In Praise of Infrastructure Privatization

Business Week has a huge cover story this week on infrastructure investment. "The public should be nervous" about it, says author Emily Thornton, but I think she's wrong.

I've written in the past about why lottery privatizations are a good idea, and much of the same reasoning lies behind my belief that other infrastructure privatization is a good idea too. Not that you'd guess it from Thornton's article. All of the good things that the proceeds can do are described as "short-term fiscal problems"; meanwhile, the investments themselves apparently yield more than the stock market, with less risk than bonds. Sound plausible? Not really.

It's worth correcting a few misapprehensions from the piece:

"There's reason to worry about the quality of service on deals that can span 100 years."
Well, there's also reason to worry about the quality of service that the public sector will provide over the next 100 years. There's no particular reason to believe that the private sector will provide worse service; certainly the experience of private-sector infrastructure projects worldwide would indicate the opposite.
What's more, most road maintenance done in the US by the public sector is done with the aim of short-term, not long-term savings. You can fix up a road for a small number of years for less money up-front, or you can do it properly for more money and save money in the long term. Governments tend to have four-year or six-year time horizons, however: they're not particularly interested in saving future administrations lots of money. A private-sector contractor with a 75- or 100-year contract, however, will be more efficient.
And then if the private-sector contractor really fouls up, the state can simply repossess the project. That's happened before, it will happen again. It's no reason not to privatize in the first place; it's more of an insurance policy.

With the market for infrastructure still in its infancy, every deal is different. The ideal blend of up-front payment, toll hikes, and revenue sharing hasn't been found.Privatization doesn't need to be "ideal" to be a good idea. The benefits of privatization come immediately, and can make a positive difference to millions of peoples' lives. I doubt the kids whose schools aren't built with the funds that the government doesn't get from not selling its toll roads would thank anybody for holding out until the "ideal blend" was found.

Infrastructure is ultra-low-risk because competition is limited by a host of forces that make it difficult to build, say, a rival toll road. With captive customers, the cash flows are virtually guaranteed.
Right now, it certainly seems like road traffic only goes up. Whether that will continue to be the case for the next 100 years, however, is far from obvious. What's more, infrastructure prices are hugely dependent on discount rates, which right now are at all-time lows. There's a good chance that, if interest rates rise, governments will realize that they will never again have the opportunity to sell these assets for anything like the kind of money they're being offered now.
What's more, as Thornton herself notes, "Federal, state, and local governments need to spend an estimated $155.5 billion improving highways and bridges in 2007, according to transportation officials, up 50% over the past 10 years". Could anybody have foreseen that 50% rise 10 years ago? Can anybody foresee how much further such construction and maintenance costs are going to rise over the next 10 years, let alone the next 100? Governments aren't just selling assets here, they're also getting liabilities off their books as well.

The main reason to privatize is that the private sector is more efficient, and, right now, with the huge amounts of infrastructure money chasing a relatively small number of projects for sale, there's a good chance that the successful bidder will end up overpaying. That's happened in the past, too. The government gets more money than the asset is worth, and gets to spend it on the public; meanwhile, the private-sector is burdened with the liabilities for the next 100 years. Sounds like the government might be getting a very good deal, to me.