Case: 17-20333 Document: 00514472379 Page: 1 Date Filed: 05/14/2018 Exhibit 1 Case: 17-20333 Document: 00514472379 Page: 2 Date Filed: 05/14/2018 DECLARATION OF JUDGE DARRELL JORDAN1 I. Introduction 1. My name is Darrell William Jordan. I was elected by the citizens of Harris County, Texas to serve as Judge of Harris County Criminal Court at Law No. 16. On January 1, 2017, I began my term as Judge of Court No. 16. When I took office, I became a Defendant in the above-captioned lawsuit, which had been filed against the other fifteen judges and my predecessor. 2. Before I joined the legal profession, I joined the Texas National Guard. Today I am a major in the Army Reserves, and as I write this, I have been called up to active duty, serving in Austin until September. 3. Before becoming a judge, I was a criminal defense lawyer. In the eight years before being elected to the bench, I represented over 1,000 clients. 60-70% were misdemeanor cases, most of them in Harris County. As a lawyer, I appeared before each of the other County Criminal Court at Law judges. As a judge, I have continued to observe and be informed of practices of my judicial colleagues. I have seen the impact in the County Criminal Courts of two momentous events in 2017: the preliminary injunction in this case and Hurricane Harvey, both of which led to significant changes to the County’s pretrial system. While some of the confusion the Harris County legal system has experienced over the past year was the unavoidable consequence of the worst natural disaster in Houston history, much was self-inflicted by judges who are fixated on seeing that the federal court’s order fails. 4. I am familiar with the claims made by Judges 1–13 and 15 (“the Fourteen Judges”) that the preliminary injunction issued in this case “has created a law enforcement and public safety crisis” and that “[r]oughly half of all unsecured release bonds issued pursuant to the preliminary injunction have been forfeited because the arrestee failed to appear for his or her court date.” I believe these claims to be false or, at best, misleading. 5. In my opinion, ever since the preliminary injunction went into effect, the Fourteen Judges have taken steps to sabotage the federal injunction and to manipulate the bond-forfeiture statistics to make it look like people released on unsecured bond pursuant to the federal order are failing to appear in droves, and that the federal court order is “wreaking havoc” on the County. In truth, no such thing has happened. The elevated bond-forfeiture rate for people released on unsecured bail, to the extent it exists, is readily explained by the Fourteen Judges’ intentional policy decisions, many of which were made by majority vote. II. Harris County Criminal Court at Law Judges Have Discretion to Decide When to Forfeit Bonds A. A “Bond Forfeiture” Is Not a “Failure to Appear” 1 Submitted in ODonnell v. Harris County, Case No. 16-cv-01414 (S.D. Tex. 2016), No. 17-20333 (5th Cir. 2017). 1 Case: 17-20333 Document: 00514472379 Page: 3 Date Filed: 05/14/2018 6. Before addressing the judges’ policies, it is crucial to understand the terminology that the Fourteen Judges are using. The Fourteen Judges report an elevated “bond forfeiture” rate for people released on unsecured bond.2 7. But the term “bond forfeiture” is not the same as “failure to appear.” The judges do not cite, let alone analyze, failures to appear. “Bond forfeiture” is merely the rough proxy that the Fourteen Judges use for “failure to appear”—and it is easy to manipulate. It is my understanding that the judges and the County began to track failures to appear after the preliminary injunction went into effect. To my knowledge, however, the Judges and the County have not reported the failure-to-appear rates publicly or in court filings. 8. The bond-forfeiture numbers are manipulable because each individual judge decides when a defendant must appear in court and each of us has complete discretion to decide whether any given instance of non-appearance warrants us entering a “bond forfeiture” into the system. There are no rules governing our decisions as judges about whether and under what circumstances to forfeit a bond. 9. Moreover, there is no requirement that we apply the same policies to people released on surety, cash, unsecured, and personal bonds. And in practice, judges’ policies vary depending on the type of bond the person was released on. 10. Texas law is yet another reason that “bond forfeitures” are a very poor proxy for failures to appear. State statutes set forth a number of procedural rights and broad substantive defenses to forfeiture. Surety companies are aware of these laws and, as experienced, routine players in our system, they often hire lawyers to represent them in forfeiture proceedings. Because the statutory defenses are broad, surety companies are often able to evade forfeiture, even when the defendant did in fact fail to appear. 11. As a result of these defenses and surety companies’ ability to take advantage of them by hiring lawyers and fighting the forfeiture, the absence of a forfeiture of a surety bond does not mean that the person released on that surety bond appeared in court for all of her court appearances. 12. By contrast, people released on unsecured bonds or personal bonds typically do not know that such defenses to forfeiture exist, partly because the judges do not inform them of their rights. So the people who benefit from defenses to forfeiture are overwhelming people released on surety bonds. 13. For these reasons, “bond forfeiture” rates are a very poor proxy for non-appearance rates. 2 When a person is released pursuant to the federal court’s preliminary injunction order, the bond type is referred to locally as an “unsecured bond.” People also sometimes call these bonds “BROR” (which stands for “bail reform ordered release” and is the acronym the Sheriff developed for his computer system when the order went into effect in June 2017) or “Sheriff’s bonds.” Throughout this declaration, I will use the term “unsecured bond” to describe the bonds issued to people who are released pursuant to the federal court’s preliminary injunction in this case. When a person is ordered released by a judge or magistrate, the bond type is referred to locally as a “personal bond.” Personal bonds are also technically unsecured bonds: personal bonds likewise do not require an up-front payment, but the person will be required to pay the unsecured amount if the bond is forfeited. 2 Case: 17-20333 Document: 00514472379 Page: 4 Date Filed: 05/14/2018 B. The Judges Have Different Bond-Forfeiture Policies 14. We judges all do things differently in our respective courts. In my court, for example, I do not forfeit a bond if the person appears sometime during the court session (i.e. it is okay if they are not in my courtroom at 9:00 a.m. for the initial docket call). I also do not forfeit a bond if a person misses a court date and then appears the next day or the day after. If I issue a bench warrant for nonappearance and the person subsequently appears in my court voluntarily (without being arrested), I recall the warrant. 15. I have developed these policies and practices for my courtroom because I understand from my many years of working in the system that when people miss court, it is typically for mundane reasons like their car broke down, they did not have money for the bus or a cab, they were unable to arrange child care, they or a family member had a medical emergency, or they simply forgot. 16. Sometimes people must choose between appearing in court for a scheduled court date or attending a medical appointment or a shift at a new job that will be the difference between paying rent that month or being homeless. I would always prefer that person keep their job and reschedule the court date, even if they are not able to call and reschedule the court date in advance. I just ask that they appear and explain the situation to me as soon as possible. 17. In my experience—and especially in misdemeanor court—people charged with crimes do not seek to “evade justice.” Instead, they often lead chaotic lives and are struggling with addiction or to make ends meet. Court dates are not always the greatest emergency for people who are unsure whether they will have shelter that night or food for themselves and their children. 18. My bond-forfeiture policies and practices reflect my understanding of the realities faced by people charged with misdemeanor offenses. 19. Some of my colleagues’ policies and practices are different from mine. Some judges forfeit a bond if someone is even five minutes late to court, i.e. if they do not answer docket call. Indeed, some will forfeit the bond of a defendant who arrives during docket call, if her name has already been called. Other judges will not forfeit a bond the first time a person misses court and will “roll over” the case to the next day, giving the person another chance to appear, but will forfeit the bond if the person does not appear the next day. Sometimes judges give people a few days to appear. 20. Because our policies vary widely, a single “bond forfeiture” conveys very little about the failure-to-appear rate. A single bond forfeiture sometimes reflects a single non-appearance, and sometimes it reflects two or more instances of non-appearances. It all depends on the judge and her practices. And some judges do not even follow standard policies for all people in their courtrooms. 21. We also have discretion to schedule court dates in each case whenever we want. 22. Our virtually unbridled discretion to decide how frequently and on what days to schedule court dates, and to decide when to forfeit a bond has provided fertile ground for the Fourteen 3 Case: 17-20333 Document: 00514472379 Page: 5 Date Filed: 05/14/2018 Judges to manipulate the failure-to-appear numbers so that it looks like people released on unsecured bonds are failing to appear at much higher rates. III. The Fourteen Judges Have Used Their Discretion to Generate Inflated “Bond Forfeiture” Numbers for People Released on Unsecured Bond 23. Exercising their discretion to schedule court appearances and decide when to forfeit bonds, the Fourteen Judges have implemented a variety of policies that I believe are designed to inflate the bond-forfeiture numbers for people released on unsecured bonds. A. Next-day settings for people released on unsecured bonds 24. First, they require people released on unsecured bonds pursuant to the federal court order to appear in court the day after the unsecured bond is granted, even though people released on other types of bonds—including surety, cash, or personal bonds—are typically given an initial court date that is five or seven days after their release from jail. 25. This next-day-court-setting policy matters because it is extremely difficult for people to get to court the day immediately after a bond is granted. 26. One reason people miss the next-day setting has to do with the jail’s processing policies. The Harris County jail routinely releases people onto the streets of Houston very late at night, and even in the early hours of the morning, without money or transportation. These individuals—many of whom cannot afford to live in downtown Houston and do not have reliable transportation—are faced with a choice between sleeping on the sidewalk outside the courthouse so they do not miss their court date, or finding their way home and then trying to get back downtown by docket call. Understandably, many people choose to return home to have a meal, see their family, take a shower, and sleep in their own beds. If they are unable to get back to the courthouse in the morning, they will miss their court date. We who are in charge of these policies directly control and have created these missed court appearances. 27. It is predictable that people who are required to return to court the day after they are granted an unsecured bond struggle to make it there. 28. Although we are not required to forfeit bonds for people who miss these settings, the Fourteen Judges have routinely done so. As a result, many of the bond forfeitures for people released on unsecured bond are attributable to the next-day-setting policy. 29. It is my opinion that the policy of requiring next-day settings for people released on unsecured bonds can chiefly be understood as deliberate manipulation of the bond-forfeiture numbers. 30. The stated reason for the next-day settings is apparently that we must comply with a requirement of the Texas Indigent Defense Counsel (“TIDC”) that Judges determine an incarcerated arrestee’s eligibility for counsel and appoint counsel within 24 hours of the person’s request for counsel. 4 Case: 17-20333 Document: 00514472379 Page: 6 Date Filed: 05/14/2018 31. But if the requirement to appoint counsel within 24 hours of a request is intended to ensure that defendants are provided counsel to argue for release on bail, the requirement of a next-day setting need not apply if the person is released. And even if the requirement must apply, there is no discernable reason to apply different policies to people released on different types of bonds. 32. The TIDC rule appears to be a pretext for a policy designed to make it more likely that people released pursuant to the federal court order will miss court and to inflate the bond-forfeiture numbers for people released pursuant to the federal court order. The purpose is to make it look like there’s a catastrophe in Harris County. 33. I find it troubling that, although all of us know about these problems, the Fourteen Judges have never disclosed this information when they have presented the information publicly and to the courts B. Misinformation About Court Dates 34. People released on unsecured bond have been systemically misinformed of their next-day court settings dates. The new policy of requiring next-day settings for people released on unsecured bond has generated tremendous confusion among jail officials and court employees. 35. In many cases, even though judges set court dates for the day after the bond is issued, jail employees tell the people they release to return to court in seven days, which had been the practice for many years and has continued to be the practice for people released on other types of bonds. 36. As a result, some people released on unsecured bond miss court because they did not even know they were supposed to show up. 37. Despite knowing about the confusion, my colleagues forfeit bonds—and issue new arrest warrants—for people who do not appear at the next-day setting without first inquiring as to the reason the person missed court. The fact of an arrest warrant deters people from subsequently appearing in court after missing a court date; they do not want to be arrested. That is why I have a policy in my courtroom of recalling a bench warrant if someone appears in court. I want to incentivize appearance, even following a missed court date. 38. Despite knowing about this systemic misinformation and confusion, the Fourteen Judges have not disclosed this information to the courts or the public when they have proffered their data concerning bond-forfeiture rates. C. Confusion Caused by Hurricane Harvey 39. Hurricane Harvey has also contributed to confusion in the court system. 40. When the hurricane struck at the end of August, it flooded the courthouse, and the County had to set up courtrooms in various other locations. My colleagues and I have been sharing 5 Case: 17-20333 Document: 00514472379 Page: 7 Date Filed: 05/14/2018 courtrooms, and have presided over our dockets in those shared courtrooms on alternate days and at designated times of day. 41. As a result, some people who knew which date to appear and thought they knew where to go have missed court dates because the times and locations of our dockets were different than they thought.3 42. Despite knowing that this confusion has caused people to miss court, my colleagues forfeit bonds for people released on unsecured bond without inquiring into why they did not appear. D. Failure to Supervise People Released Pursuant to the Federal Court Order 43. The County and other Judges have also implemented policies that appear designed to funnel people who have the greatest likelihood of non-appearance into the group of people released pursuant to the federal court’s order. 44. Since July 1, 2017, the County has used a pretrial assessment tool to determine the amount of “risk” an arrestee poses of not appearing for court. The higher a person’s risk score, the more likely it is that, absent intervention (such as text message reminders or transportation assistance), the person will not appear. Accordingly, the higher the person’s “risk” score, the greater the person’s need for pretrial assistance to get to court. 45. Pursuant to the injunction, misdemeanor arrestees are given an opportunity to state how much money they can afford to pay to be released. If a Hearing Officer or a Judge then requires a money bail amount that exceeds the amount the person can afford, the Sheriff converts the secured amount to unsecured and releases the person. 46. Since the injunction went into effect, Hearing Officers and Judges have granted release on personal bond—with supervision by pretrial services and court reminders—to an increasing number of people deemed “low risk” by the pretrial assessment tool. 47. However, they have continued to impose unaffordable money bail amounts as a condition of release for people deemed “high risk” by the pretrial assessment tool. Prior to the injunction, these higher “risk” arrestees would have been detained due to their inability to pay the amount of money required for release. Now, because of the injunction, these arrestees may not be kept in jail due to inability to pay, and the Sheriff releases them—as required—on unsecured bail. 48. Despite knowing that higher risk arrestees will be released on unsecured bond pursuant to the injunction, my colleagues have refused to require basic pretrial supervision and court reminders to the group of arrestees released on unsecured bond. The Fourteen Judges and other Defendants are also refusing to impose non-financial conditions (unless statutorily required) that have been 3 In January 2018, the Houston Chronicle conducted an investigation (though not an exhaustive review) and identified “at least 31 case in which defendants were given wrong information about their court dates; 15 of them forfeited their bonds for failing to appear at the first setting.” See Meagan Flynn, Harris County bail system offers little help to defendants who most need it (Jan. 22, 2018), available at https://www.houstonchronicle.com/news/houstontexas/houston/article/Harris-County-bail-system-shortchanges-defendants-12516456.php. 6 Case: 17-20333 Document: 00514472379 Page: 8 Date Filed: 05/14/2018 proven to increase court appearance rates. They therefore know that these arrestees will be released with no non-financial conditions or supervision because of their decision only to require a secured financial condition of release that they know the Sheriff will not enforce. 49. At the same time, the County has been providing supervision and court reminders to other arrestees, even though they are less in need of assistance getting to court (according to the County’s pretrial assessment tool). All of these practices can dramatically skew success rates for those released on different types of bonds. E. ICE Custody 50. The federal court’s preliminary injunction requires the Sheriff to automatically convert a secured money bail amount that a person cannot afford to an unsecured amount and release the person. If the person is subject to an “ICE detainer,” ICE is notified to come and pick up the person from the jail to be transported to a federal immigration detention facility as soon as an unsecured bond is granted. 51. But there is no system in place to adequately facilitate communication between the Harris County court system and ICE. As a result, when people who are in ICE detention do not come to court for their misdemeanor case, my colleagues appear to be forfeiting their bonds. 52. The Texas Observer recently reported that between 5 and 10 percent of forfeitures are attributable to people who did not appear in court because they were in ICE detention.4 53. Despite knowing that many people miss court because they are in ICE detention, my colleagues forfeit bonds for people released on unsecured bond without inquiring into why they did not appear. F. Counting Bonds Not People 54. It is also important to note that the Fourteen Judges’ statistics purport to identify the number of bonds that have been forfeited. (I have no way at this point of independently verifying that the data actually produce the numbers they have been presenting.) But arrestees are given a bond for each open case, and therefore a single missed court appearance can translate to multiple bond forfeitures. 55. To take an example I see in my courtroom, consider a person who is homeless and is repeatedly arrested for trespassing in the location where she sleeps at night. That person will be given a bond each time she is arrested. Since the federal court order went into effect, those bonds are likely to be unsecured bonds. The Judges tend to consolidate court dates for multiple open cases. Thus, that person will typically be required to appear on a single date to address each of the open cases. If the person does not appear at a single court date, multiple bond forfeitures will be entered: one for each open case. 4 Stephen Paulsen, A Catch-22 in the Crimmigration System in Harris County (May 8, 2018), available at https://www.texasobserver.org/a-catch-22-in-the-crimmigration-system-in-harris-county/. 7 Case: 17-20333 Document: 00514472379 Page: 9 Date Filed: 05/14/2018 56. In January 2018, the Houston Chronicle conducted a review of about 125 cases filed in a roughly two-month period. The investigation revealed that 29 defendants (who were described as mentally ill by the public defender’s office) accounted for 235 bond forfeitures. See Flynn, supra n.2. 57. This means that a relatively small number of people—often people who are chronically homeless or mentally ill—are responsible for many bond forfeitures. C. What the Data Does Not Show 58. It is important to take a moment to identify specifically what the data the Judges present does not show. 59. Most importantly, they do not report the number of people who have missed court dates, or the number of court dates a particular person has missed. 60. In short, the data presented by the other judges and County is fatally flawed and meaningless. D. Conclusion 61. A few months ago I learned that arrestees who were deemed “high risk” by the County’s new pretrial assessment tool and who were released on unsecured bonds pursuant to the federal court order were being released with no supervision. I immediately contacted Pretrial Services and instructed them to provide supervision to all arrestees assigned to Court 16. I am aware Judge Fields has the same practice in his court. In the other fourteen courts, it is my understanding that those deemed high risk by the County’s pretrial assessment tool and released on an unsecured bond with no conditions are not being supervised. 62. It is my belief that this policy, along with the other policies described in my declaration, explain the elevated bond-forfeiture rate for people released on unsecured bond pursuant to the federal court order. The only explanation for these policies and practices is a win-at-all-costs mentality. The Fourteen Judges claim that defendants released on an unsecured bond pose an intolerable danger to the community. If that statement is true, then it is also true that they have contributed to the risk to the community by refusing to require Pretrial Services to supervise those who they claim pose the highest risk. 63. It is also important to remember that prior to the federal court’s order, people who could not afford to pay money bail were detained until disposition (40% of all misdemeanor arrestees annually) and 85% of them pled guilty in a median of 3.2 days: 67% of those who pled guilty were then released within one day, and 83% were released within five days, of conviction. At the end of the day, the Judges would prefer to release people after conviction, instead of before. But they ignore the fact that many of those convictions were coerced guilty pleas from people who desperately wanted to get out of jail and could not afford money bail. Their prior policies did not make our community safer. 8 Case: 17-20333 Document: 00514472379 Page: 10 Date Filed: 05/14/2018 64. Based on everything I have observed, there is no reason to believe that the federal court injunction has increased failure to appear rates or caused a public safety crisis. Rather, there is every reason to believe that the numbers cited by the Fourteen Judges are the result of deliberate policy decisions, both acts and omissions. 65. Since February, as ordered by Judge Rosenthal (Doc. 383), I have worked with Judge Fields, and conferred with counsel representing Plaintiffs, to craft a revised preliminary injunction order that conforms to the Fifth Circuit’s holdings and the factual record. District Attorney Kim Ogg and Public Defender Alex Bunin support the proposed injunction. The Plaintiffs will not object to its being entered. The County and the Fourteen Judges abandoned the process of working together to create a revised order. 66. Contrary to the Judges’ claim, the sky is not falling in Harris County. The current injunction should remain in place while Judge Rosenthal has an opportunity to consider all parties’ input on a revised order. I declare under penalty of perjury that the foregoing is true and correct to the best of my ability. May 14, 2018 ____________________________ Date ___________________________ Judge Darrell Jordan 9 Case: 17-20333 Document: 00514472379 Page: 11 Date Filed: 05/14/2018 Exhibit 2 Case: 17-20333 Document: 00514472379 Page: 12 Date Filed: 05/14/2018 DECLARATION OF COUNTY CRIMINAL COURT AT LAW NO. 14 1 JUDGE MICHAEL FIELDS I. Introduction 1. My name is Michael Fields. I was elected in 1998 by the citizens of Harris County, Texas to serve as Judge of Harris County Criminal Court at Law No. 14. On January 1, 1999, I began my term as Judge of Court No. 14.  2. Prior to beginning my service as Judge of Harris County Criminal Court No. 14, I maintained a private criminal law practice. Additionally, I served as an Assistant Attorney General for the state of Texas and as an Assistant District Attorney for Harris County, Texas. I was also appointed by the Supreme Court of Texas to serve as a member of the State Commission on Judicial Conduct, where I served for nearly seven years. 3. I am familiar with the claims made by Judges 1–13 and 15 (“the Fourteen Judges”) that the preliminary injunction issued in this case “has created a law enforcement and public safety crisis” and that “[r]oughly ​half of all unsecured release bonds issued pursuant to the preliminary injunction have been forfeited because the arrestee failed to appear for his or her court date.” I believe the data buttressing that claim can be easily misinterpreted. 4. In my opinion, the Harris County Criminal Court at Law Judges have implemented policies that, when coupled with other factors, have resulted in the artificial inflation of bond-forfeiture statistics, thereby making it look like people released on unsecured bond pursuant to the federal order are failing to appear in droves, and that the federal court order is “wreaking havoc” on the County. The truth, however, is much more nuanced. The elevated bond-forfeiture rate for people released on unsecured bail, to the extent it exists, is more fully explained, by a number of factors, including policy decisions, agency regulations and a catastrophic weather event. II. Harris County Criminal Court at Law Judges Have Discretion to Decide When to Forfeit Bonds A. A “Bond Forfeiture” Is Not a “Failure to Appear” 5. Before addressing the judges’ policies, it is crucial to, first, understand the terminology that is being employed. Primarily, there have been indications that the federal district Court's 2 order has resulted in an elevated “bond forfeiture” rate for people released on unsecured bond. 1 Submitted in ​ODonnell v. Harris County,​ Case No. 16-cv-01414 (S.D. Tex. 2016), No. 17-20333 (5th Cir. 2017). 2 When a person is released pursuant to the federal court’s order, the bond type is referred to as an “unsecured bond.” When a person is ordered released by a judge or magistrate, the bond type is referred to as a “personal bond” (though personal bonds are also technically unsecured bonds: personal bonds do not require an up-front payment, but the person will be required to pay the unsecured amount if the bond is forfeited). 1 Case: 17-20333 Document: 00514472379 Page: 13 Date Filed: 05/14/2018 6. The term “bond forfeiture” is not the same as “failure to appear.” “Bond forfeiture” is merely the rough proxy that my colleagues and I use for “failure to appear”—and it is easy to manipulate. Although the judges and the County have apparently begun to track failures to appear, they have not, to my knowledge reported those statistics to the court, only the statistics for bond forfeitures. 7. One reason the bond-forfeiture numbers are misleading is that one defendant may be charged with multiple cases. For example, a person who is homeless and is repeatedly arrested for trespassing in the location where she sleeps at night will be given a bond each time she is arrested. Since the federal court order went into effect, those bonds are likely to be unsecured bonds. As the Houston Chronicle reported in January 2018, according to the public defenders who staff the bail hearings, “[t]he mentally ill homeless generally don’t get the personal bonds. 3 They get unsecured release because they’re considered high risk.” 8. Because Judges tend to consolidate court dates for multiple open cases, a person will typically be required to appear on a single date to address each of the open cases. If the person does not appear at a ​single court date, ​multiple bond forfeitures will be entered: one for each open case. My understanding of the bond forfeiture statistics reported by my fourteen colleagues suggests that the bond forfeiture rates are being reported by case and not by defendant. As a result, a single defendant may be counted as having forfeited multiple times, when in actuality, the defendant has only failed to appear on one court date. 9. Additionally, Texas law may be another reason that “bond forfeitures” are a very poor proxy for failures to appear. State statutes set forth a number of procedural rights and broad substantive defenses to forfeiture. Surety companies are aware of these laws and, as experienced, routine players in our system, they often hire lawyers to represent them in forfeiture proceedings. Because the statutory defenses are broad, surety companies are often able to evade forfeiture, even when the defendant did in fact fail to appear. 10. As a result of these defenses and surety companies’ ability to take advantage of them by hiring lawyers and fighting the forfeiture, the ​absence of a forfeiture of a surety bond does not mean that the person released on that surety bond appeared in court for all of her court appearances.   11. By contrast, people released on unsecured bonds or personal bonds may not know that such defenses to forfeiture exist, partly because the judges are not required to inform them of their defenses to a bond forfeiture. As a result, the people who benefit from defenses to forfeiture are overwhelmingly people released on surety bonds. 12. For these reasons, “bond forfeiture” rates are a very poor proxy for non-appearance rates. Meagan Flynn, ​Harris County bail system offers little help to defendants who most need it (Jan. 22, 2018), available at https://www.houstonchronicle.com/news/houston-texas/houston/article/Harris-County-bail-system-shortchanges-def endants-12516456.php. 3 2 Case: 17-20333 Document: 00514472379 Page: 14 Date Filed: 05/14/2018 B. The Judges Have Different Bond-Forfeiture Policies 13. We judges all handle missed appearances and bond forfeitures differently in our respective courts. 14. Over the years, I have developed policies and practices for my courtroom that take into account some of the mundane reasons defendants miss court such as their car broke down, they did not have money for the bus or a cab, they were unable to arrange child care, they or a family member had a medical emergency, or they simply forgot. 15. It has been my experience that people must sometimes choose between appearing in court for a scheduled court date or attending a medical appointment or a shift at a new job that will be the difference between paying rent that month or being homeless. It is my preference that a person keep their job and reschedule the court date. 16. In my experience—and especially in misdemeanor court—people charged with crimes often lead chaotic lives and are struggling with addiction or to make ends meet. Consequently, court dates might not always be the greatest emergency for people who are unsure whether they will have shelter that night or food for themselves and their children. 17. I believe bond-forfeiture policies and practices should reflect an understanding of the realities faced by people charged with misdemeanor offenses. 18. Some of my colleagues’ policies and practices are different from mine. Some judges may forfeit a bond if someone is late to court, i.e. if they do not answer docket call at 8:30 a.m. Other judges will not forfeit a bond the first time a person misses court and will “roll over” the case to the next day, giving the person another chance to appear, but will forfeit the bond if the person does not appear the next day. Sometimes judges give people a few days to appear. 19. Because our policies vary widely, the “bond forfeiture” rate conveys very little about the failure-to-appear rate, even though the data provided by my colleagues imply that they are the same. A single bond forfeiture sometimes reflects a single non-appearance, and sometimes it reflects two or more instances of non-appearance. It all depends on the judge and her practices. III. The Bond Forfeiture Rate for Unsecured Bonds Are Impacted by Decisions Policy A. Next-day settings for people released on unsecured bonds have resulted in Inflated “Bond Forfeiture” Numbers for People Released on Unsecured Bond 20. First, the Judges have a policy of requiring people released on unsecured bonds pursuant to the federal court order to appear in court on the docket immediately following the day a person is ordered released on an unsecured bond. This requirement has resulted in an elevated number of bond forfeitures. 3 Case: 17-20333 Document: 00514472379 Page: 15 Date Filed: 05/14/2018 21. This next-day-court-setting policy matters because it is extremely difficult for people to get to court the day immediately after, or in many cases, the same morning a release is granted. 22. One reason people, likely, miss the next-day setting is because they are ​released mere hours before ​they are supposed to appear in a County Court at Law. The Harris County jail routinely releases people onto the streets of Houston very late at night, and even in the early hours of the morning, without money or transportation. These individuals—many of whom cannot afford to live in downtown Houston and do not have reliable transportation—are, apparently, faced with a choice between sleeping on the sidewalk outside the courthouse so they do not miss their court date, or finding their way home and then trying to get back downtown by docket call. Understandably, many people may choose to return home to have a meal, see their family, take a shower, and sleep in their own beds. If they are unable to get back to the courthouse in time, they will miss their court date. The net result is that local policies are resulting in missed court appearances. It is troubling that this information was not presented in the materials supplied to the Court. 23. It is my opinion that the policy of requiring next-day settings for people released on unsecured bonds has resulted in a manipulation of the bond-forfeiture numbers. 24. The ​stated reason for the next-day settings is apparently that we must comply with a requirement of the Texas Indigent Defense Counsel (“TIDC”) that Judges determine an incarcerated arrestee’s eligibility for counsel and appoint counsel within 24 hours of the person’s request for counsel. 25. But if the requirement to appoint counsel within 24 hours of a request is intended to insure that defendants are provided counsel to argue for release on bail, the requirement of a next-day setting need not apply if the person is released. And even if the requirement must apply, there is no discernible reason to apply different policies to people released on different types of bonds. 26. The TIDC rule appears to be a pretext for a policy designed to make it more likely that those released on an unsecured bond will miss court. As previously stated, this policy artificially inflates the bond-forfeiture numbers for people released pursuant to the federal court order. It is reasonable to assume the purpose is to give the impression there’s a catastrophe in Harris County. B. Misinformation About Court Dates 27. People released on unsecured bond have been systemically misinformed of their next-day court setting. The new policy of requiring next-day settings for those people released on unsecured bond has generated tremendous confusion among jail officials and court employees. 28. In many cases, even though judges set court dates for the day after the bond is issued, the computer records indicate that the released arrestees must return to court in seven days. The jail 4 Case: 17-20333 Document: 00514472379 Page: 16 Date Filed: 05/14/2018 informs the people they release that they must return in seven days, which was the practice that was in effect for many years prior to the federal district court’s injunction, and the policy that applies to people released on other bond types. 29. As a result, it is likely some people released on unsecured bond miss court because they are confused as to when they are supposed to show up. 30. I do not feel this information, concerning bond forfeiture rates, has been fully disclosed to the courts.  C. Confusion Caused by Hurricane Harvey 31. Hurricane Harvey has also contributed to confusion in the court system. 32. When the hurricane struck at the end of August, it flooded the courthouse, and the County had to set up courtrooms in various other locations. My colleagues and I have been sharing courtrooms, and have presided over our dockets in those shared courtrooms at designated times of day. 33. As a result, some people who thought they knew when to appear and where to go, have missed court because the location of their assigned court and their scheduled appearance time changed. D. Failure to Supervise People Released Pursuant to the Federal Court Order 34. Additionally, policies have been implemented which have the net result of funneling people who have the greatest likelihood of non-appearance into the group of people released pursuant to the federal court’s order.    35. Since  July  1,  2017,  the  County  has  used  a  pretrial  assessment  tool  to  determine  the  amount  of  “risk”  an  arrestee  poses  of  not  appearing for court. My belief is, the higher a person’s  risk  score,  the  more  likely  it  is  that,  ​absent  intervention  (such  as  text  message  reminders  or  transportation  assistance),  the  person will not appear. Accordingly, the higher the person’s “risk”  score, the greater the person’s need for pretrial assistance to get to court.   36. Pursuant to the injunction, misdemeanor arrestees are given an opportunity to state how much money they can afford to pay to be released. If a Hearing Officer or a Judge then requires a money bail amount that exceeds the amount the person can afford, the Sheriff converts the secured amount to unsecured and releases the person. 37. Since the injunction went into effect, Hearing Officers and Judges have granted release on personal bond—with supervision by pretrial services and court reminders—to an increasing number of people deemed “low risk” by the pretrial assessment tool. 5 Case: 17-20333 Document: 00514472379 Page: 17 Date Filed: 05/14/2018 38. However, they have continued to impose unaffordable money bail amounts as a condition of release for people deemed “high risk” by the pretrial assessment tool. Prior to the injunction, these higher “risk” arrestees would have been detained due to their inability to pay the amount of money required for release. Now, because of the injunction, these arrestees may not be kept in jail due to inability to pay, and the Sheriff releases them—as required—on unsecured bail. 39. Despite  the  fact  that  higher  risk arrestees are being released on unsecured bonds pursuant  to  the  injunction,  Pretrial  Services  has  yet  to  provide  the  identical  level of supervision to people  released on unsecured bonds as those released on pretrial release bonds.   40. Additionally,  several  of  my  colleagues  have  refused  to  require  non-financial  conditions  (unless statutorily required) that have been proven to increase court appearance rates.  41. At  the  same  time,  greater  supervision  and  court  reminders  are  being  provided  to  other  arrestees,  even  though  they  are  less  in  need  of  assistance  getting  to  court  (according  to  the  County’s pretrial assessment tool).  E. ICE Custody 42. The federal court’s preliminary injunction requires the Sheriff to automatically convert a secured money bail a person cannot afford to unsecured bail and release the person. If the person is subject to an “ICE detainer,” ICE is notified to come and pick up the person from the jail to be transported to a federal immigration detention facility as soon as an unsecured bond is granted. 43. But, to my knowledge, there is no system in place to adequately facilitate communication between the Harris County court system and ICE. As a result, when people who are in ICE detention do not come to court for their misdemeanor case, their bonds are subject to forfeiture. F. Counting ​Bonds​ Not ​People 44. It is also important, in my opinion, to note that statistics which purport to identify the number of ​bonds that have been forfeited must also indicate the number of defendants represented by that number. (I have not independently verified that the data actually produce the numbers that have been presented.) But as previously mentioned, arrestees are given a bond for each open case. Therefore, a single missed court appearance can translate to multiple bond forfeitures. 45. This results in a relatively small number of people may being responsible for a large number of bond forfeitures. IV. What the Data Does Not Show 46. It is important to take a moment to identify specifically what the data presented may ​not show. 6 Case: 17-20333 Document: 00514472379 Page: 18 Date Filed: 05/14/2018 47. Most importantly, it does not report the number of people who have missed court dates, or the number of court dates a particular person has missed. 48. It is my belief that data which does not give as full a picture as possible is unreliable. V. Conclusion 49. Based on everything I have observed, there is no way to credibly illustrate that the federal court injunction alone has increased failure to appear rates or caused any public safety crisis. Rather, there is adequate reason to believe that the numbers cited by my colleagues are equally as likely the result of policy decisions, the federal district Court’s order and the difficult circumstances caused by hurricane Harvey. I declare under penalty of perjury that the foregoing is true and correct to the best of my ability. 5/14/2018 7:31:17 PM PDT ___________________________ Judge Michael Fields ____________________________ Date 7 Case: 17-20333 Document: 00514472379 Page: 19 Date Filed: 05/14/2018 Exhibit 3 Case: 17-20333 Document: 00514472379 Page: 20 Date Filed: 05/14/2018 EXPERT ANALYSIS OF DR. STEPHEN I. Introduction 1. I have conducted a preliminary analysis of the data Defendants have produced in discovery in the above-captioned case. My analysis has focused on the data showing bond forfeitures and the judges? assertion that bond-forfeiture rates have dramatically increased since the injunction went into effect.2 The County and judges have not produced all the data necessary to analyze bond forfeitures and failures to appear throughout the entire 11 months since the injunction went into effect. I am awaiting updated data and will continue my analysis when I receive it. My analysis therefore is ongoing. 2. This report sets forth my preliminary ?ndings, which I am continuing to update. My preliminary analysis of the judges? bond-forfeiture statistics suggests that they are deliberately manipulating the system in an effort to produce a desired result: to suggest that people released on unsecured bonds supposedly evade justice. II. A ?bond forfeiture? is not a one-to-one proxy for ?failure to appear? 3. I have reviewed Exhibit A, ?led by the Fourteen Judges in the Fifth Circuit Court of Appeals on May 3, 2018, which includes a declaration by Ed Wells and a chart purporting to show bond failure rates, including the bond-forfeiture rates for four different bond types. The Judges report that between June 6, 2017 and April 30, 2018, the ?raw? bond forfeiture rates by bond type were as follows: Cash: 15.82% ?Secured? (surety bonds): 10.02% Personal: 26.67% ?Sheriff" 5? bonds (unsecured bonds issued pursuant to the federal court injunction): 49.30% 9-9 9:2? 4. These raw bond-forfeiture statistics do not provide meaningful information about comparative appearance rates for people released on different bond types. 5. Court Manager Ed Wells states in his declaration that each of the above-listed bond- forfeiture rates ?represents a failure-to-appear rate.? That is incorrect for several reasons discussed 1 Submitted in ODonnelI v. Harris County, Case No. l6-cv-014l4 (S.D. Tex. 2016), No. 17-20333 (5th Cir. 2017). 2 I am aware of the judges? claim that bond-forfeiture rates have ?skyrocketed while the preliminary injunction has been in place,? but the judges do not identify a baseline against which they are comparing current rates. To the extent the judges are comparing bond-forfeiture rates for each bond type from the months immediately following June 7, 2017 (when the injunction went into effect) to the bond-forfeiture rates they now report for the full 11 months since the injunction went into effect, the increase is to be expected: as time passes, more court dates are scheduled and there are more opportunities for each person to not appear and for a bond to be forfeited. To the extent the judges are comparing bond-forfeiture rates post-injunction to bond-forfeiture rates pre-injunction, the comparison is meaningless because there is an entirely new category of bonds post-injunction (the unsecured bonds issued pursuant to the injunction), and because this type of comparison can only be made if an identical time period were identi?ed, which the judges do not do. Case: 17-20333 Document: 00514472379 Page: 21 Date Filed: 05/14/2018 in this declaration. Among them: the bond-forfeiture statistics count bonds not people. The data show that many people have multiple bonds and that a single missed court date can result in numerous bond forfeitures. 6. Jurisdictions that rigorously monitor and evaluate their pretrial systems track failures to appear clearly by person and court date, not by the number of total cases. They follow them prospectively, as opposed to just lumping them all together. 7. 8. I used the data to analyze the relationship between bond forfeitures and failures to appear. My preliminary analysis (which involved cases with arrest dates between June and September 2017 because I have not received data for the full 1 1 months since the injunction has been in effect) shows that: a. There are numerous cases involving people released on unsecured bond who, according to the County?s data, appeared in court for their ?rst court date, but their bonds were forfeited anyway; b. There are numerous cases involving people released on surety bonds who, according to the County?s data, missed multiple court appearances, but whose bonds were never forfeited; 0. Among people released on unsecured bands who, according to the County?s data, missed their ?rst two scheduled court dates, 36% had their bonds forfeited. (1. Among people released on surety bands who, according to the County?s data, missed their ?rst two scheduled court dates, only 6% had their bonds forfeited. 9. Although there are inconsistencies in the data that I will continue to explore during my ongoing analysis, these systemic patterns suggest that people released on unsecured bonds are treated differently?and more people released on surety bonds. Based on the available data, it is likely that differential treatment by the Judges explains the different forfeiture rates between unsecured and surety bonds in these cases. 111. The Elevated Bond-Forfeiture Rate for People Released on Unsecured Bond Can Be Explained by Additional Factors A. People Released on Unsecured Bond Belong to a Group of People Who, According to the County?s Pretrial Assessment Tool, Are More Likely to Miss Court Than People Released on Other Bond Types And Are Not Receiving Pretrial Services or Supervision 10. I analyzed the data ?'om the County?s pretrial assessment tool to determine the average assessed risk of failure-to-appear (FTA) on a 6-point scale among people released on different types of bonds. Between July 28, 2017, and January 31, 2018, people released on unsecured bonds pursuant to the federal court?s order presented, on average, a substantially higher risk score (2.51) 2 Case: 17-20333 Document: 00514472379 Page: 22 Date Filed: 05/14/2018 than people released on surety bonds (1.64) or personal bonds (1.85). That is, the FTA risk score for people released on unsecured bonds was 53% higher than the score for people released on surety bonds and 36% higher than for people released on personal bonds. 11. Despite their higher ?risk? of nonappearancwand greater need for pretrial assistance to get to court?people released pursuant to the federal court?s order during this time period were much less likely to be supervised or to have non-?nancial conditions imposed. Using data showing bond supervision and conditions for cases released between June 7 and November 20, 2017, I found that only 5.6% of people released on unsecured bonds were supervised by pretrial services and only 4.4% had additional non??nancial conditions. In contrast, 67% of people released on personal bonds were supervised, and roughly 13% had additional non-?nancial conditions. B. People Are Routinely Released From the Jail in the Middle of the Night 12. Data showing release times reveal that people are often released ?'om the jail in the middle of the night. C. People Released on Unsecured Bond Are Typically Required to Return to Court the Next Day And Have More Court Settings Scheduled Close in Time 13. For people released on unsecured or personal bonds between June 7, 2017 and December 31, 2017, the median time between bond-approval and ?rst-setting was one day. About 25% of people released on unsecured bond had less than 12 hours between the bond-approved time and ?rst?setting time; 57% had less than 24 hours between the bond?approved time and ?rst-setting time. 14. For people released on cash or surety bonds during this time period, the median time between bond-approval and ?rst-setting was ?ve days. Only 4% of cash releasees and 5% of surety releasees had less than 12 hours between bond-approved time and ?rst-setting time; about 26% of cash and surety bond releases had less than 24 hours between bond?approved time and ?rst-setting time. 15. The data also show that people released on unsecured bonds are scheduled, on average, for more court settings than are people released on surety bonds, and that those court settings are scheduled closer in time for people released on unsecured bonds as compared to surety bonds. These practices dramatically increase the chances for a person to not appear. D. Summary 16. What these ?ndings mean is that the County and Judges are releasing the people most in need of services without any services or support and often in the middle of the night within hours of their ?rst hearing before a County Criminal Court at Law Judge. People released on unsecured and personal bonds have considerably less time after release until their ?rst court date than people released on surety or cash bonds. Moreover, the County and the Judges have chosen not to require additional non-?nancial conditions designed to reduce nonappearance for precisely the category of arrestees that their own risk assessment tool suggests should receive those conditions. On top 3 Case: 17-20333 Document: 00514472379 Page: 23 Date Filed: 05/14/2018 of all that, the Judges? apparently have a policy or practice of treating people released on surety bonds much more leniently than they treat people released on unsecured bonds with respect to bond forfeiture and appearance. Each of these factors on its own makes it much more likely that a person will miss a court date. Together, they make appearance extremely dif?cult. As such, comparisons made by the Judges and the County between the forfeiture rates of unsecured and surety bonds are dubious, especially given that the data is replete with errors suggesting ?aws in its integrity. IV. Bond-Forfeiture Rates Vary Dramatically Among the Judges 17. I was able to analyze bond approval and forfeiture data for June 7, 2017 through April 30, 2018. Bond forfeiture rates vary dramatically among the various judges. For example, I observed bond-forfeiture rates in certain courts that are two or three times greater than the bond-forfeiture rate in other courts. Further analysis is necessary to determine the cause of the variations, but the initial ?ndings are surprising given my understanding that cases are assigned ?randomly? across the 16 courts which would make the average characteristics of caseloads the same across the 16 courts. 18. Unless there is some other difference in the nature or timing of cases across the courts, another possible explanation for differences in forfeiture rates is that judges have different practices for setting court dates or use different standards in deciding when to forfeit bonds. V. People Who Court Are Not Evading Prosecution 19. Based on my analysis of the data involving thousands of cases, I do not believe that people released on unsecured bonds who do not appear for their initial court dates are evading justice. 20. To get a sense of what happens to these cases, I examined cases in which people were released on unsecured bonds and surety bonds between June 7 and August 15, 2017 and did not appear for their ?rst two court dates. My preliminary analysis shows that about 2/3 of people released on unsecured bonds had returned to court and resolved their cases by January 31, 2018. (The data show that a similar percentage of people released on surety bonds appeared by January 31, 2018.) 21. My data analysis is consistent with the research literature, which shows that only a very small portion of FTAs are ?willful,? meaning that the person is actively evading justice. More likely explanations are that people lacked transportation or childcare, could not take off work, were confused about the right court date, were told to go to the wrong courtroom or courthouse, or were afraid to appear because of possible outstanding warrants caused by an earlier nonappearance for one of the above reasons. 22. My analysis of case outcomes is ongoing and will become more robust as I receive data for the rest of the post-injunction time period. Case: 17-20333 Document: 00514472379 Page: 24 Date Filed: 05/14/2018 VI. Conclusions 23. The judges? bond-forfeiture rates for people released on unsecured bonds are seriously in?ated by their and the County?s policy decisions and practices. This in?ation fatally undermines any conceivable utility from them. 24. The elevated bond-forfeiture rate in Harris County, to the extent that it exists, does not re?ect that unsecured bond is ineffective or any less effective than secured bond. The data does not provide a basis for drawing the conclusion that the Judges suggest in their Motion about the federal court order having any negative effect on nonappearance. The statistics suggest a deliberate manipulation of the numbers and the post-arrest system to attempt to produce a desired outcome. I declare under penalty of perjury that the foregoing is true and correct to the best of my ability. gig/QM 51/4, Stepeen Demuth, Date Case: 17-20333 Document: 00514472379 Page: 25 Date Filed: 05/14/2018