At the risk of sounding “doctrinaire” again, I would like to respond to Rodrik’s criticisms. In order of his points:
1. I am willing to admit that signaling is not perfect. Gains from trade can be left unexploited through this mechanism. However, (a) this does not mean that self-enforcing mechanisms cannot be scaled up, and (b) unexploited gains from trade are also possible when government enforces contracts instead. One might argue that the magnitude of these losses are larger under self-enforcement than under government enforcement, but once again, this very much depends upon what kind of government we’re talking about. The fact that, in theory, a perfectly knowledgeable and benevolent state that is interested only in enforcing contracts to the utmost of its ability can enable the gains from trade to be maximized is irrelevant. Let’s look at what existing governments in many parts of the world actually do. For instance, is government more of a help or a hindrance to individuals’ abilities to realize the gains from exchange in, say, Africa, or Eastern Europe? This is an empirical question the answer to which does not provide us with as much reason to be sanguine about government enforcement as Rodrik seems to suggest.
2. As a large literature points out, the very reason private international arbitration associations emerged in the first place was because international traders could not in most cases use national courts to enforce international commercial agreements. This inability stems from several factors, which have been discussed by others at length elsewhere. I will recap a few here:
First, parties to international commercial agreements are interested in avoiding the home court of the other party. Parties fear being subjected to unknown laws, having a decision rendered in an unknown language via unknown procedure, being subjected to law or procedure they disagree with or feel is inappropriate for their case, or they fear that a state court will favor their adversary if he is a citizen of that nation.
Even where this fear does not exist, there is still the problem of figuring out how to render a legitimate and binding decision consistent with the law when parties from multiple countries, each with its own separate legal system and jurisdiction, are involved. Indeed, the whole problem is that when this is the case, there is no “the law.” Which party’s national law should apply to the case? Which party’s national legal procedure should be followed? There is no “correct” answer to these questions, which is precisely why judgments rendered in such cases have no formal force.
Second, and closely related, there is an important question as to which state court, if any, has jurisdiction in the matter of an international commercial dispute. Competing claims to jurisdiction are problematic. But equally troublesome is the unwillingness of either state court to decide the dispute because neither feels that it is equipped to adjudicate an international matter.
Third, the decisions of state courts regarding matters of international commerce are difficult to enforce. In many cases state courts do not recognize foreign judgments. Even when they do it is difficult to seize the assets of the loser if he is not from the country where the court’s decision is made. If I have an international commercial dispute with someone from China and I am somehow able to convince him to have the dispute decided in America (by the way, how might I do this?) and the American court decides he owes me $75,000, how am I to go about enforcing this award? My trading partner’s assets are in China, not America. And the American court’s jurisdiction and power to enforce its will extends no further than the borders of the United States. I could rely on the Chinese government’s goodwill to send me a check. But even in the unlikely case it displays this goodwill, notice that the legal judgment of the deciding national court (in my example, America), which Rodrik places great emphasis on, has no force and thus nothing to do with me getting paid.
Rodrik’s claim that private international arbitration decisions could be regularly enforced in national courts is equally mistaken. As I pointed out in my last comment, until 1958 this was simply not possible. This is precisely the reason why the New York Convention was created—to try and make it possible to enforce international arbitration decisions in national courts. If national courts were already doing this, it is strange that the NYC was created in 1958 with this purpose. As a side note, it is peculiar to describe the NYC, which has been dubbed the “cornerstone of current international arbitration” as “esoteric.”
3. I have not argued, nor do I know of anyone who suggests, that government is always concerned with rent-seeking, corruption, etc. My point is simply that pretending that government is not often concerned with these things is mistaken and will lead to mistaken policy conclusions. The alternative assumption one could make is that government is composed of individuals who are rarely concerned with these things and are usually altruistic and well-behaved. Unless we are to think that somehow the individuals who compose government have a different nature than those in the private sector, this would mean that private sector actors are also usually altruistic and well-behaved. But if this is so, we return, albeit in a different context, to the question posed in my initial essay: Why do we need government?
 van den Berg, Albert Jan (1981). The New York Arbitration Convention of 1958. Deventer: Kluwer Law, p. 1.