According to a recent study by GoEuro, the U.S. ranks 19th of 20 countries surveyed in high-speed rail. One might quibble with how they calculated it, but nevertheless, the U.S. has basically no high-speed rail, and an ageing and limited standard rail system. At present, people are transported long distances principally in private automobiles and aircraft, while approximately 70% of freight is moved by trucks. While there are some in-city heavy rail systems, this post focuses on long-distance transportation.
Passenger rail come in two basic flavors, standard and high-speed although the line between the two is a bit fuzzy. Standard rail runs at speeds of up to ~100 mph (apologies to non-U.S. readers for the units), averaging 50-60 mph, while high-speed rail is usually defined by top speeds in excess of 150 mph, with systems such as the French TGV averaging speeds between 150 mph and 170 mph depending on the route. Note that even standard rail typically travels at near freeway speeds, while high-speed rail vastly outstrips the speeds safely attainable by consumer vehicles. Freight rail is usually only run on standard tracks, but there’s no intrinsic reason it couldn’t be run on high-speed rail if that turned out to be cost-effective (I doubt it).
Next, we can look at the relative energy efficiencies of various modes of transport: The US Transportation Energy Data Book indicates that moving 1 passenger 1 mile by train requires 2,435 BTU, compared to an average of 3,600 BTU for private vehicles. The difference in energy costs for moving a ton of freight is even more drastic: trucking cargo takes 3,357 BTU per ton-mile, compared to only <i>289</i> for freight rail. This measure is fuel-neutral, but even assuming that rail lines continue to be run entirely on diesel, the reduction in greenhouse emissions from switching back to rail freight would be enormous, and it would be far easier to retrofit a relatively small number of locomotives for alternate energy sources than millions of semi trucks.
But wait, I hear you cry, what about the cost of it all? Well, I’ll answer you that: Dual track standard rails cost around $3 million per mile to lay in flat, relatively unoccupied terrain, about the same as a mile of 2 lane (for obvious reasons)blacktop highway in similar conditions (Costs go up rapidly in mountainous or populated regions, but both modes increase by similar factors. In many cases the actual cost for either mode run as high as $25-30 million.). High-speed rail costs far more per mile of track , running around $30 million per mile, but offers considerable benefits by comparison. Leaving aside the benefits in greenhouse gas reduction, the cost of laying the rails or roads is only a fraction of the total costs, and not just to the public purse. While ‘fiscal conservatives’ scream about tax money going to rail lines, they ignore the extra costs to the taxpayer of using that money for roads instead: the cost of having to own a car to get anywhere*. There’s the price of the vehicle, there’s insurance, gas, maintenance, repairs, and it’s all coming out of your personal pocket with no way to spread the impact. If transit proposals raise your tax bill by less than ~$8,000 per year, you’ll still save money over owning a car yourself (~$13,000 if you drive an SUV).
An addition problem faced by passenger rail in the U.S. (And a not insignificant problem for freight rail) is that a large portion of existing U.S. rails are privately owned by freight lines, and Amtrak (the U.S. passenger rail service) has to rent space on them and defer to freight trains, which wreaks havoc with their scheduling. This is only one of many areas in which the U.S. persists in the extremely ill-advised practice of allowing for-profit companies to own and operate infrastructure, to the general detriment.*
In sum, the correct course of action would be to a)nationalize all existing railroads in the U.S., and b) sink a very large sum of money into expanding the rail network, and c)build high-speed rail lines connecting every population center in the country above 200,000 people.
*This is worth an article itself, probably, and will also be mentioned a lot in other infrastructure articles.