Retired air traffic controller and blogger Don Brown had a great item yesterday. “President Obama’s budget includes cutting a ‘long-range navigation system’. . . . It took me a day or two to figure out they were talking about LORAN — Long-Range Aid to Navigation. Now, you tell me, did anyone (besides me) mention that LORAN was supposed to be the backup system for GPS? . . . Well, actually it was the DOD (you know, the folks in charge of GPS) via the JPDO that was warning everyone. I just happened to include it in my blog. Remember?”
Then, Brown links to a James Fallows post reporting on a GAO finding that the “U.S. Air Force, which runs the GPS satellites, has not managed to get new ‘IIF’-model satellites ready in time to replace the ones that are wearing out.” Brown: “So, we’re going to eliminate the system that GPS made obsolete, just in time for the GPS system to become unreliable. I assume the thought that the left hand of the government doesn’t know what the right is doing pops into your mind also.”
This reminded me of a new article by Jonathan Last about the natural process of knowledge decay in institutions. He documents the $92 million process of reconstructing the formula for “Fogbank,” a substance included in late-Cold War nuclear warheads whose formula had since been lost to history. Knowledge is lost, distorted, housed in human capital. Think of what happens when the admin in your office who’s been there for twenty-five years retires. No matter how much she writes down, she still takes away (and does not replace) a vast amount of institutional memory and knowledge.
And when you’re dealing with an institution as large and complex as the U.S. government, these kind of gaps are bound to happen frequently. The FAA thinks, “GPS, it’s a great system, let’s make it the basis of air traffic control.” (This is not a post on merits of Next Gen, FYI.) The Pentagon doesn’t bring a replacement for the current GPS system on line fast enough. And the OMB, eager to find redundancies in the federal budget, cuts the GPS back-up without knowing about delays in deploying its successor system. This is natural in giant institutions. We should probably be more surprised when systems actually work.