« Is Reform Possible in France? | Main | Time to Build »

Thursday, April 27, 2006

The Political Economy of Trade Policy

For those of you either opposed to or in favor of free trade, I thought I would try to help the debate along by looking at the welfare and political arguments for and against open borders. This material is from Chapter 9 of Krugman and Obstfeld's International Economics, "The Political Economy of Trade Policy." For example, after talking about the advantages of free trade, the chapter states:

Although economists often argue that deviations from free trade reduce national welfare, there are, in fact, some theoretical grounds for believing that activist trade policies can sometimes increase the welfare of the nation as a whole. ... We need to realize that economic theory does not provide a dogmatic defense of free trade, something that it is often accused of doing.

Another reason for presenting this material is to give a better sense of how economists approach these questions and present them in undergraduate classes, and to show that we may not be as closed-minded on the trade issue as many of you assume. Here are some arguments for and against free trade:

Chapter 9 The Political Economy of Trade Policy

In 1981 the United States asked Japan to limit its exports of autos to the United States. This raised the prices of imported cars and forced U.S. consumers to buy domestic autos they clearly did not like as much. While Japan was willing to accommodate the U.S. government on this point, it was unwilling to do so on another - a request that Japan eliminate import quotas on beef and citrus products - quotas that forced Japanese consumers to buy incredibly expensive domestic products instead of cheap imports from the United States. The governments of both countries were thus determined to pursue policies that, according to the cost-benefit analysis developed in Chapter 8, produced more costs than benefits. Clearly, government policies reflect objectives that go beyond simple measures of cost and benefit.

In this chapter we examine some of the reasons governments either should not or, at any rate, do not base their policy on economists' cost-benefit calculations.

The examination of the forces motivating trade policy in practice continues in Chapters 10 and 11, which discuss the characteristic trade policy issues facing developing and advanced countries, respectively.

The first step toward understanding actual trade policies is to ask what reasons there are for governments not to interfere with trade-that is, what is the case for free trade? With this question answered, arguments for intervention can be examined as challenges to the assumptions underlying the case for free trade.

The Case for Free Trade

Few countries have anything approaching completely free trade. The city of Hong Kong, which is legally part of China but maintains a separate economic policy, may he the only modern economy with no tariffs or import quotas. Nonetheless, since the time of Adam Smith economists have advocated free trade as an ideal toward which trade policy should strive. The reasons for this advocacy are not quite as simple as the idea itself. At one level. theoretical models suggest that free trade will avoid the efficiency losses associated with protection. Many economists believe that free trade produces additional gains beyond the elimination of production and consumption distortions. Finally, even among economists who believe free trade is a less than perfect policy, many believe free trade is usually better than any other policy a government is likely to follow.

Free Trade and Efficiency

The efficiency case for free trade is simply the reverse of the cost-benefit analysis of a tariff. Figure 9-1 shows the basic point once again for the case of a small country that cannot influence foreign export prices. A tariff causes a net loss to the economy measured by the area of the two triangles; it does so by distorting the economic incentives of both producers and consumers. Conversely, a move to free trade eliminates these distortions and increases national welfare.


A number of efforts have been made to add the total costs of distortions due to tariffs and import quotas in particular economies. Table 9-1 presents some representative estimates. It is noteworthy that the costs of protection to the United States are measured as quite small relative to national income. This situation reflects two facts: (1) the United States is relatively less dependent on trade than other countries and (2) with some major exceptions, U.S. trade is fairly free. By contrast. some smaller countries that impose very restrictive tariffs and quotas are estimated to lose as much as 10 percent of their potential national income to distortions caused by their trade policies.

Table 9-1 - Estimated Cost of Protection,
as a Percentage of National Income
Brazil (1966) 9.5
Turkey (1978) 5.4
Philippines (1978) 5.4
United States(1983) 0.26

Sources: Brazil: Bela Balassa, The Structure of Protection in Developing Countries (Baltimore: The Johns Hopkins Press, 1971): Turkey and Philippines, World Bank. The World Development Report 1987 (Washington: World Bank, 1987): United States: David G. Tarr and Morris E. Morkre, Aggregate Costs to the United States of Tariffs and Quotas on Imports (Washington D.C.: Federal Trade Commission, 1984).

Additional Gains From Free Trade1

There is a widespread belief among economists that calculations of the kind reported in Table 9-I, even though they report substantial gains from free trade in some cases, do not represent the whole story. In small countries in general and developing countries in particular, many economists would argue that there are important gains from free trade not accounted for in conventional cost-benefit analysis.

One kind of additional gain involves economies of scale. Protected markets not only fragment production internationally, but by reducing competition and raising profits, they also lead too many firms to enter the protected industry. With a proliferation of firms in narrow domestic markets, the scale of production of each firm becomes inefficient. A good example of how protection leads to inefficient scale is the case of the Argentine automobile industry, which emerged because of import restrictions. An efficient scale assembly plant should make from 80,000 to 200,000 automobiles per year, yet in 1964 the Argentine industry, which produced only 166,000 cars, had no less than 13 firms! Some economists argue that the need to deter excessive entry and the resulting inefficient scale of production is a reason for free trade that goes beyond the standard cost-benefit calculations.

Another argument for free trade is that by providing entrepreneurs with an incentive to seek new ways to export or compete with imports, free trade offers more opportunities for learning and innovation than are provided by a system of "managed" trade, where the government largely dictates the pattern of imports and exports. Chapter 10 discusses the experiences of less-developed countries that discovered unexpected export opportunities when they shifted from systems of import quotas and tariffs to more open trade policies.

These additional arguments for free trade are for the most part not quantified. In 1985, however, Canadian economists Richard Harris and David Cox attempted to quantify the gains for Canada of free trade with the United States, taking into account the gains from a more efficient scale of production within Canada. They estimated that Canada's real income would rise by 8.6 percent-an increase about three times as large as the one typically estimated by economists who do not take into account the gains from economies of scale.2

If the additional gains from free trade are as large as some economists believe, the costs of distorting trade with tariffs, quotas, export subsidies, and so on are correspondingly larger than the conventional cost-benefit analysis measures.

Political Argument For Free Trade

A political argument for free trade reflects the fact that a political commitment to free trade may be a good idea in practice even though there may be better policies in principle. Economists often argue that trade policies in practice are dominated by special-interest politics rather than consideration of national costs and benefits. Economists can sometimes show that in theory a selective set of tariffs and export subsidies could increase national welfare, but in reality any government agency attempting to pursue a sophisticated program of intervention in trade would probably be captured by interest groups and converted into a device for redistributing income to politically influential sectors. If this argument is correct, it may he better to advocate free trade without exceptions, even though on purely economic grounds free trade may not always be the best conceivable policy.

The three arguments outlined in the previous section probably represent the standard view of most international economists, at least in the United States:

  1. The conventionally measured costs of deviating from free trade are large.
  2. There are other benefits from free trade that add to the costs of protectionist policies.
  3. Any attempt to pursue sophisticated deviations from free trade will be subverted by the political process.

Nonetheless, there are intellectually respectable arguments for deviating from free trade, and these arguments deserve a fair hearing.

National Welfare Arguments Against Free Trade

Most tariffs, import quotas, and other trade policy measures are undertaken primarily to protect the income of particular interest groups. Politicians often claim, however, that the policies are being undertaken in the interest of the nation as a whole, and sometimes they are even telling the truth. Although economists often argue that deviations from free trade reduce national welfare, there are, in fact, some theoretical grounds for believing that activist trade policies can sometimes increase the welfare of the nation as a whole.

The Terms of Trade Argument for a Tariff

One argument for deviating from free trade comes directly out of cost-benefit analysis: For a large country that is able to affect the prices of foreign exporters, a tariff lowers the price of imports and thus generates a terms of trade benefit. This benefit must be set against the costs of the tariff, which arise because the tariff distorts production and consumption incentives. It is possible, however, that in some cases the terms of trade benefits of a tariff outweigh its costs, so there is a terms of trade argument for a tariff.

The appendix to this chapter shows that for a sufficiently small tariff the terms of trade benefits must outweigh the costs. Thus at small tariff rates a large country's welfare is higher than with free trade (Figure 9-2). As the tariff rate is increased, however, the costs eventually begin to grow more rapidly than the benefits and the curve relating national welfare to the tariff rate turns down. A tariff rate that completely prohibits trade (tp in Figure 9-2) leaves the country worse off than with free trade; further increases in the tariff rate beyond  tp, have no effect, so the curve flattens out.


At point 1 on the curve in Figure 9-2, corresponding to the tariff rate t0, national welfare is maximized. The tariff rate t0, that maximizes national welfare is the optimum tariff. (By convention the phrase optimum tariff is usually used to refer to the tariff justified by a terms of trade argument rather than to the best tariff given all possible considerations.) The optimum tariff rate is always positive but less than the prohibitive rate (tp) that would eliminate all imports.

What policy would the terms of trade argument dictate for export sectors? Since an export subsidy worsens the terms of trade, and therefore unambiguously reduces national welfare, the optimal policy in export sectors must be a negative subsidy, that is, a tax on exports that raises the price of exports to foreigners. Like the optimum tariff, the optimum export tax is always positive but less than the prohibitive tax that would eliminate exports completely.

The policy of Saudi Arabia and other oil exporters has been to tax their exports of oil, raising the price to the rest of the world. Although oil prices fell in the mid-1980s, it is hard to argue that Saudi Arabia would have been better off under free trade.

The terms of trade argument against free trade has some important limitations, however. Most small countries have very little ability to affect the world prices of either their imports or other exports, so that the terms of trade argument is of little practical importance. For big countries like the United States, the problem is that the terms of trade argument amounts to an argument for using national monopoly power to extract gains at other countries' expense. The United States could surely do this to some extent, but such a predatory policy would probably bring retaliation from other large countries. A cycle of retaliatory trade moves would, in turn, undermine the attempts at international trade policy coordination described later in this chapter.

The terms of trade argument against free trade, then, is intellectually impeccable but of doubtful usefulness. In practice, it is emphasized more by economists as a theoretical proposition than it is used by governments as a justification for trade policy.

The Domestic Market Failure Argument Against Free Trade

Leaving aside the issue of the terms of trade, the basic theoretical case for free trade rested on cost-benefit analysis using the concepts of consumer and producer surplus. Many economists have made a case against free trade based on the counterargument that these concepts, producer surplus in particular, do not properly measure costs and benefits.

Why might producer surplus not properly measure the benefits of producing a good? We consider a variety of reasons in the next two chapters: These include the possibility that the labor used in a sector would otherwise be unemployed or underemployed, the existence of defects in the capital or labor markets that prevent resources from being transferred as rapidly as they should be to sectors that yield high returns, and the possibility of technological spillovers from industries that are new or particularly innovative. These can all be classified under the general heading of domestic market failures. That is, each of these examples is one in which some market in the country is not doing its job right-the labor market is not clearing the capital market is not allocating resources efficiently, and so on.

Suppose, for example, that the production of some good yields experience that will improve the technology of the economy as a whole but that the firms in the sector cannot appropriate this benefit and therefore do not take it into account in deciding how much to produce. Then there is a marginal social benefit to additional production that is not captured by the producer surplus measure. This marginal social benefit can serve as a justification for tariffs or other trade policies.


Figure 9-3 illustrates the domestic market failure argument against free trade. Figure 9-3a shows the conventional cost-benefit analysis of a tariff for a small country (without terms of trade effects). Figure 9-3b shows the marginal benefit from production that is not taken account of by the producer surplus measure. The figure shows the effects of tariff that raises the domestic price from Pw to  Pw + t. Production rises from S1 to S2, with resulting production distortion indicated by the area labeled a. Consumption falls from D1 to D2, with a resulting consumption distortion indicated by the area b. If we considered only consumer and producer surplus, we would find that the cost of the tariff exceed its benefits. Figure 9-3b shows, however, that this calculation overlooks an additional benefit that may make the tariff preferable to free trade. The increase in production yields a social benefit that may be measured by the area under the marginal social benefit curve from S1 to S2 indicated by c. In fact, by an argument similar to that in the terms of trade case, we can show that if the tariff is small enough the area c must always exceed the area a+b and that there is some welfare-maximizing tariff that yields a level of social welfare higher than that free trade.

The domestic market failure argument against free trade is a particular case of a more general concept known in economics as the theory of the second best. This theory states that a hands-off policy is desirable in any one market only if all other markets are working properly. If they are not, a government intervention that appears to distort incentives in one market may actually increase welfare by offsetting the consequences of market failures elsewhere. For example, if the labor market is malfunctioning and fails to deliver full employment, a policy of subsidizing labor-intensive industries, which would he undesirable in a full-employment economy, might turn out to he a good idea. It would he better to fix the labor market, for example, by making wages more flexible, but if for some reason this cannot be done, intervening in other markets may be a "second-best" way of alleviating the problem.

When economists apply the theory of the second best to trade policy, they argue that imperfections in the internal functioning of an economy may justify interfering in its external economic relations. This argument accepts that international trade is not the source of the problem but suggests nonetheless that trade policy can provide at least a partial solution.

How Convincing Is the Market Failure Argument?

When they were first proposed, market failure arguments for protection seemed to undermine much of the case for free trade. After all, who would want to argue that the real economics we live in are free from market failures? In poorer nations, in particular, market imperfections seem to be legion. For example, unemployment and massive differences between rural and urban wage rates are present in many less-developed countries (Chapter 10). The evidence that markets work badly is less glaring in advanced countries, but it is easy to develop hypotheses suggesting major market failures there as well - for example, the inability of innovative firms to reap the full rewards of their innovations. How can we defend free trade given the likelihood that there are interventions that could raise national welfare?

There are two lines of defense for free trade: The first argues that domestic market failures should be corrected by domestic policies aimed directly at the problems' sources the second argues that economists cannot diagnose market failure well enough to prescribe policy.

The point that domestic market failure calls for domestic policy changes, not international trade policies, can be made by cost-benefit analysis, modified to account for any unmeasured marginal social benefits. Figure 9-3 showed that a tariff might raise welfare despite the production and consumption distortion it causes because it leads to additional production that yields social benefits. If the same production increase were achieved via a production subsidy rather than a tariff, however, the price to consumers would not increase and the consumption loss b would be avoided. In other words, by targeting directly the particular activity we want to encourage, a production subsidy would avoid some of the side costs associated with a tariff.

This example illustrates a general principle when dealing with market failures: It is always preferable to deal with market failures as directly as possible, because indirect policy responses lead to unintended distortions of incentives elsewhere in the economy. Thus, trade policies justified by domestic market failure are never the most efficient response; they are always "second-best" rather than "first-best" policies.

This insight has important implications for trade policymakers Any proposed trade policy should always be compared with a purely domestic policy aimed at correcting the same problem. If the domestic policy appears too costly or has undesirable side effects, the trade policy is almost surely even less desirable-even though the costs are less apparent.

In the United States, for example, an import quota on automobiles has been supported on the grounds that it is necessary to save the jobs of autoworkers. The advocates of an import quota argue that U.S. labor markets are too inflexible for autoworkers to remain employed either by cutting their wages or by finding jobs in other sectors. Now consider a purely domestic policy aimed at the same problem: a subsidy to firms that employ autoworkers. Such a policy would encounter massive political opposition. For one thing to preserve current levels of employment without protection would require large subsidy payments, which would either increase the federal government budget deficit or require a tax increase. Furthermore, autoworkers are among the highest-paid workers in the manufacturing sector; the general public would surely object to subsidizing them. It is hard to believe an employment subsidy for autoworkers could pass congress. Yet an import quota would be even more expensive, because while bringing about the same increase in employment, it would also distort consumer choice. The only difference is that the costs would be less visible, taking the form of higher automobile prices rather than direct government outlays.

Critics of the domestic market failure justification for protection argue that this case is typical: Most deviations from free trade are adopted not because their benefits exceed their cost but because the public fails to understand their true costs. Comparing the costs of trade policy with alternative domestic policies is a useful way to focus attention on how large these costs are.

The second defense of free trade is that because market failures are typically hard to identify precisely, it is difficult to be sure about the appropriate policy response. For example, suppose there is urban unemployment in a less-developed country: what is the appropriate policy? One hypothesis (examined more closely in Chapter 10) says that a tariff to protect urban industrial sectors will draw the unemployed into productive work and thus generate social benefits that more than compensate for its costs. Another hypothesis says, however, that this policy will encourage so much migration to urban areas that unemployment will, in fact, increase. It is difficult to say which of these hypotheses is right. While economic theory says much about the working of markets that function properly, it provides much less guidance on those that don't; there are many ways in which markets can malfunction, and the choice of a second-best policy depends on the details of the market failure.

The difficulty of ascertaining the right second-best trade policy to follow reinforces the political argument for free trade mentioned earlier. If trade policy experts are highly uncertain about how policy should deviate from free trade and disagree among themselves, it is all too easy for trade policy to ignore national welfare altogether and become dominated by special-interest politics. If the market failures are not too had to start with, a commitment to free trade might in the end be a better policy than opening the Pandora's box of a more flexible approach.

This is, however, a judgment about politics rather than economics. We need to realize that economic theory does not provide a dogmatic defense of free trade, something that it is often accused of doing.

Income Distribution and Trade Policy

The discussion so far has focused on national welfare arguments for and against tariff policy. It is appropriate to start there, both because a distinction between national welfare and the welfare of particular groups helps to clarify the issues and because the advocates of trade policies usually claim they will benefit the nation as a whole. When looking at the actual politics of trade policy, however, it becomes necessary to deal with the reality that there is no such thing as national welfare; there are only the desires of individuals, which get more or less imperfectly reflected in the objectives of government.

How do the preferences of individuals get added up to produce the trade policy we actually see? There is no single, generally accepted answer, but there has been a growing body of economic analysis that explores models in which governments are assumed to be trying to maximize political success rather than an abstract measure of national welfare.

[If there is interest, I will continue with this section...]

1 The additional gains from free trade that are discussed here are sometimes referred to as "dynamic" gains, because increased competition and innovation may need more time to take effect than the elimination of production and consumption distortions.

2 See Harris and Cox, Trade, Industrial Policy, and Canadian Manufacturing (Toronto: Ontario Economic Council, 1984); and, by the same authors, "Trade Liberalization and Industrial Organization: Sonic Estimates fur Canada," Journal of Political Economy 93 (February 1985), pp. 115-145.

    Posted by on Thursday, April 27, 2006 at 12:04 AM in Economics, International Trade, Policy, Politics, Regulation | Permalink  TrackBack (0)  Comments (10)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference The Political Economy of Trade Policy:


    Feed You can follow this conversation by subscribing to the comment feed for this post.