Showing posts with label law enforcement. Show all posts
Showing posts with label law enforcement. Show all posts

Monday, May 13, 2024

Why Did Chicago Shoot Down ShotSpotter?

ShotSpotter is an acoustic gunshot-detection system marketed by the public-safety technology firm SoundThinking and used by well over 100 cities in the U. S.  In some ways, it sounds like a law-enforcement dream come true.  Before ShotSpotter, a citizen who reported hearing gunshots could report them, but usually had no idea where the sound came from.  In an area covered by ShotSpotter, police can now often pinpoint the source of the gunshot with an accuracy in the range of 2 to 8 meters (about 6 to 26 feet).  What's not to like about ShotSpotter?

 

A lot, it turns out, at least if you're the mayor of Chicago.  Back in February, the office of Mayor Brandon Johnson, who won his first election campaign promising to end the use of ShotSpotter, announced that the city would not renew its ShotSpotter contract.  Understandably, having spent millions of dollars on the system, Chicago Police Superintendent Larry Snelling defends ShotSpotter.  He was quoted in an Associated Press report as saying, "If we're not utilizing technology, then we're falling behind in crime fighting." 

 

Mayor Johnson and other critics have three main charges against the way ShotSpotter is used.  They say it's "inaccurate."  That is of course a relative term.  In a technical paper published on the ShotSpotter website, the location accuracy was tabulated in a simulated test in a typical environment, with the results cited above (2 to 8 meters, typically).  With regard to false positives (saying there was a gunshot when there was actually some more benign sound such as a car backfiring) and false negatives (missing a true gunshot), the company claims that typically 96% of gunshots are detected correctly.  So although no system is perfect, ShotSpotter engineers seem to have achieved a remarkable success rate in a highly challenging acoustic environment by using advanced signal-processing techniques to enhance the accuracy of what is basically a time-of-flight location system. 

 

Another accusation leveled against the system as typically deployed in an urban setting is that it is racially biased.  A survey by Wired Magazine based on a leaked document giving the secret physical locations of about 25,000 ShotSpotter sensors backs up this accusation.  Wired found that about three-fourths of the neighborhoods where at least one ShotSpotter sensor was deployed were non-white, with an average income of about $50,000 a year.

 

When confronted with these results, SoundThinking senior vice president of forensic services Tom Chittum said that given a limited number of sensors, the company chooses to deploy them in areas that are "likely disproportionately impacted by gun violence."  In other words, they place sensors where they are most likely to pick up gunshot sounds.  For reasons that have nothing to do with ShotSpotter but are deeply rooted in historical and cultural factors, these neighborhoods tend to be poorer and where minority groups live. 

 

The third accusation is harder to refute:  that law-enforcement personnel "misuse" the data provided by ShotSpotter.  Critics cite cases in which police officers are deployed to a Shotspotter-indicated location and find bystanders who they then arrest and charge with violations unrelated to gun use.  Sometimes this leads to cases such as a Chicago grandfather arrested after a ShotSpotter location led officers to him, in which the accused was later released after a judge found insufficient evidence to convict him.

 

Undoubtedly, ShotSpotter has also assisted in the capture and conviction of real criminals.  Otherwise it seems hard to believe that police forces all over the country would spend hundreds of millions of dollars on it, unless they are all playing keep-up-with-the-Jonesville-police-department and making sure they have the latest technology just because it's there. 

 

In the case of Mayor Johnson dumping Chicago's ShotSpotter system, there has been no love lost between the mayor and the police force in general.  In April, Chicago's Fraternal Order of Police—their union—endorsed a drive to recall Mayor Johnson.  And in numerous ways, Mayor Johnson has made it no secret that he is highly critical of how the police do their job.

 

Not everyone in a neighborhood plagued by drug abuse, crime, and violence dislikes the police.  Lots of ordinary citizens would like to see more of the police than they do, and they are probably in favor of anything which helps the police do their job of fighting crime, including ShotSpotter.  After all, most of the sensors are on private property, and the owner's permission has to be granted for the sensor to be installed.  If there was a groundswell of opposition to ShotSpotter, you would think the company would have problems installing their sensors.

 

One of the most basic functions of a city's government is law enforcement.  But over the past few years, especially during the George Floyd riots of 2020, we have been treated to the spectacle of supposedly responsible officials proposing to defund entire police departments.  In the cities where movements in that direction actually gained headway, the results were in keeping with what common sense would predict:  soaring rates of crime and an exodus of both residents and businesses. 

 

District attorneys who make blanket announcements that certain types of crime will no longer be prosecuted find that those exact kinds of crimes proliferate.  How this policy is supposed to benefit society (unless you consider "society" to be restricted to the class of petty criminals) is not clear.

 

As a technology, ShotSpotter works about as well as the state of the art permits, given the challenging environment it works in.  But technology is always about more than technology.  The social environment and the way ShotSpotter results are used has led to a perception in some circles that it is just another way to beat down black people and other persecuted minorities.  Many police personnel are of the same race and culture of the people they are sworn to defend.  It is profoundly demoralizing to be told that a dangerous, tedious job which you do to the best of your ability is not only not appreciated by the highest official of the city, but actively criticized.  With regard to law enforcement and the use of ShotSpotter, Chicago is clearly a house divided.  And we know what happens to a house divided sooner or later:  it cannot stand.

 

Sources:  The Wired report on ShotSpotter sensor locations was published on Feb. 22, 2024 at https://www.wired.com/story/shotspotter-secret-sensor-locations-leak/.  I also referred to the Associated Press article "Chicago to stop using controversial gunshot detection technology this year" at https://apnews.com/article/shotspotter-chicago-gunshot-technology-mayor-f9a1b24d97a1f1efb80296dbe9aff1ed.  I referred to ShotSpotter Inc.'s technical note TN-098, "Precision and accuracy of acoustic gunshot location in an urban environment," dated Jan. 2020.

 


Monday, February 02, 2015

Who Wazed the Sheriff?—Traffic Apps and Law Enforcement


Google's traffic app called Waze allows users to tell each other about traffic-related issues such as construction zones, tie-ups, and speed traps.  It uses a phone's GPS system to locate an icon on a map of the area that everyone using Waze can see.  Google bought Waze from its Israeli developers for a billion dollars in 2013, and it is now one of the most popular free apps on Apple's rankings.  But the National Sheriffs' Association (NSA) is not happy about it.

In a widely publicized statement, the NSA's Deputy Executive Director John Thompson said "we are . . . concerned this app will have a negative effect on saving lives and with public safety activities."  The app's little police icons can show locations of speed traps and other law-enforcement operations.  The sheriffs cite recent ambush attacks on law enforcement, such as the killing of New York City patrolmen Rafael Ramos and Wenjian Liu last December, as examples of hostile actions that could be aided by Waze.

An Associated Press report of the reactions to the NSA statement includes a response by a Google spokesperson, who pointed out that most users tend to drive more carefully when they believe law enforcement is nearby.  Free-speech advocates oppose any restrictions on locating law-enforcement operations via Waze as long as the operations are clearly visible on public property. 

Traffic-law enforcers face a problem that is in some ways paradoxical.  Sometimes they want to be highly visible simply because of the deterrent effect on most law-abiding citizens.  After all, the overarching goal of law enforcement is to encourage obedience to the law.  This goal would be achieved with respect to speed laws if everyone obeyed the speed limits.  And most drivers (but not all) who become aware of a potential speed trap will slow down.  So letting folks know that Smokey is hiding over that next ridge on the interstate will probably lead to fewer speeders, which is what we want, isn't it?  That doesn't take into account the other aspect of the paradox, which is that sometimes traffic cops want to hide, too.

I think it may be significant that the National Sheriffs' Association, but no other major law-enforcement organization, has come out in opposition to Waze.  In small towns in rural areas, and in larger Western counties where the main law enforcement is by sheriffs and not town or city policemen, a considerable fraction of the sheriff's office revenue may come from speeding tickets.  If a deputy has found a nice concealed location where drivers who are just passing through frequently get ticketed for speeding, the last thing he or she wants is for this prize fishing hole to show up on Waze.  Local circumstances such as these can create perverse incentives which encourage law enforcers to rely on a certain number of speeders to show up, just to keep them in business. 

The problem of publicizing law-enforcement operations and locations should not simply be brushed off.  You can imagine a months-long sting operation by police that would climax in a stealthy approach to a crime organization's secret hideout.  But if some clueless driver comes along and posts a lot of cop icons on Waze, and one of the crooks happens to be looking at his phone at the time, the whole operation could come unglued, with dire consequences up to and including bloodshed.

Back in the slow-media days when newspapers were the main forum of public information about law enforcement, reporters would sometimes get wind of secret police operations in advance.  It was a part of the journalistic code of ethics not to spill such beans when it would cause major problems to the police, even though it would make a scoop that would sell papers.  Editors have sat on such hot news many times until after the police have had time to spring their traps.  While such measures could have been viewed as press self-censorship, most observers would agree that it was done in the public's best interest in most cases.  The public's right to know is not absolute, and must be tempered by other considerations such as the safety of law-enforcement officials when publicity would put their lives at risk.

But this is 2015, not 1935, and the age of citizen-journalists.  Instead of fedora-wearing photographers armed with big Graflex cameras, we have baseball-capped passersby armed with iPhones linked to Facebook and Waze.  We can no longer count on the reasoned restraint of professional journalists who can view the larger picture and weigh the consequences of their actions in the long run.  If a Waze user sees a cop and posts the sighting on Waze, the user has no idea whether the cop is there for a routine speed trap or for more specialized and delicate reasons. 

So far, there have been no major incidents to my knowledge in which Waze data on law enforcement personnel locations has led to a major miscarriage of justice or harm to an officer.  But in the present atmosphere of tension between police and many citizens, I can understand why the National Sheriffs Association is touchy about the popularity of Waze, and why they have asked Google to do something about it.

Unfortunately for the NSA, chances are not good for that to happen.  While Google could conceivably run interference between the raw data coming from observers and the displays of police icons, it would be a resource-intensive and probably manual process, which would slow down the edited displays and diminish what is one of the main attractions of Waze in the first place:  its timeliness.  Public-access apps that let the public post information directly depend on that same public not to lie or manipulate their inputs in a nefarious way.  Fortunately for law enforcement, and everybody else, most people at most times are simply trying to get along and help others when it's not too much trouble.  Waze helps them do that, and it looks like the speeding-deterrent effects of posting speed-trap locations will outweigh the possible negative consequences, at least in Google's view.  And in this case, unless some more powerful force intervenes, it's Google's view that counts.

Sources:  The Associated Press article by Eileen Sullivan describing reactions to the National Sheriffs' Association press release was carried by numerous outlets such as the Chicago Tribune at http://www.chicagotribune.com/business/sns-bc-us--police-tracking-app-20150128-story.html.  The NSA statement itself can be found at http://www.sheriffs.org/content/waze-concerns-sheriffs.  I also referred to Wikipedia's article on Waze.