http://phila.gov/phils/carchive.htm

This grant of $6,409 supported the first phase of a cooperative pilot project between the Department of Records, the Clerk of Quarter Sessions, and the First Judicial District of Pennsylvania to sample and rehouse criminal court records of Philadelphia County housed at the Philadelphia City Archives. The project was promoted to the public in the following ways:

1. Stakeholders? Group

Project staff explicitly sought to promote involvement of constituencies with long-term interest in the preservation and management of records of criminal court cases in Philadelphia County by forming a committee of stakeholders including representatives from the First Judicial District, the Philadelphia District Attorney?s Office, the Pennsylvania Historical and Museum Commission, scholarly researchers, a statistical consultant who provided his services pro bono, and interested lay people to critique proposed sampling methodologies at various stages of testing (Appendix A). The Department of Records encouraged participation by posting an announcement on the Urban History listserv. The group convened with the Department of Records? staff to discuss the project on May 6, June 10, and November 14, 1997. Individuals who responded to the initial listserv posting but were unable to attend meetings remained on the mailing list to receive project updates. Stakeholders provided continuing input as specific methodological questions arose.

Although the primary focus of the pilot grant has been to devise a method for sampling criminal court records in Philadelphia County--with the practical emphasis on records which have fulfilled or which will soon fulfill their minimum retention requirements--project personnel have been mindful of ways in which their work speaks toward the more generalized need for guidelines in managing court records, particularly the burgeoning mass of criminal court records which are characteristic of twentieth-century urban communities. Care has been taken to document the sampling pretests which were conducted and to develop written guidelines and forms for implementing the criminal court records sampling procedure which was selected for Philadelphia County as described (See section B.3, "Selection of Sampling Strategy," in this report). The procedural guidelines form the basis for a manual on criminal court records sampling which is being used in training staff and volunteers who are performing the sampling on the current project. The project staff also plans to publish a summary of the sampling methodology and accompanying documentation on the City of Philadelphia web site (http://phila.gov/phils/carchive.htm) and in appropriate records management literature.

The project staff issued a press release describing the project (Appendix B) that was distributed by the Office of the City Representative to the local media and the professional archival and records management associations. The content of the press release also was electronically distributed to the archives and records management listservs. This external promotion served to attract substantial interest in the development of the stakeholder advisory group.

The project was further publicized in a feature story entitled, "How Many Crooks Can Stand By Their Records," by Ron Avery in the December 23, 1997, Philadelphia Daily News (Appendix C). Author of a recently published popular history of crime in Philadelphia, Avery explained the work of the sampling project to the public and publicized our need for volunteers to assist with the sampling and background research. The article attracted several volunteers to the project. Announcements of volunteer and student internship opportunities with the project were distributed through the Department of Records and to professors at colleges and universities who had previously expressed interest in placing undergraduate students in internships at the Archives.

2. Describe Project Results and List Final Work Products and Project Results

The purpose of this project was to develop a sampling methodology which had as its end the reduction of the total volume of the County?s court records in order to make it feasible to use these records for research, to facilitate their preservation, and to reduce storage costs. The City Records Center holds roughly 26,000 cubic feet of criminal court papers dating from 1838 to 1987. The intention was to concentrate our efforts on those dating from 1940, a period during which the number of criminal court cases rose dramatically and which has contributed substantially to the bulk of these records. The City Archives determined that criminal court records before 1940 were sufficiently manageable in terms of volume to permit a 100 percent retention of the pre-1940 files.

The grant enabled the Department of Records to hire a project staff person, Gail E. Farr, a certified archivist with background in public records management and computer skills, to coordinate the work of the project. Her duties have included the following:

1. Original Assumptions

A preliminary sampling strategy had been formulated when this project grant began in the spring of 1997. Generally speaking, the Department of Records proposed to sample the following categories: (A) a random selection of all the criminal court case files from 1940 for permanent retention; (B) files pertaining to charges of homicide or manslaughter (as mandated by the County Records Manual, all court records relating to first or second degree homicides must be retained permanently); and (C) files pertaining to spectacular or notorious crimes in Philadelphia history as identified on the basis of reference sources such as the local chronology section of the Bulletin Almanac--a compilation of newspaper headlines--and the annual reports of the District Attorney's Office. These sampling methods needed to be tested and sampling ratios worked out.

Devising a sampling method which would enable the project staff to capture the entire corpus of documentation pertaining to a crime without necessitating intensive research in dockets was one of the major challenges the project staff anticipated. In appraising the court papers, the project staff found that individual instances of crime more often than not result in multiple indictments. For example, a robbery may not result in an indictment only for robbery but also for carrying a concealed deadly weapon, assault and battery, conspiracy, receiving stolen goods, and other associated charges. Conversely, one bill of indictment may be issued against several persons, e.g., four persons charged for the same robbery. However, in Philadelphia County, up through 1958, the bills of indictment were filed as individual tri-folded documents. The project staff, therefore, was unable to use the "1 in nth" interval sampling method of pulling documents which has been used in various other archival applications because it would result in the omission of part of the record of any given crime episode in which there were multiple charges. The project staff also recognized the need to do two sets of file pulls in any one year: one pull to segregate selected bills of indictment and another pull to segregate the files for notes of testimony of proceedings regarding those indictments which led to a trial. In Philadelphia County, the bills of indictment and notes of testimony were filed in separate containers up through 1958. The sampling for the years from 1958 looked to be a simpler process because the court began filing all bills of indictment for related charges and all related notes of testimony for those charges in a single packet--or group of packets--with the bill of indictment numbers labeled clearly on the outside of the packets.

The project archivist investigated the feasibility of several possible sampling approaches for selecting Group (A), the random sample group, including sampling by year, stratification by type of crime, and the fat-file approach. The approach which would provide the most useful information from a research standpoint would include the documentation regarding the principle charge in a criminal episode together with the documentation on any associated charges, but determining a feasible method for identifying these bodies of interrelated material proved difficult. First the project staff attempted a modified 1 in nth model by counting through the docket books, selecting every tenth indictment number, compiling a list of selected indictment numbers and the numbers of indictments for any associated charges, and then pulling and segregating the corresponding bills of indictment and notes of testimony. On closer examination, the project staff eliminated the need to do the preliminary research in the docket because the bill numbers of any associated charges were plainly marked on the face of the bills. This method of selecting every 1 in nth bill and associated charges was not adopted, however. The advisory committee pointed out that it favored the selection of the larger crimes; the more indictments in a crime, the greater would be the likelihood of that crime being selected. Similarly, it was noted that there are many categories of crimes which ordinarily result in only one charge which meant the project staff had to find safeguards to ensure that these were not inadvertently excluded from the historical record. They also noted the difficulties for researchers seeking to do statistical analysis of files selected by the means proposed which would require weighting.

The advisory committee suggested that the project staff start by going through the boxes of bills of indictment; use the notations on the faces of the bills to identify all the groups of related charges; count the number of groups of charges; and then choose a percentage of the groups of charges for permanent preservation. This approach would ensure the equal likelihood of selecting the crime involving one bill of indictment and the crime involving fifty indictments. The project staff found this method to be extremely labor intensive: to pre-sort and select random sample group (A) for all 12 months of a year and then collect sample groups (B) homicide, and (C) spectacular/notorious cases for that year took over 100 hours per year sampled.

• Sample Group A

The advisory committee urged us to find a compromise which would enable us to random-sample crime episodes (not bills per se) in order to preserve the integrity of the sampling method. The most straightforward method for doing so was to reduce the size of the sampling ratio: for example, instead of doing the time-intensive pre-sorting of all bills of indictment throughout an entire year, the project staff investigated the possibility of selecting every other box, every tenth box, the first box of every month, etc., and conducting the intensive pre-sorting of the bills into crime cluster groups on these selected segments only. The method which proved most workable was to randomly select a sequence of months distributed throughout the year, identify the clusters of associated cases in these months, and use a random number table to select a portion of the cases in the selected months for permanent retention in Sample Group A. This method offered the appropriate guarantee of randomness while coming within range of the most time-economical method of sampling bills of indictment, which, given the particular character of the filing system used in the Court of Common Pleas for the period in question, was regarded as conceptually unsound by researchers participating on the review committee. A comparison of time estimates is shown in Appendix D outlining Plan A and Plan B. Fortunately, the choice of sampling crime episodes, as opposed to bills of indictment, ceases to be an issue after 1958. In that year, court officials began filing all the bills of indictment and notes of testimony pertaining to a crime episode in the same file or binder so that they are in effect "pre-sorted" for us to sample from.

• Sample Group B

The selection of Sample Group B posed no major problems. For the years 1940-1958 the project staff compiled a list of these charges from the docket index. This work was done by volunteers using a form made up for compiling the information (see Appendix E: Data Sheet for Sample Group B: Homicide/Manslaughter). On average, the project staff found it took a volunteer 8-10 hours to compile a list of charges for murder and homicide for one year.

• Sample Group C

With the help of volunteers, project staff surveyed the Bulletin Almanac and completed data sheets for each crime listed in the local chronology section from 1940 through 1977, when the Almanac ceased publication. With volunteer assistants, the Project Archivist was also able to compile data sheets for all of the specific crimes mentioned in the annual reports of the District Attorney's Office for those years when reports were available: 1952-72 and 1982-83. (See Appendix F: Data Sheet for Sample Group C: Spectacular/Notorious Cases). Through these means, the project staff identified around 300 spectacular and notorious cases for the time period.

Overall the project staff found these sources useful in identifying a wide cross-section of activities which resulted in criminal charges and related documentation. These included activities in the criminal underworld; crimes committed against famous people; crimes against ordinary persons; daring crimes such as kidnaps and jailbreaks; arrests related to the civil rights movement and anti-war activism; charges related to race riots; drug possession and sale; and white-collar crime including graft in city departments. Additionally, the District Attorney's Office reports were particularly useful in highlighting cases which were important from the standpoint of crime-fighting, such as increases in arrests and convictions due to new legislation or technologies aimed at lawbreakers.

Conversely, the project staff had difficulty using the Almanac in that the headlines generally listed only the crimes, not the defendants, which meant the project staff had no way of locating the records through the name index which is the most direct route. However, by checking the dates in the newspaper itself, one could almost always find the news coverage of the crime and obtain the defendant's name with minimal searching. Indeed, the newspaper coverage was often quite comprehensive in listing the names of persons charged. The project staff have therefore built some time for conducting newspaper name searches into our estimate for collecting Sample Group C. Another concern with Sample Group C was that it did not allow for the systematic compilation of information regarding individuals who were notorious in the crime world. The advisory group directed us to additional references such as the reports of the Pennsylvania Commission on Crime as a source for compiling a list of names of individuals involved in organized crime in Philadelphia. The project staff have subsequently assigned a volunteer to go through the docket indexes and compile lists of indictment numbers so that the Spectacular/Notorious sample will include documentation of spectacular careers in crime as well as individual spectacular crimes (See Appendix G, Data Sheet for Sample Group C, Name Index File). Finally, the project staff have expanded our definition of Sample Group C to include crimes which occurred infrequently as indicated in the volume-by-volume search of docket indexes being done for Sample Group B (Homicide/Manslaughter). For example, there are relatively few indictments for abortion in the period from 1940 up until abortions were legalized in the 1970s: 25 to 50 per year. These cases are being identified for inclusion with Sample Group C along with documentation for other kinds of charges which occurred relatively infrequently. These include riot, embezzlement, extortion, violation of election laws, practicing medicine without a license, and keeping a ferocious dog.

2. Conclusion

As a result of this investigation, the Philadelphia Department of Records has developed a systematic plan for reducing the bulk of improving access to criminal court records for Philadelphia County. During the project, we have developed a plan for handling more than 20,000 cubic feet of records (1866 ? 1986); see attached table, Appendix H. It is estimated that the sampling , when fully implemented, will reduce the bulk of records from 1940 to 1986 from 19,268 cubic feet to 2,582 cubic feet or an overall reduction of almost 85 percent. Additionally, the Department of Records plans to rebox and refolder the sampled material and produce an inventory. As an additional end product, we will produce a manual on court records sampling which can serve as a model for other county records agencies in Pennsylvania

3. Appendices

Appendix A. Stakeholders? Group

Appendix B. Press Release

Appendix C. Philadelphia Daily News article

Appendix D. Plan A and B Time Estimates

Appendix E. Data Sheet for Sample Group B: Homicide/Manslaughter

Appendix F. Data Sheet for Sample Group C: Spectacular/Notorious Cases

Appendix G. Data Sheet for Sample Group C: Name Index File

Len Armstrong, Deputy Director for Active Criminal Records, Court of Common Pleas, Criminal Justice Center-2nd Floor, 1301 Filbert Street, Philadelphia, PA 19107

Press Release: "Pennsylvania Historical and Museum Commission Awards Criminal Court Records Sampling Grant to the Philadelphia Department of Records," NEWS ? Office of the City Representative, February 20, 1997.

Text of article: "How Many Crooks Can Stand By Their Records," by Ron Avery, Philadelphia Daily News, December 23, 1997, p. 13, and accompanying photograph

STRATEGY A: AECONOMY PLAN@

Data Sheet for Sample Group B: Charges of Homicide or Manslaughter and Related Charges

Data Sheet for Sample Group C: Spectacular/Notorious Cases

Year:_______________ Source of Reference: Bulletin Almanac:_____ District Attorney?s Annual Report: _____ Docket index______Newspaper____ (give reference)_____________________Police Detective File______________ Other:______

(Month/Date) listed in entry : /

Summary of Case:

Spectacular/Notorious Case Research By (Name):_____________________________ Date:_____________________

Data Sheet for Sample Group C: Spectacular/Notorious Cases?NAME INDEX FILE

Summary of why notorious/significant:

Research/Summary by (Initials): _________ Date:___________

Indictments Identified in Index:

Year Month Indictment #(s) Charge(s) Box# Date Pulled

(Please continue on other side if necessary)

Docket research by: ________ Date: _______________ Pulled by (Initials) _______ Date______