Death Stars Bad, Droids Good

Wednesday, May 7th, 2014

When Dan Ward introduced his daughter to the Star Wars movies, she came to this conclusion: “Daddy, they shouldn’t build those Death Stars any more. They keep getting blown up.”

Dan Ward is a Lieutenant Colonel in the US Air Force, an acquisition officer, and he agrees that Death Stars are overrated military investments:

In addition to the observation that Death Stars “keep getting blown up,” which prevents them from being very effective, I also noticed they are inevitably over budget and behind schedule. Then I read an interview in which George Lucas himself anointed the humble astromech droid R2-D2 as “the hero of the whole thing.” The conclusion for how the military spends its money was clear: Death Stars bad, droids good.

This counterintuitive finding was not entirely unexpected. Since the early 2000s, I’d been developing and experimenting with an approach to innovation based on restraint — tight budgets, short schedules, small teams, and highly focused objectives. This was a relatively unusual approach for an Air Force officer to take, given my service’s longstanding preference for spending decades and billions developing enormous, multi-role systems. Indeed, these mega-projects are the most prestigious and the surest path to promotion in the military, which helps explain why we build so many of them.

And yet, while they may be impressive to work on, Death Stars contribute very little to the fight, partly because they’re always behind schedule and partly because they explode on a regular basis. This pattern is not limited to the movies.

The Army’s Comanche helicopter (22 years + $7 billion = zero helicopters) and the Joint Tactical Radio System (15 years and $6 billion before it was cancelled) are just two recent examples. But as the Government Accountability Office helpfully notes, these huge projects overall tend to “cost more, take longer to field, and often encounter performance problems” — not unlike the Empire’s moon-sized battle station. This means there are economic as well as operational reasons not to build them, which perhaps explains why I get a very bad feeling whenever I’m around one.

Real-life performance data shows that the most important and high-impact technologies are not the gold-plated, over-engineered wonder weapons that turn majors into colonels, colonels into generals, and young Jedi apprentices into Sith Lords. Instead, data suggest the real winners are humble, simple, low-cost products made by small, rapid innovation teams — the type of projects that don’t attract much attention from the press or from the brass because all they do is get the mission done without any fuss.

Defense analyst Pierre Sprey has written extensively about these “cheap winners” and “expensive losers,” a pattern which also showed up in my career.

He goes on to mention some real-world projects. Oddly, he never once mentions Lockheed’s famous Skunk Works.


  1. Toddy Cat says:

    The “Military Reformers” like Sprey almost always overstate their case and cherry-pick their data, which leaves them open to criticism, but they do have some valid points and insights. Even if one believes in complexity with regard to weapon systems, there’s no doubt that we pay too much and get too little.

  2. I agree, Toddy Cat. I find Sprey’s analyses refreshing; I’m especially interested in his strong pro-APC, anti-IFV perspective. He, and those like him, do tend to overdo things in the manner you describe. The M48 vs. M1 comparison in the linked document is especially groan-worthy, for instance.

    I have some personal experience with the DOD’s preference for big, complex, contractor-fattening systems of marginal value. For the last couple of years I’ve been repeatedly trying to interest the Army or a contractor in developing a concept I originated that would represent a cheap, easy-to-develop/deploy, and robust way to solve a serious problem and greatly improve the combat performance of certain systems. There was essentially no interest expressed, and earlier this week I saw that the concept is now being developed/tested by another country’s armed services.

  3. Toddy Cat says:

    “There was essentially no interest expressed, and earlier this week I saw that the concept is now being developed/tested by another country’s armed services.”

    That’s gotta drive you nuts. With any luck, it’s not being developed by a potential adversary…

  4. Thankfully not. Yet.

  5. Toddy Cat says:

    I believe that it was Zhukov who said, “Complexity should always be done in the simplest way possible”. And yeah, “Yet” is probably the proper phrasing. Who knows, the country in question might fail to recognize gay marriage or take steps to stamp out cishet privilige, thus rendering them our enemy.

  6. I had forgotten about that Zhukov quote, but it is very apropos.

    I had actually intended my “yet” to imply that it was not yet under development by a potential adversary of the US but likely soon would be, rather than your (amusing) interpretation that the country in question was not yet an adversary but might become one. In this case that seems rather unlikely, but in general you make a good point.

  7. Mike in Boston says:

    “Cheap winners, expensive losers” also makes for a good investment strategy: it seems to me that when low-margin producers of small, low-cost equipment start coming out with bigger, higher-end stuff, they generally eat the lunch of the slow-moving, high-cost, high-margin incumbents.

Leave a Reply