Get Informed

Subscribe to our newsletters for regular updates, analysis and context straight to your email.

Close Newsletter Signup

Baltimore’s Latest Plan To Clamp Down On Crime: Tricking And Trapping Youth

Once again, police are vilifying kids.

Children from the Pleasant Hope Baptist Church attend a rally and news conference held by the Campaign for Justice, Safety and Jobs in Baltimore, Maryland. During the rally, the advocacy group called for a ‘six-point plan for police reform in Baltimore.’
Win McNamee/Getty Images

Baltimore’s Latest Plan To Clamp Down On Crime: Tricking And Trapping Youth

Once again, police are vilifying kids.


With the number of lethal and nonlethal shootings, robberies, and assaults on the rise in Baltimore, police and government officials are clamoring for ways to make sense of what is happening and stem the violence. There’s been public finger pointing and posturing, but there isn’t a consensus about who among them should be blamed for the current dilemma. Some say the responsibility falls on judgesSome say it is the fault of State’s Attorney Marilyn MosbyOthers say Mayor Catherine Pugh is the problem. But there is at least one narrative that the Baltimore Police Department, city leaders, and some community members seem to agree on: that crime isn’t just an adult problem but a juvenile one as well.

To hear them tell it, youth in Baltimore have gone stark raving mad — robbing, assaulting, breaking property, and swearing in front of children. They are miscreants who are “out of control” and committing a “rash of violent juvenile attacks” and “wreaking havoc.” They are kids who must be stopped. But a new approach to tackling these so-called violent culprits points to a much more unnerving problem: that Baltimore is reviving disastrous narratives about kids that prevailed during the 1990s.

During a public safety event early this month, Baltimore Police Commissioner Kevin Davis announced a new decoy unit of undercover, “young-looking officers” that will patrol the streets for misbehaving youth. “I’m convinced that these juvenile offenders travel in packs. I’m convinced that they look for people who appear vulnerable to them,” he said.

At a press conference days later, Davis noted that many violent crimes are committed by repeat offenders who have been arrested and released. “They either need to be interrupted with a jail cell or some other type of intervention. But the way we’re doing it now is not working,” he said. The commissioner has also decried the number of kids transferred from adult court to juvenile court. In addition to the new decoy unit, more officers will be stationed in South Baltimore and the Inner Harbortwo of five neighborhoods with the most white residents.

“For the rest of the year, you will see an enhanced police presence, and that is not going to go away,” Davis reportedly told an applauding crowd.

Not everyone is clapping, though.

“It, for me, sounds like a return of the myth of the superpredator,” juvenile public defender Jennifer Egan told In Justice Today. “Baltimore police, media, and public officials have started using the same fear-mongering terms — roving gangs, violent youth, brazen attacks — the exact same language that [John] Dilulio, criminologists, and police used in the 1990s to vilify black youth in urban centers.”

While this year has seen an uptick in juvenile arrests for assault, carjacking, and robbery, BPD hasn’t said definitively that more crime is actually happening. What is clear, though, is that juvenile crime has been trending downward for years. Between 2014 and 2016, juvenile complaints fell 38 percent. Juvenile homicide rates have also plummeted for decades. Despite a recent spike in murders and shootings in Baltimore, kids haven’t been committing them, Egan says. Yet the rhetoric surrounding the new police unit and recent assaults does not reflect this reality, according to the public defender. Instead, it is re-upping racialized narratives that previously painted kids as monsters and led to their mass incarceration.

“Kids are not mini adults,” Egan said. “All kids are kids. Another way to describe ‘brazen daylight attacks’ is ‘immature, impulsive, and unsophisticated.’” And the Supreme Court agrees with her. More than once, it has ruled that kids are reckless, irrational, susceptible to peer pressure, and impulsive — precisely because their brains aren’t fully developed. As such, they cannot be treated like adults.

The latest narratives also ignore the fact that youth are underfunded, under-resourced, and heavily policed already, says 16-year-old activist Chelsea Gilmer of the Baltimore Algebra Project, an organization that fights for education and human rights of local youth. For instance, bus stops have turned into their own “gated communities” due to barricades set up by militarized police officers, according to both Gilmer and Egan. Gilmer says patrol officers park their cars near the bus stops and brandish pepper spray and guns as a form of intimidation. “It’s really crazy to me that there’s systems put in place to target us and put us in institutions to keep us from society. But there’s not a lot of systems put in place to help us,” she said.

Egan and criminal defense lawyer Jason Downs agree that any violence committed by young people in Baltimore is not a symptom of depraved, monstrous youth but a result of systemic neglect. Downs points to the new $35 million juvenile detention facility that received three times more funding than a local job program for the city’s youth. “The idea of targeting youthful offenders by creating a unit full of ‘young looking officers’ is, at best, the equivalent of putting a Band-Aid on gunshot wound to stop the bleeding,” he said. “Until we seriously invest in deterring children from the criminal justice system, our City will not be truly safe.”

Reviving the superpredator myth while ignoring educational neglect, unsafe housing, lead paint, and other forms of societal abuse, does the exact opposite, says Egan. “Right now I’m very worried that people are trying to score political points on the backs of children who they have personally failed.”

This article was published in partnership with The Root.

America Had Fewer Executions In 2017 Than In 23 Of The Last 25 Years

Why Execution Numbers Continue To Fall Off A Cliff

David McNew / Getty Images

America Had Fewer Executions In 2017 Than In 23 Of The Last 25 Years

Why Execution Numbers Continue To Fall Off A Cliff


Midday on Tuesday, November 28, the Texas Court of Criminal appeals stayedthe execution of Juan Castillo, who was convicted and sentenced to death as the trigger-man in a 2003 robbery-murder on a San Antonio lovers’ lane. At Castillo’s trial, another inmate testified that Castillo had confessed his role in the crime. That wasn’t true; in what has become a familiar scenario, the inmate recently signed an affidavit declaring that that he made the confession up to curry favor with the State in his own case. The stay was significant for another reason: Castillo was the last American inmate scheduled for a 2017 execution. Every execution concludes a distinct story worth telling, but the end-of-year statistics generate a blaring headline: execution activity continues to fall off a cliff, and is concentrating in a small cohort of capitally-active jurisdictions.

America’s execution activity is dwindling. There were twenty-three executions in 2017. To put that in perspective, there were ninety-eight executions in 1999. Those numbers are not outliers; they are representative of the execution activity during each of the two eras. Execution activity between 1999 and 2001 was well over triple what it was between 2015 and 2017 (eighty-three to twenty-four). Execution activity bottomed out in 2016, when American jurisdictions only executed twenty people. Texas, the most active death penalty jurisdiction in the country, executed thirty-five people in 1999. In 2017, it executed seven, the lowest number since 1996. One might be inclined towards an obvious explanation — American jurisdictions are sentencing fewer people to death, so they are executing fewer of them. That theory, however, fails to account for the fact that death rows are crowded enough to keep the execution chambers quite busy without a new supply of death-sentenced inmates, even in jurisdictions experiencing a slowdown.

That slowdown has to do with a few other things. First, there’s growing judicial acceptance of science about wrongful convictions. Whereas previous generations of judges were conditioned to think of lengthy post-conviction litigation almost exclusively as a means of stalling, there is growing awareness that — particularly in death cases — there may be serious innocence-related errors capable of being corrected. “Death qualified” juries pruned of community members having philosophical objections to executions tended over-credit eyewitness testimony, shoddy forensic evidence, and incentivized-witness accounts. That change in judicial attitudes corresponds with what seems to be a change in the practices of State’s Attorneys, who are relatively more willing to accede to proceedings involving new evidence.

Second, even when capital sentences are lawful, problems with execution method seem to be throttling the execution rate. California has America’s largest death row but hasn’t executed anyone since 2006 because it lacks a lawful lethal injection protocol. As drug manufacturers have moved to prevent their products from being used in executions, a number of American jurisdictions have struggled to obtain an alternative supply. As a result, states seeking to execute inmates are being forced to litigate things like the acceptability of different lethal injection sequences, the use of lethal injection drugs from questionable foreign sources or compounding pharmacies, the use of lethal injection drugs nearing or past an expiration date, the availability of specialists necessary to deliver the drugs without causing excessive pain and suffering, the acceptability of abrupt changes in execution protocol, and the secrecy sometimes used to conceal attributes of the process. One need look no further than Arkansas to see just how spooked states are about lethal injection litigation — it tried to execute eight people in eight days because its batch of midazolam was about to expire.

Third, there’s lots of discussion about how expensive securing a death verdict is, and that’s true. The less-discussed truth is that executing the condemned also costs a small fortune. The number of constitutional challenges available to death row inmates is greater than what is available to those convicted of noncapital crimes, and they usually get superior post-conviction representation. Every state except for Alabama requires the appointment of post-conviction counsel in capital cases, and 18 U.S.C. § 3599 guarantees legal representation, including two lawyers, to capital inmates in federal proceedings. The State can no longer assume its ability to steamroll an in-over-his-head insurance lawyer working his first capital case. If it wants to execute someone, then it’s going to have to pay to do it. And if the entity that decides whether to move forward with executions is closely affiliated with the entity that bears the cost of the litigation, rising costs should reduce execution activity.

The data discloses a national decline in execution activity, but it also shows substantial geographic concentration. This year, there were eight states that had executions. A little over thirty percent of those executions were in Texas, actually down somewhat from the historical average of thirty-seven percent. (The Texas number partially reflects the fact that Harris County, America’s execution juggernaut, elected a new District Attorney that expressly campaigned on restrained use of the death penalty.) Between 1996 and 2000, American executions occurred in twenty-eight states and represented death sentences from 200 counties. Between 2011 and 2015, the figures dropped to fourteen states and 104 counties. In short, the geographic footprint of American execution activity is half of what it was just twenty years ago.

One might look at the shrinking execution footprint and just conclude that the shrinkage is entirely attributable to states that stop executing people. Nope. Even within tried-and-true retentionist jurisdictions, fewer and fewer counties are responsible for executions. Take the five jurisdictions that have imposed the most death sentences in the last twenty years. Fifty-seven Texas counties sentenced inmates executed between 1996 and 2000; that figure fell to twenty-seven in the period between 2011 and 2015. The story is the same in Oklahoma (fourteen to nine), Virginia (twenty-seven to three), and Missouri (fourteen to ten). Only Florida bucks the trend (ten to fourteen).

Others, including Justice BreyerRobert J. Smith, and Brandon L. Garrett, have documented, explained, and evaluated the concentration of death sentences. The causes of concentrating sentencing and execution activity, however, are distinct. I’ve devoted considerable energy to teasing out the differences. Certain localities produce disproportionate execution activity because the practice resides in what I call their “local muscle memory.” Speaking quite generally, a jurisdiction’s muscle memory is the correlated exercise of discretion by multiple local stakeholders, which is heavily influenced by the locality’s history of capital punishment practice. In other words, inertia matters — but more on that in a moment.

There is another, even more basic driver of concentration: money. The counties largely responsible for the executions in concentrating states are large and urban — i.e., they have big budgets. As the per-inmate cost of execution increases, so does the budgetary hit. Rural counties that would have to pay for the litigation necessary to execute an inmate are simply getting priced out of the execution market. The geographic distribution of Florida executions would seem superficially inconsistent with that hypothesis, but there is a wrinkle. In Florida, the county has no role in deciding whether to go ahead with the execution of a death-sentenced inmate. Unlike in Texas — where the District Attorney in the sentencing county must move for an execution date — in Florida, the Governor decides.

Moreover, it is precisely these large jurisdictions that are already the most proficient at the death penalty; they are the ones with the superior institutional muscle memory. The State’s Attorneys in these counties reside near the top of a steep learning curve that requires them, among other things, to secure execution dates, to defend against eleventh-hour collateral litigation in state and federal court, and to support the state attorney general and prison warden. The State’s Attorney offices in these counties are also the offices most capable of transmitting institutional knowledge from one generation of prosecutors to the next. What Professor David Dow said of Texas is equally true of virtually any capitally active county: “[It] executes so many people because it executes so many people. . . . [K]illing people is like most anything else; the more you do it, the better you get. If killing people were like playing the violin, [it] would have been selling out Carnegie Hall years ago.”

More in Explainers

The Truth About Predictive Policing and Race

90-Day Crime Map, Oakland PD
City of Oakland

The Truth About Predictive Policing and Race


Sunday, the New York Times published a well-meaning op-ed about the fears of racial bias in artificial intelligence and predictive policing systems. The author, Bärí A. Williams, should be commended for engaging the debate about building “intelligent” computer systems to predict crime, and for framing these developments in racial justice terms. One thing we have learned about new technologies is that they routinely replicate deep-seated social inequalities — including racial discrimination. In just the last year, we have seen facial recognition technologies unable to accurately identify people of color, and familial DNA databases challenged as discriminatory to over-policed populations. Artificial intelligence policing systems will be no different. If you unthinkingly train A.I. models with racially-biased inputs, the outputs will reflect the underlying societal inequality.

But the issue of racial bias and predictive policing is more complicated than what is detailed in the op-ed. I should know. For several years, I have been researching predictive policing because I was concerned about the racial justice impacts of these new technologies. I am still concerned, but think we need to be clear where the real threats exist.

Take, for example, the situation in Oakland, California described in the op-ed. Ms. Williams eloquently writes:

It’s no wonder criminologists have raised red flags about the self-fulfilling nature of using historical crime data.

This hits close to home. An October 2016 study by the Human Rights Data Analysis Group concluded that if the Oakland Police Department used its 2010 record of drug-crimes information as the basis of an algorithm to guide policing, the department “would have dispatched officers almost exclusively to lower-income, minority neighborhoods,” despite the fact that public-health-based estimates suggest that drug use is much more widespread, taking place in many other parts of the city where my family and I live.

Those “lower-income, minority neighborhoods” contain the barbershop where I take my son for his monthly haircut and our favorite hoagie shop. Would I let him run ahead of me if I knew that simply setting foot on those sidewalks would make him more likely to be seen as a criminal in the eyes of the law?

These are honest fears.

If, as the op-ed suggested, Oakland police used drug arrest statistics to forecast where future crime would occur, then its crime predictions would be as racially discriminatory as the arrest activity. In essence, the crime prediction simply would be replicating arrest patterns (where police patrol), not drug use (where people use drugs). Police patterns might, thus, be influenced by socio-economic and racial factors — not the underlying prevalence of the crime. This would be a discriminatory result — which is why it is quite fortunate that Oakland is doing no such thing. In fact, the Human Rights Data Analysis Group (HRDAG) report that Ms. Williams cites is a hypothetical model examining how a predictive policing system could be racially biased. The HRDAG researchers received a lot of positive press about their study because it used a real predictive policing algorithm designed by PredPol, an actual predictive policing company. But, PredPol does not predict drug crimes, and does not use arrests in its algorithm, precisely because the company knows the results would be racially discriminatory. Nor does Oakland use PredPol. So, the hypothetical fear is not inaccurate, but the suggestion that this is the way predictive policing is actually being used around Oakland barbershops is slightly misleading.

Do not misunderstand this to be a minimization of the racial justice problems in Oakland. As Stanford Professor Jennifer Eberhardt and other researchers have shown, the Oakland Police Department has a demonstrated pattern of racial discrimination that impacts who gets stopped, arrested, and handcuffed — and which suggests deep systemic problems. But, linking real fears about racially unfair policing to hypothetical fears about predictive technologies (which are not being used as described) distorts the critique.

Similarly, the op-ed singles out HunchLab as a company which uses artificial intelligence to build predictive policing systems:

These downsides of A.I. are no secret. Despite this, state and local law enforcement agencies have begun to use predictive policing applications fueled by A.I. like HunchLab, which combines historical crime data, moon phases, location, census data and even professional sports team schedules to predict when and where crime will occur and even who’s likely to commit or be a victim of certain crimes.

The problem with historical crime data is that it’s based upon policing practices that already disproportionately hone in on blacks, Latinos, and those who live in low-income areas.

If the police have discriminated in the past, predictive technology reinforces and perpetuates the problem, sending more officers after people who we know are already targeted and unfairly treated.

This statement certainly has accurate and concerning elements to it. Systems designed on past crime reports (not just arrests) will focus police on poor communities of color in a way that might exacerbate over-policing. This self-fulfilling prophesy problem is something I and others have flagged as it relates to constitutional law, racial bias, and distortions of police practices. It is a worthy concern that directly impacts concerns about racial inequality.

But, here are two cautions to the op-ed’s analysis. The first is the need to be precise in identifying what technology you are critiquing. HunchLab is a place-based predictive policing technology. HunchLab does not do any person-based predictions, so the claim that HunchLab predicts “even who’s likely to commit or be a victim of certain crimes” is wrong and conflates place-based technologies with person-based systems like the Chicago Police Department’s “strategic subjects list” (a.k.a. the “heat list”). They are different systems with different concerns. Place-based predictive policing technologies (which forecast sites of criminal activity) and person-based predictive systems (which forecast individuals at risk for violence) are different in practice, theory, and design, and should not be lumped in the same analysis even if they both ultimately might negatively impact communities of color.

Second, technologies like HunchLab do not blindly follow the data. Instead, they recognize — reflecting the op-ed’s concern — that sometimes the policing remedy does more harm than good. So, in response to a concern about over-policing, HunchLab might re-weight the severity of crimes in their predictive models to avoid policing in an aggressive and unthinking manner. For example, if certain crimes are deemed less threatening to community order, HunchLab recommends less deterrence-based policing in those areas to avoid unnecessary police-citizen contact. The company specifically tweaks the model to avoid unnecessary contact that might increase racial tension. Not all systems do this, and some racially discriminatory impacts will remain, but HunchLab — which was singled out — does attempt to avoid the problem.

I am not minimizing the risks of predictive policing technologies on creating new forms of racial bias. In fact, I have made a scholarly career making similar critical arguments — and just wrote a book — The Rise of Big Data Policing: Surveillance, Race, and the Future of Law Enforcement –that explains how race is a central problem with these new technologies. But, my point is that if you are going to criticize racial bias in predictive policing, you have to do it on the merits of the where the dangers actually lie. In my mind, the more immediate focus of concern should be the Chicago Police Department’s “heat list.” The heat list disproportionately includes men of color in its ever-expanding listpartially because the algorithm uses arrests as the inputs (which again, are more of a function of policing practices than crime realities).

I am also not minimizing the genuine fear that generated Ms. Williams’ op-ed. As anyone who has studied the history of policing in America knows, surveillance technologies tend to be used against communities of color first and in a disproportionate manner. Georgetown University Law School recently hosted a conference on the “color of surveillance” (the “color” unsurprisingly is predominantly black and brown). This reality must shape our critical reaction to any new technology which risks replicating old biases.

In the final analysis, op-eds and news stories about the dangers of racial bias in predictive technologies may be helpful (even when not completely precise), because they generate a societal fear that the companies will respond to through their products. If citizens react to the potential problems, the technology companies can respond and offer ways to solve those problems. In fact, new entrants into the predictive policing field have begun advertising themselves as offering solutions to the problem of racial bias in the data.

But, the danger of creating fear, without engaging the complexity of each technology, means that otherwise worthy arguments can be too easily defeated. Honest conversations are hard. Nuanced debates are less “headline worthy.” But, the future of policing depends on truthfully examining the technologies with an eye toward improving the capabilities and not demonizing the companies (unless they actually deserve it).

Racial bias is a real and future concern for predictive policing, but we need to be thoughtful in our criticisms and precise in our analysis. The future of big data policing depends on it.

More in Podcasts