Tuesday, October 23, 2007

Is Voting on Taxes for Specific Services a Good Idea?

In addition to voting on Measure 49 and Measure 50, I am also being asked to vote on a measure, 02-64, that would levy $0.90 per assessed $1,000 to support services in Benton County, most notably the County Sheriff's department and justice system. Last year, Corvallis residents were asked to support a levy to support local schools. Measure 50 is, of course, a special tax on cigarettes to support children's health care access. And so, time and again, we see voters being asked to support a tax for a specific public service or infrastructure project.

Some celebrate this type of measure-driven public policy arguing that it gives voters more control over what they want government to do. Economics tell us, however, that this is a poor way to provide public goods - goods that have at least partial non-excludability and non-diminishability. As an example of public goods consider police protection: the fact that I enjoy police protection does not exclude any other person from enjoying it and the fact that I may need police assistance from time to time does not in any substantial way diminish what is available to others. This is true of other services (fire, ambulance, street cleaning, etc.) and infrastructure (roads, bridges, libraries, etc.). The reason that voting for the funding of specific government functions is a poor way to do things can be easily seen in this simple example I have been teaching undergrads for 10 years. The story goes like this: suppose a town beside a river is considering building a bridge over the river. There are only five residents in this town and they each would value having that bridge differently. Suppose their own private valuations for having the bridge are: $800, $600, $300, $200, $100. In other words one person would be willing to pay up to $800 to have the bridge built, the next $600, etc. Note that the total benefit to the town of a bridge is $2000. Suppose that the bridge would cost $1600 to construct. Socially then constructing the bridge would lead to $400 in surplus. But will it be built if an equal tax were to be voted on? No. The proposed tax would be $320 and only the top two valuation residerns would vote yes. The tax would fail a popular vote 3 to 2. What about asking for voluntary contributions? It is likely that each individual would contribute little, hoping that others will contribute more. So, in the end the community is likely to loose the $400 in surplus a bridge would provide. What if the government had a general fund from which it could draw the $1600? Given the $400 surplus, this would be an easy choice to make - build the bridge, everyone is better off.

Now there are ways to make the tax progressive that might help. (But note that I have said nothing about income or wealth - just valuation: so the $800 person might be the poorest but work directly across the river and without the bridge has to travel 50 miles each day) Sill there is a good chance that socially beneficial projects or services will not be provided. A bigger problem arises in the from of free-riders. Since I can enjoy a public good if you pay for it, what is best for me is to enjoy it without contributing (like OPB for example). It is these same motivations that provides the rationale for government provided services in the first place. Allowing voters to decide on them al-a-carte style makes no sense: it makes it likely that public goods will be underprovided. This is why a goverment staffed with skilled beaureaucrats is the model we use - it allows us to escape many of the problems of free market provision of public goods.

No comments: