Managing major capital projects: lessons learned from Milwaukee's 3rd district police station and data communications center.
Most local government public works departments are well prepared to direct and oversee their own capital improvement projects, including street construction and repair, water and sewer construction and replacement, street lighting, and standard office and warehouse buildings. However, as capital projects extend into new communications or other advanced technologies, face "fast track" schedule requirements, or involve relatively new engineering design features, the standard budgeting and project management techniques often fall short.
Such was the case when the City of Milwaukee set out to build a new district police station and public safety data communications center. Because of scope creep, cost overruns, and project delays, one city alderman requested that the comptroller conduct a management audit of the project. This article reviews the findings of that audit, highlighting what went wrong and why and recommending specific actions local governments can take to avoid the problems experienced by Milwaukee. (1)
In 1993, the police chief and his managers identified two increasingly urgent needs: (1) the need to replace and relocate an aging, inadequate police station in District 3 on the near north side and (2) the need to replace failing, unreliable data and phone communications and defendant identification systems. In 1996, the two projects were combined into a single project. The project ultimately expanded to include a new multi-level parking structure, an enhanced 911 system, two separate computer-aided dispatch systems, a new records management system, a citywide emergency operations center, and various off-site improvements.
Four city departments--the Police Department, the Fire Department, Public Works, and City Development--were involved in carrying out the project. City Development was to negotiate with existing property owners to acquire the site. Public Works was to administer the building construction and communications, and oversee the work of private vendors. The Police Department was to make final scope decisions and oversee the work of City Development and Public Works. The Fire Department was to work with the Police Department on shared user needs.
Public Works retained a construction manager to review the work of the general contractor, including any proposed change orders. Building construction began in 1999. As the project proceeded, a multi-level parking structure adjacent to the main building was added, as were the installation of emergency generators and the renovation of a radio shop at locations far removed from the project site. The building and parking structure were completed in late 2001.
In the end, Milwaukee had constructed a state-of-the-art police station and data communications center. According to the Police Department, the newly installed technology will significantly improve police response times and overall public safety operations. However, as of October 2003, certain technology components, including the computer-aided dispatching systems, were not fully operational. Moreover, the original 1994 project budget of $20.8 million had ballooned to an estimated $64.1 million by the end of 2003, leading to serious concerns about the city's ability to effectively manage major capital projects.
WHAT WENT WRONG
The final audit report, released in October 2003, found significant weaknesses in virtually every area of project management--project planning and definition, project budgeting, and ongoing oversight. Because the city did not clearly define the scope of the project before getting started, the project got bigger and bigger as time wore on. The completed project included a number of items that were not part of the original plan, including a multi-level parking garage, department-wide records management, offsite emergency generators, and a renovated radio repair shop. As a result, the project budget increased and the timeline was repeatedly extended, as shown in Exhibit 1.
Cost estimates for major pieces of the project were often not supported by adequate documentation or analysis. The original budget estimate reflected the costs of two separate building projects, not a combined project, while little attention was paid to estimating the cost of the technology components. Even though major scope expansion decisions were made throughout the course of the project, the budgeting process provided little if any opportunity for the Common Council to make meaningful "go-no go" decisions.
No city department took responsibility for controlling project costs. The general construction contract was let prior to the city acquiring control of the project site, leading to delays and some avoidable contractor costs. AS the project proceeded, cooperation problems arose among city departments. The building and technology components of the project were never properly coordinated. The Police and Fire departments could not reconcile their respective dispatching needs, necessitating separate dispatching systems that cost $12 million more than the joint dispatching plan. Although the Police Department ultimately assumed control of the technology components of the project, this was initially a source of disagreement between the Police Department and Public Works. There was no mechanism in place to resolve such disagreements as they arose during the project.
Public Works contracted for a construction manager for the project. Although the construction manager's contract was well done, the department in many cases failed to enforce the contract requirements. Project accounts were not properly structured, preventing the issuance of timely, accurate financial reports. Aside from periodic verbal briefings, senior management did not receive regular financial status reports, and there were no formal reports on the physical completion status of the project.
STRENGTHENING PROJECT MANAGEMENT
The purpose of the audit was not only to identify what went wrong during this particular project, but, more importantly, to recommend steps for improving project management in the future. The audit recommendations included four major changes to be applied to large capital facility and information systems development projects:
1. Formalize project budget proposal requirements
2. Revise ordinances to clarify project authority and responsibility
3. Prepare a memorandum of understanding in advance of complex projects
4. Institute and enforce project status reporting standards
The overriding goal of these changes is successful, on-time, and on-budget projects. The balance of this section discusses in some detail each of the four major recommendations.
Formalize project budget proposal requirements. To properly evaluate proposed projects, the city must have a set of well-defined parameters that promote meaningful assessment and comparative analysis. What follows is a checklist of essential information that is to accompany any future capital project budgets proposed by city departments.
* Statement of project scope. The statement of project scope is a complete description of the proposed project that defines what is and what is not part of the project. Specific work products (deliverables) that will result from the project should also be identified.
* Statement of needs and expected benefits. The statement of needs and expected benefits justifies why the project should be funded. The statement should quantify to the extent possible the needs that will be met and the benefits that will be realized by the project. Industry standards and other relevant supporting documentation should be included.
* Cost estimate. This is perhaps the most critical element of the project budget proposal. The cost estimate should include not only an estimate of all financial and non-financial project costs, but also an explanation as to how the estimate was developed. The cost estimate often involves the work of consultants and engineers, and should include any supporting documentation.
* Funding sources. All sources of project funding must be identified, and together they must be sufficient to pay for the entire project. The approval status of any non-city grant funding sources should be specified.
* Project work plan and schedule. The work plan should describe the sequence and schedule of major work tasks, highlighting interim or "milestone" deliverables over the course of the project. The work plan should also identify the units and personnel responsible for each of these deliverables, as well as an estimate of the resources required to complete them.
* City department responsibility matrix. The responsibilities of each city department involved in the project should be outlined in detail, including any use of outside vendors. Procedures for coordinating the work of different departments and for resolving conflicts should also be specified.
* Statement of project assumptions. All significant assumptions underlying the various components of the project proposal should be identified. Examples include estimated bid prices, labor rates, key staff and materials availability, subsoil conditions, private financing, completion of allied project components, timing of land acquisition, etc.
* Application of city standards. Any city government standards (such as those related to engineering, office or operating space, data processing, data security, etc.) to be applied to the project should be identified and explained.
* Project risk profile. The project proposal should identify any inherent risks and explain how those risks might affect the city's ability to successfully complete the project on time and within budget.
When the project budget proposal cannot be reliably completed as outlined above because of the project's magnitude or complexity, the effort should be divided into two or more project phases, each funded sequentially over two or more years. For example, a large building project could be funded for design only in year one, followed by construction funding in year two after all of the design work has been completed and evaluated. The design would provide the information on construction features, benefits, quantities, and cost needed for an informed construction funding decision for year two. Where both design and construction approval are required in the same budget year, the Council could decide to appropriate an amount estimated to complete both design and construction, but only release design funds initially. Contingent upon a favorable review of the completed design, construction funding would be released later in the year.
Revise city ordinances to establish citywide standards and assign responsibilities accordingly. City departments whose needs are the driving factor behind a major capital project should have overall responsibility and accountability for that project. Sponsoring departments are in the best position to define project functions, features, and requirements, and to control project budgets. As such, they should assume overall responsibility for project success and cost.
Furthermore, existing city standards must be applied consistently to capital projects. Common Council ordinances and policies assign to specific departments certain responsibilities for building construction and information technology projects. By ordinance, Public Works has responsibility for the design and construction of city buildings and municipal communication systems. However, the audit found that Public Works did not assume these responsibilities for the district police station project, playing more of a supporting role to the Police Department. There was no indication that Public Works imposed standards on the design and construction of the new building or its communications systems.
The audit recommended that the city modify current ordinances to:
* Reassign to the sponsoring department the overall responsibility for and control of a budgeted major capital project.
* Assign to Public Works the responsibility for establishing and enforcing engineering design and materials standards for capital projects, as well as office space standards, bid letting, and code compliance.
* Retain in the Information Technology Management Division the responsibility for establishing and monitoring information systems standards, including security standards (subject to more stringent industry standards applicable to specific city information systems).
Prepare a memorandum of understanding in advance of complex projects. The fact that the district police station project was behind schedule and over budget can be blamed to some degree oil the lack of coordination among city departments. To improve coordination and cooperation, the audit recommended that participating departments consider entering into a written memorandum of understanding outlining their responsibilities. In effect, this is an intra-governmental contract among departments. The memorandum of understanding should be completed prior to the commencement of the project and should address the following issues:
* Assign responsibilities for each major project activity, including change order approval and processing.
* Specify the documentation needed to support billing by city departments and outside contractors.
* Outline a formal method of dispute resolution among all parties.
Institute and enforce project status reporting standards. The principal objective of project status reporting is to provide timely, meaningful information on the financial and physical completion status of a project against the project plan, budget, and schedule. For Milwaukee's district police station project, senior management did not receive regular and reliable project status reports. Project managers held daily progress meetings, but did not regularly report to the Common Council, the Police Department, or the Fire Department.
The audit recommended that the Budget and Management Division establish project status monitoring standards and guidelines for use by city departments managing capital projects. Under these guidelines, the department sponsoring the project would be responsible for status reporting over the course of the project. Reporting on both financial and physical completion status would be provided periodically throughout the course of the project--weekly for day-to-day project managers and monthly or bi-monthly for elected officials. Ideally, a "roll-up" concept would be applied to status reporting, allowing stakeholders to view project status at different levels of detail. Exhibit 2 is a checklist of the key features of an effective project status reporting system.
Exhibit 2: Key Features of an Effective Project Status Reporting System 1. Includes all costs associated with the project, including all in-house and outside contractor/vendor costs. 2. Provides formal periodic reports to management. Frequency depends on level of management responsibility. Reporting frequency generally ranges from weekly to monthly and is in addition to day-to-day project status meetings. 3. Includes an executive summary status report at least monthly for top management and elected officials. This report should provide the following information: * Current financial and physical completion status * Progress over the last reporting period * Problems encountered since the last report and how those problems were or are being addressed * Planned activities and accomplishments over the coming reporting period 4. "Roll-up" reporting in which project status information is available at various levels of detail, from component-specific information (for example, the status of masonry work) to a summary of all project work segments for top management and elected officials. 5. "Critical path" scheduling and monitoring techniques to continuously identify those tasks that, if delayed, will affect the overall completion date of the project. 6. Comparisons of project budget apportioned over the expected project life versus actual expenditures to date, and total project budget versus estimated cost at completion. Such comparisons should be provided for the total project and for major project work segments. 7. Frequent on-site examination by city management to monitor adherence to city design, materials, and quality standards, and to determine the actual extent of physical completion for each project component or work segment. Such information is used to provide reliable estimates of physical completion versus budget expended. 8. Utilize American Institute of Architects chart of accounts for all major construction projects. This will minimize confusion when bids are prepared and work is charged against the project. 9. Accurate, timely accumulation and application of all project charges to: * The project's cost accounting and billing system * The proper city capital project budget line items * The city's financial accounting system
Various project management software aids are available to assist in developing a reliable status reporting system. The key is to lay out a plan and schedule for each major project component, then closely monitor component spending and physical completion status against that plan. While the government's financial system can provide the necessary information on the project's financial status, physical completion can only be measured with the aid of ongoing estimates of work remaining by each responsible component manager. These completion estimates should be verified with frequent on-site examinations by independent government engineer-observers.
The audit concluded by recommending that prior to the commencement of any capital project, the sponsoring department should establish an appropriate set of project accounts consistent with the American Institute of Architect's standard chart of accounts for construction projects, if applicable. Project accounting should include all project expenditures and revenue sources. The audit also emphasized the importance of establishing procedures to ensure the timely reconciliation of project activity and budget balances to the city's financial system.
The sheer complexity of this project and the variety of city and private sector entities involved made the prospect of completing the facility on time and within budget a significant challenge. Still, the project budgeting and management failures noted in the audit caused substantial avoidable costs and delayed the completion of the project for many months.
After the audit was released in October 2003, the Common Council's Finance and Personnel Committee discussed the audit findings and recommendations. The police chief, budget director, and public works commissioner responded by letter, and the commissioner and budget director further commented on the audit findings and recommendations before the Committee. These officials largely agreed with the major audit findings, including the lack of control over scope. The Common Council passed a resolution adopting the audit's major recommendations and directing the affected departments to make the necessary policy, procedure, and system changes.
While local situations will no doubt vary, the technical and organizational complexity of this project holds valuable lessons for all governments undertaking unique or complex projects. Clearly, an investment in "front end" planning is required. Scope needs to be fully discussed, clarified, and documented, and departmental and vendor responsibilities must be outlined and properly communicated. A monitoring and reporting system needs to be in place before the project begins, and frequent field examinations should be conducted during the project to verify physical completion status. Finally, because change and unforeseen problems are inevitable during these types of projects, procedures should be established beforehand to deal with issues as they arise. Properly implemented, the reforms discussed in this article can be expected to enhance a government's ability to deliver major capital projects on time and on budget.
Exhibit 1: Chronology of Increases in Project Scope and Cost ($ millions) Date Activity 1994 New MPD 3rd District Police Station proposed 1994 Convert old station to Data Communications Center Dec 96 MPD states intent to combine 3rd District Station and Data Communications Center in single building Feb 97 MPD updates project estimate Sep 97 DPW proposes alternatives at 46th and Lisbon Avenue Apr 98 DPW updates estimate for 49th and Lisbon Avenue Oct 98 DCD updates site acquisition estimate Dec 98 DPW consultant updates estimate, including staff parking Sep 99 DPW consultant pre-construction bid estimate Oct 99 MPD estimates Radio Shop upgrade 1999 MPD estimates emergency generators at the police administration building 1999 MFD proposes 2000 budget for technology Late 99 Additional cost for PAB generators Early 00 Additional cost for site acquisition May 00 DPW updates Radio Shop estimate Late 00 MFD proposes additional 2001 technology budget Mar 01 DPW updates Radio Shop estimate Jut 01 Architect estimates adding open-records facility. 00-01 Additional cost for new building 00-01 Additional cost for infrastructure work Jan 02 MPD and MFD decide to develop separate computer-aided dispatch systems 2002 MFD proposes additional 2003 technology budget 2003 Additional cost for Radio Shop renovation 02-04 Additional cost for MPD technology Date Cost Total Cost 1994 $8.8 $8.8 1994 $12.0 $20.8 Dec 96 -- $20.8 Feb 97 $9.6 $30.4 Sep 97 $1.4 $31.8 Apr 98 $0.5 $32.3 Oct 98 $0.4 $32.7 Dec 98 $4.0 $36.7 Sep 99 $0.9 $37.6 Oct 99 $0.3 $37.9 1999 $1.2 $39.1 1999 $1.5 $40.6 Late 99 $0.7 $41.3 Early 00 $0.9 $42.2 May 00 $0.8 $43.0 Late 00 $2.5 $45.5 Mar 01 $0.3 $45.8 Jut 01 $0.3 $46.1 00-01 $6.2 $52.3 00-01 $2.6 $54.9 Jan 02 -- $54.9 2002 $2.5 $57.4 2003 $1.0 $58.4 02-04 $5.7 $64.1
(1.) The entire audit report, including department comments, is available on Milwaukee's Web site (http://www.ci.mil.wi.ns/citygov/comptrol/ 3rdDistrictFinalAuditReport.pdf).
MICHAEL DAUN is director of financial services for the Office of the City of Milwaukee Comptroller W. Martin Morics.