United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued March 20, 2020 Decided July 17, 2020
No. 19-5212
ASSOCIATION FOR COMMUNITY AFFILIATED PLANS, ET AL., APPELLANTS
v.
UNITED STATES DEPARTMENT OF THE TREASURY, ET AL., APPELLEES
Appeal from the United States District Court for the District of Columbia (No. 1:18-cv-02133)
Charles A. Rothfeld argued the cause for appellants. With him on the briefs was Andrew J. Pincus.
Douglas N. Letter, General Counsel, U.S. House of Representatives, Todd B. Tatelman, Deputy General Counsel, Megan Barbero, Associate General Counsel, Adam A. Grogg, Assistant General Counsel, Elizabeth B. Wydra, Brianne J. Gorod, and Ashwin P. Phatak were on the brief for amicus curiae U.S. House of Representatives in support of appellants.
Chad I. Golder was on the brief for amici curiae American Medical Association, et al. in support of appellants. 2 Kelly Bagby and Dara S. Smith were on the brief for amici curiae AARP, et al. in support of appellants.
Joseph R. Palmore and James Sigel were on the brief for amici curiae National American Cancer Society, et al. in support of plaintiffs-appellants.
Daniel Winik, Attorney, U.S. Department of Justice, argued the cause for appellees. With him on the brief was Alisa B. Klein, Attorney.
Robert Alt and Ilya Shapiro were on the brief for amici curiae The Buckeye Institute, et al. in support of defendants- appellees.
Monica Derbes Gibson was on the brief for amicus curiae Louisiana Commissioner of Insurance James J. Donelon in support of appellees and in support of affirmance.
Lawrence G. Wasden, Attorney General, Office of the Attorney General for the State of Idaho, Brian Kane, Assistant Chief Deputy, Megan A. Larrondo, Deputy Attorney General, and Anthony F. Shelley were on the brief for amici curiae State of Idaho, et al. in support of appellees and in support of affirmance.
Before: ROGERS, GRIFFITH, and KATSAS, Circuit Judges.
Opinion for the Court filed by Circuit Judge GRIFFITH.
Dissenting opinion filed by Circuit Judge ROGERS.
GRIFFITH, Circuit Judge: Since 1996, federal law has exempted “short-term limited duration insurance” (STLDI) from most federal health insurance regulations. For nearly two 3 decades, the Departments of Treasury, Labor, and Health and Human Services (the “Departments”) defined STLDI as plans with an initial contract term of less than one year. When Congress enacted the Patient Protection and Affordable Care Act (ACA) in 2010, it retained the STLDI exemption and left untouched the Departments’ longstanding definition. As a result, the ACA allowed insurers to sell STLDI plans to healthy individuals at a discount without complying with certain of the statute’s pricing and coverage rules. In 2016, the Departments became concerned that STLDI plans were drawing healthy people out of the risk pool for ACA-compliant insurance, causing premiums to rise. So they capped the length of such plans at three months. But over the next two years, premiums for ACA-compliant plans continued to soar while enrollment dropped off. The Departments reversed course with the goal of increasing the availability of more affordable insurance. The Association for Community Affiliated Plans (ACAP), along with other plaintiffs, challenged this reversal. The district court granted the Departments summary judgment, and we affirm.
I
A
Congress first carved out an exception for STLDI in the Health Insurance Portability and Accountability Act of 1996 (HIPAA), Pub. L. No. 104-191, § 102(a), 110 Stat. 1936, 1973 (codified at 42 U.S.C. § 300gg-91(b)(5)). By defining “individual health insurance coverage” to “[ex]clude short- term limited duration insurance,” Congress exempted STLDI plans from many of HIPAA’s standards. Id. Congress delegated the task of defining STLDI to the Departments. See 42 U.S.C. § 300gg-92 (permitting the Departments to “promulgate such regulations as may be necessary or appropriate to carry out the provisions of this subchapter”). In 4 1997, the Departments defined STLDI as coverage that expires “within 12 months of the date the contract becomes effective,” subject to renewal with the insurer’s consent. Interim Rules for Health Insurance Portability for Group Health Plans, 62 Fed. Reg. 16,894, 16,958 (Apr. 8, 1997). Seven years later, the Departments reaffirmed that definition in a final rulemaking. See Final Regulations for Health Coverage Portability for Group Health Plans and Group Health Insurance Issuers Under HIPAA Titles I & IV, 69 Fed. Reg. 78,720, 78,748 (Dec. 30, 2004).
When Congress enacted the ACA in 2010 to “expand coverage in the individual health insurance market,” King v. Burwell, 135 S. Ct. 2480, 2485 (2015), it incorporated by cross- reference HIPAA’s definition of “individual health insurance coverage,” including its exclusion of STLDI, see Pub. L. No. 111-148, § 1551, 124 Stat. 119, 258 (2010). As a result, STLDI policies were not subject to many of the ACA’s key reforms, which applied only to “individual health insurance coverage.”
Those key reforms included a combination of carrots and sticks that encouraged consumers to purchase more comprehensive coverage and ensured that they had the financial means to do so. The ACA’s “guaranteed issue” and “community rating” provisions prohibited insurers from denying coverage or charging higher premiums based on an individual’s race, gender, or health status. See 42 U.S.C. §§ 300gg, 300gg-1(a). Recognizing that these provisions could cause premiums to skyrocket by drawing older and sicker Americans into the risk pool, Congress required everyone to purchase “minimum essential coverage,” or else pay a tax penalty. 26 U.S.C. § 5000A. Congress hoped that this “individual mandate” would induce young, healthy people to enter the market. However, Congress appreciated that comprehensive insurance might be too expensive for some, so 5 it exempted low-income individuals from the penalty, id. § 5000A(e)(1), (5), and provided tax-credit subsidies to those purchasing insurance through government-run “Exchanges,” id. § 36B. Finally, Congress required that all plans offered on the Exchanges provide “essential health benefits,” including emergency services, prenatal care, and prescription drug coverage. 42 U.S.C. §§ 18021(a)(1)(B), 18022(b)(1), 18031(d)(2)(B)(i).
More than 85% of those purchasing insurance on the Exchanges do so using federal tax credits. See King, 135 S. Ct. at 2493; Wu Decl. ¶ 6, J.A. 91. These credits effectively cap the amount of money a person can expect to pay toward her insurance. For example, a single person whose income is equal to the poverty line will receive a subsidy sufficient to allow her to purchase insurance for no more than 2% of her income. See 26 U.S.C. § 36B(b)(3)(A)(i). If insurance prices go up, subsidies do too. As a result, subsidized individuals are largely insulated from ballooning premiums.
Because the ACA directed the states to expand their Medicaid coverage, Congress assumed that those below the federal poverty line would be covered and did not make them eligible for federal subsidies. But after NFIB v. Sebelius, 567 U.S. 519 (2012), held that the ACA’s Medicaid expansion must be deemed optional to be constitutional, 2.3 million Americans were left unable to afford insurance in states that declined to expand their Medicaid programs, resulting in what’s now called the Medicaid coverage gap. See Kaiser Family Foundation, The Coverage Gap: Uninsured Poor Adults in States that Do Not Expand Medicaid (Jan. 14, 2020), https://www.kff.org/medicaid/issue-brief/the-coverage-gap- uninsured-poor-adults-in-states-that-do-not-expand-medicaid. 6 When the Exchanges opened in 2014 and premiums started to rise, consumers seeking cheaper insurance turned to STLDI policies. These policies can be purchased at a fraction of the cost because they are exempt from the ACA’s community-rating, guaranteed-issue, and essential-health- benefits requirements. But you get what you pay for. STLDI plans offer skimpier coverage and higher deductibles. They often expose consumers with undiagnosed preexisting conditions to the risk of cancellation. And because they don’t qualify as “minimum essential coverage,” they don’t satisfy the individual mandate, meaning that those insured under STLDI plans may be subject to the tax penalty.* Still, for those in the Medicaid coverage gap or otherwise unable to afford an ACA- compliant plan, a barebones STLDI policy is better than nothing.
In 2016, the Departments became concerned that these policies were drawing healthy Americans out of the risk pool for ACA-compliant insurance, causing premiums to rise. To discourage people from purchasing STLDI policies as their primary insurance, the Departments revised the definition of STLDI to cover only plans that expired “less than 3 months after the original effective date of the contract.” Excepted Benefits; Lifetime and Annual Limits; and Short-Term, Limited-Duration Insurance, 81 Fed. Reg. 75,316, 73,326 (Oct. 31, 2016). By capping STLDI plans at three months and prohibiting renewals, the Departments hoped to minimize the use of STLDI as a “primary form of health coverage,” reducing “adverse[] impact[s] [on] the risk pool for Affordable Care Act-compliant coverage.” Id. at 75,317-18. Although some commenters pointed out that the rule wouldn’t prevent insurers
* The penalty now has no bite, because Congress reduced it to $0, effective January 1, 2019. See Tax Cuts and Jobs Act of 2017, Pub. L. No. 115-97, § 11081, 131 Stat. 2054, 2092 (2017). 7 from stringing together four three-month-long STLDI policies to create year-round coverage, the Departments decided that a prohibition on such bundling would be too difficult to enforce. Id. at 75,318. Thus, even under the 2016 Rule, insurers could— and did—market STLDI policies in year-round blocks.
Despite the Departments’ efforts, premiums in the individual health insurance market continued to soar. Between 2016 and 2017, average premiums shot up 21%, while Exchange enrollment of unsubsidized adults fell by almost the same percentage (1.3 million in total). Short-Term, Limited- Duration Insurance, 83 Fed. Reg. 38,212, 38,214 (Aug. 3, 2018). Acknowledging the burdens that these rising premiums created, the Department of Health and Human Services sought comments on how to expand affordable coverage options. Id. at 38,213. Several commenters suggested revitalizing the STLDI market. Id.
In 2018, the Departments proposed returning to the original definition of STLDI. Short-Term, Limited-Duration Insurance, 83 Fed. Reg. 7,437, 7,446 (Feb. 21, 2018). Following a comment period, the Departments issued a final rule defining STLDI as coverage with an initial contract term of less than one year and a maximum duration of three years counting renewals. 83 Fed. Reg. at 38,243. The Departments also expanded disclosure requirements, directing insurers to include a disclaimer that STLDI policies may “exclu[de] . . . coverage of preexisting conditions,” may not provide certain “health benefits,” and may not trigger a special enrollment period if coverage expires mid-year. Id.
Two main reasons were given for the new rule: (1) increasing access to affordable health insurance, especially among the uninsured, and (2) increasing consumer choice. The Departments explained that although the 2016 Rule “was 8 intended to boost enrollment in individual health insurance coverage . . . , it did not succeed in that regard,” so “expansion of additional coverage options . . . [was] necessary.” Id. at 38,214. They reasoned that the new rule would “expand[] access to additional, more affordable coverage options for individuals, including those who might otherwise be uninsured, as well as to those who do not qualify for [premium tax credits],” such as those in the Medicaid coverage gap. Id. at 38,216. The Departments acknowledged that expanding the availability of STLDI “could have an impact on the risk pools for individual health insurance coverage[] and could therefore raise premiums.” Id. at 38,217. However, they predicted that this effect would be modest, as subsidized enrollees were shielded from the effect of rising premiums. Moreover, because subsidies were available only on the Exchanges and “the individual subsidized premium [was] so low,” they anticipated that most “healthy lower-income individuals [would] remain in [their ACA-compliant] plans.” Id. at 38,235-36.
The Departments estimated that approximately 100,000 uninsured people would enroll in STLDI plans in 2019 and approximately 500,000 people would swap their ACA- compliant plans for STLDI plans, producing a 1% increase in unsubsidized premiums. Id. at 38,236. By 2028, the Departments projected that 200,000 previously uninsured individuals would enroll in STLDI plans, and 1.3 million individuals would shift from ACA-compliant plans to STLDI plans. Id. This would lead to a 5% increase in unsubsidized premiums. Id. The Congressional Budget Office and the Urban Institute both projected that the share of new STLDI enrollees who were previously uninsured would be somewhat higher (35% and 40% respectively). See id. at 38,237-38. 9
B
ACAP challenged the STLDI Rule, alleging that it was contrary to law and arbitrary and capricious. The district court held that ACAP had competitor standing because its members—private insurers selling plans on government Exchanges—faced growing competition from the STLDI market. On the merits, the district court granted the Departments’ motion for summary judgment, holding that the STLDI Rule was a reasonable interpretation of HIPAA and the ACA and that the change from the 2016 Rule to the current STLDI Rule was not arbitrary and capricious. “We review the district court’s grant of summary judgment de novo,” applying the familiar standards of the Administrative Procedure Act. Alpharma, Inc. v. Leavitt, 460 F.3d 1, 6 (D.C. Cir. 2006).
II
ACAP argues that the STLDI Rule is contrary to law because it is inconsistent with HIPAA’s plain text and an unreasonable interpretation of that text in light of the ACA’s structure and purpose. We are not persuaded.
Recall that the phrase “short-term limited duration insurance” does not appear in the ACA. Instead, the ACA incorporates by cross-reference HIPAA’s definition of “individual health insurance coverage,” which in turn is defined to exclude “short-term limited duration insurance.” See Pub. L. No. 111-148, § 1551, 124 Stat. at 258; 42 U.S.C. § 300gg-91(b)(5). ACAP argues that the Departments’ definition of STLDI is inconsistent with the text. We evaluate that definition under Chevron USA, Inc. v. Natural Resources 10 Defense Council, Inc., 467 U.S. 837, 842-43 (1984). Because the phrase “short-term limited duration insurance” is ambiguous, we defer to the Departments’ interpretation so long as it is “based on a permissible construction of” HIPAA and the ACA. Id. at 843. It is.
ACAP argues that the Departments’ definition involves an unreasonable interpretation of “short-term” for two reasons.
First, ACAP argues that the ACA’s definition of “short coverage gaps” restricts the Departments’ discretion to define “short-term” as used in HIPAA and incorporated by cross- reference into the ACA. Noting that the ACA exempts from the individual mandate persons who experience “short coverage gaps” of “less than 3 months,” 26 U.S.C. § 5000A(e)(4)(A), ACAP maintains that “Congress presumptively intended [the ACA’s] definition of short—as meaning a period of less than 3 months—to apply to the interpretation of . . . [HIPAA’s] phrase short-term coverage.” ACAP Br. 54 (internal quotation marks omitted). In other words, whatever “short-term” originally meant under HIPAA, it must now mean three months.
We cannot agree that Congress intended to amend HIPAA, a statute written over a decade before the ACA, in such a roundabout way. “[W]e will not understand Congress to have amended [a prior] act by implication unless there is a positive repugnancy between the provisions of the preexisting and newly enacted statutes, as well as language manifesting Congress’s considered determination of the ostensible change.” U.S. Ass’n of Reptile Keepers, Inc. v. Zinke, 852 F.3d 1131, 1141 (D.C. Cir. 2017) (internal quotation marks omitted). Congress knows how to impose time limits—after all, 11 it defined “short coverage gaps” as “less than 3 months”—but it didn’t do so for STLDI plans.
Second, ACAP responds that even if the ACA doesn’t limit “short-term” insurance to three months, the Departments’ definition still contradicts the plain text of HIPAA. “Short- term” means “occurring over or involving a relatively short period of time.” Short-Term, WEBSTER’S THIRD NEW INTERNATIONAL DICTIONARY 2103 (1981). As ACAP sees it, an STLDI policy must be “meaningfully shorter than the standard annual insurance term,” and a 364-day policy is not “meaningfully shorter” than a 365-day one. ACAP Br. 51.
But there is nothing unreasonable about the Departments’ definition. Consider, for example, how federal tax law defines capital gains. A “short-term capital gain” is a gain derived from an investment held for less than one year. 26 U.S.C. § 1222(1). A “long-term capital gain” is a gain derived from an investment held for one year or more. Id. § 1222(3). A 364-day investment is not “meaningfully shorter” than a 365-day one, yet the gains from each investment fall into different categories. So too here, it’s perfectly reasonable to describe a 364-day policy as “short- term,” even if a 365-day policy would not be.
ACAP would impose an artificial limitation on the Departments’ discretion by requiring STLDI policies to be not just “shorter” than the standard term but “meaningfully” so. This limitation finds no support in the text and strikes us as unworkable. Can the Departments cap STLDI plans at nine months? Ten months? Eleven months? Without further guidance from Congress, we will not place amorphous restrictions on the Departments’ authority to define such an open-ended term. It suffices to say that the Departments have the discretion to define STLDI to include policies shorter than the standard policy term. 12
ACAP next argues that the Departments’ definition is not properly confined to “limited duration” plans. It would seem that a plan that cannot be renewed beyond three years is, quite literally, “limited” in “duration.” Nevertheless, in an effort to evade the phrase’s ordinary meaning, ACAP suggests that “limited duration” actually means “nonrenewable.” ACAP Br. 56. One of HIPAA’s central reforms was to guarantee renewability of most “individual health insurance coverage.” 42 U.S.C. § 300gg-42(a). STLDI plans are exempt from that guarantee because they are exempt from HIPAA’s definition of “individual health insurance coverage.” Id. § 300gg-91(b)(5). From this lack of a guarantee of renewability, ACAP infers a prohibition. But nothing in HIPAA prevents insurers from renewing expired STLDI policies. Indeed, from 1997 to 2016, renewals were allowed with the insurer’s consent.
ACAP responds that if “limited duration” does not mean “nonrenewable,” then it’s redundant of “short term.” Not so. Under the Departments’ definition, “short-term” refers to the initial contract term, while “limited duration” refers to the policy’s total length, including renewals. This reasonable reading gives independent meaning to each term.
In any event, the Departments didn’t pick the three-year limitation out of a hat. They matched the duration of STLDI policies to that of similar types of temporary insurance, such as COBRA. See 83 Fed. Reg. at 38,221 (noting that COBRA “requires certain group health plan sponsors to provide a temporary continuation coverage option for a minimum of 18, 29, or 36 months”); see also id. (explaining that the Federal Employees Health Benefits Program permits temporary continuation of coverage for up to three years). Congress 13 granted the Departments wide latitude to define STLDI, and while the Departments retain the flexibility to narrow their definition in the future, nothing in the text forecloses their current interpretation.
ACAP next argues that the STLDI Rule is “irreconcilable with the structure and policy of the ACA,” ACAP Br. 25, and will ravage the government Exchanges. We disagree.
ACAP’s core contention is that the STLDI Rule contravenes the spirit of the ACA. ACAP contends that “Congress’s plan was to create a single, ACA-compliant individual market.” ACAP Br. 42 (emphasis added). ACAP says that the STLDI Rule is unreasonable because it facilitates the development of a parallel, “shadow” market for plans that do not provide comprehensive coverage. ACAP Reply 3. But the exception for STLDI is baked into the statute itself. By its own terms, the ACA exempts STLDI plans from the provisions requiring insurers to provide certain benefits, see 42 U.S.C. §§ 18021(a)(1)(B), 18022(b)(1), 18031(d)(2)(B)(i), and to treat all purchasers as members of a single risk pool, see id. § 18032(c). Contrary to ACAP’s portrayal, the Departments did not fashion a new category of insurance out of whole cloth to evade the ACA’s restrictions; they simply crafted rules to clarify which policies fall within the exception Congress created.
And the Departments reasonably defined the contours of that exception. On the day that Congress enacted the ACA, HIPAA had excluded “short-term limited duration insurance” from the definition of “individual health insurance coverage” 14 for over a decade. And for all that time, the Departments had defined the term almost exactly as they do today. That is powerful evidence that the modern STLDI Rule is consistent with the ACA. After all, “[w]here Congress ‘adopts a new law incorporating sections of a prior law, Congress normally can be presumed to have had knowledge of the interpretation given to the incorporated law, at least insofar as it affects the new statute.’” Gordon v. U.S. Capitol Police, 778 F.3d 158, 165 (D.C. Cir. 2015) (quoting Lorillard v. Pons, 434 U.S. 575, 581 (1978)).
ACAP argues that there’s “no evidence that Congress was even aware of the Departments’ interpretation . . . when it enacted the ACA.” ACAP Br. 48. But if there were ever “reason to assume[] congressional familiarity with the administrative interpretation at issue,” Public Citizen, Inc. v. HHS, 332 F.3d 654, 669 (D.C. Cir. 2003), it is here, where “[d]espite the ACA’s sweeping reforms,” Congress “left intact and incorporated” the STLDI exception, Central United Life Insurance Co. v. Burwell, 827 F.3d 70, 72 (D.C. Cir. 2016).
ACAP objects that Congress would’ve spoken more clearly had it had intended to empower the Departments to permit the sale of a primary insurance product outside of the ACA-compliant marketplace. Riffing on Justice Scalia, ACAP accuses the Departments of trying to squeeze a “regulatory elephant” into a “statutory mousehole.” ACAP Br. 40 n.15 (citing Whitman v. Am. Trucking Ass’ns, Inc., 531 U.S. 457, 468 (2001)). But a legislative provision authorizing the Departments to define an entire category of insurance not subject to ordinary federal standards is no “mousehole.” And a regulation that has only modest effects on the government Exchanges is no “elephant.” 15 Nevertheless, ACAP insists that Congress likely did not expect insurance companies to market STLDI as primary insurance. Instead, ACAP says, Congress must have assumed that STLDI would be sold as temporary coverage that did not compete with ACA-compliant plans. The dissent goes further, suggesting that Congress “decided not to allow consumers to purchase plans offering less than minimum ‘essential health benefits’ as their primary form of coverage.” Dissent at 6 (emphasis added). The problem with this argument is that Congress expressly elected not to set up a Hobson’s choice between purchasing ACA-compliant insurance and forgoing coverage altogether. Cf. 42 U.S.C. § 18032(d)(3)(A) (“Nothing in this title shall be construed to restrict the choice of a qualified individual to enroll or not to enroll in a qualified health plan or to participate in an Exchange.” (footnote omitted)). To be sure, Congress hoped that most individuals would purchase ACA- compliant plans as their primary insurance, and it provided incentives to encourage them to do so. It increased the availability of such plans through the community-rating and guaranteed-issue provisions, provided subsidies to low-income adults, and imposed a penalty on those who failed to maintain “minimum essential coverage.” But it did not foreclose other options.
For example, in addition to STLDI, Congress left in place exceptions for “fixed indemnity” insurance, which pays out a set amount for predetermined events such as hospitalization. Id. § 300gg-91(c)(3)(B). As with STLDI, “many individuals found it cost-effective to forego minimum essential coverage (even despite the penalty) in favor of these fixed indemnity policies.” Central United Life Insurance, 827 F.3d at 72. As we have previously acknowledged, the ACA permits that choice, id. at 72-75, even as it nudges individuals toward choosing more comprehensive insurance. ACAP sees these alternative options as loopholes that the Departments should have closed, 16 but the Departments need not rewrite the law to fit ACAP’s preferences.
ACAP frames the ACA as relentlessly pursuing one goal: maximizing the number of individuals with comprehensive health insurance. But “no legislation pursues its purposes at all costs.” See Albany Eng’g Corp. v. FERC, 548 F.3d 1071, 1076 (D.C. Cir. 2008) (quoting Rodriguez v. United States, 480 U.S. 522, 525-26 (1987)). And like most statutes, the ACA pursues multiple competing missions, among them expanding coverage, decreasing premiums, and maximizing quality. The STLDI Rule reasonably balances those goals by expanding coverage to the uninsured, including those in the Medicaid coverage gap, at the expense of higher unsubsidized premiums for comprehensive insurance. Balancing the costs and benefits of expanding the length of STLDI policies is the Departments’ bailiwick. And whatever choice we might have made in their shoes, we cannot substitute our judgment for theirs.
ACAP next objects that the Departments cannot adopt an interpretation of STLDI that would lay waste to one of the ACA’s key reforms: the Exchanges. Although we agree that the Departments may not adopt a definition of STLDI that “would destabilize the individual insurance market . . . and likely create the very ‘death spirals’ that Congress designed the Act to avoid,” King, 135 S. Ct. at 2493, the Departments reasonably predicted that the Rule’s impacts on Exchange enrollment and premiums would be limited. And experience has borne out that prediction.
We defer to “reasonable agency prediction[s] about the future impact of [the agency’s] own regulatory policies.” Louisiana Energy & Power Auth. v. FERC, 141 F.3d 364, 370 17 (D.C. Cir. 1998). Here, the Departments reasonably concluded that the Rule’s potential effects on premiums would be relatively small. Compare 38 Fed. Reg. at 38,236-38 (predicting a 5% increase), with King, 135 S. Ct. at 2493 (predicting as much as a 47% increase). And the Departments reasonably predicted that the Rule’s potential effects on Exchange enrollment would be blunted by federal subsidies. The vast majority of individuals purchasing plans on the Exchanges receive subsidies and are thus “largely insulated from premium increases.” 83 Fed. Reg. at 38,213. Because subsidies “are available only for [ACA-compliant] plans offered on [the] Exchanges” and the out-of-pocket cost to subsidized individuals is “so low,” the Departments anticipated that most “lower-income individuals [would] remain in [their ACA-compliant] plans.” Id. at 38,235-36.
This prediction was shared by the Congressional Budget Office and several nongovernmental organizations, including opponents of the STLDI Rule. See id. at 38,325-28. As even a report commissioned by ACAP acknowledged, “the concept of a death spiral . . . is less applicable” to the Exchanges because the subsidies soak up premium increases. See Wakely Consulting Group, Effects of Short-Term Limited Duration Plans on the ACA-Compliant Individual Market 3, http://www.communityplans.net/wp-content/uploads/2018/04/ Wakely-Short-Term-Limited-Duration-Plans-Report.pdf; see also ACAP Comment at 5, J.A. 393 (citing this report).
Experience confirms these predictions were reasonable. Following the promulgation of the STLDI Rule, premiums for benchmark Exchange plans actually fell by 1.5% in 2019. See Wu Decl. ¶ 18, J.A. 94-95. And in 2020, premiums for those same benchmark plans dropped another 4%. See Press Release, Centers for Medicare & Medicaid Services (Oct. 22, 2019), https://www.cms.gov/newsroom/press-releases/premiums- 18 healthcaregov-plans-are-down-4-percent-remain-unaffordable -non-subsidized-consumers. Similarly, participation in the Exchanges was not obviously correlated with the new Rule. Indeed, enrollment went up in some states that permitted the sale of year-long STLDI policies and down in others that restricted its sale to shorter time periods. See Wu Decl. ¶¶ 21-22, J.A. 95-96. Because the Departments reasonably (and, as it turns out, correctly) predicted that the STLDI Rule would not result in a premium-driven mass exit from the Exchanges, we reject ACAP’s argument that the Rule is invalid based on speculation about its potential, unrealized effects.
III
Finally, ACAP argues that the STLDI Rule is arbitrary and capricious. Once again, we disagree.
First, ACAP says that the Departments failed to consider the impact of the STLDI Rule on the Exchanges and relied on factors that Congress had not intended them to consider. But the Departments expressly acknowledged that expanding the length of STLDI plans “could have an impact on the [Exchange] risk pools” and “could therefore raise premiums.” 83 Fed. Reg. at 38,217. They concluded, however, that such an impact would be relatively minor and that the need to expand affordable coverage options, especially for those who could not afford ACA-complaint insurance, “substantially outweigh[ed]” that impact. Id. We therefore reject ACAP’s assertion that the Departments failed to consider the Rule’s effects or acted outside of their discretion to balance the statute’s competing policy goals.
Next, ACAP argues that the Departments failed to adequately explain their departure from the 2016 Rule. “Agencies are free to change their existing policies as long as 19 they provide a reasoned explanation for the change.” Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125 (2016). The agency “need not demonstrate to a court’s satisfaction that the reasons for the new policy are better than the reasons for the old one.” FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 (2009). “[I]t suffices that the new policy is permissible under the statute, that there are good reasons for it, and that the agency believes it to be better . . . .” Id.
The Departments amply met this obligation. As the Departments explained, the 2016 Rule “did not succeed” in “boost[ing] enrollment in individual health insurance coverage.” 83 Fed. Reg. at 38,214. Instead, average monthly enrollment dropped by 10%, and average monthly premiums increased by 21% from 2016 to 2017. Id. Acknowledging that expanding the availability of STLDI plans would draw some individuals out of comprehensive plans into skimpier STLDI plans, see id. at 38,236, the Departments reasoned that the change would be beneficial because it would “reduce the fraction of the population that is uninsured,” id. at 38,228. Especially given the advent of the Medicaid coverage gap, it was reasonable for the Departments to strive to create cheaper coverage options for those who might otherwise go uninsured.
Last, ACAP argues that the STLDI Rule could produce coverage gaps for consumers whose STLDI policies expire mid-year. Adults who lose their ACA-compliant coverage qualify for a special enrollment period. 45 C.F.R. § 155.420. But a person who loses STLDI coverage typically must wait until the next open enrollment period to obtain ACA-compliant coverage on the Exchanges.
The Departments reasoned that the 2016 Rule exacerbated the coverage-gap problem because three-month STLDI plans were often not long enough to tide people over to the next open 20 enrollment period. See 83 Fed. Reg. at 38,217. For example, an individual who lost coverage in February and was not entitled to a special enrollment period would have to wait until November to enroll. Allowing STLDI policies to run for just under one year ensures that individuals can always purchase a policy to fit their need for temporary coverage.
ACAP responds that as long as individuals only use STLDI to bridge gaps between two ACA-compliant policies, there need never be a coverage-gap issue under the 2016 Rule. But the reality is that even under the 2016 Rule, many individuals were purchasing STLDI as their primary insurance. For those people, the 2016 Rule created more volatility because they could be “subject to re-underwriting” every three months, could see a “greatly increased” premium, could be denied a new policy “based on preexisting medical conditions,” and “would not get credit” toward any deductible on a new plan “for money spent toward the deductible during the previous 3 months.” 83 Fed. Reg. at 38,218. Finally, to ensure that persons considering purchasing an STLDI policy in lieu of an ACA- compliant one would be aware of the risk of coverage gaps, the Departments required insurers to include a disclaimer that the loss of STLDI coverage may not trigger a special enrollment period. Id. at 38,243. Under our deferential standard of review, that is sufficient to respond to commenters’ concerns.
IV
The dissent would invalidate the STLDI Rule as “inconsistent with the [ACA’s] statutory scheme.” Dissent at 6. But the dissent never says what that scheme requires. The dissent acknowledges that Congress expressly exempted STLDI policies from the statute’s requirements, leaving in place the Departments’ longstanding regulatory definition. Id. at 3. The dissent does not suggest that the ACA required the 21 Departments to initiate a rulemaking to change that definition. Nor does the dissent adopt ACAP’s more extreme textual argument that the ACA required the Departments to cap STLDI policies at three months. And while the dissent presumably would not have taken issue with the 2016 Rule, that rule also did not prevent individuals from purchasing STLDI plans as their primary coverage. 81 Fed. Reg. at 75,318.
Boiled down, the dissent’s objection to the STLDI Rule is a prudential one—STLDI plans aren’t good for consumers, so they should be restricted as much as possible. But so long as the Departments have acted within the bounds of their statutorily delegated authority, that policy judgment is theirs to make. When Congress delegates decisionmaking authority to an agency, it sacrifices control for flexibility. Delegation empowers a comparatively nimbler actor to respond to changed circumstances and unanticipated consequences. Sometimes (perhaps often), the agency will have to make policy tradeoffs in real-world settings that Congress did not imagine. That is exactly what happened here. In 2016, the Departments changed the definition of STLDI to respond to concerns about increasing premiums and decreasing enrollment. Two years later, confronted by still-increasing premiums and the Medicaid coverage gap, the Departments decided that expanding affordable coverage options was the way to go. If Congress disagrees with that decision, it can take back the reins. Or if a new Administration comes to power with a different vision of how the ACA’s competing policy goals should be balanced, it can revisit the Departments’ choice. But as judges, our role is narrow: to ensure only that the Departments reasonably exercised the policymaking authority granted to them and not to us. Because the Departments satisfied that constraint, we leave the STLDI Rule in place. 22 V
Having concluded that the STLDI Rule is neither contrary to law nor arbitrary and capricious, we affirm.
So ordered. ROGERS, Circuit Judge, dissenting: Today the court upholds a Rule defining “short-term limited duration insurance” (“STLDI”) to include plans that last for up to three years and function as their purchasers’ primary form of health insurance, in stark contrast to the gap-filling purpose for which such plans were created. Because STLDI plans are exempt from the requirements of the Patient Protection and Affordable Care Act (“ACA”), insurers offering them can cut costs by denying basic benefits, price discriminating based on age and health status, and refusing coverage to older individuals and those with preexisting conditions. As a result, they leave enrollees without benefits that Congress deemed essential and disproportionately draw young, healthy individuals out of the “single risk pool” that Congress deemed critical to the success of the ACA’s statutory scheme. 42 U.S.C. § 18032(c)(1). The Supreme Court has instructed courts to interpret the ACA’s provisions in a manner “consistent with . . . Congress’s plan.” King v. Burwell, 135 S. Ct. 2480, 2496 (2015). Because the Rule flies in the face of that plan by expanding a narrow statutory exemption beyond recognition to create an alternative market for primary health insurance that is exempt from the ACA’s comprehensive coverage and fair access requirements, I respectfully dissent.
I.
The ACA is a comprehensive statutory scheme that Congress enacted to address certain problems that had existed for decades in the health insurance market. See King, 135 S. Ct. at 2485. First, insurers competed for consumers by selling low-cost but skimpy plans that offered less than comprehensive coverage. For example, before the ACA, 75% of non-group health plans did not cover delivery and inpatient maternity care, 38% did not cover mental health services, and nearly 20% limited their coverage of prescription drugs. Amicus Br. of Am. Med. Ass’n et al. 13. Second, insurers further competed on price by denying coverage to individuals who were likely to 2 incur greater medical expenses, particularly those with preexisting medical conditions, or charging such individuals higher rates. Amicus Br. of Nat’l Am. Cancer Soc’y et al. 17. This practice disproportionately affected older people and women; the prevalence of preexisting conditions increases with age, id., and before the ACA, insurers routinely denied coverage on the basis of such preexisting conditions as pregnancy, a previous Cesarean section, or a history of surviving domestic abuse, Amicus Br. of U.S. House of Representatives 7. Additionally, in a practice known as age rating, insurers frequently charged higher premiums based solely on an individual’s age, sometimes by as much as eleven times the rates they charged younger people. Amicus Br. of AARP et al. 13.
The ACA addressed these problems through a particular “series of interlocking reforms” designed to promote fair access to comprehensive, affordable coverage. King, 135 S. Ct. at 2485. As to fair access, the ACA’s central provisions include “guaranteed issue” and “community rating” requirements, which mandate that insurers accept everyone who applies for coverage and limit price discrimination, respectively. 42 U.S.C. §§ 300gg(a)(1), 300gg-1(a). For example, insurers may not take preexisting conditions or gender into consideration when setting premiums, and age rating may not exceed a factor of three to one. Id. §§ 300gg(a)(1)(A)(iii), (a)(1)(B). As to comprehensive coverage, Congress required all individual plans to provide “essential health benefits,” id. § 300gg-6(a), including preventive care, prescription drugs, maternity and newborn care, mental health services, emergency services, and hospitalization, id. § 18022(b)(1). As to affordability, Congress offered tax credits to qualifying individuals. 26 U.S.C. § 36B. Further, Congress understood from failed healthcare reform efforts at the state level that guaranteed issue and community rating requirements 3 have the unintended consequence of encouraging adverse selection. King, 135 S. Ct. 2485–86. That is, when insurers are required to accept anyone who applies for coverage and to charge the same premiums regardless of health status, consumers have an incentive to wait to purchase insurance until they become ill, which drives premiums higher. Id. To “minimize this adverse selection and broaden the health insurance risk pool to include healthy individuals, which will lower health insurance premiums,” 42 U.S.C. § 18091(2)(I), Congress required most people to maintain “minimum essential coverage,” 26 U.S.C. § 5000A(a), and required insurers to consider all enrollees in the individual market “to be members of a single risk pool,” 42 U.S.C. § 18032(c)(1).
Congress provided for certain limited exemptions from the ACA’s requirements, including the exemption of “short-term limited duration insurance.” Id. § 300gg-91(b)(5). As the Departments of Treasury, Labor, and Health and Human Services (“Departments”) acknowledged in the preamble to the challenged Rule, STLDI was a well-understood insurance product that existed before the ACA and “was primarily designed to fill temporary gaps in coverage that may occur when an individual is transitioning from one plan or coverage to another plan or coverage.” Short-Term, Limited-Duration Insurance, 83 Fed. Reg. 38,212, 38,213 (Aug. 3, 2018) (“2018 Final Rule”); see also Excepted Benefits; Lifetime and Annual Limits; and Short-Term, Limited-Duration Insurance, 81 Fed. Reg. 75,316, 75,317 (Oct. 31, 2016) (“2016 Final Rule”). As the product’s name suggests, STLDI never was intended as a long-term form of primary insurance coverage. STLDI plans therefore did not compete with ACA-compliant plans for enrollees, because these short-term, stop-gap plans served a different purpose than long-term coverage. 4 Following the ACA’s enactment, some insurers began to offer STLDI plans “in situations other than those that the exception from the definition of individual health insurance coverage was initially intended to address,” namely, as purchasing individuals’ “primary form of health coverage.” 2016 Final Rule, 81 Fed. Reg. at 75,317. Because STLDI is not subject to the ACA’s requirements, those who enroll in STLDI plans may not receive “meaningful health coverage,” and because STLDI issuers can cut costs by discriminating based on health status, these plans may disproportionately attract healthier individuals, “thus adversely impacting the risk pool for Affordable Care Act-compliant coverage.” Id. at 75,317–18. To prevent the ACA from being undermined in this manner, the Departments defined STLDI as a health insurance plan lasting no longer than three months, taking into account any extensions. Id. at 75,326 (amending 45 C.F.R. § 144.103).
On January 20, 2017, the day President Trump took office, he issued an executive order announcing his administration’s intention “to seek the prompt repeal of the Patient Protection and Affordable Care Act.” Exec. Order No. 13,765, Minimizing the Economic Burden of the Patient Protection and Affordable Care Act Pending Repeal, 82 Fed. Reg. 8351, 8351 (Jan. 24, 2017). After failing to persuade Congress to repeal the statute, the President issued a new executive order, which observed that “STLDI is exempt from the onerous and expensive insurance mandates and regulations” of the ACA and therefore was “an appealing and affordable alternative to government-run exchanges.” Exec. Order No. 13,813, Promoting Healthcare Choice and Competition Across the United States, 82 Fed. Reg. 48,385, 48,385 (Oct. 17, 2017). The President therefore directed the Departments to consider proposing regulations “to expand the availability of STLDI” within sixty days. Id. at 48,386. 5
Following this directive, the Departments promulgated a Rule designed to facilitate the use of STLDI as “an affordable alternative” to ACA-compliant insurance. 2018 Final Rule, 83 Fed. Reg. at 38,229. The Departments acknowledged that STLDI was a product “that was primarily designed to fill temporary gaps in coverage that may occur when an individual is transitioning from one plan or coverage to another plan or coverage.” Id. at 38,213. Nevertheless, they determined that it also should be offered as “an additional choice” to “exist[] side-by-side with individual market coverage” that must comply with the ACA’s requirements. Id. at 38,218. With this objective in mind, the Departments redefined STLDI to include any plan with an initial contract term of less than twelve months and a total duration of no longer than thirty-six months including renewals or extensions. Id. at 38,243 (amending 45 C.F.R. § 144.103).
II.
In administering the ACA, the Departments “are bound, not only by the ultimate purposes Congress has selected, but by the means it has deemed appropriate, and prescribed, for the pursuit of those purposes.” MCI Telecomms. Corp. v. Am. Tel. & Tel. Co., 512 U.S. 218, 231 n.4 (1994). The ACA not only sought to expand access to affordable health insurance, but it did so in a particular manner: Congress deemed certain health benefits essential, prohibited discrimination against individuals with preexisting conditions, and ensured that healthier and less healthy individuals would share a single risk pool. The exemption from these requirements for STLDI plans addressed a well-understood insurance product that existed at the time to fill gaps in coverage, as the Departments have acknowledged. 2018 Final Rule, 83 Fed. Reg. at 38,213. Apparently unsatisfied with the statutory scheme that Congress devised, 6 the Departments fashioned this limited exemption into an alternative class of primary health insurance that need not comply with the ACA’s statutory requirements. I would hold that the Departments impermissibly defined “short-term limited duration insurance” in a manner inconsistent with the statutory scheme and would remand the Rule for further proceedings consistent with the ACA’s structure. See Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842– 43 (1984).
The Rule departs from the ACA’s structure in several significant ways, recreating the problems that existed in the American health insurance market before the statute’s enactment and that the statute was designed to solve. First, the Rule promotes the use of STLDI plans to circumvent the coverage requirements that Congress deemed essential. The Departments state that the Rule “empowers consumers to purchase the benefits they want and reduce overinsurance.” 2018 Final Rule, 83 Fed. Reg. at 38,228. But Congress expressly decided not to allow consumers to purchase plans offering less than minimum “essential health benefits” as their primary form of coverage. 42 U.S.C. § 300gg-6(a). In contravention of Congress’s judgment, 71% of recently studied STLDI plans do not cover outpatient prescription drugs, 43% do not cover mental health services, and none cover maternity care. Amicus Br. of Am. Med. Ass’n et al. 17.
Unsurprisingly, failing to provide minimum essential benefits allows STLDI issuers to charge approximately half the cost of an average, unsubsidized ACA-compliant plan available through the Exchange, 2018 Final Rule, 83 Fed. Reg. at 38,236, but at the expense of allowing consumers to gamble on plans that may not offer adequate protection against unforeseen medical expenses. For example, STLDI plans generally do not cover oncology drugs for patients diagnosed 7 with cancer, which cost approximately $10,000 per month on average. Amicus Br. of Am. Med. Ass’n et al. 16 n.27 (quoting Rachel Schwab, Coming up Short: The Problem with Counting Short-Term, Limited Duration Insurance as Coverage, CTR. ON HEALTH INS. REFORMS, GEORGETOWN UNIV. HEALTH POLICY INST., June 7, 2019). Yet approximately 40% of Americans will develop cancer at some point in their lifetimes, and needless to say, most cancer diagnoses are unexpected. Amicus Br. of Nat’l Am. Cancer Soc’y et al. 8, 25. Further, individuals who purchase STLDI may not realize that their plans contain such limitations; reports indicate that STLDI brokers often use aggressive and misleading marketing tactics, Amicus Br. of AARP et al. 18–19, and they can advertise more extensively than brokers of ACA-compliant plans, because they are not subject to the ACA’s requirement that insurers must spend at least 80% of premiums on clinical services and quality improvements, as opposed to other costs such as marketing, 42 U.S.C. §§ 300gg-18(a), (b)(1)(A)(ii); see Amicus Br. of Am. Med. Ass’n et al. 27.
Second, because STLDI plans need not comply with the ACA’s guaranteed issue and community rating requirements, insurers can further cut costs by discriminating based on preexisting conditions, age, or any other factor. While this may seem to benefit those individuals who qualify for STLDI plans, cancellation may occur retroactively, resulting in abrupt and unexpected loss of coverage. Amicus Br. of AARP et al. 15; Amicus Br. of Am. Med. Ass’n et al. 22–23. For example, one Arizona woman who enrolled in STLDI was hospitalized with an abdominal infection a few weeks after receiving emergency surgery for diverticulitis. Amicus Br. of Am. Med. Ass’n et al. 22. Her insurer treated the diverticulitis as a preexisting condition and canceled her plan, leaving her with $97,000 in medical bills. Id. at 22–23. In this respect, as in terms of their less than comprehensive coverage, STLDI plans may “benefit 8 insurance companies more than the patients who purchase them.” Id. at 27 (quoting Shelby Livingston, Short-Term Health Plans Spend Little on Medical Care, MODERN HEALTHCARE, Aug. 6, 2019).
Third, not only does the use of STLDI as primary health insurance leave enrollees without congressionally mandated protections, but it also fractures the “single risk pool” that Congress deemed critical to the success of the ACA. 42 U.S.C. § 18032(c)(1). “The Departments acknowledge[d] that relatively young, relatively healthy individuals in the middle- class and upper middle-class” would be “more likely to purchase” STLDI, which “could lead to adverse selection and the worsening of the individual market risk pool.” 2018 Final Rule, 83 Fed. Reg. at 38,235. As a result, the Departments estimated that unsubsidized premiums for those who remained in the risk pool for ACA-compliant coverage available through the Exchanges—disproportionately, older or less healthy individuals—would increase by 1% in 2019 and 5% in 2028. Id. at 38,236. In other words, the Rule draws younger, healthier consumers out of the market for ACA-compliant insurance, with the predicted result of higher premiums for those who remain in the risk pool. It thereby directly undermines a central purpose of the ACA’s “major reforms,” namely to “minimize . . . adverse selection and broaden the health insurance risk pool to include healthy individuals, which will lower health insurance premiums.” King, 135 S. Ct. at 2493 (alteration in original) (quoting 42 U.S.C. § 18091(2)(I)). It is difficult to imagine a starker conflict between a statutory scheme and a rule that purports to administer it.
III.
None of the court’s attempts to defend the Rule as consistent with the ACA is persuasive. First, the court places 9 considerable weight on the similarity between the 2018 Final Rule and a prior rule defining “short-term limited duration insurance” that was in effect when the ACA was enacted, suggesting that this similarity is “powerful evidence” that the Departments’ interpretation is consistent with the statute. Op. 14. To the contrary, there was no reason for Congress to expect that consumers would begin purchasing STLDI plans as their primary form of health insurance, considering that when Congress enacted the ACA, STLDI was simply a product used to fill gaps in coverage, as the Departments have acknowledged. See 2018 Final Rule, 83 Fed. Reg. at 38,213.
Second, the court surmises that for individuals who otherwise would go uninsured, “a barebones STLDI policy is better than nothing.” Op. 6. Although the Departments justified the Rule in part as an effort “to reduce the number of uninsured individuals,” 2018 Final Rule, 83 Fed. Reg. at 38,218, their own data reflect that this was not the primary anticipated effect of the Rule. Rather, the vast majority of new enrollees in STLDI plans were expected to switch from existing coverage. The Departments estimated that by 2028, enrollment in STLDI plans would increase by 1.4 million, while “the total number of people with some type of coverage” would increase by only 0.2 million. Id. at 38,236. That is, only approximately one in seven individuals enrolling in STLDI by 2028 otherwise would be uninsured. The central issue, then, is not whether an STLDI plan is better than nothing, but whether such a policy is an appropriate substitute for a plan offering the comprehensive coverage and fair access that Congress deemed essential. Unless Congress amends the ACA’s central provisions or repeals the statute, that decision is not left to the Departments or to individual consumers.
Third, the court brushes aside the Departments’ own estimate that the Rule would increase premiums for ACA- 10 compliant coverage by 5% within a decade by stating that this predicted impact, confirmed by experience since the Rule took effect, is “relatively small.” Op. 17. By this logic, the Executive Branch may incrementally chip away at a statute by promulgating rules that undermine the statutory scheme, so long as the effect of each regulatory action is sufficiently modest. When an agency prioritizes its own policy objectives over those that Congress enacted, as occurred here, this court necessarily must conclude that the agency’s action was arbitrary and capricious. See Gresham v. Azar, 950 F.3d 93, 104 (D.C. Cir. 2020).
In sum, “[e]ven under under Chevron’s deferential framework, . . . reasonable statutory interpretation must account for both ‘the specific context in which . . . language is used’ and ‘the broader context of the statute as a whole.’” Util. Air Regulatory Grp. v. EPA, 573 U.S. 302, 321 (2014) (third alteration in original) (quoting Robinson v. Shell Oil Co., 519 U.S. 337, 341 (1997)). The Departments’ Rule fails to account for the specific context in which the term “short-term limited duration insurance” was used at the time of the ACA’s enactment, namely to refer to a well-understood insurance product used to fill gaps in coverage, not to serve as an individual’s primary form of health insurance. The Rule further fails to account for the general context of the ACA’s scheme by undermining the particular “series of interlocking reforms” included in the statute to ensure fair access to comprehensive, affordable medical coverage and recreating the same problems in the health insurance market that the ACA was designed to solve. King, 135 S. Ct. at 2485. Accordingly, I respectfully dissent.