Smarter Cities grant helps St. Louis get smarter on crime
This article first appeared in the St. Louis Beacon, Jan. 31, 2012 - In the criminal justice system, the people of St. Louis are represented by more than just two separate but equally important groups.
Sure, there are the police who investigate crime and the attorneys who prosecute the offenders, but there are also the courts, the jailers and the probation and parole system.
This is the story of how they are trying to work more closely together, to keep crime in the city on a downward slide.
It begins with a report from IBM, which said that the city should work more closely with criminologists at the University of Missouri-St. Louis to make sure that all of those anti-crime agencies not only are working toward the same goal but also are talking to each other along the way.
To help that coordination effort, the report said the city should hire a chief performance officer and make the best possible use of data not just to solve crimes but stop them before they occur. You might call it the "Moneyball" approach to protecting the city.
After all, says Mayor Francis Slay, that is the whole goal: to assure St. Louisans that despite so-called authoritative crime statistics that cast the city in a negative light, where they live is safe and getting safer.
"You can have good schools and you can clean up the snow real well," Slay says. "But if people don't feel safe in their homes, nothing else matters."
Smarter Cities Challenge
After sending a team of public safety experts to St. Louis for three weeks, IBM released a report last year as part of its Smarter Cities Challenge, recommending how the city's crime-fighting agencies could work together more efficiently.
James Lingerfelt, IBM's senior consultant for public safety, said that in meetings with the Slay administration, personnel at various agencies, community organizations and others, three themes emerged:
A lack of objective metrics on crime
A lack of accountability, making sure everyone knew who was responsible for what
The need for a consistent view of the people who enter the criminal justice as their cases move through the various agencies.
Lingerfelt praised Slay for taking the initiative and applying for a grant for St. Louis, which was the first city chosen for the program in 2011. He said the good working relationship between the mayor's office and Police Chief Dan Isom helped make the visit as productive as possible.
The role that IBM can play in helping the city coordinate the activities of agencies from police through parole is to show how data can help everyone do their jobs better, he said.
"It goes a lot deeper than counting the number of crimes in a given area and trying to predict what crimes would be committed in the future," Lingerfelt said. "You look at economics and recidivism and make predictions where the crimes might occur based on certain people who have committed crimes in the past and were released into the community. You can look at the efficacy of training programs and community programs and their impact on the problem.
"You can take huge amounts of data and come up with some conclusions. This simply wasn't possible in the past. We're taking the principles from areas like finance and purely scientific applications and applying them to the business of public service. It's pretty exciting."
A particularly valuable part of the process, he said, was the partnership reached by the city, the police department and the UMSL department of criminology and criminal justice, to provide a broader view of what too often has been isolated in separate buckets.
"In terms of what St. Louis is trying to do," he said, "you are really breaking new ground by looking at using the whole system so you have a single view. It's something I've not seen in my experience, and I'm excited to see how far it will go.
"Lots of folks are using the tools, but I don't think they're getting the potential out of them. We've seen recommendations accepted by a city, then not much is done with them. That is not the case in St. Louis. We are looking forward to great things."
An Emphasis on Perfomance
One key recommendation that the city adopted quickly was to hire a chief performance officer for coordinating the broader view of crime. Eddie Roth, a lawyer and former member of the city's Board of Police Commissioners, was hired away from his job as an editorial writer for the Post-Dispatch and installed deep in the warren of offices in the mayoral suite at City Hall.
Through his window looking out on Tucker Boulevard, Roth can see the municipal courts building to the left and the city jail to the right, along with the skywalk that connects the two. If he could crane his neck outside and look around the corner, he could see police headquarters.
The interior view is telling as well. On the wall directly opposite his desk, Roth has a map of the city's nine police districts, with numbers of how many people patrol each district and the number of crimes against people and property. Homicides are listed last, in red.
Just as the skywalk makes it easier for the courts and the jail to work together, Roth hopes to use that data and the tools provided by IBM to help all parts of the city's criminal justice system coordinate their efforts. For too long, he said, each agency has looked at people moving through the system a little differently, and their views have not been shared.
"At each stop along the way," he said, "some fields of information about a person drop off and some fields of information are added on. The practical consequence is the prosecutor may not know of police interest in a certain individual when she is deciding on charges, or if the court is considering probation or the release of an arrestee on bail."
Roth said a more efficient use of the numbers will give a better answer to a key question: "Is this an especially bad guy who in the triage of the system deserves more attention, or is this a guy who is less of a threat?"
Slay said that getting a better answer to that question will help the city make the best use of its crime-fighting budget.
"We spend over $250 million a year on crime and law enforcement," the mayor said, "trying to reduce crime in the city. That is a lot of money. That's what this is all about -- making sure we are spending the money in the most efficient way."
Inefficiencies that exist can be blamed more on structure than any animosity among components of the system, Slay added. They have to share data, and they have to speak the same language.
"Police, prosecutors, courts, probation and parole are all doing their job," he said, "and they may be doing it very well. But they are not all answering to the same authority. They are doing the job in silos and not sharing information as much as they could.
"We know that a relatively small number of people are committing a disproportionate share of the crime. You've got thugs out there who are victimizing a lot of people. You have to know who they are so we can catch them, and once they are caught you have to make sure they don't slip through the cracks, if the left hand doesn't know what the right hand is doing."
Public safety has become a particularly sensitive subject in St. Louis because of periodic surveys that place the city atop lists that no one can be proud of. Despite repeated efforts to discredit such reports, Slay says assuring people that the city is safe is a big part of his job.
"Nobody is suggesting that we don't have a challenge here with crime, just like other cities do," he said. "But those kinds of comparisons of cities are apples to oranges. They are not meaningful, and they are unfair. Measuring against yourself is a lot better indicator, and violent crime and property crime here are at their lowest level since 1967."
Three Early Answers
For Richard Rosenfeld, a prominent member of UMSL's criminology department, the university's commitment and the willingness of the police department to become involved in the partnership are big steps.
Rosenfeld has been released from teaching duties this semester for the project, and a graduate student will be embedded with the police department for 20 hours a week, developing the data that can be used to evaluate law enforcement strategies. Other UMSL staff members are on board as well.
Right off, Rosenfeld said there are three ways that authorities can help take a bite out of crime: "hot spot," "focused deterrence" and Project Hope.
The hot spot approach identifies areas where crime has been concentrated in recent days and deploys resources there that make police presence very visible.
"There are two ways police can respond to crime in an area," he said. "They can show up there and arrest persons wanted on warrants, then search for and arrest persons for whom they have good evidence were involved in serious crime. That presumably not only takes those folks off the street, at least for a time, but it also perhaps deters others from engaging in serious crime in that area.
"A better way is simply to show up and drive very slowly through the area, passing by certain intersections very frequently. If you can avert crimes short of making an arrest, that is desirable. What we don't know is how effective it is. It's not clear whether police presence alone averts crime compared with police showing up, stopping people and arresting them."
With focused deterrence, police shine a bright spotlight on people who have been known to be involved in crime in the past, and they offer not just the threat of punishment but the hope for a more positive direction.
"What do you do with those people up to the point they commit their next crime," Rosenfeld asked. "You can either prevent those crimes or arrest the people after the crimes occur. If you call those folks into a meeting, you can encircle them with people who can pull policy levers to make life easier or more difficult for them.
"The message is quite direct and quite simple: We're watching you. If you screw up again, you're off the street, and we'll pull all the levers we can to make sure that's the last time. If you want out, we have service providers who can help you. It's a carrot and stick approach."
Project Hope, which originated in Hawaii, is a kind of tough-love approach for people already on probation or parole to keep them on the straight and narrow.
"Every time they screw up -- violate a condition of community supervision, flunk a drug test, don't show up to an appointment with their probation or parole officer -- the idea is to deliver very swift and very certain sanctions for each one of those violations," Rosenfeld said.
"The sanctions would at the beginning be relatively mild compared to what would happen with repeated violations. Maybe at first they would have additional supervision requirements or even a short stay in jail. By delivering these sanctions swiftly and with certainty, a good bit of research suggests that people's behavior will change and they will comply."
Using data to help make each of those strategies succeed is becoming increasingly popular, Rosenfeld said. But, he added, just as the "Moneyball" approach doesn't sit well with old-time baseball types, it won't necessarily work for everyone, on either side of the equation.
"Think about a young officer who went into police work because he was committed to the work, not just to get a salary," he said. "They want to put bad guys away. They want to go up against them, solve crimes, prevent crimes perhaps, but in the process find the guys committing those crimes and clear the streets of them.
"That is a time-honored approach of police work. Compare that approach to just showing up in a neighborhood that is affected by crime. You're not going to satisfy some officers who are in the business just to arrest criminals. So one can apply the moneyball analogy here, but I wouldn't carry it too far. There is plenty of room in modern policing to engage in both of those strategies."
The chief has a Ph.D.
Working with UMSL will be nothing new for Isom, who has a bachelor's, master's and Ph.D. from the campus. He thinks the academic approach should help complement more traditional police techniques and result in a safer city.
"There are a lot of different questions that could be answered that could provide us with a better template on how we could use our resources more efficiently," Isom said.
"All law enforcement agencies do a rudimentary analysis of cause and effect of what they are doing. But most police officers are not college professors, so it's not formulated in a way that it's an academic research question that has a concrete evaluation component to it. I think that's what this relationship provides."
As far as the value of data, Isom uses the example of something that anyone who has watched a cop show on TV is familiar with.
"Most people who know a little bit about law enforcement know about pin maps," he said. "They were a manual way of tracking crime trends in neighborhoods. But it was very tedious.
"You have a map of the city of St. Louis and a pile of police reports on robberies, and you might have different color pins for different times of day and stick those on the board and see if you can spot a pattern. You can imagine how much time that would take. Computers do that within seconds now."
Like Rosenfeld, Isom says the moneyball analogy can be taken only so far, but the numbers can be a valuable tool.
"If we look a hot spot policing, for example, what we hope that UMSL will do is establish some data points at the beginning of interventions, data points they could use to do more scientific research on whether or not our interventions are having an effect on the neighborhoods we are targeting," he said.
"Anecdotally, we certainly believe that is the case. We will see a pattern in crime, focus our resources there for a period of time and often see a decline in crime in those areas. In-depth research can provide more efficiency to our operations. What exact measure resulted in the decline? Which ones did not have an effect? Potentially, if we can get a lot more detail to the analysis, we'll know how much time to dedicate to a certain area to get the desired effect of reducing crime."
What Will Success Look Like?
That lower crime rate is what everyone involved says is the ultimate goal. But how will they know whether to judge the partnership a success?
Isom says the attitudes of St. Louisans are one legitimate measure.
"The reality is that if crime is reduced by 50 percent," he says, "the public perception of the community is going to change, and the numbers will too, so you won't be able to suggest that the city of St. Louis is crime-ridden. That would also produce a sense of comfort about the community as well.
"What I would hope to see is that we have some structures in place where we were able to evaluate different strategies and see if they produce even more crime reduction, more of a sense of safety in neighborhoods, than we have now. The main goal of all of this is to reduce crime, reduce fear in neighborhoods and provide more protection. I think that will be reflected in two ways: Is crime going down, and do people feel safer? Those are the two measures of success."
Rosenfeld adds that there is little sense in establishing for a specific numerical goal.
"Those kinds of announcements are really quite silly, frankly," he said. "Serious crime has been declining in the city of St. Louis. We want to make sure that the decline persists, and we want to make the decline even steeper.
"We're headed in the right direction, and during a period of flat and declining crime rates, we have the opportunity to try new things. When the crime rate is skyrocketing, every available hand on deck is devoted to preventing the next crime. We have an opportunity now, and we certainly want to be able to take advantage of it."
And the view from City Hall? Slay appreciates the reduction in crime, but he says flatly, "We can do better. It's not enough. If you are a victim, or someone you know is a victim, or there is a crime committed on your block, that's one too many crimes. We are working to make sure every neighborhood is a safe neighborhood."