There is a fine case to make for high-speed rail: It connects metro areas, shrinks the distance between key cities, cuts down on congestion, and helps businesspeople move quicker and cheaper than airplanes. And there is a fine case to make against high-speed rail: It's super-duper expensive and there's little indication that Floridians are clamoring for a train system that will cost their state tens of billions of dollars. That's how I see the debate, anyway.

But I think Cato's Tad DeHaven is confusing the issue when he writes:

As is the case with Amtrak, HSR can't compete with more efficient modes of transportation like automobiles and airplanes without massive subsidies.

But automobiles and airplanes already get along with massive subsidies. In 2009, the Department of Transportation ran a $70 billion budget, more than half of which went to the federal highway system. The airline industry only got through the post-9/11 years with considerable help from the U.S. government, and taxpayers pay more than 90 percent of the cost of some small-airport flights

So saying that HSR can't compete with other heavily subsidized modes of transportation is a bit like saying that Jeff Gordon can't compete with other NASCAR drivers without getting behind the wheel of a fast car. It's true that he cannot, but NASCAR drivers aren't running, either.

That leaves the case for roads and trains built by private companies. The case for private infrastructure is compelling when you start and end with the principle that private companies build higher quality products more efficiently. But let's consider what is the purpose of a road. It is to facilitate commerce to grow a city's economy. (Imagine how effective you would be if it took three hours to get to work every day.*) That means that a highway is more valuable than the money it technically "makes" in taxes and fees.

If you're the government, the benefit of good infrastructure is a stronger city economy, and the benefits of a stronger city economy loop back to the government in the form of more people, more business, and higher tax revenue. If you're a private company that owns a road, however, you don't see those wider benefits. You only get the money you collect in the form of, say, tolls. What's more, as a private company you have an obligation to send some of your profit to stockholders in the form of dividends, which means you have to raise more money than is needed to simply keep up the roads. (Ryan Avent identifies some other challenges to private infrastructure.)

So it seems to me that private infrastructure would almost certainly raise the cost of roads for drivers, even if we take into consideration general taxes, because private companies would have to overcharge to make a profit on a product whose benefits are mostly invisible to them. This would make roads a lot more expensive. If we want fewer drivers, then maybe that's a good thing. But if our goal is, in the tradition of Adam Smith, to facilitate commerce, raising the cost of getting somewhere (anywhere) is a bad thing.

*If you live in Northern Virginia, you don't have to try very hard to imagine.