March 26, 2003

silhouette3.JPG From the desk of Jane Galt:

How much is the war going to cost?

I've seen a number of claims like this one from Eric Alterman:

The first $75 billion is just a downpayment. Expect to pay hundreds of billions in the short-term, trillions in the long run. Expect it to come out of your schools, your police forces, your highways, your future and your children�s future

Anyone who's sat through a budget meeting knows that almost everyone overestimates their successess, underestimates their costs; it's easier to go back for money later, when you can wave a nice hunk of sunk costs around, than say up front that you think whatever it is you're proposing will be expensive as hell.

But trillions? US GDP is roughly $10 trillion. Alterman is saying that over the long run, this war is going to cost us at least 20% of GDP. That's nuts, and it's not the first time I've seen those sorts of numbers around.

Reality check: the entire US military budget is in the range of $350b. Saying that this war will cost trillions in any term short enough for us to care about (I mean, he's probably right, if we use a timescale of several hundred years, but that's not very useful), is saying that this war is going to cost nearly as much as the entire military budget, year in and year out, for decades. For reference, the next six months are estimated to cost $60b on military spending. (I'm excluding the humanitarian and domestic segment of the budget submitted by the President.) Even with a fudge factor of 50%, that's $90b over the next six months, $180b a year. At that rate, assuming you do absolutely no discounting at all, it would take us over 10 years to get to $2t, thus meeting the "trillions" criteria. Which is madness. By that logic, we were spending as much on WWII in 1953 as we were in 1943. If you don't know, military spending during WWII was over 50% of GDP; it was in the 10% range during Korea, and dropped sharply thereafter. This while we were still occupying Japan, still garrisoning Germany, had a mandatory draft, and were building up for the Cold War. Even if you attribute the entire cost of the Cold War to WWII, and none of it to Stalinist imperialism, you still don't get the kind of numbers required to make the occupation cost as much as the battle. The difference is even more stark now, for you must remember that we have an all-volunteer army, which gets paid whether or not they're in Iraq. The extra, non-labor cost of the war is heavy on things like ordnance, which we won't be expending once we control the country.

Where do they get these numbers? With gems like this from James Galbraith, son of the amiably paranoiac pop-economist John Kenneth Galbraith.

RECENTLY AS we debated the war now underway in Iraq, seven Nobel laureates joined 150 other US economists (including myself) to call for careful consideration of the costs of war in Iraq. When economists talk about costs, what do we mean? First, we mean budget costs -- for gasoline, equipment, and explosives -- that begin at about $100 billion1. This figure is based on an assumption that the war goes well. It the assumption is wrong, the numbers will go up fast. The history of warfare -- from Europe in 1914 to Vietnam in the 1960s -- is littered with gross underestimates of budget costs.

We also mean the material costs, which are sometimes overstated in war -- bombs may fall on empty fields or on rubble and damage can look worse than it is. In Iraq, though, the civilian population is already stressed. Even modest material damage -- to the water, to the electric grids and the health system -- could bring on humanitarian disaster. There are risks of sabotage, not least to the oil fields. And there will be some damage, inevitably, to the archeological heritage of Iraq and especially Baghdad.2

The human costs are beyond accounting. No matter the number of casualities, every dead soldier, on either side, every dead civilian, is a human being who could have lived a productive and perhaps happy life. Every injured person will carry a burden of pain. We need not demean the grief ahead by trying to give it money value.3

The uncertainty costs are more prosaic but just as hard to calculate. How much business investment, how much production, how much trade have we already lost -- not only in the United States but in the world economy -- because of the fear and uncertainty surrounding this war?4 What effect will war have on global economic decision-making, consumer and market confidence, global energy prices? How much more lies ahead?

The reconstruction costs are imponderable. One estimate for the cost of rebuilding Iraq runs to $2 trillion.5 But will the US actually do the job? What if it takes two years and a 100,000 troops? Five years and 200,000 troops? What if the oil fields are shut down in the meantime?

The follow-on costs arise from the military situation we may face after the war ends. Will peace and democracy break out in Iraq? Will the war lead to peace, democracy, and demilitarization across the Middle East, as some believe? Or will there be rebellions, revenge killings, and proxy wars across Iraq, Syria, Iran, Saudi Arabia, and even Egypt? Not to mention in Israel and Palestine.

The diplomatic costs lie in the damage already done, and more may lie ahead -- to relationships with Europe, Russia and other countries. One may count also the cost of disillusion, of much of the world's population, with the American ideal.

The opportunity costs are those that arise every time we make a decision to do one thing rather than another. By choosing to go to war, we are choosing to do less to solve our problems at home. We face a crisis in every state and local budget in this country -- in every school, every welfare program, and every part of public health care. We face a crisis of trust in our corporations, and a crisis of confidence in the profitability of future business investment. American households are facing in slow motion a crisis of household debts. Little will be done about any of this, so long as we are preoccupied with war.

Finally, the apocalyptic costs should be considered. There is the risk, already unfolding, that North Korea will produce atomic bombs. There is also the risk that Iran will buy a few of them or make some of its own. There is the risk that we will shortly face one, two, or perhaps more nuclear powers who regard us -- and not entirely without reason -- as a mortal threat to their existence. There is the risk that we may make a catastrophic mistake in our response.

Once the real costs have been considered, the economic conclusion is not controversial. It is that collective security -- the kind provided by strong alliances, the rule of law, and the United Nations Security Council -- is the only real security. It is certainly the only kind that we, or any other country, can afford. Perhaps war is sometimes necessary. But it is never really cheap enough.

He offers vague possibilities, making no attempt to quantify them, much less calculate their probability, even though the alleged subject of the article is the economic cost of the war.

He conflates all sorts of costs into one big amorphous bundle. He only looks at costs on one side; for example, discussing the cost in lives of the war, without discussing the cost in lives of Saddaam's regime and the sanctions that are the likely alternative to the war. If we kill 300 Iraqi civilians and 300 American troops ousting Saddaam, and Saddaam's secret police are murdering 1,000 people a year, and 5,000 people a year are dying from the humanitarian crisis brought on by sanctions, it is not a net "cost" in human lives.

Likewise, he examines only the negative consequences the current uncertainty might have on the economy, without mentioning that, for example, a successful war might boost the consumer confidence dampened by fears of terrorism, or that lowered security risk in the Middle East might result in both lower oil prices, and higher investment in highly oil-dependent industries.

He offers unsourced references for large numbers -- "One estimate for the cost of rebuilding Iraq runs to $2 trillion" -- in order to give his claims a false patina of precision.

He cites any number of highly speculative, unquantitative "costs" in terms of US prestige and other such intangibles that have nothing to do with economic costs. He posits "opportunity costs" of not doing things that many of us don't want to spend federal money on in the first place. An opportunity cost is a precise economic term: it means the next-best alternative use for your money. You can't claim that our failure to institute national health care is an opportunity cost of the war when such a thing would cost far more than the money being spent on the war, and when it's something that we probably wouldn't be doing anyway.

Finally, he blathers about a possible nuclear holocaust as a result of this invasion, a subject on which he -- and I -- are thoroughly unqualified to comment. I will nonetheless point out that a scenario in which North Korea nukes us or one of their neighbors in order to revenge Saddaam seems slightly less likely than my winning Fear Factor Miss USA. Then he announces that he has proven -- through the miracle of economic science! -- that the only solution we can afford is, surprise, the one dictated by Galbraith's brand of warmed-over 50's liberal internationalism.

Thus, Eric Alterman is enabled to claim that the cost to the US taxpayer will be over $2t, even though most of the larger costs cited by Galbraith aren't going to be borne by Americans either directly or indirectly, but by Iraqi oil.6 That's the oil that will be able to flow freely for the first time in ten years because of this war -- and the revenue from which will flow to the Iraqi people for the first time in a decade.

The war will certainly cost more than the $60b and change that the President is asking for. But it is not going to run us several trillion dollars (though even if it did, that would work out to less than 0.1% of GDP over the next 20 years.) I don't know how much more, and neither does anyone else, although I'm sure the military has better guesses than I could make. It's important to think about the economic cost of the war -- the pro-war side has mostly dropped the ball on this, and it's an important calculation when we consider whether or not to go. But making up ridiculous numbers in order to support your predisposition isn't helpful -- and when the war doesn't cost us $2t, people are going to remember that the next time you talk about the costs of a program you don't like.

1 Note the number is an inflation of my already inflated number.

2 Note that he does not attempt to quantify any of these costs, nor compare them to the current humanitarian disaster from sanctions.

3 "We need not demean the grief ahead by trying to give it money value" is a funny statement for an economist, since quantifying the value of a human life is routine calculations. It's especially funny in an article titled "What Economic Price This War?".

4 Almost none. Investment is deferred consumption. It can, more or less, be invested in six months just as well: your banker doesn't say, "Well, since you're not investing today, I think I'll just take this hunk of cash out back and burn it." Trade doesn't seem to be suffering. Not all production can be made up, but that amount seems likely to be small.

5 This is a scurvy writing trick: "Someone, somewhere has made a Wild Assed Guess that is Extremely High, but I'm not going to tell you who, because then you'd realize that they're a total idiot and this number is meaningless."

6 Am I suggesting that the Iraqis should pay for occupation expenses? Nope. We can afford it, and there's something repellent about making impoverished Iraqis pay for a war foisted on them by an evil dictator. But most of that $2t, if it is any sort of a real number, will be stuff for Iraqis: roads, schools, hospitals, government buildings, power plants and sewers and all the good stuff that lets us live like citizens of the 21st century. That stuff should come out of Iraqi oil revenues.

Posted by Jane Galt at March 26, 2003 03:23 PM | TrackBack | Technorati inbound links

Please, close the small tags! (I think you left off the slash on the close-small tag, giving double-small.)

Posted by: Devilbunny on March 26, 2003 03:38 PM


The $2T figure (it's actually $1.9T) that Gailbraith cites is from Yale economist William Nordhaus who seems to specialize in the economics of BIG issues and BIG numbers. Nordhaus's report is *here.* Instead of railing on Gailbraith who is nothing really more than a messenger (one of thousands in fact, since Nordhaus has been all over the news since January), why don't we discuss the source?

As an aside, If you remember, Nordhaus *once echoed two of your favorite economists Kevin Murphy and Robert Topel on healthcare by writing, "the value of increasing longevity over the past century could be as large as the value of growth in all other good and services over the same period." (for those of you who don't know, Kevin Murphy and Richard Topel are Chicago economists who wrote that "every 20% reduction in the occurance of heart disease and cancer was worth $10T or the entire US economy."

Since Nordhaus, Murphy and Topel have been in alignment on other issues in the past (and you apparently agreed with their findings then) -- I wonder, what would Murphy and Topel say about the cost of war in Iraq (and would you agree with it if it came from a Chicago economist instead of a Yalie)?

Posted by: Matt Johnson on March 26, 2003 03:51 PM

If it's the Nordhaus number, why doesn't he source it?

But if it is, he's citing it wrong: the Nordhaus number isn't the cost of the war to the American taxpayer, as cited by Alterman, nor the "cost of rebuilding Iraq", as cited by Galbraith. Nordhaus places the occupation at a lower bound of $75b and an upper bound of $500b, roughly 1/4 of what Galbraith claims. If we add in reconstruction, at $30b to $75b, and humanitarian aid, at $1b to $10b, we're still only at an upper boundry of 1/3 of Galbraith's figure.

Posted by: Jane Galt on March 26, 2003 04:11 PM

The question I have is what does the $60 billion mean? What is being spent on this war that would not otherwise have been spent on the military in some way? The Bush Administration already made a $380 billion budget request for the Pentagon; is the $60 billion over and above that, part of that or some combination of the two?

The bottom line is how separable the cost of Operation Iraqi Freedom from military spending in general? Can the accounting be so clear cut?

Posted by: Damien Smith on March 26, 2003 04:22 PM

For an economist to talk about dollar cost and opportunity cost as if they are additive is wrong. If you spend $100 billion, you can't say "not only did it cost $100 billion, but we could have spent $100 billion on health care, so the total cost is $200 billion."

The liberal economists who are trying to make an "economic case" against the war make me sick. Its as disgusting as if the pro-war side were touting the benefits of giving American companies business opportunities in Iraq.

Posted by: Arnold Kling on March 26, 2003 04:32 PM

OK. So what might be a fair estimate of the cost of the war?

Obviously making one is hard because it relies on assumptions about the post-war occupation. No one can be sure what that will look like or how long it will last, or how much of it Iraq's oil will pay for -- and different assumptions will lead to widely different cost estimates.

So, might as well do a SWAG.

The first Gulf War had an incremental cost of about $61 billion in 1992 dollars, and only lasted a month and a half. That war included no occupation.

So, SWAG #1: $61B in 1992 is about $76B today, inflation adjusted (I have not sourced this number, I'm guesstimating).

This cost for Gulf War I doesn't match what we're hearing about expected costs this time around.

Megan writes: "For reference, the next six months are estimated to cost $60b on military spending."

I'm assuming that that number comes from the administration, but it is a bit surprising that the 1.5 month Gulf War cost us ~$76B while less than this is now supposed to last us for six months, despite a more ambitious enterprise this time around.

So let's split the difference. Instead of $76B lasting us for 1.5 months in GWI or $60B for 6 months in the new projection Megan mentions, let's say it's 3.75 months and $68B.

In other words, if the war were to last an entire year, it would cost about $172B.

But we're assuming a less than four month war.

So, SWAG #2: The war costs $68B.

Then you've got the occupation.

SWAG #3: The occupation lasts 5 years.

We were in Japan for 7 years before real power-sharing (1945-1952), though in fact we stayed for decades and are, of course, still there.

SWAG #4: Each year of occupation will cost $30 billion.

That is the midpoint of Congressional Budget Office estimates of $12 - $48B in annual costs.

"Further, the incremental cost of an occupation following combat operations could vary from about $1 billion to $4 billion a month."

Assuming that the annualized cost of a war would be $172B as we said before, then occupation would cost less than 20% of this.

What do you get in total? About $218 billion over 5 years for war and occupation. $68B for war and $150B for occupation -- approximately a 12% annual increase in our military budget.

I have no idea what reconstruction / humanitarian aid might add to this figure. But someone could certainly do an analysis of that with the same (lack of ) thoroughness that I just did for war and occupation.

This is a complete wild ass guess, but hey, you need to hang your hat on something.

Posted by: Jim on March 26, 2003 04:51 PM

Jane, which report are you reading? The one I have reads, "However,if the United States has a string of bad luck or misjudgments during
or after the war,the outcome could reach the $1.9 trillion of the high case." (page 77)

And look! Here's Nordhaus in the press, touting the nubmer:

Here he is again:¬Found=true

and again:

oh my! not again:

not another one:

Shall I continue?

Posted by: Matt Johnson on March 26, 2003 04:55 PM

why are many of you assuming that occupying iraq for five years will be similar to the occupation of japan and germany? both of these countries had some history of democratic processes and were pretty homogenous. with the explosive mix of shia majority and sunni minority, do we actually believe that the occupation will be as "smooth" (relatively speaking), as the experiences of germany and japan. please, if there is any real trouble in maintaining the post war peace in iraq, up your figures for the cost of iraqi occupation.

Posted by: cas on March 26, 2003 05:10 PM

The high boundary of the Nordhaus report included things like costs to Iraqi civilians, which are not part of the cost to the US taxpayer, which is what Alterman was talking about. Nor was it part of the cost of rebuilding Iraq, which is what Galbraith cited. The upper boundary of invasion, occupation, infrastructure rebuilding, and humanitarian aid is under $700b -- and those are worst-case, not likely scenario numbers, whereas Alterman proclaims that the war "will" cost the taxpayer trillions more, which is extremely unlikely.

But even at that, Nordhaus does the same thing: although he talks a great deal about the suffering of the Iraqis, he estimates the costs as if the alternative were for Saddaam to suddenly have a change of heart, wave his magic wand, and turn the nation into a representative democracy with robust civil and economic institutions on the road to prosperity. He too cites the costs to civilians without subtracting the cost to them, in lives and money, of Saddaam's regime. And he factors in the spectre of an oil shock like those of the 1970's, which I consider basically impossible, because oil production is more diversified, and because the population boom in the oil-producing nations and subsequent declines in per-capita GDP would make such a move impossible to carry out, even if Venezuela and Mexico would be willing to join, which I'm pretty much 100% sure they wouldn't given their current economic and geopolitical situations.

Jim's numbers are as good as anything I'm likely to come up with -- and they track somewhere near Lindsay's upper number of $200b. But at some point, these things become very hard to sort out. Is the Cold War a cost of WWII, or of Soviet imperialism, or something else? Is our occupation of Japan a lingering cost of WWII, or a strategic benefit, allowing us to project the potential force in the area necessary for the world's only superpower? A lot of the decision on how to allocate costs for this war will be a judgement call between items made necessary by the war, and things independantly desireable made possible by it. Obviously, those who support the war will generally call those expenses as unrelated to the war, while those who oppose it will attribute most of those expenses to it. Those are legitimate differences. But the numbers cited by those two aren't within the boundaries of legitimate difference.

Posted by: Jane Galt on March 26, 2003 05:35 PM

Jane, if Nordhaus isn't saying the cost to taxpayers is $2T (as you suggest), then wouldn't you agree the larger news story here is that 5 of the top news organizations in the world are running articles saying otherwise?

As an economic blog, I'd think you'd be way more interested in blowing the lid off of that one than merely criticizing Alterman and Galbriath (who have already clearly established an anti-war bias).

I stand by my assertion that they're using Nordhaus's number.

Posted by: Matt Johnson on March 26, 2003 05:52 PM

Matt, you can read the report yourself. It assumes that all costs of reconstructing Iraq are borne by the United States, rather than Iraqi oil revenues. And Nordhaus himself says he made no attempt to estimate any beneficial impact of the war.

More importantly, over 50% of the $1.9t worst case scenario is impact on oil markets and macroeconomic impact on the US economy. These do not qualify either as government expenditures, which was the subject of Alterman's post, or the cost of rebuilding Iraq, which is how Galbraith presented it. Nor is the worst case scenario even close to likely enough for Alterman to be declaring that the war "will" cost trillions, even if all of Nordhaus's numbers were applicable, which they aren't. In short, if that is the source of their numbers, then they are guilty of gross incompetence.

Posted by: Jane Galt on March 26, 2003 06:04 PM

Matt, it would be nice if they would at least assert just once where they got the number. The problem isn't that no one can find where they are getting their numbers from, but that we have to guess where they are coming from.

And it isn't exactly earth shattereing news that the mainstream press fucked up when reporting on economic issues.

Posted by: Joe on March 26, 2003 06:07 PM

Has anyone allowed for the fact that the cost of occupying Iraq must be offset by the way America has finally stopped occupying Germany?

All the costs of having a division or two sitting in Iraq may not greatly exceed the costs of having them sit in Germany, where they would still be if not for this war.

Posted by: Patrick on March 26, 2003 06:27 PM

I have a feeling occupying Iraq will not be like occupying Germany is today.

Today's German occupation hardly deserves the word. Though we do have bases there, the last time our forces fought German forces was fifty years ago. In Iraq, we will be fresh off a military conquest. Qualitatively, I'd think that means we'll face a different sort of project there than we do in Germany today (even if it resembles the initial occupation of Germany in the 1940s -- a point which is itself arguable).

Posted by: Jim on March 26, 2003 06:41 PM


I can't speak for Alterman but Galbraith was using Nordhaus's number. How do I know? I asked him. Jane's seen the email and she knows it too. Not only does Galbraith say it isn't his number, he says that he included the example only to illustrate that certain people (without singling out Norhdaus) have expanded the cost of war tremendously; nor does he endorse the number.

In what started out as a criticism of Galbraith and Alterman, Jane has clearly picked up on a distinction that no one else is reporting -- that Alterman clearly doesn't get either: that Nordhaus is arguing that the true cost of war isn't just man, material and effort.

The economist writes, "Mr Nordhaus estimates that the total cost of a war to America could range between $100 billion and $1.9 trillion, spread over a ten-year period. That could be as much as 2% of American GDP for every year of the decade."

I am left to conclude that Alterman has taken a widely published number such as Nordhaus's $1.9 Trillion and twisted into something to further his own interests -- not to mention the number of mainstream press who have done exactly the same thing!

And this isn't a story?

Posted by: Matt Johnson on March 26, 2003 08:42 PM

Only a trillion�? I'm going to estimate at least a quadrillion bucks, and you can quote me on that. If you give me a percentage, that is.

Posted by: David Perron on March 26, 2003 11:31 PM

"More importantly, over 50% of the $1.9t worst case scenario is impact on oil markets and macroeconomic impact on the US economy. These do not qualify either as government expenditures, which was the subject of Alterman's post, or the cost of rebuilding Iraq, which is how Galbraith presented it."

Government expenditure or not, it sure looks like a cost to me. Maybe the alternatives are worse, but still.

Posted by: Jason McCullough on March 27, 2003 01:46 AM

From past observation of Alterman, even if the cost is 1/100th of what he claims, and offset by sales of Iraqi oil, he'll just claim that the real costs are being hidden by the tame conserative media and phantom right wing goon squads, after which he will go into another multi-month whine about the "selected" President that will only be interrupted by the time-honored operation of the 22nd Amendment. At that point, he might have to get a real job.

Posted by: M. Scott Eiland on March 27, 2003 04:32 AM

First of all, most of those costs come from things we now know won't happen, since we've secured most of the oil fields and ports: it was predicated on total destruction of the Iraqi oil fields and an OPEC retaliation, the latter of which any economist should have recognized is most unlikely given the current realities of that organization, and the excess Russian capacity. And second of all, Alterman was discussing government expenditures. Expropriation of one number to use as another, related number, is how half of the awful statistics we see get created. If you wait a couple weeks, I suspect you'll be able to find someone who has taken Alterman's estimate that the war is going to cost the taxpayer "trillions", added in the macroeconomic effect and an oil shock, and calculated that we can expect the war to completely level the US economy.

Posted by: Jane Galt on March 27, 2003 07:39 AM

Jane: Still got an open <small> tag... it's the one at the end of the 4th footnote. Cheers!

Posted by: Old Grouch on March 27, 2003 09:03 PM


Posted by: Kimberly on September 9, 2003 11:13 AM


Posted by: Kimberly on September 9, 2003 11:13 AM

Comments are Closed.