Making decisions based on sunk costs

I really like this blog, but too often he falls into the trap of focusing on sunk costs.

In this post, he blasts the Times for recommending cutting the F-22 program. While he states that more analysis should be done, focusing on what is given up and what is gained...the focus on sunk R&D costs is particularly awful.

In the defense world, many are worried about what they've spent in the past getting the system ready for production. But it doesn't matter, or it shouldn't matter in the analysis. Those costs are gone, whether you continue buying the F-22 or not. The analysis MUST focus on the marginal costs and compare them to the marginal benefits.

The hard choices facing defense really do lie in procurement. Part of this is the cost of these unproven systems (EMALS in CVN-78, DDG-1000, LCS, CG(X)) but also in a strategy of what defense needs. It always seems that the analysis is done on some perceived threats and meeting them, damn the costs of doing so. But what this ignores is that you can't guard against everything. Not only would it be extremely costly, but it is impossible. What defense really needs is for someone to come in with a figure for procurement and then say, "here's the money we can spend, what can we defend against?" and then let DoD figure out what they can get to meet some threats.