One Surge Glory

CBS News has another doom+gloom Iraq poll, but some of the trendlines are surprising. For eleven months, CBS has been asking this question:

From what you have seen or heard about the situation in Iraq, what should the United States do now? Should the U.S. increase the number of U.S. troops in Iraq, keep the same number of U.S. troops in Iraq as there are now, decrease the number of U.S. troops in Iraq, or remove all its troops from Iraq?

Eleven months ago, only four percent of people said "increase troops." Thirty-six wanted to keep troops at the 120,000 level. Twenty-eight percent wanted to decrease troop numbers, 29 percent wanted to start pulling them out.

What's happened since then? Well, the war has gotten more and more unpopular. Disapproval of Bush's policy was in the low 60s, and now it's in the high 60s. Concurrently, support for a troop increase was going up. In the last poll before the midterm, the numbers were: 16 percent for increase, 27 for status quo, 26 for decrease, 24 for withdrawal. In the latest poll, the curve has inverted from that original poll. Now "increase" is the second most popular option for Iraq at 26 percent. "Remove all" is at 28, basically unchanged from a year ago. "Keep the same" and "decrease" have plunged, and all those fence-sitters have become pro-increase.

So as Iraq continued its steady political and strategic implosion, more and more people (but to fair, a maximum of 1/4) decided it was time to double down. W, as the kids say, TF? One answer's in this Foreign Policy article by Daniel Kahneman and Jonathan Renshon.

Imagine, for example, the choice between:

Option A: A sure loss of $890

Option B: A 90 percent chance to lose $1,000 and a 10 percent chance to lose nothing.

In this situation, a large majority of decision makers will prefer the gamble in Option B, even though the other choice is statistically superior. People prefer to avoid a certain loss in favor of a potential loss, even if they risk losing significantly more. When things are going badly in a conflict, the aversion to cutting one's losses, often compounded by wishful thinking, is likely to dominate the calculus of the losing side. This brew of psychological factors tends to cause conflicts to endure long beyond the point where a reasonable observer would see the outcome as a near certainty. Many other factors pull in the same direction, notably the fact that for the leaders who have led their nation to the brink of defeat, the consequences of giving up will usually not be worse if the conflict is prolonged, even if they are worse for the citizens they lead.

The article is about leaders, not voters, but the argument applies. The problem for out-of-Iraq thinkers/politicians is to convince voters that the choice is between a sure loss of $890 and a sure loss of $10,000, no chance of winning. Oh, and to convince the voters that they should never trust this croupier again.

Comments

[JasonL]

Give it a rest, Weigel. What these studies reveal is that people think the situation sucks now but it is not obviously unwinnable.

People don't act like the answer is obvious because the answer isn't obvious. No matter what side of this you are on, you should at least recognize that much.

[Andy]

I think the money metaphor is all the more powerful when one considers the full decision-making process for the average policy-maker in Washington. A necessity for a career in policy is the belief that you can provide the right answers and, as a result, many people stubbornly believe that it's not only a 10 percent chance of losing nothing. That the troop increase, when coupled with "doing the right thing" (even if they're not 100 percent sure what this mythical change of policy is) will solve things.

In other words, if you can keep the problem on the table long enough, someone will be able to find a solution. After all, that's why they're in a position of such power, right? They find the solutions. So, I think, in a lot of calculations, the long-term chance of success feels much higher than any realistic assessment would warrant.

[Fluffy]

I'd like to be further educated on the statistical example, because I still think B sounds better.

I've already lost $890, if that's my other choice. I have a 9 in 10 chance to lose another $110, but a 1 in 10 chance to win back $1000. Basically you're selling me a $1000 lottery ticket with a 10% win chance for $110. If the ticket cost $100, it would be fairly priced for its level of risk, so I'm only really statistically at risk for the additional $10. That sounds like it's worth taking the chance, unless I'm VERY risk-averse.

[JasonL]

To follow up on Andy's insightful comments. The hang around approach has appeal for the politician because there are few cases of unmitigated disaster or unmitigated victory and it is even rarer that your actions are the sole relevant ones.

In this case, I think there is a sense that if we hang around, will the Iraqis figure something out they can live with? It just isn't obviously a wrong thought.

[tros]

I think the gambling metaphor is very apt. It's just sad that they see the war as something that needs to be fixed instead of ended. Seriously, what on Earth is going to make us benefit from this? It's like they are flooring the gas pedal of the meat grinder one more time before they shut it off.

I am really curious, though, are there any possibly tangible benefits that could be acheived besides controlling the entire Middle East out of Iraq (ROFL)?

[Timothy]

Fluffy: You're misunderstanding. Given the choice of A or B, which do you choose? A lot of people pick B, because they might lose nothing, despite the EV being $10 worse (.9x1000=900). However, I'm not sure that's all that germane to the point in Iraq.

I think voters are suffering from the Sunk Cost Fallacy. They've already lost X, and they seem to think that X matters when it comes to deciding whether or not to continue. What's retarded about that whole thing is that X is already gone and unrecoverable, so the real question is whether the "surge" is a better alternative at the margin than the other choices. I'm going to go with "no", but getting other people to understand that would require them grasping simple economic concepts. I'm not holding my breath.

[black_box]

It is not at all surprising that more people would support a surge now than they would many months ago, even with the war's growing unpopularity. Now that GWB has declared the surge to be his chosen strategy some people will naturally be influenced by him and that is what is reflected in the polls.

[Timothy]

Let me clarify:

EV (A) = -$890
EV (B) = -$900 (.9*-$1,000 = -$900)

So option B is statistically worse, but people still tend to choose it because they have a small chance of success.

[thoreau]

I'm quite aware of expectation values, but I also consider the marginal utility of each scenario. Relative to the guarantee of losing $890, losing another $110 is not the end of the world in my current situation. OTOH, avoiding the loss of $890 is something worth going for, if the worst case scenario is only an extra $110.

It's not about wanting to take the chance, it's more about the marginal costs and benefits of each scenario relative to the "baseline" scenario of a guaranteed loss.

For instance, suppose somebody offers me the choice between $10 guaranteed or a 1 in 1000 shot at $5000. I'll take the chance, even though the expectation value of the second case is $5, because an extra $10 won't do much for me. (If I were homeless, of course, my choice would be different.)

OTOH, if somebody offered me a guarantee of $2000, or a 1 in 1000 shot at $4 million, I'll take the $2000, even though the expectation value of the second scenario is $4000. $2000 could pay off some bills for me, and put me in a significantly better situation.

I realize this is far afield from the original context of the surge, but I don't care.

[Andy]

Tros -

In an ideal world there are a HUGE range of benefits that could be gained from total success in Iraq. If you are willing to fiat both the solution for the violence AND election of a moderate democratic government, there are a galaxy of potential benefits. On the less tangible side, you would have a model for democracy in the Middle East, a moderating influence for global Muslim opinion, and a success story for America in a region that needs one desperately. More tangibly, you'd have a counterbalance to Iranian regional hegemony, a reliable source of oil, and an American base for operations (military, diplomatic, covert, etc.) in the Middle East. Of course, that chain of events was unlikely when we got involved, let alone now.

So we're left more with avoiding the negative: Iraq was serving, at least to some extent, as a counterbalance to Iran before we got involved, so if we could at least allow it to gain a level of some strength and stability it could hopefully aspire to that role again soon in the future. Also, Saddam feared the influence of radical Muslim theology, so Iraq was (ironically) one of the few places terrorists were totally unwelcome, whereas if we leave it in chaos it would almost certainly become a breeding ground that will make Afghanistan look like the poppy field in the Wizard of Oz. Finally, total collapse basically destroys American credibility in the Middle East, both as a military power and a reliable ally. It could very well leave the same sort of legacy of fear that American failure to act during the Hungarian Revolution left in the Cold War, leaving states unwilling to confront radicals and terrorists because of uncertainty of American commitment to supporting such moves.

Of course there is, in my opinion, almost zero chance of the current course of action accomplishing any of the above. But don't think that there's NO policy impetus for wanting a stable, or stableR, Iraq. Or even just for trying to delay the inevitable collapse jsut a few months.

(And on a side note, I think blackbox has the answer here, this data reflects three major groups: people who hadn't considered the option before, but have heard enough about it in the in the press to be willing to give it a chance, people who still are totally unwilling to admit Bush could be wrong about anything, and (I would guess by far the smallest group) people who support a troop increase (though that probably means something very different than "the surge" as we know it) as a reasonable policy alternative.) I would guess that most of the third group was in the four percent previously, and new additions are slim enough so as to be irrelevant. So you're basically just looking at those still drinking the kool-aid (a distressingly high number to be sure) and those who hadn't even considered it before, and figure we oughta give it a chance before we give up on the whole thing (which I think is where Timothy's Sunk Cost Fallacy comes heavily into play. I would wager many of them don't really care how great the chances of success are, they just hate to see all the effort so far wasted without trying a few more things.)

[Ironchef]

So much criticism against the running of the war has been that there were too few troops - especially early on.

Maybe some of the surge supporters are just people who held that view.

[Timothy]

I suppose there's something to be said for the marginal utility aspect, I hadn't really thought about it.

I wonder how this game would look if you imposed any old diminishing marginal utility of wealth. That is, any old utility of wealth function with a downward sloping derivative...that sounds a lot more pompous than I wanted it to, but there it is. It might come out that choosing the sure loss doesn't look better from an Expected Utility perspective.

Anyway, to talk about the surge again for a second, I really do think the issue is more one of using sunk costs in the calculation than one of making irrational decisions regarding the EV.

[John]

Maybe people just want to win and have looked at the situation and concluded that the cost of loosing is greater than the cost of trying to win? Even if they think the war was a mistake, it doesn't necessarily follow that they want to pull out. If people think that going forward the costs of loosing are higher then that of staying, then it is perfectly rational for them to support a troop increase regardless of their opinion of the war in general. Maybe they are wrong in thinking that and Weigel is certainly free to tell the world why he thinks that. Of course he doesn't and instead just assumes he is right and everyone else is stupid.

The assumption behind the post, like all Weigel posts is that trying to win is pointless and that the costs of quitting could never outweigh the costs of continuing. I am not really sure what the point of this post was beyond Weigel having the opportunity to show everyone his deeply held beliefs and how he thinks anyone who disagrees with him is completely unreasonable. Good for you Dave.

[semm]

Weigel, you just arent 'getting it.' Most people don't think Iraq is hopeless, they just want more effective leadership in getting the situation under control. There is nothing inconsisant about this.

I know you think nothing can be done to improve the situation there, but you happen to be in the minority on that.

Not that I think strategy-wise the best way to go is polling (laugh), but there you have it.

[STEPHEN THE GOLDBERGER]

The problem with the option A and B choice is that "winning" the war in Iraq and creating a stable democracy there is not "losing nothing" but supposedly the wonderful dream we're shooting for. The idea is that we won't just break even there.

At least I hope that's the goal, but now it's probably becoming just to get out while saving face.

In any case I hate expected Utility arguments because it's usually done using a false metric for payoffs (i.e. every dollar has a different value to every person, one's millionth dollar means far less than one's only dollar, etc.)

[Tony]

The popular choice in the game where you face a sure vs. probable loss only seems irrational if you ignore the diminishing marginal value of money.

In the extreme case, if all you have is $1200 for the whole year, and you will starve to death if even $800 of that is taken away, of COURSE you will choose the 10 percent chance of survival. Going for the maximum expectation value is a game for wealthy people whose lives are not substantially effected by it.

The application of that principle to Iraq is left as an exercise to the reader.

[Andy]

I think the marginal utility HAS to be a factor at some point, given the chance between being given 10 million dollars and having a 25% chance to win 100 million, I think almost anyone would take the 10 million, even though the EV is massively higher. With 10 million, you can retire, build a house, buy a boat, etc..

But looking at things from an economist point of view, I don't think that matters. You have to be concerned solely with which option is really, mathematically better. This is especially important from a policy perspective. Any given person can override the "right" answer to decide that it exceeds their tolerance for risk, or that the marginal utility just isn't there, but when dealing with infinite decisions for infinite people, you have to keep making the smart bet to win, if that makes any sense.

That, I think, is the central issue here: it's not the policy makers are the house, they're the gambler who's making a decision with all of our pooled money. They're overruling common sense on a hunch (because personal pride and the like is involved) - basically, they've walked up to the roulette table and put it all on 33 black. We can debate all day whether that's personally the choice we'd make, but from a mathematical standpoint it's a terrible decision. I think that's the central issue here.

Basically, the White House is Lucas, and we're the cast of Empire Records. I can only hope this means that the endgame of the war in Iraq is Liv Tyler in her underwear, or at the very least a rooftop duet with Coyote Shivers.

[thoreau]

In any case I hate expected Utility arguments because it's usually done using a false metric for payoffs (i.e. every dollar has a different value to every person, one's millionth dollar means far less than one's only dollar, etc.)

Exactly.

[Lamar]

"Weigel, you just arent 'getting it.' Most people don't think Iraq is hopeless,"

Isn't the point the idea that Americans will gamble on a loser until there is no hope left, i.e., make poor choices because there is a lottery's chance at getting back to zero? I am against the war, and I can't rule out a successful surge. I wouldn't bet 3 cents on it, though. But I hope it works.

[John]

"I wouldn't bet 3 cents on it, though. But I hope it works."

Good for you Lamar. I am not sure how much I would bet either. Moreover, even if it does work it probably won't be becuase of the surge but more because of other long term trends finally coming to fruition in time for the surge to get credit. There are never short easy sollutions to problems like this and when the sollution does come it is ussually when people least expect it and for reasons other than what they think.

[Andy P]

we're all stupid little monkeys when it comes to thoreau's baseline option (baseline fallacy is more apt). The reason why people want a surge is because it was promoted to "baseline" option status.

As Bryan Caplan points out at econlog "war raises the social ostracism costs of pacifism." Once Bush upted the ante to surge, the baseline of ostracism was moved.

[torrentprime]

"Now that GWB has declared the surge to be his chosen strategy some people will naturally be influenced by him and that is what is reflected in the polls."

Exactly. We need to remember the millions of people (and this applies to both sides of the aisle) that take their talking points from various sources in the media, blogs, the news, etc. along partisan lines. When the Bush critics were talking about how Bush screwed up by not sending enough troops in the first invasion and post-"victory" occupation, the reflexive red-state move was to not only say that no more troops were needed, but to categorically deny that any more even should or could go. Couple that with anti-war activists who wanted (want) the troops home yesterday, and you end up with a very small number of people who believed in more troops.
Now that Bush and the right-wing news/info machine is saying "surge it, baby" all of a sudden millions of people believe that 21,000/48,000 troops are precisely what is needed. Couple that with the "America loves a winner" and the whole miraculous come-from-behind victory thing, and it's no wonder that that the simple concept of "send in the cavalry" is now so highly rated.

[Herb Schaffler]

This reminds me of the person who at the midway throws darts to break baloons to win a teddy bear. Let's say the teddy bear is worth $10. The more the person loses in trying to win the bear, the more he keeps at it. He can't bear the thought of having thrown away $9 to win a $10 bear so he keeps trying. He finally wins the bear after spending $20. Was the cost worth it?

Bush has too much invested in the Iraq war to give up, otherwise those 3000 deaths of American soldiers and $350 billion dollars will have been wasted. So he keeps trying something new like the surge in hopes it might work. The cost may be to great like $20 for a $10 teddy bear. The Shiites will join in an alliance with Iran. The war will result in more terrorism against the US, endangering American citizens. And don't forget about the innocent Iraqis who were killed. Was that worth it? One estimate is as large as 900,000.

[Timothy]

Andy: You've made some very good points to chew over. Do you have a link to Bryan's post about social costs? I haven't been keeping up with econlog.

[The Real Bill]

Caplan's post is still on the main page. It's titled, "Asymmetric Sell-Outs."

[Dan T.]

The most puzzling thing about the discourse on the Iraqi war is that oil is almost never mentioned at all, much less as the number one factor behind our Middle East policy.

It's becoming clear to me that Americans are really in denial about the true price of our lifestyles. 9/11 should have awoken us a little bit but it doesn't appear that it did.

[ChrisO]

One major problem with the gambling analogy: The gamble has a distinct potential payoff. The only "payoff" possible in Iraq (as in Vietnam) is the absence of a negative. In other words, the only possible positive outcome in Iraq is that nothing bad will happen over a long period of time. You could never find a distinct point in time at which things will definitely be resolved.

I think this is reflected in the uncertainty in public opinion over the war and moves the U.S. govt should make. There's no tangible reward, but simply a continued slog.

The administration would be on better footing politically if it indicated that this surge was a prelude to a withdrawal, even though such an announcement would harm the effectiveness (if any) of the surge.

[ChrisO]

The most puzzling thing about the discourse on the Iraqi war is that oil is almost never mentioned at all, much less as the number one factor behind our Middle East policy.

Because oil clearly was NOT the point. If it had been, our actions would have been much different. The "green zone" would not be downtown Baghdad, but the primary oil facilities. Assuming rational actors, the best course to secure the oil would have been to quickly install a Shiite strongman in Baghdad, since most of the oil sites and facilities are in Shiite territories.

[bubba]

So option B is statistically worse, but people still tend to choose it because they have a small chance of success.

It's the difference between minimizing losses, and striving for success.

Or the difference between being French and American.

It also fails to account for the impact your decision has on observers. Fighting the school bully rarely ends well that day, but it is still the recommended course of action, at least in the US.

[bubba]

Because oil clearly was NOT the point. If it had been, our actions would have been much different. The "green zone" would not be downtown Baghdad, but the primary oil facilities. Assuming rational actors, the best course to secure the oil would have been to quickly install a Shiite strongman in Baghdad, since most of the oil sites and facilities are in Shiite territories.

More like we'd have cut a deal with Saddam, ending sanctions in exchange for guaranteed access to oil. Like the French.

[joe]

Herr Bubba, we need to get our troops out of Stalingrad now, or disaster!

Why, so we can look like a bunch of sissies?

[Dan T.]

Because oil clearly was NOT the point. If it had been, our actions would have been much different. The "green zone" would not be downtown Baghdad, but the primary oil facilities. Assuming rational actors, the best course to secure the oil would have been to quickly install a Shiite strongman in Baghdad, since most of the oil sites and facilities are in Shiite territories.

Oil is the point of our entire Middle East policy, and the theory is that a stable, US-friendly Iraq was a good way to stabilize the region. In the long run, that makes more sense than installing a dictator who might become our enemy tomorrow as they so often do. BTW, the Green Zone is in Baghdad because that's where the most modern infrastructure was - we certainly don't have the resources to construct an entirely new capital city!

[John]

"Herr Bubba, we need to get our troops out of Stalingrad now, or disaster!"

Yeah Joe because there is such a danger of 100s of thousands of American troops surrendering and be lead off to captivity. Let's at least be honest about the costs.

[TrickyVic]

"""So much criticism against the running of the war has been that there were too few troops - especially early on.

Maybe some of the surge supporters are just people who held that view."""

I disagree, maybe some, but for the most part, those who were critical of too few troops from the start, do not support the surge because it's still too few troops.

The biggest problem we have is that we are not in control, therefore it is up to the Iraqis to "fix" the problem. It is obvious that many Iraqis do not want the sectarian fighting stopped, they endorse it. Where does that leave us?

My prediction has been that the surge will be useless because the bad guys will leave until a better time arrives, and/or take their fight outside of Baghdad. I'm seeing some evidence of this. We did a major sweep in an area that intel said was loaded with bomb makers and found 6 AK-47s and no insurgence.

The nutty adventures of keystone cops

[josephdietrich]

Even if we could pull this out and make this somehow work (i.e., ponies for all Iraqis who, because of the steely resolve of the extra GIs, have decided to lay down arms and live in peace), would it be worth the $1-2 trillion dollars we we're going drop on it? I say not.

Invading Iraq was a waste of our tax dollars. Rebuilding Iraq is a waste of our tax dollars.

By drawing this out, we are just wasting more money. Even if it works.

[Herb Schaffler]

"Oil is the point of our entire Middle East policy"

What about Israel?

[joe]

Good point from Dan T.

When people talk about this being a war for oil, they're not suggesting that we invaded so that we could bring their crude back to the states as war booty.

They're tallking about having friendly governments in place, so that our oil companies and consumers can enjoy higher profits and lower costs by avoiding instability and hostile actions.

When we did "regime change" in Hondoras on behalf of United Fruit, we didn't have our troops harvest the bananas and ship them back, so we could sell them at a profit.

[Gasping Libertarian]

What does ANY of this have to do with Libertarians? This site has become so predictably dull.

[Andy]

Really? You can't see how a debate about escalating funding for a war of choice would be relevant to a libertarian discussion group?

[bubba]

What about Israel?

If it weren't for oil, we'd have let Israel play to win.

So, I'd wager that oil the cause of our Israel problem, not the other way around.

[Herb Schaffler]

"So, I'd wager that oil the cause of our Israel problem"

The cause of our Israel problem is AIPAC. Politicians in both parties are too tied to AIPAC.

[clone12]

Let's go long, with someone elses' lives.

What could go wrong?

[Windypundit]

If you're in a boxing match, you either try to win or you get out of the ring. Either one is better than just taking your opponent's punches, which seems like what we've been doing in Iraq lately.

If, like many, you don't think Iraq is winnable, then only withdrawl makes sense. But if you think Iraq is winnable, either withdrawing or the surge (if it's a winning move) is better than what we've been doing.

[Matt]

Weigel: "The voters aren't saying what I think they should say. That means they're obviously irrational!" Maybe they're victims of false consciousness as well..... The Kossification of Mr. Weigel continues apace.

[Genghis Kahn]

I remember somebody writing an article once, years ago, about Isreal and Palestine. There was supposedly this survey taken, and the outcome was that if the Palestinians were given enough money for the loss of their homes and homeland, they'd be happy enough (on an individual basis) giving up the whole fight and moving somewhere else in the ME.

There was a pretty persuasive argument that even if the Palestinians wanted an unreasonably large lump sum payment, it still would have been much cheaper in the long run to just "buy them out".

Of course, there was never anyone willing to put up enough money to back such an idea. And such an idea would never let Hama and Fatah keep their King-of-the-Hill acts. So it never went anywhere.

But I've wondered more than once if such an approach might work in Iraq. Conclusion: not likely.

Iraq will not stablize until a bunch of people kill a bunch of other people. When one group breaks the backbone of the other, then things may settle down.

If we're serious about settling things down in Iraq then we should (my opinion) take one of three approaches:

1) up the troop level to at least half a million, and commit to stay until we've literally pacified the place. Like, Roman style they way they pacified Iberia (several times over).

2) Look for some prospect of a "buy out" solution. One that we could live with. Deal is, if the Iraqis take the money and then keep this piss-ant contest going, the US gets to kick serious ass and civilians be damned. Everybody knows going in that this is the deal.

The US would never follow through with this. The Iraqis probably wouldn't either.

c) Figure out what kind of gasoline we've got to pour on this fire, and torch the place. I mean get them fighting now, and hard. Let them be the ones who make a disaster of their homeland.

The US, meanwhile, finds some way to sit on the sidelines and wait it out. All we do is pour a little more gas on their if seems the flames are dying down to quickly.

You have to understand that in their culture, a once-and-for-all duke it out war is not the norm. They're used to 300 year pissing contests (with or without the wind).

Let them -- indeed encourage them -- to beat each other senseless. When they're done, we'll find it much easier to broker a "solution" to achieving a stable country.

We'd pay hell trying this last option. And I'd get called an idiot by Wiegel three times a day. But if I was betting on a solution that was going to work, I'd put more stock in encouraging the big "ethnic cleansing" they seem to want.

Of course we'd have to be Machiavellian about it. You don't TELL them what you're up to. You have to let them think it's all their idea anyway.

We need to let them know that if they start shooting each other, we aren't going to stop them.

Ugh. This is not a good option either. I'm just trying to poke around and look for some other way to crack this nut.

I give up. Bring them home.

[Genghis Kahn]

Andy,

I can only hope this means that the endgame of the war in Iraq is Liv Tyler in her underwear

If you think this is a possible outcome, you might convince a whole lotta people to support a much bigger surge.

[jb]

Genghis: The problem wasn't the Palestinian willingness to move--after 1948 and 1967, the majority of them moved. The problem was that the other Arab states made them second-class citizens, so they didn't assimilate and instead created a national consciousness. Even now most Palestinians would move for free, if some place would take them. The West Bank and Gaza are s***holes, and the inhabitants know that--the problem is getting somewhere else to treat them like citizens.