During her 18 years in office, District Attorney Lynne M. Abraham insisted she would not "do justice by the numbers." And the Philadelphia court system's administrator, David C. Lawrence, similarly said, "We're not in the research business."
That meant overall conviction rates for violent-crime cases in the Philadelphia courts were unknown - until The Inquirer drew on raw court data to report that nearly two-thirds of all defendants accused of violent crimes walked free of all charges.
Abraham's successor, District Attorney Seth Williams, has now obtained a $492,000 grant to equip his office with software that, for the first time, will be able to routinely produce conviction data.
Eventually, his office plans to make the results publicly available on what it says will be a greatly retooled office website.
"Seth's view is that this stuff should be open and transparent," said Sarah Hart, deputy district attorney for performance and policy, a new post. "I suspect it won't always be pretty, but it's important we shine a light on it."
Once a software-design firm is selected, Hart wants staff prosecutors to be interviewed so the system can be customized to address their headaches. Getting the system up and running could take two years or more, she said.
Hart, 55, who began working in the District Attorney's Office three decades ago fresh out of law school at Rutgers University, learned early on that prosecutors spent much of their time in paperwork hell - drafting subpoenas, photocopying records, annotating files.
Little has changed, but she has set out to cut the office's red tape and use sophisticated social-science techniques to guide its performance.
In the future, Philadelphia prosecutors should be able to update their cases, issuing subpoenas and swapping information with defense lawyers during pretrial discovery, with a few keystrokes.
More significant, the district attorney and the public will know precisely how criminal cases turn out. All of this is a big shift from the policies of Abraham and the court system.
M. Elaine Nugent-Borakove, president of the nonprofit Justice Management Institute and an expert on prosecutorial policy, said Hart's project appeared to be on the cutting edge in a field where data are sometimes scarce or not shared with the public.
"That kind of initiative is a model that we would like other justice agencies to be following," she said.
The $492,000 federal grant for the software was routed through the state Commission on Crime and Delinquency and approved last month.
Lawrence, the courts administrator, has said that although an analysis of the outcomes of court cases might be "academically or philosophically interesting," it was not essential.
Similarly, neither the District Attorney's Office under Abraham nor the Public Defender's Office kept track of convictions.
The Inquirer published the results of its analysis in December in a series titled "Justice: Delayed, Dismissed, Denied."
The newspaper derived conviction rates by tabulating and analyzing court data on all 31,000 violent-crime prosecutions in 2006 and 2007. The paper's finding of an abysmal conviction rate closely tracked federal studies, based on data samplings, that repeatedly found Philadelphia's felony conviction rates to be among the lowest in the nation.
Most cases collapsed far ahead of trial, the research showed. Factors included witness fatigue and outright fear, the overbooking of police witnesses into simultaneous court hearings, and gamesmanship by defense lawyers who seek delay after delay.
But it has been difficult to determine which factors were the most significant.
As Hart said, "There are a lot of cases that fall out of the system. At the current time, we don't capture information about the reasons why they fall out."
Hart said the office's new software would remedy that, producing data showing not only how many cases fail to end in conviction, but the precise reasons.
The office will seek to scientifically measure such factors as threats to witnesses, the number of times a witness has to show up in court per case, and much more.
With a new "attrition analysis" in hand, Hart said, the office will be able to fix key breakdown points.
Eventually, Tasha Jamerson, Williams' communications director, hopes to make the data public on a retooled website. Jamerson also envisions the site's offering updates and status reports so victims and witnesses can follow their cases.
In the meantime, Hart's policy shop already has begun analyzing case failures to see why they collapse. Her team also plans to conduct a scientific sampling this summer of past prosecutions to dig more deeply into the conviction rate and more speedily produce usable results.
Ellen Greenlee, director of the Defender Association of Philadelphia, which provides representation for those too poor to hire a lawyer, said her staff needed to upgrade its technology as well.
"We're still on paper," she said. "For the system to function at all, it's necessary for all the stakeholders to communicate. They can't let us lag behind."
That gets no quarrel from Hart, who has backed a plan to have prosecutors and defense lawyers jointly work out a way to electronically speed up the release of people on parole.
Hart joined the District Attorney's Office in 1979 and left in 1995 to serve as chief counsel for the Pennsylvania prison system. In 2001, she moved up to direct the National Institute of Justice during George W. Bush's presidency.
In that post, she helped line up $1 million in funding to assemble the team of DNA experts who identified victims of the 9/11 attacks.
She returned to Philadelphia as a prosecutor in 2006.
Hart's main challenge in her two stints in the District Attorney's Office has been to fight efforts to release inmates from local jails in response to lawsuits alleging overcrowding.
Until Abraham and Hart lobbied successfully for a federal law to restrict such lawsuits, the city routinely set free scores of prisoners over the protests of prosecutors.
According to federal surveys, Philadelphia has one of the nation's most severe fugitive problems. At last count, 47,000 defendants in the Philadelphia courts were at large.
Hart traces the problems to a street culture of lawlessness that took hold when the courts were under a federal judge's oversight.
"That's when the fugitive rate in Philadelphia shot through the roof," Hart said. "The number tripled."
More recently, Hart was instrumental in winning changes in Pennsylvania law that have greatly cut the local prison population, clearing the way for hundreds of prisoners to be transferred to state institutions.
Defense lawyer David Richman, who represents inmates in the long-running prisons-court fight, said he had found Hart "a terrific adversary, friendly, cordial, civil."
"She's very able and has a terrific background in the field of criminal justice," he added. In her new role, he said, "she's in a position to make a tremendous contribution."
Since her promotion after Williams took office in January, Hart has joined other top lieutenants in helping to advance his agenda. Among other priorities, she has been helping to develop his plan for so-called zone prosecutions, in which teams of prosecutors would be assigned to pursue cases grouped by neighborhood.
Hart also wants to use social science to help the office prioritize prosecutions.
She has already reached out to Richard Berk, a pioneering criminologist at the University of Pennsylvania and developer of a computerized way to identify high-risk parolees for the city Probation Department.
Berk and Hart are exploring ways to use the same computer model to identify which low-risk defendants can safely be diverted to treatment programs and which dangerous ones need to be hammered with, in Hart's social-science argot, "long, incapacitating sentences."
In large part, Hart said, her job is to figure out ways to apply new methods to old prosecutorial problems.
"We've always been about the business of figuring out risk," she said. "We just haven't done it very scientifically."