COLUMBUS, Ohio - In a public park and forced to leave.
Cuffed and questioned for whistling while waiting for a bus.
Pulled over for spending too much time at a gas station.
Some black drivers and pedestrians in Cincinnati say they've been unfairly stopped and questioned by police.
"It seems to be if you are a minority, you're a target and you're automatically doing something wrong," said Michelle Cameron, a black resident who lives in the Cincinnati neighborhood of Westwood.
Cameron's perception isn't rare for blacks in Ohio's three largest cities. An analysis of hundreds of thousands of police stop data confirms a racial disparity does exist across the state.
Certain police actions, such as pulling over a driver or issuing a ticket, are more likely to occur in predominantly black neighborhoods than in white ones.
And black drivers are also much more likely to be pulled over than whites, according to a review of 315,281 stops from 2009-2017 in Cincinnati, 128,157 in Columbus from 2012 through 2016, and 47,079 in Cleveland in 2016 and 2017.
But the analysis also found that some disparities were likely the result of police practices versus outright discrimination.
Reporters compared police stops in census block groups that were at least 75% white and compared them to ones that were at least 75% African American. Highway stops were not included. Here are key findings:
In Cincinnati, police made 120% more total stops per resident in predominantly black areas.
In Columbus, police made 84% more total stops per resident in neighborhoods that were at least 75% black.
Cleveland issued 26% more tickets per resident in predominantly black areas.
Once ticketed or stopped, police arrested blacks at a much higher rate than whites. Blacks made up 74% of all Cincinnati stop arrests, 70% of all Cleveland arrests during ticketed stops and 59% of traffic stop arrests in Columbus.
Police pulled over African Americans at about the same high rate in a time-controlled sample of stops made either in daylight or darkness. This suggests they pull over drivers regardless of race.
But more stops in minority neighborhoods suggests the location of patrols may be a factor in the racial disparities.
Cincinnati Police Chief Eliot Isaac said the findings concern him.
"We need to look at the reasons why in predominantly African American communities there (is) more police contact," he said.
The racial makeup of police stops on highways was closer to the overall proportion of the population, suggesting that blacks and whites commit traffic offenses at the same rate.
In 2016, the population of Cincinnati was 49% non-Hispanic white, 43% black, and the rest Hispanic, Asian, mixed, or unknown.
Yet black people made up 57% of traffic stops between 2012 and 2017, both on foot and by vehicle, the data analysis shows. Whites made up 41% of stops.
Considering the racial makeup of Cincinnati, blacks were stopped at a 58% higher rate than whites.
And once stopped, 19% of black drivers pulled over are arrested, compared to 9% of whites.
Blacks made up 75% of all arrests, compared to 24% for whites.
"This type of disparity is what must be explained, and so far the city has failed to do so," said Al Gerhardstein of the racial disparity in Cincinnati.
The Cincinnati civil rights lawyer helped write the Collaborative Agreement. The landmark pact was created nearly two decades ago in part to improve police-community relations after a riot in 2001. It has been praised as a national model, though some say the agreement, which exists today by mutual consent, is neglected in some respects.
And the Cincinnati Police Department has been criticized in the past for failing to adequately track possible police bias.
Dan Hils, president of the union representing Cincinnati police, said by email that past examinations have "confirmed that our officers do not use race as the basis for our stops."
"Our people are deployed so that they can provide service where it is needed most," Hils added
Today, the department does not maintain formal policy that guides decisions about where to deploy officers, according to Lt. Steve Saunders, a police spokesman. But Isaac said such decisions are guided by recommendations from the International Association of Chiefs of Police.
More than half of all sworn Cincinnati officers, about 575, are regularly on patrol to enforce traffic violations, though they have other duties as well. Areas in the city with more violent crime and more calls for service are generally more likely to see a higher police presence, Saunders added.
Cincinnati police did implement an implicit bias training program for all law enforcement personnel. The program is expected to be completed by April, according to a city spokesman.
Isaac does not believe widespread bias exists in the department. But that's not to say an officer never engaged in biased activity, he added.
He pointed to social conditions such as poverty as factors that bring police to certain communities.
"If we address our poverty issues in this city," Isaac said, "you will see these disparities disappear."
What the analysis doesn't prove
A higher percentage of blacks stopped doesn't necessarily mean that police targeted African Americans, though, based on the data analysis.
One theory has it that if widespread racial profiling exists in a city, black drivers will be stopped more often in daylight than in darkness, when it is harder for police to see drivers. That's not occurring, though, in Ohio's three largest cities, based on the data analysis.
Reporters analyzed street stops between 5:16 p.m. and 9:41 p.m. (excluding twilight) over several years in Cincinnati. The stops occurred in either daylight or darkness because of the changing seasons and daylight savings time.
In darkness, when it would be harder to see the race of a driver, white drivers were approximately 33.1% of stops, black drivers 65.7%, and others 2.2%.
In daylight, rates were nearly the same. Whites were still pulled over at a lower rate. Of those stopped, 33.5% were white, 64.7% were black, and 1.8% were Asian, mixed, or unknown.
For highway stops though, the opposite occurred: whites were pulled over at a rate higher than their city population. In that same time period, in both daylight and darkness, about 68% of those stopped were non-Hispanic whites, about 27% were black, and the rest Asians, Hispanic, or unknown.
In other words, when it may be difficult to tell the race of drivers (because they are whizzing past on the freeway), police stopped many more white drivers than black ones.
For neighborhood stops though, the data shows police consistently pulled over more black drivers, even if they probably couldn't see them, possibly because they were more likely to already be in African American neighborhoods.
The daylight-darkness findings assume people drive the same way at all times, though, which a 2019 study suggested might not be the case. This analysis also doesn't account for the racial makeup of people driving into and out of the city.
Cop complaints made by blacks rarely sustained
Yet some black drivers in the city believe they have been profiled.
Drivers like Michelle Cameron, the black Westwood resident out driving one evening in 2017 in Price Hill.
The 40-year-old Cameron had some time before her shift as a security guard at a downtown office building, so she pulled into Mount Echo Park, hoping to take in the view and complete some paperwork.
A Cincinnati police officer ruined any chance for a quiet moment, Cameron said, pulling up beside her and asking her to leave because the park had closed.
But Cameron knew the park was actually open and that she hadn't done anything wrong (she wasn't ticketed). Even so, she felt she had no choice but to drive away.
She reported the officer to the Citizen Complaint Authority, an agency that investigates accusations against police. Five months passed before the officer was questioned. The agency is currently short-staffed.
Findings by the agency are voted on by a board of seven citizens. The city manager signs off on conclusions.
The Citizen Complaint Authority reviewed Cameron's stop, finding the officer had broken department policy by not turning on her body camera. But discrimination, harassment and improper stop allegations were not sustained due to a lack of "sufficient evidence."
A not sustained ruling indicates the agency couldn't prove an allegation one way or the other.
"Something should have happened to that (officer)," Cameron said. "I tried to expose wrong."
The cop in Cameron's case told an agency investigator that there are "frequent heroin users in the park," but didn't mention any cause for suspicion of Cameron or her vehicle, according to a record of the complaint.
Cameron felt racially profiled, she said. She is hesitant to drive at night now, fearful she may be humiliated again.
"I felt defeated," she said. "I felt violated. ... I felt low."
Reporters obtained 70 stop-related complaints submitted to the Citizen Complaint Authority from 2015 through May of 2019. Older cases exist, but a city spokesman said some full complaint investigation reports have likely been disposed of as part of regular record maintenance.
None of the 14 formal allegations of discrimination contained in the cases, including Cameron's, was upheld by the agency. Investigators found nine such accusations were "not sustained," meaning they could not determine whether discrimination occurred. Five allegations were deemed unfounded.
Three out of four of the complaints were filed by black people.
Data available through the city's website provides case outcomes since 2014, showing discrimination charges are rarely sustained by the Citizen Complaint Authority.
The agency received 149 stop-related complaints and 84 discrimination complaints in more than five years, according to an analysis of the city data. Some cases contain more than one complaint.
Of the more than 225 complaints, just one stop complaint and three discrimination complaints were validated.
The actual number of instances of perceived discrimination is likely higher, based on a review of the 70 cases obtained by reporters.
For example, in a 2017 complaint, a black man said he felt profiled, and the claim was recorded on an officer's body camera. The man had parked and was walking when officers handcuffed him under suspicion of drug possession. The man later accused an officer of telling him, "You the big man around here? We're going to get to know you," according to the complaint.
The authority didn't consider discrimination in the case despite mentioning the man's comment about feeling profiled in its report.
And in a case last year, an officer claimed he heard a black woman whistle while investigating a prostitution complaint in the West End. The officer handcuffed the woman.
The authority didn't consider discrimination in the case. It determined her alleged whistling constituted enough reasonable suspicion to clear the officer of wrongdoing.
The woman, 45 years old at the time, was waiting for a bus to go babysit her granddaughter, according to the complaint report.
Fear of the 'gateway' stop
Cincinnati officers' conduct has raised concerns from the Citizen Complaint Authority in some cases but rarely sustained charges of wrongdoing.
In 2015, a black driver and his passengers were stopped and ordered out of their vehicle by officers working an off-duty detail. One officer later told the agency that he'd smelled marijuana. No drugs were recovered.
The officer also said later that the stop was initiated because the vehicle had been parked too long at a gas station and had a cracked windshield, according to a record of the complaint. The occupants had been "just sitting" for about 10 minutes at the station, the officer said.
Both officers were cleared of wrongdoing.
In 2017, an officer thought he smelled marijuana through an open vehicle window occupied by a black man.
After the man entered a nearby house, the officer yelled for him to come out, telling him he was under arrest. More officers arrived, and a woman and her minor son were forced from the house as officers searched it. Nothing was found. The agency determined the search violated Cincinnati police policies, according to the case record.
The man was charged with obstructing official business and fined $310 but wasn't charged with a drug-related offense.
About a year earlier, in a complaint report regarding a different case, the agency wrote that increasingly, Cincinnati officers "believed" they saw or smelled marijuana and thus conducted a vehicle search. The agency recommended Cincinnati police review the issue to determine if more training or new policy would help decrease unsuccessful searches.
Kim Neal, the director of the Citizen Complaint Authority, said that after the agency raised the issue with Cincinnati police, similar incidents decreased.
An independent review team assessing the city's effort toward bias-free policing suggested tracking arrests by officers ultimately declined for prosecution.
When asked about whether the suggestion had been implemented, a city official said by email that the implementation phase didn't begin until February of this year. The final report by the review team was submitted in January of 2018.
"(The independent review) recommendations are multi-faceted and cannot be addressed all at once," wrote Jason Cooper, who heads criminal justice initiatives within the Cincinnati city manager's office.
Cooper added that experts and community leaders are helping to guide the creation of a broad strategy for data collection and analysis for the "ongoing evaluation of progress towards bias-free policing." Recommendations from the independent review are being considered.
A plan will be shared with an oversight group in February.
Gregory Levy said he views stops as a "gateway," a means for police to investigate whether a person has committed a crime - mainly drug crimes.
In January, he tried to stroll across a street near the Hamilton County Courthouse. As he crossed Court Street at Main, an officer hailed him, demanding he stop, Levy said.
The officer asked if he'd noticed the "don't walk" signal flashing as he stepped into the crosswalk, Levy said.
"I said, 'No, I didn't see it. (There are) no cars coming.' I got the music in, walking across the street," Levy said. "It was some petty stuff."
He was charged with a crosswalk violation, an offense that was ultimately dismissed, according to court records.
And in March 2017, Levy was driving along Central Parkway with a friend, both "jigging to the music" in his Mercury Cougar. A Cincinnati officer pulled him over. The officer asked if any drugs were in his car, Levy said.
Levy was ticketed for a loud muffler and driving with a suspended license, suspended because he'd fallen behind on child support, according to court records.
The loud muffler charge was dismissed, but he was fined $260 on the other charge.
"Black people get targeted because they (police) are looking to find drugs all the time," Levy said. "It's like we're trapped."
In Columbus: 'I'm a magnet for the police'
Robert Jackson said the same day as his release from prison, in mid-September, he was at the downtown Columbus Greyhound station when he saw two men arguing. He thought he should leave before police arrived. As he walked away, an officer stopped and questioned him, he said.
"He automatically assumed it was us," Jackson said, referring to himself and several other black men who had also left the scene.
Jackson felt profiled, but his case did not appear in the Columbus stop data, which didn't include pedestrian stops.
"I was incredulous," Jackson said. "Like, wow, I haven't even been out of prison an hour and here I am encountering the police. I was telling my daughter's mother yesterday: I'm a magnet for the police."
In Columbus, a similar racial disparity as in Cincinnati emerged in the police stop data.
Reporters looked at vehicular police stops from 2012 to 2016. In 2016, the population was 28% black, 58% non-Hispanic white, and the rest Asian, Hispanic, or unknown.
Yet blacks made up a much higher share of those pulled over, compared to the population. Police pulled over blacks 44% of the time, whites 48% of the time, and 8% for others.
Considering the racial makeup of Columbus, blacks were stopped at an 89.8% higher rate than whites.
Between 5:08 and 9:06 p.m. in Columbus, (excluding twilight) street stops for blacks were 48% in the darkness, and the same rate in daylight. Non-hispanic whites made up 45 of darkness stops, and 47% of those stopped in the day. (Asians, Hispanics, and mixed made up 7% in darkness, and 5% during the day.)
On the highways, during that same time period, non-Hispanic whites ranged between 57% of those stopped in darkness, to 61% in daylight. During that same time period, Blacks were pulled over at roughly the same rate, from 34% in darkness, 33% in daylight. (Other races made up the remaining 6-to-8%.)
Again, police made more stops altogether in black neighborhoods compared to white ones. In Columbus, police made 84% more total stops per person in areas that were at least 75% black.
Columbus' Police Spokesman, and Diversity Inclusion Liaison Sgt. James Fuqua, said the force tries to allocate officers by where they get a lot of calls on a particular day. There are no particular rules governing which areas officers are required to be in, and no guidelines on how to provide equal coverage.
"We always encourage people to reach out to their community liaison if there's an issue," he said.
Police previously added different shifts, particularly more nights and weekends, in 2010. In 2017, they proposed redistricting, but never adopted the proposal.
Reporters later shared the results of their analysis with the police force and asked if they would make any changes. Fuqua said Police Chief Thomas Quinlan, who had served as interim police chief from April 2019 until being named permanently on Tuesday, has plans to shift resources.
In Columbus, blacks were arrested more often as well. About 5.5% of blacks pulled over were arrested, compared to just 2.9% of whites.
Blacks were also searched at a much higher rate in Columbus. Of blacks pulled over, police searched 9.6% in Columbus. In contrast, only 4.8% of non-Hispanic whites were searched. Hispanics had the highest search rate, at 13.7% although they are a very small slice of the Columbus population overall.
However, police arrested suspects after searching them at about the same rate, 61% for blacks and 60% for whites.
Altogether in Columbus, of all those arrested, 59% were black, compared to 34% white. Those searched were 60% black, and 33% were white. (Cincinnati data did not include if a search was conducted for any stop. And though requested, none of the cities provided data on whether contraband was actually found.)
Former Columbus Police Chief Kim Jacobs disagreed with the idea that some areas of the city are overpoliced.
"That's not how we look at calls for service," she said. "We're not looking at the race of the victim or suspect."
Jacobs said instead of putting officers in different locations, a better solution would be better training for officers, so they know how to respond appropriately in a situation. She said officers don't take the severity of a particular crime into account, they simply respond to all calls.
More tickets, more arrests for Cleveland blacks
In Cleveland, reporters looked at 47,079 traffic tickets in 2016 and 2017 and saw a similar pattern to Columbus and Cincinnati.
According to Census estimates, in 2017 Cleveland's population was 40% white, 50% black, 2% Asian, and the remaining 8% Native American, Pacific Islander, multiple races or unknown. Yet blacks made up 59% of the traffic tickets given out by police. Whites comprised 36% of those ticketed.
Considering the racial makeup of Cleveland, blacks received traffic tickets at a 31.1% higher rate than whites.
Reporters looked at traffic tickets issued between 4:58 p.m. and 9:05 p.m. (excluding twilight) and noted if each stop happened in daylight or darkness, because of the changing seasons and daylight savings time.
When looking just at highway stops, police ticketed white drivers 59% of the time in darkness, and 56% of the time in daylight, compared to 37% of blacks stopped at night, and 38% during the day. (Others made up 4% at night, and 6% the day respectively.)
When looking at neighborhood stops in that same timeframe, though, the opposite happened: Black drivers made up 59% of stops in darkness and 60% of those stopped during the day. Whites made up 36% of stops in darkness, and 35% of stops during the day. (Asians, Native Americans, and others made up 4% at night, and 5% during the day.)
In other words, when police couldn't see drivers whizzing past them, on a highway, they consistently pulled over a high number of white drivers. On surface streets, they consistently pulled over more black drivers, whether they could see them or not.
Again, the answer could lie in the initial placement of officers. Cleveland issued 26% more tickets per person in neighborhoods that were at least 75 black.
In addition to being ticketed at a higher rate, blacks were more likely to be arrested as well. Nine percent of all blacks pulled over are arrested, whereas only 6% of all whites pulled over are arrested. Altogether, blacks made up 70% of all traffic ticket recipients who were also arrested. Whites represented 28%, and the remaining 2% were Asian, Arab or unknown.
When initially asked about the placement of officers, Cleveland's Detective David Gallagher said officers are deployed strategically throughout the five neighborhood districts and within specialized units based on personnel analytics and crime analysis. He also said their deployment and staffing plans are constantly reviewed and updated.
Reporters later shared the results of their analysis with the police force and asked if they will make any changes. Sgt. Jennifer Caiccia said police staffing levels are based on need, populations, and the crime rate.
"The department already implemented training across the board in implicit bias," she said.
Reporters from the nonprofit newsroom Eye on Ohio, The Cincinnati Enquirer and researchers from Stanford University's Big Local News program examined police stops to assess how the three largest communities in Ohio use public safety resources and to identify potential bias in policing. The Plain Dealer of Cleveland contributed to this story.