Fill worksheet

Subscribe to the Series

ON A SPRING AFTERNOON IN 2014, Brisha Borden was running late to pick up her god-sister from school when she spotted an unlocked kid’s blue Hu�y bicycle and a silver Razor scooter. Borden and a friend grabbed the bike and scooter and tried to ride them down the street in the Fort Lauderdale suburb of Coral Springs.

Just as the 18-year-old girls were realizing they were too big for the tiny conveyances — which belonged to a 6-year-old boy — a woman came running after them saying, “That’s my kid’s stu�.” Borden and her friend immediately dropped the bike and scooter and walked away.

But it was too late — a neighbor who witnessed the heist had already called the police. Borden and her friend were arrested and charged with burglary and petty theft for the items, which were valued at a total of $80.

Compare their crime with a similar one: The previous summer, 41-year-old Vernon Prater

Machine Bias There’s software used across the country to predict future criminals. And it’s biased

against blacks.

by Julia Angwin, Je� Larson, Surya Mattu and Lauren Kirchner, ProPublica May 23, 2016

Donate

Machine Bias: Investigating the algorithms that control our lives.

[email protected]

Subscribe

Read the Documents Northpointe document collection

Sentencing reports that include risk assessments

Get the Data Read about how we analyzed the risk

assessments algorithm

Download the full data used in our analysis

was picked up for shoplifting $86.35 worth of tools from a nearby Home Depot store.

Prater was the more seasoned criminal. He had already been convicted of armed robbery and attempted armed robbery, for which he served �ve years in prison, in addition to another armed robbery charge. Borden had a record, too, but it was for misdemeanors committed when she was a juvenile.

Yet something odd happened when Borden and Prater were booked into jail: A computer program spat out a score predicting the likelihood of each committing a future crime. Borden — who is black — was rated a high risk. Prater — who is white — was rated a low risk.

Two years later, we know the computer algorithm got it exactly backward. Borden has not been charged with any new crimes. Prater is serving an eight-year prison term for subsequently breaking into a warehouse and stealing thousands of dollars’ worth of electronics.

Scores like this — known as risk assessments — are increasingly common in courtrooms across the nation. They are used to inform decisions about who can be set free at every stage of the criminal justice system, from assigning bond amounts — as is the case in Fort Lauderdale — to even more fundamental decisions about defendants’ freedom. In Arizona, Colorado, Delaware, Kentucky, Louisiana, Oklahoma, Virginia, Washington and Wisconsin, the results of such assessments are given to judges during criminal sentencing.

Rating a defendant’s risk of future crime is often done in conjunction with an evaluation of a defendant’s rehabilitation needs. The Justice Department’s National Institute of Corrections now encourages the use of such combined assessments at every stage of the criminal justice process. And a landmark sentencing reform bill currently pending in Congress would mandate the use of such assessments in federal prisons.

In 2014, then U.S. Attorney General Eric Holder warned that the risk scores might be injecting bias into the courts. He called for the U.S. Sentencing Commission to study their use. “Although these measures were crafted with

📄

📄

📓



Two Petty Theft Arrests

Donate

the best of intentions, I am concerned that they inadvertently undermine our e�orts to ensure individualized and equal justice,” he said, adding, “they may exacerbate unwarranted and unjust disparities that are already far too common in our criminal justice system and in our society.”

The sentencing commission did not, however, launch a study of risk scores. So ProPublica did, as part of a larger examination of the powerful, largely hidden e�ect of algorithms in American life.

We obtained the risk scores assigned to more than 7,000 people arrested in Broward County,

Florida, in 2013 and 2014 and checked to see how many were charged with new crimes over the next two years, the same benchmark used by the creators of the algorithm.

The score proved remarkably unreliable in forecasting violent crime: Only 20 percent of the people predicted to commit violent crimes actually went on to do so.

When a full range of crimes were taken into account — including misdemeanors such as driving with an expired license — the algorithm was somewhat more accurate than a coin �ip. Of those deemed likely to re-o�end, 61 percent were arrested for any subsequent crimes within two years.

We also turned up signi�cant racial disparities, just as Holder feared. In forecasting who would re-o�end, the algorithm made mistakes with black and white defendants at roughly the same rate but in very di�erent ways.

The formula was particularly likely to falsely �ag black defendants as future criminals, wrongly labeling them this way at almost twice the rate as white defendants.

White defendants were mislabeled as low risk more often than black defendants.

Could this disparity be explained by defendants’ prior crimes or the type of crimes they were arrested for? No. We ran a statistical test that isolated the e�ect of race from criminal history and recidivism, as well as from defendants’ age and gender. Black defendants were still 77

Borden was rated high risk for future crime after she and a friend took a kid’s bike and scooter that were sitting outside. She did not reo�end.

V E R N O N P R A T E RV E R N O N P R A T E R

R I S K : 3

B R I S H A B O R D E NB R I S H A B O R D E N

R I S K : 8

Donate

percent more likely to be pegged as at higher risk of committing a future violent crime and 45 percent more likely to be predicted to commit a future crime of any kind. (Read our analysis.)

The algorithm used to create the Florida risk scores is a product of a for-pro�t company, Northpointe. The company disputes our analysis.

In a letter, it criticized ProPublica’s methodology and defended the accuracy of its test: “Northpointe does not agree that the results of your analysis, or the claims being made based upon that analysis, are correct or that they accurately re�ect the outcomes from the application of the model.”

Northpointe’s software is among the most widely used assessment tools in the country. The company does not publicly disclose the calculations used to arrive at defendants’ risk scores, so it is not possible for either defendants or the public to see what might be driving the disparity. (On Sunday, Northpointe gave ProPublica the basics of its future-crime formula — which includes factors such as education levels, and whether a defendant has a job. It did not share the speci�c calculations, which it said are proprietary.)

Northpointe’s core product is a set of scores derived from 137 questions that are either answered by defendants or pulled from criminal records. Race is not one of the questions. The survey asks defendants such things as: “Was one of your parents ever sent to jail or prison?” “How many of your friends/acquaintances are taking drugs illegally?” and “How often did you get in �ghts while at school?” The questionnaire also asks people to agree or disagree with statements such as “A hungry person has a right to steal” and “If people make me angry or lose my temper, I can be dangerous.”

The appeal of risk scores is obvious: The United States locks up far more people than any other country, a disproportionate number of them black. For more than two centuries, the key decisions in the legal process, from pretrial release to sentencing to parole, have been in the hands of human beings guided by their instincts and personal biases.

If computers could accurately predict which defendants were likely to commit new crimes, the criminal justice system could be fairer and more selective about who is incarcerated and for how long. The trick, of course, is to make sure the computer gets it right. If it’s wrong in one direction, a dangerous criminal could go free. If it’s wrong in another direction, it could result in someone unfairly receiving a harsher sentence or waiting longer for parole than is appropriate.

The �rst time Paul Zilly heard of his score — and realized how much was riding on it — was during his sentencing hearing on Feb. 15, 2013, in court in Barron County, Wisconsin. Zilly had been convicted of stealing a push lawnmower and some tools. The prosecutor recommended a

Donate

year in county jail and follow-up supervision that could help Zilly with “staying on the right path.” His lawyer agreed to a plea deal.

But Judge James Babler had seen Zilly’s scores. Northpointe’s software had rated Zilly as a high risk for future violent crime and a medium risk for general recidivism. “When I look at the risk assessment,” Babler said in court, “it is about as bad as it could be.”

Then Babler overturned the plea deal that had been agreed on by the prosecution and defense and imposed two years in state prison and three years of supervision.

CRIMINOLOGISTS HAVE LONG TRIED to predict which criminals are more dangerous before deciding whether they should be released. Race, nationality and skin color were often used in making such predictions until about the 1970s, when it became politically unacceptable, according to a survey of risk assessment tools by Columbia University law professor Bernard Harcourt.

In the 1980s, as a crime wave engulfed the nation, lawmakers made it much harder for judges and parole boards to exercise discretion in making such decisions. States and the federal government began instituting mandatory sentences and, in some cases, abolished parole, making it less important to evaluate individual o�enders.

But as states struggle to pay for swelling prison and jail populations, forecasting criminal risk has made a comeback.

Dozens of risk assessments are being used across the nation — some created by for- pro�t companies such as Northpointe and others by nonpro�t organizations. (One tool being used in states including Kentucky and Arizona, called the Public Safety Assessment, was developed by the Laura and John Arnold Foundation, which also is a funder of ProPublica.)

There have been few independent studies of these criminal risk assessments. In 2013, researchers Sarah Desmarais and Jay Singh examined 19 di�erent risk methodologies used in the United States and found that “in

Two Drug Possession Arrests

Fugett was rated low risk after being arrested with cocaine and marijuana. He was arrested three times on drug charges after

D Y L A N F U G E T TD Y L A N F U G E T T

R I S K : 3

B E R N A R D P A R K E RB E R N A R D P A R K E R

R I S K : 1 0

Donate

most cases, validity had only been examined in one or two studies” and that “frequently, those investigations were completed by the same people who developed the instrument.”

Their analysis of the research through 2012 found that the tools “were moderate at best in terms of predictive validity,” Desmarais said in an interview. And she could not �nd any substantial set of studies conducted in the United States that examined whether risk scores were racially biased. “The data do not exist,” she said.

Since then, there have been some attempts to explore racial disparities in risk scores. One 2016 study examined the validity of a risk assessment tool, not Northpointe’s, used to make probation decisions for about 35,000 federal convicts. The researchers, Jennifer Skeem at University of California, Berkeley, and Christopher T. Lowenkamp from the Administrative O�ce of the U.S. Courts, found that blacks did get a higher average score but concluded the di�erences were not attributable to bias.

The increasing use of risk scores is controversial and has garnered media coverage, including articles by the Associated Press, and the Marshall Project and FiveThirtyEight last year.

Most modern risk tools were originally designed to provide judges with insight into the types of treatment that an individual might need — from drug treatment to mental health counseling.

“What it tells the judge is that if I put you on probation, I’m going to need to give you a lot of services or you’re probably going to fail,” said Edward Latessa, a University of Cincinnati professor who is the author of a risk assessment tool that is used in Ohio and several other states.

But being judged ineligible for alternative treatment — particularly during a sentencing hearing — can translate into incarceration. Defendants rarely have an opportunity to challenge their assessments. The results are usually shared with the defendant’s attorney, but the calculations that transformed the underlying data into a score are rarely revealed.

“Risk assessments should be impermissible unless both parties get to see all the data that go into them,” said Christopher Slobogin, director of the criminal justice program at Vanderbilt Law School. “It should be an open, full-court adversarial proceeding.”

that.

Black Defendants’ Risk Scores

Donate

Proponents of risk scores argue they can be used to reduce the rate of incarceration. In 2002, Virginia became one of the �rst states to begin using a risk assessment tool in the sentencing of nonviolent felony o�enders statewide. In 2014, Virginia judges using the tool sent nearly half of those defendants to alternatives to prison, according to a state sentencing commission

C ou

nt

Risk Score

1

100

0

200

300

400

500

600

2 3 4 5 6 7 8 9 10

White Defendants’ Risk Scores

1

100

0

200

300

400

C ou

nt

Risk Score

500

600

2 3 4 5 6 7 8 9 10

These charts show that scores for white defendants were skewed toward lower-risk categories. Scores for black defendants were not. (Source: ProPublica analysis of data from Broward County, Fla.)

Donate

“I’m surprised [my risk score] is so low. I spent �ve years in state prison in Massachusetts.” ( Josh Ritchie for ProPublica)

report. Since 2005, the state’s prison population growth has slowed to 5 percent from a rate of 31 percent the previous decade.

In some jurisdictions, such as Napa County, California, the probation department uses risk assessments to suggest to the judge an appropriate probation or treatment plan for individuals being sentenced. Napa County Superior Court Judge Mark Boessenecker said he �nds the recommendations helpful. “We have a dearth of good treatment programs, so �lling a slot in a program with someone who doesn’t need it is foolish,” he said.

However, Boessenecker, who trains other judges around the state in evidence-based sentencing, cautions his colleagues that the score doesn’t necessarily reveal whether a person is dangerous or if they should go to prison.

“A guy who has molested a small child every day for a year could still come out as a low risk because he probably has a job,” Boessenecker said. “Meanwhile, a drunk guy will look high risk because he’s homeless. These risk factors don’t tell you whether the guy ought to go to prison or not; the risk factors tell you more about what the probation conditions ought to be.”

Sometimes, the scores make little sense even to defendants.

James Rivelli, a 54-year old Hollywood, Florida, man, was arrested two years ago for shoplifting seven boxes of Crest Whitestrips from a CVS drugstore. Despite a criminal record that included aggravated assault, multiple thefts and felony drug tra�cking, the Northpointe algorithm classi�ed him as being at a low risk of reo�ending.

“I am surprised it is so low,” Rivelli said when told by a reporter he had been rated a 3 out of a possible 10. “I spent �ve years in state prison in Massachusetts. But I guess they don’t count that here in Broward County.” In fact, criminal records from across the nation are supposed to be included in risk assessments.

Less than a year later, he was charged with two felony counts for shoplifting about $1,000 worth of tools from Home Depot. He said his

Donate

crimes were fueled by drug addiction and that he is now sober.

NORTHPOINTE WAS FOUNDED in 1989 by Tim Brennan, then a professor of statistics at the University of Colorado, and Dave Wells, who was running a corrections program in Traverse City, Michigan.

Wells had built a prisoner classi�cation system for his jail. “It was a beautiful piece of work,” Brennan said in an interview conducted before ProPublica had completed its analysis. Brennan and Wells shared a love for what Brennan called “quantitative taxonomy” — the measurement of personality traits such as intelligence, extroversion and introversion. The two decided to build a risk assessment score for the corrections industry.

Brennan wanted to improve on a leading risk assessment score, the LSI, or Level of Service Inventory, which had been developed in Canada. “I found a fair amount of weakness in the LSI,” Brennan said. He wanted a tool that addressed the major theories about the causes of crime.

Brennan and Wells named their product the Correctional O�ender Management Pro�ling for Alternative Sanctions, or COMPAS. It assesses not just risk but also nearly two dozen so-called “criminogenic needs” that relate to the major theories of criminality, including “criminal personality,” “social isolation,” “substance abuse” and “residence/stability.” Defendants are ranked low, medium or high risk in each category.

As often happens with risk assessment tools, many jurisdictions have adopted Northpointe’s software before rigorously testing whether it works. New York State, for instance, started using the tool to assess people on probation in a pilot project in 2001 and rolled it out to the rest of the state’s probation departments — except New York City — by 2010. The state didn’t publish a comprehensive statistical evaluation of the tool until 2012. The study of more than 16,000 probationers found the tool was 71 percent accurate, but it did not evaluate racial di�erences.

A spokeswoman for the New York state division of criminal justice services said the study did

Two DUI Arrests

Lugo crashed his Lincoln Navigator into a Toyota Camry while drunk. He was rated as a low risk of reo�ending despite the fact

G R E G O R Y L U G OG R E G O R Y L U G O

R I S K : 1

M A L L O R Y W I L L I A M SM A L L O R Y W I L L I A M S

R I S K : 6

Donate

not examine race because it only sought to test whether the tool had been properly calibrated

to �t New York’s probation population. She also said judges in nearly all New York counties are given defendants’ Northpointe assessments during sentencing.

In 2009, Brennan and two colleagues published a validation study that found that Northpointe’s risk of recidivism score had an accuracy rate of 68 percent in a sample of 2,328 people. Their study also found that the score was slightly less predictive for black men than white men — 67 percent versus 69 percent. It did not examine racial disparities beyond that, including whether some groups were more likely to be wrongly labeled higher risk.

Brennan said it is di�cult to construct a score that doesn’t include items that can be correlated with race — such as poverty, joblessness and social marginalization. “If those are omitted from your risk assessment, accuracy goes down,” he said.

In 2011, Brennan and Wells sold Northpointe to Toronto-based conglomerate Constellation Software for an undisclosed sum.

Wisconsin has been among the most eager and expansive users of Northpointe’s risk assessment tool in sentencing decisions. In 2012, the Wisconsin Department of Corrections launched the use of the software throughout the state. It is used at each step in the prison system, from sentencing to parole.

In a 2012 presentation, corrections o�cial Jared Hoy described the system as a “giant correctional pinball machine” in which correctional o�cers could use the scores at every “decision point.”

Wisconsin has not yet completed a statistical validation study of the tool and has not said when one might be released. State corrections o�cials declined repeated requests to comment for this article.

Some Wisconsin counties use other risk assessment tools at arrest to determine if a defendant is too risky for pretrial release. Once a defendant is convicted of a felony anywhere in the state, the Department of Corrections attaches Northpointe’s assessment to the con�dential presentence report given to judges, according to Hoy’s presentation.

In theory, judges are not supposed to give longer sentences to defendants with higher risk scores. Rather, they are supposed to use the tests primarily to determine which defendants are eligible for probation or treatment programs.

Prediction Fails Differently for Black Defendants

that it was at least his fourth DUI.

Donate

W H I T E A F R I C A N A M E R I C A N

Labeled Higher Risk, But Didn’t Re-O�end 23.5% 44.9%

Labeled Lower Risk, Yet Did Re-O�end 47.7% 28.0%

Overall, Northpointe’s assessment tool correctly predicts recidivism 61 percent of the time. But blacks are almost twice as likely as whites to be labeled a higher risk but not actually re-o�end. It makes the opposite mistake among whites: They are much more likely than blacks to be labeled lower risk but go on to commit other crimes. (Source: ProPublica analysis of data from Broward County, Fla.)

But judges have cited scores in their sentencing decisions. In August 2013, Judge Scott Horne in La Crosse County, Wisconsin, declared that defendant Eric Loomis had been “identi�ed, through the COMPAS assessment, as an individual who is at high risk to the community.” The judge then imposed a sentence of eight years and six months in prison.

Loomis, who was charged with driving a stolen vehicle and �eeing from police, is challenging the use of the score at sentencing as a violation of his due process rights. The state has defended Horne’s use of the score with the argument that judges can consider the score in addition to other factors. It has also stopped including scores in presentencing reports until the state Supreme Court decides the case.

“The risk score alone should not determine the sentence of an o�ender,” Wisconsin Assistant Attorney General Christine Remington said last month during state Supreme Court arguments in the Loomis case. “We don’t want courts to say, this person in front of me is a 10 on COMPAS as far as risk, and therefore I’m going to give him the maximum sentence.”

That is almost exactly what happened to Zilly, the 48-year-old construction worker sent to prison for stealing a push lawnmower and some tools he intended to sell for parts. Zilly has long struggled with a meth habit. In 2012, he had been working toward recovery with the help of a Christian pastor when he relapsed and committed the thefts.

After Zilly was scored as a high risk for violent recidivism and sent to prison, a public defender appealed the sentence and called the score’s creator, Brennan, as a witness.

Brennan testi�ed that he didn’t design his software to be used in sentencing. “I wanted to stay away from the courts,” Brennan said, explaining that his focus was on reducing crime rather than punishment. “But as time went on I started realizing that so many decisions are made, you know, in the courts. So I gradually softened on whether this could be used in the courts or not.”

Donate

“Not that I’m innocent, but I just believe people do change.” (Stephen Maturen for ProPublica)

Still, Brennan testi�ed, “I don’t like the idea myself of COMPAS being the sole evidence that a decision would be based upon.”

After Brennan’s testimony, Judge Babler reduced Zilly’s sentence, from two years in prison to 18 months. “Had I not had the COMPAS, I believe it would likely be that I would have given one year, six months,” the judge said at an appeals hearing on Nov. 14, 2013.

Zilly said the score didn’t take into account all the changes he was making in his life — his conversion to Christianity, his struggle to quit using drugs and his e�orts to be more available for his son. “Not that I’m innocent, but I just believe people do change.”

FLORIDA’S BROWARD COUNTY, where Brisha Borden stole the Hu�y bike and was scored as high risk, does not use risk assessments in sentencing. “We don’t think the [risk assessment] factors have any bearing on a sentence,” said David Scharf, executive director of community programs for the Broward County Sheri� ’s O�ce in Fort Lauderdale.

Donate

Broward County has, however, adopted the score in pretrial hearings, in the hope of addressing jail overcrowding. A court-appointed monitor has overseen Broward County’s jails since 1994 as a result of the settlement of a lawsuit brought by inmates in the 1970s. Even now, years later, the Broward County jail system is often more than 85 percent full, Scharf said.

In 2008, the sheri� ’s o�ce decided that instead of building another jail, it would begin using Northpointe’s risk scores to help identify which defendants were low risk enough to be released on bail pending trial. Since then, nearly everyone arrested in Broward has been scored soon after being booked. (People charged with murder and other capital crimes are not scored because they are not eligible for pretrial release.)

The scores are provided to the judges who decide which defendants can be released from jail. “My feeling is that if they don’t need them to be in jail, let’s get them out of there,” Scharf said.

Scharf said the county chose Northpointe’s software over other tools because it was easy to use and produced “simple yet e�ective charts and graphs for judicial review.” He said the system costs about $22,000 a year.

In 2010, researchers at Florida State University examined the use of Northpointe’s system in Broward County over a 12-month period and concluded that its predictive accuracy was “equivalent” in assessing defendants of di�erent races. Like others, they did not examine whether di�erent races were classi�ed di�erently as low or high risk.

Scharf said the county would review ProPublica’s �ndings. “We’ll really look at them up close,” he said.

Broward County Judge John Hurley, who oversees most of the pretrial release hearings, said the scores were helpful when he was a new judge, but now that he has experience he prefers to rely on his own judgment. “I haven’t relied on COMPAS in a couple years,” he said.

Hurley said he relies on factors including a person’s prior criminal record, the type of crime committed, ties to the community, and their history of failing to appear at court proceedings.

Two Shoplifting Arrests

After Rivelli stole from a CVS and was caught with heroin in his car, he was rated a low risk. He later shoplifted $1,000 worth of tools from a Home Depot.

J A M E S R I V E L L IJ A M E S R I V E L L I

R I S K : 3

R O B E R T C A N N O NR O B E R T C A N N O N

R I S K : 6

Donate

ProPublica’s analysis reveals that higher Northpointe scores are slightly correlated with longer pretrial incarceration in Broward County. But there are many reasons that could be true other than judges being swayed by the scores — people with higher risk scores may also be poorer and have di�culty paying bond, for example.

Most crimes are presented to the judge with a recommended bond amount, but he or she can adjust the amount. Hurley said he often releases �rst-time or low-level o�enders without any bond at all.

However, in the case of Borden and her friend Sade Jones, the teenage girls who stole a kid’s bike and scooter, Hurley raised the bond amount for each girl from the recommended $0 to $1,000 each.

Hurley said he has no recollection of the case and cannot recall if the scores in�uenced his decision.

Sade Jones, who had never been arrested before, was rated a medium risk. ( Josh Ritchie for ProPublica)

The girls spent two nights in jail before being released on bond.

Donate

“We literally sat there and cried” the whole time they were in jail, Jones recalled. The girls were kept in the same cell. Otherwise, Jones said, “I would have gone crazy.” Borden declined repeated requests to comment for this article.

Jones, who had never been arrested before, was rated a medium risk. She completed probation and got the felony burglary charge reduced to misdemeanor trespassing, but she has still struggled to �nd work.

“I went to McDonald’s and a dollar store, and they all said no because of my background,” she said. “It’s all kind of di�cult and unnecessary.”

Julia Angwin is a senior reporter at ProPublica. From 2000 to 2013, she was a reporter at The Wall Street Journal, where she led a privacy investigative team that was a �nalist for a Pulitzer Prize in Explanatory Reporting in 2011 and won a Gerald Loeb Award in 2010.

Je� Larson is the Data Editor at ProPublica. He is a winner of the Livingston Award for the 2011 series Redistricting: How Powerful Interests are Drawing You Out of a Vote. Je� ’s public key can be found here.

Lauren Kirchner is a senior reporting fellow at ProPublica. Surya Mattu is a contributing researcher. Design and production by Rob Weychert and David Sleight.

Sign up to get ProPublica’s investigations delivered to your inbox.

[email protected]

Subscribe

185 Comments ProPublica Login1

t Tweet f Share Sort by Best Recommend  15

Donate

•

FastClock • 3 years ago

Due process requires that in criminal proceedings the accused be able to confront and cross examine the witnesses against him. How does one cross examine a computer algorithm that judges are required to use? I hope the constitutional challenges come soon. 32 △ ▽

•

JL • 3 years ago> FastClock

Depends on what you mean by "required to use". All sentencing risk assessment being used right now are advisory - meaning that they do not commit a judge to a course of action based on the score. A similar challenge was brought up in VA and failed because judges have a right to ignore the risk assessments scores completely. △ ▽

•

uegidock • 3 years ago> JL

And it's only used during sentencing right? It's one thing to challenge your accuser, but once you're found guilty it seems like the rules change. △ ▽

•

Lars Carlson • 3 years ago

Isn't the very concept of 'predictive law enforcement' both entirely unconstitutional and utterly immoral? Who is allowing this? What kind of Orwellian nightmare is 'law enforcement' foisting on us? Here's some 'predictive law-enforcement' for you: investigate every single cop on the force and every single taxpayer making more than $500k a year. You'll find crime there with no problem, no problem at all. 24 △ ▽

•

poor_richard2 • 3 years ago • edited> Lars Carlson

The "for-profit" prison industry generally pressures Republican legislatures to write these laws. Its far more profitable than slave plantations, and removes Democratic voters en masse from voter registration rolls. 13 △ ▽

•

hogwash_bill • 3 years ago> poor_richard2

Front-running Democrats are in their pockets too. Hillary received $133,000 from for-profit prisons, the same amount as Marco Rubio. Only Bernie held the moral high ground. 11 △ ▽

Share ›

Share ›

Share ›

Share ›

Share ›

Share ›

Donate

•

Jm • 3 years ago> poor_richard2

Not only are Democrats complicit in this with Republicans, but Bill Clinton himself largely opened the floodgates for the modern prison industry. All of them should be fed to sharks in the ocean, at least then some good would come from their remaining existence. 5 △ ▽

•

DPsombra • 3 years ago> poor_richard2

I've pretty much always likened how we treat our prisoners to the "new slavery." But, it's cool! The Kardashians are on the air, so we don't have to think about these things! 2 △ ▽

•

dougtheavenger • 3 years ago> poor_richard2

Hillary sucks billionaire donor along every day for money. △ ▽

This comment is awaiting moderation. Show comment.

•

Jm • 3 years ago> richstacy

Left and right are BOTH wrong. You have been brainwashed and are fighting against yourself. WAKE UP. 1 △ ▽

•

Larry Strong • 3 years ago> Lars Carlson

Welcome to the Minority Report. And yes, the police are often the biggest criminals of all. 8 △ ▽

Nixak*77* • 3 years ago • edited> Larry Strong

My Brief Assessment of this 'Unbiased' 'Minority 'pre- crime' Report'-

This so-called 'unbiased' [Oh really?] Assessment report, as a computer based 'predictor' of the likelihood that an offender may offend again, asks 137 questions- which IMO could / should be cut-down to just 1/2 or even 1/3 of that. A number of questions are thinly disguised racially biased, & many are quite

Share ›

Share ›

Share ›

Share ›

Share ›

Donate

•

see more

redundant [IE: it asks about gang affiliation at-least 3 or 4 different Xs]. In fact many or most [or even all] of some entire sections are questionable at best: IE: Under Family Criminality: Asks if you were raised by both parents or NOT, or if you're adopted or raised in foster care [obviously this is biased vs Black people], then if your parents &/or even siblings were ever arrested, & then asks again if they were ever convicted [= redundant].

3 △ ▽

•

MikeCody • 3 years ago> Lars Carlson

Every decision made concerning bail, probation, or parole is an example of predictive law enforcement. In theory, these tools reduce the ability of the decision makers to decide based on the fact that the defendant is the County Commissioner's nephew. Like any computer based scoring system, the algorithms need tuning over time as statistical analysis finds flaws. △ ▽

•

jakee308 • 3 years ago> Lars Carlson

Well they're not actually predicting crime as much as they're predicting whether someone is likely to commit a crime again.

And there are studies that show some people will continue to commit crimes no matter how much or how often they get encarcerated.

But yes, this is the first step on a path to predict who will be a criminal and who will not. They've got some pilot programs going in Great Britain doing just that. The Brits actually like to lock people up more than the US but they don't lock them up for as long so it doesn't show up as much. Plus they're a smaller country so the raw numbers are lower. △ ▽

Nigel Tolley • 3 years ago> jakee308

Not sure where you get that idea. The UK, like most places, has realised that prison is the last resort. You go to prison and your life is affected forever - you get hooked on drugs, you lose your house, it destroys the futures of your kids. Hence prison is a last resort. Also, magistrates can only go up to 3 years (might still

Share ›

Share ›

Share ›

Donate

•

, g y g p 3 y ( g be 2) in prison, maximum. They have to pass the case up the tree to a professional judge in more serious cases. 3 △ ▽

This comment is awaiting moderation. Show comment.

•

Nigel Tolley • 3 years ago

> notgivingname2amachine

No, it isn't perfect. The UK government passes bad laws, same as any other. But a lot fewer die with random violence. △ ▽

•

oh_yea_cuz_england_is_so_great • 3 years ago

> Nigel Tolley

England also has quadruple the violent crime rate of America 2,000 per 100,000 vs. 466. and also leads Europe. It is NOT the model of an effective criminal justice system. △ ▽

Nigel Tolley • 3 years ago

> oh_yea_cuz_england_is_so_great

It may not be perfect, but consider that I know precisely one person who has been murdered, and the murder rate here is 1.0 per 100,000, vs 3.8 per 100,000 in the USA. As for reported violence? If you never see violence, you remember it. If it is everyday, you forget the majority.

But enough statistics, have an "anecdata".

I work a "high risk" job. One I feel pretty sure would've got me into a lot of violence in the USA - indeed, I know that warrants are largely served by SWAT in the US now. In the course of ~10 years doing warrant entries, breaking in to thousands of homes and businesses a year, without anything more than a bit of paper, a Warrant Officer (unarmed, not even a stab vest) & an engineer (for the gas or electric meter), I have experienced violence. We've had a few "close calls" - a guy with guns, a few with axes, etc., many drug farms (nearly all grow houses), & yes, I've been assaulted. I've been assaulted

Share ›

Share ›

Share ›

Donate

•

& yes, I ve been assaulted. I ve been assaulted twice.

Twice. In total, over thousands of jobs where *we literally bang on your door until you answer*, and if you don't, I pick the locks and we let ourselves in. And if you are inside, we tell you we have a warrant and refuse to leave.

We call the police if we feel we need to, I think 3 times in one week is the most, & we sometimes do 8 jobs a day.

So no. Actual "what an American would call violence" violent crime is tiny over here. And our murder rate is a quarter yours. 4 △ ▽

•

Barry • 3 years ago

Wow. Unbelievable that an inaccurate computer algorithm attempting to predict future crime should be allowed to even be seen by a judge. This is unconstitutional.

Criminals should be punished for what they did, not what they might do. 8 △ ▽

•

Nixak*77* • 3 years ago • edited> Barry

'Failed' Polygraphs are NOT allowed as evidence against a defendant, for this very reason. Though They supposedly have a 80% 'success' rate of predicting if someone's Lying- That means they have a 20% FAILURE Rate [false positives / false negatives or undetermined] = Well below the legal standard of ' Guilt Beyond a Reasonable Doubt'!

As such this so-called 'unbiased' 'Minority {pre-crime} Report' which is even MORE UNRELIABLE than polygraphs are, should have also NEVER been allowed in court- effectively used as 'evidence' against defendants!!! 2 △ ▽

•

Christopher Perrien • 3 years ago> Barry

Good point, though recidivism I am sure plays into most judges' judgements. 1 △ ▽

Barry • 3 years ago • edited> Christopher Perrien

That is perfectly OK. Turning that decision making

Share ›

Share ›

Share ›

Share ›

Donate

•

over to a computer that proves inaccurate is not OK. The criteria for applying past crimes to punishment should be simple and transparent. 4 △ ▽

•

disagree • 3 years ago> Barry

I'll accept that only if they get the algorithm to 98-99%. And even then I think a jury should still be required to sign off on it. 2 △ ▽

•

Kal E • 3 years ago • edited> Barry

If you are tasked with determining whether or not someone is allowed to be released under bail, or forced to wait in jail until the case comes up, and you are asked to consider risk of flight (i.e., skipping town) or re-offending, what would you take into consideration? As you go through that thought process, imagine programming a machine to do the same and take the same variables into deciding- that's it. That's the boogey man that most people reacting with worry don't understand. It's actually not that scary or complicated. △ ▽

•

Barry • 3 years ago> Kal E

What is scary is that the results of this program have been shown to be inaccurate and racially biased (even after controlling for different rates of crimes between certain races). 3 △ ▽

•

Kal E • 3 years ago> Barry

Even scarier is when 10,000 judges across the country make decisions where no one can see their "algorithm" and bias- and we just let them continue to perpetuate injustice. I prefer an algorithm that everyone can see, study, and work to fix. It's easier to fix and test the algorithm than to train and hope judges don't bring bias into decision-making. △ ▽

dwpittelli • 3 years ago> Kal E

OK, it may generally make sense to penalize people for their own past bad behavior. But should we increase someone's penalty because his dad went to prison when he was a child?

Share ›

Share ›

Share ›

Share ›

Share ›

Donate

•1 △ ▽

•

Kal E • 3 years ago> dwpittelli

It depends. First, we would need to establish, through rigorous research, that there is causation between those two variables. Assuming that there is, and that adding that variable to our model increases its predictive strength, then we can consider it on ethical grounds and decide if it should be added to the model. Regardless, it should be a matter a public policy and entirely open (no proprietary secret models); not a willy-nilly free market for predictive models with every city/county/state deciding what to do on their own, with no federal oversight/controls- which is what we have now. △ ▽

shinobi42 • 3 years ago • edited

I have been working as a statistician for 10 years, which is relevant when I say that predictive models are always wrong. They are even more wrong when it comes to predicting human behavior.

Surprisingly, this model seems to be doing pretty well on an overall level. 70% for a consumer model would be amazing. The problem is, this isn't a consumer model, this is a model that is impacting the lives of individuals. And it is wrong, analysts know it is going to be wrong some percentage of the time.

I have some general ideas how the strong racial bias could have happened from a modeling perspective. Often there are hidden relationships in the data that are difficult to see, which is why the concept of "Correlation does not imply Causation" is so critical to all modeling endeavors. For instance the percentage of the black population with family members in prison is higher than the non black population. It is likely that race SHOULD have been considered in order to control for the pre exsisting racial bias in our current system. There are many other possible reasons for the model to behave this way.

But the reasons behind this racial bias are not the point. This is a model that the modelers should never have built, in my opinion. Or at the very least it should never have been deployed this way. Handing a simplistic analysis like this off to non analysts without a very strong sense of it's error and overall limitations is just a set up for failure.

S ti th ti i t "C di t thi ?" B t "Sh ld

Share ›

Share ›

Donate

•

Some times the question is not "Can we predict this?" But "Should we?" 7 △ ▽

•

Nixak*77* • 3 years ago • edited> shinobi42

If you actually look at the questionnaire on which this {non}'Predictive' {mis}'Analysis' is based, you'll see it's quite redundant as well as biased [IE: asks about gang affiliation at- least 3Xs - 4Xs]. It has whole sections where IMO much/most if not ALL the questions are of dubious relevance [IE: one section asks mainly about boredom]. IMO the sections asking about family ties [were / are you raised in a 2 Parent home], employment & education status, & if one lives in a hi-crime area- are all [blatantly] Biased against Black people [= 'clever' ways of 'determining' if someone's most likely Black without actually asking the question outright].

Thus IMO this questionnaire is at-least 2Xs - 3Xs too long- it's that redundant, biased &/or dubiously relevant [see my above comment where I go into more detail about this]. 1 △ ▽

•

RCA • 3 years ago • edited> shinobi42

And the alternative is what? Release every person unless we are 100% certain they will re-offend? Lock everyone up? Lean heavily on flawed human intuition, which is often less accurate and more biased than measures like this? △ ▽

•

shinobi42 • 3 years ago> RCA

Build a system that is designed to reduce recidivism, (Especially through drug rehab and occupational programs) and also use punishments that fit the crime committed? 1 △ ▽

RCA • 3 years ago • edited> shinobi42

And in the real world we still need to exercise some sort of judgement about who we release, under what conditions, etc. Which side do we err on?

Clearly no system is perfect, but some are better than others. Even if you are willing to accept (much) higher crime rates in your community to reduce incarceration, the reality is that many people are not willing to make these choices.

Share ›

Share ›

Share ›

Share ›

Donate

•

p p g You cannot simply will them away. Given this reality it seems to me that there are good welfare enhancing arguments for accepting a model like this, even knowing that they are imperfect (both type I and type II errors), because they are the least bad choice that we have. You know what they say about making the perfect the enemy of the good..... △ ▽

•

shinobi42 • 3 years ago> RCA

I personally think that making one person's life harder because a flawed computer model said they are a certain way, is always wrong. These aren't just statistics, they are actual people.

I certainly think criminals should still be punished and incarcerated, but I think the entire system as a whole needs further reform. That is not my area of expertise. 3 △ ▽

•

Nixak*77* • 3 years ago> RCA

FYI: This 'Minority {pre-crime} Report' seems to be used by judges to justify sentencing people to extra-time [or NOT], NOT by parole review boards &/or judges to decide if those already doing time may / should be eligible for early release. △ ▽

RCA • 3 years ago • edited> Nixak*77*

I know but the issues are very similar. The algorithm isn't doing anything that judges and other humans haven't been doing for ages using their own heuristics and the like (keep in mind that this is only being used as a guide)

Incidentally, propublica's claims of bias here are unfounded. I know two people that independently analyzed it and I saw their work- - no sign of bias (if properly analyzed) and the prediction actually looks quite good in the sample propublica provided (obviously quite predictive, linear trend between risk and probability of offending, no sign of b/w bias, etc) I doubt the median judge would

Share ›

Share ›

Share ›

Donate

•

etc). I doubt the median judge would outperform on his own. △ ▽

•

Nixak*77* • 3 years ago • edited> RCA

FYI: I looked at this 137 questionnaire myself & IMO it has multiple instances of NOT only [NOT so hidden] biases, redundancies & even dubiously relevant questions, but that's so even for some of its entire sections!! As such IMO this questionnaire is at-least 2Xs to 3Xs too long [See my 1st comment(s) above]. △ ▽

•

RCA • 3 years ago • edited> Nixak*77*

I haven't spent much time looking at the questionnaire, but the actual risk scores generated by the algorithm show no signs of bias against blacks. In fact, at any given recidivism risk score blacks are at least as likely to reoffend (often more likely though these diffs aren't usually significant). If the questionnaire is as "biased" as you claim, how do you explain that? △ ▽

•

Nixak*77* • 3 years ago • edited> RCA

This article has given several detailed cases of comparatively similar cases for Black vs white 'offenders', which clearly shows some degree of racial bias. I don't know how many actual cases this Propublica investigative team checked into [I think they gave links to their full assessments of cases, as they did the actual questionnaire], but I looked at this questionnaire myself, & I stand by my assessment which I go over more fully in my 1st comment(s) above. △ ▽

This comment is awaiting moderation. Show comment.

Share ›

Share ›

Share ›

Share ›