Election Day had barely drawn to a close when then-President Trump began his protests that the election had been “stolen.”1 He claimed, among other things, that voting software had been compromised; that ballot boxes had been stuffed; and that voter fraud had been rampant.2 By January 6, 2021, Trump and his allies had filed at least sixty-two lawsuits aiming to overturn election results in swing states that Trump had narrowly lost.3 No judge agreed with the claim that the election had been “rigged” or plagued by widespread fraud.4 Nevertheless, Trump’s assertions continued to spread through social media and conservative news outlets.5 When some Trump Administration officials publicly disavowed the allegations, they were fired or quickly resigned.6
Trump’s “big lie” continued to find significant support among Republican voters. One March 2021 poll, for example, found that fifty-five percent of Republicans believed that Trump’s electoral defeat “resulted from illegal voting.”7 That same month, another poll claimed that sixty-six percent of Republicans either completely or mostly agreed that “the 2020 election was stolen from Donald Trump.”8 Whether these respondents actually believed their answers (versus just expressed their partisan loyalty) is an open question. But to those that did believe the claim, the Biden Administration represented an illegitimate power grab.9
These recent events underscore the relationship between perceived election integrity and the felt legitimacy of the incoming administration.10 To be sure, it cannot be the case that a losing candidate’s baseless claims threaten the objective legitimacy of the winner. The observation for now is simply the positive correlation between valid elections that are recognized as such and acceptance of a transition in power — precisely why Trump’s meritless claims pose such a democratic threat.11 Indeed, this nexus was perhaps felt most keenly after the Supreme Court’s decision in Bush v. Gore,12 which essentially decided the 2000 presidential election.13 Following that election, there was a substantial partisan divide in the perceived validity of Bush Administration decisions.14 Electoral legitimacy, however, has both subjective and objective dimensions. Subjective legitimacy arises from the perceptions of those affected by an election, whereas objective legitimacy stems from the “actual properties” of the election.15 One objective account, for example, looks at the extent to which elections are inclusive, facilitate policy-directed voting, and allow effective aggregation.16 Some of these criteria overlap with other objective conceptions that emphasize compliance with “international conventions and universal standards about elections reflecting global norms.”17 While objective and subjective legitimacy can arise from different sources, they are interrelated: if elections are objectively marred, for example, voters may refuse to consent to them going forward.18 In this manner, valid elections require both subjective and objective legitimacy.
Both of these components of electoral legitimacy, in turn, are critical premises of Professor Cristina Rodríguez’s bold Foreword. There, Rodríguez argues that democratic elections justify rapid and wholesale regime change. In her words, democracy features the “reciprocal values of accepting losses” — a subjective state of mind — and “seizing wins in the political process.”19 As such, in Rodríguez’s view, the law should not treat the policy whiplash that can come with new presidential administrations as something that demands judicial skepticism.20 Rather, courts should enable, even facilitate, the energy and dynamism that a new government promises.21
Importantly, Rodríguez’s account is grounded in a distinctive view of democracy: one that “mak[es] the government work for the people.”22 Put differently, Rodríguez ultimately wants responsive government, one that can actually accomplish its promised outcomes, not one hamstrung by judges that are political actors in their own right.23 Her view does not hinge on the well-worn idea that the President represents the national polity and is therefore more politically accountable.24 To the contrary, Rodríguez wants to “decenter” the President and focus our attention instead on the regime as a whole25 — the slew of “people and ideas” that come into government to effect change.26
As a result, Rodríguez discounts the traditional countervailing concerns about stability and reliance interests in favor of her bold vision for a government that actually delivers. After all, she argues, “pluralism and disagreement demand acceptance that preferences other than one’s own may prevail.”27 This condition only holds when the means through which disagreements are settled — elections — are valid and recognized as such. Otherwise, losers do not, in fact, accept that the other side has prevailed.
To her credit, Rodríguez recognizes the threats that pitched battles over voting rights and immigration pose to her theory.28 Indeed, she writes that her argument “presumes reasonably fair terms of competition”29 — a more objective conception of electoral legitimacy. But her justified criticisms narrowly target the role of a politicized Supreme Court in election-related matters.30 She objects, for instance, to the Court’s efforts to “disconnect[] the Voting Rights Act from its history and narrow[] its reach,” thus “making it less likely that the law will be of use in combatting far more serious threats to democracy now emerging from state legislatures.”31
But the threat to electoral legitimacy also arises from another source, an inescapable fact about federal elections: the President plays an integral role in their administration.32 States are primarily responsible for running elections, but the Constitution grants the federal government the authority to preempt certain state regulations with respect to federal elections, as well as to combat voter discrimination.33 These statutes sometimes grant the President direct authority, but more commonly the President exercises his traditional administrative oversight over their execution.34 That is, he supervises and directs the interpretation and implementation of election-related statutes. Unsurprisingly, Presidents have historically done so in ways that benefit their parties.35 As a result, Presidents themselves also pose a danger to democratic ideals.
This Response argues that Rodríguez’s call for full-throated regime change has limits in the realm of election administration. If an incoming administration could freely implement election-related policies to entrench itself in power, doing so would undermine the democratic legitimacy of regime change itself. Indeed, when politicians attempt to keep themselves in power, public confidence in elections suffers.36 Common methods of entrenchment are almost always deeply unpopular, as with partisan gerrymandering,37 or divisive, as with restrictions on the electorate.38 As such, courts and other actors should not blindly defer to a new regime’s attempts to influence future elections.
At the same time, Rodríguez convincingly endorses the attitudinal view of judges as political actors themselves.39 Accordingly, there is also an important role for presidential self-restraint. Part I below provides some background on the historical role that Presidents have played in federal election administration. Part II explores how to think about the appropriate baseline for legitimate federal elections given the risks of partisan entrenchment. Given congressional paralysis on voting rights, Part III then considers how courts and, more importantly, the President should respond to these temptations — namely, by engaging in a kind of electoral forbearance. Forbearance in election administration, in other words, would help justify the kind of robust regime change that Rodríguez seeks.
I. Elections and the Executive Branch
By virtue of his position as head of the executive branch, the President plays an important role in federal election administration. Because Congress has passed a number of statutes that necessarily require discretion to execute, the President makes administrative decisions that can influence electoral results. Whether these choices are outcome-determinative is difficult to assess empirically,40 but at the very least, the perception and potential that they are is clear.41 Moreover, the President’s authority in this arena is only likely to increase in coming years, given the erosion of norms against politicization.42 The general polarization of the country, in turn, only heightens the stakes involved.43
Federal elections, as mentioned, are administered by a number of federal agencies in collaboration with state and local governments, which take primary responsibility on the ground for the “time, place, and manner” of the elections.44 Congress, for its part, has exercised its constitutional power to “make or alter” these regulations selectively — perhaps most notably in the areas of campaign finance,45 antidiscrimination,46 ballot provision,47 vote-counting technology, and voter registration by both domestic48 and overseas voters.49
Congress, in turn, has delegated many of these election-related responsibilities to various agencies, both executive and independent in structure.50 The Department of Justice (DOJ), for example, has played a key role in enforcing various sections of the Voting Rights Act (VRA).51 The Census Bureau within the Department of Commerce runs the decennial census,52 which impacts congressional apportionment and state redistricting efforts. Independent agencies like the Federal Election Commission (FEC) and Election Assistance Commission (EAC) administer the Federal Election Campaign Act and Help America Vote Act, respectively.53
The President has the most direct control over single-headed executive agencies, which are generally led by his Senate-confirmed appointees who are removable at will.54 As a matter of norms and design, these agencies are most easily influenced by the White House. By contrast, the FEC and EAC are even-numbered commissions that are often mired in deadlock.55 They are more difficult for the President to influence, though he can help to facilitate inaction.56 The broader observation is that different “regimes,” in Rodríguez’s sense of the term, have various mechanisms — some more effective than others — to influence election administration on the ground. These regimes include the President at the top, but also his appointees and career staff below.
To illustrate the role of the executive branch, consider its historical approach to section 2 of the VRA. The provision bans electoral structures that result in certain minority groups “hav[ing] less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.”57 In this sense, section 2 is a nationally applicable prohibition against voting practices and procedures — such as redistricting — that discriminate on the basis of race, color, or language minority status. It prohibits not only election-related practices that are intended to be racially discriminatory, but also those that are shown to have a racially discriminatory impact.58 For these reasons, there is likely a “causal relationship” between “the practices governed by section 2” and “electoral outcomes.”59
The VRA allows the Attorney General, as well as private citizens, to bring suit to obtain court-ordered remedies.60 Across different administrations, the DOJ has strategically engaged in enforcement as well as changed its interpretation of section 2 through litigation.61 There is debate about whether the agency has done so for partisan gain,62 but the point here is to illustrate some of the levers that the executive branch possesses in election administration. For example, in an amicus brief, the Reagan Administration interpreted an amendment to section 2 to preclude a violation if a racial minority achieved proportional representation in a single election.63
President George H.W. Bush’s Administration was particularly active in interpreting section 2 through litigation. As both a litigant and amicus curiae, the Bush Administration argued that section 2 is broad enough to cover judicial elections.64 The Clinton Administration, in turn, argued that (1) the use of a single election commissioner instead of a multimember commission may be challenged under section 2;65 (2) proportional representation should be assessed on a statewide basis for section 2 vote dilution claims;66 and (3) section 2 required “the creation of a majority-minority district in east-central Georgia.”67 Both President George W. Bush and President Trump also advanced their interpretations of section 2 through litigation. The second Bush Administration used an amicus brief to argue that section 2 did not require the adoption of certain districts in Texas that would allegedly provide additional electoral control to racial minorities.68 The Trump Administration argued in favor of a strong presumption of legislative good faith in section 2 intentional vote dilution claims, especially when a legislature adopts a court-ordered interim redistricting plan.69
Different administration priorities are also reflected in different enforcement decisions.70 The Bush Administration, for instance, pursued the “first-ever § 2 case against an African-American defendant for discriminating against white voters.”71 By contrast, President Obama’s DOJ instead instructed Voting Section attorneys to “focus on ‘traditional civil rights’ cases and . . . political equality for racial and ethnic minorities.”72 More recently, the Trump Administration brought only one new section 2 case.73 In this manner, the executive branch can influence election administration through litigation and enforcement, as well as other traditional tools such as appointments decisions and guidance documents.74
II. Toward a Procedural Baseline
Critics of executive branch efforts to influence federal elections often deride these efforts as “politicized.”75 By this adjective, commentators generally mean that agency decisions are made by political appointees in a nontechnical manner to achieve some partisan outcome.76 Legal judgments, however, rarely admit of one objectively correct or manifestly clear answer.77 Indeed, in Rodríguez’s estimation, they are often little more than thinly veiled attempts at political maneuvering.78 As a result, most executive branch interpretations are likely to have some partisan valence. After all, government lawyers are socialized to serve the needs of their client, which, in this case, includes the administration in power.79
The question, again, for Rodríguez’s theory of regime change is how to draw the line between permissible, perhaps even inevitable,80 partisan interpretation and impermissible self-entrenchment in federal election administration.81 On the one hand, her theory posits that new regimes should be allowed to engage in full-throated policy change. On the other hand, it also requires some notion of legitimacy to justify the new government’s dynamism in the first place. Elections are subjectively and objectively illegitimate, however, when they reflect not democratic preferences but rather successful efforts to entrench power.82 Under these circumstances, the critical link to democratic responsiveness — much celebrated by Rodríguez — is missing.
But entrenchment must be assessed against some baseline and these “normative baselines . . . are themselves hotly disputed.”83 Nevertheless, it is important to consider the criterion against which to evaluate an administrative decision. Existing scholarship has proposed baselines of a hypothetical majoritarian mechanism84 or, alternatively, the costs required to achieve the status quo in the first place.85 One natural way to think about the former is in terms of elections that yield the majoritarian winner of a national popular vote. Doing so aligns with the intuitive notion that the candidate or party receiving the most votes should win.86
Presidential elections, however, are not majoritarian in large part because of the Electoral College; some have even called these elections “antimajoritarian.”87 Indeed, candidates can win the presidency despite losing the national popular vote.88 Constitutional designers intended this outcome to reflect federalism values, among others.89 In this manner, presidential elections attempt to combine visions of the relevant electorate at both the national and individual state levels.90 In fact, none of the elected constitutional actors — including Senators and Representatives — likely channel majoritarian preferences either.91 The majoritarian criterion therefore does not align with the constitutional structure of American federal elections by design.92
Given our current institutions, the question then remains how to think about the appropriate baseline for evaluating election administration decisions that inevitably have some partisan effect. Consider now the alternative baseline of the costs that were necessary to attain the status quo. In this view, entrenchment occurs when “a political arrangement is now more difficult to change that it was to create in the first place.”93 If the new burdens on voting, for example, make it harder for the pre-existing electorate to change the new election rule, then this would constitute entrenchment.94
A major challenge with this baseline, however, is the difficulty of measuring the relevant costs — both the baseline costs as well as those imposed by the new election rule. Among other things, one would need to isolate the burdens narrowly caused by the administrative decision, while controlling for a number of variables that are likely to be shifting at the same time. Moreover, this definition also does not answer the question of how high the new burdens must be to be deemed “entrenchment”: Is any new cost sufficient or must it result in outright vote denial? Finally, this approach also ignores the many ways in which election administration decisions can functionally entrench interests as well.95
Given these conceptual and empirical difficulties, another possibility is to turn from a more substantive criterion to a more procedural one: an approach that focuses instead on the process through which an agency decision is made. In this view, there is a baseline procedure that should be followed when election administration decisions are made, departures from which detract from the decision’s legitimacy. This procedure could be a more formal version of what I have elsewhere referred to as a “reciprocal hierarchy” within agencies — the notion that administrative decisions should be the result of a structured deliberation between civil servants and political appointees.96 This internal procedural formality is arguably more important in the election context than in ordinary policymaking.
More specifically, an ideal bureaucratic process contains both means of internal control to facilitate accountability, as well as bottom-up input from expert civil servants with institutional memory and experience.97 This concept arguably underlies competing accounts of broader administrative legitimacy. The civic republican emphasis on deliberation, for instance, depends heavily on a back-and-forth between career staff and appointees to vindicate its vision.98 The pluralist account similarly requires a pathway from civil servants and interest groups to political decisionmakers within the agency.99 Finally, expert-driven justifications also require channels for information and data to help inform final decisionmaking by agency heads.100 To be clear, the claim here is not that appointees must adopt the views of civil servants — they can (and often should) reject them altogether; rather, it is simply that consideration of their views is an important part of an ideal administrative process.
This internal deliberative procedure can take various forms. For example, many agencies currently use what are known as “clearance procedures” before a decision is finalized.101 These require different offices within an agency to review documents in a specified order before the agency head signs off.102 In this manner, agency heads have a robust means of aggregating information and diverse views. Some agencies formalize these procedures even further, for example, by requiring written recommendations from civil servants, rather than more informal, verbal ones.103 These more formal procedures arguably discipline underlying rationales as well as make it more difficult for political appointees to reverse the decisions without some explanation.104 Moreover, if an appointee is deceptive about the basis for a decision or offers a rationale clearly contrary to law, there will now be a written record more susceptible to leaks or available to serve as evidence in a whistleblower action or litigation.105
Take, for instance, the DOJ, which had such a system in place for enforcing the VRA. As background, the Civil Rights Division within the agency has a dedicated Voting Section traditionally containing about thirty-five to forty career attorneys at any given time, though that number has recently increased.106 For years, the staff attorneys had offered written recommendations in major VRA cases — a procedure “meant to insulate such decisions from politics.”107 During the Bush Administration, however, political officials at the DOJ reportedly prohibited career attorneys from making recommendations in writing.108 This new policy against written recommendations was also followed by rare reversals of career staff preclearance recommendations by high-level political officials.109 This dynamic of silencing internal dissent appears to have continued into the Obama Administration110 but has since been disavowed.111
Going forward, election administration decisions made without the requisite internal deliberation should arguably be treated as more suspect by the courts and Congress, that is, more likely to reflect pure efforts at partisan entrenchment. They should be rejected even under Rodríguez’s own theory of regime change. They should not, that is, receive the deference that she otherwise calls for in the context of other agency decisions.112 Indeed, this perspective aligns with Rodríguez’s broader regime-based analysis. In her words, democratic responsiveness is a function of “appointees and civil servants”113 who are “collaborative and bring to bear sometimes rivalrous, but more often complementary, forms of reasoning and approaches to decisionmaking.”114 When these conditions are violated in the realm of election administration, however, the regime itself becomes further untethered to democratic preferences.
III. Scrutiny and Self-Restraint
In this manner, internal administrative process can serve as a check on partisan efforts to entrench power. When this process has been compromised, judicial or political monitors would be justified in scrutinizing the resulting electoral policy more closely. In this sense, internal and external review can serve as substitutes of sorts.115 The internal procedure itself is likely to check the most blatant attempts at self-entrenchment in a world where such effects are otherwise difficult to measure. In addition, it should also result in more transparent and well-reasoned public justifications, qualities that would enhance electoral legitimacy.
But when there are signs that internal deliberation has been subverted or is otherwise lacking, courts in particular should evaluate the underlying legal and policy rationales more closely under Chevron116 and arbitrariness review.117 Chevron, of course, requires courts to defer to agency interpretations of ambiguous statutes.118 Perhaps unsurprisingly, Rodríguez celebrates the doctrine’s ability to promote the “democratic evolution of the law” 119 in that it allows “the government to adapt to changing circumstances and changing times . . . by giving agencies the freedom to choose among plausible interpretations to advance their policy goals.”120
However, as argued here, such flexibility cannot be left unchecked in the context of election administration. In these circumstances, courts should perhaps apply a form of heightened Skidmore121 deference instead, which requires judges to be more independently persuaded by the administration’s legal argument.122 Under Skidmore, courts traditionally look at a number of factors such as the “thoroughness” of the agency actor’s consideration, the reasoning’s “validity” and “consistency,” and, more generally, any factors which give an interpretation “power to persuade, if lacking power to control.”123 When applying these factors in the context of election administration, judges should adopt a more skeptical stance.124 To be sure, there are difficult boundary problems — what constitutes election-related versus ordinary administration?125 — but courts can confront these difficulties on a case-by-case basis.
To illustrate the proposed approach, consider an interpretation of VRA section 2 advanced by the Reagan Administration. Under the plain text of the amended section 2, a trial court must assess section 2 violations based on a “totality of the circumstances.”126 Despite this, the Reagan DOJ essentially argued that a single factor — electoral success in any recent election — should be dispositive.127 The government’s amicus brief did not even attempt to reconcile this with the statutory text, nor the legislative history which made it clear that Congress intended the amendment to codify the “results test” as previously applied by the Court.128 The validity of this argument, however, is clearly suspect under a less deferential version of Skidmore review.129
Now consider arbitrary and capricious review under the Administrative Procedure Act.130 The doctrine demands that agencies provide technocratic reasons to justify their decisions.131 When election administration is at stake, however, the logic above suggests that courts should apply a harder look than usual. In fact, one could argue that judges have already been adopting this principle. Consider the Court’s recent application of the standard in Department of Commerce v. New York.132 Commerce Secretary Wilbur Ross sought to add a citizenship question to the decennial census on the grounds that doing so was necessary to enforce the VRA.133 Media accounts quickly circulated, however, detailing objections from career staff and irregularities in the internal decisionmaking process.134 A coalition of states, cities, and counties sued the agency all the way to the Supreme Court.135
Chief Justice Roberts, writing for the majority, found that the agency had not violated traditional arbitrariness review.136 Nevertheless, he remanded the agency’s decision on the grounds that its explanation was pretextual.137 In other words, Roberts refused to credit a clearly manufactured rationale, even if a traditional arbitrariness analysis might have done so.138 In this manner, the majority applied a kind of heightened scrutiny to a decision that had been marred by aberrations in internal agency procedure. Note, in line with Roberts’s approach, that the argument here does not require a wholesale rejection of partisan influence on election administration. Instead, it can be consistent with Rodríguez and others’ view that agency actions may result from permissibly political reasons.139 In the realm of election administration, however, those reasons must be disclosed transparently rather than concealed behind pretext. Such disclosure, in turn, would enhance the legitimacy of the decision.
At the same time, Rodríguez rightly notes that judges themselves are partisan actors, going as far as to characterize them as “agents of regime change.”140 In her estimation, the current Court reflects the efforts of a decades-long attempt by the “Reagan coalition” to consolidate power in the judiciary.141 Because judges possess life tenure, they represent a kind of “partisan entrenchment run amok” that threatens to delegitimize courts themselves.142 But Rodríguez does not go as far as to reject judicial supremacy, nor does she attempt to resolve how to manage the conflicts between the two kinds of regimes (political and legal) — simply noting that they will be a continuing “source of debate.”143
In light of Rodríguez’s valid concerns — particularly at a time when courts are increasingly minoritarian144 and Congress is moribund on voting rights reform145 — perhaps the ideal constraint on partisan entrenchment would come from the executive branch itself. That is, an incoming President could embrace robust regime change by agreeing to restrain himself in the arena of election administration and letting his policies be judged fairly at the polls. Adopting this stance publicly, in turn, could go a long way toward increasing the objective and perceived legitimacy of his administration. Establishing robust internal procedures to govern election administration would be an example of just that. The President could make his precommitment more credible by adopting these self-constraints through notice and comment, which increases the potential that they would be judicially enforceable.146 Like Odysseus tying himself to the mast, this restraint would only further empower a President’s administration147 — exactly the kind of robust regime change that Rodríguez, in this year’s Foreword, seeks.
** Professor of Law, University of Chicago Law School. Many thanks to Lisa Marshall Manheim for helpful comments and to Rob Clark for superb research assistance.