Toward a Model Act for the Prevention and Remedy of Erroneous Convictions Michael J. Saks, Lauri Constantine, Michelle Dolezal, Jennifer Garcia, Carrie Titus, Gary Horton, Tom Leavell, Jonathan Muntz, Londa Rivera, Jennifer Stewart, Fredrica Strumpf & Holly VanderHaar I. INTRODUCTION When a person who has been wrongly convicted and imprisoned is later excluded as the perpetrator by DNA typing, 1 that person is thereby exonerated more convincingly than had ever been true of cases of exoneration in the past.2 While returning innocent individuals to freedom from wrongful imprisonment has unquestionable value, DNA exonerations have created the potential for making far more important improvements in our criminal justice system. Indeed, DNA typing has opened an exceptionally valuable window through which to view the criminal justice system. By analyzing the now-visible cases of wrongful conviction, we can discover what went wrong in those cases, identify patterns of causes of error among the cases, and make specific repairs to the investigation and trial process. In making those repairs, we can prevent erroneous convictions (not only correct them after the harm has been done), and we can prevent errors in many kinds of cases (not only those in which the perpetrator’s DNA has been left at the crime scene). However, the opportunity to take advantage of the unique view from this window is time-limited. The window will soon close, and the view it affords will fade. Some of the erro- Professor Saks is a member of the faculty at the Arizona State University College of Law. The other authors are students at the Arizona State University College of Law and members of the seminar drafting the Model Act. Other me mbers of the seminar are S. Michael Levin and Robert Pastor. 1. Much of the DNA exoneration work has been done by Barry Scheck and Peter Neufeld and the numerous “Innocence Projects” (some of them in prosecutors’ offices) that they have inspired. Many of these cases have been described in BARRY SCHECK ET AL., ACTUAL INNOCENCE: FIVE D AYS TO EXECUTION AND OTHER DISPATCHES FROM THE WRONGLY CONVICTED (2000). 2. Barry Scheck, Peter Neufeld, and Jim Dwyer are by no means the first to discuss the problem of wrongful convictions, but DNA typing is the first method to expose those errors with virtual certainty. Whether the errors brought to light by lawyers and researchers of the past century were truly errors has always been debatable. After DNA exonerations, the systematic flaws of the criminal justice system have risen to the plane of the indisputable. 669 670 NEW ENGLAND LAW REVIEW [Vol. 35:3 neous convictions that occurred before the advent of DNA typing can now be re-evaluated with DNA evidence. This has made some of the mistakes of the past visible. Nonetheless, this unique situation has already started to disappear. In cases where DNA typing can be performed it will routinely be performed, and the post-conviction DNA exoneration cases that today are almost commonplace will disappear. That will be a good thing, as far as the subset of cases involving biological markers is concerned. However, the opportunity to see basic flaws in the criminal justice process will disappear, and the flaws themselves will remain. With the exception of those excluded by DNA typing, the resulting erroneous convictions, wrongful incarceration, and the occasional execution of innocent persons will necessarily continue unabated. The most important thing that can be done with the rare and timelimited view afforded by DNA exonerations is to identify the systemic flaws in the criminal justice system that produce errors and work to cure those flaws. By taking advantage now of the fact that defects in the investigation and trial process have been made temporarily visible, we can make repairs that will broadly and permanently improve the workings of the system. II. SEMINAR GOALS AND PROCESS The work described in this article is being performed by a group of Arizona State University law students in a seminar course. Their ultimate product will be a Model Act embodying a comprehensive set of criminal justice system reforms. The goal of these reforms is to reduce the probability of an erroneous conviction, without reducing the probability of a correct conviction. That is, each reform is calculated to prevent innocent people from being convicted while preserving the ability of the system to correctly identify and convict guilty persons. The seminar has proceeded through the following steps. During the fall semester, members of the seminar: (a) researched existing studies and existing and proposed law related to the criminal justice system’s problem areas; (b) met to discuss the findings of their research; and (c) debated policy preferences concerning reforms. These meetings sometimes included visits from persons with special knowledge of the topics under discussion.3 During the spring semester, the seminar: (d) will transform its selected policies into a draft Model Act, complete with commentary; (e) circulate the draft Model Act to interested parties around the nation, and invite their written comments; (f) hold a one-day hearing to which 3. Susan Narveson, Director of the Phoenix Police Department’s crime laboratory and president-elect of the American Society of Crime Laboratory Directors and Scott Bales, former Assistant United States Attorney and currently Solicitor General of Arizona. 2001] TOWARD A MODEL ACT 671 interested persons could testify to the “commission” about the Model Act; then (g) revise the draft act, in light of the comments received, into final form; and (h) publish the final version of the Model Act. The Model Act will then be available to states that wish to adopt all or some of its provisions. III. MAJOR CAUSES OF ERRONEOUS CONVICTIONS The DNA exoneration cases highlight the leading causes of erroneous convictions. Through examination of the evidence in cases that ultimately resulted in DNA exonerations, it is possible to discover what evidence or other conduct in the original trial was erroneous or otherwise flawed. The following Table is updated from the data presented in Actual Innocence: Five Days to Execution and Other Dispatches from the Wrongly Convicted4 and is based on a set of DNA exclusion cases. 5 Some cases involved two or more erroneous sources. The data suggest that the leading causes of erroneous convictions are forensic science errors and eyewitness identification errors, followed by police and prosecutorial misconduct, and finally a variety of other causes. Accordingly, the aim of the Model Act is to reduce the likelihood that these shortcomings will vitiate the investigation, prosecution, and trial of future cases. TABLE 1 FACTORS LEADING TO WRONGFUL CONVICTIONS IN 81 CASES Erroneous Forensic Science Fraudulent, Tainted Evidence Mistaken Eyewitness Identification Police Misconduct Prosecutorial Misconduct Bad Lawyering False Confessions Informants/Snitches False Witness Testimony 4. 5. 53 25 60 26 32 23 15 14 14 See SCHECK ET AL., supra note 1. See id. at 263. (The table is entitled “Factors Leading to Wrongful Convictions in 62 U.S. Cases.”). See id. The statistics in the table are accurate since the table was corrected by omitting one line of data which contains some cases already counted in other lines of the table. See Interview with Peter Neufeld, CoDirector, The Innocence Project, Cardozo School of Law, in Boston, Mass. (Nov. 14, 2000) (on file with the author). 672 NEW ENGLAND LAW REVIEW IV. [Vol. 35:3 POLICIES TO REDUCE THE INCIDENCE OF ERRONEOUS CONVICTIONS The following sub-sections summarize the policy preferences adopted by the seminar members at the midway point in their deliberations. Some of their details may change by the end of the process of drafting, but they provide a reliable summary of the major thinking of the seminar about how to fix what is flawed. The subsections that follow are organized in the approximate chronological sequence of crime investigation and litigation. A. Eyewitness Identification Procedures: Interviewing of Witnesses Police officers receive little training in how to elicit accurate information from witnesses, and are rarely taught the most modern and empirically tested methods of interviewing.6 Accordingly, the Model Act will require that police departments provide initial and ongoing training of all personnel in proper and effective methods of interviewing. This training would be tailored to an employee’s role in relation to witnesses and to the circumstances under which the officer encounters witnesses. The training would be provided in a manner that ensures that the employee has acquired the appropriate skills to deal with such encounters. The Department of Justice Guidelines already prescribe that, in interviewing witnesses, the officer engage in no prompting, ask only open-ended questions, and that the witness refrain from guessing. All interviews related to felony investigations, whether in the field or at the police station, will be audio or video recorded. The specific method of recording would depend upon what type of technology is available and feasible for the purpose of the interview. The recordings would be preserved and possibly used later by investigators, prosecution and defense attorneys and in certain circumstances by judges and jurors as well, to assist in determining the course of events and the individuals involved in them. Whether a witness’s recollections are accurate or not is important to the quality of the investigation that is guided by those statements. To a greater or lesser extent, the taped statements themselves will indicate the internal consistency, clarity, and perhaps the validity of a witness’s statements. It would also be helpful to ask witnesses for a statement of certainty about key aspects of their statement, such as their description of the perpetrator. Such an inquiry would ensure that investigators relying on 6. See Gary Wells et al., From the Lab to the Police Station: A Successful Application of Eyewitness Research, 55 AM. PSYCHOL. 581, 583 (2000) (noting such common errors as asking closed-ended questions, interrupting witnesses in mid-narrative, and asking questions in a rigid, predetermined order). 2001] TOWARD A MODEL ACT 673 the statements would be less likely to focus on weak leads or incorrect suspects. B. Eyewitness Identification Procedures: Lineups and Photospreads Procedures for maximizing accurate identifications and minimizing erroneous identifications have been extensively empirically tested by cognitive and social psychologists who have studied the eyewitness identification process.7 Indeed, these have been more thoroughly and more scientifically studied than any other aspect of the criminal justice process, including much of forensic science.8 The role of eyewitness error in DNA exonerations prompted the U.S. Department of Justice (DOJ) to create a working group to develop guidelines based on the findings of eyewitness research in order to reduce the rate of wrongful convictions attributable to eyewitness errors. Most, but not all, of the major research conclusions concerning how to construct the most accurate eyewitness identification procedures have already been incorporated into the Guidelines recently promulgated by the DOJ.9 The Model Act will embody not only the provisions adopted by the DOJ Guidelines, but all of the procedures found to be effective. In accord with the best available research, photospreads will have a functional size that includes at least five foils, in addition to the suspect, and live lineups will have at least four. Each foil will resemble the description of the perpetrator in all significant respects. Each lineup or photospread will contain only a single suspect. Lineups or photospreads will be sequential. That is, instead of presenting all lineup members at once, they are presented one at a time, and each candidate is then judged to be or not to be the perpetrator. All of the candidates that have been selected for the lineup or photospread are shown to the witness, even if the witness chooses one early in the sequence.10 The information given to eyewitnesses as they approach the task has important effects on the accuracy of their decisions. As a result, they need to be properly prepared for the task. Lineups and photospreads will be preceded by giving witnesses instruc- 7. See, e.g., Gary L. Wells, What Do We Know About Eyewitness Identification?, 48 AM. PSYCHOLOGIST 553 (1993). 8. See generally DAVID FAIGMAN ET AL., MODERN SCIENTIFIC EVIDENCE: THE LAW AND SCIENCE OF E XPERT TESTIMONY (1st ed. 1997 & Supp. 2000). A comparison of the chapter on eyewitness identification with those on the forensic identification sciences offers a sobering lesson about the research foundations on which these fields and their experts stand. 9. See TECHNICAL W ORKING GROUP FOR EYEWITNESS E VIDENCE, U.S. DEP’ T OF JUSTICE, E YEWITNESS EVIDENCE : A GUIDE FOR LAW E NFORCEMENT (1999). 10. The procedures contained in this section are also contained in the Department of Justice Guidelines. See id. at 29-31. 674 NEW ENGLAND LAW REVIEW [Vol. 35:3 tions which explain that previous advice or instruction should be disregarded, that the perpetrator may or may not be in the lineup, that excluding the innocent is as important as including the guilty, and that if the perpetrator is not in the lineup the investigation will continue in an effort to find the actual perpetrator.11 Furthermore, information given to eyewitnesses after they make their selections from lineups and photospreads has important effects on their beliefs about and later testimony regarding the identification they have made. Regardless of whom the witness chooses, a statement of certainty will be taken immediately following the choice and without any feedback being given to the witness. The statement of certainty is to be given in the witness’s own words, and will be audio or video recorded, or written by and signed by the witness. Thus, the witness will receive no information about suspects (or the responses or reports of other witnesses) before any identification procedure, and minimal feedback afterwards. 12 Although there will be no prohibition on informing witnesses, especially victims, about the progress of an investigation, the commentary to the Model Act will advise that good practice dictates that prospective trial witnesses be contaminated by such information as little as possible. The law enforcement officer conducting the lineup or photospread will be blind as to who the suspect is or, in the alternative, the identification procedure shall be videotaped. Blind administration ensures that the officer does not inadvertently give cues to the witness as to which member of the lineup is the suspect, and in turn that the selection of a suspect is entirely the choice of the witness. If properly conducted blind, there is no need for videotaping because cues cannot be given to the witness. If not conducted blind, the procedure needs to be videotaped so that there is a chance to evaluate whether improper cues were given. Show-ups will not be specifically regulated, but the commentary will summarize the problems of show-ups, and suggest that where multiple witnesses are available, they should not all be contaminated by being asked to participate in a show-up. At trial, witnesses who are to be asked to identify or to exclude the defendant as the perpetrator will be sequestered from each other. C. Interrogation of Suspects Presently, much of the training regarding how to conduct an interrogation is aimed at eliciting confessions from suspects, and some of that pro- 11. The procedures contained those contained in the Department of 12. The procedures contained those contained in the Department of in this section are essentially the same as Justice Guidelines. See id. at 31-33. in this section are essentially the same as Justice Guidelines. See id. at 33-37. 2001] TOWARD A MODEL ACT 675 cess involves chicanery of one sort or another. 13 The Model Act will not prohibit any of these interrogation procedures, which have become standard for most law enforcement agencies. However, it will require videotaping of all interviews or interrogations in their entirety of potential or established suspects. The recordings will be automatically time-and-date stamped. Furthermore, the video recording may prevent the most coercive methods from even being used.14 Equally important, because anything done during the interrogation will be accessible to lawyers and judges, the videotape will ensure that the voluntariness and validity of the interrogation and any resulting confession can be assessed. The admissibility of any confession would be subject to a test of its reliability or validity. Upon motion contesting the admissibility of a confession, or upon the court’s own motion, the burden falls on the proponent of admission to prove that the confession is reliable. Commentary will explain that this provision is in addition to the existing constitutional requirement of voluntariness. In making the decision concerning admissibility of a confession, the court will have the duty to carefully consider whether the confession is internally consistent, whether the questioners fed information to the suspect either by using leading questions or otherwise, and whether the statements are corroborated by independent evidence, taking into account the totality of the circumstances. Given the reliability test of admissibility and the importance of the videotape in evaluating that reliability, the absence, for any reason, of a videotape creates extreme difficulties for the judge called upon to evaluate admissibility. Consequently, for interrogations that take place in a police station, the absence of a substantially complete videotape will render that confession inadmissible.15 Confessions made “on the street” in the aftermath of a crime and which have been audio taped according to the guidelines discussed above (which require automatic taping of all felony investigations), will be admissible despite the lack of a complete video recording. Nonetheless, when technology permits, and recording becomes sufficiently easy and inexpensive, the “on the street” rule can be amended to require the same completeness as the “in the police station” rules. 13. See generally, FRED E. INBAU, ET AL., C RIMINAL INTERROGATION AND CONFESSIONS (3d ed. 1986) (Chapter six entitled “Tactics and Techniques for the Interrogation of Suspects Whose Guilt Seems Definite or Reasonably Certain – The Nine Steps to Effectiveness,” is especially important.); ROBERT F. ROYAL & STEVE R. SCHUTT, THE G ENTLE ART OF INTERVIEWING AND INTERROGATION : A PROFESSIONAL MANUAL AND GUIDE 115-30 (1976). 14. These coercive methods include physical abuse, threats of harm, threats to implicate innocent relatives, or other unlawful threats or promises. 15. Several state supreme courts have made electronic recording of interrogations a precondition for admissibility. See, e.g., Stephan v. State, 711 P.2d 1156 (Alaska 1985); State v. Scales, 518 N.W.2d 587 (Minn. 1994). 676 NEW ENGLAND LAW REVIEW [Vol. 35:3 Since enforcement of any set of specific rules regarding the humane treatment of suspects will be difficult or impossible, and because our concern is coercion of innocent suspects, members of the seminar believe that prevention of the worst coercive treatment will be best accomplished by the requirement of videotaping the entire interrogation.16 At the same time, requiring videotaped interrogations will also protect the police, since a videotape can provide evidence to defend against claims of brutality and other unlawful conduct. Similar to the training requirements for the interviewing of witnesses, requirements will be established for the training of those who interrogate suspects. Moreover, given the limited reliable research that exists regarding what are in fact effective and non-coercive techniques for interrogating suspects, a provision of the statute will mandate that the state support valid empirical research on these questions. The findings of all such research must be incorporated into subsequent training. D. Forensic Science Table One suggests that forensic science may surpass eyewitness error as the leading cause of erroneous convictions. A growing body of literature documents a tendency of crime laboratory personnel not only to improperly and unethically view themselves as advocates for one party to litigation, but to go to such lengths as to fudge and fabricate their “scientific” test results.17 A second, and perhaps even more surprising develop- 16. After Miranda v. Arizona, 384 U.S. 436 (1966), if Miranda warnings have been given, the circumstances in which a court will find that a suspect co nfessed involuntarily are rare. Cases where claims of involuntariness were denied include McCall v. Dutton, 863 F.2d 454 (6th Cir. 1988) (wounded suspect interrogated with drawn weapons); United States v. Kelley, 953 F.2d 562 (9th Cir. 1992) (handcuffed suspect confessed while suffering from heroin withdrawal); Sumpter v. Nix, 863 F.2d 563 (8th Cir. 1988) (low-I.Q. suspect with psychological problems interrogated for more than seven hours); and Moore v. Dugger, 856 F.2d 129 (11th Cir. 1988) (low I.Q. suspect had not had food or sleep for 25 hours at the time he confessed). 17. See, e.g., Paul C. Giannelli, The Abuse of Scientific Evidence in Criminal Cases: The Need for Independent Crime Laboratories, 4 VA. J. SOC. POL’ Y & L. 439 (1997) (detailing a lengthy set of fraudulent forensic science; noting esp ecially the extensive fraud committed by Fred Zain, Dr. Ralph Erdmann, and Dr. Louise Robbins); JOSEPH L. PETERSON ET AL., FORENSIC EVIDENCE AND THE POLICE 114 (1984) (finding that of the laboratories studied, on average fewer than 10% of lab reports dissociated a suspect from the crime); Andre A. Moenssens, Novel Scientific Evidence in Criminal Cases: Some Words of Caution, 84 J. C RIM. L. & CRIMINOLOGY 1, 17 (1993) (explaining that “[a]ll [forensic science] experts are tempted, many times during their careers, to report positive results when their inquiries come up inconclusive, or indeed to report a negative result as p ositive.”). See also INSPECTOR GENERAL, U.S. DEP’ T OF JUSTICE, THE FBI LABORATORY: A N 2001] TOWARD A MODEL ACT 677 ment, is the growing realization that some forensic sciences have scientific foundations that are infirm to the point of being nonexistent.18 Remarkably little data has been gathered supporting the claims made in the many fields of forensic individualization science. This has become apparent as challenges have been brought, mostly in the federal courts, under Daubert v. Merrell Dow Pharmaceuticals, Inc.19 and Kumho Tire Company, Ltd. v. Carmichael,20 and the proponents of some of these fields have been unable to meet the rule’s new requirement of validity. 21 When traditional forensic INVESTIGATION INTO LABORATORY PRACTICES AND A LLEGED M ISCONDUCT EXPLOSIVE -RELATED AND O THER CASES (1997). In this report: IN [MAU Chief James] Corby reviewed approximately 200 cases and found many serious flaws in Rudolph’s [former senior explosives residue e xaminer in the FBI laboratory] work. . . . Corby noted that Rudolph had failed to follow his own explosives residue protocol, had formed co nclusions and prepared dictation without a basis, had failed to run standards or confirmatory tests, had offered opinions to fit the case scenario or findings of other units whether or not supported by his own analyses, had failed to label charts properly, and, where data [was] present in the file, had sometimes made technical errors. Id. at 41. 18. See, e.g., United States v. Hines, 55 F. Supp. 2d 62 (D. Mass. 1999) (handwriting); United States v. Fujii, No. 00 CR17, slip op. at 3-4 (N.D. Ill. 2000) (handwriting); United States v. Mitchell, Crim No. 96-407-1 (E.D. Pa. 2000) (fingerprints); Williamson v. Reynold, 904 F. Supp. 1529, 1557-58 (E.D. Okla. 1995) (hair); People v. Linscott, 566 N.E.2d 1355 (Ill. 1991) (hair); Simon Cole , The Myth of Fingerprints, in LINGUA FRANCA: THE REVIEW OF ACADEMIC LIFE, Vol. 10, No. 8, Nov., 2000, at 54. See generally, Michael J. Saks, Merlin and Solomon: Lessons from the Law’s Formative Encounters with Forensic Identification Science, 49 HASTINGS L.J. 1069 (1998); DAVID FAIGMAN ET AL., supra note 8 (discussing forensic identification science in chapters 15-26). See also NATIONAL INSTITUTE OF JUSTICE, SOLICITATION: FORENSIC FRICTION RIDGE (FINGERPRINT ) EXAMINATION VALIDATION STUDIES, Mar., 2000. It is important to note that: The uniqueness of friction ridge patterns, be they fingerprints, palmprints, or bare footprints, has long been accepted by the scientific community and by the courts. The reason for this widespread acce ptance perhaps lies in the fact that fingerprints were first introduced at a time in our history when society was less demanding of proof and more trusting of authority. As a result of widespread recent challenges that friction ridge identification, as well as other forms of personal identif ication evidence, lack a proven scientific foundation, efforts have been ongoing to supply today’s demands of validation and verification that is . . . required by . . . Daubert v. Merrell Dow Pharmaceuticals. Andre Moenssens, Validating Friction Ridge Examination Techniques Proposals Solicited (visited March 23, 2001) <http://www.forensic-evidence.com/site/ID/ ID_fpValidation.html>. 19. 509 U.S. 579 (1993). 20. 526 U.S. 137 (1999). 21. Or “evidentiary reliability,” as the Daubert Court preferred to call it. 678 NEW ENGLAND LAW REVIEW [Vol. 35:3 identification sciences and their inferential methodologies are compared to DNA typing, the shortcomings of the traditional forensic individualization “sciences” become apparent.22 The goal of crime laboratory reform is to ensure that forensic science contributes to determining the truth in a given matter using sound scientific methods. To achieve that goal, the organizational structure of forensic laboratories needs to be redesigned and their culture needs to be reoriented. Under the Model Act, laboratories will be uncoupled from police departments and replaced with independent laboratories. Management and control of the labs will be placed in the hands of real scientists with real scientific training, whose first loyalty is to performing reliable scientific work. These independent labs will be funded by the state, and their budgets will be separate from those of police and prosecution agencies. Furthermore, these laboratories will be managed by scientists rather than police officials and administered as independent agencies under an authority (the Forensic Laboratory Commission) separate from and independent of police or prosecution agencies. Also, records and staff will be equally available and responsive to both prosecution and defense lawyers, consistent with the laboratories’ independent status.23 A Forensic Laboratory Commission will be created that has general supervisory, policy-making, and rule-making authority for the state’s forensic laboratory system. The Commission will be composed of a majority of scientists who are not forensic scientists, but come primarily from industry, academia, medical settings, and other government laboratories. The other members will represent a balance among prosecution, defense, and the judiciary. Its powers will include setting standards for hiring, workload, resources, and quality control methods including rules for certification of examiners, accreditation of labs, blind proficiency testing, external scientific audits of the lab’s work, etc. The results of quality control testing will be on file with the Commission and available as public records. The Forensic Laboratory Commission will also ensure that the state’s labs develop and follow protocols that have been properly validated. Specialties that cannot provide adequately validated protocols may be reduced to an investigatory role, which can be used to produce investigative leads, but not be offered as proof at trial. 24 The Commission will represent the state’s forensic laboratories to the legislature for all purposes, including Daubert, 509 U.S. at 590-91. 22. Michael J. Saks & Jonathan J. Koehler, What DNA “Fingerprinting” Can Teach the Law About the Rest of Forensic Science, 13 CARDOZO L. REV. 361, 361 (1991). 23. This is intended to solve existing problems of discovery and defense counsel’s need for forensic science consultation and other assistance. 24. For example, the FBI has already reduced “voiceprint” identification to such a role. See FAIGMAN, supra note 8, Vol. II, at 194. 2001] TOWARD A MODEL ACT 679 budgetary matters. Furthermore, misbehavior by laboratory personnel such as fudging, falsification, fabrication, perjury, etc. will be subject to administrative penalties within a framework established by the Commission. Where necessary and appropriate, criminal charges may be brought against such employees for obstruction of justice, perjury, and the like. Finally, the Model Act might adopt the provisions of the new Federal Rules of Evidence 702, which essentially codifies Daubert, and governs the admission of expert evidence at trial.25 E. Indigent Legal Representation In many states, legal representation of the indigent is characterized by inadequacy: too few lawyers for the caseload, insufficient skill, and inadequate resources. A recent study of judicial review of death penalty cases found that thirty-seven percent of the reversals of convictions were due to inadequate assistance of counsel.26 Under the Model Act to the extent possible, indigent legal representation will be accomplished through the establishment of a centralized, statewide, public defender system. The public defender organization will be empowered to organize itself into subunits for effective and efficient functioning. This may be done by geographic units, by trial and appellate units, or specialized subject matter areas. In circumstances where a problem of geographic remoteness, seasonal fluctuations in caseloads, or conflicts of interest exist, the statewide public defender is authorized to hire temporary assistant public defenders or to contract with private attorneys, who must meet the same training and experience standards as permanent assistant public defenders.27 Furthermore, the office of the public defend- 25. The recently revised FED. R. EVID. 702, states: If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is sufficiently based upon reliable facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case. Id. 26. See James S. Liebman et al., Capital Attrition: Error Rates in Capital Cases, 1973-1995, 78 TEX. L. REV. 1839, 1850 (2000). 27. To deal with the special problem of conflicts of interest, two quite different approaches might be considered. Kentucky law maintains th at attorneys from separate jurisdictional offices of the same statewide public defender agency are sufficient to avoid the conflict. See People v. Wilkins, 268 N.E.2d 756 (1971); RULES OF PROFESSIONAL CONDUCT Rule 1.10, paragraph 3 of Commentary; KBA 680 NEW ENGLAND LAW REVIEW [Vol. 35:3 er must adhere to certain minimum standards, including caseload limitations.28 It will have resources commensurate with those of the government, such as support staff, investigators, and experts. 29 A permanent commission at the state level will oversee the state public defender’s office. This commission will be within the judicial branch, and will determine what type of system should exist in each county. 30 The commission will develop the necessary standards for providing effective representation for all systems within the state, including required qualifications and performance standards and checks. Its responsibilities would include representing the public defender’s office to the legislature on all matters, including budgetary issues. Among its powers will be the authority to release funds that have been encumbered to support public defender services. F. Prosecutorial Misconduct The problem of prosecutorial misconduct in erroneous convictions consists largely of either the suppression of exculpatory evidence or the presentation of false evidence – either of which might be done inadvertently or deliberately. Since most cases of prosecutorial misconduct are not further investigated, instead of imposing mandatory sanctions to punish offenders after the misconduct has occurred, the Model Act will mandate an “open file” discovery policy. The general rule would be that all evidence, and other relevant information, which is in the possession or control of the government is to be shared with the defense. While several states (among them Florida and New Hampshire 31) already impose broad criminal discovery obligations on prosecutors, these rules often make dis- Ethics Opinion 321; ALI RESTATEMENT (THIRD) OF THE LAW GOVERNING LAWYERS, § 203 (d)(iv)). Colorado, by contrast, has established an alternate d efense counsel, which provides legal representation in cases where the state public defender has a conflict of interest. A nine-member commission appointed by the Supreme Court of Colorado appoints and discharges those who serve as alternative defense counsel. COLO. REV. STAT. § 21-2-101(1) (2000). 28. They might choose to follow the caseload maximums recommended by the ABA: “400 misdemeanors per attorney per year, or 200 juvenile cases per attorney per year, or 200 mental commitment cases per attorney per year, or 25 appeals per attorney per year.” THE ABA STANDARDS FOR CRIMINAL JUSTICE § 55.3 Commentary (3d ed. 1992). 29. The necessity for experts depends upon the nature of the crime laborat ories that exist in a state. Where crime laboratories are independent and equally available to prosecutors and defenders, there will be much less need for additional expert support services. 30. See Terry Brooks, Efforts Increase to Improve State Systems, 15 C RIM . JUST. 59, 59 (2000). 31. See N.H. R. S UPER . C T. 98; FLA. R. C RIM . P. 3.220. 2001] TOWARD A MODEL ACT 681 tinctions between evidence that is to be disclosed and that which may be withheld. The Model Act will eliminate these distinctions, making it unnecessary to define what evidence the government is required to disclose. Generally, all evidence must be disclosed. However, the Model Act’s open file policy will exclude attorney work product, which remains nondiscoverable under the usual rules of criminal procedure.32 Notwithstanding the general open file policy, any party that believes special circumstances exist that should preclude mandatory disclosure may seek a protective order from the court. These circumstances may include, but are not limited to, an unacceptable risk of harm to a witness or an ongoing investigation with multiple defendants when not all arrests have been made. Whenever the court grants such a protective order, it will also issue an order ensuring that the defense has appropriate opportunity to gather any exculpatory evidence that may exist. The court and parties must find a satisfactory way to meet the legitimate goals of the protective order while not defeating the purposes of mandatory disclosure. When violations of the open file discovery policy occur, substantial penalties will be available against the offending persons, regardless of whether the evidence withheld had or would have had any effect on the outcome of the underlying case. However, these penalties will not include overturning the verdict and ordering a new trial unless the prosecutorial misconduct is determined to have caused actual prejudice. G. Post-Conviction Procedures The emphasis on finality in existing law has meant that even the most convincing evidence of innocence simply could not be considered by the courts of many jurisdictions post-conviction. While there is good reason not to permit endless appeals on hair-splitting legal issues, the law needs to make room for inventions that can radically alter the factual conclusions of a case. Accordingly, the Model Act will contain a provision for the filing of a petition alleging “actual innocence,” to be filed with the state’s intermediate court of appeals. If merited, the court will have the power to set aside the conviction and order a new trial. These petitions may be filed without limit as to time and without limit as to the number of such petitions. They must specify new evidence or new analyses of old evidence, which would substantially support a claim of actual innocence. The action permitted under this provision is limited to factual evidence bearing on claims of actual innocence; there is no requirement to show any constitutional violation. Moreover, the filing of these petitions would not impair any other substantive or procedural rights that may be available. 32. See, e.g., FED. R. CRIM. P. 16(a)(2). 682 NEW ENGLAND LAW REVIEW [Vol. 35:3 Since this rule may, at least initially, create a surplus of filings, courts will be authorized to engage in a two-stage review process. This process will allow a judge to review a petition and decide whether it ought to be dismissed summarily or whether further proceedings should be scheduled to more thoroughly consider the facts alleged in the petition. The goal is to achieve a balance between the value of finality and the societal and individual interest in freeing an innocent person from wrongful punishment. In addition, the Model Act will contain rules requiring the preservation of evidence for a period of time that would allow DNA – or other innovations that might be developed long after conviction – to be employed. H. Post-Release Rights Under the Model Act, when a judicial order vacating a conviction on grounds of actual innocence is issued, a claim for compensation may be initiated by the wrongly convicted person or by the court on its own motion, at any time within two years following the order vacating the conviction.33 These damages will be supplemental to tort and civil rights actions that may be available. Compensation is to be granted if the claimant establishes innocence by clear and convincing evidence. This high standard helps to ensure that compensation is awarded only to the actually innocent, and not to persons released only on technical legal grounds. The elements of damages shall include general damages (pain, suffering, loss of enjoyment of life, etc.) taking into account the length and conditions of the unjust confinement; lost earnings and vocational services; costs of rehabilitation or counseling needed to assist reintegration into society and taking into account family circumstances; and attorneys’ fees. No punitive (exemplary) damages will be available under the Model Act. 34 There is no cap on these damages, but the court may order structured payments rather than one lump sum payment. These damages shall be available even if the state was without fault.35 Furthermore, the court may issue an order completely expunging the wrongly convicted person’s criminal justice system records that came into being as a result of the wrongful conviction. 33. See generally Adele Bernhard, When Justice Fails: Indemnification for Unjust Conviction, 6 U. CHI. L. SCH. ROUNDTABLE 73 (1999). Many of the ideas in this section are drawn from the work of Adele Bernhard. 34. Exemplary damages would be available under 42 U.S.C.A. § 1983 (1994). 35. For example, if the erroneous conviction were entirely the fault of a perjurious witness. In such an instance, the state will have no right of subrogation against the perjurious witness, but can prosecute that witness criminally. 2001] TOWARD A MODEL ACT 683 I. Sanctions for Violations The Model Act will contain a schedule of penalties for violation of its provisions. These will range from letters of reprimand to fines and imprisonment. These penalties are intended to promote compliance with the Act’s provisions. However, the only way to achieve the benefits sought by the Model Act will be for its provisions to be taken seriously and that will only occur if they have teeth.