In my previous post ^[1] on the US Long Range Strike–Bomber (LRS-B), I promised to return to the likely cost-effectiveness of the future system. But first I’m going to back-track to examine the possible costs in a bit more detail. We’ll get around to cost-effectiveness in part three.

I noted that the USAF was estimating that it can build 100 LRS-Bs for the same cost as the 21 B-2s it acquired in the 1980s and 1990s. The finance industry is required to say that ‘past performance is not a reliable indicator of future returns’. Sadly, that isn’t true of Pentagon weapon system programs. Back in the early days of The Strategist I showed how Augustine’s laws of defence acquisition could be used to accurately cost the new Ford-class aircraft carriers ^[2], despite previous optimistic projections.

Now I’m going to change my tune and state that Augustine’s laws won’t work for the LRS-B. But first let’s note that they’ve done pretty well so far on predicting US bomber costs and the long-term trend is a well-established one. The graph below shows the cost of American bombers from 1929 to 1985, plotted in then-year dollars. (If we adjust for inflation, the graph flattens a little, but the growth is always well above inflation, at times as much as 10% per year, equivalent to a doubling of unit cost every seven years.) And Augustine’s data set includes some medium bombers. If we include only ‘heavy’ bombers (another relative term), the trend is stronger still.

Had we used Augustine’s growth law in 1990 to predict the cost of the B-2, we would’ve got the apparently stupid answer that each aircraft would cost $4 billion (all costs in US$) when delivered in 2000. Of course that turned out to be an overestimate—but not by as much as would’ve seemed likely at the time! Each B-2 cost $800 million flyaway off the production line, but the extraordinary R&D costs saw each of the delivered aircraft cost the American taxpayer $2.6 billion in total (program cost)—a totally unsustainable amount. I’ve plotted those costs on the graph, and the program cost isn’t far off the historic trend line.

^[3]

The reason I say that this law won’t hold for the LRS-B is simply that it can’t. Running the trend out to 2025 gives a figure of around $100 billion per aircraft. Given that the system blinked at the cost of the B-2, that’s just not plausible, even if the US economy was in a happy place and the perceived threat was high. I’ve also plotted the projected $550 million (adjusted to 2025 dollars) unit cost of the LRS-B and the optimism involved in that estimate is clear. To my mind, it’s likely that costs will be substantially higher than the projected figure, or the LRS-B isn’t going to provide all of the capabilities the USAF wants in a future air warfare environment, or both.

I suppose we should also allow for the possibility that the LRS-B will be so cleverly designed that it’ll buck the long-term cost trend and be a stunning performer. In which case we’ll have to wonder what all of the talented aircraft designers of the past were thinking when they produced the cutting-edge systems of the day at the cutting-edge prices of the day.

Much more likely is that the design of the LRS-B will be a compromise between state-of-the-art performance and cost. While it would ideally be large, thus giving it a substantial range, endurance and payload, there are suggestions ^[4] that it’ll be significantly smaller than the B-2. That’s a fairly direct way to keep costs down but a smaller aircraft will reduce the ‘throw weight’ of weapons it can deliver, a significant drawback in what’s meant to be a system that’ll help American forces defeat anti-access systems which rely on numbers as part of their effectiveness. Precision delivery is great, and precise small weapons can ruin a lot of enemy materiel, but strikes against hardened targets require a certain amount of heft. And targets like mobile missile launchers are relatively cheap to build—if a billion-dollar bomber can be defeated by hardening targets and building more million-dollar mobile missile systems, you have to wonder if it’s worth building.

If it’s to be an effective weapon system well into the 21st century, it’ll need a wide range of active and passive defensive measures, enabling it to penetrate contested airspace and strike even well-defended targets. Even then, it’s going to need to be able to deliver stand-off weapons with ranges of hundreds of kilometres, and possibly more in future. That raises the possibility of further increasing the stand-off distance and using non-penetrating launch systems.

Andrew Davies ^[5] is senior analyst for defence capability and director of research at ASPI.