LaHood’s Cost-Effectiveness Rule

It was with some trepidation that the Antiplanner finally took the time to carefully read the Department of Transportation’s final rules for major transit capital grants. Long-time readers may recall that the Antiplanner is concerned about the cost-effectiveness of these grants, and urged the Department to strengthen those requirements–without much hope that the Obama administration would pay any attention.

The law requires the Department to take cost-effectiveness into account when it considers applications for funds for streetcars, light rail, and other “New Starts” transit projects. But the Federal Transit Administration had always given this only token consideration until Bush’s second Secretary of Transportation, Mary Peters, put some strict limits on just how expensive projects could be if they were to get any federal funds.

Secretary LaHood chafed at these limits, particularly because they prevented any funds being given for streetcars. So he announced in 2010 that he was going to get rid of the limits. On behalf of the Cato Institute, the Antiplanner commented on LaHood’s proposal to make the change and then commented on the draft rules.

My main argument was that cost-effectiveness is not a simple yes-or-no question; instead, it is a gradient that requires consideration of a wide range of alternatives. Even under Mary Peters, the DOT process treated the alternatives analysis and the cost-effectiveness analysis as two separate steps: first develop alternatives and reject the ones you don’t like; second do a cost-effectiveness analysis on the preferred alternative.

That is not how cost-effectiveness works. Without the context of alternatives, cost-effectiveness is a meaningless number. The Department even agreed with this in its reply to my comments that was bundled with the draft rules. However, the draft rules did not change the process.

Instead, the issue for the Federal Transit Administration (and LaHood) was how cost-effectiveness was measured. Under rules that had been set before Peters, cost-effectiveness was calculated in terms of cost per hour of travel time that a transit project saved. Peters’ twist was to say that any project that cost more than $25 per hour saved was automatically ruled out.

The transit industry hated this rule. New Starts were a big pot of money, and historically the agencies that developed the most ridiculously expensive projects got the largest shares of the pot. They certainly didn’t want to be held to any standards.

Some agencies persuaded Congress to exempt a few of the biggest boondoggles–Dulles rail; BART to San Jose; Portland’s commuter-rail fiasco; and San Francisco’s Central Artery–from the rule. Many other agencies had to go back and trim their project costs so that they would come in at, say, $24.95 per hour. A few agencies even admitted that their projects were so costly that even the federal government wouldn’t contribute to them, hoping local voters would agree to fund the entire scams themselves–but for some reason this rarely worked.

Placating the industry, last year’s draft rules replaced travel time with a measure based simply on the cost per transit rider carried by the project. In other words, where the old rules credited a project for reducing congestion (and thus saving travel time), the new rules only credited the project for carrying transit riders even if it made travel worse for everyone else. In fact, making congestion worse might actually favor a project because it could lead to more transit riders. Of course, the draft rules have no threshold, so a project could be funded even it if cost a million dollars per rider.

In addition, Peters’ rule for streetcar projects, known in the law as “small starts,” was that transit agencies had to show that streetcars were more cost-effective than buses. The industry hated that rule too, and because of it most federal dollars for streetcar projects came from stimulus funds, which were under a different set of rules, not transportation funds. Of course, the draft rules deleted this requirement as well.

LaHood was personally responsible for the other big change in the draft rules: adding new criteria for making transit grants, specifically, “livability,” “environmental justice,” and “multi-modality.” The Antiplanner argued that the Department of Transportation was not legally authorized to add these new criteria; but the Department said that a phrase in the law, “other factors,” gave it that authority.

The final rules make two significant changes from the draft. First, in an ironic turn of events, the Senate Democrats who wrote much of the most recent transportation bill, Map-21, left out the “other factors” phrase, so the final rules had to cut LaHood’s dream of giving away money based on “livability,” “environmental justice,” and “multi-modality.”

The new law also deleted mention of an alternatives analysis in recognition of the fact that this is redundant with the National Environmental Policy Act, which requires an alternatives analysis in the environmental impact statements for major projects. The Department responded by simply eliminating any mention of alternatives at all in the rules. An entire lengthy section of both the old rules and the draft rules, ยง611.209, was completely deleted.

Legally, the environmental impact statements should include a wide range of alternatives, an idea affirmed by a recent court decision over a rail project in Honolulu. Potentially, the comparison of the cost-effectiveness of the alternatives could be done there. But don’t count on it: the Department has never required transit agencies to consider much more than a “no action” alternative in their environmental impact statements, and (as the people who objected to the Honolulu EIS learned) citizen enforcement of this law can be very expensive.

In short, the new rules–which go into effect on April 9–come down to this: if your transit agency can calculate how much it will cost to carry each rider, it will be eligible for federal funding. It won’t matter what the cost is, nor will it matter if other alternatives (such as buses instead of streetcars or light rail) could carry those riders for far less money. Naturally, this doesn’t bode well for those who think that taxpayer dollars should be spent only where they can do the most good.


2 thoughts on “LaHood’s Cost-Effectiveness Rule

  1. bennett

    The problem with the rules as set up by LaHood or Peters is that it gives so much leeway to agencies to self-police. In the context of how streetcar projects are planned and implemented, the consideration of “alternatives” is a complete joke. The decision to build a streetcar is made eons before any “alternative” is cooked up. Once it comes time to develop alternatives it is done in a way to make the predetermined decision look as good as possible.

    I personally like streetcars and use them when I visit places that have them. However, as a proponent of public transit I would like to see more resources go to transit and mobility projects that help the cohorts that need it most and put streetcar projects (which from what I can tell are mostly about real estate and development) on the back burner.

  2. MJ

    In other words, where the old rules credited a project for reducing congestion (and thus saving travel time), the new rules only credited the project for carrying transit riders even if it made travel worse for everyone else.

    Another major problem with this approach is that it doesn’t focus on how many new riders a project attracts. In other words, a rail line that simply replaces a highly productive bus line would be viewed favorably, even if it wasn’t really adding any new transit users. One can see how this approach would be particularly favorable to streetcar projects, which rarely attract many new riders due to their slow operating speeds, and are often concentrated in the densest neighborhoods of urban areas. Light rail projects would also benefit, though to a lesser extent.

    My favored approach would be to dispense with cost-effectiveness analysis entirely and go back to using straightforward benefit-cost criteria to evaluate these projects. You would know whether you were measuring the right things, you would not be subject to arbitrary thresholds, and all of the assumptions that went into the analysis could be stated upfront and then fairly debated.

Leave a Reply