Advancing Criminal Reform Through Ballot Initiatives

Nicholas Ansel^*

Full Article.

I. Introduction

2020 brought the world’s attention to bear on American policing and the American carceral regime. After the murder of George Floyd on May 25, 2020, protests erupted in every state in America.^[1] By early July, experts estimated that “about 15 million to 26 million people in the United States ha[d] participated in demonstrations,” representing “the largest movement in the country’s history.”^[2] As of mid-October, Black Lives Matter demonstrations still continue in many cities.^[3]

America now cages more people than any other country in the history of Earth, incarcerating African Americans “at a rate six times that of South Africa during Apartheid.”^[4] The murder of George Floyd represented a crystallization of the American criminal project’s treatment of black and brown communities, of marginalized communities, of indigent persons.

While the general contours of this crisis are consistent nationally, Americans now know that the proposed remedies must be diffuse. Policing is scattered among over 18,000 law enforcement agencies, many governed by countless local and municipal governments.^[5]Likewise, criminalization and incarceration are almost entirely accomplished at the hands of state governments, each having established unique criminal codes and policies.^[6]

So how should those involved in the criminal reform movement approach changing the punishment bureaucracy? If one were to find herself engrossed in the relevant academic literature, she might learn to form broad coalitions, lobby political bodies, strategically vote for representatives, engage media in her cause, and fund more academic studies and expertise in criminal policy.

While all this work has been invaluable, this formula omits a crucial—perhaps the crucial—vehicle for criminal reform in America. If, instead, reformers were to focus on how bold reform policies are being implemented, they may end up turning their attention somewhere else: direct democracy.

In many states, fundamental reforms have been driven in a patently American way—directly by the people. Particularly over the last decade, reformers have placed a growing number of policies directly on the ballot with compelling rates of success.[7] As it turns out, placing trust in the people has been an extremely successful way to implement changes like policing reform, drug decriminalization, search and seizure protections, felon enfranchisement, jury unanimity, and bail reform.[8] Coalition-building is certainly constitutive of transformative change, but maybe it is time that reformers appreciate the success their movement has already had in the minds of everyday Americans.

This Comment rebuts the traditional narrative that leaving criminal policy in the hands of average citizens necessarily results in an overly punitive system.^[9] It argues that direct democracy is a critical component of criminal reform because it serves as an effective vehicle for passing bold reform policies. Unique advantages inhere in direct democracy: it places reformers at the helm of policymaking without having to negotiate or dilute their goals, it bypasses the ruling bureaucrats who shaped mass incarceration and are thus unlikely to truly reform it, and it puts reform policies directly before the actual people who are most affected by the lived experience of mass incarceration. And, to borrow the popular idiom, direct democracy provides the movement with a greater ability to strike while the iron is hot.

The Comment follows several lines of investigation to support its core assertion. Part II focuses on direct democracy, examining the process of getting a measure to the ballot, including the express and unspoken requirements for ensuring an initiative makes its way to the voters. It then reviews a series of restrictions state legislatures have placed on the initiative process. Part III examines criminal-focused ballot initiatives that have been proposed since 2000, finding that these measures are passing with increased frequency and cover a broad range of criminal injustices, while punitive‑focused pro-criminalization measures are becoming more limited in scope. Finally, Part IV discusses how criminal reform is well-suited for direct democracy because it embraces its theoretical underpinnings while dispelling its traditional critiques.

II. Understanding Direct Democracy

Before discussing direct democracy in the criminal context, this Comment analyzes direct democracy conceptually, as well as how it formally operates in modern society. This Part treats that material by dividing it into two Sections. Section A briefly discusses the foundational theory of direct democracy in America before examining its several forms and how a measure actually gets onto the ballot. Section B focuses on how state legislatures have sought to restrict direct democracy and undermine some enacted measures. Each Section represents an important piece in analyzing whether direct democracy is a suitable mechanism for criminal reform.

A.    The Process

American direct democracy largely developed at the turn of the twentieth century in the American West, spurred by the Progressive movement and the enormous power exerted by corporate interest groups on state legislative bodies.^[10] Direct democracy originated and took shape in America primarily as a tool “for the majority to uproot powerful corporate interests.”^[11] One might view direct democracy as an attempt to trifurcate state legislative bodies, serving as an additional check on representative legislative power.

There are several forms of direct democracy in America. The most common and widely understood form is the initiative, whereby citizens propose a law or a constitutional amendment.^[12] There are two forms of initiatives: direct and indirect.^[13] A direct initiative sends a qualified proposal directly to the ballot for the voters to decide.^[14] An indirect initiative is first sent to the legislature, which has an opportunity to act on that measure.^[15] If the legislature fails to act, revises the proposal, or rejects it, the initiative then goes directly to the ballot for the voters to decide.^[16]

Another form of direct democracy is the referendum, whereby voters may repeal a law already enacted by the legislature.^[17] Finally, legislative referrals (otherwise known as legislative measures or legislative propositions) involve the legislature placing a proposal onto the ballot for a public vote.^[18]

Today, twenty-six states and the District of Columbia have some form of the initiative or referendum processes for statewide elections.^[19]The sheer quantity of states that employ direct democracy fails to adequately capture its widespread effect in America. While used by only half of the states, “over 70 percent of the population now lives in either a state or city where these tools of direct democracy are available.”^[20]

When a group of citizens decide to propose an initiative or referendum, it must follow several procedural requirements for their proposed measure to qualify for the ballot.^[21] The most critical requirement is gathering enough signatures.^[22] While the percentages vary from state to state, usually signatures amounting to between 3% and 15% of the votes cast in the state’s previous gubernatorial election are required.^[23]The signature requirement ensures that measures have adequate public support.^[24]

Gathering enough signatures can be rather difficult, especially depending on the size of the state in which a ballot measure is sought.^[25] In order to facilitate this process, organizers generally form a campaign^[26] that then hires petition circulators to collect signatures. Since the inception of direct democracy, paid circulators have been a staple of these campaigns.[27] In the early twentieth century, there were already frequent critiques of petition peddlers.^[28] Several states even banned the use of paid circulators, until those bans were ruled unconstitutional in 1988.^[29] Today, a vast subculture of professional circulators exists to support campaigns.^[30]

States have implemented varying requirements for these circulators, including: in-state residency, prohibiting payment by signature collected, registration with the state, and swearing by affidavit as to the validity of the signatures they collect.^[31] Because signatures can be invalidated for many reasons—including the lack of a matching signature, signatures by non‑residents, or unverified signatures—far more signatures must be collected than the concrete requirements prescribed by each state, and successful ballot campaigns routinely collect far more than the threshold requirements.^[32]

Beyond the organizational tasks, campaigns must create the substance of the initiative.[33] A title need be given to the ballot measure for both the petition phase—where signatures are collected—and the ballot phase—where voting occurs.^[34] A summary of the initiative, usually subject to a word limit, needs to be drafted for both petitions and the ballot.^[35] Since most voters will never read the actual legislation, the results of these drafting processes are instrumental in the voting process.^[36]

Finally, the actual legislation must be drafted. This is, of course, only true when citizens are proposing an initiative, not a referendum.[37]There has been considerable scholarship devoted to critiquing the initiative process for lacking the deliberative qualities of formal governing bodies.^[38] Today, though, this process is hardly haphazard. Successful ballot campaigns are typically comprised of robust collectives that are able to focus their attention on drafting a singular piece of legislation.[39] It is extremely common for these campaigns to hire experienced teams of drafters comprised of lawyers and even former legislators, or to hire outside firms that specialize in the initiative drafting process.^[40]And, where a campaign lacks those resources, many states provide resources like the services of a legislative drafting body or a review period during which the campaign may adopt any recommended changes.^[41]

Ballot campaigns must also be keenly aware of potential legal challenges to procedural deficiencies in their ballot measures. It is common for measures to be challenged for including vague or deceptive language,^[42] failing to gather enough signatures or follow filing requirements,^[43] or not adhering to state‑specific guidelines like single-issue requirements.^[44] If vague language exists, instead of decertifying measures, courts can intervene to re-draft the ballot language in a less confusing manner or order state officials to do so.^[45]

However, once the voters have approved a measure, it is the state that is tasked with defending any substantive legal challenges to the enacted measure, not the ballot campaign.^[46] Once the voters enact or reject a measure, the sponsors’ role in the initiative process is over.[47]

B.    Legislative Resistance to Direct Democracy

Generally, legislators aren’t quite keen on the voters bypassing their authority. Accordingly, several states have recently enacted restrictive requirements for ballot initiatives resulting in less initiatives finding their way onto the ballot.[48] From 2006 to 2015, the number of measures that found their way onto the ballot consistently declined.^[49] For a variety of reasons (i.e. low turnout in 2014 and a perception of legislative inaction), 2016 saw a significant increase in citizen initiatives.^[50] That increase prompted legislatures across the country to pass more restrictive policies in order to curb the reach of direct democracy.^[51] According to the Ballot Initiative Strategy Center, “[Ten] states adopted legislation to make it more difficult to put citizen-led measures on the ballot” between 2017 and 2018.^[52]

Some states have worked to make it more difficult to gather the requisite signatures for ballot measures by creating geographic distribution requirements. In Michigan, the Legislature passed a bill capping signatures collected from a single district at 15% so that organizers must evenly distribute their signature collection efforts across the state.^[53] Idaho passed a similar but more expansive bill requiring “signatures from 10 percent of registered voters in 32 of Idaho’s 35 districts,” and reducing the timeframe for gathering signatures from eighteen to six months.^[54] Legislators in Arizona also passed a geographic distribution requirement, setting signature thresholds for all thirty districts.^[55]Other states have introduced bills seeking to raise the bar for passing initiatives. In Missouri and Florida, bills were introduced raising that threshold to 66% for voters to approve any initiative.^[56]

One of the most restrictive and strange requirements, though, was passed prior to the 2016 initiative resurgence by the Arizona Legislature. A 2014 bill added a “strikeout law” stipulating that all circulators can be subpoenaed to appear in court and verify the signatures they collected.^[57] Oftentimes circulators are from out-of-state, so if they are unable to get to the court hearing, the signatures they collected are invalidated.[58] In recent litigation surrounding a clean energy initiative, 1,180 petition circulators were subpoenaed to court to verify the signatures they collected.^[59] In what the Arizona Supreme Court called “[a] five-day trial of extraordinary logistical complexity,” the court was only able to call fifty of those circulators to testify.^[60] Nevertheless, courts have upheld this restriction on the initiative process.^[61]

In 2017, the Arizona Legislature passed House Bill 2404, which added a new spate of ballot-measure requirements.^[62] Now, ballot measures must strictly comply with all statutory regulations, a much more rigid standard than the prior substantial compliance.^[63] In other words, ballot measures must follow to the tee every statutory requirement or face decertification.

States are not just resisting the initiative process by imposing strict procedural requirements—they are also directly fighting enacted initiatives by passing legislation to undermine or dampen their effects. While twenty‑six states permit ballot initiatives, eleven of those states also grant their legislatures the authority to overturn them.^[64] In fact, from 2017 to 2018, over one hundred bills were introduced across the country attempting to reverse ballot initiatives.^[65] Among those hostile states is Maine, where the Legislature has fought to weaken progressive initiatives passed by voters in 2016.^[66] In fact, the highest number of legislatively altered bills over the last decade occurred in 2016, when fourteen initiatives were subsequently altered by state legislatures.^[67] One high-profile example is the Florida Legislature’s distortion of the felon enfranchisement initiative.^[68] Only two states, California and Arizona, require voter approval prior to any legislative alteration of voter‑enacted initiatives.^[69]

The key takeaway regarding legislative alteration is that not all ballot measures are created equal. Some measures may have more staying power, depending on whether and to what extent they are insulated from legislative interference. Nevertheless, this feature of direct democracy paints a fascinating picture of the legislative dialogue that occurs between citizens and elected representatives.

III. Criminal Reform on the Ballot

In order to analyze the effectiveness of criminal reform ballot measures, this Comment produces an empirical study by creating a compendium of all ballot measures that predominantly involve criminal law since 2000.^[70] Section A parses the data on all ballot measures since 2000 with significant criminal implications, finding that criminal reform ballot measures have been extremely successful nationwide, especially recently. These measures cut across an array of issues affecting criminal law and mass incarceration. Section B investigates pro-criminalization measures since 2000, culling data from every ballot measure to show that they are appearing with less frequency and represent a shift in punishment theory.

A.    Criminal Reform Measures

Twelve years ago in an article deriding direct democracy, Professor Erwin Chemerinsky made the following remark: “[W]hen was the last time that the voters passed an initiative to increase the rights of prisoners, or increase the rights of criminal defendants?”^[71] Since the publication of that article, the world has changed; voters have come to the ballot time and again to pass criminal reform policies.

Historically, several of the earliest ballot initiatives were attempts to enact criminal reform policies—namely abolishing the death penalty. The early twentieth century saw a wave of abolitionist sentiment fueled by progressivism that led to many states abolishing the death penalty.^[72] Most of those states had very small populations of non-white persons, making them “conducive to lenient or less severe criminal penalties.”^[73] Arizona’s first Governor, George W.P. Hunt, led the charge of abolition in Arizona.^[74] He met fierce opposition from the state Legislature, which favored capital punishment and tried to pass laws circumscribing his powers.^[75] Unable to achieve abolition through the Legislature, a citizen initiative was passed abolishing the death penalty in Arizona in 1916 by a mere 152 votes.^[76] After several heinous crimes were widely covered by the newspapers in 1917,^[77] Arizona reinstated the death penalty by legislatively referred initiative just two years later in a resounding vote.^[78]

The history of ballot initiatives concerning the death penalty has been extremely unfavorable to reformers. In the twenty-eight ballot initiatives from 1912 to 2016, voters cast ballots in favor of the death penalty 89% of the time.^[79] The only two initiatives that entirely abolished the death penalty, in Arizona (1916) and Oregon (1914), were overturned by subsequent ballot initiatives (Arizona in 1918 and Oregon in 1920).^[80]

Death penalty reform has failed on the ballot in recent years as well. California voters rejected abolishing the death penalty twice, in 2012 and 2016, and instead passed a ballot initiative requiring a five-year period for post-conviction challenges.^[81] Voters in Nebraska and Oklahoma also came out in favor of the death penalty in 2016.^[82]

With respect to all other criminal reform initiatives and referenda, reformers have had increasing success over the past two decades. These measures have covered a startling array of reforms including the legalization of marijuana, the decriminalization of drug convictions, policing reform, property forfeiture, DNA evidence admissibility, jury unanimity, felon voting rights, forced inmate labor, three-strikes laws, search protections for electronic devices, bail reform, addiction rehabilitation funding, decriminalization of assisted suicide, and gambling.^[83]

A significant proportion of these measures, 47 of 103, were related to medical marijuana, decriminalization of marijuana possession, or marijuana legalization.^[84] These measures, while at surface level seem to only address a tangential criminal reform issue, are actually addressing one of the most expansive overcriminalization issues in America—just last year over 600,000 people were arrested for simple marijuana possession.^[85]

Some changes were led by state legislatures turning to the ballot. From 2000 to 2009, voters in Virginia approved permitting the state supreme court to consider exculpatory DNA evidence post-conviction, Pennsylvania voters expanded the right to confront witnesses, and New York voters approved allowing prisoners to perform volunteer work for nonprofit organizations.^[86] And in Rhode Island, voters approved the restoration of voting rights upon release.^[87]

In that same decade, citizens were able to pass only a few criminal reform initiatives. Measure 3 in Oregon and Initiative B in Utah both restricted criminal property forfeiture, and Proposition 36 in California required probation and treatment for possession of certain controlled substances.^[88] All three measures passed in 2000.^[89]

Few of the criminal reform measures, outside of marijuana measures, met success from 2000 to 2009. Reform ballot measures failed across the country, from mandatory minimums to funding treatment programs to changing three‑strikes laws.^[90] Still, when accounting for all criminal reform measures, 48% of criminal reform measures that were proposed passed between 2000 and 2009.^[91]

That history makes the story of criminal reforms on the ballot over the last decade astonishing. From 2010 to 2020, 81% of reform measures passed when put before the voters.^[92] Several of these measures have come by way of legislative referral. Colorado removed a constitutional provision permitting forced inmate labor, Missouri passed an initiative protecting electronic communications and data from unreasonable searches and seizures, and New Mexico and New Jersey passed bail reform, all through legislative referrals.^[93] It was Louisiana’s Amendment 2, though, that garnered the most public attention. Louisiana permitted non-unanimous jury convictions until the people voted by a two-thirds margin to require unanimous juries in 2018.^[94] Passage of these measures suggests that sometimes even formalistic reform issues are neither too foreign nor too complicated for the voters.

Voters also came to the polls over the last decade to support gambling. New Jersey’s Public Question 1 (2011) and Colorado’s Proposition DD (2019) legalized sports betting.[95] Both measures were legislatively referred.[96] Arkansas passed Issue 4—legalizing gambling—in 2018 by a citizen‑initiated measure.^[97]

The breadth of the criminal reform measures passed since 2010 through the initiative is hard to overstate. What’s more, the incidence of citizens resorting to ballot initiatives to effectuate reforms has been steadily increasing alongside each ballot cycle.^[98] For example, in 2012, California voters enacted a measure “modifying the state’s three strikes law to require a life sentence only when a defendant’s current conviction is for a violent or serious offense”; the initiative also permitted retroactive resentencing of those already incarcerated.^[99] That same year, Washington, Colorado, and Massachusetts each passed marijuana initiatives.^[100] Then in 2014, California passed Proposition 47, reclassifying most nonviolent property and drug crimes as misdemeanors, and Alaska, Oregon, and Washington, D.C., passed marijuana initiatives.^[101]

2016 was a banner year for criminal reform initiatives. California passed Proposition 57, which increased parole and good-behavior opportunities for felons convicted of nonviolent crimes and “required [juvenile courts] to hold hearings before transferring a child to the adult court.”^[102] Oklahoma passed two initiatives. The first reclassified certain property offenses and simple drug possession as misdemeanor crimes, and the second used the money saved by reclassifying those offenses to fund rehabilitative programs.^[103] In 2016 alone, marijuana initiatives passed in Arkansas, California, Florida, Maine, Massachusetts, Montana, Nevada, and North Dakota.^[104]

In 2018, more ballot initiatives were aimed at bold criminal reform. Louisiana voted to require unanimous juries; Florida restored the voting rights of felons and expanded the retroactive force of criminal rights; and Washington passed an initiative creating a good-faith test in reviewing police use of deadly force, requiring de-escalation and mental health training, and requiring law enforcement officers to provide first aid.^[105] In addition, Michigan, Missouri, Oklahoma, and Utah passed marijuana initiatives.^[106]

Finally, in 2020, twenty-one decriminalization measures were placed on ballots, and nineteen passed.^[107] Four more states—Arizona, Montana, New Jersey, and South Dakota—legalized marijuana.^[108] Mississippi joined as well, adopting an expansive medical marijuana program.^[109] Voters in Mississippi rejected the legislative alternative to the marijuana initiative, which would have limited medical use to terminal illnesses.^[110] And drug decriminalization reached new terrain in 2020: the District of Columbia voted to deprioritize law enforcement for entheogenic plants, and Oregon fully decriminalized possession of controlled substances, opting instead to treat addiction and substance abuse as public health matters requiring treatment, not criminalization.^[111] The success of Oregon’s approach, shepherded forward by voters, may forecast a sea change in America’s drug criminalization regime.

Of course, 2020 measures also focused on issues beyond drug decriminalization. Michigan voters passed—with 88.75% support—a requirement to obtain a search warrant prior to searching an individual’s electronic data and communications.^[112] California voters restored the voting rights of all persons not presently incarcerated.^[113] One of the only two measures to fail, California’s Proposition 25, paints a complicated picture of how voters are approaching nuanced issues of criminal reform.^[114] The Legislature adopted a measure to replace cash bail with a risk assessment system in an effort to reduce the bail system’s outsized influence on indigent defendants.^[115] Then the American Bail Coalition referred that legislation to the ballot, where it was rejected.[116] At the same time, progressive organizations—like the ACLU and the Human Rights Watch—campaigned against the risk assessment alternative arguing that it relies on racially biased algorithms and can lead to even more pretrial detention.^[117] So even though California voters came out in overwhelming favor of other criminal reform measures, with this Proposition those same voters did the hard work that the state Legislature was unwilling to do, assessing the nuances of complicated reforms and public advocacy, and deciding that cash bail needed a better solution.^[118]

Though not the focus of this Comment, local reform measures in 2020 pushed the boundaries of possible change. Among numerous successful measures, Philadelphia voters banned stop and frisk, Measure J in Los Angeles redirected 10% of county funds toward community investment, San Francisco voters eliminated a minimum threshold on the number of police officers, and many cities created and expanded police oversight boards.^[119] Almost all of these local measures passed with at least 75% support.^[120]

The proverbial cherry on top, however, is that only four statewide criminal reform ballot measures—not relating to marijuana—have failed since 2011.^[121] So while several reform measures were rejected in the preceding decade, these measures have been much more effective over the last decade.

Overall, 71% of measures proposing criminal reforms have passed since 2000.^[122] When accounting for iterative measures—those denied within a state in one election but later passed by a subsequent measure—the numbers are starker: 85% of criminal reform measures passed on the ballot.^[123] And, as discussed above, the rate of adoption of reform measures is increasing. These increases are particularly noteworthy in light of the overall reduction in state ballot measures over the last decade—there were 15% fewer ballot measures after 2010 than there were in the preceding decade.[124] Perhaps the most interesting data concerning criminal ballot measures are where reform measures originate: 73% were citizen-initiated.^[125]

In aggregate, the data illustrate why ballot measures are such effective mechanisms: criminal reform initiatives are passed at high rates, and they are appearing with more frequency on ballots, despite a reduction in the initiative process writ large.^[126] As an added bonus, courts typically broadly construe voter-enacted criminal reforms.^[127] The sheer rate at which citizens are bringing these issues to the ballot, particularly in recent years, reveals the transformative potential of initiatives in bringing about criminal reform.

B.    Pro-Criminalization Measures

The case for criminal reform on the ballot becomes stronger still by analyzing trends in pro-criminalization measures. While criminal reform measures over the past two decades have run the gamut, addressing a broad variety of reforms, pro-criminalization measures have waned in the past decade and have modestly shifted their focus to more informed theories of punishment.

Pro-criminalization measures have historically met overwhelming success. It would be remiss not to acknowledge how ballot measures contributed to our carceral state. Mass incarceration was at least partially fueled by California’s three-strikes law, passed by the citizen-initiated Proposition 184 in 1994^[128]—not to mention the spate of tough-on-crime initiatives that emanated from public sentiment and political fervor in the 1980s and 1990s.^[129] At the same time, ballot initiatives are far from solely responsible for our current carceral state.

Still, even since 2000, 82% of pro-criminalization measures have passed when put to the voters.^[130] Prior to 2010, a wide range of pro-criminalization measures passed on the ballot. Several of these measures dealt with expanding sentences for offenders:

• Proposition 18 in California, expanding the special circumstances that can aggravate a first-degree murder conviction, and mandating that any such conviction will result in a sentence of death or of life in prison;
• Proposition 21 in California, expanding the crimes for which juveniles must appear in adult court, increasing certain sentences, and reducing case confidentiality;
• Proposition 301 in Arizona, permitting those convicted of methamphetamine use or possession to be sentenced to jail or prisons terms; and
• Measure 57 in Oregon, increasing sentences for certain repeat property crimes, including theft against the elderly.^[131]

A second category of these measures established new crimes: animal cruelty and cockfighting, payment to ballot canvassers by signatures collected, and barring unmarried cohabitants from providing foster care.^[132]

Third, a number of measures prior to 2010 created significant procedural hurdles for criminal defendants. Several of these measures regarded denying bail to categories of defendants: those charged with sexual assault crimes (Proposition 103 in Arizona), those undocumented persons who had been charged with enumerated “serious” offenses (Proposition 100 in Arizona), those who had violated a condition of release (Proposition 4 in Texas), and those who had violated certain court orders (Proposition 13 in Texas).^[133] Two Hawaii ballot measures were passed that permitted the initiation of a felony criminal charge by written information (Question 3 and Question 4).^[134] Other measures dealt with expanding the grounds and uses of civil forfeiture (Measure 12 in Louisiana and Measure 53 in Oregon) and requiring DNA collection for certain offenses (Proposition 69 in California).^[135] In addition, states enacted measures that revised procedural protections for criminal defendants, including:

• Question 1 in Indiana, requiring that criminal appeals from life sentences or extremely long terms must follow the same path through the court system as civil appeals;
• Amendment 2 in Pennsylvania, allowing the Legislature to enact laws circumscribing the testimony of children in criminal proceedings;
• Proposition 7 in Texas, reducing the number of jurors in misdemeanor cases to six; and
• Question 3 in Hawaii, rendering privileged communications between a crime victim and a health professional inadmissible.^[136]

Fourth, states enacted a series of more generalized measures leveling miscellaneous infirmities at those accused or convicted of crimes. For example, Massachusetts passed Question 2, limiting the voting rights of felons while they are incarcerated; Washington passed SJR 8212 authorizing state-operated inmate labor programs by private entities; and California and Oregon passed victims’ rights initiatives.^[137] All of these measures, passed between 2000 and 2009, represent the far-reaching pro-criminalization ethos endorsed by voters during that decade.

However, after 2010, that ethos seemed to subside as fewer pro‑criminalization measures made it to the ballot and passed. Between 2000 and 2009, thirty-seven pro-criminalization measures appeared on ballots; that number decreased to thirty-four between 2010 and 2020.^[138]

The measures that did pass shifted in focus from a holistic punitive ideology to a more informed theory of criminality. Take, for example, the few particularly substantive pro-criminalization measures that passed this decade:

• Initiative 1501 in Washington, increasing criminal identity‑theft penalties;
• Amendment 2 in Alabama, recognizing rights of unborn children;
• Measure 73 in Oregon, requiring an increased minimum sentence for certain sex crimes and repeat DUI offenses;
• Proposition 35 in California, increasing prison terms for human traffickers, requiring them to register as sex offenders, and requiring sex offenders to disclose internet browsing history; and
• Amendment 2 in Missouri, permitting the admissibility of relevant evidence of prior criminal acts in cases of sex crimes committed against minors.^[139]

While several of these enactments are grossly punitive, these measures (with the exception of Alabama’s Amendment 2)^[140] show a concerted shift in focus toward more harmful societal ills with widespread effects, such as sex crimes, human trafficking, DUIs, and identity theft. Other laws that passed in this decade also evidence this public shift, including laws aimed at curbing animal cruelty and the misconduct of public officials.^[141]

A much narrower range of procedural pro-criminalization measures have passed since 2010. HJR 4220 in Washington was a response to highly publicized shootings of officers; it authorized the denial of bail only in cases where the public is at risk.^[142] In addition, measures in Louisiana and North Carolina involved the circumstances in which defendants may waive their right to a jury trial.^[143] Each was legislatively referred and passed prior to 2015.[144]

The bulk (54%) of pro-criminalization measures passed since 2010 were Marsy’s Law Initiatives.^[145] These measures, spurred by the eponymous measure California passed in 2008, were aimed at expanding the rights of crime victims in the criminal adjudicatory process. While each was uniquely tailored, they generally focused on expanding the role of victims throughout prosecution, conviction, and postconviction, and changing the process of parole hearings and parole revocation.^[146] Marsy’s Laws represent the largest block of pro-criminalization measures over the last decade.^[147]

What can be easily overlooked in the pure statistical analysis of pro‑criminalization bills is that, when reviewing the actual legislation, the measures represent a modest victory in the fight for criminal reform. While many measures between 1980 and 2009 contributed to the atmosphere of overcriminalization, since 2010 the ideology undergirding that movement has waned. Almost 60% of these measures were legislatively referred;^[148] initiatives hardly even press pro-criminalization issues anymore. Recently passed measures shifted their attention to represent a reshaping of what we ought to punish, including animal rights, #metoo advocacy, and public accountability.[149] Courts might be shifting too; six pro-criminalization bills were invalidated by courts over the last decade.^[150] The bulk of pro‑criminalization ethos that survived this past decade is accounted for by victims’ rights measures.

Taking each Section of Part III together, reformers should be encouraged that reform measures pass at an exceedingly high rate and cover a large swath of issues in criminal law. What’s more, measures to enact criminal reform are gaining traction. In the decade spanning 2000–2009, 46% of criminal-related ballot measures were decriminalization reform measures.^[151] Since 2010, 68% of ballot measures regarding criminalization were reform efforts.^[152]

Meanwhile, pro-criminalization measures have generally shifted toward a more progressive view of criminalization, shedding some of the gross excises of the tough-on-crime sentiment that fueled prior decades. Nevertheless, reformers must always vigilantly guard against the novel and pernicious ways that twenty-first century pro-criminalization efforts may take shape.

IV. Why Direct Democracy Is a Strong Vehicle for Criminal Reform

This Comment seeks not to resolve the intractable debate over the merits of ballot initiatives generally. Instead, it makes the claim that direct democracy is uniquely situated for advancing criminal reform. Specifically, criminal reform measures are necessary because: (1) criminalization is a deeply personal issue, which demands direct advocacy by affected persons; (2) criminal reform inherently advances minority rights; (3) broad public support exists for reform; (4) legislative action has been insufficient to tackle the criminalization crisis; and (5) initiatives can serve as bellwethers for bold policy. The biggest hurdle facing initiatives is (6) funding.

A.    Voter Competence

Are citizens capable of legislatively redressing criminal injustice? Critics of citizen-led legislation decry the lack of a sufficiently deliberative process.^[153] Proponents of direct democracy dispute that account,^[154] arguing that voters are empowered by the transformative informational access of modern technology,^[155] and that the voters themselves are neither more nor less deliberative than standard legislators.

When it comes to crafting legislation, scores of theorists have derided the initiative process for lacking the institutional safeguards of traditional lawmaking.^[156] In reality, modern initiatives are often drafted by specialized attorneys, outside law firms, and former legislators who have a keen grasp of how to draft legislation.^[157] So many stakeholders spend so much time drafting a single initiative that successful initiatives are often extremely well-crafted; whereas with traditional legislation, a legislative council composed of a select group of people is tasked with drafting and revising hundreds, if not thousands, of pieces of legislation every year.^[158]

Perhaps it is precisely this untethering of the traditional levers of power that makes the initiative process appealing to criminal reform groups.^[159] By bringing a range of affected voices together, criminal reform initiatives can manage to capture the range of experiences of scholars, legal advocates, activists, and reformers by mediating their input into targeted reforms.

When reformers pursue lobbying efforts or draft model legislation for representatives, there are more opportunities for criminalization advocates, such as obstinate district and county attorneys or prison profiteering enterprises—who have privileged access to the legislative process—to dilute the original goals. Unlike legislative action, initiatives restrict those pro‑criminalization perspectives from being factored into the actual legislation, instead directing those parties to pursue public advocacy in the form of ballot materials or advertisements against the measures.^[160] By exclusively targeting perspectives on the reform spectrum, initiatives can enhance the deliberative process without diluting the goals of reform.

When it comes to voting, criminal law is unique; it does not fall into the traditional categories of ballot measures because it involves more straightforward issues than, for example, the complexities of zoning.^[161] Criminal issues generally involve public values, so it is less clear that legislative bodies are privy to an obvious information advantage.^[162]

Part of the appeal of criminalization on the ballot is that its effects are deeply personal to so many people. Because over seventy million Americans have criminal records,^[163] mass incarceration’s destructive effects have evolved into a majoritarian issue perfectly suited for direct democracy. Voters may be more highly attuned to criminalization than legislators, whose status^[164] shields them from the pervasive effects of mass incarceration, and alternatively, from the traumatic effects of crime itself.

B.    Minority Rights

Considerable scholarship has been dedicated to critiquing direct democracy as a vehicle for disadvantaging minority communities.^[165]While it is true that many types of ballot measures have worked to the disadvantage of minority communities, this is not the case for criminal reform. Mass incarceration is a phenomenon that disproportionately affects minority communities.^[166] The criminalization of many types of drugs—such as peyote, marijuana, and crack cocaine—was fueled by associations of those drugs with minority communities.^[167] When ballot measures are aimed at criminal reform, they inherently protect minority communities who have been disproportionately affected by mass incarceration.^[168] Initiatives are better suited to factor in the lived experiences of marginalized voices affected by mass incarceration by giving those locked out of the legislative process the power to override legislators when laws treat their communities unfairly.

Florida’s Amendment 4 helps illustrate this phenomenon. When Desmond Meade realized his prior conviction prevented him from voting for his wife’s legislative candidacy, he began campaigning across Florida to re‑enfranchise former felons.^[169] He joined others who had felony convictions to create the Florida Rights Restoration Coalition, the group that created and worked to pass Florida’s historic Amendment 4.^[170] In a sense, it was only through the initiative process that Meade and 1.4 million others regained a political voice.[171] But the initiative process offered more; it put those affected by the carceral regime at the center of political change and gave them the opportunity to create an organization and to build coalitions that offer lasting political power. Where reforms originate is just as important as the reforms themselves. Initiatives are often the only mechanism that can put marginalized voices affected by mass incarceration at the center of politics, giving those locked out of the legislative process the power to override legislators when laws treat their communities unfairly.

C.    Public Support

But the ability to craft focused legislation is toothless without a showing of actual support for criminal reform. Polling has found support for criminal reform stretches across the political spectrum. As a broad category, the amount of support for criminal reform ranges from 75% of Americans at the lower end^[172] to as high as 90%.^[173] Even when presented with the nuances of criminal reform, Americans are undeterred. Eighty-seven percent disapprove of mandatory minimums and express support for alternatives to incarceration.^[174] Eighty-five percent believe that the goal of the criminal justice system should be rehabilitation, as opposed to punishment or deterrence.^[175] And 90% are in favor of removing barriers to employment, education, and other opportunities for the formerly incarcerated.^[176]

The strong support for criminal reform cuts across a range of issues, from racial bias recognition to mental health alternatives to incarceration, and from mandatory minimums to reducing the overall prison population.^[177] What’s more, public support remains high not just nationally but also at the state level, including states with the initiative process.^[178]

Better than all this polling, though, is results. Over the last two decades, citizens in states across the country have come to the polls to pass a wide array of criminal reform initiatives. More than 80% of the proposed reforms were passed.^[179] Both polling and the successful initiatives underscore the broad-based support for reform.

D.   Legislative Inaction

Somehow widespread public support among diverse cross sections of the American public has not translated into fundamental reforms at the legislative level. Arizona’s most recent legislative session is one among countless examples. The 2019 session yielded almost no movement on criminal reform; only one of the seventeen bills proposed actually passed.^[180] Many bills were killed without a single hearing, there were allegations that Arizona’s District Attorney exercised outsized influence in preventing legislation, and in some instances the Governor’s veto thwarted legislation.^[181]

In contrast, Oklahoma’s reform politics are worth examining. Kris Steele, a long-serving member of the Oklahoma House of Representatives and former Speaker of the House, used his position to shepherd criminal reform bills through the Legislature.^[182] After passage, however, the Legislature refused to fund those plans. So instead of seeking change through traditional legislative action, he decided that “[t]he best way . . . to ultimately change the path that we have been on forever was to bypass the political gridlock.”^[183]

He then led a coalition of criminal justice reform groups in placing criminal reform measures directly on the ballot in 2016.[184] Those measures sought to reclassify certain drug and property crimes as misdemeanors and redirect any savings to fund rehabilitation programs.^[185]Both measures passed in Oklahoma with almost 60% support from the voters.^[186]

Though merely anecdotes, both stories present an important piece of the argument for ballot initiatives. It is much more difficult to influence all the various levers of institutional power than to capture the will of the popular majority. Since elected representatives aren’t myopically committed to a single issue, counting on critical reforms by placing an extra degree of separation between the voter and the reform only complicates criminal reform efforts. Legislatures have not embraced the mantle of criminal reform; it may be wise to put more trust in our people than in our politicians.

E.    The Bellwether Advantage

Representative government is geared toward compromise, not transformative politics. This is all too familiar to the reform movement. For example, New York attempted to enact a progressive bail reform program, which “would have reduced the jail population by at least [forty] percent.”^[187] After pressure from the typical pro-criminalization contingent, the Legislature balked and significantly walked back its original proposal three months later.^[188] Big changes are hard to achieve through legislatures captured by institutional actors and averse to programmatic change.

But mass incarceration is an urgent crisis that demands more than piecemeal legislation. Initiatives, by design, are audacious policies; they arise where the legislative process fails. This allows them to act as a bellwether for progressive criminal policy. It also takes time before an initiative can be revisited by the people, giving the policy more space to breathe and the populace more time to reflect. When a state adopts a criminal reform policy through the initiative process, its effectiveness can serve as a powerful signal to other states.^[189] As transformative ballot measures are implemented, tested, and proven highly effective in forerunner states, they can serve as a bellwether for other states by triggering later policy changes elsewhere.^[190] By following the models of marijuana legalization, felon enfranchisement, and sanctuary programs at the city level, the reform movement can use direct democracy to generate more national attention for the urgency of bold criminal reform.

F.    Money, Money, Money

The strongest critique of using ballot measures to pursue criminal reform is the cost. Political expense is not unique to direct democracy. Initiatives are naturally expensive projects helmed by a robust supervisory committee and campaign effort. The sixty-eight citizen-initiated measures placed on the ballot in 2018 cost, in total, around seventy-five million dollars.^[191] Simple math will tell you that, on average, citizen-initiated measures cost over a million dollars each.^[192] This number is highly dependent upon the specific state; a campaign will need more money in California than, say, Oklahoma.^[193]

The costs begin with a filing fee and organizational start-up costs, but the single-most expensive aspect of initiatives is signature gathering.^[194] Initiative campaigns must hire hundreds, if not thousands, of professional signature gatherers to reach the threshold necessary for appearing on the ballot.^[195] Initiative campaigns must also anticipate the costs of pre-ballot litigation in the event that any group challenges their initiative for failure to meet statutory requirements.^[196] Consequent to these costs, the critique goes, there are limits on who can use the initiative process.^[197]

Funding is about priorities. Right now, the reform movement is hardly cash poor. After the George Floyd protests, the Minnesota Freedom Fund—a bail fund—received thirty million dollars in donations.^[198] It was so much money they struggled to spend even 1%.^[199] That same influx of money could have placed a reimagined bail system on ballots in half the states in America, with change to spare. Technological advancements and public attention have enhanced the ability for prospective groups to attract fundraising;^[200] focusing more funds on initiatives might maximize the return on investment.

V. Toward a Conclusion

2020 may be an inflection point in this movement. People organized, lined up outside the halls of power across our nation, and demanded justice. They did this despite the devastating challenges and risks associated with the COVID-19 outbreak. Several ballot campaigns were forced to postpone their 2020 measures.^[201] Public closures made it rather difficult to gather signatures in time for the election. This makes it inspiring that so many criminal reform measures still qualified for the ballot and passed. If anything, the profound turnout in the 2020 election raised the stakes of future initiatives that will now require even more signatures.

At this very moment, ballot campaigns across the country are likely preparing a new slate of initiatives to bring directly to the people in upcoming elections. Maybe 2020 marked the crossing of a Rubicon in the fight for criminal justice, and we will see even bolder reforms passed by the next election cycle. Regardless, there is much work to be done. After all, lasting criminal reform depends on the power of people, not politicians.

Black Lives Matter

APPENDIX A^[202]

Death Penalty Ballot Measures

APPENDIX B^[203]

Decriminalization Ballot Measures

APPENDIX C^[204]

Pro-criminalization Ballot Measures

     *      J.D. Candidate, 2021. This is dedicated to my family. To friends. To my editors and colleagues at the Arizona State Law Journal. To the Movement for Black Lives. And to the courage required to organize and demand better from our public institutions.

     [1].     Janie Haseman, Karina Zaiets, Mitchell Thorson, Carlie Procell, George Petras & Shawn J. Sullivan, Tracking Protests Across the USA in the Wake of George Floyd’s Death, USA Today (June 18, 2020, 3:48 PM), https://www.usatoday.com/in-depth/graphics/2020/06/03/map-protests-wake-george-floyds-death/5310149002/[https://perma.cc/VR8K-VSA4].

     [2].     Larry Buchanan, Quoctrung Bui & Jugal K. Patel, Black Lives Matter May Be the Largest Movement in U.S. History, N.Y. Times (July 3, 2020), https://www.nytimes.com/interactive/2020/07/03/us/george-floyd-protests-crowd-size.html [https://perma.cc/4R9P-JKSS].

     [3].     See, e.g., Black Lives Matter Protesters and Counter Protesters Clash at Prescott Rally, AZCentral (Oct. 11, 2020, 7:30 AM), https://www.azcentral.com/picture-gallery/news/local/arizona/2020/10/11/black-lives-matter-protesters-and-counter-protesters-clash-prescott-rally/5948563002/ [https://perma.cc/PJW9-B2SC].

     [4].     Alec Karakatsanis, The Punishment Bureaucracy: How To Think About “Criminal Justice Reform,” 128 Yale L.J.F. 848, 850 (2019) (citing Lani Guinier & Gerald Torres, The Miner’s Canary: Enlisting Race, Resisting Power, Transforming Democracy 263 (2002)).

     [5].     Duren Banks, Matthew Hickman & Tracey Kyckelhahn, U.S. Dep’t of Just., National Sources of Law Enforcement Employment Data 1 (Oct. 4, 2016), https://www.bjs.gov/content/pub/pdf/nsleed.pdf [https://perma.cc/E4WR-WVJY].

     [6].     Jeff Welty, Overcriminalization in North Carolina, 92 N.C. L. Rev. 1935, 1937 (2014) (“[T]he vast majority of criminal prosecutions in the United States happen in state courts.”).

     [7].     Press Release, Pub. Pol’y Inst. of Cal., Initiative Process: Money Doesn’t Buy Success at Ballot Box, https://www.ppic.org/press-release/initiative-process-money-doesnt-buy-success-at-ballot-box/ [https://perma.cc/T6FN-HTNF].

     [8].     Appendix B.

     [9].     See, e.g., Rachel Elise Barkow, Prisoners of Politics: Breaking the Cycle of Mass Incarceration 1 (2019) (“[T]he typical voter lacks the requisite data and knowledge to make the best decisions in these areas.”).

     [10].   See generally Nathaniel A. Persily, The Peculiar Geography of Direct Democracy: Why the Initiative, Referendum and Recall Developed in the American West, 2 Mich. L. & Pol’y Rev. 11, 21–40 (1997).

     [11].   Id. at 32.

     [12].   What Are Ballot Propositions, Initiatives, and Referendums?, Initiative & Referendum Inst., http://www.iandrinstitute.org/quick-facts.cfm [https://perma.cc/YUZ5-3NTN].

     [13].   Id.

     [14].   Id.

     [15].   Id.

     [16].   See id.

     [17].   Id.

     [18].   See Recall (Political), Ballotpedia, https://ballotpedia.org/Recall_(political) [https://perma.cc/ELA9-ARCS]. Another form of direct democracy is the recall, where people vote on whether to recall a public official from their office. Id. The recall process is outside the focus of this Comment.

     [19].   States with Initiative or Referendum, Ballotpedia, https://ballotpedia.org/States_with_initiative_or_referendum [https://perma.cc/5RSL-WUJM]. For an analysis of which states use certain direct democracy processes, see Initiative and Referendum States, Nat’l Conf. of St. Legislatures, https://www.ncsl.org/research/elections-and-campaigns/chart-of-the-initiative-states.aspx [https://perma.cc/BZ7P-A2LW].

     [20].   John G. Matsusaka, Direct Democracy Works, 19 J. Econ. Persps. 185, 185–86 (2005).

     [21].   For example, states have individualized procedural requirements regarding how to notify the state that a group is pursuing a ballot measure and regarding the forms and language that must be included on petitions. See, e.g., Laws Governing the Initiative Process in Arizona, Ballotpedia,https://ballotpedia.org/Laws_governing_the_initiative_process_in_Arizona [https://perma.cc/7G88-KD83].

     [22].   See generally Signature Requirements, Ballotpedia, https://ballotpedia.org/Signature_requirements [https://perma.cc/3KDQ-TTWE] (listing the various signature requirement for each state).

     [23].   Some states set their thresholds on the basis of votes cast in the previous presidential election; others look to the prior secretary of state election; and still others formulate the percentage based on the number of citizens or registered voters in their state. Id.

     [24].   Richard J. Ellis, Signature Gathering in the Initiative Process: How Democratic Is It?, 64 Mont. L. Rev. 35, 44 (2003).

     [25].   In 2000, the same 5% signature threshold required 420,000 signatures in California, while requiring only 20,000 signatures in Montana. Id. at 46.

     [26].   See, e.g., 3 Cal. Fair Pol. Pracs. Comm’n, Ballot Measure Committees: Campaign Disclosure Manual (2020), http://www.fppc.ca.gov/content/dam/fppc/NS-Documents/TAD/Campaign%20Manuals/Manual_3/Final_Manual_3.pdf [https://perma.cc/‌9C7P-V2RZ].

     [27].   Ellis, supra note 24, at 49 (citing James Barnett, The Operation of the Initiative, Referendum, and Recall in Oregon 59–60 (1915)).

     [28].   Id.

     [29].   Meyer v. Grant, 486 U.S. 414, 414 (1988) (overturning Colorado’s ban on paid petitioners).

     [30].   See Dan Frosch, Professional Petitioners Aid Ballot Initiatives, N.Y. Times (Oct. 23, 2010), https://www.nytimes.com/2010/10/24/us/politics/24petition.html [https://perma.cc/‌DXB3-UWVG]; Ellis, supra note 24, at 37.

     [31].   Laws Governing Petition Circulators, Nat’l Conf. of St. Legislatures, https://www.ncsl.org/research/elections-and-campaigns/laws-governing-petition-circulators.‌aspx [https://perma.cc/3D77-G3UN] (2012); see also Petition Circulator, Ballotpedia, https://ballotpedia.org/Petition_circulator [https://perma.cc/J3KD-TM5Q].

     [32].   “To ensure the measure qualifies for the ballot, the campaign’s signature goal should [be] 33–100% higher than the number of actual signatures required.” Greenbelt All., Successful Citizens’ Initiatives: A Guide to Winning Local Land-Use Ballot Measure Campaigns 5 (2002), http://www.greenbelt.org/wp-content/uploads/2012/01/Successful-Citizens-Initiatives.pdf [https://perma.cc/E4KW-GRZM].

The Washington Secretary of State urges initiative campaigns to gather “as many signatures as possible” due to potential invalidations. Frequently Asked Questions About Circulating Initiative and Referendum Petitions, Wash. Sec’y of St., https://www.sos.wa.gov/elections/initiatives/faq.aspx [https://perma.cc/2Z9D-BJQY].

     [33].   See Greenbelt All., supra note 32, at 16.

     [34].   In several states, the proponent may draft the petition title, sometimes subject to the approval of a state officer such as the Attorney General, Secretary of State, or Board of Elections. More commonly, proponents present their initiative, whereupon those state officials draft the title. A proponent can, in almost all initiative states, challenge the given title. See Overview, Nat’l Conf. of St. Legislatures, https://www.ncsl.org/research/elections-and-campaigns/preparation-of-a-ballot-title-and-summary.aspx [https://perma.cc/6GPZ-Q8TK].

     [35].   See id.

     [36].   See id.

     [37].   See Initiative and Referendum Process, Nat’l Conf. of St. Legislatures, https://www.ncsl.org/research/elections-and-campaigns/initiative-and-referendum-processes.‌aspx [https://perma.cc/JM6P-XQST].

     [38].   See, e.g., Matsusaka, supra note 20, at 186.

     [39].   See, e.g., Cheryl Miller, Sacramento Firm Drafts Ballot Initiative Challenging Landmark Labor Law, Law.com (Oct. 29, 2019, 5:54 PM), https://www.law.com/therecorder/2019/10/29/sacramento-firm-drafts-ballot-initiative-challen‌ging-landmark-labor-law/ [https://perma.cc/EF4H-S9LZ]; What We Do, Ballot Initiative Strategy Ctr., https://ballot.org/ [https://perma.cc/A7EY-G6BM].

     [40].   See, e.g., Miller, supra note 39; Ballot Initiative Strategy Ctr., supra note 39.

     [41].   See, e.g., Laws Governing the Initiative Process in California, Ballotpedia, https://ballotpedia.org/Laws_governing_the_initiative_process_in_California [https://perma.cc/‌8LDV-2NNQ] (“Proponents may seek the assistance of the Office of the Legislative Counsel in drafting the measure prior to filing it . . . .”).

     [42].   See, e.g., Stiritz v. Martin, 2018 Ark. 281, at 3–8, 556 S.W.3d 523, 527–29 (holding that an initiative legalizing gambling was not misleading); Dep’t of State v. Hollander, 256 So. 3d 1300, 1302 (Fla. 2018) (reversing a lower court ruling that the ballot language for Florida’s Marsy’s Law measure was misleading); Prevent Sanctuary Cities v. Haley, No. 74966, 2018 WL 2272955, at *4–6 (Nev. May 16, 2018) (ruling that a Nevada initiative was “confusing and misleading to voters”); Conway v. Martin, 2016 Ark. 322, at 4–11, 499 S.W.3d 209, 212–15 (certifying a medical marijuana initiative because the language was not misleading and instead would “give voters a fair understanding of the issues presented”).

     [43].   Benca v. Martin, 2016 Ark. 359, at 13–16, 500 S.W.3d 742, 751–52 (striking a medical marijuana initiative from the ballot because signature irregularities and missing canvasser information resulted in an insufficient number of signatures); Birks v. Dunlap, No. BCD-AP-16-04, 2016 Me. Bus. & Consumer LEXIS 9, at *32–34 (Apr. 8, 2016) (overturning the Secretary of State’s invalidation of signatures because the State had impermissibly invalidated all signatures by some circulators due to some signature variance).

     [44].   League of Women Voters of Pa. v. Boockvar, No. 578 M.D. 2019, 2019 Pa. Commw. Unpub. LEXIS 623, at *6–7, aff’d, 219 A.3d 594 (Pa. 2019) (enjoining the Secretary of State from certifying the results of the Marsy’s Law Initiative while the court determines whether the measure satisfies the single subject requirement); Mont. Ass’n of Cntys. v. State, 2017 MT 267, ¶¶ 1–3, 389 Mont. 183, 185, 404 P.3d 733, 735 (preventing Marsy’s Law from reaching the ballot because it violated the single-issue and separate-vote requirements).

     [45].   See, e.g., Steele v. Pruitt, 2016 OK 87, ¶¶ 17–19, 378 P.3d 47, 53–54; see also Bob Salsberg, Pot Legalization Question Cleared for Massachusetts Ballot in November, Ins. J. (July 8, 2016), https://www.insurancejournal.com/news/east/2016/07/08/419525.htm [https://perma.cc/BA8H-U8KR] (explaining that where the current title and statement of Massachusetts’s Marijuana Legalization Initiative were “clearly misleading,” the court ordered state officials to change the wording but allowed it to proceed to the ballot).

     [46].   Hollingsworth v. Perry, 570 U.S. 693, 693–94 (2013) (holding that ballot sponsors did not have Article III standing to appeal an adverse ruling when the state refused to appeal).

     [47].   See id.

     [48].   Vann R. Newkirk II, American Voters Are Turning to Direct Democracy, Atlantic (Apr. 18, 2018), https://www.theatlantic.com/politics/archive/2018/04/citizen-ballot-initiatives-2018-elections/558098/ [https://perma.cc/KP2V-MCLA].

     [49].   Id.

     [50].   Id.

     [51].   Id.

     [52].   Ari Berman, After Voters Passed Progressive Ballot Initiatives, GOP Legislatures Are Trying To Kill Future Ones, Mother Jones (Dec. 20, 2018), https://www.motherjones.com/politics/2018/12/after-voters-passed-progressive-ballot-initiatives‌-gop-legislatures-are-trying-to-kill-future-ones/ [https://perma.cc/9L36-MVH3].

     [53].   Zachary Roth, The Assault on Direct Democracy, Brennan Ctr. for Just. (Mar. 27, 2019) [hereinafter Roth, The Assault on Direct Democracy], https://www.brennancenter.org/our-work/analysis-opinion/assault-direct-democracy [https://perma.cc/BXR8-DP9A]. This policy was subsequently overturned by the Court of Claims in Michigan, which held that Michigan “cannot restrict the number of signatures” that can be gathered from a single district. Cheyna Roth, Court Rules on Ballot Initiatives Law, Mich. Radio: NPR (Sept. 27, 2019), https://www.michiganradio.org/post/court-rules-ballot-initiatives-law [https://perma.cc/78TA-6F9C].

     [54].   Cynthia Sewell, GOP Senators Broke Rank To Vote Against Idaho’s ‘Ballot Bill,’ but It Narrowly Passed, Idaho Statesman (Mar. 22, 2019, 2:37 PM), https://www.idahostatesman.com/news/politics-government/state-politics/article228288169.‌html [https://perma.cc/6GTM-2TND].

     [55].   Ryan Randazzo, Arizona Lawmakers Vote To Make It Harder for Citizens To Change Laws Via Ballot Measures, AZCentral (May 9, 2019, 5:59 PM), https://www.azcentral.com/story/news/politics/arizona/2019/05/08/citizen-initiatives-arizona-legislature-approves-bill-add-barriers-ballot-measures/1142430001/ [https://perma.cc/T98U-YTZA].

     [56].   Roth, The Assault on Direct Democracy, supra note 53.

     [57].   Howard Fischer, Federal Lawsuit Challenges Law To Regulate Signature Gatherers, Ariz. Capitol Times (July 15, 2019), https://azcapitoltimes.com/news/2019/07/15/federal-lawsuit-challenges-law-to-regulate-signature-gatherers/ [https://perma.cc/MNB3-UFRH].

     [58].   Id.

     [59].   Id.

     [60].   Leach v. Reagan, 430 P.3d 1241, 1244 (Ariz. 2018); Statement from NextGen Climate Senior Advisor Jamison Foser on NextGen’s Participation in a Lawsuit To Overturn Arizona’s Undemocratic Strikeout Law, NextGen Am., https://nextgenamerica.org/statement-on-arizonas-undemocratic-strikeout-law/ [https://perma.cc/LPB5-JKNH].

     [61].   Stanwitz v. Reagan, 429 P.3d 1138, 1140 (Ariz. 2018); Howard Fischer, Judge Lets Arizona Law on Initiative Petitions To Stand, Ariz. Capitol Times (Dec. 16, 2019), https://‌azcapitoltimes.com/news/2019/12/16/judge-lets-arizona-law-on-initiative-petitions-to-stand/ [https://perma.cc/FP78-3U72].

     [62].   Howard Fischer, Arizona Lawmakers Advance Bill To Limit Voter Initiatives, Ariz. Daily Star (Mar. 1, 2017), https://tucson.com/news/local/arizona-lawmakers-advance-bill-to-limit-voter-initiatives/article_9f951b03-ac64-5dcc-bdb1-af739b197362.html [https://perma.cc/‌6N76-MDKA].

     [63].   Id. To be certain, it is unclear whether the strict compliance requirement is constitutional because courts in Arizona have traditionally applied substantial compliance, not strict, when reviewing citizen-initiated measures for procedural deficiencies. In the only two cases in which this strict compliance requirement has been challenged, Arizona courts have declined to rule on its constitutionality. See Stanwitz, 429 P.3d at 1142; Madonna v. State, No. 1 CA-CV 17-0550, 2018 WL 2111441, at *2–3 (Ariz. Ct. App. May 8, 2018) (ruling that appellants lacked standing).

     [64].   Berman, supra note 52.

     [65].   Id.

     [66].   David A. Graham, The Legislators Working To Thwart the Will of Voters, Atlantic (Mar. 27, 2017, 9:57 AM), https://www.theatlantic.com/politics/archive/2017/03/state-legislatures-overturn-ballot-initiatives/519591/ [https://perma.cc/9DYQ-CFSA].

     [67].   Legislative Alterations of Ballot Initiatives, Ballotpedia (Apr. 16, 2019), https://ballotpedia.org/Legislative_alterations_of_ballot_initiatives [https://perma.cc/BT2A-ZJY9] (select the “Alterations by state” tab located within the last section of the page). For a thorough analysis of the various mechanisms and statistics regarding legislative alteration of initiatives, see id.

     [68].   After the voters enacted Amendment 4 permitting voting by former felons, the Florida Legislature passed a law requiring that all fines and fees associated with their convictions be paid prior to being eligible to vote. After robust litigation, the Eleventh Circuit ultimately upheld the legislative restriction as constitutional. Jones v. Governor of Fla., 975 F.3d 1016, 1025 (11th Cir. 2020) (en banc).

     [69].   Ballotpedia, supra note 67. North Dakota, Washington, Nebraska, and Arkansas each require a two-thirds vote of the legislature to amend initiatives. In addition, North Dakota, Washington, Alaska, Nevada, and Wyoming each require the elapse of a specified period of time, typically two to three years, prior to any alteration. Id.

     [70].   Because they do not constitute the core of mass incarceration, this study largely omitted from consideration ballot measures regarding election reform, which may include the imposition of some criminal sanctions, unless they are explicitly focused on criminalization. It also excluded measures that tangentially include a potential criminal sanction but are peripheral to core issues involved in criminalization. The primary focus here is criminal-related ballot measures since 2000 (apart from the analysis of death penalty measures, which reviews all historical measures).

It is worth acknowledging the inherent difficulties involved in any such statistical approach, including: multi-issue or dual-purpose measures with competing interests, lack of voter alternatives, tangentially criminal measures, qualitatively ambiguous measures, and the array of measures that were proposed but failed to make it onto ballots. Still, statistical review of these selected measures provides a worthwhile starting point in analyzing their effectiveness and frequency.

     [71].   Erwin Chemerinsky, Challenging Direct Democracy, 2007 Mich. St. L. Rev. 293, 297; see also Tanya M. Larrabee, Vote “No” on Criminal Justice Ballot Measures, 42 New Eng. J. Crim. & Civ. Confinement 207, 208 (2016) (“[C]riminal justice matters should not be placed on an election ballot.”).

     [72].   See John F. Galliher, Gregory Ray & Brent Cook, Abolition and Reinstatement of Capital Punishment During the Progressive Era and Early 20th Century, 83 J. Crim. L. & Criminology 538, 538–39 (1992).

     [73].   Id. at 542.

     [74].   Id. at 551.

     [75].   The Arizona Legislature passed laws restricting the Governor’s power to issue pardons and reprieves, laws that Governor Hunt then vetoed. Id.

     [76].   Id. at 552.

     [77].   First, there was the lynching of a murderer and rapist who had boasted that the state could not execute him. Then, on Christmas Eve, several men murdered a local proprietor, having discussed that the worst possible sentence they faced was life imprisonment. One of the men danced while being led away to serve his sentence. The Tucson Citizen published an article shortly thereafter advocating for the return of capital punishment and its deterrent effect. Id. at 563.

     [78].   Id. at 562.

     [79].   Appendix A.

     [80].   Galliher et al., supra note 72, at 541.

     [81].   Appendix A.

     [82].   Id.

     [83].   Appendix B.

     [84].   Id.

     [85].   See Karakatsanis, supra note 4, at 892.

     [86].   Appendix B.

     [87].   Id.

     [88].   Id.

     [89].   Id.

     [90].   Id.

     [91].   See id.

     [92].   See id.

     [93].   Id.

     [94].   Id.; Louisiana Amendment 2, Unanimous Jury Verdict for Felony Trials Amendment (2018), Ballotpedia, https://ballotpedia.org/Louisiana_Amendment_2,_Unanimous_‌Jury_Verdict_for_Felony_Trials_Amendment_(2018) [https://perma.cc/K5RB-LAK8].

     [95].   Appendix B.

     [96].   Id.

     [97].   Id.

     [98].   Sarah Holder, Where It’s Legal To Reverse the Vote of the People, Bloomberg: CityLab (Oct. 12, 2018, 7:00 AM), https://www.citylab.com/equity/2018/10/where-its-legal-to-reverse-the-vote-of-the-people/572386/ [https://perma.cc/JE3M-CRHX].

     [99].   Erik Luna, Prosecutorial Decriminalization, 102 J. Crim. L. & Criminology 785, 805–06 n.111 (2012).

     [100]. Appendix B.

     [101]. Id.

     [102]. Marcy Mistrett & Jeree Thomas, A Campaign Approach To Challenging the Prosecution of Youth as Adults, 62 S.D. L. Rev. 559, 568 (2017) (citing California Passes Proposition 57; Direct File Abolished, Pac. Juv. Def. Ctr. (Nov. 9, 2016), https://www.pjdc.org/california-passes-proposition-57-direct-file-abolished/[https://perma.cc/‌WM8Z-74V6]).

     [103]. Appendix B.

     [104]. Id.

     [105]. Id.

     [106]. Id.

     [107]. See id. Two Alabama “stand your ground” measures were omitted from the analysis because they could not be fit in either decriminalization or pro-criminalization categories. See Alabama 2020 Ballot Measures, Ballotpedia, https://ballotpedia.org/Alabama_2020_ballot_measures [https://perma.cc/T9T6-66QY]. From one perspective, these measures support removing criminal penalties for certain conduct; alternatively, they support the private imposition of the most serious punishment against certain offenders.

     [108]. Appendix B.

     [109]. Id.

     [110]. See Mississippi Ballot Measure 1, Initiative 65 and Alternative 65A, Medical Marijuana Amendment (2020), Ballotpedia, https://ballotpedia.org/Mississippi_Ballot_Measure_1,_Initiative_65_and_Alternative_65A,_Medical_Marijuana_Amendment_(2020) [https://perma.cc/UN5D-CNGY].

     [111]. Appendix B.

     [112]. Michigan Proposal 2, Search Warrant for Electronic Data Amendment (2020), Ballotpedia, https://ballotpedia.org/Michigan_Proposal_2,_Search_Warrant_for_Electronic_‌Data_Amendment_(2020)_ [https://perma.cc/L68L-S6BW].

     [113]. Appendix B.

     [114]. See id.

     [115]. See id.; Samantha Michaels, California Votes To Keep Cash Bail, Mother Jones (Nov. 4, 2020), https://www.motherjones.com/crime-justice/2020/11/california-votes-to-keep-cash-bail/ [https://perma.cc/P5ND-ETVD].

     [116]. California Proposition 25, Replace Cash Bail with Risk Assessments Referendum (2020), Ballotpedia, https://ballotpedia.org/California_Proposition_25,_Replace_Cash_Bail_‌with_Risk_Assessments_Referendum_(2020) [https://perma.cc/UF3F-GEL8].

     [117]. Michaels, supra note 115.

     [118]. Id.

     [119]. See Ayanna Alexander, Voters Back More Police Oversight After George Floyd’s Death (1), Bloomberg (Nov. 5, 2020, 2:44 PM), https://about.bgov.com/news/voters-nationwide-ok-police-oversight-measures-after-floyd-death/ [https://perma.cc/P5FL-CY83].

     [120]. Id.

     [121]. In 2020, California voters rejected the controversial bail reform Proposition, and Oklahoma failed to pass State Question 805, which would have prohibited the use of sentence enhancements for past, non-violent felony convictions. See Appendix B. Ohio failed to pass Issue 1 in 2018, which would have made drug possession and use misdemeanors and prohibited incarceration for those on parole for such convictions. See id. And Idaho’s Proposition 1, in 2018, which would have permitted betting on horse races at video terminals, also failed to pass. See id.

     [122]. See id.

     [123]. See id.

     [124]. Id.; See List of Ballot Measures By Year, Ballotpedia, https://ballotpedia.org/List_of_ballot_measures_by_year [https://perma.cc/3PD6-42Y4] (aggregating data by reviewing ballot initiatives in each year from 2000 to 2020).

     [125]. See Appendix B.

     [126]. Roger A. Fairfax, Jr., From “Overcriminalization” to “Smart on Crime”: American Criminal Justice Reform—Legacy and Prospects, 7 J.L. Econ. & Pol’y 597, 598 (2011).

     [127]. In California and Arizona, this is often because after repeated interference, voters insulated their ballot measures from most legislative alteration. See, e.g., State v. Jones, 440 P.3d 1139, 1144 (Ariz. 2019) (construing Proposition 203 broadly to preclude a conviction based on hashish); People v. Page, 406 P.3d 319, 320, 325 (Cal. 2017) (holding that though Proposition 47 did not explicitly refer to the Vehicle Code, it was designed to allow resentencing for such charges); People v. Conley, 373 P.3d 435, 439 (Cal. 2016) (applying the decriminalization‑focused Proposition 36 broadly to the benefit of criminal defendants). But even in the absence of strong state constitutional protections, the Massachusetts Supreme Court has repeatedly upheld the broad social policies attached to the ballot measure that decriminalized marijuana. Joseph N. Schneiderman, Smokers, Dealers and Growers: The Supreme Judicial Court Hashes Out the Marijuana-Reform Ballot Initiative, 58 Bos. Bar J. 24, 24 (2014) (citing cases).

     [128]. Matt Taibbi, Cruel and Unusual Punishment: The Shame of Three Strikes Laws, Rolling Stone (Mar. 27, 2013, 11:00 AM), https://www.rollingstone.com/politics/politics-news/cruel-and-unusual-punishment-the-shame-of-three-strikes-laws-92042/ [https://perma.cc/‌5VNW-VML6].

     [129]. Judith Greene, Getting Tough on Crime: The History and Political Context of Sentencing Reform Developments Leading to the Passage of the 1994 Crime Act, inSentencing and Society 1, 11–28 (Cyrus Tata & Neil Hutton eds., 2002), https://www.justicestrategies.org/sites/default/files/Judy/GettingToughOnCrime.pdf [https://‌perma.cc/SSC3-DBXC].

     [130]. Appendix C.

     [131]. Id. After Measure 57’s passage, the Oregon Legislature passed a bill counteracting that measure, and the Oregon Supreme Court upheld the legislative bill over the measure, permitting the reduced sentencing scheme. State v. Vallin, 434 P.3d 413, 421–22 (Or. 2019).

     [132]. Appendix C. The Arkansas Supreme Court invalidated the initiative preventing unmarried cohabitants from fostering children for violating the right to privacy and lacking an individualized assessment. Ark. Dep’t of Hum. Servs. v. Cole, 2011 Ark. 145, at 26, 380 S.W.3d 429, 443.

     [133]. Appendix C. Arizona’s Proposition 100 and 103 were both invalidated by courts. Lopez-Valenzuela v. Arpaio, 770 F.3d 772, 775 (9th Cir. 2014) (en banc); State v. Wein, 417 P.3d 787, 789–90 (Ariz. 2018).

     [134]. Appendix C.

     [135]. Id.

     [136]. Id.

     [137]. Id. A federal district court struck down several aspects of California’s Proposition 9, otherwise known as Marsy’s Law. Valdivia v. Brown, No. CIV. S-94-671, 2012 U.S. Dist. LEXIS 8092, at *21, *27–28, *30 (E.D. Cal. Jan. 24, 2012) (failure to provide prompt hearing).

     [138]. See Appendix C.

     [139]. Id.

     [140]. In any event, Amendment 2 was overturned as unconstitutional. Kim Chandler, Judge Dismisses Wrongful Death Lawsuit over Abortion, Associated Press (Sept. 3, 2019), https://apnews.com/fce7dc5bfd3a491aab191ad5e6b46596 [https://perma.cc/KL5E-SYTM].

     [141]. Measure 100 in Oregon and Initiative 1401 in Washington were both aimed at animal cruelty. See Appendix C. Proposal 2 in New York authorized the revocation of pensions for officials convicted of felonies related to their official duties, and Amendment 1 in Louisiana prohibited felons from holding office until five years after completion of their sentences. Id.

     [142]. Id.

     [143]. Id. The Tenth Circuit Court of Appeals upheld a district court ruling that Oklahoma’s ban on judicial consideration of international and Sharia law, passed by State Question 755, represented disfavored treatment of Muslims in violation of the Establishment Clause. Awad v. Ziriax, 670 F.3d 1111, 1133 (10th Cir. 2012).

     [144]. Appendix C.

     [145]. Again, in the last decade, eleven out of these fifteen measures were legislatively referred. Id.

     [146]. Ryan S. Appleby, Note, Proposition 9, Marsy’s Law: An Ill-Suited Ballot Initiative and the (Predictably) Unsatisfactory Results, 86 S. Cal. L. Rev. 321, 322 (2013).

     [147]. There has been at least some debate on the propriety and criminalization posture of these measures. Compare id. (arguing that Marsy’s Laws are violative of due process, lead to an increase in the number of those incarcerated, and present ex post facto violations), and Sophie Quinton, ‘Marsy’s Law’ Protections for Crime Victims Sound Great, but Could Cause Problems, Pew Rsch. Ctr.: Stateline (Oct. 12, 2018), https://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2018/10/12/marsys-law-protections-for-crime-victims-sound-great-but-could-cause-problems [https://perma.cc/HZ6B-NTLZ] (explaining that crime victims are already sufficiently protected by state laws and these initiatives will unduly influence prosecutors and cause inconsistent interpretations and due process violations), with Kathryne M. Young, Parole Hearings and Victims’ Rights: Implementation, Ambiguity, and Reform, 49 Conn. L. Rev. 431, 434–35 (2016) (arguing that while there needs to be a better tailoring of victims’ rights proposals to impactful and meaningful representation, victims’ rights should be decoupled from traditional punitive measures).

     [148]. See Appendix C.

     [149]. Id.

     [150]. See supra notes 132, 133, 137, 140, and 143. This might be because courts are more skeptical of punitive measures approved by popular vote. Cf. Julian N. Eule, Judicial Review of Direct Democracy, 99 Yale L.J. 1503, 1507 (1990) (“[J]udicial review of direct democracy frequently calls for less rather than more restraint.”).

     [151]. Appendix B; Appendix C.

     [152]. Appendix B; Appendix C. When discounting Marsy’s Law Initiatives as neither reform nor pro-criminalization measures, this number jumps from 68% to 79%.

     [153]. Elizabeth Garrett, Who Directs Direct Democracy?, 4 U. Chi. L. Sch. Roundtable 17, 22, 33 (1997). Voters, critics say, are “easily baffled by the complexities of the issues” and are often unduly influenced by advertising campaigns that can be highly deceptive. Shaun Bowler, When Is It OK To Limit Direct Democracy?, 97 Minn. L. Rev. 1780, 1783 (2013); Chemerinsky, supra note 71, at 299.

     [154]. Bowler, supra note 153, at 1784; see also Shaun Bowler & Todd Donovan, Demanding Choices: Opinion, Voting, and Direct Democracy 165–73 (2000).

     [155]. Technological advancements in communications have made for a more informed and participatory electorate. Jane S. Schacter, Digitally Democratizing Congress? Technology and Political Accountability, 89 B.U. L. Rev. 641, 643 (2009); see also Matsusaka, supra note 20, at 186.

     [156]. Critics largely point to process to make this point: in legislative chambers, laws are first drafted by professional staff and make their way through committees to the entire legislative body. Along the way, the language of the legislation is changed as legislators deliberate over key issues. This results in political compromise reflecting the will of the majority of voters. In contrast, it is argued, ballot initiatives need not be vetted, and often result in extreme positions that do not represent the electorate at large. See, e.g., Chemerinsky, supra note 71, at 297–98.

     [157]. See infra Part II.A.

     [158]. See Kim Eckart, New, UW-Developed Data Tool Tracks State Legislative Process, from First Draft to Final Law, Univ. Wash. News (Feb. 20, 2020), https://www.washington.edu/news/2020/02/20/new-uw-developed-data-tool-tracks-state-legislative-process-from-first-draft-to-final-law/ [https://perma.cc/4ZRT-3NY7].

     [159]. Labeling legislative bodies as deliberative without scrutinizing their true behaviors is antithetical to the highly suspect view direct-democracy critics take on voter deliberation. See Andrew Stengel, Lawrence Norden & Laura Seago, Still Broken: New York State Legislative Reform 27 (2008),http://www.brennancenter.org/sites/default/files/legacy/publications/Still.Broken.pdf [https://‌perma.cc/ZMS6-N9X3]. Legislators, like voters, can be influenced by advertising materials, outside money, and special interests. Bowler, supra note 153, at 1787.

     [160]. See Bowler, supra note 153, at 1788.

     [161]. See id. at 1784.

     [162]. Matsusaka, supra note 20, at 193.

     [163]. Americans with Criminal Records, Half in Ten, https://www.sentencingproject.org/wp-content/uploads/2015/11/Americans-with-Criminal-Records-Poverty-and-Opportunity-Profile.pdf [https://perma.cc/3TKV-5FF9].

     [164]. See, e.g., By the Numbers: Wealthiest State Lawmakers in Florida, News Serv. Fla. (Sept. 10, 2019), https://news.wjct.org/post/numbers-wealthiest-state-lawmakers-florida [https://perma.cc/Z5YV-Z8PA].

     [165]. Scores of ballot measures have systematically targeted racial and language minorities, LGBTQ persons, and other political minorities. Some initiatives expressly curtailed marriage and other rights for same-sex couples or deprived undocumented immigrants from access to public benefits. Chemerinsky, supra note 71, at 294–97; see also Todd Donovan, Direct Democracy and Campaigns Against Minorities, 97 Minn. L. Rev. 1730 (2013) (discussing initiatives curbing minority rights in housing, discrimination, desegregation, and employment contexts); David Schultz, Liberty v. Elections: Minority Rights and the Failure of Direct Democracy, 34 Hamline J. Pub. L. & Pol’y 169, 171 (2012) (“[T]he Progressive Era experiment with direct democracy has failed . . . .”).

     [166]. See generally Karakatsanis, supra note 4.

     [167]. See, e.g., Adryan Corcione, America’s War on Drugs Has Been Racist for a Century, Leafly (Oct. 5, 2018), https://www.leafly.com/news/politics/americas-war-on-drugs-has-been-racist-for-a-century [https://perma.cc/8GB5-35VL].

     [168]. But see Kristen Clarke, Opinion, Cannabis Industry Shouldn’t Expand Until We Fix Marijuana’s Racial Inequities, Injustices, USA Today (Feb. 22, 2019, 6:00 AM), https://www.usatoday.com/story/opinion/2019/02/22/marijuana-legalization-exposes-cannabis-industry-racial-injustice-incarceration-minorities-column/2836449002/ [https://perma.cc/Z5KD-VHA8] (stating that minorities who were directly affected by marijuana laws are cut out of the entrepreneurial benefits of its legalization and denied expungement of their records).

     [169]. Kathryn Fink & Kaity Kline, Blocked from the Ballot: Meet Desmond Meade, Who Has a Felony Conviction, WAMU (Mar. 10, 2020), https://the1a.org/segments/blocked-from-the-ballot-individuals-with-felony-charges/ [https://perma.cc/JL2X-ZXGT].

     [170]. Karen Duffin, Who Gets To Vote in Florida?, NPR: Planet Money (Oct. 28, 2020, 7:30 PM), https://www.npr.org/2020/10/26/927846676/who-gets-to-vote-in-florida [https://perma.cc/9PSP-CJ7R]; About Desmond Meade, Fla. Rts. Coal., https://floridarrc.com/desmond-meade/ [https://perma.cc/T6NF-F2AT].

     [171]. Fla. Rts. Coal., supra note 170.

     [172]. Overwhelming Majority of Americans Support Criminal Justice Reform, New Poll Finds, Vera Inst. of Just.: Think Just. Blog (Jan. 25, 2018), https://www.vera.org/blog/overwhelming-majority-of-americans-support-criminal-justice-reform-new-poll-finds [https://perma.cc/QT5K-Y243]; Poll Shows Americans Overwhelmingly Support Prison, Sentencing Reforms, Comm. on the Judiciary (Aug. 23, 2018), https://www.judiciary.senate.gov/press/rep/releases/poll-shows-americans-overwhelmingly-support-prison-sentencing-reforms [https://perma.cc/6LGB-Z7N5].

     [173]. 91 Percent of Americans Support Criminal Justice Reform, ACLU Polling Finds, ACLU (Nov. 16, 2017), https://www.aclu.org/press-releases/91-percent-americans-support-criminal-justice-reform-aclu-polling-finds [https://perma.cc/YPC8-M8A4].

     [174]. Vera Inst. of Just., supra note 172.

     [175]. Id.

     [176]. Id.

     [177]. ACLU, supra note 173; see also Matthew Clarke, Polls Show People Favor Rehabilitation over Incarceration, Prison Legal News (Nov. 6, 2018), https://www.prisonlegalnews.org/news/2018/nov/6/polls-show-people-favor-rehabilitation-over-incarceration/ [https://perma.cc/G7YU-ZS5Z] (collecting studies).

     [178]. See, e.g., Press Release, FWD.us, New Poll Shows Arizona Voters Support Bold Criminal Justice Reforms (Feb. 8, 2019), https://www.fwd.us/news/new-poll-shows-arizona-voters-support-bold-criminal-justice-reforms/ [https://perma.cc/6UQ8-FWSQ] (finding that over 80% of registered voters in Arizona support reform and “only one percent want to spend additional tax dollars on jails and prisons”).

     [179]. This figure reflects the iterative passage figure. See supra text accompanying note 123.

     [180]. Vanessa Ontiveros, Despite Passage of Arizona Criminal Justice Reform Bill, More Work Needs To Be Done, Advocates Say, Ariz. Daily Star (Aug. 8, 2019), https://tucson.com/news/local/despite-passage-of-arizona-criminal-justice-reform-bill-more-work/article_909d9829-b64d-5fab-b005-61f5485d9341.html [https://perma.cc/5GDD-4824].

     [181]. Id.; Steven Hsieh, One Bill Shows Why Arizona Hasn’t Passed Any Criminal Justice Reform This Year, Phx. New Times (Apr. 15, 2019, 10:56 AM), https://www.phoenixnewtimes.com/news/one-bill-shows-why-arizona-criminal-justice-reform-this-year-11269619 [https://perma.cc/DS9Y-SDAM]; see also 2019 #MSLEG Tracking: Criminal Justice Reform, ACLU Miss., https://www.aclu-ms.org/en/legislation/2019-msleg-tracking-criminal-justice-reform [https://perma.cc/KK8R-75QY] (showing that almost all criminal reform bills died in committee); Steven Zeidman, The State Legislature’s Criminal Justice Reform Failures, N.Y. Daily News (July 8, 2019, 12:29 PM), https://www.nydailynews.com/opinion/ny-oped-the-state-legislatures-criminal-justice-reform-failures-20190708-hs6rk7xw3zcpze7z567k7dl3vy-story.html [https://perma.cc/SB74-YJN9].

     [182]. Graham, supra note 66.

     [183]. Id. (quoting Kris Steele, member of the Oklahoma House of Representatives).

     [184]. Id.

     [185]. Id.

     [186]. See K. Querry, Oklahoma Voters Pass Criminal Justice Reform State Questions, KFOR (Nov. 8, 2016, 9:34 PM), https://kfor.com/news/oklahoma-voters-pass-criminal-justice-reform-state-questions/ [https://perma.cc/MD8S-HSPW].

     [187]. Taryn A. Merkl, New York’s Latest Bail Law Changes Explained, Brennan Ctr. for Just. (Apr. 16, 2020), https://www.brennancenter.org/our-work/analysis-opinion/new-yorks-latest-bail-law-changes-explained [https://perma.cc/YU4S-TVZ8].

     [188]. See id.

     [189]. See Press Release, All. for Safety & Just., Leading into Democratic Presidential Debate, Advocates Point to CA as Model of Criminal Justice Reforms that Candidates Have Proposed (Dec. 19, 2019), https://allianceforsafetyandjustice.org/press-release/leading-into-democratic-presidential-debate-advocates-point-to-ca-as-model-of-criminal-justice-reforms-that-candidates-have-proposed/ [https://perma.cc/MF7N-QKUC].

     [190]. Matsusaka, supra note 20, at 194.

     [191]. Ballot Measures Cost Per Required Signatures Analysis, Ballotpedia, https://ballotpedia.org/Ballot_measures_cost_per_required_signatures_analysis [https://‌perma.cc/UN55-8UHP].

     [192]. See id.

     [193]. See id.; see also Meghan McCarty Carino, How California Initiatives Went from ‘Power to the People’ to a Big Money Game, Cap Pub. Radio (Nov. 1, 2018), http://www.capradio.org/articles/2018/11/01/how-california-initiatives-went-from-power-to-the-people-to-a-big-money-game/ [https://perma.cc/FNF2-H48K]; Chemerinsky, supra note 71, at 298.

     [194]. Carino, supra note 193.

     [195]. Id.

     [196]. See supra Part II.B.

     [197]. See Chemerinsky, supra note 71, at 298.

     [198]. Sigal Samuel, Racial Justice Groups Have Never Had So Much Cash. It’s Actually Hard To Spend It., Vox (June 19, 2020, 11:00 AM), https://www.vox.com/future-perfect/2020/6/19/21294819/minnesota-freedom-fund-donations-police-protests [https://‌perma.cc/BY4B-J96A].

     [199]. Id.

     [200]. Digital reproducibility allows donation-gathering to be infinitely reproducible at little to no marginal cost; the internet has buoyed this small donor revolution. See Joe Ready, Small Dollar Donors Are Winning! (for Now), U.S. Pub. Int. Rsch. Grp. (Aug. 21, 2019), https://uspirg.org/blogs/blog/usp/small-dollar-donors-are-winning-%C2%A0now [https://‌perma.cc/5JFE-U2A6].

     [201]. Sarah Holder, How Coronavirus Is Killing Off Ballot Measures, Bloomberg: CityLab (Apr. 9, 2020, 9:00 AM), https://www.citylab.com/equity/2020/04/coronavirus-election-campaign-ballot-measure-signature-voter/609232/ [https://perma.cc/78TE-25VG].

     [202]. Death Penalty on the Ballot, Ballotpedia, https://ballotpedia.org/Death_penalty_on_the_ballot [https://perma.cc/Y9YC-D6KX]; Ballotpedia, supra note 124.

     [203]. Ballotpedia, supra note 124.

     [204]. Id.