In its 2017 “State of Pretrial Justice in America” report, the Pretrial Justice Institute (“PJI”) praised Arizona’s pretrial justice reform efforts. Arizona was one of nine states awarded a “B” grade (only one state—New Jersey—received an “A” grade). Arizona was also named a “state to watch.” Other commentators and news outlets have highlighted Arizona’s efforts and argue that it is one of the jurisdictions at the forefront of bail reform.
Although we agree that Arizona has taken some key steps toward creating a just and effective bail system, more needs to be done before the state becomes a true leader in pretrial justice. Arizona remains a state with high numbers of unconvicted people in its jails. Of the approximately 14,000 people incarcerated in Arizona’s jails in 2015, more than 10,000 of them were awaiting trial or other court hearings. In order to effect true pretrial justice reform and reduce the number of people detained in jail pending court hearings, this Article assesses Arizona’s pretrial justice reforms to date and suggests some ways for the State to move to the forefront of pretrial justice reform. Part II highlights the problems with pretrial justice in the United States and explains the impact of pretrial detention on individual defendants and the criminal justice system as a whole. Part III briefly reviews the structure of bail and pretrial detention in the United States. Part IV describes the structure and process of bail and pretrial release decision-making in Arizona. Part V discusses recent reform efforts and successes in Arizona. And Part VI describes the steps that the state still needs to take.
II. The Problems with Pretrial Justice in America
Each year, jails in the United States handle roughly 10.7 million admissions. In 2018, jails housed an average of 738,400 people on any given day. Just over 248,500 (33.7%) of these people were serving a custodial sentence after a criminal conviction. Put differently, two-thirds of the people in U.S. jails were being held in pretrial detention. More recent official data are not available, but researchers at the Vera Institute of Justice estimate that over the past year, the average daily jail population has increased to 758,400 people—the highest number of jail inmates in a decade. Only 160,000 of these people, representing less than a fifth of the total number, were convicted of a crime and serving a sentence of incarceration.
The raw numbers of unconvicted persons in jails do not even begin to capture the toll the U.S. money bail system takes on people accused, but not convicted, of criminal offenses. A study of bail in New York City found that “[e]ven when bail is set comparatively low—at $500 or less, as it is in one-third of nonfelony cases—only 15 percent of defendants are able to come up with the money to avoid jail.”
The effects of not being able to post bail go beyond the loss of liberty while awaiting trial. Indeed, . . . roughly half of all nonfelony cases in New York City end with an acquittal; in contrast, the conviction rate skyrockets to 92 percent for pretrial detainees. [It appears that pretrial detention is] so unpleasant that it pressures those accused of crimes to plead guilty in order to escape the conditions of confinement. [In short, bail] serves as both a mechanism “for locking people up” prior to any criminal conviction and for inducing guilty pleas [even though] neither could be further from the intended emancipatory purpose of bail when the concept first came into practice in England.
Disparities attendant to bail outcomes between those who live in poverty and those with financial means raise serious Fourteenth Amendment concerns. Indeed, courts have recently invalidated bail processes based on fixed bail schedules on Due Process or Equal Protection grounds. But this sorry state of affairs should also concern people for several other reasons, most notably its devastating effects on public budgets and its impact on public safety.
The financial costs of pretrial detention are astounding. According to the PJI,
pretrial detention costs taxpayers $38 million per day, or $14 billion per year—an amount that could support the employment of 250,000 elementary school teachers, the provision of free or reduced lunch for 31 million children, or the provision of shelter and services for the country’s 50,000 homeless veterans, and homelessness prevention services for the 1.4 million veterans who are at risk of becoming homeless. [Further], the costs of pretrial detention far exceed the costs of alternatives to incarceration, including pretrial supervision.
Second, contrary to the widely accepted myth that bail helps to keep communities safer, the opposite appears true. Consider that a comprehensive study of defendants in Harris County, Texas, found that “despite the initial incapacitation” attendant to pretrial detention, within one month, “the average number of new charges for [pretrial] detainees [exceeded] that of their similarly situated counterparts who were released” on bond. Moreover, within eighteen months, that disparity increases such that those detained pretrial had a 22% increase in subsequent misdemeanor charges and nearly a 33% increase in subsequent felony charges compared to those who were released. A study in Maricopa County, Arizona, similarly reported that even pretrial, low-risk defendants who were detained for four or more days had 49% greater odds of new criminal activity within twelve months of release. Moreover, these odds skyrocketed to being 78% more likely to recidivate for those held in pretrial detention for a month. These outcomes are likely a function of several factors, including detention leading to “job loss, disrupted interpersonal relationships, or other collateral consequences that change the relative attractiveness of crime in the future. . . . Pretrial detainees may also make new social ties or learn new skills through their interactions with other jail inmates that change their propensity for crime.”
Beyond the criminogenic effects of pretrial detention on those who could be released pending trial, a system of money bail also allows some defendants of financial means to pose a risk to public safety. Few people seem to be concerned about people who can buy their freedom pending trial, even though their money potentially allows them to evade prosecution.
Over the past few years, however, advocates from across the political spectrum have begun advocating for bail reform. Some of these efforts have taken the form of constitutional challenges to money bail systems. In other local jurisdictions, grassroots reform has led to the election of chief prosecutors who pledged not to seek bail in certain types of cases. In other states, activists have prompted legislatures to take action. Notably, New Jersey dramatically reduced its reliance on bail in 2017. Not only did the statutory reform create a presumption in favor of pretrial release, but it also limited pretrial detention to those who pose either an unacceptable flight risk or a danger to their community. Select other states have enacted legislation restricting money bail. And in still other states, judiciaries have taken the lead. Arizona, as discussed in Part V, is one of those states because the legislature declined to take action, prompting Scott Bales, who was the chief justice of the Arizona Supreme Court until 2019, to act.
III. The Structure of Bail and Pretrial Detention in the United States
The Eighth Amendment to the U.S. Constitution prohibits “excessive” bail but does not guarantee a right to bail. Nonetheless, the Eighth Amendment’s limits on judicial discretion for setting the amount of bail are virtually meaningless. Pretrial release is dependent both on state constitutions and statutory law.
Early case law on bail expressed two “related principles concerning the purpose of bail: (1) partially as a function of the presumption of innocence, bail is not supposed to be used in a manner that keeps those accused of crimes in pretrial detention”; and (2) bail is supposed to facilitate appearance at future court proceedings, not prevent the commission of other offenses. But these views have morphed over time. Quite the contrary, the Court unequivocally determined that the presumption of innocence “has no application to a determination of the rights of a pretrial detainee during confinement before his trial has even begun.”
More significant was the passage of the Bail Reform Act of 1984, upheld by the U.S. Supreme Court in United States v. Salerno, which specifically sanctioned “the previously unspoken practice of considering the dangerousness of a defendant” in making pre-trial release decisions. The Act mandated pretrial detention for those accused of certain felonies whenever a judicial officer finds there is clear and convincing evidence that “no condition or combination of conditions will reasonably assure the appearance of the person as required and the safety of any other person and the community.”
The explicit sanctioning of preventative detention is not the same as setting a high bail amount, because the former “guarantees incarceration and marks the first return to explicit remand after a century and a half of expansion of the right to bail. . . .” As a result, judges to this day concern themselves not only with a criminally accused person’s flight risk, but also with the danger that person poses to the community if released on bail. This is true in both the federal and state systems. Troublingly, however, predictions of future dangerousness typically amount to little more than unreliable prognostications.
What remains unclear is what is meant by the prohibition on excessive bail, which is one of the least litigated provisions of the Bill of Rights. In practice, it does not appear to put any restrictions on judges’ ability to set high bail. Judges can deliberately set bail that a defendant is unable to pay in order to effectively detain them; indeed, they often do so.  And judges rarely consider a defendant’s ability to pay when setting bail.
More recently, challenges to bail schedules and other determinations that do not take into account the particular defendant’s ability to pay have been challenged under the Fourteenth Amendment, relying on its dual guarantees of Due Process and Equal Protection. Although none of these decisions have reached the highest court in the land, some lower courts have embraced these arguments.
IV. The Structure and Process of Bail and Pretrial Release in Arizona
As in most states, bail in Arizona is governed by both constitutional and statutory provisions. In addition, the Arizona Rules of Criminal Procedure play an important role in how bail and pretrial release decisions are made, particularly as a result of changes that have taken place over the last few years. The purposes of the bail system are threefold: (1) to assure the appearance of the accused; (2) to protect against the intimidation of witnesses; and (3) to protect the safety of the victim, any other person, or the community.
A. Constitutional Provisions
Arizona’s state constitution provides a right to bail “by sufficient sureties.” But the right is qualified by four exceptions, all of which require a baseline showing of sufficient evidence such that “the proof is evident or the presumption great.” First, bail is prohibited for capital offenses. Second, bail is prohibited when a person was already admitted to bail on a separate felony charge and is subsequently charged with a new felony. Third, bail is not to be granted in felony cases when the defendant “poses a substantial danger to any other person or to the community,” and no conditions of release may be imposed that would “reasonably assure . . . safety.”
B. Statutory Provisions
Statutory law in Arizona mirrors the provisions in the state constitution, although after successful challenges to the constitutional provisions concerning sexual conduct with a minor and child molestation, the statute was amended. To qualify as nonbailable, at the time either of those offenses are alleged to have occurred, the defendant must now either have been “at least eighteen years of age and the victim was under thirteen years of age,” or alternatively, that “the victim was thirteen or fourteen years of age and the person was at least ten years older than the victim.” The statute also operationalizes the standard of proof for the constitutional exception prohibiting bail in select felony cases. Specifically, the statute directs courts to deny bail to a defendant
if the person is charged with a felony offense and the state certifies by motion and the court finds after a hearing on the matter that there is clear and convincing evidence that the person charged  poses a substantial danger to another person or the community or  engaged in conduct constituting a violent offense, [and] that no condition or combination of conditions of release may be imposed that will reasonably assure the safety of the other person or the community.
As used in this specific statutory provision, a “violent offense” is limited only to “dangerous crime[s] against children” or terrorism. The statute also provides that the accused’s membership in a “criminal street gang . . . may give rise to the inference [of] . . . substantial danger” to others or the community that no conditions could mitigate.
In practice, the judicial officer handling the initial appearance determines whether the offense charged is bailable under the constitutional and statutory schemes previously described. But even if an offense is deemed nonbailable, the court is free to release someone on bail if the proof evident or presumption great standard is not satisfied. Moreover, the court is vested with the authority to make exceptions in particular cases if the defendant “is in such physical condition” that confinement would endanger the life of the accused.
Under Simpson v. Owens, formal hearings are required to determine whether the “proof is evident or the presumption great” that the defendant committed a nonbailable offense. Though such hearings should not necessarily be conducted at the same time as a preliminary hearing, they should be held as soon as practicable. Consistent with Salerno, defendants in Arizona state proceedings are entitled to counsel during Simpson hearings, may testify and present information, and may cross-examine witnesses. The trial court must only admit evidence that is material to the question at issue. The state bears the burden of persuasion to prove that “all of the evidence, fully considered by the court, makes it plain and clear to the understanding, and satisfactory and apparent to the well-guarded, dispassionate judgment of the court that the accused committed” one of the designated nonbailable offenses. The proof presented must be “substantial” but need not meet the “proof beyond a reasonable doubt” standard. And the judge is required to make factual findings in making the determination.
C. The Release Decision in Bailable Offenses
If an offense is bailable, Arizona law requires judicial officers to release a defendant pending trial “on his own recognizance or on the execution of bail in an amount specified by the judicial officer,” who makes pretrial release determinations. By statute, the commissioner, magistrates, and judges who make these decisions are statutorily required to consider fifteen factors, including the nature and circumstances of the offense, the defendant’s criminal history, their family ties, and their appearance history. To assist them in making their release determination, some Arizona counties, including the most populous county, Maricopa, use bail schedules to some degree; however, as we subsequently explain, the use of bail schedules appears to have been restricted by recent reforms.
As a result of changes made to the Arizona Rules of Criminal Procedure, the structure of bail hearings has changed significantly over the last three years. Most of these changes were implemented in a December 2016 order by Chief Justice Scott Bales, which went into effect in April 2017. The most important change is that there is now a presumption of release on recognizance in most cases.Release on recognizance is required unless “the court determines that such a release will not reasonably assure the defendant’s appearance” as required or protect other persons or the community from risk posed by the person. In such a case, only the “least onerous [condition or] conditions” are to be imposed.
By statute, courts can impose a wide variety of conditions. For instance, the court may place the defendant in the custody of another person; restrict the defendant’s travel, associates, and place of abode; and prohibit certain activities, including consuming alcohol or drugs or possessing a deadly weapon. In some circumstances, the court may also order electronic monitoring or prohibit contact with victims. The court retains the ability to amend its order to include additional or different conditions. And if, after a hearing, a court finds that a defendant willfully violated conditions of release, release may be revoked entirely.
The Rules of Criminal Procedure differentiate between mandatory and discretionary conditions of release and between monetary and non-monetary conditions. In determining whether a discretionary condition is necessary, the court must now consider the results of a risk assessment tool approved by the Supreme Court. In addition, if an indigent defendant is detained pretrial after being charged with a misdemeanor, they are “entitled to a court-appointed attorney . . . for the limited purpose of determining release conditions.”
Although cash bail has not been abolished in Arizona, it is now much more difficult to impose money bail, particularly a bond that is unaffordable, because Rule 7.3 limits the imposition of “a monetary condition that results in unnecessary pretrial incarceration solely because the defendant is unable to pay the [bond].” And even if a court imposes a monetary condition, it must impose “the least onerous” of the types of bonds, with priority given to an unsecured appearance bond. Commercial bond remains an option for defendants and their families, but as we discuss in Part V, the changes made by the Supreme Court will likely result in a serious hit to the bail bond industry.
D. Pretrial Decisions in Practice
Unfortunately, there is little information on how pretrial release hearings play out in practice. Unlike in other jurisdictions, there are no court monitoring programs in Arizona, and little research has been conducted. Further, scant data on jail populations or other indicators are made available at the state or county level. However, the story of one defendant is illustrative and shows that the reforms may not be working as intended.
Yolanda Suniga was in an abusive relationship with her boyfriend. She had called the police several times to report domestic violence incidents, but nothing was ever done about it. Her relationship with this violent man ultimately led Suniga to be arrested in March of 2018 on charges stemming from the following two incidents.
On November 4, 2017, Suniga and her boyfriend argued in their cHiar while parked at a gas station. He punched her in the face and threatened to take her children, who were in the back seat of the car. She responded by exiting the vehicle and getting a pair of scissors from the trunk of the car. She then used the scissors to scratch her boyfriend. She called 911. Police observed injuries consistent with her having been hit. An independent witness heard Suniga crying for help and observed her holding her face before she retrieved the scissors. And surveillance video confirmed that the boyfriend had hit Suniga in the face. Importantly, the surveillance footage also showed that when Suniga exited the vehicle, the boyfriend moved from the passenger seat “to the driver’s seat—an indication that he was acting on his threats of kidnapping” the children. Nonetheless, Suniga was eventually charged with aggravated assault, while the boyfriend was not arrested; instead, he was listed as the victim, even though he was the initial aggressor.
On December 10, 2017, Suniga stood on a light rail platform after exiting a train. She was with her abusive boyfriend at the time. She began talking to a man who, like Suniga, had just exited the train. She accepted a cigarette from the man. According to the State, Suniga eventually pointed at the man’s backpack. Surveillance video of the event in question shows this characterization to be questionable; it could just as easily be interpreted as showing that Suniga merely dropped her left arm during conversation. Immediately after this ambiguous gesture, Suniga’s boyfriend moved behind her and began arguing with the other man. The boyfriend then moved toward the man. As the victim noticed the boyfriend’s actions, he began to move toward Suniga’s left side. They bumped into each other. She stepped out of the way as the situation between the man and her boyfriend escalated. The boyfriend hit the victim and then took his backpack. The boyfriend was subsequently charged with dangerous aggravated assault and aggravated robbery. Based on these ambiguous facts, even though in an interview with police the boyfriend denied that Suniga had aided him in any way, she was eventually charged with aggravated robbery via accomplice liability.
At her initial appearance, during which she was not represented by counsel, the court set Suniga’s bail at $35,000. This amount was set at $5,000 for the assault with the scissors and $30,000 in consideration of Suniga’s aggravated robbery charge and the boyfriend’s violent aggravated assault and robbery, which were erroneously listed as charges against Suniga on the documents upon which the court relied during the initial appearance. Suniga could not afford to post bail, so she was held in pretrial detention while she was pregnant with her abusive boyfriend’s child at the time.
Arizona law required that Suniga be released on her own recognizance unless the court determined that such release would not reasonably assure the defendant’s appearance or protect the victim.Suniga had no other open cases, was not on probation or parole at the time, and had no prior convictions. Importantly, Suniga was not a flight risk. Prior to her arrest, she had been actively engaging in court-ordered drug testing, parenting classes, and supervised visitation to regain custody of her children. Thus, she had a demonstrated interest in staying in Arizona, where she had lived her whole life and could continue to live locally with her mother, to complete these requirements to get back her children. Moreover, clean drug tests are a specifically listed factor favoring pretrial release. But even if the court felt bond was necessary to ensure her future appearance, the court was prohibited from imposing a monetary condition that resulted “in unnecessary pretrial incarceration solely because the defendant is unable to pay the imposed monetary condition.” Yet, despite being presented with this additional information, including the jail’s own records showing that Suniga was losing weight while incarcerated, two different judges denied motions to modify Suniga’s bail amounts and order her released under the supervision of pretrial services, even with electronic monitoring.
Even though the aggravated assault with the scissors was defensible on self-defense and crime prevention grounds, and the accomplice robbery charge was defensible on a “mere presence” theory, Suniga ultimately opted to waive her defenses and pleaded guilty because she was losing weight while in jail, which, in turn, was adversely affecting the health of the child she was carrying.Nonetheless, she was forced to give birth while in custody because the state demanded incarceration as a term of the plea. She never got to touch her newborn because the child was immediately taken away from her. She served two more months during which time both she and baby were deprived of bonding opportunities. Even with the associated term of incarceration, pleading guilty proved to be a quicker way for Suniga to get out of custody than it would have been if the two cases had been tried.
Suniga’s case demonstrates that financial conditions of release are still imposed on Arizonans who cannot afford their bail amounts. In Suniga’s case, this is especially egregious for six reasons. First, less restrictive alternatives could have been ordered, especially since Suniga had a stable residence with her mother in which she could have resided if she had been released. Second, between the time of her initial appearance and the pendency of the motions to modify her conditions of release, there had been a substantial and appropriate reduction in her charges. Specifically, Suniga’s boyfriend was indicted on violent crime charges associated with the robbery, whereas Suniga was ultimately only indicted as a non-violent accomplice. Third, the weight of the evidence on the accomplice robbery charge was not great in light of her mere presence at the scene, her boyfriend’s statement indicated she was uninvolved, and the surveillance video showed no active participation on her part. Fourth, the weight of the evidence on the aggravated assault charge was also not great, especially in light of the fact that she was acting to prevent her boyfriend from absconding with her children. Fifth, Suniga posed no danger to her assault “victim,” as her boyfriend was being held in pretrial detention on charges stemming from his acts of violence during the robbery, and the robbery victim was living out of town. And finally, Suniga’s pregnancy coupled with the fact that she only had ties to the Phoenix area created added incentives to comply with release conditions if they had been ordered. She would not have been at risk of non-appearance if released to pretrial services. Clearly, the reforms already implemented are not enough, and additional changes are needed to improve the state of pretrial justice in Arizona.
V. Assessing Reforms to Date
Early reforms in Arizona focused on the expansion of pretrial services agencies across the state and piloting actuarial risk assessment tools. Prior to 2014, only half of Arizona counties had pretrial services agencies. In 2014, the Arizona Supreme Court amended the Code of Judicial Administration to include a section focused on evidence-based pretrial services.  This section provided guidelines for establishing and operating pretrial services and also ordered judges in counties with pretrial services agencies to begin assessing each individual defendant’s actual risk of danger or flight before setting bail. A number of jurisdictions had piloted the Public Safety Assessment (“PSA”), a pretrial actuarial risk assessment tool developed by the Laura and John Arnold Foundation, and based on its success, it was approved for all superior courts in Arizona in June 2015.
Two years later, former Arizona Chief Justice Scott Bales established a Task Force, one of the key goals of which was to “recommend best practices for making release decisions that protect the public but do not keep people in jail solely for the inability to pay bail.” The Task Force was also asked to identify technology and best practices about court date notifications to reduce failures to appear.
The Task Force issued a report that included sixty-five detailed recommendations, many of which relate specifically to pretrial justice. The Task Force emphasized the harms caused by pretrial detention, even short periods, and highlighted the empirical data showing that “detaining low- and moderate-risk defendants causes harm and higher rates of new criminal activity.” To date, the legislature has taken no action on any of the Task Force recommendations. However, the Arizona Supreme Court approved a number of key changes to the state’s Rules of Criminal Procedure to effect some of the recommendations.
A. Reducing the Use of Money Bail and Commercial Bond
Although the Task Force recommended that Arizona eliminate cash bail, the state has not yet done so. But perhaps anticipating that its recommendation would not be accepted, the Task Force also recommended that Arizona “[e]liminate the requirement for cash surety to the greatest extent possible.” In support of this recommendation, the Task Force criticized the heavy reliance on commercial bail agencies and recommended that where a financial bond was imposed, preference be given to cash deposited with the court.
These recommendations were implemented, and as previously described, courts must now release defendants charged with bailable offenses on their own recognizance unless doing so would not assure their appearance or would fail to protect the community. Moreover, in cases where a court determines that a monetary condition of release is necessary, it must impose the “least onerous” type of bond. If followed, these changes would help those without the financial resources to pay bail, who often either remain in jail pending trial or have to utilize the services of a bail bond agent. Most importantly, a move toward release on recognizance and unsecured bonds should reduce the number of people in jail pending trial. In addition, it would reduce reliance on the commercial bail industry, which, across the nation, acts as a gatekeeper for more than two million defendants released on bail each year, earning it an estimated $2.4 billion annually.
Arizona defines a bail bond as “any contract that is executed by a surety insurer for the release of a person who is . . . confined for any actual or alleged violation of . . . criminal law where the released person’s attendance in court . . . is guaranteed.” Bail bonds agents serve as a broker between a defendant and an insurance company, which underwrites their appearance. Commercial bail agents typically charge a non-refundable, up-front fee of 10% of the bail amount. In theory, bond agents are supposed to pay courts the complete bail amount when defendants do not subsequently appear for judicial proceedings. But nationally, bail forfeiture is exceedingly rare, with some states reporting that commercial bail companies actually pay only 1.7% to 12% of forfeitures they owe.
But even in the relatively uncommon circumstances when forfeiture payments occur, bail agents rarely pay the full bond amount when a defendant fails to appear. Some courts are willing to accept dramatically lower amounts in satisfaction of a bail bond. Moreover, even after such payments, many states broadly authorize courts to remit all or a portion of the funds to sureties. As a result, the insurance companies that back bail bonds tend to pay less than 1 percent of their bail-related revenue on forfeiture losses—a remarkably low figure in comparison to property and auto insurance companies, which usually pay between 40 and 60 percent of premium revenue on claim losses.
Moreover, the commercial bail industry has a long history of corruption. Even as the industry has professionalized, these problems have not gone away. In addition, while many bail agents see themselves as providing a service to people who might otherwise remain in jail, a recent study of a bail bond company in a large urban county illustrates the predatory practices of bail companies.
Not surprisingly, the changes to Arizona’s bail rules have negatively impacted Arizona’s commercial bond industry, which correctly foresaw a decline in their business. One bail agent described calls for service as having “significantly dropped” since the rule went into effect. Given the problems with the industry, we see this decline as a positive step.
B. Individualized Hearings that Consider a Defendant’s Ability To Pay
The Task Force also criticized the use of uniform bond schedules and recommended that release decisions be individualized and based on the risk posed by a defendant. The Supreme Court amended Rule 7.3(b)(2) to implement this recommendation. Individualized decision-making like this is crucial to ensure that only individuals who pose a real risk of failing to appear or to public safety are detained pending trial. Although Arizona has never required judges to follow bail schedules, which typically are a list of offenses with a presumptive bond amount for each offense, many counties do have them, particularly for traffic and low-level offenses. Their impact has not been studied in Arizona specifically; research from California and Texas suggest that judges follow them reflexively.
The use of bail schedules often goes hand in hand with a failure to consider an individual’s ability to pay the bail set in their case. And arguably this is why so many people across the United States are sitting in jail: simply because they cannot afford to get out. Arizona’s decision to require judges to consider ability to pay in deciding whether to set money bail, and if so, how much bail to set, is a move in the right direction.
VI. More Reforms Are Needed
A. Cash Bail Must Be Abolished
The Task Force’s report recognized that “money bond is not required to secure appearance of defendants” and, therefore, recommended that it be completely eliminated. Without money determining release decisions, a system where poor people are locked up simply because they are poor should disappear. Such a system, which still exists in Arizona despite the efforts of the state’s supreme court, is problematic for a number of reasons.
First, using money to determine release means that many poor defendants are locked up simply because they or their families cannot afford even a small amount of bail. Most of these people, accused of relatively minor offenses, likely pose little to no risk of either failing to appear or reoffending. As the Arizona Supreme Court’s Task Force noted, “The traditional money bail system has little to do with actual risk, and expecting money to effectively mitigate risk, especially risk to public safety, is historically unfounded.”
Second, if an individual is unable to pay to be released from jail, there is a litany of negative consequences that can endure beyond the resolution of the initial case. Pretrial detention deprives individuals of their liberty and subjects them to harsh conditions of confinement before they are convicted of the crime(s) with which they are charged. Although the Supreme Court has determined that pretrial detention is “regulation,” not “punishment,” given how many defendants experience jail, it is difficult to see that distinction as anything other than a semantic one. “[E]ven a brief stay in jail can be destructive to individuals, their families, and entire communities.” Jails “are often overcrowded, with poor physical and mental health care,” and they have “fewer educational and vocational opportunities than prisons.” Further, “many jails have higher death rates than the national average, and the suicide rate in jails is three times the rate in prison.” The high-profile cases of Sandra Bland and Kalief Browder are illustrative. Bland committed suicide after three days in detention following a highly publicized arrest in 2015 stemming from a traffic stop. She could not pay the $500 bail that would have allowed her to be released pending trial. Browder was incarcerated in Rikers Island for three years, nearly two of which were in solitary confinement. Correctional staff and jail inmates alike repeatedly assaulted Browder. He tried to commit suicide several times while incarcerated. A year after he was released, after prosecutors dropped the charges against him, he killed himself.
Given the conditions pretrial detainees experience, it is not surprising that many plead guilty simply to get out of jail. Some, like Yolanda Suniga, do so out of desperation even though the charges they face are defensible; some others even plead guilty to charges of which they are completely innocent. And fifty years of research clearly demonstrates the negative effects of pretrial detention on case outcomes. Controlling for factors like offense type, charge severity, and criminal history, people who are detained are more likely to be convicted (typically as a result of guilty pleas), more likely to be sentenced to jail or prison (rather than to probation), and to receive a longer incarceration sentence. The impact of these outcomes on mass incarceration in the United States is beyond doubt.And the consequences of this sorry state of affairs fall overwhelmingly on the poor and minorities, exacerbating existing inequalities in the criminal justice system.
The American Bar Association Standards for Pretrial Release recommend the use of “unsecured” bonds or release on conditions that will help assure court appearance. When conditions of release are imposed, it is essential that judges order the least restrictive conditions that are reasonably calculated to assure appearance at subsequent court dates and to protect public safety. Pretrial supervision is much more effective than secured bonds in achieving both of these goals. Indeed, pretrial detention has only minimal impact on court appearances. To be sure, “[p]retrial detainees almost always appear in court, whereas it is undisputed that some defendants granted pretrial release do not. But there are fewer in the latter group than most people think.” And although studies show that marginal released defendants are 15.6% more likely to fail to appear, pretrial detention remains a costly and burdensome way of securing appearance. Consider that Washington, D.C., has all but abolished money bail. There,
[A] mere 4% of people have financial conditions attached to their release and roughly 88% are freed with no monetary obligations. Of those released, 90% made all scheduled court appearances and 98% were not rearrested for a violent crime. A well-funded pretrial agency, which monitors defendants and reminds them of upcoming court dates, helps this system function smoothly, as does the robust public defender program that supplies representation for indigent defendants during pretrial hearings.
A bill pending before the Arizona State Legislature would codify this recommendation by amending Arizona Revised Statutes title 13, section 3967(A) to mandate “release[ ] pending trial on [the defendant’s own] recognizance or on the execution of [an unsecured bond].” Although this would be a step in the right direction, the additional reforms are needed—especially in light of Arizona’s dubious reliance on a risk assessment instrument that appears to be racially biased.
B. Risk Assessments Are Not Necessarily the Answer
With the notable exceptions of New York and Ohio, all states that have reformed their pretrial justice systems in the last few years have relied on actuarial risk assessment tools. The Fair Justice for All Report recommended a system of pretrial release in which “[l]ow-risk defendants are released on their own recognizance or with unsecured appearance bonds”; “[m]oderate-risk defendants are released to Pretrial Services with specific release conditions imposed to mitigate the risks presented”; and “[h]igh-risk defendants are held in custody as preventive detention when no condition or combination of conditions of release can reasonably assure the appearance of the person [at trial] or will endanger the safety of any person or the community.” At first blush, this approach seems appealing. But deeper examination of the assessment of risk suggests this is a deeply flawed approach.
Actuarial risk assessment tools purport to be objective tools that accurately predict whether someone poses a particular type of risk. In the pretrial context, the risks being measured are typically the risk of failing to appear and the risk of committing a new offense while on release. Although tools like these are appealing to policymakers who wish to give judges an ostensibly objective way of making pretrial release decisions, as many scholars have concluded, actuarial pretrial risk assessment instruments (“APRAIs”) suffer from many problems.
It is important to remember that to be useful, APRAIs need to predict distinct types of risk, including a defendant’s risk of failing to appear at a court hearing, a defendant’s risk of flight from a jurisdiction, and a defendant’s risk of committing a new crime while awaiting trial. Unfortunately, most tools fail to do this. Some tools “only provide a [general] pretrial failure risk score, which is a combined outcome of missing a court appearance or being rearrested.” Further, even those tools that do provide separate scores for risk of failure to appear and risk to public safety do not adequately distinguish between different kinds of risk. For example, there are different types of failure to appear ranging from true flight by fugitives who intentionally evade justice, to the far more common inadvertent failure to appear that occurs when defendants forget about a hearing or are unable to get transportation to court.
In addition, when a person is evaluated as high risk of reoffending or a high public safety risk, it is not always clear what that means. Sometimes that is measured as probability of arrest, even for a minor offense. Other times it is defined as a risk of arrest for a violent crime, as is the case with the Arnold Foundation’s Public Safety Assessment (“PSA”) that is used in Arizona. However, “because pretrial violence is exceedingly rare,” APRAIs “cannot identify people who are more likely than not to commit a violent crime.” Most people designated as a high risk to public safety do not get arrested for a violent crime while awaiting trial. Indeed,
[i]f these tools were calibrated to be as accurate as possible, then they would predict that every person was unlikely to commit a violent crime while on pretrial release. Instead, risk assessments sacrifice accuracy and generate substantially more false positives (people who are flagged for violence but do not go on to commit a violent crime) than true positives (people who are flagged for violence and do go on to be arrested for a violent crime).
APRAIs are only as good as the data on which they are built. It is well documented, however, that there are significant problems with the criminal history data that are used to predict future behavior. The inclusion of criminal history in risk assessment instruments “produces a ‘ratchet effect’ by oversampling from minority populations who are already disproportionately represented in the criminal justice system” not only because their communities are overpoliced and, as a result, they are arrested at higher rates, but also because people of color face more serious charges, are convicted at higher rates, and face more severe sentences than White people. Sandra Mayson referred to this problem as “bias in, bias out.”
[I]f the thing that we undertake to predict—say arrest—happened more frequently to black people than to white people in the past data, then a predictive analysis will project it to happen more frequently to black people than to white people in the future. The predicted event . . . is thus the key to racial disparity in prediction.
Thus, it is not surprising that the overestimation of risk mentioned above is more pronounced for monitories.
In sum, APRAIs are tainted by systemic bias. Moreover,
they do not predict the future behavior of any particular defendant. Nonetheless, their actuarial identification of people who share characteristics with other “high-risk” or “low-risk” defendants is clearly better than chance—which is really all statistical significance means. But . . . chance is not the outcome to which risk assessments should be compared. Rather, the proper comparison point is how risk assessments perform when compared to individualized determinations of risk that judges have made for time immemorial without the assistance of risk scores . . . . [But] very little research has ever conducted such comparisons, and the handful of studies that have done so are seriously methodologically flawed. For example, . . . a study of the federal pretrial risk assessment instrument that evaluated pretrial officers’ assessments of “a single fictitious case” . . . had “no actual case outcomes with which to measure ‘accuracy.’” . . . Surely, more research is needed to discern which APRAIs are both valid and reliable. Similarly, we need more empirical research on whether, and to what extent, particular risk assessment algorithms can increase public safety while mitigating, rather than perpetuating, racial, ethnic, and socioeconomic disparities. Algorithms honed by machine learning may help to actualize such a goal one day.
The Task Force recognized the concerns about racial bias with APRAIs in general but noted that “no issues have been found with the PSA instrument to date.” It therefore recommended not only the continued use of the instrument in the superior courts of Arizona but also its expanded use in the state’s lower courts. Although some jurisdictions in the United States have seen positive outcomes after introducing APRAIs, these tools are rarely introduced in isolation; thus, it is impossible to tell whether they or the other changes instituted led to the increased release rates and other reported positive outcomes. But even more importantly, since the time the Task Force made its recommendation on the use of the PSA, more empirical data have revealed the biased nature of such tools. It now seems clear that current checklist-style APRAIs, like the PSA, fall short of protecting public safety and equal justice under law. Thus, as previously stated, even the PJI—which had previously been at the forefront of advocating the adoption of APRAIs—has abandoned that position because these instruments deepen racial inequities.
We join the PJI in calling for Arizona to stop using APRAIs to determine risk and instead address “people’s needs . . . through community-based support.”
C. Expanding the Right to Counsel
The Sixth Amendment “guarantees a defendant the right to have counsel present at all ‘critical’ stages of the criminal proceedings.” But the precise moment at which adversarial proceedings begin is not clear-cut.
In Rothgery v. Gillespie County, the U.S. Supreme Court held the Sixth Amendment right to counsel “applies at the first appearance before a judicial officer at which a defendant is told of the formal accusation against him and restrictions are imposed on his liberty.” But that conclusion is not clear because “the Rothgery Court went to great lengths to distinguish between the technical attachment of the right to counsel at the initial appearance and the separate question of whether counsel must actually be present during that proceeding.” Thus, defendants may not have counsel present during their first appearance in court, but rather have counsel appointed after a probable cause hearing.
Like Yolanda Suniga, defendants in Arizona are routinely unrepresented by counsel at their initial appearances in court. But such representation is essential to obtaining release.
A study in Baltimore, for example, evaluated an eighteen-month experiment in which lawyers provided representation to indigent arrestees accused of nonviolent crimes. Those who were represented by counsel were 2.5 times more likely to be released on their own recognizance (34 percent) than those who were unrepresented (13 percent). For those whom bail was ordered, counsel was successfully able to have the amount of bail set to something the accused could afford for 59 percent of clients, compared to only 14 percent of the unrepresented being successful at securing affordable bail for themselves. Additional studies in other jurisdictions have reached similar conclusions.
Dozens of other advocacy groups and scholars agree with the American Bar Association’s long-standing recommendation that counsel be “provided to the accused as soon as feasible and, in any event, after custody begins, at appearance before a committing magistrate, or when formal charges are filed, whichever occurs earliest.” And the money saved by reducing the number of people jailed pending trial could be used to pay for providing representation during initial appearances.
The Task Force recognized the importance of legal advice, noting the negative impacts of pretrial detention and highlighting the problem with assigning counsel based on the possibility of a jail sentence:
In misdemeanor matters, a prosecutor may charge a person and specify that jail time will not be requested as part of the sentence. Such a declaration makes the defendant ineligible for a court-appointed lawyer. If such a person is required to post a financial bond but cannot pay it, the unconvicted defendant likely will remain incarcerated for a longer period than if he or she were found guilty of the offense. This certainly constitutes incarceration and should make the person eligible for the appointment of an attorney.
Nonetheless, it declined to recommend the automatic provision of counsel at initial appearance and instead recommended the following: “Encourage the presence of court-appointed counsel and prosecutors at initial appearance hearings to assist the court in determining appropriate release conditions and to resolve misdemeanor cases.”
The Arizona Supreme Court did not even go this far, although it did slightly expand the right to counsel, giving indigent defendants an attorney “for the limited purpose of determining release conditions at or following the initial appearance, if the defendant is detained after a misdemeanor charge is filed.” However, this change did not help Yolanda Suniga, and her case is illustrative. An attorney could have corrected the record at her initial appearance such that bail could have been set based on the charges she faced, rather than those also attributive to her boyfriend. Conditioning the right to an attorney on a defendant already having been detained fails to protect defendants from the negative consequences of even a short period of detention.
D. Court Notification
All counties in Arizona should adopt the Task Force’s series of recommendations to promote court appearances, including implementing a court reminder system, using either email, text, or phone. Automated court notification programs appear to be a reliable method of reducing failures to appear. “For example, a study in Jefferson County, Colorado, found that a phone call reminder reduced the rate of failure to appear from 21 to 12 percent. Similarly, Multnomah County, Oregon, saw a 31 percent reduction in failures to appear when it implemented an automated reminder system.” More recently, a randomized control trial of court reminders in New York found that when individuals who were issued desk appearance tickets (a form of custodial arrest used mostly in misdemeanor cases that allows the arrestee to leave the precinct and return for arraignment at a later date) were given a reminder phone call, they were more likely to appear at arraignment. The study concluded that overall, failures to appear “at the first court appearance was reduced by as much as 47 percent for those individuals who received both a three day and same day phone call . . . compared to those who did not receive any phone-call reminder.” The Justice for All Task Force noted similar success in Arizona courts. Given this research, other jurisdictions are moving in this direction.
There are some important reforms that can be made to improve the state of pretrial justice in Arizona. First and foremost, monetary bail should be eliminated because there is no evidence that money bail effectively incentivizes court appearances or protects public safety, as the District of Columbia’s system evidences. Its most recent data show that almost 94% of defendants remained in the community while awaiting their case resolution, 88% of whom made all scheduled court appearances, and less than 14% of defendants were rearrested for a new crime while awaiting trial, most of these for non-violent offenses.
Second, it is important that APRAIs be reliable and properly validated, but most are not. Moreover, APRAIs should not be used if they perpetuate racial, ethnic, and socio-economic disparities.The PSA that Arizona uses fails in this regard, as do most other APRAIs that use variables so statistically correlated with race, ethnicity, and low socio-economic status that they are proxies for race, ethnicity, and poverty.
Third, even if Arizona continues to use pretrial risk assessment instruments, they ought not be used mechanistically. We need to use “better, more sophisticated decision-making with regard to what ‘high risk’ means and how to respond to it using the least-restrictive-means principle.” 
[T]he default response to risk need not be coercion. What if it were support instead? Risk, after all, is neither intrinsic nor immutable. It is possible to change the odds. In the short term, a supportive, needs-oriented response to risk would mitigate the immediate racial impact of prediction. If a high-risk classification meant greater access to support and opportunities, a higher false-positive rate among black defendants would be less of a concern. In the long term, a supportive response to risk might help to counter the social conditions that drive crime, for the benefit of all.
Probation agents and others who monitor defendants released on pretrial conditions can and should be the key to providing such responses.
Fourth, “conditions imposed on released defendants are limited to those that will increase the likelihood that they will appear at scheduled court dates.” Court date notification tools are particularly effective at reducing failures to appear. In addition, it is important that any costs of pretrial supervision not be passed on to defendants, as this would defeat the purpose of eliminating money bail.
In closing, “with alternative methods to manage risk, money can be virtually eliminated from the bail process without negatively affecting court appearance rates or public safety.” Doing so “would be a major step toward a system that does not punish people simply for being poor”—especially in a state like Arizona in which one in seven people live below the poverty level—the fifth highest poverty rate in the United States. As the Task Force recognized, “[T]hose without means should not be disparately punished because they are poor.”
* The authors thank Michael Serota, Visiting Assistant Professor and an Associate Deputy Director of the Academy for Justice at Arizona State University’s Sandra Day O’Connor College of Law, for inviting them to be a part of this special issue of the Arizona State Law Journal. The authors also gratefully acknowledge the research assistance of Andi Lefor, former law student at Arizona State University’s Sandra Day O’Connor College of Law who served as a Note and Comment Editor of the Arizona State Law Journal, provided to us for this Article. Additionally, we thank the attorneys at the Maricopa County Public Defender’s Office, whose names are being kept in confidence, for sharing their insights on pretrial injustice in the county’s court system. And, finally, the authors convey their sincere appreciation to Yolanda Suniga for her willingness to allow us to share her story in this Article.
** Professor and Associate Director, Arizona State University School of Criminology and Criminal Justice; Affiliate Professor, Sandra Day O’Connor College of Law. Dr. Fradella earned a B.A. in psychology from Clark University, a master’s in forensic science, and a J.D. from The George Washington University, and a Ph.D. in justice studies from Arizona State University.
*** Associate Professor, California State University, Long Beach, School of Criminology, Criminal Justice, and Emergency Management. Dr. Scott-Hayward earned a Bachelor of Civil Laws (International) from University College Dublin, a master’s in social science from the University of Chicago, and a Ph.D. in law and society from New York University.
. See, e.g., Benjamin Barber, Cash Bail System Faces Challenges in Southern Communities, Facing S. (Nov. 22, 2019), https://www.facingsouth.org/2019/11/cash-bail-system-faces-challenges-southern-communities [https://perma.cc/JE8G-K8YB] (“Several states including Arizona, California, and New Jersey have already reformed or entirely eliminated cash bail . . . .”); John Buntin, The Fight To Fix America’s Broken Bail System, Governing (Oct. 2017), http://www.governing.com/topics/public-justice-safety/gov-bail-reform-texas-new-jersey.html [https://perma.cc/2DAX-TZH9] (“Arizona . . . lowered the state’s pretrial incarceration rate by encouraging judges to utilize nonjail alternatives and pay closer attention to defendants’ financial circumstances when setting bail bonds.”).
. See Administrative Order, No. 2016-16 (Ariz. Mar. 3, 2016), http://www.azcourts.gov/Portals/22/admorder/Orders16/2016-16.pdf [https://perma.cc/BZG3-QGK6]; see also Task Force on Fair Just. for All, Report and Recommendations of the Task Force on Fair Justice for All (2016) [hereinafter Fair Justice for All Report], https://www.azcourts.gov/Portals/74/TFFAIR/Reports/FINAL%20FairJustice%20Aug%2012-final%20formatted%20versionRED%20(002).pdf [https://perma.cc/PA99-FVDE].
. See Jacob Kang-Brown et. al., Vera Inst. of Just., People in Jail in 2019, at 1 (2019), https://www.vera.org/downloads/publications/people-in-jail-in-2019.pdf [https://perma.cc/XP4X-XJXD]. The Vera Institute of Justice did not, however, report the percentages of those persons who were convicted or held in pretrial detention.
. Nick Pinto, The Bail Trap, N.Y. Times Mag. (Aug. 13, 2015), https://www.nytimes.com/2015/08/16/magazine/the-bail-trap.html [https://perma.cc/8USL-ZZWP]; see also Christine S. Scott-Hayward & Henry F. Fradella, Punishing Poverty: How Bail and Pretrial Detention Fuel Inequalities in the Criminal Justice System 129–51 (2019) (describing the consequences of pretrial detention).
. Scott-Hayward & Fradella, supra note 14, at 5 (citing Mary T. Phillips, N.Y.C. Crim. Just. Agency, A Decade of Bail Research in New York City, Final Report 116 (2012), https://www.prisonpolicy.org/scans/DecadeBailResearch12.pdf [https://perma.cc/EX3W-TY8Q]); see also Pinto, supra note 14 (“Faced with the prospect of going to jail for want of bail, many defendants accept plea deals instead.”). For a summary of the history of bail from the Anglo-Saxon period to the present, see Scott-Hayward & Fradella, supra note 14, at 11–28; see also June Carbone, Seeing Through the Emperor’s New Clothes: Rediscovery of Basic Principles in the Administration of Bail, 34 Syracuse L. Rev. 517, 574 (1983); Timothy R. Schnacke et al., Pretrial Just. Inst., The History of Bail and Pretrial Release (2010), https://cdpsdocs.state.co.us/ccjj/Committees/BailSub/Handouts/HistoryofBail-Pre-TrialRelease-PJI_2010.pdf [https://perma.cc/4DFL-ZK7M]. For an in-depth history of the origins and purposes of bail, see William F. Duker, The Right to Bail: A Historical Inquiry, 42 Alb. L. Rev. 33, 119 (1977).
. See, e.g., Pugh v. Rainwater, 572 F.2d. 1053, 1057 (5th Cir. 1978) (ruling the plaintiffs’ claims were moot, the court noted that while “[u]tilization of a master bond schedule provides speedy and convenient release for those who have no difficulty in meeting[ ] its requirements . . . [t]he incarceration of those who cannot, without meaningful consideration of other possible alternatives, infringes on both due process and equal protection requirements”); see also Nicholas P. Johnson, Cash Rules Everything Around the Money Bail System: The Effect of Cash-Only Bail on Indigent Defendants in America’s Money Bail System, 36–37 Buff. Pub. Int. L.J. 29, 97 (2019).
. See, e.g., ODonnell v. Harris County, 251 F. Supp. 3d 1052 (S.D. Tex. 2017), aff’d in part, rev’d in part, 892 F.3d 147 (5th Cir. 2018) (holding bail schedules violated equal protection); Thompson v. Moss Point, No. 1:15cv182LG-RHW, 2015 WL 10322003 (S.D. Miss. Nov. 6, 2015) (entering declaratory judgment against secured bail schedules as applied to those who could not afford to pay); Walker v. City of Calhoun, No. 4:15-cv-0170-HLM, 2017 WL 2794064 (N.D. Ga. June 16, 2017) (granting preliminary injunction against detention of people arrested for misdemeanor or ordinance violations who are unable to afford bond), vacated and remanded to, 901 F.3d 1245 (11th Cir. 2018), cert. denied, 139 U.S. 1446 (2019); cf. Pierce v. City of Velda City, No. 4:15-cv-570-HEA, 2015 WL 10013006 (E.D. Mo. June 3, 2015) (entering declaratory judgment, pursuant to a settlement agreement, that holding someone in pretrial detention because the person is too poor to post a monetary bond violates equal protection); In re Humphrey, 228 Cal. Rptr. 3d 513, 541 (Cal. Ct. App. 2018) (refraining from holding bail schedules unconstitutional, the court cautioned that “unquestioning reliance upon the bail schedule without consideration of a defendant’s ability to pay, as well as other individualized factors bearing upon his or her dangerousness and/or risk of flight, runs afoul of the requirements of due process for a decision that may result in pretrial detention”), rev. granted, sua sponte, 417 P.3d 769 (Cal. 2018), restoring precedential effect of Part III of the appellate decision, 2020 WL 5269846 (Cal. Aug. 26, 2020) (en banc) (mem.); see also Christine S. Scott-Hayward & Sarah Ottone, Punishing Poverty: California’s Unconstitutional Bail System, 70 Stan. L. Rev. Online 167, 178 (2018), https://review.law.stanford.edu/wp-content/uploads/sites/3/2018/04/70-Stan.-L.-Rev.-Online-167-Scott-Hayward-and-Ottone.pdf [https://perma.cc/R8S7-9ZEP] (arguing that California’s reliance on bail schedules is unconstitutional).
. Scott-Hayward & Fradella, supra note 14, at 147 (citing Supervision Costs Significantly Less than Incarceration in Federal System, U.S. Cts. (July 18, 2013), http://www.uscourts.gov/news/2013/07/18/supervision-costs-significantly-less-incarceration-federal-system [https://perma.cc/W3DA-KUCJ]).
. See, e.g., Anna Aizer & Joseph J. Doyle, Jr., Juvenile Incarceration, Human Capital, and Future Crime: Evidence from Randomly Assigned Judges, 130 Q.J. Econ. 759, 763 (2015) (finding that incarceration has a criminogenic effect using data from a large, urban county in the United States); Arpit Gupta, Christopher Hansman & Ethan Frenchman, The Heavy Costs of High Bail: Evidence from Judge Randomization, 45 J. Legal Stud. 471, 492 (2016) (reporting that assigning money bail, which increased the likelihood of pretrial detention, increased the likelihood of future criminal behavior between 6% and 9%); Christopher T. Lowenkamp, Marie VanNostrand & Alexander Holsinger, The Hidden Costs of Pretrial Detention 19–24 (2013), https://craftmediabucket.s3.amazonaws.com/uploads/PDFs/LJAF_Report_hidden-costs_FNL.pdf [https://perma.cc/66YD-GSSQ]; Michael Mueller-Smith, The Criminal and Labor Market Impacts of Incarceration 2–3, 24 fig.4 (Aug. 18, 2015) (unpublished manuscript), http://sites.lsa.umich.edu/mgms/wp-content/uploads/sites/283/2015/09/incar.pdf [https://perma.cc/FFG3-4G7R] (reporting that pretrial detainees experienced a 6.0% and 6.7% increase in the likelihood of being charged with a new misdemeanor and felony, respectively); Rafael Di Tella & Ernesto Schargrodsky, Criminal Recidivism After Prison and Electronic Monitoring, 121 J. Pol. Econ. 28 (2013) (reporting, using data from Argentina, reduced reoffending rates for defendants released pretrial on electronic monitoring compared to those held in pretrial detention). But cf. Charles E. Loeffler, Does Imprisonment Alter the Life Course? Evidence on Crime and Employment from a Natural Experiment, 51 Criminology 137, 154 (2013) (concluding post-conviction incarceration is not criminogenic).
. Paul Heaton, Sandra Mayson & Megan Stevenson, The Downstream Consequences of Misdemeanor Pretrial Detention, 69 Stan. L. Rev. 711, 762 (2017) (reporting empirical data from more than 380,000 defendants in Harris County, Texas, that pretrial detention in jail has a criminogenic effect).
. Id. at 762–66, 793–94. Note, however, that Heaton and colleagues did not find any long‑term impact on public safety; specifically, they could identify no impact on either rearrests or new convictions within two years after the bail hearing.
. Arthur W. Pepin, Conf. of State Ct. Adm’rs, Evidence-Based Pretrial Release 2 (2013), https://www.ncsc.org/__data/assets/pdf_file/0015/23802/Evidence-Based-Pre-Trial-Release-Final.pdf [https://perma.cc/MAG5-L6YC].
. The case of Robert Durst, who was profiled in the HBO documentary The Jinx, serves as a notable example. The Jinx: The Life and Deaths of Robert Durst (HBO Documentary Films television series 2015); see also Elizabeth Yuko, Robert Durst Murder Trial Begins in Los Angeles, Rolling Stone (Feb. 11, 2020, 12:46 PM), https://www.rollingstone.com/culture/culture-news/robert-durst-hbo-jinx-murder-trial-starts-los-angeles-950794/ [https://perma.cc/3X3R-YQXK] (noting that Durst was never charged in his wife’s disappearance and was only convicted of evidence-tampering and bail-jumping in the killing and dismemberment of his neighbor). For a broader discussion of such risks, see Pepin, supra note 26, at 2.
. Alexa Van Brunt & Locke E. Bowman, Toward a Just Model of Pretrial Release: A History of Bail Reform and a Prescription for What’s Next, 108 J. Crim. L. & Criminology 701, 703–05 (2018) (referring to recent efforts away from monetary bail as the “‘third wave’ of bail reform in the United States” in light of initial reforms in the 1960s and the emergence of pretrial detention in the name of public safety that occurred in the 1980s).
. E.g., ODonnell v. Harris County, 251 F. Supp. 3d 1052 (S.D. Tex. 2017), aff’d in part, rev’d in part, 892 F.3d 147 (5th Cir. 2018). For a listing of dozens of cases filed by reformers, see Challenging the Money Bail System, C.R. Corps, https://www.civilrightscorps.org/work/wealth-based-detention [https://perma.cc/35E5-43F2] and Ending American Money Bail, Equal Just. Under L., https://equaljusticeunderlaw.org/money-bail-1 [https://perma.cc/Q8VB-8W38].
. See, e.g., Larry Hannan, Bail Reform Embraced by Cook County State’s Attorney, Appeal (July 5, 2017), https://theappeal.org/bail-reform-embraced-by-cook-county-states-attorney-3b9f45839ee/ [https://perma.cc/J73G-VPT6]; Andrew Parent, Philly DA Larry Krasner To End Cash Bail for Two Dozen Non-Violent Crimes, Philly Voice (Feb. 21, 2018), https://www.phillyvoice.com/da-larry-krasner-wont-seek-cash-bail-certain-offenses/ [https://perma.cc/QLR9-7TEG]; Evan Sernoffsky, San Francisco DA Chesa Boudin Ends Cash Bail for All Criminal Cases, S.F. Chron. (Jan. 29, 2020, 8:43 AM), https://www.sfchronicle.com/crime/article/San-Francisco-DA-Chesa-Boudin-ends-cash-bail-for-14996400.php [https://perma.cc/F9YS-NB7R].
. 2014 N.J. Laws 467 (codified at N.J. Stat. Ann. § 2A, c. 162 and scattered sections of § 2B (2020)); Winning Bail Reform in New Jersey, Drug Pol’y All., https://www.drugpolicy.org/new-jersey/winning-bail-reform [https://perma.cc/GNB8-Y9T7].
. Glenn A. Grant, N.J. Admin. Off. of the Cts., 2018 Report to the Governor and the Legislature 45 (2019), https://njcourts.gov/courts/assets/criminal/2018cjrannual.pdf [https://perma.cc/D6J6-76EP] (reporting that since pretrial justice reform went into effect in New Jersey, crime rates in the state fell, particularly for violent crime; the rate of alleged new criminal activity remained “virtually the same”; court appearance rates dropped by approximately 3%, but overall dispositions remained relatively constant, suggesting no significant increases in absconding; and most notably, jail populations decreased by nearly 45%, translating into 750,000 fewer jail beds in a year, while racial and ethnic disparities in pretrial detention were reduced significantly, especially among Black women).
. Peter Krouse, Ohio Supreme Court Proposes Bail Reforms that Don’t Include Risk Assessments, Cleveland.com (Jan. 25, 2020), https://www.cleveland.com/news/2020/01/ohio-supreme-court-proposes-bail-reforms-that-dont-include-risk-assessments.html [https://perma.cc/8BY7-PFJQ].
. Bales was replaced as chief justice in June 2019 by Robert Brutinel, who has committed to continue bail reform. Howard Fischer, New State Chief Justice To Continue Push for ‘Bail Reform,’ Ariz. Capitol Times (June 23, 2019), https://azcapitoltimes.com/news/2019/06/23/new-state-chief-justice-to-continue-push-for-bail-reform/ [https://perma.cc/38VH-U674].
. As we explained our book, “the Supreme Court has never decided a case that required it to rule squarely on whether the Excessive Bail Clause restrains only the federal government or, alternatively, if it also applies to the states.” Scott-Hayward & Fradella, supra note 14, at 23 (citing Scott W. Howe, The Implications of Incorporating the Eighth Amendment Prohibition on Excessive Bail, 43 Hofstra L. Rev. 1039, 1043 (2015)).
However, in McDonald v. City of Chi., Ill.—the 2010 case in which the Court held the Second Amendment was incorporated and made applicable to the states by the Due Process Clause of the Fourteenth Amendment—the Court, in dicta, included a footnote in which it listed the provisions of the Bill of Rights that had previously been held to be incorporated against the states.
Id. (citing McDonald v. City of Chicago, 561 U.S. 742, 765 n.12 (2010)). But because that case did not directly address the issue, this “oddity . . . does not inspire confidence that the Justices had reached a momentous civil-rights decision.” Howe, supra, at 1084. Accordingly, law professor Samuel Wiseman posited that McDonald is the case that incorporated the Excessive Bail Clause. Samuel R. Wiseman, McDonald’s Other Right, 97 Va. L. Rev. Brief 23, 26 (2011).
. Scott-Hayward & Fradella, supra note 14, at 19 (citing United States v. Feely, 25 F. Cas. 1055, 1057 (Va. Cir. Ct. 1813)); see also Duker, supra note 15, at 69 (citing Workman v. Cardwell, 338 F. Supp. 893, 898 (N.D. Ohio 1972), aff’d in part, vacated in part, 471 F.2d 909 (6th Cir.); Reynolds v. United States, 80 S. Ct. 30, 32 (1959); DeAngelis v. South Carolina, 330 F. Supp. 889, 894 (D.S.C. 1971)).
. Ex parte Milburn, 34 U.S. (9 Pet.) 704, 710 (1835); see also Duker, supra note 15, at 69 (citing United States v. Foster, 79 F. Supp. 422, 423 (S.D.N.Y. 1948) (stating that “[c]learly, it is not the function of bail to prevent the commission of crimes between indictment and trial”).
. Scott-Hayward & Fradella, supra note 14, at 27 (footnotes omitted); see, e.g., Arthur R. Angel et al., Preventive Detention: An Empirical Analysis, 6 Harv. C.R.-C.L. L. Rev. 300, 369–70 (1971); John S. Goldkamp, Questioning the Practice of Pretrial Detention: Some Empirical Evidence from Philadelphia, 74 J. Crim. L. & Criminology 1556, 1587 (1983); Rudy A. Haapanen, Selective Incapacitation and the Serious Offender: A Longitudinal Study of Criminal Career Patterns(1990). See generally Bernard E. Harcourt, Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age (2007).
. Duker, supra note 15, at 90. But see Howe, supra note 37, at 1058 (arguing that the excessive bail clause would only have merited incorporation if it “confers a right to bail in some circumstances and regulates the permissible purposes of bail and, thus, the measure of excessiveness”).
. See, e.g., Richard A. Webster, The Judge Whose Bail Requirements Leave Cash-Strapped Defendants in Jail, Guardian (Feb. 3, 2020, 6:00 AM), https://www.theguardian.com/us-news/2020/feb/03/the-judge-whose-bail-requirements-leave-cash-strapped-defendants-in-jail [https://perma.cc/VK5C-6VEG].
. Ariz. Const. art. II, § 22(A)(1). This constitutional provision also contains a prohibition on bail “when the proof is evident or the presumption great” that the defendant committed sexual assault, sexual conduct with a minor under the age of fifteen, or child molestation. But these provisions were all declared unconstitutional on due process grounds between 2017 and 2018. See State v. Wein, 417 P.3d 787, 796 (Ariz. 2018) (invalidating provision making sexual conduct with a minor under fifteen years old nonbailable); Simpson v. Miller, 387 P.3d 1270, 1278–79 (Ariz. 2017) (invalidating provision making sexual misconduct with a minor under the age of fifteen nonbailable); Chantry v. Astrowsky, 395 P.3d 1114, 1115 (Ariz. Ct. App. 2017) (invalidating provision making child molestation unbailable).
. Ariz. Rev. Stat. Ann. § 13-3967(A) (2020). The state’s discretionary pretrial release scheme has withstood constitutional scrutiny for compliance with the Fourteenth Amendment’s Due Process Clause. Samiuddin v. Nothwehr, 404 P.3d 232, 238 (Ariz. 2017).
. § 13-3967(B). These factors include the views of the victim; the nature and circumstances of the offense charged; whether the accused has a prior arrest or conviction for a serious offense or violent or aggravated felony; evidence that the accused poses a danger to others in the community; the results of a risk or lethality assessment in a domestic violence charge that is presented to the court; the weight of evidence against the accused; the accused’s family ties, employment, financial resources, character, and mental condition; the results of any drug test submitted to the court; whether the accused is using any substance if its possession or use is illegal; whether the accused violated certain laws prohibiting possession and use of methamphetamine; the length of residence in the community; the accused’s record of arrests and convictions; the accused’s record of appearance at court proceedings or of flight to avoid prosecution or failure to appear at court proceedings; whether the accused has entered or remained in the United States illegally; and whether the accused’s residence is in this state, in another state, or outside the United States. Id.
. Two public defenders in Maricopa County, the state’s most populous county, who were interviewed by the authors argue that commissioners refer to an unofficial schedule when making initial determinations in county jails. Telephone Interview with Anonymous, Att’y, Maricopa Cnty. Pub. Def. Off. (Feb. 13, 2020) [hereinafter Interview]; Telephone Interview with Anonymous, Att’y, Maricopa Cnty. Pub. Def. Off. (Feb. 19, 2020). Moreover, some counties have adopted advisory bail schedules for misdemeanor and traffic offenses to promote “greater consistency in the setting of bond amounts for the same or similar criminal offenses and in the imposition of sanctions for civil traffic offenses by the municipal and justice courts.” Admin. Order, No. 2015-002 (Ariz. Super. Ct. Jan. 5, 2015), https://www.superiorcourt.maricopa.gov/SuperiorCourt/AdministrativeOrders/AdminOrders/Admin%20Order%202015-002.pdf [https://perma.cc/UNP7-9B4Q].
. See Order Amending Rules 3, 6, and 7, and Abrogating Forms 6 and 7, Rule 41, Rules of Criminal Procedure, and Substituting New Forms 6 and 7 in Their Place, No. R-16-0041, (Ariz. Dec. 14, 2016), https://www.azcourts.gov/Portals/20/2016%20December%20Rules%20Agenda/R_16_0041.pdf [https://perma.cc/L6JP-YULM].
. See Ray Stern, Arizona’s Low-Income, Low-Level Crime Suspects To Be Released Without Bail, Phx. New Times (Feb. 22, 2017, 8:30 AM), https://www.phoenixnewtimes.com/news/arizonas-low-income-low-level-crime-suspects-to-be-released-without-bail-9102227 [https://perma.cc/XZM4-T386]; Megan Cassidy, Bond Companies ‘Extremely Worried’ as Arizona Moves Away from Cash Bail Bonds, AZCentral (June 22, 2017, 1:21 PM), https://www.azcentral.com/story/news/local/arizona/2017/06/21/arizona-courts-back-away-cash-bail-system-bond-companies-worried/400209001/ [https://perma.cc/5GE9-B58H].
. Indeed, unlike other states that make court data easily accessible online, conducting empirical research on judicial processes in Arizona is quite difficult because “each county can determine what degree of access is provided, with policy determination[s] often left to the county clerk, often because individual counties maintain their own case-flow management systems.” Council for Ct. Excellence, Remote Public Access to Electronic Court Records: A Cross-Jurisdictional Review for the D.C. Courts 9 (2017), http://www.courtexcellence.org/uploads/publications/RACER_final_report.pdf [https://perma.cc/8NXS-HWJC].
. All of the information reported in this section regarding Ms. Suniga’s case comes from a telephone interview with an attorney employed by the Maricopa Public Defender’s Office (Feb. 13, 2020), as well as from pleadings filed in Ms. Suniga’s case. See Indictment, State v. Suniga, No. CR2018-113466-001 DT (Ariz. Super. Ct. Apr. 3, 2018) [hereinafter Suniga Indictment] (attaching Mesa Police Department reports by Officer T. Cook in case number 20173080155) (on file with authors)); Indictment, State v. Griffen, No. CR2018-113466-001 (Ariz. Super. Ct. Mar. 22, 2018) [hereinafter Griffen Indictment] (attaching Tempe Police Department reports by Officer E. Santana, J. Forsen, and R. Valencia in case number TE 2017-146820) (on file with authors)); Motion To Modify Conditions of Release, State v. Suniga, CR2018-113466-001 DT (Ariz. Super. Ct. Apr. 12, 2018), ECF No. 9253984 [hereinafter Apr. Motion] (on file with authors); Motion To Modify Conditions of Release, State v. Suniga, CR2018-113466-001 DT (Ariz. Super. Ct. June 18, 2018), ECF No. 9442594 [hereinafter June Motion] (on file with authors).
. Minute Entry at 2, State v. Suniga, No. CR2018-001602 (Ariz. Super. Ct. Apr. 30, 2018) [hereinafter Apr. Minute Entry]; see Minute Entry at 2, State v. Suniga, No. CR2018-001602 (Ariz. Super. Ct. July 2, 2018).
. Interview, supra note 80; Minute Entry at 1, State v. Suniga, No. CR2018-001602 (Ariz. Super. Ct. Aug. 16, 2018); see, e.g., Janice F. Bell et al., Jail Incarceration and Birth Outcomes, 81 J. Urban Health 630 (2004), https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3455933/pdf/11524_2006_Article_331.pdf [https://perma.cc/LSZ8-EN83] (“Relative to controls, women incarcerated during pregnancy had progressively higher odds of low birth weight and preterm birth through age 39 years.”). For a review of the law concerning the sentencing and incarceration of pregnant women, see Christine S. Scott-Hayward, Correctional and Sentencing Law Commentary: Regulating Pregnancy, 56 Crim. L. Bull. 553 (2020).
. Pretrial Just. Ctr. for Cts., Snapshot of Pretrial Justice Reform: Arizona 1 (2015), https://www.ncsc.org/__data/assets/pdf_file/0021/1596/pretrial-justice-brief-3-az-final.ashx.pdf [https://perma.cc/8AWH-HB8B].
. Ariz. Code Judicial Admin. § 5-201 (2014), adopted by Admin. Order, No. 2014-12 (Ariz. Jan. 10, 2014), https://www.azcourts.gov/Portals/22/admorder/Orders14/2014-12.pdf [https://perma.cc/5WGY-GE8D].
. Admin. Order, supra note 5. Chief Justice Robert Brutinel, the Republican who succeeded Democrat Bales, stated the following about his intention to continue on the bail reform path: “[T]he courts have created ‘risk-analysis’ instruments to help judges determine who can be released. What’s next . . . is automating that system to make it available . . . to all levels of the court system.” Fischer, supra note 36.
. Id. at 32–33. When cash bail is deposited with a court, the full amount will be refunded if the defendant appears at subsequent court dates; however, if the defendant does not appear for other court dates, the money may be forfeited, the court may issue a warrant for the defendant’s arrest, and the defendant may even be charged with a separate crime for failing to appear. Ariz. Rev. Stat. Ann. § 13-3968 (2020).
. § 13-3969; see also Fair Justice for All Report, supra note 5, at 32. Note that Arizona law sets forth a number of restrictions on bail bonds agents. See §§ 13-3968 to -3985. Many of the restrictions were put into place to curb the powers of bounty hunters who track down bail jumpers for bail bonds agents. See Ann L. Merry, S.B. 1257: Arizona Regulates Bounty Hunters, 31 Ariz. St. L.J. 229 (1999).
. Bryce Covert, This Deep Red State Just Ended Cash Bail, Appeal (Jan. 10, 2018), https://theappeal.org/alaska-ends-its-reliance-on-money-bail [https://perma.cc/GKB7-S3G5]. According to a 2017 report, the more than 25,000 individual bail bonds businesses in the United States are typically underwritten by “under-the-radar subsidiaries” of large multinational insurance companies. Color of Change & ACLU Campaign for Smart Just., $elling Off Our Freedom: How Insurance Corporations Have Taken over Our Bail System 9 (2017) [hereinafter $elling Off Our Freedom], https://www.aclu.org/sites/default/files/field_document/059_bail_report_2_1.pdf [https://perma.cc/GSE9-F7NV]. Indeed, only nine large insurers underwrite a majority of the roughly $14 billion in bail bonds issued in the United States each year. Id. at 7.
. State of N.J. Comm’n of Investigation, Inside Out: Questionable and Abusive Practices in New Jersey’s Bail-Bond Industry (2014), https://www.nj.gov/sci/pdf/BailReportSmall.pdf [https://perma.cc/YBJ9-2H94]; Off. of the Legis. Auditor Gen., A Performance Audit of Utah’s Monetary Bail System (2017), https://le.utah.gov/audit/17_01rpt.pdf [https://perma.cc/78VV-ZWCL].
. See $elling Off Our Freedom, supra note 161, at 36; Travis Fain, Conspiracy Theories, Criminal Investigations Plentiful in NC Bail Bonds World, WRAL.com (July 13, 2018, 3:20 PM), https://www.wral.com/conspiracy-theories-criminal-investigations-plentiful-in-nc-bail-bonds-world/17333869/ [https://perma.cc/DF5B-SRW2].
. Joshua Page, Desperation and Service in the Bail Industry, Contexts (June 19, 2017), https://contexts.org/articles/bail/ [https://perma.cc/SF3C-9HK9]; see also Chris Blaylock, I’m a Bail Bondsman Who’s Tired of Being Demonized, Marshall Project (June 1, 2017, 10:00 PM), https://www.themarshallproject.org/2017/06/01/i-m-a-bail-bondsman-who-s-tired-of-being-demonized [https://perma.cc/Z2G2-PZVY].
. Joshua Page, Victoria Piehowski & Joe Soss, A Debt of Care: Commercial Bail and the Gendered Logic of Criminal Justice Predation, Russell Sage Found. J. Soc. Scis. (2019), https://www.rsfjournal.org/content/rsfjss/5/1/150.full.pdf [https://perma.cc/2RGY-HS99].
. See, e.g., Ottone & Scott-Hayward, supra note 51, at 34–35; see also ODonnell v. Harris County, 251 F. Supp. 3d 1052, 1110 (S.D. Tex. 2017) (finding that even though the bond schedule was just one of a series of factors to be considered, hearing officers followed it in 90% of cases), aff’d in part, rev’d in part, 892 F.3d 147 (5th Cir. 2018).
. Fair Justice for All Report, supra note 5, at 32 (quoting Timothy R. Schnacke, Nat’l Inst. of Corrections, Money as a Criminal Justice Stakeholder: The Judge’s Decision To Release or Detain a Defendant Pretrial 27 (2014), https://nicic.gov/money-criminal-justice-stakeholder-judge-s-decision-release-or-detain-defendant-pretrial [https://perma.cc/P3U5-AZ8R]).
. Ram Subramanian et al., Vera Inst. of Just., Incarceration’s Front Door: The Misuse of Jail in America 2 (2015), http://www.safetyandjusticechallenge.org/wp-content/uploads/2015/01/incarcerations-front-door-report.pdf [https://perma.cc/3558-UL9A]; see also Gerstein v. Pugh, 420 U.S. 103, 114 (1975) (noting that “pretrial confinement may imperil the suspect’s job, interrupt his source of income, and impair his family relationships”).
. Scott-Hayward & Fradella, supra note 14, at 132; see also Vera Inst. of Just., The State of Justice Reform: The State of Jails: A Shifting Landscape (2018), https://www.vera.org/state-of-justice-reform/2018/the-state-of-jails [https://perma.cc/SC2H-5373] (citing Zhen Zeng, U.S. Dep’t of Just., Bureau of Just. Stat., Jail Inmates in 2016, at 1 (2018), https://www.bjs.gov/content/pub/pdf/ji16.pdf [https://perma.cc/A56U-R3NR] (documenting overcrowding)).
. Scott-Hayward & Fradella, supra note 14, at 132; (citing Margaret Noonan, U.S. Dep’t of Just., Bureau of Just. Stat., Mortality in Local Jails, 2000–2014—Statistical Tables (2016), https://www.bjs.gov/content/pub/pdf/mlj0014st.pdf [https://perma.cc/63LV-ZL6R] (documenting deaths in jail custody)).
. See George C. Klein, On the Death of Sandra Bland: A Case of Anger and Indifference, 8 J. Police Emergency Response 3 (2018), https://journals.sagepub.com/doi/pdf/10.1177/2158244018754936 [https://perma.cc/BJ9B-NNF6].
. See, e.g., John Raphling, Opinion, Plead Guilty, Go Home. Plead Not Guilty, Stay in Jail, L.A. Times (May 17, 2017, 4:00 AM), https://www.latimes.com/opinion/op-ed/la-oe-raphling-bail-20170517-story.html [https://perma.cc/XK4F-RZE6]; Joshua Vaughn, Pleading Guilty To Get Out of Jail, Appeal (June 6, 2019), https://theappeal.org/franklin-county-pennsylvania-bail-jail-high-jail-population/ [https://perma.cc/7M7F-SNNK].
. See, e.g., Elisha Fieldstadt, An Innocent Man Pleaded Guilty to a Drug Charge To Get Out of Jail. It’s More Common than You Think., NBC News (Oct. 16, 2019, 1:46 PM), https://www.nbcnews.com/news/us-news/innocent-man-pleaded-guilty-drug-charge-get-out-jail-it-n1067321 [https://perma.cc/4PL7-Q42V]; Penelope Gibbs, Why Do the Innocent Plead Guilty?, Just. Gap (Sept. 11, 2019, 8:27 AM), https://www.thejusticegap.com/why-do-the-innocent-plead-guilty/ [https://perma.cc/BDS8-6W55].
. See Léon Digard & Elizabeth Swavola, Vera Inst. of Just., Justice Denied: The Harmful and Lasting Effects of Pretrial Detention 1 (2019), http://www.safetyandjusticechallenge.org/wp-content/uploads/2019/04/Justice-Denied-Evidence-Brief.pdf [https://perma.cc/2BJG-XS6C].
. See, e.g., Jacob Kang-Brown et al., Vera Inst. of Just., The New Dynamics of Mass Incarceration (2018), https://www.vera.org/downloads/publications/the-new-dynamics-of-mass-incarceration-report.pdf [https://perma.cc/D3ZH-PL7Z].
. See Stephen Demuth, Racial and Ethnic Differences in Pretrial Release Decisions and Outcomes: A Comparison of Hispanic, Black, and White Felony Arrestees, 41 Criminology 873, 901 (2003); Heaton et al., supra note 21Error! No bookmark name given., at 745–51; Charles M. Katz & Cassia C. Spohn, The Effect of Race and Gender on Bail Outcomes: A Test of an Interactive Model, 19 Am. J. Crim. Just. 161 (1995); Megan T. Stevenson, Distortion of Justice: How the Inability To Pay Bail Affects Case Outcomes, 37 J.L. Econ. & Org. 511 (2018). See generally Scott-Hayward & Fradella, supra note 14, passim.
. Fair Justice for All Report, supra note 5, at 35 (citing Christopher T. Lowenkamp & Marie VanNostrand, Exploring the Impact of Supervision on Pretrial Outcomes(2013), https://craftmediabucket.s3.amazonaws.com/uploads/PDFs/LJAF_Report_Supervision_FNL.pdf [https://perma.cc/3JXX-9XGM]).
. Will Dobbie, Jacob Goldin & Crystal S. Yang, The Effects of Pretrial Detention on Conviction, Future Crime, and Employment: Evidence from Randomly Assigned Judges, 108 Am. Econ. Rev. 201, 226 (2018); see also Shima Baradaran Baughman, Costs of Pretrial Detention, 97 B.U. L. Rev. 1, 27–30 (2017) (estimating that fine-tuning pretrial detention decisions through careful cost-benefit analyses could save up to $78 billion annually).
. See Cassie Miller, The Two-Tiered Justice System: Money Bail in Historical Perspective, S. Poverty L. Ctr. (June 6, 2017), https://www.splcenter.org/20170606/two-tiered-justice-system-money-bail-historical-perspective[https://perma.cc/94NU-SX4C].
. Sarah L. Desmarais & Evan M. Lowder, Pretrial Risk Assessment Tools: A Primer for Judges, Prosecutors, and Defense Attorneys 5 (2019), http://www.safetyandjusticechallenge.org/wp-content/uploads/2019/02/Pretrial-Risk-Assessment-Primer-February-2019.pdf [https://perma.cc/KG5P-CG9T].
. See id.; Open Letter from Chelsea Barabas et al., Technical Flaws of Pretrial Risk Assessments Raise Grave Concerns (July 16, 2019), https://dam-prod.media.mit.edu/x/2019/07/16/TechnicalFlawsOfPretrial_ML%20site.pdf [https://perma.cc/7W8M-Q6WA].
. For example, nearly thirty prominent researchers from MIT, Harvard, Princeton, New York University, the University of California, Berkeley, Columbia, and the Algorithmic Justice League signed an open statement of concern about dangers of using APRAIs as part of pretrial justice reform efforts. See Open Letter from Chelsea Barabas et al., supra note 207, at 1; see also, e.g., Bernard E. Harcourt, Risk as a Proxy for Race: The Dangers of Risk Assessment, 27 Fed. Sent’g Rep. 237 (2015); see also Ted Gest, Civil Rights Advocates Say Risk Assessment May ‘Worsen Racial Disparities’ in Bail Decisions, Crime Rep. (July 31, 2018), https://thecrimereport.org/2018/07/31/civil-rights-advocates-say-risk-assessment-may-worsen-racial-disparities/ [https://perma.cc/WB9Z-5MJ2]. For an in-depth examination of the pretrial risk assessment instruments and their limits, see Scott-Hayward & Fradella, supra note 14, at 77–128.
. Technical Flaws, supra note 207, at 3; see also Note, Bail Reform and Risk Assessment: The Cautionary Tale of Federal Sentencing, 131 Harv. L. Rev. 1125 (2018); Nicole Gonzalez Van Cleve & Lauren Mayes, Criminal Justice Through “Colorblind” Lenses: A Call To Examine the Mutual Constitution of Race and Criminal Justice, 40 Law & Soc. Inquiry 406 (2015).
. Mayson, supra note 218, at 2224 n.23 (“The computer-science idiom is ‘garbage in, garbage out,’ which refers to the fact that algorithmic prediction is only as good as the data on which the algorithm is trained.”).
. Matthew DeMichele, Peter Baumgartner, Michael Wenger, Kelle Barrick, Megan Comfort & Shilpi Misra, The Public Safety Assessment: A Re-Validation and Assessment of Predictive Utility and Differential Prediction by Race and Gender in Kentucky 49–52 (2018), https://pdfs.semanticscholar.org/d3e5/3bfc7c64d7fa1eef6bf1cfbc17d2d5a37e1b.pdf?_ga=2.202141335.700504761.1599377281-1811880916.1599377281 [https://perma.cc/9X4L-3KSM].
. Scott-Hayward & Fradella, supra note 14, at 127 (emphasis added) (quoting Sonja B. Starr, The New Profiling: Why Punishing Based on Poverty and Identity Is Unconstitutional and Wrong, 27 Fed. Sent’g Rep. 229, 232 (2015)); see id.(criticizing J.C. Oleson, Scott W. VanBenschoten, Charles R. Robinson & Christopher T. Lowenkamp, Training To See Risk: Measuring the Accuracy of Clinical and Actuarial Risk Assessments Among Federal Probation Officers, 75 Fed. Prob. 52 (2011)); see also Jon Kleinberg, Himabindu Lakkaraju, Jure Leskovec, Jens Ludwig & Sendhil Mullainathan, Human Decisions and Machine Predictions, 133 Q.J. Econ. 237, 241 (2018) (analyzing more than 758,000 pretrial releasees in New York City and concluding that properly designed algorithms that utilize machine learning—a more powerful and complex tool for prediction than current risk assessment instruments—“can be a force for racial equity”).
. Fair Justice for All Report, supra note 5, at 36. Indeed, to address the potential for bias in the PSA, the Task Force simply “[e]ncourage[d] the Arnold Foundation to conduct periodic reviews to revalidate the Public Safety Assessment (PSA) tool as to its effect on minority populations.” Id. at 37.
. Pretrial Just. Inst., Updated Position on Pretrial Risk Assessment Tools 1 (2020), https://www.pretrial.org/wp-content/uploads/Risk-Statement-PJI-2020.pdf [https://perma.cc/CVL5-SSFQ] (“We now see that pretrial risk assessment tools . . . can no longer be a part of our solution for building equitable pretrial justice systems.”).
. Id. (first citing Douglas L. Colbert, Ray Paternoster & Shawn Bushway, Do Attorneys Really Matter? The Empirical and Legal Case for the Right of Counsel at Bail, 23 Cardozo L. Rev. 1719, 1752–53 (2002); and then citing Ernest J. Fazio Jr., Sandra Wexler, Thomas Foster, Michael J. Lowy, David Sheppard & Juliet A. Musso, Nat’l Inst. of Just., Early Representation by Defense Counsel Field Test: Final Evaluation Report 210 (1984), https://www.ncjrs.gov/pdffiles1/Digitization/97596NCJRS.pdf [https://perma.cc/4KJK-JG8J]); see also Colbert et al., supra, at 1720; cf. Marian R. Williams, The Effect of Attorney Type on Bail Decisions, 28 Crim. Just. Pol’y Rev. 3, 13 (2017) (reporting that although Florida defendants “with public defenders were more likely to be denied bail and less likely to be released, they also benefited from lower bail amounts and non-financial release options.”).
. Am. Bar Ass’n, ABA Standards for Criminal Justice Providing Defense Services 77 (3d ed. 1992); see also, e.g., Alexander Bunin, The Constitutional Right to Counsel at Bail Hearings, 31 Crim. Just. 23 (2016); Colbert et al., supra note 235, at 1782–83; Sandra G. Thompson, Do Prosecutors Really Matter? A Proposal To Ban One-Sided Bail Hearings, 44 Hofstra L. Rev. 1161 (2016).
. The Const. Project, Don’t I Need a Lawyer? Pretrial Justice and the Right to Counsel at First Judicial Bail Hearing 40–41 (2015), https://constitutionproject.org/wp-content/uploads/2015/03/RTC-DINAL_3.18.15.pdf [https://perma.cc/85VS-CF5C].
. Scott-Hayward & Fradella, supra note 14, at 73 (citing Timothy R. Schnacke, Michael R. Jones & Dorian M. Wilderman, Increasing Court-Appearance Rates and Other Benefits of Live-Caller Telephone Court-Date Reminders: The Jefferson County, Colorado, FTA Pilot Project and Resulting Court Date Notification Program, 48 Ct. Rev. 86, 89 (2012); $elling Off Our Freedom, supra note 161, at 47).
. New Text Message Reminders for Summons Recipients Improves Attendance in Court and Dramatically Cuts Warrants, NYC (Jan. 24, 2018), https://www1.nyc.gov/office-of-the-mayor/news/058-18/new-text-message-reminders-summons-recipients-improves-attendance-court-dramatically [https://perma.cc/BTW3-PL2L].
. Russell Ferri, N.Y.C. Crim. Just. Agency, Desk Appearance Tickets and Appearance Rates—the Benefits of Court Date Reminders 6 (July 2019), https://www.nycja.org/assets/DAT-Notification-Research-Brief-07152019.PDF [https://perma.cc/XJX4-WZCQ].
. See, e.g., Chao Xiong, In Hennepin County, Text and E-Mail Reminders of Court Dates Reduce Number of Warrants, Star Trib. (Oct. 4, 2019, 7:52 PM), http://www.startribune.com/hennepin-county-ereminder-program-for-court-rolls-out-statewide/562200402/ [https://perma.cc/LJE4-X5ZD].
. However, if Arizona retains a money bail system, it is important to ensure that defendants are guaranteed a determination of their ability to pay the money bail set. Such procedures should include notice to the defendant that bail determinations must be individualized, a presumption that defendants who are indigent are unable to pay money bail, a hearing on the record at which the defendant has the right to counsel, as well as the right to prompt review. See Crim. Just. Pol’y Program, Harvard L. Sch., Moving Beyond Money: A Primer on Bail Reform (2016), http://cjpp.law.harvard.edu/assets/FINAL-Primer-on-Bail-Reform.pdf [https://perma.cc/C3MC-F4RL].
. Pretrial Servs. Agency for D.C., Washington, DC Pretrial Facts and Figures (2018), https://www.psa.gov/sites/default/files/Pretrial%20Facts%20and%20Figures%20-%20Updated%203.2018.pdf [https://perma.cc/US8E-LKHJ].
. Renata Cló, Poverty Rate Falls in Arizona, but Still Exceeds National Average, Cronkite News (Sept. 13, 2018), https://cronkitenews.azpbs.org/2018/09/13/poverty-rate-falls-in-arizona-but-still-exceeds-national-average/ [https://perma.cc/3U7H-BXG3]; Griselda Zetino, Arizona Is Home to the Nation’s Fifth-Highest Poverty Rates, KTAR News (Jan. 22, 2018, 11:29 AM), https://ktar.com/story/1910747/arizona-is-home-to-the-nations-fifth-highest-poverty-rates/ [https://perma.cc/6ENN-94ZU].