As many Verdict readers know, the Supreme Court this term will hear arguments in Moore v. Harper, a congressional-districting dispute from North Carolina that raises the so-called “Independent State Legislature (ISL)” theory. The theory is being invoked by Republican elected legislators in North Carolina to challenge the rulings of the North Carolina Supreme Court; that court had invalidated the state legislature’s congressional districting as violative of the state constitution’s ban on excessive partisan gerrymandering, and (invoking its powers to remedy state constitutional violations) replaced the legislature’s handiwork with a districting plan that complied with state-constitutional limits.
As I have explained in great detail elsewhere (most recently on this site here), ISL theory contends that under Article I of the U.S. Constitution (and also, even less plausibly, under Article II, governing presidential elections) each state’s ordinary elected legislature enjoys a federal entitlement to have its enactments concerning federal-election logistics take full effect notwithstanding anything in the state constitution that creates and bounds the legislature itself. Building from the words in Article I that say congressional election logistics “shall be prescribed in each State by the Legislature thereof,” the logic of ISL would deny the ability of state constitutions to define each state legislature’s contours and procedures, and to identify enforceable substantive limits it must respect.
ISL theory fails for many, many reasons, but here are three big ones: (1) As an originalist matter, the historical and ideological backdrop against which Articles I and II of the federal Constitution were adopted made clear that state legislative enactments were inferior to state constitutions, and state courts had the power and duty to exercise judicial review to enforce state constitutions; (2) relatedly, and again as an originalist matter, “legislature” in Articles I and II was understood, both in the period right before and right after federal constitutional ratification, to mean “lawmaking system” rather than any specific entity or body of persons; and (3) the Court has repeatedly and recently repudiated ISL in four key cases, one of them decided unanimously (on this point) in 2019. (For more background on these, and many other, reasons why ISL has no merit, readers can consult a recent law review article I co-wrote with my brother Professor Akhil Amar that was published four months ago. Readers should also be on the lookout for an amicus brief Akhil and I will be soon be filing that includes additional compelling material that goes well beyond the law review article and all my columns on this site.)
Enter Professors Will Baude and Michael McConnell (B/M), both accomplished constitutional law scholars, who wrote an essay in The Atlantic last week seeking to offer what they call a “commonsense middle ground” between the ISL proponents and ISL critics (like me and Akhil.)
Like ISL’s critics, B/M reject the broad premise of ISL that ordinary legislatures are immune from state constitutional regulation and state judicial review, because ISL:
ignore[s] the fact the state legislature is created and governed by the people of the state through their state constitutions. Even in matters of election law. . . there is no reason to think the Framers of the federal Constitution intended to liberate state legislatures from the ordinary constraints of state constitutional law.
But B/M also reject the broad rejection of ISL because they believe folks like Akhil and I:
would rewrite the Constitution to replace the term legislature with the term state. Given the specific references to legislature, executive, judges and conventions in various parts of the Constitution, it is not plausible to think that these institutional designations can be disregarded as meaningless. Rather, the Framers of the Constitution appear to have believed that a power so central to democratic governance should be vested in the branch of the government most representative of the people. [This reasoning tracks the language and reasoning of Justice Gorsuch when he wrote in a shadow-docket case in 2020 from Wisconsin that “The Constitution provides that state legislatures—not federal judges, not state judges, not state governors and other state officials—bear primary responsibility for setting election rules.”]
The B/M solution? “A state constitution may limit a legislature’s power over federal elections, but it may not give that power to somebody else.”
They elaborate:
Because state legislatures derive their lawmaking power from their own people, their authority is limited to what their state constitution gives them. When the federal Constitution gave state legislatures additional authority, it took them as it found them, as created by state constitutions rather than a new free-floating entity. State legislatures are not independent of their constitutions.
But the claim that state courts may hold state legislatures to state constitutional limits does not mean . . . [state courts] can replace the legislature. The federal Constitution’s text explicitly empowers one of these branches to regulate federal elections, not the other.
The upshot for B/M is that while a state court can invalidate statutes regulating federal elections that violate the state constitution and remand back to the legislature, it cannot engage in fashioning the regulation itself.
The B/M position, although it is not styled as such, is essentially a reprisal of the position Chief Justice Roberts (CJR) took in his dissent in the seminal 2015 Arizona Legislature v. Arizona Independent Redistricting Commission (AIRC) case—upholding a voter-initiated independent commission to draw congressional districts in Arizona—in which he tried to distinguish between state devices that supplement the ordinary state legislature and those that supplant it.
I very much appreciate the instinct B/M (and Chief Justice Roberts) may have to seek a middle ground in a very divisive legal area in a very divided country. (Indeed, I have elsewhere suggested that the Court in Moore could affirm the North Carolina Supreme Court on the ground that the state legislature affirmatively enlisted state courts in enforcing state constitutional provisions in federal elections, in which case radical ISL arguments need not be considered.) But middle-ground solutions (like all other solutions) have to be principled and workable, and on examination the B/M stance—like the CJR dissent in the AIRC case—turns out to be neither.
First, the premise of B/M is misplaced; they argue that ISL cannot be completely rejected because to do so would be to ignore the text of the Constitution. To repeat, they say words like “legislature,” “executive” and “judges” were presumably used intentionally, and to distinguish each one from the other. And sometimes this may be so (as I have argued in the context of the Seventeenth Amendment.) But it is not always the case—sometimes, as in Article I, “legislature” does not refer to an entity, but is a shorthand for a system of lawmaking that is defined by, and evolves under, a state constitution.
A close analogy may help illustrate the folly of B/M’s textual argument. The word “Congress” appears in the federal Constitution over 60 times. Context makes clear that the word often describes the House and Senate, but not the President (as when the Constitution discusses the “sessions” of Congress, or the President’s provision of information to Congress). But more often, “Congress” means House and Senate acting with the President via lawmaking. And this is true whether or not the document specifies that “Congress” shall act “by law.” Indeed, in many important contexts, “Congress” is empowered without the “by law” qualifier, and we still understand the reference to “Congress” to mean not Congress the entity (House and Senate) but the federal lawmaking system (Congress plus the President.) For example, Article II says “[t]he Congress may determine the Time of chusing the Electors,” and shortly afterward says “the Congress may by Law provide for the Case of Removal, Death, Resignation or Inability, both of the President and Vice President” (emphasis added). And yet no one argues that as to selecting the time of picking presidential electors, Congress can act without presidential involvement. Does that make “meaningless” (to use B/M’s word) the term “by law” wherever it is in fact used? No, it just illustrates that context, structure and history must inform the meaning of each term in various places of the Constitution. And just as context, structure and history make clear that “Congress” (even without the “by law” qualifier) is almost always best read to mean a lawmaking system rather than a particular institution, so too for the Constitution’s various references to a given state’s “legislature.” In context, this word generally (as in Articles I and II) means a state’s lawmaking system—as the Supreme Court has repeatedly held in a century-old line of cases from Ohio ex rel Davis v. Hildebrant (in 1916), to Smiley v. Holm (in 1932) to AIRC (2015).
And for those wondering why “legislature” rather than the more capacious “state” might have been used in Article I, here is one explanation. “Legislature” offered a comforting textual continuity with the Articles of Confederation whose language precisely mirrored that of Article I of the Constitution in this regard and whose interpretive track record by the states during the period leading up to the Constitution made clear that “legislature” did not refer to a particular body but rather to the state’s lawmaking system. Relatedly, the word “legislature” also essentially referred to each state’s particular extant off-the-shelf lawmaking apparatus. This was particularly important because use of “legislature” facilitated the upcoming 1789 federal election (one of the Founders’ many goals) and avoided needless wrangling by removing one source of uncertainty; in some states, executives and judges whom no one considered part of the lawmaking system nonetheless participated in various state appointments and elections. Article I made clear they would not participate in the first federal elections.
That “legislature” referred to the entire extant lawmaking apparatus created by the state constitution rather than the set of elected representatives who met in session in the state capital is nicely illustrated by the fact that two states in 1787—Massachusetts and New York—made use of governors and judges (via a gubernatorial veto and a veto by a Council of Revision made up of executive officials and judges, respectively) in their lawmaking systems. Crucially, both of those states right after 1787 regulated federal elections with the involvement of these governors and judges, even though, per ISL and B/M, these officers necessarily fall outside the Article I “legislature.”
And to add even more insult to textual injury, if “legislature” is read flatfootedly à la B/M and the North Carolina Republicans, then mere supplementation of ordinary modern bicameral “legislatures” would violate the (flatfootedly-defined) “legislature’s” power to “prescribe.” Why? Because “prescribe” means “dictate,” “ordain,” “call the shots.” As a textual matter, it does not mean “merely be involved in” or “assent to.” B/M (and CJR) have offered no explanation for why judicial remand in the name of the state constitution (which they would permit), is consistent with the legislature’s power to prescribe. After all, under the “legislature as entity” reading, “legislatures” are (according to B/M) different from “judges.”
And also different from governors. B/M say “legislature” and “executive” are different things, but then have no explanation for why gubernatorial vetoes can be part of the process, not just in Massachusetts in 1787, but for all states in the twentieth century (as the Court unanimously held in Smiley and as has been uncontroversial everywhere for a century). Thus, not only would the 2015 Arizona case fall under principled application of the B/M wooden textual approach, but, and quite radically, so too would Smiley and Hildebrandt (which involved popular referendum, a device that falls outside the narrow “legislature as entity” concept). The middle ground looks not so inviting after all.
Moving from text to logic, if a state constitution can leave districting to the elected legislature but prescribe every jot and tittle of the criteria and process the legislature must use—B/M concede that state legislatures’ “authority is limited to what their state constitution[s] gives them,” and offer no limits on the limits state constitutions can impose, much less any principled limits on the limits—the state constitution must also be allowed to generously empower other institutions for districting purposes, whether it be independent commissions or the courts themselves. Is there any difference between a court drawing maps, on the one hand, and the court saying to the legislature, “there is only one detailed kind of map that can be drawn consistent with the state constitution, and so, legislature, here it is—you enact it.”
That brings me to another key point, the fact that B/M’s suggestion that courts can never draw district lines simply proves too much. Courts provide remedies all the time (when they strike down parts of statutes) that result in the implementation of regulations that never passed the legislature in the precise form in which these regulations become operative. That is a generic aspect of constitutional remedies—the law that gets enforced after the exercise of judicial review is often something the legislature never actually adopted (but instead something the judiciary reasonably believes the legislature would have adopted had it known it could not enforce that which it did adopt.) I would think that B/M agree that judicial review, practiced by both federal and state courts, is perfectly in keeping with republican-government democratic principles.
If state courts are impermissibly usurping legislative powers whenever they redo districts that run afoul of state constitutional constraints, then they are impermissibly usurping legislative powers in many other settings too. Indeed, under the B/M approach federal courts would be violating Article I when they redo districts that run afoul of federal constitutional constraints. (And a rejoinder here that the federal Constitution trumps state laws by virtue of the Supremacy Clause of Article VI does not do the trick, since a state constitution also trumps state laws under the Supremacy Clause’s ordering of the various kinds of laws, as Chief Justice Marshall himself observed in the closing passage of Marbury, where he invoked the Supremacy Clause and observed that a “principle . . . essential to all written Constitutions, [is] that a law repugnant to the Constitution is void, and that courts . . . are bound by that instrument.”)
Finally, the aversion to a large role for state courts here that B/M display (an aversion understandable given the way we think of federal courts), simply isn’t fair to state courts and the role they occupy in our constitutional system. From the 1700s to the present moment, the relationship between state legislatures and state courts has been very different from the relationship between Congress and federal courts.
At the Founding, various state judges had powerful legislative roles. New York’s judges, sitting in a Council of Revision with the state executive, had a veto over ordinary legislation. Also, state legislators at times had judicial roles; influenced the British House of Lords, some states at the Founding, as the Court noted in 1798 in Calder v. Bull, vested judicial duties in the state legislature’s upper chamber or the legislature as a whole. The super-strict distinction that ISLers (and B/M to a certain extent as well) rely upon—a sharp delineation between state legislatures and state courts—simply did not exist.
Thus, the U.S. Constitution does not generally prevent a state from giving lawmaking and adjudicative power to the same body. In some Revolution-era colonies and states, the senate in effect doubled as the supreme court. Going forward, the Constitution allowed a state to have two supreme courts or two legislatures. (Today, Texas has two supreme courts and many states split legislative power between an ordinary legislature and a special initiative/referendum process.) The Constitution also allows a state to make its supreme “court” its supreme “legislature” or vice versa, as the Court’s members recognized long ago in Calder.
Even now, state courts often perform functions akin to those discharged by ordinary legislatures. As Justice Louis Brandeis explained in one of the twentieth century’s most iconic cases, Erie Railroad Co. v. Tompkins, “whether the law of the State shall be declared by its legislature in a statute or by its highest court in a decision is not a matter of federal concern.” This is why B/M’s suggestion that the (often gerrymandered) elected legislature is inherently the “branch of government most representative of the people” doesn’t go very far.
Perhaps because of one or more of these compelling reasons, CJR himself has backed away from his AIRC dissent (a dissent apparently influenced by a fake historical claim regarding a bogus so-called Pinckney Plan). In the Court’s recent Rucho v. Common Cause ruling, the Court doubled down on AIRC in an opinion authored by the Chief Justice himself and joined by both of the remaining AIRC dissenters (Justices Clarence Thomas and Samuel Alito), plus Justices Neil Gorsuch and Brett Kavanaugh.
Let me double-back to what I said earlier—I greatly appreciate B/M’s instinct to help identify middle ground. But as Yogi Berra is reported to have said (and this Yogi-ism actually makes sense in certain contexts), “when you come to a fork in the road, take it!” In Moore v. Harper, the Court can’t, in any kind of remotely coherent way, split the ISL baby. Either “legislature” means a collection of certain persons (somehow magically and ahistorically defined other than by each state’s constitution) or it means “lawmaking system as defined and updated in each state’s constitution.” I think the latter is clearly the right path, and the one the Court need take if it rules on the merits of ISL.