Every morning when Detective Jerry Thompson arrives at work, he checks the jail roster to see who was arrested last night and who’ll be appearing before a judge later in the morning for a bond hearing. Assaults, DUIs, misdemeanors, felony drugs -- Thompson passes over these. He’s looking for just one type of arrest: domestic violence.

Thompson is one of two detectives who works domestic violence cases in High Point, N.C., a city of more than 107,000 people in Guilford County. For years, it has had the highest rate of domestic violence in the state. Roughly a fifth of the police department’s calls for service -- some 5,000 a year -- involve domestic disturbances. The most dangerous involve “intimate partner violence” -- a boyfriend beating up his girlfriend, a husband assaulting his wife or, occasionally, the other way around. These kinds of assaults are associated with a variety of unfortunate outcomes. Abused women have a higher risk for substance abuse, are more likely to commit suicide and have a greater chance of being murdered. In High Point in 2008, a third of the city’s murders involved people killing their intimate partners.

Guilford County’s cities and towns aren’t alone in having a problem with domestic violence. The crime is one of the most pervasive public safety issues nationwide. According to a 2010 survey by the Centers for Disease Control and Prevention, 6 percent of women and 5 percent of men reported being stalked, raped or assaulted in the previous 12 months, primarily by past or present partners. Studies suggest that one out of seven female murder victims is killed by a current or former intimate partner.

Communities across the country have responded in a number of ways. They have expanded and strengthened social services. Many police departments now engage in mandatory arrest, which is taking offenders into custody whenever there are indications that a domestic assault has occurred. Risk assessment tools have been improved. Judges routinely remand offenders to mandatory counseling.

Like other progressive police departments, the High Point Police Department embraced many of these initiatives. In practice, though, its efforts and those of the broader criminal justice system were lackluster. Investigations were perfunctory. Faced with victims who were hard to find or who did not want to press charges, detectives gave up on investigations into “minor” incidents. Typically, there was no follow-up with victims. The police department didn’t work with advocacy groups. Implementation, says High Point Police Chief Marty Sumner, “was poor.”

In 2011, the department decided to reinvigorate its approach. Through an array of positive incentives, community engagement and warnings of jail time, it sought to deter first-time abusers from assaulting their partners again and to dissuade chronic offenders from continuing or escalating their pattern of assaults.

At the core of High Point’s approach is an intervention known as focused deterrence, a crime reduction strategy developed in Boston in the early 1990s as a way to stop gun violence among gangs. Under the strategy, officers would target a specific criminal behavior committed by a small number of chronic offenders, such as gang members; offer them various forms of assistance, such as help earning their GED; and threaten them with sanctions and punishment if the behavior did not stop. In the late 1990s, High Point became one of the first jurisdictions to replicate Boston’s approach. As a result, the city’s violent crime rate fell by nearly half in a year’s time. In 2002, High Point tried focused deterrence on open-air drug markets with notable successes. Then the city decided to use the approach to try to curtail domestic violence.

High Point had been experiencing three to five intimate partner homicides a year. Since the intervention began five years ago, it has had only two (one involved someone new to the city and the other a couple passing through town). Its re-arrest rates for domestic violence have fallen to the low- to mid-teens, far below the 20 to 34 percent experienced by other police departments.

Susan Herman, the New York Police Department’s deputy commissioner for collaborative policing and the former director of the National Center for Victims of Violent Crime, describes the High Point model as the most promising idea to reduce domestic violence she has seen in more than a decade. “This new strategy, backed by research and carefully implemented in partnership with victim advocates, is showing impressive results,” she says.

As word of High Point’s success has spread, other police departments, from Detroit to New York to Tampa, have visited the city. Last summer, the U.S. Justice Department’s Office on Violence Against Women announced plans to replicate High Point’s model in three or four additional cities across the country.

Yet even supporters of the High Point approach to domestic violence concede that its implementation is not a simple or straightforward matter. It takes a certain skill set and experience to launch and run a focused deterrence program. Moreover, injecting the power of the state into volatile domestic relationships comes with risks. One danger stands out above all others: the possibility that offenders will react to a deterrence message not by curtailing their abuse but by escalating it.

Protecting victims should always be the first priority, Herman says. But fear of unintended consequences should not bind communities to a deadly status quo. “Our field was stuck,” she says. “There had been no new ideas in years. We were all operating out of fear -- fear that anything we did could make matters worse.”

To minimize the risk of unintended consequences, High Point proceeded with great care. It took two years to think through the possible outcomes and put safeguards in place. Sumner and his team were convinced that the approach would be valuable, yet they were aware that they were piloting a program “that could potentially make it worse for the victims.” In putting the plan into practice, the city brought more than 15 years of experience with focused deterrence to the task. Which raises a key question: Are police departments that have never done focused deterrence prepared to try it when the risks are so high?

Focused deterrence was originally developed by David Kennedy, one of the country’s most celebrated criminologists. In the mid-1990s he had helped law enforcement officials in Boston design a strikingly effective approach to youth gun violence. Police, probation officers and prosecutors would identify the most dangerous offenders. Then, working together with gang intervention workers, local clergy and people from other service agencies, they delivered a message to those offenders: We care about you. We’re going to help you. However, if you resort to gun violence, we will come down on you, hard.

Some police officers believed that chronic criminals were irrational and unreachable -- that a program like this could not possibly work. Kennedy thought they were rational and that their long criminal records were precisely what made them vulnerable. Many were on probation or awaiting upcoming trials. If offenders ignored the message, authorities could move trials up and ask for stiffer sanctions, step up probation enforcement, or let federal authorities take the lead. “Pulling levers,” Kennedy called it. He referred to the overall approach as focused deterrence.

The strategy worked. Youth homicides in Boston stopped completely for 17 months and the overall homicide rate fell by nearly one-half. Operation Ceasefire, as the Boston intervention was called, became one of the most acclaimed violence reduction initiatives in the country.

Unfortunately, success proved to be a more formidable adversary than youth homicide. A scramble for credit broke out. Kennedy, who served as a coordinator for Operation Ceasefire, guiding and codifying its work, became a focus of resentment by some members of the group. Eventually, the intervention broke down. It was a pattern that repeated itself in other cities. As a result, focused deterrence in 2002 was a strategy that was at once acclaimed and also seen as difficult to sustain.

In 2002, the Hewlett Foundation looked to the technique as a possible answer to other types of crime. It asked Kennedy to see if focused deterrence could be used to deter domestic violence. Picking specific gang members and discouraging them from using guns was one thing. Because they were so criminally active, they were exposed. Law enforcement had many levers it could pull. Deterring abusive men and women from hurting their partners, though, seemed like quite another thing.

“One of the pieces of conventional wisdom about domestic violence is that it’s different than other kinds of violence,” Kennedy says. It happens everywhere and is unpredictable.

Focus seemed impossible. That was why advocates concentrated on expanding services for victims and working to change societal attitudes.

Kennedy keyed in on a particular subset of domestic violence: intimate partner homicides. He was intrigued by what he learned. One study of abusers in Massachusetts found that three-quarters of the 18,000 men who had restraining orders issued against them had a prior criminal history. Nearly half had been arraigned for a violent crime and a quarter had drug offenses or drunk driving citations. People who actually killed their partners had long criminal records. In most cases, Kennedy realized, they had worked their way up to homicide. What if, he wondered, instead of showing offenders they could get away with abusing their partners, the criminal justice system taught them the opposite?

Kennedy was excited about using focused deterrence against intimate partner violence. But when he presented the idea at a special forum on domestic violence convened by the Hewlett Foundation at the Urban Institute in Washington, D.C., he got a frosty response from seasoned domestic violence advocates. Eventually, Kennedy turned his attention elsewhere.

Then, in the summer of 2008, High Point experienced two horrifying episodes of domestic violence just two weeks apart. In each case, a man known to have abused his partner in the past killed her and then killed himself. “It was very rare for us to have a murder/suicide, and we’d had two in two weeks,” recalls Sumner, who at the time was deputy chief and led the major crimes unit.

Every year Sumner reviewed crime trends in order to set priorities for the upcoming fiscal year. This time around, he took a close look at domestic violence and was startled to find that a third of the city’s homicides involved intimate partners. “If we’re not working on that,” Sumner told then-Chief Jim Fealy, “I don’t know what we should be doing.”

Fealy agreed. So Sumner reached out to Kennedy, who sent him a copy of his 2002 paper. Sumner then put together a more complete picture of intimate partner homicide in High Point, pulling files on the 17 intimate partner homicides that had occurred in High Point over the course of the preceding five years. What Sumner discovered mirrored Kennedy’s 2002 findings. “Literally, all of the guys fit this chronic offender pattern -- every single one of them,” says Sumner.

Worse, police, personnel from other agencies and advocates had known every victim was in trouble. In every instance, the victim looked for protection and had not gotten it.

Faced with this evidence, Chief Fealy gave the go-ahead to try focused deterrence against domestic violence. In September 2009, the High Point Police Department set up a meeting to hash out the idea, inviting representatives from 25 agencies, including the district attorney’s office, the U.S. attorney’s office, and every local and state victim advocate group they could find. It began as interagency meetings often do -- with agencies on the defensive. Everyone acknowledged the severity of the problem, but no one was willing to own it. Instead, agencies talked about how they couldn’t do their jobs well because some other agency was not doing its job well. The discussion continued in this way until Kennedy, who was sitting in the back of the room, raised his hand and posed a question. “Are these offenders resisting our best efforts?” he asked.

“Everybody looked around at each other,” recalls Sumner. “The consensus was no.” Police investigations were lacking, as were connections with advocates and service providers. Probation had a supervision program for high-risk domestic violence offenders, but if the offender never appeared to meet with his probation officer, the agency changed the offender’s designation to “inactive” and ended supervision requirements -- without even telling the judge. The group realized the situation needed to be changed and resolved to try an offender-focused domestic violence program.

The first step was to understand the problem. The department applied for and won a grant to document exactly how prevalent domestic violence was. With that $25,000 in hand, the police department hired researchers from the University of North Carolina-Greensboro (UNCG) to review 10 years of data. Their findings reinforced Kennedy and Sumner’s prior research. In the first decade of the aughts, 1,030 people had been charged with a domestic violence-related offense. These were not one-off offenders. On average, each had 10 charges on his record. In other words, these were people deeply involved in the criminal justice system. There were levers to pull.

While the researchers continued their study, a working group began to develop a strategy. The group agreed that a first-time offender should be treated differently from a dangerous, chronic offender. Ultimately, the department decided to set up a four-tier system. When patrol officers responded to a call about a domestic disturbance and determined that one party was an aggressor, he or she would become a “D-class” offender. The next day a different patrol officer would deliver a customized letter, signed by the chief, notifying them that they’d been added to a watch list. A similar intervention had been carried out in Yorkshire, England, in the 1990s, and had significantly reduced recidivism. High Point hoped to achieve a similar outcome with a follow-up visit.

Offenders who were arrested would receive a stiffer dose of deterrence, starting with a visit in jail from Detective Thompson or his counterpart. There, the detectives would deliver a letter from the police chief, informing them that they were now listed as “C-class” offenders and that if they offended again there would be a variety of unpleasant consequences -- higher bails, tighter supervision on probation, more vigorous prosecution. The names of C-class offenders were added to the police alert system so that officers would know they were dealing with a C-class offender if they stopped that person and ran their name through the system. A police liaison to the court would flag C-class offenders for prosecutors and judges so that cases against them could be moved up and bail set higher. “Basically, we want them to see that the system treats them differently because they are on the list,” says Thompson.

People with a history of domestic violence arrests or “C-classers” who went on to commit another assault went into the “B-class.” They got a heavier dose of deterrence in the form of a “call-in” -- a meeting at City Hall with a domestic violence task force, a group that included police, prosecutors, social services and nonprofit representatives, and community representatives. They would be offered assistance, such as rehabilitation services or job training, but they would be given another message as well: Re-offense would have serious consequences. Prosecutors and police would make convicting them and securing a long jail or prison sentence a priority. People who ignored the call-in warning or who had cases pending and a history of the most dangerous types of violence went into the “A-class.” These were the offenders prosecutors were trying to put away.

What High Point was creating was unprecedented. This was not just a new risk assessment system. It was an attempt to put in place a comprehensive plan for deterring intimate partner domestic violence -- for teaching abusers that they could not get away with repeated violent acts against their partners.

High Point’s efforts also came with risks. Two scenarios in particular gave the working group nightmares. The first was that an offender might receive a warning, perhaps from a detective visiting him in jail or during a call-in, and then turn around and kill his victim. The second nightmare scenario, says Kennedy, “was that he would have her chained in the basement unable to talk to her friends or get to a telephone. That would look to us like success because she wasn’t calling anymore.”

The first test of the program came in February 2012, when the task force did its first call-in with B-level offenders. The group had identified 12 candidates, four of whom were men living with the partner they had abused. To avoid the nightmare scenarios, the task force created safety plans for these and other at-risk women living with abusive partners. The women were asked to identify someone that advocates could call -- a colleague at work, a neighbor, a family member -- who would know if the victims were all right even if the service providers couldn’t reach them directly. No one knew if these measures would be enough.

Sumner, who had by that time been promoted to chief, shared Kennedy’s concerns. Like everyone involved in the plan, he recognized that an intervention of this sort “could potentially make it worse for our victims.” That was one of the reasons the department and its partners spent two years mapping out exactly how the intervention would work. At the same time, Sumner believed that if High Point did nothing, “surely we would have three to five domestic‑related homicides this year.”

The call-in was held at City Hall on the evening of Feb. 21, 2012. A dozen men were ushered into the meeting room where the city council usually convenes. There, the working group explained the program to them. A community activist and a minister described how they wanted to help. Community speakers -- including an entire chapter of a motorcycle club in full regalia -- spoke of their abhorrence for domestic violence. Then police laid out what would happen if the domestic assaults continued -- vigorous enforcement, sky-high bails, federal cases, prosecution rather than plea bargains.

Kennedy, who has attended hundreds of call-ins around the country, was struck by the offenders’ demeanor. The men were surly, angry to have been called out. One was openly scornful. No one displayed the charm that even some drug dealers have. No one asked about services. Interacting with these offenders felt nastier than dealing with violent gang members.

The next day, victims services contacted the women who were still living with their abusers. They were asked how their partners received the message, what they said when they got home. The response from all the victims was similar, Sumner reports. “It was, ‘Oh, he got the message. He didn’t like it, but he heard you, and the most important thing you told him was that this was not being driven by me at all. It was completely you and him, the task force and him and his behavior. Y’all are holding him accountable, and I had nothing to do with it.’”

This was exactly the response the task force had hoped for. Still, everyone worried about retaliation by the offenders. It never came. Instead, recidivism for domestic abuse declined dramatically. Twelve months into the program, only 9 percent of listed offenders in High Point had attacked again, compared with 20 to 34 percent of abusers nationwide. Indeed, recidivism rates are so low that High Point hasn’t had to schedule a call-in for B-class offenders since September 2014.

Given High Point’s results, it’s not surprising that its intervention has begun to attract interest from other governments. But a focused deterrence intervention targeting domestic abuse requires intense, ongoing dedication. Spending a morning with Thompson at the Guilford County jail illustrates why.

After reviewing the arrests of the previous night, Thompson meets with High Point’s newest C-class offender. He’s a young male, age 23, wearing an orange jail jumpsuit and flip-flops. He looks sleepy. Thompson introduces himself and strikes a sympathetic note. “I read the report about what happened,” Thompson begins. “I wasn’t there. You know the truth. I don’t know the truth. I do know you were drinking some.”

The man looks noncommittal.

“Look, it has to stop,” Thompson continues. “You are a convicted felon already. Domestic violence stuff in High Point, years ago it was different. Now things are different. They got your name flagged. If you get charged again, it’s going to be different than you’ve ever seen before. There’s a guy in here now on a $100,000 bond still waiting for a court date.

“You got to think about your kids,” he continues, changing track. “You have an eight-month-old? You want your child to be visiting you in prison?”

The man mumbles that his girlfriend is “crazy.” He volunteers that he may ask a judge to take out a restraining order on her.

“You have got to figure out some way to deal with it where you don’t get locked up,” says Thompson. “You got to think different. You are a grown man now.”

Thompson leaves, pleased with the encounter. He feels like he got through. “Whoever the detective is, he has to figure out who the person he is taking with is,” he says of the meeting. “You have to be serious about it, but also try to get on their level to a certain extent.” How you have the conversation, he believes, is critically important.

An hour later, Thompson is at the biweekly task force meeting, which includes representatives from at least eight agencies and nonprofit groups. Today, they’re talking about a chronic abuser who has ignored earlier deterrence messages and whose partner has steadfastly refused to testify against him. This time, the man assaulted her at a gas station -- one that had video cameras. The victim is refusing to testify, says Assistant District Attorney Walter Jones, but with the video, his office believes it can make the case. It’s an example of what a determined interagency group can do. It also illustrates that such collaboration is resource-intensive. Doing focused deterrence for domestic violence is, Sumner acknowledges, “a heavy lift,” even for a department with 18 years’ experience doing such intervention.

That points to another major question about focused deterrence. High Point had used the strategy against gun violence, open-air drug markets and shoplifting before attempting to deter domestic violence. That history of success helped greatly with buy-in; the experiences gained made interagency collaboration easier. It is less clear that other police departments will be able to address domestic violence with focused deterrence if this is the first time they’ve tried it. “The operational piece of this is very complex and requires quite a bit of commitment and attention to detail,” says UNCG’s John Weil, one of the program evaluators. “You can’t drop any particular piece and have this be effective.”

Indeed, Weil and his co-evaluator Stacy Sechrist, who have fostered other focused deterrence interventions across the state, say the departments that have used the intervention with the greatest success think of it as more of a mindset or approach than a program. “People bring a program mentality to this, but they are not programs,” says Weil. “There is no beginning and no end.”

Despite these challenges, Sumner believes the model can be replicated. An effort in the nearby town of Lexington, N.C., has already yielded similar results.

It’s not just his three years overseeing the program that makes Sumner believe a focused deterrence approach to domestic violence can and should be tried elsewhere. It’s personal. “I witnessed [domestic violence] firsthand. My mom was a DV victim for probably 10 years,” he says. “It was one of those situations where my dad’s peers weren’t putting any pressure on him, the courts wouldn’t put any pressure on him, there was nobody outside the home putting any pressure on him to control himself. I think he would have responded to this.”