Joe Stiglitz on how to correctly budget for a war
In 2008, Nobel prize-winning economist Joe Stiglitz, alongside economist Linda Bilmes, published the ‘The Three Trillion Dollar War: The True Cost of the Iraq Conflict.’ The book — which Stiglitz summarized in a 2008 speech (pdf) — argued that the way we account for war tends to vastly underestimate the actual cost of conflict because it misses everything but the direct, short-term expenditures made on troops and munitions. As we begin our intervention in Libya — an intervention for which we’ve seen little-to-nothing in the way of cost estimates — it seemed a good time to return to Stiglitz’s original arguments about how to budget for war, and why it’s important to do so. A lightly edited transcript of our discussion follows.
Ezra Klein: Let’s start with the macroeconomics. What does war do to the broader economy?
Joe Stiglitz: The macroeconomic effects were huge in the Iraq war. Libya is more complicated because, one way or another, there was going to be an eruption. For Iraq, our book assumed the war added $5 to $10 to the price of oil. The reason we did that was we thought no one would object to numbers that were so conservative. The price had gone from $23 per gallon to $140. Lots of things are always going on, but it was not coincidental that the price started going up very strongly after the war. And corroborating our view was that prior to the war, the futures markets had expected the price to remain around $23 a gallon.
One of the other things we argued in the book — remember, this was before the financial crisis — was that one of the reasons for our lax monetary policy, which we thought was distorting the economy and contributing to a bubble, was that purchasing power was being taken out of the economy to buy oil and the Federal Reserve had to compensate for that and it did it through looser monetary policy.
And the other major costs that aren’t included in our initial accounting?
The big points we raised in Iraq and Afghanistan — and have updated in congressional testimony (pdf) — are the disabilities, the health-care costs that follow the war, and the equipment you have to replace after the war.
I’ve seen you make the point that, as an economist who won a Nobel prize for studying the the role of information in decision-making, skipping this honest accounting process deprives us of crucial information we need to make good decisions about matters of war. Could you expand on that a bit?
This is particularly the case for wars of choice, which most recent wars have been. When we were attacked in Pearl Harbor, we weren’t going to go through a cost-benefit analysis of responding. The Iraq War, however, was a war of choice. And if you’re going to approach that with any degree of rationality, you have to think about what it’s going to cost you. To do that, you want an inclusive cost estimate, particularly because you know the government is going to try and narrow the cost down. They’re going to talk about only the cost of the bombs, whereas we emphasized that there are a whole other set of costs — taking care of the disabled, replacing equipment, etc. — that any good accounting system would take into account.
One objection people often raise to this line of thinking is that the president needs to move fast when making decisions about whether to go to war, and if we normalized a budgeting process, he’d lose that crucial flexibility.
Two points. One, the Iraq War was a war of choice and so we didn’t have to rush in. We weren’t being bombed. But the second point is that even if you don’t go through this analysis before you embark, as you’re fighting the war, these issues arise over and over again. During the war, we made a lot of penny-wise and pound-foolish decisions. To not spend on body armor at the beginning clearly increased our spending in the long run. Same for mine-resistant vehicles, which are as much as a million dollars apiece. They’re expensive, but compared to not buying them, they’re cheap. A large fraction of the disabilities would’ve been avoided if they’d had those rather than Humvees. If you’d had the accounting based on the full cost of health care, you would’ve known they were a bargain.
The other objection is that you can’t put a price on a human life. Iraq was advocated on at least partially humanitarian grounds, and the intervention in Libya came in the context of a more acute humanitarian crisis. So can you really try to make these projections when so many lives hang in the balance?
I’m sympathetic to the view that the economics are not everything. We didn’t try to value the benefit of winning the war, for instance. But whatever our judgments about the benefits, we have to be realistic about the costs. I don’t think any of us thought tallying the costs would provide the answer to whether we should do it. But not to have those costs is a vital mistake. In the case of valuing lives, that was one of the things we really sought to emphasize: If you started valuing lives of just our troops, you would realize the cost of the Iraq war was enormous. We don’t value the lives of our troops in the way we value lives in the Environmental Protection Agency or the Department of Transportation. If we started using those numbers, we’d get a very different total cost — and even more so if we began valuing Iraqi lives.
You’ve written that some of the rules you came across while researching your book shocked you. In particular, you brought up the odd way we pay for contractor insurance. Can you go into a bit more detail on that?
We require contractors to have insurance. But no private guy will pay for the insurance, so we pay for the insurance ourselves. Interestingly, the bad actor in all this was AIG. The government went and contracted with AIG, paying them the premiums, and then after the contractors got injured, AIG claimed that the injuries were excluded under their policy. So they’d exclude, for instance, death from combat but take a very broad definition of combat, where everyone in the combat zone was in combat. So they’d be getting these premiums and there’d be almost no one who they’d admit qualified for the benefit. So we were paying the premiums to AIG and, in most cases, they were denying the claims.