1.1 The Risk-Based Approach to Unmanaged Legacy Records
The Risk-Based Approach (RBA) is an expedited method of disposition recommended by Library and Archives Canada (LAC) as a tool to tackle the problem of the unsustainable accumulation of paper records stored by federal institutions. Unmanaged legacy records are records that have never been classified, have little or no supporting documentation, are generally not covered by a valid Records Disposition Authority and are not currently being accessed by program staff or for ATIP requests. Essentially, these are records that have been "lost" to government and to Canadians.
1.2 Extent of the "Paper Mountain" of Legacy Records
LAC estimates that GC institutions hold approximately 17 million linear feet of records, comprising the "paper mountain" of legacy records including significant amounts of unmanaged legacy records. Storage and access to legacy paper records costs the GC approximately $337 million annually.1 The Gomery Commission Report and the Auditor General's 2003 report both acknowledged that inadequate recordkeeping is a serious problem common to many departments.
1.3 LAC Response to the Situation: Developing the RBA Methodology
A pilot project to test the RBA was conducted with Health Canada as a partner in 2004-06. The pilot showed that the RBA methodology was sound, and the team recommended the approach be used by all institutions attempting to impose order on unmanaged legacy records. A Step-by-Step Guide to Criteria-based Disposition of Unmanaged Legacy Records was prepared in anticipation of a further test with a different partner to refine the methodology.
1.4 Refining the RBA Methodology through the Assessment Project with Natural Resources Canada
In December 2007, Marj Akerley, Chief Information Officer at NRCan, and Jean-Stéphen Piché, Director General of Government Records Branch, LAC signed a MOU which described the responsibilities of each institution in further testing the RBA method on selected NRCan records. The objective of this Assessment project is a one-time Records Disposition Authority (RDA), to be approved by September 30, 2008.
1.5 Targeting the Earth Science Sector Records of NRCan Stored at the Bentley Warehouse
Approximately 1,300 boxes from the Earth Sciences Sector (ESS) that met the definition of unmanaged legacy records were identified at the Bentley Warehouse. As of 2001, ESS maintained its own records office that filed paper records in its own subject-based file classification system. This inconsistent approach to recordkeeping resulted in unmanaged records left at the Bentley warehouse; these made an excellent target for further testing.
1.6 Testing and Refining the RBA Methodology at the Bentley Warehouse
From 21 January to 15 February 2008, six agency contractors sorted files and captured data according to the RBA method and the Disposition Rules, under the supervision of Denis Papineau, Information Analyst, Leah Sander, LAC archivist responsible for NRCan, and Sarah Stacy, Project Manager. The team reported weekly to Danielle Blais, IM Lead, Retention & Disposition Program, Susan Lance, A/Manager of Records Management, NRCan and the LAC Assessment Committee chaired by GRB.
This project examined the contents of every container according to an agreed upon set of Disposition Rules. In addition, the records were pre-assessed to ensure they met the definition of unmanaged legacy business records. The project was documented with information captured at the box level; e.g., file titles, provenance and condition of records. Based on this information, the files were triaged into one of the following three categories, with three corresponding disposition recommendations for the records:
|Records have ongoing business value||Records are retained and incorporated into the department's existing records classification structure|
|Records have outlived their business value but have archival value||Records are transferred to Library and Archives Canada according to a one-time RDA|
|Records have no ongoing business value and no archival value||Records are destroyed according to a one-time RDA|
1.7 Results of the Pilot Project with NRCan
The RBA method is an effective way of treating unmanaged legacy records. It is a cost-effective alternative to traditional disposition because the pilot project with NRCan cost $ 25.00 per container to triage using the RBA method. By contrast the traditional method costs $104.762 per container.
The RBA method is faster because each contractor spent an average of 12 minutes to examine each container while the traditional method averaged 200 minutes per container. Using the RBA method, each contractor averaged 11.2 boxes/per day and 2.44 boxes/per day using the traditional method.
This pilot project also found that the records were broken into these overall disposition decisions: 60% of had no further business value and were recommended for destruction; 27% had continuing business value to NRCan; and 13% had archival value and will be transferred to LAC. LAC observed that the risk-averse culture of GC institutions resulted in a larger portion of the records as being identified as "potential business value," despite the Disposition Rules.
1.8 Benefit of the RBA Approach
The RBA method salvages records of ongoing value while saving the institution the money associated with traditional disposition and ongoing storage. This is called the "risk-based" approach because its emphasis on box, rather than file level, description means the work is expedited. Therefore, the institution has agreed to accept the possibility of losing the odd file of potential value in order to treat the entire run of targeted boxes. These formerly "lost" records may be disposed of through a one-time RDA issued by LAC if the records have no further business value. Or, the records may be incorporated into the recordkeeping system if they have continuing business value and made available to staff.
The RBA method is a feasible tool to offer to GC institutions to address their backlogs of unmanaged legacy records. LAC recommends the use of this method on records that meet the definition of unmanaged legacy business records. However, LAC must insist on certain pre-conditions, such as a signed agreement to release records to the RBA process and the commitment of staff throughout the life of the project.
The RBA project shares the goal, along with other assessment projects in the Legacy Cluster, of reducing the paper mountain through innovative but practical options. The project has findings in common with other assessment projects regarding capacity, including the lack of recordkeeping skills and competencies across the public service and the need to educate public servants on risk management and risk acceptance.
2.1 Statistical information
Table 1: Comparing RBA Method VS Traditional Disposition3
|Time taken for records technicians to treat each container||RBA Method||Traditional Disposition|
|Average time||12 minutes||200 minutes|
|Shortest amount of time||1 minute|
|Longest amount of time||28 minutes|
|Cost per box|
Table 2: Breakdown Of Files By Disposition Decision
|Breakdown by disposition decision|
|Destroy: no enduring value||60%|
|Retain for business value||27%|
|Archival: Transfer to LAC||13%|
Table 3: Cost of Project
|Cost of Project|
|Cost of contractors (records technicians from agency):||$27,168.93|
|Cost of equipment and supplies:||$5,335.21|
Table 4: Immediate Cost Savings To NRCan
|Total cost of treating ca. 1300 boxes using RBA Method in NRCan assessment project||Projected cost if ca. 1300 boxes had been treated using traditional disposition|
|Immediate cost savings to NRCan: $ 103,683.86|
The condition of the records indicated that the RBA may be an effective method of identifying government records at risk, especially archival records, enabling LAC to invoke section 13.3 of the LAC Act in the future. Bentley Warehouse, inadequate for long-term storage of records, housed the records for a period of 10 to 30 years. For the most part, the records were transferred to this site when a staff member moved out of his/her office. Because the records had never been incorporated into the central recordkeeping system, it was difficult to determine the origin and significance of the records. In some cases, loose papers were simply dumped into boxes. The contractors captured file data where it existed, but many files were not fully labelled. In general, the condition of the records could only be described as "fair." The Bentley Warehouse had experienced a burst water pipe, and a large number of boxes were water damaged and mouldy. There was also evidence of rodent activity in the containers. As the Condition of Archival Records in Federal Institutions (CARFI) final report recognized, "the [second] highest source of risk to archival records is from storage environments that are not designed for the preservation of archival records."4The Bentley Warehouse demonstrated five out of six risk factors associated with inadequate storage facilities according to the CARFI survey.
Major Finding 1: Departments need to recognize that leaving records in substandard storage results in deterioration and may result in the loss of valuable information.
As already discussed, the RBA methodology does not require that records are classified or described at the file level, as traditional disposition does. Focussing instead on the efficient use of resources, the RBA's minimum threshold for description is at the box level, with disposition decisions made at the file level. The thesis of the RBA method is that it is not cost-effective to expend extra resources on records that have received minimal attention over an extended period.
Despite the emphasis on efficiency, the RBA method offers quality assurance factors such as: the production of documentation for the material treated, the verification of disposition decisions by the GC institution's disposition experts and on-site time given by the portfolio archivist. Nevertheless, the RBA project challenges GC institutions to suspend existing ways of thinking and adopt a "risk based mentality".
Making contact with operational program staff, who would sign off on the disposition recommendations at the conclusion of the RBA project, was found to be difficult. In part, this was because NRCan has two separate organizational units responsible for disposition and records management work: the staff member onsite during the RBA project was not part of the same unit that would obtain sign-off for disposition. As a result, no contact was made with program staff during the project, as was recommended as part of the RBA methodology established by Consulting and Audit Canada during the first pilot project. Attempts to make a presentation to NRCan senior management on the RBA project before its start met with little success because of conflicting schedules.
Recommendation 1: The MOU should be signed by the senior official responsible for Information Management in the GC institution, the senior official from the relevant program area(s) as well as the DG of GRB, LAC, before the project commences.
The RBA project undertaken in partnership with NRCan followed the methodology established by LAC, Consulting and Audit Canada and Health Canada quite closely. A few improvements to the RBA methodology are noted below:
The GRB Working Group on the Risk-Based Approach developed a comprehensive definition of "unmanaged business legacy records" which ensured that this method is used only on the most appropriate records. The definition is:
"records that receive no records management attention aside from storage. They are characterized by a lack of documentation, which can include a lack of file lists and/or identification for disposition, and fall outside of the classification system. The lack of documentation for these records also means that the institutions that hold them (whether on-site or in private storage) are unable to meet reference or ATIP requests in a timely or efficient manner; it also means that these records do not meet the criteria for storage in a Federal Records Centre. They may not be covered by a Records Disposition Authority and have fallen out of the institution's central recordkeeping system."
The records identified for the project by NRCan were pre-assessed by the LAC team to ensure they met this definition. This proved to be a very useful phase, as NRCan had initially suggested records from another sector which proved to be inappropriate for the project. The pre-assessment consisted of:
Because the RBA project was of short duration but required some training of the records technicians, we equipped them with a binder that contained a list of the disposition rules and other FAQs about the project. We found this to be useful. We also recommend that LAC create a list of contractors with experience in the RBA approach, since familiarity with the methodology would save training time for any future RBA projects.
Recommendation 2: Create a list of qualified contractors that have experience with RBA work.
Batching was a very useful mechanism used in the Health Canada RBA study. Batching boxes chopped the volume of material to be treated into manageable chunks and was an easy way to distribute boxes to the contractors. Even without undertaking detailed research, visual clues from the boxes provided enough information to group the material into batches. Batching by provenance worked well in this project, as it was found that contractors could identify more homogenous groups of records quicker than if the records were from a variety of unknown or different sources.
The Health Canada Study recommended devoting part of the first week of work to making an inventory of all boxes to be treated in the RBA project. Because the boxes at Bentley were stored two deep and two high on warehouse shelving, we did not do this because we wanted to reduce the number of times the contractors would have to move a particular box. Instead, we labelled the batches and box numbers as we progressed, rather than upfront. In hindsight, it would have been useful if an inventory (and final count) of all the boxes to be used in the project was undertaken at the beginning. This would have enabled us to manage our time more effectively. Instead we had to guess how many more boxes were to be processed during the project.
The batches were numbered AA01, AA02, etc: AA indicated the bay of shelving, and 1,2 indicated the number of batch within the bay. The Health Canada RBA report recommended using a barcode scanner for tracking the movements of boxes. However, our experience suggests that this would be unnecessary.
The Health Canada project developed a relational database for use in their RBA project. However, it was reported that the database was complex and did not necessarily make data capture easier or more efficient. For this project, data was captured on an Excel template. This proved to be a quick and easy way to capture the data. It was also a useful tool for NRCan as they input data into their recordkeeping system using Excel.
See Appendix A for a sample of the Excel template.
The Interim Report from the Health Canada RBA project stated that a sampling method was used. It was learned later that every box was opened, but only every third box was described. According to those working on the HC Study, this lack of documentation was part of the reason they were unable to obtain sign off from program staff.
After discussing different sampling methods, a decision was made to open and document every box for the RBA project with NRCan. Because the material was unmanaged, data capture was not always as simple as jotting down sample file titles. the file type most frequently encountered during this project was the "Program working file". Some files only had handwritten titles. In some cases, loose files were dumped into boxes and contractors did not know how to describe these beyond stating that they had found a box containing a jumble of loose papers. While the LAC team encouraged the contractors to provide as much description as possible, this was not always obtained. In addition, NRCan did not always consult with the contractors as they worked. As a result, NRCan had difficulty in verifying the decisions and had to review several batches after the project concluded.
Because of the possibility that data entry skills may falter, the contractors were asked to make photocopies so that each batch had sample documentation on file. For larger batches, we requested that the contractors make photocopies for every second box. This was useful and assisted the archivist in researching the significance of the records, and for the department staff in identifying the file types more specifically. It was also meant to prevent opening the boxes after the triaging work was completed. It was felt that some of the photocopies were too minimal (one sheet per box) and that more pages from a file needed to be copied (file cover, file title page, several sample pages from the file).
Recommendation 3: Each batch have adequate documentation that consists of photocopies (file cover, file title page, several sample pages from a file) from every second box
Because some recent (late 1990s - early 2000s) batches were located, one team offered to create box-lists for NRCan, to assist in the re-integration of the records into their classification system. The box list consisted of batch no., file number, file title, date range, if this was possible to determine. This was a good innovation on the part of the team member, and will lessen the burden for GC institutions in re-integrating files found to have business value.
Recommendation 4: Where appropriate, make box listings for files that will be re-integrated into the institution's classification structure.
Despite the increased level of documentation as compared with the Health Canada study, NRCan IM/records management staff were unsatisfied with the quality of data entry done by contractors and wanted to map each file title to the records classification structure. This is close to the traditional approach and is likely unnecessary.
Major Finding 2: The RBA method requires IM/RM staff to accept a certain measure of risk for files of uncertain origin and value. While there is less documentation than what would be produced by the traditional approach, the documentation produced is adequate for purposes of the RBA methodology.
The Health Canada study asked that contractors make a decision between two options: retain or dispose. However, the NRCan project requested the contractors make three: retain for ongoing business value, retain for transfer to archives, and destroy. This was an attempt to clarify the "destination" of the records once they were triaged. This also anticipated further steps, such as the boxing of archival records in archival boxes, which was done onsite by the contractors when possible. This worked well.
Recommendation 5: Triaging into three categories - retain for ongoing business value, retain for transfer to archives, and destroy - is the best way to sort the boxes of records.
Disposition rules are essentially simplified retention schedules meant to be used for the RBA project. The Health Canada Interim Report recommended that the disposition criteria be expanded for future use (p. 21). After much discussion, sixteen file types were created. This pilot found that the expanded disposition rules may have been overly complex, and that some of the new criteria were not useful, for example: "environmentally damaged." This is not really a file type and the teams found it simpler to just indicate this in a notes field. Additionally, information relating to media, taken from the Terms and Conditions documentation, were made available to records technicians. However, no "lumps" of media were located (outside of related paper files) that would warrant looking at these rules. In the future, the portfolio archivist should be consulted about this media.
LAC had difficulty communicating the crucial role of the disposition rules, which are meant to inform all the decision-making during the records triage. NRCan representatives continued to discuss the need to develop retention schedules for all the files to be treated during the RBA project. In reality, it would only be necessary to develop retention schedules for any records identified to be of ongoing business value, as these would be reincorporated into their records classification structure.
The category of "business value" for records proved to be problematic during the assessment project. LAC collaborated with NRCan to establish rules for material of business value by building on the parameters established in the Health Canada study. For example, recent records and scientific reports and data were identified as having ongoing business value. However NRCan staff did not in fact apply the disposition rules when reviewing the documentation. Instead, records management staff continually "overrode" the rule they themselves helped to craft to include more records under the "business value" banner. For example, older program files dated from the 1970s and 1980s.
Major Finding 3: LAC faces challenges in recommending recordkeeping tools such as the RBA because of the risk-averse culture prevalent in GC institutions.
As well, by identifying all scientific data as having ongoing business value, large quantities of materials such as hand-plotted graphs with little or no contextual information were retained though the significance of these records was basically unknown. Consultation with program staff did not take place during the project. Program staff were not to play a role in the RBA project until the documentation was completed. Perhaps because of the specialized nature of scientific records, NRCan may have been guessing at what their business value might encompass instead of consulting with the appropriate area. In a way, the focus of LAC and NRCan was different. NRCan was concerned about finding records of business value while LAC was focussing on those records that could be disposed of.
See Appendix B for list of Recommended Disposition Rules.
Recommendation 6: Adopt the disposition rules listed in Appendix B.
It was suggested that the pilot with NRCan be taken as an opportunity to test the RBA method on electronic records. However, very few non-paper records were located.
Going into the project, the team wanted to keep in mind different ways legacy records might intersect with electronic records. One example was print-outs - paper records that may be the only evidence of an electronic database or file. We had assumed that if the print-out contained information of interest or value, we would investigate whether the information was captured elsewhere (for example, was it maintained in a more contemporary electronic version with the department, or had it been transferred to LAC already?). However, we did not come across many such records that obviated such an investigation. Another issue kept in mind was obsolete formats and while a small quantity of data tapes were found, their containers were quite mouldy, and one of the labels said, "erased."
Still, electronic legacy projects could benefit from the expertise gained during the RBA assessment project. Electronic legacy projects likely share common values and a common goal: attempting to extract records of value in an efficient manner from an unmanaged environment. The RBA methodology asks that contractors search for "red flags" to identify files of value: date ranges, key words and other attributes. It is feasible to think that this approach could be used in an electronic environment e.g., using open text and indexing to extract records from an electronic repository such as the shared drive. Another suggestion is that a thesaurus be created to identify keywords that would enable searching technologies to identify records of long-term value. Working groups that RBA staff could contribute to at LAC groups include the Shared Drive working group and the Digitization Standards Working Group.
Overview from the Project Charter
|Role||Participant (s) and Title|
|Project Champion (LAC)||Jean-Stéphen Piché, DG, GRB, LAC|
|Project Directors||Anne Allard, Director - RKI, GRB, LAC
Peter Cowan, Director, IM Division, NRCan
|Project Lead (LAC)||Judith Roberts-Moore, Mgr., RKI, GRB|
|Project Manager(LAC)||Sarah Stacy, Sr. Project Officer, RKI, GRB|
|LAC- NRCan Project Team:|
|IM/RM Expert (NRCan):||Susan Lance, A/Head, Records Management, NRCan|
|Disposition Experts (NRCan)||Danielle Blais, Retention & Disposition Co-ordinator, NRCan
Denis Papineau, Retention and Disposition, NRCan
|Disposition Expert (LAC):||Leah Sander, Archivist, Archival Operations Division|
|Supervisor (NRCan)||Simon Philips, Records Officer, NRCan|
|Records Technicians (6)||Agency Contractors|
Although the work plan developed by Consulting and Audit Canada recommended hiring a supervisor, the partners opted not to. At the end of the project, we realized that it would have been advantageous to have hired a supervisor from the contract agency.
The pilot with NRCan demonstrated that the records technicians are capable of understanding and applying the disposition criteria. Decisions may have taken some consideration initially (for example, the first instance a particular file type was encountered) but because similar files types were found together in this project, technicians were then able to move quickly through runs of similar files with a low error rate.
Crucial skills for records technicians include:
Disposition Expert (GC institution)
Perhaps the most important overall role, the individual appointed as Disposition Expert must commit to playing an active, on-site role during the triage process but also be available to review the documentation after the triaging has been completed. This person, whom we expect to be a staff member from the department's information management/records management area, should be on-site every day. This person must fully understand the goals of the RBA project and its methodology and should have past supervisory experience.
Responsibilities of the Disposition Expert
Crucial skills for this role include:
If the Disposition Expert performs these duties, the workload for the archivist will be lessened. The Disposition Expert's role is to provide quality assurance to the project by ensuring the data is correct. It is highly advised that one person take on this role (and ownership of the RBA project), and that the duties not be parceled out to other staff in other organizational areas of the Department. If the person appointed as Disposition Expert does not carry out his/her responsibilities, the project will likely fail.
Major Finding 4: Disposition Expert is a key player in any RBA project, must have appropriate experience and be available on a full time basis for the duration of the project.
The LAC archivist spent a much larger amount of time than estimated on-site during the pilot project at NRCan (almost every day during the first week, and 2 days per week afterwards). A higher than expected amount of potentially archival records found amongst NRCan's unmanaged legacy records played a role in increasing the archivist's workload. Describing the project as "an exercise in micro-appraisal," the archivist found that because the records were so de-contextualized, research time was needed to arrive at an informed decision.
The project manager was a LAC staff member and was responsible for the planning of the project as well as the smooth operation of the actual activities at the warehouse. The project manager trained the project team members in the RBA methodology, answered questions and generally provided advice on all aspects of the project. This individual was on site four days a week in order to ensure the progress of the project. It is recommended that for future RBA projects, a LAC staff member be the project manager to ensure that the methodology is adhered to and that the Disposition Rules are respected.
Equipment needs are simple: tools to access boxes (ladder, carts), equipment to capture data (a photocopier, laptops, general office supplies), file folders, boxes, plastic wrap, pallets, and because the environment can be dirty, hand sanitizers, masks and gloves.
The Bentley Warehouse did in fact have a small amount of space where boxes could be placed temporarily, but it was tight. The benefit was that the contractors were working within close reach of the records and did not need to transport them very far. The NRCan staff were able to negotiate with their neighbour (DND) in the Bentley facility to store records on skids in their loading area temporarily as there was insufficient free shelf space in the NRCan portion of the warehouse.
During this project, the technicians were directed to segregate all boxes for destruction, placing the boxes on skids and wrapping them in clear plastic. Boxes to be retained for business value were put back on the shelves, and records that were identified as having archival value were segregated and re-boxed in archival containers. An RBA project needs enough space to split records into three categories and store them for the duration of the project.
The potential need for boxes to be moved more than once always existed and in some cases, skids wrapped in plastic were unwrapped and searched. This was felt to be an inevitable aspect of the project. This was the best solution the LAC team could devise in order to easily locate files as they were being treated.
The costing for the assessment project with NRCan, based on the figures available for the Health Canada RBA Study, was found to be generous. Overall, the RBA project with NRCan required less effort than estimated.
|Estimated Project Member||Actual||Estimated Total No. of Days||Actual|
|Project Director (NRCan)||6.3||25|
|Project Manager (LAC)||65.7||60|
|IM/RM Expert (NRCan)||39||-|
|Disposition Experts (2 - NRCan)||39.9
|Disposition Expert (LAC)||39||10|
|6 Records Technicians (Agency Contractors)||234||116.25|
|Total Team Effort||438||148.75|
|Total Effort||501.9 days||210.75|
The costing for the assessment project with NRCan, based on the figures available for the Health Canada RBA Study, was found to be generous. Overall, the RBA project with NRCan required less effort than estimated.
The RBA method is a "get out of jail free card," that allows GC institutions to clean up neglected pockets of records within the "paper mountain." It is not meant to be applied repeatedly by the same institution; this may only encourage poor recordkeeping habits. The assessment project with NRCan also demonstrated that:
8.1 Steps and Prerequisites for implementing an RBA project
LAC must also agree to:
The Risk Based Approach to the disposition of unmanaged legacy business records is a part of a strategic suite of tools used to "tame" the paper mountain of legacy records. This assessment project allowed LAC to improve upon the existing methodology and to demonstrate that it is a cost-efficient tool appropriate for unmanaged legacy records. While traditional disposition remains the "gold standard" and the recommended tool for disposition, this is a costlier approach. Records that lack context and may be physically compromised are better served by the RBA. The RBA method supports the New Storage Model goal of providing safe, secure and accessible storage for government records.
An ongoing challenge with the RBA is the risk-averse culture of GC institutions. GC institutions must recognize the RBA method as a risk mitigation tool and understand that implementing the RBA is superior to having no strategy in place at all for unmanaged business records. Educating GC institutions about risk mitigation is necessary and will support the Recordkeeping Regime.
9.1 Summary of major findings:
9.2 Summary of recommendations:
1 "FRC-Hybrid New Storage Model Accommodation Strategy." Presentation by consultant Ron Mason to the Fast Track Project Secretariat, 13 March 2007.
2 "Risk-Based Disposition Project". Presentation to MGI Fund Day, January 2006, slide 10. All costs related to the traditional method come from this document.
3 Ibid, p 18.
4 CARFI final report, p. 81.
5 Please note that the figures for NRCan staff have been estimated.