Economics of Intellectual Property

By jrincayc in Media
Tue Jan 08, 2002 at 04:48:36 PM EST
Tags: Software

Information wants to be free. Free copying of music will destroy music. Those two statements don't fit together very well. Discussions of intellectual property and its ownership often come up on the Internet. However, the basics of the economics of intellectual property are rarely understood, by either side. Economics supports both arguments and neither argument.

First, I will explain in detail how economics works with traditional physical goods. I want you to understand trade so we can properly examine trade in the context of intellectual property without using analogies. Then I will describe trades in the context of intellectual property and problems with maximizing society benefit with intellectual property. Lastly, I will discuss possible solutions to the problems of maximizing society benefit with intellectual property.

Surplus

The basic idea used to evaluate the benefit of a trade in economics is surplus. Consumer surplus is how much more the consumer values the good than they have to pay for it. Producer surplus is how much more the producer can sell the good for than they had to pay to produce it. For the purposes of this article, producer surplus can be considered the same as profit. Summed together they are called social gain . Here is a simple example:

In other words, to figure out if a trade is economically beneficial, you calculate the consumer surplus and add it to the producer surplus. Or you take the value to the consumer minus the cost to produce. Of course things get more complicated when there are more people involved. For example:

Notice none of the additional people in this example contributed to social gain. If Bob and Dave had traded one bushel for $45 and Alice and Carol had traded one bushel for $10 then producer surplus would be ($45 - $40) + ($10 - $10) = $5 and consumer surplus would be ($45 - $45) + ($15 - $10) = $5 with a social gain of $10. Not coincidentally this equals $35 + ($15-$40) since $-25 is the social gain if Carol and Dave had traded in the above example. Since the surpluses of this trade are negative it will not happen as long as everyone has the freedom to choose trading partners. Alice can underbid Dave so if there is only one trade it will be with Alice. Bob can outbid Carol so if there is only one trade it will involve Bob. (Of course if you put Alice and Carol on one side of a concrete wall and Dave and Bob on the other, then Alice would trade with Carol, and Dave would trade with Bob, and the concrete wall would have lowered the social gain to $10.) Surpluses allow two things:

  1. A method to determine if a trade will occur since trades will only happen when consumer surplus and producer surplus are not negative (CS >= 0 and PS >= 0).
  2. A method to determine how beneficial a trade is to society since it is the consumer's value minus the producer's cost. (SG = CV - PC). *

* Note that this assumes that all costs are taken into account. If there is some additional cost that is not paid by the producer or the consumer then a trade might still be harmful to society. This caveat would apply to things like burning a ton of coal with the attendant environmental costs.

Another important point is that social gain is maximized in this example merely by each of the individual parties trying to maximize their individual surpluses. This is an important point since freedom to choose trades and individual maximizing surpluses will maximize social gain as well as an all knowledgeable benevolent dictator could. Social gain is better maximized by selfish individuals with limited knowledge than altruistic individuals with limited knowledge. Most markets operate this way. Intellectual property markets are not one of the market types that maximize social gain with selfish individuals with limited knowledge.

So far I have done all the examples with money. If this bothers you, it can also be done with other scarce quantities. The most common alternative ways use time or energy. It is generally harder to do. An example would be:

Social gain can be calculated without ever using money. In this case the social gain comes from the fact that Alice and Bob take different ratios of time to grow beans and potatoes. Social gain is usually calculated with money for reasons of simplicity. Other factors like energy or time are then converted into money to do the calculation.

Hopefully by now you understand the concept of surpluses quite well. I will give you two more examples to help bridge the gap between normal goods and intellectual property.

This first bridging example talks about some of the details of social gain when combined with entry costs.

The big thing to notice here is that sometimes there is a large cost associated with entering some market. This can sometimes cause the initial producer surplus to be negative. The trade may still take place if the total producer surplus is positive with more trades.

One more example, this time dealing with theft.

Theft of physical property is bad since it almost always produces a negative social gain, creates negative surpluses, and decreases future production and consumption. That is why there are laws regarding theft. These produce enough of a disincentive to keep theft at a low level.

And then there was Intellectual Property

Now that we have all these wonderful new pieces of jargon, let's look at another example, only this time it involves the spiffy piece of software called Baz:

And thus society gains since social gain is positive. Note that producer cost is very different in this example. The first copy of Baz costs $25 dollars to produce. The rest of the copies cost $0 to produce. The first copy of Baz sold has negative social gain since the producer cost is $25 and consumer value is $15 so social gain is -$10. This later evens out when more copies are created. As with the carpet example it is important to look at the total producer cost and compare it to the total consumer value to determine the social gain. This feature will have important implications.

But wait, Dave is a pirate

And thus society gains even more because Dave pirates the software. Objection. Dave pirating Baz is bad since it does not increase Trent's income (producer surplus) . True, but it increases consumer surplus and so is beneficial to society. Pirating can never decrease social gain directly since the cost to produce that copy is $0. Since the producer cost is $0 the short term social gain of piracy is always positive (PC = $0, CV > 0, SG = CV - PC = CV so SG > 0). In other words, pirating software that would not be bought is economically beneficial.

But wait, Alice is a pirate

And thus society gains less because Alice will pirate the software thus causing Trent to not produce the software since the producer gain is too little. Objection. Trent will produce Baz anyway out of the goodness of his heart. Maybe, but, first of all, businesses don't have hearts, and also Trent probably has other goals like getting food and writing another program. (Trent will probably not care as much about social gain and so will work on what he wants to write, instead of what is more needed. (Ever looked at how many IRC clients and bots there are listed on freshmeat ?)) So piracy can harm society when the software would otherwise be purchased but the producer never produces the software in the first place since expected piracy levels are too high.

The Double Edged Sword of Piracy

If there is no piracy and the price is fixed, then there will be lost social gains because the software producer will not sell to people who value the product less than the price (i. e. the price is higher than they are willing to pay). Giving the consumer the product would increase social gains since the cost to produce is less than the consumer's value of the software. In other words, social gain is not maximized when the price of some intellectual property is greater than zero (when PC = $0 and SG = CV - PC then SG = CV so each time the intellectual property is used SG increases).

On the other hand, if there are no restrictions on piracy then some software will not be produced that could increase social gains. One of the earlier statements of this is Bill Gate's open letter to hobbyists . No restrictions on piracy harms society since software that could increase social gains will not be produced as the producer surplus is negative.

In a nutshell, piracy always increases social gain in the present and may decrease social gain in the future.

So the problem is that both rampant piracy and no piracy harm society. There are no practical perfect solutions to this problem.

Possible Solution 0: All Knowing Benevolent Dictator

Let's play an economist trick. Assume that there is an All Knowing Benevolent Dictator (AKBD) who knows how much a given piece of software will take to write and how much each subject in the country values the software. Now we will revisit the previous example with this individual.

The important thing to notice is that the AKBD has maximized the social gain. Since the AKBD knows the producer cost and the consumer value the AKBD can create any project that increases social gain. Also since the AKBD knows how much each consumer wants the product the AKBD can choose some scheme to divide the cost among the consumers so that none of them pay more than they would be willing to pay.

Of course, this solution has a serious problem. There is no AKBD. The knowledge needed is difficult to find out. Finding the consumer value is tricky. When consumers are required to pay based on their own estimation they will tend to underestimate so that they pay less. This is even worse with large numbers since it is a form of prisoners dilemma. On the other hand if a consumer is not required to pay based on their estimation, then they will tend to exaggerate it so as to increase the chance that the software is written.

Another problem with implementing this in real life is that it seems unfair. In the above example Alice and Bob pay more than Carol and Trent. If everyone trusts the person setting the price then this works but rarely is this the case. For example the DVD region encoding can increase social gain by allowing some people to pay less, but quite a few people don't trust the people in control (and there are good reasons not to trust them, but that is another article).

This solution works theoretically but in practice is almost useless. It might work if some piece of software were specialized enough and valuable enough that all the interested parties could come together in a meeting and work out a deal. However, if there were forgotten parties then the social gain would not be maximized. In most cases it would not work properly anyway since the knowledge is spread among too many individuals.

Possible Solution 1: Free Software

Let's assume that the software has been produced. Using the four freedoms defined by the The Free Software Definition of the GNU Project will maximize consumer surplus. (Note that freedom refers to the ability to act without certain restraints, not lack of price.) They are:

Freedom 0
The freedom to run the program, for any purpose.
Freedom 1
The freedom to study how the program works, and adapt it to your needs. Access to the source code is a precondition for this.
Freedom 2
The freedom to redistribute copies so you can help your neighbor.
Freedom 3
The freedom to improve the program, and release your improvements to the public, so that the whole community benefits. Access to the source code is a precondition for this.

Freedom 0 allows the consumer to get some benefit out of the program. Any limitation on personal use of the program will decrease the value of the program and so decrease consumer value.

Freedom 1 allows the consumer to improve the program to meet personal needs and desires. This increases the benefit of the program to that consumer and so increases consumer value. Access to the source code also allows the consumer to more accurately know what the program does. This will improve the value of the program if the program does what the user expects.

Freedom 2 allows consumers to distribute the software to other users and thereby create more consumer surplus by providing the software to other consumers who would not otherwise have it.

Freedom 3 produces increased consumer value by leveraging freedoms 1 and 2 to increase the program's value and distributing it.

Combined together these freedoms will maximize total consumer surplus since they allow the price to be zero and allow consumers to maximize their personal consumer value.

The biggest problem with free software is that it is very hard to correctly reward the creators of it. The freedoms prevent the typical method of charging each user. The first users can be charged for the software but after that both they and the originator are free to give it away. This decreases producer surplus and thus we would expect that less software is written as free software than the optimum amount since the producer can only get part of the value of the software back. This creates an underproduction problem.

On the other hand free software has increased consumer value. If the producers of the software can get some of this increased consumer value, a higher producer cost can be supported so there will be social gain and more software will be produced.

The increased consumer value increases the amount of free software versus non-free software. The fact that the producer receives a decreased proportion of the consumer value decreases the amount of free software versus non-free software. These two effects counteract each other and it is hard to tell which is more dominant. However, since there is much more non-free software than free software, the decrease in producer portion of the surplus would seem to be the dominant effect.

However, for many things the increased consumer value sufficiently exceeds producer cost (CV >> PC) so free software will be profitably produced.

The main cause of free software is that some consumers value it enough that they will produce it themselves. TeX is a good example of this. Donald Knuth valued high quality typesetting so much that he was willing to devote over a decade of his life to creating TeX for this purpose. His own value of TeX was great enough that the cost of producing it was less than his personal consumer value for it. Making it free software increased his value of it since it enabled him to see more mathematical papers beautifully typeset. If the producer is also a consumer, and the value to that entity exceeds the production cost, then free software can be produced that way. Producer surplus is already positive so there is no need to increase it monetarily, especially if charging money will be a hassle. Altruistic desires can also increase personal consumer value enough that it will be created and released.

Freedoms 1 and 3 are the most important freedoms for the production of free software. These freedoms allow any consumer to become a producer if they so choose. This creates an opportunity for social gain any time there is an improvement to the software where the producing cost is less than that particular consumer's value of the improvement. With freedom 3 the improvement can then be sent outward to whomever wants it, increasing the consumer value of the whole software. This can cause continual improvements to free software that are much less likely to occur in non-free software.

Free software maximizes consumer surplus for every piece of it that is produced. However, it still has the problem of underproduction. There are some advantages to free software that help it to be produced. Unfortunately it is unlikely to produce all the software that is needed do to its underproduction problem.

Possible Solution 2: Government Incentives

Sub Solution: Copyright law reform

The purpose of copyright is to increase social gain. It does this by:

  1. Providing protection from piracy so that producers have improved odds of getting producer surplus.
  2. Expire so that the works that are produced maximize their social gain by becoming free after a certain amount of time.

Unfortunately copyright law is doing the second purpose pretty poorly. The 95 year time limit in the United States is far too long to maximize social gain. After 95 years very very few works still have social gain left to be gained by the freedom. Also there is no requirement that the full source code be provided at the expiration of the copyright. This means that for a piece of software it will take 95 years to get freedoms 0 and 2, and freedoms 1 and 3 will never happen unless the publisher wishes them to.

A law that would improve social gain without significantly decreasing producer surplus would be something along the lines of:

This law would substantially improve the social gain. The reregistration requirement will ensure that there are usable copies still in existence when the copyright expires. The source requirement ensures that freedoms 1 and 3 can be used. The reregistration requirement also ensures that when the creator or publisher is no longer interested in the work, the work soon gains freedoms 0 through 3, giving the task of maximizing consumer surplus to the public.

The only downside to these reforms is that it slightly increases producer cost since it adds a registration and reregistration requirement and the Library of Congress is required to store more stuff. However, if the registration cost (medium and mailing cost) is higher than the producer wants to pay, that is probably a good indication that the copyright should expire since that means their remaining producer surplus is low, so the public should get a chance. The Library of Congress is already in charge of storing all the books copyrighted so storing more stuff is reasonable.

Sub Solution: Other incentives

The most important thing that the government can do is make sure that the laws are reasonable to allow creation of intellectual property. Laws that allow fair use, reverse engineering, parodies, and derivations permit the creation of reasonable derivatives of existing intellectual property and improve social gain. Also, the laws need to be flexible enough to uphold a license like the GNU General Public License so that experimentation can happen that might find new methods of intellectual property creation. As long as the laws allow free intellectual property to be created some will be. Laws that protect non-free intellectual property at the expense of free intellectual property need to be examined very closely to make sure that they are beneficial. Laws such as the SSSCA would be very damaging to social gain (and freedom in general) if passed.

Another way that the government could help would be to help lower the producer cost of free intellectual property. One key way of helping support free electronic intellectual property would be to provide free web/ftp hosting through the Library of Congress for free intellectual property and expired copyright works. This increases the distribution of free intellectual property and provide an incentive for making intellectual property free so it can get the free hosting.

The last method is direct funding of the development. Since free intellectual property has more consumer value than non-free intellectual property, much of the intellectual property that the government develops should be free. Grants to produce free software can often benefit society since they raise social gain by the amount that the consumer value is increased.

Solution Conclusion

Governments can help free intellectual property development by making sure that laws are supportive of such property creation. Since free intellectual property has higher consumer value it should be preferred to non-free intellectual property. Both types of intellectual property should be supported by law since each can be valuable in different situations. Intellectual property that the government creates should be free due to the higher consumer value and therefore higher social gain.

Possible Solution 3: Escrowed Multiway Auctions and Other Untried Ideas

One final possibility for funding free software would be a escrowed multiway auction. Basically someone would propose a software project and provide requirements and a deadline for it. Either a software developer could propose a product that they want to work on, or a user could propose an idea that they would like to use. From then on potential users could put money up to be escrowed to pay to anyone who developed the software. If the software is successfully developed by the deadline, the first team to develop a solution would get the money. If the deadline passes, the potential users get their money back, possibly even with interest.

One key to working this properly is copious communication. Developers need to know how much they can get for a project and what the project is. Developers need to communicate how they are progressing and that they plan on trying to obtain the reward money. Keeping two development teams from trying for the same prize is one of the possible problems. Developers also need to communicate if they think the payment is too low. Ample communication is the key to eliminating the problems of this type of solution.

Theoretically, this solution can maximize social gain since anyone who gains consumer value can put money into the project reasonably risk free and there is no incentive to lie about their consumer value. Overproduction does not happen since there will not be sufficient money. Underproduction will not happen since potential users can continue to add money (up to their consumer value) to the reward until some team decides to go for it.

There are problems to this approach. If two or more teams end up working on different solutions to the problem, somebody won't get the prize money. (Publicly available software subgoals and the teams that have met them would help to lessen this problem by showing the progress of various teams.) This will take somewhat longer to produce software than a standard contract since there is more complication in figuring out which team is capable of producing the software and will produce the software. Also, splitting up the money within the team will be problematic (how much should Linus get? how much should Alan get? ...).

This solution might work, or it might not. Either way, I think it is an interesting idea to get people thinking about solutions to the problem of maximizing the social gain of intellectual property.

Conclusion

Intellectual property is problematic. The only practical solution to maximize consumer value is to give it the four GNU freedoms. The primary method for producing it requires it to have a definite non zero cost and forbid copying. Maximizing social gain with either is difficult. More thought must be given to how to do this. The two simple solutions of no copyright and absolute copyright both fail to maximize social gain. I have never heard of any practical alternative solution that is in use and maximizes social gain.



Thanks to Elizabeth Cogliati for the extensive proofreading (any mistakes left are my own fault.)

Copyright 2002 Josh Cogliati

Permission is granted to anyone to make or distribute verbatim copies of this document as received, in any medium, provided that the copyright notice and permission notice are preserved, and that the distributor grants the recipient permission for further redistribution as permitted by this notice.

Permission is granted to distribute modified versions of this document, or of portions of it, under the above conditions, provided also that they carry prominent notices stating who last altered them.