Defence may be a public good. Military spending isn't.

Public goods are, as defined by economists, non-rival, meaning that the cost of an additional person using the good is zero, and non-excludable, meaning that it is technologically impossible, or prohibitively costly, to prevent people from enjoying the goods.

When something is a pure public good, private markets will fail to supply it in adequate quantities. By definition, firms cannot exclude people who don't pay for the good, hence generating revenues is somewhat problematic, and firms will tend to either go out of business, or not go into business in the first place.

Since it is widely accepted that governments have a role in providing pure public goods, the definition of a public good is highly political. For example, to argue that HPV vaccines are a pure public good is to argue that private markets will under-supply the vaccines, so government intervention (and, most especially, government finance) is necessary.

National defence is frequently cited as the paradigmatic example of a pure public good - see, for example, here, here, here, and here. Indeed, it is the one public good sanctioned by the Fraser Institute (here) and the Liberty Fund (here).

Economic thinking about national defence frequently goes something along these lines:

What's the difference between national defence and pizza?...To begin, one big difference between the two commodities is that two people cannot consume a piece of pizza simultaneously - if I eat the piece, you can't. In contrast, your consumption of the protective services provided by the army does nothing to diminish my consumption of the same services.  A second major difference arises because I can easily exclude you from consuming my pizza, but excluding you from the benefits of national defence is all but impossible. (It's hard to imagine a situation in which a foreign army is allowed to overrun your home but not mine.) National defence is an example of a pure public good.

                   Rosen, Wen and Snoddon, Public Finance in Canada, 4th ed'n, p. 50.

The core problem with an argument like Rosen's is that it confounds the abstract  - defence, protective services -- and real world military spending.

Yet the connection between spending money on the military and providing defence against external or internal threats is something to be established with empirical evidence, not assumed.  For example, during the Cold War, millions were spent on huge underground bunkers to house the government in event of nuclear attack. These VIP-only bunkers contributed little or nothing to the safety of average Canadians.

Military spending can be worse than useless. Not all governments, and not all armies, are nice people bent on making the world a better, cleaner, safer place. Some are interested in suppressing internal dissent, capturing resources, or other forms of rent-seeking behaviour. For those oppressed by their nation's army, spending on the military is a public bad, not a public good.

Even if it possible to establish a direct connection between military expenditures and the well-being of the public, it turns out that national defence is, as often as not, rival. Soldiers cannot be in two places at the same time. If the army is fighting floods in Manitoba, it is not providing peacekeeping services in Kabul, or quelling riots in Montreal.

The national-defence-as-public-goods argument displays a naivete about the ways of the world. "It's hard to imagine a situation in which a foreign army is allowed to overrun your home but not mine." Really? Imagine an divided society, where the foreign army overruns the homes of ethnic group A, but not those of ethnic group B, because the foreign army and ethnic group B share the same racial/linguistic/religious or other heritage.

I'm not the first person to make these kinds of arguments. Murray Rothbard expressed similar skepticism back in the 1960s. But they haven't stuck. I guess "national defence is a public good" is a better slogan than "it's complicated."

Woodrow Wilson is credited with saying

Changing the college curriculum is like moving a graveyard – you never know how many friends the dead have until you try to move them.

I continue to teach student about public goods. One reason is that I'm the kind of person who respects graveyards, and doesn't like them being tampered with. Another is that I find public goods fun - concept of public goods has enormous power when applied to everyday social interactions - who takes out the garbage, who cleans the bathroom, for example - and is a powerful metaphor for new technologies, such as downloading and pirating.

But if I have to read about national defence and pizza one more time.....

Comments

[Dan Kervick]

So there are bad kinds of military spending. Then we should try to stop doing that kind of spending.

The judicial system is a public good. But of course much actual spending on the judicial system might be wasteful or even positively harmful: gold-plated toilets for judges; bonuses for incompetent police officers who don't do their jobs and oppress citizens; emergency jail space that is never used, etc.

I don't see how your points undermine Rosen's claim or show that it is naive. The claim in the text seems to be only that national defense is a public good, not that every kind of spending carried out in the actual or purported pursuit of that good is worthwhile.

[Daniel Kuehn]

re: "But if I have to read about national defence and pizza one more time....."

I'm not sure why... you made a pretty good case that its the inputs and certain non-defense outputs that are the issue, not national defense.

Anyway, I prefer thinking about externalities and social costs and benefits for just this reason. Most cases are complicated but that doesn't change their "public-ness". Externalities makes talking that out much easier.

[Determinant]

The Diefenbunker is a wonderful museum. It's privately owned by a non-profit foundation/history club, it's not a government museum. But it is interesting to see and well done.

Second, the national defence argument is in some ways backwards. National Defence through the Armed Forces is a government good, it is the defining role of the government, the source of both order (riot suppression, putting down rebellions) and its continued existence. Policing as we think of it is a development of the order role of the armed forces, in the 19th Century it was thought better to have a separate agency do law-and-order work instead of calling out the troops. That's why the RCMP has such a militarized structure, it was inspired by the Royal Irish Constabulary which was an "order and rule" force and very military in itself.

The point is that Defence became a public good when Parliament and democracy became the foundation of government legitimacy in Canada, which happened in the 1850's. Same thing in the UK where the change happened in the 1830s, and similar changes happened in mainland Europe. The armed forces may be the Crown's servants in name, but they are servants of the public in practice.

Since we maintain that government must be responsible to the people and democratic, defence and use of the armed forces as a public good follows from that.

[Jacques René Giguère]

Quelling iots in Montreal?
Frances, Frances, Frances... once again being caught reading Maclean's...
The real face of Montreal protest is not the one on that rag frontpage but the Anarchopanda

https://www.google.com/search?hl=en&q=anarchopanda

a philosopher in a panda suit who gives hugs to the riot squad ( City of Montreal cops accept, the Sureté du Québec refuse, a nice paper waiting to be written).
BTW, there are demonstration in about 100 cities and villages.
We had one here last Tuesday. Conspicuously more than 50, just for the principle...
Radio-Canada web site had a map pinpointing demonstration sites, including the "pots and pans" but was told by the gunmint this was infringing the law. So much for freedom of the press.
Sorry for the temporary interruption, we now return to normal programming.

[Frances Woolley]

Jacques Rene - if you'll pardon the anglicism, that's a fair cop. Should I change it to riots in Vancouver?

Daniel: "I prefer thinking about externalities and social costs and benefits for just this reason" I agree, and I often think I should move the graveyard and cut the public goods chapter entirely. But then again, public goods are so useful for thinking about new media, institutions, rules, social interactions, etc. On the pizza thing - perhaps you haven't taught pub econ as often as I have!

Determinant: "The Diefenbunker is a wonderful museum. It's privately owned by a non-profit foundation/history club"

And is thus a great example for all of those who figure the private sector does a good job of providing public goods! With regards to your other point: to say "national defence is a public good when it serves a legitimate, democratically elected government" is very different from saying "national defence is a public good".

[Phil]

Rothbard is generally great, but sometimes unconvincing, national defense being a case in point. Even David Friedman acknowledges in The Machinery of Freedom that privatization of defense is unfeasible.

With regards to the government bunkers, the reason they didn't provide any services to the people was that they were not used. Well, US nuclear missiles were not used, and still provided the social benefit of deterrence. In the same way, you can make a good argument that gov. bunkers are a form of deterrence: if your enemy knows that your government can retaliate from these bunkers even after he has launched a nuclear attack, then he may think twice before pushing the red button. I don't want to suggest that this means that the construction of these bunkers was justified. The point is that it's an empirical question of costs/benefits.

The notion that tyrannical governments can use military spending as a public bad to repress their own people is of little use. It's a safe bet that the populace, being routinely fired at, already knows this. That's probably why they're trying to overthrow their regimes, so that they can start managing their resources more rationally.

That "soldiers cannot be in two places at the same time" is does not demonstrate the non-rivalry of defense spending. Non-rivalry means, just as you properly said, that wherever these soldiers are, the cost of an additional person being protected is zero.

Finally, the situation where a majority allows the ethnic persecution of a minority by a foreign army does not alter the fact that defense spending is non-excludable, because that minority is still being protected when a foreign army hostile to the majority attacks the nation. It's just that in this case, the majority uses the public good against the minority. Seen from another angle, this shows that fundamentally, no good is truly non-excluable, because you can always kill a citizen.

[Frances Woolley]

Phil: " Even David Friedman acknowledges in The Machinery of Freedom that privatization of defense is unfeasible."

Tell that to Boeing or MacDonald-Douglas.

Even if one accepts that public *finance* of defense is better than private finance, or that it is a good idea for democratically elected governments to have a monopoly on the use of force, *it does not follow* that national defense is a public good.

"the reason they [bunkers] didn't provide any services to the people was that they were not used."

Nope the reason they didn't provide any services to the people was that there was only space for about 500 people in them. Even if the bunker was still up and running, it be of absolutely no benefit to me personally in the event of a nuclear attack because I would be dead.

"Non-rivalry means, just as you properly said, that wherever these soldiers are, the cost of an additional person being protected is zero."

Yup, and if the lighting is on in my living room, the cost of an additional person enjoying lighting is zero. If the TV is playing, the cost of an additional person watching it is zero. In this trivial sense, most goods can be considered public goods.

[Bob Smith]

When I first read your title, I thought you were going to make a very different (and rather Zen) point. Accepting that National defense is a public good (although you make good points that it need not be), but observing that the military spending that achieves it is not. It's a self-evident point, but one that is frequently, and conveniently for the parties involved, omitted - public goods may be financed by the government, but they're provided by private parties (soldiers, suppliers, arm's makers), for their private benefit (that's a bit harsh, many of the soldiers I know serve out of a strong sense of duty to their country - they aren't mercenaries - but on the other hand they aren't working for free and are reasonably well-paid).

It's a important distinction (especially in the era of austerity), that providers of "public goods" (which are often nothing of the sort) like to fudge. When governments propose cutting funding for, say, education, health or defense, we're those are opposed on the grounds that they are public goods. That may (or may not) be true, but it's beside the point, the question is whether or not you can get the same benefit from those goods at a lower cost.

I remember the UofT TA union used to run ads to the effect that their (our) working conditions were students learning conditions. True enough, but the counter-argument was pretty easy "sure, but if we don't give you your 5% increase are you going to be a 5% worse instructor?".

[Bob Smith]

In terms of your observation that defense may be rivalrous in consumption (I.e., an army can retreat and leave part of a country at the mercy of an invader), it's a fair point in some (non-trivial) circumstances.

I wonder if the conception of defense being a non-rivalrous in consumption owes something to the decidedly Britanic (and anglosphere more generally) roots of modern economics. If you're a land power with a large (and post-1790s conscript) army (say, Napoleonic France or post 1870 Germany) the abandon the frontier, defense in depth, strategy makes sense (and has been used with some success over the years).

On the other hand, if you're an island with a large Navy and tiny professional army, than defense seems pretty non-rivalrous. Either your navy can stop your enemies from landing on your coast or they can't. If they can't you lose. For such a country, defense is an all or nothing proposition. And that would be a fair description of both Britain from the 18th century onward and the US prior to WWII.

Against that background, it's easy to see why economic theorists living in those countries might have picked defense as the obvious example of a public good.

[Min]

Frances Woolley: "Public goods are, as defined by economists, non-rival, meaning that the cost of an additional person using the good is zero, and non-excludable, meaning that it is technologically impossible, or prohibitively costly, to prevent people from enjoying the goods.

"When something is a pure public good, private markets will fail to supply it in adequate quantities."

I think that I am having a little trouble with the definition. It sounds like speech is a public good. The cost of another person's speech is not exactly zero, but that's a good approximation. And the cost of shutting someone else up is often prohibitive. OTOH, private markets seem to supply more than enough of it. ;)

Somehow, though, I suspect that speech would not be considered a public good by economists.

?????

[Frances Woolley]

Min - as was said earlier, there is considerable overlap between public goods and externalities, and some would argue that it would be more useful to get rid of the concept of public goods entirely, and just talk about externalities instead. Speech is, perhaps, an example of this.

Where speech be can widely disseminated and made available at effectively zero cost, e.g. the Ted lectures http://www.ted.com/talks the question of whether or not it can be thought of as a public good is an interesting one. Is it excludable or not given modern technology? Even if it is a public good, are there alternatives to government provision? These are the kinds of things that it's possible to talk about once one gets past meaningless abstractions like "defence."

Bob - interesting idea on the Anglo-Island hypothesis. The idea of collective consumption goods, now known as public goods, is usually attributed to Paul Samuelson - he wrote a paper on the subject shortly after WWII. In some ways I'd see it more as a Cold War idea, where nuclear weapons act as deterrence, protecting whole nations (or, sometimes, whole continents) from nuclear attack.

Bob - "the military spending that achieves it is not" - that is part of what I was trying to get at with this post.

[Shangwen]

As Bob says, it is self-evident that defence spending is not a public good. Indeed, the lopsidedness of the defence industry, due mostly to either having one customer or one type of customer, and a legal barrier to selling many products to private citizens, makes you wonder how "private" these firms are. Sure, they compete against each other, have boards, are publicly traded, etc. Yet they do not behave and participate in the economy in the same way as Wal-Mart, Applebee's, or Disney. If you have a customer with almost unlimited resources, you aren't playing the same game.

I would compare this, of course, to the health care regime, and the difficulty in really identifying what the labels public and private mean. (And no, health care is not a public good. Otherwise there would be no wait lists.) Is OHIP a public entity, when a huge chunk of its payments are to private (non-civil-servant) doctors? Are those doctors "private" when their sole payer is the government and there is no price system? The BC College of Nurses and the Ontario College of Physicians are technically public entities, accountable to the provincial government. Do you think the government gives them any marching orders, or have their leaders been punished for not licensing more and cheaper surgeons? They are essentially private (self-directed) entities whose self-interested behavior imposes massive externalities.

And why is defence always seen in terms of protecting the state, rather than protection of private individuals? In a time of conventional war, the actions of the military yield unevenly distributed benefits. Defence departments also initiate war, and they pick and choose their targets as they (privately) see fit. If you are on the wrong side of "national defense", you really want that good to be excludable.

[Frances Woolley]

Shangwen: "And no, health care is not a public good. Otherwise there would be no wait lists." Absolutely - year after year I give MC questions along the lines of "A hip replacement is a good example of a public good: (a) true (b) false (c) uncertain" (it's better to use a specific procedure to avoid questions about the public benefits of things like vaccinations). Every year there are people who get it wrong.

The thing is, people go from the specific to the abstract in health care, just like they do in national defence. Sure, a hip replacement might be a private good, like a bunker for the prime minister of Canada (you bet that's excludable, they had people with serious weaponry standing by to exclude the great unwashed). But people go immediately to abstract goods like "a health population" "a strong economy" "national defence" to argue for publicness.

[Shangwen]

Here's a double whammy (and some great medical advice): If you want the best health care in the world, go to an American military hospital. Admittedly, there are some complicated, perhaps unpleasant, steps to getting there, but once you're in you'll get the best care, newest technology, most effective drugs, and highest staff ratios of any medical system in the developed world. I've visited a few of them, and my jaw never fails to hit the floor. This is national defense and health care in glorious union.

The access to a massive budget is not the core issue. The issue is that national defense, health, lighthouses (don't forget those), etc., may in some sense have "publicness", but the institutions we construct to deliver them are decidedly not, at least as economic actors. By institutions, I mean not only bricks-and-mortar, but the regulations, payment systems, restricted labor markets, political capture, and rent-seeking through which all of these are delivered. As a result, even in peacetime in a country with very high overall health, we have small numbers of people earning great wealth from militarism and medical care.

And that robotic pathology lab in Virginia...ye gods.

[Peter N]

A spending category is an attempt at a shotgun marriage of philosophy and accounting. You can't look at a single defence spending transaction and determine unequivocally how much defense it provided or how defensey it was.

You can't look an individual transaction for investment goods and say how much investment it bought or how investy it was. Ditto for consumer goods, and even government spending (and the 3 categories each have significantly different accounting rules).

Economists tend to ignore this problem. But it crops up everywhere. Take saving = investment. This may be pretty in a model, but it involves two levels of conceptual mapping. First the model has to be an accurate representation of functions of the categories (sort of a God's eye view), and second the categories need to map onto the system of national accounts which would require massive disaggregation and reorganization. Otherwise saving is just a residual term where discrepancies go to die, a sort of GDP graveyard.

What economists seem to actually do is disaggregate those portions of the accounts where they feel they have to (and different economists have different feelings) and take the rest as gospel.

It might be possible to meaningfully do wholesale disaggregation, but I've never heard of anyone doing it.

Then, of course, you'd have to calculate error bounds on the individual disaggregated sub-categories and propagate them through the model. And, you have to deal with imputations and hedonic corrections under non-homothetic preferences.

I think it's fair to say this would be an interesting and and definitely (almost definitively) non-trivial project.

[Peter T]

I think defence challenges economics not because of its "public good" (or not) nature, but because it's inherently collective. There's no way to reduce the defence of a state or a tribe to the defence of each of the individuals involved. The collective will have its essential areas to protect (perhaps the capital or the major industrial areas), its key networks to maintain, and will call on some selection of its population to die and some others to suffer so that the whole can survive. And its historically characteristic of people that they will mostly accept this - even those who are most likely to die.

Defence has occasionally been provided by private interests. It always failed when more than trivially tested, because the individuals involved could not be paid enough to die: there was no felt collective to defend. It's an area where you have to re-think your conceptual individualism.

[W. Peden]

Peter T,

I don't think that a collectivist conception of defence is going to get us very far, at least in the non-idealised cases that form the motivating problem of this post. For example, in the Allied democracies during WWII (I pick on them to show that this isn't a democracy vs. authoritarianism issue) there was considerable internment of "alien" citizens, e.g. the internment of Japanese-Americans and Japanese-Canadians. This was a defence measure: it what sense was it protecting the liberty of American citizens or Canadian citizens as a collective?

What seems to be most problematic here is that the word 'public' doesn't have an unambigious reference. Is spending on underground bunkers for the protection of leaders spending on the public? Is spending on putting citizens into camps for the protection of other citizens spending on the public?

Say I'm a devoted Arab nationalist living in Sinai in 1973. Which side is spending on my defence in my view: the Israelis, who control the territory on which I live, or the Egyptians, who I see (and the Egyptians see) as my liberators? At the very least, this is a concrete example of a case where I would regard the Egyptian army in a very different way from a hypothetical Jewish Israeli neighbour. What's the collective being defended here? And by whom?

Why suppose that collectives correspond with national boundaries at all? Many non-Spaniards fought and died in the Spanish Civil War on both sides. George Orwell didn't risk his life for Spain, he risked his life for socialist internationalism and he was later willing to risk his life for Britain in WWII; like most of us, he regarded himself as a member of many collectives and so his behaviour cannot be analysed in terms of his membership of any one group, but rather as a result of his individual psychology.

As for economic analyses of military spending as if all military spending was defence spending, it's an interesting example of the power of propaganda. If all military departments still had names like "The Department of War" or "The War Office", then it would be apparent that spending by the War Office and spending on defence are two distinct categories.

Also, there is absolutely a difference between the purchase of armed forces and the provision of armed forces. Thousands of years of mercenary armies attest to that and it didn't always fail when more than trivially tested, otherwise people would have been insane to go on recruiting mercenary armies. The French Foreign Legion is still going strong and they clearly are willing to die. There were Vikings willing to sail to the other side of the world (Byzantium) which was hugely risky in itself, just in order to risk their lives for pay and glory.

It's also anachronistic to see all past non-mercenary armies as fighting for "the collective" i.e. it imposes modern nationalistic notions on, say, Medieval peasants. At the extreme, if my options are living at a subsistence-level (with a very real threat of death due to insufficient caloric intake) or going to risk my life fighting for some distant King, I may well take the latter option out of self-interest (I might be able to plunder some gold) or divine duty (or both). If I go to fight in the Crusades, it might be for plunder or it might be to get into heaven or to do my duty to God (or all three). If I fight for the Laird who owns the run rigs I cultivate, it might be out of a sense of loyalty to the Laird or it might be because the alternative was even more risky.

In these cases where the goal was minimising my risk of dying due to the dangers of peasant life, in what sense am I different from a mercenary? And isn't it a well established historical fact that men- especially those living close to the margins of death- have been willing to risk their lives for plunder and/or glory? If so, why assume that mercenaries would be unwilling to do so?

[Peter N]

If you hear phrases like "how can you put a price on..." or "no sacrifice is too great for...", then you know this is an area that will be an infinite source of agency problems and people ready to exploit them.

What is defense exactly? Where is it supposed to be done and how? Who pays for the inevitable externalities? What is to be the allocation of defense capital?

History supplies an enormous number attempted solutions to problems of defense just compare Poland around 1600, Switzerland, the Ottoman Empire, the Byzantine Empire, England before 1640 and the WW II Soviet Union. These are all generally considered successful systems, but a public good analysis of them is beyond my powers.

[Shangwen]

Peter T: I get what you are saying, but your comment strengthens Frances' point. With the exception of nuclear deterrence, defence is not "inherently collective" but traditionally majoritarian, aimed at a common view of the legitimacy of the state and, far worse, the legitimacy of imperial contests between governments (e.g., "bringing democracy" to a country). But what the institutions of national defence deliver is inherently private because unequal. If 97% of people live outside the National Capital Region, how is it possible for "the collective" to give it priority for protection? I doubt the victims in the Twin Towers died praying for the safety of Air Force One.

As for the collective requiring "some selection of its population to die and some others to suffer so that the whole can survive", this may reflect a prevalent belief, but I think this assumption goes back to an era before the 20th century, when people assumed (correctly perhaps) that a wall of soldiers would suffer massive casualties but ultimately stop invasion. But the massive drop in combat death ratios since WWI--and the huge increase in combat casualty survival rates--has changed all that. Police injury and death rates are now higher than the military equivalents. The US Air Force has not a single program in development involving manned flight; it's all drones. The reality of serious personal risk for military members is in rapid decline--though its mystique is not.

Obama signed a law allowing indefinite detention of civilians without cause; other governments have handed over millions to Karzai and his cronies. So yes, it is frighteningly collective, because so few people have objected to these things.

[Frances Woolley]

I was going to write a long reply to Peter T but W Peden, Peter N and Shangwen have posted such excellent comments I have little to add.

W. Peden: "the internment of Japanese-Americans and Japanese-Canadians. This was a defence measure: it what sense was it protecting the liberty of American citizens or Canadian citizens as a collective?"

I don't know much about the Japanese-American internment, but one of the interesting things about the Japanese-Canadian internment was the extent to which it was an excuse to appropriate Japanese land and property. There was a thriving Japanese community in what is now Richmond, BC. Japanese-Canadian families had lived there, sometimes for years/generations, and had market gardens, fishing boats, and all sorts of other thriving small businesses. All of that property was confiscated and sold for next to nothing.

Shangwen: "The US Air Force has not a single program in development involving manned flight" - yup, and we're buying F-35s. For whose benefit?

[W. Peden]

Frances Woolley,

Interesting. I wouldn't want to have to give an analysis of the relationship between those Japanese-Canadians and the government in terms of protection of property rights.

[Determinant]

Shangwen: "The US Air Force has not a single program in development involving manned flight" - yup, and we're buying F-35s. For whose benefit?

Canadian military contractors. The terms of the US/Canada Defence Purchasing Agreement, which date to 1941, mean that the purchases of each country in the other country have to be balanced by subsequent purchases in the other direction.

SNC Lavalin, IIRC, has a munitions factory in Saskatchewan geared to supplying ammunition to the US Army in order to provide such countervailing transactions.

I've worked at military suppliers; the market is very artificial.

[Shangwen]

Determinant: "the market is very artificial". No kidding. Or in other words, it's a command economy.

Despite disagreeing with Peter T, I think his post raises an interesting problem: if people oppose a public good (e.g., they would welcome invasion, which happens; the lighthouse wrecks their view), and some of the non-excluded still experience disutility, it's hardly a public good then.

[The Keystone Garter]

"And that robotic pathology lab in Virginia...ye gods."

How is it robotic?? Gene sequencing machines?

[Peter T]

W Peden makes some good points, but I don't think they dislodge the notion that defence (or attack, for that matter) is about collectives. Collective here is short-hand for the group the people involve feel part of for the purposes of defence - roughly, the one they are willing to die for. People have lots of identities, but for the last century or so, this particular one has mostly been the nation-state, although various causes and non-state ethnicities have continued to play a part. It can historically be the clan, tribe, regiment, mercenary company or whatever. It will exclude (fail to defend) people outside the group - Japanese-Americans in an example given above, and protect the key people (so government gets the bunkers, but it could be the bodyguard dying for the shaman or the king). Only in a democracy is it plausibly majoritarian - the peasants may be the majority, but they may be seen as expendable by the elite (but the peasants themselves will identify with and fight for their village, if not the state). So the decisions are more like the automatic decisions of your body facing injury - the key systems are preserved, not the largest bits. Which only makes sense if the reasoning is about the group and group survival. Since group identities shift around, merge or break down, the details are messy. But it's always about "us".

Mercenary companies when effective built strong collectives - they were mostly little tribes, complete with women and children. When they didn't - when people were just hired as individuals, they didn't work. There's a lot of history on this - too much to cite here.

And I don't know the people in the Twin Towers, but the group who rushed the cockpit on Flight 93 were surely not hoping to survive themselves, nor were their immediate families (if not on the flight) in danger. They were defending some "us" (same for hijackers of course).

The take for Frances post is that the public here has to analysed as a people thinking/feeling their way through a group-level issue, not as an aggregate of individuals summing utilities.

[Determinant]

Determinant: "the market is very artificial". No kidding. Or in other words, it's a command economy.

To a large extent, yes. But unless we wish to produce a jet fighter, drone or artillery shell entirely in a government-owned factory (which has been done often enough but is has drawbacks) then we are going to rely on capitalist firms to supply armaments at a profit. During WWI and especially WWII "Excess Profits Taxes" were implemented in limit war profiteering. It was politically popular and financially expedient to do so. Historically private firms were relied upon to supply "surge" capacity when needed.

Second, as I noticed in my work, saying what is military and what is not in an industrial, technological age is well nigh impossible. That winch to be fitted on a warship was worked on with tools from Canadian Tire and lubricated with Esso oil, for example. The list is endless.

But the world is not a Night Watchman State; laws are not self-enforcing, power differentials exist and force is a method to resolve disputes.

Oh, and the purpose of the Diefenbunker was to enable the core apparatus of the state to survive and re-establish some sort of order after a nuclear catastrophe. It was the same with the Greenbrier Resort bunker for the US Congress. Interestingly the Diefenbunker had a large vault for the Bank of Canada's gold reserves. One may take this as proof that Public Servants are not libertarians.

[K]

The elephant in the room, of course, is that *globally* defense is an unmitigated net *bad*, not a good. Whenever we spend to improve our own defense, the negative externality on the rest of the world is therefore even greater than the benefit we receive. Global defense spending is literally more than 100% wasteful. Even finance can't compete with that kind of inefficiency.

[Determinant]

No argument. At least Costa Rica gets it right, they have no armed forces and are a long-standing, stable, solid respected democracy.

[Frances Woolley]

K: "The elephant in the room, of course, is that *globally* defense is an unmitigated net *bad*, not a good."

This is an excellent point, and one that I don't think I've ever seen made in textbook discussions of public goods. The standard argument is that, without government intervention, public goods like "defence" are underprovided, but in the context of a global arms race, military spending is likely over-provided.

Determinant: "Oh, and the purpose of the Diefenbunker was to enable the core apparatus of the state to survive and re-establish some sort of order after a nuclear catastrophe." Yup, and there were two problems with this idea. First, the location of the Diefenbunker was so well known that it would almost certainly have been a target in the event of a nuclear war. Second, on the idea of re-establishing order after a nuclear catastrophe: watch the little (and achingly sad) animated story When the Wind Blows, there seems to be a youtube copy here.

Peter T: "I don't think they dislodge the notion that defence (or attack, for that matter) is about collectives"

Agreed, defence is about collectives. I'd go on to say that collectives are fascinating, which is why I like teaching and studying public goods.

But military spending is not equivalent to defence. If you have time, go on the World Bank website, and download the military spending numbers. You'll see that some of the top countries in terms of military spending as a percentage of GDP are places like Saudi Arabia. Now is that about defence (it might be, the middle-East has had its share of wars over the years), or it about strengthening the US/Saudi relationship, or it just about having some really cool jets?

As an aside - I've heard it said that no one would ever try to invade the United States, because the sheer number of guns in circulation would make occupying the country impossible. That's a form of defence that isn't counted as military spending.

[K]

Thanks, Frances.

"in the context of a global arms race, military spending is likely over-provided"

Is there any other context? Since there are no external (to the world) threats, *all* defense spending since the beginning of civilization has been to gain relative advantage. There is no absolute advantage possible without an external threat. So when you say it is "likely" over-provided, I'm not sure I understand what possibility you are leaving open. Alien invasion? :-)

[Frances Woolley]

K "I'm not sure I understand what possibility you are leaving open"

The possibility that someone will come up with a convincing counter-arguments, thus making me look foolish!

[Bob Smith]

Frances,

Only a Canadian could think that the Canadian prime minister would merit a soviet nuke - keeping in mind that in the late 50/early 60s the soviet inter-continental nuclear ability was quite limited, and not terribly accurate (nuking Ottawa, easy, nuking a particular bunker, not so much).

For the same reasons, the Diefenbunker was built at a time when a nuclear war was still potentially "winable" (at least from a North American perspective)- i.e., the sovs could be wiped out at a cost of "only" 10-15 million American lives (obviously, this is "winning" in a very Charlie Sheen use of the word).

Over the course of the 60s with the expansion of the soviet icbm arsenal (the sovs still wouldn't be able to waste the diefenbunker - the sovs never managed to develop particularly accurate nukes- but there wouldn't be much left above ground afterwards), that situation would change, but at the time the prospect of having to govern a post-apocalyptic country that was more than a collection of radioactive carbon flakes, mutant roachs and mad-max biker barbarians was real.

[Shangwen]

K, your point on the global context is a great one. That should be in a textbook.

As I was reading these comments, it struck me how many other parallels there are. For example, "public safety" is equally vague and meaningless though it may be considered a public good, but only if you accept that the public (or the collective) needs to be continually redefined by removing some humans from the public, in which case it is perfectly ok to ignore problems like prison violence and rape (including violence by staff). I'm not opposed to just imprisonment, but if anyone thinks safety is a public good then our disregard for institutional violence shows how wrong that is.

[Determinant]

The Diefenbunker staff state in their tours that the bunker could not withstand a direct nuclear blast from a typical warhead. It was designed to withstand a blast lobbed in the direction of Ottawa and also to keep the radioactive cloud off the senior government.

The whole place has an air of unreality, it is very serious and very silly all at the same time.

[Peter N]

@Frances Woolley

This is in response to your reply to me at the money illusion, since that thread is dead, and the topic is more likely to be productive here (one can hope).

Many of the arguments here and in other economics blogs seem to arise from terminology problems. you can't get anywhere with different people using different definitions. If there are subtleties involved this deadly. I saw this with the Steve Keen discussion. People were confusing accounting identities with dynamics. Macroeconomists seem to be rather casual about this, and often it doesn't matter, but not always. Confusions of stocks and flows can cause a lot of trouble, though "Economics Is The Science Of Confusing Stocks With Flows." is a bit extreme.

Another oddball problem is historical revisionism. Walras wasn't a proponent of Walras law. Say changed his mind about Say's law, Hicks didn't think very much of the IS-LM model ("a classroom gadget") or equilibrium models, in general. History gets lost or revised in every field, but the problem seems almost willful in economics.

Take money. It has a number of related everyday meanings, a formal definition, and certain special uses common to particular persuasions of economists.

From Wikipedia

"Money is any object or record that is generally accepted as payment for goods and services and repayment of debts in a given socio-economic context or country.[1][2][3] The main functions of money are distinguished as: a medium of exchange; a unit of account; a store of value; and, occasionally in the past, a standard of deferred payment.[4][5] Any kind of object or secure verifiable record that fulfills these functions can serve as money."

I would agree with this.

"The money supply of a country consists of currency (banknotes and coins) and bank money (the balance held in checking accounts and savings accounts). Bank money usually forms by far the largest part of the money supply. [6][7][8]"

"In economics, money is a broad term that refers to any financial instrument that can fulfill the functions of money (detailed above). These financial instruments together are collectively referred to as the money supply of an economy. In other words, the money supply is the amount of financial instruments within a specific economy available for purchasing goods or services. Since the money supply consists of various financial instruments (usually currency, demand deposits and various other types of deposits), the amount of money in an economy is measured by adding together these financial instruments creating a monetary aggregate."

Two different definitions in the same article! And neither is completely correct! The first is poorly defined for the US, where sweep accounts are common.It's also more than M1, but less than M2. And "the amount of money in an economy is measured by adding together these financial instruments creating a monetary aggregate" isn't right either for the Federal Reserve aggregates (where it's pretty close) or the Divisia aggregates, where it is both wrong and misleading. The advantage of the Divisia aggregates is that they are not simple sums, but instead are the result of Barnett's work applying index theory.

You seem to be using money in the sense of M2, but I'm not exactly sure.

I also commonly see "high powered money". This appears to have been a synonym for M0, but in the era of excess reserves, I don't know what it is other than something to be multiplied in the money multiplier story.

As for the point I was trying to make:

Defaults cause losses for bond holders, while the gain for the debtor is notional, since it represents a sunk cost.

Banks used mortgage based securities as assets rather than loans, because this improved their CAR (since the AAA rated and insured securities were "riskless") and credit default swaps were inexpensive (as they were from AIG).

Many institutions were required to hold only AAA rated securities. If the securities lost their AAA rating, they had to sell. Forced sales -> price declines -> doubts about the asset class-> more sales->mark to market accounting losses by the bondholders

Now if the bondholders are banks, they may be forced over their CAR and be required to raise capital or sell assets. The FDIC will close banks that exceed their CAR for more than a short time.

This can produce something like 2007-2008.

The way debt enters into this is through zombification. Assets with large unrealized capital losses become immobilized, because the realization of the loss would be aversive. Since this aversion easily spreads to apply to the incumbent politicians, governments have a strong temptation to try to postpone or cushion the resolution of the debt, because such resolution will enrage either debtor, creditor or voter, depending on who is going to suffer the loss.

However if the loss is not acknowledged, this unpleasantness can be postponed, until a rising economy floats some boats, inflation erodes the real value of the debt, the voter is tricked into paying the debt through obfuscation and political smoke and mirror or the politician retires - in short, "maybe the horse will sing".

This plan "succeeded" in Japan, and is now failing in the euro-zone.

As I've said before creditors are seeking a real return, when the future economy can, by definition only give them a nominal one. They want to be made whole for the difference between the real and nominal returns. That this is economically the equivalent of pulling yourself up by your bootstraps, doesn't seem to persuade.

And, of course, it isn't impossible to make some of the creditors whole at the expense of the debtors, other creditors, innocent bystanders and the economy in general. They don't care if the pie is smaller, as long as the real value of their slice is larger.

The Greeks can't afford to pay everybody, but, as we've seen, some creditors can improve their positions relative to others while the country falls apart. Of course those of the Greeks who might be considered the most responsible, have probably already left the country with their money.

Any resolution of this charade would require the creditors to appear greedy and shift the moral focus from the sins of the Greeks to their suffering. This gets close to anger at the victim for making you feel guilty, which is, unfortunately, all too common.

It hasn't quite come to this in the US, but you have only to look at how the Resolution Trust Corporation dealt with the S&L crisis to see what a solution might look like.

Andrew Mellon was right that the rot had to be purged from the system. He just had the wrong idea of what was rotten and what the cure should be.

[K]

Frances,

Nothing to feel foolish about. Saying occasional minor wrong things is a necessary hazard of writing deep, thought provoking stuff in a public space. Others with apparently hyper-fragile egos delete embarrassing comments or turn off comment sections *whole sale* when caught making blunders. Almost, literally none, other than you, will actually admit to ever making a mistake. That is a great credit to your intellectual honesty and dedication to open-minded public discourse, which is why you are such a huge part of what makes this blog one of the most important sites on the web for me. Thanks!

Shangwen,

Still digesting your comment. You have put your finger on a horrible truth. Blacks, women, gays, children, prisoners, non-citizens in general, and God-knows who else: there is a vast expanse of humanity whose legal agency has at various times been denied, sacrificed at the altar of our grand delusions of allocative efficiency. Thank God I'm one of the "humans."

[Frances Woolley]

K - you're making me blush!

Peter N - hopefully Nick Rowe will see this and reply, this is really more his area than mine.

[Bob Smith]

"I'm not sure I understand what possibility you are leaving open. Alien invasion? :-)"

Doesn't the "precautionary pricincipal" require us to develop policies to address that posibility?

On a serious level, Frances is probably right that some level of global defense spending is probably optimal, in much the same way that some level of police spending is likely to be optimal. Just as crime wouldn't disappear if you spent nothing on the police, conflict wouldn't disappear if countries spent nothing on their armed forces or militaries. You wouldn't have world wars, but as the last half-century has amply demonstrated, you don't need conventional wars, large armies, technology or meaningful spending to perpetrate bloodbaths, chaos and conflict (think Cambodia, Sudan, Somalia, or Rwanda for examples low-tech mass-slaughter, or Al Qaeda, Somali pirates (or pirates of any nationality since man first set sail, lest I be accused of picking on Somalis) for examples of non-nation state, non-military sources of conflict). Think of it as policing outside the boundaries of the traditional nation state.

I know, I know, on one level some of those are terrible examples, because they occcured despite significant global military spending. On the other hand, those are examples where, for a variety of reasons (none of them particularly honourable) external military power was not applied to resolve/mitigate those conflicts. Those are probably cases where you would want a "benovolent" defense provider to intervene. And others are good examples of military force being used to provide stability (of sorts) - notably the Vietnamese invasion of Cambodia (the Vietnamese had their own reasons, but I'll give them credit for the result) or the gradual, more or less thorough, suppression of piracy over the 18th and 19th centuries.

Granted, though, this is probably a case where competition is costly - if there were a single defence "provider", you could probably get the same level of security for a fraction of the cost (the example of the suppression of piracy might be a good one - the Royal Navy arguably had a dominant share of the "anti-piracy" markets, sparing other countries the cost of providing their own anti-piracy defense. Hmm, another anglocentric example of a "public good"). I suppose the counterargument is that you might not want a single provider of "defense", for much the same reason you don't want a single provider of cars or computers. That would only be more efficient if the "defense" provider is generally benovolent, it would certainly not be the case otherwise (the Royal Navy would have been far less efficient a "defense" provider if, in addition to hunting pirates, it also preyed on the merchant shipping of other nations). As in any other market, the efficiency case for monopoly over competition (or vice-versa) in defense is highly fact specific.

And then there are the aliens.

[Shangwen]

Yes, there are the aliens among us, though I am really more concerned about zombies and vampires. Suppose we treat public safety as the domestic equivalent of national defense—that is, a public good that will non-specifically preserve the population’s well-being in the face of some risk agent. And indeed we hear very similar rhetoric about safety in domestic matters: needing police and prisons to keep the streets safe; needing occupational licensing to keep consumers safe; the importance of keeping the very poor wretched so that others can have minimum wages, etc. And if over-providing national defense means economic harm elsewhere, over-providing job safety for Canadians means impoverishing potential immigrants.

The truly private nature of goods under the guise of publicness can indeed be brutal. The only reason so few people care about the horrors of prison is that most have accepted the idea that anyone in prison must be a disgusting animal, and so politicians have little to gain (indeed, something to lose) by advocating for more humane and thus costlier prisons. Considering that, as per this report, two-thirds of all prison rapes are committed against prisoners by staff, and not by other prisoners, there is a huge price to pay for “public safety”. Apparently the constitution not only authorizes the government to establish a criminal code and penal system, it also authorizes the government to destroy people.

[K]

Bob,

I wasn't thinking of the Somali pirates as external to the global economy. I was merely pointing out that for them to make defense (attack) investments and for the rest of the world to invest in defending against them, is, at best, a zero sum game. You seem to make an a priori separation of the world into "aliens" (a collection non-state and nominally illegitimate state actors that we don't like) and "countries" defending against said aliens. While emotionally appealing, in order for this to be a useful distinction, it requires suspension of disbelief, that state actors, like the aliens, aren't simply using their "defensive" capability in order to maximize their own economic gain. I know that to compare the military actions of superpowers with certain non-state actors is an impermissible form of public discourse, so I won't be going down that road. Suffice it, perhaps, to point out that to ascribe morally superior, non-optimizing motives to state actors is not very "economist" of you.

As for the *real* aliens, two points. First, there's not a chance in hell of humans mounting an actual defense against an interstellar space faring civilization minded to annihilate us. Imagine how easy it would be to nuke earth from space. Second, aliens are "people" too. Why stop at homo sapiens when computing the cost of externalities?

[Bob Smith]

Shangwen, on the subject of zombies, did you read this story: http://www.thestar.com/news/world/article/1201635--man-shot-eating-another-man-s-face-caused-by-excited-delirium-say-miami-police?bn=1 ? I call your attention to the fact that it is apparently the second such incident in the last month. Creepy. Yet another reason to avoid miami.

K,

Well, let me put it to you this way. Is hiring the police a zero sum game? Because on a global scale, devoting resources to defense is akin to hiring police nationally. Yes is a correct answer, but only if you assume away criminality. No is probably the correct answer if you start with the assumption that people are inclined to commit crimes. Scaled-up to the global level, the analogy is the same. Military spending is socially wasteful, but only if you believe that a world without conflict is possible (despite the vast human experience to the contrary).

As for the suggestion that I "ascibed morally superior, non-optimizing motives to state actors", I gotta ask, did you read my post? I pointed out that in many of the examples I cited, state actors DIDN'T intervene for reasons that were not "particularly honourable" (Rwanda and Sudan stick out in that regard), or that when they did they did so for their own reasons (Vietnam in Cambodia, the Royal Navy in the 18th and 19th centuries, although it would be churlish to complain about people who do the right thing for the wrong reasons).

More to the point, the thrust of my post was that the optimal level of military spending that a benovolent provider of "defense" would engage in is likely to be non-zero. (For much the same reason that no nation state, or at least no nation state of any side, engages in police spending that is non-zero). That isn't a statement about what countries actually do (we don't have a single benovolent provider of defense), it's a statement about what a benovolent provider of defense SHOULD do. Of course, we might disagree about the extent of that role (although when it comes to Rwandan Genocidaires, Sudanese Janjawee, the Khmer Rouge, Somali (and other) pirates, or Al Qaeda terrorists, that's not really a collection of people for whom "live and let live" really comes to mind), but unless you think that role is nothing, you're conceding the (really quite narrow) point.

[Bob Smith]

As for "real" aliens, you're assuming that only aliens minded to annihilate us would be dangerous.

The history of "alien" encounters on earth suggests the dangers in that thinking. Aliens may inadvertently kill humans through disease or polution (smallpox killed more native Americans than European muskets ever did). Aliens might not even realize that they're annihilating us (the arrival of humans in the Americas coincided closely with the extinction of large mammals in the Americas, ditto with the disappearance of the dodo, the impact of human pollution on certain animal populations might also be analogous - humans are typically fairly clueless about their impact on other species, why would aliens be any different?).

Furthemore, even aliens with hostile intentions might be deterred. And, again, the experience of "alien" encounters on earth is telling.

First, it's conceivable that aliens arriving on earth might be technologically sophisticated in one respect - say inter-steller travel, but technologically backwards in other respects - weapons. There aren't a lot of good "technological" examples of this on earth (technological diffusion generally paralleling human conflict), but one theory of Neaderthal extinction posits that that the Neanderthals, while bigger, stonger and having larger brains than homo sapiens, found themselves outclassed by homo sapiens who could run faster (due to anatomical differences) and had lower energy demands. In effect, the Neaderthal "technology", while better than that of the homo sapiens in many respects was crucially behind the homo sapien "technology" in crucial areas.

Second, it's possible that alien travellers, even if technologically more sophisticated than humans might not have sufficient resources to conquer a well defended planet. To give an analogy, on Columbus' first trips to North America he travelled with 3 ships and 80-odd men. He may have had technological advantages over the peoples he encoutered, but had they turned their mind to it, they could have burned his ships and slaughtered him and his men wholesale. In retrospect, they probably should have. It isn't implausible to think that our first encounter with an alien would involve an alien Columbus, with us playing the role of the Tainos people. (And while, for Columbus, the Americas were a few weeks away, for any inter-stellar space traveller help might be aeons away).

Third, aliens with a taste for primate might think that Humans are good eating, but they might not be inclined to hunt well-defended humans and move on to eat the monkey people of Omicron-Persei 8, for much the same reason that a pack of wolves might avoid an alert herd of cariboo and move on to eat a dying Bison. It's not that wolves can't kill Cariboo that form part of an alert herd (if hungry enough, they'll try), it's just that it's hard to do and carries risks that the wolves might be killed or injured. The aliens might well be able to slaugher humans en masse, if so inclined, but we just might not be worth the trouble (until they run out of monkey people).

Finally, what we perceive as being hostile, the aliens might not. Think of 19th century Europeans rounding-up speciments for zoos (or think of the Crocodile Hunter, in a modern incarnation). They may have what they see as the best of intentions - quite understandably, we might disagree with that characterization. We might not be able to communicate with one another, but they aren't going to exterminate us just because the odd alien hunter gets gored (or gunned down).

[K]

Bob,

On your first comment: I agree. There really are aliens (like that face-eating zombie guy). The socially optimal level of defense is non-zero (but tiny in relation to the actual size of the military/police state). 

"It isn't implausible to think that our first encounter with an alien would involve an alien Columbus"

I love this tangent! 

Given that the universe is 14bn years old, it seems unlikely that we'll be technologically separated from them by a few hundred years of development. A few billion strikes me as more likely. And I don't think they'll be looking to eat us. As a good friend once pointed out, if they're looking to destroy us, a likely purpose might be to save any number of the thousands of species we annihilate on any given day, some of which might be of more interest to them than we are.

[Frances Woolley]

Actually, there has been some semi-serious analysis of alien attack in international relations, see, for example, Cynthia Weber's book on international relations theory (which contains a long-ish discussion of the morality of an Independence Day-style aggressive counter-attack) and, especially, Daniel Drezner, who has spent some time pondering topics such as How do you say "realpolitik" in Klingon".

[Wonks Anonymous]

Bryan Caplan has a pacifist argument against national defense as public good:
http://econlog.econlib.org/archives/2009/10/a_simple_proof.html

[K]

Wonks Anonymous,

Nothing new under the sun. No doubt the same argument appears in ancient texts, of which I'm unfortunately too illiterate to be aware.

[Billiam Smith]

Frances,

I don't see that any of the examples you used demonstrate that "national defence" is not a public good:

- The Diefenbunker may have only physically protected a relatively small number of government officials, but how can you argue that keeping the machinery of government running through a national crisis is not a public good? Of course the officials in the bunker might be relatively safer than people not in it, but the fact that they are enabled to continue doing their work is indeed a public good.

- Your envisioning of Myanmar-style military also doesn't do anything to counter the public good argument: the key distinction is that if a military is being used to oppress the public, it is not engaged in "national defence", it's engaged in "repression." I don't suppose anyone has ever argued that repression is a public good.

- When you argue that a military that is being used to fight floods can't also fight in Kabul or quell riots elsewhere, you're really discussing a resourcing issue, not something inherent in national defence. Of course a limited national defence resource can't respond to every emergency at the same time, but the fact that it is prepared to respond to the most urgent ones is clearly a public good. To use your preferred examples, one garbage collector can't collect garbage for a whole city, and the work of the person who cleans the bathroom disproportionately benefits the first person to use it after it has been cleaned. Does this mean these are not public goods?

- Your example of a foreign army overrunning the homes of one ethnic group but not another also doesn't demonstrate that national defence isn't a public good. Regardless of the outcome for the areas invaded, the "national defence" army is defending both ethnic groups equally. All you've proven is that invading armies aren't a public good, but again I can't imagine anyone claiming otherwise.

The key hole in your argument (as has been suggested by others) is that you conflate "national defence" with "military spending" (which may also be a error on the part of the people you're argument against). Naturally, a military can be used for all sorts of nefarious things, but if it is providing "national defence", which may happen proactively abroad as well as reactively at home, then it is indeed a public good.

[Shangwen]

My final comment on this: The US government did, early on, require every man between 18 and 45 to own a gun at his own expense. But not broccoli.

[RPLong]

FWIW, I have been to many countries with competing security forces offering defense, and they take a variety of forms: mafia, security guards, government-sanctioned militias, non-sanctioned militias, Taliban, and so on. When you look at the full cross-section of available options, government monopoly is clearly the one that offers the most stability.

Due to this inherent (social) instability, I believe society evolved into a preference for government military monopolies. It's not that they're more efficient, nor is it really that "defense is a public good." Most likely it is purely a matter of stability.

[Frances Woolley]

RPLong - Yup, there's a post to be written on the *real* reasons why defence is provided by government.