Cornelius Bentley III was gunned down in a motel room on Chef Menteur Highway in New Orleans East on Sept. 5. Young and African-American, he was the type of person Mayor Mitch Landrieu is trying to save with his signature murder reduction campaign.

As police search for Bentley’s killer, a new study from researchers at the University of Cincinnati claims that the mass “call-ins” and indictments of gang members that form the centerpiece of Landrieu’s NOLA for Life program led to a 27 percent drop in killings of young black men in the program’s first 17 months.

With homicides up sharply this year, however, one of the study’s authors said the program may have reached a point of diminishing returns.

The peer-reviewed paper from University of Cincinnati professors Nicholas Corsaro and Robin Engel found an overall 17 percent decline in New Orleans murders, over and above a drop in similar U.S. cities, from November 2012 to March 2014.

That decline was led by a 32 percent drop in group or gang member-related killings and the 27 percent decline in killings of black men aged 20 to 29.

The study found no significant effect from the program on homicides involving older black men or white, Hispanic and female victims.

“The greatest effect was on young, African-American-victim, shooting-related homicides,” Corsaro said. “The NOLA for Life intervention has reduced firearm-related violence, at least during the time we looked at.”

There are several important caveats to the paper’s conclusions, however.

First, Corsaro’s co-author, criminologist Engel, helped design the New Orleans Police Department’s gang violence reduction program as a consultant, and she thus has an interest in its success.

Second, the paper’s authors acknowledge that comparing New Orleans against other U.S. cities is an imperfect research method. It would have been better, they say, to pit different parts of New Orleans against each other, giving some neighborhoods the NOLA for Life treatment but not others — but the drive to reduce murders everywhere in the city trumped the researchers’ “dream” study.

Third, at least one number suggests NOLA for Life may have hit a brick wall: As of Friday, homicides have jumped 18 percent in 2015, compared with the same period last year.

Corsaro said that trend is not surprising. Throughout the country, many cities have implemented “focused deterrence” programs like NOLA for Life’s call-ins and indictments. Those programs, modeled on the work of criminologist David Kennedy, offer gang members a choice: They can either accept a city’s offer of social workers and job placement, or they can expect swift police action against entire gangs when even a single member is associated with acts of violence.

In New Orleans, groups of gang members have been hauled before judges, prosecutors, social workers and the mayor himself to receive ultimatums. They also have been indicted in large groups. The most recent indictment, unsealed in federal court on Aug. 28, targeted five alleged members of the Young Melph Mafia in Central City.

One weakness in the University of Cincinnati research is that it does not distinguish between the effect of the call-ins versus the indictments, said Jeff Asher, who helped implement NOLA for Life in his former job as a city crime analyst. He has concluded that indictments lead to immediate drops in violence but that the call-ins do not produce significant gains.

Overall, said Daniel Webster, a professor who studies gun violence at Johns Hopkins University, focused deterrence has “a good track record for reducing gun violence, especially among gangs. But we know less about long-term effectiveness and sustainability.”

As time goes on, the targets for call-ins and gang indictments often become less obvious. The most flagrant violent offenders already have been hauled in for warnings before city officials or shipped off to jail.

“It’s really hard to sustain” such a program, Corsaro said. “It’s tricky because you bring in your most violent groups first, so the effects tend to be immediate.”

Meanwhile, police often revert to simply answering calls for service rather than focusing on targeted anti-gang measures.

The city says NOLA for Life is alive and kicking. Charles West, who as director of the city’s Innovation Delivery Team leads the overall NOLA for Life effort, said there has been one call-in so far this year and another is planned. Last year, there were three.

West said that what the city calls “group member murders” — a broad definition of gang-related killings — are down about 23 percent this year. Police Superintendent Michael Harrison has suggested that other types of murders, including domestic violence killings, account for the overall jump despite that decrease.

One part of the academics’ paper, however, points to difficulty in persuading police commanders to adopt a wide definition of gang-related killings. The academics wanted it to be as broad as possible, but they said in the paper that in their personal experience, NOPD officers often were reluctant to label killings as gang-related.

If police are simply marking fewer murders as gang-related, that could deflate the number of gang-related killings recorded.

West said group violence-related killings are down to about 30 percent of the city’s murders, compared with 50 percent when NOLA for Life kicked off in earnest in 2012.

He rejected the suggestion that police are coding fewer killings as gang-related. “NOPD has put in place some formal steps (for identifying gang violence),” he said. “I don’t think they’ve reverted to an old way.”

New Orleans is not the only — and far from the worst — city to experience an uptick in killings this year. A recent New York Times article listed an eye-popping 76 percent increase in Milwaukee, 60 percent in St. Louis and 56 percent in Baltimore. In the 60 biggest cities across the country, according to a more thorough survey by the website 538, there has been a roughly 16 percent increase in killings.

The NOPD and city officials are quick to point out that even though homicides have risen thus far in 2015, the number of shootings has dropped, from 211 at this point last year to 175 this year.

Criminologists say it is hard to explain what might be causing the murder jump and caution that short-term crime trends often are misleading.

In New Orleans, the tempo of killings slowed in the latter part of the summer. Experts also note that even after the nationwide spike, murder rates are far lower than the historical highs of the 1990s.

“Yes, you’ve seen this increase in violence across the country,” West said. “But we aren’t at all taking that as an excuse.”