Don R. Willett, Circuit Judge

Appeal from the United States District Court for the Western District of Texas, USDC No. 5:21-CV-844, USDC No. 1:21-CV-786, Xavier Rodriguez, U.S. District Judge Uzoma Nkem Nkwonta, Christopher D. Dodge (argued), Elias Law Group, L.L.P., Washington, DC, for Plaintiffs—Appellees LULAC Texas, Vote Latino, Texas Alliance for Retired Americans, Texas AFT. Jonathan Backer (argued), Alisa Philo, U.S. Department of Justice, Civil Rights Division, Appellate Section, Washington, DC, Bonnie Ilene Robin-Vergeer, Senior Attorney, U.S. Department of Justice, Civil Division, Appellate Section, Washington, DC, for Plaintiff—Appellee United States of America. Lanora Christine Pettit (argued), William Francis Cole, Esq., Office of the Texas Attorney General, Solicitor General Division, Austin, TX, William Thomas Thompson, Assistant General, Lehotsky Keller Cohn, L.L.P., Austin, TX, Colleen Tran, Texas Office of the Attorney General, Government Relations, Austin, TX, for Appellants. Before Richman, Chief Judge, and Wiener and Willett, Circuit Judges. Don R. Willett, Circuit Judge:

After the Texas Legislature amended the Election Code in 2021, the United States and others sued, alleging the changes were racially discriminatory. When the plaintiffs sought discovery from individual, non-party state legislators, those legislators withheld some documents, citing legislative privilege. The district court largely rejected the legislators' privilege claims, and they filed this interlocutory appeal. We REVERSE.

I

The Texas Legislature recently amended the Election Code as it relates to voter registration, voting by mail, poll watchers, and other aspects of election integrity and security. The United States, LULAC Texas, and dozens of other plaintiffs sued (together, "Plaintiffs"). They argued that the Legislature acted with racially discriminatory intent, and thus that the amendment violates the Constitution and the Voting Rights Act. The district court consolidated many of the suits. Plaintiffs then sought discovery from individual, non-party legislators related to the circumstances surrounding the amendment's proposal and passage. The legislators produced some documents, but they withheld others, citing legislative privilege. Plaintiffs moved to compel production. The district court rejected most of the legislators' privilege claims and ordered them to produce about 220 documents. But the district court stayed that order while the legislators pursued this interlocutory appeal.

II

Our appellate jurisdiction generally extends only to "final decisions of the district courts." But as the Supreme Court explained in Cohen v. Beneficial Industrial Loan Corporation, courts have "long given" this restriction "a practical rather than a technical construction." Under that construction, we have jurisdiction over " 'a narrow class of decisions . . .' immediately appealable as collateral orders even if no final judgment has been rendered." Orders are immediately appealable under this rule only if they "(1) conclusively determine the disputed question, (2) resolve an important issue completely separate from the merits of the action, and (3) [are] effectively unreviewable on appeal." "[T]he decisive consideration is whether delaying review until the entry of final judgment 'would imperil a substantial public interest' or 'some particular value of a high order.' " Another constraint is that we do not apply this rule case-by-case or in an "individualized" manner. Instead, "our focus is on 'the entire category to which a claim belongs' " and on "the class of claims[ ] taken as a whole." The class at issue in this appeal consists of orders denying non-party state-legislators' assertions of legislative privilege. We agree that the orders in this class are immediately appealable.

This class satisfies "the three traditional Cohen conditions." The order's conclusiveness is apparent, among other reasons, because "failure to comply with it may result in sanctions against" the legislators. The class also "involves important questions" such that "the cost of allowing immediate appeal" is justified. The importance derives from the purpose of legislative privilege, which is not to protect against disclosure in general, but to foster the "public good" by protecting lawmakers from "deterrents to the uninhibited discharge of their legislative duty." Requiring legislators to negotiate protective orders or to suffer contempt proceedings would diminish that protection. For the same reason, this class of claims is not "adequately vindicable" at a later stage of the litigation. For one thing, litigation itself distracts lawmakers from the job that voters sent them to do. They cannot get that time back. But even setting that aside, once the legislators produce documents, an appellate court cannot "remedy the improper disclosure of privileged material in the same way [it can] remedy a host of other erroneous evidentiary rulings: by vacating an adverse judgment and remanding for a new trial in which the protected material and its fruits are excluded from evidence." As non-parties, the legislators cannot move for a new trial. And even if they could, a new trial cannot retract privileged information that has been shared into the public domain. For all those reasons, we have interlocutory jurisdiction under Cohen.

The class of claims also satisfies Mohawk's formulation, which allows interlocutory jurisdiction when delaying review would harm "a substantial public interest" or "some particular value of a high order." Delaying review for this class of claims would imperil both of those interests. The public has a substantial interest in ensuring that elective office remains an invitation to draft legislation, not defend privilege logs. Freedom from constant distraction is a high-order value. That is especially so for this class, which consists solely of claims from non-parties who "lack appellate remedies available to the contenders in litigation."

It should come as no surprise, then, that our jurisdiction has long extended to adjacent classes of claims. In Overby v. United States Fidelity and Guarantee Company, we considered a dispute between a bank and a surety company. The surety company sought to discover certain documents in the bank's possession, but the Acting Secretary of the United States Treasury filed a "formal claim of privilege" over the documents. We held that the district court's denial of privilege was immediately appealable under the collateral order doctrine, "even though the main suit between the bank and the surety ha[d] not been concluded." Why? Because "[a]fter such production, there would be no further point to the claim of privilege, it would be irretrievably breached and beyond the protection of an appellate court."

Likewise, in Carr v. Monroe Manufacturing Co., we held that "discovery orders may be appealable where a governmental privilege is asserted and the government is not a party to the suit." We reasoned that "[i]n such cases," our exercise of jurisdiction is warranted because "the asserted governmental interest may be 'irretrievably breached' by disclosure, and the government has no remedy on appeal from a final judgment in the original action." The governmental interest is no less breached "where the matter sought to be discovered is held by one of the parties on the government's behalf." And just days ago, in Jackson Municipal Airport Authority v. Harkins, we held that "appellate jurisdiction exist[ed]" over a class of claims involving legislators who were themselves parties in that case.

Plaintiffs argue that Mohawk bars interlocutory jurisdiction. There, the Supreme Court held that the collateral-order doctrine does not cover a private litigant's assertion of attorney-client privilege. This case is different in all three respects: here we have governmental (not private) non-parties (not litigants) asserting legislative privilege (not attorney-client privilege). Indeed, Mohawk "express[ed] no view on" how the collateral-order doctrine applies to "governmental privileges." We have held, too, that "Mohawk does not speak to the predicament of third parties, whose claims to reasonable protection from the courts have often been met with respect." Nothing in Mohawk forbids jurisdiction here. Rather, Mohawk is relevant because it explains how to determine which classes of claims are immediately appealable. That test favors jurisdiction over this class of claims. Plaintiffs see Mohawk differently, but we disagree, as explained below.

For instance, Plaintiffs suggest that "[t]he absence of . . . constitutional values" means that no "high order" issues are at stake here. But constitutional values cannot be the litmus test, for we already "allow[ ] immediate appeal of orders that unseal a nonparty's confidential business documents." And even if some constitutional value were required, federalism and comity both fit the bill, and both are at stake when a federal court orders state lawmakers to produce documents. Plaintiffs also say that the legislators' privilege claims are inseparable from the underlying case's merits, because the order compels documents that "bear directly on" the merits. That argument misstates Cohen's second factor, which requires separate "issue[s]." Here, the underlying merits issue is whether the amendment violates federal law, while the issue in this appeal is whether the legislators can claim privilege. Those are separate issues.

Finally, Plaintiffs point us to decisions from our sister circuits. We doubt that Overby and Carr leave us much room to import new law. Rather, we agree with the Eleventh Circuit, which is bound by our pre-1981 precedent, and which has held that Mohawk was "not the kind of 'clearly on point' Supreme Court precedent" that would justify disregarding the line of Fifth Circuit cases allowing immediate appeal of orders directing non-parties to produce material for which they assert a governmental privilege. Mohawk, that court concluded, "said nothing about . . . [legislative] privilege[ ] . . . or any governmental privilege for that matter. Nor did it alter the scope of the collateral order doctrine." Our jurisdiction here is secure.

III

We review discovery orders "for abuse of discretion." "The district court's legal conclusions should be reviewed de novo, and its factual findings should not be disturbed unless they are clearly erroneous."

"[L]egislative privilege . . . is an evidentiary privilege[ ] governed by federal common law, as applied through Rule 501 of the Federal Rules of Evidence." We begin by defining the privilege's scope—that is, the many actions and documents that are within "the legislative process itself" and that the common-law privilege therefore traditionally protects. We next hold that the legislators here did not waive the privilege by communicating with individuals who are outside the Legislature. We conclude by explaining why the privilege does not yield in this case.

A

State lawmakers can invoke legislative privilege to protect actions that occurred within "the sphere of legitimate legislative activity" or within "the regular course of the legislative process." "[T]he privilege is not limited to the casting of a vote on a resolution or bill; it covers all aspects of the legislative process." As part of that process, lawmakers routinely "[m]eet[ ] with persons outside the legislature—such as executive officers, partisans, political interest groups, or constituents—to discuss issues that bear on potential legislation." "Consequently, some communications with third parties, such as private communications with advocacy groups, are protected by legislative privilege . . . ." These cases teach, and we agree, that the legislative privilege's scope is necessarily broad.

While our analysis begins with the privilege's scope, it does not end there. Records are not protected from production just because they are within the privilege's scope. Instead, like other privileges, the legislative privilege is "qualified" by exceptions that serve "the normally predominant principle of utilizing all rational means for ascertaining the truth."

The district court properly concluded that the documents at issue "are subject to legislative privilege." Likewise, on appeal, the parties agree that the legislators have properly invoked the privilege. That is, they agree that the documents fall within the privilege's scope. For their part, the legislators rely on the privilege for each of the disputed documents. Plaintiffs, too, do not argue that the documents are non-legislative. Instead, they argue only that the privilege either "was waived" or "must yield."

B

The legislators did not waive the legislative privilege when they "communicated with parties outside the legislature, such as party leaders and lobbyists." The district court's contrary holding flouts the rule that the privilege covers "legislators' actions in the proposal, formulation, and passage of legislation." An exception for communications "outside the legislature" would swallow the rule almost whole, because "[m]eeting with 'interest' groups . . . is a part and parcel of the modern legislative procedures through which legislators receive information possibly bearing on the legislation they are to consider."

Even glimpsing at Plaintiffs' objections to the privilege log reveals as much. For example, Plaintiffs say that one of the legislators waived the privilege for hand-written notes that reveal his mental impressions—because the notes appear on a document that he received from a third party. Plaintiffs also say that the legislators waived the privilege for correspondence that they solicited from constituents, or for advice that they solicited from the office of the Secretary of State. A privilege that protected so little of the lawmaking process would not rightly be called "legislative." We see Plaintiffs' "waiver" argument not as an exception, but as an indirect attack on the privilege's scope. We reject that approach.

Plaintiffs' contrary arguments lack merit. For instance, Plaintiffs say that a non-waivable privilege would be anomalous. But this argument errs by assuming that the legislative privilege is non-waivable. On the contrary, "legislative privilege as to certain documents is waived when the Legislator publicly reveal[s] those documents." But that is not what happened here. The privilege log shows that the legislators did not send privileged documents to third parties outside the legislative process; instead they brought third parties into the process. That decision did not waive the privilege. The very fact that Plaintiffs need discovery to access these documents shows that they have not been shared publicly. On the other hand, if the legislators had shared the documents publicly, then they could not rely on the privilege to prevent Plaintiffs from introducing those documents as evidence. But here, where the documents have been shared with some third parties—but haven't been shared publicly—the waiver argument fails.

Plaintiffs also point us to district court opinions holding that the privilege's purpose is to buttress the separation of powers by protecting lawmakers from the other branches' threats of prosecution and conviction. Under this view, a federal court's meddling in the state legislature does not strictly violate the separation of powers, and therefore there is no cause for alarm. But check and balances are not the privilege's only purpose. It also serves the "public good" by allowing lawmakers to focus on their jobs rather than on motions practice in lawsuits.

Plaintiffs also criticize the legislators for drawing on caselaw involving either the Constitution's Speech or Debate Clause or legislative immunity (rather than legislative privilege). As for the first point, the legislative privilege that protects state lawmakers "is similar in origin and rationale to that accorded Congressmen under the Speech or Debate Clause." Even if the federal privilege yields to fewer exceptions than the state privilege, we see no reason to differentiate between state and federal lawmakers when determining what counts as "legitimate legislative activity." In other words, the legislative privilege's scope is similar for state and federal lawmakers—even if the privilege for state lawmakers has more exceptions. So too for legislative immunity, which the Supreme Court has often analyzed in parallel to legislative privilege. Both concepts involve the core question whether a lawmaker may "be made to answer—either in terms of questions or in terms of defending . . . from prosecution." While the parallel between them may not run to the horizon, we follow the Supreme Court's lead in drawing on both strands even though this case involves a privilege from disclosure rather than an immunity from suit or liability.

C

Nor is this one of those "extraordinary instances" in which the legislative privilege must "yield[ ]." The legislative privilege gives way "where important federal interests are at stake, as in the enforcement of federal criminal statutes." According to the Supreme Court, "in protecting the independence of state legislators, Tenney and subsequent cases on official immunity have drawn the line at civil actions." We have thus recognized that the legislative privilege "is qualified." While "important federal interests" may be at stake in criminal as well as "extraordinary" civil cases, the qualifications do not subsume the rule. The privilege would be of little value if legislators "could be subjected to the cost and inconvenience and distractions of a trial upon a conclusion of the pleader, or to the hazard of a judgment against them based upon a jury's speculation as to motives."

This holds true even when constitutional rights are at stake. "The claim of an unworthy purpose does not destroy the privilege." A court proceeding that probes legislators' subjective intent in the legislative process is a "deterrent[ ] to the uninhibited discharge of their legislative duty." The Supreme Court explained in Tenney that "[t]he reason for the privilege is clear." "In order to enable and encourage a representative of the public to discharge his public trust with firmness and success, it is indispensably necessary[ ] that he should enjoy the fullest liberty of speech, and that he should be protected from the resentment of every one, however powerful, to whom the exercise of that liberty may occasion offense." "Regardless of the level of government, the exercise of legislative discretion should not be inhibited by judicial interference . . . ." "[I]t [i]s not consonant with our scheme of government for a court to inquire into the motives of legislators," and courts are not to facilitate an expedition seeking to uncover a legislator's subjective intent in drafting, supporting, or opposing proposed or enacted legislation.

Even for allegations involving racial animus or retaliation for the exercise of First Amendment rights, the Supreme Court has held that the legislative privilege stands fast. In Bogan v. Scott-Harris, a city employee sued the mayor, a city council member, and other city officials under 42 U.S.C. § 1983, contending that they eliminated her position out of "racial animus and a desire to retaliate against her for exercising her First Amendment rights in filing the complaint against [a colleague]." The city employee won a partial jury verdict. The First Circuit affirmed the judgment against the Mayor and the city council member. But the Supreme Court reversed. It held that the city officials were absolutely immune from suit under § 1983, because "the common law accord[s] [them] the same absolute immunity it accord[s] legislators at other levels of government." In discussing "absolute legislative immunity," the Supreme Court also noted that it had previously extended that immunity to an interstate regional planning agency because of the purposes of legislative immunity and the importance of such immunity in advancing the "public good."

Further, the Ninth Circuit has held that "plaintiffs are generally barred from deposing local legislators, even in 'extraordinary circumstances.' " In Lee v. City of Los Angeles, the plaintiffs contended "that race was in fact the overriding motivation behind [the redrawing of a city's voting district] boundaries." The plaintiffs sought to depose local officials, but the Ninth Circuit held that the legislative privilege applied. That court "recognize[d] that claims of racial gerrymandering involve serious allegations," and also that the "[d]efendants ha[d] been accused of violating that important constitutional right," because " '[a]t the heart of the Constitution's guarantee of equal protection lies the simple command that the Government must treat citizens as individuals, not as simply components of a racial . . . class.' " Still, the Ninth Circuit declined to recognize an "exception whenever a constitutional claim directly implicates the government's intent," because "that exception would render the privilege 'of little value.' "

Plaintiffs' attempt to require state legislators to produce documents is far closer on the continuum of legislative immunity and privilege to the suits under 42 U.S.C. § 1983 at issue in Tenney and Bogan than it is to the criminal prosecution under federal law at issue in Gillock. In Tenney, there was an "important federal interest[ ]" at stake: the vindication of civil rights under § 1983. The issue in Tenney was "whether state legislators were immune from civil suits for alleged violations of civil rights under 42 U.S.C. § 1983." But despite the important federal rights that § 1983 aims to vindicate, the Supreme Court explained that the legislative privilege did not yield to those interests. "We cannot believe that Congress—itself a staunch advocate of legislative freedom—would impinge on a tradition so well grounded in history and reason by covert inclusion in the general language of § 1983." The Supreme Court held that "a state legislator's common-law absolute immunity from civil suit survived the passage of the Civil Rights Act of 1871." By the same token, a state legislator's common-law absolute immunity from civil actions precludes the compelled discovery of documents pertaining to the state legislative process that Plaintiffs seek here.

Finally, Plaintiffs' reliance on Jefferson Community Health Care Centers, Inc. v. Jefferson Parish Government is misplaced. That decision stated that "[w]hile the common-law legislative immunity for state legislators is absolute, the legislative privilege for state lawmakers is, at best, one which is qualified." But that case provides no support for the idea that state legislators can be compelled to produce documents concerning the legislative process and a legislator's subjective thoughts and motives. There, a Louisiana parish sought to evict a health care center from property that the parish owned. We rejected the city council members' arguments for a legislative privilege that would have barred the court from even deciding whether to issue injunctive relief. We said only that "[a]t any rate, even assuming that the councilmembers' reasons for passing the resolutions are privileged in the sense that they cannot be directly compelled to disclose them, this evidentiary privilege cannot bar the adjudication of a claim." We held only that a claim for injunctive relief could proceed. That holding says nothing about cases like this one.

We REVERSE.