Bryan Caplan  

Steelmanning the Iraq War

The West was right about China... Reminder on Troy Speech Today...
The Iraq War started 15 years ago today.  I always opposed it, for my standard pacifist reasons.  But here is a case for the Iraq War that would have intellectually and morally impressed me at the time.  To be clear: Though I'm the author, I strongly disagree with this speech.  Still, I'd enjoy talking to someone who sincerely believed it.

You can treat what follows as a steelmanning exercise.  (It's not really an Ideological Turing Test because as far as I know, no prominent advocate of the Iraq War would agree with it).  Alternately, you can treat it as mirror: Actual war-makers are blameworthy insofar as they fall short of the standards it exemplifies.

My fellow Americans,

In World War II, over 400,000 American soldiers lost their lives over the course of four years.  It was a tremendous and tragic loss.  But it was absolutely worth it.  The sacrifice of the fallen is the foundation of the amazingly peaceful and prosperous world in which we live.  Yes, we take their achievement for granted.  But the achievement was so great that it would have been worth paying a far steeper price.
Now our nation and the civilized world face another grave challenge.  We saw it plainly in the terrorist attacks of September 11, 2001.  But those attacks are only a symptom of a festering threat to the peace and prosperity of the world.  What is that threat?  Though I fear to alienate possible allies, the best name for that threat is: Muslim tyranny.  Whether Sunni, Shiites, or "secular," the Muslim world is almost entirely ruled by governments that have little respect for democracy, and even less for human rights.  After years of study - and careful analysis of DARPA-sponsored prediction markets - I conclude with heavy heart that Muslim tyranny will not fix itself.  Indeed, its theory and practice is spreading and intensifying, threatening Central Asia, Africa, and even Europe.

For now, I freely admit, Muslim tyranny poses little military threat to the civilized world.  But the same was once true for Communism and fascism.  These threats could and should have been removed in their infancy, sparing mankind countless horrors.  While we cannot undo the mistakes of the past, we can avoid repeating them.  As your leader, I say we must.

Make no mistake about it: Our mission will be painful and long.  If you are not prepared to lose a million American lives to achieve lasting victory, we should not go to war.  If you are not prepared for a hundred-year occupation, we should not go to war.  If you are not prepared for a thousand domestic retaliatory terrorist attacks, we should not go to war.  If you are not prepared for the war to spread far beyond the borders of Iraq, we should not go to war. 

I do not seek enthusiastic but short-lived support; indeed, fickle support is more dangerous than thoughtful opposition.  Instead, I ask each of you to visualize the immense and lasting suffering our country and the world are going to endure if we follow my lead.  Indeed, I ask you to visualize the vast numbers of innocent lives our war will destroy.  Think of all the children the United States and its allies burned to cinders in World War II.  To win, we will have to do the same.  Nothing can justify such atrocities - except a high probability of making Muslim tyranny history. 

Why start with Iraq?  By the standards of the region, Saddam Hussein's Baathist regime is "secular."  But it is a ghastly tyranny, and its Islamic roots insulate it from the life-giving ideas of human rights and democracy.  Furthermore, it is extremely diplomatically isolated.  Militarily, we can defeat them with ease - and turn Iraq into a model for the rest of the Muslim world.

It would be criminal to invade Iraq without meticulously describing our model in advance.  So let me share it.  While our goal is to bring human rights and democracy to Iraq, human rights will come first.  Democracy will only come when human rights in Iraq are democratically sustainable.  This distinction is crucial, because Muslim tyranny has deep cultural roots.  Saddam Hussein is not personally popular Iraq, but he's a lot more popular than the ideals of liberty.

What does all this mean in practice?  Let me be blunt.  We will give Iraq full democracy once gay couples can walk the streets of Baghdad holding hands.  We will give Iraq full democracy once ex-Muslims can sleep soundly in Mosul after publicly preaching atheism on Iraqi television.  We will give Iraq full democracy once violence between Sunni and Shia in Iraq is as common as violence between Catholic and Protestant in the U.S.  And not before. 

As in post-war Germany and Japan, we will hold elections.  But only candidates who embrace our model will be allowed to run - and any elected official who refuses full cooperation with the American military occupation will be summarily removed.  Our enemies will no doubt call this "imperialism."  I say this is bigotry on their part; if American rule is the only credible way to protect human rights in Iraq, people of all nations should support American rule.

Many advised me not to use the phrase "Muslim tyranny."  But I honestly couldn't think of a better one.  All of the major religions have, at one point, provided an ideological foundation for tyranny.  But Islam is the only major religion that continues to serve this function.  We're going to end that once and for all.  Freedom of religion is a basic human right - but imposing your religion on others is not.

I'm sure many of you are thinking, "He's asking a lot."  You're right.  To repeat, Iraq is only the beginning.  In the best-case scenario, the many surrounding Muslim tyrannies will see that we mean business, and earnestly launch domestic reforms.  I welcome such developments with open arms, but we should not count on anything of the sort.  Instead, we should expect Muslim tyranny to get worse before its gets better.  Fortunately, to repeat, they are militarily no match for us.  Our only scarce resource is resolve.  As long as we are willing to lose a million American lives over the next century, we will do for the Muslim world what we did for Germany and Japan: bring human rights to their people and security to the world.

I know that our enemies will selectively quote this speech to make me seem like a monster.  And I know that my predecessors gave little heed to the innocent foreign lives they took in pursuit of victory.  Shame on them!  So let me say this: If I could end Muslim tyranny without killing a single person, I would gladly do it.  Any leader who wishes to spare his people the horrors of war can do so by immediately unconditionally surrendering to us.  If that prospect frightens you, look how we treated the people of Germany and Japan when World War II ended.  For us, there is no victory until we turn our most wretched enemies into flourishing friends.  I fondly look forward to the day when Disneyland is packed with Iraqi tourists.

In the darkest days of World War II, Winston Churchill told the British people, "I have nothing to offer but blood, toil, tears, and sweat."  I admire these sentiments, but I know that our enemy is not yet at the gate.  There is a far worse course than doing nothing: Invading Iraq in anger, then abandoning it in frustration.  But our best option is to excise Muslim tyranny now when it's weak, instead of waiting for this political cancer to spread.  My fellow Americans, are you with me?

COMMENTS (21 to date)
Jim writes:

Often wonder how future historians will timeline the American empire?

Starts: American Revolution
Warms up: Civil War (or Northern aggression sold as changing names of slaves to sharecroppers)
Sizzles: WW1 (Wilson’s War to end all wars)
Cooks: WW2 (or Great Patriotic War as victorious Ruskies call it)
Boils over: Vietnam War (“American” War as Vietnamese call it)
Evaporate: the Christian Crusade era of 2000
Blows up: when petrodollar blows out like an oil well trying to fund it and full Weimar Republic inflation kicks
Peaceful ending: crypto currencies take over and nation states evaporate, now seen as barbaric ant colonies.

DWAnderson writes:

That is pretty good, but I would add the following:

On September 11 we suffered a grievous loss, one that caused hundred's of billions of dollars in not only in physical damage, but in fear and losses imposed by security measures taken in response to that fear. Those security measures like new airport security and greater surveillance also cost us some significant liberty and benefits of an open society.

We could sit back and play defense at home and abroad. We could increase security at home. We could continue to contain threats abroad. But that is asymmetricly costly for us in dollars and in ill will abroad.

Besides, defense is not the American way and it yields the initiative. Better to strike a quick blow now to create a model state in the heart of the Middle East. There are risks, but the potentila reward is worth it. A free Iraq would serve to inspire movements toward freedom in its neighbors eventually resulting in far greater freedom for hundreds of millions.

Note: that was my thought at the time. Since then I have come to believe that I overestimated the odds of success (but once burned, I not all that confident in my revised assessment).

Mark Bahner writes:
My fellow Americans, are you with me?

Who in the U.S. would answer affirmatively, given these previous statements:

If you are not prepared to lose a million American lives to achieve lasting victory, we should not go to war. If you are not prepared for a hundred-year occupation, we should not go to war. If you are not prepared for a thousand domestic retaliatory terrorist attacks, we should not go to war. If you are not prepared for the war to spread far beyond the borders of Iraq, we should not go to war.


Mark Z writes:

I think the use of Germany and Japan as case studies is a pretty good strategy. I'd probably emphasize the idea that the benefits of liberal democracy in Iraq will be reaped by countless generations to come, and that almost any cost paid by one generation will be dwarfed by improving the quality of life for every generation of Iraqis henceforth; could even appeal to 'domino theory' and argue that a flourishing Iraqi democracy will precipitate the rest of the region to change as well, as Iraq's neighbors yearn to share in their prosperity.

Curiously lacking in your speech, though, are economic arguments for the war: opening up Iraq's economy to the world, establishing secure property rights, etc. and all the benefits those could hypothetically bring.

Lastly, another argument I'd consider making (perhaps inconsistent with your sanguine speech): the war will be worth it even if it fails to make Iraq a better place, because by punishing the Iraqi government for its abuses, we will deter authoritarianism, anti-Americanism, and crimes against humanity in the future.

Dave D writes:

To the best of my knowledge, the average rate of Iraqi civilian deaths during the U.S. occupation is lower than the average rate of Iraqi civilian deaths during the Hussein regime. That alone is a (slight but incomplete) justification for the war.

Now I have to ask Professor Caplan: how are you against both immigration restrictions and foreign intervention? Do you believe that immigrants from tyrannical nations won't bring their tyrannical ideas with them?

Matthias Görgens writes:

Dave D, that's a good point, but not quite there yet: conditions have improved in other countries as well with passage of time without any invasion.

(Just steady progress thanks to human effort and technology.)

BC writes:

This doesn't seem like steelmanning to me, as it doesn't really present a stronger case than Iraq War supporters' actual case. According to Wikipedia, US deaths in Iraq so far number 4491, so the prediction of 1M American lives lost seems too high. Granted, we did not achieve anywhere near the victory that Caplan calls for, but the gap between 1M and less than 5k is huge.

Caplan bases the estimate from WW2, but that was between evenly matched powers. Casualties are much lower when one side is much stronger than the other. Some estimate, for example, that the dropping of the atomic bomb on Hiroshima and Nagasaki saved hundreds of thousands or millions of lives. Had the US developed the bomb at the beginning instead of end of WW2, might the US casualty count have been much less then 400k? One could imagine that even the German and Japanese casualties would have been significantly reduced. Hiroshima killed between 90-146k and Nagasaki 39-80k. Germany lost 6.6M-8.8M and Japan 2.6M-3.1M in all of WW2 (civilian+military), so asymmetric strength ends a war more quickly, resulting in fewer casualties. (I'm not suggesting that the US should have nuked Iraq, just drawing an analogy between the asymmetries in strength between a nuclear-armed US vs. non-nuclear Japan-Germany in WW2 and US vs. Iraq and how that asymmetry reduces casualties compared to evenly matched non-nuclear US vs. Japan-Germany in WW2.)

Scott Sumner always mentions that, in monetary policy, a credibly committed central bank actually has to do less than an uncommitted bank. Does Caplan mean that the strongest case for war would involve the US credibly committing to being willing to sacrifice 1M lives so that it would end up sacrificing much fewer? If so, there were probably other ways of demonstrating committment, e.g., building a permanent military base in Iraq like the ones we have in Japan. I guess you could call that "100-yr occupation" if you want, but I'm not sure that we are considered to be occupying Japan, South Korea, Germany, etc. right now.

Peter Gerdes writes:

Seems to me the argument that was actually given (alongside the BS about WMDs) was pretty decent and not too far from what you said here.

Basically the argument was: Look, Iraq is a miserable tyranny that treats its people in awful ways and makes them suffer and creating negative externalities in the region. It seems plausible to think that the Iraqi people are educated enough and have sufficent institutions to make a more humane government work.

I mean it turned out that it didn't work out but estimating the probabilities here is really really hard and other than ideology its not at all clear that the pro-war argument was actually epistemically in the wrong rather than unlucky.


On a different point if Sadam was able to keep peace via brutality and you believe that this peace was preferable to the persisting violence after the invasion doesn't that suggest that the US or its proxies (once they realized what was happening) should have adopted those brutal measures?

Shane L writes:

I agree with BC. If anything, I would suggest this was an unusually weak argument.

Invading Iraq with the explicit aim of permanently destroying large parts of their culture would terrify and possibly draw together other peoples around the world.

Pew surveys find that majorities in Asia/Pacific (65%), Africa (87%) and the Middle East (92%) think homosexuality is morally unacceptable. If the author was planning to occupy Iraq, without elections, until "gay couples can walk the streets of Baghdad holding hands" it implies that much of the world would be at risk of arrogant American imperialism, not just the Muslims. Picture a pan-global anti-American alliance and probable nuclear proliferation.

The conventional arguments for war in 2003 were less ambitious, to my memory, and less intimidating to the rest of the world. Several Muslim-majority countries joined the invasion. If the message instead had been: "we're coming for all of you - Asian or African; Muslim, Christian, Hindu or Buddhist - until you conform to our beliefs", I would expect a fierce near-global opposition to coalesce.

Miguel Madeira writes:

A strong argument against this imaginary position could be "In many ways, this is not what Saddam already is?", a non-democratic government that repressed political opposition but were relative "liberal" (at least, for regional standards) in cultural and private live issues (same people say that, if you stayed away from politics, you have the same freedom in Iraq then in the West).

And, don't forget that, in 2003, the US already had occupied the Afghanistan - if you already are in charge of a Muslim country, why not to try to end Muslim tyranny there instead of conquering another country?

Nathan Smith writes:

While it's wise to prioritize human rights over democracy, public displays of affection are a very minor aspect of human rights at best. Freedom of religion and free speech are the really essential human rights, along with some sort of protection of property rights and exchange.

The war in Iraq as it was actually fought was a just and beneficent undertaking. An openly imperialist war seeking to occupy the Islamic world indefinitely to impose the Sexual Revolution would not have been.

Brent Buckner writes:

I think Den Beste was making a few of those arguments in 2003. (c.f. )

Andrew_FL writes:

This is like some hawkish alt right case for and approach to Iraq. Weird that you think this is a stronger case than the actual case.

Andy Hallman writes:

Hi Bryan. This is a neat idea, but I have a few criticisms.

1) I'd suggest using a country's own recent history as a predictive guide instead of very different countries like Germany and Japan in the 1940s. I notice the word "Iran" does not appear in your argument, even though Iran tried something similar in the 1980s. After Iraq invaded Iran, Iran pushed them back and even went into Iraq, hoping that the majority Shia would support their fellow Shias in Iran by overthrowing Saddam. It didn't happen. That suggests that even when invaders are of the same religion and even the same sect, they're still unwelcome. Americans would be forcing even more foreign customs on the Iraqis than Iranians.

2) I would change the thought experiment so we're asking questions like "Under what circumstances would an invasion be just?" instead of "Given a particular invasion, how can we make it look good?" Doing it the first way allows you to answer the question honestly in your own voice rather than pretending to hold views you really don't.

JohnB writes:

Another thing missing is the alternative to crushing the enemy now. If we don't do this horrible and costly action, then what even more horrible and more costly future will occur?

Right now it reads as a request for a horrible and costly war which isn't necessary.

JohnB writes:

Another thing missing is the alternative to crushing the enemy now. If we don't do this horrible and costly action, then what even more horrible and more costly future will occur?

Right now it reads as a request for a horrible and costly war which isn't necessary.

Dan Hill writes:

Your steelman is made of an inferior quality of steel and rusting badly. I'm not sure there is any scenario based on realism rather than wishful thinking that would lead to the conclusion to proceed with the invasion.

The analogy with post-WWII Germany and Japan lacks one critical thing - something to keep the occupied population in line. In the case of (West) Germany and Japan it was fear of the Communist powers. No such parallel exists in Iraq, which is why we had such a vicious insurgency (something that many predicted would happen in occupied Germany). Plus neither country ever had anything like the Sunni-Shite divide in Iraq as a destabilizing force.

Floccina writes:

I was against the war at the but the strongest argument for it was that the Government of Saddam Hussein was kill people at high rate.

In January 2004, Human Rights Watch stated: "Having devoted extensive time and effort to documenting [Saddam's] atrocities, we estimate that in the last twenty-five years of Ba'th Party rule the Iraqi government murdered or 'disappeared' some quarter of a million Iraqis, if not more."[23][24] The 1988 Al-Anfal campaign resulted in the death of 50,000-100,000 Kurds (although Kurdish sources have cited a higher figure of 182,000), while 25,000-100,000 civilians and rebels were killed during the suppression of the 1991 uprisings.[10][25] In addition, 4,000 prisoners at Abu Ghraib prison were reportedly executed in a particularly large 1984 purge.[26] Far fewer Iraqis are known to have been executed during other years of Saddam's rule. For example, "Amnesty International reported that in 1981 over 350 people were officially executed in Iraq ... the Committee Against Repression in Iraq gives biographic particulars on 798 executions (along with 264 killings of unknown persons, and 428 biographies of unsentenced detainees and disappeared persons)." Kanan Makiya cautions that a focus on the death toll obscures the full extent of "the terror inside Iraq," which was largely the product of the pervasive secret police and systematic use of torture.[22]
Plucky writes:

I supported the Iraq war at the time and to this day do not consider it a mistake (I do think that there were a number of terrible choices made in the next 3 steps of the decision tree, but that's another discussion)

There are three main criticisms I'd have of your argument

1) A similar argument could have been made of a number of ghastly tyrannies in various times and places. You could just as easily have put this argument forward for invading Cuba at various times and plausibly argued that there'd be both a higher chance of success and a lower cost in lives vs Iraq. As powerful as the US military is, it still faces a budget constraint and so any pro-invasion argument must make clear why specifically Iraq and why specifically then. Your steelman argument is pretty perfunctory on that point.

You are correct that 9/11 is key, but go the wrong direction. Liberal (classical sense) societies can withstand terrorism on the scale of car bombs well enough, as evidenced by the British experience with the IRA, Spain with the Basque terrorist groups, and assorted communist terrorists in Italy and Germany. What liberal societies may not be able to withstand is terrorism designed to inflict maximum damage, either by targeting critical infrastructure or by use of WMDs (note that this does not just mean nukes- a serious anthrax attack could make a major city uninhabitable for months). 9/11 demonstrated two things: 1)such terrorism was a very live threat and 2)the organizational capability to achieve that kind of attack required the tacit support of a state or state-ish entity that exercised de-facto sovereign control of territory. The corollary those two realizations is that any state in the world that tolerated organized terrorist groups with such aims represents an intolerable security threat, regardless of their (in)ability to engage in military expansionism in the mode of 20th century fascism and communism. That such a consequential security threat could originate from such a remote, backwards, and conventionally weak country as Afghanistan was the big change of 9/11.

The trifecta of critical state threat was 1)overt mortal hostility to the US, 2)historical use or support for terrorism as state policy, and 3)possession or obvious desire for possession of WMDs. Thus Bush's naming of Iraq, Iran, and North Korea as the "axis of evil", as only they hit all three criteria. Syria, Pakistan, and Libya c. 2003 hit 2) and 3) but not 1). Cuba hit 1), arguably 2), but not 3). FARC-controlled territory in Colombia only satisfied 2). Between Iraq, Iran, and North Korea, Iraq was obviously the least difficult target for invasion/overthrow and one for which there was no realistic prospect of internal regime collapse, as there was/is for Iran and NK. Unlike NK Iraq was also not considered a buffer state in its sphere of influence by any great military power.

The nature of the security threat did not require that post-invasion Iraq become Switzerland to resolve it. Improvement to the level of non-hostile South American junta would have been adequate. That sort of limited improvement would not be worth a million lives but still would be worth some. Which brings up point 2)

2) While you are correct to emphasize the length of time required to establish a non-tyrannical government and more importantly the social foundations of a free society, you are setting the bar way too high on social changes, both in terms of what is achievable and what justifies continued military occupation. Furthermore, as other commenters have noted, it sets a much more terrifying precedent of imposing cultural values by force than does military invasion to resolve international security threats.

Japan c. 1960 was not exactly a socially progressive place. Should we have maintained the military occupation for another 20 years just to ensure women got a fair shake in the job market? Of course not. Had such a change been imposed militarily it would not necessarily be as durable, because the fact of foreign imposition would have reduced its perceived legitimacy. Legitimacy comes from self-determination and consent. Once a country has reached the point where everyone has agreed to resolve disputes within an agreed political system rather than by attempting to overturn it, then continued military occupation loses its moral justification. You're an economist-think at the margin. You reach the point of diminishing returns on continued occupation far before you achieve the full slate of social changes you want to see. Contemporary OECD-level physical security for outgroups is admirable goal, but if you only get to the level of, say, rural India, would it still be morally acceptable to maintain a military occupation? I'd say not. Paternalistic imperialism does have a place in the world, but it's not that big a place.

3) Again, a problem with your argument is any consideration of thinking at the margin. Granted, it's formulated as a typical political argument and marginal thinking is rare in those, but a steelman should be an argument that comes closest to convincing you, and you're an economist entirely comfortable with at-the-margin logic. Suppose that your tradeoffs when invading/occupying Iraq are (I'm not saying this is actually the case, just for the sake of argument) 1,000 deaths to create Yemen, 5,000 to create early-90s Colombia, 50,000 to create Pinochet's Chile, 500,000 to create postwar Japan, and 1,000,000 to create a part-Singapore/part-Switzerland/part-Dutch/part-Vermont Libertarian paradise, all with commensurately increasing lengths of occupation. Even if the Libertarisn paradise is your ideal preferred outcome it's not obvious the sacrifice in lives is justifiable past the margin of Pinochet. Obviously those numbers are all made up, but when considering a lengthy occupation you should have some sort of expectation as to what the lives/time/resulting outcome tradeoffs looks like.

Now your 1,000,000 American lives number was just as made up as mine were, and it's simply preposterous. The actual number in Iraq totaled around 5-6,000. In the 90's, Clinton was operating on casualty expectations of 10,000-20,000 when considering war with North Korea over its nuclear program. Multi-year wars against real militaries in Vietnam and Korea involved deaths on the order of 50,000. 1 million deaths would be larger than the entire active-duty Army. Using a number that large would also make hash of the (correct) argument that the initial military invasion of Iraq would be comparably easy. It's not a bad-but-in-the-ballpark bad estimate, it's bad by 2 orders of magnitude.

Now, it's not entirely clear from the context, but that 1 million number seems to imply substantial domestic civilian deaths suffered in response to retaliatory terrorist attacks. If that's the case it gets the security argument 180 degrees backwards. The security case for invading was that mass-casualty terrorist attacks were something the US was already exposed to, and that by invading Iraq we would reduce rather than increase the risks of suffering them by (see point 1) denying terrorist groups territory in which to freely organize and a potential state sponsor. A version of that argument was put forward back in 2003 under the name of "flypaper theory", positing that given a certain amount of ineradicable violent hatred of the US it was better such would-be terrorists be engaged in Iraq by professional military rather than policed domestically with all the surveillance-state methods the latter strategy would necessitate. Of course as it turned out we got the Iraq invasion, some amount of surveillance state, and a campaign of systematic drone assassinations, so unless you have access to a lot of highly classified information there's probably not a good way to delineate which (or in what proportion) gets to claim success for the lack of successful major terror attacks since 9/11. Given that the surveillance state and drone assassinations do not offer the prospect of freeing a country from a monster like Saddam Hussein it's not obvious those are morally superior even if they directly kill far fewer people.

The bottom line is that the nation-building, society-changing goal was, and properly was, secondary to the security goal. It existed and was made and believed in (including by myself), but only because the security argument for deposing Saddam Hussein militarily would create the obvious problem of "and then replace him with... what?" The nation-building goal was from both a moral and strategic perspective the proper answer to that question, but it was not the fundamental reason to invade in the first place. We didn't fight WW2 for the purpose of creating free and prosperous West Germany and postwar Japan, we did it to destroy what preceded them. Neither was the creation of a free and prosperous Iraq the purpose of the invasion (nor would it have sufficed morally had Iraq not also been an international security threat), but conditional on invasion to remove Saddam Hussein as a security threat such a secondary goal was obviously superior to the alternatives.

TMC writes:

Plucky wins this one, hands down.

Tom West writes:

Plucky, I'm struggling to see how Iraq satisfied

(2) historical use or support for terrorism as state policy

except in the most peripheral of ways (unless you mean internally terrorizing the populace).


Return to top