1
00:00:00,843 --> 00:00:03,158
In 2007, I became the Attorney General

2
00:00:03,158 --> 00:00:05,159
of the State of New Jersey.

3
00:00:05,159 --> 00:00:07,439
Before that, I'd been a criminal prosecutor,

4
00:00:07,439 --> 00:00:10,120
first in the Manhattan District Attorney's office,

5
00:00:10,120 --> 00:00:12,770
and then at the United States Department of Justice.

6
00:00:12,770 --> 00:00:14,980
But when I became the Attorney General,

7
00:00:14,980 --> 00:00:15,925
two things happened

8
00:00:15,925 --> 00:00:18,866
that changed the way I see criminal justice.

9
00:00:18,866 --> 00:00:20,589
The first is that I asked what I thought

10
00:00:20,589 --> 00:00:23,082
were really basic questions.

11
00:00:23,082 --> 00:00:26,246
I wanted to understand who we were arresting,

12
00:00:26,246 --> 00:00:27,433
who we were charging,

13
00:00:27,433 --> 00:00:29,730
and who we were putting in our nation's jails

14
00:00:29,730 --> 00:00:31,146
and prisons.

15
00:00:31,146 --> 00:00:32,794
I also wanted to understand

16
00:00:32,794 --> 00:00:34,123
if we were making decisions

17
00:00:34,123 --> 00:00:36,641
in a way that made us safer.

18
00:00:36,641 --> 00:00:39,893
And I couldn't get this information out.

19
00:00:39,893 --> 00:00:43,250
It turned out that most big criminal justice agencies

20
00:00:43,250 --> 00:00:44,552
like my own

21
00:00:44,552 --> 00:00:46,934
didn't track the things that matter.

22
00:00:46,934 --> 00:00:50,252
So after about a month of being incredibly frustrated,

23
00:00:50,252 --> 00:00:52,223
I walked down into a conference room

24
00:00:52,223 --> 00:00:54,113
that was filled with detectives

25
00:00:54,113 --> 00:00:56,895
and stacks and stacks of case files,

26
00:00:56,895 --> 00:00:58,071
and the detectives were sitting there

27
00:00:58,071 --> 00:01:00,305
with yellow legal pads taking notes.

28
00:01:00,305 --> 00:01:01,845
They were trying to get the information

29
00:01:01,845 --> 00:01:03,109
I was looking for

30
00:01:03,109 --> 00:01:05,154
by going through case by case

31
00:01:05,154 --> 00:01:07,591
for the past five years.

32
00:01:07,591 --> 00:01:08,705
And as you can imagine,

33
00:01:08,705 --> 00:01:11,795
when we finally got the results, they weren't good.

34
00:01:11,795 --> 00:01:13,003
It turned out that we were doing

35
00:01:13,003 --> 00:01:15,023
a lot of low-level drug cases

36
00:01:15,023 --> 00:01:16,498
on the streets just around the corner

37
00:01:16,498 --> 00:01:18,766
from our office in Trenton.

38
00:01:18,766 --> 00:01:20,233
The second thing that happened

39
00:01:20,233 --> 00:01:23,907
is that I spent the day in the Camden,
New Jersey, Police Department.

40
00:01:23,907 --> 00:01:25,794
Now at that time, Camden, New Jersey,

41
00:01:25,794 --> 00:01:28,446
was the most dangerous city in America.

42
00:01:28,446 --> 00:01:32,273
I ran the Camden Police
Department because of that.

43
00:01:32,273 --> 00:01:34,385
I spent the day in the Police Department,

44
00:01:34,385 --> 00:01:37,111
and I was taken into a room
with senior police officials,

45
00:01:37,111 --> 00:01:38,786
all of whom were working hard

46
00:01:38,786 --> 00:01:42,320
and trying very hard to reduce crime in Camden.

47
00:01:42,320 --> 00:01:43,869
And what I saw in that room,

48
00:01:43,869 --> 00:01:46,114
as we talked about how to reduce crime,

49
00:01:46,114 --> 00:01:49,973
were a series of officers with a
lot of little yellow sticky notes.

50
00:01:49,973 --> 00:01:51,804
And they would take a yellow sticky

51
00:01:51,804 --> 00:01:52,823
and they would write something on it

52
00:01:52,823 --> 00:01:54,622
and they would put it up on a board.

53
00:01:54,622 --> 00:01:56,717
And one of them said, 
"We had a robbery two weeks ago.

54
00:01:56,717 --> 00:01:58,504
We have no suspects."

55
00:01:58,504 --> 00:02:03,531
And another said, "We had a shooting in this neighborhood last week. We have no suspects."

56
00:02:03,531 --> 00:02:06,114
We weren't using data-driven policing.

57
00:02:06,114 --> 00:02:08,156
We were essentially trying to fight crime

58
00:02:08,156 --> 00:02:10,683
with yellow post-it notes.

59
00:02:10,683 --> 00:02:12,818
Now both of these things made me realize

60
00:02:12,818 --> 00:02:16,069
fundamentally that we were failing.

61
00:02:16,069 --> 00:02:19,192
We didn't even know who was
in our criminal justice system,

62
00:02:19,192 --> 00:02:22,427
we didn't have any data about
the things that mattered,

63
00:02:22,427 --> 00:02:24,995
and we didn't share data or use analytics

64
00:02:24,995 --> 00:02:27,146
or tools to help us make better decisions

65
00:02:27,146 --> 00:02:29,149
and to reduce crime.

66
00:02:29,149 --> 00:02:31,127
And for the first time, I started to think

67
00:02:31,127 --> 00:02:33,283
about how we made decisions.

68
00:02:33,283 --> 00:02:34,680
When I was an assistant D.A.,

69
00:02:34,680 --> 00:02:36,550
and when I was a federal prosecutor,

70
00:02:36,550 --> 00:02:38,296
I looked at the cases in front of me,

71
00:02:38,296 --> 00:02:40,922
and I generally made decisions based on my instinct

72
00:02:40,922 --> 00:02:42,614
and my experience.

73
00:02:42,614 --> 00:02:44,273
When I became Attorney General,

74
00:02:44,273 --> 00:02:45,912
I could look at this system as a whole,

75
00:02:45,912 --> 00:02:47,377
and what surprised me is that I found

76
00:02:47,377 --> 00:02:49,635
that that was exactly how we were doing it

77
00:02:49,635 --> 00:02:51,938
across the entire system,

78
00:02:51,938 --> 00:02:54,339
in police departments, in prosecutors's offices,

79
00:02:54,339 --> 00:02:57,139
in courts, and in jails.

80
00:02:57,139 --> 00:02:59,336
And what I learned very quickly

81
00:02:59,336 --> 00:03:02,969
is that we weren't doing a good job.

82
00:03:02,969 --> 00:03:04,985
So I wanted to do things differently.

83
00:03:04,985 --> 00:03:07,182
I wanted to introduce data and analytics

84
00:03:07,182 --> 00:03:09,231
and rigorous statistical analysis

85
00:03:09,231 --> 00:03:10,631
into our work.

86
00:03:10,631 --> 00:03:13,601
In short, I wanted to moneyball criminal justice.

87
00:03:13,601 --> 00:03:15,628
Now, moneyball, as many of you know,

88
00:03:15,628 --> 00:03:17,197
is what the Oakland A's did,

89
00:03:17,197 --> 00:03:19,170
where they used smart data and statistics

90
00:03:19,170 --> 00:03:20,469
to figure out how to pick players

91
00:03:20,469 --> 00:03:22,313
that would help them win games,

92
00:03:22,313 --> 00:03:24,298
and they went from a system that was based

93
00:03:24,298 --> 00:03:25,386
on baseball scouts

94
00:03:25,386 --> 00:03:27,153
who used to go out and watch players

95
00:03:27,153 --> 00:03:28,714
and use their instinct and experience,

96
00:03:28,714 --> 00:03:30,533
the scouts' instincts and experience,

97
00:03:30,533 --> 00:03:32,246
to pick players, from one to use

98
00:03:32,246 --> 00:03:35,068
smart data and rigorous statistical analysis

99
00:03:35,068 --> 00:03:38,439
to figure out how to pick players
that would help them win games.

100
00:03:38,439 --> 00:03:40,237
It worked for the Oakland A's,

101
00:03:40,237 --> 00:03:42,456
and it worked in the State of New Jersey.

102
00:03:42,456 --> 00:03:44,529
We took Camden off the top of the list

103
00:03:44,529 --> 00:03:46,700
as the most dangerous city in America.

104
00:03:46,700 --> 00:03:49,855
We reduced murders there by 41 percent,

105
00:03:49,855 --> 00:03:52,837
which actually means 37 lives were saved.

106
00:03:52,837 --> 00:03:56,577
And we reduced all crime in the city by 26 percent.

107
00:03:56,577 --> 00:03:59,816
We also changed the way
we did criminal prosecutions.

108
00:03:59,816 --> 00:04:01,821
So we went from doing low-level drug crimes

109
00:04:01,821 --> 00:04:03,463
that were outside our building

110
00:04:03,463 --> 00:04:05,805
to doing cases of state-wide importance,

111
00:04:05,805 --> 00:04:08,963
on things like reducing violence
with the most violent offenders,

112
00:04:08,963 --> 00:04:10,821
prosecuting street gangs,

113
00:04:10,821 --> 00:04:14,229
gun and drug trafficking, and political corruption.

114
00:04:14,229 --> 00:04:16,731
And all of this matters greatly,

115
00:04:16,731 --> 00:04:18,676
because public safety to me

116
00:04:18,676 --> 00:04:21,212
is the most important function of government.

117
00:04:21,212 --> 00:04:23,510
If we're not safe, we can't be educated,

118
00:04:23,510 --> 00:04:24,858
we can't be healthy,

119
00:04:24,858 --> 00:04:25,738
we can't do any of the other things

120
00:04:25,738 --> 00:04:28,327
we want to do in our lives.

121
00:04:28,327 --> 00:04:29,504
And we live in a country today

122
00:04:29,504 --> 00:04:32,638
where we face serious criminal justice problems.

123
00:04:32,638 --> 00:04:36,299
We have 12 million arrests every single year.

124
00:04:36,299 --> 00:04:38,342
The vast majority of those arrests

125
00:04:38,342 --> 00:04:41,170
are for low-level crimes, like misdemeanors,

126
00:04:41,170 --> 00:04:43,088
70 to 80 percent.

127
00:04:43,088 --> 00:04:45,079
Less than five percent of all arrests

128
00:04:45,079 --> 00:04:46,974
are for violent crime.

129
00:04:46,974 --> 00:04:49,029
Yet we spend 75 billion,

130
00:04:49,029 --> 00:04:50,447
that's b for billion,

131
00:04:50,447 --> 00:04:54,574
dollars a year on state and local corrections costs.

132
00:04:54,574 --> 00:04:57,415
Right now, today, we have 2.3 million people

133
00:04:57,415 --> 00:04:59,315
in our jails and prisons.

134
00:04:59,315 --> 00:05:02,111
And we face unbelievable public safety challenges

135
00:05:02,111 --> 00:05:04,050
because we have a situation

136
00:05:04,050 --> 00:05:06,625
in which two thirds of the people in our jails

137
00:05:06,625 --> 00:05:08,702
are there waiting for trial.

138
00:05:08,702 --> 00:05:10,837
They haven't yet been convicted of a crime.

139
00:05:10,837 --> 00:05:12,956
They're just waiting for their day in court.

140
00:05:12,956 --> 00:05:16,504
And 67 percent of people come back.

141
00:05:16,504 --> 00:05:18,471
Our recidivism rate is amongst the highest

142
00:05:18,471 --> 00:05:19,526
in the world.

143
00:05:19,526 --> 00:05:21,359
Almost seven in 10 people who are released

144
00:05:21,359 --> 00:05:23,286
from prison will be rearrested

145
00:05:23,286 --> 00:05:27,241
in a constant cycle of crime and incarceration.

146
00:05:27,241 --> 00:05:29,823
So when I started my job at the Arnold Foundation,

147
00:05:29,823 --> 00:05:32,559
I came back to looking at a lot of these questions,

148
00:05:32,559 --> 00:05:34,213
and I came back to thinking about how

149
00:05:34,213 --> 00:05:36,304
we had used data and analytics to transform

150
00:05:36,304 --> 00:05:39,180
the way we did criminal justice in New Jersey.

151
00:05:39,180 --> 00:05:41,063
And when I look at the criminal justice system

152
00:05:41,063 --> 00:05:42,780
in the United States today,

153
00:05:42,780 --> 00:05:44,466
I feel the exact same way that I did

154
00:05:44,466 --> 00:05:47,085
about the State of New Jersey when I started there,

155
00:05:47,085 --> 00:05:50,313
which is that we absolutely have to do better,

156
00:05:50,313 --> 00:05:52,560
and I know that we can do better.

157
00:05:52,560 --> 00:05:53,941
So I decided to focus

158
00:05:53,941 --> 00:05:55,835
on using data and analytics

159
00:05:55,835 --> 00:05:58,519
to help make the most critical decision

160
00:05:58,519 --> 00:06:00,125
in public safety,

161
00:06:00,125 --> 00:06:02,146
and that decision is the determination

162
00:06:02,146 --> 00:06:04,681
of whether, when someone has been arrested,

163
00:06:04,681 --> 00:06:06,596
whether they pose a risk to public safety

164
00:06:06,596 --> 00:06:08,122
and should be detained,

165
00:06:08,122 --> 00:06:10,478
or whether they don't pose a risk to public safety

166
00:06:10,478 --> 00:06:12,115
and should be released.

167
00:06:12,115 --> 00:06:13,881
Everything that happens in criminal cases

168
00:06:13,881 --> 00:06:15,806
comes out of this one decision.

169
00:06:15,806 --> 00:06:17,302
It impacts everything.

170
00:06:17,302 --> 00:06:18,652
It impacts sentencing.

171
00:06:18,652 --> 00:06:20,677
It impacts whether someone gets drug treatment.

172
00:06:20,677 --> 00:06:22,876
It impacts crime and violence.

173
00:06:22,876 --> 00:06:24,813
And when I talk to judges around the United States,

174
00:06:24,813 --> 00:06:26,772
which I do all the time now,

175
00:06:26,772 --> 00:06:28,578
they all say the same thing,

176
00:06:28,578 --> 00:06:31,685
which is that we put dangerous people in jail,

177
00:06:31,685 --> 00:06:35,210
and we let non-dangerous, non-violent people out.

178
00:06:35,210 --> 00:06:37,443
They mean it and they believe it.

179
00:06:37,443 --> 00:06:39,176
But when you start to look at the data,

180
00:06:39,176 --> 00:06:41,640
which, by the way, the judges don't have,

181
00:06:41,640 --> 00:06:43,252
when we start to look at the data,

182
00:06:43,252 --> 00:06:45,486
what we find time and time again,

183
00:06:45,486 --> 00:06:47,652
is that this isn't the case.

184
00:06:47,652 --> 00:06:49,333
We find low-risk offenders,

185
00:06:49,333 --> 00:06:53,047
which makes up 50 percent of our
entire criminal justice population,

186
00:06:53,047 --> 00:06:55,446
we find that they're in jail.

187
00:06:55,446 --> 00:06:57,932
Take Leslie Chew, who was a Texas man

188
00:06:57,932 --> 00:07:00,816
who stole four blankets on a cold winter night.

189
00:07:00,816 --> 00:07:03,411
He was arrested, and he was kept in jail

190
00:07:03,411 --> 00:07:05,464
on 3,500 dollars bail,

191
00:07:05,464 --> 00:07:08,240
an amount that he could not afford to pay.

192
00:07:08,240 --> 00:07:10,828
And he stayed in jail for eight months

193
00:07:10,828 --> 00:07:12,893
until his case came up for trial,

194
00:07:12,893 --> 00:07:16,798
at a cost to taxpayers of more than 9,000 dollars.

195
00:07:16,798 --> 00:07:18,626
And at the other end of the spectrum,

196
00:07:18,626 --> 00:07:21,077
we're doing an equally terrible job.

197
00:07:21,077 --> 00:07:22,649
The people who we find

198
00:07:22,649 --> 00:07:24,668
are the highest risk offenders,

199
00:07:24,668 --> 00:07:26,689
the people who we think have the highest likelihood

200
00:07:26,689 --> 00:07:29,117
of committing a new crime if they're released,

201
00:07:29,117 --> 00:07:31,929
we see nationally that 50 percent of those people

202
00:07:31,929 --> 00:07:34,041
are being released.

203
00:07:34,041 --> 00:07:37,215
The reason for this is the way we make decisions.

204
00:07:37,215 --> 00:07:39,740
Judges have the best intentions

205
00:07:39,740 --> 00:07:41,970
when they make these decisions about risk,

206
00:07:41,970 --> 00:07:43,360
but they're making them subjectively.

207
00:07:43,360 --> 00:07:45,583
They're like the baseball scouts 20 years ago

208
00:07:45,583 --> 00:07:47,637
who were using their instinct and their experience

209
00:07:47,637 --> 00:07:50,871
to try to decide what risk someone poses.

210
00:07:50,871 --> 00:07:52,125
They're being subjective,

211
00:07:52,125 --> 00:07:54,937
and we know what happens
with subjective decision-making,

212
00:07:54,937 --> 00:07:57,266
which is that we are often wrong.

213
00:07:57,266 --> 00:07:59,202
What we need in this space

214
00:07:59,202 --> 00:08:01,800
are strong data and analytics.

215
00:08:01,800 --> 00:08:03,331
What I decided to look for

216
00:08:03,331 --> 00:08:06,321
was a strong data and analytic risk assessment tool,

217
00:08:06,321 --> 00:08:08,870
something that would let judges actually understand

218
00:08:08,870 --> 00:08:11,314
with a scientific and objective way

219
00:08:11,314 --> 00:08:12,930
what the risk was that was posed

220
00:08:12,930 --> 00:08:14,663
by someone in front of them.

221
00:08:14,663 --> 00:08:16,266
I looked all over the country,

222
00:08:16,266 --> 00:08:18,470
and I found that between five and 10 percent

223
00:08:18,470 --> 00:08:19,537
of all U.S. jurisdictions

224
00:08:19,537 --> 00:08:22,393
actually use any type of risk assessment tool,

225
00:08:22,393 --> 00:08:23,940
and when I looked at these tools,

226
00:08:23,940 --> 00:08:25,892
I quickly realized why.

227
00:08:25,892 --> 00:08:28,367
They were unbelievably expensive to administer,

228
00:08:28,367 --> 00:08:30,110
they were time-consuming,

229
00:08:30,110 --> 00:08:32,680
they were limited to the local jurisdiction

230
00:08:32,680 --> 00:08:33,739
in which they'd been created.

231
00:08:33,739 --> 00:08:35,287
So basically, they couldn't be scaled

232
00:08:35,287 --> 00:08:37,726
or transferred to other places.

233
00:08:37,726 --> 00:08:39,901
So I went out and build a phenomenal team

234
00:08:39,901 --> 00:08:42,900
of data scientists and researchers

235
00:08:42,900 --> 00:08:43,587
and statisticians

236
00:08:43,587 --> 00:08:46,339
to build a universal risk assessment tool,

237
00:08:46,339 --> 00:08:48,732
so that every single judge in
the United States of America

238
00:08:48,732 --> 00:08:53,380
can have an objective, scientific measure of risk.

239
00:08:53,380 --> 00:08:54,714
In the tool that we've built,

240
00:08:54,714 --> 00:08:57,382
what we did was we collected 1.5 million cases

241
00:08:57,382 --> 00:08:59,465
from all around the United States,

242
00:08:59,465 --> 00:09:01,710
from cities, from counties,

243
00:09:01,710 --> 00:09:02,635
from every single state in the country,

244
00:09:02,635 --> 00:09:04,320
the federal districts.

245
00:09:04,320 --> 00:09:06,370
And with those 1.5 million cases,

246
00:09:06,370 --> 00:09:08,310
which is the largest data set on pretrial

247
00:09:08,310 --> 00:09:10,870
in the United States today,

248
00:09:10,870 --> 00:09:12,273
we were able to basically find that there were

249
00:09:12,273 --> 00:09:15,302
900-plus risk factors that we could look at

250
00:09:15,302 --> 00:09:18,800
to try to figure out what mattered most.

251
00:09:18,800 --> 00:09:20,466
And we found that there were nine specific things

252
00:09:20,466 --> 00:09:22,484
that mattered all across the country

253
00:09:22,484 --> 00:09:25,478
and that were the most highly predictive of risk.

254
00:09:25,478 --> 00:09:28,982
And so we built a universal risk assessment tool.

255
00:09:28,982 --> 00:09:30,827
And it looks like this.

256
00:09:30,827 --> 00:09:33,223
As you'll see, we put some information in,

257
00:09:33,223 --> 00:09:35,390
but most of it is incredibly simple,

258
00:09:35,390 --> 00:09:36,992
it's easy to use,

259
00:09:36,992 --> 00:09:39,761
it focuses on things like the
defendant's prior convictions,

260
00:09:39,761 --> 00:09:41,786
whether they've been sentenced to incarceration,

261
00:09:41,786 --> 00:09:43,957
whether they've engaged in violence before,

262
00:09:43,957 --> 00:09:46,397
whether they've even failed to come back to court.

263
00:09:46,397 --> 00:09:48,912
And with this tool, we can predict three things.

264
00:09:48,912 --> 00:09:50,596
First, whether or not someone will commit

265
00:09:50,596 --> 00:09:52,315
a new crime if they're released.

266
00:09:52,315 --> 00:09:53,979
Second, for the first time,

267
00:09:53,979 --> 00:09:55,840
and I think this is incredibly important,

268
00:09:55,840 --> 00:09:57,517
we can predict whether someone will commit

269
00:09:57,517 --> 00:09:59,689
an act of violence if they're released.

270
00:09:59,689 --> 00:10:01,391
And that's the single most important thing

271
00:10:01,391 --> 00:10:03,244
that judges say when you talk to them.

272
00:10:03,244 --> 00:10:04,889
And third, we can predict whether someone

273
00:10:04,889 --> 00:10:06,802
will come back to court.

274
00:10:06,802 --> 00:10:10,680
And every single judge in the
United States of America can use it,

275
00:10:10,680 --> 00:10:13,554
because it's been created on a universal data set.

276
00:10:13,554 --> 00:10:15,994
What judges see if they run the risk assessment tool

277
00:10:15,994 --> 00:10:18,559
is this: it's a dashboard.

278
00:10:18,559 --> 00:10:21,500
At the top, you see the new criminal activity score,

279
00:10:21,500 --> 00:10:23,860
six of course being the highest,

280
00:10:23,860 --> 00:10:26,970
and then in the middle you
see "elevated risk of violence."

281
00:10:26,970 --> 00:10:27,793
What that says is that this person

282
00:10:27,793 --> 00:10:29,655
is someone who has an elevated risk of violence

283
00:10:29,655 --> 00:10:31,584
that the judge should look twice at.

284
00:10:31,584 --> 00:10:32,813
And then, towards the bottom,

285
00:10:32,813 --> 00:10:34,734
you see the "Failure to Appear" score,

286
00:10:34,734 --> 00:10:36,850
which again is the likelihood

287
00:10:36,850 --> 00:10:38,787
that someone will come back to court.

288
00:10:38,787 --> 00:10:41,537
Now I want to say something really important.

289
00:10:41,537 --> 00:10:43,865
It's not that I think we should be eliminating

290
00:10:43,865 --> 00:10:46,416
the judge's instinct and experience

291
00:10:46,416 --> 00:10:48,112
from this process.

292
00:10:48,112 --> 00:10:49,294
I don't.

293
00:10:49,294 --> 00:10:51,900
I actually believe the problem that we see

294
00:10:51,900 --> 00:10:53,940
and the reason that we have
these incredible system errors,

295
00:10:53,940 --> 00:10:56,933
where we're incarcerating
low-level, nonviolent people

296
00:10:56,933 --> 00:11:00,229
and we're releasing high-risk, dangerous people,

297
00:11:00,229 --> 00:11:02,998
is that we don't have an objective measure of risk.

298
00:11:02,998 --> 00:11:04,313
But what I believe should happen

299
00:11:04,313 --> 00:11:06,867
is that we should take that
data-driven risk assessment

300
00:11:06,867 --> 00:11:09,908
and combine that with the
judge's instinct and experience

301
00:11:09,908 --> 00:11:12,897
to lead us to better decision-making.

302
00:11:12,897 --> 00:11:16,415
The tool went state-wide in Kentucky on July 1st,

303
00:11:16,415 --> 00:11:19,720
and we're about to go up in a
number of other U.S. jurisdictions.

304
00:11:19,720 --> 00:11:22,220
Our goal, quite simply, is that every single judge

305
00:11:22,220 --> 00:11:24,472
in the United States will use a data-driven risk tool

306
00:11:24,472 --> 00:11:26,548
within the next five years.

307
00:11:26,548 --> 00:11:28,900
We're now working on risk tools

308
00:11:28,900 --> 00:11:31,230
for prosecutors and for police officers, as well,

309
00:11:31,230 --> 00:11:33,762
to try to take a system that runs today

310
00:11:33,762 --> 00:11:36,772
in America the same way it did 50 years ago,

311
00:11:36,772 --> 00:11:38,762
based on instinct and experience,

312
00:11:38,762 --> 00:11:40,586
and make it into one that runs

313
00:11:40,586 --> 00:11:43,286
on data and analytics.

314
00:11:43,286 --> 00:11:45,930
Now, the great news about all this,

315
00:11:45,930 --> 00:11:46,762
and we have a ton of work left to do,

316
00:11:46,762 --> 00:11:48,527
and we have a lot of culture to change,

317
00:11:48,527 --> 00:11:50,366
but the great news about all of it

318
00:11:50,366 --> 00:11:52,710
is that we know it works.

319
00:11:52,710 --> 00:11:54,740
It's why Google is Google,

320
00:11:54,740 --> 00:11:56,802
and it's why all these baseball teams use moneyball

321
00:11:56,802 --> 00:11:58,598
to win games.

322
00:11:58,598 --> 00:12:00,350
The great news for us as well

323
00:12:00,350 --> 00:12:02,123
is that it's the way that we can transform

324
00:12:02,123 --> 00:12:04,492
the American criminal justice system.

325
00:12:04,492 --> 00:12:06,924
It's how we can make our streets safer,

326
00:12:06,924 --> 00:12:09,470
we can reduce our prison costs,

327
00:12:09,470 --> 00:12:11,290
and we can make our system much fairer

328
00:12:11,290 --> 00:12:12,969
and more just.

329
00:12:12,969 --> 00:12:14,870
Some people call it data science.

330
00:12:14,870 --> 00:12:17,356
I call it moneyballing criminal justice.

331
00:12:17,356 --> 00:12:19,175
Thank you.

332
00:12:19,175 --> 00:12:23,175
(Applause)