The Park Service wants to build and rebuild housing for its employees. The Forest Service proposes to remove excess fuels from private forestlands. The Anchorage transit system promises to create 12 jobs at a cost of $729,000 each. Amtrak wants $30 billion to $40 billion to build a new high‐speed rail line from New York to Washington.
Even a trillion‐dollar infrastructure bill won’t cover more than a small fraction of these wish lists from the states, let alone the current $50 billion. Here are four rules responsible legislators should use to decide which projects to fund:
(1) At least half the cost of any project must be spent within nine months after any infrastructure bill passes. It is not enough for projects to be shovel‐ready. Many ready‐to‐go projects will take years to complete. Projects that follow this rule will maximize primary benefits.
(2) Projects must be largely completed within a year to produce secondary benefits that are just as, if not more, important than the primary ones. Transportation projects, for example, will have the greatest stimulative effect if they lead to significant unsubsidized development. Projects following this rule will maximize these secondary benefits.
(3) User fees must cover all operating and most capital costs. Not all infrastructure is created equal. Bridges to nowhere are infrastructure. The dirigible tower on the Empire State Building is infrastructure. To produce secondary effects, infrastructure must be useful to people.
The best test of infrastructure value is whether users are willing to pay for it. The ideal stimulus package would not be grants but low‐interest loans to be eventually repaid out of user fees. At the very least, user fees should cover half of construction costs and all future operational costs. Funding projects out of user fees also avoids deficit spending.
Many, if not most, wish‐list projects fail this test. House Transportation Committee Chairman James Oberstar, Minnesota Democrat, wants to increase transit’s share of federal surface transportation funding from 15 to nearly 30 percent. But transit riders pay only a third of the operating costs and none of the capital costs of transit, while highway users pay 80 to 90 percent of highway costs. This suggests transit will not have anywhere near the stimulative effect of highway spending.
Even if it is publicly owned, most infrastructure is privately used. However, years of taxpayer subsidies have loosened the ties that should connect users and providers of infrastructure. Any infrastructure bill should encourage state and local governments to return to user‐fee‐funding of infrastructure projects.
(4) Attempts to achieve secondary objectives, such as reducing greenhouse gas emissions, must be cost‐effective. For example, McKinsey & Co. estimates the United States can reduce greenhouse gases 50 percent by 2030 if it invests in projects that cost no more than $50 per ton of reduced emissions. Funding a project that costs $5,000 per ton means forgoing investments that could have eliminated hundreds of times more emissions.
Coordinating traffic signals saves people time and fuel and reduces greenhouse gas emissions for about $10 per ton. Converting buses to biodiesel costs about $200 a ton. Replacing diesel buses with hybrid electrics costs more than $1,000 per ton. Rail transit, when it saves emissions at all (and don’t forget to count emissions from construction), costs more than $5,000 a ton.
The Federal Highway Administration estimates three‐fourths of the nation’s traffic signals are poorly coordinated with nearby signals. No city that still has an uncoordinated traffic signal should invest a dime in rail transit.
Some economists argue that past efforts to use fiscal stimuli to recover from recessions have rarely succeeded. Perhaps it was because the governments involved failed to follow the above rules. But if Congress follows these rules and the infrastructure bill still fails to stimulate the economy, at least it will cost‐effectively produce other benefits that users want and need.