This is an archived article that was published on sltrib.com in 2015, and information in the article may be outdated. It is provided only for personal research purposes and may not be reprinted.

Body cameras. De-escalation training. Civilian oversight.

These are some of the strategies proposed by police watchdogs and policy advocates in response to months of controversy, nationwide and in Utah, about shootings by police.

But nobody can say whether deadly force is actually on the rise — or measure, beyond individual departments, whether a given policy change reduces fatal encounters with law enforcement.

That's because, bottom line, no one knows how many people are shot by police.

"It is the most awesome power that the state possesses," said David Klinger, professor of criminology at the University of Missouri-St. Louis. "For citizens not to know how often the state kills them, or shoots them and they survive, or shoots and misses — how are we supposed to monitor and judge the appropriateness of the conduct of state agents if we don't know how often they do something?"

Federal efforts to collect data have produced inaccurate results.

The FBI's annual Uniform Crime Report gathers nationwide crime statistics from more than 18,000 police agencies, representing 98 percent of the U.S. population. But police agencies often do not submit the supplementary information that shows the number of people killed by police in justifiable homicides.

Most Utah agencies did not report officer-involved homicides, and those agencies that did vastly underreported them.

From 2007 to 2012, the most recent year for which full FBI stats are available, Utah police agencies reported 18 justifiable homicides by law enforcement. However, The Salt Lake Tribune has identified 59 homicides by law enforcement officers, deemed justified by prosecutors, during that time period. A Tribune review from 2010 through 2014 showed police use of force is the second leading contributor to the state's homicide tally, topped only by domestic violence.

The paper's findings of underreporting are borne out by other reviews of FBI data. The Wall Street Journal recently gathered internal data on homicides by police during the same time period from 105 of the nation's largest departments and found that more than 500 of at least 1,800 police killings were missing from the FBI tally.

The U.S. Centers for Disease Control and Prevention tracks "legal intervention" deaths, using death certificates and autopsy findings. But experts long have considered the totals to be incomplete because medical examiners may not always note police involvement on death certificates.

In yet another federal effort to count deaths at the hands of law enforcement, the U.S. Bureau of Justice Statistics (BJS) called on states to report use-of-force fatalities as part of its "Arrest-Related Deaths" program, launched in 2003. Utah's tally was more accurate in this report than in the FBI data, accounting for 49 use-of-force homicides of 53 identified by The Tribune from 2003 to 2009. The Utah Commission on Criminal and Juvenile Justice gathered and submitted the figures for the BJS report, but rather than relying on police agencies to volunteer information, the commission culled media reports and contacted police for background details, said Ben Peterson, director of research for the CCJJ.

But when an audit found too many deaths missing in the BJS national data, the program was discontinued, said Andrea Burch, a statistician for the U.S. Department of Justice. No data after 2009 will be available.

Peterson said that, to his knowledge, the raw numbers submitted by the state to the justice bureau have never been requested by any agency hoping to study the use of deadly force by police in Utah.

Trends & training • That is not to say Utah's police are disinterested in tracking their own use of force. West Jordan police and the Utah Department of Public Safety, in hope of identifying trends and risk factors, have adopted software to log an array of variables when they use any kind of physical force.

If pepper spray is used, for instance, officers must input where they aimed the spray and what its effect was. If a Taser is deployed, they can record where on the subject's body the electrodes landed. Officers may list the duration of a restraint hold; injuries; the size, race, sex and age of a subject; and whether the force was successful in subduing a person.

"Sometimes [force] works, sometimes it doesn't, and sometimes it works and stops working," said West Jordan police spokesman Dan Roberts.

"We hope, with this new database, we will then be able to go back and say, '42.7 percent of officers do this. Or 39 percent of the times we do this, this happens,' " Roberts said. "It helps us to adjust policy, to adjust training. It helps us to reinforce training and make corrections."

DPS implemented its software about two years ago and is in the process of entering data back to 2000, said Robert Anderson, lieutenant over professional standards. For troopers, one of the most valuable stats could be that of timing, Anderson said.

"Do we need to do more verbally?" Anderson asked. "Are we going to force too quickly?" And at what point in a stop is conflict likely to escalate?

For example, troopers' experiences suggest that during DUI arrests, drivers often stay calm until the blood draw, when suspects may become combative, Anderson said. It could be helpful to learn whether the data bear that out and whether the department can prevent escalation by, say, ensuring the trooper has backup at that point.

"Where is there more potential for conflict?" Anderson said. "If we know that, we can teach and train officers on what to look for."

But the problem with limiting data to individual agencies, Klinger said, is that the deadliest expressions of force don't occur regularly enough for smaller agencies to gather big pools of statistics.

For instance, a Tribune review of shootings by police shows that West Jordan officers have shot three people in the past 10 years, two of them fatally; troopers for Utah Highway Patrol, the state's largest agency, have shot seven people, four fatally.

Among the state's next largest police agencies, Salt Lake City police have shot 21 people, 14 fatally; officers with Unified Police, previously the Salt Lake County Sheriff's Office, have shot 17 people, nine fatally; West Valley City police have shot 19 people, 12 fatally.

"We want to get national data so we can see what types of patterns are going on, what types of situations are emerging, what situations are waning," Klinger said. "Once we start getting the big info, the big numbers, we can find out what's going on in different regions and different states."

The Peace Officer Standards and Training (POST) division, which oversees the basic training of all police recruits in Utah, incorporates observations from the field into the curriculum, said director Scott Stephenson. But, he acknowledged, those observations come as anecdotes, not stats.

"A lot of it is informal," Stephenson said. "We employ about 80 to 100 adjunct officers that help teach our curriculum. We ask them what's been seen on the streets. … Sometimes we see trends in what the subject did and how the officer reacted. [We ask] how can we infuse this into training?"

Quantifying details in those stories could help isolate particular training needs, he noted, but reliance on data could hinder officers' ability to react when events deviate from the norm.

"There is not one situation that's the same; there is not one [shooting] that's the same as another," Stephenson said. "We have to train for the improbable. I'd be very cautious in using statistics exclusively to drive training."

Brian Buchner, president of the National Association for Civilian Oversight of Law Enforcement, said there is too much knowledge to gain from widespread data collection to delay it any longer.

"The absence of data makes it difficult for individual agencies, the policing profession and others to really understand what these incidents are, what agencies are more likely to have shootings, what factors influence shootings or the lethality of shootings, who is involved — so you can reduce them," Buchner said. "They're negative outcomes. Even righteous, justified shootings have a negative impact — to the person who was shot, to the officer, to the communities. You want to reduce the shootings that occur. It's hard to do that without understanding them."

Changing the rules • Apart from the inaccurate FBI stats on justifiable homicides by police, no system exists for agencies to report their shootings, fatal and nonfatal. Developing a mechanism to collect that data could be a challenge.

"There's a lot that has to happen," Buchner said. "Just getting agencies to report their crime data took a lot of time and money and pressure by the federal government. Left to their own devices, individual agencies are not inclined to always collect and report those data, and they're particularly sensitive about reporting officer-involved shootings and other lethal uses of force."

Klinger was more optimistic.

"It's not that difficult," he said. "You make a requirement, you decide which fields of information you want to get and put it into a report."

The Legislature codified this approach last year, requiring Utah police agencies to report all SWAT team and special task-force deployments, as well as forcible entries to execute search warrants.

Lawmakers began examining such operations in the wake of the deadly raid on the Ogden home of Matthew David Stewart in 2012. The Weber-Morgan Narcotics Strike Force served a knock-and-announce warrant that turned into a gunbattle in which Agent Jared Francom was killed and five other officers were wounded. Stewart, who was being investigated for marijuana cultivation, was also wounded. The same strike force has fatally shot at least two other subjects while executing no-knock search warrants: Todd Blair, 45, was killed in 2010 at his Roy home, and William Florez, 51, was killed in 2007, when the team raided his Ogden home.

A coalition of civil-liberties groups pushed for the requirement, modeled on a temporary Maryland law, after the American Civil Liberties Union of Utah tried to gather data for a report on the militarization of police and was unable to get consistent figures.

"It's hard to run any statistical analysis on data that is not standardized," said Connor Boyack, president of the Libertas Institute of Utah, which proposed the bill.

"In these high-risk police encounters, there was no transparency," Boyack said. "Lawmakers were very eager to see data. Rather than relying on an anecdotal story that blows up in the news, how about we get some data from across the state so we can really analyze when and where and whether this use of force is a concern."

When a Utah police agency executes a forcible entry or deploys a special tactical team, it is required to report certain information such as date, location, arrests, property seized and any shots fired by the officers. Agencies are to begin reporting 2014 incidents to the CCJJ by April.

The measure, however, does not require agencies to report all lethal force, focusing on force used only in raids or by tactical groups. Boyack said Libertas is preparing to argue the law be expanded to collect statewide data on all shootings in 2016, after the existing reporting requirement has had "time to get established and prove itself."

"It should be expanded to include all officer-involved shootings because no officer-involved shootings database exists," Boyack said. "In the wake of these national and Utah shootings, we think it's reasonable to say we need some tracking of when and where and why these things are happening."

Buchner hopes the motivation to examine shootings by police helps get similar national measures off the ground.

"There are police policy think tanks and research institutions that haven't been able to get momentum until now, to explore the idea of creating national police shooting data," Buchner said. "It's a herculean task, but I would argue that with something as important as the use of lethal force by police, it's worth investing time, energy and resources into it."

Tribune online • Fatal shootings database

O The Salt Lake Tribune has compiled a database of 87 fatal shootings by law enforcement since 2005. The database at local.sltrib.com/charts/shootings/policeshootings.html names shooting victims and officers, and has narratives of alleged events. Data were gathered from police, prosecutors and media reports; court documents; families of the deceased and open-records requests.