McCulloch v. Marbury

34 Const. Comment. 263
PDF

Some are born great, some achieve greatness, and some have greatness thrust upon them.

—William Shakespeare, Twelfth Night, II.V

Introduction: Marbury’s Greatness

Marbury v. Madison is a great case.[1] That much is undeniable. It is “widely regarded today as the most important case in American constitutional history.”[2] Some enthusiasts go farther; in 1901, an Arkansas judge proclaimed it “the most important event in our history” apart from the formation of the government.[3] While the Supreme Court has never gone that far, it tends to invoke Marbury in high-profile and contentious cases: in the hands of the Justices, Marbury is a confidence-booster, an important support for the exercise of judicial review.[4] If there is one proposition that constitutional scholars and judges of all stripes can agree on, it is that “Marbury’s place in the constitutional canon is secure.”[5]

But why? How did Marbury become great? What does it stand for? And—perhaps most important—does it deserve its lofty standing?

Earlier articles have explored some of these questions. There has been substantial recent scholarship on the process by which Marbury became a great case and the construction of its meaning—the proposition it stands for in contemporary legal discourse.[6] We build on that research to create a platform from which to explore the last question. In line with Voltaire’s suggestion that if God did not exist, it would be necessary to invent him, it turns out that Marbury became great because it was needed. People in the legal community—judges, scholars, and advocates—felt that a great case was needed to promote aggressive judicial review. And so they made one.

The use of Marbury as support for judicial review—and more particularly, as support for judicial invalidation of government acts—did not follow immediately or ineluctably from the decision itself. Instead, as we describe in Part I, it appears to have begun during the Lochner era, as advocates of aggressive judicial review of economic regulation looked for support for their preferred stance. After the Lochner jurisprudence went down to defeat in the New Deal era, Marbury too diminished in prominence. It returned in the Warren Court era, beginning with Cooper v. Aaron and continuing thereafter. In some cases, we find it deployed by opponents of judicial invalidation, or by both sides of the decision. Everyone agrees that Marbury is important; everyone wants to have Marbury on their side—even if they cannot agree on what it means.

This much has been said before. The novel move we make is to offer a normative assessment of the canonization of Marbury. Our argument is that Marbury cannot sustain the burden its role demands. The early ascension of Marbury came as part of an era of judicial review now roundly rejected by the Supreme Court and commentators alike.[7] Marbury has come back, miraculously untainted by its association with Lochner.[8] But, as we argue in Part II, Marbury cannot escape Lochner, for Lochner (in its symbolic version) is simply the dark side of Marbury, no more separable than its shadow.[9]

Marbury does enunciate or endorse (some form of) judicial review. But it has nothing to say about the crucial question of how judicial review should be exercised: when it should be aggressive and when it should be deferential.[10] (This opacity or lack of content is why the Lochner Court, the Warren Court, and the Rehnquist/Roberts Court can all invoke Marbury for support, even though they have very different theories about the proper exercise of judicial review.) Indeed, if the question is about the proper level of deference rather than the existence or non-existence of judicial review, Marbury is not a great decision at all. As we argue in Part III, it is a terrible one, worse even than Lochner.[11]

Where does that leave us? It would be nice if there were a canonical early decision that enunciates the principle of judicial review and at the same time says something about when it should be aggressive and when it should be deferential. It would be nice if that opinion were written by the great Chief Justice John Marshall. And for teaching purposes—since the constitutional canon is constructed in large part through the selection of cases for casebooks—it would be nice if the decision illustrated both aggressive and deferential review.

Fortunately, such a decision exists. As we explain in Part IV, it is McCulloch v. Maryland.[12] The last part of the Article examines McCulloch, explaining why it succeeds where Marbury fails. McCulloch, we conclude, should replace Marbury in our constitutional canon as the symbol of judicial review. Whether this will occur in judicial opinions is, of course, up to judges, and we concede that law review articles may have limited impact on judicial practice. (It does, however, appear that law professors and legal activists are responsible for creating the Marbury to which judges now appeal.[13]) Thus we have an additional suggestion, which law professors can implement on their own. McCulloch should replace Marbury as the first case that students encounter in their constitutional law courses.

As a starting point, we turn now to the story of Marbury. How did it become the decision it is today?

  1. Marbury’s Rise

In its fledging years, Marshall’s now-famous opinion was rarely cited in cases or treatises. This is likely because judicial review, the principle that eventually elevated Marbury to greatness, was widely accepted and commonly used in the years prior to Marbury.[14] That is not to say there was no debate over the operation of judicial review and its scope – there was. But Marbury actually says very little about those issues.[15] The issue for which Marbury has become famous—whether the power of judicial review exists at all—had been largely settled. The suggestion, occasionally seen in celebrations of the opinion, that Marbury created judicial review or was the first instance of its exercise is simply wrong.[16] More striking, the arguments that Marbury marshals in favor of judicial review were not novel. Marshall’s argument—and even his rhetorical flourishes—borrow from Alexander Hamilton’s Federalist 78 to an extent that would support charges of academic dishonesty in an undergraduate term paper.[17]

Given that Marbury said nothing novel about judicial review, it is not surprising that up through the mid-nineteenth century the attention it garnered was for other aspects of its discussion. Marbury was cited largely for its analysis of the availability of mandamus and the scope of original jurisdiction.[18] When the Supreme Court eventually found another occasion to exercise judicial review to invalidate an act of Congress in the infamous Dred Scott case, it did so without mention of Marbury.[19] After a half-century drought, Dred Scott opened the floodgates of judicial review, as nineteen decisions invalidated acts of Congress between 1857 and 1893.[20] In none of those cases, however, did the Court cite Marbury. Presumably, doing so seemed neither necessary nor valuable to the opinion writers. When, then, did Marbury take the stage as support for the exercise of judicial review?

Surprisingly, the answer originates in another infamous case, which former Supreme Court Chief Justice William Howard Taft believed did more to damage the Court’s prestige than any other, Pollock v. Farmers’ Loan & Trust Co. (1895).[21] Pollock marked the Supreme Court’s first invocation of Marbury when using judicial review to invalidate a federal statute.[22] The majority opinion in Pollock, a 5-4 decision, elicited strong rebukes from the dissenting Justices,[23] incited controversy and led to both Democrats and Populists adopting anti-Court and anti-Marbury rhetoric in the elections of 1896.[24] Ultimately, Pollock was reversed by a constitutional amendment (the sixteenth).[25]

Despite the reaction, opinion-writers apparently absorbed the lesson that Marbury helped: after Pollock, invocations of Marbury increased notably.[26] While Pollock’s opponents “regarded Marbury as a “usurpation” of legislative authority, and judicial review as an unwarranted power,” its defenders hailed Marbury as fundamental to American liberty.[27] By the early twentieth century, the conservative faction, including much of the elite American bar, orchestrated the canonization of Marbury and deification of its author, John Marshall, whose 1901 commemoration featured several of the Pollock majority Supreme Court Justices.[28] In lockstep, throughout the early 1900s, constitutional treatises, once nearly devoid of Marbury’s judicial review holding, began expounding Marbury’s primacy.[29] As noted by Davidson M. Douglas, “Constitutional law treatises published after 1900 bore a very different quality with respect to judicial review and the importance of Marbury in comparison with their nineteenth-century predecessors. Almost without exception, the status of Marbury is significantly elevated. Most early twentieth-century treatises devoted a separate section to a discussion of the case.”[30]

By the 1920s, nearly every treatise exalted (or inflated) Marbury’s status, as it became the leading case for learning about the Court’s role in overturning unconstitutional laws.[31] Indeed, in some Constitutional texts, Marbury made up the entire discussion of judicial review.[32]

The same pattern exists within constitutional law casebooks. As late as 1894, prominent casebooks’ discussions of judicial review either omitted Marbury or included it as simply one of several sources without attaching special significance to it.[33] But beginning in the twentieth century, constitutional law casebooks recognized and facilitated Marbury’s ascendancy, often singling it out as the case on judicial review, or marking it as first among equals.[34] Although there remained a bloc of intransigent scholars refusing to acquiesce to Marbury’s newfound role (and arguing against judicial review itself), [35] the Supreme Court itself took the pro-Marbury side.

In the first decade of the twentieth century, the Court cited Marbury three times to justify judicial power to overrule laws — more than in the entire nineteenth century.[36] By 1911, a unanimous Court was invoking Marbury as foundational: “When may this court, in the exercise of the judicial power, pass upon the constitutional validity of an act of Congress? That question has been settled from the early history of the court, the leading case on the subject being Marbury v. Madison.”[37]

In 1926, Chief Justice Taft, that bitter critic of the Pollock decision, accepted the prevailing wisdom as to Marbury. It was a “great constitutional authority,” Taft pronounced, “one of the great landmarks in the history and construction of the Constitution of the United States, and … of supreme authority … in respect to the power and duty of the Supreme Court and other courts to consider and pass upon the validity of acts of Congress enacted in violation of the limitations of the Constitution.”[38] Taft’s embrace of Marbury marks its successful separation from the wreckage of Pollock: though Taft presumably still thought Marbury had been misused in the past, he recognized its utility for the future.

There was one more snare for Marbury to avoid. Courts and commentators had, in effect, created Marbury (at least, the Marbury we know) in an effort to support aggressive Lochner-era judicial review. That jurisprudence did not get reversed by constitutional amendment, as Pollock had, but it did go down to defeat. Yet again, Marbury survived. It was not invoked as frequently during the New Deal era—unsurprisingly, since the New Deal Court was quite deferential to the federal government—but the case never became a symbol of judicial overreach.[39] Indeed, when Marbury came back, it came back stronger.

Taft had called the opinion itself the supreme authority on judicial review. In 1958, in Cooper v. Aaron, it became associated with judicial supremacy: a unanimous Court described Marbury as “declar[ing] the basic principle that the federal judiciary is supreme in the exposition of the law of the Constitution ….”[40] (One searches Marbury in vain for this proposition. It may well be the province of the judicial department to say what the law is, but that simply tells us what courts do. It does not tell us what effect the views of other branches of government should have on those judicial statements or what effect the judicial statements should have on the views of other branches.[41]) While the invocation of Marbury had already (somehow) begun to support not just the exercise of judicial review but the conclusion that a particular law was unconstitutional, “Cooper v. Aaron transformed Marbury into the modern symbol of judicial power, and elevated a different dimension of John Marshall’s argument into the standard judicial and legal rhetoric of the late twentieth century.”[42]

After Cooper, as after Pollock, academic critiques of Marbury became more common. Some targeted the case’s poor sequencing, reasoning, and result-oriented nature, while others targeted its subsequent outsized influence and popular misinterpretation – many did both.[43] However, the academic criticism did not dent Marbury’s popularity. If anything, it fed the fascination. Marbury’s popularity grew, along with what Jack Rakove has called the “emphasis on [its] talismanic power.”[44]

Cooper—likely because of its own prominence, controversy, and boldness—established Marbury as the Supreme Court’s go-to citation when it felt a need to buttress its authority. In particular, it is Cooper that first focused on Marbury’s statement that “[i]t is emphatically the province and duty of the judicial department to say what the law is.”[45] After Cooper, the percentage of the Court’s citations to Marbury that refer to it for the power of judicial review more than doubled.[46] This uptick created a self-reinforcing pattern: the more the Court invoked Marbury, the more powerful it became. Yet at the same time, it became increasingly devoid of content. If there is a proposition for which Marbury stands in this era of post-Cooper Supreme Court opinion writing, it seems to be little more than “we are right to invalidate this law.” [47]

That pattern has (largely) continued from the Warren Court to the Rehnquist and Roberts Courts. No matter which Court is invalidating a government act, no matter which constitutional provision is at stake, no matter the ideological valence of the opinion, Marbury can be invoked in support of invalidation.[48] Because Marbury has become both inviolable (everyone agrees it is a great decision) and largely empty of content (the Supreme Court seldom even attempts to investigate what it might mean, beyond “we are right to invalidate this law”), some enterprising Justices have even begun to deploy it against invalidation.

Perhaps no case better reveals Marbury’s role in contemporary jurisprudence than U.S. v. Windsor. [49] Windsor invalidated the federal Defense of Marriage Act, which provided that the federal government would not recognize same-sex marriages valid under state law. In Windsor, two separate opinions cite Marbury. First, the Kennedy majority invokes Marbury for the standard purpose of judicial self-aggrandizement, warning that failing to decide the constitutional question presented would undermine Marbury’s holding that “[i]t is emphatically the province and duty of the judicial department to say what the law is.”[50] Justice Scalia, in dissent, faults the majority for its “exalted conception of the role of [the Supreme Court] in America” and goes on to suggest that Marbury supports judicial review only in the presence of actual controversies.[51] Justice Roberts made a similar move in King v. Burwell, arguing that Marbury demonstrates that the power of courts is merely to interpret (“to say what the law is”) rather than to make law.[52]

What, then, is Marbury today? It is a canonical decision, that much is clear.[53] It stands, almost invariably, for expansive judicial authority. In particular, the Supreme Court tends to use it to justify invalidation of government acts—even though the existence of judicial review is a question quite distinct from its proper exercise or the correct outcome in a particular case. Because Marbury gets cited for a proposition (“we are right to invalidate this law”) about which it says nothing, it is hardly surprising that Justices have started trying to press it into service to oppose invalidation as well.

But they don’t need to. There is already a canonical citation for that proposition. Judicial review has a dark side. And, we will see, Marbury does too.

  1. Marbury’s Shadow

The existence of judicial review is now accepted; indeed, as we have suggested, it was accepted before Marbury. And most scholars (and unsurprisingly, almost all judges) now agree that judicial review is a good thing. But the institution is different from the exercise: while the institution of judicial review is good, particular exercises of judicial review may be good or bad. And while there is general agreement that courts should invalidate laws that violate the Constitution, there is also agreement that they should not invalidate laws simply because the Justices disagree with the policies behind them. The legal world associates the first type of invalidation with a heroic Supreme Court defending American values against majoritarian oppression and invokes it with a citation to Marbury. We associate the second type of invalidation with an activist Court imposing its will on the American people, and we invoke it with a citation to a different case: Lochner v. New York.[54] Where Marbury is deployed as rhetorical shorthand for the argument that judicial invalidation is good and necessary to our system of government, Lochner is deployed as rhetorical shorthand for the opposite point: judicial invalidation is suspect, activist, and corrosive of our constitutional democracy.

Lochner’s history is in many ways similar to that of Marbury. Understood at first as a case about the scope of state police power (which is what it was actually about[55]), Lochner was gradually transformed into a symbol for something different—judicial activism, or the imposition of judicial policy preferences under the guise of constitutional interpretation.

Lochner, of course, invalidated a New York statute that prohibited the employment of bakeshop workers for more than sixty hours per week. It was not, at the time, widely reviled: it was a notable decision with both supporters and detractors, but there was little indication that it would become the anathema it is today.[56] The dispute between the majority and the main dissent (written by Justice Harlan and joined by Justices White and Day) centered on the factual question of whether bakery employment was unhealthy—not a particularly dramatic question, nor one raising issues of appropriate judicial behavior.[57]

Nor was Lochner considered especially pathbreaking or influential in its own time: prior to West Coast Hotel v. Parrish[58] (the case generally considered to have overruled Lochner), the Supreme Court cited Lochner in only eleven cases.[59] For Lochner’s first decade, the few cases that did cite it did so in connection with other cases, specifically Adair and Allgeyer (which were both mentioned far more regularly in this period) or in opinions by Justice Holmes, author of the now famous dissent in Lochner, or both. In fact, as early as 1917 “Lochner seemed to be dead and buried for good,”[60] with little in the way of fanfare, epitaph, or continuing relevance. It had a brief resurgence in the 1920s, but with the triumph of the New Deal and the acquiescence of the Supreme Court, Lochner’s willingness to scrutinize “ordinary commercial transactions” went into eclipse.[61]

Lochner was rejected by 1937, but many rejected cases are simply forgotten: they do not live on as symbols or join the anti-canon.[62] How did Lochner acquire its anti-canonical status? The answer is much the same as it was for Marbury: opponents of aggressive judicial review needed a symbol, so they made one. They had some help: Justice Holmes, rather than addressing the case within the accepted doctrinal framework of state police power analysis, boldly denounced the majority as motivated by an enthusiasm for laissez-faire economics: “This case is decided upon an economic theory which a large part of the country does not entertain.”[63]

As the Great Depression gripped the country, both intellectuals and the general public turned against laissez-faire economics, and Lochner became a convenient symbol of that. In addition to the Holmes dissent, there was the 1937 publication of Felix Frankfurter’s The Commerce Clause Under Marshall, Taney, and Waite. That work advanced the general progressive narrative that, by deciding cases based on economic preferences rather than law, due process jurisprudence in cases like Lochner had usurped the authority of the people to govern themselves.

Frankfurter was a friend and acolyte of Holmes, and it is not surprising that the first explicit Supreme Court invocation of Lochner as a symbol of inappropriate judicial behavior is a dissent by Justice Frankfurter, in Winters v. New York.[64] Frankfurter did this again, in a concurrence in American Federation of Labor v. American Sash & Door Co., where he characterized the Lochner era as one in which “economic views of confined validity were treated by lawyers and judges as though the Framers had enshrined them in the Constitution” and invoked the Holmes dissent (which he now called a “famous protest”).[65]

Just as conservative forces promoted Marshall and Marbury at the turn of the century, Frankfurter championed Justice Holmes and his Lochner dissent, in the process turning Lochner into a symbol of activism.[66] Similar invocations followed: in 1952, for instance, a majority opinion by Justice Douglas cited Lochner as one of a number of cases in which the Court improperly sat “as a super-legislature to weigh the wisdom of legislation [or] to decide whether the policy which it expresses offends the public welfare.”[67] By the early 1960s, the understanding of Lochner as second-guessing legislative policy decisions was well established.[68]

But to make Lochner truly anti-canonical, it would take a high-profile and controversial case. Much as Cooper elevated Marbury, Griswold v. Connecticut[69] raised Lochner’s profile. Lochner appears in three of the opinions written in Griswold. First, in a move that would become common, the Douglas majority tries to buttress its legitimacy by disavowing Lochner. “Overtones of some arguments suggest that Lochner v. State of New York, should be our guide,” Douglas announces. “But we decline that invitation […]. We do not sit as a super-legislature to determine the wisdom, need, and propriety of laws that touch economic problems, business affairs, or social conditions.”[70] Second, in another move that would be repeated, the Black dissent accuses Justices White and Goldberg of adopting the “long-discredited” Lochner methodology and argues that if they had their way they “would reinstate Lochner.”[71] Third, the Stewart dissent levies similar charges against Justices Harlan and White, again invoking Lochner by name.[72] The idea of “the Lochner era” as a period of judicial misbehavior emerged by 1970; in 1977, the Supreme Court used the phrase for the first time. [73]

When another high-profile Due Process case came along, Lochner again took center stage. In Roe v. Wade, Justice Blackmun began with the ritualistic denunciation of Lochner and then proceeded to candidly balance the competing interests to determine a wise solution to the abortion dilemma.[74] For the majority, Lochner had become simply a symbol of bad judicial behavior, drained of any substantive content.

Other opinions in Roe, with somewhat more justification, described the majority as repeating the sin of Lochner: usurping the authority of the legislature to make policy choices.[75] Lochner thereafter was linked to Roe, at least in the minds of opponents of Roe. As John Hart Ely wrote in 1973:

The Court continues to disavow the philosophy of Lochner. Yet as   Justice Stewart’s concurrence admits, it is impossible candidly to        regard Roe as the product of anything else. That alone should be enough to damn it. Criticism of the Lochner philosophy has been      virtually universal and will not be rehearsed here. I would, however,   like to suggest briefly that although Lochner and Roe are twins to be       sure, they are not identical. While I would hesitate to argue that one    is more defensible than the other in terms of judicial style, there are    differences in that regard that suggest Roe may turn out to be the more dangerous precedent. [76]

Robert Bork put it more succinctly: “Who says Roe must say Lochner . . . .”[77]

Linking Lochner and Griswold, and, similarly, linking Lochner and Roe, makes some obvious sense. All three are opinions applying the Due Process Clause. (One might, as some do, add in Dred Scott.[78]) However, as we will suggest, the “substantive due process” of the Lochner era was in fact quite different from the doctrine that emerged in the 1960s. More important for Lochner’s fate, though, the invocation of Lochner did not stop with substantive due process. After Roe, Justices began to invoke Lochner as a bogeyman with increasing regularity and in a broader range of circumstances. By 1987, Cass Sunstein could proclaim that “for more than a half-century, the most important of all defining cases has been Lochner v. New York.”[79]

Since everyone agreed that Lochner was bad, it was available to everyone as a device to condemn opponents. Justices supportive of national power used “Lochner as an epithet” when disagreeing with new federalist opinions.[80] On the other side, conservative judges invoked Lochner when decrying the recognition and protection of unenumerated rights.[81] Almost everyone, as David Bernstein put it, “condemn[s] Lochner for improper ‘judicial activism.’”[82]

This understanding has shifted a bit in the academy in recent decades.[83] But the Supreme Court holds firm to its Lochner-as-activism model. In Obergefell v. Hodges, the landmark case requiring all states to grant and recognize same-sex marriages, Chief Justice Roberts’ dissent cited Lochner sixteen times, arguing that the majority repeats its sins. (For good measure, Roberts also associated the majority with Dred Scott.[84]) What are those sins? “[C]onfus[ing] our own preferences with the requirements of the law,” Roberts explains, “unprincipled … judicial policymaking,” “elevat[ing] their own policy judgments to the status of constitutionally protected ‘liberty,’”[85] “converting personal preferences into constitutional mandates,”[86] acting on “naked policy preferences,”[87] … you get the idea. Lochner stands for activism, yes, but anything a commentator dislikes can become activist. In the massive Obamacare decision, NFIB v. Sebelius, a concurrence/dissent in part criticizes other opinions (including that of Chief Justice Roberts) by saying that they “bear a disquieting resemblance to … long-overruled decisions” such as Lochner.[88]

Marbury is universally considered good; Lochner is universally considered bad. (At least, this is true as a matter of judicial rhetoric. In the scholarly literature, there are valuable reappraisals of both decisions.[89] But in judicial opinions, it is still more or less the case that Marbury means “we are right to strike down this law,” while Lochner means “you are wrong to strike down this law.”) In that sense, they could hardly be more different. But while they do represent opposite poles of a phenomenon, it is the same phenomenon. Lochner is the anti-judicial review version of Marbury, created and anti-canonized for exactly the same reasons on the opposite side of the debate. Both are citations that have virtually no content: Marbury means that judicial invalidation is good and appropriate, Lochner that it is bad and inappropriate. But what determines whether a particular exercise of judicial review should be associated with Marbury or with Lochner? How can those who want to don the mantle of Marbury escape the taint of Lochner?

The cases themselves, as popularly understood, offer no answer.[90] Delving a little deeper into the cases, which we will do in the next Part, offers us some insight—but not an answer. What it does, in fact, is to show that neither Marbury nor Lochner—neither the stylized versions invoked in judicial decisions nor the actual decisions as revealed and analyzed by scholars—can help distinguish good judicial review from bad. Lochner, from this perspective, is simply Marbury’s shadow: the dark version that follows and cannot be escaped. But the point can be made more strongly still. If we look closely at the cases in terms of appropriate judicial behavior, Marbury is actually far worse than Lochner. It is Marbury that should be the epithet.

III. Escaping Lochner

The stylized versions of Lochner and Marbury found in judicial opinions are not helpful guides. One tells us that invalidation is good, the other that invalidation is bad, but neither tells us how to distinguish the good from the bad. (Lochner as epithet suggests something bad—decisions based on policy preferences rather than the Constitution—but of course the other side will claim it is following Marbury, and invocations of Lochner never tell us how to detect the forbidden judicial policymaking.)
But distinguishing good from bad exercises of judicial review is precisely what we need to do.

No matter their political persuasion, legal thinkers tend to favor aggressive judicial review in some kinds of cases and deferential review in others. For instance, liberals tend to like judicial invalidation of abortion restrictions, laws that burden racial minorities, and (some kinds of) restrictions on speech and religious exercise. Conservatives tend to like judicial invalidation of expansive federal legislation, laws that burden whites, and (some kinds of) restrictions on speech and free enterprise.[91]

What this means is that analysis of judicial review should proceed at the retail, rather than the wholesale level. It is productive to talk about individual exercises of judicial review: whether the level of deference is appropriate, whether the decision is correct. It is much less productive to talk about judicial review as an institution—the institution is obviously established beyond the possibility of removal, and as the preceding paragraph noted, there is essentially no one who does not think it appropriate in some cases.

So the issue really comes down to particular decisions. Can we approach that issue fruitfully by starting with the general institution of judicial review? Not if we are simply asking whether judicial review should exist, which is the question Marbury addresses. Judicial opinions that cite Marbury as support for the invalidation of particular government acts are engaged in non sequitur or misdirection. The academic obsession with the countermajoritarian difficulty is another example of the error of conflating the institution with its exercise. Alexander Bickel, who deserves credit (or blame) for elevating the problem to the first rank of academic concerns, was thinking about a particular decision (Brown), not the whole range of judicial review.[92] But in that case, he should have asked how Brown could be justified in a democracy, not judicial review. (No answer, Bickel wrote in that book, is what the wrong question begets.[93]) Neither Brown nor Dred Scott, to take a quite different example, has much in common with a decision invalidating a federal statute that purported to eliminate Wyoming’s Senate representation. There is not much to say about judicial review that would be relevant to all of them.

John Hart Ely did a better job. He understood that a theory of judicial review can be useful if it seeks to explain not whether judicial review should exist but when deferential review is appropriate and when courts should be aggressive. Armed with a particular justification for judicial review—protecting the democratic process—he went on to develop a theory about when courts should be willing to intervene in normal politics and when they should stay their hands.[94] Ely found support for his approach in the Supreme Court’s decision in Carolene Products.[95] And that kind of analysis, we will suggest, points a way to dissolve the opposition between Lochner and Marbury—ultimately, to find a case that has something to say about both aggressive and deferential judicial review.[96] Identifying and analyzing that case is the task of Part V. Here, we set out the requirements for escaping Lochner and explain why Marbury fails to do so.

What was bad about Lochner—what makes it wrong? It was not just the exercise of judicial review; as citations to Marbury indicate, judicial review has a heroic side. The problem, at least according to contemporary judicial opinions, was that the Lochner Justices based their decision on their own notions of good policy rather than the Constitution, that they manipulated the legal materials presented to them in order to produce a decision motivated by other factors. To fall on the Marbury side of judicial review, then, a decision presumably must be based on a faithful application of the Constitution and be motivated solely by an attempt to reach the legally correct answer.

It turns out, however, that under these criteria, Lochner is the good decision and Marbury the bad one. We consider them in that order.

  1. Rehabilitating Lochner

There are many different ways to understand Lochner. Most critics, as noted above, understand it as the imposition of judicial policy preferences under the guise of constitutional interpretation. Others see it as an attempt to elevate freedom of contract to the status of a fundamental right, which the state can restrict only with an unusually persuasive justification.[97]

But understanding Lochner in terms of modern fundamental jurisprudence—in terms of rights that are protected by some form of heightened scrutiny—is a mistake. What the Court was actually doing, recent scholarship suggests, was patrolling the boundaries of state legislative authority: making sure that the government was keeping within the boundaries of its power.[98] (If you read the Lochner opinions—excepting the Holmes dissent—they make relatively clear that this is what was going on.[99])

To decide whether this is an appropriate job for judges under the federal constitution—an occasion to invoke the stylized Marbury rather than Lochner—we would need to answer three questions. First, is there a good argument that legislative acts going beyond the state’s police power violate the federal constitution, and in particular the due process clause? Second, are there limits to the police power that judges can discern? And last, do judges have sufficient advantages in discerning and enforcing those limits that aggressive judicial review is justified?

On the first question, the argument is quite straightforward—considerably more straightforward than the idea that the due process clause protects fundamental substantive rights, which John Hart Ely called an oxymoron akin to “green pastel redness.”[100] The starting point is the premise that an act that exceeds the powers of the legislature is not a law. This is a point commonly made in our constitutional jurisprudence.[101] It is, in fact, one of the justifications for judicial review given in Marbury.[102] Another classic early statement—which focuses more on the inherent limits of legislative power and provides an account of those limits—is given by Justice Chase in Calder v. Bull.[103] Since such an act is “no law,” an attempt to use it to restrain individuals, and a fortiori an attempt to punish them for violating it, deprives those individuals of liberty without due process of law.[104]

As to the second question, the argument is again relatively straightforward. Are there judicially-identifiable limits on the powers that the peoples of the states delegated to their governments? Justice Chase’s opinion in Calder v. Bull suggests so. “The purposes for which men enter into society will determine the nature and terms of the social compact; and as they are the foundation of the legislative power, they will decide what are the proper objects of it: The nature, and ends of legislative power will limit the exercise of it….”[105] If Calder v. Bull is not a sufficiently distinguished pedigree for the argument, it can be traced back further: this is the political theory of the Declaration of Independence.[106] And as the twentieth century dawned, this was the theory that courts used to hold that redistributive legislation exceeded the police power.[107]

The last question turns out to be the troubling one for Lochner. Are judges good at identifying the boundaries of the police power? It was generally accepted that legislatures could regulate to protect the public health, safety, and morals, and also to promote the public interest.[108] Cases about the first three of these tended not to be especially controversial. Holden v. Hardy, in which the Court upheld a maximum hours law for miners, is an example.[109] Judges might actually not be as good as legislatures at determining the factual question of whether some job is deleterious to the health, and so cases such as Holden display a fair degree of deference.[110]

Lochner does not. The majority and the dissent offer dueling authorities on the question of whether baking is unhealthy, but in the face of a conflict of experts, deference suggests that the legislature can take either side. The Lochner majority refused to defer—not so much because it believed itself better able than the New York legislature to assess the health impacts of baking, but because it suspected that there was an illegitimate purpose at work. “It is impossible for us to shut our eyes to the fact that many of the laws of this character, while passed under what is claimed to be the police power for the purpose of protecting the public health or welfare, are, in reality, passed from other motives.”[111]

What was the illegitimate goal? As Justice Harlan’s dissent conceded, “[i]t may be that the statute had its origin, in part, in the belief that employers and employees in such establishments were not upon an equal footing . . . .”[112] The Court believed that the New York legislature was trying to redress an imbalance in bargaining power. That, in the minds of the Lochner Court, was an impermissible purpose. A law that tried to redistribute in order to promote equality was “partial legislation,” akin to taking from one person and giving to another. That was not protecting natural rights but infringing on them. It was not something people would give their government the power to do—and so it was no law at all.

We can describe Lochner­-era substantive due process as follows. The Court was acting to keep the legislature within the bounds of the powers people might reasonably be thought to have delegated to it. In particular, it was acting to prevent unfair favoritism of the sort that people, behind a veil of ignorance, would not condone.

So stated, the Lochner jurisprudence is not unreasonable. It is certainly not based on simple judicial policy preferences. (Lochner says this explicitly, in a passage that critics seldom quote or take seriously: “This is not a question of substituting the judgment of the court for that of the legislature. If the act be within the power of the state it is valid, although the judgment of the court might be totally opposed to the enactment of such a law.”[113]) It does, however, require some other principles to be workable. And what happened to Lochner was that those necessary principles went away.

The most basic premise Lochner required was that redistribution was never in the public interest. This turns out to be false for two reasons. First, the alternative to redistributive government intervention might not be continuation of the status quo: it might be economic collapse. In such cases, redistribution does not simply make one group better off at the expense of another (the move Lochner forbids); it makes everyone better when compared to the consequences of inaction.[114] Second, through the process of repeated bargaining with shifting coalitions, a series of redistributive laws may in the end make everyone better off.[115]

Perhaps more seriously, the Lochner regime also required a neutral baseline from which courts could determine whether the government was engaging in impermissible redistribution. At the time of Lochner itself, the common law could supply that baseline, with courts examining statutes for impermissible state favoritism. But once the common law was understood as state law—as much the product of state action as a statute—its status as a neutral baseline collapsed.

Lochner-era jurisprudence could not survive the loss of its baseline and its anti-redistributive principle. Without those, it does indeed look like courts second-guessing legislative policy choices. But we should remember that some sort of policy analysis is involved in the decision of most constitutional questions—in applying the familiar tiers of scrutiny, for instance, courts must decide both how important some government objective is and how well the government act will serve that goal. Lochner, asking whether the statute bore a real and substantial relation to the protection of health, was not doing anything unusual by modern standards. And, perhaps more important, the Justices in the majority do seem to have been applying reasonable constitutional principles in good faith. The same cannot be said of John Marshall and Marbury.

  1. Reassessing Marbury

This Part reassesses Marbury, but not by pointing out anything novel.[116] The most amazing thing about Marbury in our contemporary legal culture is that its faults—well-known and serious—have done so little to dim its luster. Everyone knows the story we are about to tell—yet somehow it seems to have almost no impact. We are not sure why this is so, but our guess it that it is connected to two things: the lack of a ready substitute for Marbury that could do the same work of buttressing the exercise of judicial review, and perhaps a sneaking suspicion on the part of some who invoked Marbury that the decisions they were seeking to protect shared some its faults. We start with the story.

Marbury, everyone knows, had its origin in the bitter election of 1800.[117] Going in, the Federalist party held the Presidency (in the person of John Adams) and Congress. But they lost control of both branches; Thomas Jefferson won the Presidency and Federalist officeholders lost seats in Congress. Under the constitutional structure then in force, the electoral winners would not take office until March 4. And during the months between the election and the actual transfer of power, the Federalists did what they could to strengthen and secure their hold on the one branch of the federal government they would retain: the judiciary.[118] With the Judiciary Act of 1801, the lame-duck Federalist Congress shrank the size of the Supreme Court to five, in the hopes of denying Jefferson an appointment. The Act also created 16 new circuit judgeships. Last, and crucial to Marbury, it created new justice of the peace positions for the District of Columbia.

William Marbury was nominated to one of these positions by President Adams. He was confirmed by the still-Federalist Senate on March 3. Adams signed commissions, including Marbury’s, late into the night. John Marshall (who was still serving as Adams’ Secretary of State, despite having assumed the position of Chief Justice on February 1, 1801) affixed the great seal of the United States. His brother James delivered a number of the commissions—but not Marbury’s.

Taking office the next day, Jefferson discovered the signed and sealed commissions. He instructed James Madison, his Secretary of State, not to deliver some of them, including Marbury’s. Marbury filed suit in the Supreme Court, requesting that the Court issue a writ of mandamus to compel the delivery of the commission.

This put Chief Justice Marshall in a delicate position. If he issued the writ, Jefferson would likely ignore it. (He had already ordered Madison not to defend the suit.) Executive defiance of a Supreme Court order would set a troubling historical precedent, from which the fledgling Court might never recover. Yet acquiescing in Jefferson’s power play would also elevate the power of the Executive relative to the Court. How could he find a way out?

This is the way constitutional law professors generally set the stage for Marbury v. Madison, to the extent that they situate it historically at all.[119] Even this historical account is a little bit misleading, in that it omits the fate of the circuit judges and the Supreme Court. The circuit judges never took the bench; their offices were eliminated by the Republican Congress in the Judiciary Act of 1802. The Court acquiesced in that purge, so Marshall had already backed down on an issue of greater practical import than Marbury’s position as justice of the peace.[120] As to the Court itself, the Judiciary Act of 1802 undid the Federalist reduction in membership and restructured the Supreme Court’s sessions, effectively eliminating the 1802 Term entirely. Again, the Court accepted these moves.

These facts complicate the context of Marbury, but perhaps it all simply heightens the drama. And Marshall did find a clever solution. Marbury had a right to the commission, Marshall ruled.[121] Mandamus was an appropriate remedy.[122] But because the Constitution gave the Court appellate rather than original jurisdiction in this case, he could not issue it.[123]

And, of course, one more thing. Congress had attempted to give the Court original jurisdiction, Marshall wrote.[124] But because that statute conflicted with the Constitution, it was void and courts were authorized, indeed required, to treat it as a nullity.[125]

Adding in the fillip about judicial review was an especially nice touch. Discussing the merits allowed Marshall to rebuke Jefferson, which must have been satisfying but had little practical effect. Enunciating the doctrine of judicial review did: it strengthened the Supreme Court vis-à-vis the other branches of the federal government. And doing so in an opinion that ended by refusing to issue the writ of mandamus made it impossible for Jefferson to push back: there was no order to defy.

So Marbury is a clever opinion in terms of finding a resolution to Marshall’s dilemma. But is that the proper frame of analysis? Chief Justice Roberts’ much-mocked description of judges as umpires rather than players may be simplistic and naïve—because judicial decisionmaking is more complicated than calling balls and strikes—but it captures something about judicial impartiality that is indeed desirable.[126] Judges should try to find the legally correct answer to the question posed by a particular case. Cleverly manipulating legal materials to reach a result preferred for non-legal reasons is not good judicial behavior.

It is important, then, to consider the legal soundness of Marbury—not the argument for judicial review, which was largely unnecessary and also largely derivative of Hamilton—but the exercise of judicial review to invalidate part of the Judiciary Act of 1789.

Because these issues have been treated in detail by other scholars, we discuss them only briefly. Our aim, again, is not to disturb the received wisdom but to emphasize that most scholars believe that Marshall’s approach to the case was not simply that of an umpire seeking to determine the legally correct answer to a question. The question we are interested is why, given the dominant understanding, Marbury is considered a great case rather than an anti-canonical one. To set the stage for that analysis, though we must discuss the dominant understanding. We develop it by asking a simple question: what are the errors of Marbury?

First, it seems relatively plain that Marshall should not have participated in the case at all, given his personal involvement in the events giving rise to Marbury’s claim.[127] (Lest one think that these events were insignificant, the reported version of Marbury reveals that the Court heard testimony from various witnesses, including James Marshall, about the precise details of the signing and sealing of the commissions—that is, about what John Marshall himself had or had not done.[128]). Second (a relatively minor quibble), Marshall should have addressed the jurisdictional issue first.[129]

Third, more significant, Marshall’s reading of the Judiciary Act of 1789 is questionable. The relevant section, section 13, reads as follows:

And be it further enacted, That the Supreme Court shall have exclusive jurisdiction of all controversies of a civil nature, where a state is a party, except between a state and its citizens; and except also between a state and citizens of other states, or aliens, in which latter case it shall have original but not exclusive jurisdiction. And shall have exclusively all such jurisdiction of suits or proceedings against ambassadors, or other public ministers, or their domestics, or domestic servants, as a court of law can have or exercise consistently with the law of nations; and original, but not exclusive jurisdiction of all suits brought by ambassadors, or other public ministers, or in which a consul, or vice consul, shall be a party. And the trial of issues in fact in the Supreme Court, in all actions at law against citizens of the United States, shall be by jury. The Supreme Court shall also have appellate jurisdiction from the circuit courts and courts of the several states, in the cases herein after specially provided for; and shall have power to issue writs of prohibition to the district courts, when proceeding as courts of admiralty and maritime jurisdiction, and writs of mandamus, in cases warranted by the principles and usages of law, to any courts appointed, or persons holding office, under the authority of the United States.[130]

This section begins with four grants of jurisdiction—exclusive jurisdiction over certain suits involving states, original but not exclusive jurisdiction over others, exclusive jurisdiction over suits against certain foreign defendants, and original but not exclusive jurisdiction over certain other suits involving foreign defendants. Each of these uses the phrase “shall have … jurisdiction.” The next sentence discusses the trial of issues of fact. The section then switches back to a grant of appellate jurisdiction, then—in the same sentence—uses for the first time the phrase “shall have power.” It is this last phrase that gives the Court the power to issue writs of mandamus to federal officials. Is it, as Marshall claims, a grant of original jurisdiction over cases seeking mandamus against federal officials?

That seems like quite a stretch. For one thing, the section seems to distinguish quite clearly between jurisdiction (the authority to decide a case) and the power to issue particular writs (the authority to grant a particular remedy). Section 14 of the Act, notably, goes on to give federal courts the power (again, using the phrase “shall have power”) to issue other kinds of writs. These grants of power are clearly not grants of jurisdiction, since the grant also includes “all other writs … which may be necessary for the exercise of their respective jurisdictions”—a provision which would be nonsensical if simply requesting a writ created jurisdiction. For another, the grant of power to issue writs of mandamus to federal officeholders is the second part of a sentence that starts with a grant of appellate jurisdiction. It seems most natural to suppose that this second half of the sentence is a grant of power that can be exercised pursuant to the jurisdiction conferred in the first half. [131]

Marshall did not make up the idea that this section granted him jurisdiction over Marbury’s claim. Marbury’s lawyer, Charles Lee, addressed jurisdiction and mentioned section 13. But Lee’s main argument was that issuing a writ of mandamus to a federal officeholder was an exercise of appellate jurisdiction, just like issuing the writ to an inferior judge. Both, he claimed, involve “superintendence of the inferior tribunals and officers, whether judicial or ministerial. In this respect there is no difference between a judicial and a ministerial office.”[132] Lee references Section 13 of the Judiciary Act primarily as confirming this understanding.[133]

Supposing, however, that Congress did intend to grant the Court jurisdiction over suits seeking mandamus and that this jurisdiction counts as original rather than appellate under Article III, is that unconstitutional? Article III does put federal questions (the basis for Marbury’s suit) into the appellate category. But it further notes that the Court shall have appellate jurisdiction “with such Exceptions and under such Regulations as the Congress shall make.”[134] Charles Lee’s argument quotes this language, suggesting that “Congress is not restrained from conferring original jurisdiction in other cases than those mentioned in the constitution.”[135] The text is certainly not clear.

At this point—having encountered a constitutional question without a clear answer—we might pause to ask what a judge’s stance should be. Granted that judicial review will be exercised—a fact about which there is no dispute—should it be aggressive or deferential? Here we might consider various factors.[136] How much do we trust Congress? Are the members of Congress who enacted this bill pure and intelligent? Did they debate its provisions and consider constitutional objections? How well do we think they understood Article II? Is the power that Congress claims here susceptible to misuse? If it is, can we trust Congress to exercise it wisely and not abuse it? What harms will follow if the Court wrongly invalidates a permissible regulation of jurisdiction—and, on the other side, what harms follow if it wrongly allows an impermissible one to stand?

We do not purport to claim that these are all of the questions relevant to deference, nor do we claim to have an answer to even the questions raised here.[137] What we do claim is that when a constitutional question is not clear, the most important thing the Court must do is decide what degree of deference is owed the judgment of the other relevant government actors—the legislature that enacted a bill, the President who signed it into law. Which factors go into that decision is a question about which reasonable people can disagree. How the factors should be balanced against each other when they point in different directions is another such question. What we argue here is simply that those questions must be considered. Most of the tests that the Court has created over the years—and in particular the tiers of scrutiny—are best understood as based fundamentally on the decision to defer (rational basis) or to be aggressive (heightened scrutiny).[138]

Marbury does not consider these questions. Marbury makes no attempt at all to decide whether deference is appropriate. Instead, having manufactured a conflict,[139] Marshall resolves it in a brisk and superficial fashion: if the Constitution calls some cases original and others appellate, Congress can’t change that. Why not? Because allowing Congress to move cases between the two categories renders the constitutional division “mere surplusage.”[140] But that is clearly not true: Article III might set out an irreducible core of original jurisdiction to which Congress can add but not subtract, or a default rule might be provided simply for convenience.[141] In any event, this element of Marbury is not even an accurate statement of current law. While the Court has adhered to the idea that Congress may not grant original jurisdiction in cases where Article III confers appellate jurisdiction, it has allowed Congress to grant appellate jurisdiction over cases that Article III puts in the original category.[142]

What we have just given is the commonly accepted version of Marbury. Given that, how should the case be evaluated? Most commentators find a lot to praise. The opinion is a “political masterstroke,”[143] a “masterwork of indirection,”[144] a “virtuoso performance.”[145] The Supreme Court, of course, regularly cites Marbury as justification for its invalidation of state and federal laws, and Justice Breyer, in a recent book, called the opinion “brilliant,” a “judicial tour de force,” and “worthy of the Great Houdini.”[146]

This is high praise. Still, with the presumable exception of Supreme Court citations, it often sounds more like aesthetic judgment than endorsement of legal reasoning or judicial behavior. R. Kent Newmyer offers one of the most extreme examples, writing that “[l]ike a great work of art, Marbury yields different meanings to different viewers at different times—which may be the true mark of greatness.”[147] There’s a lot of truth there as far as what makes a work of art great, but it begs the question of whether judicial opinions should be judged by the same criteria. And more urgently, it does not explain why one of Marbury’s meanings—in fact, the dominant one we have described—is that this is how good judges behave.

That Marbury is considered good judicial behavior, especially but not exclusively by the Supreme Court, is relatively clear. The Court would hardly turn to an example of judicial misbehavior (like the conventional understanding of Lochner) to support its authority in controversial cases. And scholars do seem generally to share this view. Louise Weinberg, for instance, pronounces that “[i]n Marbury, a great father of our country bequeathed to us his greatest legacy and our most precious inheritance—the inestimable treasure of an enforceable Constitution.”[148]

But what is it that these commentators are praising? Marbury, as we have described it, and as it is commonly understood, is a decision that creatively manipulates legal materials to reach a decision deemed desirable for extra-legal reasons. On the conventional account, then, Marbury is activist in just the way Lochner is supposed to be: it seeks to strengthen the judiciary because that is considered good policy. One might defend Marbury by pointing out that the policy it advances (judicial power) is a desirable and accepted constitutional value while the one Lochner advances (laissez-faire economics, on the crude account, or a ban on redistribution for its own sake, on the more sophisticated one) is not. Viewed this way, both decisions are examples of what Jack Balkin and Sandy Levinson have called “high politics,” with the difference that Marbury’s preferred value has stood the test of time.[149]

But Marbury is not only, or even primarily, high politics. Marshall’s reasons are not just a desire to strengthen the federal judiciary (something that may strike most federal judges as anodyne[150]); they also include a desire to advance the fortunes of a particular political party: the Federalists. By favoring the judicial branch because it is controlled by the Federalists, Marbury is partisan to a degree that has almost never been replicated by the Supreme Court.[151]

Supporting the attempt of a defeated political party to rearrange the playing field for future battles, resisting the peaceful transfer of power that democratic elections are supposed to produce … these are bad things. (After the election of 2018, defeated parties in states including Wisconsin and North Carolina attempted to alter the powers of state government to reduce the impact of their loss. These efforts were widely condemned.[152])

Amazingly, it gets even worse. Susan Low Bloch recounts the details of Marbury and goes on to suggest that perhaps the whole thing was a collusive suit: perhaps Marshall had a plan, and Marbury was in on it.[153] “[I]t is at least possible,” she argues, that Marshall “knowingly withheld the delivery of some commissions to set the stage for this profoundly important case” and that “Marbury and his colleagues knowingly chose to sue in the Supreme Court” because Marshall “might have foreseen that he could use the case in the manner he eventually did and might have let it be known that suing in the Supreme Court was the preferred course.” This is an astonishing chronicle of judicial misbehavior. But what conclusion does Bloch draw? That Marshall in Marbury pulled off “a remarkable feat for which we are all in his debt.”[154]

The suggestion that Marbury was a collusive suit may or may not be true. Even without that, we know enough to know that, in the words of Michael Stokes Paulsen, Marbury is “rotten to the core.”[155] The question that remains is the one he raises: “How could we possibly celebrate such a despicable opinion as the cornerstone of American constitutional law?!”[156]

There are some answers. We have tried above to describe how (and when) Marbury rose to prominence. Around the turn of the 20th century (happily coincident with Marbury’s centennial), advocates of aggressive judicial review needed a symbol and made one. And perhaps they needed one as contentless and political as Marbury is: what they were trying to support, remember, was the era of judicial review now reviled as activism, the Lochner era. The harder question is how Marbury survived the death of Lochner. The answer there, we have suggested, is that it managed to latch on to Warren Court cases that law professors like. And again, the content-free nature of Marbury made it easy to use: Marbury can be deployed in support of any invalidating decision.[157]

But the Warren Court is over. Law professors, being generally liberal, are no longer so enthusiastic about the aggressive judicial review practiced by the Rehnquist and Roberts Courts. Why does Marbury persist?

One answer is that there is always a constituency for aggressive judicial review. More recently, conservatives have embraced Marbury. Liberals have a hard time rejecting it because it is still associated with civil rights cases. And although everyone seems to understand the problematic features of the actual decision, rejecting Marbury as symbol is hard. Marbury has no explanation of when aggressive judicial review is good and when it is bad; indeed, it seems entirely unaware of the question. Hence, there is nothing in the opinion to attack, except judicial review in general. Making arguments about why Marbury does not support invalidation in a particular case is like having a fistfight with fog.

This is perhaps the greatest problem, that Marbury and Lochner seem to have set the terms for debates about the exercise of judicial review. The two sides hurl empty slogans at each other, alternating their invocations of the two meaningless cases as expediency demands.[158] And there is no way to make progress, because we have no replacement. There is no case explaining when aggressive judicial review is appropriate and when the Court should defer. Or at least no early case. Or at least no case with an opinion by the great John Marshall.

Except there is.

  1. Replacing Marbury

Can we dislodge Marbury? Remove it from the rhetoric of the Supreme Court, as a symbol of the propriety of aggressive judicial review and judicial supremacy in constitutional interpretation? Remove it from the first-year constitutional law course, as an introduction to judicial review? Some scholars doubt it. “Marbury v. Madison is so firmly enshrined as the dramatic founding moment of the doctrine of judicial review,” writes Jack Ralove, “that it is difficult to imagine how it could ever be displaced.”[159]

But times change. Marbury as the avatar of judicial review came from nowhere; we can hope it will return there.[160] Honestly confronting its flaws is one step in that direction. But Marbury exists as a symbol not because it is misunderstood but because some symbol is needed. People will always want a case that stands for the idea that aggressive judicial review is appropriate.

Still, unless they are acting in bad faith, they might want a case that actually says something about that idea. And as we noted in the introduction, such a case exists. An early case, a Marshall opinion, a case regularly and approvingly cited by the Supreme Court: McCulloch v. Maryland.

McCulloch arises out of another political struggle, the fight over the Bank of the United States. After considerable debate about whether Congress had the power to charter a national bank, the First Bank of the United States was incorporated in 1791. Without sufficient political support to renew it, its charter expired in 1811. The War of 1812, however, demonstrated the need for a national bank, and the Second Bank of the United States was incorporated in 1816.

Political opposition remained, and some states, including Maryland, imposed taxes on the Second Bank in an attempt to destroy it.[161] James McCulloch, the cashier of the Baltimore branch of the Second Bank, refused to pay the tax. The state of Maryland sued him in its own courts and, perhaps unsurprisingly, prevailed. McCulloch then took the case to the United States Supreme Court.

John Marshall confronted two questions, conventionally called McCulloch I and McCulloch II. First, did Congress have the power to create the bank? Second, did Maryland have the power to tax it?

Neither of these questions had a clear answer based on the words of the Constitution. The power to charter a bank, or to create corporations generally, was not specifically enumerated in Article I, Section 8. But perhaps, as Hamilton had argued, the creation of a bank was an appropriate means to one or more of the ends enumerated in Section 8. As for Maryland’s tax, states certainly did have the power to tax corporations operating within their borders. But the structure of our federal system might impose limits on what states could do to instrumentalities of the federal government.

McCulloch is considered a great case, and it is a staple of constitutional law casebooks. [162] But it is generally understood to stand primarily for the idea that the federal government has broad powers.[163] It is not generally considered an important case about the exercise of judicial review. And those who do focus on its discussion of judicial review tend to read it primarily for the proposition that the Constitution is to be interpreted in a flexible manner.[164]

But that reading misunderstands McCulloch. Marshall does not suggest that the Constitution should always be interpreted in a particular way. “[I]t is a Constitution we are expounding” means that text alone will often not supply an answer. But that is just to say that many constitutional questions will not be clear—and the consequence of that, as we have argued earlier, is that the Court must then decide whether or not to defer to the views of the government actor whose conduct it is reviewing.

The genius of McCulloch—what makes it the perfect case to start a constitutional law course—is that it addresses this issue. It addresses it in the context of two different questions (McCulloch I and McCulloch II), and it offers a plausible explanation of why deference is appropriate in one, but not the other.

As to the power of Congress to create a bank, Marshall makes several observations about deference. Congress debated the issue, he points out; its Members were pure and intelligent. (They were, of course, the same people who drafted the Judiciary Act of 1791, a factor given not even a mention in Marbury.) Those considerations are reasonable—in deciding whether to defer to what Jefferson called “the wisdom of the legislature” one might want to know whether the wisdom exists and has been employed—but they have not ended up playing a significant role in Supreme Court decisionmaking.[165] He also offers two other considerations, however, which have.

First, Marshall observes that the question of whether Congress can create a bank is one “in the decision of which the great principles of liberty are not concerned, but the respective powers of those who are equally the representatives of the people are to be adjusted. . . .”[166] This is a distinction between cases about individual rights (the great principles of liberty) and federalism cases (those that determine whether a particular power belongs to the states or the federal government, who are equally the representatives of the people), with the implication being that courts should be more deferential on issues relating to federalism.

Marshall does not explain why he believes that more aggressive review is justified in individual rights cases, but it is not difficult to construct arguments in support of this principle. Consideration of the costs of error gives one: if the Court mistakenly allows Congress to use a power that should have been left to the states, it is not clear that any individual is worse off. There is some departure from our scheme of dual sovereignty, but that injury may not be severe if Congress can be trusted not to use its powers to regulate matters the states can handle on their own. (And at this point the first argument blends into the second one, discussed below.) If, on the other hand, the Court mistakenly allows the government to abridge some individual right, the cost is obvious: it is the violation of that right.

Second, Marshall considers the extent to which Congress may be trusted not to abuse its power. (The point is made explicitly in McCulloch II, but the idea informs the discussion of deference in McCulloch I.) “In the legislature of the Union alone,” he observes, “all are represented. The legislature of the Union alone, therefore, can be trusted by the people with the power of controlling measures which concern all, in the confidence that it will not be abused.”[167] Because the states are represented in Congress, that is—because Senators are selected by state legislatures[168]—Congress may be trusted to observe the appropriate lines between state and federal authority.

That is an argument for deference. McCulloch II presents the opposite side of the coin, particularly the representation argument. Counsel for Maryland argued that just as Congress could be trusted with expansive powers, so could the state. But there is an obvious difference, answered Marshall: “that which always exists, and must always exist, between the action of a whole on a part, and the action of a part on the whole… .”[169] Because Maryland will receive all the benefits of a tax imposed on the Bank of the United States, while most of the burden will fall on out-of-staters with no voice in the Maryland legislature, Maryland has obviously distorted incentives.

That is an argument for aggressive judicial review, based on what we could call defects in democracy. The Maryland legislature may well have superior institutional competence in deciding what level of taxation is appropriate given the services the state provides to the Bank. But there is every reason to think that it will not, in fact, do its best to identify a reasonable level and will instead tax too highly.

So McCulloch offers us a perfect example of how to exercise judicial review: one issue on which deference is appropriate, and one issue on which it is not. But did McCulloch’s model of judicial review fade away? In particular, if like many (though not all) modern judges and law professors we would like to justify a large part of the Warren Court jurisprudence but not revive Lochner, does McCulloch offer anything?

Of course it does. Far more than Marbury—which has no theory about when aggressive review is good or bad and was, remember, first canonized precisely to defend LochnerMcCulloch tells us what the difference is. Lochner now looks to us like judicial second-guessing the wisdom of legislative policy choices. That is unjustified on the McCulloch theory: very bad policies may well be unconstitutional (most of the doctrinal tests implementing the Due Process and Equal Protection clauses actually have elements of policy assessment in them[170]), but legislatures are better at making those choices and there is no reason, with ordinary economic regulation, to suppose they will not use their superior competence.[171]

And what about the Warren Court? John Hart Ely famously argued that most of the controversial Warren Court decisions could actually be understood as responses to what we have called defects in democracy: circumstances in which the benefits of a law flow to a politically powerful group and the burdens fall on a weaker one. Most of the decisions, he said, could be justified under the theory of Carolene Products footnote four.

But where does Carolene Products get its theory? From McCulloch! Carolene Products explicitly cites McCulloch (more precisely, McCulloch II) as a case supporting its suggestion of heightened scrutiny for laws that burden the interests of discrete and insular minorities.[172]

The point bears emphasis. Carolene Products marks the Court’s abandonment of Lochner and its renunciation of the idea that judges can second-guess legislative determinations about what the public interest requires. But in footnote four, it suggests that some exercises of aggressive judicial review can still be justified. And for that principle it does not cite Marbury­—wisely so, since Marbury says nothing about that question and had recently been used to support Lochner. It cites McCulloch II, which gives exactly the theory that is required.[173]

McCulloch, it turns out, does what the Court and commentators used Marbury for in the civil rights era. It supports the Warren Court, both particular decisions such as Brown and and the general jurisprudential stance: deference with respect to the limits of federal power, greater suspicion with respect to individual rights, and less deference to state legislatures. And unlike Marbury, it does so with coherent argument rather than slogan and convenient elision; unlike Marbury, it cannot be invoked to support any act of judicial invalidation. And unlike Lochner, it offers firm ground to criticize a court that fails to defer when it should. There is thus a particular irony that in turning to Marbury instead of McCulloch, defenders of Warren Court jurisprudence elevated a bad model of judicial review (partisan activism) rather than a good one (differential deference).

Do you want to criticize New Federalism decisions like Lopez or Shelby County?[174] Don’t say they’re like Lochner; say that this is McCulloch I. Do you want to support Warren Court decisions like Brown, or more modern ones like Obergefell? Don’t say they’re like Marbury; say that they are McCulloch II. Do you want to claim supremacy for the judiciary? McCulloch actually has a powerful statement to that effect—unlike Marbury, which really offers only the word “emphatically.” Of constitutional questions, McCulloch says this “by this tribunal alone can the decision be made. On the supreme court of the United States has the constitution of our country devolved this important duty.”[175]

McCulloch does everything that judges can appropriately ask Marbury to do, and does it much better.[176] It does not have the historical stain of being invoked in support of Lochner and Dred Scott.[177] And it is not an undertheorized, disingenuous, rankly partisan example of judicial misbehavior. It is an example of how judicial review should be done. McCulloch, not Marbury, is what the Court should appeal to if it wants to justify the exercise of judicial review.

What about law professors? Perhaps the worst thing about starting a constitutional law course with Marbury is that it promotes a certain vision of the judicial role—not judicial supremacy so much as judicial exclusivity.[178] In Marbury, constitutional analysis seems to take place only within the Court. There is no suggestion that anyone else might have thought about the possible conflict between the Judiciary Act and Article III. That is part and parcel of Marbury’s neglect of the question of deference: if judges are the only ones who think about the Constitution, of course their views are the only ones that matter.

We do not know what, if anything, the drafters of the Judiciary Act of 1789 said about the constitutional question Marshall found in Marbury. But we do know what Congress and other governmental actors said about the question in McCulloch I. James Madison gave a speech to the House of Representatives in which he argued against the constitutionality of the First Bank—clear evidence that Congress took seriously its obligation to consider the constitutionality of its bills.[179] When, despite Madison’s arguments, Congress passed the bill, President Washington asked his cabinet for opinions about its constitutionality—evidence, similarly, that the President has an obligation to consider constitutionality of legislation before signing.[180] Most of those opinions analyzed the issue on a blank slate, as if it were being considered for the first time. But Thomas Jefferson, interestingly, also discussed the significance of the fact that two houses of Congress had voted for the bill. Unless a bill is clearly unconstitutional, he advised Washington, “a just respect for the wisdom of the legislature would naturally decide the balance in favor of their opinion.” That is a prescription for deference—but not always. The veto, Jefferson continues, is for cases where Congress is “clearly misled by error, ambition, or interest…”[181]

Here, as in McCulloch, are the beginnings of a theory about when deference to legislative judgment is or is not appropriate. And because deference plays a role in the President’s decision, we can see the emergence of the gap between the limits the Constitution imposes and the limits particular actors will enforce—we can see that the President might not veto a bill he believes is unconstitutional, or one that is in fact unconstitutional. We can see the emergence of the distinction between doctrine and meaning, or operative propositions and decision rules,[182] which is (in our view) essential to understanding the role of the Supreme Court in relation to other government actors.[183]

Would it be effective to present McCulloch this way, not simply as a case but as what you could call a case study, demonstrating how different government actors interpret the Constitution both independently and in relation to each other? Would it help students to start their analysis of judicial review with a case that exercises that power in a sophisticated and institutionally thoughtful way, rather than one that relies on superficial sloganeering?

We think so. And it has been done: what we have just described is the approach taken by Sandy Levinson’s casebook, used by one of us for fifteen years. Levinson, of course, has vowed not to teach Marbury again. While his coauthors might not go so far, they do presumably agree with the use of McCulloch to demonstrate how constitutional interpretation is or should be done. That adds Paul Brest, Jack Balkin, Akhil Amar, and Reva Siegel to the list of professors who think that Marbury can, at least, lose its spot as the first case assigned in Constitutional Law.

Should Marbury remain in the canon at all? As judicial rhetoric in support of the invalidation of a government act, we think the answer is no. As we have said already, Marbury does not justify aggressive, rather than deferential, judicial review because it does not engage the question of deference at all. And the constitutional analysis Marbury performs is perfunctory, unconvincing, and disingenuous. It may be strategically clever, but partisan wile is the last thing judges should want to associate their opinions with.

As far as teaching goes, we think the value of Marbury is also quite overstated. The opinion does, of course, offer arguments in favor of judicial review. But presenting those to students is, in our view, counterproductive. First, it misleads students about history, by suggesting that judicial review was controversial and in need of establishment. As scholars have pointed out, this was not so.[184] (Strikingly, the Judiciary Act of 1789 itself explicitly contemplates judicial review, since it provides for Supreme Court review of state high court decisions invalidating federal statutes or sustaining state statutes against federal constitutional challenges.[185]) Second, and perhaps more significant, it focuses students’ attention on the existence of judicial review rather than its exercise—it suggests, that is, that we should think about how the institution can be justified, rather than the question of when judicial review should be aggressive and when it should be deferential. This is precisely the mistake that Alexander Bickel made.[186] The fact that a generation of scholars followed him in setting their lances for the countermajoritarian difficulty does not make it any less a windmill.

Still, as long as Marbury retains its talismanic role in judicial opinions, it might be going too far to omit it from the first-year course entirely. Our modest proposal is that Marbury be presented later—and that, when it is presented, it should be portrayed neither as the establishment of judicial review nor as a shining moment in Supreme Court history. Instead, it should be revealed for what it is. That is, a sort of Frankenstein’s monster, a decision whose many unconvincing elements are a consequence of a fundamentally inappropriate approach to constitutional decisionmaking—but, for all that, a symbol which, having sparked to life, we may now be incapable of killing.[187]

         *    Professor of Law, University of Pennsylvania Law School.

       **    JD 2019, University of Pennsylvania Law School. We thank Sandy Levinson for helpful comments on an earlier draft. The opinions expressed in this Article are those of the authors alone.

        [1].    Judicial statements to this effect appear as early as 1825, see People ex rel. Ewing v. Forquer, 1 Ill. 104 (Ill. 1825), and were picked up by the U.S. Supreme Court in 1893. See Noble v. Union River Logging R.R. Co., 147 U.S. 165, 171 (1893). Forquer, interestingly, displays some ambivalence. After describing the Marbury Court as “a tribunal filled with as enlightened and as able jurists as ever graced the judgment-seat in this or any other nation” and the opinion itself as “conspicuous for its luminous displays of deep research and constitutional learning,” the Illinois Supreme Court went on to note that the opinion “has not given universal satisfaction.” 1 Ill. 104 at 106. The next decision, Johnson v. United States, 13 F. Cas. 868 (D. Me. 1830) (Story, J.), calls Marbury great and explains that it is “great, not only from the authority which pronounced it, but also from the importance of the topics which it discussed … .” Id. at 873. References become more simply laudatory towards the end of the 18th century. See, e.g., Ex parte Lusk, 2 So. 140, 144 (Ala. 1887) (referring to “the great case of Marbury v. Madison” without elaboration). When the Supreme Court first used the phrase, in Noble, it did not explain why Marbury was great. It invoked Marbury to support a distinction between discretionary executive acts, which were unreviewable, and ministerial acts, which could be reviewed. (As discussed below, this was Marbury’s main early significance. Its emergence as a great case about judicial review came later. See, e.g., Ragland v. Anderson, 100 S.W. 865, 868 (Ky. Ct. App. 1907) (“It has never been doubted in this country since the great case of Marbury v. Madison [citation omitted] that an act of the legislative part of the government which is contrary to the Constitution is void, and will be so held by the courts whenever brought to their attention.”).)

        [2].    Davison M. Douglas, The Rhetorical Uses of Marbury v. Madison: The Emergence of a “Great Case,” 38 Wake Forest L. Rev. 375, 376 (2003).

        [3].    Id. (quoting 3 John Marshall: Life, Character and Judicial Services 130 (John F. Dillon ed., 1903).

        [4].    The first invocation of Marbury in support of invalidation of a federal statute appears to be Pollock v. Farmers’ Loan & Trust Co., 157 U.S. 429, 583 (1895). Marbury gets a passing mention in Mugler v. Kansas, 123 U.S. 623, 661 (1887), where the Court ended up upholding a state law. In more recent history, Marbury gets prominent mention in Cooper v. Aaron, 358 U.S. 1, 17–18 (1958) and thereafter appears with high frequency in controversial decisions where the Court (or individual Justices) might naturally seek support for the exercise of judicial review. See, e.g.,           Janus v. American Federation of State, County, and Mun. Employees, 138 S. Ct. 2448, 2486 n.27 (2018); Ziglar v. Abbasi, 137 S. Ct. 1843, 1874 (2017) (Thomas, J., concurring); Nation Labor Relations Board v. Noel Canning, 573 U.S. 513, 525 (2014); United States v. Nixon, 418 U.S. 683, 703–05 (1974); Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 567 (1985) (Powell, J., dissenting); Thompson v. Oklahoma, 487 U.S. 815, 833 n.40 (1988); James B. Beam Distilling Co. v. Georgia, 501 U.S. 529 (1991); Clinton v. Jones, 520 U.S. 681, 703 (1997); Vieth v. Jubelirer, 541 U.S. 267, 277 (2004); King v. Burwell, 135 S. Ct. 2480, 2496 (2015); Sanchez-Llamas v. Oregon, 548 U.S. 331, 353–54 (2006); United States v. Windsor, 570 U.S. 744, 762 (2013); Citizens United v. Federal Election Commission, 558 U.S. 310, 365 (2010). In some of these decisions, Marbury is cited as support for not invalidating a challenged law—a consequence, we will suggest, of the fact that Marbury provides no guidance as to when judicial review should be deferential and when it should be aggressive. See infra Part III.

        [5].    Amanda Rinderle & Keith E. Whittington, Making a Mountain out of a Molehill? Marbury and the Construction of the Constitutional Canon, 39 Hastings Const. L.Q. 823, 824 (2012). See also, e.g., Jack M. Balkin & Sanford Levinson, The Canons of Constitutional Law, 111 Harv. L. Rev. 963, 1010 (1998) (describing Marbury as “[t]he crown jewel in the constitutional canon”).

        [6].    In particular, see Douglas, supra note 2; Whittingon & Rinderle, supra note 5. See also Robert M. Casale, Revisiting one of the Law’s Great Fallacies: Marbury v. Madison, 89 Conn. Bar. J. 62 (2015).

        [7].    See, e.g., Jamal Greene, The Anticanon, 125 Harv. L. Rev. 379 (2011) (identifying Lochner as one of four anti-canonical cases); Richard A.Primus, Canon,, Anti-Canon, and Judicial Dissent, 48 Duke L.J. 243 (1998). Lochner does have its defenders. See, e.g.,  David E. Bernstein, Lochner v. New York: A Centennial Retrospective, 83 Wash. U.L.Q. 1469 (2005).  We will suggest that Lochner’s bad reputation—like Marbury’s good one—is attributable more to subsequent mythologizing than the actual decision. See infra Part III.

        [8].    Most of what we say in Parts I and II is relatively well known among law professors. See Casale, supra note 6, at 62 (stating that “scholars have known for years that this [conventional] understanding of Marbury [as establishing judicial review] is wrong”). We are, though, unsure about the extent to which professors or judges are aware of the historical connection between Marbury and Lochner. The scholarship that develops this connection, notably Robert Lowry Clinton, Marbury v. Madison and Judicial Review (1989), is now relatively old; however, most legal scholarship still treats the two cases as bearing their conventional symbolic meaning—that is, as opposites. See, e.g., Richard A. Primus, Canon, Anti-Canon, and Judicial Dissent, 48 Duke L. J. 243, 245 (1998) (contrasting Lochner and Plessy with Marbury and McCulloch). Judges, too, tend to invoke Marbury and Lochner as opposites—at least since Lochner acquired its modern meaning as a symbol of illegitimate judicial activism. See, e.g., United States v. Lopez, 514 U.S. 549, 566, 605 (majority relies on Marbury while dissent characterizes decision as akin to Lochner); Compare In re Senate Resolution No. 2 Concerning Constitutionality of House Bill No. 6, 31 P.2d 325, 331 (Colo. 1933) (citing Marbury and Lochner together to support judicial invalidation).

        [9].    Marbury cannot escape Lochner, we will argue, because while everyone agrees that judicial review is a good thing (the proposition that Marbury actually argues for), Marbury is actually invoked to support not the existence of judicial review but the invalidation of government acts. And everyone also agrees that some kinds of invalidation—the “judicial activism” now associated with Lochner—is bad. But because Marbury says nothing about how judicial review should be exercised, it provides no way to distinguish between good Marbury invalidations and bad Lochner ones. See infra Part III.

      [10].    For an argument about the centrality of deference, see generally Kermit Roosevelt III, The Myth of Judicial Activism (2006).

      [11].    Marbury is worse, we will argue, for two main reasons. First, unlike Lochner, it has no plausible explanation of why it chooses the aggressive version of judicial review. And second, unlike Lochner, it is a bad faith exercise in partisan political maneuvering. See infra Part III.

      [12].    17 U.S. 316 (1819).

      [13].    See infra Part I.

      [14].    See William Michael Treanor, Judicial Review Before Marbury, 58 Stan. L. Rev. 455, 555 (2005) (“The fact that judicial review was exercised much more frequently than previously recognized in the years before Marbury helps explain why Marshall’s assertion of the power to exercise judicial review in the case elicited so little comment and also highlights the consistency between Marbury and the prior body of case law.”)

      [15].    Not nothing—as we note later, Marbury distinguishes between unreviewable discretionary authority and reviewable ministerial acts, a distinction which was considered the most important part of the opinion for almost a century. It does not, however, say anything about when the Court’s review should be deferential and when it should be aggressive.

      [16].    See Alexander M. Bickel, The Least Dangerous Branch 8 (1963) (stating that Marbury established judicial review as an institution “‘done’ at a given time and by a given act.”); Robert L. Clinton, The Strange History of Marbury v. Madison in the Supreme Court of the United States, 8 St. Louis U. Pub. L. Rev. 13, 17–18 (1989).

      [17].    Compare Marbury v. Madison (“It is emphatically the province and duty of the Judicial Department to say what the law is. Those who apply the rule to particular cases must, of necessity, expound and interpret the rule. If two laws conflict with each other, the Court must decide on the operation of each.”), with The Federalist No. 78 (Alexander Hamilton) (“The interpretation of the laws is the proper and peculiar province of the courts. … It therefore belongs to them to ascertain its meaning, as well as the meaning of any particular act proceeding from the legislative body. If there should happen to be an irreconcilable variance between the two, … the Constitution ought to be preferred to the statute ….”).

      [18].    See e.g., William Alexander Duer, Constitutional Jurisprudence of the United States (rev. ed. 1856). What may be the first decision to call Marbury “great,” People Ex rel. Ewing v. Forquer, 1 Ill. 104(Ill. 1825), is focused on the availability of mandamus and displays some ambivalence even as to that. After describing the Marbury Court as “a tribunal filled with as enlightened and as able jurists as ever graced the judgment-seat in this or any other nation” and the opinion itself as “conspicuous for its luminous displays of deep research and constitutional learning,” the Illinois Supreme Court went on to note that the opinion “has not given universal satisfaction.” 1 Ill. 104 at 106. See generally Douglas, supra note 2; Rinderle & Whittington, supra note 5; Robert Lowry Clinton, Marbury v. Madison and Judicial review 120 (1989).

      [19].    60 U.S. 393 (1857). Had Marbury not existed, late-nineteenth century proponents of aggressive judicial review might have greater difficulty making their case, since Dred Scott quickly became part of the anti-canon: cases that discredit their associated doctrines. And had Dred Scott anticipated the Lochnerites by itself appealing to Marbury, it might have created a similar effect. One of the surprising things about Marbury, though, is the extent to which it has avoided guilt by association—it has not been tarnished by its invocation in Pollock v. Farmers’ Loan & Trust Co., 157 U.S. 429, 583 (1895) and during the Lochner era. The most surprising thing to us, though, is that Marbury has avoided liability for its own guilt: it is held out as a great case despite being a terrible example of the proper role of courts in a constitutional democracy. See infra Part IV.

      [20].    See Monongahela Navigation Co. v. United States, 148 U.S. 312 (1893); Callan v. Wilson, 127 U.S. 540 (1888); Baldwin v. Franks, 120 U.S. 678 (1887); Boyd v. United States, 116 U.S. 616 (1886); The Civil Rights Cases, 109 U.S. 3 (1883); United States v. Harris, 106 U.S. 629 (1883); Trade-Mark Cases, 100 U.S. 82 (1879); United States v. Fox, 95 U.S. 670 (1878); United States v. Reese, 92 U.S. 214 (1876); United States v. R.R. Co., 84 U.S. 322 (1873); United States v. Klein, 80 U.S. 128 (1872); Collector v. Day, 78 U.S. 113 (1871); United States v. DeWitt, 76 U.S. 41 (1870); Justices v. Murray, 76 U.S. 274 (1870); Hepburn v. Griswold, 75 U.S. 603 (1870); Alicia, 74 U.S. 571 (1869); Reichart v. Felps, 73 U.S. 160 (1868); Ex parte Garland, 71 U.S. 333 (1867); Gordon v. United States, 69 U.S. 561 (1865); Scott v. Sandford, 60 U.S. 393 (1857). See Douglas, supra note 2 (listing cases) at 395 n.85.

      [21].    See Bruce Ackerman, Taxation and the Constitution, 99 Colum. L. Rev. 1, 5 (1999) (quoting 1 Archibald Butt, Taft and Roosevelt 134 (1930)). A contemporaneous review of Pollock begins by setting itself up as a parallel to critiques of Dred Scott and describes the decision as “overrul[ing] in effect three direct adjudications made by [the Supreme Court] … and thereby cripple[ing] an important and necessary power and function of a coordinate branch of government [via] an opinion … that is contrary to what has been accepted as law for nearly one hundred years.” Francis R. Jones, Pollock v. Farmers’ Loan and Trust Company, 9 Harv. L. Rev. 198, 198 (1895). In addition to being widely viewed as poorly reasoned and incorrect, Pollock is one of the few Supreme Court decisions to be reversed by a constitutional amendment (the sixteenth). For discussion of its anti-canonical status, see Jamal Greene, The Anticanon, 125 Harv. L. Rev. 379, 389 (2011).

      [22].    Pollock v. Farmers’ Loan & Trust Co., 157 U.S. 429, 583 (1895). Before this, the Supreme Court did discuss Marbury to support the Court’s power to review the constitutionality of state laws in Mugler v. Kansas, though it ultimately decided the state law in question was constitutional. 123 U.S. 623, 661 (1887).

      [23].    Justice Edward White, dissenting, argued that the decision was inconsistent with the idea of the Court as an institution independent of its membership—that is, that it looked like a political opinion rather than a judicial judgment. 157 U.S. 429 at 652 (White, J., dissenting) (arguing that [t]he fundamental conception of a judicial body is that of one hedged about by precedents which are binding on the court without regard to the personality of its members”); Id. at 651 (arguing that “[i]f the permanency of [the Court’s] conclusions is to depend upon the personal opinions of those who, from time to time, may make up its membership, it will inevitably become a theater of political strife, and its action will be without coherence or consistency”).

      [24].    See Douglas, supra note 2, at 395–97.

      [25].    See U.S. Const. amend. XVI.

      [26].    See id. at 379.

      [27].    Robert L. Clinton, The Strange History of Marbury v. Madison in the Supreme Court of the United States, 8 St. Louis U. Pub. L. Rev. 13, 18. (“When Progressive reformers attacked the courts in the late-nineteenth and early-twentieth centuries, one of the focal points of that attack was Marbury.”); Id (“Supporters of the courts hailed the decision as having established a principle which, in its effect if not in its purpose, was as important a ‘charter of American liberty’ as the Declaration of Independence and the Constitution itself”). Notably, both sides committed a characteristic error by treating the dispute as if it were about the existence of judicial review rather than its proper exercise. See infra Part IV.

      [28].    See Douglas supra note 2, at 398–401.

      [29].    See id. at 404.

      [30].    See id. for specific examples.

      [31].    See id.

      [32].    See id. (citing Charles E. Martin, An Introduction to the Study of the American Constitution: A Study of the Formation and Development of the American Constitutional System and of the Ideals Upon Which it is Based with Illustrative Materials 111–13 (1928)).

      [33].    See Douglas, supra note 2, at 405.

      [34].    See id.

      [35].    See id. at 407.

      [36].    See Fairbank v. United States, 181 U.S. 283, 285 (1900); Dooley v. United States, 183 U.S. 151, 173 (1901) (Fuller. C.J., dissenting); Muskrat v. United States, 219 U.S. 346, 357 (1910).

      [37].    Muskrat v. United States, 219 U.S. 346, 357 (1911).

      [38].    Myers v. United States, 272 U.S. 52, 139 (1926).

      [39].    From 1926 to 1958, when Cooper v. Aaron was decided, the Supreme Court cited Marbury’s judicial review component only five times. Id.; Adamson v. California, 332 U.S. 46, 90 (1946) (Black, J., dissenting); United States v. Commodities Corp., 339 U.S. 121, 124 (1949); Touhy v. Ragen, 340 U.S. 462, 468 (1950); Textile Workers Union of Am. v. Lincoln Mills of Ala., 353 U.S. 448, 464 (1956) (Frankfurter, J., dissenting).

      [40].    Cooper v. Aaron, 358 U.S. 1, 18 (1958).

      [41].    See Clinton, supra note 16, at 14 (“Careful scrutiny of the Marshall opinion … reveals no explicit declaration of the Court’s authority to issue ‘final’ proclamations on constitutional issues generally, so to ‘bind’ coordinate departments to the judicial declaration.”).

      [42].    Keith E. Whittington & Amanda Rinderle, Making a Mountain out of a Molehill – Marbury and the Construction of the Constitutional Canon, 39 Hastings Const. L.Q. 823, 860 (2012).

      [43].    See, e.g., Alexander Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics 1–14 (1962); Clinton, supra note 8; William van Alstyne, A Critical Guide to Marbury v. Madison, 18 Duke L.J. 1 (1969); see generally Douglas, supra note 2.

      [44].    See Jack N. Rakove, The Origins of Judicial Review: A Plea for New Contexts, 49 Stan. L. Rev. 1031, 1064 (1997) (“Marbury has become a story that has to be told less because it is really necessary to know how or where or why judicial review originated, but because agreement on a common point of departure makes it easier to frame and dispute the issues that matter today … [w]hether or not Marbury is the right point of historical departure does not really matter if its heuristic value is so clearly accepted.”)

      [45].    Cooper v. Aaron, 358 U.S. 1, 18 (1958) (emphasis added) (quoting Marbury, 5 U.S. 137, 177.) The “emphatically” sentence appears to have been quoted once before, in the middle of a lengthy quotation from Marbury, in Fairbank v. United States, 181 U.S. 283, 284–86, (1901). But Cooper is the first case to read that language as support for judicial supremacy.

      [46].    Rinderle & Whittington supra note 5, at 836.

      [47].    Id. at 838 (noting that “[t]he correlation between citations to Marbury and cases striking down federal laws is strong in the twentieth century”).

      [48].    See, e.g., Ziglar v. Abbasi, __ U.S. __, 137 S. Ct. 1843 (2017); King v. Burwell, __ U.S. __, 135 S. Ct. 2480 (2015); United States v. Windsor, 570 U.S. 744 (2013); Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519 (2012); Citizens United v. Federal Election Commission, 558 U.S. 310 (2009); District of Columbia v. Heller, 554 U.S. 570 (2008); Boumediene v. Bush, 553 U.S. 723 (2008); United States v. Morrison, 529 U.S. 598 (2000); Bush v. Gore 531 U.S. 98 (2000); Chicago v. Morales, 527 U.S. 41 (1999); Printz v. United States, 521 U.S. 898 (1997); Washington v. Glucksberg, 521 U.S. 702 (1997); Boerne v. Flores, 521 U.S. 507 (1997); Miller v. Johnson, 515 U.S. 900 (1997); United States v. Lopez, 514 U.S. 549 (1994); Planned Parenthood v. Casey, 505 U.S. 833 (1992); Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528 (1984); INS v. Chadha, 462 U.S. 919 (1983); Miss. Univ. for Women v. Hogan, 458 U.S. 718 (1982); United States v. Nixon, 418 U.S. 683 (1974); Bivens v. Six Unknown Named Agents of Fed. Bureau of Narcotics, 403 U.S. 388 (1971); Goldberg v. Kelly, 397 U.S. 254 (1969); Griswold v. Connecticut, 381 U.S. 479 (1965); Heart of Atlanta Motel v. United States, 379 U.S. 241 (1964); Bell v. Maryland, 378 U.S. 226 (1963).

      [49].    570 U.S. 744 (2013).

      [50].    Id. at 762.

      [51].    Id. at 778 (Scalia, J., dissenting); id. at 787 (focusing on Marbury’s description of “[t]hose who apply the rule to particular cases”).

      [52].    __ U.S. __,135 S. Ct. 2480, 2496 (2015).

      [53].    “Canonical” here means something different from “great.” Marbury’s canonical status is demonstrated by the number of judicial decisions that treat it as a great decision and meaningful support for the judicial invalidation of government acts. We do not believe it should be treated that way, but mistaken or strategic canonization is still canonization.

      [54].    198 U.S. 45 (1905).

      [55].    See infra Part IV.

      [56].    See David E. Bernstein, Lochner v. New York: A Centennial Retrospective, 83 Wash. U.L.Q. 1469, 1500–1505 (2005) (discussing contemporaneous reaction to Lochner).

      [57].    Compare Lochner, 198 U.S., at 59 (stating that “[w]e think there can be no fair doubt that the trade of a baker … is not an unhealthy one”) with id. at 70–71 (arguing that baking is unhealthy) (Harlan, J., dissenting); See Howard Gillman, De-Lochnerizing Lochner, 85 B.U. L. Rev. 859, 860 (2005) (stating that Lochner “does not appear to be especially dramatic” and “did not cause much of a stir in the political system”).

      [58].    300 U.S. 379 (1937).

      [59].    Burnet v. Coronado Oil & Gas Co., 285 U.S. 393 (1932); Meyer v. Nebraska, 262 U.S. 390 (1923); Adkins v. Children’s Hospital of the District of Columbia, 261 U.S. 525 (1923); Prudential Ins. Co. of Am. v. Cheek, 259 U.S. 530 (1922); Coppage v. State of Kansas, 236 U.S. 1 (1915); Chicago, B. & Q.R. Co. v. McGuire, 219 U.S. 549 (1911); Atlantic Coast Line R. Co. v. Riverside Mills, 219 U.S. 186 (1911); Laurel Hill Cemetery v. San Francisco, 216 U.S. 358 (1910); Welch v. Swasey, 214 U.S. 91 (1909); Muller v. Oregon,  208 U.S. 412 (1908); Adair v. United States, 208 U.S. 161 (1908).

      [60].    David E. Bernstein, Lochner v. New York: A Centennial Retrospective, 83 Wash. U. L.Q. 1469 (2005) at 1506; See also id., at 1511–1512 (noting that Lochner “did not yet have the symbolic resonance that it later acquired”).

      [61].    The phrase is from United States v. Carolene Products, 304 U.S. 144, 152 (1938).

      [62].    An attempt to define the constitutional anti-canon finds Lochner joined by only three other cases: Dred Scott v. Sanford, Plessy v. Ferguson, and Korematsu v. United States. See Jamal Greene, The Anticanon, 125 Harv. L. Rev. 379 (2011).

      [63].    Id. at 75 (Holmes, J., dissenting).

      [64].    333 U.S. 507, 527 (1948) (Frankfurter, J., dissenting). Earlier, Frankfurter had quoted the Holmes dissent for the aphorism that general propositions do not decide concrete cases—part of the elevation of the Holmes dissent into the canon, but not a suggestion that the Lochner majority had behaved improperly. See Harris v. United States, 331 U.S. 145, 157 (1947) (Frankfurter, J., dissenting). In his Winters dissent, though, he deployed Lochner in its modern meaning: he accused the majority of repeating the sin of Lochner by “confusing economic dogmas with constitutional edicts.”

      [65].    335 U.S. 538, 543 (1949) (Frankfurter, J., concurring).

      [66].    See Howard Gillman, De-Lochnerizing Lochner, 85 B.U. L. Rev. 859, 860 (2005) (discussing Frankfurter’s role in the creation of Lochner as a symbol of activism).

      [67].    Day-Brite Lighting Inc. v. Missouri, 342 U.S. 421, 423 (1952).

      [68].    See, e.g., Poe v. Ullman, 367 U.S. 497, 517 (1961) (describing Lochner era as one in which “the Court struck down social legislation when a particular law did not fit the notions of a majority of Justices as to legislation appropriate for a free enterprise system”).

      [69].    381 U.S. 479 (1965).

      [70].    Griswold v. Connecticut, 381 U.S. 479, 481–82 (1965). The distinction that Douglas draws here, interestingly, does not seem to be between faithful, mechanical enforcement of the Constitution and policy-driven activism but rather between “laws that touch economic problems, business affairs, or social conditions” and those that affect “an intimate relation of husband and wife”—suggesting that the Court may indeed determine the wisdom of such laws. Id. at 482.

      [71].    Griswold, 381 U.S. at 514, 524 (Black, J., dissenting)

      [72].    Id. at 528 (Stewart, J., dissenting).

      [73].    See David E. Bernstein, Lochner v. New York: A Centennial Retrospective, 83 Wash. U. L.Q. 1469 (2005); Moore v. E. Cleveland, 431 U.S. 494, 502 (1977) (the first Supreme Court case to use the phrase “Lochner era”).

      [74].    Roe v. Wade, 410 U.S. 113, 117 (1973) (“We bear in mind, too, Mr. Justice Holmes’ admonition in his now-vindicated dissent in Lochner v. New York”); id. at 165 (“This holding, we feel, is consistent with the relative weights of the respective interests involved”).

      [75].    Interestingly, one of these opinions (that of Justice Stewart) is a concurrence: Stewart notes that he thought Lochner was dead, but since it was resurrected in Griswold he is willing to second-guess legislative policy choices and agrees with the Court’s balancing of interests. Id. at 168 (Stewart, J., concurring). The Rehnquist dissent makes the more standard move of accusing the Court of improperly assuming the power “to examine the legislative policies and pass on the wisdom of these policies.” Id. at 174.

      [76].    John Hart Ely The Wages of Crying Wolf: A Comment on Roe v. Wade, 82 Yale L.J. 920, 935–46 (1973).

      [77].    Robert H. Bork, The Tempting of America 32 (1990). The quoted sentence continues “and [Dred] Scott,” which is a similar attempt to yoke Roe to an anticanonical case. Supporters of abortion rights viewed the matter differently, although some of them also found fault with Roe’s reasoning. See, e.g., Laurence H. Tribe, Foreword: Toward a Model of Roles in the Due Process of Life and Law, 87 Harv. L. Rev. 1 (1973) (viewing abortion as a liberty issue appropriately addressed via substantive due process and attempting to distinguish Lochner); Ruth Bader Ginsburg, Some Thoughts on Autonomy and Equality in Relation to Roe v. Wade, 63 N.C. L. Rev. 375 (1985) (viewing abortion as an equality issue); Sylvia A. Law, Rethinking Sex and the Constitution, 132 U. Pa. L. Rev. 955 (1984); Catharine A. MacKinnon, Reflections on Sex Equality Under Law, 100 Yale L.J. 1281 (1991); Reva Siegel, Reasoning from the Body: A Historical Perspective on Abortion Regulation and Questions of Equal Protection, 44 Stan. L. Rev. 261 (1992).

      [78].    See Bork, supra note 77..

      [79].    Cass R. Sunstein, Lochner’s Legacy, 87 Colum. L. Rev. 873, at 873 (1987).

      [80].    Bernstein, David E., Lochner v. New York: A Centennial Retrospective 83 WASH. U. L.Q. 1469, 1523 (2005) (“Indeed, to Justices on the “left” of the Rehnquist Court, the majority’s decisions expanding the scope of the Eleventh Amendment and limiting the scope of the Commerce Clause are reminiscent of Lochner.”)

      [81].    Id.

      [82].    “Activism” itself tends to be an empty epithet, what one of us has called “little more than a rhetorically charged shorthand for decisions the speaker disagrees with.” Kermit Roosevelt III, The Myth of Judicial Activism 3 (2006).

      [83].    Sunstein’s article was one of the first reappraisals of Lochner, followed by the careful and historically informed scholarship of Barry Cushman, Howard Gillman, and G. Edward White, among others. See, e.g., Barry Cushman, Rethinking the New Deal Court: The Structure of a Constitutional Revolution (1998); Howard Gillman, The Constitution Besieged: The Rise and Demise of Lochner Era Police Powers Jurisprudence (1993); G. Edward White, The Constitution and the New Deal (2000).

      [84].    See Obergefell v. Hodges, 135 S. Ct. 2584, 2617 (2015).

      [85].    Id. at 2617

      [86].    Id. at 2618.

      [87].    Id. at 2621

      [88].    567 U.S. 519, 623 (Ginsburg, J., concurring/dissenting).

      [89].    On Lochner, see sources cited supra note 83. On Marbury, see, e.g., van Alstyne, supra note 43; Michael Stokes Paulsen, Marbury’s Wrongness, 20 Const. Comment. 343 (2003).

      [90].    The popular understanding of Lochner does tell us that judges should not decide cases by writing their policy preferences into law. But of course no judge has ever described her decision in this way. Lochner itself disavows the practice: “This is not a question of substituting the judgment of the court for that of the legislature. If the act be within the power of the state it is valid, although the judgment of the court might be totally opposed to the enactment of such a law.” 198 U.S. 45, 56–57 (1905).

      [91].    See, e.g., Mark A. Graber, Desperately Ducking Slavery: Dred Scott and Contemporary Constitutional Theory, 14 Const. Comment. 271, 274 (1997). John Hart Ely, quoting Graber, identifies himself as a “democratic” constitutional theorist who would sustain both abortion restrictions and affirmative action, but Ely of course had his own areas where courts should be aggressive: laws that restricted the democratic process. See John Hart Ely, Democracy and Distrust (1980).

      [92].    See Bickel, supra note 43; Sylvia Snowiss, Text and Principle in John Marshall’s Constitutional Law: The Cases of Marbury and McCulloch, 33 J. Marshall L. Rev. 973, 1011–1012 (2000) (“Bickel was motivated immediately to find support for Brown, and, more generally, for the practice of judicial review”).

      [93].    See Bickel, supra note 92, at 103. John Hart Ely used this quote, juxtaposed with another from Bickel about the task of identifying fundamental values, as epigraph for his Harvard Foreword. See John Hart Ely, Foreword, On Discovering Fundamental Values, 92 Harv. L. Rev. 5 (1978).

      [94].    See Hart Ely, supra note 91.

      [95].    See Hart Ely, supra note 93.

      [96].    For Ely, Carolene Products was that case. But it is unlikely that Carolene Products can replace Marbury as the first case that constitutional law students encounter: it is neither recent enough to be ripped-from-the-headlines relevant nor old enough to be foundational. Moreover, the distinction that Carolene Products draws—between ordinary economic regulation and laws that conflict with the text of the Constitution, restrict the political process, or burden the interests of discrete and insular minorities, is too complex for an introduction to judicial review and may not be fully adequate on its own terms. See generally Bruce Ackerman, Beyond Carolene Products 98 Harv. L. Rev. 713 (1985). Also, as we suggest in Part V, Carolene Products cannot replace Marbury because Carolene Products embodies a Reconstruction vision of judicial review, and constitutional law courses generally start with the Founding Constitution.

      [97].    See, e.g., David Strauss, Why Was Lochner Wrong?, 70 U. Chi. L. Rev. 373 (2003); David Bernstein, Rehabilitating Lochner (2011). The author’s view about the importance of freedom of contract then informs the normative assessment; Strauss is a critic of Lochner, while Bernstein finds more to like.

      [98].    For this reinterpretation of Lochner, see, e.g., Barry Cushman, Rethinking the New Deal Court  (1998); Howard Gillman, The Constitution Besieged: The Rise and Demise of Lochner Era Police Powers Jurisprudence (1993); G. Edward White, The Constitution and the New Deal (2000). Our earlier expositions of this argument can be found at Kermit Roosevelt III, Valid Rule Due Process Challenges: Bond v. United States and Erie’s Constitutional Source, 54 Wm. & Mary L. Rev. 987, 990–97 (2013); Kermit Roosevelt III, Forget the Fundamentals: Fixing Substantive Due Process, 8 U. Pa. J. Con.L. 983 (2006).

      [99].    Justice Peckham explains, for instance, that the state can restrain liberty in the exercise of the police power—as, for instance, to protect the health of miners. The flaw he finds in the New York maximum hours law is not that it conflicts with a fundamental right, but rather that it exceeds the scope of legislative authority. “The limit of the police power has been reached and passed in this case.”

    [100].    John Hart Ely, supra note 91, 18..

    [101].    See, e.g., Montgomery v. Louisiana, __ U.S. __, 136 S. Ct. 718, 731 (2016) (noting that “[an] unconstitutional law is void, and is as no law”) (quoting Ex parte Siebold, 100 U.S. 371, 376 (1880).

    [102].    See Marbury v. Madison, 5 U.S. 137, 177 (1803) (stating that “a legislative act contrary to the constitution is not law”). For further discussion of the argument’s role in early debates over judicial review, see Larry Kramer, The Supreme Court 2000 Term Foreword: We the Court, 115 Harv. L. Rev. 4, 50–52 (2001); See also Akhil Reed Amar, Sixth Amendment First Principles, 84 Geo. L.J. 641, 664 (1996) (arguing that a statute authorizing unreasonable searches (in violation of the Fourth Amendment) “was null and void—ultra vires”).

    [103].    See Calder v. Bull, 3 U.S. 386 (1798) (“There are acts which the Federal, or State, Legislature cannot do, without exceeding their authority. . .. An ACT of the Legislature (for I cannot call it a law) contrary to the great first principles of the social compact, cannot be considered a rightful exercise of legislative authority.”) For a similar statement from the Lochner era, see Citizens’ Savings & Loan Ass’n v. City of Topeka, 87 U.S. 655, 664 (1874) (“To lay with one hand the power of government on the property of the citizen, and with the other to bestow it upon favored individuals to aid private enterprises and build up private fortunes, is none the less a robbery because it is done under the forms of law and is called taxation. This is not legislation.”).

    [104].    Some evidence that the lines the court was patrolling were limits of state power rather than boundaries of federal rights comes from the fact that before the 14th Amendment, the Court issued some Lochner-style opinions in the exercise of its diversity jurisdiction. What the 14th Amendment’s Due Process Clause did, on this understanding, was not to set limits on state police power—those limits existed as a matter of general constitutional law, or what the Court sometimes called “the universal law of all free governments.” Green v. Biddle, 21 U.S. 1, 38 (1823). Instead, it created a federal right enforceable through federal question jurisdiction. For a discussion of the origins and evolution of this jurisprudence, see Michael G. Collins, Before Lochner: Diversity Jurisdiction and the Development of General Constitutional Law, 74 Tul. L. Rev. 1263 (2000); Michael G. Collins, October Term, 1896—Embracing Due Process, 45 Am. J. Legal Hist. 71 (2001).

    [105].    Calder v. Bull, 3 U.S. 386 (1798).

    [106].    The preamble of the Declaration asserts that people form governments to secure their natural rights: governments that instead invade those rights are malfunctioning sufficiently to justify revolution.

    [107].    See generally John V. Orth, Taking From A and Giving to B: Substantive Due Process and the Case of the Shifting Paradigm, 14 Const. Comment. 337 (1997).

    [108].    See Lochner v. New York, 198 U.S. 45, 53 (1905) (describing police powers as those which “relate to the safety, health, morals, and general welfare of the public”).

    [109].    169 U.S. 366 (1898).

    [110].    In Holden, the Court explicitly noted the need for deference: “a large discretion is necessarily vested in the legislature, to determine, not only what the interest of the public require, but what measures are necessary for the protection of such interests.” 169 U.S. at 392 (internal quotation omitted).  As Lochner put it, justly, “This court has recognized the existence and upheld the exercise of the police powers of the states in many cases which might fairly be considered as border one, and it has, in the course of its determination of questions regarding the asserted invalidity of such statutes, on the ground of their violation of the rights secured in the Federal Constitution, been guided by rules of a very liberal nature, the application of which has resulted, in numerous instances, in upholding the validity of state statutes thus assailed.” 198 U.S. at 54. The Court went on to note that it had sustained a Sunday closing law, which does not look like the application of modern heightened scrutiny. See id. at 56 (discussing Petit v. Minnesota, 177 U.S. 164 (1900)).

    [111].    Id. at 64.

    [112].    Id. at 69 (Harlan, J., dissenting).

    [113].    198 U.S. at 56–57.

    [114].    This point comes out quite clearly in Home Building & Loan Ass’n v. Blaisdell, where the Court sustained a moratorium on foreclosures (seemingly a law that benefited homeowners at the expense of mortgage holders) on the grounds that not only “the concerns of individuals or of classes were involved,” but rather “the economic structure upon which the good of all depends.” 290 U.S. 398, 442 (1934).

    [115].    See Roosevelt, supra note 10, at 992.

    [116].    Or not very much. A review of the Marbury literature suggests that many authors have not paid much attention to the argument presented by Marbury’s lawyer, Charles Lee, and we do discuss the light Lee’s argument sheds on the decision.

    [117].    The facts of Marbury can be found in any number of places; one thoughtful source is Sanford Levinson & Jack M. Balkin, What are the Facts of Marbury v. Madison?, 20 Const. Comment. 255 (2003).

    [118].    As Thomas Jefferson put it, the Federalists “retired into the judiciary as a stronghold.” 10 The Writings of Jefferson 302 (Andrew A. Lipscomb & Albert E. Bergh eds., 1903).

    [119].    See, e.g., Jerome A. Barron, et al., Constitutional Law: Principles and Policy Cases and Materials 11–23 (8th ed. 2012) (discussing Marbury without mention of Stuart v. Laird); Calvin Massey, American Constitutional Law: Powers and Liberties 7 (2001) (mentioning Stuart v. Laird but not disclosing its outcome); Kathleen M. Sullivan & Noah Feldman, Constitutional Law 9–11 (18th ed. 2013) (offering substantial historical context but omitting Stuart v. Laird). See generally Sanford Levinson, Why I Do Not Teach Marbury (Except to Eastern Europeans) and Why You Shouldn’t Either, 38 Wake Forest L. Rev. 553, 557 (2003) (discussing treatment of historical context in casebooks).

    [120].    See Stuart v. Laird, 5 U.S. 299 (1803).

    [121].    Marbury v. Madison, 5 U.S. 137, 154–162 (1803).

    [122].    Id. at 168–73.

    [123].    Id. at 174–76.

    [124].    Id. at 173.

    [125].    Id. at 176–80.

    [126].    See, e.g., Carol S. Steiker & Panetti v. Quarterman, Is there a “Rational Understanding” of the Supreme Court’s Eighth Amendment Jurisprudence?, 5 Ohio St. J. Crim. L. 285, 285 (2007).

    [127].    See, e.g., van Alstyne, supra note 43.

    [128].    5 U.S. 137, 142–46 (1803).

    [129].    See van Alstyne, supra note 89.

    [130].    Judiciary Act of 1789, 1 Stat. 73 (1789). Pfander points to a different version of Section 13, which uses a colon rather than a semicolon to connect the two parts of the last sentence and capitalizes the word “And.” James E. Pfander, Marbury, Original Jurisdiction, and the Supreme Court’s Supervisory Powers, 101 Colum. L. Rev. 1515, 1540 (2001). Getting the reading of Section 13 right in some absolute sense is ancillary to our project—as noted, the question is not what the “real” Marbury is, but why we assess “our” Marbury the way we do. But we note that even this version makes the grant of power the second part of a sentence about appellate jurisdiction.

    [131].    See, e.g., Akhil Reed Amar, Marbury, Section 13, and the Original Jurisdiction of the Supreme Court, 56 U. Chi. L. Rev. 443, 445 (1989). This article sums up the § 13, and other arguments opposing the soundness of Marshall’s opinion.

    [132].    Marbury, 5 U.S. at 146. James Pfander highlights this point, arguing that it supports Marshall’s interpretation of Section 13 as providing original jurisdiction whenever mandamus is sought against a federal official. See Pfander, supra note 130. But what it more naturally supports is Lee’s argument that this mandamus is in fact the exercise of appellate jurisdiction (which might explain its presence in a sentence that starts out talking about appellate jurisdiction)—not Marshall’s argument that it is an attempt to expand original jurisdiction. See id. at 1567–1568 (noting this argument).

    [133].    Marbury, 5 U.S. at 148.

    [134].    U.S. Const. Art. III, § 2, cl. 2.

    [135].    Marbury, 5 U.S. at 147–148.

    [136].    See generally Roosevelt, supra note 10, at 22–36 (discussing factors relevant to decision about deference); See also McCulloch v. Maryland, 17 U.S. 316, 401–402 (1819) (same).

    [137].    For those that do have answers, we suggest that most would support judicial deference. The First Congress presumably understood Article III quite well, since several Members of Congress had participated in the drafting. And allowing Congress to take a case from the appellate category and make it original does not seem an outcome that threatens the rights or well-being of individuals, or the structure of our government, or the ability of the Court to perform its function in that system.

    [138].    See Roosevelt, supra note 10.

    [139].    See Pfander, supra note 130, at 1517 (noting that “[u]nder the dominant narrative, Marshall has been said to have manufactured the conflict between section 13 and Article III”).

    [140].    Marbury v. Madison, 5 U.S. 137, 147–48, 174–75 (1803).

    [141].    See, e.g., Michael Stokes Paulsen, Marbury’s Wrongness, 20 Const. Comm. 343, 355 (2003); Akhil Reed Amar, Marbury, Section 13, and the Original Jurisdiction of the Supreme Court, 56 U. Chi. L. Rev. 443, 445 (1989) (stating that since “alternative readings would leave the words of Article III with some ‘operation,’ it seems fair to criticize the Marbury Court for overstating the logical force of its preferred reading”).

    [142].    See Ames v. Kansas, 111 U.S. 449 (1884).

    [143].    Hart & Wechsler, The Federal Courts and the Federal System 69 (1953).

    [144].    Robert G. McCloskey, The American Supreme Court 40 (1960).

    [145].    Christopher L. Eisgruber, Marbury, Marshall, and the Politics of Constitutional Judgment, 89 Va. L. Rev. 1203, 1204 (2003).

    [146].    Stephen G. Breyer, Making Our Democracy Work 12, 16 (2010). See also, e.g., Paul J. Larkin, Jr., Essay: Philemon, Marbury, and the Passive-Aggressive Assertion of Legal Authority, 29 B.Y.U. J. Pub. L. 241, 241–242 (2014) (“a brilliant legal and political tour de force”).

    [147].    R. Kent Newmyer, John Marshall and the Heroic Age of the Supreme Court (2001).

    [148].    Louise Weinberg, Our Marbury, 89 Va. L. Rev. 1235, 1412 (2003).

    [149].    See Sanford Levinson, Why I Still Won’t Teach Marbury (Except in a Seminar), 6 U.Pa. J. Con. L. 588, 599 (2004); Jack M. Balkin & Sanford Levinson, Understanding the Constitutional Revolution, 87 Va. L. Rev. 1045, 1062–63 (2001).

    [150].    But see Paulsen, supra note 89, at 356 (stating that “[i]t is flat-out wrong to knowingly misuse judicial authority in a specific case in order to advance judicial power generally”).

    [151].    The Supreme Court often reaches results that align with the preferences of one or another political party and often the Justices voting in those cases vote to promote the preferences of the party of their appointing President. But there are plausible innocent explanations for this—political parties have constitutional visions and appoint Justices who share them. See generally Jack M. Balkin, Bush v. Gore and the Boundary Between Law and Politics, 110 Yale L.J. 1407, 1408–09 (2001). The Justices very seldom seem to be seeking to promote the power of a party for its own sake. The most notable exceptions are perhaps decisions about voting rights and regulations, which have a clear partisan effect, and of course the resolution of the 2000 presidential election. See Shelby County v. Holder, 570 U.S. 529 (2013); Crawford v. Marion County Election Bd., 553 U.S. 1818 (2008); Bush v. Gore, 531 U.S. 98 (2000).

    [152].    See, e.g., James Hohmann, The Daily 202: Lame-duck power grab in Wisconsin showcases the GOP’s embrace of zero-sum politics, The Washington Post (Dec. 4, 2018), https://www.washingtonpost.com/news/powerpost/paloma/daily-202/2018/12/04/‌daily-202-‌lame-duck-power-grab-in-wisconsin-showcases-the-gop-s-embrace-of-zero-sum-politics/5c05f6ff1b326b60d12800fe/?utm_term=.f0d27616082b.

    [153].    See Susan Low Bloch, The Marbury Mystery: Why did William Marbury Sue in the Supreme Court?, 18 Const. Comment. 607, 625–27 (2001).

    [154].    Id. at 627.

    [155].    Paulsen, supra note 89, at 357.

    [156].    Id. Paulsen’s answer to this question is to back off from the analysis that suggests Marshall deliberately manufactured a conflict between the Judiciary Act and Article III: “I would prefer to believe, perhaps against the evidence, that John Marshall simply announced in good faith a lot of legal propositions that I … think are mistaken.” Id. at 357. Agreeing with Paulsen’s assessment of the evidence, we have less difficulty believing the strategic narrative. In any case, the dominant narrative presents Marbury as a carefully crafted strategic decision, so, for the purposes of evaluating its place in the constitutional canon, we evaluate it as such.

    [157].    If one is trying to support Supreme Court decisions that one thinks are not actually simply attempts to reach the legally correct outcome, then Marbury has obvious appeal. As long as judges are doing what you want, the judge who skillfully manipulates legal materials to produce justice despite the law is a heroic figure. It seems likely that some law professors viewed the Warren Court this way. See Laura Kalman, The Strange Career of Legal Liberalism 6 (1996); Barry Friedman, The Birth of An Academic Obsession: The History of the Countermajoritarian Difficulty, Part Five, 112 Yale L. J. 153, 237–54 (2002). Herbert Wechsler, for instance, notoriously stated that although he approved of Brown on policy grounds, he could not justify it as constitutional law. See Friedman at 251. The problem for such professors came when judges started manipulating the materials to reach ends they did not like: having used Marbury to support activism on the left, they had no remaining ground from which to criticize the activism of the right.

    [158].    The list of cases in which one side invokes Marbury while the other invokes Lochner is substantial. Notable such decisions include NFIB v. Sibelius, 567 U.S. 519 (2012); McDonald v. City of Chicago, 561 U.S. 742 (2010); Kelo v. City of New London, 545 U.S. 469 (2005); College Sav. Bank v. Florida Prepaid Postsecondary Educ. Bd., 527 U.S. 666 (1999); Alden v. Maine, 527 U.S. 706 (1999); Washington v. Glucksberg, 521 U.S. 702 (1997); Seminole Tribe of Florida v. Florida, 517 U.S. 44 (1996); United States v. Lopez, 514 U.S. 549 (1995); Planned Parenthood v. Casey, 505 U.S. 833 (1992); and Griswold v. Connecticut, 381 U.S. 479 (1965).

    [159].    Jack N. Rakove, The Origins of Judicial Review: A Plea for New Contexts, 49 Stan. L. Rev. 1031, 1036 (1997).

    [160].    There are some law professors who do not teach Marbury. See Levinson, supra note 119, at 557. And Levinson’s casebook presents McCulloch as its first case, moving to Marbury only in the context of political struggles in the early Republic. See Paul Brest, et al., Processes of Constitutional Decisionmaking (7th ed. 2018).

    [161].    Evidence that destruction was Maryland’s goal can be found in the nature of the tax, which was large and discriminatory, worded to apply only to the Bank of the United States. Additionally, in McCulloch, Maryland argued that the Bank was unconstitutional—an argument that would not help it collect the tax but would, if it succeeded, destroy the Bank.

    [162].    See David S. Schwartz, Misreading McCulloch v. Maryland, 18 U. Pa. J. Con. L. 1, 8 (2015) (“McCulloch is presented as a ‘principal case’ in every current constitutional law casebook…”).

    [163].    See generally id., at 8–13 (describing conventional view of McCulloch as “aggressive nationalism”).

    [164].    Ian Bartrum, for example, pairs Marbury and Lochner as two sides of judicial review. He then considers McCulloch as an illustration of flexibility in interpretation. The next section of his article returns to the topic of judicial review; it is about representation and democracy. But rather than McCulloch, it goes directly to Brown and John Hart Ely. See Ian Bartrum, Constitutional Value Judgments and Interpretive Theory Choice, 40 Fla. St. L. Rev. 259 273–77 (2013). Sylvia Snowiss goes even further, arguing that Marbury is an example of principled judicial review and McCulloch is something more like judicial lawmaking. See Sylvia Snowiss, Text and Principle in John Marshall’s Constitutional Law: The Cases of Marbury and McCulloch, 33 J. Marshall L. Rev. 973 (2000); Stephen B. Presser, Marbury, McCulloch, Gore and Bush: A Comment on Sylvia Snowiss, 33 J. Mar. L. Rev. 1157, 1159 (2000).

    [165].    They have played some role. In the wake of United States v. Lopez, 514 U.S. 549 (1995), some scholars suggested that the Court should examine the process of congressional lawmaking, and in other federalism cases the Court has observed that the presence of relevant congressional findings may provide support for a law.

    [166].    McCulloch v. Maryland, 17 U.S. 316, 401 (1819).

    [167].    Id. at 431; see also id. at 435 (noting that “[t]he people of all the States, and the States themselves, are represented in Congress”).

    [168].    See U.S. Const. art. I. This process has, of course, been altered by the 17th Amendment.

    [169].    McCulloch, 17 U.S. at 435–436.

    [170].    Courts must decide, for instance, how important the government’s asserted goal is: legitimate, important, or compelling? They must also decide how closely tied to that goal the chosen means is: rationally, substantially, or necessarily.

    [171].    That said, you could actually make a pro-aggressive judicial review argument from McCulloch. Cases about economic regulation are individual rights cases, so they involve the great principles of liberty. And if economic regulations are pretextual—if they are not good faith attempts to advance the public interest but rather seek some impermissible motive—the pretext passage of McCulloch seems to authorize invalidation. See 17 U.S. at 423. This argument does not tell us what is an impermissible motive—Lochner’s anti-redistribution principle is no longer available—and it does not tell judges not to defer. But it may illustrate that McCulloch, too, can contain different meanings for different audiences. One could even argue that the 17th Amendment changes the calculus for review of the limits of federal power: now that state legislatures no longer select Senators, the Senate is much less of a check on unnecessary federal legislation.

    [172].    See United States v. Carolene Products, 304 U.S. 144, 152 n.4. (1938).

    [173].    There is an important difference between McCulloch and Carolene Products, of course. McCulloch was concerned that states would not give due weight to the interests of out-of-staters, a concern that existed at the Founding and is reflected in various provisions of the original Constitution, most notably though not exclusively Article IV. Carolene Products raises the concern that states will not give due weight to the interests of politically weak groups within the state. That concern is largely absent from the Founding; it emerges in the aftermath of the Civil War and enters the Constitution with the Reconstruction Amendments. Carolene Products, really, is the Reconstruction version of McCulloch.

    [174].    In Lopez, Chief Justice Rehnquist’s majority opinion invoked Marbury, see 514 U.S. at 566, while Justice Souter’s dissent responded with Lochner, id. at 605 (Souter, J., dissenting).

    [175].    McCulloch v. Maryland, 17 U.S. 316, 401 (1819).

    [176].    McCulloch does not, of course, support aggressive judicial review in cases where there is no plausible argument that aggressive review is appropriate. But that is perhaps the best reason to prefer it to Marbury.

    [177].    See Snowiss, supra note 92, at 1007 (noting that when Dred Scott was attacked, “[d]efense of judicial review drew on Marbury”). By contrast, in Dred Scott itself, McCulloch is used by Justice McClean’s dissenting opinion to criticize the majority. See 60 U.S. 393, 542 (1857) (McClean, J., dissenting).

    [178].    We have said that we do not think Marbury in fact contains a statement of judicial supremacy. Others agree. See, e.g., Eric Segall, Why I Still Teach Marbury (and So Should You): A Response to Professor Levinson, 6 U. Pa. J. Con. L. 573, 583 (2004). But judicial supremacy is not quite our concern. Judicial supremacy is generally taken to be the idea that the Supreme Court has the last word on the meaning of the Constitution. Accepting that for present purposes, there are still two important questions to be considered. First, what is the significance of the fact that some other governmental actor believes that what it is doing is consistent with that meaning? (The Supreme Court interprets the Equal Protection Clause to prohibit laws that stigmatize a disfavored group, for instance: how much does it matter that a legislature believes its law is not stigmatizing? More or less than it matters that Congress believes an activity in the aggregate substantially affects interstate commerce?) Second, to what extent are other government actors bound to follow not just the meaning of the Constitution as the Supreme Court has interpreted it, but the doctrine the Court has created to implement that meaning. (The Supreme Court uses rational basis review to assess state discrimination on the basis of age, for instance: can Congress decide that some state law stigmatizes the elderly and is therefore unconstitutional even if the Court would uphold it?). See generally Roosevelt, supra note 10 (distinguishing between doctrine and meaning). Both these questions involve constitutional decisionmaking outside the courts, but neither presents the question of judicial supremacy in its conventional form.

    [179].    See Brest, et al., supra note 160, at 29.

    [180].    See id. at 32–37.

    [181].    Id. at 34.

    [182].    This terminology is used by Mitchell N. Berman, Constitutional Decision Rules, 90 Va. L. Rev. 1 (2004), and Kermit Roosevelt III, Constitutional Calcification: How the Law Becomes What the Court Does, 91 Va. L. Rev. 1649 (2005).

    [183].    Eric Segall, defending Marbury’s inclusion in the first year curriculum, argues that McCulloch contains “little … about the Court’s relationship to the other branches of the national government,” while Marbury discusses “the Supreme Court’s role in relation to the other branches.” Segall, supra note 178, at 584. We think this is wrong; in fact, backwards. McCulloch does discuss the legislative process leading up to the passage of the Bank bill and its significance; preceding that discussion with Madison’s speech and the opinions of Washington’s Cabinet shows the operation of all three branches. By contrast, Marbury offers only simplistic slogans about the rule of law that ignore the role and competencies of the other branches.

    [184].    See, e.g., Michael J. Klarman, How Great were the “Great” Marshall Court Decisions, 87 Va. L. Rev. 1111, 1113–15 (2001); William Michael Treanor, Judicial Review Before Marbury, 58 Stan. L. Rev. 455 (2005).

    [185].    See § 25; See generally Mark A. Graber, Establishing Judicial Review: Marbury and the Judiciary Act of 1789, 38 Tulsa L. Rev. 609 (2003).

    [186].    See supra text accompanying notes 92–93.

    [187].    One way to do this, we suggest, is to present McCulloch as a model of good judicial behavior and plausible choices about deference and Marbury as an example of bad behavior and failure to address the issue of deference.