The military failure machine — Crooked Timber

Nicholas Kristof has a column in the NYT putting forward the heretical idea that the US should spend less on the military and more on diplomacy and education. The argument is obviously right as far as it goes, but it leaves one big question unasked. An obvious reason for the focus on military spending is that Americans have massive confidence in their military and much less in their education system, particularly the public school systems.

Yet judged by results, the opposite should surely be the case. Why is this so?

The US military has fought five large-scale wars in the past fifty years, resulting in a draw in Korea[1], a defeat in Vietnam, and three inconclusive outcomes in Iraq (twice) and Afghanistan. That’s a record that makes the worst inner-city public school look pretty good. At least the majority of students, even at the worst schools, end up more or less literate.

The US military does an excellent job in defeating anyone silly enough to put a conventional army in the field against it. But, as a result there aren’t many adversaries so silly (even Saddam didn’t expect war when he invaded Kuwait and did his best to avoid it in 2002-03). Potential opponents either try to acquire nukes or fight with IEDs and suicide bombers.

Kristof is right that even where the use of military power is successful in its own terms, it is unlikely to be cost-effective – his striking observation on this is that the cost of one US soldier in Afghanistan is the same as that of 20 schools. Similarly, Greg Mortensen observes that sending back 243 troops would be enough to finance the entire Afghan higher education system [2].

But the striking thing about military expenditure is that its failure rate is so high. More or less by definition, it’s impossible for both sides to win an armed conflict, but it’s certainly possible (and probably the par outcome) for both sides to lose. So, the US success rate since 1950 is probably about what would be expected. As I’ve mentioned previously, US experience of war (apart from the Civil War) before 1950 was by contrast exceptionally favorable – even the War of 1812 was claimed as a win

Moreover, in all sorts of respects the self-image of the US (as a land of opportunity and social mobility, a generous giver of foreign aid, a beacon of democracy in a generally undemocratic world and so on) seems in most respects to have been set in concrete by 1950. The failure to learn anything from a string of military failures and disappointments seems to fit with this.

I’m talking here mostly about the views of the American public, but these views are even more predominant among the policy elite and the Foreign Policy Community. I don’t think this is primarily because either the elite or the capitalist class they might be regarded as representing benefit from wars. It’s true that there is not much of a penalty for advocating disastrous wars, but as long as you steer clear of a handful of topics, there is not much of a penalty for anything in the US policy elite, once you are regarded as “serious”. And while some businesses obviously benefit from, and lobby for, war, there are plenty more who would prefer to make money trading with putative enemies like Iran and Iraq.

At least, the majority of Americans regard the Iraq and Afghan wars as mistakes where the costs have outweighed the benefits. If that (correct) judgement could be generalised into a recognition that military force rarely generates unequivocal victory, and is rarely worth the cost even when it does, arguments like those of Kristof might begin to prevail.

fn1. In fact, it would probably be more accurate to break the Korean War into two parts: a brief and victorious defensive war in 1950 in which the North’s invading army was thrown back across the border, and a counter-invasion of the North which resulted in a disastrous defeat, and three years of bloody struggle ending in the status quo ante. October 1950 marks the point when US military policy (at least as regards large-scale international conflicts) shifted from reluctant involvement in wars started by others to an increasing preference for pre-emptive military action.

fn2. I think this is an overestimate. Mortensen is estimating the cost of keeping a US soldier in the field at $1 million a year, but taking account of support costs and deferred costs, it’s probably closer to $5 million, which implies that withdrawing a single platoon would be enough.

37 thoughts on “The military failure machine — Crooked Timber

  1. John Quiggin :
    PML, a couple of brief points. I agree that the 1812 War (sic) was in reality a loss for both sides – I was merely pointing out that, since the Americans believed they had won, it added support to the idea that the US was naturally victorious.

    Please do not imply that I agree with a position that I did not state and do not in fact hold. The British not only won that war in the narrow technical sense of totally preventing any rewards to aggression, through a curious and fortuitous concatenation of circumstances it actually improved their situation as compared with what would otherwise have obtained: regiments were not disbanded according to usual practice with the apparent end of the Napoleonic Wars, but were on hand for the Waterloo Campaign (some troop ships made it back just in time). I will agree that this is unusual – but I most emphatically do not agree that that particular episode made Britain worse off.

    On your “not universal” point, that is why I used the word “rarely” which allows for (fairly trivial) exceptions such as those you’ve noted.

    Are you suggesting that you will insist that counter-examples are rare unless you are surfeited with them? But blog replies do not allow that.

    And the Goa example, which you’ve used before IIRC, is not a very good one. If India had done nothing, the Portuguese would have pulled out in 1975 anyway, so the (almost bloodless) takeover only brought forward the inevitable by a decade or so. Had there been any serious risk of Portuguese resistance, an invasion would have had costs that exceeded the benefits. Since there wasn’t, it wasn’t really a war at all.

    Again you (taking the charitable view) misunderstand. I went to some trouble to provide a whole family of counter-examples, and here you address just one. And of course that particular one was cost-effective (and, over time, would have been even if the Portuguese commander had obeyed his orders to fight to the bitter end). But that’s the whole point – it’s a better way of doing things, and gives wars that pay. Claiming that cost-effective military operations aren’t “really” wars is a No True Scotsman argument.

  2. John Quiggin :
    As regards Kashmir, if you think that either Pakistan or India has gained from their actions their, relative to unilaterally conceding the demands of the other side, I have some PowerGen contracts that you might want to buy.

    India is definitely better off, as compared with that alternative; it is likely that its hold on its provinces would be weaker (the same fear that Indonesia had about letting go of East Timor – with some justification), and it would certainly have been harder to penetrate the Himalayan Kingdoms. But I presented a different alternative, what would have happened if India had allowed a border closer than Kashmir. There would still have been just as much cost from just as troublesome a border, only with less hinterland to support it.

    For Pakistan, though I did not claim that as cost effective (it isn’t), it is actually a matter of long term survival. I have sometimes encountered Indian assertions that “there is no place called Pakistan”, very irredentist. Accepting long term absorption would be cost effective for the absorbed, of course, but that’s not their calculus.

    Try some game theory. The actual win-win option, which neither side was willing to try (possibly for fear of betrayal) was an independent buffer state – what the Maharajah of Kashmir had leaned towards, until the Congress politicians leaned back.

  3. @Peter T
    Peter T asks “whats changed”

    Well maybe modern wars dont really get to appropriate so much land as before 1900? The Isrealis seem to be the only ones getting away with it (and its not much)…but Im not a war expert at all – it just seems there is not much room for expansion although you could have said the same when Hitler expanded across Europe but the Germans didnt get to keep much of that expansion in the end.

  4. Peter T :
    I think Prof Q is probably right about the relative costs and benefits of war since roughly 1945. But P M Lawrence is right that up to 1900 or so, a country could gain a great deal through successful war. It is, after all, how most empires are built, how Britain came to have 70% of the world’s merchant fleet, and how impoverished cottagers from the British Isles were able to make new and much more prosperous lives for themselves and their descendants in lands taken from others.
    So the interesting question is – what’s changed?

    Just some features of the current state of the Art of War and of available resources. The important thing is to realise that there could just as easily be a shift in another direction just as rapid, profound and consequential as that between (roughly) 1895 and 1914 – if the First World War had happened even ten years earlier, Germany couldn’t have made nitrates and wouldn’t have lasted a year. Shifts like that have happened before, e.g. at the beginning (heavy armoured horsemen, castles) and end (Swiss pikemen and their tactics, guns) end of the age of chivalry; it could never be anticipated, and so could happen again.

  5. “how Britain came to have 70% of the world’s merchant fleet, ”

    Primarily, by being good and shipbuilding and running capital markets. But they were helped by the fact that much of the US fleet was destroyed during the civil war, which rather supports my point.

    PML, you seem to be having some methodological problems. A list of (counter) examples can’t refute a claim that such examples are rare exceptions to a general rule. This is particularly true if the examples are trivial/marginal like those you give.

    What you need is a quantitative study. But, if you could provide a plausible argument on Kashmir, I’d be impressed. Bear in mind the necessity that your argument should show a negative-sum outcome. That is, if your argument purports to show that India has benefited, it should show that Pakistan has lost. Since you’ve been admirably clear about your views of my analysis, I’ll suggest that you can’t do this, and that, for all your erudition, you’ve failed basic logic.

  6. jquiggin :
    “how Britain came to have 70% of the world’s merchant fleet, ”
    Primarily, by being good and shipbuilding and [at?] running capital markets. But they were helped by the fact that much of the US fleet was destroyed during the civil war, which rather supports my point.

    That last point seems tenuous, considering that the USA simply wasn’t a major economic force until much later. However, it did play a major role in specialised areas, like whaling and the China trade. And it certainly doesn’t account for Britain’s prevailing over existing European maritime powers.

    But, curiously, Britain wasn’t “better” at shipbuilding in a qualitative sense, as regards its products, apart from a brief period in the 1870s between overtaking the French in quality and the rise of German ships that had less of a burden of past practice. The very real superiority was in a production engineering sense, from being able to make more ships more cost-effectively, and from seafaring skills acquired by practice, either or both of which may have been what you meant.

    PML, you seem to be having some methodological problems. A list of (counter) examples can’t refute a claim that such examples are rare exceptions to a general rule. This is particularly true if the examples are trivial/marginal like those you give.

    Is that first sentence projection? I never offered a list of counter-examples, for precisely the reason that it was impractical and would not prove the point; I just wondered if that was what you were after. That last sentence is a repetition of the implicit No True Scotsman argument I mentioned: any cost-effective military operations must necessarily appear “trivial/marginal” by the standards you have applied to the cases I did offer (explicitly in the Portuguese case, and implicitly to the others since you also do not accept those). I will readily agree that any military operations that have huge costs and/or downside risks are not cost-effective – but that is a tautology. To me, the important point is whether military operations can still find a place in statesmen’s repertoire, and I would suggest that the record shows that they can because from time to time they do indeed work out. So what if the occasions for them are rare? When they work, they work – and, between times, having military resources is an overhead justified by genuinely defensive reasons. And that means that a truly defensive defence policy must allow for, and seek to head off, any opportunities of that sort that others might spot, e.g. not concentrating on out and out invasions alone.

    What you need is a quantitative study. But, if you could provide a plausible argument on Kashmir, I’d be impressed.

    Um… isn’t that a tautology too? If you are not impressed, for whatever reason, it’s not plausible – to you, at any rate. As things stand, that is. Can you suggest criteria for a short reply that you would find persuasive? It would have to be short for me to manage it here, but if only a long one would serve, its absence wouldn’t be evidence against. Rather, it would leave the matter either false, or true without being provable – without you having evidence either way.

    Bear in mind the necessity that your argument should show a negative-sum outcome. That is, if your argument purports to show that India has benefited, it should show that Pakistan has lost.

    Actually, it needn’t, unless you assume that both sides use the same assessment (and that the costs cannot be passed off to third parties – that was what happened in the Middle Ages, when knights kept ransoms from their captives but their vassals had to pay their ransoms if they were captured themselves). Are you familiar with the paradox of the “two tie bet”? Two friends meet in a pub, and notice that they have each bought a new tie, each at the same price. Each thinks that his is better, so they bet on it with the barman as adjudicator, with the one with the best tie to buy another one the same for his friend. Each reasons that either he will get a tie better than his, or buy a tie just the same, so at fifty-fifty odds he can expect a gain. How can this be true for both, with a symmetrical bet? The answer lies in the fact that they aren’t using the same values.

    You will note that I explicitly pointed out that Pakistan’s calculus includes a very high value for long term national survival, which is far less material for India, and that consequently Pakistan didn’t use ordinary notions of cost-effectiveness.

    Since you’ve been admirably clear about your views of my analysis, I’ll suggest that you can’t do this, and that, for all your erudition, you’ve failed basic logic.

    Then, if I may suggest it, you have been supplying your own assumptions and reasoning in place of what I actually used. If you feel there is something wrong with what I am bringing out, the tactful approach would be to ask for clarification and not simply to assume that I am using data or arguments that don’t hold up. Every single one of the flaws you believe you have spotted are indeed flaws, some of them even fatal ones; only, they are not in what I presented or in what led me to it. For you to read them in is to make a straw man without knowing or intending it.

    For what it’s worth, I think your analysis is on the whole sound, but rests on faulty assumptions that are so close to you that they remain unexamined – and, in some cases, they do not apply. That works out like an intermittent fault in machinery, which in many ways is worse than machinery that is outright broken. Since defence issues are so important – from their downside risks – it does rather matter not to get things that work now but might fail on us when we need them.

  7. The problem with the whole “they won or lost the war thing” is always identifying wars as being between States. They seldom are in modern timies and certainly not for the USA. Wars are great money makers for some and for those people the US wars have been a fabulous success. The great trick has been to keep enough people convinced that these wars are somehow or other for the people of the USA, or for Australia for that matter.

  8. @P.M.Lawrence

    You will note that I explicitly pointed out that Pakistan’s calculus includes a very high value for long term national survival, which is far less material for India, and that consequently Pakistan didn’t use ordinary notions of cost-effectiveness.

    This rather puts its finger on the problem of subjecting war to cost-benefit analysis. Much of the prospective benefit is unquantifiable and in any event rests on counterfactuals of doubtful standing. One can wonder what benefit one puts on not being occupied by a foreign power — if that is what is asserted and clearly would ensue — and much of that is simply impossible to model.

    I don’t know why but I keep coming back to Gloria’s words from White Men Can’t Jump whenever this topic arises:

    Sometimes when you win, you really lose, and sometimes when you lose, you really win, and sometimes when you win or lose, you actually tie, and sometimes when you tie, you actually win or lose. Winning or losing is all one organic mechanism, from which one extracts what one needs

  9. “Primarily, by being good and shipbuilding and running capital markets. But they were helped by the fact that much of the US fleet was destroyed during the civil war, which rather supports my point.”

    Really? When the Dutch had the next largest (and the most efficient) merchant fleet? Not by capturing or destroying the Dutch, Spanish and French merchant fleets, and making trade other than in British bottoms risky and unprofitable? Check the Navigation Acts and look at statistics on the British share of the European trans-shipment trades.

    I think that what has changes is that effective government now needs the continuing cooperation of the governed if the cost of the garrison is not to be larger than the taxes. But, as i said, note that this is a development of modernity – we still sit on the fruits of conquest.

  10. @Hal9000
    I just call them “superfluous redundancies”. However, the use of the word “pleonastic” is especially appropriate as in my case the two words ie “superfluous” and “redundancy” are being used as true synonyms; interchangeable within the context of use. On the other hand the expression “especially egregious” is clearly not using the two words as true synoms, however much it is a true fact. For these situations “pleonastic redundancy” seems “more adequate than “superfluous redundancy”.

    What do others think – “superfluous” or “pleonastic”?

    To The Topic: military troops are extremely expensive to maintain at operational capacity; the subdivision of costs runs foul of the problem of fixed assets/costs, the variable but regularly incurred costs (ie variable costs), and to questions of risk assessment of troop in-field incapacitisation, and the subsequent rehabilition/removal, as circumstances may require.

    I guess the variable costs split broadly into theatre-specific variable costs (ie ones that any soldier would be at risk of incurring under the circumstances) and idiosyncratic variable costs (ie ones that are truly individual, applying to a given soldier, rather than being a statistical average of soldiers in a common context).

    A soldier asset that makes it far enough in-field to inflict harm-maximising experiental FMJ rounds upon the non-payrolled oppositioners (aka people attacking us who we weren’t able to buy over before the fighting started), has also made it far enough for their integrated human integrity parameter to be at severe risk of rapid decrease from its natural non-combat maximum. At this point the variable costs are of the non-idiosyncratic variety, for any soldier asset in this situation would be subject to the same risks as any other soldier asset, ie the soldier assets are largely interchangeable (within reason) in this context. Once the asset has sustained a compromising structural deviation, then we are in the realm of the idiosyncratic variable cost (new limbs, long vs short rehab, or even a death event).

    Here is where the calculus of worth (of the soldier asset) ramifies into a messy tangle of contradictory rules, obligations, ethics, and pragmatics. For example, once a minor structural compromise to a soldier asset’s integrated human integrity parameter has occurred, the question becomes one of whether it is better to retrain the newly fitted “alternatively limbed” client in the more specialised combat operations methodology for the “differently abled”, or to decommission this asset and assign a replacement.

    The question is one of risk: do we risk an already compromised soldier asset in a similar environment to the one in which the asset failed to protect integrity, or do we assign a newly formed, uncompromised soldier asset, for which the variable cost projections are perhaps more accurate as the non-idiosyncratic and idiosyncratic risk factors are quite different at this point.

    Happy New Year, civvies and grunts alike.

    Regards,
    Donald Oats

    PS: and, yes, I was making yet another feeble attempt at humour. If anyone needs a flexibly downwardly remunerated first line obscurator armed with a pen, give me a call 🙂 yeah, you, Mike R.

  11. This theory depends on the idea that that becoming involved in a war is a choice. I think that it is, but I imagine a large part of the US population sees it differently – that the US only gets involved in wars when they have been attacked, and that in those circumstances the only choice is to fight or to face annihilation. At least that’s the way the wars are sold to the population.

  12. All hail. Donald Oats!
    The man has set forth in mortal combat with and succeeeded in defeating. a malignant pleonasm.
    Or was that neologism?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s