STB Petition to Consider Benefit-Cost Analysis

January 31, 2020

Download this Public Interest Comment (PDF)


On July 8, 2019, the STB decided to delay consideration of a petition asking the board to adopt a procedural rule that would require benefit-cost analysis in certain board rulemakings. On November 4, 2019, the STB solicited further information from the public about specific methods that could be used for benefit-cost analysis of rules related to economic regulation of freight railroads. The STB is prudent to explore methods for improving its economic analysis of regulatory proposals—as several other independent agencies have done in recent years.

The board asked commenters to address a hypothetical rule that would reduce the revenue/variable cost (R/VC) percentage threshold used to infer that a railroad is not market dominant from 180 percent to 165 percent. At the time the board released its SOI, I had already drafted a comment outlining how the analytical framework in Office of Management and Budget (OMB) Circular A-4, OMB’s regulatory impact analysis guidance, could be applied in the Market Dominance Streamlined Approach and Final Offer Rate Review proceedings, and I submitted that comment both in those proceedings and in response to the SOI.

This comment briefly explains how the principles and concepts outlined in my earlier comment could be applied to the hypothetical regulation on which the STB requested comment.

Regulatory impact analysis should start with an evidence-based analysis of the problem the regulation seeks to solve. Unfortunately, the SOI does not even postulate a particular problem. I know of no economic literature that demonstrates that setting the R/VC threshold at 180 percent of variable cost creates any kind of systematic problem that can be solved by reducing the threshold to 165 percent. The use of R/VC ratios based on the Uniform Rail Costing System (URCS) to infer whether a railroad is market dominant has been roundly criticized by economists, but the solution to this problem is to stop using R/VC ratios and employ different criteria. Consequently, a sound RIA for this hypothetical rulemaking should include rate benchmarking as an alternative to using R/VC ratios as a screen to determine which rates the STB should examine more closely.

A reduction in the R/VC threshold from 180 to 165 percent could create transfers from railroads to shippers, and I suggest how the STB could generate an upper bound estimate of the size of the transfers. Any such estimate will be subject to two significant uncertainties that should be accounted for in the analysis: (1) It is uncertain what percentage of rate cases railroads will win, and (2) it is uncertain whether reducing the R/VC threshold alone will generate significantly more rate cases or give shippers significantly more leverage in their negotiations with railroads.

The social benefits of any rate reductions captive shippers receive are likely to be small, because captive shippers by definition have no good transportation alternatives and hence have a low elasticity of demand for rail service. Significant social costs would arise only if the transfers from railroads to shippers are large enough to affect railroad investment in ways that reduce the quantity, quality, or safety of rail service.