Irreconcilable Differences: Analyzing The Split In The First And Second Circuit Courts Of Appeals’ Decisions On National Bank Act Preemption Of State Interest-on-Escrow Laws

On May 5, 2026, a majority panel of the U.S. Court of Appeals for the Second Circuit (the “Second Circuit”) issued an opinion holding that the National Bank Act (NBA) preempts New York General Obligations Law (GOL) § 5‑601, which requires banks to pay interest on mortgage escrow accounts. Cantero v. Bank of America, N.A., __F.4th __, 2026 U.S. App. LEXIS 13066 (2d Cir. May 5, 2026) (“Cantero III”). The decision marks the second time the Second Circuit has concluded that GOL § 5‑601 is preempted by the NBA, having previously done so in 2022. Cantero v. Bank of America, N.A., 49 F.4th 121 (2d Cir. 2022) (“Cantero I”). The Supreme Court, however, vacated Cantero I and remanded the case after determining that the Second Circuit had not applied the proper preemption standard. Cantero v. Bank of America, N.A., 602 U.S. 205 (2024) (“Cantero II”).

The Second Circuit’s decision in Cantero III results in national banks being in the same position they were prior to the Supreme Court’s decision in Cantero II. Specifically, banks once again are facing a Circuit split, with the U.S. Courts of Appeals for the Ninth Circuit (the “Ninth Circuit”) and the First Circuit (the “First Circuit”) ruling state interest-on-escrow requirements are not preempted under the NBA, and the Second Circuit ruling they are. The Circuit Courts are purporting to apply the same governing framework, reviewing and applying the same Supreme Court precedent to similar laws, and yet they reach opposite results, underscoring material differences and uncertainty in how lower courts are, or should be, interpreting Cantero II. In addition, the First and Ninth Circuits have established interpretations that, in practice, set a higher bar that banks would need to meet in those Circuits to successfully argue that a state consumer financial law would be preempted.

We set out below a brief background and summary of the Second Circuit’s reasoning in Cantero III, followed by key areas of divergence from the First Circuit’s decision in Conti v. Citizens Bank, N.A., 157 F.4th 10 (1st Cir. 2025) that highlight differences that warrant further clarification from the Supreme Court on its Cantero II preemption standard. We also discuss what weight and relevance the Second Circuit afforded a recent related (and at the time, proposed but not final) rulemaking and preemption determination by the Office of the Comptroller of the Currency (OCC) regarding national bank powers with respect to escrow accounts and conclude by outlining the current implications of these decisions and actions for national banks.

Background Circuit Court Consideration of Interest-on-Escrow Laws

In Cantero II, the Supreme Court emphasized that courts must apply the preemption framework for state consumer financial laws required under 12 U.S.C. § 25b, which incorporates the preemption standard utilized by the Supreme Court in Barnett Bank of Marion County v. Nelson, 517 U.S. 25 (1996). In Barnett Bank, the Supreme Court held that a state law is preempted by the NBA when it prevents or significantly interferes with the exercise of a national bank’s powers. The Supreme Court instructed the Second Circuit that a Barnett Bank preemption analysis requires a “nuanced comparative analysis” that evaluates the nature and degree of the state law’s interference by considering the text and structure of the laws, Supreme Court precedent addressing federal banking preemption, and common sense. Cantero II at 220 & n.3. In vacating Cantero I, the Supreme Court held that the Second Circuit instead had applied a bright line preemption standard that only considered whether the state law sought to “control” a national banking power. Cantero II at 213, 220.

The Second Circuit, however, is not the first Federal appellate court that has applied the Cantero II-styled preemption analysis to state interest-on-escrow laws. In Conti, the First Circuit also applied Cantero II, but concluded that the NBA does not preempt Rhode Island’s escrow‑interest statute. The Supreme Court denied Citizens Bank’s petition for writ of certiorari. Citizens Bank, N.A. v. Conti, 2026 U.S. LEXIS 1670 (U.S., Apr. 20, 2026).

In addition, the Ninth Circuit has also addressed, post-Cantero II, whether the NBA preempts a similar California interest-on-escrow law. In 2022, a Ninth Circuit panel concluded that it remained bound by a 2018 Ninth Circuit decision in Lusnak v. Bank of America, N.A., 883 F.3d 1185 (9th Cir. 2018), which held that the NBA does not preempt the California law. Kivett v. Flagstar Bank, FSB, 2022 U.S. App. LEXIS 13347 (9th Cir. 2022) (Kivett I). Flagstar petitioned for certiorari, and the Supreme Court, like its disposition of Cantero I, vacated and remanded Kivett I to the Ninth Circuit for reconsideration. Flagstar Bank, N.A. v. Kivett, 144 S.Ct. 2628 (2024).

On remand of Kivett I, a Ninth Circuit panel, through a series of orders, eventually reaffirmed the determination that the NBA does not preempt the California law, but it did so on procedural, rather than on substantive, grounds. Kivett v. Flagstar Bank, FSB, 154 F.4th 640 (9th Cir. 2025) (Kivett II). Specifically, the panel majority held that, under Ninth Circuit precedent, it could overrule Lusnak only if it determined that the prior decision was “clearly irreconcilable” with Cantero II’s preemption standard. Kivett II at 644-645. The panel concluded that it could not make that determination, noting that existing Supreme Court precedent provided only limited guidance that did not compel a specific preemption result and that only the Ninth Circuit sitting en banc could overrule Lusnak. Kivett II at 647-649. Circuit Judge Nelson, writing in dissent, determined both that the analysis performed in Lusnak was “clearly irreconcilable” with the analysis required by Cantero II and that the state law was preempted. The Ninth Circuit subsequently denied Flagstar’s petition for rehearing en banc in March 2026. Kivett v. Flagstar Bank, FSB, 2026 U.S. App. LEXIS 8917 (9th Cir. March 26,2026).

The Second Circuit’s Cantero III Preemption Analysis

In reconsidering whether GOL § 5‑601 could be enforced against national banks, the Second Circuit framed the Cantero II standard as requiring courts to determine:

1. whether the state consumer financial law affects the exercise of a Federal bank power; and
2. if so, whether the law “prevents or significantly interferes” with that power, by considering (i) the nature and (ii) the degree, of that interference. Cantero III at 13.

Applying that framework, the Second Circuit first considered whether GOL § 5‑601 affected a Federal banking power, rather than merely affected banks. The court noted that Federal law expressly authorizes national banks to make real estate secured loans, 12 U.S.C. § 371(a), and confers incidental authority to banks to offer related escrow accounts under 12 U.S.C. § 24 (Seventh). Cantero III at 14. The Second Circuit then concluded that New York’s interest‑on‑escrow law interferes with these powers by requiring banks to pay a mandated minimum rate of interest-on-escrow balances. Id.

Turning to the questions of the nature and the degree of the interference, the Second Circuit held that the law “significantly interferes” with national bank powers. In reaching that conclusion, the Second Circuit first reviewed Supreme Court precedent and determined that the nature of the New York law’s interference was not analogous to those of generally applicable state laws or to those that apply a generally applicable rule to a bank that the Supreme Court considered and upheld in McClellan v. Chipman, 164 U.S. 347 (1896) (state fraudulent conveyance law), and Anderson National Bank v. Luckett, 321 U.S. 233 (233 (1944) (common law escheatment principles). Cantero III at 21.

The Second Circuit instead characterized the nature of the New York law’s interference as more like state laws that the Supreme Court held were preempted due to their interference with an expressly- or broadly- granted power in Fidelity Federal Savings and Loan Ass’n v. de la Cuesta, 458 U.S. 141 (1982) (restrictions on broad Federal power to utilize due on sale clauses), and Barnett Bank (restrictions on banks from exercising Federal authority to act as an insurance agent). Cantero III at 21-24. The Second Circuit interpreted the Truth in Lending Act (TILA), which permits state required interest-on-escrow in limited contexts, and the Real Estate Settlement Procedures Act (RESPA), which extensively regulates escrow accounts but does not require payment of interest, as collectively implying that Congress granted national banks a broad power to set interest rates for most mortgage-escrow accounts and found no indication that Congress intended to make that power otherwise subject to state restrictions. Id.

With respect to the final element of its analysis, the Second Circuit concluded that the degree of the New York law’s interference was “significant.” Id. at 24-26. The Second Circuit noted that banks incur both operational and compliance costs in administering escrow accounts and that the interest payment requirements may result in banks deciding to offer fewer escrow accounts, to take action to recover these costs in other ways, or to reduce their level of lending. Id. The Second Circuit found this power was a core component of a bank’s authorized lending power and that the degree of interference was therefore similar to the significant interference the Supreme Court found was represented by a New York law that limited a bank’s ability to advertise its authorized savings account powers in Franklin National Bank of Franklin Square v. New York, 347 U.S. 373 (1954). Id. Presumably giving recognition of Cantero II’s direction that lower courts use common sense in conducting their analysis, the Second Circuit noted that the New York escrow law operated as a greater interference on bank powers than the advertising restrictions in Franklin, citing, among other sources, Circuit Judge Nelson’s dissenting opinion in Kivett II. Cantero III at 25.

What interpretive differences in the Second and First Circuits’ application of Cantero II account for the split?

The difference in the outcomes in Cantero III and Conti reflects the fundamental difference in the Second and First Circuits’ understandings of how Cantero II requires lower courts to apply Supreme Court precedent. Key points of disagreement between the Circuit Courts include the following:

!. Whether Courts Can Infer a Broad Incidental Bank Power Under Federal Law Post-Dodd-Frank.

As discussed above, the Second Circuit interpreted TILA and RESPA as reflecting Congress’ broad implied authorization for national banks to set the interest terms of escrow accounts, noted as a core component of a bank’s mortgage lending powers, without regard to state interest requirement (other than for those mortgages covered by TILA). Id. at 21-24. The Second Circuit viewed its approach as consistent with the Supreme Court’s analysis in Barnett Bank, Franklin, and Fidelity, where the Court noted that the text and structure of Federal laws that incorporate only some aspects of state law reflect a broad grant of a power that is not otherwise preempted by state law. Id. at 22-23. In contrast, the First Circuit in Conti read Dodd‑Frank as affirmatively recognizing a broad role for state consumer financial regulation and that neither TILA nor RESPA were applicable to the analysis and that no Supreme Court precedent had held state law preempted by silence. Conti at 21-22. In its view, Dodd-Frank now limits any inference of implicit Federal authority that prevents a bank’s operations from being subject to state laws, resulting in the NBA only preempting state laws that either directly conflict with express federal powers or that impose unusual conditions with extreme outcomes on national banks and their customers.

2. How strictly should courts apply the facts and reasoning of Supreme Court precedent.

As noted above, the First Circuit adopted a narrow reading of Supreme Court precedent, distinguishing the relevancy or applicability of cases depending on whether the state law involves an express conflict with federal powers or involves an interference with, but not a conflict with, federal powers. This approach effectively limits the precedent supporting “significant interference” preemption to First National Bank of San Jose v. California, 262 U.S. 366 (1923) (state law held preempted that imposed unusual presumptions of escheatment that may have caused customers to stop placing deposits with a bank), and Franklin.

The Second Circuit rejected the First Circuit’s narrow framing of Supreme Court precedent. The First Circuit considered Barnett Bank and Fidelity to be “generally inapposite” because they involved express grants of bank power while, in contrast, it noted that no Federal law either expressly prohibited state interest-on-escrow laws or expressly reserved escrow-interest rate decisions to banks. Conti at 20. In contrast, the Second Circuit undertook a broader comparative analysis, emphasizing that Cantero II and the limited universe of relevant Supreme Court decisions necessitate a more nuanced assessment of their reasoning. For example, the Second Circuit declined to discount decisions such as Barnett Bank, Franklin, and Fidelity, which more generally consider state law constraints on bank flexibility and effective operations. The Second Circuit also declined to follow Conti’s analytical approach because it failed to consider the “practical reality” that state law restrictions impacting pricing materially impact a bank’s operations. Cantero III at 28-29.

3. Whether banks are required to develop a factual record to support claims of “significant interference”.

The First Circuit suggested, particularly during oral argument in Conti, that banks may need to develop an evidentiary record demonstrating “significant interference.” The Second Circuit, reading Supreme Court precedent, determined that a factual record was not necessary, noting that no Supreme Court precedent has required, or even relied on, a factual record. Cantero III at 20 & n.5. In an apparent nod to Cantero II’s call for courts to assess the nature and degree of interference posed by the state law by also applying common sense, the Second Circuit concluded that impacts to bank pricing decisions almost by definition create more interference than advertising restrictions. Cantero III at 25-26. Neither the First Circuit nor Circuit Judge Perez’ dissent in Cantero III recognized this correlation as reflecting common sense evidence of the law’s practical impediments on a bank’s operations.¹

The efficacy of the OCC’s actions to support preemption remains to be seen, but a pending U.S. District Court evaluation may provide initial clues.

Earlier this year, the OCC responded both to state interest-on-escrow laws and provisions in Illinois’ Interchange Fee Prohibition Act (IFPA) that restrict the ability of national banks to receive interchange fees. The OCC issued interim final and proposed rules clarifying the authority of national banks under 12 U.S.C. § 24 (Seventh) to receive non-interest charges for their products and services, including for interchange fees set by card networks, and to set the terms and conditions of mortgage escrow accounts.² In addition, the OCC issued a proposed determination that state interest-on-escrow laws are preempted by the NBA and an interim final preemption determination that the IFPA’s fee restrictions are preempted.³

While the OCC has now finalized its escrow powers rule and its preemption determination on state interest-on-escrow laws, these actions remained only proposals at the time of Cantero III. The Second Circuit did, however, reference and leverage statements and positions of the OCC in the proposed actions in the Second Circuit’s consideration of (i) the benefits of escrow accounts; (ii) the relationship between escrow accounts and a bank’s core lending powers; (iii) the flexibility provided under Federal law for banks to set the terms of such accounts; and (iv) the potential impact of state laws. So, while the validity of the OCC’s escrow powers rule and preemption determination was not directly at issue in Cantero III, the Second Circuit did give some weight, although not overt deference, to the OCC’s interpretations and positions. After the Supreme Court’s overruling of general court deference to agency determinations in Loper Bright Enterprises v. Raimondo, 603 U. S. 369 (2024), the Second Circuit’s consideration of the OCC’s statements and positions reflects the expected review of agency action that is in line with the standard required by Skidmore v. Swift & Co., U.S. 323 U.S. 134 (1944).

Circuit Judge Perez’s dissent in Cantero III, however, criticized the OCC’s approach of simultaneously proposing a rule granting banks broad discretion over escrow terms and determining that all related state laws are then preempted because they constrain that discretion, arguing that this approach created a conflict in the manner of the state law preempted in Fidelity. Circuit Judge Perez stated that this approach is inconsistent with the “limited” grant of authority to preempt state regulations under Dodd-Frank⁴ and suggested the approach should result in little deference under the judicial review standard Dodd-Frank requires for courts directly reviewing the OCC’s preemption determinations.

While the OCC’s escrow rule and preemption determination are currently only proposals, the OCC issued an interim final rule on a bank’s authority to receive non-interest fees and charges and an interim final determination that the IFPA’s interchange fee prohibitions are preempted. The U.S. Court of Appeals for the Seventh Circuit (the “Seventh Circuit”) had accepted an expedited appeal of a U.S. District Court order that declined to enjoin enforcement of the IFPA’s interchange fee restrictions, Illinois Bankers Ass’n v. Raoul, __ F. 4th __, 2026 U.S. Dist. LEXIS 27167 (U.S, Dist. Ct., N.D. Ill., E. Div., Feb. 10, 2026). The Seventh Circuit, however, recently cancelled oral arguments that it had previously scheduled for May 13^th, and instead remanded the case back to the District Court explicitly for the lower court to address the impact of the OCC’s issuance of the interim final rule and an interim final preemption determination. Ill. Bankers Ass’n v. Raoul, 2026 U.S. App. LEXIS 13896 (7th Cir., May 8, 2026). An expedited decision from the District Court, or perhaps a temporary injunction of the IFPA’s enforcement, may be expected given the pending effective date of the IFPA later this summer.

Pending further guidance from the Supreme Court, what do these decisions mean for national banks?

Despite the clear need for further clarifying guidance from the Supreme Court, it is unlikely to come in the near term. Even if the Supreme Court were to accept the plaintiff’s expected appeal of Cantero III, it took two years from Cantero I before the Supreme Court issued its opinion in Cantero II. As reflected in this discussion of recent preemption cases, developments are moving quickly and national banks will need to consider the following:

• While payment of interest-on-escrow accounts is required in states that have such a mandate and are within the First and Ninth Circuits, but not within the Second Circuit, the outcome of challenges to similar state laws within other Circuits remains unclear for the time being. If consistent with bank risk appetite, a bank may choose to maintain its current practices with respect to the payment of interest in those states rather than seeking to make operational changes in advance of further clarity from the Supreme Court. Banks that are not paying interest in states within the First and Ninth Circuits where interest is required, however, face increased litigation risk.
• Banks may see an increase in interest-on-escrow legislative efforts in states that do not currently impose similar types of requirements, particularly states within the First and Ninth Circuits, and will need to track any legislative initiatives closely to determine account requirements and approach.
• Banks will need to consider what impact, if any, the recent Circuit Court decisions may have on other types of state laws as well, particularly for those requiring interest on other accounts or those restricting other types of bank non-interest fees and charges. Legislatures and state banking regulators, particularly for states in the First Circuit, may feel emboldened to increase legislative and rulemaking efforts applicable to national banks.
• Finally, banks should consider what the Circuit split means more broadly for their approach to, and risk tolerance for, asserting NBA preemption in the First, Second and Ninth Circuits, as well as the other Circuits which have not applied Cantero II to date. Now that another Circuit has weighed in, other Circuits may be inclined to adopt the more nuanced analysis of the Second Circuit given its influence over banking law in the United States.

Bank legal, compliance, and risk teams need to keep bank senior management updated on fast moving developments in case law and new legislation. The Financial Regulatory Advice & Response Team monitors these developments closely and advises institutions across the full range of state law requirements, associated preemption considerations, and what these cases mean and appropriate bank response measures. Please reach out to a member of our team if you have further questions.

Footnotes

1. Conti at 25 (“Citizens has not developed any substantial argument about the practical effects that arise from the enforcement of the Rhode Island statute….”); Cantero III (Perez, J., dissenting) at 40 and n.7.

2. Notice of Proposed Rulemaking: Real Estate Lending Escrow Accounts, December 23, 2025 (https://occ.gov/news-issuances/news-releases/2025/nr-occ-2025-133a.pdf) and Interim Final Rule: National Bank Non-Interest Charges and Fees (https://occ.gov/news-issuances/news-releases/2026/nr-occ-2026-32a.pdf). The OCC finalized the proposed rulemaking on escrow accounts on May 1, 2026 (https://occ.treas.gov/news-issuances/news-releases/2026/nr-occ-2026-37a.pdf).

3. Notice of Proposed Rulemaking, Preemption Determination: State Interest-on-Escrow Laws, December 23, 2025 (https://occ.gov/news-issuances/bulletins/2025/bulletin-2025-50.html) and Notice of Proposed Rulemaking, Order Interim Final Order Preempting the Illinois Fee Prohibition Act, December 23, 2025 (https://occ.gov/news-issuances/federal-register/2026/91fr23150.pdf). The OCC finalized its preemption determination on state interest-on-escrow laws on May 1, 2026 (https://occ.treas.gov/news-issuances/news-releases/2026/nr-occ-2026-37b.pdf).

4. Both Circuit Judge Perez’s dissent in Cantero III and the First Circuit’s decision in Conti suggest a view that Dodd-Frank’s provisions require that most state consumer financial laws cannot be preempted, or stated in the reverse, that few state laws can be preempted. Conti at 21 (describing NBA preemption post-Dodd Frank as a “limited” preemption regime). In our view, this approach represents a fundamental misreading of Dodd-Frank and also dismisses express language by a unanimous decision of the Supreme Court in Cantero II. State laws, no matter how many of them, are preempted if they prevent or significantly interfere with the exercise of a bank’s powers. While Cantero II certainly rejected a reading of Barnett Bank as employing a preemption test that defaults to preempting most state laws, it also explicitly cautioned against employing a counter test that would result in few state laws being preempted. Cantero II at 221 (“By contrast, the plaintiffs would yank the preemption standard to the opposite extreme, and would preempt virtually no non-discriminatory state laws that apply to both state and national banks.”). The test is the nature and degree of the law’s interference, not consideration of the number of state laws that have been or may be preempted.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.