Daniel Hall helpfully gathers together the links to which I’ll be referring on this post. Go read them, please, but I’ll tell you the bottom line. America needs a great deal of infrastructural investment–trillions of dollars worth. We need this to keep stuff from falling apart, to keep cities functioning (and get them functioning better), and to reduce energy use and carbon emissions, among other things.

Fine. How to pay for it? Increasingly, the answer is public-private partnership, or outright privatization. How should we feel about that?

There are several clear positives to a private approach. One is that we get capital invested publicly without having to raise taxes (or more often, without having to raise taxes as much). Another is that private firms might do a better and more efficient job building and managing the infrastructure.

The downside is that absent proper incentives, private money is not going to think about social goals. And that’s fine. That’s not private money’s problem, it’s ours. If we want to go the private route and do it well, we need to make sure that we’re structuring the groundrules appropriately. And if we can’t structure the rules appropriately, then we need to say no thanks, we’ll suck it up and make the needed investments via the government.

Revenue maximization on a single stretch of toll road might have little to do with the overall goals of a metropolitan area. For one thing, road owners will have a strong incentive to move as many people near the road as possible, and will no doubt lobby along those lines. They will likely lobby against competing infrastructure (like, say, a commuter rail line). They might do all they can to discourage trucking, which might not be economically efficient. Private firms will not consider things like carbon emissions unless we build that into the rules or prices explicitly. And there may be asymmetries in what can attract private funds that determine what gets done. If roads are easy to privatize and passenger railways are hard to privatize, and we’re committed to building the next round of infrastructure with private money, guess where that money is going to be concentrated, regardless of whether or not new highways is at all what we need?

We’re not going to find ourselves in an ideal world, with a sim-city private manager trying to optimize total metropolitan output subject to certain, social welfare maximizing environmental constraints. We’re going to be handing over pieces of a complicated, interconnected system to private firms, and in doing so, we’re going to be reducing the influence of a (one assumes) social-optima oriented government on structures that display a lot of positive and negative externalities.

It would be beyond stupid for me to suggest that governments have performed these planning tasks ably in past decades. That doesn’t mean that efficiently managed private projects are going to produce better net metropolitan outcomes.