The current portfolio is smaller yet more expensive that last year’s portfolio. This is the third time in 10 years that a decrease in the number of portfolio systems was not accompanied by a corresponding decrease in expected costs…. The average total acquisition cost per program has continued to increase over the last decade. In 2008, the average cost of an acquisition program was $18.3 billion, while the comparative average cost for 2018 is $20.6 billion… In 2018, two programs entered the portfolio and six exited.
While the number of programs in the portfolio has declined since last year, the average age of the programs that have remained has steadily increased since 2012.
[The illustration below shows] how the current portfolio’s total acquisition cost is apportioned between older programs that were initiated prior to 2010, newer ones initiated since 2010, and the F-35 Lightning II Joint Strike Fighter—DOD’s largest acquisition program, initiated in 2001.
That was all from a May 2019 GAO Report, Weapon Systems Annual Assessment. A couple observations:
First, the average MDAP is 14 years old. That means the average acquisition officer works on a program that he or she had absolutely no input into how it was devised and executed — largely managing the existing plans laid down. Indeed, there’s probably no one on the program who was around during the early years.
Second, only 17 percent of the *entire* MDAP portfolio by dollar value was initiated after 2010. There is such a small opportunity for leaders to affect new programs, because they are very infrequency and massive in size. The total dollar value of programs initiated after 2010 is quite a bit less than the F-35 program alone!
By the way, for 183 Major Defense Acquisition Programs contracts, the top 5 contractors received $262 billion while the rest received less than $100 billion. That share must be bigger today, since Raytheon wasn’t included in the top five and merged with United Technologies (which was in the top five, unexpectedly for me).
We took a look at this a couple of years ago, Eric, and found that there are some things going on that make it really hard to understand the trends here. The big one is that what counts as “a program” has been changing over time.
Once upon a time, an MDAP generally had a start and a finish, and it developed and fielded one end item. That end item might be a new system (e.g. F-16 or Bradley Fighting Vehicle) or a modification of an existing system (e.g. F-16C, or M1A2 Bradley).
Increasingly, that’s not how it works. Many MDAPs are now ongoing capability providers, developing and producing a sequence of systems for the same general set of missions. For example, the AMRAAM missile program started in the late 1970s, and is still going strong. It has produced 7 or 8 distinct models of radar-guided air-to-air missile, including both incremental upgrades and major redesigns. This makes the AMRAAM program look old and expensive by GAO measures, but this is an artifact of not calling it a sequence of MDAPs (as was done with the parallel AIM-9 missile series).
An even more striking example is the transition of the Evolved Expendable Launch Vehicle (EELV) program, which designed and built rocket boosters, into the National Security Space Launch program, which provides ongoing booster development and launch services. There is no plan for NSSL to ever “finish”, which makes it a bit odd to talk about its age or its cumulative size in the same way as something like F-22 or DDG-1000.
Given that it is often essentially arbitrary whether a new product gets developed within an existing program or in a new program, and that the bureaucratic incentives are to avoid starting a new MDAP if you can reuse an old one, and that more programs are transitioning from a product orientation to a service orientation, it is not surprising that extant MDAPs have been around longer and have spent more money than was true in the past, other things being equal.