federal court would be a disaster for the already crowded district courts, especially in population centers like NY, Chicago and LA where file sharing is especially popular.
Criminal prosecution is a whole other matter. When used selectively it has deterrence effect. However, a widespread war-on-drugs style rounding up of kids, college students, and adult file sharers for prosecution will have anything but the effect the RIAA, MPAA, and DOJ expect. Just as the civil suits have created a small backlash, even among the IP community, widespread criminal prosecution risks a wholesale consumer backlash against the media industry, both economically and politically.
A political backlash against, of all things, copyright law, is the last thing the media industry wants or needs, as it could lead to disfavored solutions like mandatory licenses or, worse, legislative decartelization. If you don't this is something to fear, ask the pharmaceutical industry.
Copyright legislation over the last thirty years has been entirely a creature of interest groups, as the statute has expanded well beyond its pre-1976 size, and even seasoned IP attorneys are sometimes suprised by the non-intuitive breadth and theoretical scope of copyright liability -- and this is without the proposed INDUCE act, or the proposed database protection legislation (i.e. a Smoot Hawley tariff for the 21st century).
This leaves three options: technological measures such as DRM, central liability by ISPs (revising the DMCA takedown provisions to include P2P), or, another solution, creating an administrative standard licensing/enforcement system (i.e. speeding tickets for the internet) which attach a small but relatively severe administrative fine for illegal file sharing, with defenses available, in a small, speeding-ticket type setting.
"But it is sometimes possible to deter unlawful behavior by a very slight threat of prosecution"
Hasn't work for roughly a century with prohibition. It will not work here. The RIAA has only sued 4000 people out of 60,000,000. It's a joke. No one cares. I think a legitimate argument can be made that the RIAA is still sleeping on its members rights.
Now the DOJ can cultivate an image that they too are impotent by pretending to enforce unenforcible laws.
However, I am in favor of "jacking up legal sanctions." This is the only road to true reform.
I've never used the standard P2P software much. I don't like the way it works.
Instead, I've tended to use the server2server P2P system called Usenet, which works fine, and the computer does not get shut down in the middle of a download.
Why does P2P seem to get more ink/legal attention than Usenet?
How does Usenet differ from standard P2P legally?
"Why does P2P seem to get more ink/legal attention than Usenet?
How does Usenet differ from standard P2P legally?"
P2P gets more attention because it's more widely used.
IANAL, but I would think that uploading a copyrighted work to Usenet would be a more significant CR violation than P2P sharing, because some P2P is potentially protected under fair use -- for example using P2P to access your own music collection from different locations.
It seems at least possible that activities like file sharing have multiple equilibria. Assuming a fixed enforcement effort, as more people share files the expected punishment goes down, while the benefits probably go up (network externalities). If this is true, it is possible that there is one equilibrium in which everyone shares files, one equilibrium in which no one does, and one (very unstable) equilibrium in which a certain percentage of people share files. What is called for, then, assuming we prefer the no-sharing equilibrium, is a crackdown of huge proportions, effectively collapsing all the equilibria to a no-sharing one. Once this is achieved, minimal enforcement efforts should be able to maintain the equilibrium, as it is stable against small shocks.
Interesting ideas James, but you forget how much file sharing networks have grown. In size, and also in implementation. What happens when they use encryption, and small decentralized networks. Ones that are home written to all look and act differently. Or ones written with trust in mind, by invite only, like Google's Orkut service?
Here's a big idea, how about a virus that starts up a redundant hub? Since its redundant, the files can flow several different ways, and if the network is designed to quickly add and remove hubs, it will be almost impossible to shut down. If the virus spreads and and gets re-written to take advantage of new security holes, there will be no way to stop it. Bittorrent is a powerful idea in filesharing: it is redundant, and can quickly add and remove sources and destinations of files, with file checks to ensure proper data goes through, and resuming file transfer.
The genie is out of the bottle, and the more you try to push him back in, the more he tries to stay out, and the worse methods he employs.
Great stuff . . . This has been a wonderful set of commentaries.
Luka, Ernest and Mathew have made interesting technical points, which demonstrate the challenges posed to any police crack-down, and I certainly agree with Luka's critique of the ideal world analysis of equilibria offered by James.
Raoul and Doug make very important points regarding limits of enforcement in this case. In fact, my original question was intended to lead in that direction.
For more on this and also on why people don't just make decisions based on "expected utility" analysis, see Richard Thaler's The Winner's Curse: Paradoxes and Anomalies of Economic Life, Chapter 6 (Princeton University Press, 1992).
And thanks to Richard for having started this discussion. As a bystander, I'm thoroughly enjoying the conversation it has generated.
I've posted some brief comments on Copyfutures.
It baffles me that behavior in which millions of people engage everyday is criminalized. Perhaps the law is just plain wrong.
Remember, alcohol prohibition did not work and caused more problems than it was intended to solve.
We lose respect for law in general when unjust laws exist.
I don't think it should baffle you, since it was criminal all along, and people simply started "engaging in it everyday" on their own, disregarding the law...
The problem is so fundamental. It's simply that human beings cannot be expected to enforce the law on themselves. That is the whole basis for our governmental system. There has to be a structure in place that allows the law to be enforced by someone other than the p2p users, and the RIAA and other private parties are failing to do that, as is the DOJ.
If the identities of all internet users were known, things might be different. So government-mandated identification or authentication systems for the net might be one way to approach the problem. Certainly, that's what we have for the roads, in the form of licence plates.
Some argue that encryption makes the whole point moot, but I disagree. Encryption can save a small number of sharers, but as the group opens to more people, eventually the likelihood of discovery becomes great enough for a bust. So only encrypted networks under a certain size, or carefully managed, can survive, just like organized crime rings, assuming a reasonable level of enforcement.
Wniosek taki nasuwa się z dość jednoznacznej, choć wyrażonej nie wprost odpowiedzi na pytanie o dalsze zawodowe plany, jakie zadali mu dziennikarze po jednym z pokazów jego najnowszego filmu "A Very Long Engagement".