By Walter Russell Mead

Among advocates of big government and Keynesian countercyclical stimulus, one subject keeps coming up: infrastructure. They're always arguing the short- and long-term benefits of building new highways, bridges, tunnels, urban light-rail systems, or, the Holy Grail itself, a national high-speed rail network.

After all, proponents point out, Republican President Dwight Eisenhower's support for the interstate highway system helped drive two generations of American growth. And scandal-ridden though they were, federal subsidies greatly accelerated the development of transcontinental rail lines in the 19th century.

Going back even further, state governments backed ambitious projects like New York's Erie Canal, which, when completed in 1825, created a navigable water route from the Atlantic Ocean to the Great Lakes.

No doubt some projects need doing, and President Eisenhower stood in a long line of Americans who believed that improving the infrastructure was a proper and important job of the federal government. When George Washington, Alexander Hamilton, Henry Clay, Daniel Webster, Abraham Lincoln and Theodore Roosevelt agree on something, they can't all be wrong.

The question is no longer whether government has a role to play, but what that role is in today's economy. In the 1820s and 1830s, many states jumped on the canal bandwagon after the success of the Erie Canal, but the arrival of the railroads made that investment essentially worthless. Mule-drawn canal barges don't compete with trains. Even the Erie Canal soon became more tourist attraction than working piece of infrastructure.

In today's economy there are three big problems with the conventional case for infrastructure spending. The first is that nothing is shovel-ready anymore. Not-in-my-backyard, or "Nimby," lawsuits and environmental requirements tie up significant projects for years if not decades.

In practical terms, this means that infrastructure can no longer serve a Keynesian, countercyclical function. Projects can't be launched to provide jobs during a recession because by the time a project gets all the local, state and federal approvals, wins all the related lawsuits, and makes sure that all the nearby spotted owls and snail darters are protected, the recession is ancient history.

The second problem is that the handful of large firms active in this market tag-team legislators and governors with their labor-union allies to ensure that government contracts are gold-plated. The fatal combination of legal and administrative paralysis, plus crony capitalism, drives the cost of building new projects and maintaining what we have to levels at which potentially sensible projects make no sense. The taxes and tolls necessary to pay these exorbitant costs offset the economic advantages.

But the third problem is more fundamental. The infrastructure lobby believes that, from a transportation point of view, the 21st century will be an extension of the 20th.

Thus Gov. Jerry Brown defends California's indefensible high-speed rail boondoggle by arguing that without it, increased demand for travel along the dense Los Angeles-San Francisco corridor will lead to choked highways and hopelessly congested air corridors.

Is that really where things are headed? The Internet is dramatically reducing the importance of distance in human affairs. Email has rendered the local post office nearly obsolete. Hundreds of thousands of Americans are already telecommuting, and many more have launched Web-based businesses from home. People who used to make three trips a week to the mall do more of their shopping online.

Two things seem clear about the 21st century: Internet connectivity and bandwidth are going to improve so that today's technologies behind services like Skype are going to change beyond recognition. Each generation of young people will be more accustomed to socializing and interacting online. We are going to have more, better and cheaper alternatives to traditional business and commuting travel patterns, and our society will find it more and more natural and desirable to shift from expensive, time-consuming travel in "meatspace" to doing business online.

More sophisticated computer technology will also allow us to use existing infrastructure more intensively and efficiently. This means better air-traffic control allowing more efficient use of runways, and self-driving "Google cars" (already in trial) allowing faster movement on existing road networks. We aren't going to need 20 lanes in either direction on the New Jersey Turnpike by 2050, or $100 billion high-speed rail projects, to save us from national gridlock.

The challenge isn't to move more meat; it is to move more information more effectively, and to re-engineer business practices and social organization to take full advantage of the extraordinary efficiencies that the Internet affords. The rush-hour rituals of the 20th century aren't destined to continue to the end of time. Telecommuting, flextime and mini-commutes to satellite offices will change the way we work.

Lobbying for more highways and high-speed rail misses the point. What's needed instead is support for advancing the Information Age economy. Government policy could reward companies that promote telecommuting and teleconferencing, for example, or otherwise facilitate the transition. (Greens take note: These and other business-friendly policies would reduce greenhouse-gas emissions.)

There is still much work to do to build the information superhighways we will need to compete in this century and the defense systems that can protect them against cyberattack. Government will have a significant role to play in creating a suitable regulatory structure and policy framework to accelerate this process. Yes, our existing roads, bridges and highways should be maintained, and in some cases enhanced. Even so, more physical infrastructure isn't our main need at this point.

We don't want to build the 21st-century equivalent of a new and improved national canal network. Infostructure rather than infrastructure should be the priority.

Walter Russell Mead is a professor of foreign affairs and humanities at Bard College. His blog, Via Meadia, appears at the American Interest Online.

Reprinted from The Wall Street Journal