United Parcel Service , Inc . v. Postal Regulatory Commission

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued December 5, 2023 Decided August 23, 2024
No. 22-1029
UNITED P ARCEL S ERVICE , INC .,
P ETITIONER
v.
P OSTAL R EGULATORY C OMMISSION,
R ESPONDENT
AMAZON.COM S ERVICES , LLC, ET AL.,
INTERVENORS
On Petition for Review of Orders
of the Postal Regulatory Commission
Kathleen M. Sullivan argued the cause for petitioner. On
the briefs were David M. Cooper and Steig D. Olson.
Michael Shih, Attorney, U.S. Department of Justice,
argued the cause for respondent. With him on the brief were
Brian M. Boynton, Principal Deputy Assistant Attorney
General, Michael S. Raab and Kevin J. Kennedy, Attorneys,
David A. Trissell, General Counsel, Postal Regulatory
Commission, Lauren A. D’Agostino, Deputy General Counsel,
and Reese T. Boone, Attorney.

-- 1 of 15 --

2
Eric P. Koetting, Morgan E. Rehrig, and Michael D.
Weaver, Attorneys, U.S. Postal Service, Michael F. Scanlon
and John Longstreth were on the brief for respondent-
intervenors.
Before: S RINIVASAN, Chief Judge, GARCIA, Circuit Judge,
and R OGERS , Senior Circuit Judge.
Opinion for the Court filed by Circuit Judge GARCIA.
GARCIA, Circuit Judge: The Postal Regulatory
Commission is an independent agency that oversees the United
States Postal Service. The Postal Service sells two kinds of
products: “market dominant” products such as First-Class mail,
where the Postal Service exercises monopoly power with
congressional blessing, and “competitive” products such as
package delivery, where the Postal Service competes with
private companies. Congress tasked the Commission with
ensuring that the Postal Service competes fairly in the
competitive products market. The Commission must ensure,
for example, that the Postal Service does not unfairly subsidize
its competitive products business with earnings from its
market-dominant products to the disadvantage of private
competitors.
Petitioner United Parcel Service, Inc. (“UPS”) is one such
competitor. Dissatisfied with how the Postal Service prices
competitive products, UPS petitioned the Commission to
initiate rulemaking proceedings. UPS argued that the Postal
Service underprices its competitive products by not holding
those products responsible for “peak-season” costs caused by a
spike in consumer demand for package deliveries every
December.

-- 2 of 15 --

3
The Commission denied UPS’s petition and its motion for
reconsideration. For the reasons explained below, we deny
UPS’s petition for review of those orders.
I
A
The Commission exercises regulatory authority under the
Postal Accountability and Enhancement Act (“Accountability
Act”), Pub. L. No. 109-435, 120 Stat. 3198 (2006). As we have
explained, the Accountability Act
ensure[s] (among other things) that the Postal
Service offers its competitive products on fair
terms. To that end, the Accountability Act
requires the Commission to promulgate
regulations that ensure that the Postal Service is
not using revenues from market-dominant
products subject to its monopoly power to
defray costs competitive products would
otherwise have to be priced to cover.
United Parcel Serv., Inc. v. Postal Regul. Comm’n (“UPS II”),
955 F.3d 1038, 1042 (D.C. Cir. 2020) (cleaned up).
Specifically, the Commission must promulgate regulations
to achieve three objectives. First, it must “prohibit the
subsidization of competitive products by market-dominant
products.” 39 U.S.C. § 3633(a)(1). Second, it must “ensure
that each competitive product covers its costs attributable,” id.
§ 3633(a)(2), defined as “the direct and indirect postal costs
attributable to such product through reliably identified causal
relationships,” id. § 3631(b). Third, the Commission must
“ensure that all competitive products collectively cover what
the Commission determines to be an appropriate share of the
institutional costs of the Postal Service.” Id. § 3633(a)(3).

-- 3 of 15 --

4
The Postal Service implements the first two mandates
using an “incremental-cost” approach that we broadly
sanctioned in United Parcel Service, Inc. v. Postal Regulatory
Commission (“UPS I”), 890 F.3d 1053 (D.C. Cir. 2018). The
incremental costs of competitive products are the costs “that
would disappear were the Postal Service to stop offering those
products for sale.” Id. at 1055. Under Section 3633(a)(1), the
Postal Service identifies the incremental cost of competitive
products as a whole. See id. at 1059. For Section 3633(a)(2),
the Postal Service examines the incremental cost of each
specific competitive product. See id. at 1066–69.
As for the third mandate, the Act does not define the term
“institutional costs,” but we have upheld the Commission’s
interpretation of that term to refer to residual costs that cannot
be attributed to any specific product via reliably identified
causal relationships. See id. at 1055–56. The Postal Service
thus treats as “institutional costs” all costs that are not “costs
attributable.” See id. It then requires competitive products to
cover “an appropriate share” of those institutional costs. See
United Parcel Serv., Inc. v. Postal Regul. Comm’n (“UPS III”),
96 F.4th 422, 424 (D.C. Cir. 2024). In sum, the Act effectively
“subjects each competitive product to a price floor, which must
be set high enough to cover both that product’s ‘costs
attributable’” and an appropriate share of the Postal Service’s
“institutional costs” under Section 3633(a)(3). UPS I, 890 F.3d
at 1055 (cleaned up).
In 2020, we remanded a Commission order adopting a
formula for the appropriate share of institutional costs under
Section 3633(a)(3), with instructions to better explain its
reasoning in certain respects. See UPS II, 955 F.3d at 1051–
52. On remand, the Commission revised its analysis while
readopting the same formula. UPS challenged that order in a
petition for review that was heard—and decided—by this panel
in a companion case to this one. See UPS III, 96 F.4th 422.

-- 4 of 15 --

5
We concluded that the Commission adequately addressed the
issues identified in UPS II and reasonably exercised its
statutory discretion in adopting the appropriate share formula.
Id. at 429. We therefore denied UPS’s petition for review.
B
In this case, UPS challenges the Commission’s
implementation of Sections 3633(a)(1) and (a)(2). In broad
terms, UPS believes the Commission is allowing the Postal
Service to underprice its competitive products by failing to
fully acknowledge that those products drive a yearly spike in
costs every December and, in turn, failing to raise those
products’ price floors accordingly. On May 29, 2020, UPS
petitioned the Commission to initiate rulemaking to rectify that
alleged problem. J.A. 6–49.
UPS focused on what it termed “peak-season costs.” J.A.
8. According to UPS, every holiday season the Postal Service
faces increased commercial demand to deliver packages, which
are largely competitive products. To meet that demand, the
Postal Service incurs “hundreds of millions of dollars” in
increased costs. J.A. 46. The Postal Service must, for example,
hire “tens of thousands of temporary workers,” open
“temporary delivery annexes,” pay “additional overtime
wages,” and send “carriers out on a host of additional runs to
deliver packages.” J.A. 9. In its petition, UPS’s core argument
was that competitive products “largely, if not exclusively,”
caused these costs. J.A. 21. As support, UPS presented
calculations from consultants purporting to show that, absent
the need to deliver competitive products, the Postal Service
would not incur these dramatically increased costs. Thus, UPS
urged, these peak-season costs are incremental costs of
competitive products and must be attributed to those products
under Sections 3633(a)(1) and 3633(a)(2). According to UPS,
however, the Postal Service’s models instead “systematically

-- 5 of 15 --

6
shift these costs into institutional costs, which are
predominantly covered with market-dominant revenues.” J.A.
46. UPS asked the Commission to require the Postal Service
to adopt a version of its consultants’ methodology and treat the
entire peak-season cost increase as incremental costs of
competitive products. J.A. 46–47; see J.A. 27–29.
The Postal Service acknowledged “that the existence of a
seasonal peak in volume can cause seasonal costs” and agreed
with UPS that “package volumes are an important part of that
peak.” J.A. 113. But it disagreed with the rest of UPS’s
analysis. The Postal Service noted that several other products,
including market-dominant products such as First-Class mail,
also have significantly increased volumes each December. See
J.A. 119–21. The Postal Service explained that to the extent
that the increased costs were driven by competitive products,
the Postal Service had long understood that fact and “has
performed, and will continue to perform, appropriate costing
exercises to ensure that package volumes bear the seasonal
peak costs that they cause.” J.A. 113. Its models therefore
already “accurately account for peak costs” caused by
competitive products. Id. According to the Postal Service,
UPS’s calculations attempting to show that even more of those
costs should be attributed to competitive products lacked
grounding in “both solid economic theory and actual
operational practice” and provided no basis to reject the Postal
Service’s current cost-attribution approach. Id.
On November 29, 2021, after holding a technical
conference and accepting written comments on UPS’s petition,
the Commission denied the petition. See J.A. 213–44. The
Commission concluded that “UPS has failed to demonstrate
that either the attribution of peak-season costs has become
significantly inaccurate or the calculation of [c]ompetitive
products’ incremental cost can be significantly improved by
applying the methodology UPS advocates.” J.A. 224. The

-- 6 of 15 --

7
Commission found that UPS’s calculations “contain[]
numerous errors and . . . do[] not produce any reliable estimates
of peak-season costs.” J.A. 225. And the Commission further
concluded that the existing cost-attribution framework already
accounts for those costs arising from the seasonal spike that are
properly attributable to competitive products. See J.A. 228–33.
On December 28, 2021, UPS filed a motion for
reconsideration. J.A. 245–50. On January 28, 2022, the
Commission denied that motion. J.A. 251–57.
UPS petitions for our review of both orders.
II
We will set aside the Commission’s orders if we find them
“arbitrary, capricious, an abuse of discretion, or otherwise not
in accordance with law.” 5 U.S.C. § 706(2)(A); 39 U.S.C.
§ 3663. Under that standard, we will reverse “only if the
agency’s decision is not supported by substantial evidence, or
the agency has made a clear error in judgment.” Safe
Extensions, Inc. v. FAA, 509 F.3d 593, 604 (D.C. Cir. 2007)
(quotation omitted).1
We are “reluctant to interfere with [the Commission’s]
reasoned judgments about technical questions within its area of
expertise. . . . In considering whether the orders suffer from
arbitrary and capricious decision-making, then, we ask only
whether the Commission’s exercise of its authority was
1 Although this case involves a denial of a petition for
rulemaking, the Commission does not rely on our decisions setting
out the particularly deferential standards that apply when we review
such denials. See, e.g., WildEarth Guardians v. EPA, 751 F.3d 649,
651 (D.C. Cir. 2014). We reject UPS’s petition without relying on
those standards.

-- 7 of 15 --

8
reasonable and reasonably explained.” UPS I, 890 F.3d at 1066
(cleaned up).
III
A
UPS’s primary support for its petition was a multistep
methodology for estimating peak-season costs and showing
that the Postal Service failed to properly attribute those costs to
competitive products. UPS asked the Commission to require
the Postal Service to deploy a version of that methodology
moving forward. See J.A. 46–47.
The Commission, however, explained in detail why UPS’s
methodology was flawed and could not support UPS’s claim.
First, the Commission explained that UPS erred by using in its
calculations the average annual per-unit cost of mail and
thereby assuming that the cost of producing each unit of mail
is constant throughout the year. See J.A. 226–28. In fact, the
Commission explained, such costs vary throughout the year
“because accrued costs in heavy volume months are higher
than in most other months.” J.A. 227. The Commission
observed that the Postal Service’s costing models capture that
seasonal variation, but UPS’s did not. See id.
The Commission also explained that UPS’s model
suffered from significant internal inconsistencies. For
example, the model emphasized increases in certain cost
categories but simply ignored a cost segment in its own
calculations that showed “an approximately $128 million cost
decrease for December.” Id. The Commission also noted that
although “UPS focuses on December volume increases, it [did]
not address the fact that volumes also change in other months.”
Id. The Commission credited the Postal Service’s expert, who
explained that if UPS’s methodology were applied to other
months, the cost increases UPS identified in December would

-- 8 of 15 --

9
be “offset” by January decreases, id., and that UPS’s model
would produce seemingly anomalous results in other months,
J.A. 227–28.
In this court, UPS offers no response to those criticisms of
the methodology that formed the foundation for its petition
before the agency. Indeed, by its reply brief, UPS explicitly
states that it does “not ask[] this Court to accept, or to require
the [Commission] to accept, the UPS methodology.” Reply
Brief 8. UPS, in other words, has abandoned a primary
building block of its petition to the agency: that the Postal
Service should be required to use UPS’s costing methodology.
UPS does continue to rely on certain figures and graphs
that it says support the underlying claim that all peak-season
costs are caused by competitive products. See Petitioner’s
Brief 7–11, 23–27. Many of those figures document the
undisputed fact that the Postal Service’s costs spike during the
holiday season each year. See J.A. 13–20. But in other figures,
UPS sought to demonstrate that competitive products “largely,
if not exclusively,” cause that spike. J.A. 21. For example, one
figure purports to show that market-dominant product volumes
decrease every December, while competitive product volumes
increase. See J.A. 23. Another figure overlays the Postal
Service’s city carrier costs over time with competitive product
volumes over time and shows that “competitive product
volumes are closely correlated with city carrier costs over
time” because “as competitive products volume peaks, so do
city carrier costs.” See J.A. 21.
The Commission addressed these figures and explained
why they did not support UPS’s claims. In particular, the
Commission explained that UPS had failed to defend its
graphs’ reliance on “indexed average daily volumes instead of
the actual volumes.” J.A. 230. In its petition, UPS said it
calculated these “indexed” measures by taking “monthly

-- 9 of 15 --

10
volume totals, divided by delivery days in the month, then
indexed to 100 in April 2019.” J.A. 23 n.22, 25 n.28. To
explain this choice, UPS conclusorily stated that “index
numbers or percentage changes are in fact very useful in this
context, where the substantially larger workload associated
with competitive products makes a piece-based comparison of
volume growth meaningless.” J.A. 108. The Postal Service’s
expert responded that indexing accentuates changes in “the
growth rates of products with small volumes”—such as certain
competitive products. J.A. 230 (quotation omitted). “This is
because small volumes can have large percentage changes.”
Id. (quotation omitted). As a result, the expert concluded that
UPS’s key graphs exaggerated the spike in competitive
volumes in December. Id.
The Commission agreed with the Postal Service’s expert
and concluded that the graphs did not depict volumes in a way
that permitted inferences about causality rather than mere
correlation. Id. Further, the Commission noted that UPS’s
graphs ran contrary to other evidence suggesting that market-
dominant products had a significant volume increase in
December and therefore were at least substantially responsible
for the increased costs. J.A. 229–31; see, e.g., J.A. 230 (“First-
Class Mail also has a December volume peak[,] and its share
of the Postal Service’s volume is much larger than the
Competitive products’ share.”).
On appeal, UPS provides a very limited response to the
Commission’s critique. UPS’s opening brief does not respond
to the indexing concern at all. In reply, UPS states that
indexing does not “accentuate[] changes in small-volume
products,” Reply Brief 5, because it “looks at the seasonal
change in volume for all market-dominant products together,”
id. at 6. We generally do not address arguments raised for the
first time in a reply brief. Rollins Env’t Servs., Inc. v. EPA, 937
F.2d 649, 652 n.2 (D.C. Cir. 1991). But even if that explanation

-- 10 of 15 --

11
were timely and correct, it is not clear how it would justify
UPS’s bottom-line choice to present indexed volumes instead
of actual volumes. Without a clearer explanation from UPS,
we have no reason to question the Commission’s expert
judgment that the graphs do not establish that competitive
products are solely responsible for peak-season cost increases.
That is particularly so given that UPS presented these graphs
as supplemental support for the point supposedly proven by its
more comprehensive methodology, which UPS no longer seeks
to defend.
B
As another way of raising substantially the same argument
that the Postal Service’s costing models do not attribute enough
of the seasonal cost spike to competitive products, UPS next
asserts that the Postal Service’s models problematically create
a category of “unexplained costs.” See J.A. 27–32.
“Unexplained costs” is UPS’s term for costs that it thinks are
caused by competitive products in December but are not
attributed to competitive products under the incremental-cost
approach. In its petition, UPS used its methodology to
calculate a yearly average of more than $500 million in
“unexplained costs.” See J.A. 30. As we have explained, the
Commission rejected that methodology, and UPS abandons it
on appeal. UPS nevertheless insists that the Commission itself
has conceded that there are nearly $250 million in such costs.
Petitioner’s Brief 28.
UPS misunderstands the Commission’s position. The
Commission’s order clarified that these costs “are not
unexplained” at all. J.A. 231. Instead, they are the costs that,
using the Commission’s longstanding “multi-step”
incremental-cost analysis, “cannot be specifically attributed to
products through reliably identified causal relationships” and

-- 11 of 15 --

12
are therefore treated as institutional costs using the approach
we approved in UPS I. Id. (quotation omitted).
UPS claims that the Commission “provides no rationale
for why the seasonal cost spike goes into institutional costs
rather than incremental costs.” Petitioner’s Brief 30. Not so.
Part of the seasonal cost spike is captured and attributed to
competitive products using the Postal Service’s existing
methodology. J.A. 229; see also J.A. 10, 29–31 (UPS
acknowledging the same in its petition to the Commission).
The remaining portion of the seasonal spike is treated as
“institutional costs” because those are the costs that the Postal
Service’s “multi-step approach” identifies as not capable of
being attributed to any specific products. J.A. 231. For
example, crediting the Postal Service’s expert, the Commission
explained that these costs include network costs and costs
associated with broad groupings of products—not just
competitive products. J.A. 232. These costs, based on the
Postal Service’s calculations, “would not disappear if any
groups of individual products were discontinued.” Id. UPS’s
only rejoinder is that the Postal Service’s calculations are
wrong, and that these costs would in fact “disappear if
competitive products were discontinued.” Petitioner’s Brief
30. But that assertion rests on the methodology and figures the
Commission found unreliable and reasonably declined to
credit.
UPS also raises a technical argument to suggest that the
Postal Service’s costing methodology is not capable of
accurately assessing which peak-season costs are attributable
to competitive products. UPS argues that because the Postal
Service relies on “annual totals” of product volumes in its
costing models, its costing models overlook that increased
package volumes in December cause unique costs “that would
not be the same if that increase were spread throughout the
year.” Id. The Commission explained, however, that the

-- 12 of 15 --

13
costing methodology uses not only annual totals but also
“distribution keys” that are calculated on a quarterly basis. J.A.
228 (quotation omitted). Distribution keys are a method of
identifying “the share of each” category of costs “that each
product is responsible for generating,” and we approved their
use in UPS I. 890 F.3d at 1056. In its opening brief, UPS does
not address that reliance on quarterly distribution keys. See
Petitioner’s Brief 30–31. UPS’s effort to do so on reply is both
conclusory and comes too late. See Reply Brief 10–11; Rollins
Env’t Servs., 937 F.2d at 652 n.2.
C
UPS next asserts that the Postal Service’s incremental-cost
methodology fails to assess what costs would disappear if the
Service stopped offering all competitive products and thus fails
to implement Section 3633(a)(1). UPS claims that the Postal
Service’s method of assessing incremental costs “assumes that
the Postal Service has fixed operations that do not significantly
change in response to the addition of competitive product
volumes.” Petitioner’s Brief 32. Because of this fixed-
operations assumption, UPS argued that the “Commission has
not yet evaluated the full set of costs the Postal Service could
eliminate through an efficient reorganization of its delivery
network and other aspects of its operations, if it ceased
delivering competitive products.” J.A. 10. UPS claims that the
“magnitude of this error . . . grows every year as package
delivery” encompasses a larger portion of the Postal Service’s
business. Petitioner’s Brief 33.
The Commission explained, however, that the Postal
Service’s incremental-costs approach is specifically designed
to estimate how postal costs would change if the Postal Service
stopped delivering competitive products. See J.A. 234–37.
The Postal Service measures incremental costs using an
activity-based costing methodology, which groups costs by

-- 13 of 15 --

14
Postal Service activities rather than products, and then
apportions activities’ costs among products. In UPS I, we
generally approved this methodology as a way of calculating
competitive products’ incremental costs. See 890 F.3d at 1069.
And as the Commission explained, contrary to UPS’s
argument, this approach does not just measure costs
attributable to individual products; it has been repeatedly
modified to “expand[] the category of attributable costs” to
ensure they are “calculated at the highest level of aggregation
for which they can be estimated reliably.” J.A. 236. UPS
claims that the Postal Service’s assessment of incremental costs
for competitive products collectively is only “marginally
greater” than the sum of costs for each individual competitive
product. Petitioner’s Brief 32. But if UPS is right that the
Postal Service’s approach assumes offering at least some
competitive products, there would be no difference at all. UPS
has presented no concrete basis for this court to second-guess
the agency’s judgment that the Postal Service is appropriately
calculating incremental costs.
At bottom, the Commission found that the Postal Service
is “well aware” of its obligation to account for peak-season
costs in setting prices and “perform[s] . . . appropriate costing
exercises to ensure that package volumes bear the seasonal
peak costs that they cause.” J.A. 229 (quotation omitted).
Although the Commission reasonably found that UPS had
identified no basis to upend the Postal Service’s existing
costing methodology, it nevertheless acknowledged that
certain of UPS’s technical concerns were “worthy of further
consideration.” J.A. 240. The Commission accordingly
initiated a new docket to “explore the opportunities to update”
a particular costing model critiqued by UPS, J.A. 241, asked
the Postal Service to “generate” new “datasets,” id., and
encouraged the Postal Service to expedite other improvements
in its allocation of peak-season costs, J.A. 239–42. The

-- 14 of 15 --

15
Commission’s position on the issues UPS presented is both
reasonable and reasonably explained.
IV
UPS also argues on appeal that the Commission erred by
failing to consider whether peak-season costs are institutional
costs “uniquely or disproportionately associated” with
competitive products. 39 U.S.C. § 3633(b). But UPS’s only
mention of this issue before the Commission was a footnote
stating the issue would be addressed “in greater detail” in the
separate proceeding on remand from UPS II. J.A. 12 n.5. The
issue is therefore not properly presented in this case.
Appalachian Power Co. v. EPA, 251 F.3d 1026, 1036 (D.C.
Cir. 2001) (“An objection must be made with sufficient
specificity reasonably to alert the agency.” (quoting Tex Tin
Corp. v. EPA, 935 F.2d 1321, 1323 (D.C. Cir. 1991))). In any
event, UPS raised those arguments in the companion case, and
they have been addressed. See UPS III, 96 F.4th 422.
V
For the foregoing reasons, we deny the petition for review.
So ordered.

-- 15 of 15 --