~by Steve Friess, University of Chicago Magazine, Winter, 2007
A landmark study by three Chicago GSB economists delves in great depth into exactly those questions with results sure to be at least as controversial as any other facet of the ongoing war in Iraq. Their conclusions: The war is certainly expensive for the United States, but perhaps no more expensive than an indefinite policy of containing the Saddam Hussein regime.

Considering the economic and human consequences of Hussein’s brutal reign for the Iraqi people, even a lengthy U.S.-led occupation there could be reasonably seen as something of a bargain, according to research from professors Steven Davis, Kevin Murphy, and Robert Topel and published last spring by the National Bureau of Economic Research.

One purpose of the exercise, Davis says, is to reframe the debate regarding this conflict and future ones of this nature – to provide a framework for assessing costs and weighing options.The trio approached the calculations from an ex ante perspective using data available as of the arch 2003 invasion. They offer a range of possible price tags for the war depending on the intensity of the major combat phase, the length and size of a U.S. occupation, the extent of any insurgency,U.S.outlays for the reconstruction of Iraq, and other factors.

“We started this project before the war because we were dissatisfied with the nature of the public discourse in the media and our discussions with colleagues,”Davis said. “We see it as an application of a useful analytical approach, and we tried to do it carefully even though we’re sure to miss some things ex post. It’s a tool for framing the analysis and trying to have a discussion – rather than a public rant about life-and-death issues.”

Murphy is more blunt: “Anti-war advocates complain about the cost of the war. Well, there also was a cost of no war. The alternative to invading Iraq wasn’t to do nothing. It was containment. There wasn’t much discussion of the cost of that.” That cost, according to the study, would have been $300 billion in expected presentvalue terms and ranges from $200 billion to $700 billion under some scenarios. By contrast, the cost of war has a low-end estimate of $100 billion under the best-case scenario – which clearly did not happen – and tops out at $870 billion in the most disastrous prospect they consider. The authors also analyze the economic and human costs of the war-versus-containment choice from an Iraqi perspective, concluding that war and the forcible removal of Hussein was far better for the Iraqi economy and likely to produce a smaller Iraqi death toll.

“We initially assume that the two policies are equally effective in dealing with the threats posed by a dangerous regime,” Davis, Murphy, and Topel write. “We also assume at least one policy has a positive net present-value. Given these assumptions, the choice between policy options reduces to presentvalue cost comparisons.”
Murphy acknowledges his own private views were borne out by the research. “I know enough not to guess in advance, but I wouldn’t say I was really surprised,” said Murphy, a supporter of the preemptive attack even before this analysis was conducted. “I can’t say there was any place where I said,’Wow, I wouldn’t have guessed that would or wouldn’t matter so much.’”

The Cost of Containment

The report starts out explaining how the authors arrive at their estimates for the cost of containment, which, Davis acknowledges, was harder to derive and impossible to ever prove – unlike the costs for a war that actually happens.

Indeed, there was no way to know how long a containment policy would take to produce a favorable regime change in Iraq, but the researchers had to grapple with the issue to assess the cost of containment. They note that containment of the Soviet Union lasted 44 years after World War II, until the Cold War ended peacefully, and that holding North Korea in check has been a task of more than 50 years and counting.With Hussein’s sons well placed to carry on his tyranny, they reasoned, there was little cause to expect a favorable change any time soon. The trio settled on a working figure of 33 more years for a policy of containment, based on an “optimistic scenario” with a 3 percent chance each year that the regime would peacefully evolve from threatening to benign.

The containment policy involved economic sanctions, disarmament requirements, weapons inspections, no-fly zones over the northern and southern areas of Iraq, and a naval presence to enforce trade restrictions. Citing a 2001 Department of Defense memo, they noted that containment required a U.S. military force of about 28,000 troops, 30 ships, and 200 planes, among other resources. They applied an annual per-troop cost of $226,000 based on past U.S. military interventions to obtain a $6.4 billion annual cost for personnel and personnel support.They estimated the cost of military hardware and munitions based on data from the Congressional Budget Office and other sources to arrive at a figure of $4.8 billion per year for capital costs. That sums to about $11.3 billion per year.

They also considered another calculation that relies on different data and assumptions to obtain an all-in cost per Army,Navy, and Air Force personnel. This alternative calculation yielded a much higher cost of $17.8 billion per year. The authors opted to average the two figures, yielding an estimate of $14.5 billion for the annual cost of containment. At conventional discount rates and a 3 percent annual chance of peaceful regime change, that figure implies a present-value cost of about $300 billion.

At the same time, the researchers noted that the annual cost could well be less than $14.5 billion, if some of the forces deployed for containment would have been engaged in the region anyway; that is, even if Iraq were not a threat. Thus, they also offer a lower baseline for annual containment costs of $9.7 billion, which yields a present-value cost of about $200 billion.

Still, there are other contingencies to consider. The $14.5 billion figure presumes a stable containment environment, one in which the same troop levels remain in place until the Iraqi regime peacefully evolves into one that is neither hostile nor tyrannical. Yet, the authors note, the history of containment after the first Gulf War in 1991 had its share of disturbances, including occasions where additional troops were needed to deter Hussein from strikes against neighbors. Hussein had also routinely thwarted United Nations weapons inspectors, which led to costly military actions. Finally, some national security experts feared that the erosion of international political support for the containment policy would allow the Hussein regime to re-arm and bring about a costly war with Iraq down the road. Depending on their likelihoods and costs, these contingencies raise the present-value cost of containment by anywhere from a small amount to a few hundred billion dollars.

“Our calculations show that one common argument against the war makes little sense – that because this was a war of choice, it was therefore a mistake,” Davis said. “I don’t know whether the war was a mistake, but the argument is fallacious. If you hold the view that we were going to fight a war against Iraq anyway, one that would be more expensive if we waited longer, then, yes, it’s a war of choice to go in 2003, but it’s a choice that makes sense given your view of the alternatives.”

Davis,Murphy, and Topel also consider the potential cost of another terrorist attack on American interests on the scale of 9/11. There’s no evidence of Hussein involvement in the Al Qaeda-executed 9/11 attacks, but containment involved risks that the regime would support anti-American terrorist organizations. The regime had a history of pursuing and using weapons of mass destruction and had openly supported terrorists in the past.Another terrorist attack on the scale of 9/11 could cost $100 billion dollars, the authors estimate. If the survival of the Saddam Hussein regime raised the probability of such an attack, then containment becomes a more costly policy option. “Opinions differ greatly on whether a continuation of the prewar containment policy would have raised or lowered the probability of another major terrorist attack on U.S. soil,” remarks Davis. “And these opinions color one’s view of the war.”

The Cost of War

Granted, the cost of the Iraq war – from the vantage point of March 2003 – was equally hard to pin down, so the writers opted to sketch out several possible scenarios, ranging from a two-month major combat phase and brief occupation to four months of major combat, a 10-year occupation, a major insurgency, and an ultimate retreat back to containment. In retrospect, the options that look closest to the likely U.S. experience involve three months of major combat at the outset, a significant insurgency requiring a 5- to 10-year occupation, and U.S. reconstruction and humanitarian assistance costs ranging from $36 billion to $81 billion. These outcomes produce U.S. costs in the range of $414 billion to $633 billion in present-value terms, according to the authors’ estimates.

The war cost estimates for the United States include those for military personnel, equipment, munitions, fuel and materials, humanitarian assistance, reconstruction aid, and U.S. casualties. Based on economic research on the value of a statistical life, the authors assigned a cost of $6.9 million per U.S. military fatality and lower costs for wounded soldiers that depend on the severity of the injury and the estimated lifetime costs of medical treatment.

The Cost to the Iraqi People

The authors also assess the impact of each policy on the Iraqi people and their economy. Hussein’s regime, they observe, had prompted a “catastrophic collapse in living standards” during which real GDP per capita plummeted from $9,000 in 1979 to about $1,200 in 2001. In that span, Hussein led the nation into two major wars – against Iran and against the U.S.-led coalition that ousted Iraq from Kuwait in 1991- and brought on ruinous economic and trade sanctions. Hussein diverted much of the nation’s remaining resources into an oversized military, an extensive apparatus of terror and repression, and some 50 lavish presidential palaces and other monuments.

Given that, said Murphy, antiwar advocates who point only to the financial costs and human suffering caused by the war lose credibility. “The only way to say that the war is wrong is then to say that getting rid of Hussein is a bad idea. And I don’t think even the war opponents say that getting rid of Hussein is a bad idea. They just disagree with how it happened.”

Davis,Murphy, and Topel compared the potential economic impacts on the Iraqi people from continued containment versus near-term regime change brought about by war. Assuming a 20-year transition period to per capita income levels comparable to those experienced by Iraqis before Saddam Hussein’s misrule, the authors estimate that war and forcible regime change would lead to a 50 percent improvement in economic well-being for the average Iraqi. The size of the improvement depends on the extent of the destruction caused by war, the success of reconstruction following war, and other factors.

The final consideration of the paper involves a comparison of lost Iraqi lives under war as compared to a policy of containment. At least 500,000 Iraqis died prematurely under Hussein, including 200,000 Kurds slaughtered in 1989 and 1990 and tens of thousands of Shi’ah killed during brutal repression after the 1991 war. Based on the experience from 1991 to 2002, the authors conclude that a continuation of the containment policy would have meant 10,000 to 30,000 premature Iraqi deaths per year. They estimate as many as 600,000 Iraqis would die under a continued Hussein regime of unknown duration. By contrast, the most pessimistic prewar projections the authors reported set the high end of an expected Iraqi death toll at 100,000.

“Much of the current discouragement over the war centers on the daily tragedy of Iraqi deaths, authored largely by remnants of the old regime,” observes Topel. “We rightly should care about those deaths, but we should give equal weight to the thousands who would have died at Hussein’s hands under continued containment. It’s a gruesome tradeoff, but Hussein’s past suggests that the death toll would have been larger had he remained in power. Economists always ask ‘What’s the alternative?’And that’s the alternative.”

While some may claim the authors have pro-war political agendas that are reflected in their conclusions, Davis is quick to note that aspects of their analysis can be used to criticize the Bush administration’s failure to plan properly for the aftermath of the invasion and occupation. The analysis “forces us to confront contingencies that might prove to be costly, and that apparently were not well factored into the pre-war thinking,” he said. “The United States government did not plan adequately for the postwar, in part because it did not systematically confront the possibility that war and regime change would be costly and difficult.”

That said, Murphy isn’t convinced that the expense of the current transition is much different – or higher – than it might have been had containment actually succeeded, so he questions whether the cost of bad postwar planning is relevant to the paper.

“I’m not sure the period of instability now is greater because there was a war versus, say, a Shi’ite overthrow of Hussein that might have been the means of Hussein’s removal,” Murphy said. “There was simply no way you’re going to take that much power away from the Sunni minority in Iraq and have a low-cost transition.”

Assessing Other Conflicts

Although this paper focuses on Iraq, its real usefulness may be for other conflicts, since Iraq is already playing itself out, and most Americans already have entrenched opinions on the subject.Applying a careful cost-benefit analysis to various war and containment options with respect to Iran might show that stanching the Iranian threat now would be far less expensive than at some future date, when Iran has nuclear weapons. North Korea, however, is believed to already possess nuclear weapons, so a cost-benefit analysis of a military intervention in North Korea would need to factor in the probability and cost of a full-scale nuclear war, making containment the more favorable option in economic and human terms,Murphy said.

“Our concern is that decision-making discourse wasn’t so good on either side in this case,” said Murphy, pointing to the fact that the Bush administration hung its entire reasoning for going to war on the concern that Hussein had WMDs, when the data and analysis in this paper show there were other, perhaps equally persuasive reasons. “Maybe people do these calculations behind the scenes, but I don’t think they necessarily do them well.” Indeed, Davis insists that the main point wasn’t to advocate a particular policy, and he acknowledges that other costs of the war – including the possibility that it actually raised the risk of terror attacks against U.S. interests – are unquantifiable today.

Indeed, Davis insists that the main point wasn’t to advocate a particular policy, and he acknowledges that other costs of the war – including the possibility that it actually raised the risk of terror attacks against U.S. interests – are unquantifiable today.

“As economists and social scientists, we don’t have a lot to say about that issue right now,” Davis said. “Maybe a decade or two from now, with some intensive social science research to back us up, we’ll reach a firm conclusion about that. But right now, it’s just a matter of personal judgment.”

Steven Davis is William H. Abbott Professor of International Business and Economics, Kevin Murphy is George J. Stigler Distinguished Service Professor of Economics, and Robert Topel is Isidore Brown and Gladys J. Brown Professor in Urban and Labor Economics.