Cracking cases with databases
Of the 5,000 reports of domestic violence that Portland police receive each year, the domestic violence unit is only able to investigate about 500. So how do they choose which cases to take on?
Work by Criminology and Criminal Justice professor Kris Henning is helping inform those difficult decisions.
Without additional information, police often pursue suspects who are the easiest to find, rather than the most dangerous or those most likely to revictimize their intimate partner.
By contrast, decisions made later in the criminal justice system rely on proven research that assesses a criminal’s likelihood of recidivism. Offenders with personality dysfunctions, those with an extensive criminal history outside the home, and those who are abusing alcohol or drugs are at significantly higher risk to commit more crimes. Studies have proven that focusing our limited correctional supervision and treatment resources on offenders at highest risk results in better outcomes as a system.
"We need to incorporate this same objective decision-making earlier in the criminal justice process," Henning says.
To address this, in 2007 he redesigned the incident reporting form used by Portland police, adding new questions and including a special interview section to allow for victim input. Henning also built a database that assembles information from the previous day's domestic violence reports, and ranks suspects according to the likelihood of future violence. That enables shift sergeants to better and more quickly respond to the highest risk situations.
In the three years that the Portland Police domestic violence unit has used the system, case clearance rates have risen slightly. High-risk offenders are often responsible for several open cases. A single arrest may lead to solving many crimes. Henning and Sgt. Greg Stewart from the Portland Police Bureau have had inquiries from other law enforcement agencies looking to improve their procedures.
Now, Henning is taking a similar approach to addressing burglaries, only 15 percent of which are ever solved nationwide. “The low clearance rate for burglaries results from having too many cases, limited evidence at most crime scenes, and insufficient resources for follow-up investigation,” says Henning.
Studies show that burglars tend to work close to home and will commit several break-ins in a short period of time. So, Henning developed a database that allows investigators to quickly identify burglaries with similar characteristics, including location and timeframe. Now Portland officers can more easily connect related incidents, pool the evidence, and then collar the criminals.
Read more about PSU's outstanding faculty.