The sources of international law

by Henry Farrell on August 23, 2007

As an international relations scholar (sort of; I began in comparative politics, but gradually shuffled sideways into IR) who believes that international law can be a meaningful constraint on state action, I’m somewhere between “Dan”: and “John”: on the question of whether the US should (or should want to be) bound by international law. The core insight of international relations is that international politics differs from domestic politics because there isn’t any actor with a monopoly on the use of legitimate violence to enforce the law. Thus, whatever international law there is flows from states or from organizations created by states. This doesn’t mean that international law doesn’t exist or that international law can’t have some degree of relative autonomy from states (international organizations aren’t perfect agents of states, and have some wriggle-room to shape law in ways that states might not initially have intended). It does mean that international law is fundamentally limited by the willingness or unwillingness of states to enforce it, except under relatively unusual circumstances (such as the European Union). However, within these limits, quite a lot is possible.

First of all though, it’s important to note that _even if_ you assume that states are rational, self-interested actors, there’s no _ex ante_ reason to believe that they won’t obey the law. As Chris “notes”:, Megan McArdle’s “argument”: collapses if you relax the assumption that states are simple self-interested utility maximizers. But you don’t even need to do that – there is a whopping big missing middle in McArdle’s argument between the “states are self-interested rational actors” and the “therefore international law won’t stop them invading countries like Iraq when it’s in their perceived interest to do so” bits. It may be perfectly within states’ rational self-interest to create international law, and to abide by it, even when it isn’t in their immediate self interest to do so. As Thomas Schelling famously remarked, the right to be sued is a very valuable right indeed. Actors that are completely unconstrained are _ipso facto_ not able to give credible commitments to others. This actually limits their effective ability to get things done in a world where there are other important players, and in particular it limits their ability to create and participate in the nice panoply of institutions that the US and its allies built up after WWII [update: “Brad DeLong”: makes much the same point]. We may reasonably expect that states are going to abjure these commitments when vital issues of national survival are at stake – but such issues weren’t really on the table in Iraq, for all the guff about mushroom clouds and smoking guns. This doesn’t mean that states such as the US are always going to obey international law, but it does mean that their compliance or non-compliance doesn’t flow in any simple or obvious way from their narrow self-interest.

Furthermore, if institutions have some degree of partial autonomy from the states that created them, then it is likely to be in the self-interest of powerful states to build such institutions in order to ensure that they can shape the rules of the game in ways that will persist after their own effective power has dwindled. This is the argument that John Ikenberry, a historical institutionalist, makes in “After Victory” (“Powells”:, “Amazon”: He claims that the US pushed for the creation of the various institutions that came to be foundations of the post WWII international economy not only in order to bind itself, and thus make it easier for its allies to trust it, but also to ensure that the game was played according to the rules that the US wanted, even if the relative power of the US came to diminish at some future point in time.

Finally, as Chris has said, it may very well be that actors’ interests and self-perceptions are much more malleable than standard economic theory would give them credit for being (the experimental research that he points to is getting considerable attention from economists these days). This means that international institutions and laws may be important not so much because they allow states to make long term commitments, nor even because they crystallize power relations over time, as because they may reshape states’ interests so that the logic of the international system is transformed over time. The classic statement of this case within the IR literature is Alex Wendt’s article “Anarchy Is What States Make It”:;2-9. Perhaps the best case of this happening in real life is the European Union. It’s inarguable that states’ interests have been transformed within the EU – it is more or less _inconceivable_ that France will go to war with Germany. EU law, insofar as it is international law, has also had a transformative effect. Even though the European Court of Justice has no armed forces to back up its decisions, they have been accepted by all member states even when they’re highly inconvenient for them. The reason, as Karen Alter and others discuss, is that domestic courts have come to accept EU law as binding, and to integrate the ECJ into their own networks of judicial decision making so that a weird kind of interpenetration has occurred between international and domestic law. It is worth mentioning, however, that the same is not true, say, of the WTO.

What this suggests is that under a _variety_ of different assumptions about the sources of actors motivations, including the standard economistic/rational choice assumptions, international law can have binding consequences, even if it flows solely from states and the actions of states. Thus, while there is controversy over the claim that international law has binding consequences, there are a lot of distinguished IR scholars who believe that it does (just as there are a lot who believe it doesn’t). The empirical evidence is mixed, providing some support for both points of view. ‘

There is also one final twist that Dan refers to in passing – much of your thinking on this is likely to be dictated by your thinking on the sources of state interests. Structural realists have argued, following Kenneth Waltz’s version of Durkheimian theory, that state interests are derived from the anarchic logic of the international system, and that domestic politics has little role to play. This approach gives short shrift to institutions – but it isn’t anywhere near as dominant as it used to be. Other IR scholars, including both Dan and I, seek to locate state interests and preferences in domestic as well as international factors – but there isn’t any very satisfactory means of determining where domestic interests come from. This is unfortunate, because much of the action in compliance/non-compliance with international institutions and international law flows from these interests and how they get translated into state preferences.

Take, for example, this “story”: in the _NYT_ today. The US faces an adverse judgement from the WTO over its domestic gambling laws, which it has been trying, without success, to wriggle out of. The problem that it faces is twofold. If it refuses to comply with the ruling, it will demonstrate its unwillingness to live up to international commitments that it has made, and will also set a dangerous precedent for countries like China, which may be similarly disinclined to comply with awkward rulings. If it does comply, it will have to push through substantial changes in US law that are going to be extremely difficult, and perhaps impossible, to get through Congress. On the one hand, the US’s rational long term interest as a unitary state and economy arguably lies in complying (the economic benefits that it receives from a trading system based on WTO rules are substantial, and it doesn’t want to endanger the system). On the other hand, specific domestic interests are going to get hurt – and these interests are powerful enough that they may have an effective veto power. This suggests the _specifics_ of compliance and non-compliance with international law, the circumstances under which states will comply against their short term interest, and the circumstances under which they won’t, often depend on domestic political factors that international relations scholars don’t understand well at all.

Update: it occurs to me that much of what I am trying to say above can be summarized as follows. The argument that states will only adhere to international law when it is in their interests to do so is a highly plausible one. But without a generally convincing account of what those interests are, it is not very helpful. We don’t have such a generally convincing account, nor are likely to do anytime soon. Instead, we have a number of competing accounts (state interests flow from the structure of the international system, state interests flow from those of powerful domestic interest groups, state interests flow from the normative structures that they are embedded in etc). Each of these can plausibly be said to capture part of the truth, but none seems like a viable candidate to provide a generally applicable theory any time soon.



John F. Opie 08.23.07 at 5:59 pm

Hi –

Part of what you are saying points to where the real problems lies: we don’t have a generally convincing account of where states interests lie, because we do not have a political body where this would be debated and then codified into the appropriate laws that would then be both interpreted and enforced.

In order to have the kind of international law that is not dependent on some sort of compliance-when-we-want-to-but-not-when-it-isn’t-what-we-want you will need to create a world government with legislative, executive and judicial components.

Until that happens, you can’t change the way that states will interact with each other, since there is no meaningful debate. If anything, the discussion of international “laws” without the accompanying civilizational parameters as outlined per ut supra means that they end up like the Kyoto Protocol, toothless and ultimately meaningless, if not even negative: if nations can flaunt the law, with impunity, then the law is not worth much.

That is the dilemma: without robust institutions, there will be no robust system of law.


Keith M Ellis 08.23.07 at 6:12 pm

This is a great post. I don’t have anything to add—I just wanted to say thanks for such a substantive post.


David 08.23.07 at 7:40 pm

Let me add my thanks, and express my gratitude. I had feared that this issue was starting to degenerate into name-calling, but, you, John, and Dan have cooperatively turned it into a really substantive and enlightening discussion, without compromising your very real differences–an excellent example of what the blogosphere *could* be.


nick s 08.23.07 at 8:08 pm

[this is good]

I’m particularly sympathetic towards the Ikenberry argument, not least because it provides a framework to judge the contrasting fates of the League of Nations and the UN.

On John’s point @1, is it too much of a nuance to say that robust institutions aren’t as important as what you might call ‘robust institutionalism’ from states? That is, it’s better to have an international framework where violations are at least acknowledged as such, rather than repudiated or treated as if they don’t exist.


Timothy Roscoe Carter 08.23.07 at 8:44 pm

john f. opie is right. And the U.S. would be better off to help create a world government because is could free itself of the domestic obligations in areas where sovereignty has been ceded to the world federation or confederation.

E.U. law at this point can no longer be seen as international law, but as the law of a confederation of states. The nations in the E.U. have effectively ceded their sovereignty in certain areas to the E.U., but the E.U. is accountable to the states, rather than to the people of Europe, as would be the case in a federation like the United States. This is my main problem with the U.S. following WTO rulings: the lack of democratic accountability for the institution. This could also be solved with a world government like Mr. opie suggests.

Timothy Roscoe Carter
Program Officer
Democratic World Federalists


Nathanael Nerode 08.23.07 at 10:21 pm

“Actors that are completely unconstrained are ipso facto not able to give credible commitments to others.”

Excellent point. The “completely unconstrained” behavior of the Bush administration — unconstrained by law, Constitution, international norms, or civilized practice — is why they are now completely incapable of making any credible commitments whatsoever. This is why the word of the US President is now worth less than nothing.


Quo Vadis 08.23.07 at 11:14 pm

It seems to me that the central issue here is accountability. To whom, to what degree and how should these international institutions be accountable? If you assume that these institutions will be run by enlightened technocrats and that the governments of the states are influenced by generally negative, self-interested forces, then institutions that are less accountable may seem preferable.

On the other hand, powerful institutions that are not accountable to those over whom they exercise authority tend to be prone to corruption and appeal even more to who enjoy exercising power than to those who want to do good. In fact, those two characteristics are typical of history’s most oppressive institutions.


Brett Bellmore 08.23.07 at 11:52 pm

‘The core insight of international relations is that international politics differs from domestic politics because there isn’t any actor with a monopoly on the use of legitimate violence to enforce the law.”

Is Webber’s conception of the state so uncontroversial that you can just assume it like that? There’s no actor with a monopoly on the use of legitimate violence in American domestic politics, either… As a matter of both tradition and constitutional law. That’s the old order for the ages.


Paul 08.23.07 at 11:52 pm

I think there’s something to the notion of international (or regional) ideals “interpenetrating” the domestic political institutions which ultimately define “national interest”.

I’m less persuaded by the argument that international law can flower because States can benefit from either making a binding pre commitments to act “legally” or from the compliance of others. Sure, both of these benefits exist, but they fall apart in a world where compliance with the law is commonly understood to be the result of a case-by-case cost benefit analysis.

In order for a pre commitment to be sufficiently binding it must be the case that I bind myself to abide by it even where it is ex post not in my long term interest to do so. Something like domestic contract law provides that commitment, because it provides a role for external sanctions, but international law looks like nothing more than a prearranged language in which to make promises backed only by reputational concerns.

Similarly, my decision to abide by a set of norms because I will benefit if others do likewise is, in a world of common knowledge of self interest, going to do nothing to achieve such compliance. Why should China, seeing me observe a rule because I will benefit if China does, change its intended behaviour? The argument only works if we presuppose some states (or some parts of states) are irrational in the sense that observing self-interested compliance will sway them toward unselfish compliance.

This whole part of the argument is circular, in that it tries to bootstrap international law into existence on the same argument for self interest which ensures in won’t dictate the outcomes of the hard cases in which it will be tested.


Brett Bellmore 08.23.07 at 11:59 pm

Anyway, I think the fundamental problem of international law is that too high a percentage of the “actors”, the states, are bad apples for all inclusive institutions to produce notably defensible results. The UN as it’s presently instituted is like giving the corner crack house a significant role in the neighborhood watch program. You can see this in real world examples, like the composition of the human rights commission.


Henry 08.24.07 at 12:31 am

Paul – I don’t think there is any very good reason why reputational factors can’t serve as an anchor for binding pre-commitments. See, for example, the David Kreps piece in Alt and Shepsle’s Perspectives on Positive Political Economy. But then, I interpret a credible commitment, perhaps idiosyncratically, as a promise that it is in my interest to keep, once I have made it (I’ll say in my defence that Russell Hardin, who’s a reasonably credible person on these issues, seems to like this definition).


paul 08.24.07 at 1:34 am


I don’t disagree – you can have a system of commitments based on reputation – I simply think that, once that’s the mechanism you’re relying on, the concept of “international law” doesn’t really have much work to do – except, as I suggested, as a language in which to phrase your rational (taking reputation effects into account) promises.

Maybe that’s a quibble based on reading too much into your specific invocation of “international law” rather than “international constraints on behaviour” or some other, broader notion of rational restraint by state actors.


Matt 08.24.07 at 3:19 am

It’s worth pointing out, I think, that WTO decisions are almost always complied with, even when there are strong domestic groups opposing such measures. Compliance is far from perfect, of course, but it’s much more common than not.


Ragout 08.24.07 at 4:09 am

Henry’s claim seems to be that countries will keep their commitments because they hope that others will reciprocate if they do and retaliate if they don’t. That’s fairly plausible in low-stakes, repeated interactions like commercial relations.

But what does this have to do with institutions? No country needs the UN in order to commit themselves not to wage aggressive war, they can just promise not to wage aggressive war. That promise is no more or less credible than signing the UN Charter. This theory doesn’t seem to give institutions anything in particular to do.


stuart 08.24.07 at 5:26 am

But what does this have to do with institutions? No country needs the UN in order to commit themselves not to wage aggressive war, they can just promise not to wage aggressive war. That promise is no more or less credible than signing the UN Charter. This theory doesn’t seem to give institutions anything in particular to do.

Surely the point of these institutions is they tend to turn lots of small bilateral agreements and conflicts and bring them into a multilateral system – whereas before the WTO lots of countries would have their tit for tat protectionist exchanges, once most people join up any action against one member tends to attract more attention from third parties – two third parties breaking agreements held between themselves only is (at the very least psychologically) different to breaking an agreement that you yourself are part of.

Equally these organisations also group things together – would Antigua be able to put anywhere near the amount of pressure on the US regarding internet gambling if it wasn’t for the WTO and the threat of them legally being able to ignore all american copyrights etc.


Kevin Donoghue 08.24.07 at 8:31 am


Didn’t the Cuban missile crisis provide an example of what institutions can do? The UN Secretary General called on both sides to back off and they did. Sure, Khruschev could have made the same decision without the UN but it would have been harder for him to do so. Given the stakes it’s not a trivial example.


Ragout 08.24.07 at 1:52 pm

I’m not so much denying that institutions can matter, as I am pointing out that you need to spell out what exactly it is that institutions do. In the case of the WTO, providing neutral judges is obviously important, if only for rallying world opinion.

Once you spell out the role of institutions, it’s easier to see their limitations. I’ve never heard of the Secretary-General’s role in the Cuban missile crisis, and I’m skeptical that he had any impact. For one thing, I don’t see how the UN Charter gave him any authority in the situation. As to Antigua, they don’t need the WTO to give them permission to ignore US copyrights, they can choose to ignore them whenever they please.


Kevin Donoghue 08.24.07 at 3:16 pm

I’ve never heard of the Secretary-General’s role in the Cuban missile crisis, and I’m skeptical that he had any impact.

Maybe the two Ks could have sorted it out without the UN and then again maybe not, but the fact is that the deal which U Thant proposed was the deal that was done:

“These proposals are essentially identical to the agreement which RFK, speaking for the president, would propose to the Soviet ambassador late on 10/27 and Khrushchev would accept early on 10/28. However, the understanding concerning the removal of US missiles from Turkey would remain secret for several decades and was not included in the public announcement on 10/28.”


anotherirprof 08.24.07 at 8:17 pm

There is no evidence that the United States built institutions to restrain itself or because it believed it would decline. Ikenberry is simply wrong in this. The United States built Bretton Woods to tear down barriers. It built NATO, an alliance, because the French demanded it. At no stage did anyone say, “boy, it would be a good idea to create these things because otherwise we’ll be too scary to others”. After Victory is just cheap assertion without proof. So why repeat the canard. It’s misleading and causes confusion when we discuss multilateralism today. Folks like Ikenberry say our objective should be to restrain ourselves. Like that’s going to sell? The only sellable argument for multilateralism and law is that it will enable us to do things that we could not do alone.


James 08.24.07 at 8:38 pm

International Law lacks the emotional resonance within the United States that it has with other first world countries. This is largely due to the willingness of political actors to conflate non-binding statements with actual ratified treaties in an effort to establish these positions as domestic law. This type of action causes all International Law to be looked on as suspect.


Henry 08.24.07 at 9:38 pm

anotherirprof – I’m not an expert on the post-Cold War settlement, but I did think that Ikenberry made a convincing sounding case. If he didn’t, isn’t there room for a useful corrective article? And I think that the fundamental point is that “restraining ourselves” and being enabled to “do things that we could not do alone” can go hand-in-hand (this, as I think I mentioned, is less Ikenberry than Schelling). This said, I am convinced that another of the famous self-restraint stories – the North/Milgrom/Weingast one isn’t anywhere near as purposive as the authors make out (Terry Moe makes this case in a very good Perspectives on Politics piece) and am willing to be convinced on the facts …

paul – one day I hope to write an article which will set out better my ideas about the relationship between reputation and institutions (including laws). Short version: I think that standard game theoretic models of reputation leave out the interesting stuff (Kreps is a partial exception here – but he doesn’t try to formalize his most important insights). For reputational arguments about sanctioning etc to work, there have to be common understandings about what does, or doesn’t constitute a “bad act.” Institutions provide these understandings. Adjudicatory institutions further provide a realm of authoritative interpretation and fact finding that can cement reputational effects. But the key is that I think that without the kinds of common understandings generated by either informal institutions or formal ones such as laws, reputational mechanisms won’t work. This mostly gets left out of the famous work by Greif etc etc.


Ragout 08.25.07 at 4:04 am

For reputational arguments about sanctioning etc to work, there have to be common understandings about what does, or doesn’t constitute a “bad act.” Institutions provide these understandings. Adjudicatory institutions further provide a realm of authoritative interpretation and fact finding that can cement reputational effects.

Henry, These seem like very good points, but if they were in your original post, I didn’t see them.

The role of a neutral arbitrator seems important in trade disputes, or other relatively important matters, but less so in matters of war and peace. You sort of said that international law isn’t going to matter for really important decisions, but you also said that international law can play a role in deterring events like the Iraq War.

Kevin also says that international institutions have an important effect on events more significant than trade disputes. He links to a document, claiming that it shows that UN Secretary-General U Thant “proposed” a deal in the Cuban missile crisis, a deal that the two sides later accepted. But actually, the document says that U Thant just passed on a Soviet proposal that RFK had already discussed with the Soviet ambassador. It’s very explicit that he was just reporting a Soviet proposal: “[UN Secretary General U Thant] also adds that the other side may ask the US to dismantle the missile sites in Turkey and Italy as part of a settlement.”


Kevin Donoghue 08.25.07 at 1:59 pm

Ragout, the conversation between Stevenson and JFK takes place on Oct. 26th and refers to earlier discussions between Stevenson and U Thant. My impression is that RFK’s discussions with the Soviet ambassador begin later that day and conclude on 27th. Maybe I’m wrong about that but I certainly don’t see where “the document says that U Thant just passed on a Soviet proposal that RFK had already discussed with the Soviet ambassador.” He isn’t “just reporting a Soviet proposal”, he is trying to broker a deal; it’s not the same thing.

However by that time it was quite clear that the Soviets wanted a deal, so that part of the story is not as crucial to the point I’m trying to make as the events which led up to it. U Thant’s most important moves were on Oct 24th, when he called for a moratorium on military action, and Oct 25th, when he appealed to Khrushchev to stop his ships short of the quarantine line. (Khrushchev agreed on 26th.) My impression is that the Americans suggested to U Thant that he make that second request, since it would be easier for Khrushchev to accept it coming from him. That’s the important point: the existence of an internationally respected broker gave Khrushchev a way to back down with dignity. Maybe he would have backed down anyway, we’ll never know.

We may be wandering away from the topic of Henry’s post. What prompted my first comment was this remark of yours: “No country needs the UN in order to commit themselves not to wage aggressive war, they can just promise not to wage aggressive war.” It’s not that simple. In a showdown what governments often need most is a way to get out of a corner where they have been trapped by their own rhetoric, domestic opinion etc.; institutions like the UN can come in very handy in such cases. You say that “a neutral arbitrator seems important in trade disputes, or other relatively [un?]important matters, but less so in matters of war and peace” but actually I think the reverse is true. People expect their leaders to compromise on commercial matters, so they are likely to do so in any case, with or without an arbitrator. The pressure to “take a stand” is greater where war is concerned. So states do have an interest in promoting institutions dedicated to finding peaceful solutions.


engels 08.25.07 at 11:58 pm

James’ #20 is a nice example of a form of argument so common among American conservatives and centrists that it would be nice to have a name for it.

“Laudable project X enjoys far less popular support in the US than it does in other civilised countries and this is the fault of… its deeply marginalised supporters within the US.”

A bit counterintuitive perhaps, but still…

Comments on this entry are closed.