ACA Cost-Sharing Damages to Insurers to Be Reduced by Tax Credits

COMMUNITY HEALTH CHOICE, INC.,
Plaintiff-Appellee
v.
UNITED STATES,
Defendant-Appellant

MAINE COMMUNITY HEALTH OPTIONS,
Plaintiff-Appellee
v.
UNITED STATES,
Defendant-Appellant

United States Court of Appeals
for the Federal Circuit

Appeal from the United States Court of Federal Claims in No. 1:18-cv-00005-MMS, Chief Judge Margaret M. Sweeney.

Appeal from the United States Court of Federal Claims in No. 1:17-cv-02057-MMS, Chief Judge Margaret M. Sweeney.

Decided: August 14, 2020

WILLIAM LEWIS ROBERTS, Faegre Drinker Biddle & Reath LLP, Minneapolis, MN, argued for plaintiff-appellee in 19-1633. Also represented by JONATHAN WILLIAM DETTMANN, NICHOLAS JAMES NELSON.

DANIEL WILLIAM WOLFF, Crowell & Moring, LLP, Washington, DC, argued for plaintiff-appellee in 19-2102. Also represented by STEPHEN JOHN MCBRADY, SKYE MATHIESON, CHARLES BAEK, CLIFTON S. ELGARTEN.

ALISA BETH KLEIN, Appellate Staff, Civil Division, United States Department of Justice, Washington, DC, argued for defendant-appellant. Also represented by MARK B. STERN, ETHAN P. DAVIS.

STEPHEN A. SWEDLOW, Quinn Emanuel Urquhart & Sullivan, LLP, Chicago, IL, for amicus curiae Common Ground Healthcare Cooperative. Also represented by DAVID COOPER, New York, NY; J. D. HORTON, ADAM WOLFSON, Los Angeles, CA.

Before DYK, BRYSON, and TARANTO, Circuit Judges.

DYK, Circuit Judge.

Today in Sanford Health Plan v. United States (“Sanford”), No. 19-1290, we hold that the United States failed to comply with section 1402 of the Patient Protection and Affordable Care Act (“ACA”), Pub. L. No. 111-148, 124 Stat. 119, 220–24 (2010) (codified at 42 U.S.C. § 18071) — which requires the government to reimburse insurers for “cost-sharing reductions.” We hold that section 1402 “imposes an unambiguous obligation on the government to pay money and that the obligation is enforceable through a damages action in the Court of Federal Claims [('Claims Court')] under the Tucker Act.” Sanford, No. 19-1290, slip op. at 3.

In these cases, following our decision in Sanford, we affirm the Claims Court's decisions as to liability. As in Sanford, we conclude that the government is not entitled to a reduction in damages with respect to cost-sharing re-ductions not paid in 2017. As to 2018, we address an issue not presented in Sanford: the appropriate measure of damages. We hold that the Claims Court must reduce the insurers' damages by the amount of additional premium tax credit payments that each insurer received as a result of the government's termination of cost-sharing reduction payments. We reverse and remand for further proceedings with respect to damages.

BACKGROUND

I

In 2010, Congress enacted the ACA, which includes “a series of interlocking reforms designed to expand coverage in the individual health insurance market.” King v. Burwell, 135 S. Ct. 2480, 2485 (2015). “[T]he Act requires the creation of an '[e]xchange' in each State — basically, a marketplace that allows people to compare and purchase insurance plans.” Id. Insurance plans sold on the ACA exchanges must provide a minimum level of “essential health benefits” and are referred to as “qualified health plans.” See 42 U.S.C. § 18031. The ACA defines four levels of coverage: bronze, silver, gold, and platinum, which are based on the percentage of essential health benefits that the insurer pays for under each type of plan. Sanford, No. 19-1290, slip op. at 4. For example, under a silver-level plan, the health insurance provider pays for 70 percent of the actuarial value of the benefits, and either the insured or the government pays the remaining 30 percent. Id.

Under most health insurance plans, the insured individual must bear two types of costs. First, the insured must pay a monthly premium to maintain coverage. Second, the insured must pay an additional fee — called “cost-sharing” — when medical expenses are incurred. Deductibles, coinsurance, and co-payments are examples of such fees. See 42 U.S.C. § 18022(c)(3)(A)(i). The ACA includes two sections, 1401 and 1402, that reduce the premiums and cost-sharing for low-income insureds by government payments to the insurers. These sections “work together: the [premium reductions] help people obtain insurance, and the cost-sharing reductions help people get treatment once they have insurance.” See Cmty. Health Choice, Inc. v. United States, 141 Fed. Cl. 744, 750 (2019) (quoting California v. Trump, 267 F. Supp. 3d 1119, 1123 (N.D. Cal. 2017)). These sections apply to taxpayers with a household income of between 100 percent and 400 percent of the federal poverty line. See 42 U.S.C. § 18071(b)(2); 26 U.S.C. § 36B(c)(1)(A); Sanford, No. 19-1290, slip op. at 5, 7. The statute refers to them as “applicable taxpayer[s]” in the case of section 1401, 26 U.S.C. § 36B(c)(1)(A), and “eligible insured[s]” in the case of section 1402, 42 U.S.C. § 18071(b).

Premium reductions. Under section 1401, each “applicable taxpayer” enrolled in an ACA exchange plan at any level of coverage is entitled to a “premium assistance credit amount” (“premium tax credit”) to offset part of the monthly premiums of the enrollee entitled to the premium tax credit. 26 U.S.C. § 36B. The ACA specifies a formula for determining the amount of premium tax credits, which depends on the applicable taxpayer's household income, but not on the monthly premium or the coverage level for the applicable taxpayer's plan. The premium tax credit cannot exceed the actual monthly premium for the individual's plan. See id. § 36B(b)(2). The government pays these premium tax credit amounts directly to insurers. See Sanford, No. 19-1290, slip op. at 8; 31 U.S.C. § 1324. Thus, the amount of the premiums charged by the insurers to the insured is effectively reduced.

Premium review. The ACA includes various measures for regulating insurance premiums. Section 1003 of the ACA establishes a “premium review process” that requires insurers to report their premium rate increases to the Secretary of Health and Human Services (“the Secretary”) and state regulators. See 42 U.S.C. § 300gg-94 (codifying ACA section 1003). State authorities can review the proposed rates. However, “[t]he rate review process does not establish federal authority to deny implementation of a proposed rate increase; it is a sunshine provision designed to publicly expose rate increases determined to be unreasonable.” See Bernadette Fernandez, Vanessa C. Forsberg & Ryan J. Rosso, Cong. Rsch. Serv., R45146, Federal Requirements on Private Health Insurance Plans 9 (2018). If a state regulator finds that an insurer's premium rate increases are “excessive or unjustified,” it is required to recommend that the Secretary “exclude[ ] [the insurer] from participation in the [state] [e]xchange.” 42 U.S.C. § 300gg-94(b)(1)(B).

Following the enactment of the ACA, states have taken a varied approach to premium rate review programs. Some, but not all, states have reserved the express authority to approve or deny premium rate increases. See Mark Newsom & Bernadette Fernandez, Cong. Rsch. Serv., R41588, Private Health Insurance Premiums and Rate Reviews 15 (2011) (“There is substantive variation in state regulation of health insurance rates.”). In states where there is no express approval requirement, insurers are still required to notify state regulators of premium increases above a certain threshold. See 42 U.S.C. § 300gg-94(a)(2); Fernandez et al., Federal Requirements on Private Health Insurance Plans at 9. The damages issue here does not turn on whether the states have required express approval of premium increases.

Cost-sharing reductions. Section 1402 of the ACA requires insurers to reduce the insured's “cost-sharing” payments and requires the Secretary to “make periodic and timely payments to the [insurer] equal to the value of the [cost-sharing] reductions.” 42 U.S.C. § 18071(c)(3)(A). The section applies to “eligible insured[s]” enrolled in silver-level plans offered on the exchanges. Id. § 18071(a), (b). Eligibility under section 1402 is tied to eligibility under section 1401, and the amount of cost-sharing reductions is directly tied to the household income of the eligible insured. See Id. § 18071(c), (f)(2); Sanford, No. 19-1290, slip op. at 7 n.2.

II

On October 12, 2017, the Secretary announced that the government would cease payment of cost-sharing reduction reimbursements. Sanford, No. 19-1290, slip op. at 11–12. The suspension of cost-sharing reduction reimbursements did not relieve the insurers of their statutory obligation to “offer plans with cost-sharing reductions to customers,” meaning that “the federal government's failure to meet its [cost-sharing reduction] payment obligations meant the insurance companies would be losing that money.” California, 267 F. Supp. 3d at 1134. The solution for the insurers was to increase premiums. These states “began working with the insurance companies to develop a plan for how to respond” “in a fashion that would avoid harm to consumers.” See id. The resulting plan involved the tax credit provision of section 1401 of the ACA.

Under section 1401, the government is required to subsidize an amount equal to the lesser of (1) the monthly premium for the applicable taxpayer's plan and (2) the difference between the monthly premium for the “applicable second lowest cost silver plan [(the 'benchmark plan')] with respect to the taxpayer” and a statutorily-defined percentage of the eligible taxpayer's monthly household income. 26 U.S.C. § 36B(b)(2) (codifying ACA section 1401(b)(2)). This percentage generally varies from 2% to 9.5% based on the eligible taxpayer's income relative to the federal poverty line. Id. § 36B(b)(3)(A). These payments are guaranteed since, unlike the cost-sharing reduction payments situation, there is a permanent appropriation for premium tax credits. See Sanford, No. 19-1290, slip op. at 8.

In effect, if the insurers increased the monthly premium for their benchmark silver plans, each insurer would receive an additional dollar-for-dollar increase in the amount of the premium tax credit for each applicable taxpayer under its silver plans, all while keeping the out-of-pocket premiums paid by each applicable taxpayer the same. See California, 267 F. Supp. 3d at 1134. But premium increases for silver-level plans would have an effect on other plans as well: the insurers would also receive additional tax credits for applicable taxpayers that were enrolled in bronze, gold, and platinum plans, whether or not the premiums for those plans were increased. Id. at 1135. Even if the insurers kept premiums the same for those other plans, they would receive additional tax credits. See id.

Because of the government's refusal to make cost-sharing reduction payments, most states agreed to allow insurers to raise premiums for silver-level health plans, but not for other plans. Cmty., 141 Fed. Cl. at 755; Me. Cmty. Health Options v. United States, 143 Fed. Cl. 381, 390 (2019). “As a result, in these states, for everyone between 100% and 400% of the federal poverty level who wishe[d] to purchase insurance on the exchanges, the available tax credits r[o]se substantially. Not just for people who purchase[d] the silver plans, but for people who purchase[d] other plans too.” Cmty., 141 Fed Cl. at 755 (quoting California, 267 F. Supp. 3d at 1135). And the insurers received “more money from the premium tax credit program, . . . mitigat[ing] the loss of the cost-sharing reduction payments.” Id. This practice was referred to as “silver loading.” Id.

This was, however, not a perfect solution. The premium tax credits could only offset premium increases for applicable taxpayers, i.e., insureds with a household income of between 100 percent and 400 percent of the federal poverty line. Thus, people having a higher household income would be paying significantly more in premiums for their silver-level plans since they did not receive premium tax credits. See California, 267 F. Supp. 3d at 1137. States took a varied approach to this issue. Although this does not appear to be the case in Texas or Maine, some states negotiated with insurers to offer off-exchange, silver-equivalent plans at the pre-silver-load premium rates. Id. Such off-exchange policies were not subject to the ACA's premium tax credits or cost-sharing reduction requirements. In other states, noneligible individuals could still switch to bronze, gold, or platinum plans (which did not have premium rate increases). Id.

III

Community Health Choice, Inc. (“Community”) and Maine Community Health Options (“Maine Community”) are health insurance providers that sell qualified health plans in Texas and Maine, respectively. See Cmty., 141 Fed. Cl. at 756; Me. Cmty., 143 Fed. Cl. at 391.¹ Both insurers offered cost-sharing reductions, as required under section 1402, to insured individuals,² and “as with every other insurer offering qualified health plans on the exchanges, stopped receiving these payments effective October 12, 2017.” Cmty., 141 Fed. Cl. at 756.

The two insurers involved here filed separate actions in the Claims Court, asserting that they were entitled to recover the unpaid cost-sharing reduction reimbursements for 2017 and 2018.³ The insurers asserted two theories of liability.⁴ First, the insurers alleged that “in failing make the cost-sharing reduction payments . . ., the government violated the statutory and regulatory mandate” of the ACA. Id. Second, the insurers alleged that the government's nonpayment constituted a “breach[ ] [of] an implied-in-fact contract.” Id.

On the insurers' motions for summary judgment, the Claims Court “conclude[d] that the government's failure to make cost-sharing reduction payments to [the insurers] violate[d] 42 U.S.C. § 18071 [(codifying ACA section 1402)] and constitute[d] a breach of an implied-in fact contract.” Id. at 770. The Claims Court concluded that each insurer was entitled to recover as damages the full amount of un-paid cost-sharing reduction reimbursements for both 2017 and 2018. The Claims Court was “unpersuaded by the [government]'s . . . contention that [the] insurers' ability to increase premiums for their silver-level qualified health plans to obtain greater premium tax credit payments, and thus offset any losses from the government's nonpayment of cost-sharing reduction reimbursements,” precluded or reduced the insurers' damages. Id. at 760.

The government appealed the Claims Court's decisions to this court, challenging the decisions as to both liability and damages. We have jurisdiction under 28 U.S.C. § 1295(a)(3).

On April 27, 2020, the Supreme Court issued its decision in Maine Community Health Options v. United States, 140 S. Ct. 1308 (2020), holding that section 1342 of the ACA (“[t]he Risk Corridors statute,” id. at 1329), which states that the government “shall pay” money to insurers offering “unprofitable plans” on the ACA exchanges, id. at 1316, created a “money-mandating obligation requiring the Federal Government to make payments under [section] 1342's formula,” id., at 1331, and that health insurance providers were entitled to “seek to collect [such] payment through a damages action in the [Claims Court],” id.

Today in Sanford, following the Supreme Court's decision in Maine Community, we hold that the government violated its obligation to make cost-sharing reduction payments under section 1402; “that the cost-sharing-reduction reimbursement provision imposes an unambiguous obligation on the government to pay money[;] and that the obligation is enforceable through a damages action in the [Claims Court] under the Tucker Act.” Sanford, No. 19-1290, slip op. at 3.

DISCUSSION

I

As noted, the government argues that section 1402 did not create a statutory obligation on the part of the government to pay cost-sharing reduction reimbursements and that its failure to make payments did not violate the statute. Our decision in Sanford resolves these issues in favor of the insurers here. Sanford, No. 19-1290, slip op. at 18. Because we affirm the Claims Court's decisions as to statutory liability, and the damages are the same under either theory of liability (as discussed below), we need not address the insurers' implied-in-fact contract theory.

II

The government nonetheless argues that, even if section 1402 created a statutory obligation, the insurers are not entitled to recover the full amount of the unpaid 2017 and 2018 cost-sharing reduction payments as damages. We find no merit to the government's argument that the insurers' 2017 damages should be reduced. Like the insurers in Sanford, Community and Maine Community did not raise their silver-level plan premiums in 2017 or receive increased tax credits for that year from the elimination of the cost-sharing reduction payments. Here, as in Sanford, we see no basis for a 2017 damages offset and affirm the Claims Court's award of 2017 damages. See Sanford, No. 19-1290, slip op. at 9, 12.

III

We turn to the 2018 cost-sharing payments. Neither the Supreme Court in Maine Community nor our decision in Sanford resolves this question. The government asserts that, beginning in 2018, both insurers raised the premiums for their silver-level plans “to account for the absence of direct reimbursement for cost-sharing reductions,” resulting in the receipt of increased premium tax credits. See Gov't Suppl. Damages Br. 12–14. It argues that the Claims Court erred when it failed to credit the government with “economic benefits” flowing from the increased tax credits when awarding damages. Id. at 15.

The government's theory is based on an analogy to con-tract law — specifically, the rule that “a non-breaching party is not entitled, through the award of damages, to achieve a position superior to the one it would reasonably have occupied had the breach not occurred.” LaSalle Tal-man Bank, F.S.B. v. United States, 317 F.3d 1363, 1371 (Fed. Cir. 2003). The government argues that silver loading was a direct result of the insurers' mitigation efforts, i.e., increasing premiums for silver-level plans, and that the insurers' recovery must be reduced by the additional payments the insurers received in the form of tax credits.

The Claims Court rejected these arguments in both cases on the same ground, holding that there was no “statutory provision permitting the government to use premium tax credit payments to offset its cost-sharing reduction payment obligation,” and that “[t]he increased amount of premium tax credit payments that insurers receive[d]” was not a “substitute[ ]” for its “cost-sharing reduction payments.” Cmty., 141 Fed. Cl. at 760. At oral argument, the parties agreed that the Claims Court's decisions rejected the government's mitigation theory on the merits. On appeal, the insurers similarly argue that the “[g]overment cannot invoke deductions not set forth in the statute itself.” Appellees' Suppl. Damages Br. 4–5.

A

In addressing the mitigation issue, it is important to distinguish between two different types of statutes providing for the grant of federal funds: those that impose an “affirmative obligation[ ]” or “condition[ ]” in exchange for federal funding, and those that do not. Pennhurst State Sch. & Hosp. v. Halderman, 451 U.S. 1, 17, 24 (1981). The Supreme Court has previously “characterized . . . [the former category of] Spending Clause legislation as 'much in the nature of a contract: in return for federal funds, the [recipients] agree to comply with federally imposed conditions.” Barnes v. Gorman, 536 U.S. 181, 186 (2002) (third alteration in original) (quoting Pennhurst, 451 U.S. at 17). On the other hand, the latter category of statutes does not involve contract-like obligations. See id. at 186; Pennhurst, 451 U.S. at 17; Sossamon v. Texas, 563 U.S. 277, 290 (2011).

Section 1402 belongs in the first category of Spending Clause legislation because it imposes contract-like obligations: in exchange for federal funds, the insurers must “'participat[e] in the healthcare exchanges' under the statutorily specified conditions.” Sanford, No. 19-1290, slip op. at 18 (quoting Me. Cmty., 140 S. Ct. at 1320); see also Nat'l Fed'n of Indep. Bus. v. Sebelius, 567 U.S. 519, 576 (2012) (analyzing the Medicaid provisions of the ACA as Spending Clause legislation). Specifically, in exchange for “the [insurer] . . . reduc[ing] the cost-sharing under [silver plans] in the manner specified in [section 1402(c)]” and “notify[ing] the Secretary of such reductions,” “the Secretary shall make periodic and timely payments to the issuer equal to the value of the reductions.” 42 U.S.C. §§ 18071(a)(2), (c)(3)(A); see also Cmty., 141 Fed. Cl. at 768 (“[T]he cost-sharing reduction program is less of an incentive program and more of a quid pro quo.”).

Under these contract-like Spending Clause statutes — where the statute itself does not provide a remedial frame-work — a contract-law “analogy applies . . . in determining the scope of damages remedies” in a suit by the government against the recipient of federal funds or by a third-party beneficiary standing in the government's shoes. Barnes, 536 U.S. at 186–87; see also Gebser v. Lago Vista Indep. Sch. Dist., 524 U.S. 274, 287 (1998) (“Title IX's contractual nature has implications for our construction of the scope of available remedies.”). In Barnes, the Court considered the government's damages remedies available under Title VI in a suit charging the federal funds recipient with failure to comply with its obligations. The Court explained that, when the statute “contains no express remedies, a recipient of federal funds is nevertheless subject to suit for compensatory damages . . . and injunction . . . forms of relief traditionally available in suits for breach of contract.” Barnes, 536 U.S. at 187 (citations omitted). Thus, “[w]hen a federal-funds recipient violates conditions of Spending Clause legislation, the wrong done is the failure to provide what the contractual obligation requires; and that wrong is 'made good' when the recipient compensates the Federal Government or a third-party beneficiary (as in this case) for the loss caused by that failure.” Id. at 189. On the other hand, forms of relief that are “generally not available for breach of contract,” such as punitive damages, are not available in suits under such Spending Clause legislation. Id. at 187–89.⁵

The same, we think, is true when an action for damages is brought against the government, under this type of Spending Clause legislation. The available remedy is de-fined by analogy to contract law where the statute does not provide its own remedies for government breach.⁶ We have indeed previously applied the contract-law analogy to limit damages in suits against the government under the Back Pay Act, 5 U.S.C. § 5596, another money-mandating statute.⁷ Our predecessor court held that in suits brought for improper discharge for federal employment, damages had to be reduced by the amount earned by the federal employee in the private sector under a mitigation theory.⁸ See Craft v. United States, 589 F.2d 1057, 1068 (Ct. Cl. 1978) (“Unless there is a regulation or a statute that provides otherwise, cases in this court routinely require the deduction of civilian earnings [from a back pay award] on an analogy to the principle of mitigation of damages.”); Laningham v. United States, 5 Cl. Ct. 146, 158 (Ct. Cl. 1984) (“This rule has been utilized as an analog to the private contract law principle of mitigation of damages.”); see also Motto v. United States, 360 F.2d 643, 645 (Ct. Cl. 1966); Borak v. United States, 78 F. Supp. 123, 125 (Ct. Cl. 1948).

Here the contract-law analogy applies because the statute “contains no express remedies” at all with respect to the government's obligation. Barnes, 536 U.S. at 187. While the ACA provides specific remedies for failure of the insurers or insured to comply with their obligations, see 42 U.S.C. §§ 300gg-22, 18081(h), “the [ACA] did not establish a [statutory] remedial scheme” for the government's non-compliance, Me. Cmty., 140 S. Ct. at 1330. Section 1402's silence as to remedies in this respect suggests that “forms of relief traditionally available in suits for breach of con-tract” are appropriate. Barnes, 536 U.S. at 187; see also Me. Cmty., 140 S. Ct. at 1330. We therefore look to government contract law to determine the scope of the insurers' damages remedy.

With respect to contract claims, the government is “to be held liable only within the same limits that any other defendant would be in any other court,” and “its rights and duties . . . are governed generally by the law applicable to contracts between private individuals.” United States v. Winstar Corp., 518 U.S. 839, 892, 895 (1996) (first quoting Horowitz v. United States, 267 U.S. 458, 461 (1925), and then quoting Lynch v. United States, 292 U.S. 571, 579 (1934)).

B

The traditional damages remedy under contract law is compensatory in nature. Restatement (Second) of Contracts § 347 (1981); Barnes v. Gorman, 536 U.S. at 187–90.

The fundamental principle that underlies the availability of contract damages is that of compensation. That is, the disappointed promisee is generally entitled to an award of money damages in an amount reasonably calculated to make him or her whole and neither more nor less; any greater sum operates to punish the breaching promisor and results in an unwarranted windfall to the promisee, while any lesser sum rewards the promisor for his or her wrongful act in breaching the contract and fails to provide the promisee with the benefit of the bargain he or she made.

24 Samuel Williston & Richard A. Lord, Williston on Con-tracts § 64:1 (4th ed. 2020); see also 11 Joseph M. Perillo & Helen Hadjiyannakis Bender, Corbin on Contracts § 55.3 (2020) (“[I]t is a basic tenet of contract law that the aggrieved party will not be placed in a better position than it would have occupied had the contract been fully per-formed.”).

Thus, courts have uniformly held — as a matter of both state and federal law — that a plaintiff suing for breach of contract is not entitled to a windfall, i.e., the non-breaching party “[i]s not entitled to be put in a better position by the recovery than if the [breaching party] had fully performed the contract.” Miller v. Robertson, 266 U.S. 243, 260 (1924); Bluebonnet Sav. Bank, F.S.B. v. United States, 339 F.3d 1341, 1345 (Fed. Cir. 2003) (“[T]he non-breaching party should not be placed in a better position through the award of damages than if there had been no breach.”); LaSalle, 317 F.3d at 1372 (“[T]he non-breaching party is not entitled, through the award of damages, to achieve a position superior to the one it would reasonably have occupied had the breach not occurred.” (citing 3 E. Allan Farnsworth, Farnsworth on Contracts 193 (2d ed. 1998)).⁹

This concern to limit contract damages to compensatory amounts is embodied, in part, in the doctrine of mitigation, which ensures that the non-breaching party will not benefit from a breach. The mitigation doctrine has two aspects. First, the non-breaching party is expected to take reasonable steps to mitigate his or her damages. Restatement (Second) of Contracts § 350 cmt. b (“Once a party has reason to know that performance by the other party will not be forthcoming, . . . he is expected to take such affirmative steps as are appropriate in the circumstances to avoid loss by making substitute arrangements or otherwise.”). Under common-law principles, the injured party may not recover damages for any “loss that the injured party could have avoided without undue risk, burden or humiliation.” Id. § 350(1); 3 Dan B. Dobbs, Law of Remedies § 12.6(1), at 127 (2d ed. 1993) (“[T]he damage recovery is reduced to the extent that the plaintiff could reasonably have avoided damages he claims and is otherwise entitled to.”); Roehm v. Horst, 178 U.S. 1, 11 (1900) (explaining that a plaintiff for breach of contract is entitled to “damages as would have arisen from the nonperformance of the contract at the appointed time, subject, however, to abatement in respect of any circumstances which may have afforded him the means of mitigating his loss” (quoting Frost v. Knight, L.R. 7 Exch. 111 (1872))). We need not determine whether this first aspect of the mitigation doctrine applies here — such that the insurers were obligated to increase premiums to secure increased premium credits.

Rather, here we look to a second aspect of the mitigation doctrine, which recognizes that there must be a reduction in damages equal to the amount of benefit that resulted from the mitigation efforts that the non-breaching party in fact undertook.¹⁰ Kansas Gas & Elec. Co. v. United States, 685 F.3d 1361, 1366 (Fed. Cir. 2012) (“[M]itigation efforts may result in direct savings that reduce the damages claim.”); Restatement (Second) of Contracts § 350 cmt. h (explaining that the calculation of mitigation should reflect “[a]ctual efforts to mitigate damages”);11 Corbin on Contracts § 57.11 (explaining that, in the case of a buyer breaching a contract for the sale of goods, the rule “measures the seller's damages by the contract price less the market price — the price actually obtained . . . by a new sale”).

For example, in Kansas Gas and Electric, the government breached a contract to dispose of the plaintiff utility companies' nuclear waste. Kansas Gas & Elec., 685 F.3d at 1364. Anticipating that the government would breach the contract, the utility companies began a “rerack project” to increase its storage capacity and mitigate the effects of a government breach. Id. We held that the plaintiffs were entitled to the costs of its rerack project taken in mitigation of the government's breach. Id. at 1365, 1371. We also held, however, that the plaintiffs' recovery was to be reduced by the “real-world benefit” realized by the plaintiff's rerack project. Id. at 1367–68. Namely, “[w]hile conducting the rerack, the [plaintiffs] both . . . used racks that could support higher enrichment fuel assemblies,” which “allowed [them] to achieve the same energy output from [their] reactor with fewer fuel assemblies,” thereby increasing the efficiency of their plant. Id. at 1364.

The plaintiffs argued that the efficiency benefits of the rerack project were “too remote and not directly related to the breach because the decision to 'pursue more highly enriched fresh nuclear fuel' was an 'independent business decision' and influenced by . . . market price[s].” Id. at 1367. We rejected that argument, holding that the rerack project was “part and parcel of the [plaintiffs]' mitigation efforts.” Id. We stated that “[t]he long-term benefit of fuel cost savings [influenced by market forces] does not sever its connection to the [plaintiffs]' mitigation efforts,” and that the appropriate inquiry was whether, “[b]y enhancing the racks to accommodate high-enrichment fuel assemblies, the [plaintiffs] mitigated the [g]overnment's breach in a way that produced a benefit.” Id. at 1368. We concluded that the plaintiffs' damages were correctly reduced “by the amount of the benefit received in mitigating the [g]overnment's partial breach of the . . . [c]ontract.” Id.

Here, each insurer mitigated the effects of the government's breach by applying for increased premiums and, as a result, received additional premium tax credits in 2018 as a direct result of the government's nonpayment of cost-sharing reduction reimbursements. Notably, the government does not argue that it is entitled to offset the premium increases in the damages calculation, but it does argue that it is entitled to offset the additional payments made by the government in the form of premium tax credits.

The insurers appear not to dispute that if the elimination of cost sharing-reduction payments directly triggered increased premium tax credits, an offset would be appropriate under a contract theory. But they argue that the premium tax credits were not “direct benefits” of the breach because they depend on actions by the insurers — the decision to pursue increased premiums. These payments were not, in the appellees' phrasing, received in the “first step.” We think the relationship is no less direct because the insured's tax credits did not automatically flow from the elimination of cost sharing reduction payments, and the insurers played a role by securing the increased premiums that in turn resulted in the increased tax credits.

There is thus a direct relationship between cost-sharing reductions and premiums, and between premiums and tax credits. The text of the ACA recognizes the relationship between premiums and cost-sharing reductions. Section 1412 of the ACA provides for the “[a]dvance determination and payment of premium tax credits and cost-sharing reductions.” 42 U.S.C. § 18082 (codifying ACA section 1412). Section 1412(a)(3) states: “the Secretary of the Treasury makes advance payments of [premium tax] credits or [cost-sharing] reductions to the [insurers] . . . in order to reduce the premiums payable by individuals eligible for such credit.” Id. § 18082(a)(3). As we noted in Sanford, this section may be understood to indicate that the statute recognizes that, without cost-sharing reduction reimbursements, “insurers might otherwise seek higher premiums to enable them to pay healthcare providers the amounts enrollees are not paying due to cost-sharing reductions.” Sanford, No. 19-1290, slip op. at 22.

The Claims Court's findings show that the premium tax credits flowed directly from the insurers' mitigation efforts. As the Claims Court found, the plaintiffs themselves recognized this connection. They negotiated for increased premiums (leading to the increased tax credits) in direct response to the cessation of cost-sharing reduction payments:

The Trump administration's termination of cost-sharing reduction payments did not come as a surprise to insurers: “Anticipating that the Administration would terminate [cost-sharing reduction] payments, most states began working with the insurance companies to develop a plan for how to respond. . . . And the states came up with an idea: allow the insurers to make up the deficiency through premium increases. . . .” California, 267 F. Supp. 3d at 1134–35. . . . In other words, by raising premiums for silver-level qualified health plans, the insurers would obtain more money from the premium tax credit program, which would help mitigate the loss of the cost-sharing reduction payments.

Cmty., 141 Fed. Cl. at 754–55 (first alteration in original); id. at 755 n.10 (noting that “increasing silver-level qualified health plan premiums would not harm most consumers who qualify for the premium tax credit because the credit increases as the premium increases”).

The practice of silver loading — and the resulting premium tax credits received by each insurer — “was a direct consequence of the government's breach” of its obligations, and “indeed was an extreme measure forced” by the government's nonpayment. LaSalle, 317 F.3d at 1372. The government's payment of the premium tax credits is directly traceable to the premium increase, and the premium increase is directly traceable to the government's breach. The insurers “received a benefit as a direct result of their mitigation activity.” Kansas Gas & Elec., 685 F.3d at 1368. The argument for an offset is particularly strong here be-cause the insurers received direct payments (rather than indirect benefits, such as efficiency gains) from the government due to their mitigation efforts.

The insurers argue, however, that there are two exceptions to the mitigation principle that defeat the government's claim to an offset: (1) the prohibition on so-called “pass-through” defenses and (2) the collateral source rule. As to the “pass-through” defense, the insurers argue that the government, as a breaching party, may not claim mitigation of damages when the non-breaching party “passe[s] through” its losses to its customers. Appellees' Suppl. Damages. Br. 15 (citing Hughes Commc'ns Galaxy, Inc. v. United States, 271 F.3d 1060, 1072 (Fed. Cir. 2001)).¹¹ The insurers assert that the cases stand for the proposition that mitigation may only be considered in the “first step,” and that “later-step” recoveries such as pass-through are “irrelevant” to the calculation of damages. Id. at 10. But this is not a case where a third-party customer pays for the insurers' losses, as was the case in Hughes.¹² The complexity of the process cannot obscure the underlying economic reality that the government is paying at least some of the increased costs that the insurers incurred as a result of the government's failure to make cost-sharing reduction payments. See Gov't Suppl. Damages Br. 24 (“[T]he government is not urging that [the] plaintiffs' damages should be reduced merely because [the] plaintiffs passed on their cost-sharing reduction expenses to customers. The crucial point is that [the] plaintiffs . . . passed these expenses on to the government itself, which by virtue of the ACA's structure is paying the cost-sharing reduction expenses . . . in the form of higher premium tax credits.”).

The government's claim is not that damages should be reduced because the insurers passed on the increased costs to their customers, but that “the insurers . . . obtain[ed] more money from the premium tax credit program, which would help mitigate the loss of the cost-sharing reduction payments.” Cmty., 141 Fed. Cl. at 755 & n.10. The pass-through exception, to the extent that it is applicable to con-tract damages, does not apply here.

Second, the insurers invoke the collateral source rule, arguing that the additional premium tax credits were collateral benefits that should not be credited against their damages. The collateral source rule is a generally recognized principle of tort law that “bars a tortfeasor from reducing the damages it owes to a plaintiff 'by the amount of recovery the plaintiff receives from other sources of compensation that are independent of (or collateral to) the tortfeasor.'” Johnson v. Cenac Towing, Inc., 544 F.3d 296, 304 (5th Cir. 2008) (quoting Davis v. Odeco, Inc., 18 F.3d 1237, 1243 (5th Cir. 1994)); see, e.g., Chisholm v. UHP Projects, Inc., 205 F.3d 731, 737 (4th Cir. 2000); Fitzgerald v. Expressway Sewerage Constr., Inc., 177 F.3d 71, 73 (1st Cir. 1999). Thus, the collateral source rule bars a reduction of damages due to “insurance policies and other forms of protection purchased by [the] plaintiff,” Johnson, 544 F.3d at 305, or unemployment benefits in the case of a wrongful-discharge case, Craig v. Y & Y Snacks, Inc., 721 F.2d 77, 83 (3d Cir. 1983).

As with the insurers' pass-through argument, their collateral source rule argument fails. We are aware of no authority, and the insurers cite none, holding that the collateral source rule applies to contract damages, and the prevailing authority rejects any such limitation. See, e.g., United States v. Twin Falls, 806 F.2d 862, 873 (9th Cir. 1986) (“We have found no authority to support the application of the collateral source rule in the contracts field.” (collecting cases rejecting the application of the collateral source rule to contract-based damages)), overruled on other grounds as recognized by Ass'n of Flight Attendants v. Horizon Air Indus., Inc., 976 F.2d 541, 551–52 (9th Cir. 1992); Star Ins. Co. v. Sunwest Metals Inc., 691 F. App'x 358, 361 (9th Cir. 2017) (noting that “California courts have declined to extend the collateral source rule to contract-based claims” and that contract damages rules are “[u]nlike” those in tort damages); LaSalle, 317 F.3d at 1372 (declining to apply the collateral source rule to government con-tracts). In any event, even if that rule applied here, the “source of compensation” is the not “independent” of the government. The source is the government itself. See Phil-lips v. W. Co. of N. Am., 953 F.2d 923, 931 (5th Cir. 1992) (“The [collateral source] rule is intended to ensure that the availability of outside sources of income does not diminish the plaintiff's recovery, not make the tortfeasor pay twice.”). The collateral source rule does not bar the reduction in damages.

We conclude that additional premium tax credits were received by Community and Maine Community in 2018 as a direct consequence of their mitigation efforts following the government's nonpayment of 2018 cost-sharing reduction reimbursements, and the Claims Court was required to credit the government with such tax credit payments in determining damages.

IV

Determining the amount of premium tax credits paid to each insurer is necessarily a fact-intensive task. Be-cause the Claims Court rejected the government's mitigation theory on a limited summary judgment record, it did not address these calculation issues. And as the insurers conceded in their briefing before the Claims Court, to the extent that the insurers' premium changes are “relevant . . . to [the] quantum,” they involve “factual questions that cannot be resolved on [the existing motion for summary judgment].” Community Reply in Supp. of Mot. for Summ. J. 15, Cmty. Health Choice, Inc. v. United States, No. 18-cv-00005, 141 Fed. Cl. 744, ECF No. 20 (Nov. 30, 2018); Maine Community Mot. for Summ. J. 1, Me. Cmty Health Options v. United States, No. 17-cv-02057, 143 Fed. Cl. 381, ECF No. 31 (Apr. 8, 2019) (adopting “all of the arguments regarding benefit year 2018 raised by . . . Community . . . in [its] brief[ ]”). We therefore remand to the Claims Court for a determination of the amount of premium increases (and resultant premium tax credits) attributable to the government's failure to make cost-sharing reduction payments. This will require either new summary judgment motions or a trial.

We note that three principles will govern the remand proceedings.

First, as the insurers argue, some of the silver-level premium increases (and resulting tax credits) may be caused by other factors, such as market forces or increased medical costs. To the extent that this is the case, the government's liability is not reduced by the tax credits attributable to these other factors.

Second, as previously mentioned, increasing the premium rates for silver plans resulted in an increase in premium tax credits for all plans on the exchange. In some states, state regulators have also allowed insurers to re-coup part of their lost cost-sharing reduction reimbursements by increasing premiums for other, non-silver plans on the exchange. In these circumstances, the tax credits for these other plans (attributable to the silver plan premium increase) are still caused by the elimination of cost-sharing reduction payments and will, of course, reduce the government's liability. But we do not address whether in situations where, as here, there have been no premium increases for other plans, the government's liability should be reduced for the increased tax credit payments with respect to other plans. We leave that issue to the Claims Court in the first instance.

Finally, the insurers will bear the burden of persuasion with respect to the amount of the tax-credit increase attributable to the loss of cost-sharing reduction reimbursements. Other circuit courts and state courts applying state law are inconsistent as to which party bears the burden of persuasion with respect to the amount of mitigation.¹³ But in the federal context the rule is clear. The plaintiffs bear the burden of proof:

[A] non-breaching plaintiff bears the burden of per-suasion to establish both the costs that it incurred and the costs that it avoided as a result of a breach of contract. The breaching party may be responsible for affirmatively pointing out costs that were avoided, but once such costs have been identified, the plaintiff must incorporate them into a plausible model of the damages that it would have incurred absent the breach.

Bos. Edison Co. v. United States, 658 F.3d 1361, 1369 (Fed. Cir. 2011) (citing S. Nuclear Operating Co. v. United States, 637 F.3d 1297, 1304 (Fed. Cir. 2011)); see also Sys. Fuels, Inc. v. United States, 666 F.3d 1306, 1312 (Fed. Cir. 2012) (collecting cases). Here, the government has affirmatively pointed out the insurers' avoided costs (due to in-creased premium tax credits). Therefore, it was the insurers burden to incorporate those benefits into their damages calculations. Energy Nw. v. United States, 641 F.3d 1300, 1309 (Fed. Cir. 2011) (explaining that, to establish damages, “a plaintiff [must] show what it would have done in the non-breach world, and what it did post-breach”). We think that this allocation of the burden of proof is particularly appropriate here because the insurers were already required by section 1003 of the ACA to provide “justification[s]” for premium rate increases. 42 U.S.C. § 300gg-94(a)(2). Thus, Community and Maine Community — having previously justified their silver-level premium increases — are “in the best position to adduce and establish such proof.” S. Nuclear, 637 F.3d at 1304 (quoting 11 Corbin on Contracts § 57.10 n.15 (2005)).

According to the insurers, they cannot be expected to bear this burden of proof by comparing “each insurer's financial picture now in relation to what it hypothetically might have been if [the cost-sharing reduction reimbursements] had been timely paid.” Appellees' Suppl. Damages Br. 9. Specifically, the insurers argue that they cannot “submit a hypothetical model establishing what their costs would have been in the absence of breach.” Id. at n.9 (quoting Gov't Suppl. Damages Br. 8). Given the explicit arguments that the insurers here have made for rate increases, we doubt that proof will be as difficult as the insurers' claim. In any event, as we have discussed, our cases make clear that the plaintiff seeking to recover damages must “prov[e] causation by comparing a hypothetical 'but for' world to a plaintiff's actual costs.” Energy Nw., 641 F.3d at 1306 (quoting Yankee Atomic Elec. Co. v. United States, 536 F.3d 1268, 1273–74 (Fed. Cir. 2008)). The insurers here cannot avoid their burden to prove damages.

V

Although we do not address the Claims Court's holding with respect to the insurers' implied-in-fact contract theory, the same damages analysis would apply to that claim as well, since, as the Claims Court recognized, a claim for breach of an implied-in-fact contract is subject to the same damages limitations as an ordinary contract. See Cmty., 141 Fed Cl. at 767–70 (analyzing damages for breach of an implied-in-fact contract under “[t]he general rule in com-mon law breach of contract cases” (quoting Estate of Berg v. United States, 687 F.2d 377, 379 (Ct. Cl. 1982)); see, e.g., Lindquist Ford, Inc. v. Middleton Motors, Inc., 557 F.3d 469, 481 (7th Cir. 2009), as amended (Mar. 18, 2009) (“[A]n implied-in-fact contract is governed by general contract principles.”); Hill v. Waxberg, 237 F.2d 936, 939 (9th Cir. 1956) (explaining that “the general contract theory of compensatory damages should be applied” in an action for breach of an implied-in-fact contract). There is thus no need on remand to separately address the insurers' implied-in-fact contract claim.

AFFIRMED IN PART, REVERSED AND REMANDED IN PART

COSTS

No costs.