Sw. Airlines Co. v. Fed. Energy Regulatory Comm'n ( 2019 )


Menu:
  •  United States Court of Appeals
    FOR THE DISTRICT OF COLUMBIA CIRCUIT
    Argued April 12, 2019                  Decided June 14, 2019
    No. 18-1134
    SOUTHWEST AIRLINES CO. AND AMERICAN AIRLINES, INC.,
    PETITIONERS
    v.
    FEDERAL ENERGY REGULATORY COMMISSION AND UNITED
    STATES OF AMERICA,
    RESPONDENTS
    SFPP, L.P.,
    INTERVENOR
    Consolidated with 18-1136, 18-1137, 18-1138
    On Petitions for Review of Orders of the
    Federal Energy Regulatory Commission
    Steven A. Adducci argued the cause for petitioners. With
    him on the briefs were Thomas J. Eastment, Gregory S.
    Wagner, Matthew D. Field, and Richard E. Powers Jr.
    Anand R. Viswanathan, Attorney, Federal Energy
    Regulatory Commission, argued the cause for respondents.
    With him on the brief were Robert J. Wiggers and Robert B.
    Nicholson, Attorneys, U.S. Department of Justice, James P.
    2
    Danly, General Counsel, Federal Energy Regulatory
    Commission, Robert H. Solomon, Solicitor, and Elizabeth E.
    Rylander, Attorney. Robert M. Kennedy Jr., Attorney, Federal
    Energy Regulatory Commission, entered an appearance.
    Charles F. Caldwell argued the cause for intervenor. With
    him on the brief were Daniel W. Sanborn, Michelle T.
    Boudreaux, and Sabina D. Walia.
    Before: TATEL, MILLETT, and KATSAS, Circuit Judges.
    Opinion for the Court filed by Circuit Judge TATEL.
    TATEL, Circuit Judge: The Federal Energy Regulatory
    Commission uses a streamlined “indexing” method to ensure
    that when oil pipelines raise their rates, the resulting charges
    remain reasonable. Every summer, the Commission calculates
    an “index” that reflects inflation between the previous two
    calendar years, and pipelines may, through an expedited
    process, rely on that index to increase their rates. If a pipeline’s
    customers believe that a particular rate increase, though index-
    compliant, is still too high, then they may challenge that rate in
    a proceeding before the Commission. These consolidated cases
    concern the kind of evidence the Commission deems relevant
    to such proceedings. In 2014, a group of customers filed
    complaints against the 2012 and 2013 index-based rate
    increases implemented by pipeline-owner SFPP, L.P. The
    Commission, departing from its previous practice, dismissed
    those complaints by relying on data generated after the
    challenged increases went into effect. Because the Commission
    failed to provide sufficient reasons for changing its policy, we
    vacate the challenged orders and remand for the Commission
    to explain or reconsider its decision to take into account post-
    rate-increase information.
    3
    I.
    For over a century, oil pipelines have been subject to
    regulation as common carriers under the Interstate Commerce
    Act. See Act of June 29, 1906, Pub. L. No. 59-337, § 1, 34 Stat.
    584, 584 (extending the Interstate Commerce Act’s definition
    of “common carriers” to include oil pipelines). For most of this
    time, the pipelines’ federal regulators—first the Interstate
    Commerce Commission and now the Federal Energy
    Regulatory Commission—used complex “fair value” or “cost-
    based” ratemaking methodologies, Ass’n of Oil Pipe Lines v.
    FERC, 
    83 F.3d 1424
    , 1428–29 (D.C. Cir. 1996) (internal
    quotation marks omitted), to prevent pipelines from unlawfully
    charging “unjust and unreasonable” rates, 49 U.S.C. app.
    § 1(5)(a) (1988). In the Energy Policy Act of 1992, however,
    Congress directed the Federal Energy Regulatory Commission
    to “streamline [its] procedures” and reduce “unnecessary
    regulatory costs and delays” by “establish[ing] a simplified and
    generally applicable ratemaking methodology for oil
    pipelines.” Pub. L. No. 102-486, §§ 1801(a), 1802(a), 106 Stat.
    2776, 3010.
    As a result, an “indexing” scheme has replaced cost-of-
    service proceedings as the Commission’s primary tool for
    regulating pipeline rates. See Revisions to Oil Pipeline
    Regulations Pursuant to the Energy Policy Act of 1992, Order
    No. 561, 58 Fed. Reg. 58,753, 58,754 (Nov. 4, 1993)
    (explaining that the “Commission believes that indexing of oil
    pipeline rates will eliminate the need for much future cost-of-
    service litigation”). Emphasizing that “the hallmark of an
    indexing system is simplicity,” the Commission explained that
    pipelines (also called “carriers”) could use the new method to
    “adjust [their] rates . . . for inflation-driven cost changes
    without the need [for] strict regulatory review of the pipeline’s
    individual cost of service.” 
    Id. at 58,758.
    By permitting the
    “nominal level of rates to rise” with “general economy-wide
    4
    costs,” the Commission stated, “indexing, conceptually,
    [would] merely preserve[] the value of just and reasonable rates
    in real economic terms.” 
    Id. at 58,759.
    The nuts and bolts of indexing work like this: For every
    “index year,” which runs from July 1 to June 30, the
    Commission publishes no later than June 1 an index “based on
    the change in the final Producer Price Index for Finished Goods
    (PPI-FG) . . . for the two calendar years immediately preceding
    the index year.” 18 C.F.R. § 342.3(c), (d)(1), (d)(2). So, for
    example, the Commission recently calculated the index for the
    twelve-month period spanning July 1, 2019, to June 30, 2020,
    by comparing the 2018 PPI-FG to the 2017 PPI-FG. See
    Revisions to Oil Pipeline Regulations Pursuant to the Energy
    Policy Act of 1992, Notice of Annual Change in the Producer
    Price Index for Finished Goods, 167 FERC ¶ 61,122, at 1 (May
    10, 2019). Once an index is set, each pipeline then computes
    its own maximum allowable rate—its so-called ceiling level—
    “by multiplying the previous index year’s ceiling level by the
    [Commission’s] most recent index.” 18 C.F.R. § 342.3(d)(1).
    A pipeline may “at any time” increase its rates “to a level which
    does not exceed [its] ceiling level.” 
    Id. § 342.3(a).
    The Commission recognizes that, though efficient, an
    indexing scheme based on “economy-wide costs” may at times
    produce rates significantly out of step with individual
    pipelines’ financial realities. Revisions to Oil Pipeline
    Regulations Pursuant to the Energy Policy Act of 1992, 58 Fed.
    Reg. at 58,759. For this reason, the Commission permits
    pipeline customers (also called “shippers”) to “challenge
    existing rates, even if such rates are below the applicable
    ceiling levels, if [those customers] reasonably believe such
    rates are excessive.” 
    Id. at 58,754.
    These index-based rate
    challenges come in two varieties: protests, which address
    proposed rates, and complaints, which address “existing rate[s]
    5
    or practice[s].” 18 C.F.R. § 343.1. In both types of proceedings,
    the challenger must “allege reasonable grounds for asserting
    . . . that the rate increase is so substantially in excess of the
    actual cost increases incurred by the carrier that the rate is
    unjust and unreasonable.” 
    Id. § 343.2(c)(1).
    How the
    Commission evaluates those allegations, however, depends on
    whether the shipper brings its challenge in the form of a protest
    or a complaint.
    Because protests proceed extremely quickly—they must
    be filed within fifteen days of a rate’s publication, see 
    id. § 343.3(a),
    and the Commission has only thirty days from the
    rate’s filing date to “determine whether to . . . initiate a formal
    investigation,” 
    id. § 343.3(c)—the
    Commission evaluates
    protests with a “quick snapshot approach” called the
    “percentage comparison test,” BP West Coast Products, LLC
    v. SFPP, L.P., 121 FERC ¶ 61,141, at PP 6–7 (2007). Using
    annual cost data found on page 700 of the pipeline’s “Form No.
    6,” the Commission performs the percentage comparison test
    by computing “the change in the prior two years’ total cost-of-
    service data.” SFPP, L.P., 163 FERC ¶ 61,232, at P 4 (2018);
    see also 18 C.F.R. § 357.2 (detailing oil pipelines’ annual
    reporting obligations). “[I]f there is [a] 10 percent or more
    differential between” the percentage-point change in the
    pipeline’s costs and the percentage-point change in its
    proposed rate, then “the Commission will investigate [the]
    protested indexed rate change.” SFPP, 163 FERC ¶ 61,232, at
    P 4; see, e.g., North Dakota Pipeline Co., 163 FERC ¶ 61,235,
    at P 11 (2018) (the Commission would investigate a July 2018
    index-based rate increase of 4.41% when costs declined by
    15.5% from 2016 to 2017).
    In contrast to protests, complaints are subject to a two-year
    statute of limitations, see 49 U.S.C. app. § 16(3)(b) (1988), and
    the Commission enjoys a “more extended time frame in which”
    6
    to consider them, BP West Coast Products, 121 FERC
    ¶ 61,141, at P 7. As a result, in the context of complaints only,
    the Commission interprets the regulatory phrase “substantially
    in excess of the [pipeline’s] actual cost increases,” 18 C.F.R.
    § 343.2(c)(1), to “provid[e] for the review of either a
    percentage increase or a dollar increase” in costs, BP West
    Coast Products, 121 FERC ¶ 61,141, at P 5; see also BP West
    Coast Products LLC v. SFPP, L.P., 123 FERC ¶ 61,121, at P 6
    (2008) (explaining that “the Commission only applies a
    percentage test when reviewing a protest and normally applies
    that test for complaints,” but will “use[] a dollar comparison
    . . . under . . . limited circumstances”). To determine whether a
    pipeline’s “dollar increase” is excessive, the Commission
    applies the so-called substantially exacerbate test, under which
    a complaint “must show (1) that the pipeline is substantially
    over-recovering its cost of service and (2) that the indexed
    based [rate] increase so exceeds the actual increase in the
    pipeline’s cost that the resulting rate . . . would substantially
    exacerbate that over-recovery.” BP West Coast Products, 121
    FERC ¶ 61,141, at PP 5, 10. Put simply, an index-based
    increase might produce a rate “substantially in excess of the
    [pipeline’s] actual cost increases,” 
    id. at P
    5, if the pipeline’s
    revenues are already significantly higher than its costs and if
    its rate increase amplifies that over-recovery.
    This case began in June 2014, when several shippers (the
    “Shippers”) filed timely complaints alleging that SFPP’s 2012
    and 2013 index-based rate increases failed the substantially
    exacerbate test. Claiming that SFPP was already over-
    recovering its costs at the time it applied its rate increases in
    2012 and 2013, the Shippers, citing page 700 data showing that
    SFPP’s costs had decreased between the two years preceding
    each rate increase, argued that the new, higher rates “would
    substantially exacerbate” those over-recoveries. 
    Id. at P
    10.
    Specifically, the Shippers alleged that (1) SFPP experienced a
    7
    4.48% decrease in costs between 2010 and 2011, yet in 2012 it
    implemented a rate increase that would exacerbate its 2011
    over-recovery of $18,368,119 by at least $6.9 million; and (2)
    SFPP experienced another 0.56% decrease in costs between
    2011 and 2012, yet in 2013 it implemented a rate increase that
    would exacerbate its 2012 over-recovery of $14,323,805 by at
    least $7.15 million.
    The Commission dismissed the complaints in December
    2016. See Hollyfrontier Refining & Marketing LLC v. SFPP,
    L.P., 157 FERC ¶ 61,186, at P 1 (2016). Its logic was simple:
    “[n]otwithstanding the application of the 2012 and 2013 index
    increases,” the Commission explained, “SFPP’s Page 700s on
    file at the time of the complaints show[ed] that the difference
    between SFPP’s costs and revenues declined from . . . 2011 [to]
    2012 [to] 2013.” 
    Id. at P
    9. Consequently, because “the 2012
    and 2013 index increases did not, in fact, substantially
    exacerbate the pre-existing difference between SFPP’s
    revenues and costs,” the Commission concluded that the
    complaints “fail[ed] the second part of the ‘substantially
    exacerbate’ test.” 
    Id. In dismissing
    the complaints, the Commission “reject[ed]
    the . . . Shippers’ contention that [it] should only evaluate the
    complaints based upon the two years prior to each index
    increase, i.e., (a) 2010 and 2011 Page 700 data for . . . [the]
    2012 index increase and (b) 2011 and 2012 Page 700 data for
    . . . [the] 2013 index increase,” 
    id. at P
    10, and instead chose to
    consider “the facts available at the time . . . the complaints”
    were filed in June 2014, 
    id. at P
    9. Acknowledging that it had
    “previously held that the only relevant data for evaluating an
    index rate change are the data from the two years prior to the
    index change,” the Commission stated that it had “applied this
    policy when investigating . . . protest[s] within 15 days of the
    challenged indexed rate filing.” 
    Id. at P
    10. In this proceeding,
    8
    by contrast, the “Shippers waited two years after the 2012 rate
    increase and one year after the 2013 index increase to file their
    complaints,” so, according to the Commission, “[t]his case
    present[ed] different circumstances” than the Commission had
    encountered before. 
    Id. The Commission
    denied the Shippers’ request for
    rehearing in March 2018. See Hollyfrontier Refining &
    Marketing LLC v. SFPP, L.P., 162 FERC ¶ 61,232 (2018).
    Reiterating that its December 2016 order had “interpret[ed] the
    Commission’s rate complaint regulations . . . in a context that
    the Commission had not previously had occasion to address”—
    that is, a “situation where additional Page 700 data was
    available to shed light on the allegations contained in the . . .
    Shippers’ complaints”—the Commission explained that “when
    shippers delay challenging [index-based] rates for one or two
    years, a different process may be employed to take into account
    data that became available prior to the complaint.” 
    Id. at P
    P 13–14, 16. The Commission “elected to use that data”
    because, in its view, “it would be inefficient and inequitable to
    ‘ignore evidence that was available at the time the . . . Shippers
    filed their complaints’ when that information ‘undermines the
    basis of the . . . Shippers’ claim.’” 
    Id. at P
    14 (quoting
    Hollyfrontier Refining & Marketing, 157 FERC ¶ 61,186, at
    P 10). The Shippers timely filed petitions for review.
    II.
    One of the most fundamental principles of administrative
    law is that agencies must give reasons for their actions. The
    Administrative Procedure Act directs courts to enforce this
    obligation by “hold[ing] unlawful and set[ting] aside agency
    action[s]” that are “arbitrary” or “capricious,” 5 U.S.C.
    § 706(2)(A), a task that requires courts “not to substitute [their]
    judgment for that of the agency” but rather to ensure that
    whatever the agency has decided, it has “examine[d] the
    9
    relevant data and articulate[d] a satisfactory explanation” for
    its actions, Motor Vehicle Manufacturers Ass’n v. State Farm
    Mutual Automobile Insurance Co., 
    463 U.S. 29
    , 43 (1983).
    Although “[w]e will . . . uphold a decision of less than ideal
    clarity if the agency’s path may reasonably be discerned,” we
    may not “attempt . . . to make up for such deficiencies”
    ourselves by “supply[ing] a reasoned basis for the agency’s
    action that the agency itself has not given.” 
    Id. (internal quotation
    marks omitted).
    “A full and rational explanation” becomes “especially
    important” when, as here, an agency elects to “shift [its]
    policy” or “depart[] from its typical manner of” administering
    a program. Great Lakes Gas Transmission Ltd. Partnership v.
    FERC, 
    984 F.2d 426
    , 433 (D.C. Cir. 1993). The agency “need
    not demonstrate . . . that the reasons for the new policy are
    better than the reasons for the old one,” FCC v. Fox Television
    Stations, Inc., 
    556 U.S. 502
    , 515 (2009), but it must at least
    “acknowledge” its seemingly inconsistent precedents and
    either offer a reason “to distinguish them” or “explain its
    apparent rejection of their approach,” Tennessee Gas Pipeline
    Co. v. FERC, 
    867 F.2d 688
    , 692 (D.C. Cir. 1989). This is not
    an especially high bar: “it suffices that the new policy is
    permissible under the statute, that there are good reasons for it,
    and that the agency believes it to be better, which the conscious
    change of course adequately indicates.” Fox 
    Television, 556 U.S. at 515
    . But however the agency justifies its new position,
    what it may not do is “gloss[] over or swerve[] from prior
    precedents without discussion.” Greater Boston Television
    Corp. v. FCC, 
    444 F.2d 841
    , 852 (D.C. Cir. 1970).
    Until this case, when considering protests and complaints
    alike, the Commission always relied exclusively on data from
    the two calendar years preceding the challenged rate to
    determine whether the increase was “substantially in excess of
    10
    the actual cost increases incurred by the carrier.” 18 C.F.R.
    § 343.2(c)(1). But in the orders at issue in this case, the
    Commission dismissed the Shippers’ complaints based on
    financial data generated after the challenged index-based rates
    had taken effect. The Commission concedes as much. See Oral
    Arg. Rec. 21:12–26 (conceding that the Commission is
    unaware of any previous protest or complaint case in which it
    considered post-rate-increase information).
    What the parties dispute is just how far the Commission
    has journeyed from its previously trodden path. The
    Commission tells us that it has done nothing more than
    consider the best available information—information which,
    given the “unusual circumstance” occasioned by the Shippers’
    “filing delay,” happens to include the “more recent and more
    representative data” generated by SFPP’s 2012 and 2013
    index-based rate increases. Respondent’s Br. 13–15. The
    Shippers see things differently. In their view, the
    Commission’s “decision does not simply take into account
    updated evidence,” but rather “reflects a fundamental change
    in the standard for evaluating an index-based rate complaint.”
    Petitioners’ Br. 31. For two reasons, the Shippers have the
    better of the argument.
    First, the Commission has explained that it relies on pre-
    rate-increase information not because it lacks more recent
    evidence, but rather because prior-year data reflects precisely
    what indexing is supposed to measure: cost changes in the
    previous year. For example, in a 2009 complaint proceeding
    the Commission distinguished “general rate” cases from index-
    based challenges by explaining that, “[i]n contrast” to “a
    general rate case” that “looks forward,” “[t]he indexing method
    looks backward to the prior year.” Tesoro Refining &
    Marketing Co. v. SFPP, L.P., 129 FERC ¶ 61,114, at P 17
    (2009). “[T]he indexing methodology . . . is based on
    11
    annualized cost changes between two calendar years,” the
    Commission elaborated, and as such it “relies on actual
    historical costs, not those that may be projected or updated in a
    general rate case.” 
    Id. at P
    P 17–18. Similarly, in a 2012 order
    denying a pipeline’s request to reopen and supplement the
    record with post-rate-increase information, the Commission
    reaffirmed that “[t]he only relevant evidence in indexing cases
    is the change in the pipeline’s cost-of-service in the two years
    preceding the index increase.” SFPP, L.P., 140 FERC
    ¶ 61,016, at P 34 (2012). To be sure, the Commission could
    have rejected the updated information because, as it now
    argues, “continual additions” “once a proceeding has
    commenced . . . would be inconsistent with . . . streamlined
    ratemaking.” Hollyfrontier Refining & Marketing, 157 FERC
    ¶ 61,186, at P 10. But that is not the reason the Commission
    gave. Instead, it declared that the proffered “later-developed
    data [was] irrelevant” because “[w]hen ruling on a proposed
    index increase, the Commission confines its inquiry to
    comparing the year-to-year change in costs . . . for the two
    preceding years.” SFPP, 140 FERC ¶ 61,016, at PP 34, 42.
    Indeed, despite this court’s equivocal dictum on the matter, the
    Commission’s past practice demonstrates that it has always
    employed a backward-looking approach in indexing
    proceedings. Compare United Airlines, Inc. v. FERC, 
    827 F.3d 122
    , 133 (D.C. Cir. 2016) (stating that “whether [the
    Commission’s] indexing mechanism is retrospective or
    prospective is unclear”), with Oral Arg. Rec. 21:12–26 (failing
    to identify any indexing proceeding in which the Commission
    considered post-rate-increase information).
    Second, in at least three previous complaint cases, the
    Commission focused solely on pre-rate-increase information
    from the preceding two years even though post-rate-increase
    information was presumably available at the time the
    complaints were filed. In one case, the Commission dismissed
    12
    complaints filed in December 2006 and January 2007 against a
    pipeline’s 2005 and 2006 index-based rate increases because,
    as it concluded, the pipeline permissibly “indexed its . . . rates
    on July 1, 2005, to reflect that its costs in 2004 exceeded its
    2003 costs” and then imposed a “July 1, 2006[,] index-based
    increase” on the basis of “Page 700 [data] for the calendar year
    2005 reflect[ing] an increase in costs” from 2004. BP West
    Coast Products LLC v. SFPP, L.P., 118 FERC ¶ 61,261, at
    PP 8–9 (2007). In another case, the Commission explained that
    it would evaluate a complaint filed in 2007 against a 2005
    index-based rate “by comparing the costs incurred [by the
    pipeline] in the calendar year preceding the index year with the
    prior year”—that is, by comparing “the pipeline’s costs in 2004
    with the costs incurred in 2003.” BP West Coast Products LLC
    v. SFPP, L.P., 119 FERC ¶ 61,241, at P 9 (2007). And in still
    another case, the Commission dismissed a 2007 complaint
    against a 2006 index-based rate because the pipeline had
    demonstrated with its “revised 2005 FERC Form No. 6” that
    its “July 2006 . . . index based increase[] did not substantially
    exacerbate its current over-recovery.” Tesoro Refining &
    Marketing Co. v. Calnev Pipe Line, LLC, 121 FERC ¶ 61,142,
    at P 7 (2007).
    These three decisions—all cited by the Commission in its
    December 2016 order or by the Shippers in their request for
    rehearing—belie the Commission’s contention that there was
    something particularly unusual about the Shippers’ “delayed
    filing” and, as a result, that their complaints “presented
    different circumstances” than the Commission “previously had
    occasion to address.” Respondent’s Br. 13, 19 (internal
    quotation marks omitted). True, as the Commission now points
    out, because no party to any of those proceedings expressly
    asked the Commission to consider such updated data, these
    decisions do not hold that post-rate-increase information is
    irrelevant. See Oral Arg. Rec. 21:28–56 (arguing that “the
    13
    question was never teed up”). But regardless of what issues the
    litigants raised in those cases, the fact remains that the
    Commission repeatedly used pre-rate-increase data to answer
    the question at hand—whether the rate increases were
    “substantially in excess of the actual cost increases incurred by
    the carrier,” 18 C.F.R. § 343.2(c)(1)—even though post-rate-
    increase information was available. Challenged by litigants or
    not, an agency’s “settled course of behavior embodies [that]
    agency’s informed judgment that, by pursuing that course, it
    will carry out the policies committed to it by Congress.”
    Atchison, Topeka & Santa Fe Railway Co. v. Wichita Board of
    Trade, 
    412 U.S. 800
    , 807 (1973) (plurality opinion).
    Consequently, the Commission’s consistent practice, whether
    adopted expressly in a holding or established impliedly through
    repetition, sets the baseline from which future departures must
    be explained. See 
    id. (plurality opinion)
    (explaining that
    adjudications “generally provide a guide to action that the
    agency may be expected to take in future cases”).
    Taken together, these cases demonstrate that when the
    Commission announced its decision to “consider the data that
    [becomes] available” “[w]hen shippers delay . . . in filing a
    complaint,” Hollyfrontier Refining & Marketing, 162 FERC
    ¶ 61,232, at P 18, it was adopting a policy inconsistent with its
    earlier course of conduct. Of course, the Commission is free to
    “depart from a prior policy or line of precedent” so long as it
    “acknowledge[s] that it is doing so and provide[s] a reasoned
    explanation.” Louisiana Public Service Commission v. FERC,
    
    772 F.3d 1297
    , 1303 (D.C. Cir. 2014). But the explanation
    offered in the challenged orders misses this mark.
    The Commission’s sole justification for its change of heart
    boils down to this: “it would be inefficient and inequitable to
    ignore evidence that was available at the time the . . . Shippers
    filed their complaints.” Hollyfrontier Refining & Marketing,
    14
    162 FERC ¶ 61,232, at P 14 (internal quotation marks omitted).
    This justification, however, begs a very important question: is
    the available evidence also relevant evidence? As the Shippers
    point out, if “the index is designed to recover cost increases for
    the period prior to the increase,” then “[d]ata relating to periods
    after the effective date of a proposed index rate increase are
    irrelevant.” Petitioners’ Br. 23. In other words, by assuming
    that any available post-rate-increase information is relevant to
    its inquiry, the Commission has reinterpreted—without
    acknowledgement or explanation—the phrase “actual cost
    increases incurred by the carrier,” 18 C.F.R. § 343.2(c)(1), to
    include not only costs incurred before the rate’s filing, but also
    costs incurred before the complaint’s filing. See Oral Arg. Rec.
    31:50–32:44 (conceding that the challenged orders changed the
    Commission’s interpretation of “incurred”). And that
    reinterpretation, in turn, calls into question the purpose of
    indexing itself. Are index-based rate increases designed to
    compensate pipelines for cost increases actually incurred in the
    previous calendar year, costs likely incurred in the current
    calendar year, or, depending on the type of proceeding, both?
    The Commission tells us that we “need not address” the
    purpose of indexing in order “to resolve this case.”
    Respondent’s Br. 21. To an extent, we agree: we would stray
    too far from our judicial function were we to venture a guess
    ourselves. See Ass’n of Oil Pipe 
    Lines, 83 F.3d at 1431
    (explaining that because “ratemaking . . . involv[es] complex
    industry analyses and difficult policy choices,” courts should
    be “particularly deferential to the Commission’s expertise”).
    But we are not convinced—at least on this record—that the
    Commission may, consistent with its obligation to engage in
    reasoned decision making, allow the question to go
    unanswered. We shall therefore vacate and remand the
    challenged orders so that the Commission, should it choose to
    maintain its new policy of considering information that
    15
    becomes available between a pipeline’s rate increase and a
    shipper’s complaint, can offer a reasoned explanation that
    either persuasively distinguishes or knowingly abandons its
    prior inconsistent practice. See Tennessee Gas 
    Pipeline, 867 F.2d at 692
    (remanding orders in which the Commission had
    “neither acknowledge[d]” apparently inconsistent “precedents
    nor purport[ed] either to distinguish them or to explain its
    apparent rejection of their approach”). Though expressing no
    opinion on how the Commission should apply the substantially
    exacerbate test going forward, we emphasize that however the
    Commission chooses to proceed, it must explain its actions in
    a way that coheres with the rest of its indexing scheme—
    namely, the manner in which it establishes yearly indexes and
    the methods it uses to evaluate challenges to index-based rates.
    In short, the Commission must provide a reasoned explanation
    that treats like cases alike.
    One final matter requires brief mention. In addition to
    arguing that the Commission departed from its prior practice
    without adequate justification, the Shippers claim that, on the
    same day the Commission issued its March 2018 order, it
    issued a different order that undermines the evidentiary basis
    for dismissing the Shippers’ complaints. But given that we are
    vacating the March 2018 order, we need not reach this
    alternative ground for granting the petitions for review. The
    Shippers are free to raise this argument on remand.
    III.
    For the foregoing reasons, we grant the petitions for
    review and vacate and remand the Commission’s December
    2016 and March 2018 orders for further proceedings consistent
    with this opinion.
    So ordered.