« "Can't Say No" Spending | Main | Congratulations to Paris Hilton »

October 18, 2007


We spend more on the military for three primary reasons. Some have a lot in common with why we spend so much on health care.

Firstly, because we can do so without suffering too much. The US is wealthier and thus spends more per capita on cars, homes, toys, sports equipment and in many other categories. We also spend more than we strictly "need" to on the military and on health care, even though marginal benefits can be considered lower the more we spend.

Secondly, because of the free rider effect. Since the start of the cold war, the US has played world policeman and alternate defender of other countries. Europe and Japan especially responded by reducing their military spending and using the US military to make up the perceived difference. Why should they pay for as much R & D and world-wide force levels if the US is going to pickup the tab? This has turned into a habit and has continued past the fall of the Soviet empire. So if you compare the relative defense expenditures of another rich nation like Japan to that of the US, of course Japan is going to be lower when the US is actually paying for most of Japan's perceived defense "needs".

Thirdly, the military spending/utility curve isn't smooth. This isn't a grapic medium and it's a little hard to illustrate without a graph, but think of military spending levels as falling into multiple utility zones. Above a certain level (US spending), you get to push anyone else you want around militarily and get most others to go along with what you want in non-military areas as well. This is the 2.5 major wars spending level that the Pentagon likes to convince congress to maintain. Below a certain level of spending, you are spending less than a reasonable estimate of the maximum forces you will face at one time. If you calculate that level properly (which is another huge debate), there isn't much more utility to be gained from additional spending, except perhaps as a means of insuring future military preparedness in the face of future spending cuts. Let's call this level "US".

Just below that spending level there is a sharp drop-off in your military influence, because you suddenly have to be much more careful in your world actions. Still, you are safe, but have much less world influence. There isn't much loss in the utility gained from your spending until you reach the level where you can no longer defend yourself from your most likely enemies. Let's call that level "EU".

There is very little additional utility in a nation spending just above the "EU" level to increase their spending unless they have decided to jump all the way to the "US" level of spending.

Dropping below that level in spending all the way down to the level where your military can't realistically defend the nation from attack, but can serve as a creditable deterrant to others (your most likely foe may win the war, but it'll cost them more than they're likely willing to pay), is a little smoother utility curve, because you can argue that a little more deterrant value is useful for a little more spending.

Aggressors are sometimes not completely rational actors, so anyone who can afford to spend at least the EU level instead of lower levels will do so unless they can get their defense in another manner. Japan's spending is really low, for example, because they rely on the US to assist them. If the US overnight dropped that commitment, Japan would embark on a massive military spending binge.

So cutting the military spending in half, while actually a good idea in terms of economic efficiency, would have to be accompanied by a policy shift that makes the US no longer provide free rider benefits to the rest of the world. At that point, instead of 1/2, it would make the most sense to cut all the way to the EU point.

Of course, an alternative proposal is that since we're providing world policeman services, we should explicitly start charging for them and make the military pay it's way more. That was the model used in the first gulf war, where we provided the bulk of the military and other nations primarily provided the cash to fund it all. If we made that arrangement more explicit and accepted, so that Japan paid X amount to us so that they didn't have to have their own military, we'd benefit from not having to spend the cash ourselves, while at the same time reaping all the policy benefits of being the most powerful military in the world. The US could provide it's "international arbitration" services along the same lines.

The main obstacle to that proposal seems to be other nation's national pride and their level of trust/distrust of the US.

1) As others above have observed, we are the de facto world's policeman, and others are free-riding on that. Also, capital is cheap, whereas in a country where the modal family has two children, life is expensive.
2) We expect to be able to put boots on the ground anywhere in the world, and rapidly. That means we have a lot of logistical tail to teeth. Many other militaries have a better ratio of teeth to tail, but fall back on us when they need to move and support troops.
3) We don't know what threats we will face in the future and have to maintain a broad spectrum capability. Many other countries have much more narrowly defined missions.
4) As Harold Brown (one-time Sect. of Defense) said in at least one speech (the one I heard), "I wish there was line in the Budget for "Waste and Fraud". It would then be easy to cut it out." Much of the waste in the military budget is mandated by Congress. Try closing a base, or cancelling a weapons system that the military doesn't want but that is built in some senator's state, and see what happens. Or try buying some vessels or planes built oversees and see how Congress reacts. All of these can be done, but it is a fight.
5) Corollary: As a former infantry lieutenant, I have a deep belief in boots on the ground, but Congress prefers big ticket items that have to built in plants in Congressional districts, not lots of cheap Soldiers.
6) Service - Congress dynamics. There was a proposal some years ago to build aircraft carriers that were half the size, and half the effectiveness of the current carriers. The Navy ran from that, fearing that rather than replacing the existing force two for one, Congress would go for a one-for-one replacement. They are both carriers, after all.

Or as Pogo once put it, "We have met the enemy, and it is us."

Sharper's long post is extremely important. I will try to say almost the same thing shorter.

The benefits of military spending depend on what your potential enemies are spending. If you have no military you must juggle your international policies so that no potential invader wants to bother. Costa rica does this well.

If an enemy can defeat you, you might still deter them but you mustn't taunt them. Presumably, the more you spend the more deterrent you get and the more independence you have.

At some point you have enough deterrent that no one seriously considers attacking you. Sharper calls this the "EU level". You're safe but nobody's afraid of you. At this point spending a little more is useless.

Only one level of spending above that is useful, what Sharper calls the "US level". When no one but you can successfully attack nations who're below the EU level, you can influence every nation you can "project power" onto -- they must persuade you not to attack them. But they can build militaries to deter you.

Sharper ignored the case of two superpowers. Arms race. Both increase spending until one can't keep up, and that one must attack now before it get worse, or else give up. Not inevitable but it's the way to bet. Nobody much wins while there are 2 superpowers. Third parties can sometimes deter one by invoking the other; sometimes that fails and they might even get invaded by both.

Sharper describes nations that trust a superpower to save them as "free riders" but I call them "client states". If they can't deter the superpower and even depend on the superpower to deter others, they are slaves -- though maybe pampered slaves. They must pay whatever tribute the master demands or else develop a deterrent force.

Not as short as I wanted but I hope it's clear.

J, Sharper, and others: Let us agree that spending at the US level gives a nation more "influence" on other nations. The question is how that could possibly be worth the cost - I want to see an itemized list, with dollar values, of the "tribute" that we get because of our extra military spending.

Robin, there's no possible way to count costs or benefits of our current military spending. If some nation agrees to an unfair trade agreement out of fear, how will we know unless the CIA tells us? How can we count the benefit compared to a fair agreement? And how can we quantify their anger about it?

Some war supporters make this argument: *Somebody* will have a US-level military. They will largely control the world, and nobody can make them stop without a world war, maybe a nuclear war, and they will start with a large advantage. Who in the world can we trust with that, except our own government?

Moving right along, the next argument goes -- if we have unquestioning dominance, everybody else should stay at an EU-level. They get nothing by having a second-best military; any time they project power we may choose to stop them. But if they try to be first then we get an arms race which is expensive for both sides and for the world.

If we're so strong that nobody else tries, we can be benevolent. Get in an arms race and the enemy decides how fast we expand our military -- and maybe we can't play Mr. Nice Guy.

And if we come in second in that race it's very bad. Lose a world war and you get no choice what happens to you.

Say we spend a lot of money to have the world's second-best navy, and we get into a serious dispute with the world's best navy. Our navy goes to the bottom of the sea and our investment is gone. Whatever it costs to have the best navy, the cost-benefit for a weaker navy is worse.

I'm somewhat swayed by these arguments. They seem to describe the world as it exists, and while we act this way the rest of the world has to go along. So that's our world. I hope we can do better for ourselves and for the world. But this is the world Thucydides lived in. Over 2 thousand years of technological improvement and the politics are exactly the same. A strong argument that we'll never change and that we have to use the same diplomacy with nukes that we used before gunpowder, or for that matter before iron. And still I hope that maybe now it can be different. It would take something new. Not just cut our military in half so China gets to arms-race levels quicker.

Count the costs and the benefits? We can't. Whatever we do, one cost is we might get genocided. Or maybe we go extinct with all the large mammals. We don't know how to minimise that. When we can't quantify that potential cost, it's trivial that we can't quantify most of the others.

g, regarding Social Security, regarding the payroll tax and the employer portion, I think that standard economic theory would tell you that, if it were eliminated, the money would go to employees in the form of higher pay. I would extrapolate that that would increase consumption, which would go directly to the bottom line of GDP, minus savings and imports.

I don't doubt that there would be redistributive effects amongst the elderly. Yes, stockholding retirees would make out way better than non-stockholding retirees. In that sense, perhaps it is slightly different than the Hansonian health care example, where pretty much everyone would have the same probability of being better or worse off (50-50, right?).

Don't forget about pensions, also. Even if retirees don't own stocks directly, many recieve pensions that benefit from a rising stock market. For example, do you think that a GM retiree would rather have a lower SS check in return for getting a GM pension? Those are the kind of tradeoffs that might be possible with a Hansonian cut in SS benefits.

Again, retirees as a group COULD very well be better off by cutting social security benefits in half. I think that that is the interesting take away. Don't get caught up in the status of the individual retiree. And keep in mind that I said "could".

Also, if I was incorrect regarding retirees being the demographic with the highest amount of wealth, they still are a very wealthy group. Anything that increases returns on investments like stocks and bonds are going to make retirees even wealthier.

Remember that Martin Feldstein projected that, if Social Security were privatized in 1980, by the mid 1990s the US economy would have been 50% larger than it was. The deadweight loss from the payroll tax is huge.

Buzzcut, I believe you have listened to propaganda about SS, propaganda that has little relation to reality.

Follow the money. When SS taxes are collected, the US government sells bonds that the SS money buys. The US government then spends the money. So, what would happen if we stopped collecting SS? The US government would either have to cut back its spending a whole lot, or else it would have to issue regular bonds to pay for the spending, or it would have to raise taxes. You are looking only at the first case. If government spending was cut back with no reduction in services, then everybody would be happy. But if the government issued more bonds to replace the SS money then the result would be about the same as we have now, except that we would have to persuade people to buy the bonds instead of just tax them. If it was new taxes then it would particularly hit the people who pay the new taxes who might not be the same as the ones who pay SS now.

People talk like SS is in financial trouble. It isn't particularly, except for the problem that the US government might choose to default on its SS bonds. The government has already spent the money and might choose not to pay it back. That's the problem.

And if we particularly cared about SS, the obvious solution is to instead buy bonds from countries whose credit is better than the USA, countries that we can be confident will pay their debts. And we could require the USA to start paying back existing SS bonds so we can invest the money in safer places. But the issue isn't how to keep SS safe, the issue is how to fund the US government this year.

Marty Feldstein was, basicly, lying. If SS had been privateered in 1980 and if government spending had been cut back proportionately with no reduction in services, then yes, the US economy would probably have been 50% larger by the mid 1990s than it was. And if Marty Feldstein had spent those years working three fulltime jobs each of which paid as well as his current job, he would have been more than twice as rich as he was not counting non-labor income, kickbacks, bribes, etc.

Those are some pretty strong words. Lying? And you know that based on what, exactly?

Feldstein could be wrong. But lying? I don't think so.

Feldstein's paper is available at NBER.

Correct me if I'm wrong, but the SS "surplus" is not anywhere near 1/2 of SS revenues. Thus, if we cut those revenues in half, with a corresponding cut in benefits, we'd still be ahead of the game, economy wise.

J and Buzz, this is not the place to argue about social security.

Sorry. But you could end the argument with your authority. Would a cut in SS benefits be Hansonian?

Who cares whether it would be "Hansonian"? What matters is whether its effects would be good. Robin's a very clever chap, but I don't see any reason to think he can pronounce authoritatively on that question. (And I bet he doesn't either.)

I care.

Buzz, state pension plans are a combination of transfer and forced saving. The forced saving has little long term effect, as people can compensate via other savings choices. The transfer is hard to evaluate, as it is not clear what would measure success for them.


I think you're missing part of the point. You've got the cart before the horse.

If you take a typical American two car family, you can easily make an economic/financial case that cutting their car costs in half will be of major economic benefit, especially over the long run. You can also point out that it's not as difficult as they might think to only be able to use one car and perhaps rent a car for specific situations that absolutely require a second car, such as one driver needing to drive out of town for a long trip.

However, you can't start with "What if we cut your car expenses in half?" and then expect them to see the "obvious" benefits to losing the second car. You have to start by convincing them that they're fine with one car. Once they're ok with that idea, you can easily convince them to cut their car expenses in half, because at that point it will make sense to them.

In a similar fashion, you'll have to convince the American people that it's just fine for the US to not be in control of what happens militarily outside of North America. Convince them to not demand we do something the next time some dictatorship invades someone else. Convince them that it's ok if China, or Iran, or North Korea, takes over whichever neighbors it wants to. Once you've accomplished that task, it becomes a simple matter to cut the US defense budget down to the EU level.

The reason for current US military spending isn't just economic and defense benefits for the US, although those do exist. The reason is what US voters expect the US to do with it's military, which is a much larger role. Change the expectations and the budget can be easily cut.

You can't change the expectations of the voters by cutting the budget in half. You have to cut the budget in half by changing the expectations of the voters. That task is of course complicated by the possibility that the voters may prefer whatever you want to call the utility they get when the US is a lone superpower to the utility of the smaller federal budget.

For an extensive and detailed analysis of where we are and some of the near-future possibilities in US military strength and relations from a world economic and social standpoint, I'd suggest "The Shield of Achilles", by Philip Bobbitt.

And for the record, there are two adults and four children in my household and we do fine with one car in a semi-rural area. :)

I have posted Greg Cochran's comments http://entitledtoanopinion.wordpress.com/2007/11/07/greg-cochran-on-military-spending/here.

Whoops, sorry about that. It's here.

A lot of unnecessary spending in the military: bands, ceremonies, General's secretaries, government employees hired to do nothing except take a pay check. Sure not a lot of money however if added up I bet it could make a difference - how can we get people in power to take a look at this?

The military is also a place where more boundedly rational people can have simpler environments in which to operate, in a society that doesn't seem to much fund those programs otherwise (the other spaces are prisons and mental institutions). Something to keep in mind when halving military spending: we may have to do something with that less productive have other than allowing them to increase social risk as they operate unmediated in the civilian population.

In similar vein, I think the military's strongly deviant benefits and job security might be capitalists buying out violent but economically unintelligent people, in the similar way that highly paid legal work for law firms mostly to protect against other lawyers might be capitalists buying out excellent arguers/propagandists who also have limited economic aptitude or interest. May also be a factor to be considered when making adjustments to size of the military population or its benefits.


The military is actually reasonably selective, and while it does recruit substantially from the bottom half (although for only a minority of its headcount) entrance requirements largely exclude the bottom third, so it's not strongly competing with prisons and mental institutions.

To be clear, I didn't mean that the military recruits substantially from the less productive half of the American population, I meant that presumably it would be the lesss productive half of the US military that would be fired/attritioned out of the military in Robin's proposal.

Also, I don't think I reasonably implied that the military strongly competes with prisons and mental institutions for its recruits. Quite the contrary, I mentioned the latter two to illustrate how stark the options are for more boundedly rational citizens: folks who have trouble maintaining the economic planning and foresight necessary to maintain an apartment, save for retirement, or have a good work performance if they face penalties no more severe than fire-at-will. I suspect we'd have a consequentially better society with a stronger (but opt-outable) paternalist structure. The military currently seems to me to be the least extremely limiting/stigmatizing of the three public sets of institutions that provide relatively strong paternalism for adults that need it.

The Cato Institute calls for a 50% cut to the U.S. military:

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30