A major turning point in the thinking of economic policy makers in industrial countries occurred at the beginning of the 1980s. The change was motivated by a desire to shift from the existing environment of low growth, high and rising unemployment, high inflation, and large fiscal deficits to conditions of stronger and sustained non-inflationary growth. The reorientation of policies that followed the reassessment of the early 1980s included two main elements, both of which were based on a recognition that, beyond the short run. Macroeconomic performance is strongly influenced by the supply side of the economy.

A major turning point in the thinking of economic policy makers in industrial countries occurred at the beginning of the 1980s. The change was motivated by a desire to shift from the existing environment of low growth, high and rising unemployment, high inflation, and large fiscal deficits to conditions of stronger and sustained non-inflationary growth. The reorientation of policies that followed the reassessment of the early 1980s included two main elements, both of which were based on a recognition that, beyond the short run. Macroeconomic performance is strongly influenced by the supply side of the economy.

First, measures to remove market distortions (micro-economic, or “structural” reforms) were introduced in order to raise potential output and increase the flexibility of the economic system. Second, a medium-term orientation of demand management policies was formulated in order to stabilize the macroeconomic environment in Which private decisions are made and to release resources from the public sector that could be used more productively by the private sector. Essentially this step involved a public commitment to fiscal consolidation and to a more stable and restrained monetary policy. The reduction of public sector spending and deficits was expected to lower real interest rates and to encourage-private investment, which had been crowded out by rapidly expanding government expenditures and deficits. At the same time, monetary policy was to eschew frequent shifts and to focus instead on the long-term objective of price stability.

Over the past decade, substantial progress has been achieved in many areas in removing distortions and increasing flexibility. In particular, financial markets throughout the industrial world have been substantially liberalized since the late 1970s. Measures to deregulate domestic goods markets and to stimulate competition through the privatization of state-owned enterprises also have been implemented in many countries. Nevertheless, much remains to be done to improve the functioning of labor markets and to eliminate other structural obstacles to noninflationary growth. Indeed, in the critical area of international trade policies, the past few years have witnessed an intensification of non-tariff barriers. The following two sections conclude that bold implementation of structural reform could make a significant contribution toward the resolution of some of the main macroeconomic problems of the day.

Structural Policies: Theory and Empirical Evidence

Structural Reform and Macroeconomic Theory

Structural policies can be defined as actions that raise the productive capacity or increase the flexibility of the economy. Because structural policies typically aim at improving efficiency by reducing distortions that hinder the free functioning of individual markets, they are often referred to as “microeconomic policies.” In addition to their microeconomic and sectoral implications, structural policies can have significant effects on key macroeconomic variables such as prices, budgetary positions, interest rates, and external current account balances. It should be noted at the outset that the desirability of structural measures should not be assessed primarily in terms of these effects: structural policies are justified inasmuch as they lead to a more efficient allocation of resources and to higher growth of productive capacity. At the same time, the macroeconomic implications of structural policies should be recognized and taken into account in the overall design of economic policy. This section summarizes the major macroeconomic effects of structural policies both in the country adopting the measures and in the rest of the world.

Macroeconomic Effects of Increases in Potential Output

Structural policies can raise potential output in three major ways, (i) They can increase total factor productivity—that is. The level of output that can be produced with given levels of capital and labor—by improving the allocation of resources (e.g., through trade liberalization or other measures that reduce factor market distortions and enhance competition) or by improving administrative efficiency or promoting technological change, (ii) They can also increase measured total factor productivity by improving the quality of factors of production (for example, the efficiency of labor could be enhanced by improving the educational system), (iii) Structural policies can increase private saving and investment and improve their temporal and intertemporal allocation by eliminating financial market rigidities (such as interest-rate ceilings) or dealing with distortions in the tax system that discourage private saving or raise the cost of capital.

The macroeconomic effects of a typical structural measure can be analyzed within the framework of the economy’s aggregate demand and supply schedules.2 The permanent increase in potential output resulting from a structural measure can be interpreted as an outward shift in the aggregate supply schedule. In the general case of a downward sloping aggregate demand schedule, this shift will result in an expansion of actual output and a decline in the price level. In turn, the rise in production will stimulate consumption expenditure (as permanent income rises) and might also raise investment through a temporary accelerator effect.3 Moreover, the decline in prices will involve a rise in real wealth which could lead to a further increase in consumption and aggregate expenditure.

These domestic macroeconomic effects of structural measure can have both demand-side and supply-side effects on the external current account and therefore on partner countries. On the demand side, the rise in expenditure would increase the demand for imports in the home country and this would tend to worsen the current account. At the same time, however, the expansion of productive capacity increases the supply of exports in the home country, thus improving the external position. Which effect will predominate in the short run cannot be determined a priori. Whether or not the current account will improve on a net basis will depend upon whether the rise in output exceeds the rise in expenditure (in other words, whether the rise in domestic saving exceeds the rise in domestic investment).

The decline in the price level in the home country will involve a rise in the real money supply and, if output is sluggish, this will result in an excess supply of money. This, in turn, will lead to a drop in the domestic interest rate and, given foreign interest rates, to a temporary depreciation of the exchange rate. In the longer run. The excess supply of home goods will need to be absorbed, which suggests a real depreciation of the home country’s exchange rate.

The rise in potential output in the home country is also likely to have supply-side effects on potential output in foreign countries. The real exchange rate appreciation abroad implies a decline in foreign consumer prices and a fall in nominal wages. As a result, the real product wage at full employment fails and foreign potential output rises. Potential output abroad may also rise as a result of technological improvements in the home country that are embodied in imports of new capital goods, thus raising productivity in the foreign country. As a general rule, these supply-side effects always involve “positive transmission” abroad of increases in potential output.

Structural measures could also affect investment and the current account by raising the rate of return on capital in the home country. If capital is internationally mobile, a higher rate of return on capital would result in a rise in investment and a temporary deterioration in the home country’s current account (reflecting, for example, imports of capital goods required to raise the domestic capital stock to the desired level), which will be financed by an inflow of foreign capital. As the capital stock rises in the home country, the rate of return on capital will fall back to the point where the excess demand for real capital in that country will disappear.

The impact of specific structural measures on international transactions will depend upon the effects of these measures on the balance between saving and investment. For example, the removal of distortions that have discouraged domestic investment in surplus countries could stimulate domestic demand and imports, and thus reduce the gap between domestic savings and investment. Alternatively, tax reforms that eliminate disincentives to save in deficit countries could diminish the reliance on foreign saving and thus help to reduce the external current account deficit.

Improvements in Degree of Economic Flexibility

In addition to affecting the level of potential output, structural policies may influence the “flexibility” of the economic system. Flexibility refers to the extent to which, and the speed with which, (i) prices respond to changes in market conditions and (ii) the supply and demand for goods, services, and factors of production respond to changes in prices (including wage rates and interest rates). By enhancing the flexibility of key prices, and particularly wages, structural policies can influence the way the economy reacts to exogenous disturbances—such as technological innovations or relative price shocks; they can thus help to reduce the margin of unused resources that would result from an adverse disturbance.4

In this connection, it is useful to distinguish between nominal and real wage flexibility.5 Nominal wage flexibility reflects the elasticity of the nominal wage with respect to the aggregate price level. Real wage flexibility concerns the responsiveness of the real wage to unemployment. Both concepts refer to the short-run response of the economy to exogenous disturbances or shocks.

In an economy with perfect real wage flexibility, the labor market is well protected against disturbances stemming from both the demand side and the supply side of the economy. However, the labor market in an economy with perfect nominal wage flexibility is well protected only from the effects of autonomous shifts in demand. Therefore, if real wage flexibility is less than perfect, the desirable degree of nominal wage flexibility depends on the relative importance of demand (nominal) and supply (real) shocks. Gray (1976) has shown in the context of a closed economy that less nominal flexibility is desirable if shocks are predominantly related to the supply side, Aizenman and Frenkel (1985) have extended Gray’s analysis to the case of a single open economy and showed that the optimal degree of nominal wage flexibility depends on the demand policies being pursued. Thus, ceteris paribus, the difference in the types of shocks experienced may help to explain why employment in some industrial countries (particularly in Europe) seems to have performed relatively well in certain periods (e.g., the 1950s and 1960s when demand shocks may have predominated), but rather poorly in other periods (e.g., the 1970s when supply shocks originating in commodity markets predominated), even though the process of wage determination remained broadly unchanged.

From the perspective of macroeconomic policy, the promotion of real wage flexibility would seem to be particularly important since it plays a fundamental role in the process by which an economy adjusts to eliminate the unemployment that results from past shocks. An increase in flexibility may significantly improve an economy’s “sacrifice ratio”—that is, the short-run cost in terms of unemployment which will be involved in achieving a given reduction in the rate of inflation.6 More importantly, harmonizing the degree of flexibility among countries at a high level may significantly lower the external adjustment costs arising from common macro-economic shocks, thus facilitating the international adjustment process. If the degree of flexibility differs among countries, countries with low real wage flexibility tend to increase the costs of adjustment both for themselves and for the countries with relatively high wage flexibility.

Empirical Evidence on Effects of Structural Policies

In a now famous remark, James Tobin once said that “It takes an awful lot of Harberger triangles to fill an Okun gap.”7 By this he meant that the gains in welfare resulting from the reallocation of resources brought about by structural reforms in a fully employed, developed economy often appear to be relatively small compared to the gaps between actual and potential output that are typically observed over the course of the business cycle. While a review of the empirical literature suggests that this may be the case for some types of structural policies, there is evidence that the gains resulting from eliminating labor market distortions and barriers to international trade could be substantial.

The evidence from the empirical literature may be divided into five categories in which interest in structural reform has recently been focused: trade policies, tax reform, financial market deregulation, deregulation and privatization of goods markets, and labor market reform. The evidence relating to these five areas can be summarized as follows:

Empirical studies based on traditional models of international trade (assuming perfect competition and constant returns to scale) suggest that the welfare gains that can be expected from the elimination of trade restrictions are generally small and subject to a wide margin of uncertainty. It should be noted, however, that a permanent rise in real GNP of a fraction of a percentage point could be significant: indeed, the present discounted value of such gains could amount to several percentage points of current GNP. Moreover, recent studies based on the “new international trade theories” (which allow for imperfect competition and increasing returns to scale and may therefore be more realistic) suggest that the potential welfare gains from reducing trade harriers are significantly larger (as much as several percentage points of GNP annually in smaller economies, which have the most to gain from external trade liberalization) than was suggested by earlier studies.8

Potential welfare gains from tax reform may also be large, although the range of estimates is quite broad.9 Even though they are not directly comparable, welfare gains generally tend to be somewhat lower than output gains because the welfare gains reflect the substitution of work for leisure. (A reduction in leisure has no impact on measured GNP although it would obviously reduce net welfare.) Changes in corporate taxation can also have appreciable effects on the cost of capital and therefore on private investment. In some cases, however, the change in the cost of capital may be small because lower corporate tax rates are combined with the elimination of certain investment incentives. The main objective of these reforms—to “level the playing field” by rationalizing the tax treatment of various assets—could lead to significant efficiency gains over time.

The quantitative impact of financial market deregulation on the economy is rather uncertain. Significant gains are often assumed to have accrued from measures such as the elimination of interest rate ceilings and the authorization of a broader range of financial instruments. In general, little empirical work has been done on the benefits or on the associated adjustment costs of these measures, both of which may be significant. In the European context, however, substantial empirical work has been done on the prospective financial integration of the European Community (EC) to be completed by 1992. This work points toward significant gains to be realized (see Cecchini (1988)).

Deregulation and privatization of competitive industries have produced important productivity gains in the industries concerned (airlines, railroads, trucking, and telecommunications, for example). However, the macroeconomic effects of these measures are often limited, given the relatively small shares of these industries in aggregate output and employment. It is unclear whether reductions in the role of government in industries that are natural monopolies will raise or lower productivity.

Structural policies may produce the largest potential gains in output for many countries through labor market reform. Estimates of the rise in non-accelerating inflation rates of unemployment (NAIRUs) between the early 1970s and early 1980s range from 1 percent of the labor force or less in the United States and Japan, to 5 percent or more in many European countries. While some of the increases in NAIRUs during that period reflected demographic factors, there is no doubt that the interaction of labor market rigidities with the disturbances that occurred during that period also played an important role. Even partial reversals of these increases in NAIRUs through labor market reform would imply annual output gains of several percentage points of GNP in some countries.10

Recently, several important studies have attempted to quantify the potential gains that might accrue from comprehensive programs of structural reform. These studies include estimates of the effects of the reduction in internal barriers in Europe proposed for 1992, of implementation of some of the structural policy recommendations of the Mayekawa Report in Japan, of reform of the Common Agricultural Policy (CAP) in Europe, and of alternative structural reform measures in the Federal Republic of Germany.11 The estimated effects of structural reforms in all of these studies are quite large, although they appear to depend crucially on the specific technical assumptions used to derive the results.

Policy Implications of Structural Reform

Relationships Between Structural and Macroeconomic Policies

As mentioned above, structural policies affect the economy by raising the level of potential output and by increasing the flexibility of the economy. It is useful to retain this distinction in discussing the relationship between demand management and structural policies.

Objectives and time horizon. The way structural and macroeconomic policies affect economic performance can differ in several important ways. Policy objectives typically include both the objective of expanding potential output and the concern with maintaining capacity utilization at a high level without generating inflationary pressures (Adams, Fenton, and Larsen (1987)). Structural policies that raise potential output relate primarily to the objective of achieving sustained growth over the medium to long term: these policies often affect the level of demand indirectly, but this is not their primary purpose. Structural policies that affect the flexibility of the economy tend to influence the cyclical behavior of the economy and its response to exogenous disturbances.12 By way of comparison, monetary policy is primarily an instrument of demand management, although its ultimate objective, price stability, is essential to the achievement of sustained growth. Fiscal policies span both short- and long-run objectives: they directly affect the level of aggregate demand and therefore the degree of capacity utilization; but they also affect national saving and capital formation and therefore the long-run capacity of the economy to generate output.

Lags. Policy analysis usually distinguishes between inside lags (the legal and institutional delays involved in implementing a change in policy) and outside lags (the delay involved between implementation of a policy and its effects on the level of economic activity). Monetary policy is usually considered to encompass relatively short inside lags and relatively long outside lags, while the reverse appears to be the case for fiscal policies. Several empirical studies suggest that structural policies often involve relatively long lags of both types, with outside lags typically being longer than for demand management policies.13

Reversibility. Structural policies should be viewed as unidirectional measures: it would make little sense to reintroduce impediments to the efficient functioning of markets (and thus reduce welfare) in order to meet short-run output, inflation, or current account objectives. In comparison, monetary and fiscal policies are to some extent reversible as tools of demand management, although short-term adaptations in these policies must be consistent with longer-term objectives of growth, budgetary consolidation, and price stability. Simultaneous implementation of structural and demand management policies may enhance the credibility, and therefore the effectiveness, of economic policies (see below).

Uncertainty. Empirical estimates of the macroeconomic effects of structural policies are subject to a considerable margin of error—perhaps larger than the margin of error relating to the effects of fiscal or monetary policies. In such circumstances, substituting structural policy measures for fiscal adjustment, for example, could imply increasing the degree of uncertainty in the achievement of a given macroeconomic target.14 In particular, it may be desirable to direct both fiscal and structural measures to the achievement of higher growth, in order to reduce the degree of uncertainty associated with achievement of that objective.

Given the different objectives, time horizons, and economic effects of structural, fiscal, and monetary policies, it is natural to think of the roles of these policies as being complementary. Indeed, there is often little or no scope for substitution between them. For example, an expansionary monetary policy cannot be used to increase potential output. In a fully employed economy, monetary policy has only a very limited and temporary effect in generating output: instead, it is likely to lead to higher inflation and thus interfere with sustained growth. Conversely, structural policies have no role to play in demand management, among other reasons because of their irreversibility. Because fiscal and structural policies can both contribute to higher growth of potential output there would appear to be a degree of substitutability between them.15 In practice, however, the scope for a trade-off between structural and fiscal measures in the achievement of growth objectives would seem to be limited because of the differences in lags, and, more important, in the degree of uncertainty associated with the effects of both policies.

The complementarity of structural and demand management policies is particularly apparent in the second major role of structural reform—namely, that of increasing the flexibility of an economy. Complementarity in this sense means that increased flexibility reduces the adjustment costs that may be associated with appropriate macroeconomic policies; such costs may be reduced in several ways.

First, increasing the flexibility of an economy may significantly reduce the costs in terms of unemployment that arise from external adjustment. For example, in the case of a country with a persistent current account deficit and high inflation, unemployment may increase as a result of a fiscal correction aimed at reducing the external deficit and curbing inflation. The greater the flexibility of the economy, the faster will the needed adjustment take place, and the lower the total cost in terms of lost output and jobs.

Second, macroeconomic adjustment often involves changes in the terms of trade, which will tend to shift resources between sectors of the economy. Any transitory structural unemployment resulting from those resource shifts will be reduced in proportion to the flexibility of the economy.

Third, harmonizing the degree of flexibility among countries at a high level will make individual countries’ adjustment to common shocks more uniform. Common macroeconomic shocks can lead to substantial external imbalances in the medium term when the degree of flexibility differs among countries. Harmonizing flexibility at high levels would help to minimize both the magnitude and the persistence of external adjustment problems that may result from a common shock.

Finally, lowering the costs of macroeconomic adjustment will make it easier for countries to implement the required changes in demand management policies at an early stage. Although structural reform may not solve the adjustment problem, it does facilitate the task of demand policies in bringing about adjustment more quickly and at lower cost. In summary, the major role of structural reform is to provide an environment that promotes economic efficiency and growth, while helping to reduce adjustment costs.

Structural Reform and Policy Coordination16

Global structural reform affects both macroeconomic developments and the ability of demand management policies to deal with these developments. The present magnitude of current account imbalances in the world economy, for example, might not have been possible without the prior integration of world capital markets. Although a better global allocation of capital results from the integrated market, this integration complicates the task of macroeconomic management and underscores the need to consider macro-economic and structural policies simultaneously.

The integration of world capital markets, which links interest rates among countries, impinges upon both the domestic and international transmission effects of fiscal and monetary policies. Fiscal crowding out of interest-sensitive components of domestic demand (particularly investment), appears to have been partly shifted to the external sector. To the extent that an expansionary fiscal policy leads to currency appreciation, therefore, fiscal crowding out of investment via the interest rate has been partly replaced by the crowding out of net exports, as the incidence of fiscal policy shifted from the interest-sensitive components of domestic demand to exchange rate-sensitive sectors.17

In contrast, it has been suggested that the impact of monetary policy on the price level may have increased as a result of the integration of world financial markets (Fukao and Hanazaki (1986)). An expansionary monetary policy, for example, may now result in a smaller decline in real interest rates (because it is likely to lead to capital outflows), but the associated depreciation of the exchange rate will affect prices of traded goods directly and through the second round effects stemming from the increase in output. Since prices and output may be affected more rapidly through exchange rate changes than through interest rate adjustments, lags associated with monetary policy may also have been shortened.

In many countries the effects of a restrictive monetary policy have shifted gradually away from credit rationing toward market-based price rationing through interest rates as a result of the elimination of interest rate regulations. At the same time the closer integration of instruments and segments of financial markets now leads more quickly to generalized movements across the term structure of interest rates.18 Also, the greater substitutability of domestic and foreign assets related to world financial integration has resulted in a linkage among long-term interest rates across countries, independently of the stance of monetary policy.

The large cross-border flows that link interest rates also influence exchange rates, thereby shifting the incidence of monetary policy (like that of fiscal policy) from the interest-sensitive components of domestic demand to the exchange rate-sensitive sectors. Which channel of transmission is dominant depends mainly on the size and openness of the economy—the smaller and more open the economy, the more the exchange rate effects lend to dominate. In either case, the increased integration of financial markets implies that the conduct of an independent monetary policy is constrained (given the objectives of controlling inflation and the current account) by the response of capital and currency markets. Therefore, this integration further reduces the usefulness of fine tuning.

In the current environment, fiscal and monetary policies influence current account balances (and changes in domestic spending relative to income) via incipient capital flows and exchange rate changes. If the maintenance of stable exchange markets together with a sustainable pattern of current account balances among countries is considered an important objective, then closer coordination of monetary and fiscal policies may be necessary than would be the case in the absence of an integrated world capital market.

Another fundamental implication of a more integrated world capital market is that decisions about real investment take place in a global context. In this setting market forces will tend, over time, to link expected after-tax returns to capital across countries. The efficient allocation of world resources requires, however, that the before-tax returns be equalized. Accordingly, there is also a strong case for coordinating capital taxation across countries in order to “level the playing field” internationally as well as domestically.19 Similarly, it may also be desirable to coordinate broader aspects of tax reform, since the effects of such reforms depend critically on the pre-existing pattern of current account imbalances among countries.20

More outward-looking industrial and agricultural policies in industrial countries would not only provide domestic benefits but would also contribute to faster growth in the export revenue of the developing countries and therefore help to alleviate the debt problem. To the extent that concerted action among industrial countries makes progress in these areas more likely, there is a clear case for coordination of some aspects of structural policies. This concerns particularly agricultural policies, which are being discussed in the context of the current round of multilateral trade negotiations. Nevertheless, in view of the severity of the distortions now affecting output and trade in farm products and the length of time that may be needed to conclude international arrangements, it would be desirable for countries to start taking concrete steps to reduce these distortions, even on a unilateral basis.

Scope and Sequencing of Structural Reform

The implementation of measures to remove market distortions must deal with at least three practical questions. First, how widespread should structural reform be—should it cover only a few markets, or many? Second, does it matter which markets are deregulated first and which last? And third, should reform be gradual or rapid?

Economic theory suggests that welfare will be raised unambiguously if all market distortions are removed. However, the welfare implications of partial reform are ambiguous. In a “second best” situation, the impact of a particular measure would depend on the circumstances in other sectors of the economy. If there are distortions elsewhere in the economy, then the removal of distortions in one sector may well lead to a reduction in aggregate welfare. Therefore, analysis of the effects of reform, if limited to the specific markets directly affected, may yield misleading results. While certain considerations (such as the degree of rent-seeking activity and the greater incentive effects of competition) suggest that the freeing up of any market would yield welfare benefits, the overall implication of the “theory of second best” is that structural reform should be comprehensive in scope in order to ensure a “first best” solution in the long run (Krueger (1984)). However, in a world with short-term adjustment costs there are economic and political reasons why immediate and concurrent reform of all markets may be neither feasible nor desirable (Edwards (1984)).

If all markets cannot be liberalized simultaneously, the issue of sequencing arises. Sequencing refers to the order in which individual markets should be liberalized in the transition towards equilibrium. Almost all of the studies on sequencing concern major and fundamental “regime” changes, and deal with the transition from a highly distorted, financially repressed, and inflation-ridden economy into a market-oriented, open economy.21 Because most of the structural reforms now being considered in the industrial countries are relatively minor compared to such comprehensive stabilization-cum-liberalization programs, the implications of the sequencing literature for the present study may be somewhat limited.22 In addition, the sequencing issue has been analyzed mainly from an individual country perspective, with little consideration given to effects on other countries.

Notwithstanding these caveats, one of the most important implications of the sequencing literature is that governments should begin by freeing those markets where prices tend to be less responsive—for example, labor markets should be deregulated before goods markets, and both labor and goods markets before financial markets. A rigid labor market could stifle adjustment to trade liberalization if workers fail to shift out of declining industries and into expanding ones, with unemployment as the inevitable result. A corollary of this principle is that the current account of the balance of payments should be liberalized before the capital account, given that adjustment tends to be more rapid in asset than in goods markets.23 The difference in the speed of adjustment in these markets has several implications.

First, in an uncertain world opening the current account first (by liberalizing trade in goods and services) provides policy makers time to examine the market’s reaction and to adjust policies. When the capital account is opened up, the initial reaction is likely to be fast and the resulting flows large.

Second, it is easier and probably less costly from a social viewpoint to reverse incorrect portfolio decisions than to reverse suboptimal real investment decisions. If the capital account were opened first, portfolio decisions would be made on the basis of price signals reflecting undistorted market conditions. However, real investment decisions would still be carried out according to distorted signals as long as the current account remained subject to restrictions. Because of these distortions, the social cost of these investments is likely to exceed the private cost. These real investment decisions would become uneconomic and would have to be reversed once the trade account was liberalized.

Third, the cost of a distortion depends both on the distortion itself and on the volume of transactions that takes place in the presence of the distortion. If the capital account were opened up first, the cost of the remaining distortion affecting the trade account would be proportional to the volume of capital flows, which is likely to be quite large in view of the high speed of adjustment in asset markets (Frenkel (1982)).

It may be useful to evaluate structural reform in industrial countries during the past decade against the background of these theoretical guidelines for the sequencing of structural reform. In virtually every country domestic capital markets have been deregulated and international capital controls have been lifted before the distortions affecting goods markets and labor markets were eliminated. From a theoretical perspective, this implies that liberalization has occurred in reverse order. According to some analysts, this sequence helps explain why some comprehensive reform programs have not lived up to earlier expectations and why it is urgent to complete the process through liberalization of current account transactions and reform of labor markets.24

With regard to the issue of timing, there are some conceptual grounds for thinking that a relatively rapid removal of distortions may be the least costly way of proceding: new signals in place would prevent continued resource misallocation in response to altered signals before the liberalization process is complete. Moreover, rapid implementation of all aspects of structural reform would reduce uncertainty about the commitment of the authorities to fully implement such measures, allowing economic agents to make decisions sooner. On the other hand, gradualism in structural adjustment would allow any short-run. Transitional costs to be spread over time, and it would allow mistakes to be corrected in a less costly way. The balance of pros and cons has suggested to several analysts that the best reform strategy to adopt is a preannounced timetable of reform measures, thereby allowing investors to anticipate early the eventual pattern of incentives (Frenkel (1983)).

Of course, preannouncement of structural reform measures will be effective only to the extent that the policy process has credibility. If the private sector expects a subsequent reversal of measures, it may decide to postpone adjustment, or it may even respond in ways that are counterproductive. In a fundamental sense, lack of credibility has effects that are quite similar to those of market distortions (like a production tax)—it impairs the beneficial effects of an otherwise desirable policy. For example, a measure to liberalize trade that is expected to be reversed in the near future may fail to stimulate investment in the tradable goods sector and may even result in destabilizing and inefficient capital market speculation if capital markets are free of controls.25 Recognition of this possibility does not imply that capital controls should be retained or reintroduced, but rather that reform efforts in other areas should be accelerated.

Political Economy of Structural Reform

Even if the problems of optimal sequencing and credibility could be solved, there would remain important problems of implementation related to the political economy of structural reform. The analysis of these problems has given rise to a vast literature. A comprehensive survey of this literature is beyond the scope of this paper, but some of the important themes that emerge deserve to be examined. These themes suggest that various structural distortions have different political and economic roots, thus pointing to different solutions to the problems of structural reform policy in various areas.

The first theme stems from that part of trade theory that deals with rent seeking. Early work suggested that calls for protection might be based on the factor intensity of industries: this suggestion was based on the now familiar Heckscher-Ohlin trade model, which implies that protection of an industry would increase the relative return to the factor of production used intensively in that industry. The factor with the strongest political representation would then seek protection for the industry that used it intensively. Further work has emphasized the importance of factor immobility in explaining pressures for protection; when factors are immobile, the costs of adjustment could exceed the gains from resource reallocation.

The rent-seeking approach was further developed by a more sophisticated breakdown of factors of production and a link to the institutions of voting. Because lower-skilled workers faced greater and perhaps unrecoverable adjustment costs when relocation was required, they would use their majority position in the political process to seek protection. A more general version of this approach evolved into a blend of location theory and political representation theory: when factors of production (labor, land, and capital) are regionally concentrated and factor mobility is low, the immediate interests of local residents may conflict with the long-run interests of society. One policy implication of these theories is that enhancing education, retraining, and the mobility of labor would help to reduce the political incentives to seek protection.

Although these economic theories of structural distortions highlight the incentives for agents to seek protection, they fail to explain why some interests succeed in achieving this goal and why others fail. To address this aspect of the political economy of distortions, a body of literature has developed around the concept of “organizing costs” for interest groups.26 The essential insight is that small, focused interest groups are less costly to organize than large, diffuse ones. Therefore, small, concentrated groups are likely to be overrepresented in the political process. Small groups (or coalitions of small groups) may benefit from the introduction of market distortions that in effect impose taxes on the (less organizable) remainder of society. Two ways of avoiding such rigidities have been suggested. The first is to organize political institutions that make successful coalition forming too costly.27 The second is to provide better information to the general public about the advantages of free trade, and more generally of distortion-free markets, and about the costs of distortions and rigidities, so that the coalitions that favor introduction or maintenance of distortions can be defeated.

One difficulty with the approach to distortions based on organizing costs is that it assumes that public sector decision makers (politicians and public sector employees) have no interests of their own. Challenging the realism of this assumption, an alternative theory of distortions has developed that emphasizes the economic interests of the public sector.28 The public sector is viewed as an interest group in itself, and uses its administrative powers to acquire economic benefits, sometimes at the expense of the general public. When these actions fail, for example, by spawning a political crisis such as a tax revolt, structural reforms must be undertaken. These theories predict that quick reform may be preferred to gradual reform, for three reasons. First, quick reform prevents anti-reform groups from organizing effectively. Second, the public sector-—politicians in particular—may have a higher discount rate than society at large, which implies that results must be seen rapidly for reform to be successful. And third, confidence in the sustainability of reform, which is essential for success, is easier to maintain when major actions are taken at an early stage.

Each of these theories of how distortions are introduced and maintained can contribute to the understanding of how market distortions arise and how they can be eliminated. For example, theories about rent seeking point to the importance of labor mobility. Structural reforms that enhance factor mobility are likely to alleviate pressures for protection in the future. Also, organizing cost theories point to reforms of political institutions that would make coalition formation more difficult. Economic reforms that raise market competition are likely to increase political competition as well, because competing economic interests cannot be sponsored easily by single politicians or small political organizations. The result of political competition is that the coalitions that tend to introduce distortions would be harder to form in the first place. Finally, public sector interest theories point to the benefits of competition between political jurisdictions and to the importance of acting quickly when opportunities for reform arise.

Need and Scope for Further Structural Reform

Political difficulties notwithstanding, the policy agenda in the industrial countries over the medium term will certainly include further structural reform. The policy recommendations that follow from the analysis presented in this paper reinforce those recently outlined by the staff in the World Economic Outlook (October 1988). These recommendations are summarized below.

The area that holds the greatest promise for structural reform is that of labor markets. Rigidities in these markets are a central factor behind the persistence of high rates of unemployment, especially in Europe. Indicators of labor market conditions (including estimates of real wage gaps and of the natural rate of unemployment) and model simulations suggest that the output gains from labor market reform could easily be as high as several percentage points of GNP in some countries. Improvements in real wage flexibility could also bring significant gains by reducing the output and employment gaps resulting from external shocks. Moreover, as the analysis in Bayoumi, Feldman, Wattleworth, and Woglam (below) indicates, these gains would be shared by both the country instituting the reforms and its trading partners.

The removal of border (tariff and non-tariff barriers) and non-border (explicit and implicit domestic subsidies) trade restrictions should also be given high priority.29 Empirical studies based on the “new international trade theories” (e.g., incorporating the effects of economies of scale and imperfect competition) suggest that welfare and output gains from these reforms can be large. These gains would accrue both to the initiating country and to its trading partners, which underscores the importance of international coordination of trade policies in the context of the Uruguay Round of GATT negotiations, Moreover, a concerted move toward freer world trade in all goods and services would facilitate the international adjustment mechanism by lowering the disproportionately large costs of adjustment currently borne by the nonsubsidized, and more vigorous, sectors of the economy.

Because tax systems affect virtually all markets for goods, services, and factors in industrial countries, they impinge directly on microeconomic efficiency and overall economic performance. Considerable progress has been achieved in reforming tax systems in many industrial countries, notably by lowering personal and corporate income tax rates while broadening the tax base. However, tax systems in several countries continue to include important elements (including certain tax preferences) that distort decisions to work, save and invest, for example by tilting the allocation of saving in favor of real estate at the expense of business fixed investment. Removal of these distortions would make a significant contribution toward raising economic efficiency and growth. Moreover, international differences in taxation among countries could lead to distortions in the international pattern of saving and investment, suggesting the need to complement tax reform at the national level with efforts to harmonize tax policies.

Much has already been achieved in the area of financial market deregulation. The benefits of these reforms include increased competition and the associated wider array of financial services that are available at lower cost, as well as the increased efficiency and speed with which savings are transformed into productive uses. The high degree of integration and speed with which the world capital markets adjust was amply demonstrated in the global stock market decline of October 1987. The relatively rapid speed with which capital markets adjust magnifies the cost of the remaining distortions in goods and factor markets—in terms of their effects on both output and current account balances—and reinforces the ease for accelerating the completion of comprehensive structural reform of all markets.


  • Adams, Charles, Paul Fenton, and Flemming Larsen, “Potential Output in Major Industrial Countries,” Staff Studies for the World Economic Outlook, World Economic and Financial Surveys (Washington: International Monetary Fund, 1987).

    • Search Google Scholar
    • Export Citation
  • Adams, Charles, Paul Fenton, and Flemming Larsen, “Differences in Employment Behavior among Industrial Countries,” Staff Studies for the World Economic Outlook, World Economic and Financial Surveys (Washington: International Monetary Fund, 1986).

    • Search Google Scholar
    • Export Citation
  • Aizenman, Joshua, and Jacob Frenkel, “Optimal Wage Indexation, Foreign Exchange Intervention, and Monetary Policy,” American Economic Review (Nashville, Tennessee), Vol. 75 (June 1985).

    • Search Google Scholar
    • Export Citation
  • Argy, Victor, and Joanne Salop, “Price and Output Effects of Monetary and Fiscal Policy Under Flexible Exchange Rates,” Staff Papers, International Monetary Fund (Washington), Vol. 26 (June 1979).

    • Search Google Scholar
    • Export Citation
  • Bean, C.R., P.R.G. Layard, and S.J. Nickell, “The Rise in Unemployment: A Multi-Country Study,” Economica (London), Vol. 53 (1986).

  • Brainard, William C., “Uncertainty and the Theory of Policy,” in Targets and Indicators of Monetary Policy, ed. by K. Brunner (San Francisco: Chandler Publishing, 1969).

    • Search Google Scholar
    • Export Citation
  • Buchanan, James, and Gordon Tullock, The Calculus of Consent (Ann Arbor: University of Michigan Press, 1962).

  • Calvo, Guillermo, “Notes on Credibility and Economic Policy,” (Washington: unpublished, abridged version of author’s “Costly Trade Liberalizations,” in International Monetary Fund, Staff Papers, September, 1988).

    • Search Google Scholar
    • Export Citation
  • Cecchini, Paolo, 1992: The European Challenge (London: Gower, 1988).

  • Commission of the European Communities, “The Economics of 1992,” European Economy, Vol. 35 (March 1988).

  • Edwards, Sebastian, The Order of Liberalization of the External Sector in Developing Countries (Princeton: Princeton Essays in International Finance No. 156, December 1984).

    • Search Google Scholar
    • Export Citation
  • Fleming, J. Marcus, “Domestic Financial Policies Under Fixed and Under Floating Exchange Rates,” Staff Papers, International Monetary Fund (Washington), Vol. 9 (November 1962), pp. 36979.

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., “The Order of Economic Liberalizaion: Lessons From Chile and Argentina: A Comment” in Economic Policy in a World of Change, ed. by K. Brunner and A.H. Meltzer (Amsterdam: North-Holland, Carnegie-Rochester Conference Series on Public Policy No. 17, 1982).

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., “Panel Discussion on Southern Cone,” Staff Papers, International Monetary Fund (Washington), Vol. 30 (March 1983).

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., and Assaf Razin, Fiscal Policies and the World Economy (Cambridge, Massachusetts: MIT Press, 1987).

  • Frenkel, Jacob A., and Assaf Razin, (1988a) “Budget Deficits under Alternative Tax Systems,” Staff Papers, International Monetary Fund (Washington), Vol. 35 (June 1988).

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., and Assaf Razin, (1988b), “International Effects of Tax Reforms,” (Washington: International Monetary Fund Working Paper, WP/88/62, July 12, 1988).

    • Search Google Scholar
    • Export Citation
  • Fukao, Mitsuhiro, “Balance of Payments Imbalances and Longterm Capital Movements: Review and Prospects,” (Paper prepared for the Fourth EPA International Symposium, March 15-17, 1988).

    • Search Google Scholar
    • Export Citation
  • Gray, Joanna, “Wage Indexation: A Macroeconomic Approach,” Journal of Monetary Economics (April 1976).

  • Harberger, Arnold C., “Monopoly and Resource Allocation,” American Economic Review, May 1954.

  • Harberger, Arnold C., and M. Hanazaki, “Internationalization of Financial Markets: Some Implications for Macroeconomic Policy and for the Allocation of Capital” (Paris: Organization for Economic Cooperation and Development, Working Paper No. 37, November 1986).

    • Search Google Scholar
    • Export Citation
  • Hernández-Catá, E., “Issues in the Design of Growth Exercises,” (Washington: International Monetary Fund Working Paper, WP/88/65, July 20, 1988).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (1988a), “Indicators of Structural Policies and Performances,” Annex III in World Economic Outlook, World Economic and Financial Surveys (Washington: International Monetary Fund, October 1988).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (1988b), World Economic Outlook: Prospects and Issues (Washington: World Economic and Financial Surveys, October 1988).

    • Search Google Scholar
    • Export Citation
  • Isard, Peter, “The Implications of Fiscal Conditions and Growing Internationalization for Monetary Policies and Financial Market Conditions” (Washington: International Monetary Fund Working Paper, WP/88/52, June 29, 1988).

    • Search Google Scholar
    • Export Citation
  • Isard, Peter, and Liliana Rojas-Suarez, “Velocity of Money and the Practice of Monetary Targeting: Experience, Theory, and the Policy Debate,” in Staff Studies for the World Economic Outlook (Washington: International Monetary Fund, July 1986).

    • Search Google Scholar
    • Export Citation
  • Kelly, Margaret, Naheed Kirmani, Miranda Xafa, Clemens Boonekamp, and Peter Winglee, Issues and Developments in International Trade Policy, Occasional Paper 63 (Washington: International Monetary Fund, December 1988).

    • Search Google Scholar
    • Export Citation
  • Khan, Mohsin and Roberto Zahler, “The Macroeconomic Effects of Changes in Business to Trade and Capital Flows: A Simulation Analysis,” Staff Papers, International Monetary Fund (Washington), Vol. 30, No. 2 (June 1983).

    • Search Google Scholar
    • Export Citation
  • Kincaid, G. Russell, “Policy Implications of Structural Changes in Financial Markets,” Finance and Development (Washington: International Monetary Fund and World Bank, March 1988).

    • Search Google Scholar
    • Export Citation
  • Krueger, Anne O., “Problems of Liberalization” in World Economic Growth, ed. by A. Harberger (San Francisco: ICS Press, 1984).

  • Krueger, Anne O., “How To Liberalize a Small Open Economy” in The Economics of the Carribean Basin, ed. by M. Connolly and J. McDermott (New York: Praeger, 1985).

    • Search Google Scholar
    • Export Citation
  • Lal, Deepak, “The Political Economy of Economic Liberalization,” World Bank Economic Review Vol. 1 (Washington: 1987).

  • Madison, James, The Federalist Papers No. 10, (New York: New American Library Edition, 1961; originally published in 1787).

  • Mayer, Thomas, “Economic Structure, the Exchange Rate and Adjustment in the Federal Republic of Germany: A General Equilibrium Approach,” (Washington: International Monetary Fund Working Paper, WP/88/79, September 1, 1988).

    • Search Google Scholar
    • Export Citation
  • McKinnon, Ronald I., “The Order of Economic Liberalization: Lessons from Chile and Argentina” in Economic Policy in a World of Change, ed. by K. Brunner and A.H. Meltzer (Amsterdam: North-Holland, Carnegie-Rochester Conference Series on Public Policy No. 17, 1982).

    • Search Google Scholar
    • Export Citation
  • Mueller, Dennis C., The Political Economy of Growth (New Haven: Yale University Press, 1983).

  • Mundell, Robert A., International Economics (New York: Macmillan, 1968).

  • OECD, “Empirical Research on Trade Liberalization with Imperfect Competition: A Survey” (Paris: OECD, CPE/WP1(88)7, Annex 1, September 27, 1988).

    • Search Google Scholar
    • Export Citation
  • Olson, Mancur, The Logic of Collective Action (Cambridge, Massachusetts: Harvard University Press, 1965).

  • Pechman, Joseph, World Tax Reform: A Progress Report (Washington: The Brookings Institution, 1988).

  • Rosenblatt, Julius, Thomas Mayer, Kasper Bartholdy, Dimitrios Demekas, Sanjeev Gupta, and Leslie Lipschitz, The Common Agricultural Policy of the European Community, Occasional Paper 62 (Washington: International Monetary Fund, November 1988).

    • Search Google Scholar
    • Export Citation
  • Stockman, Alan C., “The Order of Economic Liberalization: Lessons From Chile and Argentina: A Comment” in Economic Policy in a World of Change, K. Brunner and A.H. Meltzer (Amsterdam: North-Holland, Carnegie-Rochester Conference Series on Public Policy No. 17, 1982).

    • Search Google Scholar
    • Export Citation
  • Takenaka, Heizo, and Naoko Ishii, “Simulation of Policy Coordination Under Structural Adjustment: An Estimate of the Economic Impact of the Mayekawa Report,” (Tokyo, unpublished, 1987).

    • Search Google Scholar
    • Export Citation
  • Tanzi, Vito and A. Lans Bovenberg, “Economic Interdependence and the International Implications of Supply-Side Policies,” (Washington: International Monetary Fund Working Paper, WP/88/71, August 4, 1988).

    • Search Google Scholar
    • Export Citation
  • Watson, Maxwell, Donald Mathieson, Russell Kincaid, David Folkerts-Landau Klaus Regling, and Caroline Atkinson (1988), International Capital Markets: Developments and Prospects, World Economic and Financial Surveys (Washington: International Monetary Fund, January 1988).

    • Search Google Scholar
    • Export Citation

The authors gratefully acknowledge many useful comments and suggestions from their colleagues throughout the Fund, particularly Charles Adams. Guillernio Calvo. Max Corden. And Jueob Frenkel.


The terms “structural reform” and “structural policy” cover a wide variety of measures (ranging from financial markei deregulation to import liberalization), which can have widely different effects on specific variables and sectors. The conclusions derived in this section should therefore be interpreted as broad generalizations which must be modified in individual cases to take into account the specific effects of particular structural measures.


As discussed below, investment might also rise as a result of an increase in the rate of return on capital.


It may be noted that increasing price flexibility in one market, while preserving rigidities in other markers, might have undesirable consequences. For example. Full exchange rate flexibility combined with price rigidity could—although it need not—result in exchange rate overshooting. This is a special case of the more general proposition that the partial removal of structural rigidities could have the effect of lowering welfare. (See below.)


These concepts are discussed in Adams, Fenton, and Larsen (1986).


In the longer run, the cost of allowing inflation to persist at a high rate is likely to exceed the short-run costs of reducing inflation to an acceptable level.


A “Harberger triangle” measures the loss in social welfare resulting from the misallocation of resources due to market distortions. The “Okun gap” measures the difference between potential and actual output.


For a comprehensive review of the new trade theories, see OECD (1988).


For an assessment of recent tax reforms in industrial countries, see Pechman (1988).


One study (Bean and others (1986)) that disaggregated the rise in European NAIRUs concluded that the most important factors related to job search intensity (which highlighted the role of unemployment benefits) rather than to higher non-wage labor costs (such as health insurance or pension benefits); the opposite appeared to be true in the rest of the industrialized world. Even in Europe, however, the link between the duration of unemployment and unemployment benefits appears to depend on the state of the labor market, with the relationship being weaker in periods of slack. Empirical studies of the effects of recent labor market reforms have had difficulty in establishing firm statistical support for the relationship between specific policy measures and actual wage behavior.


Accordingly these policies can enhance the effectiveness of—and reduce the adjustment costs associated with—appropriate demand management policies.


Two qualifications are necessary (see Hernández-Catá (1988)). First, long structural policy lags appear to be partly related to the private sector’s expectations about the durability of structural measures—the higher the perceived probability of policy reversal, the more delayed the response. Second, initial conditions are important. Where the operation of market forces has been suppressed for an extended period, immediate adjustments following a structural policy change could imply rapid economic effects.


See Hernández-Catá (1988). The introduction of uncertainty about the relationships between economic policies and objectives introduces the need for “redundancy” in the use of policy measures. Brainard (1969) has shown that Tinbergen’s principle of equality between the numbers of instruments and targets breaks down in the presence of uncertainty; it will normally be efficient to use all the available instruments in order to maximize the probability of achieving policy objectives, particularly if there is a cost involved in changing policy instruments.


Structural measures can have significant fiscal implications (for example, measures to stimulate private saving and investment or to reduce tariff protection could directly reduce government revenue). It is equally clear that fiscal policies can have significant implications for the efficiency of resource allocation as well as for the aggregate levels of saving and investment. Thus, fiscal and structural measures often are closely interrelated, and it is important to take these interrelations into account in evaluating the impact of these measures on economic growth.


Much of the material in this subsection draws on the work of Fukao and Hanazaki (1986) and Kincaid (1988). For an analytical treatment and a survey of the issues discussed in this subsection, see Watson and others (1988) and Fukao (1988).


In addition to the degree of financial liberalization, the domestic and international incidence of fiscal measures also depends on the timing of changes in specific taxes or expenditure components. In an intertemporal setting, the entire profile of interest rates, real exchange rates, and temporal and intertemporal allocations of output and demand depend on the sequence in which various budgetary measures are introduced (Frenkel and Razin (1987 and 1988a)).


Financial market liberalization has had important implications for the stability and control of the monetary aggregates. For a survey of these issues see Isard and Rojas-Suarez (1986) and Watson and others (1988).


See Frenkel and Razin (1988b) and Tanzi and Bovenberg (1988). p. 14. Isard (1988) also emphasizes the important link between fiscal policy changes that affect after-tax returns on capital located in different countries and exchange rate movements.


For example, the shift from an income to a consumption tax will broaden the tax base if a country has an external deficit (i.e., where expenditure exceeds income) and will narrow it in the case or an external surplus. Consequently, a revenue-neutral tax reform would require a reduction in tax rates in the former case and an increase in tax rates in the latter case See Frenkel and Razin (1988b).


Surveys of the literature in this area include McKinnon (1982), Frenkel (1982 and 1983), Edwards (1984), and Krueger (1984 and 1985).


There are very few empirical studies of the sequencing issue in industrial countries. For a simulation study of alternative sequencing of structural reforms in a “typical” Latin American economy, see Khan and Zahler (1983).


For two dissenting views that reach the opposite conclusion, see Stockman (1982) and Lal (1987). Stockman argues that the essence of a successful liberalization program is policy credibility, and moreover, that credibility can be established only by simultaneously making major reductions in controls in all parts of the economy. Lal also emphasizes policy credibility and political difficulties in the reform process. We discuss the credibility issue below and the political economy of structural reform in a subsequent section.


The difficulties associated with inadequate attention to the sequencing issue have been summarized by Kincaid (1988) as follows:

… rising protectionism—particularly nontariff barriers—among industrial countries and structural rigidities in the nonfinancial sectors of these economies have hampered the ability of goods and labor markets to respond to changes in macroeconomic policies and external developments. Meanwhile, keener competition, reduced market segmentation, and new technology have increased the speed at which financial markets adjust to these same factors. A greater divergence in the adjustment of these two elements could increase volatility—overshooting of interest rates and exchange rates—as financial markets react more sharply than before to compensate for rigidities elsewhere in the economic structure. Indeed, changes in exchange rates may need to be greater to have the same aggregate impact on the trade account, while the incidence of adjustment in the trade account would fall more heavily on the segments of the traded-goods sector that had not been accorded special protection.


Calvo (1988) offers an intuitive example in which the authorities announce that all tariffs will be permanently abolished, but the private sector expects that such a policy will be modified in the future and that a new set of tariffs will be put into place. If there is no substitution in production and consumption, but consumption goods can be stored at low cost, the lack of credibility of the tariff reform will trigger speculative accumulations of inventories. Although this activity may be profitable from the point of view of the private sector, it will not necessarily be desirable for the society as a whole because the accumulation of inventories may crowd out the acquisition of other types of assets with a higher social rate of return. Note also that (as discussed earlier in the text) the social cost of this misallocation is a function of the structure of capital markets. In a regime of perfect capital mobility, unlimited funds could he obtained from abroad for speculation in inventories: if capital movements were controlled internationally and domestically, a lesser amount of capital could be lied up in this form.


Pioneering works in this field are those of Buchanan and Tullock (1962), and Olson (1965). A compendium of articles that applies these ideas to comparison of growth performance across countries is provided by Mueller (1983).


This was the essence of Madison’s argument in favor of a federal government for the United States; see Madison (1787).


These ideas have their modern starting point in Buchanan and Tullock (1962). A recent summary is given in Lal (1987).


A thorough review of the issues in the trade area, including the role of non-tariff barriers, is contained in Kelly and others (1988).