CJA S SUPERVISED RELEASE PROGRAMS AND MANHATTAN PROGRAM START-UP: CASE SCREENING AND PARTICIPANT SELECTION PROCESS

Transcription

1 Jerome E. McElroy Executive Director CJA S SUPERVISED RELEASE PROGRAMS AND MANHATTAN PROGRAM START-UP: CASE SCREENING AND PARTICIPANT SELECTION PROCESS Freda F. Solomon, Ph.D. Senior Research Fellow and Project Director FINAL REPORT April Duane Street, Third Floor, New York, NY (646) The mission of the New York City Criminal Justice Agency, Inc., is to assist the courts and the City in reducing unnecessary pretrial

2 CJA S SUPERVISED RELEASE PROGRAMS AND MANHATTAN PROGRAM START-UP: CASE SCREENING AND PARTICIPANT SELECTION PROCESS Freda F. Solomon, Ph.D. Senior Research Fellow and Project Director Jonathan Carmona Senior Research Assistant Geraldine Staehs-Goirn IT Programmer Analyst Wayne Nehwadowich IT Deputy Director for Programming April 2014 This report can be downloaded from: NYC Criminal Justice Agency, Inc. When citing this report, please include the following elements, adapted to your citation style: Solomon, Freda F CJA s Supervised Release Programs and Manhattan Program Start-Up: Case Screening and Participant Selection Process. New York: New York City Criminal Justice Agency, Inc.

3 ACKNOWLEDGEMENTS This report reflects the combined efforts of many both within and outside of the New York City Criminal Justice Agency, Inc. (CJA). Jerome McElroy, CJA s Executive Director, has long advocated introducing an alternative between ROR and money bail into the City s Criminal Courts. That goal has become a reality due to the vision and tireless efforts of Mari Curbelo, CJA s Director of Court Operations, who crafted and implemented CJA s successful Supervised Release programs currently operating in two counties. She also assembled, along with Andrea Barrow, Director of Manhattan s Supervised Release program, the courtroom team who not only screen and advocate for prospective program clients, but also carefully document their work. Wayne Nehwadowich, IT Deputy Director for Programming, has been responsible for creating the program s computerized information system, and is indispensable to our research efforts. Geraldine Staehs-Goirn, IT Programmer Analyst, created the data file that forms the basis of the information analyzed in this report, combining items from the program s database with additional variables from CJA s main computerized information system. I greatly appreciate all the hard work of Jonathan Carmona, Senior Research Assistant. He has taken the lead, assisted by Steven Corrente, Supervised Release Records Management Assistant, in the review of program screening forms, overseeing and assisting in their data entry, and performing quality control for the Manhattan program data. He also created the analytic data set for this research, performed some of the analysis, and assisted in the writing of the report. There are no adequate words of thanks for all of the assistance provided by Joann DeJesus, Executive s Manager of Special Projects. The range of her responsibilities and contributions to all of us involved with the Supervised Release programs, so ably performed, are too numerous to list. CJA s Supervised Release programs have been made possible by the commitment of the New York City Office of the Criminal Justice Coordinator (now renamed the Mayor s Office of Criminal Justice) to the development and funding of these programs. Within that Office special thanks are due to Michele Sviridoff, Deputy Coordinator for Policy and Planning for her ongoing assistance, and to its Counsel, Jordan M. Dressler and Gerald T. Foley, Assistant Director of Program Management and Development. I am indebted to many of my Agency colleagues who gave so generously of their time to review and comment on earlier drafts of this report. However, I alone am responsible for any errors. Thanks always are due to Annie Su, the Research Department s Administrative Associate, who oversees all the final details that lead to the production and distribution of research reports. Freda F. Solomon, Ph.D. Senior Research Fellow and Project Director

4 TABLE OF CONTENTS ACKNOWLEDGEMENTS INTRODUCTION... 1 CJA S SUPERVISED RELEASE PROGRAMS... 3 Charge Criteria... 3 CJA Release Recommendation, Criminal Conviction History and Related Items... 3 Defendant Access and Interview... 5 Community Ties... 6 Assignment to the Supervised Release Program... 6 Data Collection During the Screening Process... 7 IMPLEMENTING THE MAHATTAN SUPERVISED RELEASE PROGRAM: THE FIRST SIX MONTHS, April 8 through October 8, SUMMARY... 22

5 INTRODUCTION The primary mission of the New York City Criminal Justice Agency (CJA), Inc. is to provide pretrial services to defendants prosecuted in the City s adult Criminal Court system. For defendants held in custody pending Criminal Court arraignment CJA uses a validated risk-assessment instrument to make a release recommendation based on statistical probabilities of failure to appear (FTA) if released on recognizance. New York remains one of only several states which rely solely on factors related to FTA as the basis for pretrial release decisions. In addition, New York judges must choose between personal recognizance or bail for pretrial release in almost all cases continued at Criminal Court arraignment. 1 For a number of years CJA has advocated for offering community supervision as an alternative to money bail for defendants posing a moderate risk of failing to appear if released on unsupervised personal recognizance. After extensive consultation with the New York City Office of the Criminal Justice Coordinator, and with their support, an experimental pretrial alternative-to-detention (ATD) program was designed to offer the option of supervised release as a bail alternative at Criminal Court arraignment in a limited number of non-violent felony cases with a high likelihood of having bail set. Criteria for defendants in these cases were agreed to with a goal of creating a program model that would achieve jail displacement for the target population along with controls to prevent the inappropriate substitution of community supervision for traditional release on recognizance (ROR). To accomplish this requires an extensive screening process for prospective program clients in charge-eligible cases. The screening elements include an assessment of CJA s pretrial release recommendation and review of the defendant s criminal history record in order to exclude those with the lowest and highest risks of pretrial misconduct. Strength of community ties is another key factor necessary to ensure successful supervision. In August 2009 the New York City Criminal Justice Agency (CJA) introduced an experimental Supervised Release (SR) program In the Queens Criminal Court. Based on the success of that program the New York City Office of the Criminal Justice Coordinator contracted with CJA to develop a similar three-year demonstration project in the New York County (Manhattan) Criminal Court, which was implemented in April For an excellent review of New York s practices in comparison with other jurisdictions see Mary T. Phillips, New York City s Bail System A World Apart, New York City Criminal Justice Agency RESEARCH BRIEF series No. 30 (2012), New York: New York City Criminal Justice Agency, Inc. available at

6 2 The interval between CJA s start-up of its Queens and Manhattan programs saw renewed national, state and local attention on issues of bail and pretrial release. In May 2011, the Office of Justice Programs of the U.S. Department of Justice in conjunction with the Pretrial Justice Institute convened a symposium to assess the accomplishments and challenges in bail reform in the decades since the 1964 National Conference on Bail and Criminal Justice. A key focus of this symposium was the assessment that throughout state court systems access to money, rather than risks of pretrial misconduct, continue to be the basis for pretrial release decisions. 2 The symposium s renewed call for bail reform was underscored in the 2013 State of the Judiciary address delivered by New York State s Chief Judge Jonathan Lippman on February 5, In his address he called for major changes in the State s bail statute and the expansion of community supervision as an alternative to money bail, citing CJA s Queens program. 3 The interest in CJA s supervised release programs frequently results in requests for more information about the criteria and screening processes used, and is the subject of the first section of this report. The second section of this report follows the process of how, when and by whom decisions were made during the first six-month start-up period of the Manhattan program. It is designed to offer readers an illustration of the how the program s screening process works in an applied setting. Introducing anything new into an established workgroup environment such as arraignment courtrooms always will be expected to be approached with some wariness and caution among all stakeholders. And in the case of CJA s Manhattan Supervised Release (MSR) program this was magnified by the amount of heightened scrutiny when the program was implemented against the backdrop of renewed interest in pretrial release decision making. Therefore the data presented in part two should be viewed in this context, with the expectation that some changes are likely to occur as the program matures and becomes more established. 2 A summary of the proceedings from this conference are available on the Pretrial Justice Institute s web site, 3 The full text is available from the new York State Office of Court Administration web site,

7 3 CJA S SUPERVISED RELEASE PROGRAMS There are many criteria used by program court staff in the process of screening prospective cases and defendants for the opportunity to be offered supervised release. Figure 1 provides an overview of the decision-making modes developed to implement the program s design. It sequentially identifies each of the key decision points and criteria used as program court staff reviews cases and defendants appearing on the court calendars of each covered Criminal Court arraignment session. The remainder of this section of the report examines each stage in the decision model in greater detail. Charge Criteria Court staff review the calendar for each covered session to identify cases awaiting arraignment on selected non-violent felony charges. These are the charges in each program location for which planning research showed a high likelihood of bail setting at Criminal Court arraignment. In both the Queens and Manhattan program locations this includes most felony drug charges such as the B-felony charges of sale or possession of narcotics and property crimes such as grand larceny. In Manhattan, but not Queens, eligible charges also include fraud/theft charges such as possession of a forged instrument or identify theft. Defendants in these cases are proactively screened further. However, charge-eligible cases that appear to involve domestic violence are excluded at the outset. CJA Release Recommendation, Criminal Conviction History and Related Items CJA serves as the City s pretrial services agency, using a statistically validated risk-assessment instrument to make a release recommendation based on the likelihood that defendants will fail to appear if released on recognizance. Defendants with the lowest FTA risk are Recommended for ROR. Research during the planning phase showed that among the cases of defendants Recommended for ROR, those without other docketed cases on the DCJS criminal history report (i.e. rap sheet) at the time of the screened case have the greatest likelihood of being released, all other things being equal. Defendants in charge-eligible cases in this subset of Recommended for ROR category are excluded from further consideration as one of the safeguards against SR becoming a substitute for recognizance release. (In the report tables this subset is shown as Recommended for ROR, first arrest and referred to in text as the cases of defendants Recommended for ROR without a criminal history.) Defendant screening can continue in all other charge-eligible cases, including those Recommended for ROR with criminal histories.

8 4 FIGURE 1 SUPERVISED RELEASE DECISION MODEL STEP 1: Find Felony Cases Ineligible STEP 2: Screen Charges Eligible Recommended for ROR and first arrest indicator is yes (except by referral) No interview and defendant bypassing Central Booking STEP 3: Locate & review CJA Interview Form for Release Recommendation and prior arrest indicator Recommended with prior arrests Moderate FTA Risk High FTA Risk (evaluate warrant history for mitigating factors) All other recommendation categories Too many convictions STEP 4a: Review number of criminal convictions from CJA Interview Form or rap sheet No more than 1 felony and/or 6 misdemeanors Too many or too recent pretrial warrants Adult VFO conviction w/in 10 years Violent history Hold on defendant STEP 4b: Use rap sheet and complaint to review all charges and relevant criminal history items including any warrant history None or no recent pretrial FTA Return within 30 days No violent criminal history Refused Unable to locate defense attorney STEP 5: Seek defense permission for defendant Interview. Unable to locate defendant Defendant unwilling Defendant unable to participate in interview (e.g. language barrier) No local address/phone Cannot verify contact info STEP 6: Interview defendant Verify local address and phone contact info Obtain supplemental information as needed Not raised in court: ROR Remand Bail set & made STEP 7: Inform defense of program willingness STEP 8: Application made to court Rejected Accepted

9 5 The numbers of adult felony and/or misdemeanor criminal convictions, if any, are examined. Defendants are rejected by the program if they have more than six convictions to misdemeanor crimes and/or more than one adult non-youthful offender (YO) felony conviction. In addition, the felony conviction cannot be to a Violent Felony Offense (VFO) if it occurred within the past ten years. Defendants in charge-eligible cases are considered criminal history ineligible by the program staff during the initial screening process if: > They are Recommended for ROR and are without a criminal history. > There is more than one adult non-yo felony conviction. > There is a felony conviction for a non-yo adjudicated VFO crime within the past ten years. > There are more than six convictions to misdemeanor crimes. Defendants who satisfy the initial criminal history criteria are further screened for any history of bench warrants for FTA in prior cases. This is especially salient in cases of defendants considered at high FTA risk and not recommended for unsupervised recognizance release. Defendants can fall into this CJA risk assessment category if there ever was an FTA warrant issued in a prior case and/or based on the absence of sufficient community ties information at the time of the initial pre-arraignment interview. For defendants with any prior or active warrant, court staff examines the total number of warrants, if any, and the circumstances of the most recent warrant. For example, is the most recent warrant for a post-arraignment pre-disposition FTA versus a warrant for failing to appear for a Desk Appearance Ticket arraignment or a postadjudication FTA (e.g. failure to pay a fine); how long ago did this occur; did the defendant return within 30 days of the issuance of a bench warrant for FTA? > Defendants with many or a recent pattern of pretrial FTA are rejected from further consideration by the program. If there is any doubt, the court staff consults with the Program Director or Manager. Also researched by court staff is whether there may be other types of criminal history circumstances that could preclude program participation. > Other types of criminal history factors include an ICE or Parole hold, or the nature or number of open cases. Defendant Access and Interview Program staff will seek to contact and interview defendants in charge-eligible cases who have not already been disqualified by the program for continued screening based on the above factors. The staff will then speak to the defense attorney to verify if there is a significant likelihood of bail being set in the case, and if so, explain the program participation requirements and gauge the defense attorney s interest in supervised release in lieu of bail. This gatekeeping function by the defense, along with

10 6 using research-based criteria and incorporating courtroom experience, is a critical piece of the program design to avoid net widening. Defendants are excluded from further screening if the defense attorney refuses access. In addition, the defendant him/herself must be able and willing to be interviewed by program staff, and willing to participate in the program if offered. Community Ties To make community supervision possible, the program is able to accept willing defendants only if they have program-verified ties to the local-area community. During the defendant interview process information already collected by CJA during its prearraignment interview for release-recommendation purposes will be reviewed with, and additional information obtained from, the defendant. Defendants who are street homeless or who cannot provide a verifiable residence in the New York City area must be rejected by the program. The CJA pre-arraignment interview process is a high volume operation conducted within a narrow time frame. In addition to obtaining criminal history information for virtually all cases of defendants held for arraignment, CJA Operations staffs attempt to interview all of the defendants held prior to arraignment in a Central Booking facility. Even when defendants are able to provide contact information, CJA interview staff rarely will be able to make more than one attempt at phone verification before the interview forms must be completed and added to the Court package complaint, rap sheet, warrants, CJA interview form with release recommendation. However, after the court papers are assembled there can be a many hours lag before defendants appear for the arraignment. The SR program staff uses a portion of this interval to accomplish defendant interviews and verification attempts. Assignment to the Supervised Release Program For proactively screened cases, if program court staff is satisfied that there are sufficient verified local ties the defense attorney will be informed that the defendant will be accepted by the program if it is offered by the Court. However, in these cases it is up to the defense attorney to raise the program option to the Court. In the adversarial courtroom setting the prosecution is first to speak in regard to pretrial release or bail and it is expected that the defense will raise the program option only when there is a bail request. This is another way in which the program is designed to achieve jail displacement. However, there is nothing in the program model that would prevent the prosecution and defense to discuss the possibility of the SR option prior to arraignment, or for the prosecution to raise or consent to SR once raised.

11 7 Sometimes program staff will review cases and defendants upon referral. These are almost always cases raised on the record at the arraignment in which the program already excluded/rejected the case or defendant, or did not have the opportunity to complete the screening process in advance of arraignment. Once raised on the record in cases in which the program is prepared to accept the defendant, the ultimate decision rests with the Criminal Court arraignment judge. Data Collection During the Screening Process Program court staffs document their review of every charge-eligible case and defendant by completing a screening form which is reviewed and entered into a dedicated computerized program information system. This form includes information about the nature of the charges, CJA release recommendation, defendant criminal history, other criminal history items, information about community ties collected during a defendant interview, and interactions with all participants in the courtroom workgroup, such as defense attorneys, as applicable. The form also contains a summary section for identifying program cases or the source and reasons for rejections for nonparticipant cases. The identifiers in the program s information system can be linked to the Agency s main database for court processing and case outcome information and other supplemental information about cases and defendants collected during the Agency s pre-arraignment interview. IMPLEMENTING THE MANHATTAN SUPERVISED RELEASE PROGRAM: THE FIRST SIX MONTHS, April 8 through October 8, 2013 The information captured for each stage of the screening process in the program s database, augmented by CJA database information, is discussed in this section. At the outset, Figure 2 illustrates some of the results of the screening process for the Manhattan Supervised Release (MSR) program during its first six-month start-up period. Among other items it shows the distribution of screened cases among three general categories cases with defendants found ineligible during initial criminal history screening, cases rejected/excluded after additional screening and the program client cases. It also shows some of the characteristics among the three groups of cases. The criteria and decision making processes limit the number of defendants eligible to participate in CJA s Supervised Release program. During the first six-months of the MSR program, court staff screened 2,718 cases. Defendants in 218 of these cases became program participants and a combined total of 2,500 were excluded or rejected.

12 8 FIGURE 2 MSR SCREENING, APRIL 8 OCTOBER 8, 2013 CHARGE ELIGIBLE SCREENED CASES N=2,718 INELIGIBLE CRIMINAL HISTORY N=1,366 NON PARTICIPANT CRIMINAL HISTORY ELIGIBLE CASES N=1,134 PROGRAM PARTICIPANT CASES N=218 PROGRAM EXCLUSIONS DUE TO: RECOMMENDED for ROR & FIRST ARREST IS YES=352 TOO MANY PRIOR CONVICTIONS=955 PRIOR FELONY IS AN ADULT VFO=52 ELIGIBILITY COULD NOT BE DETERMINED=7 CJA RECOMMENDATION RECOMMEND ROR& NOT FIRST ARREST=184 MODERATE FTA RISK=381 HIGH FTA RISK=384 BENCH WARRANT=92 INCOMPLETE INTERVIEW=55 OTHER CATEGORIES=11 MISSING INTERVIEW=27 REJECTED BY PROGRAM=670 DEFENSE=262 DEFENDANT=19 JUDGE/COURT=137 NOT RAISED=46 CJA RECOMMENDATION RECOMMENDED for ROR & FIRST ARREST IS YES=10 RECOMMENDED for ROR & NOT FIRST ARREST=38 MODERATE FTA RISK=63 HIGH FTA RISK=86 BENCH WARRANT=12 INCOMPLETE INTERVIEW=4 OTHER CATEGORIES=3 MISSING INTERVIEW=2 CHARGE TYPE DRUG=686 PROPERTY CRIME=451 HARM=53 FRAUD/THEFT=156 MISCONDUCT=13 OBSTRUCT JUSTICE=6 OTHER=1 CHARGE TYPE DRUG=452 PROPERTY CRIME=382 HARM=62 FRAUD/THEFT=206 PROSTITUTION=1 MISCONDUCT=18 OBSTRUCT JUSTICE=12 OTHER=1 CHARGE TYPE DRUG=99 PROPERTY CRIME=81 HARM=10 FRAUD/THEFT=24 MISCONDUCT=1 OBSTRUCT JUSTICE=1 OTHER=2 (VTL) CHARGE SEVERITY A FELONY=4 B FELONY=536 C FELONY=42 D FELONY=471 E FELONY=307 A MISDEMEANOR=6 CHARGE SEVERITY A FELONY=10 B FELONY=323 C FELONY=49 D FELONY=423 E FELONY=325 A MISDEMEANOR=4 CHARGE SEVERITY B FELONY=67 C FELONY=11 D FELONY=92 E FELONY=48

13 9 The majority of defendants in screened non-participant cases will be excluded or rejected by the program. Of the 2,500 non-participant cases the program staff had to exclude 1,366 (54.6% of non-participant cases) during the initial screening phase based on criminal history related eligibility criteria: In 352 of these cases defendants without any criminal history were Recommended for ROR. In 955 cases the defendants had more than a single felony conviction and/or more than six misdemeanor convictions. An additional 52 cases were rejected when it was found that the single felony conviction was for a VFO within the past ten years. There were seven cases of defendants that bypassed the Central Booking facility so that there was insufficient information to determine eligibility. There were 1,134 non-participant cases that received further evaluation and of these the program rejected/excluded 670 cases (59.1% of the non-participant criminal history eligible cases). Among the most common reasons for this were: Other criminal history issues such as extensive warrant histories, or the nature or number of open cases; A lack of verifiable or verified local community ties or residences too far away from the program site; and, Cases arraigned before the program was able to complete its review. Defense attorneys are effective gatekeepers and reduce contact with defendants likely to be released at arraignment. There were in total 262 non-participant cases in which defense attorneys refused a program interview with defendants who so far had met the program s criteria. There were 220 of these cases continued at arraignment in which defense counsel refused access based on an expectation of ROR. In over 90% of these cases the defendant received ROR. Defendants almost always will consent to the interview and the program option if offered. There were only 19 instances in which the defendant refused to be considered, and in eight of these cases the defendant expressed a preference for bail. It is expected that the program option will be raised on the record in cases in which the District Attorney s (DA) Office is seeking bail. In addition, the Assistant District Attorney (ADA) may choose to consent or object to the program or even raise the program as an alternative to bail.

14 10 The program option was raised on the record in a total of 384 cases, including not only those rejected by judges but also some cases subsequently rejected by the program, defense or defendant. The position of the DA s office was known in all but 33 of these cases. Among the 218 cases of program participants the ADA position was recorded in 199 of the cases. Bail was requested in 196 of the cases; in two cases the DA requested bail but would consent to SR as a bail alternative; and in one case consented outright to the program in lieu of any bail request. The DA s office placed an objection to the program on the record in 16 of the program participant cases. Among the 166 cases rejected after the program was raised on the record the DA s position was known in 152 cases. Bail was requested in 146 of the cases with consent to the program in six cases. Among the non-participant cases in which the program was raised on the record, the DA s office placed an objection to the program in lieu of bail in eleven of these cases. Bail was set in nine of the cases, with an ROR release in the other two cases. Judges make the final decision regarding supervised release. There were 137 instances in which judges rejected the supervised release option. In 31 (22.6%) of these cases the judge released the defendant on personal recognizance, set bail in 105 (76.7%) of the cases among which two cases had defendants able to post bail at the arraignment, and remanded the remaining case. Judges infrequently provide an on-the-record explanation for choosing bail over supervised release in cases in which MSR was prepared to accept the defendant. Defendant criminal history, other open cases or current charges were most often cited among the small number of cases for which a judicial rejection reason was given. In most instances the program alternative will be raised on the record by defense attorneys in instances in which the program has vetted defendants and found them qualified. However, there can be instances in which the program alternative will be raised in cases in which the program already may have excluded/rejected the case or did not have the opportunity to fully assess the defendant s eligibility. There were 92 cases in which SR was raised on the record as a result of a referral during the arraignment process. Judges were the most frequent referral source, accounting for 71 cases. Defense attorneys made 17 on-the-record referrals, with the DA s office and defendants themselves each accounting for two cases.

15 11 Among the 71 cases in which SR was raised on the record by judges, the program accepted defendants in 62 of the cases and rejected nine. In most of these cases the defendants were found to have met the CJA recommendation and criminal history criteria but were disqualified based on unverified community ties, other criminal history issues, personal circumstances, or because the defense attorney had expected ROR. In all of these cases negotiations and information provided after the program was raised in court were sufficient to resolve the impediment to program acceptance. There were eight cases in which defendants were Recommended for ROR without an adult criminal record and in each of these cases the defendants were accepted when it was clear that without the program the judge would set bail. There was one case in which the defendant had a VFO conviction dated 9 ½ years earlier. The program accepted the defendant upon the consent of both the ADA and Program Manager. The most common reason for program rejection of judicial referrals was because of insufficient community ties. Among the 17 defense attorney referrals the program ultimately accepted defendants in 12 of the cases and rejected defendants in the other five. Defendants in all but one of these cases were criminal history eligible but had initially been disqualified by the program due to a potential conflict with court appearances in pending cases or personal circumstances. There was a single case in which the defendant was Recommended for Release without a criminal history but an exception was made by the Program Director based on the defense referral. In total there were two cases in which the ADA initiated the on-the-record discussion of Supervised Release. In one case the defendant had been rejected by the program because of a potential conflict with court appearances in pending matters, but agreed to accept the defendant who was otherwise qualified for the program. There was one case in which the ADA raised SR in the bail application in the event the judge was considering an ROR. The judge in this case rejected SR and chose ROR because the defendant had no criminal history. Sometimes the SR option will not be raised on the record even though the program was prepared to accept the defendant. There were 46 cases of eligible defendants in which the program option was not raised on the record. There was an ROR in 39 of these cases. Among the other seven cases, bail was set in five, one case had a remanded defendant and the remaining case was disposed at the arraignment.

16 12 In most instances SR is raised in the course of the arraignment process without the need to second call the case. Overall 83 of the 384 cases raised on the record had a second call. There were 62 participant cases with a second call, 47 of which were judicial referrals. There were 18 rejected cases with a second call, 10 of which were judicial referrals. The crime types of the first complaint charge in the cases differ among the participant and non-participant categories. The first complaint charge in the overwhelming majority of all screened cases falls into the drug, property or fraud/theft categories, but these are found in different proportions among the categories. Drug charges make up the largest percentage of cases of defendants excluded due to an excessive criminal history. The second largest percentage of drug cases is found among program participants, 45.4%. Charge-eligible cases in the property crime category have the largest percentage of the cases excluded because defendants without criminal records are Recommended for ROR. Property crimes make up the second largest category of charges for program participants, (37.2%) only somewhat greater than the percentage of propertycrime category cases found in the eligible but rejected category (33.7%). Fraud/theft category charges make up the greatest percentage of cases of defendants without criminal records excluded due to an ROR recommendation, 23.0%, and the smallest percentage of the cases of defendants rejected for excessive criminal records, 7.3%. Fraud/theft category charges make up the third largest percentage of Manhattan s participant cases, 11.0%. Arraignment Charge Characteristics Participants N=218 Criminal History Eligible but Rejected N=1,134 Not Criminal History Eligible N=1,007 (Excludes 7 cases) Recommended for ROR with No Criminal Record N=352 Charge Type N % N % N % N % Drug Property Harm Fraud/Theft

17 13 Misconduct Obstruct Justice Other There are some differences in the distribution of charges within crime categories among the participant and non-participant categories. There are differences in the distribution of charges in the drug category. The B-felony narcotic sale charge (PL ) is a far larger percentage of the top arraignment charges among cases of defendants with excessive criminal histories, 53.4%, and the smallest percentage, 21.2%, of program participant drug cases. B-felony drug possession (PL ) has the greatest percentage of drug charges among participant cases, 46.5%, and also among the cases of defendants excluded when Recommended for ROR without a criminal record, 45.1%. The greatest percentage of cases arraigned on the D-felony possession charge (PL ), 15%, is found among program participant cases in comparison with the other eligibility categories. Arraignment Charge Characteristics Participants Eligible but Rejected Not Criminal History Eligible Recommended for ROR with No Criminal Record Drug Charges N % N % N % N % PL PL PL PL PL PL PL PL PL PL All Other Drug Total Drug

18 14 There are differences in the first complaint charges found in the cases in the property-crime category. The greatest percentage of property cases in each eligibility category is the E- felony grand larceny (PL ) charge, but this charge makes up a smaller percentage of participant cases (35.8%) in comparison with the other categories. The largest percentage of cases arraigned with the top charge of D-felony grand larceny (PL ) is found in the Recommended for ROR with no criminal record (32.5%) and program-participant categories (27.2%). This charge makes up a far smaller percentage of cases of those not eligible due to an excessive criminal history (11.0%). The third largest percentage of participant cases (17.3%) is those with a D-felony burglary (PL ) top arraignment charge. This charge makes up only 4.2% of the cases excluded in the Recommended with no criminal record category and 30.0% of the cases of defendants with excessive criminal histories. Arraignment Charge Characteristics Property Charges Participants Eligible but Rejected Not Criminal History Eligible Recommended for ROR with No Criminal Record N % N % N % N % PL PL PL PL PL PL PL Attempted PL All Other Property Charges Total Property The third largest category of cases has a charge in the fraud/theft category, and here too there are differences in the percentage distributions of charges within this crime type by eligibility category. The largest percentage of cases in each eligibility category has a D-felony possession of a forged instrument charge (PL ) but the percentage varies from 44.4% among Recommended for ROR with no criminal record and 45.8%

19 15 among participant cases, to 58.1% among the cases of those excluded due to an excessive criminal history. A quarter of program participant cases (6 of the 24) have a top arraignment charge of D-felony identity theft (PL ), a charge that is less than 5% of the cases in the other categories. Arraignment Charge Characteristics Fraud/Theft Charges Participants Eligible but Rejected Not Criminal History Eligible Recommended for ROR with No Criminal Record N % N % N % N % PL PL PL PL PL PL PL All Other Fraud Charges Total Fraud/Theft The severity of the first complaint charge differs among the participant and nonparticipant categories. Severity differences among eligibility categories to a large extent reflect differences in charge composition among the crime categories. D-felony charges make up the greatest percentages of cases of program participants and defendants excluded when Recommended for ROR without criminal records (42.2 and 41.8 percent respectively). The greatest percentage of cases with B-felony severity charges is found among defendants excluded due to excessive criminal records (47.0%). B-felony cases make up the second largest percentage of program participant cases (30.7%).

20 16 Arraignment Charge Characteristics Participants N=218 Eligible but Rejected N=1,134 Not Criminal History Eligible N=1,007 (Excludes 7 cases) Recommended for ROR with No Criminal Record N=352 Charge Severity N % N % N % N % A felony* B felony C felony D felony E felony A Misdemeanor** *Cases with a non-violent A-felony charge were not excluded from screening during the first weeks of the program s implementation. **These are referrals from early in the program. Other differences are found in comparing the participant and non-participant defendants and cases. There are differences in defendant characteristics among program participants and non-participants. Defendants excluded because of their extensive criminal histories are proportionately older, male and non-hispanic black in comparison with participants and with non-participants in the other rejection categories. The average (mean) age of non-participants excluded due to an excessive criminal history is 42 with a median (midpoint) age of 44. Program participants and those excluded from consideration due to being Recommended for ROR and not having a criminal record have the most similar age distribution. The average age for participants is 30 with a median of 26 years; the average age for the excluded Recommended for Release group is 30 with a median age of 27. The representation of women defendants differs among the categories. The smallest percentage of female defendants, 8.2%, is found among the cases excluded due to extensive criminal histories. The largest percentage of female defendants is found in the cases excluded due to an ROR recommendation with no criminal histories, 33.5%. Program participants have the second largest group of female defendants, 28.0%.

21 17 In every category the overwhelming majority of screened cases have defendants of color. The greatest percentage of cases with non-hispanic white defendants, 19.9%, is found among the cases excluded due to an ROR recommendation with no criminal history. Less than a tenth of program participants are non-hispanic white defendants. Demographic Characteristics Participants N=218 Eligible but Rejected N=1,134 Not Criminal History Eligible N=1,007 (Excludes 7 cases) Recommended for ROR with No Criminal Record N=352 Age Group N % N % N % N % Mean age Median Age Sex N % N % N % N % Male Female Ethnicity N % N % N % N % Black Hispanic White (Non- Hispanic) Other Unknown

22 18 Criminal conviction characteristics differ among the categories examined. The proportions of cases of defendants with New York State adult convictions to crimes of misdemeanor and/or felony severity differ among some of the categories. Over three-fifths of program participants have no New York State adult convictions to crimes of misdemeanor or felony severity, very similar to those found criminal history eligible but rejected after additional screening. Over four-fifths of defendants in the cases excluded because of their criminal histories have convictions to both misdemeanor and felony crimes. Criminal Conviction History Criminal Convictions No Prior Convictions Misdemeanor Only Participants N=218 Eligible but Rejected N=1,134 Not Criminal History Eligible N=1,007 (Excludes 7 cases) Recommended for ROR with No Criminal Record N=352 N % N % N % N % Felony Only Both Screened cases have arrests from throughout Manhattan s precincts. In addition, because all Special Narcotics Office cases are prosecuted in New York County, there are some screened cases with arrests outside of Manhattan. The greatest numbers and percentages of screened cases have arrests in the 14 th (Midtown South), 25 th (primarily East Harlem), 13 th (lower Midtown Manhattan), 18 th (Midtown North) and 9 th (East Village) precincts, although the percentages vary among eligibility categories. The greatest number and percentage of program participants were arrested in the 14 th precinct (11.0%), with a tie between the 9 th and 25 th precincts for the second largest numbers. Among eligible but rejected cases the greatest percentage came from arrests in the 14 th precinct (8.8%) with the second largest group being arrests in the 18 th precinct. Among those with excessive criminal records, the greatest numbers and percentages of arrests also were in the 14 th precinct (12.4%), with the second largest group coming from the 25 th precinct.

23 19 In the recommended for ROR with no criminal record category the greatest number and percentage of arrests originated in the 18 th precinct (11.6%), with 14 th precinct arrests constituting the second largest group. PRECINCT OF ARREST PARTICIPANT ELIGIBLE BUT REJECTED NOT CRIMINAL HISTORY ELIGIBLE RECOMMENDED FOR ROR- NO CRIMINAL HISTORY TOTAL N % N % N % N % N % BKLYN QUEENS SI BRONX Total

24 20 The volume and composition of arrests differ across neighborhoods and their police precincts, and this is reflected in screened arrests. For example, precincts with large numbers of retail stores have a comparatively higher proportion of property arrests than will be found in more residential Manhattan neighborhoods. (Data not shown) The greatest numbers and percentages of arrests in the drug category occurred in the 25 th precinct, covering the East Harlem neighborhood. Property crime arrests proportionately were greatest in the 14 th and 18 th precincts covering areas of Midtown Manhattan with busy commercial areas. The greatest proportions of fraud/theft category arrests occurred in the 18 th, 19 th, 14 th and 5 th precincts. Defendants in almost all screened cases with a known residence zip code report an address within the City s five boroughs. Manhattan zip codes are the most numerous and comprise somewhat under half (45.4%) of the defendants in screened cases, followed by Bronx and Brooklyn zip codes. The greatest percentage of defendants with Manhattan zip codes is found among cases rejected due to excessive criminal records (53.1%) followed by program participants (46.8%); the smallest percentage is found among cases in the excluded category of Recommended for ROR with no criminal record (39.5%). Among non-manhattan City residents, defendants with Brooklyn zip codes are the most likely to be found among program participants (23.8%) and those excluded as Recommended for ROR with no criminal record (22.9%). In comparison to Brooklyn, there are smaller percentages in these two categories among cases of defendants with Bronx zip codes, and larger percentages of defendants with Bronx zip codes among those rejected due to excessive criminal records. Zip Code of Residence Participants N=218 Eligible but Rejected N=1,134 Not Criminal History Eligible N=1,007 (Excludes 7 cases) Recommended for ROR with No Criminal Record N=352 Location N % N % N % N % N % Brooklyn Manhattan Queens Staten Island Bronx Subtotal New York City Total

25 21 Westchester Long Island New Jersey Connecticut Everywhere Else Total, known zip code Unknown zip code or homeless Total Among the screened cases with Manhattan residents, the only strong relationship between zip codes and arrest precincts appears among cases in the drug category. There is a high concentration of program screened drug cases from precincts in northern Manhattan neighborhoods (e.g. Harlem, Washington Heights) and in lower Manhattan areas covering the Lower Eastside and East Village, with majorities of Manhattan arrestees for drug crimes also having zip codes that correspond with these neighborhoods.

26 22 SUMMARY CJA created a supervised release program, designed to provide judges at Criminal Court arraignment an alternative to money bail and pretrial detention, after extensive consultation with City criminal justice policy makers and other stakeholders. The program was first implemented on an experimental basis in the Queens County Criminal Court in August Based on the demonstrated success of this initial program, a similar three-year demonstration project was implemented in Manhattan in April There are many restrictions placed on the characteristics of eligible felony cases and defendants which require intensive screening by program court staff. A consequence of these requirements is that the program option only will be made available in a limited subset of all screened cases. There are several key screening segments. At the outset, once charge-eligible cases are identified, the CJA recommendation and criminal conviction histories of defendants are reviewed. The program excludes defendants Recommended for ROR without criminal histories because they have the lowest FTA rates and also are the most likely to receive ROR. At the other end of the spectrum are the cases of defendants with multiple criminal convictions or a felony conviction for a VFO crime within the past ten years. Defendants in this category are the most at risk for pretrial misconduct for both FTA and re-arrest. For the cases that remain eligible for further screening the next segment is court staff examination of other criminal history factors that could preclude program participation, and strength and verifiability of community ties needed for community supervision. Within this segment the program requires the consent of both defense counsel and defendants for a face-to-face interview. For defendants who pass all of the screening criteria the final step is presenting the program option on the record during the Criminal Court arraignment process. Because the program is designed to offer the Court community supervision as an alternative to money bail, the expectation is that the program option only will be raised in cases in which the District Attorney s Office is requesting bail. How all of these elements operate in an actual courtroom setting can be seen in following the screening of the case and defendant selection process during the first six months of the implementation of the Manhattan Supervised Release (MSR) program. However, because the data reported cover only a start-up period, some changes may occur over time as the program becomes more familiar to all the stakeholders and more routinely integrated into the arraignment process.