There is a tendency among practitioners to treat theoretical analysis with a measure of derision; theoreticians tend to aim for theoretical coherence regardless of practical difficulties. The most workable balance, however, lies somewhere in between. Theory untempered by pragmatism is but shallow idealism; pragmatism unguided by theory is incoherent and short-sighted. In many ways the problem of subsidies in international trade suffers from the worst of both worlds. Much analysis is done in the abstract and in the world of pure economic theory; practical negotiations and applications of international trade rules are so politicised and are conducted with such overwhelming cynicism as to mock the very idea that a coherent philosophy underlies the exercise. It probably does not. As Cantin and Lowenfeld observe with respect to the Canada-US FTA and NAFTA,
even reading one chapter of each agreement is enough to make one forget what the agreement was about, as one searches for clues to rules within rules, exceptions to exceptions, and bewildering cross references. 1
The observation is equally applicable to most other trade agreements.
There is, of course, very little use decrying the cynicism of politics if the theory itself is not realistic and does not lend itself to pragmatic application. Therefore, my concern here is with the extent to which the theory of free trade, and specifically its analysis of the role of subsidies in international trade and commerce, takes note of the realities of modern economics. For too long, misguided by abstract theory and exasperated by politics, trade negotiators have groped in