Legacy and Accountability


Part of me wished that Chief Justice John Roberts had voted with the majority to overrule Roe v. Wade.1 While I support the constitutional right to choose, if the right is restricted, having it fall by a 6-3 vote would have helped preserve Ruth Bader Ginsburg’s legacy. Justice Roberts’ willingness to uphold substantial parts of Roe makes it clearer that Ginsburg’s choice not to retire when a pro-choice president held office was a but-for cause of the Court eliminating this foundational constitutional right.

There were many points in the past when Justice Ginsburg might have prudently retired, but a particularly pivotal moment was a lunch she had with President Obama at the White House in July 2013.2 At the time, the Justice was eighty years old and had already served on the Supreme Court for twenty years. At the lunch, The New York Times reported that Obama “did not directly bring up the subject of retirement,” but he did mention “the looming 2014 midterm elections and how Democrats might lose control of the Senate.”3

At the time of this lunch, Justice Ginsburg had survived two different bouts of cancer: colon cancer when she was sixty-six years old and pancreatic cancer when she was seventy-six, just four years earlier.4 She knew—or should have known—that she might not survive until another pro-choice president took office. She also knew that, if she were succeeded by a Republican-nominated Justice, Roe would be at risk. Her decision not to retire while President Obama was in office was reckless.

It pains me to write those words because Justice Ginsburg continues to be a hero of mine. She deserves honor for the barriers she overcame to become a lawyer. More importantly, she deserves honor for what she accomplished as a litigator and then as a judge and a Justice. In 1973, she made the first of six oral arguments before the Supreme Court in Frontiero v. Richardson.5 This case exemplifies Ginsburg’s strategy of attacking rules that disadvantage men as a vehicle for inducing a male-dominated judiciary to see the equal protection violation of sex-based classifications.6 I recommend that readers pause to hear the powerful words of a forty-year-old Ginsburg (which can be found starting at minute seventeen of the oral arguments).7

After ascending to our highest court, Justice Ginsburg authored a number of lasting, memorable opinions. In United States v. Virginia, her majority opinion struck down Virginia Military Institute’s male-only admission policy, with words that reaffirmed and strengthened our nation’s commitment to equal protection under the law.8 In her first twenty years on the Court, she authored important controlling opinions on everything from insider trading and environmental standing to disability discrimination and sentencing guidelines.9 She also impacted the law through her dissents. In 2007, she departed from convention and read her dissent in Ledbetter v. Goodyear Tire & Rubber Co. from the bench.10 In that case, the majority overturned a plaintiff’s $360,000 wage discrimination award because the claim was made more than 180 days after the defendant’s original decision about her pay.11 Justice Ginsburg’s dissent criticized the majority for underplaying the often secret nature of salaries: “In our view, the court does not comprehend, or is indifferent to, the insidious way in which women can be victims of pay discrimination.”12 Her dissent explicitly called for the decision to be legislatively overturned, writing, “the ball is in Congress’ court . . . to correct this Court’s parsimonious reading of Title VII.”13 Two years later, Congress responded to her invitation, passing the Lilly Ledbetter Fair Pay Act.14

Given these manifold accomplishments, why on earth would I think it appropriate to cast an aspersion on this “feminist icon” who has contributed so, so much to our nation? The perversity that one who has done so much to secure constitutional rights for women would be a cause of women losing a foundational constitutional right is not in itself an adequate reason for writing. Instead, I do so for two independent reasons. First, the Ginsburg example is a cautionary tale that might help others avoid tainting their legacy by ill-considered actions late in life. The law can help or hinder us in making legacy preserving choices. This Essay will suggest ways—including information escrows—that the law could better dissuade individuals from taking actions that might tarnish and counterbalance the good they’ve done earlier in their lives. Second, this history of Justice Ginsburg’s succession highlights the need to make the Supreme Court more democratically accountable, and additionally underscores the importance of adopting “regular Presidential appointment” reforms.

Tainted Legacies

Justice Ginsburg’s refusal to retire in a timely fashion to my mind puts her in conversation with another personal hero: Ralph Nader.15 Nader’s tireless advocacy for consumer protection by itself has likely saved thousands of lives. His efforts to shine the spotlight on dangerous car designs with his 1965 publication, Unsafe at Any Speed, his recruitment of law students (“Nader’s Raiders”), and his founding of the watchdog group, Public Citizen, to scrutinize government regulation, elevates him—like Ginsburg—into my pantheon of twentieth-century public servants.16

However, his choice to run for the presidency in 2000 was a likely but-for cause of our 2003 invasion of Iraq and ensuing occupation of that country.17 My causal inference requires the following two steps: First, if Nader had not run in 2000, Gore would have won Florida and become president.18 Second, Al Gore would not have preemptively invaded Iraq on false claims of weapons of mass destruction. Nader justified his candidacy by claiming that there was no meaningful difference between Al Gore and George W. Bush,19 but Gore did not have the same inclination to complete the unfinished business of the first Bush presidency by removing Saddam Hussein from power by invading Iraq. I agree with Bernie Sanders’ view that the Iraq War, which sacrificed the lives of nearly 5,000 U.S. troops and more than 150,000 civilians20 and cost more than $2 trillion,21 was “one of the worst foreign policy blunders in the history of the United States.”22

The Iraq War (caused at least in part by Nader’s candidacy) and the impending loss of Roe (which, was caused at least in part by Ginsburg’s decision not to retire during the Obama presidency) are both devastating consequences driven by poor choices. However, in one sense, the loss of Roe is more blameworthy because it was more foreseeable. Or, in terms of tort law, more of a “proximate cause.” Nader could have foreseen the first step in the causal chain—that his presidential candidacy would throw the election to Bush—but 9/11 and the impetus to send troops into Afghanistan and Iraq were much less foreseeable. In comparison, the causal chain regarding Justice Ginsburg’s failure to retire—that she would be replaced by a conservative Justice who would provide the fifth vote needed to overturn Roe—was already on many people’s minds.23 Given this foreseeability, part of me feels as if Ginsburg herself cast the fifth vote to overturn Roe.24

A take-home lesson for us all is to ensure that our choices in later stages of our lives do not end up marring our legacy, potentially undoing—or even outweighing—our earlier achievements. Ginsburg and Nader serve as examples that highlight the importance of carefully considering our actions towards the end of our lives. These are periods in which people often have more power and resources to do harm. Having reached the peaks of their careers, they may be more prone to hubris and fail to realize when their capacities are deteriorating.

We should be particularly on guard to resist the mentality that “I alone can fix it.”25 While we should honor Justice Ginsburg’s plentiful accomplishments throughout her life, we should at the same time be willing to analyze with some critical distance the value of her contributions relative to those of her likely successor had she retired after twenty years.26 She remained in those final years a powerful voice. In the 2016 case of Whole Woman’s Health v. Hellerstedt, Justice Ginsburg argued that a TRAP law (“Targeted Regulation of Abortion Providers”) unconstitutionally burdened women’s choice, stating that, “[I]t is beyond rational belief that H.B. 2 could genuinely protect the health of women, and certain that the law ‘would simply make it more difficult for them to obtain abortions.’”27 While memorable and persuasive, her opinion was merely a concurrence.28 I am at a loss to find crucial votes or opinions in the years after that fateful White House lunch where she provided substantial value over the foreseeable actions of an Obama-nominated successor.29 It is a cruel irony that Justice Ginsburg’s decision to stay on the Court made it more likely that “women in desperate circumstances,” as she had said in Hellerstedt, “may resort to unlicensed rogue practitioners, faute de mieux, at great risk to their health and safety.”30

The concern about clinging to power or overstaying one’s welcome is not just a problem of the oligarchs. I have colleagues who cling to their tenure to the detriment of their institutions, not consciously out of selfishness, but because they hubristically overestimate their current abilities. An even more common-place phenomenon is seniors who refuse to relinquish their car keys. I had painful discussions with my eighty-three-year-old father in encouraging him to abstain from driving. Even after he had multiple fender benders, he was disinclined to give up the independence of movement that driving afforded. But his driving unreasonably endangered others, so I asked him to think about what would happen if his driving caused someone else to be seriously injured or killed. He came to understand that such an event would forever change his legacy: the way his friends and family remembered him.

Both drivers and teachers tend to have “illusory superiority.”31 One study found that 93 percent of Americans believed they were better than average drivers.32 Another found that 90 percent of college professors thought they were above-average teachers.33 It should not be surprising that these cognitive biases become more pronounced as their abilities decline with age. Empirically, there is fairly definitive evidence that driving ability tends to decline with age as seen in the following graph from the Insurance Institute for Highway Safety:34
While it is broadly understood that younger (particularly younger male) drivers are more likely to be involved in fatal accidents,35 the startling extent to which driving ability declines in later years is at times discounted by the elderly themselves. It is easy to believe that you are the exception.

Can the Law Help Improve Decision-Making?

The most obvious lesson of these histories is that individuals should be careful in the later stages of their lives not to act in ways that will tarnish their legacies. In particular, we might ask whether status quo biases cause us to remain in our current positions for longer periods of time. Steve Levitt published a study, leveraging his Freakonomics fame, in which he induced more than 20,000 people who sought help making decisions to be guided by a randomized coin flip.36 If the coin came up heads, they were instructed to take action; if tails, they were told not to take action, i.e., maintain their status quo.37 Levitt discovered that those who got heads were more likely to act than those who did not and subsequently reported greater levels of happiness.38 In other words:

“[I]ndividuals who are told by the coin toss to make a change are more likely to make a change . . .  and [are] happier six months later than those whose coin toss instructed maintaining the status quo. This finding suggests that people may be excessively cautious when facing life-changing choices.”39

The results held for important decisions where subjects asked whether they should quit their job or retire.40

Private employers might also be able to reduce the tendency of employees to overstay their welcome. To take just one example: Reverse mentoring is a process in which younger employees are paired with more senior executives “to mentor them on various topics of strategic[, technological, or] cultural relevance.”41 A potential payoff of such mentoring is to let senior employees see that younger employees are capable and that the company can be safely left in their hands.

Beyond these personal and private prophylactics, this Part explores the extent to which the law can help or hinder these behavioral lapses. Different contexts might require interventions. To begin, consider how the law might respond to the risk of minority-party candidates, like Ralph Nader, impacting the course of elections. While there is little that the law could or should directly do to discourage candidates from running, there are straightforward interventions that would reduce the impact of such a candidacy. For example, in 2020, Edward Foley worried about the analogous possibility “that Howard Schultz, the former CEO of Starbucks, might run as an independent, siphon votes from the Democratic nominee and cause President Donald Trump to win reelection.”42 Foley cogently argued that an achievable solution to the “Howard Schultz problem” would simply be for “battleground states . . . to adopt ranked-choice voting—or, if they prefer, a conventional runoff—in presidential elections.”43 Under either ranked-choice voting or a runoff system, voters who initially expressed their preference for a Nader or Schultz (or a Ross Perot in 1992) would be given the opportunity to express their preference between the two major party candidates. Ranked-choice voting in Florida in the 2000 presidential election might have led to a future that avoided the Iraq War, notwithstanding Nader’s reckless run.

Could the law also respond to the Ginsburg problem of employees choosing to continue their employment against their better judgment? I believe the answer is unequivocally “yes.” To begin, the law could allow employers to impose mandatory retirement ages. In view of the earlier graph demonstrating the increased risk of the elderly behind the wheel, it should not be surprising that pilots older than sixty-five cannot be licensed to fly commercial aircraft.44 The Age Discrimination in Employment Act of 1967 (“ADEA”) has analogous exemptions that allow mandatory retirement for firefighters, law enforcement officers, and “bona fide executives.”45 However, as amended, the Act does not allow institutions of higher education to impose mandatory retirement ages on professors.46

Short of allowing mandatory retirement ages, the ADEA hinders the ability of employers to provide the kind of feedback that employees might value when considering whether to retire. Employers must often rely on the self-awareness of their senior employees for these decisions. Justice Oliver Wendell Holmes, for example, waited until he was ninety years old to retire.47 He was “determined to stay on the Court as long as his health permitted or until his colleagues told him that ‘he was no longer as good.’”48 But for his strategy to work, a colleague must be willing to convey such an idea to him without fearing the consequences of a negative reaction. It is a daunting task for junior colleagues to tell their esteemed seniors that they are slipping. This is in some ways the quintessential “difficult conversation.” Few of us are well equipped to broach the subject, especially because it often entails some degree of transparent self-interest. Younger colleagues would essentially be asking their senior colleagues to cede some degree of control, and both parties know that retirement provides opportunities for promotion. It becomes even more difficult for institutions to foster such communications when the ADEA raises the possibility that such a conversation would itself constitute a form of age discrimination.

To mitigate this problem, employment law might do well to emulate a feature in the way we regulate driver’s licenses. All fifty states and the District of Columbia have review or reexamination processes for driver’s licenses to ensure public safety on the roads.49 Failing the process can lead to revocation or suspension of one’s driver’s license.50 In forty-six states, review by licensing authorities can be triggered by a report from family, friends, or any other citizen.51 In a handful of states, including Maryland and Texas, a review can even be triggered by anonymous referrals, or “reports from individuals who do not provide their name[s].”52 For this system to be effective, it is crucial that the target of the report (i.e., the driver) cannot ascertain the identity of the reporter. It is reasonable for the Department of Motor Vehicles (“DMV”) to require that the reporter disclose their identity to the agency to deter false reporting. For example, Connecticut requires third-party reporters to submit a notarized affidavit—“made under oath[] and subject to penalty of false statement”—attesting to their personal observation of the driver.53 It is likely better, though, for DMVs not to pass on the identity of third-party reporters to the drivers themselves.

Nonetheless, the discomfort of reporting family, friends, and colleagues is so great that DMVs and employers could also consider deploying “information escrow” mechanisms as a way to facilitate more frequent and accurate reporting. An information escrow is a mechanism that gives individuals the opportunity to deposit information with an escrow agent who will only disclose it under pre-specified conditions.54 Such escrows have been used in a wide variety of circumstances. For example, Tinder is a kind of shared-interest information escrow. When a Tinder user swipes right, the user’s interest is held in escrow and only communicated if it is matched by a similar private report of interest by the other individual.55 Information escrows can be particularly effective at overcoming first-mover disadvantages to reporting. For example, survivors of sexual misconduct are at times reluctant to be the first to report. The reporting platform, Callisto, responds to this reluctance by allowing survivors to deposit private information about the misconduct on the platform and then releases this information for potential investigation only if the platform receives a matching deposit naming the same perpetrator.56

An analogous reporting platform might notify employees or drivers that their performance is in decline.57 An escrow mechanism encourages reporting in these situations for some of the same reasons that it does in the context of sexual misconduct. An individual reporter might be more worried about resentment, retaliation, or disbelief when their report is uncorroborated by others. A pair of reports demonstrating a pattern of behavior is stronger evidence than a single report and gives individual reporters more confidence and certainty about their observations. For example, seeing a friend drive imprudently once might not convince me that their license should be suspended. But if I knew that another person had also observed a similar incident, I might feel more comfortable sending a message.58 A Callisto-like escrow mechanism does just that. An individual can deposit a concern knowing that it will only be released if it is seconded by another person.

Like the driver’s license review processes of some states, an escrow mechanism can be structured to preserve the anonymity of the reporter vis-à-vis the reported, while still giving the recipient some control over who is allowed to report. For example, a Justice might only designate a subset of clerks or other Justices to report when they believe it is time for that Justice to step down, and specify that this feedback would only be delivered, anonymously, if a certain number concurred in that judgment.59

However, the law should also be concerned with the possibility that employers create environments that inappropriately pressure workers into retirement. In response to this concern, Congress amended the ADEA in 1990 by adding the Older Workers Benefit Protection Act (“OWBPA”), which establishes specific requirements for a “knowing and voluntary” release of ADEA claims.60 To qualify as a valid waiver under the OWBPA and associated Equal Employment Opportunity Commission (“EEOC”) compliance guidelines, a waiver must satisfy the following seven factors to mitigate concerns of undue pressure.61 To wit: a valid waiver must:

(1) be written in a manner that can be clearly understood; 

(2) “specifically refer[] to rights or claims arising under” the ADEA;

(3) advise the employee in writing to consult an attorney before accepting the agreement;62

(4) provide the employee with at least twenty-one days to consider the offer;

(5) give an employee seven days to revoke his or her signature;

(6) not include rights and claims that may arise after the date on which the waiver is executed;

(7) be supported by consideration in addition to that to which the employee already is entitled.63

While OWBPA is well structured to provide a safe harbor for employers and employees to pursue jointly desired severance agreements, including incentivized early retirement packages, it does not help waive potential liability for managers or co-workers who might otherwise be inclined to send direct or escrow messages suggesting that it is time to retire. In particular, the sixth requirement, that waivers not include “claims that may arise after the date [of] the waiver,” prohibits a senior worker from waiving future claims.64 Under current law, a senior worker could say that they welcome advice from colleagues or managers about when they should retire and then turn around and sue at a later date.

The EEOC would do well to provide further guidance about the extent to which employees can make a knowing and voluntary release of ADEA claims with respect to prospective advice about retirement timing, including how an information escrow mechanism could be deployed. In making this determination, the agency could look at the variety of ways that information escrows have been used in the adoption context. When adopted children reach adulthood, many seek to learn the identity of their biological parents. “[F]orty-one . . . states have implemented some version of shared-interest escrows that allows adoptees to initiate intermediated contact.”65 Some of these states presume consent of the biological parent, “requiring the parent to file an affidavit to prevent the release of identifying information at the adoptee’s request.”66 Other states presume nonconsent, so the “identifying information is not released unless the biological parent files an affidavit in advance affirmatively consenting to disclosure.”67 A third set of states appoints “‘confidential intermediar[ies] to contact biological parents.”68 “The intermediaries only forward contact information” to the child if their biological parent indicates consent upon learning that the child has applied.69

At a minimum, the EEOC should allow an employer to create a retirement escrow with presumed nonconsent. Under such a system, an employee would receive anonymous suggestions to retire from a group of people that the employee preselected only if the employee affirmatively opted into the mechanism. An employer should be protected from liability for setting up this system and notifying its workforce that they have the option to participate if they wish.70

Providing a system that assures anonymity can be important to securing participation, Emily Bazelon, in assessing why Justice Ginsburg “refused to step down,” wrote:

Dorothy Samuels, a former legal editorial writer for The New York Times, conducted interviews for a book on Ginsburg starting in 2018. She asked friends and former clerks of the justice to look back to the period in 2013 and 2014. “I was struck by how many people I spoke with, including friends, acquaintances and former clerks, felt she should have resigned at the time and that her staying on was terribly self-centered — a view I share,” Samuels emailed me. “I was also struck that normally forceful advocates I spoke with would not express their dismay on the record while she was alive.”71

If these friends, acquaintances, and former clerks could have spoken anonymously (and possibly via an information escrow that assured their retirement advice would only be forwarded if seconded), Justice Ginsburg would have at least been able to consider candid advice from people she respected.

The central claim of this Part has been that the law can be a help or a hinderance regarding the behavioral lapses that predictably lead people to remain on the job longer than is socially and personally optimal. I have sketched ways that age discrimination law could better accommodate the ability of institutions to help guide senior employees to make better decisions. But accountability is—at its core—a personal issue, and the cautionary tales of my heroes, the notorious RBG and Ralph Nader, illustrate that we should each think carefully about how a poor decision could all too easily wipe out the good of a lifetime. With that in mind, I hereby declare that:

I welcome anonymous or non-anonymous messages from friends, family, colleagues or administrators alerting me that it is time to leave off particular activities (including, say, driving, teaching, or writing about particular topics) or even to retire.72 I intend and promise not to sue on the basis of such messages.

Even if the foregoing is not legally binding, it may still be effective in assuring others that I welcome their advice and that I expect to be held in less esteem if I complain through litigation—or otherwise—about receiving such messages.

Personal vs. Democratic Accountability

The primary thrust of this Essay has been about personal accountability. Individuals fail to be accountable to their larger life commitments when they overstay their time on the public stage. Justice Ginsburg’s choice not to retire jeopardized what she believed was a foundational constitutional right for all women in the United States. There is, however, another perspective that focuses not on the personal accountability of individual Justices, but instead on the democratic accountability of the Court itself. Justice Ginsburg may have felt that it was her duty not to play politics with the timing of her retirement, that Justices should trust the political process to, over time, produce a Court responsive to the will of the Demos.

From this point of view, a second lesson from the Ginsburg affair is the urgency of making the Court more democratically accountable. Ryan Doerfler and Samuel Moyn have written a series of pieces suggesting ways to democratize the Supreme Court.73 They consider a number of Congressional interventions—including jurisdiction stripping, supermajority rules, and a legislative overrides session—that would disempower the Court in deciding particular cases from departing from the will of the people.74 At a structural level, the composition of the Court can be made more democratically accountable simply by limiting the terms of Justices to eighteen years.

Advocates for limiting Justices’ effective service on the Supreme Court to eighteen years have marshalled multiple cogent arguments and debated multiple means of achieving their ends. Akhil Amar, for example, has proffered “18 arguments for 18 years,” which at least implicitly includes the rationale of enhanced democratic accountability.75 Framing these proposals in terms of forced retirement could focus attention on how the proposals disable Justices from strategically timing their retirement or serving beyond their best years. These same proposals might alternatively be framed as ways to enable presidents to make regular appointments. A central component of most of the eighteen-year proposals is that a president would be able to nominate a new Justice every two years, coinciding with the beginning of each new Congress.

The failure of our current system to assure regular appointments means that the political representation of Justices serving on the Supreme Court can drift far from the democratic will, as articulated through presidential elections. For example, former-President Obama served eight years and made two appointments,76 while President Trump—at the time of this writing—has served four years and made three appointments.77 In contrast, under the regular appointment proposals, Obama would have made four appointments and Trump only two. These proposals would need to handle the eventuality that a Justice dies or ends their service before finishing eighteen years of service, and the potential need to make interim appointments could cause the Court’s political composition to drift from what would otherwise occur with biennial nominations.78 But as an empirical matter, the risk of death and early retirement is exceedingly small. Indeed, Lewis Powell was the last Justice to serve fewer than eighteen years, when he retired after fifteen years and 150 days in 1987.79 So it has been more than thirty-five years since this issue has arisen, and with improvements in medicine, this risk is likely to continue to decline.80

Some have already described the ways in which regular appointments could make the Court more democratically accountable.81 However, here I provide an empirical comparison of what the composition of the Court might have looked like in the past if we counterfactually had had a system of regular two-year appointments (that is, nominations and confirmations). While counterfactual thought exercises are often speculative and contestable, they can offer useful illustrations of what could have been. In this simulation, I assume that adopting this reform would not change the winners of the presidential election, and that those presidents succeeded in having nominees confirmed and sworn into service. I also assume that a system of regularized appointments was fully in place in 1987, meaning that the composition of the Court in 1987–89 would be determined by the presidents serving in the previous years (i.e., Nixon/Ford would have appointed three, Carter would have appointed two and Reagan would have appointed four Justices). I analyze the ensuing eighteen Congresses to compare the actual composition of the Court to the hypothetical composition under a system of regular appointments every two years. The results of this analysis are shown in the following table:

This table provides some stylized evidence of the extent to which the composition of the Court has strayed from what would have been if nominations were distributed more evenly with regard to presidential wins. As summarized in the right-most column, over these nineteen Congressional sessions, Justices nominated by Democratic presidents would have occupied a majority of the Court 31.6 percent of the time (five out of nineteen sessions) with an average representation of 3.8 Justices nominated by Democratic presidents. This starkly contrasts to what actually occurred: There was never a time during this period in which a majority of Justices were appointed by Democratic presidents (and the actual average representation was only 2.6 Justices). Indeed, one would need to go back to 1969 (when Abe Fortas retired) to find a time when Justices nominated by Democrats held at least five seats on the Court). Hence, for most practicing lawyers, including myself, Republican-nominated Justices have held a perpetual majority on the Court throughout our entire lives.

While proponents of regularized appointments frame the issue as an apolitical improvement that should garner bipartisan support, as an empirical matter, the table shows that non-regular appointments have starkly favored Republicans. In addition to cementing a seemingly perpetual majority, the final row shows that, in 84.2 percent of the Congressional sessions analyzed (sixteen out of nineteen), the actual representation of Republican-nominated Justices exceeds what it would have been with regularized appointments. And this misrepresentation is likely to get even worse. While forecasting the make-up of the Court into the distant future is not feasible (given the uncertainty of future presidential elections), it is possible to project that the current 6-3 majority of Republican-nominated Justices is likely to persist at least until the next Congressional session starting in a few months, even though that is the period in which regularized nominations would have produced a six-member Democrat-nominated majority. In other words, we are months away from the largest deviation between the Court’s actual nominated composition and what we would have had with regularized appointments over the last several decades.

Notwithstanding the empirical tilt of the current system toward Republican-nominated Justices, it is important to see how much more fragile a 6-3 majority might be with regularized nomination. The current 6-3 Republican-nominated majority would last at least fifteen more years if all the current Justices serve until they reach the same age (eighty-seven years old) that Justice Ginsburg was when her death ended her service. If, during that fifteen-year period, a Republican president was elected, the two oldest currently serving Republican-nominated Justices could substantially extend the Republican Party’s power by strategically retiring during that term. In contrast, what would have been a 6-3 Democrat-nominated majority starting in 2023 would be lost just four years later if the Republican Party won the 2024 presidential election and thereby placed two new members on the Court in 2025 and 2027, returning the Court to a 5-4 majority of Republican-nominated Justices.


In thinking about why Republican-nominated Justices have continually held a majority of seats on the Supreme Court, it is natural, by focusing on the relatively recent deaths of Justices Ginsburg and Scalia, to conclude that Republicans have more successfully manipulated the existing system. As adumbrated above, Justice Ginsburg played the game poorly by gambling her life expectancy. Senate Majority Leader Mitch McConnell played the game well by refusing to hold hearings for Merrick Garland. Both Justices died during the terms of presidents from opposing parties, but only one vacancy resulted in a political shift.

Another important reason for the continuous hold of Republican-appointed Justices on the Court is simply that Republican-appointed Justices began the analyzed period with a substantial majority of the seats, and then, Justices of all political stripes had a tendency to retire when the presidency was held by the same political party that held the office when they were nominated. Indeed, during this period only three Justices ended their service by dying in office. Of those who retired, nearly two-thirds retired during the tenure of the president of the same political party.

The idea that one branch of government could be continually held by one party poses a grave risk to the ideals of our democracy. As I draft this Essay, the January 6th Committee has been holding public hearings concerning the efforts of Donald Trump to remain as president. We faced the fearful possibility that sitting Vice Presidents, simply by refusing to certify electoral ballots from states won by the opposition party, could effectively ensure their party’s perpetual control of the executive branch. That incumbent party would then be able to choose the political identity of its successor. One of the most chilling lessons of this Essay is how closely this outcome already obtains with regard to the judicial branch. Yes, it was reckless for Justice Ginsburg to remain on the Court beyond the Obama administration. Indeed, the consequences of her decision are even more serious with the present democratic deficit of the current appointment process because her mistake may continue to impact the composition of the Court even if Democratic candidates win a substantial majority of future elections.82 But the far greater wrong is that the current regime allows such pronounced and long-lived deviations of the Court’s composition from the will of the people.


  1. [1]. See Roe v. Wade, 410 U.S. 113 (1973), overruled by Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022). I drafted and submitted this Essay before Dobbs was decided and originally said, “Part of me hopes that Chief Justice John Roberts votes with the majority to overrule Roe v. Wade.

  2. [2]. Susan Dominus & Charlie Savage, The Quiet 2013 Lunch That Could Have Altered Supreme Court History, N.Y. Times (Sept. 25, 2020), [].

  3. [3]. Id.

  4. [4]. Steven Petrow, Cancer Survivor: Don’t Forget That Accolade When it Comes to Ruth Bader Ginsburg’s Remarkable Life, Stat (Sept. 22, 2020), [].

  5. [5]. Ruth Bader Ginsburg: A Reading List: Arguments before the Supreme Court, Libr. at Wash. & Lee Univ. Sch. of L. (Dec. 7, 2020, 4:01 PM), [].

  6. [6]. See id.

  7. [7]. Oral Argument at 17:00, Frontiero v. Richardson, 411 U.S. 677 (1973) (No. 71-1694), []. Though female Justices are more likely to be interrupted during oral argument than their male counterparts, advocate Ginsburg faced a different form of disparate treatment that day: She was the only lawyer arguing not to receive a single question from the bench. Tonja Jacobi & Dylan Schweers, Justice, Interrupted: The Effect of Gender, Ideology, and Seniority at Supreme Court Oral Arguments, 103 Va. L. Rev. 1379, 1381–82 (2017).

  8. [8]. United States v. Virginia, 518 U.S. 515, 556–58 (1996).

  9. [9]. See, e.g., United States v. O’Hagan, 521 U.S. 642, 647 (1997) (insider trading); Friends of the Earth, Inc. v. Laidlaw Env’t Servs. (TOC), Inc., 528 U.S. 167, 173–74 (2000) (environmental standing); Olmstead v. L.C. ex rel. Zimring, 527 U.S. 581, 587 (1999) (disability discrimination); Kimbrough v. United States, 552 U.S. 85, 91 (2007) (sentencing guidelines).

  10. [10]. See Robert Barnes, Over Ginsburg’s Dissent, Court Limits Bias Suits, Wash. Post (May 30, 2007), [].

  11. [11]. Ledbetter v. Goodyear Tire & Rubber Co., 550 U.S. 618, 643 (2007).

  12. [12]. Barnes, supra note 10.

  13. [13]. See Ledbetter, 550 U.S. at 661 (Ginsburg, J., dissenting).

  14. [14]. Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111-2, 123 Stat. 5 (2009); see also Equal Pay Act of 1963 and Lilly Ledbetter Fair Pay Act of 2009, EEOC (Apr. 29, 2014), [].

  15. [15]. For a discussion regarding Ralph Nader’s achievements, see Ralph Nader, Acad. of Achievement (Feb. 9, 2022), [].

  16. [16]. See Ralph Nader, Encyc. Britannica, []; see also Ralph Nader, Unsafe at Any Speed: The Designed-In Dangers of the American Automobile (1965); Pub. Citizen, [].

  17. [17]. See generally Timeline: The Iraq War, Council on Foreign Rels., [] (discussing the United States invasion into Iraq and ensuing occupation).

  18. [18]. See Michael C. Herron & Jeffrey B. Lewis, Did Ralph Nader Spoil Al Gore’s Presidential Bid? A Ballot-Level Study of Green and Reform Party Voters in the 2000 Presidential Election, 2 Q.J. Pol. Sci. 205, 206 (2007).

  19. [19]. See Steven Greenhouse, The 2000 Campaign: The Green Party; Nader, in Harlem, Attacks Gore and Bush With Gusto, N.Y. Times (Nov. 7, 2000), [] (“Repeating a theme of his campaign, Mr. Nader called Mr. Gore and Mr. Bush Tweedledee and Tweedledum and said the two candidates took identical positions on many issues and were both beholden to the corporations that finance much of their campaigns.”),

  20. [20]. Neta C. Crawford & Catherine Lutz, Watson Inst. for Int’l & Pub. Affs., Human Cost of Post-9/11 Wars 1 (2019), [].

  21. [21]. Daniel Trotta, Iraq War Costs U.S. More Than $2 Trillion: Study, Reuters (Mar. 14, 2013, 9:49 AM), []. An updated study released in 2021, which combines Iraq and Syria in its estimates, calculates the cost at $2.058 trillion. Neta C. Crawford, Watson Inst. for Int’l & Pub. Affs., The U.S. Budgetary Costs of the Post-9/11 Wars 3 (2021), [].

  22. [22]. Adam Taylor, Bernie Sanders Said Iraq Was ‘The Worst Foreign Policy Blunder’ in U.S. History. Really?, Wash. Post (Oct. 14, 2015, 10:30 AM), [] (internal quotation marks omitted). Of course, Nader’s decision to run was not the only sufficient cause of Bush’s paper-thin electoral college victory. And his candidacy may have had other causal impacts—including Bush’s opportunity to nominate John Robert to replace retiring Chief Justice William Rehnquist and Samuel Alito to replace retiring Justice O’Connor. But those nominations occurred in President Bush’s second term, and it is far from clear whether Rehnquist and O’Connor would have retired if Gore had been president.

  23. [23]. In both cases, the untoward consequence needed the subsequent actions of third parties, which might be deemed superseding causes that reduce the culpability of Ginsburg and Nader. For example, if Senate Majority Leader Mitch McConnell had given Supreme Court nominee Merrick Garland a fairer process, Ginsburg’s decision to remain on the Court might not have been as consequential. See generally Robin Bradley Kar & Jason Mazzone, The Garland Affair: What History and the Constitution Really Say About President Obama’s Powers to Appoint a Replacement for Justice Scalia, 91 N.Y.U. L. Rev. Online 53 (2016) (noting how the Senate Republican’s plan to block Garland’s nomination was not in accordance with historical tradition).

  24. [24]. Ginsburg is at worst one of many but-for causes for the elimination of reproductive rights in several American states. There is a dysfunctional tendency of humans (including me) to focus after the fact on a single cause (“if only this person had done that”) and occlude from our moral vision others who bear responsibility. The normative challenge for us all is to keep in mind the responsibility that Ginsburg and Nader bear while simultaneously appreciating the responsibility of others who causally contributed to the outcome.

  25. [25]. See generally Yoni Appelbaum, ‘I Alone Can Fix It, Atlantic (July 21, 2016), [] (for quoted language).

  26. [26]. Ginsburg might have been drawn to longer service in part to exercise the power granted to the Justice with the longest tenure in the majority of assigning majority opinions. It wasn’t until Justice Ginsburg was eighty-five years old—five years after her Obama lunch—that she assigned her first Supreme Court opinion. Mark Joseph Stern, A Milestone for Ruth Bader Ginsburg, Slate (Apr. 18, 2018, 3:54 PM), [].

  27. [27]. Whole Woman’s Health v. Hellerstedt, 136 S. Ct. 2292, 2321 (2016) (Ginsburg, J., concurring) (citing Planned Parenthood of Wisc., Inc. v. Schimel, 806 F.3d 908, 910 (2015)), abrogated by Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022); Linda Greenhouse & Reva B. Siegel, The Difference a Whole Woman Makes: Protection for the Abortion Right After Whole Woman’s Health, 126 Yale L.J.F. 149, 151 (2016).

  28. [28]. Hellerstedt, 136 S. Ct. at 2320–21 (Ginsburg, J., concurring).

  29. [29]. Other candidates for late-career contributions include her 2014 dissent in Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 751 (2014) (Ginsburg, J., dissenting) (criticizing “a for-profit corporation’s qualification for a religious exemption from a generally applicable law”), and her majority opinion in the 9-0 decision, Timbs v. Indiana,139 S. Ct. 682, 686–87 (2019) (holding that the Eighth Amendment’s ban on excessive fines applies to states and local governments).

    In considering her late-career contributions, it is also appropriate to acknowledge concerns with some of the intemperate ex cathedra comments that Justice Ginsburg made in her final years on the Court. Of particular concern are the repeated criticisms that she leveled at presidential-candidate Donald Trump. As a characteristic example, on CNN, she had called Trump a “faker” and said, “He has no consistency about him. He says whatever comes into his head at the moment.” Joan Biskupic, Justice Ruth Bader Ginsburg Calls Trump a ‘Faker,’ He Says She Should Resign, CNN (July 13, 2016, 7:45 AM), []. If she were a district or circuit court judge, her comments might have appropriately subjected her to discipline for violating Canon 5 of the Code of Judicial Conduct, which says a federal judge “should not . . . publicly endorse or oppose a candidate for public office.” U.S. Cts., Code of Conduct for United States Judges 18 (2019), []; see also Michael Biesecker, Ginsburg’s Public Trump Critique Raises Ethical Quandaries, AP News (July 13, 2016), []. Though she apologized for these comments, as well as for dismissively characterizing Colin Kaepernick’s protests as “dumb and disrespectful,” one might conclude that in her last years she at times did not exhibit appropriate judicial temperament. See Jessica Taylor, Ginsburg Apologizes For ‘Ill-Advised’ Trump Comments, NPR (July 14, 2016, 10:44 AM), []; Katie Couric, Ruth Bader Ginsburg on Trump, Kaepernick and Her Lifelong Love of the Law, Yahoo!News (Oct. 10, 2016), [].

  30. [30]. See Hellerstedt, 136 S. Ct. at 2321 (Ginsburg, J., concurring) (emphasis in original).

  31. [31]. See Vera Hoorens, Self-Enhancement and Superiority Biases in Social Comparison, 4 Eur. Rev. Soc. Psych. 113, 117 (2011) (providing an overview of illusory superiority).

  32. [32]. Ola Svenson, Are We All Less Risky and More Skillful Than Our Fellow Drivers?, 47 Acta Psychologica 143, 146 (1981).

  33. [33]. K. Patricia Cross, Not Can, But Will College Teaching Be Improved?, 17 New Directions for Higher Educ. 1, 1 (1977).

  34. [34]. Fatality Facts 2020: Older People: Population and Mileage Rates, Ins. Inst. for Highway Safety & Highway Loss Data Ins. (May 2022), [].

  35. [35]. The disproportionate dangerousness of young male drivers was at issue in Craig v. Boren, 429 U.S. 190, 192 (1976), where the Supreme Court struck down an Oklahoma statute that required males to be three years older before legally drinking 3.2 percent beer. Three years after she successfully argued that government discrimination against males offended the Equal Protection Clause, Ginsburg filed an amicus brief on behalf of the American Civil Liberties Union arguing, in part, that the sex-based classification was not sufficiently supported by statistical evidence–even though in the relevant age category (eighteen-twenty years old), males were more than eleven times more likely than females to be arrested for driving while under the influence. Id. at 201 (“.18% of females and 2% of males in that age group were arrested for that offense.”); Motion of American Civil Liberties Union for Leave to File Brief Amici Curiae and Brief Amicus Curiae, Craig v. Boren, 429 U.S. 190 (1976) (No. 75-628), 1976 WL 194246, at *25.

  36. [36]. Steven D. Levitt, Heads or Tails: The Impact of a Coin Toss on Major Life Decisions and Subsequent Happiness, 88 Rev. Econ. Stud. 378, 379 (2020).

  37. [37]. Id. at 383.

  38. [38]. Id. at 379–80.

  39. [39]. Id. at 378.

  40. [40]. Id. at 383 tbl.1.

  41. [41]. Jennifer Jordan & Michael Sorell, Why Reverse Mentoring Works and How to Do It Right, Harv. Bus. Rev. (Oct. 3, 2019), [].

  42. [42]. Edward B. Foley, Want to Fix Presidential Elections? Here’s the Quickest Way, Politico Mag. (May 4, 2019), [].

  43. [43]. Id.

  44. [44]. 49 U.S.C. § 44729(a) (2020) (“[A] pilot may serve in multicrew covered operations until attaining 65 years of age.”) Once a pilot reaches sixty years old, they must be paired with another pilot; the maximum age for single-pilot operations is sixty. Id. § 44729(c). The sixty-five years old standard was set by the International Civil Aviation Authority (“ICAO”) and later adopted by the FAA on June 7, 2018. Fed. Aviation Admin., Fair Treatment of Experienced Pilots Act (The Age 65 Law) Information, Questions and Answers (2019), [].

  45. [45]. 29 U.S.C. § 623(j) (exemption for firefighters and law enforcement officers); 29 U.S.C. § 631(c) (exemption for bona fide executives or high policymakers).

  46. [46]. Age Discrimination in Employment Amendments of 1986, Pub. L. No. 99-592, 100 Stat. 3342 (1986); see also 29 U.S.C. § 623(m); Monica Ferrelli, Tracing the Conception and Meaning of the Age Discrimination in Employment Act: Where Are We with Mandatory Retirement?, 12 J. Socio. & Soc. Welfare 326, 330 (1985); Nat’l Rsch. Council, Ending Mandatory Retirement for Tenured Faculty: The Consequences for Higher Education 1–4 (P. Brett Hammond & Harriet P. Morgan eds.,1991).

  47. [47]. See Theodore Voorhees, The Resignation of Justice Holmes (Part I), 63 A.B.A. J. 260, 260 (1977).

  48. [48]. Theodore Voorhees, The Resignation of Justice Holmes (Part II), 63 A.B.A. J. 426, 427 (1977).

  49. [49]. Kathy H. Lococo, Kathy J. Sifrit, Jane Stutts & Loren Staplin, Nat’l Highway Traffic Safety Admin., D.O.T. H.S. 812 402, Medical Review Practices for Driver Licensing Volume 3: Guidelines and Processes in the United States 482–90 tbl.25 (2017), [].

  50. [50]. Id. at 566–70 tbl.47.

  51. [51]. See id. at 491–95 tbl.26.

  52. [52]. Id. at 517–25 tbls.34–35.

  53. [53]. Id. at 50–52.

  54. [54]. See Ian Ayres & Cait Unkovic, Information Escrows, 111 Mich. L. Rev. 145, 150 (2012).

  55. [55]. See Ian Ayres, Meet Callisto, the Tinder-Like Platform that Aims to Fight Sexual Assault, Wash. Post (Oct. 9, 2015), [].

  56. [56]. Id. The possibility of using an information escrow to combat sexual misconduct is discussed extensively in Ayres & Unkovic, supra note 54. Ayres is currently a member of the Callisto Board of Directors. The original implementation of Callisto Campus would automatically, upon matching, disclose the identity of survivors to the school’s Title IX office for investigation. See Mission + Vision, Callisto (2022), []. The current implementation of Callisto, upon matching, discloses the information of survivor identity to a “Legal Options Counselor” who can then advise survivors on their paths forward, including a Title IX investigation by their school. Id.

  57. [57]. This possibility is briefly considered in Information Escrows: “For example, a professor might wonder if his colleagues think it is time for him to retire, or if his lectures are boring.” Ayres & Unkovic, supra note 54, at 195.

  58. [58]. Id. at 161 n.56 (“Knowledge of repeated episodes might thus make complainants both more certain of ... improper behavior.”).

  59. [59]. A mechanism of this type was referred to in my original Information Escrows article as an “insecurity escrow”:

    [A]n insecurity escrow might (1) only report a random subset of anonymous responses and (2) only report the random subset if a minimum number of responses is received. Thus, if a professor asked ten colleagues for an opinion, the escrow would only report the anonymous results if at least five colleagues responded and only send on five responses—choosing five at random from the submitted responses.

    Id. at 195–96.

  60. [60]. Older Workers Benefit Protection Act, Pub. L. No. 101-433, 104 Stat. 983 (1990); see also Oubre v. Entergy Operations, Inc., 522 U.S. 422, 428–30 (1998) (failure to comply with OWBPA’s requirements will result in an invalid waiver).

  61. [61]. 29 U.S.C. § 626(f)(1)–(2).

  62. [62]. Id.; see also American Airlines, Inc. v. Cardoza-Rodriguez, 133 F.3d 111, 117–18 (1st Cir. 1998) (finding that “advise the employee[]” means affirmatively “to‘caution,’ ‘warn,’or‘recommend’”) (citing Webster’s Third New World International Dictionary 32 (1986)).

  63. [63]. 29 U.S.C. § 626(f)(1).

  64. [64]. Id. § 626(f)(1)(C).

  65. [65]. Ayres & Unkovic, supra note 54, at 158.

  66. [66]. Id. (citing U.S. Dep’t of Health & Hum. Servs., Access to Adoption Records: Summary of State Laws (2009)).

  67. [67]. Id.

  68. [68]. Id.

  69. [69]. Id. (citing U.S. Dep’t of Health & Hum. Servs., Access to Adoption Records: Summary of State Laws 5, 5 n.9 (2009)).

  70. [70]. The law might go further and allow something akin to the “confidential intermediary” adoption system, where employees could be notified that a colleague has sent a message regarding retirement and whether they consent to receive it. It should be noted that the “confidential intermediary” adoption mechanisms presume (irrefutably) that the biological parent consents to be contacted by the intermediary. One could imagine allowing biological parents to opt out ex ante from such contact, or alternatively presuming that they do not want to be contacted unless they affirmatively indicate a contrary interest.

    As Deputy Dean of Yale Law School, in the spirit of fostering improvement, I had sent an email notifying those professors who had taught a course that students had evaluated in the bottom quintile of general teaching effectiveness. One of my colleagues asked that I never send such emails to them in the future. In retrospect, I should have considered whether it is better to presume consent or nonconsent in receiving this information, and how I could have allowed individual professors to opt in or opt out of such notice. One advantage of most student evaluation platforms is that they create a mechanism for students to send anonymous messages to their professors—including opinions that it is time for the professor to retire—while verifying that the message comes from one of the students who took a particular course.

  71. [71]. Emily Bazelon, Why Ruth Bader Ginsburg Refused to Step Down, N.Y. Times (Sept. 30, 2020), [].

  72. [72]. By Googling the phrase “how to send an anonymous email,” one can find a variety of ways to do so. Non-anonymous messages are likely to carry more weight, as they are more likely to come from a trusted source. This is one of the ways in which mechanisms that allow anonymous messages from groups of individuals pre-designated by the recipient preserve both trust and anonymity.

  73. [73]See “The Court’s Role in Our Constitutional System”: Hearing Before the Presidential Comm’n on the Sup. Ct. of the U.S. 2 (June 30, 2021) (statement of Samuel Moyn) [hereinafter, Statement of Samuel Moyn]. For fuller substantiation of all claims offered, see also Ryan D. Doerfler & Samuel Moyn, Democratizing the Supreme Court, 109 Calif. L. Rev. 1703, 1705–09 (2021); Samuel Moyn, Stop Worrying About Kavanaugh, Liberals. Start Winning the Political Argument., Wash. Post (Aug. 8, 2018, 6:00 AM), []; Samuel Moyn, Resisting the Juristocracy, Bos. Rev. (Oct. 5, 2018), []; Samuel Moyn, The Court Is Not Your Friend, Dissent (Winter 2020), []; Ryan D. Doerfler & Samuel Moyn, Reform the Court, But Don’t Pack It, Atlantic (Aug. 8, 2020), []; Ryan D. Doerfler & Samuel Moyn, Making the Supreme Court Safe for Democracy, New Republic (Oct. 13, 2020), []; Samuel Moyn, Change the Court, For Good: For Starters, the Oldest Justice Should Retire While There’s a Democratic President and Senate, N.Y. Daily News (Jan. 30, 2021, 5:30 PM), []. 

  74. [74]. See supra note 73 and accompanying text.

  75. [75]. See Testimony in Support of a Congressional Statute Establishing an Eighteen-Year Limit on Active Supreme Court Service: Hearing Before the Presidential Comm’n on the Sup. Ct. of the U.S. 5 (July 20, 2021) (statement of Akhil Reed Amar) [hereinafter, Statement of Akhil Reed Amar]; see also Akhil Reed Amar, Episode 16: 18 Arguments for 18 Years (Apr. 28, 2021), []. As to the means of achieving term limits, scholars have debated a variety of ways that term limits might be achieved without Constitutional amendment.
    COTUS-Report-Final-12.8.21-1.pdf [] (discussing the pros and cons of term-limit proposals). There is vast literature on Supreme Court term limits. See, e.g., Reforming the Court: Term Limits for Supreme Court Justices (Roger C. Cramton & Paul D. Carrington, eds. 2006) (collecting essays); Judith Resnik, Judicial Selection and Democratic Theory: Demand, Supply, and Life Tenure, 26 Cardozo L. Rev. 579, 641 (2005); Saikrishna B. Prakash, American’s Aristocracy, 109 Yale L.J. 541, 568–84 (1999) (book review). A new mechanism recently proposed by Richard Re argues that opt-in term limits could be achieved by encouraging Justices to make binding commitments to retire eighteen years after joining the Court (and that such commitments could be conditional on subsequent appointees making similar commitments). Richard M. Re, The Peril and Promise of SCOTUS Resignations, 107 Iowa L. Rev. Online 117, 117–18 (2022).

  76. [76]. Supreme Court Nominations (1789-Present), U.S. Senate, [].

  77. [77]. Id.

  78. [78]. For example, Akhil Amar’s model statutes suggests:

    If any Justice in active service shall take Emeritus status or leave the Court at a time when the total number of remaining active-service Justices shall be nine or more, no Court vacancy shall thereby be created. If, however, any Justice in active service shall take Emeritus status or leave the Court at a time when the total number of remaining active-service Justices shall be less than nine, the vacancy may be filled, upon presidential nomination and Senate confirmation and presidential issuance of a good behavior commission, by a Replacement Justice. This Replacement Justice may continue in active service until displaced by the commissioning of a Regularized Justice in due course whose addition to the Court brings the total number of remaining active-service Justices back to nine; provided that in no event may any Replacement Justice continue in active service for more than eighteen years. If at any time there shall be more than one Replacement Justice, the most junior Replacement Justice shall be the first to be displaced, the next-most junior shall be the next to be displaced, and so on. At the end of his or her active service, a Replacement Justice may elect to remain on the Court by taking Emeritus status.

    Statement of Akhil Reed Amar, supra note 75, at 12–13.

  79. [79]. See Joan Biskupic & Fred Barbash, Retired Justice Lewis Powell Dies at 90, Wash. Post, Aug. 26, 2998, at A1. Under a system of eighteen-year appointments, it is possible that Justice Powell would have remained on the Court two-and-a-half more years to fill out his expected term. We would need to go back more than forty years to find a Justice who did not survive at least eighteen years after beginning their Court service – as Abe Fortas died in 1982 just seventeen years after taking office. But see The Texts of Letters and Statements Involving the Resignation of Justice Fortas, N.Y. Times, May 16, 1969, at 20 (Fortas resigned three-and-a-half years after taking office in the shadow of an ethics scandal). See also Linda Greenhouse, Ex-Justice Abe Fortas Dies at 71; Shaped Historic Rulings on Rights, N.Y. Times, Apr. 7, 1982, at A1.

  80. [80]. Of course, we might worry about Justices strategically retiring before serving their eighteen years in order to impact the composition of the Court, but such incentives could be dampened by limiting the service of their replacement to a sufficiently short period. See Statement of Akhil Reed Amar, supra note 75, at 9.

  81. [81]. See, e.g., Re, supra note 75, at 127 (“[F]ederal judicial term limits have broad, bipartisan support because they would help regularize, democratize, and depoliticize the Court. If each presidential term generated roughly the same number of judicial appointments, then the bench would more closely mirror democratic trends ... .”) (footnote omitted).

  82. [82]. Kartikeya Kandula and I ran Monte Carlo simulations to estimate the future composition of Justices assuming that: (1) sitting Justices would never retire before serving 18 years in office or reaching the age of seventy-eight, whichever comes first, and would thereafter strategically retire as soon as a President of their same party was in office; (2) Justices who did not retire could probabilistically leave office by dying (with age-contingent death probabilities taken from actuary tables); and (3) replacement Justices would take office at age fifty (roughly the confirmation age of Justices Brown Jackson and Coney Barret). Ian Ayres and Kart Kandula, How Long is a Republican-Nominated Majority on the Supreme Court Likely to Persist?, Balkinization (July 3, 2022, 11:38 PM), []. We estimated the likely longevity of a Republican-nominated majority with different probabilities of Democratic success at winning the White House. As shown in the figure below, we found that even if Democrats have a 70 percent chance of winning future presidential elections, a Republican-nominated majority on the Supreme Court is likely to persist for more than twenty-five years:



Oscar M. Ruebhausen Professor, Yale Law School. Eni Iljazi and Jacqueline Huang provide excellent assistance.

I thank Sai Prakash and Richard Re for helpful comments.