Data released by Stanford University last month shows troopers in Connecticut operate differently from county to county when it comes to the drivers they choose to search.

According to the data, troopers who made stops in New London County from 2011 to 2015 searched black and Hispanic drivers about twice as often as white drivers.

In Hartford County, however, troopers searched black and Hispanic drivers at least three times as much as white drivers. In Tolland, they examined the cars of black drivers three times as often but were less than twice as likely to search the vehicles of Hispanic drivers. The list goes on.

Speaking via email earlier this week, Sgt. Eric Haglund, a state police spokesman, said all troopers operate under existing state and federal case law when it comes to searching a vehicle without a warrant.

According to the courts, troopers can enact such warrantless searches:

• After a lawful arrest.
• If there’s probable cause to believe the car has evidence of a crime.
• To inventory the car’s contents in accordance with standard procedure.
• If the officer has “reasonable and articulable suspicion” the car contains a weapon.

Haglund didn’t attempt to explain why the search rates might vary so much from county to county, but he pointed out that state police typically are using “bias blind technology” when it comes to making traffic stops in the first place. In other words, troopers are using radar or laser technology from a distance on speeding cars.

“Stops using this technology preclude identification of race or ethnicity,” he explained.

Haglund gave a nod to an analysis that the Connecticut Racial Profiling Prohibition Project, or CTRP3, released last week. CTRP3 has singled out Troop H in Hartford in the past, but has not found statistical disparities in other troops. It should be noted that Connecticut's troops are not broken up along county lines.

The analysis, which came out of Central Connecticut State University, affirmed that certain state police troops' differing stop rates could be explained by the characteristics of the regions they cover. It did not delve into search rates, however.

“The State Police remain committed to providing fair and impartial policing across the state and will continue to work with researchers at CCSU and elsewhere to resolve outstanding issues and further inform our work in the years ahead,” Haglund wrote.

The Stanford data is the result of a groundbreaking effort to draw information from all state police agencies into one database. So far, researchers have been able to clean up data on 60 million stops from 20 states, including Connecticut, to accomplish that goal. They additionally have incomplete data from 11 more states.

In their working paper, the researchers made it abundantly clear that it’s not good to draw conclusions from search rates alone. Perhaps certain ethnic groups drive more in certain areas. Or maybe one ethnic group is more likely than another to be carrying contraband, the paper says.

The researchers pointed to the so-called outcome test as one way to add meaning to the numbers. If searches of minority drivers are less likely to be successful, the test reasons, it might mean troopers have a lower standard for searching those drivers.

In New London County, where blacks and Hispanics were searched about twice as often as whites, troopers found contraband on 29.5 percent of Hispanics, 31.4 percent of whites and 34.8 percent of blacks.

In Tolland, for perspective, troopers found contraband on 43.8 percent of whites compared to just 15.6 percent of blacks and 18.2 percent of Hispanics, all while searching blacks and Hispanics more than twice as often.

“What you’re seeing here is that whites are searched far less often and contraband is more likely to be found,” said David McGuire, executive director of the American Civil Liberties Union of Connecticut. “That’s likely the result of police using their training and good policing to identify which white people to stop. They’re stopping white people when there’s a good reason and searching when there’s probable cause.”

He said the opposite appears to be true for minority drivers.

“For us, this is another example of bias in policing,” McGuire said. “At this point there’s not a question of whether it is occurring in Connecticut. The answer is a very loud yes.”

Notably, Stanford researchers, after controlling for age, gender, time and location, still found that blacks and Hispanics were more likely to be ticketed, searched and arrested than white drivers.

McGuire said although the search and hit rates are concerning, they’re not surprising because many other studies have shown similar rates. But he pointed out that state police have been active on CTRP3’s Advisory Board and said he hopes they’ll address Stanford’s numbers, too.

“When someone is searched without cause or is pressured into a consent search and nothing is found, that does a lot to undermine that person’s relationship with law enforcement,” McGuire said. “It ultimately leads to us being less safe.”

l.boyle@theday.com