Making Contractual Component Repair More Responsive to Air Force Needs

by John Abell, Mary E. Chenoweth

Research Brief

The U.S. Air Force is spending a large and growing share of its depot-level repair and maintenance budget on private contractors. In 1992, almost one-third of Air Force expenditures on depot-level repair and maintenance—$1.2 billion of a total $3.7 billion—went to contractors. In 1994, contractual repair apparently consumed some 40 percent of depot-level maintenance expenditures. As the Air Force's reliance on contractual repair grows, so does the need for greater responsiveness, in other words, repair that is timely, relevant, and robust. A repair system is relevant if it is always repairing next the asset that will do the combat force the most good. It is robust if it holds up under uncertain repair demands.

A recent RAND report from Project AIR FORCE indicates that contractual repair of aircraft components is sluggish and costly—a situation the Air Force can ill afford. Poor contractor responsiveness forces the Air Force to invest in more aircraft recoverable spares to fill the repair pipelines.[1] It inhibits the depots from responding quickly to urgent, unanticipated needs. In Contractual Component Repair Policy: A Key to Improving Depot Responsiveness, authors Mary Chenoweth and John Abell analyze the problem.

Evaluating the State of Contractor Responsiveness

Chenoweth and Abell analyzed 19 contracts for the repair of F-16 components at the Air Logistics Center in Ogden, Utah, examining repair flow time through a comparison of negotiated and actual times. Repair flow time refers to the elapsed time from when repair on an asset begins until it is shipped out in serviceable condition. Negotiated repair flow time represents the mean number of days the contractor and Air Force agree upon for performing a repair, excluding delays such as time awaiting parts—the time an asset may sit idle in the contractor's facility, awaiting the arrival of government-furnished material. The actual repair flow times that Chenoweth and Abell observed include time awaiting parts, although they had no means to identify or distinguish it from other sources of delay.

Figure 1. Repair Flow Times for Assets with 30-Day Negotiated Repair Flow Times

The difference between negotiated and actual repair flow times can be great. At Ogden, assets with negotiated repair flow times of 30 days took, on average, 59 days to complete. These repair flow times look even longer when we consider absolute repair flow times instead of averages. As Figure 1 shows, only 30 percent of assets with negotiated repair flow times of 30 days were repaired in 30 days or less, while a significant fraction took 90 days or more.

Even 30 days is a long time for repairing these assets. At Ogden's own repair facility, repair flow times for similar assets are 12.5 days, on average, including time awaiting parts. Organic repair at Ogden has been very responsive since the Coronet Deuce demonstrations of 1991–1992, which showed that greater emphasis on timely repairs can reduce repair flow times dramatically. Such improvements in organic operations raise the expectation that the Air Force can improve the responsiveness of contractual repair through similar policy changes.

Although Chenoweth and Abell examined only the repair of F-16 components at one air logistics center, their analysis has much broader implications because the problems with contractor responsiveness are not peculiar to one air logistics center or component type.

Sources of the Problem

There are several reasons why contractual repair takes so long. First, the contracts Chenoweth and Abell examined did not contain language or incentives to encourage contractors to accomplish the work any faster than the minimum established in the contract. Second, the government is not providing contractors with the materiel support they need to expedite repairs. When contractors do not have adequate supplies of government-furnished material, time awaiting parts increases. Third, since the air logistics centers do not routinely evaluate time awaiting parts, the information needed to track contractor responsiveness (and help improve it) is not readily available.

Determining repair requirements is inherently fraught with uncertainty. Yet to negotiate favorable contracts, the Air Force must project the quantity and mix of repairs needed over a given period. The Air Force must also project what funds will be available to pay for these repairs. The more uncertain the terms of the contract, the more the Air Force pays the contractor as insurance against uncertain funding and unpredictable workloads. However, the way the Air Force allocates the workload of "split items"—assets serviced by both the depot and the contractor—only increases uncertainty for the contractor. Traditionally, the Air Force has allocated the workload of split items to its organic facilities up to their capacities and the overflow to contractors. This creates a stable, predictable workload for the organic facility and a less predictable one for the contractor.

Contractual repair is not always relevant because it does not routinely give priority to repairing first assets that contribute the most to mission capability.

Ways to Improve Contractor Responsiveness

The Air Force can improve contractor responsiveness. First, the Air Force needs to write incentives into contracts to encourage greater responsiveness. Contracts might specify dramatically reduced repair flow times, awards for achieving repair flow times shorter than those specified, or penalties for failing to meet specified repair flow times.

Second, the Air Force needs to measure time awaiting parts and to provide contractors with adequate supplies of government-furnished material and repair parts. When useful measures of contractor responsiveness are available, the Air Force can award contracts more effectively. When better materiel support policies are in place, time awaiting parts can be reduced.

Third, the burden of uncertain repair requirements needs to be shared more equitably between contractors and organic facilities. The Air Force could achieve this by stabilizing repair requirements and funding for contractual repair, and by changing its current practice of giving the base workload of split items to the organic shop and any excess to the contractor. The rationale for the current practice is that it is more cost-effective for the contractor to handle the overflow of split items. Chenoweth and Abell believe that the opposite may be true because the organic facilities generally have greater scope of repair and are better able to cope with uncertainty.

Fourth, the Air Force should extend the use of DRIVE (a computerized decision support system now applied to organic repair) to improve shipping and repair priorities and workload allocation in contractual repair. First demonstrated at Ogden in 1987, DRIVE works on the principle that if the organic shop has a choice of what asset to repair next, it should choose the one that provides the most benefit to its customer. The same principle can probably be adapted to contractual repair.

To implement these suggestions, the Air Force could authorize an office within the Air Force Material Command and at each air logistics center to initiate policy changes. This office would represent the buyer of contractual services. Today, the responsibility for contractual repair is dispersed.

There are other steps the Air Force can take to improve contractor responsiveness. Although diverse, the measures all share a common goal: an Air Force logistics system that is more responsive and robust in an uncertain environment.


  • [1] A recoverable is usually repaired when it fails; a consumable is discarded.

Research conducted by

This report is part of the RAND Corporation Research brief series. RAND research briefs present policy-oriented summaries of individual published, peer-reviewed documents or of a body of published work.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.