This article was inspired by thoughts brought to mind by Poetrooper’s F35 article yesterday.
. . .
Here’s a conversation you might hear somewhere inside the DC Beltway:
Congressional Staffer: “So tell me, Mr. Program Manager – just what will one of those systems you’re developing for DoD cost? Senator Whositz wants to know.”
Program Manager: “Well, that depends. What cost does he want?”
Congressional Staffer: “C’mon, bud – don’t play games. What will each of the systems cost?”
Program Manager: “I’m not playing games. What cost does he want?”
Congressional Staffer: “Are you trying to get Congress to cut your program’s budget?”
Introduction.
The above might sound like an Abbot and Costello comedy routine – but it’s not. When it comes to weapons systems, a question seemingly as simple as “How much will it cost?” can be hard to answer – because DoD, in it’s “infinite wisdom”, has defined at least eight distinct costs associated with a defense system under development. Each means something different, and each is calculated differently. And several (but not all of them) can be used, along with the total projected number of systems to be produced, to come up with a valid “unit cost” for the system by dividing the cost in question by the number of systems to be produced.
And, yes – which one is quoted to the press often depends on politics and desired “spin”.
The purpose of this article is to give an overview of the ways costs for DoD systems are calculated. Just as importantly, it will also indicate what is – and what is not – included in each cost.
To help understanding, I’ll try to relate these costs to something we all presumably know all too well: the cost of owning a car.
Obligatory warning: the subject matter can be a bit soporific. Might want to grab a caffeinated beverage or two before reading further. (smile)
. . .
The eight costs DoD defines for a system under development are the Development Cost; the Flyaway/Rollaway/Sailaway Cost; the Weapon System Cost; the Procurement Cost; the Program Acquisition Cost; the Operations and Support Cost; the Disposal Cost; and the Life Cycle Cost. All of these except Disposal Cost have been around for a while – since at least the release of DoD 5000.4-M in December 1992, and perhaps before. The Disposal Cost is a relative newcomer.
So, what do these costs mean, and how are they calculated? Glad you asked.
1. Development Cost.
Development cost is pretty simple. It’s essentially the cost of virtually all activities needed to develop a new weapon system from Day 1 up to the point of production. This includes research and development (R&D); design and engineering activities; much if not the vast majority of the system’s test activities; extensive modeling and simulation (or model development), if required; development of new technology or materials, if required; construction of prototypes and test articles/systems; and other activities. In short, it’s what you gotta do before you set up the production line to make sure the system “works” (the quotes are intentional).
Development cost may be relatively small for a system using mature technology. For one that was/is using “bleeding edge” technology (like the B-2), it can end up being a huge part of the total cost of getting the system to the field – particularly if the number of intended systems is cut dramatically after the program has begun. (I’ll have more to say on that in a sidebar near the end of this article.)
To reference this to buying a car, it’s what Chevrolet (or Honda or BMW or Hyundai) spends to design the model you bought. The buyers of new cars pay this. You’ll probably never know how much of the purchase price it was for your car.
2. Flyaway/Rollaway/Sailaway Cost.
This is generally the most-favorable (e.g., lowest) cost for the system. It is the cost of producing the system, including the costs associated with setting up and running the production line. No spares, no essential associated equipment, no personnel to operate it or maintain it, no supplies or fuel. Just the end item.
This cost will tend to go down over time. The more systems are produced, the more widely the cost of setting up the production line gets distributed – and the more proficient the workforce tends to be at producing the item. Both a current and an average flyaway/rollaway/sailaway cost can be calculated. The current one is almost always the most favorable (lowest).
For buying a car, this equates to most – but not all – of what you paid for the vehicle. No insurance, no maintenance, no associated tools or other equipment you need – just part of the cost of the car. (Part of what you pay is actually the manufacturer recovering some of their development costs and the dealer’s overhead, but for commercial items this isn’t information that’s typically readily available to the consumer.)
3. Weapon System Cost.
This cost is defined as the cost of the system, plus any required equipment needed to make it fully functional. For example: if the system requires an already-existing generator or a prime mover (truck or track) or radio to make it fully mission capable – or if it requires special test equipment or tools that aren’t already in the system – then those costs get added to the “flyaway” cost to produce the Weapon System Cost.
For buying a car, there isn’t any real hardware equivalent – unless maybe you were foolish enough to buy a Volt or some other fully electric vehicle that required you to install a charging station in your garage. In that case, the cost of the charging station would be added. However, mandatory charges such as taxes, title fees, and required insurance could arguably be added here – though these IMO better fit into the “Operations and Support Cost” category.
4. Procurement Cost.
The procurement cost for an item takes the Weapon System Cost and adds a few things: initial spares and (for ships) often also includes outfitting and post-delivery costs. If memory serves, it also includes costs associated with initial training and transportation of the system to the field users; however, I’ve been away from the “game” for a bit, and no longer have access to some of my former references. I could be wrong about those latter two items.
If you tend to keep a spare set of plugs and filters on-hand, adding the cost of the first set of each you’d have to buy for a newly-purchased car might be an analog to the added costs here. If you had to travel to pick up your new car and stay overnight, those costs could also be considered an “add-on” here – as could the charges for “dealer prep” and “transportation”.
5. Program Acquisition Cost.
Program Acquisition Cost is the sum of Development Cost and Procurement Cost. As such, it’s one of the overall figures that actually is meaningful. Taking this figure, and dividing it by the projected number of such systems to be produced, gives a meaningful answer to the question, “How much will each of these systems cost?”
Unfortunately, this is why reducing the numbers to be produced often causes the overall unit cost of a system to skyrocket. If a system’s Development Cost was $1 billion, that adds $1,000,000 to the cost of each system if you plan to produce 1,000 of them ($1B/1,000 = $1M). If the number produced is cut to 20, that means each system now gets ($1B/20) = $50,000,000 added to its unit cost.
[Sidebar: This is precisely why the B-2 ended up being a $2 billion aircraft. Original plans for the B-2 called for 132 aircraft to be produced. When this ended up being cut to 21, a huge development cost of substantially over $29 billion was “split” among only 21 aircraft – yielding a Procurement Acquisition Unit Cost for the B-2 of roughly $2.13 billion EACH ($44.75B / 21). The flyaway cost for the B-2 was a bit over 1/3 of that, and would have continued to drop as more were produced. However, even in that case IMO its average cost likely would have still been somewhere well above $500M each.]
For the automobile example, this is much of what you paid for the car, with the other adds noted above for initial spare parts, tag, tax, title, etc . . . . As noted above, the development costs are hidden in the sticker price of the car; you’ll pay a pro-rata portion of them regardless, but you’ll probably never know how that was broken down.
6. Operating and Support Costs.
Operating and Support costs are simple in concept. They’re what it costs to operate the system after it’s deployed. This includes operator salaries; fuel and lubricants; maintenance; post-deployent engineering upgrades and fixes; spare parts other than initial spares, and all other costs associated with normal operations and maintenance.
For most systems other than space systems, this is actually the single biggest cost category – anywhere from 50+% to 80% a typical military system’s life cycle cost, with somewhere around 2/3 being typical for non-software systems. Space systems are the “outlier” here – they have much lower operating and support costs (and consequently higher development and procurement costs) than other military systems. That’s generally because servicing a space system in orbit is still somewhat problematic. (smile)
Since they are paid out of O&M funds (or, for military salaries, the military personnel appropriation), you virtually never hear about these costs when systems under development are discussed. That’s too bad, because they’re typically the lion’s share of the cost of any weapon system.
An automotive analog is the recurring cost of license plates/insurance/oil/gas/filters/tires/brakes/other maintenance and repair that the vehicle requires post-purchase. Drive a lot – or buy a lemon – and these costs are likely to exceed the purchase cost of the vehicle within a few years.
7. Disposal Cost.
Although it seems as if we do, we really don’t keep military systems around forever. Eventually they get retired. But you can’t just haul them off to the local dump at end of life.
Demilitarizing a former weapons system isn’t free. The cost may be low (“SGT Jones – go check to see that none of the troops stuffed anything sensitive or classified in the glovebox or under the seats of that deuce-and-a-half we’re sending to PDO or left any personal gear there.”). Or it may be high (think prepping a ship for sinking as an artificial reef – which is why many end up sold to shipbreakers for scrapping). It may actually result in a gain for Uncle Sam – e.g., the proceeds of selling a former military vehicle at auction. But it’s only coincidentally zero, so it’s a cost to be considered.
For an automotive example, think selling that old clunker in the driveway – or having to pay someone to tow it away.
8 Life Cycle Cost.
This is the true “whole enchilada” with respect to a DoD system. It’s the total cost of buying, owning, using, and disposing of a system from Day One to the day the system is completely “washed out” of DoD’s inventory. It’s the sum of Program Acquisition Cost, Operations and Support Cost, and Disposal Cost. Counter-intuitively, it’s also not one you hear that much about.
For an automobile, this is every penny you ever spent on that car. Every last cent. And for a privately owned auto, that neglects the value of your time, business use of car possibly excepted.
. . .
So What?
So, how are these used? Well, as one DoD organization frequently puts it: “That depends.” (smile)
A PM trying to defending their program will generally try to make it seem as efficiently run and as low cost as possible. That means they’ll likely focus any public statements on their system’s flyaway/rollaway/sailaway cost – typically the smallest of the costs – and will downplay others. His or her position will be that, “Future improvements will keep those other costs under control, and we believe we’ll be able to lower them over time.”
Conversely, a system critic will focus on making the system appear costly. Life Cycle Cost will be something he or she is VERY interested in determining. Barring that, they’ll focus on either Operations and Support or Program Acquisition Costs. The latter is generally available.
Finally, when it comes to the unit cost of a particular system the number of items produced can matter A LOT. Cut the numbers produced on a system with only minor Development Cost and there’s not a big effect. Cut the numbers to be produced on a system with a huge R&D cost, and you could well triple the cost of the system. That’s pretty much what happened on the B-2.
In any event, a meaningful number for the real unit cost of developing a military system under development can be obtained from two pieces of information: the Program Acquisition Cost and the total estimated number of systems to be produced. Using those two, a “best case” overall projected average unit cost for the system in question can be calculated. I say “best case” because reducing the number to be produced virtually always raises the overall unit cost of any system – the development cost gets spread among fewer units, and you see less production line improvement due to workforce learning.
This is why I’m VERY concerned with the F-35. Per info I found yesterday while considering Poetrooper’s article about the F-35, the Program Acquisition Cost for the F-35 is currently estimated at approximately $406.5 billion. The total number to be produced is currently estimated at 2,456.That yields a “best case” unit cost for the F-35 of $165+ million each. That’s twice that of the F/A-18E/F.
Reduce the number to be produced – either through Congress cutting the budget or foreign customers deciding that it’s simply too damn expensive – and we could conceivably see a unit cost of 2 or 3 times that. Add to that the fact that it simply can’t adequately replace the A-10, and, well . . . the term “white elephant” comes to mind.
Yes, that could indeed happen. It already did once with the B-2.
A weapon system that is too expensive to buy in sufficient quantity, or is too expensive to use, is worse than not having one at all. It won’t do the job – and the money wasted on such a white elephant could easily have been used to buy something useful instead.
. . .
Well, I hope this helps. Might have a detail or two wrong above (I have been away from that for a few years now), but I’m pretty sure it’s in general correct.
References:
https://dap.dau.mil/acquipedia/Pages/Default.aspx
https://biotech.law.lsu.edu/blaw/dodd/corres/pdf/50004m_1292/p50004m.pdf
https://dap.dau.mil/aap/pages/qdetails.aspx?cgiSubjectAreaID=8&cgiQuestionID=108656
