The Morality of Copying and Copyright
I’ve been doing lots of thinking about economics this past few days. I really ought to blog up a storm on it, and perhaps I will, if I don’t get too tired out talking with students about their term papers. First: copying.
Before the Talk. Today I present, informally, on my copying paper in the Bus Econ Tuesday lunch. The idea is to ask what norm would be efficient for dealing with copying and copyright. If we could somehow instill guilt and self-satisfaction into people depending on their copying behavior, how would we do it? Would we, for example, make it so that a person would feel as guilty about downloading a copyrighted song as he would about shoplifting a CD from a record store?
You could try to answer this question using any system of morality as a starting point. For example, you could take as a premise that downloading copyrighted songs is intrinsically evil, and decide whether you want to make the norm strong enough to stop this evil behavior.
There are three difficulties, though. First, our moral capacity is limited. Human nature makes it infeasible to instill the same amount of potential guilt for burping in public as for murder. Thus, we may not want to use enough of that capacity on minor evils such as copying so as to deter it altogether. Second, we will lose some human happiness as a result. Even if copying is intrinsically evil, there are situations in which it causes nobody harm and makes the copier happier. There has to be a strong reason to hurt people just to avoid the intrinsic evil. Third, it is very hard to show that downloading copyrighted songs is intrinsically evil.
Let me expand on that last point. There are a number of moral systems you might be using. You might be Christian, or Moslem, or a Stoic, or believe in following the moral intuitions of human nature, or a Rawlsian liberal, or a Kantian. None of those systems can deal very well with music copying. It is not mentioned in the Bible, and it is not hardwired into human nature by God or evolution. All of these systems include a command of “Do not steal”, but that command begs the question– what does “steal” mean? We could write copyright law to allow people to download songs for personal use or not. It’s not obvious which way is morally right by any of the systems I have mentioned. In fact, until relatively recently, songs had no protection whatsoever under copyright law, and not more than a century or two before that, there was no copyright law even for books. Thus, I don’t think moral systems give much guidance to whether we should categorize music downloading as stealing, however much guidance they give as to whether a person should steal or not.
What I will look at, therefore, is the idea of value maximization, or “economic efficiency” as it is typically but misleadingly called. What norm would make people happiest, with happiness being defined in terms of satisfying their desires given that the world has finite resources?
Let us put aside the issue of what to do about copying that can be stopped by laws, as opposed to norms. It is feasible for a law to prevent someone from opening up a store to copy and sell copyrighted books without royalties, for example, because if such an operation were to be profitable it would have to be big and public enough to be detected by the police. A law against a person borrowing a friend’s CD and copying a song from it onto his computer, however, would be impossible to enforce. (Note that there does exist such a law: you can be sued and fined, though not imprisoned , if you copy a friend’s CD [U.S. Code 17-106-1, 17-504, 17-506, 18-2319-b-3])
Since it is easy enough to enforce much of current copyright law, we need not consider the use of norms there. To be sure, norms are often useful even as an addition to law, but commercial copyright violation is so each to catch that a norm against it would be less important than, say, a norm against burglary, a much harder crime to detect.
Here are a few principles.
1. There should be no guilt over use by copying that would not have occurred if the person had to buy it, because purchase is so troublesome.
Suppose it costs $3 in time and convenience to figure out how to buy the song, and only $1 to copy it without purchase, and that the song is only worth $2 to the person. Copying is efficient and purchase is inefficient.
2. There should be no guilt over use by copying that would not have occurred if the person had to buy it, because the price is so high.
The price might be high for two reasons. First, sale is more costly to do than copying, because it requires funds transfer as well as the copying of the song, so a price equal to the marginal cost of selling would still be greater than the marginal cost of copying. Second, the price would be higher than marginal cost, because the seller has a monopoly.
Suppose the time cost of copying is $1 but the time cost of a sale is $2 for the buyer and $4 for the seller, a total cost of $6. Suppose furthermore that the seller charges a price of $8, but this particular person’s value for the song is $5. The sale would not occur, and would not occur even if the price fell to the seller’s marginal cost of $4.
3. Maybe there should be guilt if the person’s value for the song is high enough that he would buy the song if copying were not possible.
Suppose the cost of copying is $1 and the transaction costs of purchase are $2 to the buyer and $3 to the seller, that the price is $15, and the consumer’s value is $30. Then we want there to be copying if the marginal effect of the $12 profit to the seller has a big enough incentive effect on creation of new works and if the opportunity cost of using up moral capacity here is low enough.
4. There should be guilt if the cost of copying is enough higher than the transactions cost of purchase.
Suppose the cost of copying is $9, because it is hard to defeat the copy protection. The transactions cost to the buyer of buying, though, is only $1, and the transaction cost to the seller of selling is $2, a total of $3, and that the seller is charging a price of $8.
All of these are single-consumer examples. In practice, some people will violate and some will not. Thus, guilt will be incurred. Is this an additional cost?
Also, actual violations may use up moral capacity in a different way.
Also, it may be that we do not value utility generated by norm violation, but we do value utility generated by compliance. Getting rid of the norm violation converts all that consupmtion utility into socially valuble utility. (the rape analogy)
If so, we want to avoid a situation where there is lots of the illegal activity.
After the Talk. The lunch went very well. These lunches are talks of the One Equation, One Regression, or One Graph variety, with the audience doing more talking, in aggregate, than the speaker. I got lots of good comments including:
1. A survey found most people thought renting a video and then copying it was wrong. This is probably the result of seeing years of FBI warnings at the start of videos. The movie industry has consciously tried to instill a norm, and has succeeded– but perhaps with a small guilt value g. This instillation has been costly. I should compute out the length of each warning, the number of viewings, and the cost per minute of viewer’s time.
2. There was an experiment done in a parking garage to see how the existing amount of trash affected people’s willingness to thrown down an ad leaflet put on their windshield. They threw down the least if there was not zero but one leaflet already on the floor.
3. I should think about how violation affects the size of guilt, g. Maybe it is psychologically impossible to keep g positive if 90% of people are incurring it daily.
4. How does guilt differ from a prison term? Both are costly punishments. But using prison means that increasing the probability of catching violators is costly, whereas for guilt it is costless.
5. I do not include shame in the model– the disutility of somebody else seeing me copy. That is a topic for another day.
6. A more important reason for giving profits to creators, probably, than incentive for creation is incentive for dissemination. The market works to generate the information that a song is good, and the owner responds by marketing it more. That is lost if there is too much copying.
7. Larry Lessig’s books Free Culture and The Future of Ideas might be relevant, as is an old 1970s Harvard Law Review article by Stephen Breyer suggesting that copyright be abolished altogether. I should of course cite Demsetz and Posner on primitive property law.
8. I can think of private advertising as attempts to affect guilt g and self-satisfaction s. An ad for organic food could try to make me fearful of not buying it, or satisfied for buying it.
9. Is checking a book out of a public library immoral? That is the same kind of question. It is legal, certainly, but it is also true that it hurts authors.
10. On the other hand, sometimes the creator gets a benefit from copying even though no price is paid. He gets free advertising, or network externalities. I should add this to the equation and to the considerations.
11. Maybe I should put the heterogeneity in the cost of copying rather than moralness. If I have two dimensions of heterogeneity, I can generate all 3 behaviors– non-use, copying, buying– in the same equilibrium.