Highway, Infrastructure Spending Needs Benefit-Cost Analysis

At the National Journal Transportation blog I take part in, Lisa Caruso asks:

On April 26, the organization representing state transportation officials released the first in a series of reports calling for increasing the nation’s capacity in transit, rail and particularly highways. Two days later, a leading consumer watchdog group unveiled a report urging federal and state governments to strongly emphasize preserving the current interstate highway and bridge system rather than expanding it.

Unlocking Gridlock,” released by the American Association of State Highway and Transportation Officials, focuses on relieving urban congestion. While AASHTO advocates maintaining and improving the performance of the current transportation system, shifting car trips to other modes and shifting freight from trucks to rail, it says that adding new highways “will be a principal part of what is needed” to relieve congestion and foster economic growth.

When it comes to highways, the U.S. Public Interest Research Group reaches a different conclusion in its report “Road Work Ahead: Holding Government Accountable for Fixing America’s Crumbling Roads and Bridges.” U.S. PIRG believes “we must adopt strong ‘fix it first’ rules that give priority to maintenance of our existing roads and bridges, set national goals for the condition of our transportation system, and hold state governments accountable for achieving results.”

How would you set priorities for highway spending? What criteria would you use? How can federal policy better encourage states to keep existing roads and bridges in good repair? How can it give them better tools to address their urban mobility and congestion problems?

My response:

Both the PIRG and the AASHTO reports are flawed, in that neither calls for highway investment policies that require a decent return on the investment.

PIRG correctly notes that elected officials have a bias toward cutting ribbons rather than ensuring proper highway maintenance. And AASHTO correctly points to ample research demonstrating the large need for adding capacity in both urban areas (for congestion relief) and on long-distance Interstates (for goods movement). But both exaggerate their cases.

PIRG’s proposal to do no capacity expansion until every last mile of road in the country is brought up to first-class standards would waste huge sums on roads with very little traffic. But AASHTO’s uncritical endorsement of more than doubling total annual highway investment to $175 billion makes the same mistake. That number does come from the excellent 2008 Conditions & Performance Report from the U.S. Department of Transportation (DOT) —but that’s only one of several dozen scenarios in the report. It’s what we might call the DOT’s “unconstrained” scenario. We can get a more credible investment approach by reviewing the report’s other scenarios.

The Conditions & Performance Report provides two approaches to prioritizing highway investment. One is to use a benefit/cost screen, allocating investment dollars only to projects that exceed some a benefit/cost threshold value. Its “unconstrained” scenario uses a benefit/cost = 1.0, a standard far below a reasonable return on an infrastructure investment—especially when resources are limited. Reviewing the same data with a a benefit/cost screen of 1.2 yields a highway investment target of $157 billion, while a benefit/cost of 1.5 puts the annual target at $137 billion. That’s still a large increase from the current $79 billion, but is a far more defensible investment goal.

The other prioritization approach in the Conditions & Performance Report is congestion pricing—assumed to be applied immediately to all congested highway segments nationwide. While that is totally unrealistic, if it were do-able, it would further reduce the required annual investments. Pricing and a 1.0 benefit/cost screen yields $122 billion per year; benefit/cost of 1.2 with pricing cuts that to $117 billion, while the 1.5 screen reduces it further to $102 billion. While we aren’t going to price all lanes on all congested highways anytime soon, those numbers should give us the bottom range of sensible highway investment.

For purposes of the current reauthorization, I’d opt for the no-pricing approach with benefit/cost of 1.5, which gives us the $137 billion/year highway investment target. The DOT’s modeling suggests that 52% of this should be system rehabilitation, 38% of it on system expansion, and the balance for safety and other enhancements.

My point here is that serious modeling of this sort gives us a way to go beyond PIRG’s ideological anti-highway posturing—and AASHTO’s state DOT wish lists. A nation that by all realistic forecasts will still be using highways for 90 to 95% of all personal travel 25 years from now and will still be using trucks for 80 to 90% (by value) of all domestic goods movement has no realistic option but to invest in both highway rehabilitation and highway expansion. But AASHTO would be more credible if it started applying realistic benefit/cost screens to prioritize both kinds of investment.

More responses here.