The political economy of nontariff barriers: a cross-national analysis.

Much research on the determinants of trade policy has focused on the efficacy of societal and statist approaches. Societal theories typically attribute patterns of protection to variations in demands made by pressure groups, whereas statist theories emphasize the effects of the "national interest" and domestic institutions in determining the level of protection. While both approaches have gained considerable currency, debates concerning their relative merits have been heated and long-standing. Yet very little quantitative evidence has been brought to bear on this topic.

In this article, we provide some of the first results of this sort. Our findings indicate that although societal and statist approaches often are considered mutually exclusive, it is more fruitful to view them as complementary. Moreover, the interaction between factors that give rise to demands for protection and those that regulate the provision of protection by policymakers has not been treated adequately in the literature on foreign economic policy. This gap in the literature is fundamentally important, since our results indicate that the interaction between these factors is a central determinant of trade policy. Thus, analyses of commercial policy that fail to consider both societal and statist variables and the interaction between them are likely to be inadequate.

Our analysis centers on explaining cross-national patterns of nontariff barriers (NTBs). Scholars have conducted little cross-national research on trade policy and virtually none with a focus on NTBs. Instead, single-country studies of tariffs comprise much of the existing literature on the political economy of commercial policy. Yet the usefulness of societal and statist theories of foreign economic policy hinges on the ability of these theories to explain variations in protection across states, and NTBs have become increasingly pervasive among the advanced industrial countries. Because the General Agreement on Tariffs and Trade (GATT) and the World Trade Organization (WTO) limit the ability of contracting parties to impose tariffs, policymakers who view protection as an attractive means by which to meet the demands of pressure groups or advance state interests are likely to rely primarily on NTBs. Many observers have suggested that this is occurring with increasing regularity and that the recent proliferation of NTBs has done much to offset the gains in liberalization made during successive rounds of the GATT. A fuller understanding is therefore needed of the factors that account for variations in NTBs across states.

Societal approaches to trade policy

Societal (or pluralist) approaches to the study of foreign economic policy focus primarily on the effects of demands for protection by pressure groups. Societal explanations consider trade policy to be the product of competition among pressure groups and other nonstate actors that are affected by commerce. The impact of these groups on policy depends largely on their ability to organize for the purpose of articulating their demands and on the amount of electoral influence they possess. Societal approaches attribute little importance to policymakers and political institutions for the purposes of explaining trade policy. As G. John Ikenberry, David Lake, and Michael Mastanduno point out, societal theories view the state as "essentially passive; it acts as a disinterested referee for competing groups, and supplies policies to satisfy the demands of successful domestic players."(1)

Societal approaches to the study of trade policy characterize much of the literature on endogenous protection.(2) Empirical studies of this sort infer the demands for protection based on macroeconomic and/or sectoral fluctuations. Most analyses of endogenous protection conducted by political scientists have been cast at the sectoral level. A large and growing body of literature, however, centers on the macroeconomic determinants of protection. Much of this research supports the view advanced by certain societal theories that macroeconomic fluctuations strongly influence pressures for protection.(3) Therefore we focus our societal analysis of NTBs on macroeconomic factors.

Chief among the macroeconomic variables that these studies emphasize are unemployment and the real exchange rate. It is widely accepted by analysts of trade policy that high levels of unemployment contribute to demands for protection. Indeed, according to C. Fred Bergsten and William Cline, "Conventional wisdom suggests that high levels of unemployment are the single most important source of protectionist pressures."(4) Widespread unemployment increases the costs to workers of adjusting to rising import levels. Workers who are displaced by imports will find it progressively more difficult to obtain alternative employment, and when they do, downward pressure will be placed on their wages. Together these factors promote pressures to restrict the flow of imports.(5)

In addition to unemployment, variations in the exchange rate are expected to give rise to protectionist pressures. In fact, Rudiger Dornbusch and Jeffrey Frankel argue that "hypotheses concerning the exchange rate may be the most important macroeconomic theories of protection."(6)

Central to the effects of the exchange rate on demands for protection is the influence of the price of a state's currency on the competitiveness of its exports and its import-competing products. An appreciated currency, by increasing the price of domestically produced goods, threatens to undermine both exports and import-competing sectors of the economy. As C. Fred Bergsten and John Williamson point out in a related context, these developments are likely to contribute to "pressure that is generated for protectionist measures. Export-and import-competing firms and workers will tend to seek help from their governments to offset these distortions, which undermine their ability to compete, with some degree of legitimacy since the distortions are accepted - in some cases, even fostered - by those governments. Coalitions in support of trade restrictions will be much easier to form, and much broader in their political clout, because no longer will only the most vulnerable firms and workers be seeking help - and no longer will the countervailing pressures from successful exporters be as effective."(7)

Public officials in liberal democracies are expected to meet demands for protection that arise due to high levels of unemployment and an appreciated currency because these variables influence the voting behavior of constituents. There is evidence that voters cast ballots on the basis of their personal economic circumstances, especially if they are recently unemployed.(8) However, substantial evidence also indicates that voters cast ballots on the basis of macroeconomic conditions, regardless of whether they are directly affected by these conditions.(9) In fact, some studies have concluded that macroeconomic factors are more salient determinants of voting behavior than are personal economic circumstances. Other survey research further suggests that public support for protection increases during downturns in the economy and when domestic industries are under severe pressure from foreign competition.(10) As a result, public officials seeking to enhance their electoral fortunes have incentives to impose protection during periods of high unemployment and currency appreciation because such measures are likely to be popular and may blunt the short-term effects of macroeconomic pressures. These analyses therefore lead us to expect a direct relationship to exist between both the level of unemployment and the real exchange rate, on the one hand, and the incidence of NTBs, on the other hand.

Statist approaches to trade policy

While societal approaches have been especially influential in the field of political economy, they also have been criticized on a number of grounds. Especially important is the charge leveled by statists and others that societal approaches systematically underestimate the effects of two factors that regulate the provision of protection: state interests with respect to trade policy and domestic institutions.(11) Analyses that emphasize state interests generally focus on the roles of politicians and policymakers in the formation of trade policy, holding constant societal pressures. Further, as Ikenberry, Lake, and Mastanduno note, these analyses presume that the preferences of public officials "are partially, if not wholly, distinct from the parochial concerns of either societal groups or particular government institutions, and are tied to conceptions of the 'national interest' or the maximization of some social welfare function."(12)

Many statists conclude that the ability of policymakers to advance the national interest depends in large measure on the extent to which domestic political institutions render them susceptible to demands by pressure groups and other nonstate actors.(13) Policymakers who are poorly insulated from, and lack autonomy with respect to, pressure groups will face difficulty advancing the national interest unless (as discussed further below) it converges with the preferences of societal groups. Thus, one hypothesis we will test is that institutional factors that foster the insulation and autonomy of public officials bolster the ability of states to pursue trade policy consistent with the national interest.

Relative size

Clearly, the national interest with respect to trade is likely to vary across states; and it is not possible to assess adequately the influence of institutional factors on trade policy from a statist perspective unless each state's interest can be specified. On this score, many analysts have argued that a state's economic size governs its national interest with respect to trade policy.

There is ample reason to expect that larger states will display a more pronounced interest in protection than their smaller counterparts. First, international trade theory suggests that this should be the case. By virtue of their size, large states are likely to be vested with disproportionate market power.(14) They can exploit their monopoly power through the use of tariffs, as well as quotas and other NTBs that duplicate a tariff's effect.(15) If the imposition of an optimal quota elicits retaliation, the welfare of both parties will suffer.(16) This, however, only limits the incentives for a large state to impose NTBs against a state of similar size, since only states with some monopoly power have an incentive to retaliate in response to the imposition of protection. Large states retain an incentive to target small states, since the latter have no incentive to retaliate. In contrast, small states are unlikely to possess the market power necessary to benefit from optimal protection and face the prospect of retaliation by trade partners (thereby reducing foreign commerce on which they tend to be highly dependent) if they impose NTBs. Hence, on average, we expect larger states to display a greater preference for NTBs than their smaller counterparts.

Second, state size is likely to be directly related to patterns of protection due to the time period analyzed in this article. As discussed further below, the empirical analysis conducted in this study is based on the mid-1980s. In the opinion of many scholars, this was a period characterized by a moderately skewed distribution of power among a few relatively large nonhegemonic states. A number of studies have concluded that systems of this sort - as well as ones in which hegemony is declining - provide incentives for the dominant states to behave in a commercially predatory manner.(17) Based on these considerations, we expect that economic size will be directly related to the incidence of NTBs.

We measure a state's relative economic size in two ways: the ratio of its imports to total global imports and the ratio of its gross domestic product (GDP) to total global GDP. The first variable has been used repeatedly as a measure of economic size.(18) The second is also important because states with relatively large GDPs are likely to possess greater market power and to be better able to forgo commerce than are states with relatively small GDPs.(19)

Although it is clear that these two measures of relative size should be highly correlated, analyzing both allows us to determine whether our empirical results are sensitive to the measure that is used. Moreover, analyzing both measures of size is important because the ratio of national imports to global imports is closely related to the measure of trade dependence used by Ronald Rogowski; and his analysis implies that any observed effect on …