Forensics, Trial Lawyering, and Wrongful Convictionssiterepository.s3.amazonaws.com/2989/garrett_forensics_presentati… · Forensics, Trial Lawyering, and Wrongful Convictions Brandon

Forensics, Trial Lawyering,

and Wrongful Convictions

Brandon L. Garrett

Professor of Law

University of Virginia School of

Law

This project was supported by Award No. 2010-

DN-BX-K259 awarded by the National

Institute of Justice, Office of Justice Programs,

U.S. Department of Justice

2

Brandon L. Garrett

Convicting the Innocent

Where Criminal Prosecutions

Go Wrong

(Harvard University

Press 2011)

“Convicting the Innocent”

Resource Websites

Data webpage home

• http://www.law.virginia.edu/html/librarysite/garrett_innocent.htm

Archive of forensics testimony

• http://www.law.virginia.edu/html/librarysite/garrett_exoneree.htm

3

Dennis Fritz & Ronald Williamson

Both faced the

death penalty for

a brutal rape and

murder in Ada,

Oklahoma.

See John

Grisham‟s “The

Innocent Man”

4

Serology in Fritz and Williamson

• Two types of forensics were presented at their

trials: serology and hair comparisons

• The analyst claimed not to have observed any

blood type on the murder victim.

• The analyst concluded that the victim was a

non-secretor, a person whose blood type does

not show up in bodily fluids.

5

Serology testimony (Fritz)“Q: Did you notice any antigen activity whatsoever in

this sample taken from the victim?A: No.Q: And you -- so, in order for that test to be valid, you

would have to assume that the victim was a non-secretor? You have to make that assumption; do you not?

A: Basically, yes.”

“Q: So, you‟re assuming that she‟s a non-secretor even though you do not have tests to indicate that?

A: That‟s correct. That‟s the conclusion that I would with common sense come to.”

6

Serology testimony (Fritz)

• The analyst concluded that the rapist must also

have been a non-secretor, since no blood types

were observed.

“Q: Okay. If the victim was a non-secretor, and the donor was a non-secretor, and their body fluids were mixed on the swab, what would you expect to find?

A: No antigen activity.

7

Degradation, masking and non-

quantification• However, the blood evidence from the murdered

victim may have degraded. The analyst knew that possibility could not be ruled out.

• Nor was anything foreign to the victim‟s assumed type observed. The victim‟s material could mask any contribution from the donor. There was no ability at the time to quantify or isolate material from the donor.

• The evidence was inconclusive – any male could have been the semen donor.

• The analyst instead implied Fritz and Williamson, both nonsecretors, were included by the tests.

8

Hair Comparison (Fritz)

Another Oklahoma State Crime Lab analyst told

the jury that a group of hairs – a total of 11 pubic

hairs and 2 head hairs were “consistent

microscopically [with Fritz] and could have

the same source”

• The analyst claimed that the hairs were a “match.”

• “This means that they match if you want it in one word.”

9

Amplifying the ConclusionThe analyst also said it was even more powerful that both scalp and pubic hairs matched:

“Q: Okay. Does the fact that you have pubic hairs which are different than scalp hairs . . . And you find both of those sources at a crime scene, is there any significance attached to that?”A: I believe there is. A single hair match, say, for example, a single scalp hair match in a case would have a particular significance. There is, say, one scalp hair that matches a particular source. This has a particular significance that if that was the only evidence present in a case, the significance would be fairly low in my way of thinking.

10

Invalid testimony

• In fact, there is no research on what it means for hairs to “match”

• We do not know how many people share any particular microscopic hair characteristics.

• The F.B.I. had rejected the use of terms like “match” in the context of hair comparisons as potentially misleading

• And there was no evidence that it was more powerful to find both scalp and pubic hairs that “match” – one would expect similarities in hair from different parts of a person‟s body.

11

Hair comparison (Williamson)

At Williamson‟s trial, the analyst told the jury that two pubic hairs and two scalp hairs were consistent with him. But the analyst again went further and called them a “match.”

“When a hair matches, if you will, it is consistent microscopically to a known source.”

The analyst also pointed out that “a large number of individuals that their known scalp and pubic hairs were submitted in this case” and yet none matched.

12

Brady violation• There was a problem with that last statement.

• The analyst had not compared crime scene hairs with all of the suspects. (Dry-hairing?)

• In 1995, not knowing Williamson was innocent,

a federal judge found that the analyst had not in

fact ever compared the hairs of suspect Glen

Gore. He also ruled that hair comparison

testimony was too error prone, “scientifically

unreliable,” and too grounded in the experts‟

“subjective opinion” to be admissible.13

A False Negative

• Later it was determined that none of the hairs belonged to Ronald Williamson, nor to Dennis Fritz.

• Instead, Glen Gore was shown by post-conviction DNA testing to have been the actual perpetrator.

• No audit was ordered of the crime laboratory until many years later, when other exonerations revealed still more cases with flawed forensics.

14

Map of Exonerations in the U.S.

15

Timeline to Exoneration

16

DNA Exoneree Trial Materials

Transcript - 88% (207)

Pleaded Guilty - 6% (16)

No transcript - 12% (27)

17

Crime of Exoneree Conviction

Rape - 68% (171)

Murder - 9% (22)

Rape & Murder - 21% (52)

Other - 2% (5)

18

DNA Identified the Real Culprit

DNA Hit - 45%

(112)

None - 55% (138)

19

Type of Defense Lawyer

Court Appointed - 78

Public Defender - 71

Retained - 53

Unknown - 48

20

0 50 100 150

Eyewitness - 76% (190)

Forensic Evidence - 74% (185)

Informant - 21% (52)

Confession - 16% (40)

Evidence Supporting Exonerees’ Convictions

21

0 20 40 60 80 100 120 140

Alibi - 68% (140)

Third Party Guilt - 15% (30)

Experts Retained - 30% (62)

Defendant Testified - 53% (110)

The Defense Case at Trial

22

Flawed Evidence at Exoneree Trials

0

20

40

60

80

100

120

140

160

180

Eyewitnesses -

88% (141/161)

Forensic

Testimony - 61%

(93/153)

Jailhouse

Informants - 93%

(26/28)

Confessions -

95% (38/40)

No transcript Not flawed Flawed evidence

23

Challenging Evidence Post-Conviction

0

20

40

60

80

100

120

Eyewitnesses -

56% challenged

(70/124)

Forensic

Testimony - 32%

(36/112)

Informants - 34%

(16/45)

Confessions -

59% (13/22)

Not challenged Reversal Unsuccessful Challenges

24

0 20 40 60 80 100

Serology - 58% (67/116)

Hair - 39% (29/75)

Fingerprint - 5% (1/20)

DNA - 17% (3/18)

Shoe print - 16% (1/6)

Bite mark - 71% (5/7)

Voice Spect. - 100% (1/1)

Types of Forensic Testimony

Invalid Not invalid

25

Invalid and Unreliable Forensics

Non-problematic & inculpatory

- 7% (10)

Invalid Testimony - 61%

(93/153)

Vague Testimony - 12% (19)

Non-probative & non-

problematic - 20% (31)

26

A Taxonomy of Invalid Conclusions

(1) Non-probative evidence presented as

probative

e.g. serology cases involving the problem of masking

and non-quantification

(2) Exculpatory Evidence Discounted

e.g. speculating that “bacterial contamination” explains

exculpatory serology

27

(3) Inaccurate Frequency of Statistic Presented

e.g. claiming that though Type O secretors are 35% of the

population, that only 17.5% of men could have contributed

(where 35% of both men and women are Type O secretors).

(4) Statistic Provided Without Empirical Support

e.g. claiming a 1 in 10,000 chance that two hairs could

have come from someone other than the defendant.

28

(5) Non-numerical statements provided without empirical support

e.g. “It is unlikely that that hair would match anyone other than the defendant; but it is possible.”

(6) Conclusion that evidence originated from defendant

e.g. “These bullets were fired from that particular gun barrel, to the exclusion of all other firearms”

29

Vague Conclusions

Statements that evidence is “consistent with” or

“similar to” or “the same” or “could be

associated” or “matched” that from the defendant

Statements involving a “reasonable degree of

scientific certainty”

30

Invalid or Misleading

Claims about Methods(1) Unsupported or un-validated error rate

claims

e.g. “Friction ridge comparisons are reliable with rare occurrences of error.”

(2) Uniqueness claims

e.g. “The marks that a gun leaves on these items are individual characteristics. They are like a fingerprint.”

“No two fingerprints have ever been found to be the same.”

(3) Experience claims (related to error rate claims)e.g. “It is rare, indeed, that you would have one examiner saying it was fired through the gun, and another examiner saying it was not fired through the gun.” 31

Cross-examination

• Few defense lawyers had their own experts (only 21 had experts).

• Defense lawyers typically do not have access to underlying bench notes.

• Few defense lawyers asked questions about the particular invalid forensic conclusions in exoneree trials – fewer than half (46 of 93) did so.

• When they did ask questions, they did not carefully inquire into the methodological problems and often elicited more serious overstatements.

32

Almost No Cross –

The Trial of Neil Miller• “Just a few questions,” the defense lawyer said

first, before asking four questions about invalid serology testimony.

• The lawyer did not ask about the failure to detect any blood group substances foreign to the victim or the problem of masking.

• Instead, he just asked the analyst to repeat the invalid statistics, and then asked, “And the bottom line is… you couldn‟t say conclusively whether the sperm that you tested came from this man, Neil Miller, could you?

• A. I could not.” 33

Ineffectual Cross –

The Trial of Curtis McCarty• At the end of her direct, Joyce Gilchrist concluded by

stating her opinion that based on her analysis, “That he was in fact there.”

• The defense lawyer asked – “Q. Taking the areas that you have examined in this case, can you positively identify Mr. McCarty as being the donor of the hair?

• A. Again I cannot positively identify anyone on hair examinations. I can positively exclude someone.”

• Yet on redirect – “Q. Just one question, Judge. Have you ever seen in the six years that you have done hair analysis, two people with the same microscopic characteristics in their hair?

• A. No, I have not.”

• The defense had no further questions.

34

The Trial of Willie Jackson

• The analyst concluded direct by stating that “Mr. Jackson is the person who bit this lady.”

• The defense asked, “Is it your testimony… that based upon your analysis these bite marks in this case couldn‟t be made by anybody else?

A. I never said that.”

• There was no follow up – that concluded the cross.

• The State then elicited again that “there is no doubt in my mind that Willie Jackson is the individual who bit” the victim.

35

The Trial of Jerry Watkins

• B blood group substances were detected, inconsistent with both the victim and Watkins

• The analyst had testified that “sometimes bacteria will acquire a B Blood Group substance activity” and therefore Watkins could not be excluded.

• The defense lawyer asked, “Did you do any cultures? A. No sir, we do not do cultures in our laboratory.”

“Q. Now have you ever personally had a situation where bacteria mimicked a blood group activity?

A. Not myself personally, there was on other serologist in the laboratory who had this particular activity possibly.”

36

Questioning the Method –

The Trial of Larry Peterson• “Q. The best you can give me, and no disrespect,

is your judgment that the particular hairs that were questioned compared microscopically to Mr. Peterson‟s hairs; is that correct?

• A. Yes. Basically that‟s the way hair comparisons are.”

• Then on re-direct – “Q. So, my understanding is what you are telling is is that every hair that was known as a questioned hair has been identified as either belonging to the victim or belonging to Mr. Peterson?

• A. Yes.”37

Cross that Elicits Added Exaggeration –

the Trial of Glen Woodall

• “Q. Can you state objectively that that hair sample belonged to Glen Woodall?

• A. I would say…that the consistencies were 100 percent, and it is very highly likely that they came from the same individual.”

• Q. But your answer is it was highly likely. You can‟t say it did, can you?

• A. There again, from the standpoint of scientifically stating from the characterization on the examination, I would say there was nothing to show me in the examination that they originated from another individual.”

38

Failure to Provide Relevant

Statistics Until Cross• Hector Gonzales was arrested for a murder - stains on his

clothes were tested for the victim‟s blood type.

• During direct examination, the analyst explained that testing for

several genetic markers was conducted, and “most of the stain in

the jeans” were “consistent with the victim‟s type.”

The analyst did not state what percentage of the population

shared that type.

• Only on cross did the analyst acknowledge that the victim‟s ACP

type (B) was found in 54% of the New York City population,

according to the Medical Examiner‟s Office database.

• On redirect, the analyst testified that one of the stains shared the

ACP and PGM type with the victim, and 20% of New York

City‟s population shared the combined ACP/PGM type.39

The Trial of Byron Halsey

• Only on cross did the analyst make clear that the serology was inconclusive:

“Q. . . you say its a mixed stain. . . Then any male could have deposited that semen…?

A. We can‟t rule out any secretor male. And, of course, we can‟t rule out the nonsecretors, so, yes.

Q. That sort of takes in the whole population, doesn‟t it?

A. That‟s exactly right.”

40

The Trial of Larry Mayes

• On direct, the analyst stated that the defendant “is within the group that could have” been the semen donor.

• On cross, the defense asked, “Am I not also correct that there is another possibility, and that is that the swabs could contain [substances from. . .] someone from a group other than the group Mr. Mayes is in?

A. Yes.”

The analyst also explained that “You can‟t

eliminate any individual.”

41

Prosecution Closings• The conviction of Stephen Linscott was reversed where “No

one testified that [the victim] was raped by a non-secretor‟ or that the seminal material „came from a non-secretor.‟ The prosecutors imply made-up that piece of „evidence.‟”

• Contrast the case of Drew Whitley, where the defense pointed out the prosecutor falsely told the jury in closings that the analyst concluded hairs “came from” the defendant. The judge offered a vague curative instruction and noted that “[i]t‟s the jury‟s responsibility to remember things.”

• In the case of Larry Ollins, the court emphasized that though the prosecutor had incorrectly claimed serology “matched” and the defendant‟s hair was found at the scene, “[a] prosecutor is permitted a wide range of expression in comments made during closing argument.”

42