WASHINGTON – As the Department of Defense focuses on ways to reduce sustainment costs for future programs, it could look to increase the use of services contracts and improve how it handles cost estimates for software upgrades.
Speaking to reporters Oct. 21, Frank Kendall, the Pentagon's top weapons buyer, said targeting sustainment costs is the next frontier for getting the price of defense programs down.
Kendall reiterated a statement, first made in September, that a fourth round of his Better Buying Power initiatives should focus on sustainment, noting that development costs average about 10 percent of a weapon system cost, production averages another 40-50 percent, and the rest comes due for upkeep.
"By far most of the cost we bear is in the sustainment phase. We don't have as good data on it," Kendall said, adding that officials need to develop "best practices" going forward.
What could that look like? Andrew Hunter, a former Pentagon acquisition official now with the Center for Strategic and International Studies, says one area to look at is how the department treats services.
"There's a lot of opportunity to do more partnering and rethink what is the critical government role, what's the inherent government function, and what is something industry can do," Hunter said. "Especially when you're trying to get the price of software talent and other technical talent. Are you going to be able to recruit [for the] government or are you better off working more closely with industry?"
In any war on sustainment costs, the Pentagon's office of Cost Assessment and Program Evaluation (CAPE) would likely play a big part. Jamie Morin, the current CAPE director, told Defense News in an exclusive interview that he sees "progress" in current sustainment cost controls, thanks in part to Kendall's focus on services contracting.
"A lot of the success there really just depends on regular management review of requirements and performance and pricing. That is just blocking and tackling at the individual contract level that has to occur," Morin explained. "In some cases, you need to pay top dollar because you need a top-dollar service. In other cases, you can drive to the lowest-cost provider because it's an area where you can afford to take technical or performance risks."
The Software Challenge
Another area Hunter predicts the Pentagon needs to deal with is how the department performs software-cost estimates. In particular, he wonders if the Pentagon’s estimates for those upgrades are operating on realistic timetables, given that future weapon systems are expected to feature software updates regularly, as opposed to a major system update every few years.
"The idea that I’m going to need to be adding and interacting with new capabilities every six months, as opposed to every five years, that’s where we have the potential to underestimate the extent to which we’re going to need to plan for," he said.
Morin acknowledged that projecting O&S costs for software upgrades comes with "wider uncertainty" than hardware estimates, but noted that some of the same considerations would apply across the board.
"You say you are going to put, each year, X number of people worth of software engineering and do it to fielding capability. Or you look at programs that have modular and spiral approaches in software that have continued to [be fielded] and you get a sense of what is the annual expenditure associated with that. Then you work it across, balance it across all the differences between the two or the multiple programs."
At a broad level, CAPE is working to improve how it predicts and handles sustainment costs for new programs – but Morin cautioned that the nature of those costs means analysts won’t know immediately if the new approach is working .
"This is one of those things where our Washington tempo doesn’t really support the adequate distance from the problem to understand it," Morin explained. "So we are making decisions now on programs as they go through acquisition milestones with much more visibility and rigor in estimating future O&S costs, but those won’t actually be realized, in many cases, for five or ten years."
"It’s incredibly unsatisfying, but I just have to counsel patience," he added.
Perhaps the most famous O&S cost estimate in history is that of the F-35 joint strike fighter. A Pentagon estimate looking out 50 years into the future predicted a $1 trillion cost for the lifecycle of the plane, a number that instantly lodged itself into headlines and continues to haunt the program even as costs come down.
For Morin, the trick to a long-term cost estimate is less trying to nail the prices of commodities like fuel, which is essentially impossible to do over a decades-long period, and more about understanding the potential points where program costs could be out of sync with economic growth.
"Are you using a material in building this [system] that you will need to replace with parts on this that is so rare, that the fact that you are buying thousands of this item is going to drive its price to spiral up faster than other commodities? Are you building something that is so software intensive, in a highly specialized way, that you are going to drive up wages among experts in this kind of software more than wages for programmers in general? Those are the things you have to think through a little bit," he said by way of example.
"The precise numbers, bottom line, are probably precisely wrong, but the process of doing the estimates identifies management levers for you and so can really help you converting broad affordability of the portfolio," Morin added.
That question of labor rates will apply to high-end software development, Hunter predicted.
"In some specialized areas DOD could be very much driving demand for the workforce and you could find yourself driving up prices in little niche technology areas," Hunter said.
Aaron Mehta was deputy editor and senior Pentagon correspondent for Defense News, covering policy, strategy and acquisition at the highest levels of the Defense Department and its international partners.