For months, a police car sat outside Elmer Avenue Elementary School, trying to reduce crime through its obvious presence.
But although the officer did finally get drivers to obey the 15 mph speed limit around the school, crime went up, not down.
It was one of many lessons the Schenectady Police Department learned as it began a new approach to policing in 2013.
The department is now using crime analysts to figure out where crime hot spots are and what times those crimes most often occur. Then, officers are sent specifically to those locations at those times to enforce traffic laws.
The goal is to reduce crime through deterrence.
In the first few months of the program, crime plummeted in the Mont Pleasant neighborhood, police say. The program also sharply reduced car thefts in the city and shootings dropped precipitously.
But the crime on Eastern Avenue persisted.
Officers were sure the program was working. Overall crime is down 11.4 percent this year and gun violence is down 25 percent. Confirmed shots fired are down 45 percent.
Officers are also doing thousands of field interviews in their assigned hot spot zones, and they’ve found witnesses to crimes as well as those committing other crimes.
So why wasn’t crime dropping on Eastern?
In the end, Chief Brian Kilcullen said, police realized Eastern wasn’t actually a crime hot spot.
The big crime there was shoplifting from the stores in the small business corridor. Shoplifters aren’t deterred by an officer pulling over speeders a few blocks away, Kilcullen said.
Police have also learned that their presence on the street doesn’t deter domestic violence.
So now they’ve shifted their strategy slightly, focusing on hot spots with street crimes: muggings, burglaries, car thefts, assaults.
And the more they’ve honed their crime analysis of those spots, the more those crimes have fallen.
Mont Pleasant became so safe last year that by January, no part of the neighborhood registered as a hot spot, Kilcullen said. Crime fell 35 percent over the course of 11 months.
“We had a presence. We had some success,” he said. “We had to move resources to hot spots.”
Unfortunately, when the officers left, crime came back. So officers went back to their posts this month.
But they’re not just patrolling 24 hours a day.
The crime analysis is so specific that it tells officers which intersections to be at and at what time.
At one intersection, vehicle crashes happen most often from 1:30 p.m. to 2:30 p.m. on weekdays. Crime hits between 5:30 p.m. and 9:30 p.m. on Saturday nights, but only from 7:30 p.m. to 9:30 p.m. on Thursdays.
Officers are called in on overtime to spend four hours focusing on hot spots, based on the data. They are generally paid through state and federal grants, including Project Impact.
Beat officers have the same data now and they’re encouraged to patrol the hot spots in their zone whenever they have a free moment, Kilcullen said.
Those officers are providing a flood of additional data for the system. They stop passersby to ask them whether they’ve seen anything suspicious. Those “field interviews” are given to the crime analysts and shared with every law enforcement agency in the area.
When a crime occurs, police look through the interviews to see who was nearby at that time.
Other agencies also scan the interviews for information.
Two names once caught a probation officer’s eye — police had stopped two people at 3 a.m. to ask if they’d noticed anyone breaking into cars nearby. The two had seen nothing and went on their way.
But the probation officer knew that both of them were on probation — and violating their curfew with their 3 a.m. “walk.” Additionally, they had been specifically forbidden from hanging out together. Violating those rules led to jail time, Kilcullen said.
“Sharing these reports is part of what this is all about,” he said. “We had two individuals out when they shouldn’t have been. We wouldn’t have known that.”
Now police will start meeting quarterly with other city departments, including engineering and code enforcement, to get help fighting crime.
Code enforcement could investigate code violations in crime-prone locations, Kilcullen said.
Engineering could help improve intersections to reduce accidents, he added.
“We’re doing all this research into the crashes,” he said. “We can do some things to reduce them, but there are others things that can be done.”
In the Central State Street area, three adjacent intersections had 30 percent of the zone’s crashes, according to the data. Those were Brandywine Avenue and Albany Street, Brandywine and State Street, and Brandywine and Duane Avenue.
The City Council is looking into that area now, considering better lighting and pedestrian lighting.
Officers are also patrolling there.
They’ve seen an impact: Crime in the first 11 months was down 6 percent in that area, and analysts have shifted the zone’s borders every three months as they home in on the biggest problem spots.
Police are also trying to drill down even further with their data.
Their next analysis, looking at the results of policing from this winter and spring, will track each crime in each hot spot. Officers will know exactly which crimes led to the zone becoming a hot spot, so that they can focus on the biggest problems.
The goal, Kilcullen said, is to keep improving.