Inside the System
Machine Bias, Plea Deals, False Convictions, Brutality Bonds, and Good News.
Last week I wrote about how easy it is for black youth and young men to encounter the police, or end up in the justice system. This week I’ll explore what happens next.
I talked last week about the criminal justice system’s overwhelming bias against black Americans, from practically the moment they’re born. Black youth are many times more likely to encounter police, and to enter the criminal or juvenile justice system. Once they reach adulthood, new sets of algorithms conspire to hold them back.
ProPublica writes about bias in court “risk assessment” software used by judges and prosecutors to determine who should receive harsher sentences:
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.
The software in question, COMPAS, is used by cities and and states across the country. Like the predictive policing software we looked at last week, there is very little transparency into how these software programs are designed, who designed them, and how they control for the over-policing of black neighborhoods. Instead, judges - like the landlords who use tenant screening software - see a score on a computer screen, in many cases with little supporting evidence, and things like this happen:
But Judge James Babler had seen [Paul] 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.
Proponents of risk assessment software claim that, if done properly, it can help divert more people away from jails and prisons, and encourage judges to seek alternatives in sentences. This is true, as the ProPublica sample shows. But only if you’re white:
My home city of Philadelphia has had its own controversy over risk assessment software and bias in sentencing, as Maura Ewing writes in The Appeal:
“The fundamental problem is that the tool doesn’t actually predict the individual’s behavior at all,” said Mark Houldin, policy director at the Defender Association of Philadelphia. “It predicts law enforcement behavior.” He continued: “As we know, [arrest rates] can depend on people’s race and the neighborhoods that are targeted by police activity.”
This is a key point - one that comes up often when discussing bias in algorithms - black Americans are far more likely to be arrested in the first place, and therefore software that predicts future arrests will assign higher risk scores to them.
Another way risk assessment tools exacerbate racial inequality is during pretrial detention. If a piece of software is the deciding factor in whether someone accused of a crime is released on bail or held in jail, racial bias puts more black people behind bars. The impacts of this are serious - people who are incarcerated lose their jobs, are unable to care for loved ones, and suffer other consequences.
Despite its supporters, there is little to no evidence pretrial risk assessment tools have any impact on how many people are sent to jail:
Nationally, there is scant research on how effective predictive algorithms are at reducing prison and jail populations. While the implementation of such tools are often heralded as advances, they may do little in practice. One of the few published studies examined Kentucky’s use of a pretrial risk assessment tool. It found that there was only a small increase in pretrial release, and that over time “judges returned to their previous habits.” Within a couple of years, the study found, the pretrial release rates in Kentucky were actually lower than they were before the tool was implemented, and lower than the national average.
For people who cite about the effectiveness of data and analysis, they produce very little to back up their claims. Giving judges - whose decisions can be influenced by when they last ate - a piece of software is not the answer when the entire system is designed to put more black bodies in their courtrooms in the first place.
Let’s Make a Deal
Whether they are the victim of a biased algorithm or not, African Americans accused of crimes are often at the mercy of prosecutors. It is the goal of prosecutors, especially when charging minor crimes, to secure a plea deal from defendants. This avoids time consuming trials, and adds to their case closure rate. Except, of course, that it’s usually a terrible deal for the defendants, especially if they are black:
[Carlos] Berdejó analyzed 30,807 misdemeanor cases in Wisconsin over a seven-year period and found that white people facing misdemeanor charges were more than 74 percent more likely than black people to have all charges carrying potential prison time dropped, dismissed, or reduced. And white people with no criminal history were substantially more likely to have charges reduced than black people who had no criminal history.
Another study - in New York City, which made locking up people of color its primary law enforcement focus for decades - found similar disparities:
Black defendants were 19 percent more likely than whites to be offered plea deals that included jail or prison time.
Blacks and Latinos charged with misdemeanor person offenses or drug offenses were more likely to be held in jail or prison at their arraignment.
Blacks and Latinos were both significantly more likely to be offered plea deals that included time behind bars for misdemeanor drug offenses. For misdemeanor marijuana cases in particular, blacks were 19 percent more likely to be offered a plea deal that required time behind bars.
When it came to incarceration, blacks were 15 percent more likely to be imprisoned for misdemeanor person and drug offenses, and 14 percent more likely to be imprisoned for felony drug offenses.
So, in our nation’s largest city - where successive mayors have touted their “law and order” stance - black citizens were imprisoned for misdemeanor offenses at much higher rates, and prosecutors offered them lousy plea deals that usually involved jail time, or locked them up before trial as a way to force pleas.
A plea deal can saddle people with criminal records, when in many cases the crime is minor, and they plead out so they can go home. If you’re wondering how much prosecutors and cops lean on this system to make their numbers look good - remember, the police barely solve 50% of murders in America - observe the backlash Larry Krasner faced in Philadelphia, after he was elected on a platform of refusing to prosecute or convict minor offenses, among other reforms. Police union leaders referred to his voters as “parasites” when he was elected, and the state legislature voted to take away some of his powers. In St Louis, the city rocked by riots and protest after the death of Mike Brown, prosecutors joined a police union after a progressive DA was elected, to prevent him from cleaning house and firing attorneys with a history of bias and abuse.
Prosecutors work closely with the police, and depend on them for testimony, evidence, and support securing convictions. Despite being ostensible public servants - our taxes pay their salaries - they have no such duty to the people they put in jail. It is a deeply biased system, and one tilted against black Americans.
With the entire criminal justice system stacked against African Americans, it is unsurprising that they are falsely convicted at much higher rates than other races. A 2017 review of the data tells the tale:
Black people convicted of murder or sexual assault are significantly more likely than their white counterparts to be later found innocent of the crimes, according to a review of nearly 2,000 exonerations nationwide over almost three decades.
Innocent blacks also had to wait disproportionately longer for their names to be cleared than innocent whites, the review, released on Tuesday by the National Registry of Exonerations, found. Blacks wrongfully convicted of murder, for example, spent an average of three more years in prison before being released than whites who were cleared.
There is even racial bias in releasing people found to have been wrongly convicted. We know police have a difficult time closing murders, and even when they do, the punishment is heaped upon black Americans:
When it comes to murder, black defendants account for 40 percent of those convicted of the crime, but 50 percent of those wrongfully convicted, they found. Whites accounted for 36 percent of wrongfully convicted murder defendants.
A high murder rate within the black community contributes to the high number of wrongfully convicted black murder defendants, but it alone does not explain the disparity, the authors write.
Racial bias may play a role. Only about 15 percent of all murders committed by black people involve white victims, yet 31 percent of blacks eventually cleared of murder convictions were initially convicted of killing white people, they found.
You can Google “black man exonerated” and find a half dozen stories from within the last year on the first page. You can visit the Innocence Project and read more stories of people having their lives destroyed by an uncaring system. You can read statistics like these:
As of September 2019, we have documented 365 DNA exoneration cases in the United States, including 20 death penalty cases. Although the individuals involved in these cases were innocent of the crime, approximately 25% had confessed or admitted guilt and 11% had plead guilty. The exonerees spent an average of 14 years in prison, with 10% serving 25 or more years.
Police and prosecutors wield such overwhelming power that a quarter of innocent people who were later exonerated confessed to something they hadn’t done! Law and order types perpetuate the fallacy that police and prosecutors are the only thing standing between our society and lawless anarchy. Is a system that regularly discriminates against and falsely imprisons segments of its population really something we should consider a success?
We Foot the Bill
So, what happens when police and prosecutors are found to have broken the law? Cities pay settlements. Not police departments - cities. Taxpayer dollars go to settle lawsuits brought by private citizens against the police for violating their rights. It is hard to fathom just how much money goes out the door to clean up after police brutality, but journalists keep track:
According to the Chicago Reporter, the city of Chicago spent more over $100 million on lawsuit settlements in 2018. In Los Angeles, the city paid out a total of $880 million in settlements between 2005 and 2018; the police department, at more than 40 percent, was the largest contributor to that total.
Many of these payments are small amounts of money, to settle unlawful arrests or beatings at the hands of police. The large, high-profile settlements tend to involve police shootings - like Eric Garner or Kalief Browder.
Is it surprising that police have not done anything about their mistreatment of minorities, when they are not the ones footing the bill for their abuses? Believe it or not, it’s even worse than that - cities are borrowing much of the money they pay out in police settlements, and Wall Street firms are making huge profits off it:
In our research into the use of police brutality bonds, we found that cities and counties across the United States issue bonds to pay for police brutality settlements and judgments.
In our five case study cities and county, we estimate a total of $1.87 billion in costs related to these police brutality bonds, including more than $1 billion in profit for the investors who buy the bonds.
You’re reading that right - banks and investors make more money in profit than goes to the families harmed by police violence. And the taxpayers cover the cost! Police departments are well aware of this dynamic, and many departments regularly exceed their “settlement” budgets - insane that this is even a thing - knowing that the city will go back to investors for more money to pay grieving families.
The Good News
Now, a palate cleanser. Let me state unequivocally - the protests we are seeing in support of black lives and against police violence are working. The New York Times highlights the incredible spike in support for BLM:
And around the country, things are changing.
The Minneapolis city council has announced plans to disband and rebuild the city police department.
Louisville has banned no-knock police warrants in the name of Breonna Taylor - though they still have not charged the police who killed her.
New York lawmakers have voted to repeal 50-A, a controversial law protecting police misconduct records from scrutiny.
In Seattle, residents set up the Capitol Hill Autonomous Zone, a citizen-run enclave where police are not allowed.
Minneapolis residents converted an abandoned hotel to a shelter for protesters and others displaced by the curfew.
We haven’t seen this sort of outrage and sustained political pressure since the 1960s, and public opinion hasn’t shifted this quickly in modern history. The reasons people are in the streets are extremely bad, but what we are seeing in response is like nothing we’ve seen in our lifetimes. Whatever role you can play, however you can help, please keep doing it. We can’t stop now.
Tips, raised fists, Autonomous Zone passport applications to firstname.lastname@example.org