September 10, 2006
Ex post versus ex ante benefit-cost analysis: Iraq 2003-
Now that the long-delayed Senate report on pre-Iraq War intelligence has finally put the Republican imprimatur on the well-established fact that the case for Iraqi WMD's had been hyped (as well as the absence of a Iraq-al Qaeda link), we can return our attention to a rational benefit-cost assessment for the invasion and subsequent occupation, ex ante as well as (quasi) ex post.
A well known study by Davis, Murphy and Topel (2006) used what I thought was the correct framework to assess the case for the invasion -- namely assessing the relative benefits and costs in terms of foregone alternatives. They state:
"According to our analysis, pre-invasion views about the likely course of the Iraq intervention imply present value costs for the United States in the range of $100 to $870 billion. Our estimated present value cost for the containment policy is nearly $300 billion and ranges upward to $700 billion when we account for several risks stressed by national security analysts. Our analysis also indicates that war and forcible regime change will yield large improvements in the economic well-being of most Iraqis relative to their prospects under the containment policy, and that the Iraqi death toll would likely be greater under containment."
As I mentioned in my previous posts on expenditures on Iraqi theater of operations and opportunity costs, the direct "burn rate" in billions of USD per month (bdpm) is 8, yielding a per year expenditure of around $96 billion. My calculations indicate, using a 2% discount rate as used in Davis et al. (2006), that if expenditures continue at the same nominal rate to end fiscal year 2010, and immediately drop to zero, then ex post the benefit-cost ratio of the invasion (even assuming Davis et al.'s worst-case scenario under continued containment) is less than unity. This calculation is biased in favor of the pro-invasion case in that I have not imputed dollar costs associated with the American military casualties that have occurred to date. Davis et al. rely upon a rate of $6.9 billion per 1000 fatalities, and $9 billion per 7153 casualties. As of end-August, US fatalities are 2644 and wounded at 19683 (see Figure 1 below).
Figure 1: Cumulative fatalities and wounded in Iraq, up to end-August; and linear projections based on OLS regressions (no constant). Source: Iraq Coalition Casualities, accessed 8 September 2006, and author's calculations.
Figure 1 also shows the trajectories based on a linear projection, ending at end-FY2010.
Obviously, this is neither a fully ex ante, nor a fully ex post analysis, so the jury is still out. But for those who believed that stabilizing and rebuilding Iraq would not be a quick venture (but rather somewhere in between Davis et al.'s scenarios 5 and 7), the ex ante benefit-cost ratio always looked less than one.
What about an ex ante benefit-cost assessment at time t=now? Some of the same uncertainties abound, even if some are resolved. And what about the expenditures already undertaken to date? Economic theory indicates that under conventional utility functions, and static analysis, sunk costs are irrelevant. Dynamic analysis might -- or might not -- alter the answer.
Late Addition 9/11/06 10AM Pacific
"The chief of intelligence for the Marine Corps in Iraq recently filed an unusual secret report concluding that the prospects for securing that country's western Anbar province are dim and that there is almost nothing the U.S. military can do to improve the political and social situation there, said several military officers and intelligence officials familiar with its contents.
The officials described Col. Pete Devlin's classified assessment of the dire state of Anbar as the first time that a senior U.S. military officer has filed so negative a report from Iraq.
One Army officer summarized it as arguing that in Anbar province, "We haven't been defeated militarily but we have been defeated politically -- and that's where wars are won and lost."
The "very pessimistic" statement, as one Marine officer called it, was dated Aug. 16 and sent to Washington shortly after that, and has been discussed across the Pentagon and elsewhere in national security circles. "I don't know if it is a shock wave, but it's made people uncomfortable," said a Defense Department official who has read the report. Like others interviewed about the report, he spoke on the condition that he not be identified by name because of the document's sensitivity."
Posted by Menzie Chinn at September 10, 2006 09:49 AMdigg this | reddit
I started college as an economics major and switched to history when econ turned me off.
Now I remember why.
Posted by: anonymous at September 10, 2006 02:48 PM
I thought their analysis was a good example of why amateurs in DoD accounting should just "say no" and stick to what they understand. The authors should have been embarrassed. For example, they relied upon CBO numbers, but it is very clear to me that they did not understand what those numbers meant. Worse yet, they did not know that DoD and the US Army actually do have various economic models for estimating deployment costs, sustainment costs and peacekeeping mission costs. Those models distinguish between variable costs and fixed costs. They assumed a $226,000 per soldier cost based on a misunderstanding of CBO's numbers. In fact, the numbers used in Army's FORCES and COST models are significantly lower. But the biggest goof in their analysis didn't even require the use of restricted For Official Use Only (FOUO) numbers. They erroneously assumed that 100% of the forces in the Persian Gulf theater were there in support of an Iraqi containment policy and, in the absence of a containment policy those costs would be zero. This is laughable on the face of it. Even if Saddam's regime had miraculously transformed into Sweden, I think it's a safe bet that virtually all of the "containment" forces in the Persian Gulf still would have been there at substantially (and perhaps greater) numbers. It's preposterous to think that there would have been no Persian Gulf military presence in the face of an unconstrained Iran, which is what we would have faced if Saddam had decided to reform himself and became our best friend. In addition, the Kuwait theater was serving as a training ground. The Army still would have incurred those training costs...just in a different location. Finally, the authors suggest (without much evidence) that there were no studies within DoD anticipating that the costs of the war would be much higher than commonly believed. Again, this just isn't true. What is true is that many of those analyses were not allowed to see the light of day, so perhaps we can forgive the authors for not knowing better. The ex ante costs of the war were very much higher than the Administration pretended.
Posted by: 2slugbaits at September 10, 2006 04:34 PM
anonymous: Numbers are everywhere, even in history. By the way, I started college as a history major and switched to economics...
2slugbaits: Thanks for the excellent observations. In particular, I agree it is now clear that many of the uniformed military had a more realistic appraisal of what would be the likely outcome of the invasion conducted with the force levels advocated by the policy level officials in the Administration.
Posted by: menzie chinn at September 10, 2006 09:25 PM
I think some issues are missing from the cost - benefit analyisis. For the baseline, I think we can assume that there is a X% probability (X siginificantly more than 0), that in the next decade the US would have had to intervene in Iraq anyway, but under much worse circumstances. Saddam is not immortal and he had a lot of enemies. If he had beed assasinated or toppled in a coup, the sectarian tensions might have been much worse, and possibly we would have an intervention by Turkey or/and Iran due to kurds/shias. Given the importance of the Middle East it would have been impossible not to intervene. Considering the expected present value of the costs of such a mess would raise the costs of containment.
For the benefits, there is an Y% chance (Y again siginifcantly more than 0) that Iraq can develop into a prosperous pro-western democracy. Do not laugh, take a look at Korea: the Korean war was a bloody stalemate, and South Korea was a nasty dictatorship at that time. Yet today South Korea is a respected prosperous democracy. The world in general and Koreans in particular are far better off than having the North Korean regime in entire Korea. Such an outcome would have huge benefits and its expected present value should be considered too.
Posted by: Laszlo Varro at September 11, 2006 05:40 AM
For the benefits, there is an Y% chance (Y again siginifcantly more than 0) that Iraq can develop into a prosperous pro-western democracy. Do not laugh, take a look at Korea: the Korean war was a bloody stalemate, and South Korea was a nasty dictatorship at that time. Yet today South Korea is a respected prosperous democracy.
I don't know whether to laugh or cry. The difference could not be more stark - Korea is populated with one ethnicity - Koreans, with one religion. There was none of the ethnic and sectarian strife that is tearing Iraq apart. There's simply no comparison.
Posted by: Alex at September 12, 2006 06:43 AM
An absolutely perfect post. Here is a more diffiuclt (possibly intractible) challenge for you: what would the benefits (reduction in terrorist threat) have been of following a different pollicy. Rather than waging war in Iraq, using those resources in Afganistan (or bribe the Pakistanis) to capture the A-Q leadership and put them on trial at The Hague; use USAID to promote democratic and economic development in failed (failing) states, demonstrating leadership to get a Doha development round (again using resources to buy off domestic interests blocking a deal, if necessary). In short, policies that would have raised the value of the U.S. franchise. Remember how virtually the entire world rallied around the U.S. post 911; think of the leadership opportunities that provided.
Posted by: RS-K at September 12, 2006 07:37 AM
Good post. You demonstrate why such economic analysis if simply academic. When you are talking man-made disasters and political considerations you simply cannot calculate such things in economic terms. What would have been the economic impact invading Cuba in the 1960s and killing Castro? What would have been the economic impact of Saddam getting nuclear technology from Pakistan?
Posted by: Dick at September 12, 2006 10:39 AM
"Now that the long-delayed Senate report on pre-Iraq War intelligence has finally put the Republican imprimatur on the well-established fact that the case for Iraqi WMD's had been hyped"
Hype: "Something deliberately misleading; a deception"
I didn't see any mention in the Bloomburg article you referenced to "hype" or deliberate deception. The lead sentence of your posting did appear a little heavy in hyperbole, however. It seems to me that the rest of your post could have stood quite nicely without that. Leading by example, as it were?
Posted by: anonymous at September 12, 2006 01:14 PM
Laszlo Varro: I agree with Alex; I don't believe the analogy with South Korea is apt. I do agree that there is a probability Y% > 0 for the scenario laid out, but I'd say Y% is approximately 0 for the next couple of decades.
Dick: You miss the point. I'm not arguing that everything can be reduced to dollars. But I am arguing that benefit-cost assessment (including pecuniary and non-pecuniary factors) is a vital tool for analysis, one that I think was not undertaken; rather ideology drove policymaking. Call me a technocrat if you wish.
Anonymous: I didn't put in the links because I thought that almost anybody could find plenty of documentation in support of the proposition that evidence of in favor of an active WMD program was either manufactured, or accepted in the face of much skepticism in the intelligence community. Example 1: aluminum tubes. Example 2: yellowcake as in purchases from Niger. Example 3: mobile biological laboratories. I'm certain many other readers can add many, many other instances.
Posted by: menzie chinn at September 12, 2006 10:38 PM
I did not miss your point, but you may have. By stating up front that the Republicans made WMD the primary issue made your analysis political rather than technical. There are many reasons the war on terror is being fought and the issue of WMD is still a serious consideration.
What would you calculate would be the economic cost of nuclear devices being detonated simultaneously in New York, Boston, Philadelphia, Washington DC, Chicago, LA, San Francisco, Seattle, Dallas, and Atlanta? How much should you factor into your analysis for the political consequences of complacency concerning WMD?
Posted by: Dick at September 13, 2006 05:08 AM
Dick: No, I understand that point when I wrote the piece. The question I have for you is the following. In diverting resources away from fighting groups (al Qaeda) or from stabilizing states that may revert to anarchy (Afghanistan), or for that matter securing the ports by investing more in inspections, toward fighting in Iraq (a state that was contained, had no WMD programs in operation, even if it had plans for such programs), have we raised the probability of such an attack as you mentioned. If the answer is yes, then the benefit-cost ratio for our venture in Iraq drops even further.
Posted by: menzie chinn at September 13, 2006 09:45 AM
Your examples (AL tubes, ...) are ones of 20/20 hindsight. Your lead sentence "... the case for Iraqi WMD's had been hyped ..." sounds as if you are saying that those "in power" deliberately deceived us for the purpose of initiating war. Such things may be obvious to those that have drunk the same brand of "political cool-aid", but to many of us they are opinion-- in particular, politically colored opinion.
You're clearly free to preach whatever kind of political religion you wish to in your econ blogs. My point, however, is that mixing political barbs with economics (presumed to be more scientific and thus more provable) draws attention away from the key focus of your message-- something to do with economics I believe?
Posted by: anonymous at September 14, 2006 10:24 AM
Anonymous: I'm sorry, but you are simply incorrect when you characterize the aluminum tube issue as one of 20-20 hindsight. DoE experts had clearly and unambiguously debunked the assertion that these tubes were to be used for centrifuges before the invasion.
Posted by: menzie chinn at September 14, 2006 10:27 AM
I really didn't mean to belabor this point, or drag you into belaboring yours. But here goes one last belaborment-- we could debate the AL tube issue--
who knew and who didn't-- but that leads to the "hype" in your lead sentence. Who "hyped"? If the "who" is provable, they should serve time. If it's not, it's opinion. Either way-- and more importantly-- why does it belong in an economics posting?
There are highly trained political pundits galore I can listen to all day long if I want to hear politics. I come here to learn from experts on economic matters, not politics.
My message, take it for what it's worth as feedback from a single individual, is that when a scientific person feels the need to preach politics in the middle of a scientific message, they automatically lose credibility with me on all fronts. Maybe I'm in the minority on that-- it's your decision-- isn't freedom of speech a great thing!
Posted by: anonymous at September 14, 2006 12:43 PM
I have to disagree with your. I was, before the invation, mildly in favor. And yet, at the time, it was obvious that the WMD was a canard.
For each instance of "irrefutable evidence" there were always lots of better informed persons refuting the evidence.
A perfect case in point is the issue with Ambassador Wilson and Yellow Cake.
A polititian (the President) claims that Iraq has contracted to buy something, and here's the evidence....
Then CIA sends a man (Wilson) to investigate, and he shows how every piece of evidence is fabricated.
President has an ax to grind, Wilson doesn't (at least not an obvious one).
Time after time we see the same pattern.
I, personally, thought that an invation was possibly worthwhile, if you considered the possibility of proper nation building and the certainty that the containment policy was unsustainable in the long run
I still think that the invasion was a moderatly good idea, the problem was the utter incompetence with which it was conducted, and the complete absence of a proper economic analysis beforehand.
Posted by: bellanson at September 14, 2006 12:47 PM
Thanks for the reply.
I am one of those people who believe that both Afghanistan and Iraq are critical to our war on terror so I do not believe that we have diverted forces from fighting Al Qaeda or from Afghanistan. To correctly evaluate the cost/benefit of the War on Terror we must understand that it is international in scope. We have resources working from the Philippines to Great Britain, from New York to Baghdad.
But I do need to ask you a question. Do you believe that the leadership of Al Qaeda means what they say concerning their view of the War on Terror? If you do then you must accept that they believe that they are fighting WWIII and that Iraq is the primary battlefield.
Not the excerpt below:
In the message broadcast by Al-Jazeera television, Ayman al-Zawahiri, second in command to Usama bin Laden, said that Al Qaeda now views "all the world as a battlefield open in front of us."
"It is a Jihad for the sake of God and will last until (our) religion prevails ... from Spain to Iraq," al-Zawahiri said. "We will attack everywhere."
Posted by: Dick at September 15, 2006 07:26 AM
Anonymous: After long reflection (at least by blogosphere standards), I have come to the conclusion that the definition you provided:
Hype: "Something deliberately misleading; a deception"
is as accurate a description as I could have come up with. I think the consensus of informed observers has shifting this way, and over time yet more information will come out that supports the view that the Administration cherry picked intelligence, and suppressed other information, in order to present half-truths. And full non-truths, as in the aluminum tubes instance.
As for your statement:
Who "hyped"? If the "who" is provable, they should serve time.
I disagree. Ever since "Remember the Maine", and before, hype has not been illegal, unless it is imbedded in a sworn statement. If the public believes that this is a serious offense, then the democratic process should hold the relevant official(s) accountable by removing them from office.
Dick: Of course, the conflict is international in nature. As I've said before, the metaphor of war is inappropriate in my mind; the (short-lived) DoD moniker "Global Struggle Against Violent Extremism" was, I thought, a better characterization, since we cannot "kill" every person bent on harming the U.S. But that is besides your main point, to which my response is: "When everything is a priority, nothing is a priority." Stabilizing Afghanistan, eliminating in a determined manner the Al Qaeda leadership, devoting scarce intelligence resources to tracking down foreign and domestic terrorist elements (instead of searching for WMD's in Iraq), investing in port security, investing in safeguarding chemical factories, investing in securing air cargo all should have been placed higher on the list.
Posted by: menzie chinn at September 15, 2006 10:53 AM