Qualifications and Assumptions

The conclusion that free trade is universally beneficial is a rather bold one to draw from such a simple model. Our simple model includes many unrealistic assumptions:

1. We have assumed a simple world in which there are only two countries and two goods. In the real world, there are many countries and many goods.

2. We have assumed away transportation costs between countries.

3. We have assumed away differences in the prices of resources in different countries. We have said nothing about exchange rates and simply assumed that cocoa and rice could be swapped on a one-to-one basis.

4. We have assumed that while resources can move freely from the production of one good to another within a country, they are not free to move internationally. In reality, some resources are somewhat internationally mobile. This is true of capital and, to a lesser extent, labor.

5. We have assumed constant returns to scale; that is, that specialization by Ghana or South Korea has no effect on the amount of resources required to produce one ton of cocoa or rice. In reality, both diminishing and increasing returns to specialization exist. The amount of resources required to produce a good might decrease or increase as a nation specializes in production of that good.

6. We have assumed that each country has a fixed stock of resources and that free trade does not change the efficiency with which a country uses its resources. This static assumption makes no allowances for the dynamic changes in a country's stock of resources and in the efficiency with which the country uses its resources that might result from free trade.

7. We have assumed away the effects of trade on income distribution within a country.

Given these assumptions, can the conclusion that free trade is mutually beneficial be extended to the real world of many countries, many goods, transportation costs, volatile exchange rates, internationally mobile resources, nonconstant returns to specialization, and dynamic changes? Although a detailed extension of the theory of comparative advantage is beyond the scope of this book, economists have shown that the basic result derived from our simple model can be generalized to a world composed of many countries producing many different goods.⁶ Despite all of the shortcomings of the Ricardian model, research suggests that the basic proposition that countries will export the goods that they are most efficient at producing is borne out by the data.⁷ However, once all the assumptions are dropped, the case for unrestricted free trade, while still positive, has been argued by some economists associated with the "new trade theory" to lose some of its strength.⁸ We return to this issue later in this chapter and in the next.

Simple Extensions of the Ricardian Model

Let us explore the effect of relaxing two of the assumptions identified above in the simple comparative advantage model. Below we relax the assumption of constant returns to specialization and the static assumption that trade does not change a country's stock of resources or the efficiency with which it utilizes those resources.

Diminishing Returns

The simple comparative advantage model developed in the preceding subsection assumes constant returns to specialization. By constant returns to specialization, we mean that the units of resources required to produce a good (cocoa or rice) are assumed to remain constant no matter where one is on a country's production possibility frontier (PPF). Thus, we assumed that it always took Ghana 10 units of resources to produce one ton of cocoa. However, it is more realistic to assume diminishing returns to specialization. Diminishing returns to specialization occur when more units of resources are required to produce each additional unit. Whereas 10 units of resources  Figure 4.3

Ghana's PPF under Diminishing Returns

may be sufficient to increase Ghana's output of cocoa from 12 tons to 13 tons, 11 units of resources may be needed to increase output from 13 to 14 tons, 12 units to increase output from 14 tons to 15 tons, and so on. Diminishing returns imply a convex PPF for Ghana (see Figure 4.3), rather than the straight line depicted in Figure 4.2.

There are two reasons why it is more realistic to assume diminishing returns. First, not all resources are of the same quality. As a country tries to increase output of a certain good, it is increasingly likely to draw on more marginal resources whose productivity is not as great as those initially employed. The end result is that it requires more resources to produce an equal increase in output. For example, some land is more productive (fertile) than other land. As Ghana tries to expand its output of cocoa, it might have to utilize increasingly marginal land that is less fertile than the land it originally used. As yields per acre decline, Ghana must use more land to produce one ton of cocoa.

A second reason for diminishing returns is that different goods use resources in different proportions. For example, imagine that growing cocoa uses more land and less labor than growing rice, and that Ghana tries to transfer resources from rice production to cocoa production. The rice industry will release proportionately too much labor and too little land for efficient cocoa production. To absorb the additional resources of labor and land, the cocoa industry will have to shift toward more labor-intensive production methods. The effect is that the efficiency with which the cocoa industry uses labor will decline; and returns will diminish.

The significance of diminishing returns is that it is not feasible for a country to specialize to the degree suggested by the simple Ricardian model outlined earlier. Diminishing returns to specialization suggest that the gains from specialization are likely to be exhausted before specialization is complete. In reality, most countries do not specialize, but produce a range of goods. However, the theory predicts that it is worthwhile to specialize until that point where the resulting gains from trade are outweighed by diminishing returns. Thus, the basic conclusion that unrestricted free trade is beneficial still holds, although due to diminishing returns, the gains may not be as great as suggested in the constant returns case.