CHAPTER I

Introduction

"I had," said he, "come to an entirely erroneous conclusion which shows, my dear Watson, how dangerous it always is to reason from insufficient data."

Arthur Conan Doyle, The Adventure of the Speckled Band

------------------------------

One way to view science is that it is a search for truth.1 Forensic science is no exception. As Attorney General Janet Reno emphasized, "The use of forensic science as a tool in the search for truth allows justice to be done not only by apprehending the guilty but also by freeing the innocent."2

This report describes a study that focused on the freeing of the innocent -- persons initially convicted and imprisoned but later released through post-conviction forensic use of DNA technology.

------------------------------

Purpose and Scope of the Study

The principal purpose of the study, initiated in June 1995, was to identify and review cases in which convicted persons were released from prison as a result of posttrial DNA testing of evidence. As of early 1996, researchers had found 28 such cases: DNA test results obtained subsequent to trial proved that, on the basis of DNA evidence, the convicted persons could not have committed the crimes for which they were incarcerated.

The study also involved a survey of 40 laboratories that conduct DNA testing.

This report does not probe the strengths or weaknesses of forensic DNA technology when applied to criminal cases.3 The discussion of DNA instead is limited to its use in exculpating convicted defendants serving prison sentences.

The authors do not claim to be scientific experts in DNA technology. This report cites reference materials that probe technological details more deeply than occurs on these pages.

The balance of this chapter outlines the study's design and provides basic background information on forensic DNA identification testing. Chapters II and III, respectively, present the study's findings and their policy implications. The final chapter consists of brief profiles of the 28 exculpatory cases. A glossary defines DNA-related terms, and the appendix reports DNA test results for some of the exculpated persons profiled in this report.

------------------------------

Study Design

To identify cases that met study criteria -- defendant conviction, imprisonment, and subsequent exoneration and release resulting from post-trial exculpatory DNA tests -- researchers examined legal and newspaper data bases and interviewed a variety of legal and DNA experts. Once initially identified as likely candidates for the study, cases were verified and assessed through interviews with the involved defense counsel, prosecutors, and forensic laboratory staff; through reviews of court opinions; and, in some instances, through examinations of case files.

For example, initial identification of the Glen Woodall case resulted from an automated search of newspaper data bases, which identified articles about the case in several West Virginia newspapers, the Philadelphia Inquirer, and the Cleveland Plain Dealer. An opinion by the West Virginia Supreme Court of Appeals in the appeal of Woodall's conviction (State v. Woodall, 385 S.E.2d 253, W. Va. 1989) contained the name of Woodall's defense attorney, who was called and interviewed at length and who provided materials related to the criminal case.

Those materials described improper activities by Fred Zain, once a serologist for the West Virginia State Police. A phone conversation with the West Virginia assistant attorney general handling the Zain misconduct cases resulted in the receipt of public case documents containing extensive details on Zain's activities related to the Woodall investigation and prosecution.

A review of transcripts from the criminal and, later, civil cases yielded the name of the laboratory that conducted the DNA testing that exculpated Woodall. A lengthy interview was conducted with the laboratory's forensic scientist who performed the DNA tests on the Woodall evidence. He provided documentation related to his examinations in the case.

Cases related to a special West Virginia Supreme Court of Appeals investigation into government misconduct surrounding Woodall's case (438 S.E.2d 501, W. Va. 1993; 445 S.E.2d 165, W. Va. 1994) also were reviewed.

Researchers collected information for the survey of DNA-testing laboratories through telephone interviews. An experienced crime laboratory director assisted the Institute for Law and Justice in conducting the survey.

This study, conducted in a short time period with limited funding, reflects a modest level of analysis and focuses on a relatively small number of cases. One can state with confidence, however, that as of the study's completion, the 28 cases identified represent most of the situations in the country where convicted felons had been released from prison on the basis of postconviction DNA testing.4

------------------------------

Background on Forensic Use of DNA Identification

Testing

Perhaps the most significant advance in criminal investigation since the advent of fingerprint identification is the use of DNA technology to help convict criminals or eliminate persons as suspects. DNA analyses on saliva, skin tissue, blood, hair, and semen can now be reliably used to link criminals to crimes. Increasingly accepted during the past 10 years, DNA technology is now widely used by police, prosecutors, defense counsel, and courts in the United States.

An authoritative study on the forensic uses of DNA, conducted by the National Research Council of the National Academy of Sciences, has noted that:

...the reliability of DNA evidence will permit it to exonerate some people who would have been wrongfully accused or convicted without it. Therefore, DNA identification is not only a way of securing convictions; it is also a way of excluding suspects who might otherwise be falsely charged with and convicted of serious crimes.5

Forensic use of DNA technology in criminal cases began in 1986 when police asked Dr. Alec J. Jeffreys (who coined the term "DNA fingerprints"6) of Leicester University (England) to verify a suspect's confession that he was responsible for two rape-murders in the English Midlands.7 Tests proved that the suspect had not committed the crimes. Police then began obtaining blood samples from several thousand male inhabitants in the area to identify a new suspect.8 In a 1987 case in England, Robert Melias became the first person convicted of a crime (rape) on the basis of DNA evidence.9

In one of the first uses of DNA in a criminal case in the United States, in November 1987, the Circuit Court in Orange County, Florida, convicted Tommy Lee Andrews of rape after DNA tests matched his DNA from a blood sample with that of semen traces found in a rape victim.10

Two other important early cases involving DNA testing are State v. Woodall11 and Spencer v. Commonwealth.12 In Woodall, the West Virginia Supreme Court was the first State high court to rule on the admissibility of DNA evidence. The court accepted DNA testing by the defendant, but inconclusive results failed to exculpate Woodall. The court upheld the defendant's conviction for rape, kidnaping, and robbery of two women. Subsequent DNA testing determined that Woodall was innocent, and he was released from prison (see the case profile in chapter IV for more details).

The multiple murder trials in Virginia of Timothy Wilson Spencer were the first cases in the United States where the admission of DNA evidence led to guilty verdicts resulting in a death penalty. The Virginia Supreme Court upheld the murder and rape convictions of Spencer, who had been convicted on the basis of DNA testing that matched his DNA with that of semen found in several victims. In Spencer, the defendant's attack upon the introduction of DNA evidence was limited to the contention that its novelty should lead the court to "hold off until another day any decision..."13 There was no testimony from expert witnesses that challenged the general acceptance of DNA testing among the scientific community.14

The first case that seriously challenged a DNA profile's admissibility was People v. Castro;15 the New York Supreme Court, in a 12-week pretrial hearing, exhaustively examined numerous issues relating to the admissibility of DNA evidence. Jose Castro was accused of murdering his neighbor and her 2-year-old daughter. A bloodstain on Castro's watch was analyzed for a match to the victim. The court held the following:

o DNA identification theory and practice are generally accepted among the scientific community.

o DNA forensic identification techniques are generally accepted by the scientific community.

o Pretrial hearings are required to determine whether the testing laboratory's methodology was substantially in accord with scientific standards and produced reliable results for jury consideration.

The Castro ruling supports the proposition that DNA identification evidence of exclusion is more presumptively admissible than DNA identification evidence of inclusion. In Castro, the court ruled that DNA tests could be used to show that blood on Castro's watch was not his, but tests could not be used to show that the blood was that of his victims.

In Castro, the court also recommended extensive discovery requirements for future proceedings, including copies of all laboratory results and reports; explanation of statistical probability calculations; explanations for any observed defects or laboratory errors, including observed contaminants; and chain of custody of documents. These recommendations soon were expanded upon by the Minnesota Supreme Court, in Schwartz v. State,16 which noted, "...ideally, a defendant should be provided with the actual DNA sample(s) in order to reproduce the results. As a practical matter, this may not be possible because forensic samples are often so small that the entire sample is used in testing. Consequently, access to the data, methodology, and actual results is crucial...for an independent expert review."17

In Schwartz, the Supreme Court of Minnesota refused to admit the DNA evidence analyzed by a private forensic laboratory; the court noted the laboratory did not comply with appropriate standards and controls. In particular, the court was troubled by failure of the laboratory to reveal its underlying population data and testing methods. Such secrecy precluded replication of the test.

In summary, courts have successfully challenged improper application of DNA scientific techniques to particular cases, especially when used to declare "matches" based on frequency estimates. However, DNA testing properly applied is generally accepted as admissible under Frye18 or Daubert19 standards.20 As stated in the National Research Council's 1996 report on DNA evidence, "The state of the profiling technology and the methods for estimating frequencies and related statistics have progressed to the point where the admissibility of properly collected and analyzed DNA data should not be in doubt."21 At this time, 46 States admit DNA evidence in criminal proceedings. In 43 States, courts have ruled on the technology, and in 3 States, statutes require admission (see exhibit 1).

------------------------------

Exhibit 1. DNA Evidence Admission in Criminal Trials by State

State DNA Admitted

Alabama Yes

Alaska Yes

Arizona Yes

Arkansas Yes

California Yes*

Colorado Yes

Connecticut Yes

Delaware Yes

Florida Yes

Georgia Yes

Hawaii Yes

Idaho Yes

Illinois Yes*

Indiana Yes

Iowa Yes

Kansas Yes

Kentucky Yes

Louisiana Yes

Maine No

Maryland Yes*

Massachusetts Yes

Michigan Yes

Minnesota Yes

Mississippi Yes

Missouri Yes

Montana Yes

Nebraska Yes

Nevada Statute

New Hampshire Yes

New Jersey Yes*

New Mexico Yes

New York Yes

North Carolina Yes

North Dakota No

Ohio Yes

Oklahoma Statute

Oregon Yes

Pennsylvania Yes

Rhode Island No

South Carolina Yes

South Dakota Yes

Tennessee Statute

Texas Yes

Utah No

Vermont Yes

Virginia Yes

Washington Yes

West Virginia Yes

Wisconsin Yes

Wyoming Yes

* Decision by Intermediate Court of Appeals

------------------------------

Notes

1. "Science is the search for truth -- it is not a game in which one tries to beat his opponent, to do harm to others." -- Linus Pauling, 1958. Cited in Beck, Emily Morison (ed.), Familiar Quotations, Boston: Little, Brown and Company, 1980.

2. Keynote address by Attorney General Janet Reno before the American Academy of Forensic Sciences, Nashville, Tennessee, February 21, 1996.

3. For articles debating the forensic use of DNA technology, see Thompson, William, "Evaluating the Admissibility of New Genetic Identification Tests:

Lessons from the DNA War," The Journal of Criminal Law & Criminology, 84, 1 (1993):22-104; Harmon, Rockne, "Legal Criticisms of DNA Typing: Where's the Beef?" The Journal of Criminal Law & Criminology, 84, 1 (1993):175-188; and Neufeld, Peter, "Have You No Sense of Decency?" The Journal of Criminal Law & Criminology, 84, 1 (1993):189- 202.

4. The study's results have been reviewed by many persons, including those involved in a peer review process. To date, no one has identified additional cases that, as of the study's completion in February 1996, are the type examined in this report.

5. National Research Council, National Academy of Sciences, DNA Technology in Forensic Science, Washington, D.C.: National Academy Press, 1992:156. (Cited as NRC report.) Another reference source is McKenna, Judith, J. Cecil, and P. Coukos, "Reference Guide on Forensic DNA Evidence," Reference Manual on Scientific Evidence, Federal Judicial Center (1994). This guide has a useful glossary of terms at p. 323.

6. Jeffreys, Alec J., Victoria Wilson, and Swee Lay Thein, "Hypervariable `Minisatellite' Regions in Human Nature," Nature, 314 (1985):67; "Individual-Specific `Fingerprints' of Human DNA," Nature, 316 (1985):76.

7. The first reported use of DNA identification was in a noncriminal setting to prove a familial relationship. A Ghanaian boy was refused entry into the United Kingdom (U.K.) for lack of proof that he was the son of a woman who had the right of settlement in the U.K. Immigration authorities contended that the boy could be the nephew of the woman, not her son. DNA testing showed a high probability of a mother-son relationship. The U.K. Government accepted the test findings and admitted the boy. See Kelly, K.F, J.J. Rankin, and R.C. Wink, "Methods and Applications of DNA Fingerprinting: A Guide for the Non-Scientist," Criminal Law Review (1987):105, 108; Note, "Stemming the DNA Tide; A Case for Quality Control Guidelines," Hamline Law Review, 16 (1992):211, 213-214.

8. Gill, Peter, Alec J. Jeffreys, and David J. Werrett, "Forensic Application of DNA Fingerprints," Nature, 318 (1985):577. See also Seton, Craig, "Life for Sex Killer Who Sent Decoy to Take Genetic Test," The Times (London) (January 23, 1988):3. A popular account of this case, The Blooding, was written by crime novelist Joseph Wambaugh, New York, N.Y.: William Morrow & Co., Inc., 1989.

9. Bureau of Justice Statistics, "Forensic DNA Analysis: Issues," Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, June 1991, at 4, note 8.

10. The admissibility of the DNA evidence was upheld by the intermediate appeals court, which cited the uncontroverted testimony of the State's expert witnesses. State v. Andrews, 533 So.2d 841(Dist. Ct. App. 1989). See also Office of Technology Assessment, Congress of the United States, Genetic Witness: Forensic Uses of DNA Tests, Washington, D.C.: July 1990.

11. 385 S.E.2d 253 (W. Va. 1989).

12. 384 S.E.2d 775 (1989). Additional court appeals by Spencer were rejected by the Virginia Supreme Court at 384 S.E.2d 785 (1989); 385 S.E.2d 850 (1989); and 393 S.E.2d 609 (1990).

13. Supra note 12 at 783.

14. Id., at 797.

15. 545 N.Y.S.2d 985 (Sup. Ct. 1989). Castro's case was never tried. He pleaded guilty to the murders in late 1989.

16. Schwartz v. State, 447 N.W.2d 422 (1989).

17. Id., at 427. The Minnesota Supreme Court further held that the use of statistical probabilities testimony should be limited because of its potential for prejudicing the jury. Id., at 428. The opinion was later modified in State v. Bloom, 516 N.W.2d 159 (1994).

18. Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). The test for the admissibility of novel scientific evidence enunciated in this case has been the most frequently invoked one in American case law. To be admissible, scientific evidence must be "sufficiently established to have gained general acceptance in the particular field in which it belongs."

19. Daubert v. Merrell Dow Pharmaceuticals, Inc., 113 S.Ct. 2786 (1993). The Supreme Court used this civil case to articulate new standards for interpreting the admissibility of scientific evidence under the Federal rules of evidence. This standard, while encompassing Frye, allows a court to expand its examination to include other indicia of reliability, including publications, peer review, known error rate, and more. The court also should consider factors that might prejudice or mislead the jury. For the application of Daubert to DNA technology, see Sheck, Barry, "DNA and Daubert," Cardozo Law Review, 15 (1994):1959.

20. This brief overview is not a treatise on DNA evidence admissibility in criminal cases. For more authoritative articles, see, Thompson, supra note 3; Kaye, D.H., "The Forensic Debut of the National Research Council's DNA Report: Population Structure, Ceiling Frequencies and the Need for Numbers," Jurimetrics Journal, 34, 4 (1994):369-382; Comments, "Admissibility of DNA Statistical Data: A Proliferation of Misconception," California Western Law Review, 30 (1993):145-178.

21. National Research Council, National Academy of Sciences, The Evaluation of Forensic DNA Evidence (prepublication copy), Washington, D.C.: National Academy Press, 1996:2.14.

================================

CHAPTER II

Study Findings

Findings pertaining to characteristics of the 28 DNA exculpatory cases identified during the study are discussed first. The chapter concludes with the results of the telephone survey of DNA laboratories.

------------------------------

General Characteristics Shared by Many Study Cases The 28 cases in this study were tried in 14 States and the District of Columbia. The States are Illinois (5 cases), New York (4 cases), Virginia (3 cases), West Virginia (3 cases), Pennsylvania (2 cases), California (2 cases), Maryland, North Carolina, Connecticut, Kansas, Ohio, Indiana, New Jersey, and Texas. Many cases share a number of descriptive characteristics, as noted below.

Most cases mid- to late 1980s. Most cases involved convictions that occurred in the 1980s, primarily mid- to late 1980s, a period when forensic DNA technology was not readily accessible. The earliest case involved a conviction in 1979, the most recent in 1991.

In each of the 28 cases, a defendant was convicted of a crime or crimes and serving a sentence of incarceration. While in prison, each defendant obtained, through an attorney, case evidence for DNA testing and consented to a comparison of the evidence-derived DNA to his own DNA sample. (In Nelson, the prosecutor conducted the tests.) In each case, the results showed that there was not a match, and the defendant was ultimately set free.

Sexual assault the most frequent crime. All 28 cases involved some form of sexual assault. In six (Bloodsworth, Cruz, Hernandez, Linscott, Nelson, and Vasquez), assailants also murdered their victims. All alleged assailants were male. All victims were female: most were adults, others teenagers or children. All but one case involved a jury trial. (The nonjury case, Vasquez, involved a guilty plea from a defendant who had mental disabilities.) Of the cases where the time required for jury deliberations was known, most had verdicts returned in less than 1 day, except for Kotler, which required 2 days.

Prison time served. The 28 defendants served a total of 197 years in prison (an average of almost 7 years each) before being released as a result of DNA testing. The longest time served was 11 years, the shortest 9 months. For a variety of legal reasons, defendants in several cases continued to remain in prison for months after exculpatory DNA test results. In Green, DNA testing was performed after conviction but prior to sentencing.

Many defendants also qualified for public defenders or appointed counsel. Most defendants appealed their convictions at least once; many appealed several times. Most appeals focused on trial error (e.g., ineffective assistance of counsel) or new evidence. For example, in some cases, the victims recanted their defendant identification testimony.

Prior police knowledge of the defendants. Police knew 15 defendants prior to their arrests, generally through criminal records. It is not known whether, in some cases, that may have influenced police to place suspects in photo spreads and lineups shown to victims and other eyewitnesses.

------------------------------

Evidence Presented During/After Trial: Common Attributes

The 28 cases shared several common themes in the evidence presented during and after trial.

Eyewitness identification. All cases, except for homicides, involved victim identification both prior to and at trial. Many cases also had additional eyewitness identification, either placing the defendant with the victim or near the crime scene (e.g., in Bloodsworth, five witnesses testified that they had seen the defendant with the 9-year-old victim on the day of the murder).

Many defendants presented an alibi defense, frequently corroborated by family or friends. For example, Edward Honaker's alibi was corroborated by his brother, sister-in-law, mother's housemate, and trailer park owner. The alibis apparently were not of sufficient weight to the juries to counter the strength of the eyewitness testimony.

Use of forensic evidence. A majority of the cases involved non-DNA-tested forensic evidence that was introduced at trial. Although not pinpointing the defendants, that evidence substantially narrowed the field of possibilities to include them. Typically, those cases involved comparisons of non-victim specimens of blood, semen, or hair at the crime scene to that of the defendants. Testimony of prosecution experts also was used to explain the reliability and scientific strength of non-DNA evidence to the jury.

Alleged government malfeasance or misconduct. Eight cases, as reported by defense attorneys and reflected in some judges' opinions, involved allegations of government misconduct, including perjured testimony at trial, police and prosecutors who intentionally kept exculpatory evidence from the defense, and intentionally erroneous laboratory tests and expert testimony admitted at trial as evidence. For example:

o In Honaker, the defendant's attorney alleged that the government intentionally kept exculpatory evidence from the defense, including information that two of the government's witnesses were secretly hypnotized to enhance their testimony and that the prosecution's criminalist was never told that Honaker had a vasectomy (and could not have been the source of the sperm in the victim).

o In Cruz, a supervising officer in the sheriff's department admitted, during the third trial, that he had lied about corroborating the testimony of his deputies in the earlier trials. This testimony focused on Cruz's "dream visions" of the murder.

o In Kotler, the government's serologist reportedly lied about his qualifications. In addition, Kotler's attorneys alleged that the government intentionally withheld exculpatory evidence from the defense. For example, police reports stated that the victim did not actually positively identify the defendant's picture but described him only as a "look alike." Furthermore, as recorded in police reports, the victim's description of the defendant was inaccurate for age, height, and weight. The defense was never informed about those reports.

o In cases involving defendants Glen Woodall, William O'Dell Harris, and Gerald Wayne Davis (and his father), the perjured testimony of Fred Zain, a serologist then with the West Virginia State Police, was in large part responsible for the wrongful convictions that ensued. The West Virginia Supreme Court of Appeals, in a special report on Zain's misconduct in more than 130 criminal cases, stated that such behavior included "...overstating the strength of results; ...reporting inconclusive results as conclusive; ...repeatedly altering laboratory records; ...."1 The report also noted that Zain's irregularities were "the result of systematic practice rather than an occasional inadvertent error." In addition, the report stated that Zain's "supervisors may have ignored or concealed complaints of his misconduct."2

o In Alejandro, the defendant was also wrongfully convicted by expert testimony from Fred Zain, who had moved from West Virginia to Texas and worked for the Bexar County crime laboratory. In July 1994, a Uvalde County grand jury indicted Zain for perjury, tampering with government records, and fabricating evidence. As of early 1996, charges of tampering and of fabricating evidence had been dropped, leaving three charges for aggravated perjury in effect, for which Zain reportedly seeks dismissal on statute of limitations grounds.

Evidence discovered after trial. In most of the cases in this study, DNA test results represented newly discovered evidence obtained after completion of the trials. States have time limits on filing motions for new trials on the basis of newly discovered evidence. For example, in Virginia, new evidence must be presented by motion within 21 days after the trial.3 Thus, the Honaker, Snyder, and Vasquez cases required a pardon from Virginia's governor to release the defendants from prison.

In some of the study cases, prosecutors waived time limits when presented with the DNA exculpatory results. However, prosecutors also have contested defendants' attempts to release evidence for DNA testing.

States also differ in the legislation and procedures pertaining to post-conviction appointment of counsel and to authorization to pay for the DNA testing. Many cases involved indigents.

DNA testing. The DNA testing phase of these cases also has common characteristics. Nearly all the defendants had their tests performed by private laboratories. The tests were conducted using blood from defendants, blood or blood-related evidence from victims, and semen stains on articles of the victims' clothing or on nearby items (a blanket was tested in one case). In over half the cases, the prosecution either conducted a DNA test totally independent of that of the defense or sent test results obtained by the defendant's laboratory to a different one to determine whether the laboratory used by the defense interpreted test results properly.

Eight laboratories used Restriction Fragment Length Polymorphism (RFLP) DNA testing, 17 conducted

Polymerase Chain Reaction (PCR) testing, and 2 used both tests. For one case, the type of DNA test conducted is unknown.

Preservation of evidence. In some cases, evidence samples had deteriorated to the point where DNA testing could not be performed. In Brison, the laboratory could not test cotton swabs from the rape kit but, instead, tested a semen stain from the victim's underwear. In Daye, after the appellate court affirmed the defendant's conviction and the State Supreme Court denied certification, the evidence was about to be destroyed when Daye's attorney filed to stay the destruction in order to conduct DNA testing.

The chain of custody in some of the cases also demonstrated a lack of adherence to proper procedures. Authorities on the subject note that the "mishandling of real evidence affects the integrity of the fact-finding process."4 In Dabbs, the defendant's attorneys reported that the defense was initially advised by the prosecution that the evidence (victim's underwear that contained a semen stain) had been destroyed (a conclusion based on failure of authorities to find the evidence in police or court custody). Eventually, the defense found the evidence at the county crime laboratory.

------------------------------

Results of DNA Laboratory Survey

Conducted in June 1995, the nationwide telephone survey of 40 public and private laboratories that performed DNA tests sought answers to such questions as: From the time the laboratories began DNA testing, how many cases have they handled? Of that number, what percentage yielded results that excluded defendants as sources of the DNA evidence or were inconclusive?

The 40 surveyed laboratories yielded 19 whose available data were sufficient for the purposes of this study. The 19 included 13 at the State/local level, 4 in the private sector, an armed forces laboratory, and the FBI's laboratory.

Most of the laboratories had initiated DNA testing only within the previous few years. Twelve began testing between 1990 and 1992. Three of the four private laboratories began in 1986 or 1987, while the FBI started DNA testing in 1988.

Seven of the laboratories reported using RFLP testing; four, PCR testing; and eight, both types of tests.

The 19 laboratories reported that, since they began testing, they had received evidence in 21,621 cases for DNA analysis, with the FBI accounting for 10,060 cases. Three of the 4 private laboratories averaged 2,400 each; the State and local laboratories averaged 331 each.

In about 23 percent of the 21,621 cases, DNA test results excluded suspects, according to respondents. An additional 16 percent of the cases, approximately, yielded inconclusive results, often because the test samples had deteriorated or were too small. Inconclusive results aside, test results in the balance of the cases did not exclude the suspect.

The FBI reported that, in the 10,060 cases it received, DNA testing results were about 20 percent inconclusive and 20 percent exclusion; the other 18 laboratories (11,561 cases) reported about 13 percent and 26 percent, respectively.*

Unfortunately, the laboratories were unable to provide more details. They did not maintain data bases that would permit categorization of DNA test results by type of offense and other criteria. What happened to the suspects who were excluded through DNA testing also cannot be determined. Were they released, or were they charged on the basis of other evidence, for example?

Thus, only the most general information is known about the results of DNA testing by laboratories. To obtain more detailed information would require a comprehensive research project.

------------------------------

Notes

1. Matter of West Virginia State Police Crime Laboratory, 438 S.E.2nd 501, 503 (W.Va. 1993).

2. Id., at 504.

3. Virginia Supreme Court Rules, Rule 3A: 15(b).

4. Giannelli, Paul, "Chain of Custody and the Handling of Real Evidence," American Criminal Law Review, 20, 4 (Spring 1983):527-568.

*If inconclusive cases were omitted, the exclusion rate for the FBI would be approximately 25 percent, and the average exclusion rate for the other 18 laboratories would be about 30 percent.

================================

CHAPTER III

Policy Implications

The 28 cases examined by the study raise issues that have policy implications for the criminal justice system. The most significant are presented below.

------------------------------

Reliability of Eyewitness Testimony

In the majority of the cases, given the absence of DNA evidence at the trial, eyewitness testimony was the most compelling evidence. Clearly, however, those eyewitness identifications were wrong. In one of the clearest examples of eyewitness testimony overwhelmingly influencing the jury, the Pennsylvania Intermediate Court of Appeals commented on the evidence in the Dale Brison case:

The Commonwealth's evidence consisted primarily of the victim's identification testimony. However, the victim's stab wounds in addition to the weather and reduced visibility may well have affected the victim's ability to accurately view her assailant, and thus, she may have been prompted to identify appellant merely because she remembered seeing him in the neighborhood. Moreover, the victim did not specifically describe any of her assailant's facial characteristics to the police. There was also no conclusive physical evidence, aside from a single hair sample which may have been consistent with any male of [A]frican-[A]merican descent, linking appellant to the crime.

This points conclusively to the need in the legal system for improved criteria for evaluating the reliability of eyewitness identification.

In Neil v. Biggers,3 the U.S. Supreme Court established criteria that jurors may use to evaluate the reliability of eyewitness identifications. However, the reliability of eyewitness testimony has been criticized extensively in the literature.4 In a recent interview, Dr. Elizabeth Loftus, one of the best-known critics of the reliability of eyewitness identification, commented on the role of DNA testing in exonerating innocent persons who served time in prison. Dr. Loftus noted that a significant factor is the potential susceptibility of eyewitnesses to suggestions from police, whether intentional or unintentional. As reported, Dr. Loftus stated that there is "pressure that comes from the police [who] want to see the crime solved, but there is also a psychological pressure that is understandable on the part of the victim who wants to see the bad guy caught and wants to feel that justice is done."5

Dr. Loftus has recommended more open-ended questioning of victims by the police to avoid leading questions. In addition, Dr. Loftus and others have recommended use of expert testimony regarding the pros and cons of relying on eyewitness testimony.6

Reliability of Non-DNA Analyses of Forensic Evidence Compared to DNA Testing

In many of the study cases, according to documentation examined and those interviewed, scientific experts had convinced juries that non-DNA analyses of blood or hair were reliable enough to clearly implicate the defendants. Scientific conclusions based on non-DNA analyses, however, were proven less discriminating and reliable than those based on DNA tests. These findings point to the need for the scientific community to take into account the reliability of non-DNA forensic analyses vis-…-vis DNA testing in identifying the sources of biological evidence.

In a recent habeas corpus hearing in a murder case, a U.S. district court held that expert testimony on

microscopic hair comparisons was inadmissible under the Daubert standard.7 The court cited studies documenting a high error rate and found that there are no accepted probability standards for human hair identification. The court ruled that in this case the expert's hair testimony was "imprecise and speculative, and its probative value was outweighed by its prejudicial effect."8

Competence and Reliability of DNA Laboratory Procedures

One of the lasting effects of the O.J. Simpson case will likely be greater scrutiny by defense lawyers of the prosecution's forensic DNA evidence presented in criminal cases. In the Simpson case, the defense, in essence, put the crime laboratory on trial. The National Research Council (NRC) report entitled DNA Technology in Forensic Science states:

There is no substantial dispute about the underlying [DNA] scientific principles. However, the adequacy of laboratory procedures and the competence of the experts who testify should remain open to inquiry.9

The NRC report recommends some degree of standardization to ensure quality and reliability. The report recommends that each forensic laboratory engaged in DNA testing must have a formal, detailed program of quality assurance and quality control. The report also states:

Quality-assurance programs in individual laboratories alone are insufficient to ensure high standards. External mechanisms are needed to ensure adherence to the practices of quality assurance. Potential mechanisms include individual certification, laboratory accreditation, and state or federal regulation.10

As recently reported by the American Society of Crime Laboratory Directors, 32 public DNA laboratories have been accredited. In addition, one private laboratory is accredited.11

Whether laboratories that conduct DNA tests possess the requisite qualifications has significant cost implications for the criminal justice system in terms of reducing the number of redundant DNA tests. In many cases in this study, both prosecution and defense obtained independent DNA tests of the biological stain evidence. Although independent examinations are common in areas that are more open to interpretation (e.g., mental fitness for trial), DNA testing, for exculpatory purposes, should be performed in a qualified laboratory, and the results, if they exculpate the suspect, should be accepted by both parties. Such acceptance would seem more likely if DNA tests were performed by laboratories that all parties agreed were qualified.

Preservation of Evidence for DNA Testing

In some States, sentenced felons may experience difficulty obtaining access to evidence for DNA testing. With an increasing volume of criminal cases, some police agencies destroy evidence when defendants have exhausted their appeals. Even when defendants obtain access to the evidence, it may be too deteriorated for DNA testing. In some of the study cases, insufficient evidence prevented laboratories from conducting Restriction Fragment Length Polymorphism (RFLP) testing, but Polymerase Chain Reaction (PCR) testing was still possible.

Preserving biological stain evidence and maintaining the proper chain of custody of the evidence are essential for successful DNA testing.12 At the trial stage, however, the U.S. Supreme Court has ruled that unless a criminal defendant can show bad faith on the part of the police, failure to preserve potentially useful evidence does not constitute a denial of due process of law.13 After a defendant's conviction, prosecutors are not required by constitutional duty to preserve evidence indefinitely. As noted earlier, in Daye, the evidence was about to be destroyed when his attorney filed to stay the destruction to conduct what turned out to be an exculpatory DNA test.

Training in DNA Forensic Uses

The introduction of DNA technology into the criminal trial setting is likely to create uncertainty, spawned in part by the complexity of the technology, and also to possibly generate unrealistic expectations of the technology's power in the minds of some or all of the players: prosecution, defense, judges, and jurors. The scientific complexities of the technology may influence all parties to rely more heavily on expert testimony than on other types of evidence.

As the use of DNA technology becomes more widely publicized, juries will come to expect it, like fingerprint evidence. This will place more pressure on prosecutors to use the technology whenever possible, especially as the cost decreases. Prosecutors must be trained on when to use the technology and how to interpret results for the jury.

When the prosecution uses DNA evidence, the defense will be forced to attack it through expert testimony.

The defense must rebut the persuasiveness of the evidence for the jury. As stated in the NRC report, "Mere cross examination by a defense attorney inexperienced in the science of DNA testing will not be sufficient."14 Thus, defense counsel as well as the prosecution and judiciary must receive training in the forensic uses of DNA technology.

Third-Party Consensual Sex Sources

The primary objective of the defense in using DNA testing in rape cases is to show that the defendant is excluded as the source of the semen evidence. Even when exclusion is established, the prosecution may be motivated, as in Davis, to eliminate as suspects any and all consensual sex partners as sources of semen in rape cases. During the first trial of Gerald Wayne Davis, the prosecution contended that the semen in the victim came from Davis. After DNA testing had excluded Davis as the source of the semen, the prosecution contended, in the second trial, that Davis could have still raped the victim but not ejaculated and that the semen in the victim could have come from the victim's fiancé‚ just prior to the rape. The prosecution never obtained a blood sample from the fiancé‚ because he died before the second trial.

A question under the law is whether third parties can be compelled to provide biological evidence for DNA testing. In some cases, the government refused to release defendants after exculpatory DNA results until third parties were located and tested. Kerry Kotler was held for an additional year after his exculpatory DNA test so the government could test the victim's husband. Edward Honaker was held for an additional 9 months after his exculpatory DNA test so the government could test the victim's boyfriend and "secret lover."

Multiple-Defendant Crimes

The DNA technology used to analyze biological evidence from crime scenes must not be oversold as an exculpatory tool -- it does have limitations. Multiple-suspect crimes present a particular problem for use of DNA identification as a crime-solving tool. In multiple-suspect sexual assaults without eyewitnesses, such as a rape-murder, it is possible that only one of the suspects ejaculated in, or even raped, the victim. In such cases, DNA testing of semen would seem likely to exculpate one or more of the suspects. This type of situation presents a real dilemma for police and prosecutors. Because of exculpatory DNA tests on semen and possibly other exculpatory evidence (e.g., an alibi, lack of other physical evidence), pressure mounts on prosecutors to release one or more of the suspects. The only other evidence against them may be the testimony of a suspect who is matched to the crime by DNA analysis.

In Dabbs, for example, the victim testified that she was dragged into an alley and raped by one man while two other men held her down. The police arrested Dabbs on the basis of identification of him by the victim, a distant cousin. The other alleged assailants were never identified or arrested. The DNA test showed that the semen evidence from the victim did not match Dabbs. One theory of the case, however, was that Dabbs participated in the crime but was not the rapist. The prosecutor ultimately dismissed the original indictment against Dabbs because of the DNA results and the reluctance of the victim to testify at a new trial.

Post-trial Relief

Most States have a time limit on presenting evidence newly discovered after trial, conviction, and sentencing. The reason for limiting the time to file appeals based on new evidence is to ensure the integrity of the trial process and jury verdicts. Many DNA issues in the study cases were not raised until the post-conviction stages. Absent constitutional issues, many State procedures, as in Virginia,15 may preclude consideration of new exculpatory DNA evidence at postconviction stages. Some of the study defendants, after receiving exculpatory DNA results, were released only by agreement of the prosecutor; sometimes they needed a pardon by the governor.

Some States, such as Oregon, permit judges to use discretion to waive new-evidence rules and set aside verdicts or order new trials.16 Thus, some States may allow an out-of-time motion for a new trial when newly discovered evidence clearly serves the interests of justice.17

At postconviction stages, appointment of counsel and payment for DNA testing become issues for indigents. While some appeals courts have ordered State-paid DNA testing for indigents where justified (e.g., where the overall case against the defendant is weak), other court rulings deny such relief, especially where the exculpatory value is speculative.18 As DNA testing to exculpate convicted persons becomes more widespread, States need to consider these issues.

Future DNA Forensic Uses

The momentum is growing, spurred in part by the public's education from the Simpson trial, for DNA testing in criminal cases. Juries may begin to question cases where the prosecutor does not offer "conclusive" DNA test results if the evidence is available for testing. More defense attorneys in court-appointed cases may file motions for DNA testing and request the State to pay for the tests (this issue may also be raised as a Brady motion for the prosecutor to conduct the tests).

The shift will be for more DNA testing in pretrial stages. Prosecutors should find that DNA testing is as helpful to them as to the defense in excluding suspects early in the investigation. This will enable the police and prosecution to save money in the long run by focusing investigations in more fruitful directions.

In Britain, mass DNA screening in search of suspects has, in recent years, produced arrests in several highly publicized cases. The most recent case involved the rape-murder of a 15-year-old South Wales girl.19 The South Wales constabulary obtained saliva swab samples from over 2,000 men who lived in the vicinity of the murder. Police went door-to-door inviting men to a makeshift laboratory to submit the samples. The saliva samples were used to develop DNA profiles to compare to the DNA profile obtained from the assailant's semen.

British law does not permit compulsory sampling, but the police made it clear that anyone who refused would become the subject of intense police investigation. A 19-year-old resident of the victim's neighborhood was arrested when his saliva sample was the only one of the thousands taken that could not be eliminated.

Such DNA dragnet methods, while employed sparingly in Great Britain, may increase as the ease and affordability of DNA testing improves. It is unlikely that such mass-testing methods would gain favor in the United States. Constitutional protections against self-incrimination and unreasonable searches and seizures, as well as the American public's zealous protection of privacy rights, would preclude such DNA dragnet practices from being implemented in this country.

------------------------------

Notes

1. This report does not discuss the issue of government misconduct because it is not particularized to the use of DNA technology. Beyond the limited instances noted in this report, enough examples of government misconduct in the criminal justice system exist in the popular media for government officials to be well aware of the problem.

2. Commonwealth v. Brison, 618 A.2d 420, 425 (Pa. Super. 1992).

3. Neil v. Biggers, 409 U.S. 188, 199-200 (1972) (factors include accuracy of the witness' prior description of the defendant, opportunity to view the defendant at the time of the crime, level of certainty demonstrated, witness' degree of attention, and time between the crime and the confrontation).

4. Loftus, Elizabeth, and D. Fishman, "Expert Psychological Testimony on Eyewitness Identification," Law and Psychology Review, 4 (1978):87-103 (lack of reliability on cross-racial identification); Loftus, Elizabeth, and W. Wagenaar, "Ten Cases of Eyewitness Identification: Logical and Procedural Problems," Journal of Criminal Justice, 18 (1990):291-319 (witnesses can be induced to point to the suspect after subtle suggestion on the part of the investigator); and Cutler, Brian, et al., "The Reliability of Eyewitness Identification: The Role of System and Estimator Variables," Law and Human Behavior, 11, 3 (1987):233-258 (level of stress experienced during crime may affect identification).

5. "DNA Testing Turns a Corner as Forensic Tool," Law Enforcement News (October 15, 1995):10.

6. Loftus, Elizabeth, and N. Schneider, "Judicial Reactions to Expert Testimony Concerning Eyewitness Reliability," UMKC Law Review, 56, 1 (1987):1-45; and Handberg, Roger, "Expert Testimony on Eyewitness Identification: A New Pair of Glasses for the Jury," American Criminal Law Review, 32, 4 (Summer 1995):1013-1064.

7. Williamson v. Reynolds, 904 F. Supp. 1529 (E.D. Okl. 1995).

8. Id., at 1558. The National Research Council report, DNA Technology in Forensic Science, notes that, in contrast to microscopic hair comparison, with the advent of DNA technology, the use of hair as an individual identifier will become more common. National Research Council, National Academy of Sciences, DNA Technology in Forensic Science, Washington, D.C.: National Academy Press, 1992:158.

9. DNA Technology in Forensic Science, supra note

8, at 145-146.

10. Id., at 16. In its 1996 DNA report, The Evaluation of Forensic DNA Technology (National Academy Press, Washington, D.C.), the National Research Council reaffirmed this position (page 3.12). The DNA Identification Act of 1994 (Public Law 103-322) also provides for a DNA advisory board to set standards for DNA testing.

11. Telephone conversation with Manuel Valdez, treasurer, American Society of Crime Laboratory Directors, March 8, 1996. (More than 100 public laboratories perform DNA tests.)

12. See "Oops! We Forgot to Put It in the Refrigerator: DNA Identification and the State's Duty to Preserve Evidence," The John Marshall Law Review, 25 (1992):809-836.

13. Arizona v. Youngblood, 109 S. Ct. 333, 337 (1988). The Supreme Court also stated that "police do not have a constitutional duty to perform any particular tests."

14. Supra note 9 at 160.

15. Virginia Supreme Court Rules, Rule 3A: 15(b).

16. An Oregon judge recently released Laverne Pavlinac and John Sosnovske from prison, where they had served 5 years after being convicted of murdering a young woman. The judge set aside their convictions because Keith Hunter Jesperson, a convicted serial killer, pleaded guilty to the murder for which the couple was convicted. See The New York Times, November 28, 1995:28.

17. Tuffiash v. State, 878 S.W. 2d 197 (Tex. App. 1994). This case involved perjured trial testimony from Fred Zain, the State's forensic serologist.

18. See State v. Thomas, 586 A. 2d 250 (N.J. Appl. Div. 1991); and Commonwealth v. Brison, 618 A. 2d 420 (Pa. Super. 1992). Compare to People v. Buxon,  593 N.Y.S. 2d 87 (App. Div. 1993).

19. "Crime-Solving by DNA Dragnet," The Washington Post (February 2, 1996):A21.

================================