Opinion
6:24-cv-357-JDK
11-22-2024
UNITEDHEALTHCARE BENEFITS OF TEXAS, INC., et al., Plaintiffs, v. CENTERS FOR MEDICARE & MEDICAID SERVICES, et al., Defendants.
MEMORANDUM OPINION AND ORDER
JEREMY D. KERNODLE UNITED STATES DISTRICT JUDGE
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 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 five-stars 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 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 disconnected the call.” Id. at 2; AR at 223. This is the only unsuccessful foreign-language 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 five-star 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.
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 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 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).
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 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 question. See M.D. Anderson, 985 F.3d at 478.
CMS's arguments to the contrary are unavailing. 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.
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”).
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 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 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.
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.
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.
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 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.
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.
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.
“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. Nor are they vested with the 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.
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:
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 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.
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 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. Id. at 46:5-12.
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.
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.
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