The pure-play defense companies which emerged from this process cannot generate growth for their shareholders if the procurement budget slumps, and will be under increasing pressure from more-diverse companies. That may not be a bad thing. This isn't the 1950s, when military aircraft were built in thousands and successful commercial aircraft in hundreds, and the cutting edge of almost every technology was in defense. Defense systems should harvest commercial supply chains for subsystems and materials, reducing cost and development time. One way to facilitate that will be for armed forces to buy more materiel (including components) from mixed-economy contractors.

It is time for defense customers to recognize that when it comes to big-ticket items, competition in the development stage is an expensive fiction. Again, this is not the era where you could lose one fighter competition and win another two months later. France has been managing without prime-contractor competitions since Dassault bought Breguet in 1971, and Sweden has never done it any other way. The challenge today is to identify what skills and technologies are unique and vital, and preserve them as efficiently as possible. State capitalism, industrial policy and mercantilism are naughty words today, but unfortunately are inevitable in defense.

Huge organizations, Krepinevich points out, seldom change their behaviors except in response to a major change in resources or catastrophic failure. When it comes to defense, better the former than the latter.