From the start, that promise was betrayed.
With pressure from the top to show results, too few investigators and a huge caseload, the unit became expert at getting rid of - or hiding - cases that could not be solved easily, an Inquirer investigation has found.
Year after year, hundreds of cases were dumped in "investigation of person," an obscure administrative category meant for situations where no crime was committed. Cases put there do not show up in crime statistics.
The victims typically had no idea their cases had been mothballed. Many never heard from police after an initial hospital interview. Some who tried to prod investigators said their calls were not returned.
Others said police seemed determined to poke holes in their accounts - to show them up as liars.
Roscoe Cofield, who retired last year after a 10-year career as a rape investigator, did not mince words in describing what happened to cases exiled to "investigation of person."
They got no further attention, he said.
"It meant, 'This is a nothing.' "
From the mid-1980s until last year, the squad put about a third of all the complaints it received in that bureaucratic dead zone - a "non-offense" code never intended to designate rapes or any other crimes. In 1996 and 1997 alone, police acknowledged last week, more than 2,000 cases wound up in "investigation of person."
They include two that, within the last two weeks, were linked through DNA testing to the man who has sexually assaulted five women and murdered one in the Rittenhouse Square area over the last two years.
Bureaucratic ruses such as "investigation of person" are only part of the story of how the rape squad polished its numbers at the expense of victims and the public.
In the early 1980s, police rejected as "unfounded" - or groundless - half the complaints they received, according to confidential FBI documents obtained recently by The Inquirer under the Freedom of Information Act.
That rejection rate was five times the national average.
The proportion of cases deemed "unfounded" fell when the unit, under pressure from the FBI to stop rejecting so many complaints, began disposing of tough cases in "investigation of person."
Two years ago, the FBI and departmental auditors questioned the use of the code. Sex-crimes investigators backed off "investigation of person" and again began "unfounding" large numbers of cases.
In a single year, the unfounded rate doubled - to 18 percent for 1998. That was twice the national average and the highest rate among the 10 largest cities in America.
What kinds of cases have been shelved?
The 1995 rape of a North Philadelphia woman who was bashed in the face on a playground.
The knifepoint rape in 1995 of a West Philadelphia woman in an abandoned building.
A 1997 assault on a woman who awoke in her Pine Street apartment sprawled on the floor, with her clothes stripped off, her throat bruised and swollen.
The attempted rape in 1997 of a Rittenhouse Square woman who woke up in the middle of the night to find a stranger on top of her.
After The Inquirer asked questions about the Pine Street assault, which the rape squad labeled "investigation of person," police conducted a fresh round of DNA tests two weeks ago on semen-stained underwear from the crime scene. The sex-crimes unit had handed the case over to regular detectives, who were investigating it as a burglary.
The tests showed that the woman had been assaulted by the Rittenhouse Square rapist.
The 1997 attempted rape was committed a month earlier and a few blocks away. The rape squad doubted the victim's story, put her case in "investigation of person," and stamped the file "inactive."
Ten days ago, police retested hairs from the crime scene and connected them to the serial rapist.
Police commanders do not dispute that the squad abused "investigation of person" - often referred to by its numerical code, 2701. During an interview earlier this year, Police Commissioner John F. Timoney called the practice "stupid."
But Timoney said the department would provide no information about sexual assaults that may have been miscoded before 1998, when he took over the force. Mayor Rendell said he supported the decision.
THE PUBLIC PAYS
THE PRICE OF SUBTERFUGE
Timoney said in an interview last week that his "complete reorganization" of the sex-crimes unit had ended statistical manipulation for good. He appointed a new commander, improved training, renamed it the Special Victims Unit, and assigned detectives to the squad for the first time.
"By and large, those 2701s . . . have almost disappeared," he said.
Yet the Inquirer found that the numbers game has continued in a new guise - albeit on a smaller scale.
The rape squad dramatically reduced its use of "investigation of person" last year. The wholesale dumping of complaints has been replaced by a more selective use of a different noncriminal code - "investigation, protection, medical examination."
Such statistical subterfuges exact a heavy price. The city has been misled about the unit's success at catching rapists. Crime statistics have been skewed to understate sexual assault in Philadelphia.
And in many cases, rapes have gotten little or no investigation.
When a crime report is stamped rape, it becomes an urgent matter for the department. The sex crimes unit will be expected to solve it or explain why it can't.
When a sexual assault lands in "investigation of person," it ceases to be a blinking light that demands attention and action. The department blinds itself and the public to the real extent of rape in Philadelphia.
George Pennington, a supervisor of the sex-crimes unit from 1981 to 1988, said that masking rapes behind noncrime codes, or whisking them into the "unfounded" column, was an entrenched practice - pervasive and unquestioned.
Pennington, now chief of police in Kline Township, northwest of Allentown, said bureaucratic sleight-of-hand offered a convenient way to put aside victims who didn't "fit a certain profile," who were not "people of substance."
Investigators would bury cases in code 2701, "investigation of person," to avoid traveling to a tough neighborhood to look for suspects, he said.
Many of the complaints relegated there were considered phony or too murky to sort out, Pennington said.
But sometimes, he said, the sex-crimes unit "would basically take a legitimate rape" and "code it '2701.' "
"I was part of the problem. I sent back 211s and 233s [the codes for rape and attempted rape, respectively] sometimes and told them to recode it 2701," said Pennington, who was the unit's chief statistical officer.
"They wanted the rapes down," Pennington said, referring to higher-ups. "You could get notes back from a lieutenant or captain - 'you knock it down, 2701.' To make your city look good, you would go under with sex-crime cases. Basically, it was public relations."
The shell game fooled the women's groups that lobbied hard to get the sex-crimes unit created, and that later tried to monitor its performance.
"I didn't know about this 'investigation of person,' " said Lynn Marks, one of the early leaders of the Philadelphia chapter of Women Organized Against Rape. "I think that limbo-land is not fair to citizens."
CHOKING ON BLOOD,
SHE'S RAPED ON A BENCH
Melody Madison knows limbo-land. The rape squad put her complaint there because she couldn't be reached by phone.
She is typical of the victims whose rape reports are most likely to be rejected or shelved. She has a history of drug and alcohol abuse and spent two months in prison for kicking and biting a housing authority policeman in 1990.
At 39, Madison looks older than her years. Her front teeth are broken. Her tongue stumbles. But on the night she was attacked four years ago, she said recently, "I was looking good. I was sharp. I had my white tights on."
Madison said she spent the evening of Aug. 24, 1995, at an after-hours club in North Philadelphia. She was walking home after midnight through a playground at Cecil B. Moore Avenue and 21st Street when she noticed a man from the bar following her.
The stranger offered her $10 for sex, according to a police report on the incident. Madison refused.
In a recent interview, she said the man snarled, "Bitch, you're going to give me something," and smashed her in the face.
Choking on blood and shattered teeth, Madison could not fight back. She said the man raped her on a playground bench, beneath a mural of abolitionist Harriet Tubman.
The assailant fled, leaving behind a pair of black sports shorts that he had worn over his trousers.
Madison staggered across Cecil B. Moore Avenue and called 911. She told police she had seen the man before, but didn't know him. She said he went by the name "Wild Bill."
A patrolman drove Madison to Thomas Jefferson University Hospital, a city-designated rape-treatment center. Records there show she received painkillers, an anti-pregnancy pill, and treatment for four dislocated teeth from a "blow to the mouth."
Later that night, in the 23d Police District stationhouse, supervisors classified the incident as a rape and forwarded it to the sex-crimes unit for investigation.
With her jaw wired shut and her head swathed in bandages, Madison was interviewed at the hospital at 5 a.m. by Roscoe Cofield, the veteran sex-crimes investigator.
He took down Madison's description of her attacker: black, 5-foot-6, 155 pounds, medium build, brown eyes, short black hair.
In addition to that information, Cofield had a nickname for the assailant, the name of the club where he drank, his black shorts, and documented injuries - the makings of a viable investigation.
His confidential investigation report details the various steps he took, including reviewing the crime scene and conferring with a supervisor. It makes no mention of any attempt to locate Wild Bill.
Madison has no phone. She gave police her sister's phone number and address. Cofield's report says he tried several times to phone her there after his initial interview, without success.
He coded the case "investigation of person - pending further contact and cooperation of complainant." A lieutenant approved the coding.
At that point, the investigation ended.
Four years after Cofield could not find her, The Inquirer located Madison. She is readily reached through contacts in the close-knit neighborhood at Ridge Avenue and 22d Street that she has long frequented - and where the attack occurred.
Reporters left business cards with merchants and a bartender. A week later, the bartender placed a call for Madison and put her on the phone.
Madison gave The Inquirer permission to examine her hospital records and to identify her by name in this article.
Kathleen Brown, an expert in rape forensics who teaches at the University of Pennsylvania School of Nursing, reviewed the medical records at the newspaper's request. Brown said evidence of a blow to the mouth was a strong indicator that sex was not voluntary.
"Most people would agree that being punched in the mouth is not part of consensual sex," she said.
Cofield, 56, now retired from the force, said in a recent interview that he could not remember the case. After looking at a copy of his report, supplied by The Inquirer, he said he had acted properly.
"What am I going to do? I can't track her down. . . . People's habits have a lot to do with the way you investigate," he said.
The lieutenant who approved Cofield's coding declined to be interviewed.
In several interviews, Madison told the same story she told Cofield that night in the hospital. She identified the rapist as Wild Bill and described him in detail, down to his black shorts.
Former Police Officer Joseph Repholz, who walked a beat in the area for 11 years and knew Madison well, said in a recent interview that he could have located her had the rape squad contacted him. Detectives in other units routinely reached out to him for help locating victims, suspects or witnesses, he said. That didn't happen in Madison's case.
Repholz, 64, who retired last year, said he encountered Madison regularly on his beat and knew her as a troubled but decent person. He was friendly with her even though he had arrested her once - on a theft charge that was later dismissed.
"She was good people," he said.
Repholz remembered seeing Madison's bandaged face after the attack and talking to her about what had happened.
He said he believed she was telling the truth.
In its report to the FBI of major crimes committed in Philadelphia in 1995, the Police Department did not list the attack on Madison - not as a rape, an assault or anything else.
"They should have stayed on top of it," Madison said. "It wasn't like I was lying. They seen the blood. And just like you all came over here and found me, they could've found me."
Madison's older brother, Marvin Leslie, 41, a hauling-company owner, said his sister deserved better from police.
"They just blew it off," Leslie said. "It was reported to them and . . . nothing."
He put a fist to his lips, blew sharply on his hand and quickly uncurled his fingers.
"It was just like, 'Pfft. Be gone.' "
A UNIT PRESSURED
BY THE NUMBERS
Current and former rape-squad investigators say the statistical trickery was forced on them by inadequate staffing, a steep workload, and the department's decades-old culture of manipulating crime numbers.
They describe a command structure obsessed with statistics and determined to make them shine - no matter the reality on the street. Pressure was intense, they say, to generate a favorable "clearance rate" - jargon for the percentage of cases that police solve.
One way to boost the squad's statistical performance was to keep tough cases out of the equation. When a case is "unfounded" or coded "investigation of person," it no longer counts against the clearance rate.
It is officially not a crime.
And so the motivation to find the perpetrator, or track down a phone-less victim, disappears.
The numbers strategy was wildly successful on paper. Year after year since the early '80s, the unit's clearance rate has been among the best of any rape squad in the country. In 1993, for instance, it reported "clearing" 74 percent percent of its caseload. The national average was 53 percent.
Propping up the clearance rate masked the unit's chronic understaffing. Last year, the squad's 60 investigators were assigned 4,100 cases of sexual assault and child abuse. That's 64 cases per investigator - 13 times the typical caseload for detectives in the Homicide Division.
Since 1980, rape units in Chicago, San Francisco and other major cities have come under fire for doctoring statistics. But Philadelphia's prolonged, heavy use of noncrime codes and its claim to one of the nation's best clearance rates suggest the problem is particularly severe here.
In preparing this series, The Inquirer analyzed crime data going back 20 years, police incident reports, emergency-room records, court files and FBI documents.
Reporters also obtained and studied 300 confidential investigative reports compiled by the rape unit in recent years.
The review found that from the mid-'80s until last year, the rape squad put about 30 percent of its caseload into the deep freeze of "investigation of person" or other noncrime classifications.
Last week, the department released data for 1996 and 1997. Of about 7,500 cases handled in those two years, 28 percent were classified "2701."
The numbers games began shortly after the squad was founded in 1981 amid promises from Mayor William J. Green that a new era had opened in the Police Department's handling of sexual assaults.
In 1982, the unit reported a 10 percent decline in rape. In 1983, the FBI, which compiles crime data for cities nationwide, noticed a possible explanation - the squad was "unfounding" more than half the complaints filed by women.
The bureau asked for an explanation - and noted pointedly that "unfounded" data, which at the time were not made public, soon would be. Philadelphia's "unfounded" rate fell abruptly in 1984. And cases began appearing, by the hundreds, in "investigation of person."
Chief Inspector John T. Maxwell, head of all Philadelphia detective units, explained what happened.
"We were scrutinized for having too many 'unfoundeds,' " Maxwell said in a recent interview. "That's when the department shifted from the 'unfounded' category to 'investigations.' We would use that category, because maybe comparing it to other cities or something, our 'unfounded' rate was high."
Instead of finding out why the squad was rejecting so many cases, the department simply found a new place to store them.
Code 2701 is typically used when patrol officers stop a pedestrian for questioning without making an arrest. It is a "non-offense code" - to be used in situations where no crime occurred.
Yet sex-crimes investigators found it useful. That's where they put cases that weren't going to be solved quickly - cases in which the victims were uncooperative or hard to reach or carried the baggage of mental problems, drug habits or criminal records.
Lots of other cases wound up there, too. Complaints from low-income women or those who lived in dangerous neighborhoods were frequently dumped in 2701, said Pennington, 59, the former rape-squad supervisor who is now a police chief in the Poconos.
"You'd try to take the slightest discrepancy to try to knock it down to 2701," he said. "If it was from a shady part of the city, who's going to complain? These people are from the inner city."
Complaints from prostitutes were also prime candidates for bureaucratic burial.
"Many a time, sex crimes blew it off that way," said retired Sgt. Barbara Cahill, 49, a member of the unit from 1990 to 1996.
Even when investigators believed a prostitute had been raped, Cahill said, they buried the case.
"Why? Because there's not enough time to do the job," Cahill said. "Because the investigator may believe they're chasing their tails. Because they believe the prostitute won't show for court. It's a dead end. You were putting a lot of legwork in when you wouldn't have a case."
Rape cases do fall apart in court, but not more often than other criminal cases. Of the 343 rape cases in the Philadelphia courts last year, 37 percent were dismissed by judges or withdrawn by prosecutors, court records show. For all other violent crimes, 56 percent of cases collapsed.
Code 2701 was an unofficial way to close cases that no one wanted to deal with. Carrying those cases on the books as rapes, albeit unsolved ones, would drag down the unit's clearance rate.
Commanders didn't like that, Cahill said.
"You were leaving an open case," she said. "They wanted a closed case. They wanted clearances.
"A lot of people 'went down' with rapes," she said, using police shorthand for downgrading crimes on paper.
Referring to 2701, she said: "It's an abused code, such an abused code."
FBI and Police Department rules do not permit such statistical dodges.
When a woman files a complaint, investigators are supposed to get the facts and make a judgment. Either she was raped or she wasn't. If, later, they discover information that puts the case in a new light, investigators can change the coding.
The basic guideline: Police should take a woman at her word that she was raped, unless there are compelling reasons not to.
A complaint can be "unfounded," under Police Department rules, only if investigation finds it to be "totally groundless."
CODE WAS NECESSARY
Though the department's current leadership has repudiated the heavy use of "investigation of person," former commanders of the rape squad defended it as necessary to shield the unit from unfair criticism.
They said the public and politicians were not prepared to accept that many rape complaints are false. If the unit unfounded all the complaints that ought to be rejected, there would be an uproar, they said.
They depicted "investigation of person" as their only option - a way to keep meritless complaints out of the official tally without inviting the ire of women's groups or the scrutiny of the FBI.
The former commanders also insisted that "investigation of person" was never used as a dumping ground for real rapes.
Clifford Barcliff, commander of the unit from 1983 to 1987, said Code 2701 was used exclusively for complaints that were obviously groundless. He cited a hypothetical case in which a woman reports a rape to police, only to recant when taken to a hospital.
By the book, police should "unfound" such a complaint, but doing so could provoke unwarranted criticism, said Barcliff, 57, now head of security for Montgomery County Community College in Blue Bell.
"People who are looking in from the outside are saying, like Chicken Little, 'The sky is falling, the sky is falling,' when the sky isn't falling," he said.
Retired Lt. Thomas Doyle, a supervisor under Barcliff, said "investigation of person" was for cases that were just too hard to sort out.
"You have the college student, the female, who was drinking the night before and was wet between the legs," said Doyle, 50, now director of the criminal-justice program at Philadelphia Community College. "Is that a rape? That's an 'investigation of person.' "
Capt. Eileen Bonner, commander of the unit from 1991 until last year, declined to be interviewed.
Maxwell, the department's chief of detectives, said "investigation of person" was frequently used when police believed women were filing false rape complaints to qualify for state-funded abortions.
Rape victims are eligible for free abortions but for many years had to obtain police reports documenting that they had filed complaints.
The number of women in Philadelphia who obtained Medicaid abortions on grounds of rape was small - 10 in 1993, the last year a police report was required.
"In some cases, they had cabs running because they wanted to get an abortion and they just asked us to give them a [report] number," said Maxwell, 55. "But we weren't going to code it a rape if it's not a rape.
"In some cases, it was a boyfriend, it was consensual sex, but they can't afford an abortion, so they're going to say they were raped. So we figured that, in those cases, we were coding them 2701 - 'investigation of person.' "
Susan Frietsche, an attorney with the Women's Law Project in Philadelphia, took strong exception to Maxwell's remarks. The organization mounted a successful battle in federal court to strike down the requirement for a police report in Medicaid abortions stemming from rape.
"One of the things that motivated us was that there was a pervasive belief out there that women cried rape to get a free abortion," she said. "And it's just untrue. . . . Throughout this case, the state could not present any evidence that women lied about rape.
"I'm sure that police believed it. Isn't it a whole lot more comfortable to believe that than to see that your city is unsafe for women?"
A SECOND VICTIM BRINGS
CREDENCE TO THE FIRST
Tikesha Farmer told police the man grabbed her at knifepoint in West Philadelphia, hustled her into an abandoned building near 40th and Lancaster, and raped and sodomized her.
Police were skeptical and coded the case 2701.
They changed their minds and upgraded it to rape - but only after a second woman was raped in the same building four months later, allegedly by the same man.
Farmer, a hair stylist who says she was sexually abused as a child, acknowledges drinking and drug problems.
She was 22 when she was raped on Feb. 10, 1995. Around 11 p.m., she was walking to a bar for a cousin's birthday party. She met a man approaching from the other direction. He put his arm around her neck and forced her into the vacant rowhouse at 712 N. Preston St.
There, she said, he raped her at knifepoint atop a filthy mattress. Discarded doors, debris and other bedding blocked every window.
She called 911. Patrol officers from the 16th District drove her to the rape center at Jefferson Hospital. Medical records, reviewed with her permission, show she was given a pelvic exam and a morning-after pill to prevent pregnancy.
At the hospital, she was interviewed by Sex Crimes Officer Robert Thurman. She gave a description of her assailant - a black male, about 28 years old, 5-foot-8 with a medium build and a box haircut.
There was another detail she remembered: The man had two moles on his penis.
Farmer told Thurman she had a history of mental illness. Her medical records refer to manic depression and schizophrenia. In an interview this year, she said the rape squad focused on that.
"They kept asking me, 'Was I hallucinating?' " she said.
Thurman's report summarizing his investigation says he spoke with a neighbor who lived next door to the rowhouse. The neighbor, 33, told Thurman that he saw a man and a woman who looked like Farmer inside the house smoking crack cocaine and that when he stuck his head in, they told him, "Everything was OK."
The neighbor told Thurman that he knew the man by the name Chuck.
Thurman designated the case "investigation of person." His lieutenant approved the coding.
Four months later, a second woman reported that she was forced into 712 N. Preston and raped - by a man she knew as Chuckie. The 27-year-old woman was interviewed by The Inquirer and said she wished to be identified by her first name, Neva.
Neva's case was also assigned to Thurman.
The Inquirer obtained his investigative reports on both cases.
As he questioned her at Jefferson Hospital soon after the assault, Neva told Thurman something startling, according to the investigator's report.
She said that as the assailant pinned her arms down and raped her, he said: "You're lucky," and told her he had sodomized "the other girl I had up here."
Police put Neva in a patrol car and drove around looking for the rapist. Within a few hours, police had arrested John Jones, 31, a West Philadelphia man with eight aliases, including Charles and Chuck; a long arrest record; and convictions for shoplifting and burglary.
The same night, Thurman pulled the file on Farmer's complaint. The location, the assailant's nickname, the physical description - all matched.
Thurman remembered the two moles Farmer had described. His investigative report says that Jones, by then in custody, was "observed to have" moles on his penis. The distinguishing marks were photographed.
Jones was charged with both rapes. In a written response to The Inquirer's questions, he said: "I was not involved in any rapes."
The District Attorney's Office ultimately withdrew the charges in Farmer's case after she missed several court hearings. Neva cooperated with prosecutors, but the case was postponed repeatedly by plea bargaining, the failure to bring the defendant to court from prison, and Neva's failure to appear at one hearing. Those charges, too, were withdrawn.
Prosecutors say they are convinced Jones raped Farmer and Neva and are prepared to reinstate the charges if the victims are willing.
"No one ever doubted the validity of these cases," said Mimi Rose, the veteran chief of the district attorney's Family Violence and Sexual Assault unit.
Thurman, 47, who retired from the force last year, said in a brief interview that he could recall few details of his investigation of the two cases.
Farmer ultimately got a vindication of sorts. Thurman, the day after Jones' arrest, filled out a "supplemental" investigative report and upgraded her case from 2701, for "investigation of person," to 211 - for rape.
Now it was official: Police finally believed Tikesha Farmer.
HOW TO GET HELP To report a sexual assault to police, dial 911.
For rape counseling and other support services, call the following organizations. The numbers listed are confidential, 24-hour hotlines.
Philadelphia - Women Organized Against Rape, 215-985-3333.
Bucks County - Network of Victim Assistance, 800-675-6900.
Chester County - Rape Crisis Council, 610-692-7273.
Delaware County - Women Against Rape, 610-566-4342.
Montgomery County - Victims Services Center, 610-277-5200.
Burlington County - Contact of Burlington County, 856-234-8888 or 609-267-8500.
Camden County - Services Empowering Rape Victims, 856-964-7378.
Gloucester County - People Acting to Stop Abuse, 856-881-3335.
CRIMES UNCOUNTED * Tomorrow: Police have found a new code for murky cases.
* philly.com Visit The Inquirer's Internet site to examine reports on more than 950,000 Philadelphia crimes from 1991 through 1998 - including those that police put in noncriminal categories. You can retrieve reports by type of crime, by location or by date of occurrence.
The database includes several thousand incidents from 1997 and 1998 labeled "investigation of person" and "medical examination," two noncriminal codes frequently used by the sex-crimes unit. The Web site also contains previous news articles on police "downgrading" of crime.
Point your browser to http://www.philly.com/packages/crime
* Inquirer staff writers Anne Barnard, Maria Panaritas, Alletta Emeno, Thomas J. Gibbons Jr. and Tom Torok and graphic artist Matthew Ericson contributed to this article.