David Tarr ¹
The World Bank

August, 1998

An effective trade policy is central to the integration of Russia in the international economic system and the growth that will generate: Tariff policy is the centerpiece of trade policy in a market system. Tariffs are, with very few exceptions, the only acceptable policy tool for protection under the GATT/WTO; and they are superior to alternative instruments of protection, such as quotas, licenses and technical barriers to trade. Today's Russian tariff averages 13-14% with ad valorem maximum tariffs of 30% but with some specific tariffs that have ad valorem tariff equivalents higher than 30%.

The analysis in this paper strongly suggests that there is little economic justification and there are many dangers in providing differentiated tariff protection to various sectors of industry and agriculture in Russia today. Most of the arguments in which tariffs differentiated by product are the optimal policy intervention are of little practical significance: There are few products in which Russia has monopsony power or industries in which it can gain advantage through "strategic" application of tariffs.

In many circumstances where tariffs are second best policy intruments, such as to raise public revenue or to cope with balance of payments problems, a uniform tariff rate is the most practical and efficient alternative. Where Russia may be interested in using the tariff as a bargaining instrument in multilateral negotiations, it is immaterial whether the tariff is uniform or differentiated--the issues have to do with its capacity to use the tariff as a bargaining instrument and what it bargains for. Differentiated tariff protection in support of infant or restructuring industries is typically ineffective at addressing the alleged market failure problem; governments are not very good at picking winners and there are serious dangers that the policy would be overwhelmed by requests for protection from vested interests irrespective of its economic merits.

A uniform tariff conveys a number of advantages, the most important of which is that if the tariff is uniform, the gains to industry lobbying are much smaller (and may be negative), creating a kind of free-rider problem for the lobbying industry and dramatically reduces the incentive to lobby for protection. Then: (1) the level of protection is likely to be lower; (2) there is a direct saving of resources from the reduced lobbying; (3) the reduction to the gains from lobbying for protection provides a vastly improved signal to valuable entrepreneurial talent which will thus be encouraged to create better and cheaper products; and (4) the reduction in resources devoted to lobbying will result in less corruption in government, which may have positive spillover effects into other dimensions of government activity.

The findings of this paper lead to the following policy conclusions:

• Eliminate Technical Barriers to Trade so that protection is only by tariffs.
• Sequentially Reduce the Maximum Rates Until the Maximum is 10% by 2001.
• Sequentially Reduce and Eliminate Specific Tariffs to improve transparency.
• Raise the Low Rates so the minimum is 10% by 2001 and Eliminate Exemptions except to Provide Duty Free Access to Imported Inputs Used in Exports
• Bind Tariff Rates at the WTO at Rates Closer to Applied Rates

Worldwide experience in the last 50 years demonstrates the benefits of open trade regimes: The OECD countries brought trade barriers down through successive WTO negotiations and experienced sustained growth in trade and incomes. Many developing countries governments initially felt differently and attempted to promote industrialization behind high protective barriers. But in the last ten years or so, the balance of opinion has shifted in these countries as well, as evidence accumulated that high rates of protection significantly depress economic development, and that open trade regimes are more conducive to growth.² Moreover, virtually all recent development success stories have been based on strong industrial export growth and relatively low barriers to imports--Chile, Hong Kong, Malaysia, Mauritius, and Singapore--or continually falling barriers--the Republic of Korea and Taiwan (China). Industrial sectors in these economies not only have higher export growth but generate more employment as well; and trade reforms have usually been accompanied by increased flows of foreign investment.

An effective trade policy is central to the integration of Russia in the international economic system and the growth that will generate: It provides the links with the international markets, and, together with the exchange rate, it forms the transmission mechanism through which international trade affects domestic resource allocation, the efficient and competitive restructuring of industry and agriculture, access to new and diverse technologies, improved incentives to exporters, and reduction of smuggling, rent-seeking and corruption in customs.

Tariff policy is the centerpiece of trade policy in a market system. Tariffs are, with very few exceptions, the only acceptable policy tool for protection under the GATT/WTO; and they are superior to alternative instruments of protection, such as non-tariff barriers (NTBs) like quotas, licenses and technical barriers to trade (TBTs),³ because they are less likely to lead to rent seeking and corrupt practices, and because tariffs limit the exercise of domestic monopoly power where it exists whereas NTBs do not.⁴

Russia's tariff regime has evolved over time without any explicit objectives or a strategic vision. In the aftermath of the breakup of the Soviet Union and central planning, consistent with the government's general desire to promote greater contacts with the international market, tariff barriers were not set high (Konovalov, 1994; Glaziev, 1994; Gros, 1994). Domestic industry and agriculture did not face significant competition from abroad however, because of the large undervaluation of the ruble. As the ruble appreciated in real terms, the pressure of international competition on domestic producers increased, as did their pressure on the govenrment to raise tariffs and protection in general. The result of these conflicting pressures is today's Russian tariff which averages 13%-14%,⁵ but which is rather diverse with maximum ad valorem tariffs of 30%, but with some specific tariffs that have ad valorem tariff equivalents higher than 30%.

This note examines the arguments for and against a uniform tariff structure in the Russian context. Arguments against uniformity are: terms of trade; "strategic," infant or restructuring industry considerations, revenue or balance of payments purposes, and tariffs as a negotiating tool at the WTO. Arguments in favor of uniformity are: political economy considerations; administrative convenience; and reduction of smuggling and corruption in customs. We maintain that tariff uniformity is clearly the best choice and develop a set of strategic guidelines which could help Russian policy makers in the design of future tariff policy.

There are several arguments in favor of government intervention through tariffs. Some of these arguments support tariffs as first best policies and if these arguments are accepted would call for non-uniform tariffs. These include tariffs: (a) to exploit a monopsony position and thereby improve the terms-of-trade; and (b) to maximize benefits from a "strategic" application of protection. Often governments wish to pursue other objectives than the pursuit of real income ⁶ or there may be constraints on the use of the first best instruments to achieve those objectives. Other objectives that call for the use of tariffs include: (c) as instruments for temporary protection of a specific "infant" or restructuring industry; (d) to raise budgetary revenue; (e) to reduce imports because of balance of payments problems; and (f) as bargaining tools to extract concessions from trading partners.⁷ First best policies in the pursuit of objectives c, d and e include, respectively, subsidies, indirect taxes, and devaluation and other macroeconomic policies. When the first best policies are not available, we argue below that low and uniform tariffs are preferred to a high and varied tariff structure.

One generally accepted theoretical reason for a country to impose tariffs on individual products is in order to exploit its monopsony power and thereby improve its terms-of-trade. If a country is large enough that it imports a significant share of the world’ s supply of a particular product, a tariff on that product could lower the price it must pay to world suppliers, which would improve its terms-of-trade. Consistent with this argument, the government could impose tariffs at different levels on different products to exploit the monopsony power it posseses; and the "optimal" tariff on each product would be different.

While this theoretical argument is valid, in practice, one can think of very few products where Russia possesses sufficient monopsony power for this to be a relevant consideration. Even in cases where countries have attempted to impose tariffs based on this rationale, the tariffs have been typically quite small (1-10%) because the share of world imports must be large for them to be large. Then, the actual tariffs in Russia are typically larger than the values optimal tariffs could reasonably be expected to take. For all practical purposes, tariff policy in Russia can be established without reference to this basically theoretical issue.

In recent years, a number of arguments have been developed justifying tariffs based on strategic considerations in industries with excess profits that are highly concentrated on a global basis. Among others, Brander and Spencer (1985) developed models that showed that in the presence on an international oligopoly, tariffs could increase a country’s welfare by enabling excess profits to be shifted from foreign to domestic firms. Another argument, for example by Krugman (1992), showed that given the existence of increasing returns to scale in the firm, protection to allow domestic firms to gain initial competitive advantage at the expense of foreign firms could be reinforced by internal economies of scale and would allow domestic firms to appropriate excess profits.

Despite its popularity among theorists in the 1980s, today strategic trade theory is not regarded as a significant policy choice. First, there is doubt whether excess profits really exist (except in the very short term) in many industries worldwide, and are not easily dissipated by new entrants or utilization of excess capacity. It has been shown by Eaton and Grossman (1986) that the policy conclusions of strategic trade policy models are completely reversed based on assumptions about which little is known. For example, whether firms compete in prices or quantities will reverse the optimal policy conclusion from subsidizing an industry to taxing the industry. Finally, one of the principal authors of this literature has concluded that the risks of following strategic trade theory far outweight the possible gains (Krugman, 1989, 1992), since a country might make small gains in some circumstances, but it is more likely to be misapplied and lead to large losses.

On balance, it would appear that this argument for differentiated tariffs would be even less applicable to a country like Russia, which seldom has any manufacturing firms competing oligopolistically in international markets with excess profits.⁸

The use of tariffs to effect welfare improvements over the longer run has been defended most often on infant industry grounds. The infant industry argument posits that certain industries are initially uneconomic but may become competitive (at world prices) in the long run because costs may decrease over time by virtue of learning-by-doing effects. Market failures, due to gains that are external to the firm, may prevent the development of such industries which exhibit positive discounted present values. For example, a firm may be unwilling to invest in technical know-how which may become freely available to other firms, i.e., the activities of an individual firm could generate externalities not capturable by the firm. (If there were no externalities the firm would be willing to make the investments and there would not be any need to depart from laissez-faire policy.)

Similar arguments have been made in the context of economies in transition where it is argued that if some firms which are in the process of restructuring are given protection for a time, they will be able to increase their productivity and become viable in the longer term. A firm may be faced with imperfections in its markets for inputs which raise its costs--e.g. because of an inefficient banking sector prevents from getting credit.

It is argued that temporary tariffs may be necessary to protect these infant and restructuring industries so they can generate benefits for the economy as a whole. Under the infant industry argument for protection, the optimum tariff structure would not normally be uniform, because protection would be accorded only to specific industries affected by market failure or externalities, and protection would not be warranted for other industries.

As Baldwin (1969) has explained, however, a tariff will not typically address the market failure problem so it is not better than laissez-faire policy. Consider, for example, the case of the inability of the firm to appropriate the gains from investment in technical know-how.

A duty raises the domestic price of a product, and from the viewpoint of the domestic industry as a whole, makes some investments in knowledge more profitable. But the individual entrepreneur still faces the same externality problem as before, namely, the risk that other firms in the same industry will copy, without cost to themselves, any new technology discovered by the firm and will then drive the product’s price or factor prices to levels at which the initial firm will be unable to recover the costs of acquiring knowledge [Baldwin (1969, p. 298)].

Thus, a tariff does not correct the problem. Indeed, it has been shown more generally that the best intervention is a policy that attacks the problem at the source (Bhagwati and Srinivasan, 1969). In this case, appropriate interventions directed at the source of the distortion, which could be imperfect appropriability, labor turnover, or capital market imperfections, are not tariffs, but rather measures such as the provision of information , patent protection or more effective use of instruments to allow collatoral.⁹

Sometimes a government may argue that the whole manufacturing sector is an infant. Although protection is unlikely the appropriate response, if any protection is offered for this purpose, a uniform tariff would be called for, not a diverse structure.

At the practical level, the arguments for a diverse tariff structure rest on the ability of governments to: (i) "pick the winners," that is to identify the candidates that are most likely meet the conditions justifying intervention, and choose and maintain the appropriate level for the policy variable (tariff, subsidy); (ii) be immune to the pressures from vested groups that inevitably arise once the willingness to grant special status is established; and (iii) prevent any protection granted from becoming permanent. The empirical evidence in both developed and developing countries during the past three decades casts doubt on most governments’ ability to meet these conditions. Endorsement of a more general approach--with little differentiation in the level of assistance--thus emanates from a wider skepticism about the practical merits of targeting of any kind, see Westphal (1990) and Krugman (1989, 1992).

Trade taxes are not optimal instruments to achieve a revenue objective because they significantly distort production and consumption choices. Peferred instruments to raise revenue are taxes such as income taxes or commodity taxes (excise, VAT, etc.). These are preferred taxes because, since they are applied neutrally to domestically produced and imported goods, they impose less distortion or inefficiency costs. The use of tariffs to raise revenue presupposes that other trade-neutral tax instruments are not available or cannot be used beyond existing levels; in other words, domestic taxes have to be taken as given either because the tax base cannot be enlarged rapidly enough or the marginal costs of increased domestic tax collection are very high (Corden, 1974; Balassa, 1989; and Mitra, 1992).

It could be argued that given the state of public finances and tax collection in Russia today, a case can be made for using tariffs as a source public revenue. In this case then, should the tariff be diverse or uniform?

One of the best known arguments for a non-uniform structure is the inverse elasticity rule. If the economy is characterized by only final goods (and ignoring rent-seeking, administrative and smuggling costs), the most efficient way to generate the tax revenue is to impose higher tariffs on the goods with the lower elasticity of demand (Ramsey, 1927). This causes the least distortion since it diverts the least resources. The simplicity of this rule has appeal to theoretical economists. The rule becomes exceedingly complex, however, in practice due substitution effects between goods, the presence of intermediate goods and goods that cannot effectively be taxed. Application then requires not just the own elasticities, but a complete set of cross-elasticities of demand, including the substitution elasticities with untaxed goods such as household leisure and underground economy goods. In practice this information is never available for a system such as the Russian tariff structure, so information requirements make the application of Ramsey type rules impractical,¹⁰ and we know of no country that has actually tried to implement them.

A still more important reason to avoid Ramsey type rules for diverse tariffs, is that political lobbying will develop that will lead to the application of tariffs that will differ from uniformity in economically inefficient ways. We return to the lobbying argument below.

Tariffs are sometimes employed to deal with a balance of payments problem. Again, they are not the best instrument. A balance of payments problem is a macroeconomic problem. Then, as discussed above, the optimal response is to attack the problem directly through macroeconomic tools, i.e., a combination of actions to reduce domestic spending (expenditure reduction) and policies that encourage exports and discourage imports (expenditure switching). Expenditure reduction can be accomplished through fiscal or monetary tightening, which reduces domestic absorption for any given level of output. Expenditure switching, which is best accomplished through a depreciation of the real exchange rate, raises the domestic price of tradables relative to non-tradables thereby encouraging exports and discouraging imports. Across-the-board import surcharges are often applied for balance of payments reasons; this achieves the same impact on reducing imports as exchange rate depreciation, but it fails to achieve the beneficial effects on the export side. The optimal tariff structure, given that it is a surrogate for a devaluation (without export incentives), must be uniform inducing resources to flow into import competing industries in general rather than any particular import competing industry.

Finally, countries may use tariffs as a bargaining tool to extract concessions from trading partners in the context of future multilateral trade negototiations within the context of the WTO. For countries applying to accede to the WTO, the future level and structure of tariffs is an important element in negotiations for accession to the WTO. In this context it should be noted that the accession negotiations focus not on the actual level of tariffs, i.e. the "applied rate," but on the so called "bound" rate which is the maximum legal level of the tariff for each individual tariff line which a country could not excede without either a renegotiation or giving compensation by reducing the tariff level for other products.

Within the rules and disciplines of the WTO, each country has considerable scope as to how restrictive or liberal its trade regime will be. There are no specific rules as to the maximum level at which a country has to bind its tariffs at accession or its degree of uniformity. Indeed some developing countries which are already WTO members have not "bound" all their tariff lines--meaning they can legally increase these rates without negotiation or compensation. Countries thus have a strategic choice to make during the accession negotiations phase: How liberal its tariff regime will be, consistent with the overall WTO disciplines.

Some governments have chosen to try to avoid legally binding commitments to a liberal trade regime in the context of accession. Accordingly, their initial tariff "offer" includes many tariff lines which are unbound--implying that they reserve the right to set legal maximum rates in the future, as well as a large number of commodities for which the proposed bound rate is significantly higher than the presently applied rates.

There are two sets of arguments that have been advanced in favor of this strategy: The first is of an economic nature and has to do with the future tariff policy most conducive to development and is similar to the infant industry and restructuring arguments discussed earlier. The other has to do with negotiating strategy and tactics to be pursued in the context of the accession process.

According to the first set of arguments, it would be unwise for a transition economy to commit itself legally to a low, and relatively uniform tariff structure at present, as its industry has not yet undergone significant restructuring. Therefore, it would be desirable to have bindings set at a high level or even for certain items not be bound at all so that the government could raise them subsequently as part of an industrial policy or restructuring that uses protection in order to promote future industrial development.

The other set of arguments starts from the premise that at accession, applicants can not negotiate significant improvements in their own market access: The tariff levels of WTO members are bound at fixed levels and the negotiations are solely over the level of the bindings of the acceding member. Hence, it would be desirable to try to maintain significant levels of legally bound tariffs, which it can use as bargaining chips to obtain improved access in future negotiating rounds. At a tactical level, it is also argued that any initial offer would have to be at a higher level than that which is finally agreed upon for two reasons: first, it may be possible for a country which is important in international trade to use the tariff level as a bargaining tool on other issues of market access (for example, in the case of transition economies, this could be over understandings over such issues as its designation as a "non market economy" in the context of antidumping).¹¹ Second, past practice in WTO accessions suggests that acceding countries are bound to face some demands to liberalize their tariff regime by existing members, almost irrespective of the level of protection they initially propose (Michalopoulos, 1998). Thus, some negotiating margin is necessary in the initial offers to ensure that countries are not forced to liberalize their trade beyond what they consider economically justified.

There are significant dangers to such a strategy, however. First, it is doubtful that in the long term efficient restructuring of transition economy industry can be done behind high tariff walls. Indeed, as the studies cited above on targetting document, industrialization behind high rates of protection may permit the survival of uncompetitive industries that hold little promise for increased productivity and efficiency that is critical to long term sustainable development. That is, this argument is not a new argument, but the restructuring or infant industry argument in the framework of the binding commitment of tariffs at the WTO.

Regarding the use of the initial tariff offer as a bargaining tool at the WTO, the initial tariff offer could be uniform without loss of bargaining power. That is, there is no reason to believe that an initial diverse tariff offer which is based on political economy or other considerations would present a superior bargaining position at the WTO than a uniform tariff offer. Thus, the bargaining power argument is not an argument against uniformity, but an argument about the level of the tariff. Regarding the level of the initial tariff offer at the WTO, it is clear that the leverage acceding countries have is quite small: While tactically, the maintenance of some margin of flexibility in the original offer is probably wise, the offer has to be used judiciously as a bargaining tool. Past practice suggests that: (a) countries have been asked to bind all their tariffs, even though existing WTO developing country members have not done so; (b) accession negotiations have gone much faster when countries have proposed lower and relatively uniform tariff rates with bound rates close to applied ones; indeed in some cases where countries have made initial tariff offers with a large divergence between applied and proposed bound rates, WTO members have refused to negotiate at all, until a different offer was put on the table which contained smaller divergences--a situation which led to considerable delays in the accession process (Michalopoulos, 1998).

But even if a country were to succeed in negotiating a structure of bound rates significantly divergent from today's applied rates, it may have gained a pyrrhic victory: By negotiating such a structure it would create an opening for domestic interests to exert political pressure for additional protection in the future. The government would lose the "political cover" the legally binding WTO commitments offer against domestic protectionist interests which may otherwise succeed in subverting the current trade regime and making it far more protective, to the detriment of long term efficient industrialization.

Binding at closer to the current applied level does not mean that the country would be powerless to deal with legitimate problems that increased imports may create for domestic industry. Increased protection to "safeguard" against serious injury to domestic industry is permitted under WTO rules--but it is supposed to be decided on the basis of a detailed and transparent investigation to demonstrate injury which is then notified to the WTO and subjected to the scrutiny of other members. This is far more difficult to do than for a powerful domestic industry to simply seek government support to raise tariffs beyond the applied level but below the higher bound level--which a government can do without any constraint. The point about WTO is not that it prohibits protection, but rather that it permits it only according to certain rules; and that obeying these rules makes protection more transparent as well as more difficult to initiate and spread.

In sum, for a transition economy acceding to the WTO, binding tariffs at low uniform levels is likely to be helpful to the longer term development of an internationally competitive and efficient industrial structure as well as facilitate the accession process.

Most policy-oriented analyses (Harberger (1988), Balassa (1989), Thomas and Nash (1991) and Mitra (1992)) start with the assumption that if any protection is to be granted at all, it should not favor any specific industry or set of industries; in other words, uniform effective protection should be provided to all industries. The arguments supporting a uniform tariff are based on political economy considerations, lobbying, administrative and smuggling costs, and adverse experience with picking winners.

Political pressures from constituents frequently induce governments to depart from trade policies that are prefered on grounds of economic welfare. Diverse and inefficient tariffs typically arise due to a "free-rider" problem in political lobbying. Political interests who want tariffs are typically the companies or unions in the industries because the gains are concentrated in relatively few hands and they are able to capture a sufficient amount of the gains that they will devote resources to lobbying for the tariff. On the other hand, those who lose from a tariff are the consumers of the product; although there a great many more of these people, their costs are smaller and not sufficient to induce them to spend resources to lobby their government to avoid the tariff. They would prefer someone else do it for them and to "free-ride" on the efforts of similarly minded individuals. The result is that typically only the industry that gains from the tariff lobbies the government and governments sometimes yield to this one-sided pressure.

The advantage of a uniform tariff is that it makes the gains to the industry much smaller, creating a kind of free-rider problem for the lobbying industry and dramatically reduces the incentive to lobby for protection. If a country employs a uniform tariff, an industry would not receive concentrated gains from its lobbying, since if it succeeded in raising the uniform tariff, it would have to bear the costs of raising the tariff for all the other products. These costs would include the higher cost of imported intermediate inputs and the lower price of its exports from induced changes in the real exchange rate. Any gains from raising the tariff would be dispersed, and would have to be weighed against the dispersed costs that the higher tariff in other industries would impose.¹²

The fact that a uniform tariff provides reduced gains to lobbying for protection conveys several advantages: (1) most important is that the level of protection is likely to be lower for reasons discussed above; (2) lobbying for protection is unproductive activity and a waste of resources. There is a direct saving of resources from the reduced lobbying; (3) the reduction to the gains from lobbying for protection provides a vastly improved signal to entrepreneurs. Entrepreneurs need to believe that they have more to gain by creating better and cheaper products or production processes than they do by lobbying their government. Entrepreneurial talent is quite scarce and valuable and if diverted into rent-seeking the growth rate of the economy can be adversely effected; and (4) the reduction in resources devoted to lobbying will result in less corruption in government, which may have positive spillover effects into other dimensions of government activity.

Uniform tariffs convey a number of administrative advantages. First, if tariffs are uniform, there is no incentive to misclassify goods. This enables customs authorities to concentrate on assuring that the value of the imported goods is not understated, and will reduce corruption related to customs clearing. In addition, the transparency and administrative simplicity of uniformity in customs clearance procedures will lower the administrative costs of trading. For example, traders will not have to obtain information on the category under which their products will fall.

A diverse tariff structure will provide an incentive to smuggle products which are subject to a high tariff. If the tariff is uniform, the strong incentives for smuggling that are presented by the high "outlyers" of a diverse tariff structure are considerably reduced.

In practice, empirical evaluations using computable general equilibrium models of actual economies have found that movements toward uniformity of the tariff structure increase real income. In the case of Turkey, a study by Harrison, Rutherford and Tarr (1993) found that uniformity in the incentives to importers and exporters would provide more than two-thirds of the gains to the economy of going to full free trade. That is, quantitatively the most important distortion in the trade regime of Turkey was not the overall height of the tariff, but rather non-uniform import tariff and export subsidy rates.

The reason for these results is that the distortion costs to the economy increase more than proportionately with the height of the tariff, i.e., a uniform tariff of 10% is much less costly to the economy than a structure that has a 20% tariff on half the products and zero tariffs on the rest.¹³ Turkey had succeeded in lowering its overall average tariff to under 15%, as in Russia; but the bulk of the distortion costs came from the relatively few sectors where tariffs or export subsidies were quite high.

Another example is the case of Chile which has a uniform tariff of 11%, and has elected to participate in a free trade area with MERCOSUR (Brazil, Argentina, Paraguay and Uruguay). Chile refused to join the MERCOSUR customs union in part because the customs union employs a common external tariff that is not uniform. Harrison, Rutherford and Tarr (1997)¹⁴ have estimated that the free trade area with MERCOSUR is substantially better for Chile than the customs union because the non-uniformity of the MERCOSUR common external tariff would impose considerable distortion costs on Chile. Similar results, although not as strong, were found in the cases of the Philippines (Clarette, 1989) and India (Mitra, 1994).

It would of course be useful to provide an estimate of how Russia compares to international experience in tariff uniformity. Unfortunately data do not exist that would allow precise international comparisons of either tariff uniformity or escalation. Although the MFN ad valorem tariff rates are available, estimates of the ad valorem equivalence of either specific tariffs or non-tariff barriers are not. The latter data are not available because tariff equivalence estimation is very time and data intensive, and involves thousands of tariff lines in many countries.¹⁵

We have, however, performed calculations based on the ad valorem rate--estimates which should provide a lower bound biased estimate of actual tariff dispersion when the tariff equivalence of the NTBs is taken into account. Despite the arguments in favor of relatively uniform protection, the actual experience with tariffs worldwide suggests that most countries differentiate their tariffs substantially (we discuss specific country examples below).¹⁶ Typically the protection pattern involves low tariffs for unprocessed commodities and raw materials as well as capital goods, and much higher tariffs for processed final goods. The basic reason for this is the influence of vested interests in maintaining protection on the final goods produced in the country, but who also lobby for tariff free access to their inputs. When there is no domestic intermediate goods industry, or the intermediate industry is small, there is no effective opposing lobbying influence for tariffs on these intermediates; the result is low tariffs on intermediates and high tariffs on selected final goods--a situation known as tariff escalation.

The tariff escalation that characterizes many countries' trade regimes, both developed and developing, causes especially serious problems in inefficient resource allocation: This "escalating" tariff structure tends to favor final goods production at the expense of intermediates, and in the long run encourages assembling type activities. That is, intermediate goods production is discouraged because it is disfavored relative to final goods and assembling activities. Thus, because an intermediate goods industry doesn’t exist today to lobby for equal protection, incentives are established which hinder its eventual creation.

Since tariffs on inputs affect the incentives for resources to move into or out of a sector, nominal protection, or the tariff rate that applies for a particular good, will not well reflect the extent to which the sector is favored or disfavored relative to another sector. Thus, the concept of "effective protection" is frequently employed, where effective protection measures tariffs on the sector’s output relative to tariffs on its inputs and its value added. Although effective protection is not a perfect measure of the impact of the tariff regime on resource allocation, the presumption is that activities with high effective protection are better able to attract resources and therefore to grow at the expense of others. In Box 1, we provide a discussion of the concept of the effective rate of protection. In general, the effective rate of protection increases as the tariff on output increases, the tariff on inputs decreases, and the value added at world prices of the sector decreases.

Although it is well known that Chile has a uniform tariff, there are quite a few countries with tariff structures that are uniform or at least close to uniform. Two countries have uniform tariffs due to the fact they practice free trade: Estonia and Hong Kong. Another three, Bolivia, the Kyrghyz Republic and Chile have virtually uniform tariff schedules of 10, 10 and 11 percent, respectively. Singapore has a simple tariff average of 0.5% and a standard deviation of less than 3%. Azerbaijan has a 15 percent maximum tariff and Bosnia-Herzogovia is reported to be about to move towards a uniform tariff. A number of other countries, including Brunei Darussalam, Equador, Honduras and Mexico, have tariff averages (under 13%) with small variances (under 6 percent).

At the other end of the spectrum are countries such as Bangladesh, and India with tariff averages of 84 and 56 percent and tariff variances of 26% and 24%. Korea, Mexico, South Africa and Turkey have more than 10,000 tariff headings, while for most other countries average about 6,000 tariff heading. A large number of countries have granted exceptional levels of protection for a limited number of products. The list of these countries includes some of the poorest countries but also some of the most prominent OECD countries, i.e., Cameroon, Canada, China, Egypt, European Union, Hungary, India, Indonesia, Israel, Nepal, Nicaragua, Norway, Saudi Arabia, Solomon Islands, Turkey, United States.

The import weighted average tariff in Russia is about 13.6% with a standard deviation of the tariff at the 3 digit level of about 7%. Due to exemptions the collected rate of import tariff is closer to 10%.

A very important element of the Russian tariff structure are the specific tariffs. For many tariff lines a mixed system applies, in which the maximum of the ad valorem or the specific tariff is applied by customs. Because of their lack transparency, specific tariffs should be avoided.

To provide a preliminary indication of the extent of tariff escalation in Russia, we classify in Figures 1 and 2 goods into three overall activities: unprocessed (stage 1 and lightly shaded in the figures), semi-processed (stage 2 and heavily shaded in the figures) and final goods (stage 3 and unshaded in the figures). The Russian tariff structure was aggregated and a concordance was established between the three digit harmonized system code and the ISIC code for production. Averages tariffs for each of the 3 digit categories was calculated and the results by stage of processing are shown in Figures 1 and 2 for important sectors. The aggregation procedure will tend to reduce dispersion of the tariff structure since the high and low tariffs in a sector will be averaged. Thus, these numbers should be considered a lower bound estimate of the extent of tariff escalation.

Sectors in Figure 1 show unambiguous tariff escalation, where sectors in Figure 2 are less clear. We find that the Russian tariff exhibits tariff escalation in the following sectors: leather products, wood products, petroleum and coal products, iron and steel and fabricated metal products; in addition, apparel products are much more heavily protected than raw materials and textile products. On the other hand, no clear pattern of tariff escalation is present in "non-classified food products, industrial chemicals, other chemicals, petroleum refineries, rubber products, other non-metallic mineral products, non-ferrous minerals, other manufactured products.

Based on our above aggregated view of the Russian tariff structure, we conclude that it does not exhibit extensive tariff escalation. Nonetheless, it indicates that leather products, petroleum and coal products and selected others are sectors where escalation likely results in very significant rates of effective protection.

The above analysis strongly suggests that there is little economic justification and many dangers in providing differentiated tariff protection to various sectors of industry and agriculture in Russia today. Most of the arguments in which tariffs differentiated by product are the optimal policy intervention are of little practical significance: There are few products in which Russia has monopsony power or industries in which it can gain advantage through "strategic" application of tariffs.

In many circumstances where tariffs are second best policy instruments, such as to raise public revenue or to cope with a balance of payments problem, a uniform tariff rate is the most practical and efficient alternative. And in the case where Russia may be interested in using the tariff as a bargaining instrument, it is immaterial whether the tariff is uniform or differentiated--the issues have to do with its capacity to use the tariff as a bargaining instrument and what it bargains for.

In the few cases where a differentiated tariff structure could be justified--usually also as a second best--in support of infant or restructuring industries, there are very serious political economy dangers for going that route: There are very few cases in which true "infant industry" learning by doing effects are present; tariffs are typically ineffective instruments at addressing the market failure problem; governments are not very good at picking them and there are serious dangers that the policy would be overwhelmed by requests for protection from vested interests irrespective of its economic merits.

Examining the present structure of Russia's applied tariffs in the light of worldwide experience and the analysis of the various arguments and justifications for imposition of tariffs, the following general strategic guidelines are suggested:

While the tariff schedule does not show as much escalation as in many other countries, there are several sectors with tariff peaks and substantial effective rates of protection, which are likely to attract resources. There is little evidence that these are sectors in which "learning by doing" would result in significant productivity improvements over time. Priority should be given to reducing these high rates and moving the tariff schedule towards greater uniformity. The costs in the form of the inefficiencies in resource allocation rise more than proportionately with the height of the tariff.¹⁷ As a consequence the greatest gains will come from reductions in the maximum rates. In addition, very high tariffs may be prohibitive of imports, so that there will be revenue gains from reductions in the rates. Reductions of the high rates will also reduce smuggling, corruption and rent-seeking disproportionately.

Raising the low rates, those rates that are less than 10%, is more controversial, especially on intermediate and capital goods. If there is no duty drawback or its equivalent in place, increasing the tariff on intermediate goods imposes a tax on the exportable goods that use the intermediate--this is bad since there are too few resources devoted to exports due to the tariffs.¹⁸ Moreover, it may be argued that raising tariffs on imports of raw materials and intermediates penalizes "technology" imports that are critical for raising productivity. Against these potential costs, one must weigh the fact that permitting imports of intermediates and capital goods at lower rates than 10% penalizes the development of Russia's own significant intermediate and capital goods industries (relative to final goods sectors)--as well as represents a foregone opportunity to raise revenue. And with a diverse tariff structure, the political economy, corruption and smuggling problems persist.¹⁹

To reduce the added anti-export bias that raising tariffs on intermediate and capital goods imposes, Russia and many countries employ mechanisms that allow exporters duty free access to imported intermediates. This includes duty drawback procedures, temporary admission and export processing zones. Coupled with effective duty free access to imported intermediates for exporters, the welfare tradeoff from raising tariffs on intermediate and capital goods is much more likely to be positive.

In these circumstances, the elimination of exemptions provides the benefit of additional revenue to the government, reducing or eliminating the revenue loss which may accompany a reduction in tariffs. And the elimination of exemptions provides a uniform incentive environment which does not favor any one class of producer or consumer over another.

The principal problem with provision of duty free access to exporters of imported intermediates is that the administration of these schemes can be very costly, and lead to cumbersome procedures and delays when tariffs are high. Exporters complain of delays and lack of payment. When tariffs are high there is also the risk of fraudulent claims. Moreover, these schemes do not remove all the anti-export bias of tariffs.²⁰ In addition, duty drawback schemes could reduce the incentive for real import liberalization which is the first best policy choice, if the schemes are perceived as an alternative to long term reform.

In general, raising tariffs on intermediates presents a conflict between the need to provide balanced protection to intermediates and final goods (i.e., reduce effective protection on final goods) and the need to reduce the anti-export bias of exports. Duty drawback and similar schemes appear to resolve the conflict but they do so at the expense of administrative complexity. The empirical evidence suggests these schemes become ineffective and very difficult to administer at high tariff rates (in excess of 15 or 20 percent) because of leakages, delays in payment and fraudulent claims (see Mitra, 1992). The experience of Russia is consistent with the international experience in this regard and emphasizes again the importance of reducing the overall level and diversity of the tariff.

It is sometimes argued that for administrative convenience the tariff structure should be simplified into 3 to 5 tariff bands. For example, with five tariff bands, tariffs could be either 0, 10, 20, 30 or 40 percent (or 0, 5, 10, 15 or 20 percent), but values in between would be prohibited. It should be clear that tariff simplification is not tariff uniformity, and simplification will allow very high rates of effective protection. Importantly, such a system suffers from virtually all the problems of a diverse structure, including encouraging lobbying for high protection by industry groups and it will encourage misclassification at customs relative to a uniform system.

If, on the other hand, tariff simplification is used as a vehicle to move toward low and uniform tariffs by limiting the number of tariffs, and reducing both the level and the dispersion of the tariff structure in the process, then tariff simplification is a very useful step. More generally, a tariff structure that is low and has a small standard deviation will convey many of the same benefits of a low uniform structure. For example, with a sufficiently small standard deviation there will be little gains from lobbying or incentive for corruption and in customs for misclassifying. Tariff simplification by itself, however, without reducting the level or the dispersion of the tariff structure, will convey relatively small benefits from reducing administrative costs.²¹

While many arguments have been made in favor of "low" tariff levels--as well as uniform ones, there has been little effort to define "low." The OECD have on average reduced their tariffs on manufactures to less than 5%--with a few peaks, especially in textiles and leather products. The main problem in most countries is agriculture where, as a result of the tariffication following the Uruguay Round, tariff schedules are quite high--reflecting the previously high supports and protection. In Russia's case a strategic decision may be appropriate to aim at a 10% uniform tariff. Anything higher, would result in significant increases in the effective rates of protection with the associated inefficiencies. A uniform rate at 10%, without exemptions, would also be helpful in generating substantial revenue. (Sequencing is discussed below.)

Russia has apparently chosen to make an initial tariff offer which is consistent with a bargaining strategy. This strategy contains a number of risks as outlined above. Based on the preliminary consultations on the offer, some WTO members have noted informally that the negotiations are likely to be long and difficult. The Russian authorities may wish to consider showing flexibility in these negotiations and a willingness to bind at tariff levels close to the applied ones. This would not only help with WTO accession, but most important, it is likely to contribute to the longer term development of a competitive and efficient industrial structure.

Protection should only provided by tariffs. Consistent with the declared intention of the Government of Russia, it should proceed to reduce TBTs by reducing the number of goods subject to mandatory certification and simplifying certifying procedures by the end of 1998; and it should harmonize Russian standards in priority sectors with internationally accepted ones by end 1999.

A good timetable for Russia would be to impose a 20% maximum ad valorem tariff in 1999, a maximum 15% in 2000, and 10% maximum in 2001. In calculating the actual applied tariff rates, it is crucial to include all surcharges, duties and discriminatory portion of excise taxes.²²

Consistent with the declared intention of the government of Russia, specific tariffs should be eliminated; they can be replaced by ad valorem tariffs or the mixed system currently employed by the government of Russia could be employed during a transition. For sectors with a mixed ad valorem-specific tariff system (where the maximum of the two tariffs applies), it would be advisable to cut the specific tariff by 50% in 1999, by another 50% in 2000 and eliminate specific tariffs by 2001.

Based on the principles discussed above, a good timetable for Russia would be to increase the minimum tariff to 3% in 1999 (Russia will apply this policy by virtue of the agreed tariff surcharge), 6% in 2000 and 10% in 2001. At the same time exemptions should be eliminated except for policies that provide duty free access to imported inputs for exporters.

The Russian authorities should be prepared to show flexibility and willingness to bind tariff levels closer to the applied ones. To do so, is likely to be helpful to the longer term development of an internationally competitive and efficient Russian industrial structure as well as facilitate the accession process.

Atkinson, A. B. and J. E. Stiglitz (1980), Lectures in Public Economics, New York: McGraw-Hill.

Balassa, B. (1989), Tariff Policy and Taxation in Developing Countries, PRE Working Paper Series No. 281, Washington: World Bank.

Baldwin, R.E. (1969), "The Case Against Infant-Industry Protection," Journal of Political Economy, Vol. 77, pp. 195-205.

Brander, J. and B. Spencer (1985), "Export Subsidies and International market Rivalry," Journal of International Economics, Vol. 18, pp. 83-100.

Centre for International Economics (1998), "Let’s Talk Uniform Tariffs," Paper prepared for the World Bank, Canberra and Sydney, mimeo.

Clarette, Ramon (1989), "The Economic Effects of Trade Liberalization on Philippine Agriculture," Philippine Review of Economics and Business, Vol. 26 (2), 208-235.

Corden, W. Max (1974), Trade Policy and Economic Welfare, Oxford: Clarendon Press.

Dahl, H., S. Devarajan and S. van Wijnbergen (1989), "Revenue-Neutral Tariff Reform: Theory and an Application to Cameroon," The World Bank, mimeo.

Dixit, A. (1986), "International Trade Under Oligopolistic Competition," Economic Journal.

Dollar, David (1992), "Outward Oriented Countries Do Grow More Rapidly," Economic Development and Cultural Change.

Eaton, J. and G. Grossman (1986), "Optimal Trade and Industrial Policy Under Oligopoly," Quarterly Journal of Economics, Vol. 101, pp. 383-406.

Edwards, Sebastian (1993), "Openness, Trade Liberalization and Growth in Developing Countries" Journal of Economic Literature, Vol. 31.

Glaziev, Sergei (1994), "On the Origins of Russian Trade Policy," in C. Michalopoulos and D. Tarr, eds. Trade in the New Independent States, Studies of Economies in Transformation No. 13, Washington D.C.: The World Bank.

Harberger, A.C., (1988), Reflections on Uniform Taxation, Paper Presented to the 44^th Congress of the International Institute of Public Finance, Istanbul.

Harrison, Glenn, Thomas Rutherford and David Tarr (1993), "Trade Reform in the Partially Liberalized Economy of Turkey," The World Bank Economic Review, May.

Harrison, Glenn, Thomas Rutherford and David Tarr (1997), "Trade Policy Options for Chile: A Quantitative Evaluation," Policy Research Paper No. 1783, The World Bank.

Kaminski, Bart, et al. (1996), Foreign Trade in the Transition, Studies in Economies of Transformation No. 20, Washington D.C.: The World Bank.

Konovalov, Vladimir (1994), "Russian Trade Policy," in C. Michalopoulos and D. Tarr, Trade in the New Independent States, Studies of Economies in Transformation No. 13, Washington D.C.: The World Bank.

Krueger, A.M. (1984), "Trade Policies in Developing Countries," in Handbook of International Economics, Edited by R.W. Jones and P.B. Jones and P.B. Kenen, North Holland.

Krugman, P.A. (1992), "Does the New Trade Theory Require a New Trade Policy," World Economy, Vol. 15, pp. 423-442.

———— (1989), "Is Free Trade Passe," Journal of Economic Perspetives, Vol. 1, pp. 131-144.

Martin, Will and L. Alan Winters, eds. (1996), The Uruguay Round and the Developing Countries, Cambridge: Cambridge University Press.

Michaely, M., D. Papageorgiou and A.M. Choksi (1991), Liberalizing Foreign Trade: Lessons of Experience in the Developing World, Vol. 7, Blackwell, Washington: World Bank.

Michalopoulos, Constantine (1998), "WTO Accession for Countries in Transition," Policy Research Working Paper No. 1934, The World Bank.

Michalopoulos, Constantine (1998a), "The Participation of Developing Countries in the WTO," Policy Research Working Paper No. 1906, The World Bank.

Michalopoulos, Constantine and David Tarr (1996), Trade Performance and Policy in the New Independent States, Directions in Development Series, Washington D.C.: The World Bank.

Michalopoulos, Constantine and David Tarr, eds. (1994), Trade in the New Independent States, Studies of Economies in Transformation No. 13, Washington D.C.: The World Bank.

Michalopoulos, Constantine and David Tarr (1992), Trade and Payments Arrangements for States of the Former USSR, Studies of Economies in Transformation No. 2, Washington D.C.: The World Bank.

Michalopoulos, Constantine and L. Alan Winters (1997), "Summary and Overview" in Ehrenhaft et al. Policies on Imports from Economies in Transition, World Bank 1997.

Mitra, P. (1992), "The Coordinated Reform of Tariffs and Indirect Taxes," World Bank Research Observer, Vol. 7, pp. 195-218.

Mitra, P. (1994), "Protective and Revenue Raising Trade Taxes: Theory and Application to India," The Economic Studies Quarterly, Vol. 45 (3), 265-287.

Morkre, Morris and David Tarr (1980), Effects of United States Restriction on Imports: Five Case Studies and Theory, Washington D.C.: US Government Printing Office.

Panagariya, Arvind (1992), "Import Tariffs, Duty Drawbacks and Tariff Reforms," Journal of International Economics, Vol. 32, pp. 131-147.

Panagariya, A and D. Rodrik (1993), "Political-Economy Arguments for a Uniform Tariff," International Economic Review.

Ramsey, F. P. (1927), "A Contribution to the Theory of Taxation," Economic Journal, Vol. 37, pp. 47-61.

Subramanian, Arvind, Ali Ibrahim and Luis A. Torres-Castro (1993), "Optimal Tariffs: Theory and Practice," International Monetary Fund Working Paper, Washington D.C., June, mimeo.

Tarr, David (1990), "A Modified Cournot Aggregation Condition for Obtaining Estimates of Cross-Elasticities of Demand, Eastern Economic Review, Vol. 16 (3), 257-264.

Thomas, V., J. Nash and associates (1991), Best Practices in Trade Policy Reform, Oxford University Press Washington: World Bank.

Westphal, L.E. (1990), "Industrial Policy in an Export-propelled Economy: Lessons from South Korea’s Experience," Journal of Economic Perspectives, Vol. 4, pp. 41-60.

World Bank (1991), World Development Report, 1991, Oxford: Oxford University Press.

World Bank (1996), World Development Report, 1996: From Plan to Market, Oxford: Oxford University Press.

World Bank (1993), The East Asia Miracle, published for the World Bank by Oxford University Press.

World Bank (1997), Global Economic Prospects, 1997, Washington D.C.: The World Bank.

1 I would like to thank Constantine Michalopoulos for his extensive contributions to this paper, Vladimir Drebentsov and Oksana Chernukha for the analysis of the Russian tariff structure in the paper and Harry Broadman for comments and direction.

2 Sachs and Warner in a recent article (1995) have estimated that open economies have grown about one percent per year faster than closed economies, and that the difference is greater among developing countries. For a review of the extensive literature on the link between open trade regimes and economic growth see Edwards (1993). See also Dollar (1992).

3 It is the intention of the Government of Russia to reduce TBTs by reducing the number of goods subject to mandatory certification, by simplifying certification procedures, and by harmonization of Russian standards in priority sectors with internationally accepted ones.

4 See, for example, Bhagwati (1965) and Takacs (1978).

5 This excludes the temporary 3% tariff surcharge.

6 The precise technical term employed by economists is economic welfare. Real income and economic welfare will be used interchangeably in this paper.

7 Income distribution objectives would call for higher tariffs on products with higher luxury content. Income taxes would be best for the purpose of income redistribution, but might not be available. Thus use of tariffs for income distribution purposes entails efficiency and growth losses. Commodity taxation, like a sales tax which does not discriminate by source, would also involve less efficiency loss that tariffs.

8 Russia does have significant world market share in platinum, nickel and diamonds, the latter of which is characterized by oligopoly internationally.

9 For a similar view see Krueger, 1984. If for purely political, rather than economic reasons, a government wishes to achieve a minimum output level in a given industry, the best intervention is a production subsidy, since it avoids the consumption distortion costs of the tariff. A production subsidy also has the advantage that it gears the infant industry to attain international competitiveness by avoiding discrimination between sales to the domestic and export market.

10 All the own and cross-elasticities of the Slutsky substitution matrix are required; given its symmetry, the number of elasticities equals n(n-1)/2 + n, where n equals the number of tariff lines (see Atkinson and Stiglitz, 1980). If there are 1000 tariff lines (and the Russian system contains more than this), the required own and cross-elasticities to calculate the optimal tariffs exceeds 500,000. This is an impossibly large number since even for much smaller systems, approximation rather than estimation techniques must be employed (see Tarr, 1991).

11 This designation, which arguably has been detrimental to its market access is not dependent on its accession WTO and, in principle, could continue after it has become a member. See Michalopoulos and Winters (1997).

12 In addition, since some imported inputs are used by import-competing sectors, a uniform tariff leads to a lower level of lobbying since it raises the costs of the import competing sectors. See Panagariya and Rodrik (1993) for details.

13 The distortion costs of the tariff increase with the square of the tariff rate. See Morkre and Tarr (1980, chapter 2) for a derivation.

14 This study showed that a movement away from uniformity for Chile would be costly.

15 We have performed estimates based on the ad valorem data.

16 Based on unpublished calculations using the "TRAINS" database by Ulrich Reineke at the World Bank.

17 See Morkre and Tarr (1980, chapter 2).

18 The "Lerner" symmetry theorem has shown that a tax on imports imposes a tax on exports. One of the principal reasons is that tariffs appreciate the real exchange rate, which makes exporting less attractive.

19 Tariffs on intermediates also convey the benefit of reducing the incentive to produce import competing products that use the intermediate--this is good because, due to tariffs, the economy allocates too many resources to these activities. A more complete treatment of this subject may be found in Panagariya (1992).

20 For example, to avoid anti-export bias, duty drawback schemes would have to be extended to indirect exporters as well (i.e., firms which do not export themselves but which sell to exporters). Administration of such mechanism are too complicated in practice and are not recommended for Russia.

21 In their excellent paper Subramanian et al. (1993) generally argue in favor of uniformity for most of the reasons in this paper. The authors also appear in practice to be willing to accept a system with several tariff bands. For consistency, we interpret their position as supporting several bands with a low standard deviation as part of a process of moving toward uniformity.

22 That is, if a 20% excise tax applies on domestic production of an item, but 25% applies on imported items, then the difference of 5% is part of the applied tariff.