RBG Icon


Starting my clerkship on Justice Ruth Bader Ginsburg’s first day at the Supreme Court provided an exceptional opportunity to assist her as she embarked on a new chapter of her career. Although it was her first Term on the Supreme Court, the Justice’s approach to judging built on her thirteen-year tenure on the District of Columbia Circuit. Her opinions that first year provided some clues about how her jurisprudence would evolve over time, and about the legacy she would ultimately leave us.

In this Article I offer some personal reflections about Justice Ginsburg and my experience during the clerkship. In addition, I share some thoughts on her approach to criminal law and procedure—a field that has been the principal focus of my practice since the clerkship—and how it generally reflects her “gradualist” philosophy of judging, her pragmatism, and her commitment to preserving the rights of the accused.

I. Justice Ginsburg’s Approach to Judging (and Law Clerks)

At the time of her nomination to the Supreme Court, Justice Ginsburg had already had a remarkable and (to use one of her trademark phrases 1 See, e.g., Florida v. Powell, 559 U.S. 50, 53, (2010) (“In a pathmarking decision, Miranda v. Arizona, 384 U.S. 436, 471 (1966) . . . .”); Skinner v. Switzer, 562 U.S. 521, 533 (2011) (“Pathmarking here is Heck v. Humphrey, 512 U.S. 477 (1994).”). ) pathmarking career as a scholar, advocate, and jurist. Her contributions to the Constitution’s goal of creating “a more perfect union” and to equal justice under law had been momentous. She founded and led the ACLU’s Women’s Rights Project, crafting a legal strategy that led to a series of landmark Supreme Court rulings on gender equality. Her work fighting gender discrimination has been compared to Thurgood Marshall’s advocacy against race discrimination. Indeed, it was in large part because of her work as an advocate that women in my generation (and our daughters) have—and take for granted—so many opportunities that earlier generations were denied. Even if Justice Ginsburg had never sat on the Supreme Court, she would have had an important place in American history. 2 For a comprehensive look at the Justice’s advocacy work, see generally Ruth Bader Ginsburg & Amanda L. Tyler, Justice, Justice Thou Shalt Pursue: A Life’s Work Fighting for A More Perfect Union 3–11, 39–44, 49–93 (2021); Amy Leigh Campbell, Raising the Bar: Ruth Bader Ginsburg and the ACLU Women’s Rights Project, 11 Tex. J. Women & L. 157 (2002).

Not surprisingly, she sailed through the Senate process and was confirmed by a vote of 96-3 at a time when (in spite of the acrimonious Bork and Thomas hearings) it was still possible for the Senate to focus on the nominee’s qualifications.

During her swearing-in at the White House in August 1993, President Clinton described some of the reasons he had nominated Justice Ginsburg to the Supreme Court. Several points he made resonate well with my own memories of Justice Ginsburg during the clerkship. He said:

We breathe life through the values we espouse through our law. It gives to every American, including the most illiterate among us, the most totally unaware of how the legal system works, a fair measure of our ideals. There is no one with a deeper appreciation of this fact than Ruth Bader Ginsburg. . . .

She has emerged as one of our country’s finest judges: progressive in outlook, wise in judgment, balanced and fair in her opinions. She defied labels like liberal and conservative just as she did in her hearing before the Senate, to earn a reputation for something else altogether: excellence. 3 Supreme Court Justice Swearing-In Ceremony, C-SPAN, at 11:37–14:17 (Aug. 10, 1993), https://www.c-span.org/video/?48274-1/supreme-court-justice-swearing-ceremony (on file with the Columbia Law Review).

Taking the last comment first: Commensurate with her own reputation as a judge, the Justice expected one thing from her clerks: excellence. She was, perhaps, demanding in that regard, but to my mind the best kind of boss. There was no better feeling than learning that your work lived up to (or at least came close to) the Justice’s very high expectations.

She inspired us to master the details and to approach every case with analytical rigor and precision. She taught us the craft of legal writing—how to describe the facts of a case and one’s legal reasoning in a concise, simple and persuasive, but also interesting, way that keeps the reader engaged. These are lessons I still endeavor to apply today, albeit in the context of legal advocacy and brief writing.

Unlike with some other Justices, however, other expectations in chambers were more relaxed. The Justice’s unusual hours—she typically arrived in chambers shortly after lunchtime and often worked late into the night—are well known. But there were no set rules about when her clerks had to be there, as long as we got the work done and were available when she needed us. Casual dress was fine on days when Court was not in session. Her focus was on substance.

At the same time, the Justice took an interest in our personal lives. She invited us to her home to enjoy Marty’s fine cooking and performed many clerks’ wedding ceremonies (including my own). She took us to the opera. And when a law clerk or former law clerk had a baby, he or she was welcomed into the broader RBG clerk family by a t-shirt emblazoned with the Supreme Court seal and “RBG Grand Clerk.”

President Clinton’s first point, about the Justice’s appreciation for how the law expresses our ideals and should speak to all of us, also rings true. Justice Ginsburg’s understanding of those principles was reflected in her historic work as an advocate for equal justice before she became a judge, as well as how she approached cases as a judge and justice.

Of course, many of her most significant opinions were directed to affirming the principle of equal justice for all. 4 E.g., Shelby County v. Holder, 570 U.S. 529, 590 (2013) (Ginsburg, J., dissenting) (dissenting from the majority’s holding that Section 5 of the Voting Rights Act is unconstitutional); United States v. Virginia, 518 U.S. 515, 519 (1996) (holding “the Constitution’s equal protection guarantee precludes Virginia from reserving exclusively to men the unique educational opportunities [the Virginia Military Institute] affords”). But just as importantly, when it came to the parties in every case before her, she always strove to give every litigant—whether they were the President of the United States or an indigent criminal defendant—a fair hearing.

For instance, one of the things she taught us, which I think grew out of her own experience as an advocate, was that a judicial opinion should always address the losing party’s arguments forthrightly and explain why they lost. She would say, imagine you’re the losing litigant. Would you feel at least that you were heard and that your arguments got a fair shake? She believed that this was critical to the rule of law and to public acceptance of the legitimacy of judicial decisions. And her opinions certainly achieve that goal.

Finally, as President Clinton remarked, the Justice was indeed progressive in outlook while (at the time) defying ideological labels. This was true when she was an advocate and a D.C. Circuit Judge, and it was true during her tenure on the Court. She had a “gradualist” philosophy—a view that progress toward equal justice and greater liberty, when carried out by courts, is more likely to last when it is taken in small, careful steps rather than via radical changes.

The Justice famously described her views on judicial decisionmaking shortly before her nomination, when she delivered the Madison Lecture at NYU Law School in December 1992. 5 Ruth Bader Ginsburg, Madison Lecture: Speaking in a Judicial Voice, 67 N.Y.U. L. Rev. 1185 (1992). Citing Alexander Hamilton’s vision of the judiciary as the “least dangerous” branch of government, Justice Ginsburg opined that “the effective judge . . . strives to persuade, and not to pontificate,” and “speaks in ‘a moderate and restrained’ voice, engaging in a dialogue with, not a diatribe against, co-equal departments of government, state authorities, and even her own colleagues.” 6 Id. at 1186. As President Clinton said, she was progressive: She strived to expand the framers’ “distinctly limited vision” of those who counted among “We the People,” and she certainly viewed the Constitution as a living document and eschewed originalism. However, as she explained: “Measured motions seem to me right, in the main, for constitutional as well as common law adjudication. Doctrinal limbs too swiftly shaped, experience teaches, may prove unstable.” 7 Id. at 1198. In essence, the Court should not get too far out in front of the political branches—judicial decisions that are too bold or too radical are less likely to bring lasting change.

To illustrate her point, Ginsburg imagined how Roe v. Wade might have been decided differently. The Texas law the Court struck down under the Due Process Clause in Roe was extreme: It exempted from criminality only procedures to save a pregnant woman’s life and thus “intolerably shackled a woman’s autonomy.” 8 Id. at 1199. Suppose, Ginsburg observed, the Court had “stopped there” and simply declared “the most extreme brand of law in the nation” unconstitutional, rather than “fashion[ing] a regime blanketing the subject, a set of rules that displaced virtually every state law then in force.” 9 Id. Had the Court focused more on “the idea of the woman in control of her destiny and her place in society,” and refrained from embarking on such a bold effort to regulate the field, she suggested, Roe “might have been less of a storm center.” 10 Id. at 1199–200. The Justice felt that narrower decisions—such as the many rulings in the 1970s and early 1980s holding unconstitutional state and federal laws that expressly treated people differently based on their sex—led to more lasting progress, because they “did not utterly condemn the legislature’s product.” 11 Id. at 1204. Instead, the decisions “opened a dialogue with the political branches of government” and “tossed” the ball “back  into  the  legislators’  court, where  the  political  forces  of  the  day  could operate.” 12 Id. Justice Ginsburg made the same point a different way decades later when asked what advice she would give to her younger female admirers. She said: “My advice is fight for the things you care about . . . . But do it in a way that will lead others to join you.” Irin Carmon, Opinion, Justice Ginsburg’s Cautious Radicalism, N.Y. Times (Oct. 24, 2015), https://www.nytimes.com/2015/10/25/opinion/sunday/justice-ginsburgs-cautious-radicalism.html (on file with the Columbia Law Review).

II. Justice Ginsburg’s Criminal Law and Criminal Procedure Jurisprudence

This gradualism was a hallmark of the Justice’s approach to criminal law and criminal procedure as well. What follows is not a comprehensive or definitive analysis of the Justice’s views on criminal law and procedure, or a data-intensive quantitative plunge into her voting record on these issues. Instead, through a more anecdotal approach, I will delve into a few specific subjects—statutory interpretation, the Sixth Amendment jury trial and confrontation rights, and capital punishment—that are illustrative of what I believe was her general approach. 13 There are other interesting areas that I did not have space to cover here, such as habeas corpus, the Fourth Amendment, Miranda and the Fifth Amendment privilege against self-incrimination, and the Sixth Amendment right to counsel. Those are perhaps worthy subjects for a separate piece exploring other nuances of the Justice’s criminal procedure jurisprudence. In general, however, it is fair to say that one will find similar tendencies in these areas—that is, a gradual approach attentive to precedent, but one that prioritized fair process and protecting the rights of accused persons. I will discuss a few decisions on each of these subjects to illustrate how Justice Ginsburg’s opinions in this area reflect her “gradualist” approach as well as her pragmatism and concern for protecting the constitutional rights of the accused.  I also endeavor to capture a flavor of her voice as a writer.

A. Interpretation of Criminal Statutes

Justice Ginsburg’s emphasis on cautious rulings and the importance of judicial dialogue with the political branches is a recurrent theme of the opinions that she wrote or joined interpreting federal criminal statutes. Her opinions generally construed criminal statutes narrowly. She often rejected the government’s expansive interpretations of penal laws with malleable language, citing the importance of fair notice and avoiding arbitrary enforcement. In so doing, the Justice frequently invoked the strong form of the rule of lenity, which holds that when there is textual ambiguity in a criminal statute, it should be interpreted in favor of criminal defendants. 14 See, e.g., Adamo Wrecking Crew v. United States, 434 U.S. 275, 285 (1978) (describing “the familiar rule that, ‘where there is ambiguity in a criminal statute, doubts are resolved in favor of the defendant’” (quoting United States v. Bass, 404 U.S. 336, 348 (1971))). By contrast, in its weaker form, the rule of lenity is to be invoked only as a last resort. Compare Muscarello v. United States, 524 U.S. 125, 138–39 (1998) (holding that the rule of lenity applies only if a “grievous ambiguity or uncertainty in the statute” exists after resorting to all the tools of statutory interpretation (citations omitted) (internal quotation marks omitted)), with id. at 148 (Ginsburg, J., dissenting) (criticizing the majority for its treatment of the rule of lenity and noting that the “sharp division on the Court on the proper reading” of the statute itself illustrated the requisite ambiguity). She also occasionally raised federalism concerns to reject broad applications of vague criminal laws in areas traditionally governed by state and local crimes. In addition, the Justice was attentive to the importance of scienter, and usually took a defendant-friendly approach to questions about the level of mens rea the government had to prove under federal statutes. And she did not hesitate to endorse a construction that increased the government’s burden of proof, even when doing so was not the most natural reading of the statutory text. Her approach was pragmatic and not strictly or formalistically textualist. At the same time, her opinions in this area were limited in scope, reasoned from well-established techniques of statutory interpretation, and sharply focused on what needed to be decided. Finally, Justice Ginsburg was somewhat reluctant to strike down even notoriously amorphous criminal statutes as unconstitutionally vague—she preferred to salvage overbroad statutes through narrowing construction. 15 There are some notable recent exceptions to the Justice’s hesitance to invalidate statutes under the vagueness doctrine. For instance, Justice Ginsburg joined majority opinions holding that a provision of the Armed Career Criminal Act violates due process in Johnson v. United States, 576 U.S. 591 (2015), and that a similar civil immigration law was unconstitutionally vague in Sessions v. Dimaya, 138 S. Ct. 1204 (2018). The Justice also joined the 5-4 majority in United States v. Davis, 139 S. Ct. 2319 (2019), which applied Johnson and Dimaya to invalidate another analogous criminal offense. Id. at 2320.

In a few discrete areas, by contrast, the Justice was somewhat more accepting of weak government arguments, perhaps due to her bent in the particular regulatory area (e.g., securities) or concerns about how a ruling for the defendant might impact other areas of  the  law  (e.g.,  deference  to  administrative  agencies).  I  touch  on  one  example below. 16 It is notable that (apart from perhaps in death penalty cases) some of the “liberal” justices with whom Justice Ginsburg overlapped voted for the government’s position more frequently, especially on matters of statutory interpretation and the Sixth Amendment issues discussed infra. Indeed, she was frequently aligned with Justice Scalia on many criminal law and procedure issues, though as we shall see, to Justice Ginsburg practical solutions were more important than literal textualism in statutory construction.

1. Clues from the First Term: Ratzlaf, Staples, X-Citement Video. — Three decisions during my clerkship (Justice Ginsburg’s first Term on the Court) provided some indication of how she would approach the interpretation of criminal statutes. These cases illustrate her insistence on a robust scienter element to ensure that people are not prosecuted for conduct they may not have known was illegal. Her opinions and votes in these cases likewise reflect her careful and systematic reasoning and use of traditional canons of statutory construction to resolve the question presented without sweeping pronouncements that might unduly hamstring further legislative action.

Ratzlaf v. United States was one of the first Supreme Court decisions Justice Ginsburg authored, and it exemplifies her approach. 17 510 U.S. 135 (1994). The defendant wanted to pay a casino debt with $100,000 in cash. A casino employee told him that all transactions of over $10,000 in cash had to be reported to the authorities, but that the casino would accept a cashier’s check for the full amount. He went to a nearby bank and, after being informed that banks also had to report cash transactions exceeding $10,000, he purchased multiple cashier’s checks of less than $10,000 at different banks and delivered them to the casino. 18 Id.

Ratzlaf was prosecuted under a federal statute making it a crime to “willfully” “structure” transactions—i.e., to break up a single cash transaction of more than $10,000 into smaller amounts in order to avoid triggering a requirement that financial institutions must report the transaction to the Secretary of the Treasury. The antistructuring statute Ratzlaf was charged with violating, 31 U.S.C. § 5324(3), provided: “No person shall for the purpose of evading [certain reporting] requirements . . . structure or assist in structuring, or attempt to structure or assist in structuring, any transaction with one or more domestic financial institutions.” 19 31 U.S.C. § 5324(3) (1988). In turn, 31 U.S.C. § 5322(a) provided for criminal penalties of up to five years’ imprisonment for “willfully violating” Section 5324. 20 Id. § 5322(a). The question presented was, what does “willfully” mean? The government argued that proof that the defendant’s purpose was to circumvent a bank’s reporting obligation was sufficient to demonstrate willful conduct. Ratzlaf, on the other hand, admitted that he structured cash transactions and did so with knowledge of, and the intent to avoid, the banks’ reporting obligations. However, he maintained that willfulness required the government to prove that the defendant knew that it was illegal to engage in structuring to avoid the reporting requirement. With Justice Ginsburg writing for a 5-4 majority, the Court agreed with the defendant and held that the government had to prove that Ratzlaf “acted with knowledge that his conduct was unlawful.” 21 Ratzlaf, 510 U.S. at 137.

The opinion carefully parses the statutory text and structure, and how related provisions had been construed. The reasoning starts with the text, noting that Section 5324 itself requires a “purpose of evading” the reporting requirements, such that, under the government’s reasoning, the willfulness requirement of Section 5322(a) would be surplusage. 22 Id. at 139–40. Treating statutory terms as such is generally disfavored, the Justice notes, “and resistance should be heightened when the words describe an element of a criminal offense.” 23 Id. at 140­­–41. Turning to structure, the Justice found it significant that the same “willfulness” provision—which applies to “other provisions in the same subchapter,” including several provisions similar to Section 5322(a)—had been consistently interpreted to require proof of an intentional violation of a “known legal duty” or similar formulations effectively requiring knowledge of illegality. 24 Id. at 141–42.

Significantly, Justice Ginsburg rejected the government’s argument that the violation of Section 5324 necessarily “exhibit[s] a purpose to do wrong” because “[s]tructuring is not the kind of activity that an ordinary person would engage in innocently.” 25 Id. at 143–44. Not so, according to the Justice. Though the opinion eschews grand pronouncements about due process and fair notice, it is clear that such concerns underlie the Justice’s reasoning. Structuring is not “inevitably nefarious,” she explained, using examples such as a person who was trying to reduce the risk of an IRS audit, or who is fearful the bank’s reports might increase the likelihood of burglary, or who is minimizing the possibility of a former spouse learning of his wealth. 26 Id. Such examples illustrate that structuring is not so “inherently ‘bad’ that the ‘willfulness’ requirement is satisfied irrespective of the defendant’s knowledge of the illegality of structuring.” 27 Id. at 146.

Finally, while recognizing that the legislative history is a bit of a mixed bag, the Justice concluded that “we do not resort to legislative history to cloud a statutory text that is clear,” and that even if the text were ambiguous, “we would resolve any doubt in favor of the defendant” under the rule of lenity. 28 Id. at 147–48. In this case, the rule of lenity is not critical, or even strictly necessary to the outcome. Instead, Justice Ginsburg merely used it to respond to the government’s arguments that legislative history supported its position. 29 See id. at 147 n.17.

Ratzlaf is particularly notable because the Court held that the government had to prove specific knowledge of unlawfulness, despite the oft-cited maxim that “ignorance of the law is no excuse.” In two other decisions that first Term, Justice Ginsburg voted in favor of similar defendant-friendly statutory interpretations consistent with the view that judges should generally presume that Congress intended to require a sufficient level of scienter to avoid a trap for the unwary, and should hew to text, structure, and context, while bearing lenity principles in mind. These latter cases demonstrate her strong view that—at a minimum—courts should typically presume that Congress required proof that the defendant knew the facts that made what might otherwise be innocent conduct illegal.

For instance, in Staples v. United States, the Court implied a knowledge requirement into a criminal statute that was silent on the mens rea required to prove a violation. 30 511 U.S. 600 (1994). The statute imposed strict registration requirements on a defined category of “firearm,” including “machinegun,” a term defined to include any “weapon which shoots, . . . or can be readily restored to shoot, automatically more than one shot, without manual reloading.” 31 26 U.S.C. § 5845(a)(6), (b) (1988). Thus, a semi-automatic weapon that had been converted into a machinegun must be registered under the provision. Failing to register a machinegun is a crime punishable by up to ten years’ incarceration. 32 Id. §§ 5861(d), 5871. The defendant in Staples was charged with unlawful possession of an unregistered machinegun after ATF agents recovered an AR-15 semi-automatic rifle which, agents testified, fired more than one shot with a single pull of the trigger. The defendant testified that the rifle had never fired automatically when in his possession and requested a jury instruction requiring the government to prove that he “knew that the gun would fire fully automatically.” The district court rejected the proposed instruction and instead charged the jury that the government was not required to prove that the defendant knew the weapon possessed all the characteristics that subjected it to regulation. Instead, the government only had to prove that “he knows he is dealing with a dangerous device of a type as would alert one to the likelihood of regulation.” 33 Staples, 511 U.S. at 603–04.

The Court, in an opinion authored by Justice Thomas, held that the government was required to prove that Staples knew that the weapon had been converted into a machinegun. The Court emphasized that criminal statutes are to be construed in light of background common law rules, including the presumption in favor of inferring a mens rea requirement unless it is clear that Congress intended to dispense with this element. In United States v. Freed, the Court had held that in a prosecution under the same statute for possession of unregistered grenades, it was unnecessary to show that the defendant had knowledge of the registration requirement. 34 See 401 U.S. 601, 607 (1971). The government cited Freed’s statement that the statute was a public safety regulatory measure. It argued that “one would hardly be surprised to learn that possession of hand grenades is not an innocent act,” 35 Id. at 601, 609. and that machineguns are also “highly dangerous.” 36 See Staples, 511 U.S. at 610. The Court rejected these arguments and discussed at some length the “long tradition of widespread lawful gun ownership by private individuals in this country.” 37           Id. This tradition, in the Court’s view, meant that even though some guns have “perhaps even greater” “destructive potential” than hand grenades, individuals would lack sufficient notice of the likelihood of regulation, making it inappropriate for the Court to dispense with the requirement that a defendant know the weapon’s characteristics. 38 Id. at 612. Although Staples pre-dates District of Columbia v. Heller, 39 554 U.S. 570 (2008). Heller was a sea change from long-standing Supreme Court precedent that had rendered the Second Amendment largely toothless. and the majority does not expressly mention the Second Amendment, the opinion hints at a very broad notion of individuals’ rights to bear arms. 40 Staples, 511 U.S. at 610.

Justice Ginsburg wrote a separate opinion (joined by Justice O’Connor) concurring in the judgment. 41 See id. at 620 (Ginsburg, J., concurring in the judgment). Justice Ginsburg began by emphasizing that even though “knowingly” did not appear in the statute’s text, “courts generally assume (absent a contrary indication) that Congress means to retain a mens rea requirement.” 42 Id. Although she did not articulate any specific disagreement with the majority and cited portions of its opinion with approval, the Justice’s opinion is narrower and avoids the sorts of sweeping pronouncements about gun rights peppered throughout Justice Thomas’s opinion. For instance, as explained below, Justice Ginsburg’s opinion focuses on the fact that Congress and the States have chosen to regulate “only a very limited class of firearms,” thus signaling that the political branches were free to enact additional regulation in the future. 43 Id. at 622.

Justice Ginsburg pointed out that Freed’s holding that the statute does not require proof of knowledge that the firearm was unregistered “rested on the premise that the defendant indeed knew the items he possessed were hand grenades.” 44 Id. at 620–21. Thus, the government conceded that the defendant had to know that he possessed the firearm, and that he knew it was a “dangerous weapon.” However, the government maintained it did not have to prove “knowledge, beyond dangerousness, of the characteristics that render the object subject to regulation, for example, awareness that the weapon is a machinegun.” 45 Id. at 621. The government argued that its reading avoided “criminalizing ‘apparently innocent conduct’” by shielding a defendant who believed what he possessed was a toy or violin case, but which in fact was a machinegun, from conviction. The Justice found this unpersuasive because it failed to “take adequate account of the ‘widespread lawful gun ownership’ Congress and the States have allowed to persist in this country.” 46 Id. (emphasis added). Because the “Nation’s legislators chose to place under a registration requirement . . . those they considered especially dangerous,” Justice Ginsburg agreed with the majority that the “generally ‘dangerous’ character of all guns . . . did not suffice to give individuals in Staples’ situation cause to inquire about the need for registration.” 47 Id. at 622 (emphasis added). Accordingly, she found that requiring the defendant to know that the object was in fact a machinegun was the only reading of the statute which “suits the purpose of the mens rea requirement—to shield people against punishment for apparently innocent activity.” 48 Id.

Additionally, exemplifying her penchant for carefully parsing the issues and explaining why she rejected the losing party’s arguments, the Justice pointed out that the indictment charged that Staples “knowingly received and possessed firearms” and that, since the “firearm” in question was a “machinegun,” this “effectively charged that Staples knowingly possessed a machinegun.” 49 Id. at 623. Thus, the only way for the government to reconcile the indictment with the jury instruction was on the “implausible” assumption that “firearm” had two different meanings when “used once in the same charge—simply ‘gun’ when referring to what petitioner knew, and ‘machinegun’ when referring to what he possessed.” 50 Id.

Finally, in her first Term Justice Ginsburg also joined the majority in United States v. X-Citement Video, Inc. 51 513 U.S. 64 (1994). —another decision in which the Court required a higher standard of scienter than that advocated by the government—in this case rejecting even the most straightforward reading of the relevant statutory language. The statute at issue, 18 U.S.C. § 2252(a), makes it a crime to:

(1) knowingly transport[] or ship[] in interstate or foreign commerce by any means including by computer or mails, any visual depiction, if—(A) the producing of such visual depiction involves the use of a minor engaging in sexually explicit conduct; and (B) such visual depiction is of such conduct; [or]

(2) knowingly receive[], or distribute[], any visual depiction that [travels in interstate commerce] if—(A) the producing of such visual depiction involves the use of a minor engaging in sexually explicit conduct; and (B) such visual depiction is of such conduct . . . . 52 18 U.S.C. § 2252(a) (1994).

In a 7-2 ruling, the Court rejected the “most natural grammatical reading” of the statute, which suggests “knowingly” modifies only the surrounding verbs (transport, ship, receive, etc.). 53 See X-Citement Video, 513 U.S. at 68. Instead, the Court held that “knowingly” also applies to the “involves the use of a minor” element. 54 Id. at 73. In other words, the defendant is guilty only if the government proves that he knew the pornography involved the depiction of a minor. As in Staples—and the precedents Justice Ginsburg cited in her Ratzlaf and Staples opinions—the Court relied on case law disfavoring interpretations that would “criminalize a broad range of apparently innocent conduct,” which it found particularly  important  given  that  the  statute  at  issue  was  not  a  public  welfare offense. 55 Traditionally, criminal liability was reserved for conduct involving “an evil-meaning mind with an evil-doing hand.” Morissette v. United States, 342 U.S. 246, 251 (1952). The term “public welfare offenses” refers to criminal penalties inflicted for violation of more modern regulations “which heighten the duties of those in control of particular industries, trades, properties or activities that affect public health, safety or welfare.” Id. at 254. The Court also noted the importance of other constitutional avoidance concerns, given that “nonobscene, sexually explicit materials” involving adults are protected by the First Amendment, such that “the age of the performers is the crucial element separating legal innocence from wrongful conduct.” 56 X-Citement Video, 513 U.S. at 72–73.

2. Later Decisions Interpreting Federal Fraud and Obstruction Statutes. — In several significant later decisions, the Justice authored majority opinions and joined other decisions sharply curtailing the government’s efforts to deploy the mail and wire fraud statutes as expansive, virtually limitless tools. These fraud statutes have long been a favored weapon in the government’s arsenal because of their malleability. Time and again, federal prosecutors have used them broadly, including at times to punish conduct that may be unethical or distasteful, but which is neither obviously criminal nor fits traditional common law conceptions of fraud. In a series of decisions beginning in the late 1980s with McNally v. United States, 57 483 U.S. 350 (1987). however, the Supreme Court has repeatedly rebuffed such efforts and insisted on a narrow interpretation of these statutes. Unfortunately, despite such decisions, prosecutors continue to try to expand the statutes—and lower courts often acquiesce—until once again the Supreme Court intervenes.

The Justice’s opinions in this area are consistent with the defendant-protective but cautious approach of her first Term decisions.

a. Property Fraud: McNally and Cleveland. — A bit of context: McNally was decided a few years before Justice Ginsburg’s appointment. But the decision—written by her predecessor, Justice White—hits on some of the same themes the Justice herself would later espouse, such as fair notice, lenity, federalism, and putting the ball back in Congress’s court should it wish to regulate more conduct. The mail and wire fraud statutes criminalize participating in any “scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations or promises” in which mails or interstate wires are used. 58 18 U.S.C. §§ 1341, 1343 (2018) (emphasis added). Most lower courts had endorsed prosecutions for “honest services fraud” that enabled the federal government to criminalize a broad range of conduct involving  some breach of duty by a public official. The conduct could involve not only bribes or kickbacks (or solicitations or offers of such), but also less obviously wrongful behavior, such as failing to disclose some conflict of interest.

McNally concerned mail fraud charges against a former public official and a private party for participating in “a self-dealing patronage scheme” to defraud Kentucky and its citizens “of certain ‘intangible rights,’ such as the right to have [Kentucky’s] affairs conducted honestly.” 59 McNally, 483 U.S. at 352. The scheme involved steering insurance contracts to a company in exchange for a commission-sharing arrangement. The Court held that the mail fraud statute does not encompass schemes to deprive persons of intangible rights such as the right to honest services, and instead is limited by the “money or property” language to schemes to obtain money or traditional, common law property. 60 Id. at 357. Despite the statute’s disjunctive text (prohibiting a “scheme or artifice to defraud, or for obtaining money or property” by deceptive means), the Court held that Congress intended to limit the statute to “money or property schemes”; “the second phrase simply made it unmistakable that the statute reached false promises and misrepresentations as to the future as well as other frauds involving money or property.” 61 Id. at 359. The Court reasoned from the history of common law fraud and its own cases interpreting the statute, as well as the lenity principle, that “when there are two rational readings of a criminal statute, one harsher than the other, we are to choose the harsher only when Congress has spoken in clear and definite language.” 62 Id. at 359­–60. Although used in McNally to support a more lenient interpretation of the statute, this formulation of the rule of lenity sets an oddly high bar. After all, if Congress “has spoken in clear and definite language,” it is hard to imagine how there could be “two rational readings” of the statute. But as in some other cases in which the Court invokes lenity principles to interpret facially ambiguous statutory text in favor of defendants, McNally unfortunately uses imprecise language. The Court also cited federalism concerns raised by construing “the statute in a manner that leaves its outer boundaries ambiguous and involves the Federal Government in setting standards of disclosure and good government for local and state officials.” 63 Id. at 360. In light of these principles, the Court read the statute as “limited in scope to the protection of property rights. 64 Id. And, in language Justice Ginsburg would later hearken back to, the Court pointed out: “If Congress desires to go further, it must speak more clearly than it has.” 65 Id. Soon thereafter, Congress accepted this invitation and enacted an “honest services fraud” offense that spawned many additional judicial debates. 66 See 18 U.S.C. § 1346 (1988) (“[T]he term ‘scheme or artifice to defraud’ includes a scheme or artifice to deprive another of the intangible right of honest services.”). We shall return to that saga later.

Unsurprisingly, meanwhile, and largely unchecked by the courts of appeals, the government continued to push the boundaries of “money or property” fraud. In 2000, Justice Ginsburg authored Cleveland v. United States, a unanimous decision rebuffing those efforts and reinforcing McNally’s teaching that the phrase “money or property” refers to traditional, common law property. 67 531 U.S. 12 (2000). At issue was a mail fraud charge based on allegations that the defendant had made false statements on an application for a Louisiana license to operate video poker machines. He argued that the state had no property interest in the license and accordingly, that the alleged fraud could not deprive it of any “property.” 68 Id.

The Court agreed, concluding that such permits or licenses “do not qualify as ‘property’ within § 1341’s compass.” 69 Id. at 15. In her opinion, Justice Ginsburg applied McNally to hold that the statute “does not reach fraud in obtaining a state or municipal license of the kind here involved, for such a license is not “property” in the government regulator’s hands.” 70 Id. at 20. She reasoned that whatever interest Louisiana might have in its poker licenses, “the State’s core concern is regulatory.” 71 Id. It had not itself “venture[d] into the video poker business,” and its regime “resembles other licensing schemes long characterized by this Court as exercises of state police powers.” 72 Id. at 21, 24.

Justice Ginsburg’s reasoning also illuminates her preference for pragmatic, rather than strictly traditionalist, statutory interpretation. Consistent with themes alluded to in those first Term decisions and in McNally itself, the Justice emphasized that the Court was rejecting the government’s theories “not simply because they stray from traditional concepts of property,” but also because accepting them would undermine principles of federalism as well as fair notice: The government’s “reading,” Justice Ginsburg observed, “invites us to approve a sweeping expansion of federal criminal jurisdiction in the absence of a clear statement by Congress.” 73 Id. at 24. That would “subject to federal mail fraud prosecution a wide range of conduct traditionally regulated by state and local authorities.” 74 Id. The Justice pointed out that the government’s interpretation of the statute “would appear to arm federal prosecutors with power to police false statements in an enormous range of submissions to state and local authorities.” 75 Id. at 24, 26. And, once again, she invoked the rule of lenity, stressing that the Court should not choose “the harsher alternative” between two plausible readings of a criminal  statute  unless  Congress  has  “spoken  in  language  that  is  clear  and definite.” 76 Id. at 24­–25. Finally, consistent with her preference for finding a narrow construction to avoid holding a statute unconstitutional—which would leave the legislative branch less room to maneuver—the Justice reiterated McNally’s invitation to Congress to “speak more clearly” if it “desires to go further.” 77 Id. at 20. Here, unlike with McNally, Congress has not seen fit to respond, and has not expanded the definition of “property” under the fraud statutes.

b. Honest Services Fraud: Skilling and Its Aftermath. — In the wake of McNally, Congress swiftly enacted 18 U.S.C. § 1346, which provides that “the term ‘scheme or artifice to defraud’ includes a scheme or artifice to deprive another of the intangible right of honest services.” 78 18 U.S.C. § 1346 (1988). This amorphous phrasing unleashed a wave of prosecutions. The result was that many lower courts blessed the prosecutorial creation of what amounted to a federal regulatory code of behavior for state and local officials (as well as private citizens)—and filled pages of the federal reporter with a smorgasbord of divergent judicial views on just what, exactly, constituted a deprivation of “the intangible right of honest services.” All of this eventually came to a head some 20 years later, in Skilling v. United States. 79 561 U.S. 358 (2010). In this landmark decision, the Court, with Justice Ginsburg writing for the majority, rebuffed an exceedingly persuasive push to jettison the statute as unconstitutionally vague. Instead, she went to great lengths to save the statute through a narrowing construction—such great lengths, in fact, that her opinion drew a sharp rebuke from Justice Scalia (joined by Justices Thomas and Kennedy), who said her analysis was not “interpretation but invention.” 80 Id. at 422 (Scalia, J., concurring in part and concurring in the judgment in part).

The Justice’s opinion, and her debate with Justice Scalia, is a textbook example of her gradualist philosophy as well as her more pragmatic approach. She understood that the statute was constitutionally problematic but strained to preserve it due to her reluctance to overstep the judicial role (and, I suspect, because she was loath to toss out an anti-corruption statute that had become an important tool for federal law enforcement). The Justice clearly recognized that the statute’s language was facially so broad that it raised the twin concerns that the vagueness doctrine is directed to alleviate: (1) failing to provide fair notice of what the statute proscribes, which risks creating a trap for the unwary; and (2) enabling prosecutors, law enforcement officers, and juries to engage in arbitrary enforcement. 81 To satisfy due process, “a penal statute must define the criminal offense with sufficient definiteness that ordinary people can understand what conduct is prohibited and in a manner that does not encourage arbitrary and discriminatory enforcement.” Kolender v. Lawson, 461 U.S. 352, 357 (1983). Yet on its face, there was no obvious narrowing construction of the sort implicitly adopted in cases like McNally, nor was there a traditional solution such as implying a mens rea requirement, or using an arguably unnatural, but still plausible, reading of statutory language to support a more lenient textual interpretation. The result is a rather strained effort at “interpretation” that, at least to me, makes Justice Scalia’s criticism fairly persuasive. But the opinion bears the hallmark of Ginsburg’s typical, tightly reasoned work. It may deviate from textualism, but it relies upon a careful analysis of the history that led to the enactment of the statute and points to traditional canons of constitutional avoidance to save a statute that Congress plainly viewed as an important weapon for battling corruption.

Skilling itself was not actually a public corruption case. It arose from the prosecution of the CEO of Enron on multiple counts of securities fraud and other charges relating to that company’s spectacular collapse. The charges included a conspiracy count, which alleged, among other things, a scheme to deprive the company and its shareholders of “the intangible right of [Skilling’s] honest services.” 82 Skilling, 561 U.S. at 358. This charge was based upon Skilling’s participation in efforts to deceive the investing public about Enron’s true financial condition, by manipulating the company’s publicly reported financial results and making other false and misleading statements about its financial performance. The theory was that Skilling’s fiduciary breach amounted to a criminal fraud under the statute. Skilling claimed the honest-services fraud statute was unconstitutionally vague. He argued that the phrase “intangible right of honest services” does not adequately define what conduct it prohibits and invites opportunistic and arbitrary prosecutions. 83 Id. at 359, 399, 403.

Justice Ginsburg, in an opinion joined by Chief Justice Roberts and Justices Stevens, Breyer, Alito, and Sotomayor, disagreed. 84 Skilling also resolved a question about whether the district court’s refusal to grant a motion to transfer venue away from Houston violated the Sixth Amendment’s right to trial by impartial jury. The Court rejected Skilling’s argument on that question in another section of the opinion that cut back the scope of some old mid-twentieth century precedents, over the dissent of Justices Sotomayor, Stevens, and Breyer. Id. at 384. Issues of jury selection and juror bias are outside the scope of this Article, however, so my focus here is solely on the honest-services fraud issue. She emphasized that Skilling’s argument “swims against our case law’s current, which requires us, if we can, to construe, not condemn, Congress’ enactments.” 85 Id. at 403. The Circuits had divided on how best to interpret the statute, but had uniformly “declined to throw [it] out as irremediably vague,” and the Justice agreed “that § 1346 should be construed rather than invalidated.” 86 Id. at 403–04.

To determine the meaning of “intangible right to honest services,” Justice Ginsburg looked to doctrine developed in the lower courts pre-McNally and then endeavored to “pare that body of precedent down to its core” and thereby “preserve what Congress certainly intended the statute to cover.” 87 Id. at 404. Skilling argued that it was impossible to identify “a salvageable honest-services core” from the pre-McNally case law because it comprised a “hodgepodge of oft-conflicting holdings.” 88 Id. at 406–07. The Justice conceded that some applications of the honest-services doctrine had led to disagreement among the Circuits but insisted that most cases involved “bribes and kickbacks.” 89 Id. at 407. Indeed, she pointed out, McNally itself—the catalyst for § 1346—“presented a paradigmatic kickback fact pattern.” 90 Id. This history showed that “Congress intended § 1346 to reach at least bribes and kickbacks”; and because a broader reading would raise the due process concerns underlying the vagueness doctrine, the Court would confine § 1346 to bribes and kickbacks. 91 Id. at 408.

In so doing, the Court also rejected the government’s effort to add another category of proscribed conduct: undisclosed self-dealing by a public official or private employee. McNally involved actual kickbacks, not mere failure to disclose a conflict of interest. According to Justice Ginsburg, there were some pre-McNally cases involving such nondisclosure schemes, but courts “reached no consensus on which schemes qualified.” 92 Id. at 410. And regardless, any ambiguity should be resolved in favor of lenity, and it would be up to Congress to speak more clearly if it wanted to go further. 93 Id. at 411.

Justice Scalia’s concurrence in the judgment made a powerful case that the honest-services-fraud statute is vague and thus violates the Due Process Clause. In his view, the Court was “strik[ing] a pose of judicial humility” by proclaiming that its job was “not to destroy the Act . . . but to construe it,” yet actually usurping a power reserved to the legislature—the power to define federal crimes—by “transforming” an honest-services-fraud prohibition into a bribery and kickback ban. 94 Id. at 415 (concurring in part and concurring in the judgment). A statute that is unconstitutionally vague, asserted Justice Scalia, cannot be saved by “judicial construction that writes in specific criteria that its text does not contain.” 95 Id. at 416.

Unlike Justice Ginsburg, he agreed with the defendant that pre-McNally case law was all over the map. The cases were not just about bribes and kickbacks involving public officials; some courts had held that private individuals who merely participated in public decisions could be public fiduciaries. 96 Id. at 417. Moreover, none of the “honest services” decisions “defined the nature and content of the fiduciary duty central” to the offense. 97 Id. There was even disagreement as to the source of the duty—state vs. federal law? “General principles”? 98 Id. The duty itself “remained hopelessly undefined,” with many formulations using “astoundingly broad language” (“moral uprightness,” “fundamental honesty,” “fair play and right dealing,” not “contrary to public policy”). 99 Id. at 418. Similar disagreements characterized cases involving private employees. 100 Id. at 417­–19.

Justice Scalia’s view was that Justice Ginsburg’s effort to find “core” conduct in the pre-McNally cases was wishful thinking because those cases (like decisions after § 1346’s enactment) “provide no clear indication of what constitutes a denial of the right of honest services.” 101 Id. at 420 (footnote omitted). He contended that the bribery and kickback limitation was “a dish the Court has cooked up all on its own.” 102 Id. at 422. He pointed out that a limiting construction to avoid constitutional problems must be a “fairly possible” one. 103 Id. at 423. Yet no court had concluded before McNally that “‘deprivation of honest services’ meant only the acceptance of bribes or kickbacks.” 104 Id. He agreed that if possible, the Court should save the statute, but did not “believe we have the power, in order to uphold an enactment, to rewrite it.” 105 Id. at 424.

The debate between the two Justices is interesting because both of them tended, as a general matter, to construe criminal statutes narrowly, and they often (though not always, as our discussion of Yates below illustrates) voted the same way on such interpretive questions. The disagreement here, however, perhaps reflects that Justice Scalia was more of a pure textualist, to whom the idea of rewriting the statute was anathema. By contrast, Justice Ginsburg was more pragmatic and less of a literal textualist. The statute had been on the books for over twenty years, and she undoubtedly was hesitant to invalidate it after all that time, and in the face of so many court of appeals decisions rejecting attacks on its vagueness. Of course, had the Court done so, it certainly could have found a way to guide Congress toward a fix, for instance, by pointing out that a fraud statute directed at bribes and kickbacks would have passed muster. But perhaps it would have been difficult to find five votes for such a ruling.

In any event, Justice Ginsburg’s voting record in other cases involving public corruption likewise reflects her general tendency toward lenity and narrow construction of criminal offenses. For instance, in her last Term she joined a unanimous decision spurning the government’s efforts to expand the property fraud statutes as an end-run around Skilling’s limitations on honest-services fraud. That case, Kelly v. United States, involved a prosecution arising from the “Bridgegate” scandal. 106 140 S. Ct. 1565 (2020). The defendants were public officials in New Jersey Governor Chris Christie’s administration.  They took actions to retaliate against a mayor who had refused to endorse Christie’s reelection bid by interfering with access to George Washington Bridge lanes and lying about why they did so (they falsely claimed it was for a traffic study). 107 Id. at 1568. There was no bribe or kickback, so this could not be charged as honest-services fraud under Skilling. Instead, the government obtained convictions on the theory that the defendants had fraudulently deprived the agency that operated the bridge of its “property.” 108 Id. at 1572.

The Supreme Court reversed the convictions, drawing heavily on Justice Ginsburg’s opinion in Cleveland. 109 Id. at 1570–74. The Court explained that the government’s exercise of the “rights of allocation, exclusion, and control” to advance a regulatory objective “do[es] ‘not create’” a government “property interest.” 110 Id. at 1572–73 (quoting Cleveland v. United States, 531 U.S. 12, 23 (2000)). And even though the defendants’ conduct had caused the government to pay for unnecessary employee labor related to the lane realignment, that economic harm was not sufficient: A conviction under the property-fraud statutes “cannot stand,” the Court explained, if economic loss to the victim “is only an incidental byproduct of the scheme.” 111 Id. at 1573. The Court emphasized, as it had in Cleveland and McNally, that “[t]o rule otherwise would undercut this Court’s oft-repeated instruction” that federal prosecutors may not use property-fraud statutes to “set[] standards of disclosure and good government for local and state officials.” 112 Id. at 1574 (quoting McNally v. United States, 483 U.S. 350, 360 (1987)). The Court acknowledged that the defendants had engaged in bad acts, but stressed that “[t]he property fraud statutes do not countenance” reaching beyond “schemes for obtaining property” and allowing federal authorities to “enforce (its view of) integrity” in regulatory decisionmaking. 113 Id. The text of Congress’s enactments cannot support such a “sweeping expansion of federal criminal jurisdiction.” 114 Id.

Justice Ginsburg also joined other significant decisions limiting the scope of federal corruption statutes in the face of concerns that, if interpreted too broadly, they could chill public discourse and the ability of citizens to lobby public officials. These cases, like Kelly, built on the reasoning of some of her own prior opinions. For instance, in McDonnell v. United States, she joined a unanimous decision that limits the type of bribery scheme criminalized by the honest-services-fraud statute (and the Hobbs Act, an extortion statute that has been construed to cover bribery schemes involving public officials). 115 136 S. Ct. 2355 (2016). The charges arose from $175,000 in loans, luxury items, and other benefits Virginia Governor Robert McDonnell and his wife received from Jonnie Williams, the CEO of a company promoting a nutritional supplement. Williams hoped McDonnell could persuade state universities to conduct certain studies necessary for the product to be classified as an anti-inflammatory drug. 116 Id. at 2361–62. The government’s theory was that McDonnell accepted the gifts in exchange for “official acts” such as “‘arranging meetings’ for Williams with other Virginia officials,” “‘hosting’ events” for the company at the Governor’s mansion, and “contacting other government officials” concerning research studies. 117 Id.

But the Supreme Court reversed the convictions and held that “setting up a meeting, calling another public official, or hosting an event does not, standing alone, qualify as an ‘official act.’” 118 Id. at 2368. Neither of the statutes charged expressly requires an “official act.” It was undisputed, however, that a bribery scheme requires one, and that the definition contained in 18 U.S.C. § 201 (2018), a statute expressly proscribing bribes and gratuities involving federal officials, should apply. Rather, an “official act” occurs only when the official formally exercises governmental power or advises or pressures another official to exercise such power. The Court expanded reasoning it had applied in United States v. Sun-Diamond Growers of California—another decision Justice Ginsburg joined—which narrowly construed “official act” in a gratuity case. 119 526 U.S. 398 (1999). And as in some of the Justice’s opinions, the Court cited the need to avoid three “significant constitutional concerns.” 120 McDonnell, 136 S. Ct. at 2371­–72. First, the government’s expansive theory could chill public officials’ interactions with the people they serve and thus impair the performance of their civic duties. 121 Id. at 2372–73. Second, citing Skilling, the Court found that the government’s interpretation of “official act” raised a “serious” due process concern, because it would sweep indefinitely and risk “arbitrary and discriminatory enforcement”; the Court’s narrower construction was thus necessary to avoid a finding of unconstitutional vagueness. 122 Id. at 2373 (quoting Skilling v. United States, 561 U.S. 358, 402–03 (2010)). Third, the government’s construction raised the same “federalism concerns” discussed in prior cases such as McNally. 123 Id.

The McDonnell Court also expressly acknowledged that the defendant’s conduct was “distasteful,” but concluded that this did not justify accepting the government’s “boundless interpretation.” 124 Id. at 2375. In language that I believe fully captures Justice Ginsburg’s own views on these questions, the Chief Justice observed:

There is no doubt that this case is distasteful; it may be worse than that. But our concern is not with tawdry tales of Ferraris, Rolexes, and ball gowns. It is instead with the broader legal implications of the Government’s boundless interpretation of the federal bribery statute. A more limited interpretation of the term “official act” leaves ample room for prosecuting corruption, while comporting with the text of the statute and the precedent of this Court. 125 Id. at 2375. Although some commentators have criticized McDonnell and argued that the decision makes it too difficult to prosecute public corruption, federal prosecutors have ample tools with which to combat corruption, and it is important to remember that there are many state and local laws regulating the ethics of public officials. Indeed, at the time the McDonnells received the gifts in question, such gifts were not prohibited by state law; subsequently Virginia amended the laws to prohibit valuable gifts. Not all ethical lapses should be subject to harsh criminal penalties. Conflict-of-interest provisions and other good-government regulations may suffice in many cases. See generally, Vincent L. Briccetti, Amie Ely, Alexandra Shapiro & Dan Stein, How Has McDonnell Affected Prosecutors’ Ability to Police Public Corruption? What Are Politicians and Lobbyists Allowed to Do, and What Are Prosecutors Able to Prosecute?, 38 Pace L. Rev. 707, 716 (2018).

c. Obstruction of Justice. — A final noteworthy opinion illustrating the Justice’s lenity-based approach to criminal statutes is Yates v. United States, 126 574 U.S. 528 (2015). a case involving a bizarre prosecution that divided the Supreme Court quite sharply but not along typical ideological lines. Justice Ginsburg wrote the plurality opinion (joined by Chief Justice Roberts and Justices Breyer and Sotomayor); Justice Alito concurred in the judgment; and Justice Kagan (joined by Justices Scalia, Kennedy, and Thomas) dissented. As in Skilling, the Justice’s reasoning is attacked by the dissent as inconsistent with a literal textual reading of the statute. Here, her interpretation invokes traditional tools of statutory construction and avoids what seem like absurd results in light of the obvious congressional intent.

In Yates, the Court reversed a fisherman’s conviction under a federal criminal statute prohibiting the destruction of records, when, “to prevent federal authorities from confirming that he had harvested undersized fish,” he ordered a crew member “to toss the suspect catch into the sea.” 127 Id. at 531. A law enforcement officer boarded Yates’s vessel off the coast of Florida and observed three red grouper that were slightly shorter than 20 inches, at a time when the law imposed civil penalties unless such undersized fish were immediately released. The officer proceeded to measure the remaining fish and separated out those short of twenty inches (most of which were over nineteen inches) and directed Yates to leave the undersized fish in certain crates until the ship returned to port. Instead, Yates directed a colleague to toss the undersized fish overboard and replace them with full-sized grouper from the rest of the catch. 128 Id. at 532–34.

Yates was convicted of violating 18 U.S.C. § 1519. In relevant part, that statute imposes a penalty of up to twenty years of imprisonment on anyone who “knowingly alters, destroys, mutilates, conceals, covers up, falsifies, or makes a false entry in any record, document, or tangible object with the intent to impede, obstruct, or influence” a federal investigation. Section 1519 was part of the Sarbanes–Oxley Act, which was enacted “to protect investors and restore trust in financial markets.” 129 Id. Sarbanes–Oxley was prompted by the collapse of Enron amid revelations that it had perpetrated a massive accounting fraud and that Arthur Andersen LLP, its outside auditor, had instructed its employees to destroy documents when the scandal broke. (Arthur Andersen’s conviction under an older obstruction statute had been unanimously reversed by the Supreme Court, 130 Arthur Andersen LLP v. United States, 544 U.S. 696 (2005). I was co-counsel to Arthur Andersen. and § 1519 was intended to close a gap in the earlier statute that barred the conviction.)

The question presented was whether a fish qualified as a “tangible object” within the meaning of this statute. As the Justice explained:

[A] fish is no doubt an object that is tangible; fish can be seen, caught, and handled, and a catch, as this case illustrates, is vulnerable to destruction. But it would cut § 1519 loose from its financial-fraud mooring to hold that it encompasses any and all objects, whatever their size or significance, destroyed with obstructive intent. 131 Yates, 574 U.S. at 532.

Because “Congress trained its attention on corporate and accounting deception and coverups,” a “tangible object captured by § 1519” “must be one used to record or preserve information.” 132 Id.

The Justice acknowledged that the dictionary definition of “tangible object” is broad. But, she opined, “[w]hether a statutory term is unambiguous” does not depend “solely on dictionary definitions of its component words.” 133 Id. at 537. Instead, context matters, and identical language may have different meanings in different statutes (and even in different provisions of the same statute). She explained: “In law as in life, . . . the same words, placed in different contexts, sometimes mean different things.” 134 Id. This point about context was very important to Justice Ginsburg and reflects her fundamental pragmatism. 135 Another criminal law example of this aspect of Justice Ginsburg’s philosophy of statutory interpretation involves the meaning of the term “uses” in 18 U.S.C. 924(c)(1)(A) (2006), which provides mandatory minimum sentences for anyone who, inter alia, “uses or carries a firearm” in connection with a narcotics trafficking offense.  Before she joined the Court, it held that trading one’s gun for drugs constitutes such a “use.” See Smith v. United States, 508 U.S. 223, 225 (1993). When the Court later held that trading one’s drugs for a gun does not involve such a use and attempted to distinguish Smith, the Justice concurred in the judgment because she was “persuaded that the Court took a wrong turn” in Smith.  Watson v. United States, 552 U.S. 74, 84 (2007) (Ginsburg, J., concurring in the judgment).  Justice Ginsburg opined that “use” should be read, “consistent with normal usage,” “to mean use as a weapon, not use in a bartering transaction.” Id.  She cited Justice Scalia’s dissent in Smith, which explained the importance of “the distinction between how a word can be used and how it ordinarily is used.” See id. at 84; Smith, 508 U.S. at 242 (Scalia, J., dissenting). Indeed, she often quoted legal realist Walter Wheeler Cook’s statement that the “tendency to assume that a word which appears in two or more legal” settings has the same meaning in both “has all the tenacity of original sin and must constantly be guarded against.” 136 See, e.g., Georgia v. Public.Resource.Org, Inc., 140 S. Ct. 1498, 1524 n.3 (2020) (Ginsburg, J., dissenting) (quoting Walter Wheeler Cook, “Substance” and “Procedure” in the Conflict of Laws, 42 Yale L.J. 333, 337 (1933)); Wachovia Bank, Nat’l Ass’n v. Schmidt, 546 U.S. 303, 319 (2006) (same); Emerald Mines Co. v. Fed. Mine Safety & Health Rev. Comm’n, 863 F.2d 51, 54 n.2 (D.C. Cir. 1988) (same).

In her Yates opinion Justice Ginsburg also invoked “familiar interpretive guides.” For instance, neither the heading (“Destruction, alteration, or falsification of records in Federal investigations and bankruptcy”) nor the title of the section (“Criminal penalties for altering documents) suggests that § 1519 prohibits “spoliation of any and all physical evidence, however remote from records.” 137 Yates, 574 U.S. at 539–40. In addition, the government’s interpretation would render Section 1519 largely superfluous of another, broader obstruction statute. Moreover, noscitur a sociis, the notion that “a word is known by the company it keeps,” also supports a narrower reading that the term refers not to any tangible object, but only to “the subsets of tangible objects involving records and documents, i.e., objects used to record or preserve information.” 138 Id. at 544. Likewise, the opinion cites ejusdem generis, the notion that where general words follow specific ones, the general words are presumed to embrace only objects similar in nature to those specifically identified in the preceding text.

And the opinion concludes, once again, by reference to the rule of lenity.  Justice Ginsburg found this interpretive principle relevant because, she explained, the government’s reading “exposes individuals to twenty-year prison sentences for tampering with any physical object that might have evidentiary value in any federal investigation into any offense, no matter whether the investigation is pending or merely contemplated, or whether the offense subject to investigation is criminal or civil.” 139 Id. at 548. Justice Alito does not specifically identify his disagreement with the plurality in his separate opinion, but his opinion is a bit narrower—he relies only on the statute’s title and its list of nouns and verbs and does not mention the rule of lenity. Id. at 549 (Alito, J., concurring in the judgment).

Justice Kagan’s dissent, by contrast, is a more literal interpretation. Kagan relies principally on the text and ordinary meaning of “tangible object.” She provides examples of other uses of the term in statutes and rules to support her position. Although not disagreeing that context is important, Justice Kagan draws a different lesson from context, focusing, for instance, on the word “any” in “any tangible object,” and claiming support for her interpretation in the legislative history. The dissent says the canons go the other way, and that the amendment to the broader obstruction statute supports its reading. In sum, the dissent, believing it has demonstrated that “traditional tools of statutory interpretation” do not support the result, suggests that “the  real  issue”  is  “overcriminalization  and  excessive  punishment  in  the  U.S. Code.” 140 Id. at 569 (Kagan, J., dissenting). The dissent concedes that Section 1519 may give “prosecutors too much leverage and sentencers too much discretion” and is not “an outlier, but an emblem of a deeper pathology” in federal criminal law. 141 Id. at 570. Nonetheless, the dissent (echoing Justice Scalia in Skilling) insists that “this Court does not get to rewrite the law.” 142 Id.

Although Yates was somewhat sui generis, Justice Ginsburg also agreed with defendants’ narrow constructions of other obstruction statutes due to concerns about subjecting people to the risk of imprisonment for conduct that could be innocent. One example is the unanimous Arthur Andersen decision (authored by Chief Justice Rehnquist) mentioned above. Another is Marinello v. United States, a 7-2 decision written by Justice Breyer. 143 138 S. Ct. 1101 (2018). There the Court held that 26 U.S.C. § 7212(a), a statute prohibiting obstruction of the administration of the tax code, encompasses only targeted tax-related proceedings like an investigation or an audit, and not routine administrative procedures such as processing tax returns.  It is unsurprising that the Justice joined the majority here. The tax obstruction statute, under the government’s broad reading, presented another trap for the unwary that counseled restraint and a narrowing construction. Otherwise, the majority observed, the provision could be used to imprison a person who pays a babysitter forty-one dollars per week in cash without withholding taxes; leaves a large cash tip in a restaurant; or fails to keep donation receipts from every charity to which she contributes. 144 Id. at 1108. These hypotheticals are similar to the examples Justice Ginsburg mentioned in her Ratzlaf opinion over twenty years earlier.

3. Aberrations: One Example. — There are some exceptions to the Justice’s narrow construction approach. These seem to be cases in which other principles the Justice cared  about  were  at  stake.  One  of  particular  interest  to  me  relates  to  insider trading. 145 I have litigated numerous cases involving questions about what constitutes illegal insider trading fraud, including, among others, Salman v. United States, 137 S. Ct. 420 (2016); United States v. Kosinski, 976 F.3d 135 (2d Cir. 2020); United States v. Blaszczak, 947 F.3d 19 (2d Cir. 2019), vacated, No. 20-5649, 2021 WL 78043 (U.S. Jan. 11, 2021); United States v. Newman, 773 F.3d 438 (2d. Cir. 2014), overruled in part by Salman, 137 S. Ct. 420. This is an area of the law in which the Supreme Court developed what amounts to a common law crime in the early 1980s based loosely on under Section 10(b) of the Securities Exchange Act of 1934 and SEC Rule 10b-5. Justice Ginsburg wrote the majority opinion in one of the most important Supreme Court cases defining the  contours  of  insider  trading  fraud  under  these  provisions,  United  States  v. O’Hagan. 146 521 U.S. 642 (1997).

Some quick background: Section 10(b) and Rule 10b-5 together comprise a general antifraud provision that bars the use of any “manipulative or deceptive device” “in connection with the purchase or sale of securities.” 147 15 U.S.C. § 78j(b) (2018). A “willful” violation of the statute carries criminal penalties. See id. § 78ff. O’Hagan also addressed a question arising under the tender offer fraud provisions, Section 14(e) of the Securities Exchange Act and SEC Rule 14e-3, but the Court’s holding on that issue is less interesting for purposes of this Article. Neither the statute nor the Rule mentions insider trading or purports to define when such trading is illegal. In 1980, however, the Supreme Court held that “insider trading” violates Section 10(b) and Rule 10b-5 if it is fraudulent. In Chiarella v. United States, an employee of a financial printer who prepared announcements for corporate takeover bids deduced the identities of target companies before the announcements were released. 148 445 U.S. 222 (1980). Without disclosing that information, he purchased shares in those companies, which he sold immediately after the takeover bids were made public. The issue was whether he had violated Section 10(b) by failing to inform sellers that he knew of a forthcoming takeover bid. The Court held that “silence in connection with the purchase or sale of securities may operate as a fraud actionable under § 10(b)” only where there is “a duty to disclose arising from a relationship of trust and confidence between parties to a transaction.” 149 Id. at 230. The defendant in Chiarella had no such fiduciary duty because he was not a corporate insider and received no confidential information from the target company; his conviction was therefore invalid. 150 Id. at 231–33.

The Chiarella Court refused to consider the government’s alternative theory—which the jury had not passed upon—that the defendant had a fiduciary duty to the acquiring corporation through his employer (the printer of the acquiring corporation) and that by misusing the information, he breached that duty and thereby committed securities fraud. The Supreme Court ultimately considered the validity of that theory (known as the “misappropriation” theory) in O’Hagan, which involved a lawyer who improperly used his law firm’s confidential information about an upcoming acquisition to trade shares of the target company, which was not the firm’s client.

Writing for the majority, Justice Ginsburg endorsed the misappropriation theory.  She explained that a person commits fraud “in connection with” a securities transaction and thus violates 10(b)/10b-5 “when he misappropriates confidential information for securities trading purposes, in breach of a duty owed to the source of the information.” 151 O’Hagan, 521 U.S. at 652. The Justice first observed that the Court had previously recognized “fraud of the same species” in a case involving the mail fraud statute, where it had held that a fiduciary who feigns loyalty to his principal while converting the latter’s information for his personal gain defrauds the principal. (In Carpenter v. United States, the Court had affirmed mail fraud convictions of several defendants involved in a tipping and trading ring that involved misuse of a newspaper’s confidential business information, which an employee of the paper had disclosed in breach of his duty to his employer. 152 484 U.S. 19 (1987). ) However, the mail (and wire) fraud statute does not require proof that fraud is “in connection with” a securities trade. Thus, the question in O’Hagan was, even if the defendant’s trading was mail or wire fraud because he had misappropriated confidential information entrusted to his law firm, wasn’t that fraud complete before he traded the securities? And if so, how could the fraud be “in connection with” those securities? The Justice’s answer was:

[T]he fiduciary’s fraud is consummated, not when the fiduciary gains the confidential information, but when, without disclosure to his principal, he uses the information to purchase or sell securities. The securities transaction and the breach of duty thus coincide. This is so even though the person or entity defrauded is not the other party to the trade, but is, instead, the source of the nonpublic information. 153 O’Hagan, 521 U.S. at 656.

By contrast, according to her opinion, if the fiduciary instead misappropriated funds and used them to trade, this would not be “in connection with” a securities transaction. Justice Thomas, in a partial dissent joined by Chief Justice Rehnquist, found this theory “incoherent” given the seeming contradiction between the results for embezzlement of funds and the embezzlement of information. 154 See id. at 681–83 (Thomas, J., concurring in the judgment in part and dissenting in part).

Justice Scalia also penned a short dissent on this issue, and he made a point Justice Ginsburg herself might have made in some other criminal law setting. He observed that the Court’s construction of Section 10(b) and Rule 10b-5 “would be entirely reasonable in some other context,” but “does not seem to accord with the principle of lenity we apply to criminal statutes.” 155 Id. at 679 (Scalia, J., concurring in part and dissenting in part). Under that principle, he opined, “the unelaborated statutory language: ‘[t]o use or employ, in connection with the purchase or sale of any security . . . any manipulative or deceptive device or contrivance,’ § 10(b), must be construed to require the manipulation or deception of a party to a securities transaction.” 156 Id. (alteration in original).

Why was Justice Ginsburg unpersuaded by the lenity argument here? I think there are two possible explanations. First, the decision is fully consistent with her voting record in favor of plaintiffs in Section 10(b) cases. There have been a series of cases involving the scope of 10(b) in the context of private actions, which have sharply divided the Court along traditional ideological lines. The liberal Justices typically vote in favor of an expansive reading of Section 10(b), because they believe it has a remedial purpose and should be read broadly to ensure robust protection of ordinary investors and the integrity of securities markets. Their conservative counterparts, on the other hand, typically vote with defendants in these cases, because of antipathy toward implied rights of action and a more anti-regulatory, pro-business bent. For instance, during my clerkship, the Court held, by a vote of 5-4, that there is no private Section 10(b) action for “aiding and abetting,” or secondary liability. Instead, only private actions against primary violators are permitted. 157 Cent. Bank of Denver, N.A. v. First Interstate Bank of Denver, N.A., 511 U.S. 164, 191–92 (1994). Justice Ginsburg joined Justice Stevens’s dissent (as did Justices Blackmun and Souter; Justice Kennedy wrote for the majority, which included Chief Justice Rehnquist and Justices O’Connor, Scalia, and Thomas). She also dissented in subsequent cases narrowing private plaintiffs’ Section 10(b) actions, including Janus Capital Group v. First Derivative Traders, which held that only persons who “make” or have “ultimate authority over statements” are liable for false statements, 158 564 U.S. 135, 142 (2011). and Stoneridge Investment Partners, LLC v. Scientific-Atlanta, Inc., which rejected so-called “scheme liability.” 159 552 U.S. 148, 159–60 (2008).

Second, the criminal 10(b) offense requires proof of a very high level of scienter. In her O’Hagan opinion, Justice Ginsburg stated that “two sturdy safeguards Congress has provided” are “[v]ital to our decision that criminal liability may be sustained under the misappropriation theory”—namely, the requirement that a person “willfully” violated the provision and the defense to imprisonment if the defendant proves he had no knowledge of the Rule. 160 O’Hagan, 521 U.S. at 665–66 (citing 15 U.S.C. § 78ff(a) (1992)). Thus, it appears that the scienter provisions provided some comfort that the misappropriation theory was not a trap for the unwary (and indeed, all nine Justices had concluded that the defendant’s conduct was mail and wire fraud, regardless of whether it also was “in connection with” a securities transaction in violation of Section 10(b) and Rule 10b-5).

B. Sixth Amendment Jurisprudence: Two Examples

During Justice Ginsburg’s tenure on the Court, there was a sea change in the law on the Sixth Amendment’s right to jury trial and Confrontation Clause. In both areas, cases decided in the early 2000s significantly expanded the rights of the accused. Justice Ginsburg did not write many of these opinions, which closely divided the Court (albeit not along traditional ideological lines). The Justice typically voted for the defendant’s position in this area. These votes (at least early on, before the new doctrines became settled decisions on which to build) were somewhat bolder than her typical gradualist positions. For instance, the line of cases applying the jury right to sentencing eventually led the Court to revise the statute that created the Federal Sentencing Guidelines and to grant defendants significant new procedural rights. This upended decades of federal sentencing practice. At the same time, one of Justice Ginsburg’s key votes in these cases—which was decisive to the outcome—reflects a classic Ginsburg attempt to salvage, rather than jettison, a constitutionally problematic federal statute.

1. Jury Trial Right and the Apprendi Doctrine. — The Constitution guarantees defendants the right to a jury determination of whether the government has proven all elements of a criminal offense beyond a reasonable doubt. 161 See, e.g., In re Winship, 397 U.S. 358, 364 (1970) (“[T]he Due Process Clause protects the accused against conviction except upon proof beyond a reasonable doubt of every fact necessary to constitute the crime with which he is charged.”); Duncan v. Louisiana, 391 U.S. 145, 149 (1968) (holding that the Sixth Amendment right to “trial by jury in criminal cases is fundamental to the American scheme of justice” and thus applies to the States through the Fourteenth Amendment). Relatedly, an indictment must set forth each element of all crimes that it charges. 162 See, e.g., Hamling v. United States, 418 U.S. 87, 117 (1974). These are basic constitutional principles, but by the late 1990s issues about how to apply them to new and increasingly harsh sentencing regimes started to percolate. This ultimately led to a series of Supreme Court cases posing questions about when a sentencing “factor” is effectively an “element” and thus subject to the same constitutionally based procedural protections. 163 I was the principal author of several amicus briefs in this area, including New York Council of Defense Lawyers briefs cited in United States v. Booker, 543 U.S. 220, 266 (2005), and Rita v. United States, 551 U.S. 338, 373 n.3 (2007) (Scalia, J., concurring in part and concurring in the judgment).

The context in which the Supreme Court began to consider these questions is important. In the heyday of the so-called “war on drugs” in the 1980s and early 1990s, there was a flurry of federal and state legislation increasing punishment for criminal offenses—recidivism statutes and other laws that increased penalties when the government proved additional aggravating facts (such as use of a deadly weapon, racial motive, or drug weight) beyond the basic offense conduct. Some of these provisions increased the maximum penalties for the crimes; others created mandatory minimum sentences that applied when such aggravating facts were present. In addition, the federal government and some states promulgated binding sentencing guidelines that increased sentences based on similar factors. These regimes generally permitted enhanced sentences based on fact-findings by the sentencing judge under a preponderance of the evidence standard.

One key example is the Federal Sentencing Guidelines, which took effect in 1987 and were promulgated pursuant to the Sentencing Reform Act of 1984 (SRA). 164 For a more detailed discussion of the SRA, see Nancy Gertner, A Short History of American Sentencing: Too Little Law, Too Much Law, or Just Right, 100 J. Crim. L & Criminology 691, 695, 698 (2010). The Guidelines radically changed the process by which federal criminal sentences were imposed. Before the SRA, federal judges generally had vast discretion to impose any sentence up to the maximum penalty permitted by the statute of conviction. As a result, there were significant disparities in sentences similar defendants were receiving for similar crimes. One of the principal goals of the new Guidelines system was to reduce that disparity and ensure that similarly situated defendants received similar sentences, regardless of where they were sentenced and which judge sentenced them. To achieve that goal, the Guidelines created a complex system that required judges to calculate a Guidelines range (for example ten to sixteen months’ imprisonment) based on numerous sentencing factors that were supposed to reflect the severity of the offense as well as the defendant’s criminal history. Under the SRA, judges were bound to impose a sentence within this range, which the judge was solely responsible for  calculating  based  on  his  or  her  fact-findings  under  the  preponderance standard. 165 The Guidelines also provided bases for upward or downward “departures” from the range, but whether a defendant qualified for such departures was also a matter for the judge alone under the preponderance standard.

Prior to Apprendi v. New Jersey and its progeny, the Court purported to draw a line between “sentencing factors”—which did not have to be disclosed in an indictment and could increase a defendant’s sentence based entirely on findings by a judge under the preponderance standard—and “elements” subject to the constitutional requirements of findings by a jury under the beyond a reasonable doubt standard. 166 530 U.S. 466, 476–78 (2000). Within fairly loose limits, the Court deferred to legislatures on which facts were to be treated as mere sentencing factors and which were elements subject to the aforementioned constitutional protections. 167 ee, e.g., McMillan v. Pennsylvania, 477 U.S. 79, 85 (1986) (“[I]n determining what facts must be proved beyond a reasonable doubt the state legislature’s definition of the elements of the offense is usually dispositive.” (citing Patterson v. New York, 432 U.S. 197, 210 (1977))), overruled by Alleyne v. United States, 570 U.S. 99, 112 (2013). Because of the draconian impact of some of the new sentencing statutes, however, defendants began raising constitutional challenges to some of the procedures that led to harsh sentences that had been increased based on judicial fact-finding. The logical implication of these arguments was that many of the new sentencing laws, including the Federal Sentencing Guidelines, were constitutionally flawed and would have to be invalidated or rewritten to ensure that fact-finding was done by juries under the reasonable doubt standard. This could lead to major changes and set back what advocates of the SRA had championed as important sentencing reforms. Also, requiring such jury fact-finding seemed impractical, inefficient, and cumbersome. Despite the potentially significant consequences, however, from the outset Justice Ginsburg voted in favor of defendants’ positions on these constitutional arguments.

At first, however, defendants did not succeed. The Supreme Court considered and rejected  the  first  such  argument  by  a  vote  of  5-4  in  Almendarez-Torres  v. United States. 168 523 U.S. 224 (1998). The case involved a statute that increased the maximum penalty for illegally reentering the United States after deportation from two to twenty years’ imprisonment if the defendant had previously been convicted of an “aggravated felony.” 169 Id. at 226. The defendant argued that his seven-year sentence, imposed based on the judge’s finding that he had committed such a felony, was unconstitutional because the indictment provided no notice of this aggravated-felony enhancement. 170 Id. at 227–28. The majority (Justice Breyer, joined by Chief Justice Rehnquist and Justices Kennedy, Thomas, and O’Connor) held that the recidivism provision does not “define a separate crime,” but “simply authorizes an enhanced penalty.” 171 Id. at 226. Justice Breyer was involved in the drafting of both the SRA (as a Senate Judiciary Committee staffer) and the original Sentencing Guidelines (as a member of the original Sentencing Commission charged with drafting the Guidelines). His commitment to the policies underlying the SRA and the Guidelines and his role as draftsperson likely explain both his views in Almendarez-Torres and his resistance in subsequent cases to imposing constitutionally based procedural protections in the sentencing context. But Justice Ginsburg (and Justices Stevens and Souter) joined Justice Scalia’s dissent. 172 Id. Despite the traditional role of the sentencing judge (even in death penalty cases), the dissenters found it “doubtful” that the “Constitution permits a judge (rather than a jury) to determine by a mere preponderance of the evidence (rather than beyond a reasonable doubt) a fact that increases the maximum penalty to which a criminal defendant is subject.” 173 Id. at 251 (Scalia, J., dissenting). Shortly thereafter, Justice Thomas switched sides, and the dissenters’ view that constitutionality was doubtful was transformed into a holding of unconstitutionality.

The tide began to turn in Jones v. United States. 174 See 526 U.S. 227, 229 (1999). There, Justice Souter, together with Justice Ginsburg, the other Almendarez-Torres dissenters, and Justice Thomas, held that a federal carjacking statute created three distinct offenses rather than a single crime with three different penalties “to avoid serious questions about the statute’s constitutionality.” 175 Id. In a footnote, the Jones Court described the constitutional principle as follows: “[U]nder the Due Process Clause of the Fifth Amendment and the notice and jury trial guarantees of the Sixth Amendment, any fact (other than prior conviction) that increases the maximum penalty . . . must be charged  in  an  indictment,  submitted  to  a  jury,  and  proven  beyond  a  reasonable doubt.” 176 Id. at 243 n.6. The Court purported to distinguish Almendarez-Torres as a case about notice and not the jury trial right, and highlighted that the decision there rested in large part on the “tradition of regarding recidivism as a sentence factor, not as an element.” Id. at 249. In dissent, Justice Kennedy presciently pointed out that the constitutional principle that the majority articulated would cast serious doubt on the Federal Guidelines as well as sentencing systems used by many states. 177 Id. at 254 (Kennedy, J., dissenting).

One year later, the Court expressly elevated the Jones footnote to a constitutional holding in Apprendi. 178 See Apprendi v. New Jersey, 530 U.S. 466, 476 (2000) (stating that the “Fourteenth Amendment commands the same answer in this case involving a state statute” (quoting Jones, 526 U.S. at 243 n.6)). The case involved a New Jersey law that increased the maximum penalty for possession of a firearm for an unlawful purpose in cases where the government proved that the defendant had a discriminatory motive (a hate-crimes law). The defendant was prosecuted for firing several bullets into the home of a Black family; his sentence was increased pursuant to the hate-crimes enhancement after the judge found by a preponderance that he had a purpose to intimidate on the basis of the victims’ race. 179 Id. at 468–71. Since the penalty was set forth in a completely different statute from the weapons-offense law, the case squarely presented the constitutional questions addressed only indirectly in the two earlier cases. Justice Ginsburg joined Justice Stevens’s opinion for the Court, which was again divided 5-4 along the same lines as in Jones. The Court noted that “it is arguable that Almendarez-Torres was incorrectly decided, and that a logical application of our reasoning today should apply if the recidivist issue were contested,” but declined to overrule it and instead deemed it “a narrow exception” to the constitutional rule Apprendi announced:

Other than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt. . . . “[I]t is unconstitutional for a legislature to remove from the jury the assessment of facts that increase the prescribed range of penalties to which a criminal defendant is exposed. It is equally clear that such facts must be established by proof beyond a reasonable doubt.” 180 Id. at 489–90 (alteration in original) (quoting Jones, 526 U.S. at 252–53 (Stevens, J., concurring)) (citing Jones, 526 U.S. at 253 (Scalia, J., concurring)).

In the principal dissent, Justice O’Connor accused the majority of effecting a radical change, saying: “In one bold stroke the Court today casts aside our traditional cautious approach and instead embraces a universal and seemingly bright-line rule limiting the power of Congress and state legislatures to define criminal offenses and the sentences that follow from convictions thereunder.” 181 Id. at 525 (O’Connor, J., dissenting). The dissent also predicted that Apprendi would lead to the invalidation of “significant sentencing reform” implemented at the federal and state levels in the previous three decades. 182 Id. at 549.

That prediction proved to be accurate, as in subsequent cases the Court—with Justice Ginsburg’s support—would apply Apprendi to invalidate Washington State’s binding sentencing guidelines; 183 Blakely v. Washington, 542 U.S. 296, 305 (2004). uphold the federal Sentencing Reform Act and its Sentencing Guidelines by making them nonbinding; 184 United States v. Booker, 543 U.S. 220, 259 (2005). invalidate California’s determinate sentencing law; 185 Cunningham v. California, 549 U.S. 270, 274 (2007). require juries to find facts triggering mandatory-minimum sentences beyond a reasonable doubt (a holding that required the Court to overturn two prior decisions, including a post-Apprendi decision); 186 Alleyne v. United States, 570 U.S. 99, 103, 107 (2013) (overruling Harris v. United States, 536 U.S. 545 (2002), and McMillian v. Pennsylvania, 477 U.S. 79 (1986)). require juries to find facts that could trigger a death sentence (overruling a 1990 decision and later invalidating Florida’s death sentencing scheme); 187 Ring v. Arizona, 536 U.S. 584, 585 (2002) (overruling Walton v. Arizona, 497 U.S. 639 (1990)); see also Hurst v. Florida, 577 U.S. 92, 94–97 (2016) (holding that Florida’s death sentencing scheme, in which a jury recommends a sentence to the judge and the judge determines whether facts support the imposition of a death sentence, violated the Sixth Amendment in light of Ring). invalidate a statute permitting a judge to revoke supervised release and impose a new and additional prison sentence based on facts not found by a jury beyond a reasonable doubt; 188 United States v. Haymond, 139 S. Ct. 2369, 2382 (2019). and require juries to find facts that could increase fines beyond a reasonable doubt. 189 S. Union Co. v. United States, 567 U.S. 343, 346 (2012).

With only one minor exception that I am aware of, 190 See Oregon v. Ice, 555 U.S. 160, 163–64 (2009); see also infra notes 218–227 and accompanying text. Justice Ginsburg consistently voted to enforce the basic constitutional principle recognized in Apprendi. 191 Justice Ginsburg also voted in favor of the right to jury trial in other contexts. See, e.g., Ramos v. Louisiana, 140 S. Ct. 1390, 1395 (2020) (holding that the Sixth Amendment jury trial right, as incorporated against the states, requires unanimous verdicts to convict defendants of serious offenses); Lewis v. United States, 518 U.S. 322, 339 (1996) (Stevens, J., dissenting) (dissenting, in an opinion joined by Justice Ginsburg, from a 7-2 opinion holding that there is no jury trial right in cases involving only petty offenses). She wrote three post-Apprendi opinions, which I discuss below, but her most interesting vote came in a case in which she did not write an opinion: United States v. Booker, which applied Apprendi to the Federal Sentencing Guidelines. 192 543 U.S. 220, 266 (2005). The Court faced two questions in Booker: (1) Because the Guidelines were binding on judges, does the Sixth Amendment limit the severity of the sentence a judge may impose based on the facts found by the jury at trial and, (2) if so, what is the remedy? 193 Id. at 226–27. The Court divided 5-4 on both questions, but there was a different lineup of Justices on each question and Justice Ginsburg cast the deciding fifth vote on both issues. 194 Id. at 225. Justice Stevens wrote the majority opinion on the first (constitutional) question and was joined by Justices Scalia, Souter, Thomas, and Ginsburg. They held that the Apprendi doctrine applies to the Guidelines because they are binding on judges. “Any fact (other than a prior conviction) which is necessary to support a sentence exceeding the maximum authorized by the facts established by a plea of guilty or a jury verdict must be admitted by the defendant or proved to a jury beyond a reasonable doubt.” 195 Id. at 244. The constitutional majority acknowledged that “in some cases jury factfinding may impair the most expedient and efficient sentencing of defendants” but found that these considerations are “always outweighed” by “the interest in fairness and reliability” protected by the Sixth Amendment right to a jury trial. 196 Id.

Despite the constitutional majority’s determination to reject the arguments about procedural efficiency, however, with Justice Ginsburg’s vote, the remedial majority provided a practical solution that avoided those purported efficiency issues. The remedial holding was set forth in Justice Breyer’s opinion for the Court, which was joined by Chief Justice Rehnquist and Justices O’Connor, Kennedy (who all dissented from Justice Stevens’s constitutional majority holding), and Ginsburg. 197 Id. at 225. The Justices in the constitutional majority—other than Justice Ginsburg—would have left the Guidelines intact, but simply required courts to follow constitutional procedure by requiring juries to find sentence-enhancing facts under the reasonable doubt standard. 198 Id. at 284 (Stevens, J., dissenting in part). Justice Thomas agreed with “Justice Stevens’ proposed remedy and much of his analysis” but wrote a separate dissent on a technical point about severability principles and use of legislative history. Id. at 313 (Thomas, J., dissenting in part) (emphasis omitted). By contrast, the remedial majority held that in light of the constitutional holding, the provision of the SRA that made the Guidelines binding, together with another provision dependent on their binding nature (establishing de novo review of departures from the Guidelines) should be excised from the statute. The result transformed the binding Guidelines into true nonbinding guidelines, which sentencing judges must consider but could reject, subject only to review for reasonableness on appeal. The Court reasoned that “the constitutional jury trial requirement is not compatible with the Act as written” and that Congress would have preferred severance to the remedial dissent’s solution. 199 Id. at 248.

Justice Ginsburg did not write an opinion explaining her vote, but her position seems consistent with her strong support for enforcing the constitutional guarantees at stake and her gradualist philosophy of trying to salvage, rather than “condemn,” legislation—even when doing so seems to involve redrafting the statute. In that respect, the remedial majority’s fix was somewhat similar to the solution the Justice arrived at in Skilling. 200 See supra notes 84–93 and accompanying text. It enabled the Court to preserve the Guidelines regime without adding procedural requirements and to preserve the primacy of the judge’s role—as Congress contemplated—while leaving room for Congress, should it so desire, to tinker further with the sentencing process in light of the Court’s constitutional guidance. And the Booker remedy certainly gave judges greater discretion to impose more lenient sentences in cases where strict adherence to the Guidelines would have led to draconian ones. This is a result I imagine would have pleased the Justice.

To be sure, the principal purpose of the SRA was to enforce consistent sentencing results, 201 Booker, 543 U.S. at 253–54; id. at 303 (Scalia, J., dissenting in part). so one can reasonably argue that excising the provisions requiring judges to apply them defeated Congress’s intent. In the main remedial dissent, Justice Stevens argued that none of the parties or amici had advocated invalidating the two provisions, and that this solution “represents a policy choice that Congress has considered and decisively rejected.” 202 Id. at 272 (Stevens, J., dissenting in part). In dissent, Justice Scalia charged: “In order to rescue from nullification a statutory scheme designed to eliminate discretionary sentencing, [the majority] discards the provisions that eliminate discretionary sentencing.” 203 Id. Scalia critiqued the remedial majority’s modification of the standard of review—again arguing that the Court was improperly rewriting the statute. 204 Id. at 309.

Justice Ginsburg wrote three post-Apprendi decisions. The first two are fully consistent with her other votes in this area, but the last one is somewhat anomalous (and puzzling to me).

The first was Ring v. Arizona, which applied Apprendi to the death penalty. 205 536 U.S. 584, 588–89 (2002). Under Arizona’s statute, once a jury convicted a defendant of first-degree murder, a judge alone determined whether the State has proven the aggravating factors required for imposition of the death penalty. 206 Id. at 592. Ring overruled Walton v. Arizona, a 1990 decision holding that the very same capital sentencing scheme did not violate the Sixth Amendment. 207 497 U.S. 639, 649 (1990), overruled by Ring, 536 U.S. 584. Ring’s result followed logically from Apprendi—although the Court in Apprendi (and its precursor, Jones) had purported to distinguish it from Walton. 208 Apprendi v. New Jersey, 530 U.S. 466, 496–97 (2000); Jones v. United States, 526 U.S. 227, 251 (1999). As Justice Ginsburg observed in Ring, “Apprendi’s reasoning is irreconcilable with Walton’s holding”: Apprendi teaches that “the Sixth Amendment does not permit a defendant to be ‘expose[d] . . . to a penalty exceeding the maximum he would receive if punished according to the facts reflected in the jury verdict alone.’” 209 Ring, 536 U.S. at 588–89 (quoting Apprendi, 530 U.S. at 483). Her opinion, which was joined by the other Justices in the Apprendi majority (Stevens, Souter, Scalia, and Thomas) as well as Justice Kennedy, 210 Id. at 587. is unsurprising given her past votes on Sixth Amendment sentencing questions and the fact that the logic is straightforward. But whereas the Court had previously (in opinions joined by Justice Ginsburg) strained to distinguish Walton in non-death penalty cases, here instead the Justice forthrightly acknowledged that Walton could not survive Apprendi. Indeed, she cited the Apprendi dissenters’ arguments to support her logic. 211 Id. at 603 (quoting Apprendi, 530 U.S. at 538 (O’Connor, J., dissenting)). (Among other things, they had called the Court’s prior distinction of Walton “baffling.” 212 Apprendi, 530 U.S. at 538 (O’Connor, J., dissenting). ) Notably, Justice Ginsburg obtained Justice Kennedy’s vote, despite his disagreement with Apprendi. He explained in a separate concurrence that he still believed Apprendi was wrongly decided, but it “is now the law” and “no principled reading” would allow Walton “to stand.” 213 Ring, 536 U.S. at 613 (Kennedy, J., concurring). Perhaps by this point it was easier to expressly jettison Walton, or there were finally five votes to do so.

Justice Ginsburg’s opinion for the Court in California v. Cunningham is similarly unremarkable in light of the Apprendi line of cases, though it sparked the expected dissents from those not part of the group of five Justices who had adopted and extended the Apprendi doctrine. 214 549 U.S. 270, 273 (2007). At issue was a 1977 California statute that replaced an indeterminate sentencing regime with a determinate one aimed at promoting uniform and proportionate punishment. The law prescribed three levels of punishment—a lower, middle, and upper term sentence—and required the judge to impose the middle term unless there were aggravating or mitigating circumstances. The upper term could only be imposed if the judge found aggravating “facts” by a preponderance. 215 Id. at 275–77. The Court had already held that the statutory maximum for Apprendi analysis is the maximum sentence that could be imposed based upon the facts reflected in the jury verdict or admitted by the defendant; 216 See supra notes 192–195. thus, the logical extension of the doctrine dictated invalidation of the California scheme (though the dissenters  protested  that  it  was  more  like  the  advisory  federal  Guidelines  after Booker 217 See Cunningham, 549 U.S. at 293; id. at 297–98 (Alito, J., dissenting). ).

The outlier is Oregon v. Ice. 218 555 U.S. 160 (2009). This is the only case that I know of in which the Justice rejected an Apprendi-based argument and is a bit of a head-scratcher in light of her other votes. The question presented was: “When a defendant has been tried and convicted of multiple offenses, each involving discrete sentencing prescriptions, does the Sixth Amendment mandate jury determination of any fact declared necessary to the imposition of consecutive, in lieu of concurrent, sentences?” 219 Id. at 163. The Oregon law required the imposition of concurrent sentences unless the judge finds certain facts; if such facts were found, the judge was permitted, but not required, to impose consecutive sentences. 220 Id. at 165. Writing for the Court, the Justice held that this scheme passed muster under the Sixth Amendment. 221 Id. at 164. Other than Justices Ginsburg and Stevens, the original five Apprendi adherents dissented (along with Chief Justice Roberts, who joined the Court after Apprendi was settled law). Id. at 162. As the dissent pointed out, in her opinion Justice Ginsburg emphasized points that could apply equally to many of the other State sentencing regimes the Court had found wanting under Apprendi. 222 Id. at 173 (Scalia, J., dissenting). For instance, she cited:

– the “common-law tradition” of “entrust[ing] to judges’ unfettered discretion” whether to impose consecutive or concurrent sentences;

– that (unlike some states that allow unfettered discretion), Oregon provides for discretion but “constrain[s] its exercise”;

– “States’ interest in the development of their penal systems, and their historic dominion in this area”;

– that “state legislative innovations like Oregon’s seek to rein in the discretion judges possessed at common law to impose consecutive sentences at will” in order to “reduce disparities” between similarly situated defendants;

– that “other state initiatives would fall” if Apprendi were applied; and

– that this would be “difficult for States to administer.” 223 Id. at 163–64, 170–72 (majority opinion).

These were some of the very arguments rejected in Apprendi, Blakely v. Washington, and Booker. 224 See id. at 173 (Scalia, J., dissenting). Moreover, deference to common law tradition is not something the Justice typically viewed as particularly important. And it is indisputable that consecutive sentences are a greater punishment than concurrent sentences, and that “the decision to impose consecutive sentences alters the single consequence most important to convicted noncapital defendants: their date of release from prison.” 225 Id. at 174. As Justice Ginsburg said in Cunningham, “If the jury’s verdict alone does not authorize the sentence, if, instead, the judge must find an additional fact to impose the longer term, the Sixth Amendment requirement is not satisfied.” 226 Cunningham v. California, 549 U.S. 270, 290 (2007). Fact-dependent consecutive sentencing seems to fit squarely within that description.

Nonetheless, it seems that, for some reason, invalidating the statute in Ice was a bridge too far for Justice Ginsburg. Perhaps she believed that Oregon’s law was a reform designed to give judges discretion to impose more lenient sentences. Requiring juries for this particular decision would result in a backtracking that would permit easier imposition of consecutive sentences. As she noted: “All agree that a scheme making consecutive sentences the rule, and concurrent sentences the exception, encounters no Sixth Amendment shoal. To hem in States by holding that they may not equally choose to make concurrent sentences the rule, and consecutive sentences the exception, would make scant sense.” 227 Ice, 555 U.S. at 171. But such considerations could easily have led her to choose the government’s side in all the other cases, so I still find Ice a puzzling piece of the Justice’s Apprendi jurisprudence.

2. Confrontation Clause. — The Confrontation Clause of the Sixth Amendment provides that “[i]n all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.” Traditionally, the rules of evidence prohibit hearsay (statements introduced for the truth of the matter asserted, where the person who made the statement—the “declarant”—does not testify in court). 228 See Fed. R. Evid. 801(b), 802. But the rule against hearsay is riddled with exceptions. And questions naturally arise as to how hearsay—even when it fits within an exception—can be reconciled with the Confrontation Clause. After all, a defendant cannot confront or cross-examine a hearsay declarant. Before the Court’s 2005 decision in Crawford v. Washington, 229 541 U.S. 36 (2004). however, the prevailing doctrine suggested there was no constitutional problem for “traditional” hearsay exceptions. That was because under Ohio v. Roberts, the admission of hearsay that fell within a “firmly rooted hearsay exception” or bore “particularized guarantees of trustworthiness” did not violate the Confrontation Clause. 230 448 U.S. 56, 66 (1980), overruled by Crawford, 541 U.S. 36.

In Crawford, the defendant was convicted of assault and attempted murder of a man who tried to rape his wife. At trial, the State introduced a recorded statement that the wife (who did not testify, asserting marital privilege) made in response to police interrogation to show that the stabbing was not in self-defense. The trial court admitted the evidence despite a Confrontation Clause objection, on the grounds that it bore “particularized guarantees of trustworthiness” under Ohio v. Roberts. 231 Crawford, 541 U.S. at 38–41. The Supreme Court held that where statements are “testimonial”—like those of the wife in response to police questioning—“the only indicium of reliability sufficient to satisfy constitutional demands is the one the Constitution actually prescribes: confrontation.” 232 Id. at 59, 68–69. Accordingly, such statements are admissible without a live witness only if the witness is unavailable and the defendant had a prior opportunity to cross-examine that witness. 233 Id. at 59. The Court attempted to distinguish several prior precedents but expressly overruled Roberts because its test applied to all hearsay and permitted the admission of “core testimonial statements that the Confrontation Clause plainly meant to exclude.” 234 Id. at 63. Justice Ginsburg joined Justice Scalia’s opinion for a 7-2 majority.

There are hints of Justice Ginsburg’s concern about the unreliability of certain kinds of hearsay in earlier cases, including one decided during my clerkship that first Term. Williamson v. United States was resolved by interpretation of the Federal Rules of Evidence, but confrontation concerns lurk in the background. 235 512 U.S. 594, 599–601 (1994). Justice Ginsburg’s separate concurring opinion reflects her belief that one hallmark of fair criminal trials is ensuring that unreliable hearsay not subject to cross-examination is not admitted into evidence. The question presented involved the hearsay exception for statements against penal interest in Rule 804(b)(3), which was the basis for the admission of numerous post-arrest statements by a drug dealer against his co-conspirator. 236 At the time, that Rule created an exception to the rule against hearsay for “statement[s] which [were] at the time of [their] making . . . so far tended to subject the declarant to . . . criminal liability . . . that a reasonable person in the declarant’s position would not have made the statement[s] unless believing [them] to be true.” Id. at 599. The current version of the Rule has been tweaked slightly. See Fed. R. Evid. 804(b)(3)(A)–(B). The statements were partially inculpatory, since the accomplice/declarant admitted his involvement, but many of the remarks focused on painting the defendant as the “big fish.” 237 Williamson, 512 U.S. at 609 (Ginsburg, J., concurring in part and concurring in the judgment). The question before the Court was whether the entire “confession” by the accomplice was admissible under the Rule even if it contained both self-inculpatory and non-self-inculpatory parts, or whether the Rule only permitted admission of “those declarations or remarks within the confession that are individually self-inculpatory.” 238 Id. at 599 (majority opinion). By a vote of 6-3, the Court adopted the narrower reading and rejected the view that “collateral” statements are admissible even if they are not against the declarant’s interest. 239 Id. at 595, 600.

Justice O’Connor wrote the majority opinion, which was joined in full by Justice Scalia and in part by Justices Blackmun, Stevens, Souter, and Ginsburg. The section of the opinion that Justice Ginsburg declined to join concluded that some of the accomplice’s statements would have been admissible (for example, statements that he knew there was cocaine in the trunk of his car) whereas others (for example, the parts implicating the defendant) were not. 240 Id. at 604–05.

Justice Ginsburg wrote an opinion concurring in part and concurring in the judgment, which was joined by Justices Blackmun, Stevens, and Souter. She agreed that the Rule excepts from the hearsay prohibition “only ‘those declarations or remarks within [a narrative] that are individually self-inculpatory’” and emphasized the Court’s acknowledgment that statements implicating another person are untrustworthy: “A person arrested in incriminating circumstances has a strong incentive to shift blame or downplay his own role in comparison with that of others, in hopes of receiving a shorter sentence and leniency in exchange for cooperation.” 241 Id. at 607–08 (Ginsburg, J., concurring in part and concurring in the judgment) (quoting id. at 599 (majority opinion)). However, the Justice parted company with Justice O’Connor in an important respect: Justice Ginsburg believed that the accomplice’s statements “do not fit, even in part, within” the Rule 804(b)(3) exception because the “arguably inculpatory statements are too closely intertwined with his self-serving declarations to be ranked as trustworthy.” 242 Id. at 608. She noted that the declarant “was caught red-handed with 19 kilos of cocaine—enough to subject even a first-time offender to a minimum of 12 ½ years’ imprisonment”—and that since denying knowing the drugs were in the trunk of the car he was driving was not a viable option, he admitted involvement but minimized his own role and shifted blame to the defendant and another man. 243 Id. She carefully parsed all the declarant’s remarks and showed how the thrust of all of them was to “paint Williamson as the ‘big fish’”: “To the extent some of these statements tended to incriminate [the declarant], they provided only marginal or cumulative evidence of his guilt. They project an image of a person acting not against his penal interest, but striving mightily to shift principal responsibility to someone else.” 244 Id. at 609. Accordingly, Justice Ginsburg would have held that none of the hearsay statements were admissible against Williamson. 245 Id. at 610.

In light of this early concurrence and Justice Ginsburg’s concerns about the untrustworthiness of such hearsay, it is unsurprising that the Justice joined the subsequent decision in Crawford to overrule Ohio v. Roberts, and also was among the Justices most inclined to enforce Crawford in a defendant-protective way. Crawford spawned a fair amount of litigation over issues like what types of hearsay statements admitted pursuant to exceptions are “testimonial.” This led to several additional Supreme Court cases raising questions about how to apply Crawford to particular types of hearsay. Justice Ginsburg usually voted for the defendants’ positions, particularly where the issues divided the Court. Indeed, for the most part she and Justice Scalia most consistently agreed with strict enforcement of Crawford’s holding—in contrast to some other liberal Justices during this time frame, who either were hostile to Crawford’s consequences (Justice Breyer), or had more nuanced positions and read Crawford more narrowly than Justice Ginsburg (Justice Sotomayor).

Michigan v. Bryant is a good example. 246 562 U.S. 344 (2011). The trial court had admitted statements a victim  made  to  police  officers  who  discovered  him  mortally  wounded  in  a  parking lot. 247 Id. at 348. The Court held that the circumstances objectively established that the “primary purpose of the interrogation” was “to enable police assistance to meet an ongoing emergency,” and accordingly, that the victim’s description of the shooter and location of the shooting were not “testimonial” statements subject to the Confrontation Clause. 248 Id. at 349 (quoting Davis v. Washington, 547 U.S. 813, 822 (2006)). Justice Sotomayor wrote the majority opinion (joined by Chief Justice Roberts and Justices Kennedy, Breyer, and Alito, with Justice Thomas concurring in the judgment). Justice Ginsburg and Justice Scalia each wrote separate dissents. 249 Justice Kagan took no part in the consideration or decision of the case. Id. at 347.

Justice Ginsburg concluded that the victim’s statements were testimonial because the declarant’s intent was what mattered. 250 Id. at 395 (Ginsburg, J., dissenting). Even if the interrogators’ intent was dispositive, the statements would still be testimonial because the officers most likely viewed their encounter as an investigation into a past crime. 251 Id. And the majority’s decision “creates an expansive exception to the Confrontation Clause for violent crimes” and “confounds our recent Confrontation Clause jurisprudence, which made it  plain  that  ‘[r]eliability  tells  us  nothing  about  whether  a  statement  is testimonial.’” 252 Id. (internal quotation marks omitted) (quoting id. at 389, 392 (Scalia, J., dissenting)). The Justice also observed that, had the issue been properly preserved, she would take up the question of whether the “well-established exception” to the confrontation requirement “in the law we inherited from England” for “dying declarations survives our recent Confrontation Clause decisions.” 253 Id. at 395–96. (This question was left open in Crawford. 254 See Crawford v. Washington, 541 U.S. 36, 56 n.6 (2004). ) Justice Ginsburg did not preview her views on the substance of this question, however.

Giles v. California is another post-Crawford decision that reflects Justice Ginsburg’s bent in favor of enforcing confrontation rights, as well as her inherent pragmatism. 255 554 U.S. 353 (2008).  The case involved the common law “forfeiture by wrongdoing” exception to confrontation. 256 Id. at 357. That exception allowed the government to introduce hearsay statements by a witness who was unable to testify because he or she was “‘detained’ or ‘kept away’ by the ‘means or procurement’ of the defendant.” 257 Id. at 359. Justice Ginsburg joined the 6-3 majority holding that this exception applies only when the defendant deliberately prevents the witness from testifying. 258 Id. at 355, 359–60. At the defendant’s murder trial, the court admitted statements the victim had made to the police about three weeks before the murder, when they had responded to a domestic-violence incident. 259 Id. at 356–57. The California appellate courts had held that “forfeiture by wrongdoing” was an exception to the Confrontation Clause and rejected a Crawford challenge to the conviction. 260 Id. at 357.

Writing for the majority, Justice Scalia employed historical analysis and concluded that the common law forfeiture exception “applied only when the defendant engaged in conduct designed to prevent the witness from testifying.” 261 Id. at 359. The Court noted that acts of domestic violence are often intended to dissuade victims from seeking outside help and include efforts designed to prevent testimony to law enforcement. 262 Id. And where “such an abusive relationship culminates in murder,” the evidence may support the requisite evidence of intentionality; the state courts, which did not consider the defendant’s intent because they erroneously viewed it as irrelevant, were free to consider evidence of his intent on remand. 263 Id. at 377.

Justice Ginsburg joined Justice Souter’s partial concurrence, which was consistent with her own preference for a pragmatic rather than purely historical approach. 264 Justices Ginsburg and Souter joined all but Part II-D-2 of the Court’s opinion, see id. at 379 (Souter, J., concurring in part), presumably because it criticized the dissent for considering the “underlying values” served by the Sixth Amendment, rather than limiting the Amendment’s scope to “the trial rights of Englishmen,” id. at 375–76. Justice Souter emphasized practical reasons why the Court’s holding was necessary for fair trials.  That is, under the alternative approach, admissibility of the statement would turn on the occurrence of the very crime for which the defendant was on trial—“evidence that the defendant killed would come in because the defendant probably killed,” and “[e]quity demands something more than this near circularity before the right to confrontation is forfeited.” 265 Id. at 379 (Souter, J., concurring in part). He observed that history was less significant given that “today’s understanding of domestic abuse had no apparent significance at the time of the framing.” 266 Id. at 379–80.

Additionally, in three cases about the admissibility of evidence relating to forensic tests where the expert who performed the test was not subject to cross-examination at trial, the Justice took the position that evidence about the tests’ results was “testimonial” and thus inadmissible unless the accused had the opportunity to cross-examine the expert who conducted the tests. In Melendez-Diaz v. Massachusetts, the Justice joined a 5-4 decision in which the Court held that a state forensic analyst’s lab report is inadmissible when the analyst does not testify subject to cross-examination at trial. 267 557 U.S. 305, 306, 309 (2009). She also wrote the majority opinion in Bullcoming v. New Mexico, a follow-on 5-4 decision in which the Court held that the defendant had the right to confront an analyst who had prepared a “testimonial” report attesting to the blood-alcohol level of a defendant in a drunk-driving case, and that the analyst’s report could not be admitted into evidence absent an opportunity for cross-examination. 268 564 U.S. 647, 651–62 (2011). And in Williams v. Illinois, the Justice joined a dissent from the Court’s fractured decision finding no Confrontation Clause violation where an expert who was not involved in a semen analysis nevertheless was permitted to testify that DNA in semen from a rape victim matched the defendant’s DNA. 269 567 U.S. 50, 56–57 (2012). As Justice Kagan argued in her dissent, which was joined by Justices Ginsburg, Scalia, and Sotomayor, under a straightforward application of Melendez-Diaz and Bullcoming the admission of this evidence without an opportunity to cross-examine the lab analyst who prepared the DNA profile violated the Confrontation Clause. 270 Id. at 118, 121–23 (Kagan, J., dissenting). The Williams decision muddies the waters in this area, as five Justices actually disagreed with Justice Alito’s analysis in the plurality opinion. Justice Thomas agreed only that admission of the statements did not violate the Confrontation Clause, but “share[d] the dissent’s view of the plurality’s flawed analysis.” Id. at 104 (Thomas, J., concurring in the judgment). As Justice Kagan observed: “[I]n all except its disposition, [Justice Alito’s] opinion is a dissent: Five Justices specifically reject every aspect of its reasoning and every paragraph of its explication.” Id. at 120 (Kagan, J., dissenting).

Justice Ginsburg did agree with the prosecution’s argument in a few post-Crawford cases, but in those instances the result was uncontroversial, or she took a narrow position. For instance, in Davis v. Washington, the Court unanimously held that statements made during a 911 call in an ongoing emergency are “non-testimonial” and thus do not implicate the Confrontation Clause; by contrast, the Court held (8-1) that written statements in an affidavit to the police during the investigation of a crime (and not an ongoing emergency) are inadmissible unless the defendant is afforded an opportunity to cross-examine the affiant. 271 547 U.S. 813, 815, 828–30 (2006); id. at 840 (Thomas, J., concurring in the judgment in part and dissenting in part). And in Ohio v. Clark, the Court unanimously held that a three-year-old’s statements to a teacher about sexual abuse are non-testimonial and thus admissible even where the child is not subject to cross-examination. 272 576 U.S. 237, 240 (2015). Justice Alito’s opinion for the Court, however, contained quite a bit of dicta likely calculated to plant seeds for further erosion of Crawford. Accordingly, Justice Ginsburg joined Justice Scalia’s opinion concurring in the judgment, which was considerably  narrower  than  the  majority  opinion  and  criticized  the  “calculated dicta.” 273 Id. at 254 (Scalia, J., concurring in the judgment). He pointed out that at common law, young children were considered incompetent to take oaths, and that the conversation entailed a three-year-old being asked questions by teachers at school and thus didn’t “have the requisite solemnity necessary for testimonial statements.” 274 Id. at 252. This would have been sufficient to decide the case, but he wrote separately “to protest the Court’s shoveling of fresh dirt upon the Sixth Amendment right of confrontation so recently rescued from the grave in Crawford”—and in particular, aspects of the majority opinion that seemed to water down Crawford and hint at resurrecting the discarded Roberts “indicia of reliability” test. 275 Id. at 252–54. Because the Crawford doctrine tends to involve historical analysis, it does not necessarily require confrontation in every situation where cross-examination might be important to truth-seeking. For instance, many studies show that young children are particularly suggestible. See, e.g., Stephen J. Ceci & Maggie Bruck, Jeopardy in the Courtroom: A Scientific Analysis of Children’s Testimony 233 (1995) (concluding “preschool-age children [are] disproportionately more vulnerable to suggestion than either school-age children or adults”); Stephen Ceci, Amelia Hritz & Caisa Royer, Understanding Suggestibility, in Forensic Interviews Regarding Child Sexual Abuse: A Guide to Evidence-Based Practice 141, 149 (William T. O’Donohue & Matthew Fanetti eds., 2016) (explaining that while recent studies show that older children are more suggestible than younger children in certain circumstances, these are largely irrelevant to the criminal justice system).

C. Death Penalty Jurisprudence

In a 2017 discussion at Stanford, Justice Ginsburg said: “If I were queen, there would be no death penalty.” 276 Ruth Bader Ginsburg, Assoc. Just., U.S. Sup. Ct., Stanford Rathbun Lecture (Feb. 6, 2017), https://awpc.cattcenter.iastate.edu/2017/10/05/stanford-rathbun-lecture-february-6-2017 [https://perma.cc/RSN9-7CZR]. That was surely her long-held belief. Although most of her advocacy was directed toward gender equality, while at the ACLU, she worked on an amicus brief in Coker v. Georgia, in which the Court held that capital punishment was cruel and unusual in rape cases where an adult victim was not killed. 277 433 U.S. 584, 592 (1977); Brief Amici Curiae of the American Civil Liberties Union et al., Coker, 433 U.S. 584 (No. 75-5444), 1976 WL 181482; see also Sidney Harring & Jeffrey L. Kirchmeier, Scrupulous in Applying the Law: Justice Ruth Bader Ginsburg and Capital Punishment, 7 N.Y.C. L. Rev. 241, 242 (2004) (explaining that before Ruth Bader Ginsburg became a Justice, “[h]er only documented encounter with the death penalty came when she co-authored an amicus brief in a capital case as a volunteer attorney for the ACLU while a law professor at Rutgers Law School”). As a judge on the D.C. Circuit, Justice Ginsburg had not had to decide capital-punishment cases; she first faced them on the Supreme Court, where decisions often must be made in emergency, imminent life-or-death situations where defendants are seeking last-minute stays of execution. During her tenure on the Court, public support for both the death penalty and the number of executions declined a great deal, and there have been fewer of these stay motions in recent years. However, at the time of my clerkship they were not uncommon, and we were often studying motion papers filed just hours before an execution was scheduled to take place. The Justice approached these stay motions with particular care and voted for stays in many cases over the years. 278 Harring & Kirchmeier, supra note 277, at 263 & n.158 (“Considering cases without opinions, significantly, Justice Ginsburg so far has voted in the minority to grant stays of executions or deny applications to vacate stays of executions in more than 150 cases while on the Supreme Court.”).

One of Justice Ginsburg’s first death penalty opinions, a dissent penned during my clerkship, reflects her grave concerns about the fairness of capital punishment proceedings. Romano v. Oklahoma was a 5-4 decision in which the Court let a death sentence stand even though the jury was told that a prior jury had already sentenced the defendant to death for a separate crime. (After the trial, the conviction in the other case was reversed on appeal, so the issue was not academic.) 279 512 U.S. 1, 2–6, 14 (1994). In Caldwell v. Mississippi, the Court had overturned a death sentence procured after the prosecutor told the jury in closing that their decision was not final because it would be reviewed by an appellate court; this argument impermissibly suggested to the jury that responsibility for determining whether the defendant would live or die ultimately rested elsewhere. 280 472 U.S. 320, 323, 341 (1985). Justice Ginsburg (joined by Justices Blackmun, Stevens, and Souter) would have reversed the death sentence in Romano because it ran afoul of Caldwell’s principle—the jury’s consideration of evidence that a prior jury had already sentenced the defendant to death likewise raised a “grave” “risk of diminished jury responsibility.” 281 Romano, 512 U.S. at 19 (Ginsburg, J., dissenting).

As a general matter, Ginsburg’s approach to capital punishment was in keeping with her broader strategic gradualism—she preferred narrow rulings that tracked precedent and eschewed grand pronouncements that might not withstand the test of time. In argued cases, the Justice did occasionally vote to uphold death sentences, but mostly voted in favor of capital defendants. And over time, her jurisprudence did become bolder (as her decision in Ring, which overruled a decision that was only twelve years old, illustrates 282 See Ring v. Arizona, 536 U.S. 584, 588–89 (2002) (overruling Walton v. Arizona, 497 U.S. 639 (1990)); supra notes 187, 205–213 and accompanying text. ). Two of her decisions during my clerkship seem particularly cautious and guarded as compared to her votes and opinions in later capital-punishment cases, perhaps in part because this was a new area, and she was being careful to find her footing and pick her battles.

For instance, in Victor v. Nebraska, Justice Ginsburg voted with a 7-2 majority to uphold the conviction of two capital defendants who challenged the validity of the reasonable-doubt instructions at their trials. 283 511 U.S. 1, 26 (1994). The jury instructions were slightly different in the two cases, but the defendants challenged language about “moral evidence” and “moral certainty,” and argued that equating “substantial doubt” with “reasonable doubt” and a reference to “strong probabilities” understated the government’s burden. 284 Id. at 10, 19, 22. Just a few years earlier, in Cage v. Louisiana, the Court had held that another reasonable doubt instruction using “moral certainty” language violated the Due Process Clause. 285 498 U.S. 39, 40–41 (1990). Nonetheless, the Court distinguished Cage and found that the instructions in Victor passed muster. 286 Victor, 511 U.S. at 15–16.

Justice Ginsburg concurred in part and concurred in the judgment, saying she believed “the instructions adequately conveyed to the jurors that they should focus exclusively upon the evidence . . . and that they should convict only if they had an ‘abiding conviction’ of the defendant’s guilt.” 287 Id. at 23–24 (Ginsburg, J., concurring in part and concurring in the judgment). The Justice went on to critique three aspects of the instructions as “unhelpful,” though not in themselves unconstitutional: (1) the term “moral certainty”; (2) the definition of reasonable doubt as “such a doubt as would cause a reasonable and prudent person, in one of the graver and more important transactions of life, to pause and hesitate before taking the represented facts as true and relying and acting thereon”; and (3) the Victor instruction counseling that guilt may be found upon “strong probabilities, . . . provided such probabilities are strong enough to exclude any doubt of his guilt that is reasonable.” 288 Id. at 24–25. Unfortunately, despite this criticism, such “hesitate to act” instructions remain common. For instance, most district judges in the Southern District of New York, where I practice, use a form of this language in criminal trials. Justice Ginsburg instead proposed that courts employ a specific definition of reasonable doubt recommended by the Federal Judicial Center, which she found “clear, straightforward, and accurate.” 289 Id. at 27. The challenged instructions do seem quite problematic, particularly in light of the Court’s earlier Cage decision. Moreover, the Justice’s concurrence highlights reasons why some of the language she critiqued is quite misleading and could confuse juries. I suspect that if this case had come before the Court in the 2000s, rather than 1994, Justice Ginsburg might well have voted to overturn the convictions.

Another interesting case from her first Term is Simmons v. South Carolina, where the Justice voted to take a narrow approach to reverse a death sentence. 290 512 U.S. 154 (1994). In Simmons, the issue was whether the Due Process Clause was violated when a trial judge refused to instruct the jury in the penalty phase of a capital trial that under state law the defendant was not eligible for parole. 291 Id. at 156. A plurality held that “where the defendant’s future dangerousness is at issue, and state law prohibits the defendant’s release on parole, due process requires that the sentencing jury be informed that the defendant is parole ineligible.” 292 Id. The plurality opinion by Justice Blackmun, which Justice Ginsburg joined, focused on the fact that in fairness, when the State argues that a defendant should be put to death because he is a future danger to society, the defendant should be able to rebut that argument by pointing out that if the jury spared his life, he could never be released on parole under state law. 293 See id. at 164–65.

In a separate concurrence, Justice Ginsburg emphasized that when the government “urges a defendant’s future dangerousness as cause for the death sentence, the defendant’s right to be heard means that he must be afforded an opportunity to rebut the argument.” 294 Id. at 174 (Ginsburg, J., concurring). She noted her agreement with Justice O’Connor’s point in the latter’s separate concurrence in the judgment that “due process does not dictate that the judge herself, rather than defense counsel, provide the instruction”—as long as “the relevant information is intelligently conveyed to the jury. 295 Id. (citing id. at 177–78 (O’Connor, J., concurring in the judgment)). The Justice did not read Justice Blackmun’s opinion for the Court to suggest otherwise. 296 Id. at 174–75. However, this view that attorney argument would have solved the problem put her at odds with Justices Souter and Stevens, who would have required the trial court to instruct the jury about life without parole. They pointed out that “arguments of counsel generally carry less weight with a jury than do instructions from the court.” 297 Id. at 172 (Souter, J., concurring) (quoting Boyde v. California, 494 U.S. 370 (1990)). This point has particular resonance when one is considering an argument made by a criminal defense lawyer, given juries’ likely tendency to trust prosecutors more than defense lawyers. But perhaps Justice Ginsburg sought to emphasize common ground with the three conservative Justices in the majority (Chief Justice Rehnquist and Justice Kennedy joined Justice O’Connor’s concurrence in the judgment), who typically supported the prosecution’s position in death penalty cases.

Six years later, the Justice wrote the majority opinion in Shafer v. South Carolina, which strongly reaffirmed Simmons and at least implicitly supported the right to a jury instruction on the fact that a life sentence does not carry any possibility of parole. 298 532 U.S. 36, 48–51 (2001). Notably, Justice Ginsburg’s opinion in Shafer was for a 7-2 majority—and with the full, unequivocal support of the same seven who voted with her in Simmons, including the three conservative Justices who had previously espoused a narrower position. 299 Id. at 39.

Unlike other Justices who opined the death penalty was always unconstitutional, 300 See, e.g., Baze v. Rees, 553 U.S. 35, 81 (2008) (Stevens, J., concurring in the judgment) (“The time for a dispassionate, impartial comparison of the enormous costs that death penalty litigation imposes on society with the benefits that it produces has surely arrived.”); Callins v. Collins, 510 U.S. 1141, 1145 (1994) (Blackmun, J., dissenting from the denial of certiorari) (“From this day forward, I no longer shall tinker with the machinery of death . . . . It is virtually self-evident to me now that no combination of procedural rules or substantive regulations ever can save the death penalty from its inherent constitutional deficiencies.”); Gregg v. Georgia, 428 U.S. 227, 230 (1976) (Brennan, J., dissenting) (“Death . . . serves no penal purpose more effectively than a less severe punishment; therefore the principle inherent in the Clause that prohibits pointless infliction of excessive punishment when less severe punishment can adequately achieve the same purposes invalidates the punishment.”); id. at 231 (Marshall, J., dissenting) (“The death penalty, I concluded, is a cruel and unusual punishment prohibited by the Eighth and Fourteenth Amendments.” (citation omitted)). Justice Ginsburg never wrote an opinion suggesting that capital punishment is per se unconstitutional. She explained it this way:

I’ve always made the distinction that if I were in the legislature, there’d be no death penalty . . . I had to make the decision was I going to be like Brennan and Marshall who took themselves out of the loop [by saying the death penalty was always unconstitutional]. There have been some good death penalty decisions. If I took myself out, I couldn’t be any kind of contributor to those. 301 U.S. Supreme Court Justice Ruth Bader Ginsburg, Death Penalty Skeptic, Has Died, Death Penalty Info. Ctr. (Sept. 19, 2020) (alterations in original) (quoting Ed Whelan, Justice Ginsburg as Chatty Cathy—Part 3, Nat’l Rev. (Aug. 25, 2014), https://www.nationalreview.com/bench-memos/justice-ginsburg-chatty-cathy-part-3-ed-whelan [https://perma.cc/U73N-66WW]), https://deathpenaltyinfo.org/news/u-s-supreme-court-justice-ruth-bader-ginsburg-death-penalty-skeptic-has-died [https://perma.cc/5V7D-GPQP].

Indeed, her votes in favor of capital defendants were often critical to the outcome. According to the Death Penalty Information Center, Justice Ginsburg “voted in the majority of every 5-4 Court decision this century that favored capital defendants and death-row prisoners” and since 2015, “no death sentenced prisoner has ever received a stay of execution from the Supreme Court without her vote.” 302 Id. Some of her more important votes were against capital punishment for juveniles, people with intellectual disabilities, and crimes in which no victim had died. 303 See Kennedy v. Louisiana, 554 U.S. 407, 412 (2008) (ruling on crimes not resulting in the death of a victim); Roper v. Simmons, 543 U.S. 551, 555 (2005) (ruling on juveniles); Atkins v. Virginia, 536 U.S. 304, 306 (2002) (ruling on people with intellectual disabilities).

Late in her career the Justice did move in the direction of arguing that capital punishment was unconstitutional. She joined a 2015 dissenting opinion by Justice Breyer raising multiple reasons it is “highly likely that the death penalty violates the Eighth Amendment” and suggesting that the Court reconsider the question, as well as a similar 2020 Breyer dissent. 304 See Barr v. Lee, 140 S. Ct. 2590, 2592 (2020) (Breyer, J., dissenting); Glossip v. Gross, 576 U.S. 863, 908 (2015) (Breyer, J., dissenting). Perhaps if she had lived longer, Justice Ginsburg would have come to the conclusion that capital punishment is inherently unconstitutional; and perhaps she would have seen still further erosion of public support among Americans for the death penalty, which might someday make that question moot and the result longer lasting.

*    *    *

In my judgment, the cases discussed here are fairly representative of Justice Ginsburg’s jurisprudence on criminal law and procedure. They illustrate that her approach in this area is consistent with the gradualism that characterized her advocacy and judicial philosophy, as well as her pragmatism, and what President Clinton aptly described as her progressive outlook. She was attentive to the need for meaningful enforcement of the Constitution’s protections for liberty and for the rights of the accused. Her opinions reflect a desire to ensure that people have fair notice of when their behavior could be criminal and to guarantee that criminal trials are fair. And she cared more about the real-world consequences of the Court’s decisions than more formalistic notions such as rote adherence to common law principles or overly literal readings of statutory language. There were some areas, such as the Sixth Amendment jury trial and confrontation rights, in which her decisions were a bit bolder and less “gradualist.” At the end of the day, the Justice’s decisions on criminal law and procedure were aimed at moving us toward realizing the promise of the Constitution, but by means she believed were more likely to produce lasting progress. In that regard, in this area (as in others) Justice Ginsburg made significant contributions to achieving the Constitution’s goals of “Liberty,” “Justice,” and “a more perfect Union.” 305 U.S. Const. pmbl.

Justice Ginsburg with Alexandra A.E. Shapiro in 1993.
Justice Ginsburg with her clerks, including Alexandra A.E. Shapiro, in 1993.