Misunderstanding Public Goods

Via Bryan Caplan, I learned about a new paper by Frances Woolley on the difficulty of teaching the theory of “public goods”. I am very sympathetic to the paper because I feel that the term is among the most frequently misunderstood and misused terms in all economics (alongside “human capital” and “market failure”).

As I see it, the main problem with the phrase “public goods” is the strong tendency to use it synonymously with “government activities”. And I’m not only talking about how uninformed laypersons use the phrase. Even trained economists tend to see everything the government does as a solution to some public goods problem. This is a mistake. There are lots of things the government provides that are not public goods – such as education and health care services. Conversely, there are lots of public goods that are provided by private companies – think radio programs or internet search engines.

Woolley mentions that problem as well, but her main concern is that in talking about public goods we frequently mix up three distinct concepts:

  • non-rivalry: does it cost more to provide the good to two or more persons than to one?
  • non-excludability: is it feasible to exclude people from using the good?
  • public finance: who pays for the goods in what way?

To be fair to my educators at the Uni Graz, I think they did try to keep the three concepts apart. I am just not sure that every student got the message. And I think Wolley is right to point out that some of the examples textbooks use to illustrate the concept such as “law and order”, “fire protection” or “national defense” are only public goods when viewed as abstractions:

The goods and services that go into creating ‘law and order’ are not themselves pure public goods: access to the courts is rival. The explosion of gated communities and private security firms is evidence of the excludability and rivalness of police protection. Fire protection is, from a technological point of view, excludable (the fire department can refuse to put out your fire if it chooses). Coase (1974) argued that lighthouses were, historically, often privately provided and financed, and changing navigational technology is making them obsolete. Parks are partially excludable (permits are required for hiking and camping at many national parks, for example).

In light of these considerations it is pretty hard to come up with real-world examples of pure public goods – challenge for the comments section: name pure public goods. I agree, therefore, with Woolley’s plea to de-emphasize public goods in introductory economics courses and address the issues related to non-rivalry and non-excludability separately.


4 thoughts on “Misunderstanding Public Goods

  1. Thank you, Max, for offering your blog post that I consider a pure public good. If Sophie reads your post, this does not prevent Christoph from reading your post as well. That is, your post is non-rivalrous. Moreover, as long as Graz Economics Blog is freely available to everyone, your post is also non-excludable. So, both defining characteristics of a pure public good are satisfied.

    However, let me respectfully disagree with one claim, namely that there is a frequent mix-up between (a) public goods and publicly provided goods or (b) non-excludability and non-rivalry. No trained economist would ever mix-up these ideas. The reason is that there are well known (and old) results regarding these concepts. Think of (a) first (and don’t worry, I will not give a lecture in public goods here – I will just express one thought). Two specific types of market failures are externalities and public goods. The private provision of public goods causes an externality (free rider problem) and is therefore inefficient. A huge literature in public economics revolves around mechanisms correcting this inefficiency. Therefore, I doubt that a trained economist could mix-up these notions. Likewise, considering (b), there is a vast literature on non-excludability and non-rivalry. Just to give one important example from development- and growth theory. It is argued that the implementation of patents – that made the nonrival good “technological knowledge” temporarily excludable – contributed the most for the large economic growth following the patent laws. And this argument is also at the heart of modern endogenous growth theory. So again, no trained economist could ever confuse non-excludability with non-rivalry.

    Some public goods are pure ones, others are impure ones (e.g., club goods). Thinking of the real world, many pure public goods come to my mind: national defense; flood control projects; calculus; basic research and development; knowledge (Wikipedia); open-source codes of computer codes for a software application; projects to improve the air quality; acid rain (yes – a public good may be a bad, in fact). No one ever said that all public goods need to be pure public goods. But as I emphasized at the beginning, my answer to your post is a pure public good as well.

  2. This blog is not a (pure) public good, because wordpress could easily make access to the blog subject to paying a fee, for instance, or restrict access in other ways. The fact that no one is currently excluded from enjoying it does not mean it’s a non-excludable good.

    The same is true of some of the candidates for pure PGs you suggest: wikipedia could easily exclude people from using it; same goes for open-source software; not all national defense is a pure public good although some aspects like missile defense are;

    I believe you if you say that no research economist – especially those specializing in public econ – are mixing up non-rivalry and non-excludability. But Woolley’s point – and the point of my blog post – is that they are getting mixed up in teaching and in textbooks by discussing them under the single heading “public goods” rather than as conceptually and empirically distinct phenomena.

  3. Well, WordPress is an Open Source project. As long as this is so, WordPress will not restrict access to its users. Specifically, implementing fees – while technically possible – might kill the project.

    The property of “excludability” is not restricted to excludability in a purely technical sense – in my understanding. Technically, it is possible to exclude others from enjoying/experiencing every good or service. The main question, though, is at which cost (- everyone can be excluded from acid rain, to mention the above example, by bringing her or him to the moon).

    My understanding has been that a good or service is excludable if exclusion comes at a “low” cost. Clearly, we have not defined what “low” means. Therefore, we talk about degrees of excludability. A high degree of excludability (say, encoded satellite tv) is associated with a low cost of exclusion, and vice versa.

    As long as WordPress does not implement fees, my answer to your post is a pure public good. I have a guess why WordPress has not implemented fees yet. While WordPress could technically implement fees – the consequence would be the eradication of their Open Source project, that is, cost would be prohibitively high.

  4. WordPress is free because its inventors have decided that, in their judgement, the benefits from exclusion (revenue from fees they could earn) are smaller than the costs (lower popularity, whatever). In no way does that mean that the costs are prohibitively high. I could, if I wanted, blog on a paywalled platform but decided against it for the same reason. Prohibitively high costs means higher than any sane individual would be willing to pay. While I think demanding a fee for access to this blog would be unwise I don’t think it would be insane.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s