Let me give you one specific example. The leverage provided by stealth aircraft in the First Gulf War is fully illustrated by this particular attack. It was the first non-stealthy attack on a target in the Basra area. Shiaba airfield. The attack package consisted of 4 US Navy A-6s dropping bombs along with 4 Saudi Tornado bomb droppers, 5 Marine AV-8Bs jamming radars, 4 Air Force F-4Gs taking out one surface-to-air missile systems, 17 Navy F-18s taking out other SAM systems, and 4 F-18 escorts because MiGs showed up, and then 4 drones that cause the enemy radar to radiate. If you do the math, that’s a total of 41 aircraft, a total of 8 of them dropping bombs on one target.
At about the same time, I had 20 F-117s dropping bombs on 28 separate targets. That’s less than half the aircraft hitting 28 times the number of targets. So stealth — while expensive on a per-unit aircraft basis — is at least an order of magnitude less expensive when considering cost per desired effect…. One B-2 could accomplish what would have taken 75 combat aircraft during Desert Storm, and at a fraction of the cost.
That was David Deptula, former Acquisition Talk guest, on the Aerospace Advantage podcast. I see this as a renewal of the original systems analysis concept, which had always since the late 1940s looked to measure costs and effectiveness. Of course, the first RAND systems analysis was a blunder, recommending a turbo-prop rather than turbo-jet bomber aircraft. And then in the McNamara years, the analysts said nuclear-powered aircraft carriers weren’t cost effective enough. One can always say that they the pencil necks simply didn’t understand the problem, but one must wonder whether people today can do it so much better.
I think there will always be a problem clearly defining along a single parameter the right measure. Incommensurable attributes have always been an issue, because you can’t compare the cost per pound of explosives delivered on a fixed target with the range of the strike package, or sortie rate, or ability to counter air-to-air, or whatever else is of importance. Of course, cost per fixed target destroyed is definitely useful, I’m just pointing out the subjectivity of the selection and its relation to other important measures.
Another issue is the inability to calculate the actual cost effectiveness until systems are actually used in combat. Realistic operational exercises can give indication, but it still requires building and testing systems. I one-hundred percent agree good cost-effectiveness measures should be used for fielding decisions, just as Deptula suggests, but I wouldn’t extend that to R&D choices. There’s a benefit to fuzzy exploration there. For example, how could you possibly know the cost-per-target-killed of a suicide drone swarm until you build it and try it out? One of the big problems in that isn’t just the technical aspects, but the back and forth between countermeasures.
And this brings up is contingency risk. The recommendation to go with F-35s over F-15EXs, for example, makes sense so long as F-35s can operate without enemy detection or tracking. The moment the enemy has a method to track the F-35, then the calculus immediately shifts. What is most likely is this long ambiguous middle phase where the enemy has some methods to detect and a harder time tracking, but learning and improving until stealth becomes at some point irrelevant. When system choice depends on contingency risk, there’s no way to optimize to a single outcome (except choose what is hoped to be the most likely).
I’ll leave you with some great lines from Deptula.
Unfortunately Slick, the F-15EX being forced on the Air Force by cost-focused analysts in the Pentagon, is evidence that combat effectiveness is less of a consideration in acquisition decisions than simple budget accounting. I hate to be that blunt, but that’s the fact of the matter. The Pentagon is still stuck in World War II thinking. This is just going to get worse as we pursue concepts like mosaic warfare, joint all domain command and control, and others, where the types of technologies we’re going to buy are going to radically alter how we fight. These concepts, a mix of high end assets like a B-21, while incorporating smaller simpler systems that team.
… We’ve got to change, or the math will lead us to build the wrong systems because to an accountant — or DoD leader who only has business experience — the only thing that matters to them is cost. We’ve got to put effectiveness back into the cost-effectiveness equation when determining what systems yield the Department the best value.
I want you to consider how the whole DoD process was radically changed in the 1960s to make outcomes-based decisions using costs and effectiveness, and these principles have been at the heart of acquisition for decades. Yet, it still doesn’t work, and puts attention back on cost because that’s easiest to measure. The very attempt to articulate costs and effectiveness too early drives bad decision making. Quantification is necessary, but it is secondary to judgment of professionals — humans need several different models for looking at complex problems — a single model should never drive professionals’ decisions.
Leave a Reply