By Cheryl W. Thompson, Ira Chinoy and Barbara Vobejda
Washington Post Staff Writers
Sunday, December 3, 2000; A01
In the District of Columbia, more and more people are getting away with murder.
Fifteen hundred homicides have gone unsolved over the past decade. Nearly two-thirds of the homicides that occurred in 1999 remained unsolved at that year's end, the poorest performance in the last 10 years.
A year-long Washington Post investigation has found fundamental flaws in D.C. homicide cases: poor supervision of detectives scattered in districts across the city, hundreds of missing and incomplete case files, and dozens of cases closed without arrests under unclear circumstances. The disarray has persisted despite repeated promises of fixes, a precipitously dropping homicide rate and the arrival two years ago of a police chief brought in to reform a troubled department.
Nothing illustrates the depths of the department's problems more clearly than the handling of homicide case files, one of the most important measures of a department's "investigative integrity," according to an internal department memo.
Police administrators vowed three years ago to make improvements after the discovery that 600 files were missing. But police are currently scrambling to locate 377 closed cases that they cannot find in their headquarters, according to internal police documents.
Police Chief Charles H. Ramsey said many of the files are in the possession of detectives or at the U.S. attorney's office and will ultimately be found. But in the files police could locate, 136 lacked crucial documents. And Post reporters did their own examination of 100 homicide cases that had been closed "administratively" without arrests between 1996 and 1999 and found that 29 of them did not have the documents explaining why those cases were closed, in violation of police policy. Many of these files also were missing crime-scene technician reports and witness statements. One case consisted of a two-page report detailing the facts of the death and nothing else.
"Nobody's holding anybody accountable for what they're doing with these cases," said David Schertler, former chief of homicide prosecutions for the U.S. attorney's office. "The leadership in the department has really failed to put into place the kind of structure you need to have quality investigations."
D.C. police have an arrest rate that has fallen sharply lower than that of other cities The Post examined.
Last year, for example, D.C. police made fewer than 60 arrests per 100 homicides, compared with an average of more than 90 arrests per 100 homicides in four other large cities--Los Angeles, New York, Philadelphia and St. Louis--from 1995 to 1998, according to D.C. police and FBI statistics. The arrest totals include cases from prior years and reflect that more than one person can be arrested for a single homicide.
The consequences in the District can be seen on the city's streets. Although cause and effect are hard to establish, scores of suspected murderers have been left free to kill again and others have been stopped only when they themselves were killed. In 3,900 homicide cases over the past decade, at least 200 killings were blamed on persons believed by police to have committed at least one earlier homicide. At least 150 alleged killers were themselves killed. And at least 17 witnesses were killed, according to a database of police homicide cases.
A succession of D.C. police chiefs has promised to improve homicide investigations during the past decade. Four studies by outside consultants have recommended fixes. In a series of administrative shake-ups and reorganizations, seven homicide commanders have come and gone. But the department's closure rate has continued to sink.
Not long after he took over the police department in April 1998, Ramsey moved the homicide detectives into the seven district stations from their centralized headquarters downtown. Rather than specializing in homicides, the detectives were assigned to be "violent crime" investigators who are also responsible for rapes and robberies. Several detectives who had been assigned to investigate old homicide cases were redeployed to the street to increase police visibility.
A number of homicide detectives said in interviews that these moves have reduced their effectiveness, especially because homicide expertise has been dissipated and many supervisors have never investigated homicides themselves. Only 11 of the 33 supervisors overseeing homicides have experience investigating such cases.
"You get people tossed into handling homicides who know nothing about the details of homicide investigations," said Detective Michael Baylor. "They're finding their way, and while they're finding their way, things are getting lost."
Working out of seven districts, the detectives say they find it difficult to share information and concentrate their investigative strength.
"We know decentralization is absolutely wrong," said 23-year veteran Eric Gainey, a 6th District detective. "I'm sure they realize this is a bad idea, but they won't do anything about it."
When they were based together at headquarters, Gainey said, as many as 10 detectives might mass at a crime scene. Today, only two or three detectives normally work a homicide, he said. "What are you going to do with two people at the scene of a murder?" Gainey asked. "Absolutely nothing."
Although speedy investigation is a key to solving a homicide, one veteran detective said he sometimes comes to work on a weekday to find a homicide case from the weekend waiting on his desk. "Some shifts you have no homicide detectives," he said. The staff in the 4th District along Georgia Avenue NW, he said, "is spread so thin, you might not get a detective if there's a murder."
Ramsey defends his decision to decentralize homicide and says he is sticking by it.
"It's not a question of recentralization and decentralization," he said. "It's the quality of investigation that needs improving."
Last summer, a D.C. Freedom of Information Act request filed by The Post turned up scores of incomplete case files. That discovery led Ramsey to order a review of every homicide case dating back to 1990.
So far, Ramsey's reviewers, four retired detectives, have been unable to locate hundreds of files and have found more than a hundred that are missing documents, according to internal police documents. Nine cases have been reassigned to detectives for further investigation, he said.
"We take this very, very seriously," Ramsey said. "There's no excuse to have even one case that we can't put our hands on."
Ramsey also said he is launching a number of initiatives to improve the homicide department, including new protocols and procedures, added training, regular reviews by supervisors, an audit unit for case files and an upgraded computer database system.
"I intend to make sure we put something in place where we can start having the checks and balances we need to be sure things are being done in a consistent way," Ramsey said in an interview last week. "A sergeant now has no excuse for handing me a case that's two months old and has four pieces of paper in it."
But the problems Ramsey confronts run deep--consultants described them in detail in 1996--and promises to fix things have been made before.
"There was a time this city had the best homicide department probably in the country," Ramsey said in an interview with Post reporters last summer. But now, he said, the detectives are weak in fundamentals, "basic blocking and tackling," such as writing reports for the files.
"We don't do fundamental stuff," the chief said. "One of the things we keep telling our detectives is you've got to put it on paper."
Homicide detectives said that their current database for storing homicide information, the Washington Area Criminal Intelligence Information System, is not as effective as it should be because detectives use it haphazardly and often fail to enter vital information. "They do not force people to use WACIIS, and it's not secure," said one veteran detective.
In addition, the detective said, "a lot of our detectives don't write well. That's one of our biggest problems. It's gotten to the point where [prosecuting] attorneys are writing most of our affidavits and arrest warrants. A lot of detectives are glorified process servers, and the prosecutors do 90 percent of the investigating."
An illustration of the problems found in D.C. homicide cases can be seen in the file of construction worker Theodore Tanner, who died on April 12, 1997, after he was stabbed in the chest in his Northeast apartment. His girlfriend said he fell on a steak knife. But the girlfriend gave Tanner's relatives "varying stories of the incident," police reports show. And a neighbor had heard the couple arguing.
"I don't think there was any doubt in my mind that she caused the murder," Detective Jose Solloso, who investigated the case, said recently. "But we couldn't prove it in a way the U.S. attorney's office felt comfortable."
Solloso said police closed the case administratively after the U.S. attorney's office declined to prosecute for lack of evidence. But the report required to close the case is missing from the Tanner case file. And Solloso doesn't recall attending an autopsy, a procedure that could be critical in distinguishing an intentional stab wound from an accidental one. Solloso also acknowledges that he did not videotape a police interview with the girlfriend, which experts say can be crucial in determining credibility.
"This is a case that should have been closed with an arrest," said William L. Hennessy, a former D.C. police captain who oversaw the homicide unit from 1993 to 1995. "This is not a mystery."
Other problematic cases--seemingly inexplicable closures with key documents missing--were turned up by The Post and also in the review ordered by Ramsey.
Consider the closed case of Jacqueline Birch, whose body was found in Northwest Washington on Nov. 18, 1997: "There is no indication in this [case] jacket that the case was closed," detectives reviewing the files for Ramsey said in an Oct. 13 internal department memo obtained by The Post. "Furthermore, there is no indication of any offenders being questioned or arrested."
In the closed case of Eddie Waller Jr., shot several times on Sept. 26, 1997, in Northeast Washington, the reviewers wrote: "Despite the fact that there was an eyewitness to the shooting and friends and family of the victim provided information to the police about the offender(s), it appears that the shooter was never arrested. There are no investigative reports in the case jacket detailing any attempt to identify the shooter."
Former detective Carl Gregory, who retired in 1997, attributed missing documents in cases to the fact that "the homicide unit went neglected and understaffed for years. I'm shocked that we were able to close as many murders as we did."
Before homicides skyrocketed in the 1990s, things were different. Another retired detective, Joseph Quantrille, who left the homicide squad in 1989, said detectives in his era worked efficiently in teams on homicide scenes, bringing in witnesses for questioning, combing neighborhoods for evidence, attending autopsies and taking copious notes. "It really was meticulous work," he said. "It's not meticulous anymore."
Closing the Gap
Under public and internal pressure to solve more homicides during the 1990s, D.C. police relied increasingly on cases closed administratively without arrests. Police general orders permit such closures in certain circumstances--when a suspect is dead or already in jail.
By 1997, administrative closures accounted for one of every four homicide cases closed in Washington. That was at least double the rate of such closures at the beginning of the decade.
Administrative closures trouble some detectives because they are made with evidence that is never tested in court.
In examining 100 administrative closures between 1996 and 1999, The Post in several cases found little or no evidence in the files linking a suspect to the crime. In others, police appeared to have enough evidence to make an arrest but failed to do so, allowing months and sometimes years to pass, only to quickly close the case after the death of a suspect.
When Deon Wilson was killed during a robbery while walking with a friend in Southeast Washington in September 1996, detectives quickly got a description and the nickname of one of the two suspects, the triggerman, "Steve from Bowie." But there was no arrest.
Four months later, the suspect was killed in a shooting. The next month, a witness identified him, Steven Terrell, as the person he had "heard on the street" as the man who shot Wilson. Eight days later, detectives recommended that the Wilson case be closed although "no 100 percent identification was made."
Detective Baylor, who investigated the case, said, "You can have weak information, but request that it be closed [administratively without an arrest]. If you interview people who said they were told who the killer was, you can request that it be closed [administratively]. We only make the recommendations. The officials close it."
Such closures look just as good as an arrest in the closure statistics, which are the primary measuring stick for homicide detectives and police chiefs.
Retired detective Quantrille said: "You can't imagine how desperate the brass is to raise their closure rate, and they'll do whatever they have to do to close cases. It's basically criminal the way they're closing cases."
Others have questioned the department's homicide bookkeeping.
A department employee responsible for entering homicides into police records said she was told by a top administrator not to count seven slayings in the ongoing 1999 tally.
"The homicide count in '99 is off," said Evelyn Noble, a police clerk who retired in September. "It came from Assistant Chief Brian Jordan, but the order came from [Executive Assistant Chief Terrance W.] Gainer's office. [Jordan] told me not to count them in the log or put them in the database. They wanted the homicide numbers to stay low."
Ramsey said there was confusion about how to count cases where people died in one year and the medical examiner's homicide ruling occurred in a different year. He said the correct number of homicides for 1999--241--was reported to the FBI, but the figure the department made public at the end of that year was too low.
"We screwed up on the internal count," he said. "There's no doubt about that."
Jordan said he didn't order Noble or anyone to alter the homicide numbers. "I have no knowledge of any homicides not being counted," he said. "If somebody didn't put them on there, then they should be dealt with."
Assistant Chief Alfred Broadbent, who took over the homicide unit from Jordan, acknowledged that there was "some confusion" about the numbers. "It was an administrative error," he said, adding the number has since been changed. "There was no intent to hide any homicides."'A Perfect Case'
D.C. homicide detectives have a difficult job under the best of circumstances in a city that has long had one of the nation's highest murder rates. A two-year federal study of D.C. homicide investigations completed in 1996 found many problems but also noted the "remarkable job" that detectives have done under "enormous adversity."
"Noncooperative witnesses, unmanageable caseloads, and limited fiscal resources are just a few of the problems the detectives combat every day," according to the report by the National Drug Intelligence Center. "In spite of these adversities, the detectives have been able to solve a majority of homicides."
Homicide investigators have had to deal with a lot: a record murder rate fueled by the crack epidemic in 1990 and 1991, the aftermath of an attack by a gunman who walked into police headquarters and killed a sergeant and two FBI agents in 1994, a financial squeeze and the loss of a popular and innovative commander in 1995, the firing of the entire homicide command in 1997 and the loss of veteran detectives through turnover. Meanwhile, a series of reorganizations has forced detectives to move back and forth three times in the past three years between headquarters and the districts.
D.C. police also are dependent on the U.S. attorney's office, which tries their cases in court. Officers complain that prosecutors won't sign arrest warrants and will decline to prosecute cases because they are afraid to lose. During a meeting with D.C. Mayor Anthony A. Williams in February, Ramsey criticized prosecutors for demanding "perfect cases." He said they did not have the "same sense of urgency" as police.
U.S. Attorney Wilma Lewis said she was "astounded" by Ramsey's comments and said her office must maintain its high standards.
"The fact that the closure rate is low is of great concern to all of us," Lewis said. "But we won't go forward with razor-thin evidence that won't withstand the scrutiny of a judge. We're not looking for a perfect case, we're just looking for enough to get the case to court."Moving Backward
Throughout the 1990s, police officials zigged and zagged in search of solutions for the homicide investigation problem.
In October 1993, after detectives complained in articles in The Post that they didn't have time to thoroughly investigate cases, then-Police Chief Fred Thomas doubled the number of detectives and appointed a new homicide commander, the fourth in three years.
William Hennessy, the new commander, started a months-long training regimen for new detectives. He created a system of "geographic assignments" where the same group of detectives, working out of the downtown headquarters, handled every homicide in particular districts in the city, developing a sense of the neighborhoods. He promoted teamwork and esprit de corps among the unit at headquarters, gathering detectives for weekly intelligence sessions with other law enforcement agencies.
The homicide squad's clearance rate rose while the homicide rate declined. In September 1995, the Justice Department called D.C. homicide a "model" unit.
However, within a month, Larry D. Soulsby, the new police chief appointed by then-Mayor Marion Barry, moved Hennessy out of homicide, citing poor management. The transfer later led to a highly publicized dispute between Soulsby and Hennessy over the reasons for the move. Despite acrimonious charges, a lawsuit and a CBS "60 Minutes" piece, the exact reasons Soulsby transferred Hennessy have never become clear.
One thing that is clear is that with Hennessy's departure, the homicide unit went into a downward spiral. "They regressed eight years when they moved Hennessy out of homicide," said former detective Gregory.
Soulsby decentralized half of the detectives, sending them into the seven police districts to get more connected to the neighborhoods.
In September 1997, Soulsby transferred Capt. Alan Dreher, Hennessy's successor, citing a low closure rate and poor management. Soulsby also purged 17 supervisors, saying a review by the Booz-Allen & Hamilton consulting firm found missing case files and languishing investigations.
"The bottom line was a lack of quality control," Soulsby said at the time.
The consultants had stumbled upon a report by the National Drug Intelligence Center studying all 1,827 D.C. homicide cases between 1991 and 1994 and making recommendations for improvement. Buried in the report, the details of which have not been made public until now, was the stunning fact that 613 files were missing: The files "were presumed to be either at the U.S. Attorney's Office, at a task force, or in the possession of a particular homicide detective," the report stated.
The report also found numerous cases listed as closed with no indication of who was arrested, and missing autopsy reports and witness statements. "Although much of this information had probably been collected by the detectives, it was not available in the homicide files, and accordingly, no permanent record of this activity exists," the report stated.
Chief Thomas had requested the report in 1993. It was finished in February 1996, after Soulsby had become chief. But it was ignored.
After news of the report surfaced in fall 1997, Soulsby promised more accountability and vowed to reinvigorate the homicide unit with more detectives, training and equipment, including a locked file room monitored by a security camera. He promised a "significant increase" in clearances within months.
But Soulsby soon resigned under a cloud--he was discovered paying a cut-rate price for a luxury apartment in the District.
An interim police chief, Sonya Proctor, brought the detectives back to headquarters from the districts. Proctor also appointed a new homicide commander, Capt. Ross Swope, who had never investigated homicides. Proctor and Swope were gone within a year.'It's Fixable'
In April 1998, Charles Ramsey was hired as police chief from Chicago to reform a troubled D.C. force. Within months, the new chief sent the homicide detectives back out into the districts. The case files scattered with them.
"When we decentralized, officers took cases out," Broadbent said. "This has been an ongoing problem."
Last summer, Ramsey had his officers look at 100 case files requested by The Post and then ordered a widespread review of homicide files. The reviewers found the same pattern of missing cases and documents that the drug intelligence center report turned up four years earlier. The report had been ignored a second time.
"Why that wasn't followed, I have absolutely no idea," Ramsey said.
Ramsey now believes his proposed initiatives will permanently fix the problem.
"People just don't always do what you want them to do when you want them to do it," Ramsey said. "Does it look bad? Certainly. But it's fixable."
When he spoke with Mayor Williams about homicide nine months ago, Ramsey said: "We have a lot of people in the detective ranks who don't have skills to be good detectives. We look ridiculous with very low closure rates."
He was bitterly ironic. He mused that killers had a better chance of facing justice on the street than in the courts. "The closure rate of thugs is a lot higher than ours. They close cases because they just kill you."
Database editor Sarah Cohen, staff researcher Alice Crites and staff writer Peter Perl contributed to this report.
Today: Despite promises of reform, homicide investigations in the District remain seriously troubled. By the end of 1999, police solved just over one-third of the homicides committed that year, the lowest rate in a decade, as they struggled with fundamental investigative flaws, missing files and inexperienced supervision.
Tomorrow: Police increasingly rely on administrative closures--cases closed without arrests. But the files often don't say how the cases were closed, and the families of the victims often are not informed.
Tuesday: As the quality of homicide investigation declined in the District, a culture of "street justice" took hold. More than 150 homicide suspects died on the streets during the 1990s.
Wednesday: When killers aren't arrested quickly or locked up on solid evidence, they often have the chance to kill again. In the past 10 years, at least 200 homicides were committed by people District police believe had killed before.