28 February 2016

LONG RANGE STRIKE: 3 LESSONS FROM DEFENSE CONTRACTING THAT TELL US THE AIR FORCE’S NEW BOMBER IS READY TO MOVE FORWARD

FEBRUARY 26, 2016

With the Government Accountability Office’s (GAO) decision to deny Boeing’s protest of the Long Range Strike Bomber (LRS-B) contract, work is ready to begin in earnest on developing the next bomber. The GAO’s ruling focuses primarily on whether the LRS-B award to Northrop Grumman followed applicable rules, regulations, and procedures, and whether it followed the criteria laid out in the request for proposals. Needless to say, this leaves a number of important questions unanswered, including whether the Air Force correctly designed the competition and the development and construction program that will follow. Answering such questions is complicated by the fact that most information about the LRS-B remains classified, and it is hard at times to winnow fact from fiction. As I’vepreviously argued, the case for buying the LRS-B is compelling, but it is also important to address questions raised about the Air Force’s approach. Before proceeding, let me note that several of the contractors responsible for specific programs mentioned in this article are donors to the Center for Strategic and International Studies, where I work. These companies did not sponsor the research discussed in this article, and I and my research team review both our methodology and our findings with other experts and researchers to ensure the objectivity of our analysis.


The fact that the LRS-B development contract was not fixed-price, but rather cost-reimbursable has raised some eyebrows. A cost-reimbursable contract is an arrangement where the contractor is compensated for its costs, with at least some profit, even when costs grow up to a cost ceiling set in the contract. Past cost-reimbursable development efforts have seen costs balloon, and the taxpayer foots most of the bill. For example, F-35 development costs have grown 50 percent. Fixed-price contracts are often believed to avoid these pitfalls. As an example, even though Boeing has experienced an estimated 25 percent cost growth in developing the KC-46 tanker under a fixed-price contract, the Air Force did not need to increase the program’s budget. The KC-46 program is a rare case: a major acquisition program that is currently projected to cost the government less than originally planned. So clearly fixed-price contracts control taxpayers’ costs while cost-reimbursable contracts allow them to soar, right? Actually, no. The truth is that the degree of risk in the development effort, the quality of the cost estimate, and the incentives built into the contract’s terms are far more important to the likelihood of cost growth than the contract type. Here are three reasons why the LRS-B is ready to move forward without changing the contract type.

Costs Don’t Necessarily Grow More Under Cost-Type Contracts

The Department of Defense engages in more than a million contract transactions every year. The majority of these transactions are fixed-price, but a substantial number are cost-reimbursable contracts and because many of the largest contracts are cost-reimbursable, they make up about 30 percent of the Department’s contract obligations from year-to-year. This portion has been relatively stable for many years, which means there are literally tens of thousands of case studies over that last decade that allow us to compare the cost growth on cost-reimbursable contracts and fixed-price contracts. A recently completed CSIS study done for the Naval Postgraduate School examined this history, compiling all of the publicly available defense contract data from 2007 to 2013 and measuring the extent to which these contracts breached their original cost ceiling. The study looked specifically at cost breaches where the government paid more, but didn’t get additional capability in return. For smaller contracts, the extent of cost breaches on cost-reimbursable contracts slightly exceeded those on fixed-price contracts. However, for contracts over $75 million and long duration contracts over $1 million, two categories that include multibillion-dollar contracts like the LRS-B, the extent of cost growth was slightly higher under fixed-price contracts. The lack of evidence that fixed-price contracts better control costs was true even when we limited our analysis to research and development contracts and contracts for major defense acquisition programs, where the risk of significant cost growth may be higher, although fixed-price contracts for major aircraft programs may do better than others. There is clear evidence that fixed-price development contracts face an increased chance of termination across the board. Fixed-price contracts resulted in contract termination about twice as often as cost-reimbursable contracts. While termination might be the appropriate outcome for contracts experiencing major cost growth, the disproportionate number of terminations on fixed-price contracts suggests that they may have been compelled more often by a failed business arrangement than by the fundamentals of the program. Given that contract termination is usually a highly negative outcome, it becomes apparent why the Department of Defense has often opted for cost-reimbursable contracts for large development programs. These results may be counterintuitive, but they aren’t really surprising. That’s because a similar study by the department’s own Office of Performance Assessment and Root Cause Analysis (PARCA) in 2014 also found that cost growth occurs on both fixed-price and cost-reimbursable contracts, and that fixed-price contracts did not necessarily lead to better cost control.



How can cost growth even happen on a fixed-price contract? Isn’t the price fixed? Most fixed-price contracts are of short duration and are used to buy relatively simple goods and services, so cost growth isn’t an issue. But unexpectedthings happen, especially during development, and that can generate substantial additional work, including design changes and additional testing. This additional work often isn’t covered under the original contract, and the contract must be modified to require it. When this happens, via a process known as a change order, the contractor will insist on additional payment. Change orders can happen under all contract types, and the uncertain nature of development, which creates the potential for cost growth, is essentially unaffected by whether the contract is nominally fixed-price or cost-based.

Development Programs Are Risky

History bears out that development programs are risky and fixed-price development programs have often cost the government more as a result. The Department of Defense experienced several expensive misses, such as the C-5 and F-111 programs, and a few outright disasters, like the A-12 Avenger program, in attempting to do fixed-price development in the 1960s and 1980s. These experiences were so painful that Congress actually restricted the use of fixed-price contracts for development in 1988. This restriction was relaxed in 2006 and the department took advantage of the change to use fixed-price development for the KC-46 tanker program awarded to Boeing in 2011. KC-46 was considered a good candidate for fixed-price development because it was a derivative of a mature commercial design (the 767 cargo variant) and utilized military subsystems that were well understood. As previously mentioned, however, even in this example, Boeing has experienced cost growth estimated at 25 percent of the initial development cost as a result of problems with wiring bundles and fuel systems. In the KC-46 case, the Air Force has been incredibly disciplined in avoiding the need for change orders so far, and Boeing has been required to absorb all of this cost growth. This has worked only because Boeing has sufficient financial resources as well as strong market incentives to complete this project even in the face of losses. At the same time, however, this has meant that the Air Force has had to forgo opportunities to make changes to the KC-46 to avoid reopening the contract. For example, although the military standard for electromagnetic pulse hardening has changed since the KC-46 contract was awarded, the contract has not been modified to require the new standard. The KC-46 experience illustrates the importance of understanding how much risk is in a development program when selecting the contract type. When anticipated performance issues and changes in requirements are unlikely to fundamentally affect mission success, it is far more likely that the intended discipline of a fixed-price development approach can be maintained. In several earlier fixed-price development programs, however, the Department of Defense was not so fortunate. It ended up paying substantial cost overruns (exceeding 100 percent in the case of the F-111) in order to avoid program termination, ensure delivery of functional aircraft, and avoid bankrupting some of its major contractors.

The reality is that most development programs contain a great deal of uncertainty because they require doing something that has never been done before. This uncertainty can lead to errors in cost estimates, changes in requirements as programs go through development, schedule slips, and changes in quantity. A seminal RAND studyin 2008 looked closely at the causes of cost growth in major acquisition programs. The study looked at a representative sample of 35 major development programs in the 1990s and 2000s and found that cost growth averaged approximately 50 percent during development and 35 percent during procurement, excluding cost growth due to quantity changes. Development programs experienced more than twice as much cost growth as procurement programs as a result of erroneous cost estimates and roughly twice as much cost growth due to changing requirements. The only source of cost growth that was greater for procurement programs was schedule slippage, because production efficiency is highly schedule-dependent. Considering the degree of uncertainty inherent in development, there can be little doubt of the risk of cost growth, irrespective of contract type. The only thing the contract type changes is how that risk is allocated between the contractor and the government.

There Are Better Ways to Combat Cost Growth in Development

The Air Force has taken several important steps to reduce the likelihood of cost growth during the development of the LRS-B. First, the Air Force included contract terms that incentivize the contractor to control cost growth. Under these incentives, cost growth causes the contractor’s profit margin to get progressively reduced until it reaches a minimum level. However, the Air Force also pays a portion of the cost growth. This incentive structure is carefully chosen to align the interests of the contractor with the interests of the Air Force. They are both motivated to do what they can to limit cost growth, including, for example, limiting requirements changes. This is important because the RAND study identified that changes in requirements were responsible for about a third of all development cost growth. It is worth noting that the Air Force has set a high bar for approving requirements changes on the LRS-B, stipulating that they must be approved by the Air Force chief of staff.

To understand why the use of incentives can be more powerful in controlling costs than contract type, consider the very different incentives created under a firm fixed-price arrangement. The contractor is incentivized to minimize the scope of contract requirements to maximize its profits, while the government incentive is exactly the opposite. The government is incentivized to interpret contract requirements as broadly as possible to maximize the capability it gets. Rather than being jointly incentivized to control requirements, they are incentivized to argue about them. Having aligned incentives will play an important role on the LRS-B. As my colleague Todd Harrison and I haveargued previously, there will be significant bumps in the road in the development of the LRS-B. By choosing a contract structure that aligns incentives, the Air Force increases its odds of successfully navigating these challenges and making the LRS-B successful.

Research shows that cost incentives work, but as with all things in acquisition, they are not a panacea. Cost incentives work well to align the interests of the contractor and the government when cost estimates are reasonably accurate. That’s because the contractor’s performance is likely to put it somewhere between earning its minimum and maximum possible profit and by controlling costs, the contractor can earn more profit. However, once enough cost growth has already occurred to minimize the contractor’s profit, the contractor is no longer incentivized to control cost.

For this reason, another important step the Air Force has taken is to use a high-quality independent cost estimate in establishing the budget baseline for the LRS-B. While the original cost estimate for the F-35 assumed that the program could do significantly better than the cost history of previous programs indicated, the LRS-B cost estimate was based extensively on historical cost data. About a third of the development cost growth identified in the RAND study resulted simply from errors in the cost estimate. This is why Congress mandated the use of independent cost estimates in the Weapon Systems Acquisition Reform Act of 2009. By using a high-quality cost estimate for the LRS-B, the Air Force has significantly increased the likelihood that the contractor will remain incentivized to control costs. It has also reduced the likelihood of cost growth by simply setting the budget higher than the program office initially estimated in its own calculations, and also reportedly significantly higher than what the contractors initially submitted in their proposals.

The final incentive on the LRS-B program is potentially the most powerful. Industry profits are generally higher during production than they are during development. As a result, contactors are highly incentivized to complete development so they can get into production. By awarding the first five production lots for the LRS-B at the same time as the development contract, the Air Force has created an incentive for the contractor to finish the development program and get started on production. And generally speaking, the less time the development program takes, the cheaper it is likely to be. The Air Force remains in control of when this transition occurs, however, and can prevent this incentive from leading to excessive overlap between development and production.

For the reasons given here, and based on the many decades of mostly unsuccessful experience in attempting fixed-price development, it is highly unlikely that using a different contract type for the LRS-B would reduce cost growth. It likely would, however, increase the chance of program termination. Because the LRS-B is a key element of the Air Force’s future force, enabling access to highly defended airspace in conventional conflicts and enhancing U.S. strategic deterrence, that’s a tradeoff simply not worth making.

Andrew Hunter is Director of the Defense-Industrial Initiatives Group at the Center for Strategic and International Studies. He previously served in the Office of the Secretary of Defense from 2011-2014 including as Director of the Joint Rapid Acquisition Cell. He worked as a Professional Staff Member of the House Armed Services Committee from 2005-2011.

Image: Secretary of Defense Ash Carter, Secretary of the Air Force Deborah Lee James, and Chief of Staff of the Air Force Gen. Mark A. Welsh III announce the long range strike bomber contract award, Oct. 27, 2015. (Photo by Senior Master Sgt. Adrian Cadiz/Released)

No comments: