UnitedHealthcare Benefits of Texas, Inc. et al

Filing 28

MEMORANDUM OPINION AND ORDER granting in part and denying in part #12 Plaintiffs' Motion for Summary Judgment; denying #13 Opposed Motion to Supplement the Administrative Record; granting in part and denying in part #15 Defendants' Cross-Motion for Summary Judgment. Signed by District Judge Jeremy D. Kernodle on 11/22/2024. (esw)

IN THE UNITED STATES DISTRICT COURT
FOR THE EASTERN DISTRICT OF TEXAS
TYLER DIVISION
§
§
§
§
§
§
§
§
§
§
§
§
§

UNITEDHEALTHCARE BENEFITS
OF TEXAS, INC., et al.,
Plaintiffs,
v.
CENTERS FOR MEDICARE &
MEDICAID SERVICES, et al.,
Defendants.

Case No. 6:24-cv-357-JDK

MEMORANDUM OPINION AND ORDER
In this Administrative Procedure Act case, Plaintiffs challenge a decision by
the Centers for Medicare and Medicaid Services rating their health plans
unfavorably.
Before the Court are the parties’ cross motions for summary judgment. Docket
Nos. 12; 15. Also before the Court is Plaintiffs’ opposed motion to supplement the
administrative record. Docket No. 13.
The Court previously granted an expedited summary judgment briefing
schedule at the parties’ request.

Docket No. 11.

Given the substantial harm

Plaintiffs will suffer absent prompt judicial resolution of this dispute, Plaintiffs asked
the Court to rule by November 30, 2024. Docket No. 12 at 1.
For the reasons stated below, Plaintiffs’ motion for summary judgment is
GRANTED IN PART.

The Court ORDERS that Defendants shall recalculate

Plaintiffs’ 2025 star ratings without consideration of the disputed call and shall
1

immediately publish the recalculated star ratings.

See Elevance Health, Inc. v.

Becerra, 2024 WL 2880415, at *19 (D.D.C. June 7, 2024) (ordering similar relief).
Defendants’ motion for summary judgment is also GRANTED IN PART. All other
grounds for summary judgment are DENIED. Plaintiffs’ motion to supplement the
administrative record is DENIED.
I.
This is a challenge to agency action under the APA, 5 U.S.C. §§ 701, et seq.
Plaintiffs are a group of health insurance companies that provide Medicare
Advantage plans to Medicare beneficiaries. Defendant Centers for Medicare and
Medicaid Services (“CMS”) is a federal agency within the U.S. Department of Health
and Human Services. CMS annually evaluates Medicare Advantage plans, including
the ones at issue here, using a star ratings system. Docket No. 15 at 1; 42 C.F.R.
§ 422.162. CMS rates plans on dozens of metrics, assigning a score of one- to fivestars for each metric. Docket No. 15 at 1. CMS then assigns each plan with an overall
rating that is a weighted average of the plan’s scores on the various individual
measures. Id. CMS publishes the ratings in October for the upcoming year to provide
enrollees relevant information about each plan’s quality before enrolling. Id. at 5–6.
The star rating received by a health plan thus affects its number of enrollees,
revenue, and market share. Id. at 6–7; Docket No. 12 at 12–13.
Plaintiffs dispute their 2025 star ratings. Specifically, Plaintiffs challenge the
four-star rating received by their call center for its performance in processing calls
from enrollees with limited English language proficiency. Docket No. 12 at 11–12.
CMS rules require that “interpreters must be available for 80 percent of incoming
2

calls requiring an interpreter within 8 minutes of reaching the customer service
representative and be made available at no cost to the caller.”

42 C.F.R.

§ 422.111(h)(1)(iii)(B). Because Plaintiffs share a single call center, each Plaintiff’s
overall star rating is based in part on the performance of that call center. Docket
No. 12 at 10. Plaintiffs claim that the four-star rating on this single metric will cause
them substantial losses as current and future customers will enroll with other plans.
Id. at 2.
To measure a call center’s ability to process calls from enrollees with limited
English language proficiency, CMS conducts foreign-language test calls. Docket
No. 15 at 7–8. CMS then evaluates the test calls as “completed” or “unsuccessful.”
Id. at 8. It may also “invalidate” unsuccessful calls, which then do not count against
a call center’s rating. See, e.g., Docket No. 12-1 (Administrative Record, hereinafter
“AR”) at 66–67 (invalidating a different test call from Plaintiffs’ call center). To
provide plans with notice about its test call evaluation process, CMS publishes the
Medicare Parts C & D Call Center Monitoring Accuracy and Accessibility Study
Technical Notes (hereinafter, “CMS’s guidelines”).

Docket No. 15 at 7.

CMS’s

guidelines exist to “provide[] notice to [Medicare Advantage Organizations] about
how [CMS] evaluates performance” and “explain how the study will be conducted and
how CMS will use the results to calculate each plan contract’s raw score.” Id.
This case comes down to a single disputed foreign-language test call. CMS
evaluated the call as “unsuccessful” because, the agency contends, the test caller “was
unable to engage a customer service representative before the call center

3

disconnected the call.” Id. at 2; AR at 223. This is the only unsuccessful foreignlanguage test call counting against Plaintiffs’ call center. Docket No. 12 at 11.
Notably, a single unsuccessful test call prevents a call center from receiving a fivestar rating for this metric. Id. at 7. As a result, the call center received a four-star,
rather than five-star, rating for this metric, which reduced the overall star rating for
several of Plaintiffs’ plans by one-half star (e.g., 3.5 instead of 4, 4 instead of 4.5, or
4.5 instead of 5). Id. at 11–12.
The facts of the challenged call are as follows. A French-speaking test caller
dialed the correct number for the call center, used an automated system, and was
routed to a French-speaking customer service representative (“CSR”). Id. at 10;
Docket No. 15 at 13–14; AR at 66. A voice is heard beginning to say something before
being cut off immediately. AR at 196 (nine-second mark). Fifteen seconds later, the
test caller says, “Hello?” in English. Id. (twenty-four-second mark); id. at 204. Silence
follows. No dialogue occurs between the test caller and the CSR. The test caller does
not ask an introductory question. The test caller remains on the line until the CSR
terminates the call after approximately eight minutes. Id. at 198.
Plaintiffs contend that CMS’s assessment of this call as “unsuccessful” was
arbitrary and capricious because the agency violated its own guidelines, treated
Plaintiffs differently than other similarly situated health plans, failed to address
Plaintiffs’ objections, and improperly relied on a private-party contractor. Docket
No. 12. Plaintiffs thus conclude that CMS violated the APA and ask the Court to
“correct[] this wrongful agency action.” Id. at 2.

4

Plaintiffs filed this lawsuit on September 30, 2024, and moved for summary
judgment on October 15, asserting that “[t]ime is of the essence” because annual
enrollment began on October 15 and will end on December 7, 2024. Id. at 25. On
October 30, CMS cross-moved for summary judgment on the same grounds raised in
Plaintiffs’ motion. Docket No. 15. The Court held oral argument on both motions on
November 18, 2024.
II.
The Administrative Procedure Act directs courts to “hold unlawful and set
aside” agency action found to be “arbitrary, capricious, an abuse of discretion, or
otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A); Mexican Gulf Fishing
Co. v. U.S. Dep’t of Com., 60 F.4th 956, 971 (5th Cir. 2023). “Disputes arising under
the APA are commonly resolved on summary judgment, where district courts sit as
an appellate tribunal to decide legal questions on the basis of the administrative
record.” Nat’l Ass’n for Gun Rights, Inc. v. Garland, 2024 WL 3517504, at *14
(N.D. Tex. July 23, 2024) (citing Amin v. Mayorkas, 24 F.4th 383, 391 (5th Cir. 2022)).
“Arbitrary and capricious review focuses on whether an agency articulated a
rational connection between the facts found and the decision made.” Mexican Gulf
Fishing, 60 F.4th at 971 (quoting ExxonMobil Pipeline Co. v. U.S. Dep’t of Transp.,
867 F.3d 564, 571 (5th Cir. 2017)). The standard is narrow, and a court is not to
substitute its judgment for that of the agency. Motor Vehicle Mfrs. Ass’n of U.S., Inc.
v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983). But this review “is not
toothless.” Texas v. United States, 40 F.4th 205, 226 (5th Cir. 2022) (quoting Sw.
Elec. Power Co. v. EPA, 920 F.3d 999, 1013 (5th Cir. 2019)). In fact, “it has a serious
5

bite.” Id. (quoting Wages & White Lion Invs., LLC v. FDA, 16 F.4th 1130, 1136
(5th Cir. 2021)).
Courts must “consider whether the decision was based on a consideration of
the relevant factors and whether there has been a clear error of judgment.” Mexican
Gulf Fishing, 60 F.4th at 971 (quoting Texas v. EPA, 983 F.3d 826, 835 (5th Cir.
2020)). Courts “must ensure that the agency did not ‘entirely fail[] to consider an
important aspect of the problem’” and reject “an explanation for its decision that runs
counter to the evidence before the agency, or is so implausible that it could not be
ascribed to a difference in view or the product of agency expertise.” Univ. of Tex. M.D.
Anderson Cancer Ctr. v. U.S. Dep’t of Health & Human Servs., 985 F.3d 472, 475
(5th Cir. 2021) (quoting State Farm, 463 U.S. at 43).
III.
Plaintiffs contend that CMS’s assessment of the disputed call violated the APA
for several, independent reasons. The Court addresses each argument in turn.
A.
Plaintiffs first argue that CMS’s decision was arbitrary and capricious because
the agency violated its own established guidelines by evaluating the call as
“unsuccessful” and including it in the study. Docket No. 12 at 13. The Court agrees.
According to CMS’s guidelines, the agency divides test calls into four phases:
(1) Dial, (2) Connect, (3) Introductory Question, and (4) Accuracy Measure. AR at 5.
Only the second and third phases are relevant here. At phase two, CMS “determine[s]
if we can reach a live CSR at the plan who can assist us with our questions.” Id.; see
also id. at 6 (“The call is considered connected when the caller connects with the
6

CSR.”); id. at 9 (“A call is considered connected when the caller reaches the CSR
(phase 2).”). “If we can connect with a live CSR within the given time parameters,
the call moves into phase 3 (introductory question) to ensure we are speaking with a
representative in the correct department.” Id. at 6. An example of an introductory
question is, “Are you the right person to answer questions about [Plan name’s] health
benefits?” Id. If the answer is “yes,” the call moves to phase four. Id.
Thus, according to CMS’s guidelines, when a test caller is simulating an
enrollee with limited English language proficiency, the agency considers the call
“connected when the caller connects with the CSR.” Id. “The [limited English
proficiency measure] is considered completed when the [CSR or an interpreter]
provides an affirmative response to the introductory question . . . within eight
minutes.” Id. On the other hand, a call is “scored as unsuccessful” if the test caller
is “not able to connect to a live CSR at the plan during that scheduled call” or the
“CSR cannot assist us with our questions or cannot forward our call to someone who
can assist.” Id. at 5. The guidelines also specify various other reasons the agency
might score a call as unsuccessful, such as “phone line problems,” “technology
barrier,” “Call Center disconnected the call,” or “CSR tells [the test caller] to call
back.” See id. at 5, 8–9. The guidelines state that CMS “invalidates” a call for at
least two reasons—the call was placed outside business hours or an operator dialed
the wrong number.

Id. at 19, 29.

The agency acknowledges that it has also

invalidated calls for at least one additional reason. Docket No. 15 at 17 (invalidating
a call where unknown technical issues prevented the call from connecting).

7

In evaluating the call here, CMS acted contrary to its own guidelines. The
record evidence demonstrates that the call “connected,” the call lasted more than
eight minutes, and the test caller never asked the introductory question contemplated
at phase three of the call.

Thus, the call should not have been marked as

“unsuccessful” according to the guidelines. CMS’s briefing in this Court at times
suggests that the call never “connected” and thus the test caller did not need to ask
an introductory question. But CMS conceded at oral argument that the call did
“connect.” And CMS repeatedly stated below that the call “connected.” AR at 223
(conceding the test caller “connected to a CSR”); id. at 226 (same); id. at 231
(acknowledging the test caller “was able to connect with the plan” and that the caller
“was on hold and then connected with a CSR”). CMS cannot now argue otherwise.
See Louisiana v. U.S. Dep’t of Energy, 90 F.4th 461, 469 (5th Cir. 2024) (“An agency
may not advance arguments before [the Court] without first presenting them in the
administrative record.”).
Because the call “connected,” the test call necessarily moved to phase three.
Under CMS’s guidelines, a call will be evaluated as “complete” at phase three when
the CSR provides “an affirmative response to the introductory question . . . within
eight minutes.” AR at 6. Conversely, a call is marked “unsuccessful” in phase three
if the “CSR cannot assist us with our questions or cannot forward our call to someone
who can assist.” Id. at 5. The problem for CMS here is that the test caller never
asked the introductory question that’s required to evaluate a call at this phase.
Because the introductory question was not asked, Plaintiffs’ call center did not fail to

8

answer it. Thus, CMS acted inconsistently with its own guidelines by evaluating the
call as “unsuccessful.”
Agency action that is contrary to the agency’s guidelines may be arbitrary and
capricious. See, e.g., M.D. Anderson, 985 F.3d at 475–78. Indeed, CMS does not argue
otherwise here, maintaining instead that it did not act inconsistently with its
guidelines. M.D. Anderson is instructive. In that case, agency regulations required
HIPPA-covered entities to implement “a mechanism” for encryption. Id. at 477. M.D.
Anderson implemented a mechanism. But the agency penalized M.D. Anderson and
argued that it “should’ve done more—either to implement a different mechanism or
to better implement its chosen mechanism.”

Id.

The Fifth Circuit found that

penalizing M.D. Anderson—requiring the hospital to do more than “the regulation
[the agency] wrote”—was arbitrary and capricious.

Id. at 478.

The regulation

required only a mechanism, and M.D. Anderson had a mechanism, “even if it could’ve
or should’ve had a better one.” Id. Thus, “M.D. Anderson satisfied [the agency’s]
regulatory requirement, even if the Government now wishes it had written a different
one.” Id.
The same principle applies here.

CMS’s guidelines “provide[] notice to

[Medicare Advantage Organizations] about how [CMS] evaluates performance” of
test calls and “explain how the study will be conducted and how CMS will use the
results to calculate each plan contract’s raw score.” Docket No. 15 at 7. CMS cannot
later use a different metric system to evaluate a plan’s performance by, for example,
requiring a call center to engage with the test caller before hearing an introductory

9

question. See M.D. Anderson, 985 F.3d at 478.
CMS’s arguments to the contrary are unavailing.1 Although conceding the call
“connected to a CSR,” CMS claims that it evaluated the call as unsuccessful because
“[a]t no point during [Plaintiffs’] provided recording can [Plaintiffs’] plan be heard
trying to engage the French speaking caller.” AR at 223; Docket No. 15 at 14. But
CMS’s justification effectively shifts the burden to the CSR to engage with the test
caller before the introductory question is asked. And CMS fails to cite where its
guidelines impose such a burden. Rather, the guidelines contemplate only that the
caller will ask the introductory question and then measure whether the CSR answers
it appropriately. See, e.g., AR at 6 (stating that the call is “complete” when the CSR
provides “an affirmative response to the introductory question”). To be sure, the
guidelines also state that “[i]n order for the [limited English proficiency measure] to
be complete, there must be true communication, meaning the CSR must answer the
introductory question and be able to converse in the foreign language we are testing
with or without an interpreter’s assistance.”

Id.

But even that statement

contemplates that the introductory question must first be asked.

1

CMS stated in the administrative record that Plaintiffs’ call center disconnected the call, which
Plaintiffs do not dispute. AR 223. But CMS has not cited this as a justification for its decision here.
See Docket No. 15 at 22–23 (noting Plaintiffs “erroneously assume[] that CMS based its decision on
the fact that Plaintiffs’ call center disconnected the call after 8 minutes of silence” and that “CMS
nowhere stated that the reason for including the call in the study was the call center’s disconnection
of the call”). And even if CMS did rely on this fact to justify its decision, the Court would find it
inconsistent with CMS’s guidelines because the call remained connected for more than eight
minutes. See AR at 6 (stating that in phase three, the CSR must provide an “affirmative response
to the introductory question . . . within eight minutes” (emphasis added)); Docket No. 15 at 15
(noting “the 8-minute clock in which [the CSR must] contact an interpreter and complete the call”).

10

CMS claims that “[i]t would be unreasonable to expect a [caller] to start asking
questions without first confirming that they were speaking to a live person.” Docket
No. 15 at 15; see also Docket No. 21 at 4 (asserting that it is “absurd” to require test
callers to ask the introductory question “when doing so is patently futile—for
example, if the call center has already disconnected the call, or if the customer service
representative experiences technical issues preventing them from communicating or
places the interviewer on hold as soon as the call connected”). In fact, CMS contends
that the test caller here did not ask the introductory question because he believed he
was on hold. Docket No. 15 at 15. But CMS created the guidelines, identified the
phases, and specified the criteria for evaluating these calls. The responsibility for
any unreasonable or absurd outcomes resulting from the guidelines therefore lies
with the agency, not Plaintiffs. Cf. Texas v. EPA, 829 F.3d 405, 430 (5th Cir. 2016)
(“It is elementary that an agency must adhere to its own rules and regulations. Ad
hoc departures from those rules, even to achieve laudable aims, cannot be
sanctioned.” (quoting Reuters Ltd. v. FCC, 781 F.2d 946, 950 (D.C. Cir. 1986))).
In addition, CMS never made these points below. Instead, CMS noted only
that the CSR failed to engage the caller. AR at 223. The agency never said the
guidelines required the CSR to engage a caller who has not asked an introductory
question or that a caller may sit silently if he believes he’s on hold. To the extent
CMS now attempts to argue something different, it is foreclosed from doing so. See
Louisiana, 90 F.4th at 469 (holding that federal agencies are prohibited from
engaging in “convenient litigating positions” and “post hoc rationalizations” (quoting

11

Kisor v. Wilkie, 588 U.S. 558, 579 (2019))); Michigan v. EPA, 576 U.S. 743, 758 (2015)
(noting it is a “foundational principle of administrative law that a court may uphold
agency action only on the grounds that the agency invoked when it took the action”).
Accordingly, the Court finds that CMS’s determination of the disputed call as
“unsuccessful” was arbitrary and capricious in violation of the APA because CMS
failed to evaluate the call consistent with its own guidelines.
B.
Plaintiffs next argue that CMS’s decision on the disputed call was arbitrary
and capricious because the agency failed to “treat like cases alike.” Docket No. 12
at 15. Specifically, Plaintiffs point to an instance in 2023 in which CMS initially
evaluated a call from the Elevance call center as “unsuccessful,” but later invalidated
the call. AR at 216; Docket No. 12 at 15–16.
“It is a bedrock principle of administrative law that an agency must ‘treat like
cases alike.’” M.D. Anderson, 985 F.3d at 479 (quoting 32 CHARLES ALAN WRIGHT &
CHARLES H. KOCH, FEDERAL PRACTICE AND PROCEDURE § 8248, at 431 (2006)).
Unexplained inconsistency is a reason for holding agency action to be arbitrary and
capricious. Id. (citing Nat’l Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545
U.S. 967, 981 (2005)). And “an administrative agency cannot hide behind the factintensive nature of penalty adjudications to ignore irrational distinctions between
like cases.” Id. at 480.
After reviewing the Elevance case, the Court concludes that it is not similarly
situated with the case here. First, the cases were in different call phases under CMS’s
guidelines. As Plaintiffs point out, the call in the Elevance case did not “connect” and
12

thus did not progress to phase three. Docket No. 12 at 15. Here, however, the call
did “connect” and necessarily moved to the third phase. Second, the issue in each
case differed. In Elevance, CMS invalidated the call because a technological problem
occurred and because it was unclear whether Elevance’s or CMS’s technology
prevented the call from connecting. Id. at 15–16; AR at 216. Here, in contrast, the
call connected but the test caller never asked the introductory question as required
in phase three. The issue here is thus whether under CMS’s guidelines, a connected
CSR has the burden to engage with a test caller prior to the introductory question
being asked.
Plaintiffs argue that the two calls are similar because there was no evidence
here that their call center was to blame for the unsuccessful test call. Docket No. 12
at 16. But comparing the calls at this level of generality could sweep in all kinds of
calls that are not really similar at all—and ignores important distinctions. The Court
does not need to look “deeply into the weeds to see a difference,” as Plaintiffs argue.
Id. at 17. Rather, the distinctions identified by the Court provide sufficient bases for
CMS to have treated Plaintiffs and Elevance differently. See M.D. Anderson, 985
F.3d at 480 (“[A]n administrative agency cannot . . . ignore irrational distinctions
between like cases.” (emphasis added)).
Accordingly, the factual distinctions identified by the Court are sufficient to
find that CMS’s decision to treat Plaintiffs’ call and Elevance’s call differently was
not arbitrary and capricious.

13

C.
Plaintiffs further argue that CMS’s evaluation of the disputed call was
arbitrary and capricious because the agency “entirely failed to consider an important
aspect of the problem.” State Farm, 463 U.S. at 43; Docket No. 12 at 18. Specifically,
during the appeal process at the agency, Plaintiffs raised effectively the same
argument made here:

that the call connected, the test caller failed to ask the

introductory question as required by phase three, and therefore the call should not
have been evaluated as unsuccessful under CMS’s guidelines. See AR at 197–98. And
Plaintiffs contend that CMS failed to adequately consider this argument below.2
The Court agrees. Plaintiffs first objected to CMS’s evaluation of the disputed
call on July 19, 2024, asserting that there may have been a technical issue on the
caller’s side. Id. at 232–33. CMS denied the appeal, noting that the agency had no
indication of audio issues with the call. Id. at 230. The parties continued discussing
the call at a general level. Id. at 226–30. Then, on September 19, 2024, Plaintiffs
expressly made the argument they are making here—that the call should not have
been marked unsuccessful because the test caller failed to ask the introductory
question in phase three. Id. at 223–25. CMS did not address this new argument.
Instead, the agency’s response was effectively a copy-and-paste of its previous
response, with the only additional substantive reply being: “At no point during your
provided recording can your plan be heard trying to engage the French speaking

2

Plaintiffs identify two additional arguments that CMS allegedly failed to adequately address during
the agency’s appeal process. Because the Court agrees that CMS failed to adequately consider
Plaintiffs’ first argument, the Court need not address the remaining two.

14

caller.” Compare id. at 223 (CMS’s denial of Plaintiffs’ appeal on September 24,
2024), with id. at 226 (CMS’s denial of Plaintiffs’ appeal on September 16, 2024).
This is an inadequate response that “entirely failed to consider an important
aspect of the problem.” State Farm, 463 U.S. at 43. As already discussed, CMS’s
guidelines are published to provide transparency about how the agency evaluates test
calls. It follows that the guidelines are a very important aspect of any dispute over
whether a call was properly deemed “unsuccessful.” Yet CMS’s response failed to
address its own guidelines whatsoever. See Data Mktg. P’ship, LP v. U.S. Dep’t of
Labor, 45 F.4th 846, 855–58 (5th Cir. 2022) (finding agency action to be arbitrary and
capricious where the agency failed to “reasonably consider” and “reasonably explain”
previous inconsistent statements it made in an advisory opinion and a regulation).
Instead, CMS simply asserted—without citation to its guidelines—that it was the
CSR’s fault for failing to engage with the test caller. See Wages & White Lion, 16
F.4th at 1138 (“We do not defer to the agency’s conclusory or unsupported
suppositions.” (quoting United Techs. Corp. v. U.S. Dep’t of Def., 601 F.3d 557, 562
(D.C. Cir. 2010))); O’Reilly v. All State Fin. Co., 2023 WL 6635070, at *6 (5th Cir. Oct.
12, 2023) (an agency “entirely failed to consider an important aspect of the problem”
by “failing to explain how it” came to its conclusion).
Accordingly, the Court agrees with Plaintiffs that CMS’s decision was
arbitrary and capricious for failing to consider an important aspect of the problem—
that the evaluation of the call did not comply with CMS’s guidelines.

15

D.
Finally, Plaintiffs argue that CMS engaged in an unlawful delegation of its
decision-making authority to private contractors. Docket No. 12 at 22–23. CMS used
two private contractors to conduct test calls, evaluate call centers’ performance, and
make recommendations to CMS about whether disputed calls should be invalidated.
Id. at 23; AR at 33. CMS acknowledges that it delegated certain authority to private
entities but argues that it was a limited and permissible subdelegation. Docket
No. 15 at 24. Again, the Court agrees with Plaintiffs here.3
“A federal agency may not ‘abdicate its statutory duties’ by delegating them to
a private entity.” Texas v. Rettig, 987 F.3d 518, 531 (5th Cir. 2021) (quoting Sierra
Club v. Lynn, 502 F.2d 43, 59 (5th Cir. 1974)). “[T]he Constitution imposes upon
federal agencies—acting as agents of the people’s representatives in Congress—a
duty to wield delegated power unless Congress authorizes subdelegation or the
subdelegation involves no more than ministerial tasks.” Consumers’ Rsch. v. FCC,
109 F.4th 743, 775 (5th Cir. 2024) (en banc). While “Congress has broad discretion
to empower executive agencies to ‘execute’ the law,” when it comes to private entities,
“there is not even a fig leaf of constitutional justification.” Id. at 768 (citing City of
Arlington v. FCC, 569 U.S. 290, 304 n.4 (2013); then quoting Dep’t of Transp. v. Ass’n
of Am. Railroads (“Amtrak II”), 575 U.S. 43, 62 (2015) (Alito, J., concurring)). “Private
entities are not vested with legislative Powers.

3

Nor are they vested with the

Plaintiffs additionally argue that this delegation was unlawful because the private contractors had
a conflict of interest in evaluating calls. Docket No. 12 at 23–24. Because the Court finds CMS’s
delegation to be unlawful, the Court need not address any potential conflict-of-interest issue.

16

executive Power, which belongs to the President.” Id. (quoting Amtrak II, 575 U.S.
at 62 (Alito, J., concurring)). Thus, “Congress may formalize a limited role for private
parties in executing its laws, but agencies may not.” Id. at 775 (cleaned up).
“Ministerial tasks” are those that require no discretion or judgment. See id.
at 773 (citing Gaines v. Thompson, 74 U.S. (7 Wall.) 347, 353 (1868) (“A ministerial
duty . . . is one in respect to which nothing is left to discretion.”)); see also Book People,
Inc. v. Wong, 91 F.4th 318, 337–38 (5th Cir. 2024) (“An act is ministerial where the
law prescribes and defines the duties to be performed with such precision and
certainty as to leave nothing to the exercise of discretion or judgment.” (cleaned up));
Ministerial, BLACK’S LAW DICTIONARY (7th ed. 1999) (“Of or relating to an act that
involves obedience to instructions or laws instead of discretion, judgment, or skill.”).
This includes subdelegating a task such as collecting fees. See Oklahoma v. United
States, 62 F.4th 221, 229 (6th Cir. 2023). It may also include subdelegating a “trivial,
fact-gathering role.” See Consumers’ Rsch., 109 F.4th at 773. The term, however,
does not include subdelegating tasks such as deciding how much money should be set
aside to execute an agency’s policies. See id. at 773, 777.
This limitation on delegating government power to a private entity exists to
protect the “vital constitutional principle” that “liberty requires accountability.” See
id. at 775 (quoting Amtrak II, 575 U.S. at 57 (Alito, J., concurring)). To subdelegate
more than ministerial tasks to a private contractor would insulate the execution of
the law from democratic accountability. See id. Private entities—unlike executive
branch officials—are in no way accountable to the people. See id.

17

CMS’s subdelegation to private contractors here was unlawful because it
lacked congressional authorization and involved more than mere “ministerial tasks.”
First, CMS failed to cite any statutory provision authorizing the agency
to subdelegate the power at issue here.

And its counsel acknowledged at oral

argument that no such authorization exists. See Hearing Tr. 11/18/2024 at 40:15–
19. Second, CMS’s subdelegation included the power to evaluate test calls and make
recommendations to the agency. These are not merely ministerial because they
involve making choices—and the power to make a choice is a discretionary
one.
The record illustrates the discretion and judgment exercised by the
private contractors here. In response to Plaintiffs’ first appeal of the disputed call,
the private contractors sent an evaluation of the call and a recommendation to CMS:
The interviewer was able to connect with the plan and inform the plan
they needed a French interpreter. The interviewer was on hold and then
connected with a CSR but only heard the sound of a voice for a second
before cutting off. It appeared to the interviewer that they were still on
hold and tried to confirm by saying “hello” but there was no indication
from the CSR that they were connected and continued to hold until the
CSR disconnected. There were no indications of interviewer error so
they did not do a case do over. Recommend keeping outcome as is.
AR at 66 (emphasis added). CMS accepted the recommendation, repeating the same
facts provided by the private contractors as justification for its denial of Plaintiffs’
appeal.

Id. at 75.

Plaintiffs appealed to CMS again once they obtained a call

recording. CMS told the private contractors, “[Plaintiffs are] back again, let me know
if the recordings change anything.” Id. at 154. The private contractors sent another
evaluation and recommendation to CMS:

18

The plan’s recording confirms the interviewer’s experience, that they
connect to a CSR and heard someone say something and then cut out.
Recommend keeping outcome as is since the plan’s recording supports
our interviewer’s experience.
Id. at 153 (emphasis added). CMS again accepted the contractors’ recommendation,
this time repeating the contractors’ language near-verbatim when denying Plaintiffs’
appeal. Id. at 168. Both instances reveal that the private contractors not only
gathered facts for CMS, but also evaluated those facts and exercised discretion and
judgment to recommend that the call not be invalidated.

Thus, the private

contractors’ actions were more than ministerial. See Consumers’ Rsch., 109 F.4th
at 773.
CMS’s arguments to the contrary are unpersuasive.

CMS claims that it

retained final decision-making authority in evaluating the test call. But that fact
alone does not render a private delegation to be lawful. See id. at 770 (noting that
even where an agency retains final decision-making authority, it is still an
unconstitutional delegation if the agency “reflexively rubber stamp[s]” work prepared
by a private entity). Nor does CMS’s final decision-making authority reduce the
private contractors’ role to a mere ministerial one. In fact, nowhere in CMS’s briefing
does it argue that it delegated only “ministerial tasks.” Rather, CMS argues that the
private contractors’ role was limited to permissible “fact gathering” and “advice
giving,” citing a D.C. Circuit case that upheld the subdelegation of such tasks. Docket
No. 15 at 23–24 (citing Int’l Dark-Sky Ass’n, Inc. v. FCC, 106 F.4th 1206, 1216
(D.C. Cir. 2024)). But the D.C. Circuit case is inconsistent with Consumers’ Research
v. FCC, which squarely held that subdelegating any function exceeding a “ministerial

19

task” in the absence of congressional authority is improper. 109 F.4th at 775. And
here, the agency subdelegated the evaluation of and recommendation regarding test
calls, which CMS apparently agrees was not ministerial.
CMS also contends that “the record shows pushback and request for
clarification from CMS,” suggesting that the agency did not simply rubber-stamp the
contractors’ recommendations. Docket No. 15 at 24. But in evaluating the disputed
call at issue here, CMS did not “push back” or “request clarification,” instead simply
adopting nearly verbatim both the recommendation and rationale of the contractors.
See AR at 75, 168. To be sure, in a different call challenged by Plaintiffs, the private
contractors recommended invalidation, and the CMS official responded, “I am
confused as to why the second call would suggest invalidation when it appears that
the recording matches what was reported.” Id. at 153. But the private contractors’
reply further establishes that they exercised independent discretion and judgment,
stating that they “wanted to err on the side of caution.” Id. at 152. CMS then
responded, “OK I understand that” and invalidated the call. Id. Indeed, only an
entity acting with discretion would have the freedom to “err on the side of caution” as
the private contractors did here. Cf. United States v. Gaubert, 499 U.S. 315, 325
(1991) (noting a “discretionary act is one that involves choice or judgment”).
Accordingly, the Court finds that CMS engaged in an unlawful delegation by
subdelegating more than ministerial tasks to private contractors in the absence of
congressional authorization. See Consumers’ Rsch., 109 F.4th at 773–75.

20

IV.
Finally, the Court considers the proper remedy. The APA empowers courts to
“hold unlawful and set aside” agency action found to be “arbitrary, capricious, an
abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A).
The default remedy in the Fifth Circuit is vacatur of an improper agency action. Data
Mktg. P’ship, 45 F.4th at 859. But the Court retains discretion to choose another
remedy if appropriate to effectuate the judgment—injunctive, declarative, or
otherwise. See Cargill v. Garland, 57 F.4th 447, 472 (5th Cir. 2023) (en banc), aff’d,
602 U.S. 406 (2024). Ultimately, a “plaintiff’s remedy must be tailored to redress the
plaintiff’s particular injury.” Id. (quoting Gill v. Whitford, 585 U.S. 48, 73 (2018)).
Here, Plaintiffs seek (1) a declaratory judgment holding CMS’s decision to
include the disputed call was unlawful and (2) an injunction requiring CMS to
recalculate Plaintiffs’ 2025 star ratings without considering the disputed call and
immediately publish the recalculated rating.

Docket No. 1 at 17.

The parties

discussed the appropriate relief at the November 18, 2024 oral argument in this case.
Under the circumstances here, the appropriate relief differs for each of
Plaintiffs’ three successful arguments. (Because the Court finds in favor of Defendant
on Plaintiffs’ second argument, no relief is appropriate on that issue.) As to Plaintiffs’
first argument—that CMS’s evaluation of the disputed call was arbitrary and
capricious—remand with specific instructions to recalculate and republish the rating
without the disputed call is appropriate. CMS appears to agree. See Hearing Tr.
11/18/2024 at 44:9–45:5; see also Elevance Health, 2024 WL 2880415, at *19 (ordering
similar relief). On Plaintiffs’ third argument—CMS’s failure to consider an important
21

aspect of the problem—the parties agree that remand solely for proper consideration
of Plaintiffs’ argument is appropriate. See Hearing Tr. 11/18/2024 at 14:22–25; 45:6–
9.

For Plaintiffs’ fourth argument—improper delegation of decision-making

authority to private contractors—the appropriate remedy is similarly remanding
with instructions to reconsider the rating, this time ignoring the private contractors’
recommendation.4 Id. at 46:5–12.
Because the Court grants summary judgment in favor of Plaintiffs on their
first argument, and the proper relief on that argument exceeds the appropriate relief
on the other two successful arguments, remand with instructions to recalculate and
republish the rating is the appropriate remedy.
*

*

*

As explained above, the Court GRANTS in part and DENIES in part
Plaintiffs’ motion for summary judgment and GRANTS in part and DENIES in part
Defendants’ cross motion for summary judgment.
The Court therefore ORDERS that CMS’s decision to include the disputed call
(D0800225) in the 2024 Call Center Monitoring Performance Metrics for Accuracy
and Accessibility Study is DECLARED unlawful and Defendants shall recalculate
Plaintiffs’ 2025 star ratings without consideration of the disputed call and shall
immediately publish the recalculated star ratings in the Medicare Plan Finder.

4

While this argument calls into question CMS’s policy of relying on private contractors’
recommendations in the absence of congressional authorization, neither party argues that
mandating a change in that overall policy is appropriate or necessary relief here. Accordingly, the
Court will not address that policy more broadly.

22

Also before the Court is Plaintiffs’ opposed motion to supplement the
administrative record. Docket No. 13. Plaintiffs seek to add two documents with
“background information” that they assert the district court needs to determine
whether the agency considered all the relevant factors. Id. at 3–4 (citing Medina
Cnty. Env’t Action v. Surface Transp. Bd., 602 F.3d 687, 706 (5th Cir. 2010)). But the
Court has not relied on these additional documents in resolving the summary
judgment motions. See also OnPath Fed. Credit Union v. U.S. Dep’t of Treasury,
Cmty. Dev. Fin. Insts. Fund, 73 F.4th 291, 299 (5th Cir. 2023) (“Motions to
supplement the record are rarely granted.”).
Accordingly, the Court DENIES the motion to supplement (Docket No. 13).
See id. (“Supplementation of the administrative record is not allowed unless the
moving party demonstrates ‘unusual circumstances justifying a departure’ from the
general presumption that review is limited to the record compiled by the agency.”
(quoting Medina, 602 F.3d at 706)).
So ORDERED and SIGNED this 22nd day of November, 2024.

___________________________________
JEREMY D. KERNODLE
UNITED STATES DISTRICT JUDGE

23