Free software and externalities

I’ve been thinking of externalities and by far the hardest form of an externality example to come up with has to be a “positive production externality”. That is, where the production of a good or service has a beneficial spillover effect for the rest of society’s producers. I came across an older post by Jason Welker where he speaks of the difficulty in providing examples. In that post he quotes economics professor Dr. Tim Haab who had this to say about PPEs:

The problems usually come in defining a positive production externality. A benefit to someone that is not fully captured by the producers–usually difficult because producers are usually pretty funny about finding ways to recover the full benefits of their production.

The case of proprietary software provides evidence for this claim. When proprietary software is published, it uses trade secret, patent and copyright laws to keep other producers from benefiting fully from the spillover. For example, trade secrecy keeps useful source code hidden from 3rd parties. Patent laws can keep 3rd parties from implementing similar ideas in their programs. As well, non-disclosure agreements are used within companies in an attempt to stop any benefit from “leaking” to other producers. Proprietary software production is an allocatively inefficient deployment of resources (historically justified under the fallacy that quality software won’t be written otherwise) and represents a market failure.

In the case of Free software production, the PPE isn’t negated by NDAs and patent thickets. With Free software, all producers are affected by the benefit. Non-copylefted Free software brings the marginal private cost (MPC) curve closer to the marginal social cost (MSC) curve while copylefted Free software aligns the two even closer as producers are unable to prevent others from acquiring the same benefits they received. Therefore, it makes the most sense for government policies to support businesses and other institutions in a way that encourages even more development of Free software.

On a related but unfortunate note, countries like Germany are trying to correct a negative consumption externality of proprietary software. Specifically, the burden of malware. Promoting Free software also tackles this problem as Free software offers users the best defense against malicious code.


Tags: , ,

10 Responses to “Free software and externalities”

  1. jayfigh Says:

    Hi there,

    would you like to clarify to me your usage of the terms market failure and market? I guess I know what you want to say, but I would use much more words and different terms to explain this. Especially, it would be very interesting to consider the long-term effects and effects for the whole society, which is what economics is (or at least should be) about.


  2. Peter Says:

    Hi, thanks for dropping in. In this case, proprietary restrictions on software are keeping at bay the benefits the entire society of software producers stand to gain from. That is, the productive resources of the software “market” are not being put into action in an efficient manner. This inefficiency is known as a “market failure” and in this case, is caused by failing to account for a positive production externality.

  3. Peter Says:

    I should add that the difference between non-copylefted and copylefted Free software is only present when proprietary software is in the mix. That is, if there were no proprietary software, there would be no difference between the marginal private costs of non-copylefted and copylefted Free software.

  4. jayfigh Says:

    Ahh, no I see what you want to come up with. Still, the biggest problem for me was that I only knew of “market failures” as it is used in the media, which I saw as senseless.
    The real problem is the term market in mainstream economics, as there probably is no way to achieve the requirements for such a market. Because of this there is probably not much of a practical use of those theories, except seeing market failures everywhere and thus asking the government to intervene ;)

    Concerning your second comment: What causes proprietary software to be possible to exist? Is it government intervention?
    Is then your conclusion that government should support businesses to create more free software the right step or wouldn’t it be better if the government would just stop supporting proprietary software? Isn’t this what the FSF actually is asking them to do?

  5. Peter Says:

    Even if there was no government intervention, proprietary software can be produced simply by distributing binaries with no source. But government involvement can make the problem worse by legalizing software patents. One way government could help is by invalidating software patents.

    Governments should absolutely not support proprietary software businesses but there are ways to support Free software. One example would be to legislate the use of open standards and protocols. Because proprietary software is so entrenched, another positive action could be to offer subsidies to organizations (for example, schools) that adopt Free software. Since proprietary software support is a monopoly (another market failure), this would help create a competitive market.

  6. jayfigh Says:

    Yes, but there would be no legal way to forbid me to do whatever I want to do with my property, even intellectual ones. This includes decompiling, recompiling, changing and selling. These are things I can do with my “hard property” in the real world. This is the short description of what Stallman describes on as the History of the GPL and the GNU project. The reason was not the practical problem, but the legal one.
    Of course software patents, which is another part of intellectual property, make the situation worse and of course, actually the Government should just get out of the way.

    Concerning the second paragraph: I see moral problems with this. First of all, it is the choice of the user which software to use and which kind of protocol. No one can judge better than me what fits best in my situation. Anyways. I would surely agree that the Government should strictly only use open protocols and standards for communication to anyone, including businesses and individuals. I just wouldn’t want to force two individuals to use open protocols when they just communicate with each other and don’t want to use this or that protocol or standard.
    Why are subsidies necessary when Free Software virtually costs nothing, except the costs you also have with proprietary software (support, learning, etc)? I actually think that Free Software always wins, when it comes to monetary values. Of course there are other values as well, ie lazyness. Anyways, considering schooling, there is definetly no need for subsidies. As the Governments chooses what has to be learned, they just should choose diversity, which means also using Free Software, which would lessen the effect of lazyness I describe above.
    What is the cause for proprietary software support as a monopoly and thus market failure?

    What I guess when you mention schooling is that there is some reason for you as a teacher to “have to use proprietary software”, which you actually wouldn’t do if you were free to choose. If it is like this, what is then the thing which hinders you from installing Free Software and teach this?

  7. Peter Says:

    This includes decompiling

    My understanding is that if it is the intent of the designer to obfuscate their code, then it isn’t possible to decompile it. Only certain situations and languages allow this.

    it is the choice of the user which software to use and which kind of protocol

    Yes, I meant for all public affairs. I suppose “legislate” was a poor word choice as it implied telling people that they must use open standards. However, what do you think about institutions that are publicly funded? For example, a public school? Should a public school have the “freedom” to tell students that they must use proprietary software and standards? Regardless, you are right – I did not mean to imply passing a law making proprietary standards illegal as a private choice. My bad.

    Why are subsidies necessary

    I don’t know that they are, but I suggest them as a possibility simply because the world was asleep at the wheel and didn’t wake up until they had found proprietary software/standards were entrenched. The argument could be made that this entrenchment can’t be overcome by the free market alone and that subsidies could help leverage a change. Those subsidies need not be permanent. I too believe that, ceteris paribus, Free software wins – but we’re not working from scratch. But again, I don’t know that they are necessary but I can see how they could help accelerate the process of correcting the market failure.

    What is the cause for proprietary software support as a monopoly

    Source code secrecy. That is, only superficial support can be given by others. In the classic car analogy, although many free markets exist (e.g. wash your car, paint your car, clean your car, etc.), the hood is welded shut and the business to work on fixing or modifying your engine is a monopoly. Even if you are a mechanic yourself, you are helpless.

    As a teacher I am forced to use Windows to do part of my job. Our current SIS is proprietary software and runs only on Windows. However, I make the best go of it I can. I run Windows in VirtualBox on GNU/Linux and literally do all the rest of my work on GNU/Linux. As well, I teach all of my classes to students using GNU/Linux.

  8. jayfigh Says:

    Do you want to talk about the technical details of compiler/decompiler? Anyways, as an obfuscator cannot change the semantics of a program you can get the program back. Of course you still have the problem to analyse the semantics of the resulting code yourself (understand the code), but this is the same problem for complex Free Software. If there would be a (legal) use for decompilation there would be much more research in optimizing decompilers than is done right now. And, consider the fact that companies sometimes behave like double-agents, in our case: selling obfuscator and decompiler.

    Secrecy is not a real problem. Consider your example with the engine. Of course, there are companies selling special tuning and fixing for every car without being part of the monopoly. And you are not helpless, most of the time, you are just damn lazy. Actually I have seen a person, while not being a mechanic (he was a trained butcher!), fixing his own transmission, with his own hands. He did not need any help. This was just two years ago.
    In the end the question is only, if you are still competetive after you have invested money and time in the research of a certain engine. So if there is no artificial burden on you (ie licensing, patents, whatever) and the originator is cheaper than you are, then the other one is not misusing the monopoly and actually is efficient. The other one also had to invest time and money into the research, so there would be no use in saying, that a competitor should not need to pay for his own research. There would be no problem for me if it was this way and I guess neither for most other people. The only problem right now is, that it is not this way, because the possible competitor has to pay licensing fees and so on.

    I guess we pretty much agree that you emphasize a little bit more where Government “could help the cause” and I emphasize the part where Government actually “did harm to the cause”, maybe unintented.

  9. Niels Says:

    Externalities of Free Software products is an interesting subject.

    Digital products have potentially little or no rivalness and excludability. Digital products are practically non-rival and non-excludable.

    But Free Software products are even anti-rival products. An interesting term coined by Steven Weber in his The Success of Open Source from 2004.

    Check out:

    As for your discussion on supporting Free Software, I think the government, exactly for the reasons of the anti-rivalness, should support it. The question in what manner is debatable. Subsidizing seems radical, (and might even disturb a free os-community?), but supporting by using the product seems to be the minimal effort that every state should be following.

    (Every state that is not earning great tax on huge software firms, that would be countries like Brazil, South America, France and China.)

  10. Peter Says:

    Niels, thanks for sharing that link. Yes, subsidizing does seem radical and would likely bother some in the Open Source community. For example, those who use Open Source not as a business built around freedom but as an add-on to bolster the sales of a related proprietary package. They might feel threatened.

    I would imagine you are correct about the tax issue. Is there an easy way to find out how much tax each of these governments takes in from proprietary software development? Perhaps then a good approach is to offer a tax reduction (or elimination) to companies that publish Free software. What would you think about that?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: