New York Police Department Deputy Chief Michael Harrington, left, departs federal court in Manhattan on Monday after being charged with corruption. A new study shows that about 1,100 police officers, or about three per day, are charged with committing crimes every year. (Lucas Jackson/Reuters)

So far this month, two New York City police commanders have been arrested on corruption allegations, an officer in Killeen, Tex., has been accused of sexually assaulting a female driver, a Philadelphia police officer has been charged with extortion of a drug dealer, and an officer in Hono­lulu has been accused of raping a 14-year-old girl.

Such sporadic news accounts of police officers being arrested led one group of researchers to a question: How much crime do police officers commit?  No one was keeping track, much as no one was tracking how often police officers shoot and kill civilians, although both may involve use of police power and abuse of public trust.

Now there is an answer: Police officers are arrested about 1,100 times a year, or roughly three officers charged every day, according to a new national study. The most common crimes were simple assault, drunken driving and aggravated assault, and significant numbers of sex crimes were also found. About 72 percent of officers charged in cases with known outcomes are convicted, more than 40 percent of the crimes are committed on duty, and nearly 95 percent of the officers charged are men.

The study is thought to be the first-ever nationwide look at police crime, and was conducted by researchers at Bowling Green State University through a grant from the Justice Department’s National Institute of Justice. The research covered seven years, 2005 to 2011, and sought to quantify not only the prevalence of police officers arrested across the country, but also how law enforcement agencies discipline officers who are arrested and how officer arrests might correlate with other forms of misconduct.

For example, the study found that 22 percent of the officers arrested had been named as defendants in a federal civil rights lawsuit at some point in their careers, unrelated to their arrest case. The authors suggest that police agencies analyzing such suits “could potentially lead to new and improved mechanisms to identify and mitigate various forms of police misconduct.”

In the seven years of the study, the researchers compiled 6,724 cases, or about 960 cases per year, involving about 792 officers per year — 674 officers were arrested more than once. But the study has continued beyond 2011, and lead researcher Philip M. Stinson at Bowling Green said the number of cases now averages about 1,100 arrests per year.

“Police crimes are not uncommon,” Stinson concluded. “Our data directly contradicts some of the prevailing assumptions and the proposition that only a small group of rotten apples perpetrate the vast majority of police crime.” Although nearly 60 percent of the crimes “occurred when the officer was technically off-duty,” Stinson wrote, “a significant portion of these so-called off-duty crimes also lies within the context of police work and the perpetrator’s role as a police officer, including instances where off-duty officers flash a badge, an official weapon, or otherwise use their power, authority, and the respect afforded to them as a means to commit crime.”

“This is probably the tip of the iceberg,” said Cara Rabe-Hemp, a professor at Illinois State University who has studied police deviance. She said the effort is the “first-ever study to quantify police crime” and shows it is “much much more common than what police scholars and police administrators previously thought.”

To be clear, police are not committing crimes at anywhere near the level of civilians. Stinson’s data found 1.7 arrests of police per 100,000 population over the seven years of the study, where the general arrest rate in 2012 alone was 3,888 arrests per 100,000 population.

The number of arrests was “not particularly notable,” said Jim Pasco, executive director of the National Fraternal Order of Police officers’ union, “when you take into account there are between 850,000 and 950,000 law enforcement officers.” The study did not include federal law enforcement, only state and local agencies. A recent Justice Department census of sworn state and local law enforcement officers put the nationwide total in 2012 at about 750,000.

Pasco said “the level of media scrutiny of police is way up,” even in the years of the study, making recruitment of quality officers more difficult. “In that context, some departments have lowered their standards,” Pasco said. “And you get what you paid for.”

The study gathered cases by using Google news alerts that send a message whenever an item on the Internet contains specific search terms. So only cases that are uncovered by a media outlet or disclosed in a police news release are captured in the database, meaning that many arrests that aren’t reported by, or even known to, the police are excluded.

“Every profession struggles with what to do when a member of their own commits a crime,” said Chuck Wexler, executive director of the Police Executive Research Forum (PERF). “But for the police, having officers involved in crime strikes at the very core of what is important to them: public trust and integrity.”

Wexler said PERF and the Justice Department studied a series of sexual assaults committed by San Diego police officers. Officers working late shifts by themselves used traffic stops to commit sexual crimes against women they pulled over, Wexler said. “The actions of these officers impacted the whole department and resulted in major changes. Better screening for hiring, more effective supervision and early warning systems could help reduce these crimes.” He said the new study “sheds light on the extent and nature of this problem.”

The study found that more than 81 percent of the crimes were committed by patrol or detective-level officers and that nearly 85 percent were reported in metropolitan agencies.

New Orleans had the highest per-capita number of officers arrested, with 44.2 arrests per 1,000 officers, during a period that included misconduct committed after Hurricane Katrina. Milwaukee, with 36.7 arrests per 1,000 officers, and Memphis with 29.7 arrests were the cities with the highest arrest ratios.

 

There were 125 officers charged with murder or non-negligent manslaughter in the seven years of the study, and the Bowling Green researchers have followed the outcomes. Of the 125 cases, 107 have been resolved and 77 of those officers were convicted, a 72 percent conviction rate, the same conviction rate as for officers in all crimes where the outcomes are known. A Justice Department study of state court convictions for all defendants, not just police, found a conviction rate in all felony cases of 68 percent, and a 70 percent conviction rate for murder.

Stinson felt it was particularly significant that of all the officers arrested, for offenses ranging from murder to drunken driving, only 54 percent were fired, and 37.5 percent arrested for domestic violence lost their jobs.
The study also found that roughly two-thirds of all the arrests were made by an agency that didn’t employ the officer, and “in at least some cases agencies are not aware of the crimes perpetrated by their own officers.”

Although applicants for police jobs are required to disclose arrests, Stinson said all police departments should require all sworn employees to disclose their arrests or protective orders against them, “so that police agencies can document and respond to known cases of police crime.” He suggested that all law enforcement agencies conduct routine annual criminal background checks of all officers, noting that officers arrested for domestic violence sometimes are able to maintain jobs requiring them to carry guns by not notifying their agency.

“Systems designed to provide an early warning of officers who are problem-prone,” Stinson wrote, “cannot be considered complete if they are unable to identify sworn law enforcement officers who have perpetrated a criminal offense.”

Rabe-Hemp noted that data showing 54 percent of arrested officers being fired likely means that many officers are allowed to resign and retain their law enforcement certification. “When officers resign, they usually just go to a neighboring county,” she said. She said she had studied a number of arrested officers’ cases, and “you could find them bopping all over the Midwest.”

Rabe-Hemp also pointed out that while women constitute 12 to 15 percent of all police officers, they make up less than 5 percent of those arrested. “One solution to police crime should involve hiring more women as police officers,” she said.

In cases involving allegations of sexual abuse, 72 percent of the officers were fired, and more than 80 percent resulted in convictions, the study found. There were 422 reported cases of forcible and statutory rape, 352 cases of forcible fondling and 94 sodomy cases over the seven years of the study, which Stinson called “larger than expected based on the existing research.” The data search turned up 174 examples of male officers arrested in cases of “Driving While Female,” in which women drivers were harassed or assaulted. About 82 percent of those cases ended in convictions.

The study also checked each arrested officer’s name in the federal court database to see if they had ever been named as a defendant in a federal civil rights suit. Of the 5,545 arrested officers, 1,233, or 22 percent, were named as a defendant in a federal civil rights action at some point during their law enforcement career. The researchers found that “officers who perpetrate crimes while on-duty are significantly more likely to have been named as a [civil rights] defendant” than those whose crimes occurred off-duty.

Although the Bowling Green database now includes about 1,100 cases per year, Stinson thought that number underrepresented how much crime police commit, both because news articles may not capture every arrest and also because police agencies may allow officers to resign in lieu of arrest because “they don’t want to air their dirty laundry.”

Stinson noted that almost two-thirds of the police arrests were made by agencies other than the officer’s. In some cases, Stinson wrote, “the employing agency should have made the arrest and failed to do so,” in part because of officers extending each other “professional courtesy.” He noted that of the 960 drunken-driving arrests, there were “comparatively few run-of-the-mill cases of DUI,” and that arrests only occurred if something egregious happened, such as a crash, injuries or leaving the scene.

Of the drunken-driving cases with known outcomes, however, officers were convicted only 35 percent of the time, and only about 38 percent lost their jobs.