1
00:00:00,390 --> 00:00:05,100
Okay everybody to a second also seminar for the term.

2
00:00:05,280 --> 00:00:10,430
And just before I introduce to speak for today I just want to let you know that that one that

3
00:00:10,440 --> 00:00:14,910
is due to happen in two weeks time has actually been cancelled because it falls in the strike.

4
00:00:15,300 --> 00:00:19,020
So we've got one day and then we're not quite sure when the next one is,

5
00:00:19,020 --> 00:00:22,770
because I don't know whether that one following the next one is also going to be cancelled, but I'll let you know.

6
00:00:22,770 --> 00:00:31,589
But certainly the very next one has been cancelled. But today we have Professor Simon Cole who is here and who's visiting the centre.

7
00:00:31,590 --> 00:00:38,280
So you've probably seen him in the centre. He's here until the end of June, so you can harass him after the talk, but also over the next few months.

8
00:00:39,360 --> 00:00:41,579
And Simon is professor of Criminology,

9
00:00:41,580 --> 00:00:47,520
Law and Society and Director of the NIEKERK Centre for Science and Society at the University of California, Irvine.

10
00:00:47,880 --> 00:00:56,280
Simon's also the Academy to US Editor in chief of Theoretical Criminology, and he's associate editor of the National Registry of Exonerations.

11
00:00:57,180 --> 00:01:01,580
Today, Simon is going to be talking to us about that registry and about the work he does with it.

12
00:01:01,620 --> 00:01:08,620
So thank you. Thanks. Thank you, Mary, and thanks for such a great turnout.

13
00:01:10,090 --> 00:01:16,209
I was really excited to when Mary asked me to talk about the National Registry of Exonerations,

14
00:01:16,210 --> 00:01:25,130
even though she later claimed not to remember doing so because no one's really no one in academia has ever really asked me to talk about it.

15
00:01:25,130 --> 00:01:32,170
And I get asked to talk about forensic science or the Castle State, but not about this, let alone in criminology.

16
00:01:32,170 --> 00:01:36,700
So. So I'm kind of excited to talk about this and thanks for such great turnout.

17
00:01:37,660 --> 00:01:45,370
I was just going to say a little bit about myself, since you may not know me and some of my background is sort of relevant to what I'll say later,

18
00:01:45,640 --> 00:01:52,180
and then I'll talk a little bit about the context of the registry, and then I'll show you the registry.

19
00:01:52,540 --> 00:02:01,270
And then at that point three, there will have been like no intellectual or research content whatsoever in the talk.

20
00:02:01,540 --> 00:02:07,090
And so then in part four, I will try to introduce some intellectual and research content by talking about a project,

21
00:02:07,270 --> 00:02:13,210
research project that I'm doing that I would be happy to get everyone's feedback on because it's ongoing.

22
00:02:14,290 --> 00:02:17,379
So I just want to say a little bit about myself.

23
00:02:17,380 --> 00:02:23,470
I work in a it's sort of interesting in the context of the Criminology Centre, I work in the Department of Criminology,

24
00:02:23,470 --> 00:02:31,330
Law and Society in which either not one or only one faculty member has a degree in criminology.

25
00:02:32,230 --> 00:02:36,280
So about half of us have a degree in sociology and half are other.

26
00:02:37,000 --> 00:02:38,049
I'm part of the other.

27
00:02:38,050 --> 00:02:45,550
I have a great degree in science and technology studies, which is sort of social social studies of the making of scientific knowledge.

28
00:02:46,330 --> 00:02:51,310
Bruno Latour is the person whose name you would know in this field if you know anyone,

29
00:02:52,240 --> 00:02:59,490
although that's a little sad to say in the UK because there's a lot of major figures in that field that I'm from the UK cleaning stories,

30
00:02:59,530 --> 00:03:02,769
Steve Woolgar, who is still here at the business school,

31
00:03:02,770 --> 00:03:12,160
and in the early days there was an Edinburgh School of Science Studies and there was a back school of lots of stuff other than Latour.

32
00:03:13,570 --> 00:03:24,760
And I worked on forensic science as a project in scientific knowledge, very difficult to get jobs in science studies in the US.

33
00:03:24,850 --> 00:03:30,459
So I ended up using the forensic science thing to move into the field of criminology, law,

34
00:03:30,460 --> 00:03:36,610
society and my work on forensic science saying that a lot of it didn't seem to have validated itself,

35
00:03:37,030 --> 00:03:42,759
got people in the innocence, research and movement interested in my work,

36
00:03:42,760 --> 00:03:49,540
and I started getting invited to conferences about wrongful conviction and innocence.

37
00:03:49,750 --> 00:03:57,879
And you can't really go to such things where there are exonerees in attendance who are talking about their stories

38
00:03:57,880 --> 00:04:05,400
of being wrongly convicted and come away not wanting to kind of do something about the problem and get involved.

39
00:04:06,010 --> 00:04:14,820
And for a while I wanted to start an Innocence Project at UC Irvine, maybe focussed on forensic science,

40
00:04:14,830 --> 00:04:21,489
which I did start it very quietly, because once you start one of these things, you start getting mail from prisons.

41
00:04:21,490 --> 00:04:29,260
And I didn't want to be someone who was like let down people's hopes and it didn't really work.

42
00:04:29,260 --> 00:04:34,750
We didn't have we didn't have enough lawyers. I'm not a lawyer and we weren't really able to help people.

43
00:04:35,170 --> 00:04:40,120
And then the opportunity came along to work on the National Registry of Exonerations.

44
00:04:40,120 --> 00:04:47,709
It was looking for a new home, and that seemed a better fit, both for me and for UC Irvine.

45
00:04:47,710 --> 00:04:58,210
It's a research project. We don't actually help people or work with people, so we didn't need to do that and it's very much an academic project.

46
00:04:58,210 --> 00:05:05,560
So it was I was really thrilled to to be able to get involved in the Innocence Movement this, this way.

47
00:05:06,730 --> 00:05:15,940
So that's kind of how how I came to it. So let me tell you a little bit about the context of the National Registry of of Exonerations.

48
00:05:16,660 --> 00:05:27,489
So wrongful long, full conviction has like a long history of research and discussion in the United States as well as as well as here.

49
00:05:27,490 --> 00:05:33,760
And it's nice to talk about this with Carolyn Foyle here has done a lot of research on this on this topic.

50
00:05:34,540 --> 00:05:42,580
But you could go back to 1923. I don't know if if you've heard of this judge learned hand this side from having just a great name.

51
00:05:43,840 --> 00:05:50,410
She is often called the greatest US judge never to serve on the U.S. Supreme Court.

52
00:05:51,310 --> 00:05:56,590
And he has this quotation here saying, there isn't you know, we don't really need to worry about wrongful conviction.

53
00:05:56,590 --> 00:06:03,970
It doesn't actually happen. It's just a myth. And starting right around that time, not.

54
00:06:04,050 --> 00:06:09,450
Long afterwards, you start getting academics sort of debunking that claim.

55
00:06:09,540 --> 00:06:16,740
And the first one is probably Edward Borchardt is a law professor at Yale, publishes this book in 1932.

56
00:06:17,100 --> 00:06:25,770
And one of the things they do in order to debunk the claim that there are no wrongful convictions is to start compiling lists of wrongful convictions.

57
00:06:25,770 --> 00:06:29,010
And that's where you get this whole idea of lists.

58
00:06:29,310 --> 00:06:36,600
And you can kind of walk through all the history of books about wrongful conviction in the United States.

59
00:06:36,600 --> 00:06:44,880
And they almost always have a list. There's always something, some kind of a list, maybe good, maybe bad, but some kind of a list.

60
00:06:46,830 --> 00:06:53,070
So it's important to note, again, this may be obvious, but let me just note it,

61
00:06:54,090 --> 00:06:59,730
that when we talk about wrongful convictions, we're trying to sort of make lists and count wrongful convictions.

62
00:06:59,910 --> 00:07:04,530
We can't really count wrongful convictions because we don't know whether people are guilty or innocent.

63
00:07:04,530 --> 00:07:09,180
That's the whole point. All we know is whether they've been convicted or acquitted.

64
00:07:09,930 --> 00:07:14,850
So we kind of try to count wrongful convictions by counting exonerations.

65
00:07:14,850 --> 00:07:25,679
So people who have somehow been released by the legal system on grounds of innocence as a proxy for wrongful convictions.

66
00:07:25,680 --> 00:07:32,190
But of course, we know that it's a very imperfect proxy and that there must be lots and lots of people,

67
00:07:32,640 --> 00:07:39,450
multiples of the people who are exonerated, who are wrongly convicted and just aren't able to prove it and nobody finds out about it.

68
00:07:39,450 --> 00:07:48,270
So we never, never learn that. So exonerations are not the same as wrongful convictions in both directions.

69
00:07:50,190 --> 00:07:55,530
Some exonerations, maybe of guilty people. I think that's probably a really small number.

70
00:07:56,700 --> 00:08:02,129
But most wrongful convictions are probably not wrongful, not exonerations.

71
00:08:02,130 --> 00:08:10,060
So exonerations are probably a very small drop in the bucket of actual wrongful convictions that we don't know about and we never will know about.

72
00:08:11,880 --> 00:08:21,600
And how do I know this? Well, just read a few exonerations and see how lucky these people were to actually get exonerated and

73
00:08:22,020 --> 00:08:27,239
look at the fortuitously involved in each of these cases that they actually got exonerated.

74
00:08:27,240 --> 00:08:30,240
And then think of the people who didn't get those lucky breaks along the way.

75
00:08:30,420 --> 00:08:33,510
And those are the unknown, wrongly convicted.

76
00:08:35,400 --> 00:08:40,889
So this list making kind of culminates in around 1987 when Hugo Pardo and Michael

77
00:08:40,890 --> 00:08:46,920
Radelet publish this list of 350 wrongful convictions in death penalty cases.

78
00:08:48,420 --> 00:08:55,469
And they use this method in the middle here, cases in which we believe a majority of neutral observers,

79
00:08:55,470 --> 00:08:59,940
given the evidence at our disposal, would judge the defendant in question to be innocent.

80
00:08:59,950 --> 00:09:05,940
So we read the cases. We think these people are innocent and we think you, the reader, would agree.

81
00:09:06,900 --> 00:09:17,340
And this is the Reagan administration. So the attorney general didn't like this, and he sent these law professors, Paul Cosell and Stephen Markman.

82
00:09:17,370 --> 00:09:22,380
Paul Cosell is still around and says, you know,

83
00:09:23,400 --> 00:09:29,820
take down this study and they go and read the cases and they pick the weakest, the ones with the weakest claim to innocence.

84
00:09:29,820 --> 00:09:33,930
And they say, we've read this case file really carefully and we don't think this person is innocent.

85
00:09:33,930 --> 00:09:37,979
They look guilty as [INAUDIBLE] to us. And so we don't think this person is innocent.

86
00:09:37,980 --> 00:09:43,440
And if you don't and therefore you can't trust these researchers, they're anti-death penalty advocates.

87
00:09:43,770 --> 00:09:47,040
So we don't believe any of the 350 cases.

88
00:09:47,430 --> 00:09:52,980
And, you know, you can say what you want about that, but it was a somewhat effective critique.

89
00:09:53,760 --> 00:09:57,959
And then DNA happens. So DNA changes the game.

90
00:09:57,960 --> 00:10:04,380
DNA comes from the UK, from Al Jeffress at the University of Leicester.

91
00:10:04,560 --> 00:10:09,960
And the first case it was used in, actually the first case is an immigration case,

92
00:10:10,950 --> 00:10:15,779
but then it was used in a murder and it actually averted a potential wrongful

93
00:10:15,780 --> 00:10:22,230
conviction because the person suspected in these murders was confessed to the crime.

94
00:10:23,790 --> 00:10:29,909
And then that confession was disproved by analysis of the DNA, DNA evidence,

95
00:10:29,910 --> 00:10:35,930
and it identified another person as the perpetrator of this two murders of teenage girls.

96
00:10:36,000 --> 00:10:46,170
So that was the first exoneration involving DNA and also sort of begins to open up the potential for for the use of of of DNA.

97
00:10:47,160 --> 00:10:51,790
Again, you may know all of this, but I'll just quickly mention.

98
00:10:52,290 --> 00:11:01,790
DNA is used in exonerations when there is bodily fluid at the crime scene that probably has to come from the perpetrator.

99
00:11:01,800 --> 00:11:15,990
It's not that useful if not. And so and so such cases, especially in the preceding decades when DNA evidence was not that powerful and sensitive,

100
00:11:16,980 --> 00:11:22,470
the kind of cases that have this biological evidence tend to be rape murder cases,

101
00:11:22,710 --> 00:11:27,000
their murder cases where there's a semen sample because of because of a rape.

102
00:11:27,690 --> 00:11:31,770
So they overwhelmingly skew towards that kind of case.

103
00:11:33,300 --> 00:11:44,490
And so then this spreads to the United States, and you have the first DNA exoneration in 1989 of David Vasquez.

104
00:11:44,760 --> 00:11:50,580
And actually the case is remarkably similar to the Lester case.

105
00:11:50,910 --> 00:12:01,260
Again, Vasquez is this mildly mentally retarded, confessed to the crime through a dream confession,

106
00:12:01,500 --> 00:12:06,240
pled guilty, and then DNA eventually found the true perpetrator.

107
00:12:06,600 --> 00:12:12,780
And so this inspires the famous Innocence Project in the United States, where the lawyers,

108
00:12:12,780 --> 00:12:20,100
Barry Scheck and Peter Neufeld, say, see the potential of using DNA to to get people out of prison.

109
00:12:23,070 --> 00:12:27,030
But again, in order to do that, they need a very special kind of case,

110
00:12:27,030 --> 00:12:34,850
a case with biological evidence at the crime scene that's not tested at the time of the original investigation that's preserved.

111
00:12:34,860 --> 00:12:39,960
And then that's that's a big one. The legal requirements to preserve that evidence are essentially nil.

112
00:12:40,320 --> 00:12:47,370
So it's just luck whether the police decided to preserve that evidence and then it's able to be tested later.

113
00:12:47,730 --> 00:12:53,930
So those are the only cases that are eligible to become a DNA exoneration.

114
00:12:55,860 --> 00:13:06,720
But still they start litigating and they start getting people out of prison and they start getting a lot of media attention for doing so.

115
00:13:07,080 --> 00:13:13,110
And in some senses, they shock the criminal justice system that they're able to keep finding these innocent people in prison.

116
00:13:14,280 --> 00:13:18,209
And people keep saying they're going to it's going to taper off.

117
00:13:18,210 --> 00:13:24,090
They're going to run out of cases. And that keeps not happening for 20 years.

118
00:13:26,040 --> 00:13:30,150
Up to over 350 cases of DNA exonerations.

119
00:13:31,230 --> 00:13:38,400
So the importance of this for the discourse over wrongful convictions is that the DNA except DNA exonerations

120
00:13:38,400 --> 00:13:46,650
become a set of exonerations for which it is almost impossible to doubt factual innocence because you have DNA,

121
00:13:47,370 --> 00:13:54,780
usually from a semen sample, from a rape murder that doesn't match the person who's been convicted.

122
00:13:54,960 --> 00:13:58,860
It becomes kind of hard to believe that they are guilty of that crime.

123
00:13:59,220 --> 00:14:06,420
In the early cases, there were attempts to sort of reconcile these things, usually by saying, well,

124
00:14:06,420 --> 00:14:14,590
the person in prison didn't rape the victim, but maybe they were there and killed the victim while an accomplice rape them.

125
00:14:15,810 --> 00:14:26,790
But that argument became less and less tenable as more and more cases piled up until even what I would call innocence sceptics.

126
00:14:26,790 --> 00:14:34,859
The Paul Castles of the world had to admit that most of these 350 people were innocent of these crimes.

127
00:14:34,860 --> 00:14:39,540
And even if you wanted to say maybe a couple of them were guilty, you would.

128
00:14:39,720 --> 00:14:44,340
It's hard to really argue that most of them that that a large number were.

129
00:14:45,270 --> 00:14:52,590
So for the first time and these compiling of these list, you have this kind of bullet-proof list, as they call it.

130
00:14:52,890 --> 00:14:57,570
But as I mentioned already, it's a very skewed and specific set of cases.

131
00:14:57,810 --> 00:15:02,400
And if we can find our discussion of wrongful conviction into these DNA exonerations,

132
00:15:02,400 --> 00:15:11,670
it would be a very skewed and selective and unrepresented tiff discussion of both exonerations and of wrongful convictions.

133
00:15:13,680 --> 00:15:18,059
So then you get the National Registry of Exonerations,

134
00:15:18,060 --> 00:15:28,260
and this project originated from the idea of creating an encyclopaedia of wrongful convictions that Rob Horton,

135
00:15:28,260 --> 00:15:34,560
who's a journalist, very active in the exoneration movement, and Michael Radelet, a sociologist, came up with.

136
00:15:35,460 --> 00:15:47,820
And Sam GROSS, who became the founder of the registry, kind of made a comment on the proposal to create this encyclopaedia, which this was,

137
00:15:47,820 --> 00:15:56,790
I think, the early 2000s, and said, of course, you might want to have a website associated with this, with this encyclopaedia.

138
00:15:57,300 --> 00:16:02,910
And so that turned out to be that comment turned out to be sort of right.

139
00:16:02,910 --> 00:16:15,670
Precisely. At the time in technological history where a year or two later it seemed obvious that this should not be a book that would remain static,

140
00:16:15,670 --> 00:16:24,970
but that it should be a sort of living database that's on the web, and that it would be completely useless if it were a fixed text kind of book.

141
00:16:26,230 --> 00:16:31,360
And so the encyclopaedia involved evolves into the National Registry of Exonerations,

142
00:16:31,360 --> 00:16:36,400
founded by Sam GROSS at the University of Michigan and Rod Warden, who's at Northwestern.

143
00:16:38,710 --> 00:16:43,720
Northwestern drops out soon thereafter, and it becomes housed at Michigan.

144
00:16:44,620 --> 00:16:54,670
And then Sam hires Maurice Poe's Lee, a Pulitzer Prize winning journalist who used to work at the Chicago Tribune as the senior researcher.

145
00:16:54,910 --> 00:16:58,930
And Maurice is kind of a game changer for the registry.

146
00:16:58,960 --> 00:17:03,310
He starts turning the article right up, which I'll show you in a second,

147
00:17:04,600 --> 00:17:11,850
from very kind of threadbare things written by students into professional journalist kind of stuff.

148
00:17:11,860 --> 00:17:17,260
And he does it very quickly and really changes the registry.

149
00:17:17,620 --> 00:17:22,990
And then Sam starts looking, thinking about retiring and about succession.

150
00:17:23,380 --> 00:17:32,200
And UC Irvine, because of having several researchers interested in this topic, emerges as a possible home,

151
00:17:32,620 --> 00:17:38,860
and it ends up being a joint project of UCI, Michigan State and the University of Michigan.

152
00:17:40,980 --> 00:17:45,140
Okay. So we've talked about lists.

153
00:17:45,740 --> 00:17:52,250
So what we have in a sense are two different approaches to how we would make a list of exonerations.

154
00:17:52,640 --> 00:17:57,110
There's what you might call the social science model embodied by Vido and Radelet,

155
00:17:57,830 --> 00:18:03,410
where the researchers are using their judgement to decide whether people are guilty or innocent.

156
00:18:04,490 --> 00:18:11,630
But as we saw, that can be attacked by anyone who just says, I don't agree with your judgement that this person is innocent.

157
00:18:13,220 --> 00:18:18,320
Sam's approach is a deference model which he wants to avoid that problem.

158
00:18:19,490 --> 00:18:24,410
And the key to it is deference to the state about in a sense,

159
00:18:24,410 --> 00:18:33,170
the registries definition of an innocent eration is dependent upon the state exonerating the person.

160
00:18:33,320 --> 00:18:36,980
And the idea is we're not making our own judgements about any case.

161
00:18:37,010 --> 00:18:44,470
These are cases where the state itself has decided to relieve this person of the consequences,

162
00:18:44,480 --> 00:18:49,700
the punitive consequences of the case of the act for which they were convicted.

163
00:18:50,540 --> 00:18:54,980
The problem is with that is that the state's judgement might not be the same as my judgement.

164
00:18:55,430 --> 00:18:58,040
But I have to just sort of subsume that.

165
00:18:59,330 --> 00:19:07,340
You could ask questions about why we would rely on the state to expose its own failures, but of course the state is exposing its own failures a lot.

166
00:19:08,360 --> 00:19:11,810
And of course, someone could attack this as well if they wanted to.

167
00:19:12,440 --> 00:19:15,520
So that's two ways of doing it. I read this article recently.

168
00:19:15,530 --> 00:19:16,939
Maybe there's another model.

169
00:19:16,940 --> 00:19:26,870
Maybe we, instead of counting, we could just do what this judge in Japan did in this comment on the Carlos Ghosn case and just say,

170
00:19:27,140 --> 00:19:31,580
without counting, there are as many wrongful convictions as grains of sand on the beach.

171
00:19:34,310 --> 00:19:41,510
Then you no need to count them. So here's here's our short definition of exoneration.

172
00:19:47,290 --> 00:19:55,780
And here's the long definition and here's the checklist version of the long definition.

173
00:19:56,110 --> 00:20:05,740
You need to be convicted of a crime. You need to be officially cleared of all related charges based at least in part

174
00:20:05,770 --> 00:20:12,040
on new evidence of innocence without unexplainable physical evidence of guilt.

175
00:20:13,960 --> 00:20:14,380
That's it.

176
00:20:16,270 --> 00:20:25,840
And so one way I like to explain this, this is to go back to the point about the relationship between wrongful convictions and exonerations.

177
00:20:26,950 --> 00:20:32,020
Let's imagine that the criminal justice system mostly gets it right,

178
00:20:32,140 --> 00:20:40,950
and most convictions are right for lots more, and thus only a smaller number are wrongful.

179
00:20:40,960 --> 00:20:51,190
Let's just assume that for the sake of argument, exonerations are going to capture only the tip of the iceberg of wrongful convictions.

180
00:20:51,190 --> 00:21:00,100
There will be lots of wrongful convictions that we never know about that never become exonerations, huge categories of of cases,

181
00:21:00,130 --> 00:21:07,720
including especially low level crimes misdemeanours where it's almost impossible to get to go through all

182
00:21:07,720 --> 00:21:13,510
the legal process that you would need to to get exonerated and probably not worth it to you to to do so.

183
00:21:14,530 --> 00:21:21,459
And maybe with our definition of exoneration, there's a much smaller number of actually guilty people who,

184
00:21:21,460 --> 00:21:25,360
for whatever reason, meet our definition of exoneration.

185
00:21:26,200 --> 00:21:35,530
And our approach is to stick to the definition and accept these two error rates despite ourselves.

186
00:21:35,710 --> 00:21:43,810
And so if if we happen to run across someone who we're not really that convinced they're innocent, they meet the definition we put them in.

187
00:21:44,680 --> 00:21:51,940
But again, it's much more common to run across cases where we really think the person is innocent.

188
00:21:52,210 --> 00:21:59,200
And and we can't we they don't meet our definition and so we can't include them.

189
00:21:59,440 --> 00:22:04,300
And those are painful experiences, especially me being new at the registry.

190
00:22:05,740 --> 00:22:10,030
And yet that's that's the rules we've kind of set for ourselves.

191
00:22:13,190 --> 00:22:18,510
I always I always fail to click these little. You got all that red.

192
00:22:18,580 --> 00:22:25,950
Okay. So now let me try to show you a little bit of the registry itself.

193
00:22:29,210 --> 00:22:32,690
And I've got this temperamental mouse here, so bear with me.

194
00:22:40,270 --> 00:22:43,570
He's another one that's got the underline. Was it.

195
00:22:44,200 --> 00:22:49,120
Yeah. No, I think it's it's it's okay. I think I just saw the.

196
00:22:49,660 --> 00:22:54,489
Yeah, they did already. Open it. Okay. All right.

197
00:22:54,490 --> 00:22:59,830
So again, bear with me a little bit as I navigate this.

198
00:23:00,370 --> 00:23:06,590
So this is the registry. I was it.

199
00:23:06,800 --> 00:23:10,820
This is your Internet connection. Has it clicked on it?

200
00:23:11,300 --> 00:23:15,230
Yeah. Yeah, it says waiting. Yeah.

201
00:23:24,630 --> 00:23:29,660
Erica. Wow.

202
00:23:37,670 --> 00:23:40,820
This is so sad. So is that.

203
00:23:41,720 --> 00:23:45,440
That's not normal. Yeah, that's not normal. We had that before. Right, Samuel?

204
00:23:47,550 --> 00:23:51,459
We're going to put that up. But there it is. That's what I want.

205
00:23:51,460 --> 00:23:55,790
Wanted. I think so.

206
00:23:55,990 --> 00:24:00,490
Didn't you close it? Did you close it? What if I go forward?

207
00:24:04,970 --> 00:24:09,740
Yes. Okay. All right. So this is the heart of the registry.

208
00:24:10,670 --> 00:24:14,340
This is a this is a table with 2551 row.

209
00:24:14,360 --> 00:24:17,630
So we have 2551 exonerees.

210
00:24:17,930 --> 00:24:23,120
And so we have these basic demographic characteristics of each case.

211
00:24:23,390 --> 00:24:29,600
And then you can sort and filter by all of these demographics.

212
00:24:32,420 --> 00:24:38,930
So a table and a data set, that's the heart of the registry in a certain sense.

213
00:24:45,550 --> 00:24:48,580
But the other part is that any one of these cases.

214
00:24:53,630 --> 00:25:03,350
So I open it up. I get this professionally written by a journalist narrative summary of the case,

215
00:25:03,860 --> 00:25:14,599
and as you see that they're pretty substantive and compared to other lists that had been compiled before by,

216
00:25:14,600 --> 00:25:17,880
for example, the Innocence Project or the Centre on Wrongful Convictions.

217
00:25:17,900 --> 00:25:30,660
These are much more thorough, and we have two former former journalists who are writing these these things, and they're journalists, right?

218
00:25:30,680 --> 00:25:38,000
So they write this stuff really quickly and then they go on to the next one because that's that's what he used to do in newspapers.

219
00:25:38,450 --> 00:25:43,130
And it's kind of a piece of work that I'm not used to as an academic.

220
00:25:44,570 --> 00:25:51,980
And it's it's really incredible. And we're posting about 150 cases per year.

221
00:25:52,520 --> 00:25:56,000
So we're posting one every two or three days.

222
00:25:57,800 --> 00:26:02,510
And there. And so I've written a couple of these just to see what it was like, you know,

223
00:26:02,510 --> 00:26:11,540
and it took me like three months before I felt comfortable with the facts and having checked out all the sources and so on.

224
00:26:11,720 --> 00:26:20,600
And so they're checking out all the sources, all the research, and doing a much more elaborate social science coding than I showed you,

225
00:26:20,600 --> 00:26:24,200
because we have all sorts of other variables we're coding that are not on the public

226
00:26:24,200 --> 00:26:29,749
website doing all that and putting it up on the web once every two or three days.

227
00:26:29,750 --> 00:26:36,980
A new case, which I think is impressive, both for what we're doing as the registry,

228
00:26:37,190 --> 00:26:42,950
but also for the fact that someone's being exonerated in the United States once every two or three days.

229
00:26:47,820 --> 00:26:51,810
Okay. What else is on my list of things I wanted to show you very quickly.

230
00:26:52,080 --> 00:26:55,080
What's the sources of the people rising up?

231
00:26:55,590 --> 00:26:59,850
It's it's legal. It's almost entirely legal documents.

232
00:26:59,850 --> 00:27:06,900
So the legal pleadings and media sources, newspaper, newspaper articles and so on.

233
00:27:07,140 --> 00:27:11,490
We talked to the lawyers. We don't talk to the exonerees.

234
00:27:12,780 --> 00:27:21,420
So we usually be an interview with the lawyer and trying to get the lawyers, send us briefs and so on, court rulings, legal judgements and so on.

235
00:27:22,040 --> 00:27:33,809
But behind the behind all this, there's a source document folder for every single case which varies greatly in quality and sort of thoroughness.

236
00:27:33,810 --> 00:27:38,520
But that's almost entirely legal documents and newspaper articles.

237
00:27:39,450 --> 00:27:42,000
So not things that many other people want.

238
00:27:42,030 --> 00:27:49,920
Like I'm interested in forensic science, so I want the lab reports and the transcripts, you know, have them in almost every case.

239
00:27:51,720 --> 00:27:55,020
That does bring me to something I was going to say is one thing, you know,

240
00:27:55,020 --> 00:28:00,720
I sort of knew about the registry, but I didn't really know it very well until I took it over running it.

241
00:28:01,680 --> 00:28:12,509
So one thing is that to 2551 cases, it's sort of like geologic strata in terms of how the data was collected.

242
00:28:12,510 --> 00:28:19,290
As you go back in time and the data collection has gotten better and more rigorous and more reliable.

243
00:28:19,380 --> 00:28:26,670
The closer we get to the present and we don't have the resources to go back and rework the old cases.

244
00:28:27,390 --> 00:28:29,970
So you find a lot of variety in our data,

245
00:28:30,000 --> 00:28:39,210
both in terms of source documents and in terms of the coding of social science variables depending on when the case was posted.

246
00:28:40,170 --> 00:28:47,220
And you will see that when you look at the source documents on one document and these old cases and then ten in recent cases.

247
00:28:47,430 --> 00:28:53,250
So that comes up a lot. I'm.

248
00:28:56,520 --> 00:29:08,400
Let me just bring up. It's over here.

249
00:29:08,700 --> 00:29:10,860
In that case, what happened to his co-defendants?

250
00:29:17,160 --> 00:29:25,550
Evidence says he was sentenced to death and he was sentenced to life without parole because he was 17.

251
00:29:26,130 --> 00:29:29,250
Is that right? Yeah. I don't know. I don't know.

252
00:29:29,940 --> 00:29:36,540
Sorry. That should be a cop. Yeah. If he were match, then he'd be hyperlinked like this.

253
00:29:36,540 --> 00:29:39,600
Co-defendant. Right. Yeah. Okay.

254
00:29:39,750 --> 00:29:43,140
So maybe they're not exonerated. They might have taken a plea.

255
00:29:43,260 --> 00:29:49,860
Yeah. To get out or plead to one of the charges or something like that.

256
00:29:50,520 --> 00:29:54,270
So, yeah, it's not uncommon with co-defendants.

257
00:29:54,280 --> 00:29:58,680
Some of them are exonerated and some are not. So.

258
00:29:58,770 --> 00:30:06,160
And a plea would be a very common thing that they took some deal to go home and the other one didn't or wasn't offered it.

259
00:30:07,080 --> 00:30:17,910
This is a really sad thing to look at, but we're able to compile a list of the longest incarcerations for four exonerees.

260
00:30:19,110 --> 00:30:23,399
And so this just says something about sentencing practices in the state.

261
00:30:23,400 --> 00:30:30,840
It states. And it wasn't just the people at the back of the room.

262
00:30:30,840 --> 00:30:36,690
It nearly always says like, Oh, well, yeah, but it's how long they serve actually serves.

263
00:30:36,710 --> 00:30:41,880
Yeah. So. So the record holder is 45 years in prison for a crime they didn't commit.

264
00:30:41,880 --> 00:30:45,960
And the top ten are all over 35 years.

265
00:30:45,990 --> 00:30:51,810
And we keep adding these new cases like high on the list and pushing people,

266
00:30:51,990 --> 00:31:03,870
people with 30 more than 30 year incarcerations off the top ten of longest incarcerations, and some of them famous cases like this.

267
00:31:03,870 --> 00:31:08,040
Wilbert Jones, who was in Angola, the famous case.

268
00:31:11,370 --> 00:31:18,210
Okay. And then just want to show you a little bit about how the registry works.

269
00:31:24,270 --> 00:31:28,080
So the advisory board isn't that important.

270
00:31:29,370 --> 00:31:32,640
This is being test for how the registry works.

271
00:31:32,650 --> 00:31:37,860
They're very important. So, so founded by Sam and Rob Gordon.

272
00:31:39,240 --> 00:31:43,080
And then the plan, I can't really scroll very well here.

273
00:31:43,860 --> 00:31:51,990
So we have three editors kind of taking over the job from Sam Barba, Brian and Catherine Grosso, who are at Michigan State and myself.

274
00:31:54,900 --> 00:32:06,910
And just want to point and again, I can't scroll, but we have two professional journalists more closely and can do that.

275
00:32:07,500 --> 00:32:14,580
Yeah, but I'm using that wheel so that's what's now doesn't it's all right.

276
00:32:16,860 --> 00:32:25,260
So we now have not one, but two former professional journalists, Maurice Persily and Ken Waterberg, researching the cases and writing the cases.

277
00:32:26,040 --> 00:32:37,650
Jessica Paradis, who's a lawyer, who's doing the first edit and checking the coding, and then it goes to the editors.

278
00:32:37,950 --> 00:32:46,810
And then even the gal who's doing administration and development and communications and kind of running the whole centre.

279
00:32:46,830 --> 00:32:56,790
So again, it's kind of an unusual thing for me as an academic and in some ways feels more like running a newspaper than than an academic project.

280
00:32:58,110 --> 00:33:04,560
It kept going, going at a very fast pace. That tends to be driven by the journalists who are sort of used to that.

281
00:33:05,790 --> 00:33:08,880
Let's not restart it. We got it.

282
00:33:08,970 --> 00:33:13,140
All right. So I'm going to close this mercifully.

283
00:33:14,760 --> 00:33:23,520
I think there is. What's happening, though. Okay.

284
00:33:31,230 --> 00:33:35,430
Okay. All right. Now I just have to get rid of this. Rita, do we start now?

285
00:33:36,000 --> 00:33:44,220
We start tonight. It always happens in windows, doesn't it?

286
00:33:44,370 --> 00:33:47,549
Because no one maintains this thing because it's not there. It was announced.

287
00:33:47,550 --> 00:33:51,060
First you have to use a mouse. It's working. Okay. All right.

288
00:33:51,180 --> 00:33:56,820
So. So, as promised, sort of intellectual and research free so far.

289
00:33:56,850 --> 00:34:05,100
Right now, I want to talk a little bit about a project on on forensic science, because that's my academic interest.

290
00:34:06,000 --> 00:34:10,140
And so since I came on, they said, well, you can work on the forensic science part.

291
00:34:10,980 --> 00:34:19,500
So a little bit after I started, Gerry Laporte, who is the head forensic science guy at the National Institute of Justice in the United States,

292
00:34:20,130 --> 00:34:26,070
published this article on wrongful convictions and DNA exonerations.

293
00:34:26,430 --> 00:34:30,270
And he's got this little thing, so you probably can't read this in the back.

294
00:34:30,510 --> 00:34:42,210
He says, Well, the Innocence Project says that 157 of its 342 DNA exonerations, forensic science contributed to the wrongful conviction.

295
00:34:42,660 --> 00:34:49,030
That's 46%. And so they're saying bad things about forensic science.

296
00:34:50,310 --> 00:34:55,250
But I looked at the National Registry of Exonerations, and they're inconsistent.

297
00:34:55,260 --> 00:35:04,379
So 24 of those, 157 cases, they say the registry says forensic science did not contribute to that.

298
00:35:04,380 --> 00:35:09,390
It was like mistaken eyewitness identification or false confessions and so on.

299
00:35:09,960 --> 00:35:15,810
So I talked to the Innocence Project, who we work with closely, and we said, this isn't good.

300
00:35:15,810 --> 00:35:23,760
We need to kind of be consistent about this or at least figure out why we have this discrepancy and what the you know,

301
00:35:23,760 --> 00:35:31,590
what the issue is that we disagree about. So we started by looking at our definitions of forensic science.

302
00:35:31,860 --> 00:35:36,750
We call it false or misleading forensic science, forensic evidence.

303
00:35:36,900 --> 00:35:41,190
And they call it misapplication of forensic science.

304
00:35:43,110 --> 00:35:48,230
And it it certainly didn't seem like the root of the problem was the definitions.

305
00:35:48,240 --> 00:35:54,270
They are different, but there's nothing fundamentally there's no sort of real fundamental disagreement there.

306
00:35:54,750 --> 00:36:00,030
So the first thing to do was to come up with see if we could come up with an agreed upon definition.

307
00:36:00,870 --> 00:36:05,940
And both of us were open to that. And so we did.

308
00:36:06,480 --> 00:36:13,210
And so the new one we're going to call fit forensic expert evidence problem.

309
00:36:14,310 --> 00:36:18,260
And here's the very simple definition of it.

310
00:36:18,270 --> 00:36:26,729
So this is, I think, the first time I've really presented this definition to any audience other than within the internal work product.

311
00:36:26,730 --> 00:36:31,049
So I'm interested, you know, and it's not final. It hasn't gone up on any website yet.

312
00:36:31,050 --> 00:36:37,890
So I'm interested in feedback and and thoughts about it, but, but we are pretty happy with it.

313
00:36:37,900 --> 00:36:48,660
So not too much feedback. But what I'm going to try to do now is sort of explain to you some of the thinking that went into this definition,

314
00:36:48,660 --> 00:36:52,110
which actually for such a simple thing, there was a fair amount.

315
00:36:52,350 --> 00:36:57,059
So it raised a couple of questions. So one is kind of the science question.

316
00:36:57,060 --> 00:37:02,790
We did notice that we were just that some of us on both sides were not coding

317
00:37:02,790 --> 00:37:09,299
cases as forensic because the forensic science seemed like not forensic science.

318
00:37:09,300 --> 00:37:13,700
So and here's some examples on the left, hypnosis, slide detection, dog said.

319
00:37:13,710 --> 00:37:17,580
So somebody was just saying, well, that isn't forensic science, so it doesn't count.

320
00:37:19,350 --> 00:37:28,229
And the solution to this and this is why I talked about the science studies framework at the beginning,

321
00:37:28,230 --> 00:37:38,700
is because here's where I think science studies came into it, is that I'm so defining science is famously difficult in philosophy of science.

322
00:37:39,330 --> 00:37:48,090
So I wanted to avoid doing it. And in the end I kind of said, I don't care if if it's science or not,

323
00:37:48,390 --> 00:37:55,740
if the court is admitting this person as some kind of an expert, then that's fair game.

324
00:37:56,070 --> 00:38:03,180
And that's why the title of it, we have the word expert in it if you want to say it's not forensic science for whatever reason.

325
00:38:03,180 --> 00:38:12,570
But to me, forensic science is anything any kind of expertise that gets admitted into court is in some sense forensic science.

326
00:38:13,320 --> 00:38:22,440
So auto mechanics, psychologists, gang experts, cultural experts, hypnotists, if the courts are letting them in, I think it's forensic science.

327
00:38:22,440 --> 00:38:28,650
If you don't think it's forensic science, you should maybe talk to the courts about not letting hypnosis hypnotists testify.

328
00:38:29,750 --> 00:38:40,490
In court. And another way we have of saying it down here is that, you know what science is the question you're asking, not the answer you're giving.

329
00:38:41,000 --> 00:38:49,280
So if you're asking, do people with tattoos, are they more often in a gang than not in a gang?

330
00:38:49,970 --> 00:38:52,160
That's an empirical, scientific question.

331
00:38:52,190 --> 00:38:58,550
Now, whether a cultural anthropologist has empirical scientific ways of answering that question, I don't know.

332
00:38:58,790 --> 00:39:03,680
But if you want to claim to have knowledge about that question, then that's a scientific question.

333
00:39:06,650 --> 00:39:11,210
So let me give two examples of this from actual cases.

334
00:39:11,220 --> 00:39:17,440
So one is the Raymond Jennings case, and this case will be interesting to the criminology audience.

335
00:39:17,450 --> 00:39:21,950
So this this was a murder in a parking lot.

336
00:39:23,090 --> 00:39:27,310
And Jennings was the security guard on duty.

337
00:39:27,320 --> 00:39:32,570
And so eventually, for lack of other suspects, they decided he was the only one there.

338
00:39:32,570 --> 00:39:45,770
So it must it must have been him. And they got this expert witness, an FBI agent, who said that he was an expert in victimology.

339
00:39:45,800 --> 00:39:53,440
Now, this was interesting as a criminologist, because victimology is a thing in criminology, but that's not his victimology.

340
00:39:53,450 --> 00:39:57,950
His is the study of the victim to determine who might have killed her.

341
00:40:01,760 --> 00:40:09,590
So the classic one, I think is if the victim is a prostitute, you might that might lead you to some theory of how they died.

342
00:40:10,490 --> 00:40:18,200
So in this case, where the victim was someone who had come back from attending a rock video filming, he had a theory that,

343
00:40:19,760 --> 00:40:28,160
you know, that it must have been crime of opportunity by the only male nearby, which is Raymond Jennings, the security guard.

344
00:40:29,360 --> 00:40:39,500
However, later in the case, when the prosecution found evidence that she was in, you know, that she had been killed by gang members during a robbery,

345
00:40:40,460 --> 00:40:50,120
he then said, well, now that I've heard that seen that evidence, I no longer think that this was a sort of sex crime of opportunity.

346
00:40:50,990 --> 00:41:00,950
And so I wouldn't say that anymore. So to me, victimology, this guy was admitted as an expert witness.

347
00:41:00,950 --> 00:41:10,460
It certainly contributed to the conviction. Whether victimology is science or not is there's no reason to debate it.

348
00:41:11,000 --> 00:41:18,910
It's obviously a weak science if it is science, but it contributed to the conviction.

349
00:41:18,920 --> 00:41:28,100
It was expert evidence. Here's another one that's useful just to just to kind of show the variety of cases that we have in the registry,

350
00:41:28,100 --> 00:41:31,490
especially by not confining ourselves to DNA exonerations.

351
00:41:32,270 --> 00:41:35,960
They're not all. They're not all murder cases.

352
00:41:36,770 --> 00:41:45,230
Although non-high, level crimes still are very rare and under underrepresented in in the registry.

353
00:41:45,500 --> 00:41:49,940
Here's one that's like insurance fraud by a farmer.

354
00:41:51,320 --> 00:42:00,860
So very unusual case. And the prosecution got some expert to say, well, other the other farms were doing very well.

355
00:42:00,860 --> 00:42:10,969
So this guy's crop failure, for which he filed a claim, an allegedly false claim, was just bad farming and it's his fault.

356
00:42:10,970 --> 00:42:16,070
And so he's not entitled to this claim. So this is a very weird kind of expert.

357
00:42:16,070 --> 00:42:22,310
We hadn't really been prepared for crop experts. But, you know, they're an expert witness.

358
00:42:23,210 --> 00:42:27,080
If we found that there were a lot of crop experts, that would be interesting.

359
00:42:27,830 --> 00:42:31,400
There aren't. But why?

360
00:42:31,430 --> 00:42:42,220
Why rule it out? Why not collect data on it? The second difficult question was what about the of which actor do we focus on?

361
00:42:42,230 --> 00:42:50,800
So where does forensic science begin? What if the bad actor is not a forensic scientist, but a police officer or a prosecutor?

362
00:42:50,840 --> 00:42:54,230
But the problem involved forensic science.

363
00:42:55,400 --> 00:43:00,170
And I thought that we thought it was bad to not capture that.

364
00:43:00,710 --> 00:43:06,350
And so we decided what we were really trying to capture is a problem with the

365
00:43:06,350 --> 00:43:13,100
delivery of forensic science that doesn't require a forensic scientist necessarily.

366
00:43:13,100 --> 00:43:18,709
It doesn't really matter who is responsible if it's a problem in the delivery of forensic science.

367
00:43:18,710 --> 00:43:28,670
And I'll give you a case to illustrate this, the Kenny Waters case, it's actually a pretty, pretty well known case because Kenny Waters.

368
00:43:28,970 --> 00:43:32,209
Sister couldn't get a lawyer to exonerate him,

369
00:43:32,210 --> 00:43:40,520
so ended up putting herself through law school in order to then litigate her brother's case and get him exonerated.

370
00:43:41,600 --> 00:43:47,720
And there they are. And their son played in the movie by Hilary Swank and Sam Rockwell.

371
00:43:48,980 --> 00:43:54,470
So fairly well-known case because of because of the movie, which is called Conviction.

372
00:43:55,310 --> 00:44:03,470
But here, the police had fingerprints that excluded Kenny Waters from the crime scene.

373
00:44:03,740 --> 00:44:07,160
The police officer who had them retired, moved away,

374
00:44:07,340 --> 00:44:15,440
stored the prints and the report saying Kenny Waters was not the source of those prints in this incriminating location in a storage unit.

375
00:44:15,710 --> 00:44:28,130
And they were not found for years until that man went to law school and managed to obtain a court order to search this storage unit.

376
00:44:28,310 --> 00:44:31,320
So the fingerprint examiners didn't do anything wrong.

377
00:44:31,370 --> 00:44:38,390
They excluded him as the source and the police officer put them in a storage unit.

378
00:44:39,980 --> 00:44:46,400
And so I think this is a case this is something we need to know about.

379
00:44:47,000 --> 00:44:51,560
And imagine if we had 100 cases like this.

380
00:44:51,860 --> 00:44:58,370
Well, this is something forensic scientists should be interested in if if if our data shows that.

381
00:45:00,110 --> 00:45:06,409
You know, actually, a lot of the problems with forensic science are not with in the laboratory, with the forensic scientists themselves.

382
00:45:06,410 --> 00:45:10,790
It's because the cops are taking the evidence and putting locking it in storage units.

383
00:45:11,600 --> 00:45:18,290
They should know that and maybe fix that. And they didn't do anything wrong.

384
00:45:18,410 --> 00:45:25,040
So so in cases like this, I'm trying to imagine we only have one case like this.

385
00:45:25,040 --> 00:45:31,910
But what if we had 100 that would tell us something that might lead to sort of a policy response or an intervention.

386
00:45:33,420 --> 00:45:40,460
Another problem we dealt with that was probably the the trickiest problem is consistent with evidence.

387
00:45:41,060 --> 00:45:48,830
What do we do with evidence? That's literally true, but seems to be highly misleading with the juror.

388
00:45:48,830 --> 00:45:55,400
And the way we got to this was hair evidence where they say it's consistent with the suspect.

389
00:45:57,100 --> 00:46:00,620
It's sort of literally true. It's consistent with the suspect.

390
00:46:00,830 --> 00:46:07,400
But hair evidence is extremely unreliable. And it turns out it wasn't the suspect.

391
00:46:07,730 --> 00:46:12,830
It turns out we have a lot of this in in a lot of different areas, this kind of consistent.

392
00:46:12,830 --> 00:46:27,360
And I'll show you an example in a minute. So our answer to this is, again, to include it that we think we want to count these cases,

393
00:46:27,720 --> 00:46:36,390
but we no longer think that saying forensic science contributed to the wrongful conviction has an element of blame worthiness.

394
00:46:36,690 --> 00:46:40,080
There's a sense in which the analyst was telling the truth.

395
00:46:41,850 --> 00:46:47,009
They're just using a really lousy technique, which is hair evidence, microscopic hair comparison,

396
00:46:47,010 --> 00:46:50,670
or presumptive field drug tests, which I'll talk about in a second.

397
00:46:52,080 --> 00:46:57,180
But and so we became uncomfortable with, especially in the registry definition.

398
00:46:57,180 --> 00:47:03,389
Was the words false or misleading? I didn't want to use the word false or misleading for this consistent with testimony.

399
00:47:03,390 --> 00:47:06,720
I didn't think it was either false or even misleading.

400
00:47:07,230 --> 00:47:08,520
So here's an example.

401
00:47:08,550 --> 00:47:16,140
This another famous case maybe you know about the Steven Avery case, which got really famous because of the documentary series Making a Murderer.

402
00:47:17,310 --> 00:47:25,379
But in that original conviction, there was hair evidence and the serology is Sherry Culhane,

403
00:47:25,380 --> 00:47:32,460
who interestingly comes off kind of badly in the documentary as it's kind of biased.

404
00:47:33,840 --> 00:47:38,490
She gave like the most conservative hair testimony that I ever saw.

405
00:47:38,520 --> 00:47:43,500
She said it was similar and consistent with the victim's hair.

406
00:47:43,860 --> 00:47:46,950
And then she said, the hair. Many people are consistent with each other.

407
00:47:47,190 --> 00:47:49,890
She can't give a probability they're from the same source.

408
00:47:50,220 --> 00:47:56,520
Finally, when pressed, she says, All I'm saying is it's not impossible that the hairs are from the victim.

409
00:47:58,140 --> 00:48:04,080
Well, that's a true statement. That's a very philosophically cautious statement.

410
00:48:04,950 --> 00:48:08,940
So do I really want to fault her for saying that?

411
00:48:09,150 --> 00:48:17,010
Not really. But on the other hand, we know from the registry that and from the Innocence Project list,

412
00:48:17,010 --> 00:48:23,640
there's tons and tons of of exoneration cases where hair evidence contributed to the wrongful conviction.

413
00:48:24,000 --> 00:48:30,810
So this kind of testimony that it's consistent with, while true, leads to a lot of wrongful convictions.

414
00:48:30,840 --> 00:48:35,160
Why is that? Because I think they're using a really weak undiscriminating technique.

415
00:48:36,180 --> 00:48:46,260
And so by coating the hair cases, we are finding something out that this microscopic hair comparison is a really lousy forensic technique.

416
00:48:47,040 --> 00:48:50,250
But I don't think we're finding that people gave false testimony.

417
00:48:51,810 --> 00:48:57,420
But we we need to capture this and we need to know about it. And here's another example from the other end of the spectrum.

418
00:48:57,930 --> 00:49:03,719
This case also got a lot of publicity, was written up by ProPublica, Amy Allbright.

419
00:49:03,720 --> 00:49:07,140
And so she's she's from Louisiana.

420
00:49:07,140 --> 00:49:11,550
She's driving into Texas. The police pull her over.

421
00:49:12,270 --> 00:49:18,389
They see a white crumb on the floor mat of the car, not a baggie of white powder.

422
00:49:18,390 --> 00:49:26,250
A white crumb. And they do these field tests for drugs that the police carry in the back of their police cars,

423
00:49:26,250 --> 00:49:30,360
where they just do, you know, put it in a little baggie and shake it up and do a little colour test.

424
00:49:30,810 --> 00:49:36,420
And and it turns a certain colour, if it might be a controlled substance and it has,

425
00:49:36,420 --> 00:49:41,370
you know, an error rate of 25, 30%, something like that, very high error rate.

426
00:49:41,610 --> 00:49:46,469
But that's okay because this is just for the police use out in the field.

427
00:49:46,470 --> 00:49:53,430
And then that sample will go back to the laboratory and they'll do the better test with their proper chemical equipment,

428
00:49:53,670 --> 00:50:04,470
except that for these drug possession cases, while they're waiting for the lab, the suspect is going to plead guilty because they want to go home.

429
00:50:04,740 --> 00:50:08,520
And they're facing a very low level drug charge anyway.

430
00:50:08,520 --> 00:50:14,880
And they don't they don't want to lose their job and get evicted, which is what happened to Amy Allbright.

431
00:50:16,800 --> 00:50:24,960
And so, lo and behold, they test it in the laboratory and it's a cookie crumb, not a not a controlled substance.

432
00:50:25,920 --> 00:50:32,070
And because Houston, the D.A. of Houston, was concerned about these cases, too,

433
00:50:33,450 --> 00:50:37,589
she found out these cases were happening and decided to track all these people

434
00:50:37,590 --> 00:50:41,370
down and make sure they were exonerated and their criminal records were clear.

435
00:50:41,940 --> 00:50:47,249
We think this problem is happening all over the country with people pleading guilty,

436
00:50:47,250 --> 00:50:50,729
except that the D.A. isn't following up and making sure to clear their records.

437
00:50:50,730 --> 00:50:55,650
So they just have a misdemeanour on their on their record with all the consequences that go with that.

438
00:50:56,550 --> 00:51:06,900
But in Houston, they cleared it up. So we have 150 exonerations from Houston involving this presumptive field drug test.

439
00:51:07,200 --> 00:51:11,249
But again, I don't think the testimony if not that there was testimony.

440
00:51:11,250 --> 00:51:14,790
Right. There's always a report of a presumptive field drug test.

441
00:51:15,300 --> 00:51:20,250
I don't think it was false or misleading. It said. We did a presumptive test and it was positive.

442
00:51:20,520 --> 00:51:29,390
You need to do a proper test to confirm this. It wasn't false, but it produces a lot of wrongful convictions.

443
00:51:29,400 --> 00:51:36,900
So that's something we need to know about. So these consistent with statements, it's a big issue in forensic science.

444
00:51:37,020 --> 00:51:45,510
They're very common. And I think if we contribute, you know, if they contribute to wrongful convictions, we need to know about them.

445
00:51:45,930 --> 00:51:50,010
It's a little bit tricky for me because my in my work in forensic science,

446
00:51:50,290 --> 00:51:54,270
I have said that forensic scientists need to quantify the uncertainty of their tests.

447
00:51:54,570 --> 00:51:59,190
I've said if they don't, they need to tone down the certainty with which they testify.

448
00:51:59,700 --> 00:52:03,080
And specifically in the context of fingerprint identification.

449
00:52:03,090 --> 00:52:06,270
I've said, you know, I don't think you should say match.

450
00:52:06,270 --> 00:52:12,450
I don't think you should say it's his fingerprint. I don't think you should say identification if you can't come up with any statistics.

451
00:52:12,450 --> 00:52:20,300
Which would? Which is what I'd really like to see you do. Just say it's consistent with this so called fingerprint examiners.

452
00:52:20,310 --> 00:52:25,290
They should use this word consistent with. But we're also coding in these cases.

453
00:52:25,560 --> 00:52:31,290
And the way I look at this is if a discipline's going to can't come up with a statistical way

454
00:52:31,290 --> 00:52:35,720
of quantifying their uncertainty and they're going to rely on consistent with statements,

455
00:52:35,970 --> 00:52:43,379
you can do that, but you're going to have to own it. If you come up with a lot of wrongful convictions and if your technique is good,

456
00:52:43,380 --> 00:52:48,270
you probably won't produce a lot of wrongful convictions like, say, fingerprint evidence.

457
00:52:48,720 --> 00:52:53,430
If your technique is really weak, like microscopic hair comparison or presumptive drug test,

458
00:52:54,150 --> 00:52:58,230
you're going to produce a lot of wrongful convictions by giving this testimony.

459
00:52:59,340 --> 00:53:08,760
Setting aside again what what the jury might hear from somebody who says it's consistent with the defendant, you know,

460
00:53:09,010 --> 00:53:18,000
they're not going to take the Deep South probative value of that to be as low as it sort of actually is, if you think about what the statement makes.

461
00:53:18,840 --> 00:53:19,140
Okay.

462
00:53:19,500 --> 00:53:30,450
So I'm close to fiction, so I think I've kind of talked about this, the false or so we're not requiring that the evidence be false or misleading.

463
00:53:30,630 --> 00:53:36,510
So we kind of came up with this word problem. There's a problem with forensic evidence, but it doesn't mean anybody did something wrong.

464
00:53:36,510 --> 00:53:43,020
It doesn't mean anybody said necessarily said something false or misleading, but something is going wrong to cause wrongful convictions.

465
00:53:44,250 --> 00:53:49,229
So in some we made these choices. We broadened the definition of forensic evidence a little bit.

466
00:53:49,230 --> 00:53:55,680
We included more actors, and we focus on disembodied problems rather than blame worthiness.

467
00:53:56,490 --> 00:54:00,960
We're trying to avoid defining ontological concepts like science.

468
00:54:01,710 --> 00:54:07,170
In some sense, I think we're trying to let the data speak to us about where the problems lie

469
00:54:07,920 --> 00:54:12,570
and that they can function as an early detection system for stakeholders.

470
00:54:12,840 --> 00:54:16,500
It probably will help more cases than the old definitions did,

471
00:54:16,830 --> 00:54:23,580
but the balance of that is that all those cases are cases where we said a forensic scientist did something wrong.

472
00:54:24,400 --> 00:54:27,690
And so you can't assume that about these cases.

473
00:54:28,500 --> 00:54:31,620
So here's a summary of what we did.

474
00:54:31,890 --> 00:54:41,760
We look we started with Gerry Laporte, as I said, identify 24 cases where we were discrepant and then we found a bunch more that he didn't notice.

475
00:54:42,870 --> 00:54:51,390
So we ended up looking at 47 cases and this sort of shows where we said yes and they said no and they said no and we said yes.

476
00:54:52,110 --> 00:54:59,310
We ended up deciding 41 of them were had a forensic expert evidence problem and six did not.

477
00:55:00,900 --> 00:55:08,850
So if you look at all the DNA exonerations at that time, when we started the project two years ago,

478
00:55:10,140 --> 00:55:20,250
that ends up being 174 of the cases of the 351 DNA exonerations that existed at that time,

479
00:55:20,250 --> 00:55:26,160
about half of them ended up having a forensic or expert evidence problem.

480
00:55:26,910 --> 00:55:30,300
So it's actually slightly higher than Laporte saw.

481
00:55:30,930 --> 00:55:33,990
But again, we have this slightly expanded definition.

482
00:55:35,280 --> 00:55:48,000
And to date, while I was here on the sabbatical, the number is pretty similar, but we're up to 367 DNA exonerations, so it's about 49%.

483
00:55:49,770 --> 00:55:56,490
So why were there these discrepancies? Overwhelmingly, it's because somebody lacked information about a case.

484
00:55:56,700 --> 00:56:02,009
And once we traded information with the Innocence Project, there was no difficulty coding it.

485
00:56:02,010 --> 00:56:03,750
We just didn't know about this.

486
00:56:03,990 --> 00:56:11,670
And and most of those cases were serology cases where there was an issue with the interpretation of the serology evidence.

487
00:56:12,480 --> 00:56:17,730
12 cases were this consistent with issue, most of them hair, but some of them pathology.

488
00:56:19,020 --> 00:56:27,180
Then these sort of odd disciplines that may or may not be forensic and then this kind of broadened definition of of problem.

489
00:56:28,680 --> 00:56:36,540
So in some the portion of DNA exonerations to which forensic evidence contribute is still about half.

490
00:56:38,220 --> 00:56:42,030
That's partly with this expanded definition, but that's only about five cases.

491
00:56:42,030 --> 00:56:44,100
So that would only knock about 2% off.

492
00:56:45,210 --> 00:56:53,580
This is was true when he wrote and it's still true the portion of all exonerations, the ones in the National Registry of Exonerations is much lower.

493
00:56:54,360 --> 00:57:02,340
It's only about 24%. But we haven't applied this new definition across all these cases, so it might go up a little bit.

494
00:57:03,060 --> 00:57:07,320
That's not really surprising because of the nature of the DNA exoneration cases.

495
00:57:07,320 --> 00:57:11,790
These are cases where the person was exonerated through the use of forensic science.

496
00:57:11,790 --> 00:57:17,520
So it's not that surprising that they were originally convicted through the use of forensic science.

497
00:57:20,220 --> 00:57:24,180
And then there's a lot more to do now that we've done this project.

498
00:57:24,570 --> 00:57:31,950
There's I'm working on looking at the four different forensic disciplines which discipline contributed to this, to temporal trends,

499
00:57:31,950 --> 00:57:36,360
to how much this CO occurs with the other contributors like mistaken witness identification,

500
00:57:37,170 --> 00:57:42,630
other patterns, and also sort of subtypes of these problems in terms of what went wrong.

501
00:57:45,420 --> 00:57:58,110
So I hope your fans of the National Registry of Exonerations now follow us on social media now and we have stickers.

502
00:57:58,830 --> 00:58:03,660
If you're interested. And I'm happy to take questions or discussion.