Ending Voting Discrimination by Compelling Voting

Tom Butcher | March 17, 2023


This post proposes that Congress adopt a system of compulsory voting to address the problem of voting discrimination. Such a system would have considerable advantages over the Voting Rights Act (particularly after Shelby County and Brnovich) and over other recently proposed legislation because it would address both governmental impairment of the right to vote and also the significant socioeconomic factors correlated with low voter turnout. Congress should have the power to enact such a plan through its Fifteenth, Nineteenth, and Twenty-Sixth Amendment powers. First Amendment concerns about compelling speech can be avoided by providing abstention options. Any concerns about criminalization can be mitigated by carefully implementing the penalty for non-voting in a way that takes account of economic status. Ultimately, I argue that a system of compulsory voting represents the most effective way to reduce voting discrimination and civic inequality in the United States.

Keywords: voting rights; discrimination; racism; classism; constitutional law; proposed legislation

Table of Contents


I.      The Voting Rights Act and Its Shortcomings

II.    Advantages of Compulsory Voting

III.   Difficulties in Enactment and Implementation

A.      Constitutional Difficulties: Individual Rights

B.      Constitutional Difficulties: Congressional Power

C.     Concerns about Criminalization

D.     Other Implementation Concerns


Appendix: Model Statute


“If a single statute represents the best of America, it is the Voting Rights Act.”[1] So Justice Kagan proclaimed with lachrymose solemnity in the opening sentence of her Brnovich v. Democratic National Committee dissent. If those words were not yet an obituary for the Voting Rights Act (VRA), certainly they were an acknowledgment that the Act only lives under a do-not-resuscitate order. Brnovich, along with the Court’s earlier decision in Shelby County, Ala. v. Holder,[2] has left the VRA a shell of its former self. Congressional Democrats have proposed at least two bills—the For the People Act (a.k.a. H.R. 1)[3] and the John Lewis Act[4]—that would do much to restore the rights formerly protected by the VRA and to add new requirements promoting access to the ballot. However, even if enacted, these bills would still be insufficient to wholly solve the problem of voting discrimination and civic inequality in the United States.

I propose that Congress pursue a different legislative strategy altogether: make voting compulsory. Many instinctively recoil from this idea, but it has been implemented successfully in several other countries. As of 2021, at least twenty-seven democracies use some form of mandated voting.[5] In comparison to currently proposed legislation, instituting a compulsory voting scheme would offer more significant protection to voting rights by addressing to a much greater extent the social and economic causes of voting discrimination and civic inequality.

Implementing this proposal would not be a simple matter. There are serious constitutional issues at stake, including whether compulsory voting would violate the First Amendment and whether Congress has the power to enact such a plan. Additionally, a poorly designed compulsory voting scheme could end up criminalizing those who, in many cases, are already among the least privileged in this country. However, I believe that this plan, while unorthodox, would ultimately be constitutional, particularly if it includes provisions allowing one to opt out of the obligation. Concerns about criminalization can be overcome by careful attention to the implementation of the law.

In this essay, I will first elaborate on the inadequacies of the VRA and of recent legislative proposals to amend it. I will then examine a model statute for compulsory voting (located in the Appendix) and explain how a compulsory approach would better protect access to the ballot box. Finally, I will address the potential difficulties with the plan and discuss how they may be overcome.

I.          The Voting Rights Act and Its Shortcomings

Before Shelby County and Brnovich, the VRA protected access to the ballot box in two principal ways. First, § 2 of the Act prohibited any state or political subdivision from adopting any voting procedure that “results in a denial or abridgement of the right . . . to vote on account of race or color.”[6] This right of action was a powerful tool to redress wrongs, but, as Congress had learned in the years preceding the VRA’s 1965 enactment, individual suits are poor vehicles for solving the systemic problem of voter disenfranchisement.[7] “Early attempts to cope with this vile infection resembled battling the Hydra” because as soon as courts struck one restriction down, two new methods of disenfranchisement would rise to take its place.[8] To address this “whack-a-mole” problem,[9] Congress enacted a “preclearance” requirement in § 5 of the Act, whereby certain states and counties with particularly egregious histories of resisting the Fifteenth Amendment’s extension of voting rights to all races would have to obtain approval from the Department of Justice (DOJ) before enacting any change to voting laws.[10] To determine which states and counties § 5 covered, Congress created a formula, enacted in § 4(b), which the DOJ followed in enforcing the Act. The formula applied § 5 to any jurisdiction that had previously used voting tests to control the franchise or that had less than 50% voter participation as of 1972 (in the most recent reauthorization), unless the jurisdiction had since made a showing of significant improvement.[11] Over the next five decades, Congress reauthorized the VRA four separate times.[12] The most recent reauthorization, in 2006, passed the Senate 98–0.[13]

Seven years later, in Shelby County,  the Supreme Court invalidated the preclearance formula in § 4(b) with a 5–4 decision because—so it claimed—the formula was out of date.[14] Although the Court left § 5 intact,[15] that section cannot function without a coverage formula, meaning that it remains unenforceable unless and until Congress amends § 4(b) in a manner consistent with Shelby County’s holding. Section 2 remains good law, but it also remains subject to the whack-a-mole problem. Moreover, the Court has subsequently indicated in Brnovich that it will treat future § 2 cases with the utmost skepticism.[16] The result is that today states are free to enact blatantly discriminatory voting laws, and so long as they can articulate facially plausible excuses invoking “election security,” the law is likely to withstand judicial scrutiny.[17] Since Shelby County, more than thirty states have enacted restrictive voting laws.[18] 2021 alone saw more than 440 restrictive bills introduced across forty-nine out of fifty states in the country, with at least thirty-four bills becoming law.[19] Had these bills been subject to the VRA’s preclearance requirement, the DOJ almost certainly would have nixed a large portion.

This plethora of discriminatory laws clearly illustrates the problem created by the Court’s dismantling of the VRA. Yet, as strong a tool as the VRA was for protecting the right to vote, the extensive geographical breadth of the push to restrict voting also illustrates the weaknesses of the VRA’s approach. At the time of Shelby County, the VRA’s preclearance requirement applied to only nine states and some jurisdictions within six other states.[20] Thus, even if Shelby County had gone the other way, many of these bills (those from jurisdictions not covered by the § 4(b) formula) would never have been subject to the preclearance requirement. They would only have been addressed (if at all) through the “whack-a-mole” protection of § 2.

Additionally, even at its most expansive, the VRA could only address direct government action. It could do nothing about the myriad other socioeconomic factors that contribute to severe disparities in voter turnout along predictable demographic lines.[21] In the 2020 federal election, slightly more than two-thirds of adult citizens voted.[22] This paltry sum represents the highest turnout of the twenty-first century.[23] It is substantially lower than voter turnout in most, if not all, of our peer democracies—even those that do not use compulsory voting.[24] Indeed, “it is likely that no U.S. President has ever received a majority of the votes of the American adult population.”[25] Given that ours is a government supposedly constituted by “We the People,” such low turnout creates severe problems of legitimacy.

The American tradition of low turnout is not demographically neutral. In 2020, 70.9% of white non-Hispanic adults voted, but only 62.6% of Black adults voted.[26] When comparing by income, the differences become even starker. In 2020, only 47.1% of those with less than $10,000 in annual household income voted, compared to 84.8% of those with annual incomes greater than $150,000.[27] Age, too reveals significant disparities: eighteen-year-olds have the lowest voting rate by age (41.8% in 2020), while 78-year-olds have the highest rate by age (79.7% in 2020).[28] In short, the older, richer, and whiter you are, the more likely you are to vote (and also the more likely you are to be able to vote). These disparities in the demographics of voters and non-voters severely undermine our government’s claim to be a representative democracy.[29]

Some of the racial disparities in the 2020 election could stem from the restrictive voting laws passed since Shelby County. And, since Black people have lower incomes on average than white people, anything that contributes to a disparity of race in voting is also likely to contribute to disparities of income in voting (and vice-versa). In 2012, the last presidential election before Shelby County, Black adults voted at a rate of 66.2%—which was 3.6 points higher than in 2020 and, in fact, higher than that of white adults in 2012.[30] Moreover, given that 2012 had a much lower overall voting rate than 2020 (61.8% compared to 66.8%), the actual impact of the restrictive laws passed since 2013 is probably more like 9 points than 3.6.[31] However, the fact that Black turnout still reached only 66.2% even while the VRA was fully in force shows that the Court’s dismantling of the law can only partly explain today’s low turnout rate.

Consequently, any attempt to simply adopt a new preclearance formula will not be able to accomplish much more than the VRA itself was able to do. Even an expansion to the preclearance formula along the lines of the John Lewis Act[32] would likely be of only limited effect. Although adopting a new formula based on more recent voting data would make § 4(b) presumptively constitutional,[33] the current Court is more conservative than it was in 2013 and may be opposed to any federal intervention into state and local voting practices. When faced with a suit challenging a hypothetical new coverage formula, today’s Court could well follow the lead of Justice Thomas’s Shelby Countyconcurrence and find the preclearance requirement itself unconstitutional.[34] H.R. 1 takes a different approach from the John Lewis Act—it would adopt an omnibus package of election access protections—but it still fundamentally suffers from the “whack-a-mole” problem of § 2 of the VRA. Moreover, even if either of these bills completely halted government discrimination, they still would have only limited effect on the other social and economic factors that correlate with disparities in turnout.

II.        Advantages of Compulsory Voting

Unlike H.R. 1 and the John Lewis Act, a compulsory voting scheme would significantly reduce voting discrimination and civic inequality. In support of this claim I have drafted portions of a model statute (see Appendix A). My model statute is incomplete, as many of the specific details would need to be rooted in far more extensive research than is feasible in a blogpost of this sort. But its core requirement is quite simple:

§ 3.      All persons in the United States shall be required by law to vote in every governmental election in which they are legally permitted to vote.

The language of this provision establishes that, if one may vote in an election, then one must. It applies not only to federal elections but also to state and local elections.[35] Although many details of implementation must still be worked out (some of which I discuss below), this core requirement would provide notable advantages over H.R. 1 and the John Lewis Act, as well as the pre-Shelby County VRA.

First, and most obviously, this statute would increase voter turnout across the board. As political scientist Jill Sheppard notes, “the literature is almost unanimous in finding compulsion increases voter turnout at the aggregate level.”[36] Some commentators have described low voter turnout as a sort of “collective action” problem—“If left to individual choice, the level of voting theoretically will be below the socially optimal level. Like jury service, taxes, and the draft, compulsory voting is a legitimate way to solve such a market failure.”[37] Although one should not characterize the act of not voting as a simple “choice”—with no attention paid to broader socioeconomic and historical factors—the fact remains that some people do simply choose not to vote, and for a variety of reasons. Increasing the direct cost of not voting, even only marginally, is likely to have significant effects. Some studies have shown that compulsory voting laws increase voter turnout across the board by seven to sixteen points.[38] Moreover, given that the United States would start from such a low turnout baseline, the effects might be particularly strong in this country.[39]

Evidence indicates that compulsory voting disproportionately boosts turnout among groups whose voting rates would otherwise be disproportionately low. Sheppard observes that “Compulsion increases turnout among socioeconomically disadvantaged citizens, reducing the stratification between voters and non-voters evident in democracies where voting is concentrated among those citizens with high socioeconomic status.”[40] Shane Singh similarly concludes that “the voting population will be more reflective of the entire electorate in compulsory systems.”[41] Additionally, these authors’ studies focus on the socio-economic effects of imposing a penalty for non-voting, which is, unsurprisingly, likely to have a disproportionately strong effect on those who would otherwise be unlikely to vote. However, my model statute aims to further increase the effectiveness of compulsory voting in increasing turnout among the young, poor, and non-white by imposing a duty upon all governments not to restrict one’s fulfillment of the obligation to vote:

§ 3(a).  No government entity or agent shall restrict any person from fulfilling the legal obligations imposed by this section. No private entity or person shall intentionally and substantially restrict any person from fulfilling that obligation.[42]

Third, a compulsory voting obligation would solve some of the fundamental shortcomings of the VRA. By implementing a simple nationwide standard, this model statute would apply across the entire country, rather than only in a handful of preclearance states and counties. Moreover, it would obviate the need for any coverage formula whatsoever and would thereby avoid the constitutional difficulty of “depart[ing] from the fundamental principle of equal sovereignty” that led the Shelby County Court to strike down § 4(b) of the VRA.[43] Additionally, compulsory voting would likely reduce the “whack-a-mole” problem dramatically. Rather than evaluating whether a voter-restrictive policy was one of many factors that led a voter to stay home, courts would instead inquire whether it restricted a citizen from fulfilling their legal duty. When the background assumption is that citizens need to fulfill an obligation, rather than choose whether to exercise a right, then interference should become considerably easier to spot.

A plan of compulsory voting along the lines proposed in the model statute would alleviate the shortcomings of § 2 and § 5 of the VRA. Even more importantly, it could address the socioeconomic factors underlying much of voting discrimination and civic inequality in this country—factors that the VRA and recent legislative proposals could never hope to address. In doing so, it would more greatly alleviate voting discrimination and civic inequality and would have a secondary benefit of raising voter turnout across the board. This would lead to a more legitimate government that is better equipped to respond to the voices of the people who constitute it.

III.       Difficulties in Enactment and Implementation

Any federal plan of compulsory voting will need to overcome several significant difficulties of enactment and implementation. Here I am interested in legal and practical difficulties. Consequently, I put aside the most obvious problem, which is that Congress is unlikely to endorse this model statute or a similar one at any point in the near future. Instead, I focus, first, on whether a law like the model statute would be constitutional; and, second, on how to implement the law without also criminalizing those who do not comply with it.

A.        Constitutional Difficulties: Individual Rights

The constitutional difficulties with a compulsory voting scheme are potentially numerous. The first category of problems relates to individual rights. While some of these issues can be side-stepped by allowing potential voters to opt out of the obligation, the constitutional issue is still worth exploring.

Does the right to vote also imply the right not to vote? Such an implication might seem intuitively to be the case. After all, not even the most zealous gun rights advocate would hold that the Second Amendment compels people to bear arms. Nevertheless, the Supreme Court has held on numerous occasions that the right to one thing “does not ordinarily carry with it the right to insist upon the opposite of that right.”[44] Many examples occur in the context of criminal procedure. For example, the constitutional right to a public trial does not imply the right to a private trial,[45] and the right to a jury trial does not necessarily imply the right to a bench trial, or even the right to forego a trial altogether.[46]

Some examples occur in the civil context as well. For example, the right to a minimum wage does not permit an employee to agree to work for less than that amount.[47] Here the reasoning is instructive: one cannot waive the right to a minimum wage because doing so “would ‘nullify the purposes’ of the [Fair Labor Standards Act] and thwart the legislative policies it was designed to effectuate.”[48] In other words, if an employer could simply choose to hire those who were willing to waive their right to a minimum wage, then that would effectively nullify the right for all other workers.

In a similar manner, the right to vote can also be construed as a duty to vote—because the public has an interest in having the American government accurately represent the body of its constituents. As one author argues, “The individual act of voting is essential to the collective’s ability to have democratic government, and as such should not be waivable.”[49] Much like the right to jury trial imposes a duty of jury service, the right to constitute the government should also create a duty to constitute the government.

Another spin on the individual rights problem is the question of whether a duty to vote constitutes a governmental compulsion of speech in violation of the First Amendment. Unlike the rights discussed above, the right to free speech does contain the right not to speak.[50] However, this is not because the First Amendment implicitly includes its opposite; rather, free speech includes the right to stay silent because silence is itself a form of speech.[51] Here there is a close analogy to voting—one could argue that not voting is actually a means of exercising the franchise, in addition to being a form of speech. However, both difficulties can be avoided by simply allowing persons to opt out of the obligation. The model statute addresses this concern in a number of ways, the most important of which is the following:

§ 4.      All ballots in elections affected by this Act shall have an abstention option.

Additionally, the model statute contains a provision (§ 4(a)) allowing one to notify the government in advance of an intent not to vote, which serves as an absolute defense against any charge of non-compliance. By supplying these abstention options, the government avoids a quandary in which it might end up forcing one to vote for a candidate whom one cannot support. Additionally, the lightness of the penalty for not voting (discussed below) further diminishes the extent to which the government could be considered as compelling speech through this law.

B.         Constitutional Difficulties: Congressional Power

The second category of constitutional difficulties relates to congressional authority to institute compulsory voting. Section 4 of Article I of the United States Constitution gives Congress the power to control some aspects of Senate and House elections.[52] When the Constitution was ratified in 1789, it explicitly left primary control over presidential elections to the states,[53] and no provision gave Congress any power over state and local elections. Since then, however, the Fifteenth, Nineteenth, and Twenty-sixth Amendments have given Congress some power over all government elections by granting it the power to pass legislation enforcing voting rights on the respective bases of race, sex, and age.[54] As noted above, significant disparities in voting exist along racial, gender,[55] and age lines.[56] Because compulsory voting would alleviate these disparities,[57] Congress would have the power to compel states to follow the model statute.[58]

Critically, the Court has held that “The Fifteenth Amendment empowers ‘Congress,’ not the Court, to determine in the first instance what legislation is needed to enforce it”[59]—and the same logic would apply to the Nineteenth and Twenty-sixth Amendments, since they empower Congress through identical language. Granted, the Shelby County Court held that, in reauthorizing § 4(b) of the VRA, Congress exceeded the powers granted to it under the Fifteenth Amendment—but the problem there was a failure to respect the “fundamental principle of equal sovereignty among the states.”[60]Such a problem would not present itself under the model statute, which would bind all states equally.

As a fallback position, even if the Court were to find lacking Congress’s power to regulate state and presidential elections pursuant to the Fifteenth, Nineteenth, and Twenty-sixth Amendments, Congress would still retain its Article I, Section 4 power to regulate congressional elections.[61] Severability provisions in the model statute (see § 8) express Congress’s intent for the construction of the statute in the event that any provisions are found unconstitutional. These severability provisions instruct the Executive Branch to enforce the statute as written, minus any unconstitutional sections. They should ensure that Congress maintains the power to compel voting in congressional elections, even if the Court strikes down other parts of the statute.[62] Although a narrow power over congressional elections would be a blow to the ultimate goals of the model statute, even this remaining compulsion would likely prove quite effective in encouraging higher turnout rates for state and local elections. Because states and municipalities tend to hold many of their elections on the same days as congressional elections, the fact that voters would be compelled to go to the poll anyway—to vote for a Representative or Senator—means that they would also be likely to also vote for whatever other local, state, and federal matters are present on the same ballot.

C.        Concerns about Criminalization

This statute is designed to alleviate voting inequality along racial, economic, and age lines. It does so by imposing a penalty—a form of criminalization—on those who do not vote. However, this criminalization introduces a number of related problems. First, the demographics that the statute aims to help (those who are less likely to vote) are also those most at risk of criminalization from a compulsory voting scheme. Second, it is well-established that poorer and non-white people are more likely to have negative interactions with the police.[63] This structurally racist and classist reality means that enforcement of the law is likely to fall disproportionately on people of color and the poor, even given the sociological realities of the first problem. Third, the demographics most at risk of criminalization are also those that would be most disproportionately harmed by the penalties the statute imposes, owing to lower average wealth. 

These are major drawbacks to any plan of compulsory voting. Without careful implementation they could become so severe as to outweigh compulsory voting’s benefits. Consequently, the model statute incorporates several measures to eliminate or diminish this counterproductive potential result as much as possible. It does this, first, by adopting a relatively minor penalty that takes economic status into account, thus minimizing the impact of criminalization on the poor, the young, and people of color. Second, it aims, where possible, to shift the penalty for non-voting onto those who are more able to bear it, such as bosses who refuse to grant their employees time off to vote.

Countries that have adopted compulsory voting use a number of different enforcement mechanisms.[64] Some countries, such as Brazil and Peru, tie proof of voting to certain social benefits rather than imposing a direct punishment.[65] This is also the approach taken by the Selective Service System in the United States: although one can nominally be prosecuted for failing to register for the draft, in practice the punishment is restriction in access to government benefits such as driver’s licenses.[66] Such a system could conceivably work in the voting context, but I am hesitant to adopt it for at least two reasons. First, those who are less likely to vote are also more likely to be reliant on government benefits, and the processes of obtaining those benefits are already byzantine enough in this country.[67]  Second, registering for the draft is a one-time obligation that one can fulfill at any time, which means that non-compliance has an easy remedy—registering later. Conversely, elections happen frequently, but only on specific days, which means that “curing” one’s status as a non-voter could only occur at the next election.

I have instead opted in the model statute to follow an approach like Australia’s, where the government can fine those who do not vote.[68] A simple fine can calibrate the penalty for non-compliance so that it does not severely impact even the destitute but still creates enough of an annoyance to ensure that most people comply. Moreover, the statute instructs the agency administering the statute to exercise discretion in deciding whether to impose a fine (§ 7(c)) and must particularly take into account the likelihood of significant hardship resulting from enforcement (§ 7(b)). Those issued a fine are also permitted several enumerated legal defenses (§§ 7(c)(2)–(5)), and the statute expressly prohibits bringing any sort of criminal prosecution to enforce the Act (§ 7(a)). Together, these restrictions on enforcement should create a situation in which the state’s power to compel adherence to the law is actually rather minimal, but its mere possibility still substantially increases turnout.

Additionally, when people are unable to vote because larger structures have prevented their access to the ballot—for example, a boss who refuses to allow employees time to vote—then it would be unjust to impose a penalty on the non-voters. For this reason, as I noted above,[69] the model statute establishes penalties on those who “intentionally and substantially restrict any person from fulfilling” their obligation to vote (§ 3(a)). It also incorporates several measures designed to increase access to the ballot box. These provisions allow the statute, where possible, to shift the penalty for non-voting onto those creating socioeconomic forces responsible for keeping potential voters away from the polls, rather than further penalizing those who have already been disenfranchised.[70]

Finally, the minor criminalization imposed by this statute could also prove to be self-correcting. Because a scheme of compulsory voting would have the effect of increasing the governmental power of marginalized persons, the government would likely become more responsive to their needs. In time, this would hopefully reduce the number of people who would be disproportionately impacted by enforcement of the law.

D.        Other Implementation Concerns

Many details of the model statute still need to be ironed out to ensure its effective implementation. For example, the model statute contains provisions that would make election days government holidays and provide a specific right of action against employers who punish employees for visiting the polls (§§ 6(b), (c), (e)). Further provisions ensuring access to the ballot box are necessary. Discussion of these provisions and potential provisions is beyond the bounds of this essay, except to note that ensuring that one has a substantial opportunity to easily fulfill the obligation to vote is another necessary measure to ensure that the statute criminalizes as little as possible. 


Voting is what ensures that the American government is composed “of the people, by the people, for the people.” But for that very same reason, the right to vote also implies a duty to vote. Low rates of voter turnout—and, in particular, disproportionately low rates along racial, income, and age lines—pose a major threat to the legitimacy of the government by undermining the public’s ability to make the government comport with its wishes. Although the VRA and recently proposed legislation have and would reduce voting discrimination and civic inequality, they are insufficient to address the full breadth of the problem. Adopting a compulsory voting scheme pursuant to Congress’s constitutional powers under the Fifteenth, Nineteenth, and Twenty-sixth Amendments, as well as Article I, Section 4, could considerably ameliorate these issues. Although significant concerns remain regarding criminalization under such a scheme, careful implementation, guided by the principle of avoiding hardship to those who are already most marginalized in American society, should be able to almost entirely eliminate the criminalization that would ensue from enforcing the model statute. Consequently, I believe that a careful system of compulsory voting represents the most effective way to reduce voting discrimination and civic inequality in this country.

Appendix: Model Statute

§ 1.      Definitions. In this Act, the following definitions shall apply:

(a).       “Election” means any official government poll, whether for the election of officials, amending state statutes or constitutions, or any other purpose.

(b).      “Employee” shall be construed to include both employees and independent contractors.

(c).       [. . .]

§ 2.      Congressional Findings. [Among the findings to be included here would be ones showing particular racial, age, and gender disparities in voting turnout that would be resolved by enacting and enforcing this statute.] 

§ 3.      Obligation to Vote. All persons in the United States shall be required by law to vote in every governmental election in which they are legally permitted to vote.

(a).       No government entity or agent shall restrict any person from fulfilling the legal obligations imposed by this section. No private entity or person shall intentionally and substantially restrict any person from fulfilling that obligation.

(1).      For the purposes of subsection (a), it shall not be considered a restriction to close polling places at a previously posted time of 7pm local time or later, except that every eligible person who enters or stands in line to enter a polling place prior to closure shall be permitted to vote.

(b).      Any person whose obligations are restricted in the manner forbidden by subsection (a) shall have a right of action against the person or entity who violated that subsection.

(c).       The Department of Justice shall also have a right of action against anyone who violates subsection (a).

(d).      In all suits under this section, a successful plaintiff may seek compensatory, nominal, and punitive damages, as well as attorney’s fees.

§ 4.      Right of Abstention. All ballots in elections affected by this Act shall have an abstention option.

(a).       Any person may submit to the Voting Compliance Agency a notice of intent not to vote. Submission of such a notice at least two weeks before any election will be accepted as an absolute excuse from the obligation to vote in that election.

(b).      Filing an intent not to vote does not place any restrictions on the filer’s ability to vote, nor shall it be construed under any circumstances to constitute a waiver of one’s right to vote. No government entity or agent shall in any way hinder a voter on the basis of a previously filed intent not to vote. 

§ 5.      Administration. This Act shall be administered by the Voting Compliance Agency (VCA), a new independent agency to be created upon passage of this Act. The VCA shall have the authority to issue regulations consistent with this Act to ensure its effective implementation.

(a).       The head of the VCA shall consist of a board of five commissioners, each of whom shall be appointed by the President for a term of ten years, and who may only be removed from office for negligence or malfeasance.

(b).      The terms of the commissioners shall be staggered, so that one commissioner is replaced or retained every two years.

(c).       [. . .]

§ 6.      Ensuring Access to the Ballot.

(a).       No voting precinct may contain more than 500 voters, unless that precinct conducts its elections exclusively through mail voting.

(b).      Any day on which a federal election occurs shall be a federal holiday. Any day on which a statewide vote occurs shall be a government holiday within that state. Any day on which a county or municipality election occurs shall be a government holiday within that voting district.

(c).       For the purposes of subsection (b), all federal elections shall also be considered to be statewide, county, and municipal elections. All statewide elections shall also be considered to be county and municipal elections within that state.

(d).      Municipalities, counties, and states shall hold elections for their respective governments on the same day as federal elections, unless there is a compelling government interest for doing otherwise.

(e).       No employer may fire, sanction, or otherwise punish any employee for taking time off to vote. Anyone who suffers a violation of this subsection—and notwithstanding subsection 3(a)—will have standing to bring suit against the violator. A successful plaintiff may be awarded compensatory, nominal, and punitive damages, as well as attorney’s fees. 

(f).       Except for precincts that conduct elections exclusively through mail voting, all elections shall be immediately preceded by at least seven consecutive days in which in-person early voting is allowed.

(g).      Voters shall be permitted to submit ballots via mail for any election.

(1).      Mail ballots shall be distributed at least two weeks ahead of any election.

(2).      Any ballot postmarked on or before the day of an election shall be counted.

(h).      [. . .]

§ 7.      Compliance.

(a).       Under no circumstances shall any criminal prosecution occur for a violation of this Act or any other state or municipal law instituting compulsory voting, nor shall failure to pay a fine levied under this Act constitute grounds for jailing or imprisonment.

(b).      In all acts of discretion related to the enforcement of this Act, the VCA shall particularly take into account the likelihood of substantial economic hardship resulting from enforcement.

(c).       Upon suspicion that an eligible voter has failed to vote in an election, the VCA shall have discretion to issue a fine to the suspected non-voter.

(1).      The amount of the fine shall be approximately 5% of the non-voter’s monthly income, subject to the VCA’s discretion. A lesser fine shall be issued if the standard fine would impose substantial economic hardship.

(2).      Upon being issued a fine, a defendant may establish by clear and convincing evidence that they either fulfilled their legal obligation to vote or notified the VCA of their intent not to vote pursuant to subsection 4(b).

(3).      Alternatively, a defendant may provide an excuse for not voting, such as mental or physical incapacity, moral opposition, interference from a government entity or other third party, or some other legitimate reason for not having voted. The VCA shall accept this excuse unless it finds either:

(A).     That the given excuse is invalid on its face; or

(B).      By clear and convincing evidence that the given excuse is false.

(4).      Upon acceptance of a defendant’s excuse or evidence of voting, the VCA shall dismiss the fine.

(5).      All decisions of the VCA related to fines may be appealed to a Federal District Court with geographic jurisdiction over the suspected non-voter.

(d).      [. . .]

§ 8.      Severability.

(a).       In the event that the Supreme Court finds § 2 of this Act unconstitutional for lack of congressional power to control state or local elections, it is the intention of Congress that the section shall be construed to apply only to federal elections.

(b).      In the event that the Supreme Court finds § 2 of this Act unconstitutional for lack of congressional power to control presidential elections, it is the intention of Congress that the section shall be construed to apply at the federal level only to congressional elections.

(c).       In the event that the Supreme Court finds § 2 of this Act unconstitutional for any other reason, or finds any other part of this Act unconstitutional, it is the intention of Congress that the unconstitutional provisions shall be severed from the Act. Thereafter, the remaining provisions of the Act shall still be enforced as written.

§9.       Effective Date.

(a).       Sections 3, 6, and 7 of this Act will become effective six months prior to the first federal election that takes place at least two calendar years following the passage of this Act into law. 

(b).      All other sections will become effective immediately upon passage.

[1] Brnovich v. Democratic Nat’l Comm., 141 S.Ct. 2321, 2350 (2021) (Kagan, J., dissenting). 

[2] Shelby Cnty., Ala. v. Holder, 570 U.S. 529 (2013).

[3] H.R. 1, 117th Cong. (2021).

[4] Id.

[5] Compulsory Voting, Int’l Inst. for Democracy & Electoral Assistance, https://www.idea.int/data-tools/data/voter-turnout/compulsory-voting(last visited May 4, 2022).

[6] Voting Rights Act of 1965 § 2, 52 U.S.C. § 10301. 

[7] Shelby Cnty., 570 U.S. 561 (Ginsburg, J., dissenting).

[8] Id. at 560.

[9] Brnovich v. Democratic Nat’l Comm., 141 S.Ct. 2321, 2354 (2021) (Kagan, J., dissenting).

[10] Id. (Kagan, J., dissenting) (discussing VRA’s legislative history).

[11] Shelby Cnty., 570 U.S. at 538 (“Among other prerequisites for bailout, jurisdictions and their subdivisions must not have used a forbidden test or device, failed to receive preclearance, or lost a § 2 suit, in the ten years prior to seeking bailout.”); see also 28 C.F.R. app. pt. 51 (containing a list of jurisdictions covered under the § 4(b) formula as of April 15, 2011).

[12] Shelby Cnty., 570 U.S. at 538.

[13] Id. at 565 (Ginsburg, J., dissenting).

[14] Id. at 557 (faulting Congress for failing to update the formula since 1975); but see id. at 576–94 (Ginsburg, J., dissenting) (providing a significant refutation of the claim that the formula was out of date).

[15] Id. at 557.

[16] See Brnovich v. Democratic Nat’l Comm., 141 S.Ct. 2321 (2021).

[17] Id. at 2339–40.

[18] See P. R. Lockhart, How Shelby County v. Holder Upended Voting Rights in America, Vox (June 25, 2019, 7:49 PM),  https://www.vox.com/policy-and-politics/2019/6/25/18701277/shelby-county-v-holder-anniversary-voting-rights-suppression-congress.

[19] Voting Laws Roundup: December 2021, Brennan Ctr. for Just. (Jan. 12, 2021),  https://www.brennancenter.org/our-work/research-reports/voting-laws-roundup-december-2021.

[20] Jurisdictions Previously Covered by Section 5, U.S. Dep’t Just. (Nov. 29, 2021), https://www.justice.gov/crt/jurisdictions-previously-covered-section-5.

[21] See Lisa Hill, Low Voter Turnout in the United States: Is Compulsory Voting a Viable Solution?, 18 J. Theoretical Pols. 207, 210–12 (Apr. 2006) (discussing many of the factors that lead to low turnout in the United States; among the more common factors are inability to get time off from work, difficulty in registering, low motivation, and complexities of federalism).

[22] 2020 Presidential Election Voting and Registration Tables Now Available, U.S. Census Bureau (Oct. 8, 2021),  https://www.census.gov/newsroom/press-releases/2021/2020-presidential-election-voting-and-registration-tables-now-available.html.

[23] Id.

[24] See Note, The Case for Compulsory Voting in the United States, 121 Harv. L. R. 591, 591 (2007).

[25] Id. at 594.

[26] Voting and Registration in the Election of November 2020, U.S. Census Bureau (Oct. 28, 2021),  https://www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-585.html

[27] Id.

[28] Id.

[29] See Hill, supra note 21, at 209 (“Perhaps the most important value that is undermined by high levels of non-voting is political equality.”).

[30] Voting and Registration in the Election of November 2012, U.S. Census Bureau (Oct. 8, 2021),  https://www.census.gov/data/tables/2012/demo/voting-and-registration/p20-568.html.

[31] If the Black turnout had increased at the same rate as the overall vote between 2012 and 2020, then we would expect the Black turnout in 2020 to be about 71.6%. Comparing the actual Black turnout in 2012 (62.4%) with the actual Black turnout in 2020 (66.2%) yields a discrepancy of 3.6 points, whereas comparing the actual Black 2012 turnout to the expected Black 2020 turnout yields a discrepancy of 9 points.

[32] The John Lewis Act would update the coverage formula of § 4(b) of the VRA to cover, inter alia, states in which “fifteen or more voting rights violations occurred . . . during the previous 25 calendar years.” H.R. 4, 117th Cong. § 5(b)(1)(A)(i) (2021).

[33] The Court explicitly stated that “Congress may draft another formula based on current conditions.”  Shelby Cnty., Ala. v. Holder, 570 U.S. 529, 557 (2013).

[34] Id. at 557–59 (Thomas, J., concurring).

[35] I use “person,” rather than “citizen,” because some cities—such as New York—allow non-citizens to vote in municipal elections.

[36] Jill Sheppard, Compulsory Voting and Political Knowledge: Testing a ‘Compelled Engagement’ Hypothesis, 40 Electoral Stud. 300, 301 (2015).

[37] Case for Compulsory Votingsupra note 24, at 601; see also Sheppard, supra note 36, at 301.

[38] Arend Lijphart, Unequal Participation: Democracy’s Unresolved Dilemma, 91 Am. Pol. Sci. Rev. 1, 8–9 (1997).

[39] Case for Compulsory Votingsupra note 24, at 592; see also Lijphart, supra note 38, at 9 (discussing the drop in turnout from 90.2% to 60.2%—a rate roughly similar to U.S. voter turnout—after Venezuela repealed its compulsory voting law).

[40] Sheppard, supra note 36, at 301.

[41] Shane P. Singh, Compulsory Voting and the Turnout Decision Calculus, 63 Pol. Stud. 548, 565 (2015).

[42] Details concerning rights of action may be found in §§ 3(b)–(d). The different standard for private actors is intended to avoid liability for minor and unintentional restrictions that might occur in everyday life. 

Here it should be noted that H.R. 1 and the model statute are complementary. If H.R. 1 were to be enacted, the model statute would gain strength from the bevy of requirements protecting access to the ballot that the For the People Act would create. Any violation of H.R. 1’s requirements would also violate § 3 of the model statute by placing a restriction on fulfilling the obligation to vote. See also infra text accompanying note 70.

[43]  Shelby Cnty., Ala. v. Holder, 570 U.S. 529, 542 (2013).

[44] Singer v. United States, 380 U.S. 24, 34–35 (1965).

[45] Id. at 35 (citing United States v. Kobli, 172 F.2d 919, 924 (3d Cir. 1949)).

[46] Id. at 36 (holding that a defendant’s ability to waive the right to jury trial is conditional on the prosecutor and the trial judge being willing to accept a plea deal).

[47] Barrentine v. Ark.-Best Freight Sys., Inc., 450 U.S. 728, 740 (1981).

[48] Id. (citation omitted).

[49] Case for Compulsory Votingsupra note 24, at 600.

[50] See W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

[51] Id. at 634 (“[A] Bill of Rights which guards the individual’s right to speak his own mind, [does not] le[ave] it open to public authorities to compel him to utter what is not in his mind.”).

[52] U.S. Const. art. I, § 4.

[53] Id. art. II, § 1, cl. 2. Congress’s only Article II power over presidential elections is the power to set the time and date of the election. Id. cl. 4.

[54]  U.S. Const. amend. XV, § 2; id. amends. XIX, XXVI, § 2.

[55] I did not discuss gender disparities in Part I because women vote at a higher rate than men; this has been the case for decades. Participation in Congressional Elections by Sex Since 1978, U.S. Census Bureau (Oct. 8, 2021), https://www.census.gov/library/visualizations/2020/comm/participation-congress-election.html. Voting turnout is thus not a pressing site of structural gender discrimination. However, the Supreme Court has held that laws discriminating against men are unconstitutional. See Craig v. Boren, 429 U.S. 190 (1976). Thus, the fact that men vote at a lower rate may also constitute a problem that is within Congress’s Nineteenth Amendment purview to correct.

[56] See supra Part I.

[57] Note that many of these disparities may be due to disparate impact rather than discriminatory intent. However, that should not pose a problem. Although the Constitution itself often forbids only government action motivated by discriminatory intent, see Washington v. Davis, 426 U.S. 229, 239 (1976), Congress has the power to adopt legislation that uses disparate impact to determine when discrimination has occurred. The Supreme Court has expressively approved of legislation, such as certain portions of Title VII, 42 U.S.C.A. § 2000e-2, that use disparate impact in this manner. See, e.g.,Griggs v. Duke Power Co., 401 U.S. 424, 430 (1971) (unanimously holding that Title VII forbids “practices, procedures, or tests neutral on their face, and even neutral in terms of intent” that nevertheless discriminate).

[58] Congress may also be able to address inequality along other lines through the Fourteenth Amendment, although the Court’s doctrine of suspect classes may limit its power to do so.

[59] Nw. Austin Mun. Util. Dist. No. One v. Holder, 557 U.S. 193, 205 (2009).

[60] Shelby Cnty. v. Holder, 570 U.S. 529, 544 (2013) (internal quotation marks and emphasis omitted).

[61] U.S. Const. art. I, § 4.

[62] Seila L. LLC v. Consumer Fin. Prot. Bureau, 140 S.Ct. 2183, 2209 (U.S. 2020) (“When Congress has expressly provided a severability clause . . . [w]e will presume ‘that Congress did not intend the validity of the statute in question to depend on the validity of the constitutionally offensive provision[.]’”) (citation omitted).

[63] See generally Reuben Jonathan Miller & Amanda Alexander, The Price of Carceral Citizenship: Punishment, Surveillance, and Social Welfare Policy in an Age of Carceral Expansion, 21 Mich. J. Race & L. 291 (2016).

[64] Case for Compulsory Votingsupra note 24, at 611.

[65] Id.

[66] Id.

[67] See, e.g., Marla McDaniel, Michael Karpman, Genevieve M. Kenney, Heather Hahn, & Eleanor Pratt, Urb. Inst., Customer Service Experiences and Enrollment Difficulties Vary Widely across Safety Net Programs (Jan. 2023), https://www.urban.org/sites/default/files/2023-01/Customer%20Service%20Experiences%20and%20Enrollment%20Difficulties%20Vary%20Widely%20across%20Safety%20Net%20Programs.pdf.

[68] Id.

[69] See supra note 42.

[70] The model statute also allows a suspected non-voter to sue others who restrict their access to the ballot (§ 3(b)) and to present evidence of being kept from the polls as an excuse against a levied fine (§ 7(c)(3)).