This quarter, I’m a TA for my advisor’s undergraduate course, Online Communities and Crowds. We’ve been reading some of the classics of public goods production and common pool resources.

Public Goods

The distinction between private and public goods comes from economics, and was first discussed by Paul Samuelson. Private goods are things like food, land, or goods. They are “excludable” (meaning that it’s possible to prevent others from consuming them) and “rivalrous” (meaning that a good can only be consumed by one person at a time). One of the foundational insights of economics is that, via prices, markets do a very good job of producing an efficient amount of private goods. That is, people spend time and money producing things that other people want.

Public goods are goods which are non-excludable and non-rivalrous. Canonical examples include national defense or lighthouses. One ship’s use of a lighthouse does not diminish the ability for other ships to benefit. Nor can a lighthouse selectively transmit light only to those ships who have paid the “lighthouse tax”.

Mancur Olson argued that goods like this often require collective action to produce, and agents behaving rationally will underproduce them.

Building a Lighthouse

To get an intuitive sense of his argument, let’s go back to the lighthouse example. Let’s say that there are 50 ships want to sell their goods in Town A, but the port requires a lighthouse to be safe. So, all of the ships dock in Town B, and then spend $10,000 per year to send their goods by train to Town A. If there were a lighthouse, then every ship would be $10,000 better off per year (and would therefore be willing to pay $10,000 per year for the lighthouse). If building and maintaining a lighthouse costs, let’s say, $100,000 per year, then overall, the ship owners (and the consumers) would be better off by 50 * $10,000 - $100,000 = $400,000 per year. Having a lighthouse is a no-brainer. It is literally better for everyone. But, Olson argues that it won’t get built.

To see why, think about the decision that each ship owner has to make. Let’s say a lighthouse architect meets with each of the ship owners, and gives them this pitch, “There are 50 of you - if each of you pays just $2,000 per year then we can build the lighthouse and you will save $10,000 in train costs.” Our ship owner may be tempted to say yes, but then he thinks for a little bit, and decides that he’ll just wait until the other 49 owners build the lighthouse, and then he’ll save his $10,000 in train costs without having to pay a dime. All 50 of the owners will think the same way, and the lighthouse - which all of them want and which they could easily afford! - never gets built.

Marwell and Oliver expanded on Olson’s insight by exploring the conditions under which collective action might be more or less likely. For example, if the benefit-to-cost ratio increases as the project gets bigger, and if there is heterogeneity in the benefits and resources of beneficiaries, then a project is more likely to succeed. In the lighthouse example, maybe there is one ship owner who owns 9 ships, and so he is willing to spend $80,000 toward building the lighthouse, knowing that he only needs to convince one of his friends with 3 ships to pitch in $20,000, instead of getting all 50 of the owners on board.

There are, of course, ways other than collective action of producing public goods. One role of government is the coercion of citizens into supporting public goods via taxation. Norms and rules are both public goods in themselves, and are means of creating and maintaining public benefits.

Non-linearity of Public Goods

I’ve spent all of this space building up the theories of public goods production because I think that it’s a very rich framework from which to view how and why people collaborate. I want to spend the last little bit introducing a non-obvious corollary from this view of public goods, which is that changes to the cost of public goods are likely to produce non-linear changes to the amount produced.

This happens because contributions to public goods are often dependent. To go back to our lighthouse example, let’s say that the top ship owner has 10 ships (and is thus willing to pay $100,000 for a lighthouse). If the lighthouse costs $101,000 to build, then there will be no lighthouse. If the cost of concrete goes down just a small percentage and now a lighthouse costs $99,900, then he will build the lighthouse, with its attendant $400,000 net benefit. This effect can be even more pronounced when we assume, like Marwell and Oliver, that the partial creation of a public good makes it rational for many others to participate (that is, that many public goods have a “snowball effect”).

Hidden Benefits of New Technology

One implication of this is that there are public goods which we may be very close to producing, but we just don’t know it, because we are not privy to the resources and interest that others have. For example, I would argue that the marginally lower costs of collaboration provided by the Internet made Wikipedia possible, and revealed that many more people than expected find contributing to Wikipedia a net gain.

This suggests that whenever we see an improvement to communication and collaboration tools and technologies we will see an increase in collective action and the production of public goods. The magnitude of this benefit is impossible to measure beforehand, but this argument suggests that we should be cautious about stifling opportunities for collaboration, such as onerous copyright laws or allowing monopolistic internet service providers.