One of the reoccurring issues public procurement needs to deal with is cost overruns.
Historically, government procurement has suffered a great deal from cost overruns, particularly those that have resulted from the failure of items ordered to perform in accordance with specifications when they are finally delivered and installed.
These problems have been particularly pronounced in defence contracting and other areas of procurement where the items purchased tend to be at the cutting edge of technology.
Here again are a few examples from the defence procurement sector that are highly instructive:
- Some years ago, the United States Army bought 6,700 cargo trailers before testing them to determine whether they could be hitched to the trucks that were intended to be used with them. Tests ultimately revealed the trailers damaged the trucks. As a result, the trailers required extensive modifications and were not usable for an extended period.
- The THADD missile was intended to offer the army an interim capability to intercept enemy missiles. It was planned to be produced in just five years at a cost of $2.5 billion. When flight testing began, problems were found in the missile’s propulsion system, launch performance and booster separation, its computer software, seeker electronics, flight controls, electrical system and telemetry control. As a result, the system took over eight years to develop and cost $4.2 billion.
- The optics of the Hubble Space Telescope were not fully tested until it was put into orbit. At that time, it was discovered the telescope functioned at a sub-optimal level for several years and required a very expensive in-space correction, including a dedicated space shuttle mission.
A civilian equivalent may be found in the development and production of the M-6 Commuter Railcar, for the Metro Commuter Railroad and Connecticut Department of Transportation.
In that project, the development of the Master Test Plan lagged well into the production phase of railcar assembly. The completion of prototype vehicles was delayed significantly and then abandoned.
As a result of the delays in the development of the Master Test Plan and in the development of prototype M-6 vehicles, railcars were assembled often with conditionally approved or without approved engineering drawings.
In view of the manufacturer’s inability to produce prototype vehicles, it was necessary to accept M-6 railcars with significant conditions attached to that acceptance. Expensive and time-consuming retrofits were required to correct the deficiencies.
As the M-6 railcar program indicates, problems of this nature are particularly acute where an effort is being made to develop a technology at the same time as production of the finished products is taking place.
While this is a particular problem in defense contracting, due to the need to produce advanced technology weapon systems, it has surprising analogs at the municipal level, particularly in such areas as mass transit, highway construction and high prestige building contracts.
It follows that although municipalities make limited purchases of this type, in comparison to more senior levels of government, the lessons learned at those levels should not be ignored.
The causes of problems of the nature may be quickly summarized:
- In an effort to keep the final price down, testing costs are kept to a minimum. This is an obvious false economy, because it will almost certainly be more expansive to redesign when testing reveals problems late in the design process. If testing is not carried out until the item is in production, those items initially produced may have to be written off or extensively modified.
- The necessary testing requirements are not fully appreciated. As a practical matter, testing requirements increase greatly where one purchases custom products, rather than mature, off-the-shelf technology.
- Testing is often avoided by contractors due to concerns about the continued support for a project if initial tests are unfavourable.
