Jump to content

What About A Contracting Think Tank?

Posted by Vern Edwards, 10 December 2008 · 1,223 views

How about a contracting think tank?

I recently sat on a panel with Professor Emeritus Ralph Nash and Associate Professor Steve Schooner of The George Washington University Law School, and Paul Dennett, a former administrator of the Office of Federal Procurement Policy, to discuss the state of government contract administration. The consensus was that the government does not have enough people to do as good a job of contract administration as it should be doing and that the people who do it need more training. There is nothing very remarkable about that consensus, the same opinions have been expressed by nearly every observer and critic of government contracting over the past 20 years.

There is a long list of contract administration functions in FAR 42.302, but the key issue seems to be quality assurance, inspection, which FAR 2.101 defines as follows:

??Inspection? means examining and testing supplies and services (including, when appropriate, raw materials, components, and intermediate assemblies) to determine whether they conform to contract requirements.?

The critics are concerned about whether the government is getting what the contractors have promised to deliver, or, to put it another way, whether the government is getting what it is paying for. There are other concerns as well, for example: whether contractors are complying with cost accounting standards and cost principles, are fulfilling their obligations under contract clauses that implement the socio-economic programs in FAR Subchapter D, are complying with security requirements, and, generally, are obeying the laws. And there are concerns about organizational conflicts of interest, the implications of ?blending? contractor and government employees, and about contractors performing inherently governmental functions.

Now, inspection of supplies (goods) is well understood, and there are countless books on quality assurance for manufactured items and raw materials. But the government spends more on services than on supplies these days, and the inspection of services confronts us with many unsolved problems and unresolved issues. When a service produces a tangible artifact that is, in and of itself, adequate evidence of the quality of the service, then quality assurance can be handled much as it would in a contract for supplies. Thus, the quality of equipment repair services can be determined by examining and testing the items that the contractor claims to have repaired, and the quality of grounds maintenance services can be determined by examining the grounds. But how do we perform quality assurance when a service produces no tangible artifact, or no artifact that is, in and of itself, adequate evidence of the quality of the service, or when the work does not entail observable physical activity on the part of the contractor's employees? There are many such services.

Consider, for example, contracts for the services of security guards who perform entry control at government facilities. A government inspector can see people lined up at the entrance, showing their identification to the guards, and the guards looking at it and at the belongings people place in the X-ray devices. But the service work is taking place in the guards? heads, and that work cannot be observed directly. A guard may look at the identification or appear to do so, but does she see it, does she think about what she sees, and does she make the right decision based on what she sees? The government can perform tests in which it tries to get people with bad identification and contraband past the guards, but such tests have to be random. Even if the guards catch the test subject, the government cannot be sure that the guards are not letting others in who should not be admitted. Random sampling is not adequate when performance must be perfect, yet 100 percent inspection is not practical. When work takes place in peoples? heads and the output is an undocumented thought, what do you examine and test?

What about contracts for analytical services? Again, the work takes place in the heads of the service employees. Studies and reports, in and of themselves, are not adequate evidence of the quality of the underlying thought processes. While an expert reader might detect obvious errors of fact or reasoning, other errors might go undetected, and disagreements with the conclusions of the service provider are not necessarily indicative of poor quality work.

Under time-and-materials contracts, it is common practice for contracting officers to require contractors to submit their invoices to a contracting officer?s representative (COR or COTR) for ?approval.? But what does ?approval? signify? Conformity of the work to contract requirements is not a condition of payment. In any case, work under many such contracts produces no tangible artifact that is, in and of itself, evidence of the quality of the contractor?s work. As a general rule, when contract performance yields no such tangible product, the sight of a contractor employee at his or her desk does not warrant a conclusion that the employee was ?working? at the moment of observation, only that the employee was present.

In addition to the intangibility of much service output and the government's inability to directly observe the contractor's work, there are problems associated with widely dispersed locations of performance and side-by-side performance of the same duties by government and contractor personnel.

These crude examples reflect just some of the problems that are unique to service contracts. My point is not that there are no solutions to these kinds of problems or that no one has solved any of them. My point is that in many cases the government (1) has not explicitly recognized the problems, (2) has not worked systematically ether to find solutions that others have developed or to develop its own, (3) has not widely reported any solutions that it may have developed, and (4) has not developed appropriate training with respect to the problems for government quality assurance personnel. The advocates of performance-based contracting would have government inspectors use statistical quality assurance methods⎯random sampling and AQLs, methods that were developed for manufacturing. But services do not produce identical units of output. Every output is a somewhat unique response to a somewhat unique set of circumstances. We don?t strive to reduce variability in the output. In fact, we want the output to be varied as appropriate. Statistical inferences based on random sampling might not be valid for such services. In any case, Military Standard 105, Sampling Procedures and Tables for Inspection by Attributes, which described the random sampling and AQL procedures used for many services contracts was canceled because it does not reflect current thinking about quality assurance processes. It has been replaced in part by MIL-STD-1916, DOD Preferred Method for Acceptance of Product. The foreward to that document says:

?Sampling inspection by itself is an inefficient industrial practice for demonstrating conformance to the requirements of a contract and its technical data package. The application of sampling plans for acceptance involves both consumer and producer risks; and increased sampling is one way of reducing these risks, but it also increases costs. Suppliers can reduce risks by employing efficient processes with appropriate process controls. To the extent that such practices are employed and are effective, risk is controlled and, consequently, inspection and testing can be reduced? This standard complies with the DoD policy of eliminating acceptable quality levels (AQL's) and associated practices within specifications.?

So where does that leave those DOD folks who have sampling and AQLs built into their Quality Assurance Surveillance Plans? Anyway, how much should a contracting officer deduct from contract payment for that one instance out of ten thousand in which a guard let someone get in with a gun?

If you wonder where I?m going with all of this, here it is: Why don?t we have a contracting think tank to work on these kinds of issues and problems? The existing think tanks that occasionally address contracting issues, like Brookings and Rand, focus on what I call ?macro? issues⎯Does outsourcing reduce costs? How much outsourcing should the government do? What work should the government outsource? What we need is a think tank that takes on day-to-day practical issues, such as the ones I described above, and publishes its findings and recommendations for chief acquisition officers, senior procurement executives, heads of contracting activities, chiefs of contracting offices, and contracting officers. I don?t mean a policy shop. Goodness knows, we don?t need yet another policy shop. No, I mean a place where smart people spend their time thinking about ways to do contracting better than we do it now and publishing their thoughts.

Contracting is a $400 billion per year business. Per year! We have about 28,000 people working at it. When you think about all of the handwringing over a one-time $700 billion stimulus package and a one-time $25 billion rescue for the auto industry, you have to wonder why we are not rounding up first rate minds and putting them together to systematically identify contracting issues and problems and to find solutions that everyone can use.

Goodness knows⎯there would be a lot to think about.