Trade liberalisation to the World Bank (WB) and International     Monetary Fund (IMF), who have been highly influenced by the orthodox neo-liberal views, meant removing non-tariff restrictions on imports, reducing the level and spread of import tariff rates on industrial products and eliminating export taxes and subsidies.
 

This formula was thought to be applicable to all developing countries, irrespective of their level of development with the expectation that trade liberalisation would change the structure of economic incentives against import substitution industries, in favour of export activities, thus leading to outward-oriented industrialisation and development.

Even though after nearly a quarter century of implementation there is ample and growing evidence that the outward-oriented development strategies promoted by the international economic institutions under one or another guise have failed, these institutions still continue to push for liberalisation of trade in developing countries. In fact, reading the commentary by Mr Tchane gives the impression that trade liberalisation and growth of trade are not just means of achieving some developmental goals but ends in themselves. The tone of his commentary implies that either the benefits of freer and ever more trade are assumed to be unequivocal - always and everywhere - or liberalisation and more trade should be pursued for their own sake.

According to a recent scholarly publication on the matter, the general argument in favour of trade liberalisation is that it allows the expansion of the size of markets, allowing the global economy to take further advantage of the economies of scale (the argument Adam Smith put forward more than 200 years ago) and it enhances global efficiency in production and exchange. The 'factor price equalisation theorem', a very important theoretical weight bearer in neoclassical trade theory, stipulates conditions under which trade in goods and services leads to full global efficiency, substituting for the free mobility of factors.

This means that growth of trade or liberalisation might be good not for its own sake, but for the sake of improving efficiency. That efficiency improvement is global - meaning that the world as a whole benefits from more liberal trade; but that does not mean that everyone would be better-off.

Those conditions are highly restrictive and the standard argument that trade liberalisation necessarily makes all countries better off (though not necessarily all individuals within each country) rests on a set of assumptions that can hardly be satisfied in many developing countries. These assumptions include the existence of full employment, perfect competition and perfect capital and risk markets.
 

In many developing countries where unemployment is high and markets are imperfect, trade liberalisation may have different effects to those anticipated in simple theoretical models. Anyone with a bare minimum knowledge of introductory economics can easily see that markets in developing economies such as Ethiopia are grossly imperfect; and capital and risk markets are non-existent or rudimentary at best.

Those who argue for across-the-board trade liberalisation generally fail to understand that aggregation across countries and across goods and services is invalid. For one thing, developing countries differ in the products that they export and import. Thus, decisions about which goods and services to liberalise, and for which there should be restrictions on subsidies, can make a great deal of difference to the global distribution of welfare following trade liberalisation.

For example, recent review of the studies conducted to measure the welfare effects of liberalisation in different sectors concludes that agriculture, though important for developing countries, is not the only important issue, nor even the largest potential area of welfare gains for developing countries. It is by far the smallest of the average estimates across the four areas considered in the study including agriculture, services, manufacturing and labour mobility.

Notably, liberalisation of labour mobility, particularly mobility of unskilled labour, promises the greatest welfare gains. It is a great irony that, the movement of labour which promises greater economic gain for developing countries in particular is the least likely to be liberalised on the part of the developed countries in the foreseeable future. This is not because the case for liberalisation in the area is water-tight, but because the issue involves much more than economics.

Moreover, despite the presumption that agriculture is a developing country issue, the empirical studies surveyed in that study suggest that the proportion of absolute global gains accruing to developing countries is 16pc in agriculture - a far less progressive result than for the other three areas. It is another irony that much of the potential gains from liberalisation of  developing country issue  go to the developed countries themselves rather than developing countries; but at the same time, the strongest opposition for the liberalisation of the sector comes from the developed countries.
 

The second fallacy concerns the tendency of commentators to talk as if all developing countries are more or less identical or to recommend one set of policy advice for all of them. However, even among developing counties whose economy is dominated by agricultural production, we can identify distinct categories: those with subsistence agriculture; those dominated by export agriculture; and those breaking out of agriculture and increasingly centred on manufacturing.
 

Therefore, even in the field of agricultural trade, across-the-board liberalisation implies differing welfare outcomes to different developing countries. This is because the net effect of a wide-ranging agricultural reform varies across countries depending on the composition of their exports and imports of different commodities, and the price sensitivity of commodities to liberalisation. Net importers of one type of agricultural product feel the impact of liberalisation differently from net exporters, for instance.
 

Mr Tchane boldly states that "African countries would do well to reduce tariffs against the rest of the world . . ." as if the most important aspect of trade liberalisation is the average level of a country's tariff structure.
 

Tariff reduction, tariff protection or any other form of trade policy is just a means of achieving other policy objectives. In addition, tariff averages can hide very important underlying trade policy features.
 

For example, during the period in which Taiwan recorded very impressive economic growth, its tariff structure was deliberately designed to hide the fact that the government was heavily involved in implementation of interventionist industrial policies. The average tariff regime of the country appeared liberal but a closer look into the details of the country's tariff structure reveals that the government was following a fine grained approach of using tariff incentives and disincentives to promote some industries while discouraging others. It was promoting the import of inputs and machinery necessary for production of exports while discouraging non-essential imports - all the while appearing 'liberal' and 'open' on the aggregate.
 

Mr Tchane's and other similar commentators' arguments remind me of a Taoist/ Zen saying: "When the sage points to the moon, all the ignorant sees is the finger."
 

Let me elaborate my point further. As one scholar rightly argued, where the overall objectives are development, building up supply capacity and industrialisation, trade policy is a means to those ends. So are international trade, market and industrial policies, as well as foreign direct investment, and technology. Therefore, as a tool of development, trade policy is not necessarily synonymous with trade liberalisation, and success in liberalisation per se is not a guarantee of success in development.
 

In short, trade policy - not just trade liberalisation but the whole body of trade policy - should form an integral part of the overall national or continental development strategy. If the overall strategy requires that a country follows an approach of selective protection and selective liberalisation at a point in time, that is exactly what the government should try to do. If it is found that liberalisation happens to be the best policy option for a country at a specific point in time, that would be the appropriate course of action.
 

All this is because, not only trade policy but also trade itself, is just a means, not an end in itself.