More Sunk-Cost Thinking on Iraq War

Bruce Wydick, a professor of economics at the University of San Francisco, has written an interesting OpEd in USA Today about sunk costs and the Iraq war. Here is his lead:

Our inability to think clearly about sunk costs is impeding our ability to make clear decisions about our involvement in Iraq. Failing to correctly identify sunk costs (those that are irretrievable), and deal with them properly, biases our decision-making in favor of prolonging the war.

This is not the first time that someone has publicly connected war policy with the cognitive misstep that behavioral economists have made widely known as the sunk-cost fallacy. As noted earlier here, Barack Obama made the same point some eight months ago. Before economists came along to name this phenomenon, I believe it was known as “throwing good money after bad.” I don’t mean this as a statement on the war itself; I mean it as a way of looking at any investment or expenditure.

Speaking of old axioms that economists have dressed up in new language: I am surprised that, in response to Levitt’s post about why High School Musical 2 was so much worse than the original movie (if indeed it was so much worse; but I’d have to agree for the most part), nobody mentioned the simple possibility of regression to the mean. In other words: a hit is a hit because it was an anomaly, and the sequel is mediocre because anomalies are anomalous. This used to be known as: lightning doesn’t strike the same place twice.

[Addendum: after I wrote the above paragraph, one commenter came a little bit close to suggesting regression to the mean as an explanation: “‘HSM’ was lightning in a bottle; there was no chance to replicate that,” wrote Susan.]

(Hat tip: Christopher Arnold)


I would imagine that a sunk-cost analysis of the Iraq War would provide a better argument to stay in Iraq. Assuming Iraq War was mismanaged, leaving Iraq would not recover those sunk costs. However, if we start to properly manage the war, we might be able to prevent another middle eastern state from becoming a terrorist safe haven with little cost (compared to how much it would cost to rid terrorists of some other safe haven). The U.S. already has troops and equipment in Iraq, and I would imagine that the majority of the costs of saving Iraq from another dictatorship have already been irreversibly spent, or as economist like to say, sunk.

Mario Ruiz

Dear Stephen,

Today I am going to play "contrarian."

The reason to be in Iraq is a geo-political. Therefore the economic reasons to control the Oman Gulf not only are economic, but also must be at least equivalent to the sunk cost of the war.

The Texas economy has flourished under the reconstruction of the oil industry of Iraq. If leaving Iraq would be the same as leaving the Gulf on enemies hands, I can stop wondering how much would be the oil.

I am not sure Bruce is one of the first thinking in the sunk cost of the war. However it seems to me he is accounting only the direct costs, not the indirect or total effect of leaving.

This is an economical thinking does not imply US must stay or leave.

Although my area is technology, as a CEO of company I need to know economy.

Mario Ruiz



Please read the the post. You write:

"I would imagine that a sunk-cost analysis of the Iraq War would provide a better argument to stay in Iraq. Assuming Iraq War was mismanaged, leaving Iraq would not recover those sunk costs."

The point of the post is that that kind of thinking is a fallacy. As Dubner and Wydick note, sunk costs CANNOT be recovered, that is their very nature, otherwise they would not be sunk.

Phil Steinmeyer

I think a failure to separate sunk costs from avoidable (future) costs permeates thinking on both sides.

On the pro-war side, there is the obvious, "don't let the sacrifice of those who've already died be in vain" concept, alluded to by Dubner above.

But on the anti-war side, I think the fallacy is even more prevalent. The thinking is, roughly, "things have gone badly so far, and going to war was a mistake, therefore, staying at war is a mistake."

The problem is compounded by the political spoils system at play. Endorsing war going forward is tantamount, politically, for a decision to endorse the war historically, and particularly, those who led the initial war effort (i.e. Bush, and perhaps Republicans as a whole). OTOH, Bush, like other leaders who've seen their policy decisions turn out badly, has been slow and reluctant to make the necessary policy changes, because doing so would be tantamount to admitting poor decisions in the first place. It seems fairly clear to me (though I am not a particularly informed observer), that Rumsfeld's decisions had been poor, and that he should have been pushed out the door in, perhaps 2004 or early 2005. Whether you think the war was/is a mistake, so long as we are at war, we should make our best efforts with our best people. But Rumsfeld was kept on the job far too long.



Obligatory poker analogy, anyone? Bush represented a big pair (WMDs) before the flop (invasion) and bet big. We got called, we didnt have a big pair and the flop came and everyone else made a hand. Now they are betting and we need to decide whether or not to chase our money to see two more cards.

Of course I realize that poker is zero sum and the war's benefits might not show up for 50 or more years. But I would like to think that our president should be able to take down a hand or two in a high stakes game.



"if we start to properly manage the war, we might be able to prevent"

That's a big "if" and big "might." Then, consider that Iraq was not a significant "terrorist safe haven" before we invaded, you realize further fallacy in that conditional statement.

How about this...*if* we stay and manage to sort things out in Iraq, and *if* we go invade some other Middle East country, we'll get into exactly the same mess and *create* (not avoid or eliminate) yet another terrorist safe haven.



Off the topic of Iraq, another good example of economists theorising homespun proverbs is "The theory of portfolio diversification". Also known as "don't put all your eggs in one basket". Omlette anyone?


One problem with all of this: No one has ever established what would be an appropriate budget to accomplish the Iraq war's posited goals. $50 billion? $1 trillion? $3 trillion? How much is a reconstructed, stable Iraq worth to U.S. taxpayers?

Josh Weinberg

I'll preface this by saying I am a member of the military. I am also starting a graduate program at USF under Prof. Wydick this week.

While sunk costs will not change no matter what happens, variable costs change according to the chosen course of action. Therefore, variable costs should be considered by rational decision makers while sunk costs should be ignored. Wydick uses the following as an example of sunk cost thinking:

"A current withdrawal would dishonor the memory of those who have made the ultimate sacrifice for the sake of liberty."

The key word here is memory. Memories are held by people who are still here: friends or family, or even a national memory. Although those who have died are gone and will not come back, memories are not static, change often according to many circumstances and differ across a population. Therefore, by definition, "dishonored memories" is a variable cost, as the idea of dishonor and memory would change according to the course of action. Besides the fact that some may see an immediate withdrawal as "dishonoring memories" while others will see continuing the war as "dishonoring memories." Complicating things further is framing the issue in relation to "the sake of liberty."

An actual, direct example of sunk cost reasoning would be, "We should continue the war because so many have already died fighting it." Wydick probably didn't use this example because its rare. Of course, the opposite, "We should NOT continue the war because so many have already died fighting it," is used very often. This version is just as vulnerable to the sunk cost fallacy.



This post is important although it is not receiving many comments. Perhaps people just don't have much to add.


The only thing that is correct about Wydick's argument is that the question of what to do in Iraq depends only on the consequences of withdrawing versus the consequences of staying. But just as there are those who appeal to honoring the memory of those who have sacrificed, there are those who make the exact opposite mistake, arguing that if it was a bad to go in, it therefore must be good to get out.

The USA Today article says: if we were to withdraw today, we would lose X. X will only grow.

That is wrong.

As of today, we have lost X. We can't get X back, and the longer we stay, the bigger X gets.

If we withdraw today, however, X is not all we stand to lose. Precipitous withdrawal has its own consequences. I happen to believe that those consequences would be very severe, which is why I am against immediate withdrawal.


"Failing to correctly identify sunk costs (those that are irretrievable), and deal with them properly, biases our decision-making ..."

Is it really that, or is it a much simpler case of being unwilling to admit we were wrong -- and don't know how to make it right?


What is the test used to determine if the costs are indeed sunk? President Bush is unlikely to admit that the costs are sunk in fact his position is that the costs can produce benefits if we stay-the-course.

mandala oblongata

I agree with those who have said that ignoring the sunk costs (as we should) strengthens the case for staying in Iraq in some form.

Before we went in, I was strongly opposed to the invasion because I did not think the benefits were worth the likely "costs" (detracting focus away from Al Qaeda, as well as the military spending, casualties, risks of civil war, etc). Unfortunately, the costs have been even greater than expected and the benefits fewer.

However, even though it was clearly a mistake to have invaded, now that we are there, we have to decide what the best investment of our resources is going forward. If we think out efforts will likely lead to better outcomes, then we should continue to "invest." If we don't, we should start to "divest."

Of course, if your money manager constantly set your cash on fire in front of you and kept saying everything was peaches and cream, you'd probably make a switch. Too bad we can't do that.



By ignoring sunk costs, the government is more heavily encouraged to stay in Iraq. By considering all the costs (staying in Iraq as well as leaving), the value of a stable Iraq now is relatively cheap ignoring sunk costs.

This is the same situation when a city begins construction on a stadium and unforeseen costs arise. Cities don't stop construction unless the new adjusted cost exceeds the expected return. Maybe that is the case for Iraq, but the expected return from a stable Iraq was high enough in the past that we spent quite a bit. It was thought to be costly but there would be a valuable return. Has the expected return decreased that much that we should no longer pay for it?

Even ignoring sunk costs, pro and anti Iraq war arguments can still be legitimate. A disagreement in foreseen costs and foreseen returns would create a different conclusion.


Phil nails the sunk cost arguments--both the left and right make arguments that ignore the sunk costs. But let's move beyond sunk costs and think about a little game theory...

If your goal is to keep the US out of all, or at least most, wars (say because you think war is often not worth the costs), then you might advocate withdrawal from Iraq even if the marginal cost of this war is lower than the marginal benefit. Making a mess of things would presumably lower US prestige and military presence in the world, and would make us less likely to enter into similar situations in the future.

Conversely, if your goal is to increase US power and influence in the world, you might advocate staying in Iraq even if the marginal cost exceeds the marginal benefit. Winning this war will make other countries less likely to ally with enemies of the US government, will bolster the morale of US forces, will increase the willingness of hawkish politicians to wage war, etc.

I call this game-theoretic reasoning because basically what I'm doing is creating a large decision tree. The sunk costs are just as sunk as before, but I think it's worth noting that the future consequences of winning or losing this war are much bigger than just what happens in Iraq.

In comparison to these consequences, the sunk costs of this war are so small that it really matters little whether they are included in the analysis or not. Arguments about "honoring the sacrifice" of dead soldiers or getting out because "Bush LIED" are, on the surface, arguments about sunk costs, but I think they are more likely useful anchors for more complicated views which have little to do with sunk costs per se.


Robert Plamondon

We can probably estimate the continuing costs of an Iraq involvement by attempting to sell it to someone. How much would be have to pay the French to tempt them to take Iraq off our hands?

The contract would be the tricky part -- we'd have to state our goals (if any) clearly and truthfully. Otherwise the contractor might just look around, determine that there were no weapons of mass destruction in Iraq, and leave.


None of this analysis matters because the decision maker -- i.e. George W. Bush a.k.a. "The Decider" -- has NOTHING TO LOSE by staying in Iraq. He's at rock bottom and the only way he can salvage his legacy is by pulling a rabbit out of his Iraq hat.

Besides, Halliburton et. al. are still making a (financial) killing over there.

Rita: Lovely Meter Maid

I think I referred to the old "lightning doesn't strike twice" thing, as regards HSM2 when I noted that people often do not want the SOS (same 'ole sh*t). This may or may not be justified by the quality of the sequel. People are not logical. I am *overtly* often without logic, but more worrying, still, than me, are the many, seemingly sane souls prancing about with globs and heaps of *covert* illogic.


we won't leave Iraq until we establish a puppet regime to control the largest untapped oil reserve on the planet, or until popular dissent forces the politicians to pull out- the disconnect of 'sunk costs' is that the citizens lose money, while the oil companies and war profiteers make money