In this Comment, I examine the “underwrite,” a concept first developed by Professors Ethan Leib and James Brudney in their article, “Legislative Underwrites.” There are plenty of law review articles discussing “overrides” (when Congress rejects a court’s statutory interpretation decision). Theirs is the first article to explore the “underwrite” (when Congress approves a statutory interpretation). This comment explores the concept as it relates to overrides and calls for a major empirical study linking the two. Rejecting Leib and Brudney’s focus on interbranch cooperation, I present a different positive political story of the underwrite based on endogenous legislative incentives. No judge should worry, in my opinion, about Congress sitting around willy-nilly underwriting the latest Supreme Court case. Underwrites must compete with far more salient policy disputes important to members’ constituents. Finally, this Comment argues that textual underwrites deserve super-charged deference but raises questions about whether courts will be willing to embrace such a rule.
Professors Ethan Leib and James Brudney have given us an intellectual treat. There are plenty of law review articles discussing “overrides” (when Congress rejects a court’s statutory interpretation decision). Theirs is the first article to explore the “underwrite” (when Congress approves a statutory interpretation). Bravo for their discovery. The “Leib–Brudney underwrite” has been born, and, for reasons I explain, it is likely to produce a good deal of important scholarship.
Although one of the most fascinating and important parts of the Article involves interbranch dialogue in the states, I focus on the federal level. And like the authors, I concentrate on judicial statutory interpretation even if, as they recognize, underwriting of administrative agency decisions is an important topic for future inquiry. Part I explains why the underwrite concept may be more important or prevalent than Leib and Brudney suggest, based on the ties between overrides and underwrites. It is in Part II that I part company with the authors by offering a different causal theory of underwrites. No one should read Leib and Brudney and imagine that members of Congress are sitting around trying to decide whether to be helpful to the courts by underwriting the latest Supreme Court case. At the federal level, given our unique system of separated powers, underwrites are unlikely to reflect conscious or even unconscious desire for interbranch cooperation; they only exist if they serve Congress’s interests. In Part III, I tease out some problems of application that the authors themselves acknowledge, particularly when it comes to legislative evidence¾even the evidence that Leib and Brudney consider a gold standard. In Part IV, I conclude by arguing that textual underwrites may deserve judicial super-deference, albeit for reasons somewhat different from those advanced by Leib and Brudney. Nevertheless, courts will have to be convinced.
II. The Connections Between Overrides and Underwrites
Congressional override of statutory-interpretation decisions has been the subject of many fine empirical studies from the leading lights of the academy and their critics. Professor Bill Eskridge has done two comprehensive studies showing the rise and ebb of overrides over time. Professor Rick Hasen has countered with his own study, rejecting some of Eskridge’s conclusions. Political scientists have weighed in on the question as well. Not until now, however, have legal scholars seriously attempted to study the fact that Congress supports, affirms, and underwrites judicial statutory decisions. Having seen this happen during my tenure as a Senate lawyer, it does not surprise me. What surprises me is that it took so long for legal academics (including myself) to notice. Such significant discovery in the academy is rare; this discovery delights.
To the casual reader, the underwrite might seem the opposite of the override, but that conclusion is deceptive. In the override, Congress rejects the Supreme Court’s interpretation; in the underwrite, Congress accepts the Supreme Court’s interpretation.  The distinction merits emphasis here because it shows the importance of their study and provides guidance for future empirical work. Consider the following scenario: In 1982, there exists Interpretation A. Ten years later in 1992, the Supreme Court substitutes Interpretation B. Twenty years later in 2012, the Congress rejects the Supreme Court’s Interpretation B because it wants to reinstate Interpretation A. If this is correct, the 2012 Congress has simultaneously overridden Interpretation B and underwritten Interpretation A. This is significant for the status of Interpretation A. Once a mere creature of the judicial branch, it is now a rule endorsed by Congress.
Consider one of Leib and Brudney’s examples of this interaction between overrides and underwrites, the Civil Rights Act of 1991. That legislation included a number of overrides. In one instance, Congress sought to overrule Patterson v. McLean Credit Union. In Patterson, the Supreme Court held that a post–Civil War statute, 42 U.S.C. § 1981,did not protect an employee from racial harassment by her employer, since the harassment occurred after the formation of the contract. Congress balked. By 1991, President George H. W. Bush agreed that Patterson should be overruled. In overriding Patterson, however, Congress retur- ned to preexisting law. The Patterson legislative “fix” reaffirmed a well-known earlier precedent, Runyon v. McCrary, which allowed African-Americans to sue to integrate private schools under Section 1981. In Runyon, the Court rejected the argument made by the schools that the plaintiffs had to show “state action” because Section 1981 did not reach private acts of discrimination. The statutory override of Patterson un- derwrote Runyon. The text of the resulting statute affirmed the principle that racial discrimination by “nongovernmental parties” is barred by 42 U.S.C. § 1981.
The bottom line: overrides are not necessarily the opposite of underwrites in practice. They may go hand in hand, revealing “recovered precedents.” In the example above, Runyon is the recovered precedent, underwritten in the override of Patterson. I highlight this simple relationship for two reasons. First, Leib and Brudney’s study warrants a full-blown empirical study of underwrites. If one is going to do such a study, one should start by looking at overrides. To be sure, there may be “stand-alone” underwrites, but my suspicion is that the conjoined override–underwrite pair may well predominate. Second, the potential link between overrides and underwrites reinforces the importance of Leib and Brudney’s study. In many cases, scholarship has focused on overrides (the 1991 Civil Rights Act is a prime example), ignoring the underwriting aspects of the legislation. In some cases, these underwrites will be codified, as in the Patterson–Runyon case. But in other cases, the recovered precedent may be less obvious. That, in turn, raises theoretical questions about the scope of underwrites, since every override leaves the court back at the prior status quo. Leib and Brudney carefully limit their definition of an underwrite to those instances in which Congress has specifically embraced the decision in the text of the adopted statute or in its legislative history, but one wonders why that is always necessary, or whether one will find that there is a hierarchy of underwrites: super-underwrites (in the text of the statute), normal underwrites (in the legislative history), and the simple recovered precedent. Presumably, if Congress is willing to override a decision, it must be aware of the recovered law. Even if that law does not merit the super-deference of explicit underwrites, perhaps it deserves more weight than the average precedent. A future empirical study of the underwrite should take all these things into account.
III. Underwrites: A Different Causal Story
When and why does Congress underwrite? Leib and Brudney spend a good deal of time analyzing the virtues and vices of underwriting as an institutional matter. They note opportunity costs¾legislators might well be doing something better with their time¾and the potential for both increased legal uncertainty and policy conflict. Put another way, the fear is that there will be too much underwriting. On the other hand, they explain that underwriting provides legal clarity, promotes interbranch cooperation, and reduces reading tea leaves from congressional inaction. They imagine that if underwriting becomes a more regular feature of lawmaking, it may be used more often. In other words, there exists a contrasting fear of too little underwriting.
All of this may lead the reader to imagine members of Congress sitting around trying to decide whether they should underwrite Supreme Court decisions. Such a vision could cause a good many judges heartburn. One imagines the judicial mind roiled with the angry thoughts that hapless members of Congress have deemed themselves super-judges. That view is wrong, but it is important to see why. Put bluntly, Congress is not a court. The incentives of members of Congress are structured by institutional realities that are entirely different from judicial incentives. Congress is a busy, sometimes chaotic institution. Its members focus on how to solve national crises, not on how to draft a legal decision, particularly a legal decision no voter has ever heard of. This explains why there are plenty of legal issues that Congress might¾and even should¾resolve, but that Congress entirely ignores, to the chagrin of law professors and judges: the status of administrative guidance, crazy scrivener’s errors in evidentiary rules, and problems with the habeas corpus statute, to name a few. Why no congressional action? Because no one votes on those issues. Congress spends its limited time on matters important to members of Congress and their constituents. That is, after all, what the Constitution directs members of Congress to do: represent the people.
All theories provide implicit causal stories. The reader might think Congress can on a whim, or at a moment’s notice, decide to underwrite any particular Supreme Court or other judicial decision. I doubt Leib and Brudney would subscribe to that causal claim. As they explain, they originally believed that the underwrite was a rather “impractical way for a legislature to spend its time.” The authors’ cost–benefit analysis, however, appears to take an exogenous view¾from outside the legislative process¾to look at something that, in my opinion, is dominated by that process. They are writing as legal scholars, carefully attempting to tie the virtues and vices of their discovery to an abstract idea of interbranch cooperation. But this exogenous approach may lead to causal misunderstandings.
Underwrites are an endogenous phenomenon: they exist within and because of institutional processes and realities. Lawmaking is difficult, time-consuming, and full of partisan bickering. In such a context, there must be very strong incentives for Congress to underwrite. Where do these incentives come from? The need for action  and the “electoral connection.” First, the need for action: in areas of complex law, Congress may choose to borrow precedents or language¾using the putatively “neutral” language of the Supreme Court¾to effect compromise and thus achieve action in an institution where action is exceedingly difficult. For example, when Senator A and Senator B cannot agree upon a standard for voting-rights cases, they pull language from a Supreme Court case, using that case as a neutral arbiter that allows the Senators to agree upon statutory text (effectively punting the case back to the Supreme Court). Even if that produces congressional agreement, however, Professor Mayhew’s “electoral connection” thesis tells us that agreement must be at least neutral if not positive for members’ electoral interests—otherwise it will not happen.
If this causal story is correct, certain theoretical “costs” and “benefits” of underwrites should not worry us. For one, we should not be terribly concerned about “too many” underwrites; members of Congress simply do not have the time or inclination. Unless the underwrite is worth more than competing legislative agenda items, the underwrite will not happen. Members of Congress do not spend their days poring over SCOTUSblog wondering what kinds of cases they can underwrite. Nor should we worry about “too few” underwrites, because if underwriting helps to achieve compromise on a politically salient bill, underwrites will happen whether we like them or not. Consider the 1991 Civil Rights Act: liberal interest groups and voters were incensed at Supreme Court interpretations which they viewed as anti-African-American and anti-woman. Congress responded. It is a crude heuristic, but well established within the political science literature, that Congress, as a whole, and individual legislators act according to electoral incentives. In this case, those electoral incentives¾the voters¾led them to override and underwrite.
Positive political theory tells us that underwrites will exist when, to achieve action on an issue of political salience, the President and the Congress prefer the underwritten precedent to a newer precedent as a policy matter  By “policy” matter, I mean here what these political actors predict to be the preferences of their constituents.
As we know, courts, over the long haul, tend to be majoritarian institutions. When courts deviate from that principle in ways that are politically salient, then majoritarian institutions will respond. They will respond with overrides and, potentially, underwrites.
One might argue that I have called for much further work on the underwrite only to suggest that underwrites are contingent on political incentives. The two positions are not inconsistent. The first suggests that we should take a new look at the override literature because it is likely to reveal additional underwrites, and that we should consider more thoroughly the role of recovered precedents. The second suggests that both overrides and underwrites will depend upon Congress’s political incentives. For those who worry about Congress sitting as a Supreme Court, do not fear. In their minds, members of Congress have better things to do¾their job, responding to voters, or at least what they think the voters want.
IV. Underwrites: Definitional Issues
What exactly is an underwrite? Any future empirical work must grapple with serious definitional issues, as Leib and Brudney recognize. No one is likely to quibble with the proposition that an underwrite specifically mentioned in statutory text qualifies, but this apparent agreement may well dissolve in practice: even statutory underwrites can raise questions of scope and application. More obviously, many will reject any claim that underwrites in legislative history (what I shall hereafter call “legislative evidence”) should be considered at all. Any future empirical study will have to address these problems of scope and evidence.
A. Questions of scope: facts versus principle
Leib and Brudney acknowledge that identifying an underwrite may lead to new problems. Assume that we identify the 1991 Civil Rights Act as underwriting the principle that the law bars private as well as governmental discrimination. Let us assume everyone agrees that the amended statutory text supports that proposition and amounts to an underwrite of Runyon v. McCrary. Questions may still arise about the scope of the underwrite.
In fact, this is precisely what has happened with respect to Section 1981’s application to nongovernmental actors. Section 1981 protects contracting, but it also protects against racially biased deprivations of the “equal benefit of all laws.” The Circuits have split on whether the “equal benefit” clause applies in the absence of state action. One could argue that the Runyon underwrite should resolve this controversy. In 1991, Congress explicitly rejected Patterson and underwrote Runyon in Subsection (c) of the statute: “The rights protected by this section are protected against impairment by nongovernmental discrimination and impairment under color of State law.” As a matter of textual analysis, the ordinary reader would presume that “rights protected by this section” refers to the substantive guarantees in Subsection (a). But, if there were doubt, the underwrite should reinforce the claim that Congress really meant it when it said private actions count.
Critics, however, will argue that the underwrite should be limited to its facts: both Runyon and Patterson involved classic contract situations, while the “equal benefit” cases involve different factual scenarios. The application question becomes: does the underwrite apply to the general concept (non-state action) or is it limited to the particular facts (contract claims)? Of course, this standard problem in legal interpretation does nothing to undermine the underwrite concept, but it may provide critics with a reason to cry foul. A situation like the conflict above invites more research: what did Congress say about its underwrite? That means legislative evidence. And if underwrites propel interpreters into the legislative history debate, many will balk.
B. The legislative evidence problem
Leib and Brudney know that using legislative evidence in statutory interpretation is controversial. I am an arch defender. But I do not necessarily believe that committee reports are the “gold standard,” even if courts sometimes say that. Given the way that statutes are constructed over time, it is entirely possible that a committee report on Senate Bill A and House Bill A will be completely irrelevant by the time the final legislation has passed. Leib and Brudney make a passing reference to underwrites “in the air.” The following is the kind of problem they may have envisioned.
Consider this scenario, increasingly true as the complexity of legislation grows: In considering a lengthy bill, a House committee decides to underwrite Case A about Title 13 of the Bankruptcy Code. On the Senate side, there is a similar desire. The committee reports are crystal clear that Case A should govern for Title 13 cases. Fast forward to bill debate. Before the bill is introduced in the Senate, the bill’s proponents will need 60 votes. To get those votes, proponents compromise significantly on the bill’s language, taking any language relating to Title 13 off the agenda and out of the bill. The resulting bill deals only with Titles 7 and 11, and has no text to which the underwrite relates at all.
Now imagine a litigant arrives in court with the committee reports in a Title 13 case. Should the court believe that Congress underwrote Case A? Congress agreed to take Title 13 off the table. Legislative evidence takes its power from the text to which it relates, and if it relates to no text, then critics are likely to balk. One of the most insistent critiques of legislative evidence is that it comes from a few committee members, not the whole of Congress. This critique is wrong, in my view, and deliberately overstated: committee reports gain institutional legitimacy from the rules Congress creates to delegate authority to subgroups. Just as any corporation delegates its work to agents, so does Congress. The problem in this scenario is that there is no final product to which to tie the relevant agent’s decision. If a corporation left out any consideration of Title 13 in its 10-K, the internal documents on Title 13 are not relevant to interpreting the 10-K. The same rules should apply to Congress. Free-floating committee underwrites untied to actual legislation may bear some weight in difficult cases, but not the super-majoritarian imprimatur one might hope to give full-blown statutory underwrites.
V. Should courts pay attention to underwrites?
Having sung the praises of underwrites, I will end by saying I do not believe they should be praised for the reasons Leib and Brudney suggest¾cooperation between the departments. Based on my work on the separation of powers, my own view of our constitutional structure envisages the departments in competition for political and bureaucratic power. Siloed departments seek to advantage their own constituencies and policy space. They typically cooperate only when doing so enables them to further these interests. If this is correct, underwrites are not to be valued any more than overrides because of an imagined conversation between the branches. The point is that even without any intent to cooperate, there has been agreement between the judiciary and the Congress and that agreement merits super-deference. Underwrites gain power because they unite the Congress and the Court on a single interpretation, increasing the democratic legitimacy of interpretive rulings. Courts should give textual underwrites super-deferential weight.
When Congress and the Court agree upon a legal principle, it should have enormous weight for reasons grounded in democratic representation. Imagine that we redefine the departments not in terms of adjectives¾executive, legislative, or judicial¾but in terms of their constituencies. As a relative matter, the President speaks to the nation more than the representative from the Fifth District of Texas, who will care relatively more about the voters in that district. If this is correct, even if the departments do not intend to cooperate, or are hostile to each other (because, for example, they have different dominant party affiliations), rules that have the support of a broad cross section of constituencies will be more robust and stable. The most stable rules are ones accepted by the President, the House, the Senate, and the courts. Underwrites can, in theory, represent the position of each relevant constituency¾the nation, states and localities¾plus the (more abstract) constituency of principle, the courts.
Brudney and Leib are careful to explain that different kinds of underwrites may deserve different kinds of judicial deference. The separation-of-powers analysis I have just elaborated shows, at a minimum, why a textual underwrite, like the Runyon underwrite in our Section 1981 case, deserves judicial super-deference. It is unclear, however, whether courts will be ready to accept the invitation. As I mentioned earlier, the departments compete as much as cooperate. Courts remind us regularly that they “say what the law is,” even as they know that Congress and the President say precisely the same thing. A court whose membership has shifted in policy orientation relative to a recovered precedent may balk. If, for example, the court believes we are in a post-racial society, will it have any interest in rediscovering Runyon’s earlier attempts to rectify the deep racial inequality of segregated schools? That textual underwrites may deserve super-deference does not mean that courts will, in the end, grant it.
 Ethan J. Leib & James J. Brudney, Legislative Underwrites, 103 Va. L. Rev. 1487, 1511 (2017).
 Matthew R. Christiansen & William N. Eskridge, Jr., Congressional Overrides of Supreme Court Statutory Interpretation Decisions, 1967–2011, 92 Tex. L. Rev. 1317 (2014); William N. Eskridge, Jr., Overriding Supreme Court Statutory Interpretation Decisions, 101 Yale L.J. 331 (1991).
 James Buatti & Richard L. Hasen, Response: Conscious Congressional Overriding of the Supreme Court, Gridlock, and Partisan Politics, 93 Tex. L. Rev. See Also 263 (2015).
 See, e.g., Beth Henschen, Statutory Interpretations of the Supreme Court: Congressional Response, 11 Am. Pol. Q. 441 (1983); Pablo T. Spiller & Emerson H. Tiller, Invitations to Override: Congressional Reversals of Supreme Court Decisions, 16 Int’l Rev. L. & Econ. 503 (1996).
 Leib & Brudney, supra note 1, at 1520.
 Pub. L. No. 102-166, 105 Stat. 1071.
 See Christiansen & Eskridge, supra note 2, at 1492–93 (listing twelve Supreme Court decisions, nine of which were decided between 1986 and 1991, that were overridden by the 1991 Civil Rights Act).
 491 U.S. 164 (1989).
 Originally passed as § 1 of the Civil Rights Act of 1866 § 1981(a) reads as follows:
All persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts, to sue, be parties, give evidence, and to the full and equal benefit of all laws and proceedings for the security of persons and property as is enjoyed by white citizens, and shall be subject to like punishment, pains, penalties, taxes, licenses, and exactions of every kind, and to no other.
42 U.S.C. § 1981(a) (2012).
 Patterson, 491 U.S. at 179.
 See Steven A. Holmes, Critics of Rights Law Fear A Flood of Suits Over Jobs, N.Y. Times, May 27, 1990, at 8 (Bush administration supports reversal of Patterson v. McLean Credit Union).
 427 U.S. 160, 170 (1976).
 Id. at 173–74.
 In 1991, Congress added the following two subsections to 42 U.S.C. § 1981. Note that section (c) explicitly underwrites the idea that section 1981 plaintiffs can sue private actors:
(b) For purposes of this section, the term “make and enforce contracts” includes the making, performance, modification, and termination of contracts, and the enjoyment of all benefits, privileges, terms, and conditions of the contractual relationship.
(c) The rights protected by this section are protected against impairment by nongovernmental discrimination and impairment under color of State law.
42 U.S.C. § 1981 (2012) (emphasis added).
 Leib & Brudney, supra note 1, at 1520.
 Id. at 1522–26.
 I have written an entire book to prove this rather obvious point. Victoria Nourse, Misreading Law, Misreading Democracy (2016).
 One of the thorniest issues in administrative law is the status of policymaking documents, typically called “guidance,” which are deemed nonlegislative rules and thus do not require notice and comment rulemaking. See, e.g., Texas v. United States, 86 F.Supp.3d 591, 671 (S.D. Tex. 2015) (striking down Obama DAPA guidance because it was a legislative rule that had to go through notice and comment rulemaking), aff’d, 809 F.3d 134 (5th Cir. 2015), aff’d by an equally divided court, United States v. Texas, 136 S.Ct. 2271 (2016) (per curiam); John Manning, Nonlegislative Rules, 72 Geo. Wash. L. Rev. 893 (2004) (highlighting judicial manageability concerns associated with nonlegislative rules).
 Green v. Bock Laundry, 490 U.S. 504, 511 (1989) (“[A]s far as civil trials are concerned, Rule 609(a)(1) ‘can’t mean what it says.’” (citation omitted)).
 On the multiple problems with the statute, passed in the crisis atmosphere of the Oklahoma City Bombing, see Carlos M. Vázquez, Habeas as Forum Allocation: A New Synthesis, 71 U. Miami L. Rev. 645 (2017); Larry W. Yackle, State Convicts and Federal Courts: Reopening the Habeas Corpus Debate, 91 Cornell L. Rev. 541 (2006).
 Nourse, supra note 17, at 21.
 Leib & Brudney, supra note 1, at 1514.
 See Victoria F. Nourse & Jane S. Schacter, The Politics of Legislative Drafting: A Congressional Case Study, 77 N.Y.U. L. Rev. 575, 594–95 (2002).
 The “electoral connection” is David Mayhew’s fine term. See David R. Mayhew, Congress: The Electoral Connection (1974).
 See Chisom v. Roemer, 501 U.S. 380, 398 (1991) (explaining that the 1982 amendments to the Voting Rights Act borrowed the phrase “to participate in the political process and to elect representatives of their choice” from a Supreme Court case).
 See Ronald D. Rotunda, Civil Rights Act of 1991: A Brief Introductory Analysis of the Congressional Response to Judicial Interpretation, 68 Notre Dame L. Rev. 923, 924–26 (1993).
 See generally Mayhew, supra note 24, at 11–78.
 On such modeling, see Daniel B. Rodriguez & Barry R. Weingast, The Positive Political Theory of Legislative History: New Perspectives on the 1964 Civil Rights Act and Its Interpretation, U. Pa. L. Rev. 1417, 1431–37 (2003).
 See Barry Friedman, The Will of the People 14–16 (2009) (discussing the influence of the popular will on the Supreme Court).
 See 42 U.S.C. § 1981(a) (2012).
 The Eighth and Third Circuits have held that for a plaintiff to state a claim under the “equal benefit” clause, the plaintiff must allege state action. See Elmore v. Harbor Freight Tools USA, Inc., 844 F.3d 764 (8th Cir. 2016); Bilello v. Kum & Go, LLC, 374 F.3d 656 (8th Cir. 2004); Youngblood v. Hy-Vee Food Stores, Inc., 266 F.3d 851 (8th Cir. 2001); Adams ex rel. Harris v. Boy Scouts of Am.-Chickasaw Council, 271 F.3d 769 (8th Cir. 2001); Brown v. Philip Morris Inc., 250 F.3d 789 (3d Cir. 2001). The Second and Sixth Circuits have held that such a plaintiff need not allege state action. See Chapman v. Higbee Co., 319 F.3d 825 (6th Cir. 2003) (en banc); Phillip v. Univ. of Rochester, 316 F.3d 291 (2d Cir. 2003).
 42 U.S.C. § 1981(c) (2012) (emphasis added).
 Leib & Brudney, supra note 1, at 1497.
 Id. at 1515.
 See, e.g., Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgement) (criticizing the use of committee reports).
 Victoria Nourse, The Vertical Separation of Powers, 49 Duke L.J. 749, 751–52 (1999); Victoria F. Nourse, Toward a New Constitutional Anatomy, 56 Stan. L. Rev. 835, 850–52 (2004).