Defensible Disenfranchisement

I.       Introduction

The practice of felon disenfranchisement has a long and sometimes odious history. In the United States, the policy of denying criminal offenders the right to vote was adapted from the English tradition of “attainder,” as well as earlier notions of “outlawry” in continental Europe and infamia in ancient Greece and Rome.1 In these European contexts, offenders experienced a form of “civil death” that stripped them of their legal status and denied them even the most basic protections of the civil and criminal law.2 In colonial America, many people were excluded from the franchise as a matter of course, but even otherwise eligible citizens could be barred from voting based on a wide range of legal and moral transgressions.3 Among the salient features of these historical antecedents is the explicitly punitive nature of the practices and the determination not only to punish transgressors, but also to render them outsiders from the political community.

In its modern incarnation, the practice of denying the right to vote to some or all incarcerated felons or ex-felons4 remains common in the United States and survives in some form in Great Britain, Germany, and a number of other democracies.5 In the United States, felon disenfranchisement is formally regulatory, not punitive, but features the harshest restrictions on offender voting rights of any modern democracy.6 As a result, felon disenfranchisement is widely viewed as retrograde, reflecting the worst aspects of our history and our nature. Specifically, because disenfranchisement runs counter to the modern trend of extending voting and other fundamental rights, it is denounced as undemocratic and illiberal.7 In addition, because it is thought to impede the reintegration of offenders into their communities upon release from prison,8 it is deemed counterproductive to rehabilitation. Finally, despite its origins in post-Civil War Reconstruction, and motivated in particular by a determination to ensure the enfranchisement of former slaves, modern U.S. disenfranchisement policy is condemned as racist because it disproportionately affects black citizens and communities.9

Despite the gravity of these claims, the modern disenfranchisement debate is curiously one-sided. Although forty-eight states in the United States retain some form of felon disenfranchisement, the scholarly and popular literature is overwhelmingly hostile to the practice. Indeed, according to one scholar, “[c]ritics of disenfranchisement may feel a bit like a boxer entering the ring only to discover there is no opponent to fight.”10 Reflecting the prevailing criticism, a recent editorial in the New York Times registers its disapproval, declaring that “[t]he only reason not to let [ex-felons] vote is to stigmatize them or to continue punishing them.”11 And at least one federal court has expressed “skepticism as to whether any ‘non-racially discriminatory public policy rationales for disenfranchising felons’ actually exist.”12

Against this tide of opinion, I develop and defend a version of felon disenfranchisement that rejects the punitive conception in favor of a regulatory approach consistent with the values of a modern liberal democracy. I begin by considering some weak, but historically significant, arguments for disenfranchisement that have dominated the scholarly debate. The most common of these—protecting the “purity” of the ballot box, preventing “subversive” voting, and diminishing the likelihood of electoral fraud—have made easy targets for opponents of disenfranchisement and have stacked the deck against the regulatory conception. Unfortunately, the scholarly consensus against disenfranchisement based on these flawed rationales reflects almost no engagement with more plausible alternatives.13

A more promising case for disenfranchisement starts with a familiar account of political community that features liberal and republican elements. I situate defensible disenfranchisement within a liberal democratic context that highlights rights and responsibilities, citizenship and civic trust, and, in cases of breach, reassurance and reintegration.14 Contrary to the leading scholarly accounts, disenfranchisement is best conceptualized not as a form of punishment but as a means of regulating electoral eligibility in a liberal-democratic polity. On this view, offenders who commit serious felonies are subject to regulatory disenfranchisement because they have violated the civic trust that makes liberal democracy possible.

In contrast to more extreme policies, regulatory disenfranchisement is not the “civil death” of an earlier era or a modern mechanism for permanent political exclusion. Instead, defensible disenfranchisement is narrower in scope and application, temporarily denying the vote to only the most serious felony offenders and providing a meaningful opportunity for restoration of the franchise. Far from alienating offenders, as critics charge, the suspension of voting rights is meant to heighten offenders’ sense of civic responsibility by establishing the expectation of restored political participation. Indeed, precisely because the right to vote has high symbolic importance and relatively low practical value,15 disenfranchisement is ideally suited to mark the breach of civic trust that criminal wrongdoing represents without unduly disrupting an offender’s daily life. In this way, defensible disenfranchisement affirms, rather than betrays, our commitment to liberal-democratic community.

Before turning to this analysis, however, a disclaimer: Despite the scholarly and popular focus on the disparate racial impact of disenfranchisement policies, I generally set aside the racial dimension in the discussion that follows. First, because disenfranchisement is predicated on the criminal justice system, it is bound to reflect the disparities and limitations of that system as well as the wider society. Second, because the complicated issue of race transcends any single policy domain, it cannot be meaningfully addressed in isolation. Finally, although race figures prominently in virtually every modern critique, it generally is not the primary basis for rejecting disenfranchisement. It may turn out that intolerable racial consequences doom the practice in any event, but my present aim is to explore the possibility of a compelling and principled case for disenfranchisement. By limiting disenfranchisement to serious felony offenders and including a mechanism for restoration, this alternative version would drastically reduce the impact of felon disenfranchisement for all groups.

II.     Traditional Rationales

The U.S. Supreme Court has consistently held that felon disenfranchisement is constitutional, but it has never offered, or even addressed, a full-throated justification for the practice. In the twentieth century, the Court settled on a regulatory conception of disenfranchisement, casting it as a “nonpenal exercise of the power to regulate the franchise”—a way to “designate a reasonable ground of eligibility for voting.”16 In rejecting a challenge to a state disenfranchisement provision, the Court also effectively foreclosed future constitutional challenges, holding that “the exclusion of felons from the vote has an affirmative sanction in § 2 of the Fourteenth Amendment.”17 In earlier decisions, the Court deferred to state assessments of electoral eligibility, affirming provisions that denied voting rights to bigamists and polygamists pursuant to a state’s plenary authority to establish voting qualifications.18

Other courts have been more expansive, specifying what they take to be suitable grounds for disenfranchising criminal offenders as a matter of electoral eligibility. One prominent federal jurist invoked social contract theory as a basis for felon disenfranchisement, suggesting that a “man who breaks the laws . . . could fairly have been thought to have abandoned the right to participate in further administering the compact.”19 And in the nineteenth century, a state court upheld a criminal disenfranchisement provision on the ground that “one rendered infamous by conviction of felony, or other base offense indicative of great moral turpitude, is unfit to exercise the privilege of suffrage.”20

Modern commentators have focused on variations of three leading rationales—protecting the “purity of the ballot box,” preventing “subversive” voting, and minimizing electoral fraud—culled from decisions like these. In each instance, the case for disenfranchisement rests largely on an empirical concern about the pernicious consequences of permitting felons to vote. As a result, each is hostage to the facts of the matter and none stands up to closer scrutiny or countervailing considerations. But as discussed more fully below, each justification reflects elements of a fertile political tradition that points to a more defensible basis for felon disenfranchisement.

A.     Purity of the Ballot Box

A common theme in historical exclusions of criminal offenders from civic life is a concern that such persons will taint the body politic. In 1658, Plymouth Colony imposed “moral qualifications” for voting and other forms of civic participation lest “some corrupt members . . . creep into the best and purest societies.”21 In 1884, the Alabama Supreme Court, in upholding a constitutional provision that denied serious offenders the right to vote or hold public office, expressed a similar concern in Washington v. State: “The manifest purpose is to preserve the purity of the ballot box, which is the only sure foundation of republican liberty, and which needs protection against the invasion of corruption . . . .”22 According to the court, “[i]t is proper . . . that this class should be denied a right, the exercise of which might sometimes hazard the welfare of communities.”23

These accounts suggest at least two versions of the corruption argument, which George Fletcher has dubbed the “mystical” and the “fanciful.”24 According to the mystical version, the mere participation of offenders somehow infects the electoral process and the community as a whole with the taint of corruption.25 As Fletcher notes, is it difficult to credit this view, which seems to depend on a set of metaphysical commitments that are out of place in a “secular legal culture.”26 The alternative view—that offenders may pose a heightened risk to electoral integrity because they are more likely to engage in dishonest behavior—is fanciful absent evidence to substantiate the claim.27 As discussed more fully below, in the context of electoral fraud, the argument that criminal offenders threaten the integrity of elections has not been, and probably cannot be, convincingly made out.

Finally, as Pamela Karlan has observed, nineteenth century judicial decisions are of limited value in any case because they reflect an obsolete conception of the right to vote.28 In Washington, for example, the court distinguished the denial of a legal right from a policy of disenfranchisement, noting that the latter “merely withholds a constitutional privilege, which is grantable or revocable by the sovereign power of the State at pleasure.”29 In the modern context, denying the right to vote requires, at a minimum, something more substantial.

B.     Subversive Voting

The case for disenfranchisement based on the prospect of subversive or anti-social voting similarly takes both a practical and a more abstract form. As a practical matter, the argument goes, offenders can be expected to exercise the franchise in ways that undermine crime control and the rule of law. After all, their criminal conduct demonstrates a lack of respect for law, and allowing them to vote “could have a perverse effect on the ability of law abiding citizens to reduce the deadly and debilitating crime in their communities.”30 More philosophically, according to one influential strain of liberal contractarianism, criminal offenders have effectively declared war on their fellows and rendered themselves enemies of the social contract that structures relations among free and equal individuals.31 On this view, because offenders have violated the terms of the contract, they have forfeited their right to participate further in the enterprise of democratic self-governance.

The empirical claim that criminal offenders as a group are more likely to vote systematically to undermine law and order is entirely speculative. What little evidence there is suggests that offenders generally are as likely as other citizens to recognize the legitimacy and importance of criminal laws that protect persons and property.32 But even if the assumption of anti-social voting were correct, the attempt to “[f]enc[e] out” certain viewpoints from political representation is constitutionally untenable and, more fundamentally, antithetical to the democratic process.33 For “[t]he ballot is the democratic system’s coin of the realm. To condition its exercise on support of the established order is to debase that currency beyond recognition.”34 Moreover, the argument from forfeiture rests on too literal an interpretation of the contract metaphor and proves too much. We do not generally hold that a single breach of the law represents a wholesale repudiation of the social contract,35 and even the worst criminal offenders remain entitled to the protections of due process and the rule of law. Thus, like earlier notions of infamia and outlawry, permanent exclusion from the franchise for a discrete act of criminal wrongdoing represents a disproportionate response to law violation.36

C.      Electoral Fraud

The concern that criminal offenders are more likely to engage in electoral fraud reflects aspects of both the purity and subversion rationales. According to California’s highest court, a “tenable ground” for disenfranchisement is the fear that an offender “might defile ‘the purity of the ballot box’ by selling or bartering his vote or otherwise engaging in election fraud.”37 Further, “such activity might affect the outcome of the election and thus frustrate the freely expressed will of the remainder of the voters.”38 On this view, “those who break our laws should not dilute the votes of law-abiding citizens.”39

Again, there is no evidence to support the claim that criminal offenders generally are more likely to commit electoral fraud.40 Indeed, the familiar challenges of predicting future criminal conduct are only compounded if the prediction concerns the likelihood of a narrow class of offenses such as bribery or ballot-stuffing. In any case, disenfranchisement for all felons is both over- and under-inclusive as a mechanism for preventing electoral fraud. Even if we were to assume that election-law violators are more likely than others to engage in electoral fraud, the large majority of felons have committed offenses unrelated to the electoral process. At the same time, as Justice Marshall observed, “many of those convicted of violating election laws are treated as misdemeanants and are not barred from voting at all.”41 Finally, the vote-dilution argument is simply a variant of the subversive voting rationale and suffers from the same basic defect. Although a liberal democracy may establish standards for electoral eligibility, a governing majority cannot legitimately use its “power to preserve inviolate its view of the social order simply by disenfranchising those with different views.”42

Opponents of felon disenfranchisement, including Justice Marshall, maintain that denying the vote—based on viewpoint or offender status—“strikes at the very heart of the democratic process.”43 My burden is to outline a conception of liberal democracy that draws a sharp distinction between these grounds for disenfranchisement, making a case for the temporary suspension of voting rights for serious felony offenders as a legitimate aspect of electoral eligibility.

III.    Electoral Eligibility in a Modern Liberal Democracy

The standard justifications for felon disenfranchisement reflect aspects of both the liberal and republican strains in the Anglo-American tradition. The contract metaphor, explicitly invoked in connection with the forfeiture rationale, captures important insights from the liberal tradition, emphasizing individual rights, voluntary consent, and self-determination. Yet strict contractarianism alone is inadequate to account for the breadth and complexity of our political heritage and by itself yields an unappealing political conception. Similarly, the preoccupation with corruption in the purity rationale reflects a basic tenet of republicanism. And while republicanism in its most ambitious form overstates the polity’s legitimate interest in the personal virtue of citizens, it highlights an important dimension of civic responsibility on which liberal democracy depends. It is by now a commonplace that modern liberal democracies incorporate liberal and republican elements, and persistent debates about the nature and role of freedom, civic virtue, and the common good reflect matters of emphasis between (and within) these traditions. In sketching a conception of liberal democracy, I draw broadly from both traditions without attempting to identify in every instance a specific ideological source.

A.     Working Conception of Liberal Democracy

The liberal-democratic context for defensible disenfranchisement starts with a familiar account of individual rights and the duties that correlate with the rights of others. To this is added a conception of citizenship as an “office” that entails a set of responsibilities distinct from those to which others can claim a right.44 In this setting, “civic trust”45 is the starting assumption that others are fulfilling their responsibilities both as persons and as citizens. When that trust is violated, in cases of criminal misconduct, for example, we may reasonably require from violators a form of reassurance that they are worthy of our trust; that they are fit to resume office. To make this case, I sketch an account of the rights and responsibilities befitting the office of citizen in a liberal democratic polity.

1.     Rights and Responsibilities

In Locke’s influential formulation, civil society is the product of free and equal individuals recognizing the advantages of mutual cooperation and consenting to form political communities that secure their rights and coordinate their activities through the mechanisms of self-government and the rule of law.46 Contemporary liberalism emphasizes autonomy and pluralism, which provide individuals the authority and resources for determining the course of their lives according to their own conception of meaning and value. Among the definitive commitments of liberalism in all its forms is the recognition that individuals, endowed with a bundle of natural rights, can significantly alter the rights and duties that apply to them through their voluntary choices. In this way, we render ourselves liable to criminal punishment, for example, or to the authoritative enforcement of private contracts to which we have freely consented.

Less emphasized, especially in the modern context, are the responsibilities that liberal citizenship requires. These are not merely the “bilateral” duties we have as correlates of the rights of others.47 To be sure, this bilateral relationship is crucial and highlights the fact that the exercise of a right—to swing one’s fist, for example—is constrained by the rights of all others not to be punched in the nose. An altogether different set of responsibilities is captured by a “unilateral” view of the relationship between rights and responsibilities. As John Deigh observes, “some rights are such that one’s possessing them implies that one possesses specific responsibilities that set limits or conditions on the freedom those rights secure, and sustains it apart from any assumption that one possesses those responsibilities because others possess rights.”48 Thus, one may occupy a position—that of a traditional sovereign, for example—that entails duties (say, to maintain peace and security) that stem not from the rights of others but from the nature of the position one holds.49

2.     The Office of Citizen

The conception of citizens as officeholders draws on the classical notion of officium, or duty, and captures the sense that citizenship is a position of distinctive responsibility. Without attempting to develop a complete account of citizenship, I outline a minimal set of rights and responsibilities sufficient to ground the case for defensible disenfranchisement. The bottom line of this account is that the responsibilities of citizenship are broader than the most individualistic versions of liberalism and less demanding than the most ambitious forms of republicanism.50 In particular, it highlights the need for minimal civic virtue and a shared commitment to a set of public values that constitute the political community.

As an initial matter, the modern liberal democracy is committed to universal suffrage, according to which adult citizens are presumed to be qualified and entitled to vote. Although a citizen may be denied the right to vote based on a legal determination that he lacks the requisite mental capacity51—or, in most U.S. jurisdictions, because of a felony conviction—a citizen cannot be required to make an affirmative showing in the first instance that he is qualified to vote. In this context, to be qualified is simply to be an adult citizen not otherwise disqualified. Accordingly, literacy tests, property requirements, and judgments of moral character are impermissible grounds for excluding citizens from the franchise.52

The rejection of character tests, or other affirmative demonstrations of fitness, reflects a repudiation of the classical republican ideal of citizenship, at least in its strongest form.53 On that view, the role of the state is to cultivate the personal virtue of citizens according to a specific conception of the good, distinguished by active political participation, self-abnegation, and close identification of one’s own interest with that of the polity. From a liberal (and modern republican) perspective, this classical republican account of citizenship violates basic notions of liberty and equality and risks subordinating the individual to the will of the collective. In American history, the colonial preoccupation with moral purity, self-sacrifice, and individual virtue reflected this classical republican influence but gave way to a constitutional structure that de-emphasizes the importance of personal virtue for a successful democratic republic.

The rejection of strong republicanism, however, does not preclude an account of citizenship that incorporates at least a minimal degree of civic virtue. Indeed, such celebrated liberal theorists as Locke, James Madison, and John Stuart Mill stressed the value of civic virtue for the maintenance of a just and stable polity.54 For Madison, the suggestion “that any form of government will secure liberty or happiness without any virtue in the people, is a chimerical idea.”55 In the modern context, Rawls highlights the importance of “certain political virtues,” which “characterize the ideal of a good citizen of a democratic state.”56 In Rawls’s formulation of this republican ideal, “[t]he safety of democratic liberties requires the active participation of citizens who possess the political virtues needed to maintain a constitutional regime.”57

The conception of citizens as officeholders reflects the determination that citizenship amounts to something more than the self-regarding pursuit of individual interests unconstrained by anything but a rudimentary duty to obey the law. As participants in a common enterprise, citizens share a commitment to the public values that constitute the political community. With respect to voting rights, this means we must reject the view that “voting is precisely about expressing biases, loyalties, commitments, and personal values,”58 because this narrow view neglects the role responsibility that conditions the right to vote. Although a liberal polity has neither the authority nor the inclination to coercively enforce or otherwise compel civic-minded voting (or any voting at all), it operates with a reasonable expectation that citizens will exercise the franchise mindful of a broader set of concerns than their individual interests.59 In this way, citizens fulfill the responsibilities of their office by taking account of the public good when exercising their right to vote—and by trusting that others will do the same.

3.     Civic Trust

Civic trust is itself a duty of citizenship and describes the baseline expectation that other citizens are fulfilling the responsibilities of their office. Thus, in addition to the duties we owe one another as persons—to refrain from harming them or otherwise violating their rights—citizens also “owe it to each other to recognize each other as fellows: not to assume in advance that others are enemies who might attack[] and against whom [one] need[s] to guard [oneself].”60 By the same token, “we also owe certain kinds of (re)assurance to each other, to make clear to them that we can be trusted.”61 Although civic trust does not require a close-knit political community, it is unsuited to a conception of the polity populated by atomistic individuals. The presumption of trust and the prospect of restoration reflect an understanding that citizens, as members of a political community, can compromise their status without necessarily alienating themselves permanently from the community.

In the context of voting rights, civic trust operates as “a presumption that those who participate electorally are committed to the ends of such participation, namely, the good of the republic and the authorization of certain people to legislate for all.”62 From here, it may be tempting to follow a familiar path to felon disenfranchisement: “It is not unreasonable to suppose that those who have committed serious crimes may be presumed to lack this trustworthiness and loyalty. Criminals are, in the aggregate, less likely to be trustworthy, good citizens.”63 Thus, disenfranchisement is justified because offenders are, as an empirical matter, more likely to be socially irresponsible in exercising their voting rights.

By now we can recognize this as a form of the subversion rationale and promptly reject it. The disenfranchisement of serious offenders is not based on a prediction about anti-social voting or the fear of electoral fraud. Instead, it is based on the breach of civic trust that criminal wrongdoing constitutes. A serious criminal offender has violated the trust of his fellow citizens by flouting the laws collectively established for our mutual benefit. But our distrust is formal, not empirical, and follows as a matter of course from the fact of conviction. It does not require (or permit) a particularized assessment of the offender’s trustworthiness because the offense, without more, establishes the breach of civic trust.

Despite the close connection between crime and disenfranchisement, the loss of voting rights should not be mistaken for punishment. Although defensible disenfranchisement is triggered by a criminal conviction, it is justified by the breach of civic trust that serious criminal misconduct represents.64 And while punishment for serious crimes typically involves imprisonment and the separation of the offender from his geographic community, the loss of voting rights signifies his separation from the political community. As a result, disenfranchisement is not “so much a matter of meting out punishment as making a statement about the standards to which the state will hold each citizen if [he] is to retain [his] claim to be a full and equal member of the political community.”65 The right to vote is the relevant currency to mark the breach of civic trust because, unlike criminal punishment, voting is linked directly to citizenship and the rights and responsibilities of that office.

4.     Reassurance and Reintegration

Criminal offenders are subject to disqualification from the franchise because they have breached the civic trust that characterizes relations among citizens in a modern liberal democracy. By virtue of his breach, an offender must sit out the vote until he can reestablish the presumption of trust formally called into question by his criminal act. For crimes are acts “that violate the values of the common enterprise in the sense that they express a lack of even minimal commitment to those values.”66 Apart from his punishment, an offender owes a form of reassurance that signifies his restored commitment to those values, including the responsibilities of citizenship.67 In the meantime, the suspension of voting rights places him “at political arm’s length from the community.”68

We must proceed with caution, however, lest we revert to a conception of citizenship and criminality inconsistent with the values of a modern liberal democracy. In particular, we must resist the us–them mentality that casts felony offenders as irredeemable outcasts categorically incapable and unworthy of citizenship or decent treatment. For despite its roots in the ancient practices of infamia and outlawry, defensible disenfranchisement rejects “civil death” as a standard response to crime and operates with the expectation of restored political participation.

Critics have observed that disenfranchisement policies risk alienating offenders and work against the goal of offender reintegration. Because “[v]oting is a powerful symbol of political equality,”69 felon disenfranchisement “relegate[s] those convicted to the condition of second-class citizenship.”70 Thus, “[d]enial of voting rights creates permanent outcasts from society, persons internally exiled who are left without any opportunity ever to regain their full status as citizens.”71 In this way, disenfranchisement “affirms criminals’ exclusion from society rather than inviting them to integrate themselves into it.”72 In place of this counterproductive strategy, critics argue, “we should be encouraging inmates to begin thinking of themselves as useful members of society with all the attendant responsibilities.”73

Our working conception of liberal democracy supports this analysis—but only with respect to permanent disenfranchisement policies. The permanent exclusion of offenders, certainly those already released from prison, is tantamount to political exile and inconsistent with the rights and responsibilities of citizenship central to the working conception. As critics charge, a policy that denies even ex-felons the chance to restore themselves to the full office of citizen risks creating an “untouchable class”—a status intolerable in a modern liberal democracy.74

By contrast, the temporary disenfranchisement of serious criminal offenders is not only consistent with modern liberal democracy, it is uniquely suited to affirm the values of liberal democratic citizenship. Precisely because the right to vote is a powerful symbol of the office of citizen, its denial forcefully expresses the political significance of a breach of civic trust. Thus, temporary disenfranchisement highlights both the breach of trust that serious crime represents and the responsibilities of citizenship that condition the right to vote. As a result—and by design—denying the right to vote “is a symbolically serious matter . . . marking one’s temporary or permanent exclusion from the rank of full citizen, and thus from full membership of the polity.”75

The availability of restoration of the franchise in appropriate cases distinguishes defensible disenfranchisement from the extreme forms of exclusion still practiced in some American jurisdictions.76 The cycle of disqualification and restoration reflects the polity’s principled commitment to the maintenance of political community even, perhaps especially, with those who have violated its trust. The provision for disqualification marks a significant breach, and the mechanism for reassurance opens the door to reintegration. Together, they affirm the values that constitute the political community.

Despite the confident claims of disenfranchisement critics, it seems equally if not more intuitive that the temporary suspension of voting rights would promote reintegration by making salient the rights and responsibilities of citizenship. Although ex-felons with voting rights are among the least likely individuals to vote, there is some evidence to suggest that offenders believe the right to vote is meaningful.77 “According to several studies . . . regaining the vote has enormous symbolic importance because it signifies that one is again entitled to the same rights and privileges that any other member of the community enjoys.”78 If the right to vote is significant in this way, “then perhaps its temporary loss . . . will carry with them as they pursue reintegration into the community on their release.”79

The symbolic importance of voting rights contrasts with a variety of other state-imposed civil disqualifications to which a state may subject convicted felons. These range from restrictions on employment and benefits eligibility to curfews and residency exclusions.80 The dramatic practical impact these restrictions can have on offenders’ daily lives and on their prospects for rehabilitation distinguishes these disqualifications from disenfranchisement. For “[t]he more that convicted persons are restricted by law from pursuing legitimate occupations [and normal activities], the fewer opportunities they will have for remaining law abiding.”81 Because voting is, as a general matter, more symbolically than practically significant, the loss of voting rights is unlikely to produce the kind of material hardship that threatens rehabilitation.

Whether defensible disenfranchisement facilitates or retards the rehabilitation of offenders is ultimately an empirical question that we cannot presently answer. But while the case for defensible disenfranchisement is not predicated on its rehabilitative potential, the temporary suspension of voting rights strikes a responsible balance between the related goals of community affirmation and reintegration. Voting is the idiom of liberal democratic citizenship, and the temporary suspension of the franchise signifies the offender’s estrangement from the political community without unnecessarily disrupting the daily activities most likely to promote his reintegration.

B.     Defensible Disenfranchisement

The case for defensible disenfranchisement, predicated on a conception of liberal democratic community, informs but does not dictate the precise terms of an optimal disenfranchisement policy. Our working conception of political community focuses on a set of values relating to rights and responsibilities, citizenship and civic trust, and reassurance and reintegration. These values point to the key considerations relevant to a defensible policy of felon disenfranchisement—the offenses that should trigger it; whether to distinguish between felons and ex-felons in setting the length of disqualification; and the terms and conditions of restoration. In what follows I offer some preliminary thoughts about how best to structure a policy of defensible disenfranchisement, though the details are less important than the values they are meant to reflect.

1.     Triggering Offenses

The link between felony offenses and the practice of disenfranchisement, whether as a form of punishment or a matter of electoral eligibility, reflects a normative judgment about the seriousness of an offender’s criminal misconduct. More serious crimes deserve harsher punishments and, as a general matter, cast greater doubt on an offender’s commitment to a community’s public values. For much of the history of the common law, felony offenses were limited in number and were commonly equated with capital offenses.82 In these cases, the wisdom or fairness of disenfranchising felons was largely moot—and certainly the least of an offender’s concerns. Along the way, the standard punishment for most felonies became a term (or range) of years, and the category of felony offenses grew dramatically. In the United States, a felony is commonly defined as any crime punishable by death or more than one year in prison, and a typical criminal code includes dozens of relatively minor felony offenses sufficient to trigger disenfranchisement.83

The proliferation of felony offenses means that a felony designation may or may not correspond to our intuitive sense of what counts as a serious crime. In an attempt to confine disenfranchisement to only serious breaches of civic trust, we might usefully rely on a version of the traditional common law felony designations—murder, rape, arson, robbery, burglary, kidnapping, and prison escape.84 This list reflects a defensible historical judgment about the types of criminal act that represent the most profound breaches of community norms and values and generally includes those offenses that combine the highest level of culpability with the greatest degree of harm.

2.     Offender Status and Length of Disqualification

Defensible disenfranchisement is distinguished by the commitment, in principle, to restoration of the franchise in appropriate cases. This way of putting things signals that restoration must be generally available but may not apply to all offenders regardless of their circumstances. As an initial matter, we conventionally—and intuitively—distinguish between offenders still in prison and those who have been released after serving their sentences. The case for disenfranchisement seems strongest for prison inmates, perhaps because of the extended physical separation from the community and the other significant restrictions on personal liberty that prison entails. Although these considerations are not dispositive, they provide a reasonable starting point for drawing the relevant distinctions. For example, because inmates will not be meaningfully subject to most of the laws that apply to free citizens, we might think they lack the appropriate stake in electoral outcomes to justify participation.85

The intuition that imprisoned felons are more obvious candidates for disenfranchisement cuts both ways. The same considerations that support the case for denying the vote to felons serving prison terms would seem to require that we restore voting rights to ex-felons immediately upon their release. Despite the pull of this logic, it neglects at least one relevant consideration. Although defensible disenfranchisement is not based on empirical predictions about law abidingness, the requirement of reassurance is meant to have a practical dimension. Offenders provide reassurance by conforming their conduct to law for a period of time, and we might reasonably require them to assure us that their commitment is strong enough to withstand the challenges and temptations of everyday life.86 A record of good behavior in prison, with its strict regimentation and nearly total control over inmates’ lives, may or may not be sufficient to signal a restored commitment to the community’s public values.

Assuming a disenfranchisement policy that extends beyond an offender’s release from prison, we would have to decide how long the loss of voting rights should last. We have rejected permanent disenfranchisement because it is a form of internal exile inconsistent with liberal democratic values.87 Even a serious offender, after he is released from prison, should have a chance to restore his status once he has provided appropriate assurance that he is again worthy of civic trust.

Permanent disenfranchisement for ex-felons is not only inconsistent with our working conception of political community, it is also a disproportionate response to the breach of trust that serious crime represents.88 In the context of criminal justice, proportionality requires a close fit between crime and punishment. Because disenfranchisement is not a form of punishment, however, the period of disqualification does not require the same level of precision, varying according to the details of particular offenders or offenses. Instead, above a certain threshold of seriousness—established with reference to the common law felonies—a standard term of disqualification is appropriate to mark the breach of civic trust and to give an offender the chance to reassure the community that he is once again worthy of its trust. In the American political context, a four-year disqualification would ensure that an offender had to sit out a single presidential election before he would be eligible for restoration. And because presidential elections in the United States have special symbolic significance, the exclusion should be sufficient to signify the seriousness of the breach associated with the triggering offenses.

3.     Restoration

The restoration of voting rights marks the renewal of civic trust between an offender and the political community. In cases of serious crime, because the conviction itself establishes the breach of trust, disenfranchisement is automatic and does not await an empirical showing (or refutation) of an offender’s untrustworthiness. Matters are slightly different in the context of restoration. Although we do not generally require otherwise eligible citizens to establish their fitness as citizens, offenders have altered their relationship to the political community. “[T]heir distinctive normative status, which arouses reasonable anxieties in their fellow citizens, warrants requiring . . . reassurance from them.”89

Some disenfranchisement policies in the United States do provide for the restoration of voting rights.90 In many cases, however, the process is unreasonably complicated or discretionary, making restoration illusory as a practical matter.91 Instead, although mechanisms for restoration should require an affirmative effort on the part of offenders, the process should be readily accessible and operate with a presumption that voting rights will be restored. A reasonable basis for denial might be the commission of further offenses, even non-triggering felonies, during the period of disqualification. A lapsed offender would then be eligible to reapply after an additional year if he were to maintain a clean criminal record. Finally, the process should be explained to offenders upon their release from prison, should not require a fee or the assistance of an attorney, and should be ministerial rather than discretionary.

This account of the triggering offenses, length and linkage to incarceration, and the mechanism for restoration are only meant to be suggestive. It is an attempt to illustrate the relevant considerations in crafting a policy of defensible disenfranchisement and how they relate to the underlying values of a liberal democratic polity.

IV.    Conclusion

“The right to vote freely . . . is of the essence of a democratic society, and any restrictions on that right strike at the heart of representative government.”92 The expansion of voting rights in the modern era is a realization of our imperfect historical commitment to representative government and the values of equality, inclusion, and self-determination. In a liberal democracy, the right to vote signifies full membership in the political community, while the denial of universal suffrage calls into question democratic legitimacy. And because elections determine public policy, “[n]o right is more precious in a free country than that of having a voice in the election of those who make the laws under which, as good citizens, we must live.”93

Despite the profound significance of the right to vote in a liberal democratic polity, however, the likely impact of any single individual’s actual vote is notoriously small.94 Taken together, votes obviously determine elections, but an individual vote is important primarily for what it signifies about the voter’s status as a member—a citizen—of the political community. Under these circumstances, the temporary loss of a citizen’s right to vote dramatically conveys the significance of the breach of civic trust that serious crime represents without practically impairing an offender’s prospects for rehabilitation.

Defensible disenfranchisement capitalizes on this dynamic, emphasizing the expectations of citizenship and the importance of civic trust through the cycle of disqualification and restoration. As a regulatory counterpart to the institution of criminal punishment, it is continuous with a conception of punishment based on the expressive and retributive values of a liberal democratic community.95 Offenders deserve punishment for violating the criminal law, but citizens are liable to disenfranchisement for violating civic trust. Both are mechanisms by which a political community affirms its normative identity through the expression of its values, including the rights and responsibilities of citizenship.

  1. [1]. See Alec C. Ewald, “Civil Death”: The Ideological Paradox of Criminal Disenfranchisement Law in the United States, 2002 Wis. L. Rev. 1045, 1059–60 (describing exclusionary practices in ancient and medieval Europe).

  2. [2]. Id.

  3. [3]. Id. at 1061.

  4. [4]. For ease of discussion, I will refer generally to both types of offender as felons, though I later suggest some distinctions between felons and ex-felons that might be relevant to disenfranchisement policies.

  5. [5]. See Elizabeth A. Hull, The Disenfranchisement of Ex-Felons 12–13 (2006) (comparing U.S. practices to those of countries around the world).

  6. [6]. See id. at 13.

  7. [7]. See, e.g., id. at 9 (describing how state felon voting laws disproportionately take votes away from racial minority groups).

  8. [8]. See, e.g., John Kleinig & Kevin Murtagh, Disenfranchising Felons, 22 J. Applied Phil. 217, 230 (2005).

  9. [9]. See, e.g., George P. Fletcher, Disenfranchisement as Punishment: Reflections on the Racial Uses of Infamia, 46 UCLA L. Rev. 1895, 1897 (1999).

  10. [10]. Jeff Manza & Christopher Uggen, Locked Out: Felon Disenfranchisement and American Democracy 12 (2006).

  11. [11]. Editorial, Disenfranchised Felons, N.Y. Times (July 15, 2012),

  12. [12]. Pamela S. Karlan, Convictions and Doubts: Retribution, Representation, and the Debate over Felon Disenfranchisement, 56 Stan. L. Rev. 1147, 1159 (2004) (quoting Johnson v. Bush, 353 F.3d 1287, 1302 n.16 (11th Cir. 2003)); see also Fletcher, supra note 9, at 1903 (“There are so many constitutional arguments against the disenfranchisement of felons that one can only wonder at the survival of the practice.”).

  13. [13]. The philosophical literature features a handful of notable exceptions. See, e.g., Andrew Altman, Democratic Self-Determination and the Disenfranchisement of Felons, 22 J. Applied Phil. 263 (2005); Kleinig & Murtagh, supra note 8; Hugh LaFollette, Collateral Consequences of Punishment: Civil Penalties Accompanying Formal Punishment, 22 J. Applied Phil. 241 (2005); Jeffrey Reiman, Liberal and Republican Arguments Against the Disenfranchisement of Felons, Crim. Just. Ethics, Winter/Spring 2005, at 3. For a thoughtful history of the constitutional issues, see Richard M. Re & Christopher M. Re, Voting and Vice: Criminal Disenfranchisement and the Reconstruction Amendments, 121 Yale L.J. 1584 (2012).

  14. [14]. See infra Part III.A.

  15. [15]. See infra notes 75–81 and accompanying text.

  16. [16]. Trop v. Dulles, 356 U.S. 86, 97 (1958).

  17. [17]. Richardson v. Ramirez, 418 U.S. 24, 54 (1974); see also U.S. Const. amend. XIV, § 2 (providing for a reduction in state representation if voting rights are denied to adult male citizens, “except for participation in rebellion, or other crime”).

  18. [18]. See Davis v. Beason, 133 U.S. 333, 346–47 (1890); Murphy v. Ramsey, 114 U.S. 15 (1885).

  19. [19]. Green v. Bd. of Elections of N.Y., 380 F.2d 445, 451 (2d Cir. 1967).

  20. [20]. Washington v. State, 75 Ala. 582, 585 (1884).

  21. [21]. Ewald, supra note 1, at 1061 (quoting Cortland F. Bishop, History of Election in the American Colonies, in 3 Studies in History, Economics and Public Law 1, 53 (Univ. Faculty of Political Sci. of Columbia Coll. ed., 1893)) (internal quotation marks omitted).

  22. [22]. Washington, 75 Ala. at 585.

  23. [23]. Id.

  24. [24]. Fletcher, supra note 9, at 1899.

  25. [25]. Id.

  26. [26]. Id.

  27. [27]. Id.

  28. [28]. See Karlan, supra note 12, at 1150.

  29. [29]. Washington v. State, 75 Ala. 582, 585 (1884) (emphasis added).

  30. [30]. Manza & Uggen, supra note 10, at 12 (quoting Civic Participation and Rehabilitation Act of 1999: Hearing Before the Subcomm. on the Constitution of the H. Comm. on the Judiciary, 106th Cong. 44 (1999) (statement of Todd F. Graziano, Senior Fellow in Legal Studies, The Heritage Foundation)) (internal quotation marks omitted).

  31. [31]. A version of this position is often attributed to Locke. See, e.g., Ewald, supra note 1, at 1073–74 (“In Locke’s view, one who commits a crime forfeits his right to participate in the political process . . . .”); Re & Re, supra note 13, at 1598–99. However, it is not clear that this view is genuinely Lockean. Specifically, the passage cited to establish Locke’s position on forfeiture—that a criminal has “declared war against all mankind”—comes in the discussion of punishment in the state of nature before people have entered into civil society. See John Locke, Two Treatises of Government and a Letter Concerning Toleration 104 (Ian Shapiro ed., Yale Univ. Press 2003) (1689). This does not justify forfeiture in civil society. See also Reiman, supra note 13, at 10 (“[T]he passage describes the criminal as declaring war on all mankind while in the state of nature, and each of these—the state of war and the state of nature—plays a different role in Locke’s theory.”).

  32. [32]. See Ewald, supra note 1, at 1099 n.219 (citing studies); Reiman, supra note 13, at 10 (citing studies that surveyed criminal defendants and finding that “almost all of them believed that what they had done was wrong, and that the law they violated was worthy of respect”).

  33. [33]. Carrington v. Rash, 380 U.S. 89, 94 (1965).

  34. [34]. Richardson v. Ramirez, 418 U.S. 24, 83 (1974) (Marshall, J., dissenting).

  35. [35]. Indeed, according to Locke, even the state of nature is governed by a proportionality principle. See Locke, supra note 31, at 104–05.

  36. [36]. Id.

  37. [37]. Otsuka v. Hite, 414 P.2d 412, 417 (Cal. 1966).

  38. [38]. Id.

  39. [39]. Manza & Uggen, supra note 10, at 12 (quoting 148 Cong. Rec. S802 (2002) (statement of Sen. Mitch McConnell)) (internal quotation marks omitted).

  40. [40]. See id. at 13; Nora V. Demleitner, Continuing Payment on One’s Debt to Society: The German Model of Felon Disenfranchisement as an Alternative, 84 Minn. L. Rev. 753, 773 (2000).

  41. [41]. Richardson v. Ramirez, 418 U.S. 24, 79 (1974) (Marshall, J., dissenting).

  42. [42]. Id. at 82.

  43. [43]. Id.

  44. [44]. John Deigh, On Rights and Responsibilities, 7 Law & Phil. 147, 155–56 (1988).

  45. [45]. I borrow (and probably distort) this concept, and the related notion of reassurance, from Antony Duff. See RA Duff, Pre-trial Detention and the Presumption of Innocence, in Prevention and the Limits of the Criminal Law 115, 123 (Andrew Ashworth et al. eds., 2013).

  46. [46]. See generally Locke, supra note 31.

  47. [47]. See Deigh, supra note 44, at 155.

  48. [48]. Id.

  49. [49]. See id.

  50. [50]. Although the most extreme versions of these positions are probably caricatures, I invoke them simply to distinguish the moderation of the working conception.

  51. [51]. See, e.g., 42 U.S.C. § 1973gg-6(a)(3)(B) (2006).

  52. [52]. See, e.g., City of Phoenix v. Kolodziejski, 399 U.S. 204 (1970); Cipriano v. City of Houma, 395 U.S. 701 (1969); Kramer v. Union Free Sch. Dist. No. 15, 395 U.S. 621 (1969); Katzenbach v. Morgan, 384 U.S. 641 (1966).

  53. [53]. Despite the varieties of republicanism, including classical republicanism, civic humanism, and civic republicanism, I distinguish broadly here between classical (or strong) republicanism and civic (or modern) republicanism. For a careful overview of these distinctions, see Frank Lovett, Republicanism, in Stanford Encyclopedia of Philosophy (Edward N. Zalta et al. eds., 2013), available at

  54. [54]. See Christopher P. Manfredi, Judicial Review and Criminal Disenfranchisement in the United States and Canada, 60 Rev. Pol. 277, 295 (1998).

  55. [55]. James Madison, Remarks to the Virginia Convention (June 20, 1788), in 3 The Debates in the Several State Conventions on the Adoption of the Federal Constitution 537 (Jonathan Elliot ed., 2d ed. 1836).

  56. [56]. John Rawls, Political Liberalism 194–95 (1993).

  57. [57]. Id. at 205.

  58. [58]. Fletcher, supra note 9, at 1906.

  59. [59]. Cf. Rawls, supra note 56, at 215.

  60. [60]. Duff, supra note 45, at 123.

  61. [61]. Id.

  62. [62]. Kleinig & Murtagh, supra note 8, at 224 (defining “political trust”).

  63. [63]. Roger Clegg, Who Should Vote?, 6 Tex. Rev. L & Pol. 159, 172 (2001).

  64. [64]. Cf. Altman, supra note 13, at 268.

  65. [65]. Id. at 265.

  66. [66]. Christopher Bennett, The Apology Ritual: A Philosophical Theory of Punishment 165 (2008).

  67. [67]. See Duff, supra note 45, at 123.

  68. [68]. Altman, supra note 13, at 265.

  69. [69]. Manza & Uggen, supra note 10, at 18.

  70. [70]. Andrew von Hirsch & Martin Wasik, Civil Disqualifications Attending Conviction: A Suggested Conceptual Framework, 56 Cambridge L.J. 599, 624 (1997).

  71. [71]. Demleitner, supra note 40, at 775.

  72. [72]. Reiman, supra note 13, at 9.

  73. [73]. Fletcher, supra note 9, at 1907.

  74. [74]. See id. at 1898.

  75. [75]. R.A. Duff, Introduction: Crime and Citizenship, 22 J. Applied Phil. 211, 213 (2005).

  76. [76]. See Hull, supra note 5, at 6.

  77. [77]. See Manza & Uggen, supra note 10, at 151.

  78. [78]. Hull, supra note 5, at 45.

  79. [79]. Suavé v. Canada (Chief Electoral Officer), [2002] 3 S.C.R. 519, para. 183 (Can.) (Gonthier, J., dissenting).

  80. [80]. See von Hirsch & Wasik, supra note 70, at 602–03.

  81. [81]. Id. at 605.

  82. [82]. 4 William Blackstone, Commentaries *94–101.

  83. [83]. See, e.g., Manza & Uggen, supra note 10, at 8 & 291 n.6 (listing examples, including “misrepresentation of tobacco leaf weight” and “misrepresentation by refrigerator contractors” (internal quotation marks omitted)).

  84. [84]. Other possible candidates for triggering offenses might include treason, large-scale financial crimes, and genocide.

  85. [85]. Note that having a stake is at most a necessary but never a sufficient condition for inclusion in the franchise. Were it sufficient, we would have to permit non-citizens, resident or otherwise, to vote since nearly any citizen of the world could plausibly claim an interest in U.S. (and other countries’) policy outcomes.

  86. [86]. This is not to discount the unimaginable challenges facing inmates in the modern prison. It is only to suggest that the relevant challenges are those facing community members outside the restrictive and totalizing environment of a prison.

  87. [87]. See supra text accompanying note 74.

  88. [88]. Although the language of proportionality is most at home in the context of punishment, it applies more broadly to other public policy domains as well. As a matter of basic justice, taxes, licensing fees, and various forms of civil disqualification should be proportionate to whatever relevant considerations justify the policy.

  89. [89]. Duff, supra note 45, at 126.

  90. [90]. See Hull, supra note 5, at 6–8 & tbl.2 (identifying the policies of each state and noting the wide variation of policies regarding restoration of voting rights).

  91. [91]. See id. at 6.

  92. [92]. Reynolds v. Sims, 377 U.S. 533, 555 (1964).

  93. [93]. Wesberry v. Sanders, 376 U.S. 1, 17 (1964).

  94. [94]. See, e.g., Andrew Gelman et al., What Is the Probability Your Vote Will Make a Difference?, 50 Econ. Inquiry 321, 323–24 (2012) (predicting that the chance of a single American casting a decisive vote in the 2008 presidential election was 1 in 60 million).

  95. [95]. See generally R.A. Duff, Punishment, Communication, and Community (2001).


Associate Dean of Faculty, Professor of Law, Arizona State University Sandra Day O’Connor College of Law. 

I am grateful to the participants in the 2012 MANCEPT Workshop, Legitimacy, Authority and Political Obligation, and the 2013 Collateral Sanctions Workshop at the Robina Institute at the University of Minnesota Law School. Special thanks to Chris Bennett and Zach Hoskins for helpful feedback along the way. Thanks also to Courtney Burks, Andrew Johnston, and the staff of the Iowa Law Review.