Trade Theories and Realities:
Why Economists Should Study Fairness




by Steve Suranovic (1)
(published in the September 1997 issue of CHALLENGE)

In the 1996 US Republican primary campaign for President, Pat Buchanan argued that the North American Free Trade Area (NAFTA) is causing the loss of American jobs. That spring Buchanan went to a closed textile plant in South Carolina to emphasize his belief that free trade agreements cause factories to close and jobs to be lost. Buchanan argued that continuing movements to free trade will put many other middle class jobs at risk. On the same day Senator Bob Dole visited a new BMW plant in South Carolina. He argued that the new jobs at the plant are precisely the kind of opportunities one can expect with free trade. Each of these candidates were trying to convince voters that their arguments were right. Unfortunately for voters, economic theory suggests both are right.

In the continuing public debate between free trade and protectionism the lessons that economic theory teaches are often misunderstood or misrepresented. Perhaps the reason is that each side in the debate tends to emphasize the positive aspects of their position while virtually ignoring the negative aspects. For example, free trade advocates tend to emphasize the opportunities that free trade will open up for exporters and consumers. It is common to hear that free trade is a win-win situation; or that it is a positive-sum rather than a zero-sum game. Opponents of free trade, on the other hand, tend to emphasize the threats that open competition with low-wage countries will pose to our import-competing industries. More specifically they will argue that free trade will cause painful job losses to already low paid workers in our economy.

In this article I will explain why both positions are consistent with economic theory. Trade theory does not tell us that free trade is good for everyone or bad for everyone. Instead free trade (or trade liberalization) is likely to cause a redistribution of income. Some individuals will gain from trade, while others lose. The issue that must be addressed is what to do in the face of such a redistribution. Economists typically suggest applying something called the compensation principle. But, as I will argue, this principle, while valid in theory, is inadequate in practice.

To find an answer that can generate broad agreement across diverse groups requires that we ask a slightly different question. It is a question inspired by the outcome of economic models, yet, it is a question whose answer cannot be obtained using economics alone. Instead we will need to explore a concept that is already prominent in trade policy discussions; the notion of fairness.

The Effects of Free Trade

The following story describes the likely effects of trade liberalization on three representative industries and individuals in the economy.

When barriers to trade are removed, increased foreign competition in a product line, such as cotton shirts, will lower the prices of imported shirts, setting off the following chain of events. Lower import prices will stimulate domestic demand for imported shirts while reducing demand for domestically produced shirts. To minimize the drop in sales, the home firms will have to cut their prices causing revenues to fall. To reduce costs local shirt factories may reduce employee bonuses, initiate smaller wage increases, and perhaps even make wage reductions. The result is a drop in income for some employees at the home shirt factories as a direct result of free trade. Since revenues are down, profits will likely fall as well, adversely affecting the owners and shareholders in the company. These outcomes are suggested by what economists call an "immobile factor model" in trade theory.

In time, the firms may be forced to reduce output. Some workers facing uncertain wages and uncertain future employment, will begin searching for alternative employment opportunities. Some of them, like Louisa, an accountant, may leave quickly. She knows she's likely to be the first one fired in her division since she was the last one hired. She's young and has a limited history and connection with the firm. Besides she has a skill that's readily transferable to other industries. If Louisa gets laid-off or loses her job before she finds another position then she will not only suffer a loss in income but also the anxiety and uncertainties associated with searching for a job. Within a few months Louisa lands a position with a rapidly growing biotechnology firm and she winds up with a much better position than before. Not only does she get a slight increase in salary, but her opportunities for promotion within the firm are also greatly improved.

Workers like Thelma, however, have worked at the shirt factory for 35 years sewing labels onto shirts. Thelma had hoped to retire within five years but now, with the recent layoffs and wage reductions, she's not sure she'll be able to maintain her current standard of living. She's not even considering another job since she figures no one will hire a 58-year old high school dropout whose only skill is sewing. For Thelma, free trade could reduce her future stream of income and thus her standard of living for the rest of her life. These outcomes are consistent with what economists call a "specific factor model" of trade.

Of course, economic models also predict positive effects. When the price of cotton shirts, and other importables fall with the reduction in trade barriers, millions of consumers who buy these products will now have a little money left over to spend on other goods in the economy. Although for any particular consumer the savings may seem negligible, when summed-up over millions of consumers it can represent a sizeable amount of additional purchasing power. This increase in purchasing power will raise demand and prices for other products in the economy. Exactly which markets will be affected will depend on how those with extra money decide to spend it. In addition foreign cotton shirt exporters and others like them will now have larger dollar incomes which will either, 1) stimulate demand for our export goods, or, 2) increase savings in the US financial sector. If the latter occurs, more funds become available for loans, which once made, will stimulate demand and raise prices of investment goods (or other consumption goods) in the domestic economy.

Regardless of what combination of effects occur, higher demand and prices for other sectors' goods and services will stimulate the growth of these sectors. Suppose one such firm experiencing an increase in demand is a telecommunications firm that exports many of its services abroad. Initially, the firm may simply ask some workers to work overtime. Profits will begin to rise benefiting shareholders and making it easier to offer larger bonuses and more generous pay increases to current employees. However, as the increased demand continues, the firm will begin to hire new employees so that it can further increase its output and profits. Jack, who just completed an MBA degree and has an undergraduate degree in computer science, lands a good paying job with the company a few weeks before graduation. Jack joins other new employees in his company, and others like him across the country who have new career opportunities in expanding firms because trade barriers were removed in other sectors of the economy.

As some industries expand they will most likely require a different mix of labor and capital than the industries that are contracting. The telecommunications firm may require specially trained workers; people with backgrounds in computer science, engineering, finance etc. It's unlikely they'll be looking for seamstresses. Thus, some of the workers who lose their jobs in declining industries will not be in demand in the expanding sectors. This will have two long-term effects. First, excess labor demand for the technically skilled workers will permanently bid up their wages compared to pre-free trade wages. Second, excess labor supply of workers like seamstresses will permanently bid down their wages. This outcome is consistent with what economists call a "factor-proportions model" in trade.

Lessons from Trade Theory

A number of important lessons emerge from this story of trade liberalization. First, economic analysis shows that trade liberalization will result in gains in income and well-being for some individuals and losses in income and well-being for others. Second, the gains and losses will be spread out over time. Some people will lose initially, but be better-off in the long-run. (like Louisa who was temporarily unemployed but landed the job with the biotechnology firm). Others will suffer losses for the rest of their lives (like Thelma who was forced to retire early). Some may gain in the short-term and lose in the long-term; others will gain from the get-go (like Jack who landed the telecom job). Third, for many individuals losses will precede the gains. (Louisa must incur adjustment costs before she enjoys the benefits of her new position). Fourth, the losers will be easier to identify than the winners. This is because the losses occur in industries that are directly affected by the removal of trade barriers and these losses occur first. The gains generally arise indirectly, after a series of effects that begin in the import sector. Fifth, the gains will more likely accrue to the higher-skilled, younger, or more adaptable workers; the losses will accrue to the lower-skilled, older and less adaptable workers. To share in the gains, it will help to have the newer skills needed in the expanding industries. Younger workers are more likely to have these newer skills. Also workers in the declining sectors, who are motivated to take classes to learn new skills, to search aggressively and widely for alternative positions and/or to move their family to a new city or state, are more likely to fare well eventually. However, older workers, near retirement, or those who simply couldn't learn a more advanced skill, or who wouldn't even consider moving their family to another city or state are more likely to suffer long-term losses.

The Economic Case for Free Trade

Since trade models clearly conclude that some individuals will lose from freer trade, it's puzzling that the general public seems to believe that economics teaches that everyone benefits from free trade. Perhaps this is because economists always emphasize what are called the "aggregate efficiency" gains from trade. Aggregate efficiency improvements of two types can be expected with free trade; production efficiency and consumption efficiency. A production efficiency improvement means that the economy will produce more goods and services overall with the same amount of resources. A consumption efficiency improvement means (essentially) that consumers will have more satisfying choices available to them. Most economic models demonstrate that trade liberalization will raise aggregate economic efficiency. Thus, it is true to say that free trade will benefit everyone on average, or, that the country benefits in the aggregate. However, both of these statements mask the income redistribution effects caused by freer trade.

A more accurate way to relate the efficiency improvements of trade liberalization is to say that the sum of the gains to the winners will exceed the sum of the losses to the losers. This way it is clear that the net effect of trade liberalization is positive but that a redistribution of income, with some individuals winning and others losing, will also result.

The presence of the redistributive effects of free trade implies that aggregate improvements are consistent with economic indicators showing signs of distress in some sectors of the economy together with signs of prosperity in other sectors. What's more, signs of distress may arise in seemingly efficient sectors of the economy.

In general, economic models predict that some firms will experience price, production and employment increases while others firms will suffer price, output and employment reductions. Growth will occur for some firms because their productivity is higher than their foreign competitors. In this case the firms have what economists call an absolute technological advantage in production.

However, even firms that lack an absolute advantage may experience growth. This may arise for firms who are the least technologically disadvantaged relative to their foreign competitors. In this case economists would say that these firms have a comparative advantage in production. This case is especially important to understand. It implies that in less technologically advanced countries, firms can grow and the country can benefit from trade in the aggregate even if they are less efficient in the production of everything relative to its trade partners. On the other hand, for the most technologically advanced countries, trade can cause some of the most advanced firms in the world to lay off workers or even shut down completely. In the US, the textile industry often argues that free trade hurts them despite the fact that productivity in the industry is higher than anywhere else in the world. This outcome is perfectly consistent with the theory of comparative advantage which explains that aggregate efficiency can rise despite the decline of industries like textiles. These counter-intuitive results, known for almost two centuries, are still widely misunderstood or simply not believed.

Even if firms lack a technological comparative advantage, trade can cause price increases and growth for some firms. Some firms may expand because they use a mix of resources in which the country is relatively well-endowed compared to other countries. For example, freer trade for Costa Rica should cause growth of its banana industry since Costa Rica is well-endowed with the climate and soil needed to grow bananas.

Lacking favorable endowments, free trade can cause growth for firms in which foreign demand for their products is relatively higher than domestic demand and the decline of firms in which demand is relatively lower. Finally, lacking any differences in supply or demand characteristics between trade partners, growth may still occur for firms which take advantage of economies of scale in production (i.e. lower costs through larger production runs).

The adjustment process requires the reallocation of capital, labor and land from the declining industries to the expanding industries. But, since the expanding firms will use the resources more efficiently, overall output in the economy will rise. Also since many prices have fallen and more goods are available in the aggregate, consumers will have a more satisfying variety of products from which to choose. Thus economists say that free trade will benefit the country because it improves aggregate economic efficiency. However, as already noted above, the efficiency improvements will not arise without cost and they will not accrue to everyone.

The Compensation Principle in Theory and Practice

Economists sometimes deflect this criticism by invoking the "compensation principle". It says that since free trade causes the sum of the gains to exceed the sum of the losses, it is possible to redistribute income from the winners to the losers in a way that everyone benefits. The problem with the principle, however, lies with its implementation. In order to apply the redistribution principle, one must first identify precisely who will win and lose, by how much, and when. As the above account suggests many different characteristics will determine which industries expand and which contract and thus which individuals are positively and negatively affected.

Economists could be asked to numerically estimate these effects, but their results would be little more than a sophisticated guess. Their estimate may be the best guess possible, but that should not imply it be treated as a "fact".

Simply consider the economic studies conducted in the debate over the NAFTA. The key issue for politicians was clearly the effect of NAFTA on jobs. Different studies provided widely different answers. Undoubtedly some of the models and estimation techniques were better than others, but even the best estimation techniques were imprecise and inaccurate. In part this is because of the difficulty in forecasting significant economic variables like exchange rates, GDP growth rates etc., which can greatly affect the results. None of the studies predicted the collapse of the Mexican peso for example. Estimation of effects is also difficult because there is a wide variety of dynamic processes at work in large complex economies that simply cannot be captured fully in any economic model. Even in hindsight, it is extremely difficult to conclude with a high degree of confidence that particular jobs were lost because of a free trade agreement rather than because of exchange rate changes or reductions in the quality of the product, or mismanagement by the firms, or regional shifts in demand, or any other of a host of potential causes.

Even if the job estimates were precise and accurate, we would still need much more information to provide the appropriate compensation. Recall that not all of the losses to individuals result because of lost employment. Some workers will merely suffer lower wage increases over time while others will take early retirement. There simply is no accurate way to measure the value of these losses attributable only to trade liberalization throughout an entire economy.

Additionally, to provide compensation, we need to know who gains how much over time. As mentioned earlier the winners will be much more difficult to identify. Because lower import prices will cause consumers to shift spending to other products, some of the firms that benefit may not even be engaged in international trade. For other firms we would need to distinguish the percentage of their growth caused by free trade from the percentage that would have occurred otherwise.

What all this implies, of course, is that although compensation from winners to losers could potentially raise everyone's welfare, in practice, compensation is no more than a pipedream. The reality is that any movement to free trade must lead to uncompensated losses for some individuals in the economy.

A New Question

These theories and realities should lead us to the following question. Under what circumstances is it appropriate for a government policy (trade liberalization in this case) to cause uncompensated losses to some individuals? How does one justify to the 58-year old seamstress that her lifetime income must fall? How do you tell the families of steelworkers that the town they grew up in will die a slow death? How do you explain to farmers across America that the days of the family farm will end? These things matter to people because they form the basis of their hopes and dreams, for themselves, their children and their grandchildren. Would it really make sense for them to accept the economic efficiency argument that their losses are appropriate since others will gain even more than they lose? Can the losers be convinced that the government would step in and compensate them for their losses to assure that everyone will indeed gain from trade? To both of these questions the answer is unequivocally no.

Indeed it is these very arguments about the potential losses likely to arise from trade liberalization that convinces many people that protectionism is warranted. However, upon further reflection, protectionism could just as easily be discredited.

When trade barriers are raised most of the effects described above are simply reversed. That means that an increase in protection, as advocated by Pat Buchanan, will also result in a redistribution of income. Those who benefited from free trade in the story above would now lose from additional protection. Those who had lost before would now gain. Increases in protection would raise the lifetime income of Thelma the seamstress. Louisa the accountant would probably never need to leave her job at the shirt factory. She would benefit in the short-run since she would not need to search for a new job, but, in the long-run her wages will remain lower than if she had gotten that job with the biotechnology firm. Jack, the MBA graduate, won't find a job as quickly as before resulting in some additional short-term losses. Furthermore the job he takes won't pay quite as well since his skills are somewhat less in demand in the labor market. Overall, economic efficiency will probably fall with the increase in protection, meaning that the sum of the losses to people like Louisa and Jack will exceed the sum of the gains to people like Thelma.

The main difference between protectionism and free trade is that the gains from protection will be obvious while the losses will be much more obscure. Thelma will readily recognize that protection saved her job. Louisa and Jack will probably be unaware of their lost opportunities. Clearly this is one of the reasons protectionism is so popular.

Nevertheless, in one important respect, the effects of protection are really much the same as the effects of trade liberalization. Protection will also generate uncompensated losses, in the form of higher prices to consumers, or in the form of lost opportunities to some individuals in the economy. And that brings us back to the same question as before. Under what circumstances is it appropriate for a government policy (protection in this case) to cause uncompensated losses to some individuals?

Whether a government moves to free trade or chooses to raise barriers to trade, its actions will effectively take money away from some people within the country and give it to others. Trade policy actions, then, are just as much a story about who gets what domestically as it is a story about one country against another. If free trade is chosen, workers like Thelma have every reason to ask why they must suffer income losses in the name of aggregate efficiency. If more protection is chosen, (fully informed) workers like Louisa and Jack have a reason to ask why their opportunities must be limited (and their incomes reduced) in order to save other workers' jobs.

Beyond Economic Analysis to Fairness

If we are ever to answer these legitimate concerns we must begin to study the concept of fairness. One reason fairness matters is because it is precisely the expectation of uncompensated losses that inspires claims of unfairness in trade policy discussions. If everyone believed that free trade (or protection) would generate benefits for everyone, then there would never be any reason to charge that free trade (or protection) is unfair. Actions are perceived as unfair only if some group expects to suffer losses as a result of the action.

To many people fair trade is perceived as synonymous with protectionism. There is good reason for this. As explained above, free trade will generate recognizable and uncompensated losses primarily to import-competing firms. To those negatively affected this may seem unfair. Protection of these industries would recognizably eliminate these uncompensated losses. Hence, for trade to be fair to them may require protection. In contrast, protection would generate losses that are either difficult to recognize or are too small per person to rally the opposition. Thus, it is less common to hear charges of unfairness in response to calls for protection.

One of the problems with the use of fairness in trade policy discussions is that it is often applied in contradictory ways. In the US fairness has been used to argue for import protection to save jobs in low-wage industries. It has also been used to argue for the removal of import barriers in Japan which are claimed to unfairly restrict access to US exports. In each case there is the expectation of losses to some group of individuals in the US economy. But, how can it be acceptable (fair) for the US to apply import barriers to protect US jobs but not acceptable (fair) for Japan to do the same thing?

Similarly, workers in a textile plant may claim that freer trade is unfair because they will lose their jobs in the face of competition with foreign firms who pay their workers less. At the same time, though, consumers may argue that it is unfair that they must pay higher prices for the imported goods if freer trade is not pursued. Thus a single policy action can be viewed as fair by one group but unfair by another.

These contradictory uses of fairness support a popular impression that fairness is a subjective rather than an objective norm. Under this view, fairness is only used to support a groups' narrow self-interests since it seems that anytime anyone expects to suffer losses as a result of some trade policy action, they are quick to scream, "unfair!" Thus fairness may be an inappropriate basis for choosing policies since someone would always stand to lose for any deviation from the status quo. There are two arguments against this notion.

First, although charges of unfairness arise only if there are expected uncompensated losses, it does not follow that all uncompensated losses should rightfully be viewed as unfair. A simple example of this is found in the arena of sporting events. Every basketball game results in an uncompensated loss to one of the competing teams, and yet it is rarely the case that the losing team will charge that the outcome is unfair. The reason is that both teams have agreed to abide by a set of rules which when followed will guarantee that one team wins and one team loses. If the rules are not followed, as is the case when a referee makes a decisive incorrect call, then the outcome might be viewed as unfair, but otherwise, losses are expected and tolerated. This suggests that some charges of unfairness in trade policy discussions may simply be inappropriate applications, somewhat like the losing basketball team screaming unfair just because they lost.

Secondly, the reason fairness arguments are made in political circles is because the general public is convinced that it matters. Almost everyone is taught from a very young age to show compassion for others; to share, especially with those who are less fortunate; to act towards others in ways that we would like others to act towards us; to be truthful and honest and to keep our promises. These sentiments are the substance of fairness. It would actually be quite astounding if the public did not expect government policies to be guided by similar principles.

These arguments suggest that fairness needs to be taken seriously when evaluating international trade policies. The public prefers policies that are fair rather than policies that are merely efficient. However, determining what's fair is not a simple matter. In order to make progress researchers need to identify how fairness arguments are applied to trade policy discussions and begin to distinguish appropriate uses of fairness from inappropriate uses. In a sense this means developing a fair set of rules that would guide international transactions. To accomplish this, researchers will need to turn their attention to disciplines like ethics or moral philosophy.

For economists this means they must rediscover their roots. Recall that the founder of modern economics, Adam Smith, was a professor of moral philosophy at Glasgow University. His first great book was titled "The Theory of Moral Sentiments" which was published 17 years before its better-known successor, "The Wealth of Nations." Perhaps if we devoted a little attention to the issues of Smith's first book we could work towards a generally acceptable answer to the key question suggested above; under what circumstances is it appropriate for a government policy to cause uncompensated losses to some individuals? Or, in other words, under what circumstances is a government policy fair?

1. Thanks go to Robert Goldfarb and Herman Stekler for helpful comments.

©1997-1998 Steven M. Suranovic, ALL RIGHTS RESERVED
Last Updated on 1/12/98