Creative Works as a Public Good

This seems blindingly obvious now that it occurred to me, but for some reason I haven’t seen anyone talk about this. Creative works (including digital media) are a public good (and furthermore, they are a global public good), and this has some interesting implications.

Before you draw any conclusions from what I am claiming, go read the Wikipedia article on what a public good is. Firstly, a public good is non-rivalrous, that is, one person’s consumption of the good does not affect another person’s ability to consume a good. Food is a rivalrous good, national defense is a non-rivalrous good. Secondly, a public good is non-excludable. In other words, it is impossible to prevent people from using a public good. Cable television is an excludable good, whereas broadcast television is non-excludable. It’s easy to see that digital media is non-rivalrous, but it also is somewhat non-excludable as a result of easy-to-use file sharing applications. As a result, digital media becomes somewhat of a public good.

The ability of digital media to be excludable is the root issue behind DRM and the DMCA. DRM attempts to make digital media excludable by using a form of encryption to prevent media from being transferred easily between computers. However, we know that DRM (in its current incarnation) is a fundamentally flawed concept. From a technological standpoint, it’s a nearly insurmountable problem. DMCA attempts to make media excludable by using laws to make it illegal to circumvent the fundamentally flawed DRM technologies. As a result, digital media in the current market is somewhere between fully excludable and fully non-excludable. To analyze the situation, let’s examine the two scenarios at each extreme.

Suppose digital media is fully excludable. If this is the case, then each individual creative work can be thought of as an individual good in a monopolistic competition. In monopolistic competition, the price and quantity level is determined by the equation MR = MC, where MR is the Marginal Revenue and MC is the Marginal Cost. When selling something like digital media, the marginal cost of producing a single unit of a good is essentially zero. This is illustrated in the figure below.

This is not a desirable outcome. Looking at the graph below, the economic surplus of the producer and consumers is shown, using the fact that the average total cost curve would be determined by ATC = FC / Q (since there are no marginal costs). In addition, you can see that there is a significant amount of deadweight loss. In addition, this graph does not show the cost of making the digital media excludable. This would likely cause a shift in the demand curve to the left (representing consumers dislike for media that is difficult to use) and an increase in fixed costs, further decreasing the total economic surplus. The presence of deadweight loss shows that this is not Pareto efficient.

Now suppose that digital media is fully non-excludable. The scenario is much easier to analyze in this scenario. Assuming that the fixed cost is significantly large, the producer will not want to sell a single copy of the digital media, since they will never have a chance of making any money at all. The result is a figure that looks something like figure below. This is very obviously not Pareto efficient.

In reality, digital media is currently neither fully non-excludable or fully excludable, and analyzing this reality is significantly harder. One way to approach the problem is to consider pirated music as a substitute good. This results in a decrease in the demand curve for the “genuine” media. This results in a loss of economic surplus compared to perfect excludability. However, there is also a gain of economic surplus from the pirated music, so the net gain/loss of economic surplus is unclear. What is clear, however, is that this scenario is also Pareto inefficient.

So if all three of these scenarios are Pareto inefficient, how do we fix this? One Pareto efficient solution is to set the price level to 0. While this is Pareto efficient, there is no incentive for the producer to sell if they are losing their initial fixed cost investment. This is very closely analogous to the scenarios for other public goods. There are various ways to give the producers incentives, such as a tax that is used to subsidize the producers so that they produce at a market efficient level. I will postpone my proposals for Pareto efficient solutions until another blog post.