Andy Leisher didn't like what he was seeing on the security cameras from his post at the front desk of the Ramada Inn in Janesville, Wisconsin. On the closed-circuit television in front of him, Leisher watched as a man in his 30s kissed what appeared to be a teenager in the motel hot tub.
It put him on alert. "It just felt awkward," Leisher said of the scene. "She just seemed really young, and he seemed really old. Or too old to be with her."
When Leisher, a part-time pastor, confirmed that the girl was 16, he called the police. A few hours later, police arrested 31-year-old Bryan Kind, and he was charged with having sex with a child and possession of child pornography. He's pleaded not guilty.
After collecting Kind's cellphones, Janesville police also found naked photos of a girl from Maryland, and they sent the information to authorities there.
It wasn't news to the Baltimore County Police Department. About a month before the May 2017 arrest, the department closed its investigation into Kind on allegations that he had sex with a 13-year-old girl.
It went on Baltimore County's books as a success, another rape case cleared.
But Kind had walked free. He wasn't charged with any crime. The Police Department hadn't arrested him, even though it had a thick investigative file on him.
Across the country, dozens of law enforcement agencies are making it appear as though they have solved a significant share of their rape cases when they simply have closed them, according to an investigation by Newsy, Reveal from The Center for Investigative Reporting, and ProPublica, based on data from more than 60 police agencies nationwide.
They are able to declare cases resolved through what's known as exceptional clearance. Federal guidelines allow police to use the classification when they have enough evidence to make an arrest and know who and where the suspect is, but can't make an arrest for reasons outside their control.
Although criminal justice experts say the designation is supposed to be used sparingly, our data analysis shows that many departments rely heavily on exceptional clearance, which can make it appear that they are better at solving rape cases than they actually are.
Because exceptional clearance data is not readily accessible to the public, we read through hundreds of police reports and sent more than 100 public records requests to the largest law enforcement agencies in the country. We analyzed data from more than 70,000 rape cases, providing an unprecedented look at how America's police close them.
Nearly half of the law enforcement agencies that provided records cleared more rapes through exceptional means than by actually arresting a suspect in 2016, the data analysis shows.
The Baltimore County Police Department, for example, reported to the public that it cleared 70 percent of its rape cases in 2016, nearly twice the national average. In reality, the department made arrests about 30 percent of the time, according to its internal data. The rest, such as Kind's case, were exceptionally cleared.
About a dozen departments that provided data had twice as many exceptional clearances as arrests in 2016. To the public, this effectively made it seem as though they had solved three times the number of rapes that they actually had.
For example, the Oakland Police Department in California cleared 60 percent of rapes reported in 2016, according to agency data. For every case they resolved through arrest, Oakland police cleared more than three by exceptional means, data provided by the department shows.
In Hillsborough County, Florida, home to Tampa, the Sheriff’s Office cleared 12 percent of rapes in 2016 by arrest. It cleared more than three times as many by exceptional means.
In Austin, Texas, where two out of three cleared rape investigations were closed by exceptional means in 2016, Police Chief Brian Manley said the high numbers are driven by the fact that so few victims decide to cooperate with police.
"It's the unfortunate reality of sexual assault in this country," he said.
Officials from Baltimore County, Oakland and Hillsborough County declined to comment about their exceptional clearance rates.
Cassia Spohn is the director of Arizona State University's School of Criminology and Criminal Justice and a co-author of seminal research on exceptional clearance. She said agencies' overuse of exceptional clearance is "misleading at best and duplicitous at worst."
"The public is concerned about the degree to which the police are able to arrest individuals who commit serious violent crimes," Spohn said. "And if the vast majority of those clearances do not involve the arrest of a suspect, what does that say about that law enforcement agency?"
* * *
In November 2015, the stepfather of a 13-year-old girl from the suburbs of Baltimore went to the local Child Advocacy Center, run by the Police Department, to report that Bryan Kind, a 29-year-old man, was using a mobile app to chat with his stepdaughter in a way that he found alarming and inappropriate.
The case that emerged would become an example of how police departments close investigations by exceptional clearance without having fully exhausted all available avenues. It also represents a case in which the classification was used improperly.
The stepfather's report originally was made to Cpl. Stacey McDaniel of the Howard County Police Department, who took on the investigation. Over the course of nearly a year, she collected evidence and hours of videotaped interviews in an effort to support the claim that Kind was having sex with the teen.
A review of the police file shows that although the girl initially denied having sex with Kind, McDaniel was patient in developing a relationship with the girl over time. After three conversations over about seven months, the teen eventually felt comfortable sharing details about their relationship with the detective.
McDaniel learned that in the summer of 2014, when the girl was 12, she started using a mobile app called Kik to message Kind, whom she'd met online. Their online romance flourished based on what she says were shared interests, such as cooking and car shows, which the girl watched on YouTube.
They both also liked Shakespeare. "We would always relate ourselves to Romeo and Juliet. It's like, 'When you're here, oh everything's fine and I would die without you,'" the teen told McDaniel.
When the teen's parents found out about her online chatting with Kind, they took away her phone and messaged Kind to tell him to stop talking to their daughter. But Kind began sending her handwritten letters, she told police. For a while, the girl borrowed her friends' phones at school to chat with Kind until he sent her a new one, which she kept under her mattress.
Bryan Kind, 33, appears in court in Rock County, Wis., facing charges of sexual assault and child pornography.
In June 2015, Kind offered to drive 10 hours from Gladwin, Michigan, to see the girl. About a month later, the teen said he visited again and they went to Wendy's for lunch and then to a motel, where they had sex, according to the police report. Afterward, they went to a pizza place nearby and he dropped her off at her house. They met up again at the motel the next day and had sex again, she said.
Almost two months later, Kind returned to see the girl, who cut middle school to meet him. They had lunch at Applebee's and played mini golf. Then they went to the same motel as before.
In addition to earning the teen's trust, McDaniel also collected emails that Kind and the teen had sent each other and found the motel where they allegedly had sex. The detective subpoenaed the guest registry for the motel to show that Kind paid for a room in cash on the three days the girl recalled him taking her there.
But it turned out that the motel where the alleged crimes had taken place was outside McDaniel's jurisdiction. About a year into the investigation, she was forced to turn it over to the neighboring law enforcement agency, the Baltimore County Police Department. Howard County police declined to comment or make McDaniel available for an interview.
Despite receiving a thorough and documented investigation, Baltimore County police let the case falter. The case file went to Detective Dana Kaczynski, who did not interview the suspect or attempt to arrest him.
Detective Dana Kaczynski of the Baltimore County Police Department in Maryland exceptionally cleared Bryan Kind's case after it was referred from a neighboring jurisdiction.
His efforts were focused on reaching the girl’s family. The detective called her mother, who told him that her daughter was doing well, and she would speak with her family about moving forward with the investigation.
Over the next six months, Kaczynski tried calling the teen's mother several more times. He never heard back, so he sent a letter.
"As I have investigated these types of cases for many years, I understand the impact and stress on the family," the detective wrote in April 2017. "Since I have not heard from you, my intention is to suspend my investigation for now. It can, however, be reopened in the future, if desired."
About two weeks after sending the letter, the detective closed the case, and on the final page of the police report, it says the case was cleared by exceptional means.
In doing so, the department was able to present the case to the public as solved.
In recognition that crime investigation is complicated, law enforcement is able to forgo making arrests — even when officials have enough evidence to do so and know who the suspect is — if there are specific circumstances out of their control.
These exceptional circumstances include situations in which the suspect dies or already has been arrested elsewhere. In rape cases, the most common reasons police clear cases this way are because a victim won't cooperate or a prosecutor declines to take the case forward, experts say. Police also are supposed to have probable cause to make an arrest.
In the Kind case, there was a known suspect and evidence, but a victim who did not want to cooperate with police. When Baltimore County reviewed the case at our request, spokesman Shawn Vinson said the department in fact didn't have probable cause to make an arrest. This would make the case an improper use of exceptional clearance.
Instead, the department could have suspended the case or left it open for further investigation. In either scenario, it wouldn't have been able to present it as cleared.
The Baltimore County Police Department did not make Kaczynski available for comment.
Ultimately, the suspect went on to be accused of having sex with another underage girl in Wisconsin.
Tom McDevitt is a retired lieutenant with the Philadelphia Police Department who helped develop a national model for handling sex crimes. He reviewed the Kind investigative file generated by both police departments in Maryland.
He said more could have been done before closing the case by exceptional means. For example, either department could have interviewed the suspect to see whether they could elicit a confession or gotten a search warrant to seize his phone and computer.
"You've got to look at these cases, that every single one of them has a potential to be a serial rapist," he said.
Vinson acknowledged that the Baltimore County Police Department "could have taken extra steps in this case," though he would not say what those specific steps should have been.
He said the agency reopened the investigation following our queries into how it was handled. However, the victim still has not expressed a willingness to participate in the investigation or prosecution, Vinson said, so the case has been suspended.
Kind did not respond to requests for comment directed to him or his attorney about the allegations made against him in Wisconsin or Maryland. He told a detective in Wisconsin that he didn't have sex with the girl there.
Beyond the Kind case, there are wider concerns about the way the Baltimore County Police Department is handling rape cases. A lawsuit filed in September by two women alleges that the department has tried to put an overly positive spin on its clearance rates by misclassifying rape cases as either a "suspicious circumstance" or closing too many cases by exceptional means.
The Police Department has not yet filed a legal response to the allegations related to how it closes cases and declined to discuss its use of exceptional clearance, citing the pending lawsuit.
* * *
In Austin, the Police Department touts its high clearance rates to illustrate to the public and local politicians that it is effective in solving crimes.
In June 2016, Brian Manley went before the City Council to ask for more money for his department. Using slides and statistics, he showed how the department's case clearances for violent crimes, including rape, "far exceed" the national average. He also noted that clearance rates are a "very critical and appropriate measure of our performance."
At a February 2017 meeting, a city councilman asked Manley — then the interim police chief — how he should interpret police clearance rates.
"Clearance rates show the effectiveness of the work we're doing," Manley said. "So if we had a lower clearance rate than the national average, or if we saw a drop, then maybe we don't have enough resources dedicated to a certain unit."
"We want a high clearance rate," he said.
But Sgt. Elizabeth Donegan said the department's use of exceptional clearance provides a false impression to the public. She had a firsthand view of how the department operated as a supervisor of its Sex Crimes Unit for nine years.
Her superiors pressured her to close more rape cases by exceptional means, she said in an interview.
Liz Donegan, a retired sergeant for the sex crimes unit of the Austin Police Department in Texas, said she refused to change the statuses of old rape cases from "suspended" to "exceptionally cleared."
"I had been told on two different occasions from the same commander under two different lieutenants that I needed to go back in and look at these cases that were suspended and change the clearance code because we were not up to the national average of exceptional clearance in Austin," she said.
Donegan says she never changed the clearance numbers as she was asked.
"It gives a false sense to the community that this case has been thoroughly investigated and it's closed," said Donegan, who retired last year as a sergeant in the Sex Offender Apprehension and Registration Unit. "It's not truthful."
In 2011, Donegan was transferred out of the Sex Crimes Unit. The next year, the share of rapes cleared by exceptional means jumped more than 50 percent, our data analysis shows.
Donegan says there's a culture within policing that overemphasizes closing cases because these numbers are presented to the public, and they're the metric by which the department and police chief often are judged.
A nationally recognized trainer on sexual assault investigations, Donegan said clearance rates are the wrong measure for judging good police work. "All of these numbers at the end really mean nothing if we haven't done a thorough investigation," she said.
The Austin Police Department acknowledged that our analysis, which found the increase in exceptional clearances between 2011 and 2012, is accurate. The department said the rates jumped after it discovered some cases were being misclassified as suspended instead of exceptionally cleared because victims didn't cooperate with the investigation.
Austin Police Chief Brian Manley bragged to the City Council in 2016 about his department's high clearance rates for violent crimes, including rape.
"We are responsible for reporting appropriately, so it was the right thing to do to make that change," Manley said.
Manley said that just prior to our interview, he confirmed with the Texas Department of Public Safety that his department is using the correct criteria to clear cases.
State officials, however, said they had not reviewed the Austin Police Department's cases and merely provided the department with a definition of exceptional clearance.
As for Donegan's claims that she was asked to reclassify cases, Manley said he believes she is trustworthy, but there was a "difference of opinions on what the appropriate way to clear cases were," Manley said.
In the past six years, the Austin Police Department has exceptionally cleared more than 1,400 rape cases, according to our analysis. Manley confirmed the findings.
Austin's handling of sexual assault cases is currently under scrutiny after a class-action lawsuit called into question the operations of both the Police Department and district attorney's office.
Marina Conner told police in Austin, Texas, that a man slammed her head into a wall and raped her following a night out with friends. The district attorney did not prosecute the man, citing a lack of DNA evidence.
Marina Conner, a student at the University of Texas at Austin, is one of the women who brought the case. She says she was raped in 2015 in a parking garage by a man who tried to sell her drugs after a night out with friends. She was drunk, and he slammed her head against a wall before raping her, she said.
"I remember crying," she said. "I remember saying, 'Stop,' and 'No,' and I remember the pain mostly."
She had a sexual assault forensic exam done. On the form, under "Impressions from the exam," the examiner wrote "sexual assault," among other findings.
Conner reported the crime to the police. Law enforcement officials found her alleged assailant, and he told them that he'd had consensual sex with Conner.
Her case eventually was rejected by the district attorney.
The Travis County district attorney's office told Conner that it turned down her case because there was no DNA evidence found in the rape kit, she said. She said prosecutors were concerned about the "CSI effect," or the expectation by a jury that DNA evidence is necessary to get a conviction.
Margaret Moore, the district attorney in Travis County, declined to discuss Conner's case, citing pending litigation. In a statement, Moore said: "Probable cause to arrest does not equate with sufficient evidence to go forward with prosecution in any type of complaint."
When prosecutors decline cases in which police have identified a suspect and have probable cause to make an arrest, the police can exceptionally clear the case. That's how the Austin Police Department classified Conner's case in 2017. In doing so, it was able to put forward to the public that her case had been solved.
When we told Conner how her case was closed, she was outraged.
"It sounds like a good thing if you tell someone a case was cleared — it doesn't sound like I was violently raped and my rapist is still out there," she said. "Makes me feel like I am being silenced, makes me feel like they're trying to sweep rapes under the rug."
* * *
For decades, the federal government has been trying to transition to a more sophisticated crime tracking system that advocates say will give policymakers a better understanding of national trends. It also will bring greater transparency to how rape cases are cleared and why.
But the more refined system creates another way for departments to make rape cases disappear.
The National Incident-Based Reporting System now is in use by about 40 percent of law enforcement agencies, many of them small to midsize. It collects more granular data that allows users to distinguish exceptional clearances from arrests, and it requires agencies to explain why they are closing cases that way.
But this more robust system doesn't track cases classified as unfounded, which means police have deemed reports false or baseless. Law enforcement agencies have been criticized for misusing this designation for decades, leading to scandals in Baltimore, Philadelphia, New Orleans and elsewhere.
The federal task force that laid out recommendations in 1985 for a new version of the FBI's Uniform Crime Reporting system specifically called for the continued tracking of unfounded cases. Neither the Bureau of Justice Statistics nor the FBI could explain why the recommendation was not followed.
Erica Smith, a unit chief for the U.S. Department of Justice's Bureau of Justice Statistics, said she was unaware that the FBI stopped tracking a key metric police agencies have used to take rape cases off the books.
Erica Smith oversees the unit at the Bureau of Justice Statistics dedicated to increasing reporting to and implementation of the more sophisticated crime statistics system. She said dropping unfounded cases from the data collection is unacceptable and she will work to get the federal government to correct the problem.
"This was not on my radar at all," Smith said. "If we're losing a really critical piece of information in that process, I will not have done my job appropriately."
Our investigation found that the updated system routinely reports zero unfounded reports for the police agencies that use it.
When we obtained records directly from 19 law enforcement agencies, beyond what they report to the new system, we found that seven had unfounded case rates above 10 percent. That's higher than what previous research has shown to be the national average.
The Prince William County Police Department in Virginia, for example, showed no unfounded cases in the government's updated system in 2016. However, internal department records show that it classified nearly 40 percent of all rape cases as unfounded, the highest rate of any of the 19 departments from which we received records.
In an interview, Prince William County Police Chief Barry M. Barnard acknowledged his department had an unusually high rate of unfounded cases. As a result of our queries, he has asked for a formal review of 39 cases designated as unfounded in 2017.
More than 10 percent of rape cases had been misclassified as unfounded, the review found. However, he said he is confident that the cases were investigated properly.
"I think we have some work to do," he said. "When should a case be made unfounded? When should it not? When should it be cleared this way? When should it be kept active? I think we need to put those definitions into our sexual assault response policies. And we need to train our staff, and then we're going to have regular reviews of our unfounded cases."
The FBI has not responded to multiple requests for comment on the apparent flaw in its data collection system for unfounded cases.
Experts say they're concerned that the way data is collected is distorting how police approach rapes and other crimes.
"The act of data collection is shaping behavior," said David Jaros, a law professor at the University of Baltimore and a former criminal defense attorney. "So much of how sex offense cases are handled is based on the incentives of the actors dealing with them."
Jaros said there are organizational pressures for law enforcement agencies to make their crime and clearance rates look good, increasing the appeal of exceptional clearances and labeling cases as unfounded. Prosecutors' offices are equally under pressure to keep their conviction rates high, even if it means passing on more cases than they should, he said.
Corey Rayburn Yung, a law professor at the University of Kansas who studies criminal law procedure and sex crimes, said "statistics-driven policing" has created scandal after scandal when it's discovered that law enforcement agencies have gamed their numbers related to how they process rape cases.
Yung said rape is particularly susceptible to manipulation because it is underreported and doesn't have the same kind of external checks that other major crimes do. Murder can be checked against vital health statistics, for example, and robbery can be tracked independently using insurance data.
The overuse of exceptional clearance is just the latest example of how law enforcement can make rape cases go away, he said.
"This pattern has happened over and over," Yung said of rape crime statistics subterfuge. "Usually, the department cleans house, brings in new people, and it happens again."
Scripps Howard Foundation Journalism Fellows Kenny Jacoby and Sophie Chou contributed to this story.
The reporters can be reached at:
Follow them on Twitter:
CORRECTION, Nov. 30, 2018: This story originally stated a government task force recommended that the Bureau of Justice Statistics continue to track unfounded cases. While BJS was involved in the task force, the recommendations did not specify the BJS should collect the data. The Uniform Crime Report is currently administered by the FBI.