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
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