Urban Public v Private Mass Transit

Mike Munger had a pretty interesting podcast on Econtalk this week about mass transit in Santiago, Chile. An thus Russ Robert's program has inspired another one of my posts...

The case for a public light-rail or subway system is reasonable enough to me. In New York or London, with the tremendous amount of commuters in a dense area, the financials probably make sense. It makes sense for the government to do this, because eminent domain may be necessary to build a connected rail system, and competing rail lines across the city may be inefficient (although, I am a little skeptical of this second argument).

However, I think the case for a government run bus system is quite hard to make. The major American cities all started with private transport systems that were eventually pretty much all taken over by the government by the mid 20th century. So, we do not have much experience in the United States with private mass transit. However, I do not see much advantage to the government running our bus systems. The government has no particular advantage over private companies in running bus transport.

Presumably, you would get the best bus service by allowing private companies to compete as the bus companies would be directly responsible to their customers rather than indirectly to voters who make their decisions in the booth based on mass-transit. Under a private system, the price system should cause the best routes to win out, there will be differentiated service based on needs, and efficient number of buses on the streets, and the best most cost-efficient buses and systems will win out. A government monopoly would have the same problems that government monopolies have. The public system is not directly responsible to it's customers, does not set prices based on the market, has centrally created routes without competition, and would most likely be run at a loss at the public's expense. It has the problems of a government run institution and of a monopoly. There may be advantages to having one system connect across a city. I'm skeptical of this, but one winner would quickly arise in a competitive system if this were true.

There is one good argument I can think of for public involvement in the buses, and that is to ensure service to poor neighborhoods. Now a government bus system could be designed to make certain that the poor have service. To the extent that this would not happen in a purely competitive system, these routes would be a subsidy to the poor. Why not go around this and offer incentives to bus private companies that give rides to the poor? This would be more transparent and also more efficient at making sure the subsidy goes to the right group of people. Also, I would expect that an incentivized private system would do a better job at servicing the favored poor than a public system that does not have to compete for their business, and also deals with a lot of other bizarre incentives.

A private bus system with subsidies to strikes me as a Pareto improvement. I can think of no strong arguments against this idea. If you can, please post.