Economics professors are more likely to walk out of bad movies than their colleagues in the humanities. Hold onto that thought for a moment; I’ll get back to it.
First, let me apologize for not having written for some time. Our business has been, to a certain extent, stumbling through limbo these past few months, at least in the United States, and the profound uncertainty has sent me back to more fundamental research looking for answers. But just now, something really stirring has reappeared, an idea so offensive and casually destructive that I might call it an intellectual kraken. Yesterday, that is, Forbes’ military-industrial columnist hauled ashore a really foul one. Here’s how the installment starts:
One of the most unsettling facets of federal finance is the way the government devalues past investments. The political system is so focused on the next budget—and the next election—that it ignores sunk costs. Thus, every program termination is considered “savings,” without regard to the money that was spent to get the project in question to its current state.
— Loren Thompson, “How To Waste $100 Billion: Weapons That Didn't Work Out,” Forbes, 19 December 2011
The preceding emphasis is mine. I didn’t catch the column when it first appeared yesterday. Rather, a colleague in Virginia sent it with the advisory that I not read too far in, “or your head will explode.” Better, as he continued, “you don’t even need to read past the first paragraph (and in fact I’d encourage you to stop reading there to avoid wasting your time). It’s astounding that Forbes is even willing to publish him at this point.” I didn’t say it; I’m just reporting it—right?
Now, on that comment about bad movies. Such was the finding of Richard Nesbett and two colleagues in their survey of professors and students at the University of Michigan some years ago: the economists, and even their economics students, better knew how to cut bait. If the movie really is a turkey, then the time and money spent there already is no basis on which to proceed. The reference is found in Alan L. Otten, “Economic Perspective Produces Steady Yields”—there’s an apt title—in the Wall Street Journal of 31 March 1982, on p. B1. I found that reference in the first microeconomics textbook I used during my master’s program at the University of Chicago: B. Peter Pashigian’s Price Theory and Applications (McGraw-Hill, 1995). The citation is found therein on page 228, in a box labeled Why “Never Give Up” Is Not Always the Best Advice. Get the point?
Certainly an old friend in the Marine Corps did; as he related to me on seeing Loren’s column, “I got a B– in intro micro and the only two things I remember are (1) always pay down principal on your mortgage first, and (2) sunk costs must be ignored in decision-making. This is disgraceful.” For my part, in writing all this up, I didn’t need to consult notes from my intermediate or advanced microeconomics courses; by the time we had gotten there, we were rather expected to grasp at least that much.
But contrary to advice, I did continue reading, and found that the rest of the column was a series of rolling assertions about how nearly every big, legendarily underperforming American weapons program of the past decade had merited continued funding, because the government had already spent so much. And that’s just the classic sunk cost fallacy, as the textbooks relate it. Monies spent are monies gone, unless they have been capitalized in an already useful asset. The measure of which to take stock is the costs of the alternatives: how much would need to be spent if starting anew, and how much would be lost—in defense, often admittedly an intangible—if we simply decided not to have one at all.
Since it’s topical, and heck, because Loren is a consultant to Lockheed Martin, let’s take the F-35 Lightning II Joint Strike Fighter program as an example. No one in his right mind expects that the full program will come to fruition. But just yesterday, the Japanese government announced that it would order perhaps fifty of the planes. That’s not surprising, as the Japanese Air Force had been salivating about the presumably stealthier and commercially more elusive F-22 Raptor for years. The decision is also an endorsement of the survivability of the program; clearly Tokyo doesn’t think that it’s on the verge of termination. But if sequestration does come to pass, and all options are indeed on the table, killing the JSF becomes tempting, because removing that chunk of money solves much of the budget problem fast.
The calculation then does not involve how many billions have already been spent—some set of assets has been attained for those monies, but the demonstrators, test articles, and LRIP aircraft have been bought, whether half as much or twice as much was spent. The capitalized value is an accounting measure, but of little raw value for decision-making. Rather than that, we’d want to hear about the probable cost to finish the program, the relatively known cost of an equal number of F-18Es or Rafale Ms, and an assessment of what those Boeing or Dassault aircraft couldn’t do that the Lockheed aircraft presumably could have done. Note the difference in language: any assessment of the potential of the unfinished program must be probabilistically discounted, and its future costs viewed similarly within a probability distribution, because the facts are simply not all in.
Thus we find the general problem grossly misidentified. It’s not that generals, admirals, air marshals, bureaucrats, and politicians all don’t understand the concept of sunk costs. Most of them do. Plenty of them, across the world, have been through places from the Kennedy School to the ENA, and a host of less luminary institutions, at all of which I’m pretty certain the concept is taught. It’s not that military decision-makers fail to account explicitly for sunk costs, for which they should not be accounting at all. Rather, it’s that they sometimes do account implicitly for them, in a group-thinking way, without admitting what they’re doing. They conceive of the future in terms of the past, and then throw good money after bad.
The specific result, to play upon one of Loren’s particularly egregious examples, is zombie programs like the so-called Future Combat System, the bureaucratic walking dead that stumble along for years without their managers being able so much as to specify whether the gun will be 105 mm, 155 mm, or something in between. The fiscal tragedy is not that the program died after so much was spent, but that it was ever allowed to start in the first place. Programs as these don’t need mollycoddling; like the sunk cost fallacy behind them, they’re cash-devouring monsters. What they need is what a Navy program manager friend of mine once termed “forty-five caliber therapy.” Damn the sunk costs—someone needs to shoot them in the head, and twice to be sure.