Bryan Caplan  

The Puzzling Idealism of the Geneva Convention

Falling oil prices have no imp... Farewell, Art Carden...
I recently read the 1929 Geneva Convention for the first time.  The terms are shocking.  It's almost as if a quixotic pacifist deliberately wrote the agreement to ban war by stealth.  It begins with typical high-sounding diplomatic rhetoric: 

Art. 2. Prisoners of war are in the power of the hostile Government, but not of the individuals or formation which captured them. They shall at all times be humanely treated and protected, particularly against acts of violence, from insults and from public curiosity. Measures of reprisal against them are forbidden.

But the specifics are high-minded as well.  You have to leave the prisoners virtually all of their stuff:

Art. 6. All personal effects and articles in personal use -- except arms, horses, military equipment and military papers -- shall remain in the possession of prisoners of war, as well as their metal helmets and gas-masks.
You can't feed enemy POWs any worse than your own troops, or cut rations to maintain control:

Art. 11. The food ration of prisoners of war shall be equivalent in quantity and quality to that of the depot troops.

Prisoners shall also be afforded the means of preparing for themselves such additional articles of food as they may possess.

All collective disciplinary measures affecting food are prohibited.

You can't endanger sick or wounded prisoners by transporting them (unless you invoke a seemingly huge loophole):

Art. 25. Unless the course of military operations demands it, sick and wounded prisoners of war shall not be transferred if their recovery might be prejudiced by the journey.

You can't punish any prisoner offense - including escape - with anything worse than thirty days of imprisonment.  If prisoners commit multiple offenses, their thirty-day sentences are served concurrently:

Art. 50. Escaped prisoners of war who are re-captured before they have been able to rejoin their own armed forces or to leave the territory occupied by the armed forces which captured them shall be liable only to disciplinary punishment.

Art. 52. Belligerents shall ensure that the competent authorities exercize the greatest leniency in considering the question whether an offence committed by a prisoner of war should be punished by disciplinary or by judicial measures.

Art. 54. Imprisonment is the most severe disciplinary punishment which may be inflicted on a prisoner of war.

The duration of any single punishment shall not exceed thirty days.

This maximum of thirty days shall, moreover, not be exceeded in the event of there being several acts for which the prisoner is answerable to discipline at the time when his case is disposed of, whether such acts are connected or not.

Where, during the course or after the termination of a period of imprisonment, a prisoner is sentenced to a fresh disciplinary penalty, a period of at least three days shall intervene between each of the periods of imprisonment, if one of such periods is of ten days or over.

Perhaps most impressively, the convention explicitly forbids legal weaseling:

III. General Observations

The conditions stated above must, in a general way, be interpreted and applied in as broad a spirit as possible

Taken seriously, this means that even loopholes like Article 25's "Unless the course of military operations demands it" provide little wiggle room.  If you apply prisoner protections "in as broad a spirit as possible," the obvious response to anyone who claims "the course of military operations demands the endangerment of sick and wounded prisoners" is, "Demands?!  Hardly.  There are plenty of other military options."


1. Did any of the signatories of the Geneva Convention actually intend to honor it when they signed?  

2. What did the signatories' military elites think about the Convention?  Weren't generals around the world privately rasping, "What's our top priority?  Victory - or humane treatment of our enemies?!"

3. Did the Convention's unrealistically high standards nevertheless lead to better POW treatment than realistic standards?  Or is there a kind of Laffer Curve, where higher standards can actually lead to worse treatment?

Please show your work.

COMMENTS (21 to date)
David R. Henderson writes:

One data point. When my Uncle Fred Henderson was captured as a Canadian civilian and put in a German prison camp, his Red Cross packages came like clockwork. And his wife, who was captured with him and held with the other women and children in Berlin, until they could make arrangements to get back to Canada, got to visit him in the prison camp. Indeed, because he was better fed than she, he gave her his Red Cross packages.
Also, when the Swiss came to inspect the prison camp, he told them that the German doctor he was practicing under was (my uncle was a doctor who was captured on his way to being a medical missionary in the Belgian Congo), contrary to the Geneva Conventions, asking prisoners if they were Jewish. He asked the Swiss to tell the German doctor to stop. They did and the German doctor did.

mark e writes:

It sounds like it was just an over-reaction to the still fresh ultra-horrors of WWI and anything that applied a psychological salve sounded great to everyone.

Lars P writes:

The military is generally in favor of humane treatment of prisoners, since they can become prisoners themselves.

Chris Hallquist writes:

I'm no expert, but my impression is that the Geneva Conventions were taken very goddamn seriously during much (though not all) of WWII. It probably helped that the combatants were to a significant degree on equal footing, so anyone contemplating violating the Geneva Conventions had to fear that this would lead the other side to violate them as well. It's the same reason we managed to avoid using poison gas during WWII. The fact that warfare has tended to be much more asymmetric since WWII also probably explains why the Geneva Conventions have gotten much less respect since then.

I'm also not sure the Geneva Conventions were all that pacifistic especially compared to, say, the Kellogg-Briand Pact or the Nuremberg Principles.

Tim Worstall writes:

Russia did not sign the Geneva Convention (that 1929 one, the Third). The Germans (OK, Nazis) deliberately and with intent starved to death some 3 million Soviet prisoners of war.

As above various "western" prisoners of the Germans, all from countries where both Germany and the home country had signed, were relatively well taken care of. It was considered a vast outrage, a war crime even, that 50 of those who took part in the Great Escape were shot after recapture.

To prove the connection is of course difficult:

"3. Did the Convention's unrealistically high standards nevertheless lead to better POW treatment than realistic standards? Or is there a kind of Laffer Curve, where higher standards can actually lead to worse treatment?"

But yes, there was definitely a difference in treatment between prisoners from countries that had signed and those from places that had not.

Yaakov writes:

I do not really understand the question. The Geneva convention you are quoting relates to prisoners of war. In the 20th century, providing people with food, was not that big a problem and human treatment was expected. I do not see any high standards and do not see how such high standards would have interfered in the war effort. The vast majority of prisoners of war were worthless as a means of intelligence and harsh punishment of prisoners of war would be counter productive, causing the enemy soldiers to put more effort in fighting than if they knew they would receive humane treatment when surrendering.

The Jewish soldiers in the British army who were taken captive were treated fairly by the Germans according to the Geneva convention, even when the Germans knew they were Jewish. In contrast, one of the worst failures of the red cross was its unwillingness to view the civilian Jews of the countries occupied by the Germans as deserving their protection.

Tracy W writes:

The Japanese didn't sign the Geneva Convention before WWII. Their treatment of POWs was appalling. Apparently in part this was because their philosophy was that warriors would die in battle, not surrender. As numerous Japanese soldiers did surrender their Western captors found it ridiculously easy to get military intelligence from them.

Returning to the European war, two of my grandma's brothers fought there. One was captured early on and spent the rest of the war in a German pow camp, the other was a navigator on a bomber who kept flying throughout and even when I knew him was clearly much more psychologically damaged.

Nathan W writes:

There was a time when soldiers and generals readily acknowledged that opposing combatants were generally decent human beings who were merely on the wrong side of the war.

This kind of thinking led to generals who would fight wars most violently and viciously to easily extend basic compassion to captured enemy combatants, in particular under the understanding that this should be reciprocal. Surely there were many exceptions where notable individuals received "special treatment" in both good and bad ways, but the general premise that the common soldier on the other side of the lines was most likely just as much a decent human being as you or I, once removed from the madness of war, is a historic ethic of warfare that should not be forgotten. Keep in mind that this was an age (preceding centuries) of warfare where rape and pillage went from standard practice of warfare to the easiest way to get shot by your own commander.

Wars are so asymmetric these days, that it hardly crosses the mind of American soldiers or generals (for example) that they are very likely to get caught, and so easily fall into falsehoods where their own illegal treatment of prisoners is viewed as a different can of worms altogether than whatever illegal treatment is dished out by those forces with the balls to take up arms against America in their own homelands (and please note that I don't think that their way of thinking and doing is optimal).

Hadur writes:

The fact that the Geneva Convention is not even the most idealistic international agreement to come out of the 1920's (there literally was a treaty banning war) should tell you about how idealistic that decade was. It was a lot like the 1990's, with "end of history" stuff being triumphant. I'm sure at least some signatories were willing to agree to anything, because since they didn't think another war was coming they weren't actually signing anything away, in their view.

In any event, I seem to recall there being something in the Geneva Accords protecting the prisoner's right to have astronomical instruments, which I've always found to be the most laughingly idealistic.

Nathan W writes:

Widespread knowledge of American contravention of Geneva principles in Guantanamo has a) significantly reduced the reputation of America worldwide and b) has significantly reduced the conviction of its allies to be involved in whatever wars the country may wish to start and/or participate in overseas. Hawks would make America the enemy of everyone. Do not give them too much of a voice, especially when it comes to humane treatment of captured enemy combatants.

A torturee can be made to say anything. This much has been readily recognized for centuries at least. Thus, any admissions extracted using any torturous method has ZERO credibility in any court of law of credibility (by definition, I would not consider a court to have any credibility the second it admits to accepting any evidence which has been obtained by any form of torture whatsoever, including any of various forms of mind control/brainwashing, etc.).

Thus, the methods implicitly supported by the framing of this post a) make enemies, and b) produce false intelligence.

It is not naivety, but precisely the opposite, which implies that we must uphold Geneva Convention and the like, and moreover be willing to sacrifice greatly to uphold them.

Jeffrey Miller writes:

The WWI prisoner of war camps that Renoir depicts in "La Grande Illusion" (1937) always seemed to me to be reasonably humane, given that they are prisons. Many of the details he shows seem consistent with the requirements that, as you note, were written in the Geneva Convention. I don't know how accurate Renoir's depictions are, but it seems possible that the ideas expressed in the Convention, at least as regards the treatment of prisoners of war, were already well accepted at least in principal as of WWI, prior to the Convention.

Hasdrubal writes:
  • First off, how were prisoners generally treated in the First World War? Were these provisions a response to the general treatment of prisoners? Responses to specific incidents? Or were they a codification (with perhaps some embellishments) of an already standard practice?
  • Secondly, can you model prisoner treatment with rational incentives and resource constraints? I.e, are Andersonvilles more likely to happen when you have to carefully plot your route of march based on the expected farm outputs of the territories you pass through because the logistics of getting enough food to your soldiers is your primary constraint? Is the treatment of prisoners by national armies correlated with the technology of food production and transportation? Do non-conventional forces use the mistreatment of prisoners as psychological warfare or do they do it quietly? Is the _type_ of mistreatment predictable based on their resources and military tactics? Are outliers like the Japanese and Russians in WWII really outliers, or do they fit the model?
Hadur writes:

Another thing to consider: we are approaching this as citizens of the world's most powerful nation, which fights wars against much weaker powers, captures many of its opponents fighters, and has relatively few of its own fighters captured. It is easy for us to see humane treatment of prisoners as a cost, and to therefore ask "who would sign up voluntarily for such a cost!"

The Geneva Convention was written in a multipolar world where there were many nations that regularly fought against their peers, and during these wars tens of thousands of POWs might be taken on either side. Even if you won a war against your neighbor, you might have thousands of your troops captured during that war.

So in the 1920's, people would have viewed this not necessarily as a cost, but as a benefit: the substantial number of your troops that you would expect to be captured in the next war would presumably be important to you, and you would want to guarantee that they are treated well. If that means also agreeing to treat the other guy's POWs well, then so be it.

JohnBinNH writes:

It's all explaned in Michael Walzer's Just and Unjust Wars.

As noted already, the military expect they may become prisoners and so want prisoners to be well-treated. Also, if soldiers know that prisoners are well-treated, they are more likely to surrender and thus shorten the war at less cost, so treating prisoners well is in a belligerent's own best interest.

Note the difference between multi-polar and bi-polar worlds. Actors in a multi-polar world are incented to make war a less-destructive exercise because if A and B are weakened by their war, C and D will take advantage and conquer them. So both A and B want a way to wage war which mostly gives the same results as a duel-to-the-death would but which doesn't weaken either one that much.

Hence long multi-polar periods give rise to Laws of War, prisoner conventions like 'parole' and battles that are mostly about maneuvering into a position that causes your opponent to concede without much combat.

Bi-polar worlds have different incentives leading to much worse behavior. Our recent military history is bi-polar, which explains the surprise when we look back at the past.

mico writes:

I don't fully understand your surprise. Mistreating prisoners is not especially militarily advantageous and usually both sides fear more for theur own prisoners in enemy possession than the legal niceties.

Only the USSR and Japan universally mistreated prisoners. They hadn't agreed to the Convention however.

TheRightMatch writes:

I think the general tone of these comments adequately responds to the question posed by Mr Caplan.

The most puzzling part of his post is that the provision of food, protection of that provision, and resistance to excessive punishment should ever be seen as "puzzling idealism" at all!

John writes:

I'm missing how this convetion is so pacifist in design or so excessive.

I think it's clear the empirical data tends to support the imrpoves the treatment of the prisoners (other pointed that out).

Seems to me it's largely explained by:
1) The horrors of WWI
2) Addressing a principle-agent type of problem in a competitive setting.

Jonathan Wight writes:

I’m not a huge fan of John Rawls, but having just played the Veil of Ignorance game in my class, it occurred to me that his approach might help us understand the Geneva Convention. If so, it is not so utopian at all.

Behind the veil of ignorance, if you did not know whether you would be a prisoner of war or a guard, on the winning side or the losing side, what kind of social contract would you seek to establish?

A basic set of primary goods for prisoners, including food, shelter, health care, and freedom from indiscriminate violence, may be the result of such a mind experiment.

David writes:

I grew up in a military family, and discussed this with high ranking officers. The real reason that these conventions are followed (even when fighting terrorists, for the most part) is more practical:

It is better for your opponent to know that if they surrender they will be well treated. That way, they surrender rather than fight to the death.

Unfortunately, that logic doesn't apply to religious zealots. They fight to the death anyway. So that also explains why religious zealots are sometimes treated differently than military or pseudo-military organizations.

But note that very few people were placed in situations contrary to the "Geneva Conventions". And that is even before you apply the actual Geneva Conventions, which say that the rules explicitly do not apply to non-signatory countries, nor non-military combatants.

Mo writes:

@David Like those religious zealots in Saddam's army? Per Hadur, the issue is that US soldiers have little fear of being captured, so there is less incentive to care about it. In a world where the other side knows about Gitmo and enhanced interrogation, the incentive to surrender went way down.

Gerard Merchant writes:

I applaud you for delving into the Geneva Convention, however your superficial inquiry has left you severely confused.

The Geneva Convention is a part of what is known as the "Law of Land Warfare." It is joined by numerous other conventions including the Hague Convention.

There are three fundamental principles of the Law of War:

1. Military Necessity
2. Prevention of Unnecessary Suffering
3. Distinction and Proportionality

The first principle is a trump card to the other two. Nearly anything is lawful if it can be justified by military necessity. For example, the bombing of Dresden, Hiroshima, and Nagasaki were lawful because they were industrial centers vital to the war effort. The murkiest questions come mostly from anecdotes such as when a small military unit captures civilians who could potentially divulge their undiscovered positions to the enemy.

In my opinion, the Convention achieved its goal. The belligerent parties in WWII largely abided by the Convention. The US, UK, Australia and others have demonstrated the willingness to prosecute their own soldiers for violations. Some violations, such as photographing the dead, are largely ignored because they cause no real harm. But as a whole, the Convention is often "victor's justice," especially when the victors are totalitarian regimes which we would not otherwise credit for their moral and ethical standards.

The purpose of the Law of War is to two-fold:
1. End the cycle of violence and reprisal that lead to endless wars
2. Decriminalize war

Prior to the Law of War, soldiers from another land were considered criminals. If you captured them, you could do basically anything to them. Soldiers were often taken captive and ransomed to their families, particularly if they were noblemen.

The Law of War changed that. Being a soldier, commanded by your sovereign, is not a criminal act, provided that the soldier otherwise obeys the Law of War. When you kill in war, it is NOT murder. The Law of War passes NO JUDGMENT about the legality or morality of war. When Hitler invaded Poland, the Law of War had nothing to say about it.

If you are a combatant in war, you are only considered a Prisoner of War if you:
1. Are commanded by a responsible authority (not acting independently)
2. Are wearing a distinctive sign recognizable at a distance (i.e. a uniform, not a spy)
3. Are carrying your arms openly
4. Are otherwise obeying the Law of War

Terrorists, therefore are NEVER prisoners of war. This is why the US declared people "enemy combatants." The correct term should have been "unlawful combatants," but it is their status, not what they are called, that matters.

While a prisoner of war can only be punished for 30 days for escaping, if he kills a guard while escaping, he can be tried judicially for murder. Once you are a POW, you are hors de combat and can no longer lawfully fight.

Comments for this entry have been closed
Return to top