EVALUATING THE CASE. Milwaukee Journal-Sentinel columnist Patrick McIlheran suggests it's wasteful to spend money speeding up passenger trains.

States already have submitted plans for more than $100 billion in train projects. And a private group favoring high-speed rail released a study this summer of what genuine bullet trains on one route would cost. Answer: Gulp.

Because what governors are planning isn't high-speed rail. The Midwest plan is for diesel trains that can go up to 110 mph, about what steam locomotives did across Wisconsin in the 1930s. Nice, but train backers point out that's mere normal speed in Europe or Japan. "I am shocked by the timidness of America," Andy Kunz, head of the U.S. High Speed Rail Association, told a newspaper.

In presenting the arguments that way, the columnist limits the policy options. Yes, European high-speed is faster. It's also more expensive. At the margin, a restoration of the 1938 timetables ("Speeding Up The Rails") might suffice to divert traffic from the roads and the airways to make the project worthwhile. A much faster service provides very little in the way of reduced running times, particularly on the shorter than end-to-end journeys that are passenger rail's greatest strength. ("Reliability More Important Than Rapidity") That steam locomotives could achieve such timings ("Applied Hiawatha Research") on jointed rail protected by semaphore signals also suggests the current regulations governing signalling and track structure are excessive. On the other hand, freight railroads operating long freight trains without cabooses also see advantages in positive train control systems that let crews know exactly where the tail end of a stopped train is, suggesting possibilities for strengthening the railroad network.

If you really want to lure riders, say backers, trains ought to exceed 200 mph. The Midwest High Speed Rail Association hired engineers to cost out such a line linking St. Louis to Chicago. Like high-speed rail in Europe, it would use a dedicated line - no freight mixing with light, fast trains. It would bypass towns and have no road crossings. It would be electric.

And it would cost $11.5 billion. That's one 305-mile line. Did you want to go somewhere else, too?

The association, which sees this as a reasonable price to replace a 65-minute flight with a two-hour ride, isn't overestimating. At $38 million a mile, it's on the low end of what the Government Accountability Office guessed. Assume a similar price in denser southeastern Wisconsin, and this implies that real high-speed rail from Milwaukee to Madison would cost $2.6 billion. Chicago-Minneapolis would be $15 billion.

The airlines dissipate all the time advantage of a 550 mph jet over even a 100 mph train by checking passengers through security, loading and unloading through one door, diverting or delaying when a thunderstorm moves through, and setting passengers down at some distance from their destination. Railroad stations exist in central business districts and have a relatively small footprint near office parks.

Yes, well, freeways cost a fortune, too: The state's twice-a-century rebuild of I-94 southward is $2 billion.

But freeways carry many more people. Even in the European Union, rife with 200-mph trains and $5-a-gallon gas, cars carry 76% of passenger travel, trains 5.8%. That's down from 8% since 1980, even as Europe built irresistible trains. Cars' share is up.

Air, bus, and rail do compete for a relatively small share of total passenger miles. The way in which these modes compete is in providing relatively cheap additions to capacity, compared to the cost of providing sufficient roads to replace the trains and planes. A fast train is an expensive way to get to the grocer or the day-care, errands that account for some of those automotive passenger miles. On the other hand, providing additional expressway, arterial, and parking capacity in Chicago or London or Tokyo to provide for the intercity and commuter rail passengers would be more expensive than finding additional capacity on the existing tracks. On the weekend, some of the British I talked with described efforts in London and Paris and other European cities to discourage driving into the city and encourage rail use. Chicago, without any special legislation, experiences the same effect by market pricing of downtown parking and a reluctance to build new expressways. A lot of drivers shift to Metra and the Transit Authority.

And freeways bring in paying customers. Cars pay more in fuel and other taxes than is spent on building and keeping highways. The feds reckoned in 2004 that for every 1,000 miles traveled, cars paid $1.79 more in taxes than their roads consumed in public money.

Passenger trains took in $210 of net subsidy for every 1,000 passenger miles, that report found. While air-travel taxes, other economists say, cover the costs of airports and traffic control, fares don't cover trains' operating costs, much less the cost of tracks.

That's why costly roads differ from costly rails. "The users of (roads) willingly pay the capital costs as well as the operating costs," said transportation scholar Bob Poole of the nonpartisan Reason Foundation. "That just makes them a different category."

Fundamentally, road and air infrastructure is funded by taxes on road and air travelers. Rail infrastructure is funded by taxes on road travelers. Non-users pay in a way they don't for roads.

Not quite. Read the report. (Page 13)

The estimates in this report include net federal subsidies to passenger transportation for highway, air, transit, and intercity railroad transportation. Subsidies to passenger transportation by state and local government are not included. The data for highway are further subdivided into net federal subsidies to autos, motorcycles, pickups and vans; school buses; transit buses; and intercity buses. Subsidies to air transportation are also presented separately for commercial air carriers and general aviation.

Without the counties and the municipalities maintaining the side streets out of general tax revenues, drivers have a more difficult journey to the roads that receive federal matching funds (roads that they share with trucks, "It's Corporate Welfare".) Some cost estimates can be done on an incremental cost basis, others require allocation of common or joint costs. Yes, improvements in the rail network are costly. It does not have to follow that improvements in the road network are cost-effective, or that the current allocation of resources to transportation infrastructure is efficient.