Term Limits and Turmoil: Roe v. Wade’s Whiplash

A fixed eighteen-year term for Supreme Court Justices has become a popular proposal with both academics and the general public as a possible solution to the countermajoritarian difficulty and as a means for depoliticizing the confirmation process. While scholars have extensively examined the potential benefits of term limits, the potential costs have been underexplored. We focus on one cost: the possible effects of term limits on doctrinal stability. Using seven statistical models that measure potential fluctuation in Supreme Court support for Roe v. Wade had the Court been operating under term limits since 1973, we explore the level of constitutional instability that a term-limit system would engender. Our models incorporate varying degrees of each new Justice’s loyalty to the nominating president’s ideology and deference to precedent, as well as account for the Senate’s level of influence on the confirmation process under conditions including the elimination of the filibuster. The results suggest that term limits could fundamentally change the way that the law evolves and might well lead to a substantial loss in doctrinal stability.

Introduction

There’s a new kid in town trying to resolve the countermajoritarian difficulty and its consequences. Rather than proposing increasingly elaborate theories to limit the Supreme Court’s discretion, some scholars, politicians, and pundits have decided to welcome the Court to the bar of politics. Worried about the Court’s tendency to act as a political body—that is, as nine legislators in black robes decreeing policy^[1]—they propose to treat it as one by imposing term limits as a substitute for periodic elections. That way, they reason, every president will be able to appoint at least two Justices, and the Court will reflect a mix of the political views of the last three or four presidents and thus of the American people as a whole. It will, in short, no longer be so strongly countermajoritarian.

Unlike many other proposals to limit judicial discretion using theories of interpretation, the idea of term limits has no clear ideological agenda and has been supported by scholars on both the left and the right. Perhaps most striking in today’s era of partisan politics, a public-opinion poll found over 70% support among Democrats and Republicans for imposing eighteen-year term limits.^[2] Support may have increased following the Republican Senate’s decision not to grant a hearing to President Obama’s Supreme Court nominee Merrick Garland, which added to the perception that Supreme Court appointments depend on political power plays rather than on a reasoned process.^[3] Term limits, proponents argue, would eliminate the need for this type of game playing, as appointments would take place on a regularized schedule.^[4] Term limits would also lower the stakes for each nomination, perhaps inducing presidents to favor excellence over youth and ideology.

Are term limits, however, the white knight that they first appear to be? While the benefits of term limits are both intuitive and well explored by distinguished scholars,^[5] they have a potentially grave downside: they arguably undermine legal stability and detrimentally affect the incremental development of constitutional doctrine. This Essay begins assessing this underexplored and underappreciated effect on legal stability and doctrinal development by tracking the fate of one of the Court’s more controversial decisions, Roe v. Wade.^[6] We ask a simple question: What would have happened to Roe over the years if the Justices since 1973 had served under eighteen-year term limits rather than having life tenure?

The Essay answers this question using statistical models that place a new Justice on the bench every two years. The models assume that five variables influence each new Justice’s behavior: the party of the nominating president, the confirming Senate’s party composition and the degree of influence it exerts, and the new Justices’ degree of loyalty to the nominating party and deference to precedent. Using these five variables, the study models the outcomes of a hypothetical reconsideration of Roe v. Wade from 1973 through the end of President Trump’s first term (2019, the last opportunity for him to nominate a Justice under a term-limits scheme). By using probability models to look at the likely effects that term limits would have on an actual holding, we hope to give scholars and policy makers a more clear-eyed view of the effects that term limits might have on current constitutional doctrines and the way in which the law evolves.

Our results reveal that term limits are very likely to have negative consequences for stability. But the size of those consequences depends on the degree of independence that Justices exhibit from the partisan politics of their nominators and on their degree of deference to precedent.^[7] If Justices exhibit low levels of ideological conformity to the views of their nominating president and have a strong deference to precedent, a term-limit system may have minimal effects on long-term doctrinal stability. On the other hand, if Justices show moderate or strong ideological alignment with the views of their nominating president and/‌or little to no deference to precedent, term limits could result in a far less stable constitutional doctrine with major precedents being reversed and then reinstituted from Term to Term.

This Essay proceeds in three parts. Part I begins with a theoretical discussion of the potential for term limits to address the countermajoritarian difficulty, including its underexplored potential impact on constitutional doctrinal stability and how the law evolves. Subpart II(A) discusses the goals of presidents and the Senate in selecting and confirming Justices. Subpart II(B) introduces three statistical models, describing Justices who exhibit moderate, high, or low loyalty—that is, the likelihood that a nominee will vote in line with his or her nominator’s policy preferences. The models evaluate the impact on the stability of constitutional doctrine, specifically Roe v. Wade, when the most senior Justice is replaced every two years. Subpart II(C) discusses the potential stabilizing role of deference to precedent, modeling the difference between high and low deference to precedent. Subpart II(D) delves into the role of the Senate confirmation process and its effect on stability. The loyalty and deference models assume that the Senate has a moderate influence on the confirmation process. This subpart briefly discusses the effect of assuming less influence, and then models two additional scenarios: a Senate with greater influence, and a moderately influential Senate without the possibility of a filibuster. Part III concludes by discussing the broader implications of the study, the model’s limits, and directions for possible future research.

I. Term Limits as the Proposed Antidote to the Countermajoritarian Difficulty

A. The Countermajoritarian Difficulty and Prior Solutions

No matter the label applied—judicial activism,^[8] a lack of institutional legitimacy,^[9] or the “counter-majoritarian difficulty”^[10]—criticisms of the Supreme Court for usurping the will of the majority are widespread and often fervent. The Court’s recognition of a constitutional right to same-sex marriage in Obergefell v. Hodges^[11] is one of the most recent examples of a decision becoming a lightning rod for claims that the Court has overstepped its constitutional bounds.^[12] The bitter political fight over Merrick Garland’s confirmation furthered the public perception that the Court is no more than a political body, making the Court a major issue during the 2016 primaries and general election.^[13] And the political circus of Brett Kavanaugh’s confirmation showed partisan politics at its worst.^[14]

The concern over the Court’s proper role, however, is almost as old as the Constitution itself. Countless books and articles have been written about the tension created by having a Supreme Court with unelected Justices wielding the power of judicial review in a democracy.^[15] The term “countermajoritarian difficulty” can be traced back to Alexander Bickel’s 1962 book The Least Dangerous Branch.^[16] It refers to the challenge of reconciling popular governance and a democratic government with the practice of judicial review that allows nine unelected Justices to overturn the will of the majority based on their interpretation of the Constitution. The countermajoritarian difficulty has been called an “obsession,” a “preoccupation,” and “the dominant paradigm of constitutional law and scholarship.”^[17] In line with this perception of the countermajoritarian difficulty as the central problem in constitutional law, most theories of constitutional interpretation and judicial decision-making are designed to lessen the tension between judicial review and majority rule by cabining judicial discretion in some way. From process theory^[18] to originalism,^[19] from minimalism^[20] and theories of moral rights^[21] to constitutional dualism^[22] and intratextualism,^[23] scholars have proposed ways to prevent judges from substituting their own political preferences for those of the elected branches. The theory of “popular constitutionalism” even goes so far in trying to relieve the tension that it proposes revoking the Supreme Court’s power of judicial review altogether, or at least making it nonbinding.^[24] Whatever the theory, trying to solve the countermajoritarian riddle has become the legal equivalent of proving Fermat’s Last Theorem.

This swirl of competing, and unsuccessful, theories has led some scholars to propose a different sort of solution: Rather than limiting the discretion of the Justices, they would limit the time period that Justices have the ability to use (or abuse) their discretion, thus ensuring that the Justices’ political preferences mirror the political preferences of the nation as a whole.

B. The Term-Limit Solution

A number of scholars have proposed term limits for Supreme Court Justices, and the idea has support from both sides of the political aisle.^[25] One of the most recent, and most concrete, proposals comes from Erwin Chemerinsky. He proposes imposing eighteen-year term limits on Supreme Court Justices, with terms staggered to end every two years.^[26] Chemerinsky argues that eighteen years is long enough for a Justice to master the job, but not so long as to enshrine “political choices from decades earlier.”^[27] The term-limit proposal is meant to walk the tightrope between reducing the countermajoritarian tendencies of the Court and fully democratizing the judiciary.^[28]

This system would result in an appointment to the Supreme Court every two years, or two nominees per presidential term.^[29] This approach has several advantages. First, it removes the fortuity of the timing of Supreme Court vacancies and gives every president an equal opportunity to influence the composition of the Court. Under the current system, Republicans have nominated 73% of Justices since 1973 despite controlling the White House only 55% of the time. The current system has also led to great variability among presidents regarding the number of appointments: President Carter never made an appointment, and President Franklin Roosevelt appointed eight Justices serving a collective 150 years on the bench.^[30] The current system also led to the contentious decision by Republicans not to grant a hearing for President Obama’s nominee Merrick Garland on the ground that the vacancy occurred in the last year of his presidency.

In addition, term limits could relieve the pressure to nominate ever-younger Justices to increase a president’s period of influence, instead giving the president more incentive to nominate the most qualified candidate.^[31] A term-limit regime also could potentially lead to a less contentious nomination process and help eliminate the political problem that occurs when an appointment arises close to a presidential election. Proponents argue that term limits would also reduce the current partisan warfare over judicial appointments: If the opposing party knows there will be another nomination in two years, and that they might be the party nominating, that reduces the stakes and might help facilitate compromise.

Moreover, Chemerinsky argues that not only would term limits solve difficult practical issues, but that they are more consistent with the term length the Founders envisioned.^[32] The average Supreme Court tenure between 1789 and 1970 compared to 1970 to 2005 has nearly doubled, growing from fifteen years to twenty-six years.^[33] In 2015, five sitting Justices had served for over twenty years.^[34] Chemerinsky argues that today’s longer tenure may actually interfere with the democratic balance that the Founders intended to create. While the Supreme Court is meant to be insulated from direct political influences and serve as a counterweight to the majoritarian process, the nomination and confirmation process is meant to be a democratic check on the Court.^[35] As term lengths have increased, making nominations and confirmation hearings less frequent, it is possible that the Supreme Court has become more countermajoritarian than the Framers intended, leading to an imbalance and divorcing the Justices’ views from the current society’s needs and values.^[36] As a concrete example, Gerald Rosenberg has argued that the New Deal standoff between the executive and the judiciary may have been partly a result of increased tenure length that led to a Court out of step with changes in society.^[37] On this view, term limits are not a new, radical intervention, but rather a mechanism to restore the balance the Founders intended.

Term limits also enjoy a broad base of support—many scholars from across the ideological spectrum, over 70% of the public, including a majority of both Democrats and Republicans, and even a current Supreme Court Justice have expressed support for the proposal.^[38]

Yet while the potential benefits of Supreme Court term limits have garnered considerable scholarly attention, the same cannot be said of the potential costs. In particular, a term-limit regime might cause increased instability in constitutional doctrine. Like all judicial reforms, the imposition of term limits would not operate in a vacuum but would interact with judicial decision-making principles such as stare decisis, the common law evolution of judicial doctrine, and the real-world influences of politics and personal predilections. Term limits, combined with the increased politicization of the confirmation process that Chemerinsky advocates,^[39] might reduce the role of precedent in constitutional doctrine, and thus lead to an increase in doctrinal instability.

Only one study has empirically examined the potential differences between a life-tenured Court and a term-limited Court.^[40] The authors concluded that switching to a term-limit system would result in larger shifts in the ideological medians, as measured by Martin–Quinn scores,^[41] and that these shifts would mirror more general political shifts in the country.^[42] Although these findings support the argument that a term-limit regime would increase democratic responsiveness, they also hint that such a regime might increase instability, perhaps to an intolerable degree.^[43] Policy makers considering the adoption of term limits should weigh the cost to stability against the gains in democratic responsiveness. To do so, however, we need to be able to measure the likely amount of instability that term limits (in particular, the staggered eighteen-year term limit Chemerinsky proposes) would introduce into the system. The remainder of this Essay explores that question.

II. Empirically Examining the Effects of Term Limits on Supreme Court Precedent: Term Limits and the Fate of Roe v. Wade from 1973 to 2019

In the abstract it is hard to know to what extent Supreme Court term limits would undermine stability, but without that information, we cannot weigh the benefits against the costs. This study attempts to gain insight into term limits’ impact on legal stability by using a prominent, some would say notorious, Supreme Court case as a test: What would the fate of Roe v. Wade have been under an eighteen-year term-limit system? As contentious as the case is, it has thus far withstood the test of time and survived since it was decided in 1973. Would a term-limit system have changed this?

Roe v. Wade was selected for the model for several reasons. Most importantly, it has become a litmus test for judicial nominees that serves as a proxy for a Justice’s view on a variety of substantive issues and as a “marker of his or her liberalism or conservatism, indicating the way he or she views the role of a judge and the proper approach to the law.”^[44] The case has become so politicized that it was the “central question” surrounding Chief Justice Roberts’s confirmation hearing.^[45] Given that proponents of term limits view judicial decision-making as inherently political, Roe is therefore an ideal test case for assessing the effects of term limits. By studying how term limits might have affected the fate of Roe, we can gain insight into how much doctrinal instability could be produced (at least in politically salient cases) by the imposition of term limits.

All of our models posit that eighteen-year term limits were imposed on Supreme Court Justices prior to 1973, and that the first new Justice following the enactment was due to be appointed in 1975, two years after Roe v. Wade was decided. The Essay then uses statistical models to predict, biennially from 1973 to 2019 (the last year in which President Trump would replace a Justice during his first term in office), the outcome of a case considering the reversal of Roe v. Wade—or, if the model predicts a prior reversal, a subsequent case considering the reinstatement of the principles of Roe.

We recognize that any analysis of term limits must account for the diversity of theories about how Justices make their decisions. The Essay takes this into account by presenting seven distinct scenarios accounting for three factors that to varying degrees have been identified as influencing judicial decision-making: the loyalty of a Justice to the nominating president’s political values, a Justice’s deference to precedent, and the amount of influence exerted by the Senate asked to ratify the nomination. All the models also factor in the nominating president’s party and the party composition of the Senate, using the actual historical data. The strength of our approach is that it allows each reader to consider the assumptions he or she believes are most accurate and most relevant to predicting a Justice’s voting behavior and then calculate the likely impact of term limits on doctrinal stability given those assumptions. Readers, for example, who believe that a Justice’s political ideology plays only a small role should focus on a less loyal model like Model 3. Those who think Justices defer strongly to precedent should turn to Model 5. Readers in between might focus on the moderately loyal model, Model 1, which acknowledges the role of ideology but does not view it as a lockstep influence. Each model thus assigns a value to the likelihood that a Justice would vote to uphold Roe v. Wade (or a subsequent case overruling or reinstating it), depending on the party of the nominating president, the makeup of the confirming Senate, the amount of influence the Senate exerts on the appointment process, the degree that Justices favor deferring to precedent, and the degree of loyalty that Justices exhibit toward their nominating president’s party.

The study then analyzes, under each model, the effect of a term-limit plan that replaces the most senior Justice with a new Justice every two years. In other words, following Chemerinsky’s proposal, the model assumes that every two years after Roe, the then-president (in 1975, President Ford) would replace the most senior Justice (in 1975, William Brennan) with his own nominee; in 1977, then-President Carter would replace the most senior Justice (Byron White) with his nominee; and this process would continue through 2019. As the hypothetical Supreme Court changes every two years, the study then presents the newly constituted Court the opportunity to affirm or overturn Roe (or a subsequent case overruling or reinstating it) under each model’s varying assumptions about the new Justice’s voting proclivities.^[46]

While an actual Court probably would not have chosen to rehear Roe on a biennial basis from 1975 forward, these hypotheticals help inform us how the Court would likely have voted under a term-limit regime had they decided to reconsider Roe in any given year. Considering all seven scenarios from 1975 through 2019 thus provides an informed sense of the real-world impact of term limits on doctrinal stability and allows exploration of how different models of judicial decision-making and doctrinal stability intersect with a term-limit system.

The study uses the Monte Carlo method to estimate the likelihood of affirming or overturning. Briefly, the Monte Carlo method is a statistical method that relies on probability distributions to obtain numerical results and estimate the real-world probability of an event occurring.^[47] In the present study, this method involves creating 10,000 hypothetical Supreme Court votes per case, using nine Justices with the probabilities assigned according to the assumptions of each model. Each Court’s vote is then assigned a 1 or a 0. A 1 is assigned if there are more than five votes to uphold Roe v. Wade, and a 0 if there are fewer. The mean for all 10,000 runs is then used as an estimate of the real-world probability that a panel made up of those nine Justices would uphold (or reinstate) Roe.^[48]

A. The Role of Loyalty: The Nomination Process and the Increased Partisan Divide on the Court

1. The President and Loyalty.—The first set of models is focused primarily on varying degrees of a Justice’s perceived loyalty to the president nominating the Justice (as tempered by Senate confirmation). This factor reflects the assumption that the nominating president is trying to choose a Justice who will uphold the president’s own political values. Attitudinalists have long sought to provide empirical proof that ideological values are the primary driving force behind Justices’ decision-making,^[49] and while the magnitude of the effect found has varied widely and the studies themselves are subject to methodological critiques,^[50] many, if not most, constitutional scholars today agree that ideology plays at least some role in a Justice’s decision-making process.^[51] Importantly, Chemerinsky, one of the major proponents of term limits, counts himself in this camp^[52] and believes that presidents and senators weigh a Justice’s ideology heavily in the nomination and confirmation process.^[53] It therefore seems fair to evaluate the proposal using its own assumptions: that ideology plays an important role in a Justice’s decision-making, that presidents consider ideology in their nomination decisions, and that a correlation exists between the president’s and Senate’s preferred outcomes and the likelihood of the Justice ruling the same way (at least for politically salient issues).^[54]

At least one empirical study has attempted to quantify the decisions of individual Justices on a normalized scale ranging from liberal to conservative.^[55] While this study did not directly measure the correlation between a nominating president’s preferences and the outcomes of the Justice’s decisions, the results are at least suggestive of a correlation between the nominating president’s party and the Justice’s performance on the court; the two most conservative Justices were appointed by Republicans and two of the three most liberal Justices were appointed by Democrats.^[56]

And while notable examples exist of Justices going against the nominating president’s ideology,^[57] the more recent trend is toward conformity: the Court increasingly votes along ideological lines that are predictable and closely aligned with the views and preferences of political parties.^[58] Importantly, political science measurements of the ideology of the Justices have found increasing alignment between partisan political divisions and the Court’s conservative–liberal split and a greater homogeneity in the ideology of Justices nominated by the same party.^[59] An empirical study by Neal Devins and Lawrence Baum discovered an increased linkage between party and ideology since 2010, a linkage they theorize is driven, in part, by increased polarization among political elites, partisan sorting, and growing awareness, acceptance, and perhaps even an embracing of this linkage by nominators and Justices.^[60]

While this acceptance and the role of political elites can most explicitly be seen in President Trump’s campaign promise that his nominees would be picked by the Federalist Society,^[61] the leveraging of political networks to nominate judges and influence the judiciary has been a long-term strategy of the right.^[62] Devins and Baum found that when Justices are arranged from liberal to conservative according to their Martin–Quinn scores from 1937 (the first year for which Martin–Quinn scores are available) to the present, only once prior to 2010 was the Court cleanly divided by the party of Justices’ nominators, and that was when seven Justices had been nominated not only by a president of the same party, but by the same president, Franklin Delano Roosevelt.^[63] For every term since Elena Kagan joined the Court in 2010, however, there has been perfect alignment between the political party of the nominating president and where the Justices fall on the liberal to conservative Martin–Quinn spectrum.^[64] The lack of partisan divide prior to 2010 is also apparent in voting behavior. Remarkably, from 1937 to 2010, in only one of the three hundred twenty-two cases with at least two dissents that are listed as “important” by the Guide to the U.S. Supreme Court did the majority and dissenting Justices divide solely along the lines of the party of their nominating president.^[65] Since 2010, by contrast, at least seven of the Guide’s “important” decisions have been decided strictly along those party lines.^[66]

While the Court did not often divide along partisan lines until 2010, an increasing correlation between the party of the nominating president and the relative conservatism or liberalism of the Justices was evident even before that date.^[67] Jeffrey Segal, Richard Timpone, and Robert Howard compared— from Presidents Franklin Roosevelt through Bill Clinton—the nominating president’s rated liberalism in civil liberties and economics and their appointed Justices’ behavior in these domains. They found a significant correlation in both. In the civil liberty domain, Justices voted liberally in 4.22% more cases for every ten points that their nominating president was rated “more liberal” by presidential scholars, and 3.41% in the economic domain.^[68] According to their model, the presidents’ ratings could explain between 20% and 34% of the Justices’ voting behavior in these domains.^[69] For example, a Justice nominated by President Lyndon Johnson would be expected to vote liberally in about 28% more civil liberty cases than a Justice appointed by President Ronald Reagan, and a Justice nominated by President Franklin Roosevelt would be expected to vote liberally in about 22% more economic cases than a Justice appointed by President Ronald Reagan.^[70]

Our models capture partisan alignment between a Justice and the nominating president by including a variable for “loyalty,” and, as we explain in more detail later, we ran the data using three different levels: moderately loyal, more loyal, and less loyal.

2. Senate Composition and Loyalty.—Of course, even if presidents are choosing more extreme candidates or are more successful in choosing ones aligned with their preferred ideology, Justices must still be confirmed by the Senate. In other words, a Democratic president would like to appoint a Justice who would uphold Roe v. Wade 100% of the time, but his ability to do so is constrained by both the unpredictability of the Justice (loyalty level) and the composition of the Senate. The inverse is true for Republican presidents. The stronger the opposing party’s presence in the Senate, the more the president must temper his or her preferences to get the Justice confirmed.

One of the most prominent theories of Senate influence on the confirmation process is the move-the-median (MTM) model. This theory posits that both the president and the Senate have ideal ideological set points to which each party wants to get as close to as possible.^[71] If the ideology of the nominee is further from a senator’s set point than is the status quo, she should vote against the nominee. If the same is true for forty-one senators (assuming the possibility of filibuster still exists), the nominee should not be confirmed and the president must put forward a nominee closer to the Senate’s set point—even if the nominee is further from the president’s preferred set point.^[72] Recent empirical testing of the MTM model, however, suggests that presidents have been far less constrained by the Senate than this theory would suggest, with the Senate often confirming nominees whom the theory suggests should have been rejected.^[73] Another study similarly found that a Justice moves towards the president’s ideological preference even following what should have been a contested nomination.^[74] For this reason, our study weights Justices as always being at least slightly loyal to the president’s ideology no matter the composition of the Senate.

But recent history makes evident that the Senate can have—and has had—a profound impact on nominations in key, even if isolated, moments, most notably in refusing to consider the nomination of Merrick Garland. Furthermore, regardless of whether the Senate’s influence on Supreme Court nominees is supported empirically, the public and political actors view the Senate’s influence as very real and powerful. The starkest illustration of this is the recent use of the “nuclear option” by the Democrats to eliminate the filibuster for lower court nominees^[75] and then by the Republicans to confirm Justice Gorsuch to the Supreme Court after the Democrats threatened to filibuster his nomination.^[76]

The confirmation process may also be fundamentally changing as the ideological separation between the parties reaches unprecedented levels. According to the National Journal Vote Rankings, in 2013 only two House Republicans (both from New York) were more liberal than the two most conservative Democrats; in the Senate, every Democrat had a more liberal voting record than every Republican and vice versa.^[77] According to DW-Nominate Scores—a measurement of an individual congressman’s ideology calculated from his or her voting behavior—the ideological divide between the parties is the widest since Reconstruction.^[78] More importantly for this study’s purposes, this trend towards separation, which started in the early 1950s, has greatly accelerated since 1975,^[79] largely driven by a disappearance of moderates in both parties.^[80] This is of particular relevance to the nomination process, because moderates are likely to be the difference between a confirmation and a filibuster.^[81] This can be seen in a marked decrease in the percentage of confirmations, an increase in the time before confirmation, and an increase in partisan voting patterns on nominees.^[82] The Justices seem very aware of the shift as well, with Chief Justice Roberts recently stating that neither Justice Scalia nor Justice Ginsburg would be confirmed today.^[83]

Reflecting the view that the Senate has a significant influence on the president’s choice of nominee, Models 1–5 all build in a relatively strong role for the Senate in the nomination process. Under these models, a president is most likely to get a strong ideologically loyal nominee when the president’s party has a supermajority. This probability decreases by 10% if the president’s party is the majority party but lacks a supermajority, by 15% if the Senate is even, by 20% if the opposing party controls the Senate but lacks a supermajority, and by 30% if the opposing party has a supermajority. The possibility that the role of the Senate is overstated or diminishing—especially if the filibuster is permanently abolished—is addressed later in Models 6 and 7.

B. The Effects of Loyalty on Legal Stability in a Term-Limit Regime

The first three models explore the effects of changes to the level of stability based on varying degrees of judicial “loyalty.” A Democratic president is assumed to want to nominate a Justice strongly in favor of upholding Roe v. Wade, while a Republican president is presumed to want a Justice who will favor overturning it. The models simulate moderately loyal, more loyal, and less loyal Justices. The more loyal a Justice is, the more likely he or she is to vote according to the preferences of the nominating president. The model posits that more loyal Justices are 10% more likely to vote in accordance with their nominating party’s preference than moderately loyal Justices. In turn, less loyal Justices are 10% less likely to vote in accordance with their nominating party’s preferences than their moderately loyal counterparts.

Before delving into the results of the models, it is important to note the shortcomings of the probabilities assigned to each Justice. While the exact values for these numbers have not been quantitatively derived, the study attempts to account for the arbitrariness of the exact numbers, if not their general direction, by presenting three different models with varying loyalty levels. While the robustness of the effect varies, the total number of precedent switches is consistent across all three loyalty levels. This suggests that the findings of this study are robust over a range of correlation strengths between the ideologies at work in the nomination process and the Justice’s decisions. Future studies, however, should examine the minimum correlation needed to produce the precedential reversals found in this study.

1. The Moderately Loyal Justices Model (Model 1).—Table 1 summarizes the probabilities assigned to each Justice based on the party of the nominating president, the composition of the Senate, and a moderately loyal Justice. Briefly, a Justice nominated by a Democratic president with a Democratic supermajority^[84] in the Senate was assigned an 85% probability of voting to uphold Roe; a Democratic president with a Democratic majority in the Senate a 75% probability of voting to uphold; a Democratic president with a Republican majority in the Senate a 65% probability of voting to uphold; and a Democratic president with a Republican supermajority in the Senate a 55% probability of voting to uphold. Justices nominated by Republican presidents were assigned probabilities in the same fashion but in reverse: a Republican president with a Democratic supermajority in the Senate was assigned a 45% probability of voting to uphold Roe; a Republican president with a Democratic majority in the Senate a 35% probability of voting to uphold; a Republican president with a Republican majority in the Senate a 25% probability of voting to uphold; a Republican president with a Republican supermajority in the Senate a 15% probability of voting to uphold; and a Republican president with an evenly divided Senate (as in 2001)^[85] a 30% probability of voting to uphold.

Table 1. This table displays the likelihood of a moderately loyal Justice voting to uphold Roe v. Wade depending on the party of the president and the composition of the Senate (“D” = Democrat, “R” = Republican, “DD” = 60 Democrats, “RR” = 60 Republicans).

As Figure 1 below shows, even for moderately loyal Justices the results are striking: Roe v. Wade would very likely suffer legal whiplash. A greater than 70% chance exists that Roe would be overturned in 1987 (and an 87% chance in 1989), a 65% chance that Roe would then be reinstated in 2009, and a 55% chance that it would be re-overturned in 2017 following the Republicans winning the White House. In other words, a term-limit system could have dramatic implications for the Court’s zigzagging, with the Court’s stance on Roe reversing three times in only forty-six years.^[86]

Odds of Overturning Roe v. Wade with Moderately Loyal Justices

Figure 1. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with moderately loyal Justices. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

One might, of course, disagree with our assessment of judicial loyalty, in either direction. Critics of term limits might argue that Justices are actually far more likely to vote in line with the preferences of their nominating president, or be more “loyal.” Proponents of term limits might argue that the values are far too high and Justices in practice are far less “loyal.” Models 2 and 3 address these scenarios.

2. The More Loyal Justices Model (Model 2).—Table 2 below summarizes the probabilities assigned to each Justice based on an assumption that Justices are likely to be more loyal to their nominating president than in the previous model. These more loyal Justices are defined as being 10% more likely to vote in line with their nominating president’s preferences than their moderately loyal counterparts.

Table 2. This table displays the likelihood of a more loyal Justice voting to uphold Roe v. Wade depending on the party of the president and the composition of the Senate (“D” = Democrat, “R” = Republican, “DD” = 60 Democrats, “RR” = 60 Republicans). More loyal judges are considered 10% more likely to vote in line with the preferences of their nominating president than moderately loyal judges.

Figure 2 below shows that, at first glance, whether one posits a moderately loyal or more loyal Justice, the superficial results are remarkably similar; the precedent switches the same number of times over the same time span, and even in the same years. Yet an important difference arises in the overall likelihood that the Court’s decisions will in fact change.^[87] Indeed, between just 1983 and 1985 with the addition of the more loyal Justice, the Court rises from a 14% probability of overturning Roe to a 62% probability of overturning. By comparison, with the moderately loyal condition the largest change in probability in any two years was only 35%, suggesting a more limited chance of actual reversal.

Odds of Overturning Roe v. Wade with More Loyal Justices

Figure 2. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with more loyal Justices. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

In other words, while the precedent switches an equal number of times (three) for each model, the swings are much more extreme, and the likelihood of reversal is much more certain under the high-loyalty model. The magnitude of these swings could fundamentally change how jurisprudence evolves or the likelihood that the Court chooses to grant cert. for any given case. In general, constitutional law, like common law, is thought to evolve slowly through small, deliberate steps rather than rapid sea changes.^[88] In the high-loyalty condition, however, these incremental changes might be bypassed, and the law could be subject to rapid, jolting swings.

3. The Less Loyal Justices Model (Model 3).—Table 3 below summarizes the probabilities assigned to each Justice assuming less loyalty. Less loyal Justices are defined as being 10% less likely to vote in line with their nominating president’s preferences than moderately loyal Justices and 20% less likely than their most loyal counterparts.

Table 3. This table displays the likelihood of a less loyal Justice voting to uphold Roe v. Wade depending on the party of the president and the composition of the Senate (“D” = Democrat, “R” = Republican, “DD” = 60 Democrats, “RR” = 60 Republicans). Less loyal judges are considered 10% less likely to vote in line with the preferences of their nominating president than moderately loyal judges.

As Figure 3 shows, positing low-loyalty Justices appears to ameliorate, but not eliminate, the potential instability created by term limits. The model predicts the same three swings in precedent, with the last one (re-overruling the reinstated Roe) following President Trump’s second nominee during his first term. Perhaps importantly, the Court’s shifts are, however, much more gradual, with the likelihood of a switch within 10% of a coin flip for thirteen of the twenty-three hypothetical cases. This arguably is within a range that would not discourage litigants from bringing cases and within a range that one could foresee a Justice switching his or her views over time or even being swayed by particularly good lawyering.^[89] The gradual nature of the swings might also help keep litigants focused on the case law as their guide rather than solely the composition of the Court. Nevertheless, the fact that the same swings occur under this model helps to confirm the robustness of the overall conclusion that term limits are likely to increase doctrinal instability.

Odds of Overturning Roe v. Wade with Less Loyal Justices

Figure 3. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with less loyal Justices. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

4. Summary: Loyalty and Term Limits.—The results of these three models illustrate the likely instability that would occur under a term-limit regime if Justices exhibit any decision-making loyalty at all towards their nominating president’s views. Even small degrees of loyalty, or ideological alignment, between a president and his or her chosen Justice can introduce substantial doctrinal instability in a term-limit system. Perhaps most surprising, as Figure 4 shows, the degree of alignment doesn’t appear to have a large effect on the net number of swings, but the more aligned a Justice is, the sharper the swings and the more certainty there is that a reversal will occur in any given year. This difference may manifest itself in how opinions are written or whether to grant cert., and may leave more or less room for the impact of good lawyering in determining a case outcome. It also means that ideological drift (not accounted for in our model) could have a far more substantial impact on case outcomes under the less loyal model because the median Justice would be less ideologically rigid.^[90]

Odds of Overturning Roe v. Wade with Varying Degrees of Justice Loyalty

Figure 4. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with either moderately, less, or more loyal Justices. In all three cases the same horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

The loyalty models, while considering one potentially critical factor in how the Court decides cases, do not account for a potential source of doctrinal stability: stare decisis. This factor is addressed in Models 4 and 5.

C. Stare Decisis and Term Limits

Models 4 and 5 are based on each Justice’s deference to precedent. These “stare decisis” models explore the possible role of deference to precedent in providing stability to the system. In the words of Justice Kagan in Kimble v. Marvel Entertainment, LLC,^[91] “[stare decisis] promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.”^[92] Stare decisis is thus a strong candidate to temper the potential swings in doctrine caused by frequent changes in the Court’s ideological preferences brought on by a more frequent change in the Court’s composition.

It is admittedly difficult to quantify the role that stare decisis plays in judicial decisions. Some Justices may pay lip service to stare decisis in their opinions but actually give it little weight in their deliberative process. Still others may defer strongly to precedent in practice but not feel the need to espouse its virtues and therefore leave its role unstated. Some studies have endeavored to quantify the role that precedent plays in the Justices’ decisions. Bailey and Maltzman, for example, used the policy preferences of members of Congress and the president to calculate policy cut-points on Supreme Court cases which they could then use as a control for the role of policy in the decision and better isolate the role of legal factors such as deference to precedent and deference to Congress.^[93] They found a significant role for deference to precedent in thirteen of the sixteen Justices examined, but very high variability among Justices.^[94] Given such high variability, it is challenging to compute its effect on the Court as a whole and may change drastically from one Court to another depending on composition.

But while the precise role of stare decisis may be difficult to capture, we can at least gain a sense of how individual Justices’ views on the importance (or lack of importance) of stare decisis might play out in a term-limited regime. For that limited purpose, we can envision three possibilities: no deference, weak deference, and strong deference.^[95] Models 1 through 3, already discussed, function without accounting for an individual Justice’s deference to precedent, and can thus serve as the equivalent of no deference. The following models present two other possibilities: a moderate version, which assumes weak deference (Model 4), and a model with strong deference (Model 5). Both the weak-deference and strong-deference models use moderately loyal Justices to explore the possible stabilizing effects of stare decisis within a term-limit system. Under the models, a Justice with weak deference is 5% less likely to vote against precedent, and a Justice with strong deference is 10% less likely to vote against precedent—both compared to moderately loyal peers who do not account for precedent at all. By examining both ends and the middle of the spectrum, we can gain a sense of how Justices’ views on the role of precedent might play out.

1. The Weak Deference to Precedent Model (Model 4).—The weak deference to precedent model uses the probabilities of moderately loyal Justices and discounts their likelihood of voting against precedent by 5%.^[96] Thus, if Roe is in effect, the Justices appointed by a Republican president are 5% less likely to vote to overturn it, and Justices appointed by a Democratic president are 5% more likely to uphold it. The reverse is true during periods when Roe has been overruled: Justices appointed by a Democratic president are 5% less likely to vote to reinstate it, and Justices appointed by a Republican president are 5% more likely to vote against reinstating it. As shown in Figure 5, even a weak adherence to the doctrine of stare decisis appears to ameliorate some of the instability caused by term limits. Indeed, between 1973 and the present day, Roe v. Wade would be overturned, but that would be the only precedent switch over the period (compared to three switches if no effect of stare decisis is factored into the probabilities). The weak-deference model, however, does have unstable aspects: Between 2009 and 2015, the Court is very near coin-flip odds of confirming or overturning, always within 10% of an even chance. Furthermore, if a Democrat had won the 2016 presidential election, regardless of the outcome of the Senate race,^[97] precedent would have swung for a second time, and Roe would have been reinstated. Nonetheless, an adherence to stare decisis, even a weak one, does insert some degree of stability into the model.

Odds of Overturning Roe v. Wade with Moderately Loyal Justices with Weak Deference to Precedent

Figure 5. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with moderately loyal Justices exhibiting a 5% deference to precedent. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade. The solid line represents the predicted results with President Trump’s two nominees and the dashed line, predicting a reinstatement of Roe in 2019, represents the predicted results had the Democrats won the 2016 presidential election.

2. The Strong Deference to Precedent Model (Model 5).—The strong deference to precedent model also uses the probabilities of moderately loyal Justices and discounts their likelihood of voting against precedent by 10%.^[98] Similar to the high-loyalty model, an assumption of strong deference to precedent produces higher probabilities of either upholding or reversing, and thus more certainty. As seen in Figure 6, a strong deference to precedent results in ten years with less than a 5% chance of overturning Roe. After that point, however, the odds swing dramatically in the other direction. As the composition of the Court changes with a new Justice every two years, there is a twenty-year period with over a 70% probability of overturning or not reinstating Roe v. Wade. This high probability results because once a precedent switch occurs, even Justices who, based on their loyalty preference, would otherwise vote against the reigning precedent are now 10% less likely to do so. Additionally, those Justices who are already inclined to vote in the same direction as the reigning precedent are even more likely to do so, making a reversal even less likely. The changes over any two-year period, however, are significantly less than those in the high-loyalty (no deference) condition, with a largest change of 31%, as opposed to 48%.^[99] While a strong deference to precedent clearly increases stability, it’s a close call: If the Democrats had taken both the White House and the Senate in 2016 and eliminated the filibuster, Roe would have been reinstated when the Democratic president appointed her second Justice in 2019.

Odds of Overturning Roe v. Wade with Moderately Loyal Justices with Strong Deference to Precedent

Figure 6. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with moderately loyal Justices exhibiting a 10% deference to precedent. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

3. Summary: Stare Decisis and Term Limits.—Deference to precedent, not surprisingly, appears to bring substantial doctrinal stability to a term-limit system. Even a weak deference to precedent reduces the number of predicted switches of Roe from three to one, though the Court is only one unexpected 2016-election result away from a predicted second switch.^[100] In a weak deference to precedent system, however, six of the twenty-three hypothetical cases are within 10% of a coin flip (compared to only one in the strong-deference model), suggesting that weak deference creates an underlying level of instability in a term-limited regime. This could have important implications for whether cases are brought at all and, if brought, how they are argued.

Too strong a deference to precedent, however, can also cement a precedent in place to such an extent that it effectively negates the advantages of a term-limit system. For example, a deference to precedent of only 15% (the strong-deference model uses 10%) under our models would never predict a reversal of Roe v. Wade,^[101] despite thirteen Republican appointees to the Court and only ten for Democrats. In other words, precedent does appear to bring doctrinal stability to a term-limit system, but at the potential cost of the very democratic responsiveness that makes term limits appealing to many proponents.

Furthermore, it is possible that precedent would function differently in a term-limit system in a way that weakens its constraining force. Precedent usually operates on lower courts through willing compliance because fewer than 1% of lower court decisions are heard and overturned by the Supreme Court.^[102] Chad Westerland and his colleagues suggest that the persuasiveness of Supreme Court precedent on lower courts is not constant, varying greatly with how ideologically estranged lower courts view the current Supreme Court from the Supreme Court that created the precedent.^[103] Under a principal–agent conceptualization of Supreme Court precedent, precedent would have limited utility in stabilizing doctrine because a lower court is more willing to defy Supreme Court precedent if it views the current Court as ideologically distinct from the precedent-creating Court, which could happen with more frequency and more quickly in a term-limit system.^[104] Furthermore, several prominent originalists, including Justice Scalia, have argued that precedent for constitutional questions is less important than in other cases, which further cabins precedent’s potential as a counterweight against instability in a term-limit system.^[105]

D. Senate Influence and Stability in a Term-Limit System

As discussed earlier, the Senate’s level of influence in the nomination process is a hotly contested theoretical and empirical issue.^[106] Our models all assume a reasonably strong influence; a president appointing a moderately loyal Justice facing a simple majority of the opposing party in the Senate never has more than a 75% probability that his or her nominee will vote the desired way.^[107] What if the Senate has more or less influence than we posit? The next three sections discuss that possibility.

1. Less Senate Influence (Not Necessary to Model Statistically).—If the Senate has less influence than we posit, all our results will tilt toward greater instability, because the likelihood of a president successfully nominating an ideologically aligned Justice will increase: a president will be able to get more loyal nominees, or nominees closer to his or her preferred policy set point, confirmed.

2. The High Senate Influence Model (Model 6).—Another possibility is that the Senate has more influence than our previous models assumed, or that term limits may increase the influence it can exert on confirmation. To address this possibility, Model 6 posits significantly stronger Senate influence on the president’s nominee than assumed in the earlier models. We assume a moderately loyal Justice with weak deference to precedent (the middle values for those variables) but assume that the Senate exerts 10% more control when it is in opposition to the precedent. Thus, the probabilities are the same as those assigned in Model 4 (moderately loyal Justices with weak deference to precedent), but with a Senate controlled by the opposite party able to exert 10% more influence than in Model 4. To illustrate, a Justice nominated by a Democratic president with a Republican majority in the Senate has a 55% probability of voting to uphold rather than a 65% probability. If Roe is the current precedent at the time of the case this would increase to 60%, and if it were overturned at the time of the case it would fall to 50%. If the Senate is controlled by the president’s party, however, nothing should change from Model 4 because we are assuming that the president and the Senate have similar desires when it comes to upholding, overruling, or reinstating Roe.^[108] Table 4 presents these probabilities graphically and Figure 7 summarizes the results.’

Table 4. This table displays the likelihood of a moderately loyal Justice voting to uphold Roe v. Wade with high Senate influence, depending on the party of the president and the composition of the Senate (“D” = Democrat, “R” = Republican, “DD” = 60 Democrats, “RR” = 60 Republicans).

Odds of Overturning Roe v. Wade with Justices Confirmed with High Senate Influence

Figure 7. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with moderately loyal Justices and confirmed by a Senate with a lot of influence on the nominee. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

As Figure 7 shows, positing increased Senate influence on the confirmation process produces the same number of precedent swings as Model 4, which assumed less Senate influence and the same moderately loyal, weakly deferential Justices. This model, however, does illustrate how increased Senate influence could help temper the Court from swinging too rapidly. For example, while in both the moderate and high Senate influence models the odds of overturning Roe drop from 2007 to 2015 as President Obama’s appointees take the bench, in the high Senate influence model the swing is less dramatic and his nominees are less able to swing the balance of the Court than in the moderate-influence model. This might help prevent the Court from swinging too far in either direction even if one party retains control of the White House for multiple terms. On the other hand, it could also lead to more contentious confirmation hearings and more visible political fights between the president and the Senate, thereby defeating one of the hoped-for benefits of a term-limit system.

3. The No Filibuster Model (Model 7).—Finally, we examine the effects of the elimination of the filibuster. Although filibustering a nominee was an option for the minority party during most of the period under consideration, it has now been eliminated. Comparing what would have happened to Roe had the filibuster been eliminated earlier with its fate with the filibuster in place can both provide insights into the wisdom of eliminating the filibuster and make the analysis more complete.

Model 7 uses the middle values for all three variables (moderate loyalty, weak deference, and moderate Senate influence) but this time posits a Senate lacking a filibuster option for the minority party. Essentially, the filibuster’s elimination is treated as raising the likelihood of confirmation to the same level when the president’s party has a simple majority in the Senate as when it holds a supermajority. Table 5 displays the probabilities, Figure 8 presents the results, and Figure 9 presents a comparison among moderate Senate influence with a filibuster, moderate Senate influence without a filibuster, and strong Senate influence with a filibuster.

Table 5. This table displays the likelihood of a moderately loyal Justice confirmed by a Senate with moderate influence but no filibuster option voting to uphold Roe v. Wade depending on the party of the president and the composition of the Senate (“D” = Democrat, “R” = Republican, “DD” = 60 Democrats, “RR” = 60 Republicans).

Odds of Overturning Roe v. Wade with Justices Confirmed by a Senate without the Filibuster Option

Figure 8. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with moderately loyal Justices with weak deference to precedent and confirmed by a Senate without the filibuster option. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

Odds of Overturning Roe v. Wade with Varying Degrees of Senate Influence

Figure 9. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with moderately loyal Justices and confirmed by a Senate with high influence on the nominee (dotted line) compared to moderately loyal justices confirmed by a Senate with either moderate influence (solid line) or moderate influence and no ability to filibuster (dashed line). The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

As shown in Figures 8 and 9, the elimination of the filibuster may be a more symbolic than practical means of changing the Court and its ideological makeup, as it does not change the number of precedent swings and indeed produces results very close to those of a moderately influential Senate with the filibuster in place. It is also worth noting that, if anything, the previous models overstated the effects of a filibuster since the threat of a filibuster was assumed to be omnipresent, rather than threatened for a specific nominee.

In isolated circumstances, however, eliminating the filibuster can have a large effect on both a term-limit system and the present system, such as in the case of President Trump’s hypothetical second nominee or the real-world nomination of Justice Kavanaugh. With the filibuster eliminated, the Justice would increase the odds of the Court overruling Roe by 9%, compared to a Justice nominated and confirmed by the same president and a Senate with the ability to filibuster. Overall, our models suggest that if less Senate influence is desired, eliminating the filibuster may be an effective way to achieve it, since its elimination moves nominees slightly more towards the nominating president.

The desirability of this outcome may depend on one’s view of democratic accountability, one of the main goals of term-limit proponents. The president is the face of the party responsible for a future Justice, so increasing his or her influence arguably increases democratic accountability by making it clearer to the public that the president bears almost exclusive responsibility for the nominee—at least where the president’s party controls the Senate. However, because the Senate is also an elected body, in many ways the democratic influence is simply being redirected rather than increased.

III. Implications and Limitations

A. Term Limits and the Rise of Doctrinal Instability

Proposals for term limits for Supreme Court Justices have gained renewed traction as a possible way to solve the countermajoritarian difficulty, depoliticize the Court, and reinvigorate the Court’s legitimacy. While not discounting the possible benefits of term limits, this study has asked whether a term-limit regime might also lead to greater legal instability. The study’s results reveal the proposal’s substantial potential to destabilize important constitutional precedents and to change the way that constitutional jurisprudence evolves by pushing it away from gradual shifts and towards more sudden jolts.

But how likely is it that term limits would in fact have such an effect? Having explored a variety of options, we can now consider the most likely scenario. As noted earlier, the trend seems to be toward nominating Justices with a high degree of loyalty and little if any willingness to defer to precedent.^[109] Assuming that we are accurate in assessing the Senate’s influence as moderate—especially given the current absence of a filibuster, which does slightly decrease the Senate’s influence—the model that best captures these three variables (high loyalty, no deference, moderate Senate influence) is Model 2, illustrated by Figure 2 on page 142 and reproduced below for the convenience of the reader.^[110]

Odds of Overturning Roe v. Wade with More Loyal Justices

Figure 2. The odds of the Supreme Court overturning Roe v. Wade between 1973 and 2019 with more loyal Justices. The horizontal line marks the cutoff between the Court being more or less likely to overturn Roe v. Wade.

This model has a substantial detrimental effect on doctrinal stability. A term-limited Court not only changes its collective mind on abortion three times in forty-six years, but also produces extreme swings with a high likelihood of reversal. Such a level of constitutional zigzagging has never been seen in the Court’s history.

Such instability could produce a number of deleterious effects. The lack of doctrinal stability might be replaced with a different type of predictability. Even in the absence of the typical in-between cases signaling change, outcomes may actually become more predictable. However, litigants would turn to reading the tea leaves of the Court’s composition rather than looking to past precedent for guidance. This change also affects the time horizon of predictability. A case can swing from a sure winner to a sure loser over the course of a single election. This could have a drastic effect on litigants who often have to wait years before their cases reach the Supreme Court and may affect whether they decide to bring cases at all.

Perhaps most damaging, it is unclear how frequent swings would affect enforcement of Supreme Court decisions as lower courts react to sudden ideological changes in Court majorities. The lower courts’ reactions might be exacerbated by the fact that instability and doctrinal swings could also open up a Pandora’s box of retroactivity and legitimacy issues. In general, new constitutional rules of criminal procedure “do not apply retroactively to cases on collateral review, but new substantive rules do.”^[111] In Montgomery v. Louisiana,^[112] the Court clarified that substantive constitutional rules include “rules forbidding criminal punishment of certain primary conduct.”^[113] If, for example, either a woman or her doctor were to be criminally convicted under a law banning abortion during an era when Roe v. Wade was not in effect, this conviction should be overturned under Montgomery following Roe’s reinstatement. If the precedent is repeatedly overturned and reinstated, however, lower courts and policy makers could decide to simply ignore the Supreme Court’s decision, counting on a reversal after subsequent elections. In other words, the Supreme Court could lose perhaps its most valuable asset, finality, which could result in increased defiance of its decisions and reduced legitimacy.

B. Limitations and Future Research

While the study finds a distinct danger that legal instability will be a serious side effect of term limits, the study also signals possible ways to limit the dangers of the proposal. The models support the proposition that Justices who display less loyalty to the viewpoints of their nominating presidents bring greater stability to the system than those who exhibit high fidelity to their nominators. Similarly, the models support the idea that the appointment of Justices with a greater fidelity to stare decisis breeds greater stability. These results suggest that the wisdom of Supreme Court term limits depends on the characteristics of the Justices appointed to the Court, pointing to an important direction for future study. Studies hoping to shed light on the practicality of term limits should attempt to measure the weight that Justices give to precedent in their decision-making processes and the degree of loyalty sitting Justices display towards their nominating presidents’ ideologies. These questions are not new, but this study highlights a renewed need to explore them. Future studies should also attempt to discern the interaction between precedent and loyalty. Importantly, proponents of term limits seem to base their proposals on the assumption that Justices are very loyal.^[114] Indeed, this assumption is one of the foundational motivations for wanting each president to have an equal opportunity to nominate Justices.^[115]

It is also important as a qualifying note to observe that the precedential swings predicted, even if the assumptions in the model and outcomes are entirely accurate, may not manifest themselves as outright swings in precedent. Instead they should be viewed as a predicted era of support for, or opposition to, the decision in question. This could manifest in a variety of ways including case selection, a weakening of the precedent through exceptions and qualifiers,^[116] as well as outright reversal. Furthermore, many issues will not align as cleanly with politics as Roe v. Wade does. As then-Senator Obama once stated in voting against the confirmation of Chief Justice Roberts:

[W]hile adherence to legal precedent and rules of statutory or constitutional construction will dispose of 95 percent of the cases that come before a court, so that both a Scalia and a Ginsburg will arrive at the same place most of the time on those 95 percent of the cases—what matters on the Supreme Court is those 5 percent of cases that are truly difficult.^[117]

While it may be true that only politically salient cases would trigger the dramatic swings hypothesized by these models, that does not mean that substantial shifts in the Court’s overall jurisprudence would not occur during these periods on lower profile issues that still have significant ramifications for our constitutional system of government, including issues such as immunity of government actors or the breadth of executive power. Furthermore, the ambit of cases that are politically salient (the “5 percent” of cases referenced by then-Senator Obama) may be increasing. The Republican Party, for example, now espouses a litmus test that nominees must be in favor of limiting the administrative state, a vast expansion from the one-case litmus test of Roe v. Wade.^[118]

Follow-up studies should also address this study’s methodological limitations. We used the Monte Carlo method to estimate the actual probabilities of a decision, but future studies should use a model that actually computes the probability for each possible combination of Supreme Court Justices. Such studies could help to support the findings of this Essay and add to the robustness of its conclusions.

Importantly, the Essay also points to possible mechanisms to limit the dangers of term limits, including appointing Justices who are inclined to exhibit less loyalty to their appointing president’s party and more deference to prior precedent. Of course, a nominating president would likely seek Justices with just the opposite inclinations, so it is unclear how these safeguards could be implemented. Future studies should also seek to assess the extent to which these safeguards are actually utilized by Justices and the interaction between the two factors.

Conclusion

Our findings suggest that the danger of increased instability due to term limits is very real and that policy makers should take that risk into account when considering the proposal. Furthermore, this reduced stability could fundamentally change the nature of jurisprudential evolution and change the focus of litigants, policy makers, and lower court judges from precedent to the Court’s composition. Any proponent of term limits has the burden to show that strategies exist, and can be effectively utilized, to mitigate these dangers and that countervailing benefits can be sufficiently realized. To date, no proponent has carried this burden, but this study helps provide a path to do so.

Appendix

Hypothetical Supreme Court composition for every two years between 1973 and 2019 based on the party of the nominating president and the composition of the Senate. Justices are named after the year in which they were appointed and replaced the longest serving Justice on the panel.

1. .One of us has written extensively on why this view is wrong. See generally Daniel A. Farber & Suzanna Sherry, Judgment Calls: Principle and Politics in Constitutional Law (2009); Suzanna Sherry, Putting the Law Back in Constitutional Law, 25 Const. Comment. 461 (2009); Suzanna Sherry, Politics and Judgment, 70 Mo. L. Rev. 973 (2005). For purposes of this Essay, however, we take the attitudinalist model as valid; judges decide most cases based on their own political preferences moderated by strategic concerns. See, e.g., Michael J. Gerhardt, Attitudes About Attitudes, 101 Mich. L. Rev. 1733, 1739 (2003). ↑
2. .Stan Greenberg et al., Broad Bi-Partisan Consensus Supports Reforms to Supreme Court, Democracy Corps tbl.5 (May 7, 2014), https://‌democracycorps.com‌//‌wp-content‌/uploads‌/2014‌/05‌/DCorps-SCOTUS-Memo-FINAL-050614.pdf [https://‌perma.cc‌/P43T-YJEX]. ↑
3. .Jess Bravin, President Obama’s Supreme Court Nomination of Merrick Garland Expires, Wall St. J. (Jan. 3, 2017, 5:23 PM), https://‌www.wsj.com‌/articles‌/president-obamas-supreme-court-nomination-of-merrick-garland-expires-1483463952 [https://‌perma.cc‌/S6SA-2JVL]. ↑
4. .Erwin Chemerinsky, The Case Against the Supreme Court 310 (2014). ↑
5. .See, e.g., Paul D. Carrington & Roger C. Cramton, The Supreme Court Renewal Act: A Return to Basic Principles, in Reforming the Court 467, 471 (Roger C. Cramton & Paul D. Carrington eds., 2006); Sanford Levinson, Our Undemocratic Constitution 123–39 (2006); Steven G. Calabresi & James Lindgren, Term Limits for the Supreme Court: Life Tenure Reconsidered, 29 Harv. J.L. & Pub. Pol’y 769, 775 (2006); Gerald N. Rosenberg, The Broken-Hearted Lover: Erwin Chemerinsky’s Romantic Longings for a Mythical Court, 69 Vand. L. Rev. 1075, 1109 (2016). ↑
6. .410 U.S. 113 (1973). ↑
7. .The most recent Supreme Court Terms suggest that, at least today, many Justices have a high degree of ideological conformity with their nominating president and a low degree of deference to precedent. As two scholars have recently noted, partisan loyalty has been increasing: “In the past 10 years . . . justices have hardly ever voted against the ideology of the president who appointed them.” Lee Epstein & Eric Posner, If the Supreme Court Is Nakedly Political, Can It Be Just?, N.Y. Times (July 9, 2018), https://‌www.nytimes.com‌/2018‌/07‌/09‌/opinion‌/supreme-court-nominee-trump.html [https://‌perma.cc‌/S5RC-8BXL]; see also Neal Devins & Lawrence Baum, Split Definitive: How Party Polarization Turned the Supreme Court into a Partisan Court, 2016 Sup. Ct. Rev. 301, 301 (2016) (recognizing that the Court’s “clear ideological blocs that coincide[] with party lines” are historically unique). As for deference to precedent, in the 2017 Term alone, the Court overruled precedents from 1967, 1977, and 1992. See South Dakota v. Wayfair, Inc., 138 S. Ct. 2080 (2018) (overturning Quill Corp. v. North Dakota, 504 U.S. 298 (1992) and Nat’l Bellas Hess, Inc. v. Dep’t of Revenue, 386 U.S. 753 (1967)); Janus v. AFSCME, 138 S. Ct. 2448 (2018) (overturning Abood v. Detroit Bd. of Educ., 431 U.S. 209 (1977)). Whether term limits might change either of these tendencies is beyond the scope of this Essay. ↑
8. .See, e.g., Craig Green, An Intellectual History of Judicial Activism, 58 Emory L.J. 1195, 1201–20 (2009) (discussing the origins and uses of the term “judicial activism”). ↑
9. .See generally John Hart Ely, Democracy and Distrust: A Theory of Judicial Review (1980); Jeremy Waldron, The Core of the Case Against Judicial Review, 115 Yale L.J. 1346 (2006). ↑
10. .See Alexander M. Bickel, The Least Dangerous Branch 16, 203 (2d ed. 1962) (coining the term “counter-majoritarian difficulty”). ↑
11. .135 S. Ct. 2584 (2015). ↑
12. .“Our founding fathers did not intend for the judicial branch to legislate from the bench, and as president, I would appoint strict Constitutional conservatives who will apply the law as written”; “The only outcome worse than this flawed, failed decision would be for the President and Congress, two co-equal branches of government, to surrender in the face of this out-of-control act of unconstitutional, judicial tyranny.” Matthew Speiser, Here’s How the 2016 Presidential Candidates Are Reacting to the Supreme Court Ruling on Gay Marriage, Bus. Insider (June 26, 2015, 11:04 AM) (quoting Rick Perry and Mike Huckabee, respectively), http://‌www.businessinsider.com‌/2016-reaction-to-gay-marriage-ruling-2015-6 [https://‌perma.cc‌/RG58-ECGT]. ↑
13. .A Pew Research Center poll found that appointments to the Supreme Court were a top-ten election issue heading into the 2016 presidential election, with 65% of respondents reporting it would be an important factor in their voting decision. This put it above other prominent issues such as abortion and the treatment of racial and ethnic minorities. 2016 Campaign: Strong Interest, Widespread Dissatisfaction: 4. Top Voting Issues in 2016 Election, Pew Res. Ctr. (July 7, 2016), http://‌www.people-press.org‌/2016‌/07‌/07‌/4-top-voting-issues-in-2016-election‌/ [https://‌perma.cc/
    L6UQ-CF8Y]; see also Devins & Baum, supra note , at 302 (discussing the political rift created over President Obama’s attempts to appoint a successor to Justice Antonin Scalia and the importance of the appointment to the 2016 presidential election); Katie Zezima, Cruz Wants to Make 2016 a Referendum on the Supreme Court: He’s Already Done It, Wash. Post (Feb. 15, 2016), https://‌www.washingtonpost.com‌/news‌/post-politics‌/wp‌/2016‌/02‌/15‌/cruz-wants-to-make-2016-a-referendum-on-the-supreme-court-hes-already-done-it‌/?utm‌_termp.793513a5caad [https://
    perma.cc/7K55-9M2V] (discussing Senator Ted Cruz’s comments regarding the political significance of replacing Justice Antonin Scalia on the Supreme Court). ↑
14. .For public reaction to the Kavanaugh confirmation, see, for example, Zack Beauchamp, The Supreme Court’s Legitimacy Crisis Is Here, Vox (Oct. 6, 2018, 4:02 PM), https://‌www.vox.com‌/policy-and-politics‌/2018‌/10‌/6‌/17915854‌/brett-kavanaugh-senate-confirmed-supreme-court-legitimacy [https://‌perma.cc‌/V5FJ-83P7]. ↑
15. .See Barry Friedman, Dialogue and Judicial Review, 91 Mich. L. Rev. 577, 578 (1993) (discussing the degree to which constitutional scholars have fixated on the countermajoritarian difficulty). ↑
16. .Bickel, supra note 10. ↑
17. .Barry Friedman, The History of the Countermajoritarian Difficulty, Part One: The Road to Judicial Supremacy, 73 N.Y.U. L. Rev. 333, 335–36 (1998) (citing Bruce A. Ackerman, The Storrs Lectures: Discovering the Constitution, 93 Yale L.J. 1013, 1046 (1984)); Akhil Reed Amar, The Consent of the Governed: Constitutional Amendment Outside Article V, 94 Colum. L. Rev. 457, 495 (1994); Erwin Chemerinsky, The Supreme Court, 1988 Term—Foreword: The Vanishing Constitution, 103 Harv. L. Rev. 43, 61 (1989)). ↑
18. .See, e.g., Ely, supra note 9. ↑
19. .See, e.g., Robert H. Bork, The Tempting of America: The Political Seduction of the Law 143–60 (1990) (defending originalism on the grounds that it limits the countermajoritarian difficulty); Randy E. Barnett, The Gravitational Force of Originalism, 82 Fordham L. Rev. 411, 429 (2013) (asserting that originalism should supersede Justices’ political preferences). ↑
20. .See generally Cass R. Sunstein, One Case at a Time: Judicial Minimalism on the Supreme Court (1999). ↑
21. .See generally Ronald Dworkin, Freedom’s Law: The Moral Reading of the American Constitution (1996) (arguing that Justices employing a moral reading of the Constitution is consistent with democracy and majority rule since the Bill of Rights commits only to general principles which then must be interpreted and applied to specific circumstances). ↑
22. .See generally 1 Bruce Ackerman, We the People: Foundations (1991) (arguing that the Court should defer to the will of the people only during times of higher lawmaking, when the people look past fleeting concerns and express their long-term desires for the future direction of the nation). ↑
23. .See generally Akhil Reed Amar, Intratextualism, 112 Harv. L. Rev. 747 (1999). ↑
24. .See, e.g., Mark Tushnet, Taking the Constitution Away from the Courts (1999). ↑
25. .See, e.g., Paul D. Carrington & Roger C. Cramton, Introduction to Reforming the Court, supra note 5, at 5–7 (noting the wide variety of scholars in favor of eliminating life tenure of Supreme Court Justices); Levinson, supra note 5, at 123–39 (citing multiple scholars’ arguments against life tenure for Supreme Court Justices); Calabresi & Lindgren, supra note , at 775 (arguing for eighteen-year term limits for Supreme Court Justices); Rosenberg, supra note , at 1105, 1109 (describing the proposed term limit as the only reform proposed by Chemerinsky which “has the potential to make much difference in the decisions of the Court” and as “supported by data, experience, and the findings of the branch relations literature”); Greenberg et al., supra note , at tbl.5 (demonstrating that Americans support term limits for Supreme Court Justices). ↑
26. .Chemerinsky, supra note , at 310. ↑
27. .Id. at 311. ↑
28. .Id. Thomas Jefferson actually first proposed term limits back in 1822 after concluding that life tenure is inconsistent with the American republic. See Calabresi & Lindgren, supra note , at 773 (citing Letter from Thomas Jefferson to William T. Barry (July 2, 1822), in 7 The Writings of Thomas Jefferson 255, 256 (H.A. Washington ed., 1854)). ↑
29. .Scholars Calabresi and Lindgren have suggested enacting the proposal by statute, although most scholars argue that any change would have to be done either through unenforceable norms or a constitutional amendment. See Calabresi & Lindgren, supra note , at 855–56 (proposing statute). ↑
30. .Ryan C. Black & Amanda C. Bryan, The Policy Consequences of Term Limits on the U.S. Supreme Court, 42 Ohio N.U. L. Rev. 821, 835 (2016). ↑
31. .Chemerinsky, supra note , at 310 (discussing the relative youth of recent nominees). ↑
32. .See id. (discussing the increase in life expectancy since the Constitution was written). ↑
33. .Id. ↑
34. .Rosenberg, supra note , at 1109–10. ↑
35. .Chemerinsky, supra note , at 311; see also Paul M. Collins, Jr. & Lori A. Ringhand, The Institutionalization of Supreme Court Confirmation Hearings, 41 L. & Soc. Inquiry 126, 141 (2016) (discussing how changes in the confirmation hearing process have increased democratic checks on the Court). ↑
36. .Chemerinsky, supra note , at 311. ↑
37. .Rosenberg, supra note , at 1110. ↑
38. .Chemerinsky, supra note , at 310; Calabresi & Lindgren, supra note , at 775; Barry Friedman, Letter to Supreme Court (Erwin Chemerinsky Is Mad. Why You Should Care.), 69 Vand. L. Rev. 995, 1007–08 (2016); John O. McGinnis, Justice Without Justices, 16 Const. Comment. 541, 541–43 (1999); Henry Paul Monaghan, The Confirmation Process: Law or Politics?, 101 Harv. L. Rev. 1202, 1211 (1988); Saikrishna B. Prakash, America’s Aristocracy, 109 Yale L.J. 541, 570–73 (1999) (book review); Akhil Reed Amar & Steven G. Calabresi, Term Limits for the High Court, Wash. Post (Aug. 9, 2002), https://‌www.washingtonpost.com‌/archive‌/opinions‌/2002‌/08‌/09‌/term-limits-for-the-high-court‌/646134cd-8e13-4166-9474-5f53be633d7c‌/?noredirect=on [https://‌perma.cc‌/M9NM-KNC9?type=image]; Seema Mehta, Huckabee Calls for Term Limits on U.S. Supreme Court Justices, L.A. Times (Mar. 28, 2015, 3:52 PM), http://‌www.latimes.com‌/nation‌/politics‌/politicsnow‌/la-pn-huckabee-term-limits-supreme-court-20150328-story.html [https://
    perma.cc/3PRY-DC26]; Stephanie Francis Ward, Justice Breyer Says He Could Support Certain SCOTUS Term Limits, A.B.A. J. (Jan. 8, 2016, 1:15 PM), http://‌www.abajournal.com‌/news‌/article‌/justice‌_breyer‌_says‌_he‌_could‌_support‌_certain‌_scotus‌_term‌_limits [https://‌perma.cc‌/UR7C-FGRT]. ↑
39. .See Chemerinsky, supra note , at 306–10. ↑
40. .Black & Bryan, supra note , at 843 (presenting data on how term limits might have altered the Court’s composition and its legal outputs for the last half century). Importantly, their work focuses on how doctrine would actually be different, whereas our focus is on how doctrine might evolve differently. For example, their analysis asks whether Roe v. Wade would have been decided differently in 1973 in a term-limit system but fails to consider its doctrinal stability over time. ↑
41. .Martin–Quinn scores place Justices and Courts on an ideological continuum from conservative to liberal based on each Justice’s ideal policy set point. See generally Andrew D. Martin & Kevin M. Quinn, Dynamic Ideal Point Estimation Via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953–1999, 10 Pol. Analysis 134 (2002). ↑
42. .Black & Bryan, supra note 30, at 843. ↑
43. .See also Calabresi & Lindgren, supra note , at 844 (discussing how the Court could shift too rapidly in a term-limit system); Ward Farnsworth, The Regulation of Turnover on the Supreme Court, 2005 U. Ill. L. Rev. 407, 436–38 (2005) (discussing Supreme Court capture and how one party winning four elections in a row could swing the court from unanimously liberal to unanimously conservative). ↑
44. .David R. Stras, Understanding the New Politics of Judicial Appointments, 86 Texas L. Rev. 1033, 1037 (2008) (book review) (quoting Jan Crawford Greenburg, Supreme Conflict 221 (2007)). For a more general discussion of the shift from judicial competence and qualifications to political ideology as the core issue in Supreme Court confirmation battles, see generally Richard L. Hasen, End of the Dialogue? Political Polarization, the Supreme Court, and Congress, 86 S. Cal. L. Rev. 205 (2013). ↑
45. .Stras, supra note , at 1037. ↑
46. .See Appendix for a complete list of the Supreme Court panels by year. ↑
47. .See Susan R. Poulter, Monte Carlo Simulation in Environmental Risk Assessment—Science, Policy and Legal Issues, 9 RISK: Health, Safety & Env’t 7, 7–9 (1998) (describing the use of the Monte Carlo method in environmental risk assessment). ↑
48. .All statistics were run in MATLAB. All scripts run are on file with author Christopher Sundby. Note that this scenario actually presents a binomial distribution with independent probabilities. The results of the Monte Carlo model were confirmed using a Poisson Equation, another method of estimating binomial distributions with varying probabilities, which returned nearly identical results. Convergence was also checked by running several of the simulations with 100,000 trials, and nearly identical results were obtained, suggesting that 10,000 trials is sufficient to estimate the actual probability. ↑
49. .See, e.g., Jeffrey A. Segal & Albert D. Cover, Ideological Values and the Votes of U.S. Supreme Court Justices, 83 Am. Pol. Sci. Rev. 557, 558 (1989) (discussing efforts made to correlate Justices’ policy preferences and values with their decisions on the Court). ↑
50. .See, e.g., Gerhardt, supra note 1, at 1739–44 (discussing the limitations of the attitudinal model). ↑
51. .See, e.g., Wayne Batchis, Constitutional Nihilism: Political Science and the Deconstruction of the Judiciary, 6 Rutgers J.L. & Pub. Pol’y 1, 11 (2008) (“Yet, even the most accomplished of jurists openly acknowledge that judicial decisions, while primarily rooted in traditional legal analysis, must involve some consideration of external factors.”). ↑
52. .Chemerinsky, supra note , at 309 (“Ideology often matters enormously in how the Court decides cases, so let’s have a confirmation process that recognizes this.”). ↑
53. .Id. at 308–09. ↑
54. .Id. at 309. ↑
55. .See, e.g., Segal & Cover, supra note , at 560 tbl.1 (rating the Supreme Court Justices from Earl Warren to Anthony Kennedy on a scale from −1 (extremely conservative) to +1 (extremely liberal) and measuring the percent of “liberal” votes in civil liberty cases). ↑
56. .Id. at 599–60 tbl.1. ↑
57. .See, e.g., Todd S. Purdum, Presidents, Picking Justices, Can Have Backfires, N.Y. Times (July 5, 2005), https://‌www.nytimes.com‌/2005‌/07‌/05‌/politics‌/politicsspecial1‌/presidents-picking-justices-can-have-backfires.html [https://‌perma.cc‌/4RJC-USL7] (noting several presidents’ discontent with their appointments, including President Roosevelt stating that he could “carve out of a banana a judge with more backbone than that” after Justice Holmes ruled against him in an antitrust case and President Eisenhower describing his appointment of Chief Justice Warren as “[t]he biggest damn fool mistake I ever made”). ↑
58. .Devins & Baum, supra note , at 301; see also Epstein & Posner, supra note (discussing the growing consistency between how Justices vote and the political ideology of the president who appointed them). ↑
59. .Interestingly, however, the conservative–liberal divide has not similarly grown despite its increased alignment with partisan political divisions. See Tom S. Clark, Measuring Ideological Polarization on the United States Supreme Court, 62 Pol. Res. Q. 146, 146, 150 (2009) (discussing the trend of ideological polarization over time as it relates to the Court); Devins & Baum, supra note , at 319 (finding a reduction in the standard deviation in the number of conservative votes among Justices appointed by presidents from each party, a measure of conformity); Donald Michael Gooch, Ideological Polarization on the Supreme Court: Trends in the Court’s Institutional Environment and Across Regimes, 1937-2008, 43 Am. Pol. Res. 999, 1002–06 (2015) (discussing Court polarization and how it relates to congressional and presidential polarization). ↑
60. .Devins & Baum, supra note , at 303–04. ↑
61. .Laurence Baum & Neal Devins, How the Federalist Society Became the De Facto Selector of Republican Supreme Court Justices, Slate (Jan. 31, 2017, 10:12 AM), http://www.slate.com/
    articles/news_and_politics/jurisprudence/2017/01/how_the_federalist_society_became_the_de
    _facto_selector_of_republican_supreme.html [https://perma.cc/6S42-RNVK]. ↑
62. .See, e.g., Amanda Hollis-Brusky, Ideas with Consequences: The Federalist Society and the Conservative Counterrevolution 21–22 (2015) (discussing the Federalist Society’s institutional goal of implementing conservative and libertarian principles in legal decisions). ↑
63. .Devins & Baum, supra note , at 309, 313–17 (analyzing the relationship between Martin–Quinn scores, the party of the nominating president, and dissent and voting behavior in landmark decisions). ↑
64. .Id. at 301. ↑
65. .Id. at 316. ↑
66. .Id. at 316–17. There are three obvious hypotheses for the increased alignment of a Justice’s voting with the party of the president that nominated him or her: (1) Nominators have only recently begun trying harder to nominate judges whose decisions align with their political beliefs, (2) nominators have gotten significantly better at selecting judges who actually will vote in line with the nominator’s political preferences, or (3) judges have become more predictable, due to partisan considerations or otherwise, in their voting behavior. ↑
67. .One study, for instance, found a reduction in the standard deviation, a measure of variability, across time in the number of conservative votes when the Justices are grouped by the party of their nominating president. This is due, at least in part, to changes in the nomination behavior of presidents. Segal & Cover, supra note , at 561. ↑
68. .Jeffrey A. Segal et al., Buyer Beware? Presidential Success Through Supreme Court Appointments, 53 Pol. Res. Q. 557, 563–64 (2000). ↑
69. .Id. at 564 tbl.2. ↑
70. .Id. This number is calculated by comparing the most and least liberal president in each domain, taking the difference divided by 10 and then multiplied by 4.22 and 3.41 respectively. ↑
71. .Charles M. Cameron & Jonathan P. Kastellec, Are Supreme Court Nominations a Move-the-Median Game?, 110 Am. Pol. Sci. Rev. 778, 778 (2016). ↑
72. .Id. at 778. ↑
73. .Id. ↑
74. .David Cottrell et al., The Power to Appoint: Presidential Nominations and Change on the Supreme Court, 81 J. Pol. 1057, 1057–58 (2019). ↑
75. .Paul Kane, Reid, Democrats Trigger ‘Nuclear’ Option; Eliminate Most Filibusters on Nominees, Wash. Post (Nov. 21, 2013), https://‌www.washingtonpost.com‌/politics‌/senate-poised-to-limit-filibusters-in-party-line-vote-that-would-alter-centuries-of-precedent‌/2013‌/11‌/21‌/d065cfe8-52b6-11e3-9fe0-fd2ca728e67c‌_story.html [https://perma.cc/T6TZ-SEN4]. ↑
76. .Ashley Killough & Ted Barrett, Senate GOP Triggers Nuclear Option to Break Democratic Filibuster on Gorsuch, CNN (Apr. 7, 2017, 9:26 AM), https://‌www.cnn.com‌/2017‌/04‌/06‌/politics‌/senate-nuclear-option-neil-gorsuch‌/index.html [https://perma.cc/9MCL-6JKM]. ↑
77. .National Journal Vote Ratings, Ballotpedia, https://‌ballotpedia.org‌/National‌_Journal‌_vote‌_ratings [https://‌perma.cc‌/8J5S-F9ZR] (last updated July 2014) (only counting congressional members for which there was a sufficient voting record to calculate a score); Josh Kraushaar, The Most Polarized Congress Ever: 2013 Congressional Vote Ratings, Atlantic (Feb. 6, 2014), https://‌www.theatlantic.com‌/politics‌/archive‌/2014‌/02‌/the-most-polarized-congress-ever-2013-congressional-vote-ratings‌/283635‌/ [https://‌perma.cc‌/KJL3-M7EL]. ↑
78. .The Polarization of the Congressional Parties, Voteview (Mar. 21, 2015), https://‌legacy.voteview.com‌/political‌_polarization‌_2014.htm [https://‌perma.cc‌/7LH4-NAP4]; Drew Desilver, The Polarized Congress of Today Has Its Roots in the 1970s, Pew Res. Ctr.: Fact Tank (June 12, 2014), https://‌www.pewresearch.org‌/fact-tank‌/2014‌/06‌/12‌/polarized-politics-in-congress-began-in-the-1970s-and-has-been-getting-worse-ever-since‌/ [https://perma.cc/A6MS-MVU6]. ↑
79. .See The Polarization of the Congressional Parties, supra note ‌78. This effect appears to be driven primarily, though not entirely, by a stark move to the right by the most liberal 10% of the Republican party starting in 1981. ↑
80. .Id. Moderates now make up less than 5% of Congress, a drastic drop from 40% in 1980. ↑
81. .A.E. Dick Howard, The Changing Face of the Supreme Court, 101 Va. L. Rev. 231, 315 (2015). ↑
82. .Id. at 296–316. ↑
83. .Brent Martin, Chief Justice Roberts: Scalia, Ginsburg Wouldn’t Be Confirmed Today, Neb. Radio Network (Sept. 19, 2014) (quoting Chief Justice Roberts), http://‌nebraskaradionetwork.com‌/2014‌/09‌/19‌/chief-justice-roberts-scalia-ginsburg-wouldnt-be-confirmed-today-audio [https://‌perma.cc‌/RQ6Z-MRT4]. ↑
84. .A “supermajority” is defined as a filibuster-proof majority, or at least sixty senators. This distinction may be less important in the future, as the Senate has abandoned the filibuster for Supreme Court nominees (and Court of Appeals nominees). However, a supermajority might still strengthen a president’s bargaining position and facilitate confirmation of a more loyal candidate. It is also possible, of course, that the Senate might decide to reinstate the filibuster at some point in the future. The effects of eliminating the filibuster are explored in Model 7. ↑
85. .While the vice president breaks such a tie (making this group functionally the same as
    R–R), this group was included because it still suggests a weaker presidential bargaining position, which may slightly temper whom the president will nominate. Also note that all models base chamber composition on the day the Senate gavels into session, on the assumption that the president will make his nomination soon thereafter (this is important for instances where senators switched parties or had to be replaced over the course of a session). No calculations were done for a Democratic president and an evenly divided Senate because that configuration did not occur between 1975 and 2019. ↑
86. .Some, however, may not find this degree of instability unsettling. Proponents could decide that the loss in stability and predictability in constitutional precedent is outweighed by the benefits of a more majoritarian and responsive Supreme Court. This Essay does not attempt to weigh in on this normative debate, but merely provides a more complete understanding of the potential costs. ↑
87. .In the high-loyalty condition, the possibility of overruling existing precedent comes within 10% of a coin flip only once over the course of forty-six years, versus eleven times in the moderately loyal condition. ↑
88. .See, e.g., Sunstein, supra note , at 63–69 (discussing the merits of minimalism and how jurisprudence evolves). ↑
89. .A recent empirical study has found that lawyer quality can make a difference in Supreme Court decisions. Michael J. Nelson & Lee Epstein, Lawyers with More Experience Obtain Better Outcomes 7 (May 14, 2019) (working paper), http://‌epstein.wustl.edu‌/research‌/LawyerExp.pdf [https://‌perma.cc‌/YJF8-YKCM]. ↑
90. .While only three “loyalty” values are presented in the main text, simulations were run assuming values ranging from 1%–100% loyalty and are available from author Christopher Sundby. Results are very similar at levels down to 65% (three precedent changes that have probabilities of 55% or above). Below that level, the probabilities resemble a coin flip, with changes predicted very frequently but at a low level (less than 55%) of probability. ↑
91. .135 S. Ct. 2401 (2015). ↑
92. .Id. at 2409 (quoting Payne v. Tennessee, 501 U.S. 808, 827–28 (1991)). ↑
93. .See Michael A. Bailey & Forrest Maltzman, Does Legal Doctrine Matter? Unpacking Law and Policy Preferences on the U.S. Supreme Court, 102 Am. Pol. Sci. Rev. 369 (2008) (attempting to disentangle the influence of law and ideological preferences in Supreme Court decisions). ↑
94. .Id. at 379 fig.5 (finding a high degree of variation in the weight given to precedent between Justices). ↑
95. .While only three simulations are presented, simulations with deference from 0%–100% for each loyalty value from 0%–100% were run and are available from author Christopher Sundby. Once the deference level reaches 15% there are almost no reversals at any level of loyalty. We discuss this finding infra at text accompanying note 109. ↑
96. .Any decision between 45% and 55% was not counted as sufficiently likely to swing the weight of precedent. ↑
97. .Assuming neither party had won a supermajority in the Senate, which was not simulated between 2015 and 2019 in the model. ↑
98. .Once again, any decision between 45% and 55% was not counted as sufficiently likely to swing the weight of precedent. ↑
99. .Compare Figure 2 with Figure 6. ↑
100. .As noted earlier, to trigger a second precedent swing in the strong deference to precedent model, it would have taken the Democrats winning both the White House and a majority in the Senate, plus the elimination of the filibuster (or a Democratic supermajority in support of a Democratic president). ↑
101. .Model on file with author Christopher Sundby. ↑
102. .Chad Westerland et al., Strategic Defiance and Compliance in the U.S. Courts of Appeals, 54 Am. J. Pol. Sci. 891, 892 (2010). ↑
103. .Id. ↑
104. .Id. ↑
105. .See South Carolina v. Gathers, 490 U.S. 805, 825 (1989) (Scalia, J., dissenting) (“A judge looking at a constitutional decision may have compulsions to revere past history and accept what was once written. But he remembers above all else that it is the Constitution which he swore to support and defend, not the gloss which his predecessors may have put on it.” (quoting William O. Douglas, Stare Decisis, 49 Colum. L. Rev. 735, 736 (1949))), overruled by Payne v. Tennessee, 501 U.S. 1277 (1991); Jeffrey A. Segal & Harold J. Spaeth, The Supreme Court and the Attitudinal Model Revisited 76–77 (2002). ↑
106. .See supra section II(A)(2) for a discussion of the research into the role of the Senate in the confirmation process. ↑
107. .See supra Table 2. ↑
108. .The only time this might vary is if the filibuster rule interacts with a strong Senate influence to create a roadblock to the president getting his or her ideal nominee confirmed. As we discussed in note , supra, this is not currently an issue. However, to account for the possibility, we will decrease by 10% the probability of a nominee in that circumstance voting in the president’s direction (in other words, treating it just like a president facing a Senate controlled by the opposing party). ↑
109. .One recent study supports our assessment of these variables. The study suggests that presidential interest in the Court is increasing, the cost of finding ideologically reliable candidates is decreasing, and that the composition of the Court is therefore swinging toward Justices whose decisions reliably align with the politics of the nominating president. See Charles M. Cameron et al., Presidential Selection of Supreme Court Nominees: The Characteristics Approach, Q.J. Pol. Sci. (forthcoming) (manuscript at 3–4), https://‌scholar.princeton.edu‌/sites‌/default‌/files‌/jkastellec‌/files‌/cameron‌_kastellec‌_mattioli‌_characteristics‌_for‌_web‌_3.pdf [https://‌perma.cc‌/A8F5-54P2]. ↑
110. .There are other combinations of loyalty and deference that also produce three swings. As long as the level of deference is below 5%, the Court is likely to change its mind three times. ↑
111. .Welch v. United States, 136 S. Ct. 1257, 1259–60 (2016) (citing Teague v. Lane, 489 U.S. 288, 310 (1989); Schriro v. Summerlin, 542 U.S. 348, 351 (2004)). ↑
112. .136 S. Ct. 718 (2016). ↑
113. .Id. at 728 (quoting Penry v. Lynaugh, 492 U.S. 302, 330 (1989)). ↑
114. .Chemerinsky, supra note , at 309 (“Ideology often matters enormously in how the Court decides cases, so let’s have a confirmation process that recognizes this.”). ↑
115. .See, e.g., id. at 311 (“Having a vacancy every two years would give all presidents the chance to equally influence the Court.”). ↑
116. .There are numerous examples of this, including but not limited to the various exceptions to the exclusionary rule (see, e.g., Jeffrey L. Fisher, Preface: Reclaiming Criminal Procedure, 38 Geo. L.J. Ann. Rev. Crim. Proc. iii, xv (2009); Wayne R. LaFave, The Smell of Herring: A Critique of the Supreme Court’s Latest Assault on the Exclusionary Rule, 99 J. Crim. L. & Criminology 757, 758 (2009)) and restrictions on abortion rights (see, e.g., Mary Ziegler, Substantial Uncertainty: Whole Woman’s Health v Hellerstedt and the Future of Abortion Law, 2016 Sup. Ct. Rev. 77, 82 (discussing the pro-life incrementalism strategy)). ↑
117. .151 Cong. Rec. 21032 (2005) (statement of Sen. Obama). ↑
118. .See Interview with Donald McGahn, then-White House Counsel, at Conservative Political Action Conference on Judicial Selection (Feb. 22, 2018). McGahn states that Gorsuch’s views on needing to rein in the administrative state are part of a broader plan, that they discuss the views of political nominees on the administrative state, and that their selection is different than past years when it was a single-issue litmus test. For commentary on these remarks and their significance, see Adam Liptak, Trump’s New Judicial Litmus Test: Shrinking ‘the Administrative State,’ N.Y. Times (Mar. 26, 2018), https://‌www.nytimes.com‌/2018‌/03‌/26‌/us‌/politics‌/trump-judges-courts-administrative-state.html [https://perma.cc/BG2C-MHD7]. ↑