Docket No. RM26-6-000
Today’s order exists because Congress directed the Commission to implement a “simplified and generally applicable ratemaking methodology” for oil pipelines as an efficient way to uphold the Interstate Commerce Act’s just and reasonable standard.[1] So every five years we undertake a rigorous and data-driven analysis of the industry-wide cost changes that pipelines experience in an effort to recalibrate the index to current market conditions. Our aim is to reasonably align pipeline ceiling rates with the oil sector’s economic realities in an administratively streamlined fashion, as an alternative to case-by-case rate cases. This practice supports fair compensation for carriers, promotes incentives for investing in critical infrastructure, and maintains reliable price signals for oil transportation. The results of the index itself are a small part of overall consumer rates, but our regular indexing effort is an important tool within the sector. Indexing works in practice because it is transparent, repeatable, and predictable, all of which ultimately supports fair compensation for carriers and reliable price signals for shippers and investors.
It’s easy to agree on those goals behind periodic indexing. But in practice, applying a transparent indexing process is still a technical and highly record-driven task. Reasonable minds can draw potentially different outcomes from the data before us within a given cycle, and reasonable minds can take different approaches within the Commission’s overall methodology across cycles, too. I leave to the Final Rule to explain in detail why the Commission landed on the specifical technical judgments in this index iteration. I write here briefly to underscore that some variation across index cycles can be a feature, not a bug, when it comes to applying Commission expertise to a record that’s grounded in the unique circumstances and policies inherent to any five-year stretch of time.
Focusing on our decision to trim to the middle 80%: The Commission has long recognized that we must do some statistical trimming to keep outliers from distorting industry cost trends. The decision where to trim oil pipeline cost data is record driven.[2] It may not always be true that more data is better than less, but in this cycle it is. Here, using the middle 80% provides a robust and representative sample of pipeline cost experience, avoids prematurely discarding relevant observations, and better reflects the diversity of operations and barrel‑mile coverage than narrower bands would. The middle 80% of the data set comprises 94% of industry-wide barrel miles, which helps give confidence that the index is representative while omitting true outlying data that would distort the calculation. This approach also aligns with Commission practice in the most recent five‑year review cycle.
True, the Commission has not always trimmed to 80%. But data trimming is as much a record-grounded art as a math equation. Different trimming bands (50%, 80%, or otherwise) could be reasonable in different cycles depending on the nature of industry activity during the measurement period that bears on how reported cost changes are distributed. Our choice of an 80% trimming band here reflects our judgment that this record warrants broader representativeness; it doesn’t imply that narrower trimming would be unreasonable in future cycles. For example, if we see more pipeline construction and targeted expansions of existing systems in the next five years—and given our nation’s infrastructure needs, I hope that we do—those operational- and market-driven changes would likely bring more outlier pipelines to our cost change analysis. Faced with a record like that, a narrower trimming band might end up better capturing the “normal” industry experience by eliminating any non-representative cost spikes.
What matters most to my mind is consistency in methodology, not necessarily in outcome. Our goal in applying the Commission’s methodology is to fairly collect and validate cost experience, then remove outliers to avoid distortion, and finally set an index that credibly tracks central tendencies over the last five years. As industry conditions change, the precise trimming band the method yields can change, too. And in cases where the record could potentially support multiple outcomes, we’re called to apply our judgment in choosing where best to draw the line given the record before us. In this proceeding, trimming to the middle 80% meets that responsibility.
Given the facts of this case, I respectfully concur with today’s order.
[1] 49 U.S.C. app. § 1 et seq. (1988).
[2] See, e.g., Five-Year Review of Oil Pipeline Index, 133 FERC ¶ 61,228, at PP 61-63 (2010) (explaining that data trimming to the middle 50 percent of pipelines is appropriate based on the record); Five-Year Rev. of the Oil Pipeline Index, 173 FERC ¶ 61,245, at PP 25-32 (2020) (explaining that it is preferable to consider additional data and trim to the middle 80 percent of pipelines because it more fully reflects the diversity of cost experiences based on the record).