The unspoken assumption on which governments act is that the best possible situation would be one in which all other countries opened up their markets, but in which they themselves remained fully protected.
This approach has a lengthy and dubious lineage. The mercantilists in the last century, and others before them, believed that exports should be encouraged, but that domestic markets should be protected from imports, even where home producers were clearly inefficient. Although this might appear to be a sensible doctrine - after all, it is surely one of the first principles of political economy that governments should seek to help their own producers - it is bound to be poisonous to efficient economic development in the long run.
Not only does the mercantilist approach assert the primacy of the producer over the consumer, but it also asserts that today's producers should be protected at the expense of tomorrow's potential producers. Free international trade obviously enables consumers to buy from the world's cheapest producers.
Less obviously, the protection of inefficient domestic producers ties up vital resources in the wrong industries, making it more difficult for the right industries to expand. As a result, they expand somewhere else, leaving heavily protected nations ossifying with the wrong industrial structure.
Of course, all this shows up only very slowly. In the shorter term - which could easily last for a decade or so - both governments and their electorates can easily convince themselves that protection is a good thing.
This is even true of trade economists, who should know better. According to the so-called 'new' trade theory of the 1980s, countries can benefit by temporarily protecting their industries, allowing capital investment to expand before their output is exposed to the full glare of foreign competition.
The new twist to these models is that they assume imperfect competition in the product markets, which is perhaps more realistic than the perfect competition assumption of old-fashioned theory. Nevertheless, I am unconvinced that the 'new' theorists have done much more than simply re-invent the well-worn 'infant industry' argument for protection.
Economists have long recognised that the key objection to this argument is that protection is usually not removed when the industry in question ceases to be an 'infant', and many modern economists who have flirted with the new trade theory (Paul Krugman, for example) have eventually decided against protection on the grounds that democratic politicians cannot be trusted to remove the protection when they should.
It is unfortunate that these new models have been used in some circles - like the Democratic Party in the United States - to provide a respectable intellectual fig-leaf for some very retrograde protectionist thinking. But of course in Geneva these past few days, the real politik of industrial pressure groups has been far more influential than the seminar rooms of Cambridge, Massachusetts, in determining the outcome. Some form of deal still looks probable, but it will be a disappointingly watered-down affair.
The Americans will probably retain the right to introduce new 'anti-dumping' measures, ostensibly to prevent other countries selling cut-price exports into the US market, but often in reality nothing more than disguised protection.
In addition, 'voluntary' export constraints agreed under duress by Pacific Rim and other developing economies will undoubtedly continue to spread. The increase in these two types of protection in the past 20 years has been immense, and it now seems that the Uruguay Round will do little to slow their insidious progress.
This may mean that the Gatt deal will have much less effect than is sometimes suggested on the industrial structure and the level of real wages in the developed countries. This is the subject of a current fad in the financial markets, which are focusing on the impact of increasing import competition from the newly industrialising countries, especially from the Pacific.
There are some good reasons for this fad. A brilliant (and so far totally unnoticed) cameo by Professor Patrick Minford in the last report of the Treasury Forecasting Panel puts the case more cogently than anywhere else. He argues that the computer and information technology revolutions mean that innovations that once took decades to filter down to the developing countries can now be transferred almost immediately.
This, and the freedom of capital movement around the world, means that multinational companies are now free to locate where costs - especially labour costs - are the lowest. The resulting migration of industry to low-cost centres such as China and the rest of emerging Asia could have profound effects on the developed economies.
For example - and this is what has fascinated the financial markets - there should be a strong tendency for real wages around the world to be forced towards greater equality (which means upwards in Asia and downwards in Europe and the US). Many international bond managers have informed me recently that they view this as the most important disinflationary force at work in the developed economies at present.
A second effect will be on the industrial structure in the West. Just as the share of agriculture in gross domestic product and employment has plummeted since the last century, it is now argued that manufacturing industry will decline in the next couple of decades.
As Professor Minford puts it: 'Poor, often ill-educated workforces can be harnessed to produce new products by use of computer-driven procedures which eliminate human error. Imported, these products displace their equivalents produced under hugely higher labour costs in the richer countries, just as cheap food from the empire displaced British-grown food after the repeal of the Corn Laws.'
There is little doubt that these effects will be felt to some extent in the next few years. They can be successfully combated only by allowing them to occur, taking advantage of the cheap imports they entail, and putting our own labour force to work in more productive new areas.
Just as it would have been futile and damaging to retain the Corn Laws, so it would now be absurd to try to prevent this new industrial revolution taking place.
However, I must end by questioning how important these new trade forces really are. While there is little or no doubt that the structure of developed economies will shift further towards services for the foreseeable future, there is considerable doubt whether it is trade with the new Asian tigers and others that is the prime cause of these changes.
Recent work by Paul Krugman indicates that such changes in the US economy are primarily triggered not by trade but by the much stronger domestic forces of shifting patterns of demand and productivity within the economy. He points out that only 2.8 per cent of US imports come from low-wage economies, about the same proportion as in 1960.
This of course further diminishes the mercantilist case for short-sighted trade measures aimed at the developing world. The mercantilists need to be defeated - and the recent experiences of the Gatt suggest it could be a long uphill struggle.