Profiles of Tariffs in Global Agricultural Markets. Paul Gibson, John Wainio,Daniel Whitley, and Mary Bohman. Market and Trade Economics Division, Eco-nomic Research Service, U.S. Department of Agriculture. Agricultural EconomicReport No. 796
Abstract
High protection for agricultural commodities in the form of tariffs continues to bethe major factor restricting world trade. The large differences in average tariffsacross countries make it possible for farmers in one country to benefit from tariffprotection while farmers in other countries lose income because of lower pricesresulting from those tariffs. This report provides the first comprehensive analysisof agricultural tariffs and tariff-rate quotas (limits on imported goods) across alarge number of countries and commodities and finds that high average tariffs cre-ate barriers to markets for U.S. and other farmers.
Keywords: Market access, megatariffs, tariff profiles, over-quota tariffs, in-quotatariffs, tariff-rate quotas, World Trade Organization.
Acknowledgments
The authors wish to thank Adrie Custer, our technical editor, and Sue DeGeorgeand Victor B. Phillips, Jr., our graphic staff. The authors gratefully acknowledgethe reviews of Nicole Ballenger, Jean-Christophe Bureau, Mary Burfisher,Praveen Dixit, John Dunmore, Carol Goodloe, Debra Henke, Cathy McKinnell,Karl Meilke, and Sharon Sheffield.
1800 M Street, NWWashington, DC 20036-5831 January 2001
ii ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Contents
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .iii
Glossary of Trade Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .vi
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Why Tariffs Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Methodology for Developing Tariff Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . .5
Tariff Profiles by Region and Commodity . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
Tariff Rate Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16
Summing up Tariff Protection: OECD vs. non-OECD . . . . . . . . . . . . . . . . .18
Comparison of Bound and Applied Tariffs . . . . . . . . . . . . . . . . . . . . . . . . . . .20
Tariff Escalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
Agricultural Tariff Schedules of the United States, EU, and Japan . . . . . . . .24
Tariffs on Commodities of Export Interest to the United States . . . . . . . . . .30
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
Appendix: Technical Details of AVE Calculations . . . . . . . . . . . . . . . . . . . . .37
Appendix Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ iii
Summary
High protection for agricultural commodities in the form of tariffs continues to bethe major factor restricting world trade. The large differences in average tariffsacross countries make it possible for farmers in one country to benefit from tariffprotection while farmers in other countries lose income because of lower pricesresulting from those tariffs. This report provides the first comprehensive analysisof agricultural tariffs and tariff-rate quotas (limits on imported goods) across alarge number of countries and commodities and finds that high average tariffs cre-ate barriers to markets for U.S. and other farmers.
Tariffs impose costs both in the country where they are applied and on other coun-tries. Tariffs tax all products that cross a border, thus raising prices within thecountry imposing the tariff. Higher prices affect suppy because farmers respond byincreasing output, and higher prices affect demand because consumers buy less.The effects of tariffs on domestic markets can also spill over onto world markets asthe combined effect of more supply and less demand reduces imports. If the coun-try imposing the tariff is a large importer, then world prices can fall. Thus, the caseagainst tariffs has two components: the distortions created within a country byhigher domestic prices and the costs imposed on other countries by lost exportsales and lower world prices.
During the Uruguay Round negotiations, the United States and other World TradeOrganization (WTO) members began negotiations to reduce support and protectionin agriculture. These negotiations, which concluded in 1994, instituted tariffication,which is the process of converting agricultural nontariff barriers (NTBs), such asvariable import levies and import quotas, into bound tariffs (tariffs set at estab-lished rates). Tariffication resulted in a tariff-based system of border protection thatallowed for an initial set of tariff cuts. Countries were also to provide a minimumlevel of import opportunities for products previously protected by NTBs. This wasaccomplished by creating tariff-rate quotas (TRQs), which generally impose a rela-tively low tariff (in-quota) on imports up to a specified level, with imports abovethat level subject to a higher tariff (over-quota).
In 2000, WTO members agreed to submit detailed proposals on how they plan tofurther liberalize trade. These proposals include plans for negotiating the levels oftariffs and TRQs, and for negotiating policies for domestic support and export sub-sidies. Three questions need to be answered in order to understand how the alter-native proposals may affect agricultural markets:
� What is the pattern of agricultural tariffs across countries? Trade distortionsacross countries contribute to shifts in global resources, potentially at theexpense of countries with a comparative advantage in agriculture.
� How do tariffs vary across agricultural commodities? Large trade distortionsfrom high tariffs signal barriers to markets for competitive producers of spe-cific commodities.
� What does the structure of protection say about strategies in future trade negoti-ations? For example, high tariffs for most agricultural commodities suggest theneed to include all commodities in negotiations to provide the most benefits.
iv ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
This report answers these questions and provides the first comprehensive analysisof agricultural tariffs and TRQs across a large number of countries and commodi-ties. This information can help U.S. policymakers, producers, and consumersunderstand what is at stake in the ongoing WTO negotiations. Key findingsinclude:
—High average tariffs characterize agricultural markets. The global average tariffon agricultural products is 62 percent and is much higher than those on manufac-tured items. From a global perspective, high average tariffs cause demand to con-tract and supply to expand by drawing resources into agriculture, both leading tolower world prices.
—Average tariffs across 13 regions range from 25 to 113 percent, indicating thatfarmers in some countries are protected at the expense of farmers in other coun-tries. North America has the lowest regional tariff at 25 percent. Both developedand developing countries employ high tariffs, although within each group, thecountries in the non-EU Western Europe and South Asia regions tend to applymuch higher tariffs than their counterparts. Thus tariffs have the potential to trans-fer income from farmers in one country to those in another.
— Average commodity tariffs range from 50 to 91 percent, with the highest tariffsset for tobacco, meats, dairy, sugar, and sweeteners. Not only is protection high inthe dairy, sugar, and meat markets, but it is uniformly high across most countries.This structure of high tariffs likely causes a significant drop in world prices. Thus,multilateral liberalization could substantially increase world prices for these com-modities.
—The average tariff for the United States is 12 percent, among the lowest in theworld. With one of the lowest average tariffs, U.S. agriculture, as a whole, standsto gain from ambitious cuts in tariffs. Like many developed countries, however, theU.S. schedule contains some high tariffs aimed at protecting specific commodities.
—Agricultural tariffs in developing countries are considerably higher, on average,than in developed countries. This, in part, reflects the special and differential treat-
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ v
ment provided to these countries, such as lower tariff reduction commitments. But,available data suggest that many developing countries actually apply tariffs that areconsiderably below the rates they agreed to in the Uruguay Round Agreement onAgriculture.
— TRQs are associated with high tariffs and sensitive sectors, as might beexpected from their relationship with products previously protected by nontariffbarriers. The average over-quota tariff of 128 percent is double the average for allagricultural products. This results from the Uruguay Round tariffication process,which allowed the conversion of some NTBs into very high tariffs. A number ofcountries have bound their in-quota rates at extremely high levels, even though thetariffication process called for the in-quota tariff to be set at a “low or minimal”rate. The estimated average in-quota tariff of 63 percent is 1 percentage pointabove the global average for all other tariffs. While no numerical rule defined “lowor minimal,” these rates would seem to contradict the spirit of the agreement, indi-cating the need to negotiate some disciplines on these tariffs as well.
—The presence of megatariffs, defined as tariffs of 100 percent or higher, acrossall commodities and regions suggests the need to use a formula that reduceshigher tariffs at a greater rate. No imports are likely to enter under tariffs thishigh, other than the minimum market access granted under a TRQ. In cases wheremegatariffs are not associated with a TRQ, the only way to provide market accesswill be to significantly cut tariffs.
— The complexity of many countries’ tariff and TRQ schedules poses barriers tounderstanding the nature of protection. The lack of transparency associated withnon-ad valorem tariffs hides the actual level of protection being provided. This isparticularly true of compound tariffs or those based on complex technical factors.The result is difficulty in comparing protection across countries or commodities,which hinders the process of negotiating tariff reductions. One of the goals of thenext negotiations might be to increase certainty and transparency by formulatingstricter rules on the submission of tariff and TRQ schedules.
vi ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Glossary of Trade Terms
Agreement on Agriculture. Part of the UruguayRound agreement covering issues related to agricul-ture—e.g., market access, export subsidies, and inter-nal support.
Applied tariff rates. The actual tariff rate charged atthe border by an importing country, sometimes differ-ing from the bound rate. The rate is allowable underthe rules of the WTO if it is at or below the boundrate.
Articles (of the GATT). Clauses of the General Agree-ment that lay out the rules and procedures that Con-tracting Parties will observe in their conduct of inter-national trade and trade policy. Each of the 38 Articlesin the GATT deals with a different aspect of trade.
Bound tariff rates. Tariff rates resulting from GATTnegotiations or accessions that are incorporated as partof a country’s schedule of concessions. Bound ratesare enforceable under Article II of GATT. If a GATTcontracting party raises a tariff above the bound rate,the affected countries have the right to retaliate againstan equivalent value of the offending country’s exportsor receive compensation, usually in the form ofreduced tariffs of other products they export to theoffending country.
Ceiling binding. In cases where an existing tariff wasnot already bound, developing countries were allowedto establish ceiling bindings. These ceiling bindingscould result in tariffs that were higher than the existingapplied rate. The ceiling bindings took effect on thefirst day of implementation of the Agreement.
Country schedules. The official schedules of subsidycommitments and tariff bindings as agreed to underGATT for member countries.
EU (European Union). Established by the Treaty ofRome in 1957 and known previously as the EuropeanEconomic Community and the Common Market. Orig-inally composed of 6 European nations, it hasexpanded to 15. The EU attempts to unify and inte-grate member economies by establishing a customsunion and common economic policies, including CAP(Common Agricultural Policy). Member nationsinclude Austria, Belgium, Denmark, Germany, Greece,Finland, France, Ireland, Italy, Luxembourg, theNetherlands, Portugal, Spain, Sweden, and the United Kingdom.
GATT (General Agreement on Tariffs and Trade).Originally negotiated in Geneva, Switzerland in 1947among 23 countries, including the United States,GATT is an agreement to increase international tradeby reducing tariffs and other trade barriers. The agree-ment provides a code of conduct for international com-merce and a framework for periodic multilateral nego-tiations on trade liberalization and expansion.
In-quota tariff. The tariff applied on imports withinthe quota. The in-quota tariff is less than the over-quota tariff.
Market access. The extent to which a country permitsimports. A variety of tariff and non-tariff trade barrierscan be used to limit the entry of foreign products.
Megatariffs. Extremely high tariffs that effectively cutoff all imports other than the minimum access amountsgranted under the agreement. Some well-known exam-ples of megatariffs resulting from tariffication includethe base tariffs calculated for EU tariffs on grains,sugar and dairy products; U.S. sugar, peanuts anddairy products; Canadian tariffs on dairy products andpoultry; and Japanese tariffs on wheat, peanuts anddairy products.
Most-Favored-Nation (MFN) status. An agreementbetween countries to extend the same trading privi-leges to each other that they extend to any other coun-try. Under a most-favored-nation agreement, for exam-ple, a country will extend to another country the low-est tariff rates it applies to any third country. A countryis under no obligation to extend MFN treatment toanother country, unless they are both members of theWTO, or unless MFN is specified in an agreementbetween them.
NAFTA (North American Free Trade Agreement). Atrade agreement involving Canada, Mexico, and theU.S., implemented on January 1, 1994, with a 15-yeartransition period. The major agricultural provisions ofNAFTA include: 1) the elimination of non-tariff barri-ers—immediately upon implementation, generallythrough their conversion to tariff-rate quotas or ordi-nary quotas; 2) elimination of tariffs—many immedi-ately, most within 10 years, and some sensitive products gradually over 15 years; 3) special safeguardprovisions; and 4) country-of-origin rules to ensure
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ vii
that Mexico does not serve as a platform for exportsfrom third countries to the United States.
Non-tariff trade barriers. Regulations used by govern-ments to restrict imports from, and exports to, othercountries, including embargoes, import quotas, andtechnical barriers to trade.
OECD (Organization for Economic Cooperation andDevelopment). An organization founded in 1961 topromote economic growth, employment, a rising stan-dard of living, and financial stability; to assist the eco-nomic expansion of member and nonmember develop-ing countries; and to expand world trade. The membercountries are Australia, Austria, Belgium, Canada, theCzech Republic, Denmark, Finland, France, Germany,Greece, Hungary, Iceland, Ireland, Italy, Japan, Lux-embourg, Mexico, the Netherlands, New Zealand, Nor-way, Poland, Portugal, Spain, Sweden, Switzerland,Turkey, the United Kingdom, and the U.S.
Over-quota tariff. The tariff applied on imports inexcess of the quota volume. The over-quota tariff isgreater than the in-quota tariff.
Round. Refers to one of a series of multilateral tradenegotiations held under the auspices of the GATT forthe purposes of reducing tariffs or other trade barriers.There have been eight trade negotiating rounds sincethe adoption of the GATT in 1947.
Sanitary and phytosanitary (SPS) measures. Techni-cal barriers designed for the protection of humanhealth or the control of animal and plant pests and diseases.
Tariff. A tax imposed on commodity imports by a gov-ernment. A tariff may be a fixed charge per unit ofproduct imported (specific tariff), a fixed percentage ofvalue (ad valorem tariff), or some combination of both.
Tariff-rate quota. Quantitative limit (quota) onimported goods, above which a higher tariff rate isapplied. A lower tariff rate applies to any importsbelow the quota amount.
Tariffication. The process of converting non-tarifftrade barriers to bound tariffs. This is done under theUR agreement in order to improve the transparency ofexisting agricultural trade barriers and facilitate theirproposed reduction.
UR (Uruguay Round) agreement. The UruguayRound of multilateral trade negotiations, conductedunder the auspices of the GATT, is a trade agreementdesigned to open world markets. The Agreement onAgriculture is one of the 29 individual legal textsincluded in the Final Act under an umbrella agreementestablishing the WTO. The negotiation began at Puntadel Este, Uruguay, in September 1986 and concludedin Marrakesh, Morocco, in April 1994.
World Trade Organization (WTO). Established onJanuary 1, 1995, as a result of the Uruguay Round, theWTO replaces GATT as the legal and institutionalfoundation of the multilateral trading system of mem-ber countries. It provides the principal contractualobligations determining how governments frame andimplement domestic trade legislation and regulations.And it is the platform on which trade relations amongcountries evolve through collective debate, negotiation,and adjudication.
viii ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 1
Introduction
High protection for agricultural commodities contin-ues to be the major distorting feature of internationaltrade. The global average agricultural tariff,1 esti-mated at 62 percent, contrasts with the much lowertariffs for industrial products estimated by Hertel andMartin (see references). Not only are some agricul-tural tariffs extremely high, but they are also highlyuneven across countries and commodities. Clearly,substantial room exists for liberalization of agricul-tural tariffs.
Among the most important accomplishments of theUruguay Round Agreement on Agriculture (URAA)was the requirement to convert agricultural nontariffbarriers (NTBs), such as variable import levies and im-port quotas, into bound tariffs. Bound tariffs are set atrates established by the General Agreement on Tariffsand Trade (GATT) negotiations. The process, knownas tariffication, resulted in a tariff-based system of bor-der protection that allowed for an initial set of tariffcuts in the URAA. Developed countries agreed to re-duce all agricultural tariffs from their base period ratesby a total of 36 percent, on a simple average basis,with a minimum cut of 15 percent for each tariff.2
Starting in 1995, tariff cuts were to take place in equalinstallments over 6 years for developed countries and10 years for developing countries. Countries were alsoto provide a minimum level of import opportunities forproducts previously protected by NTBs. This was
accomplished by creating tariff-rate quotas (TRQs),which generally impose a relatively low tariff (in-quota) on imports up to a specified level, with importsabove that level subject to a higher tariff (over-quota).
The high tariffs currently existing in the agriculturalsector restrict trade in agricultural products and causeworld prices to fall. Research conducted by the U.S.Department of Agriculture’s Economic Research Ser-vice (ERS) has shown that tariffs and associated TRQsaccount for the largest share of global agricultural dis-tortions. Export subsidies and domestic farm programsare the other major distortions. When all three types ofdistorting policies are removed, world prices increaseby 12 percent. Tariffs account for 52 percent of theincrease in world prices (Burfisher et al). While reduc-ing tariffs is a necessary part of increasing marketaccess, other impediments to trade may also need to beaddressed. For example, factors such as sanitary andphytosanitary (SPS) measures or state trading enter-prises may also limit market access.
This report addresses three questions about tariffs thatare relevant for future negotiations on market access.
� What is the pattern of agricultural tariffs acrosscountries? Distortions across countries contribute toshifts in global resources, potentially at the expenseof countries with a comparative advantage in agri-culture. Figure 1 shows the landscape of global tar-iffs. Disparities in tariffs indicate that some coun-tries protect their agricultural sectors at the expenseof other countries. Leveling the playing field acrosscountries would help alleviate this problem.
� How do tariffs vary across agricultural commodi-ties? Global average tariffs range from 50 to 91 per-cent for the 46 commodity groups analyzed in thisreport. Large distortions from high tariffs signal bar-riers to markets for competitive producers of spe-cific commodities.
Profiles of Tariffs in Global Agricultural Markets
Paul Gibson, John Wainio, Daniel Whitley, and Mary Bohman
1 In this report, the term “tariff” refers to the import duties thatWTO members may levy on imports from other members (boundMFN tariffs based on final URAA implementation).
2 Developing countries were required to reduce their tariffs on aver-age by only 24 percent, with a minimum cut of 10 percent for eachtariff. However, in the case of previously unbound tariffs or whenconverting NTBs to tariffs, many developing countries chose theoption of offering tariff bindings with no reduction in tariff levels.Least developed countries were not required to reduce their tariffs,although they still had to replace their NTBs with tariffs and bindall tariffs.
2 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
� What does the structure of protection say aboutstrategies in future trade negotiations? In particular,in what countries do high tariffs exist for commodi-ties and food products of interest to the UnitedStates?
By answering the three questions above, this reportpaints a picture of the current pattern of market accessprotection for agriculture. It begins with an economicperspective of the ways that tariffs affect markets, fol-
lowed by the methodology behind the indicators of tar-iff impacts, and then compares different types of tar-iffs. The heart of the report identifies patterns in globaltariff and TRQ profiles across countries and commodi-ties. The report then digs deeper into the structure ofprotection for the three major agricultural players inglobal markets, the United States, the European Union(EU), and Japan. An overview of protection for com-modities of interest to the United States concludes the analysis.
Average tariffs of WTO membersFigure 1
Tariff average0 - 24%25 - 49%50 - 99%100 - 200%
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 3
Why Tariffs Matter
Tariffs impose a cost on all products that cross a bor-der, thus raising prices within the country that imposesthe tariff. Higher prices affect supplies as farmersrespond by increasing output and affect demand asconsumers buy less. Countries apply tariffs primarilyto protect domestic industries. This and other justifica-tions for tariffs are discussed in the box, Why Coun-tries Use Tariffs. The domestic market effects of tariffscan also spill over onto world markets as the combinedeffect of more supply and less demand reducesimports. If the country imposing the tariff is a largeimporter, then world prices can fall. Thus, the caseagainst tariffs has two components: the distortionscreated within the country via higher domestic pricesand the costs imposed on other countries via lostexport sales and lower world prices.
Table 1 shows how tariffs affect different parts of theagricultural economy for three categories of countries:the tariff-imposing country, exporting countries, andimporting countries. The effects are shown for the finalconsumption and the product as an input or intermedi-ate good. For simplicity, the analysis focuses on a sin-gle commodity, durum wheat, and two final products,pasta and bread.
The country imposing the tariff in table 1 realizes adrop in net economic benefits. This loss of overall eco-nomic benefits comes from two sources. First, a higherdomestic price draws resources into wheat farming,instead of other agricultural and nonagricultural uses,that might have created more value elsewhere. Forexample, capital and labor used to produce extra wheatmight be more productive elsewhere, such as produc-ing alternative crops or information technology. Sec-ond, higher wheat prices alter consumer choice andlower real income.
While tariffs increase prices in the imposing country,they can also lower world market prices. Producers inall other countries suffer from lower prices. Thus,while tariffs in one country may be seen as protectingits domestic wheat farmers, those same tariffs penalizewheat farmers in other countries. The high level of,and differences in, tariffs across countries shown infigure 1 indicate that current protection levels shiftwealth across national borders.
The higher consumer price extends to industries thatuse the product, such as manufacturers that use wheatin pasta production. For example, tariffs that increasethe price of wheat for domestic pasta manufacturerscould decrease the price for foreign pasta makers. Thehigher wheat price raises costs for domestic pastamakers and puts them at a disadvantage to foreigncompanies in both the domestic and foreign markets.
Table 1 focuses on one commodity, but the costs andbenefits from a single tariff can spill over to othercommodities as well. As stated earlier, tariffs can drawresources away from the production of other com-modities within a country. They can also alter produc-tion and consumption decisions in other countries.Lower international wheat prices could cause farmersin other countries to plant alternative crops such asbarley or rapeseed, leading to an increase in their sup-ply and a resulting fall in world price. By alteringprices, tariffs alter economic incentives, which can sig-nificantly affect how efficiently economies use theirresources. In general, one might expect that the morecomplex a country’s tariff schedule, the less likely thatany single component will have the intended effect.
To summarize, tariffs are a tool to protect domesticindustries. They transfer income from consumers toproducers and across the value-added chain. Tariffshave other unintended or spillover effects as well.
���������� ���������� ������������������������ ����� ������������
��������������������� �
��� ���������� �������� ������� ����� ����� ������� ����� �������������������������� ������������������� �������������������
���������� � � ��������� ��� �������� ���!����� � � ��� �������� ���!����� � � �"�#�� �� ����������#� ��� � $ �
%
&��� � ����� �'
� �%(�������#� ����� ������������� ����� �������������������������� ����������������#����������������������������#� ����� �����������������)'* ��������������� �������������������� ����������� �� ����������!� �+������������,�� �� ������ ��� � �������������������������������������������� ������������� �������+�������,���������#� ������������������� ��������� �����)��� ��������!��������� ���������� ������������� ��-�����������������������).�����/����� �����0������.��#���!�1.2*)
4 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
First, they decrease overall wealth by distorting pro-duction and consumption. These distortions can filterdown to prices of land and other inputs primarily usedin agriculture. Second, tariffs in one country hurt farm-ers in other countries and benefit foreign consumers.
Additional spillover effects across countries includechanges in other countries’ balance of paymentsthrough changes in export or tariff revenue and import costs.
Providing protection against competition from importsfor a specific commodity or sector is the most com-mon reason countries apply tariffs. Underlying thisreason, however, is the old mercantilist notion that acountry is better off if it exports more than it importsand that, therefore, protective tariffs will add to thenation’s prosperity. One of modern economics’ great-est contributions has been to point out the fallacy ofthe mercantilist argument by demonstrating the econo-mywide benefits from free trade.
But, the economic case against tariffs, which exposesthe distribution of costs and benefits to the economy,also helps to explain why those who benefit from tar-iffs continue to lobby ardently for protection. Thecosts, in the form of higher consumer prices, arespread out over a large number of consumers. How-ever, the benefits are concentrated on a relativelysmall group of producers of the product. Any changein tariffs simply means more to the average producerthan to the average consumer.
Several reasons are commonly used to justify apply-ing tariffs. In agriculture, concern about farm incomeas well as nonmarket benefits from agriculture (e.g.,benefits from agricultural landscapes) provide ration-ale for farm programs that often include tariffs as apolicy instrument. Tariff protection is often an inte-gral and essential element of a country’s domesticagricultural policy and can only be eliminated ifaccompanied by changes in domestic regimes. Inparticular, programs designed to raise domesticprices above world prices may be unsustainable inthe face of increased imports. While providing pro-tection to producers, tariffs also raise consumer
prices and create more distortions than direct supportfor producers. Therefore, economists find that poli-cies that directly target the policy objective, such asincome transfers to address low incomes, are moreeffective policy instruments than tariffs (Corden).
Some justifications for tariffs relate to current marketconditions. Temporary use of tariffs has been justi-fied in order to protect new or infant industries andto provide a window to become established in themarket. In practice, however, these tariffs prove diffi-cult to remove, as those that benefit come to rely onthe protection they provide. Under specific circum-stances, tariffs can be introduced or raised even whenthey are bound at zero or have low rates. For exam-ple, the WTO allows members to apply anti-dumping(AD), countervailing (CVD), or special safeguard(SSG) duties (and, in the case of safeguards, importquotas). CVDs are sometimes applied to offset subsi-dies by other countries, while ADs are applied whenforeign firms sell products below costs. SSGs can beimposed if a country experiences an increase in thevolume of imports or a drop in the price of importswhich exceed certain trigger levels. Tariffs appliedfor these three reasons represent an extremely small,but growing share of all tariffs, and WTO rules pro-vide guidelines for their application.
Governments in developing countries sometimesapply tariffs to achieve other objectives. The relativeease of taxing goods at international borders com-pared to levying income or sales taxes makes tariffsan attractive source of revenue. Managing the bal-ance of payments by restricting imports is anotherrationale developing countries use to apply tariffs.
Why Countries Use Tariffs
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 5
Methodology for DevelopingTariff Profiles
Countries levy tariffs in a number of different, andsometimes complex, forms. Most tariffs are expressedin “ad valorem” terms, or as a percentage of the valueof the imported good. However, a significant portion isexpressed in specific, or other non-ad valorem terms(see box, Tariff Formats Conceal High Levels of Pro-tection). Agriculture is somewhat unique in the extentto which non-ad valorem tariffs are still used. In theUnited States and the EU, for example, approximately44 percent of agricultural tariff-lines (categories ofproducts with tariffs) are specified in non-ad valoremterms. There are a number of reasons for this, includ-ing the increased protection that a non-ad valorem taxcan provide against large drops in import prices andthe lack of transparency associated with these rates,which helps conceal the level of protection being provided.
Tariffs are bound at the tariff-line level, which refers tothe category to which the legally established tariffapplies. The complexity of many schedules and thelack of transparency associated with this complexitymake it very difficult to compare tariffs across coun-tries or across commodity markets. The challenge inmaking the comparisons is to transform the data to acommon basis and then develop measures to summa-rize the thousands of tariff-lines that can make up aschedule. This section describes the conceptualapproach used to develop meaningful tariff profiles foreach country. The steps are presented following thesame process we used to transform the tariff-lines intostatistics that characterize each country’s tariff sched-ule. Appendix A provides technical details on these calculations.
Calculation of Tariff Ad Valorem Equivalents
The first step in developing tariff profiles is to calcu-late an ad valorem equivalent (AVE) for each non-advalorem tariff. Unfortunately, no single AVE exists fornon-ad valorem tariffs, as the calculated value dependson the choice of import price and exchange rate, bothof which can change over time. The import priceshould approximate the declared value against whichthe ad valorem tariff would have been charged.Domestic prices overstate this value because they havebeen inflated by the tariff, while the country-specificimport unit values reflect preferential import condi-tions and, thus, can be out of line with representativeworld prices and vary widely across countries (Lind-
Countries levy tariffs in a number of differentways:
� As a percentage of the value of imports (ad val-orem tariffs)
� As a monetary amount per unit of import suchas cents per liter (specific tariffs)
� As a combination of the two, such as 12.5 per-cent plus 2 cents per liter (compound tariffs)
Other factors can further complicate compound tariffs, including appending a threshold, such as, butnot less than 15 cents per liter or greater than 25cents per liter. In this case, either the ad valorem orthe specific portion of the tariff can be binding. Tar-iffs may also vary based on the time of year (sea-sonal tariffs) or be determined by complex technicalfactors (such as sugar or alcohol content).
One of the main rationales for specific duties istheir administrative simplicity, since they avoid theproblem of having to value imports. Defenders ofspecific tariffs have argued that ad valorem ratesgive an incentive to importers to underinvoice,since the size of the duty depends on the price ofthe import. Supporters of ad valorem rates havecountered that specific duties place a heavier bur-den on lower priced items within a given tariff-line and are therefore a regressive tax on con-sumers. In addition, they point to the lack of trans-parency associated with specific duties, since thead valorem equivalent is often difficult to deter-mine (Irwin).
Since calculating AVEs takes considerable timeand effort, and since the data needed to performsuch calculations are often not available, non-advalorem tariffs for agriculture are often excludedfrom calculations of average tariffs. This can resultin an average that is underestimated, since theAVE of these tariffs tends to be quite high. Basedon the AVEs calculated in this study, non-ad val-orem tariffs appear to provide significantly highertariff protection than ad valorem tariffs. The aver-age of bound tariffs specified solely in ad valoremterms is 58 percent, while the average AVE ofnon-ad valorem tariffs is 123 percent.
Tariff Formats Conceal High Levels of Protection
6 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
land). In many cases, no country-specific import valueeven exists, particularly when tariffs are so high as topreclude any trade from taking place. Representativeworld prices also present a problem since they may notconform to the specific quality or variety of the com-modity imported.
In this study, world import unit values were used as aproxy for import prices, since the global perspectivecorrects for individual country tariffs and representsthe average quality or variety of the product in ques-tion. This approach also allowed us to calculate AVEseven when the country imported none of the commod-ity in question. Unfortunately, world import unit val-ues present a drawback as well, since they are onlyavailable at the relatively aggregate levels. When coun-tries have many disaggregated tariff-lines, using theworld unit import value may underestimate the AVEfor some of these tariffs and overestimate it for others.
One characteristic of fixed, specific tariffs is that theyprovide a level of protection inversely related to prices.Therefore, in a time of low prices, the level of protec-tion provided by the tariff is higher than during aperiod of high prices. Likewise, when tariffs aredenominated in domestic currencies while prices are inU.S. dollars, a depreciation in the exchange rate willresult in a decrease in the AVE, even if the specific tar-iff and the dollar price have not changed. Thus, theAVE will vary based on the time period of the worldimport unit value used in the calculation. Prices usedto calculate AVEs in this report are based on averageworld import unit values for 1995-97. To the extentthat world import prices during this period reflectsomewhat higher world prices than prevailed at othertimes, the AVE tariffs presented here will be lowerthan AVEs calculated during a period of lower prices,such as current prices.
Country and Commodity Coverage
In order to identify patterns in protection, the next stepis to aggregate the tariffs for extremely narrowly definedproducts (a total of 91,000 tariffs across all countries)into broad country and commodity categories. The nextsection uses regional aggregations to provide a broadoverview of the differences in tariff protection. Countrycoverage of the data used includes 129 of 140 WTOmembers.3 Commodity or product groupings used in
this report cover a broad range of agricultural productstraded by both developed and developing countries. Thecommodity list used in this report covers most, but notall of the lines that fall under the WTO definition ofagriculture (see Appendix).4
The most common way to aggregate tariffs, used inthis study, is to calculate the simple, unweighted aver-age. However, drawbacks are associated with a simpleaverage. An unweighted average does not distinguishbetween “important” and “unimportant” tariffs. Sinceequal weight is given to all agricultural tariffs, akumquat tariff is as important as a wheat tariff, if eachenters as a single tariff-line item. The different levelsof commodity aggregation found in each country’s tariff schedule present another drawback. For instance,in the category “dairy,” there are 27 tariff-lines forAustralia, 75 for Canada, 183 for the United States,and 187 for the EU. If tariffs for these items are large (which they are), the higher the level of disaggre-gation, the greater the upward bias in the country average.
There are a number of alternative ways to average andaggregate tariffs across countries and commodities,none of which is without bias. Weighted averages areoften calculated in an attempt to emphasize certain tar-iffs over others. Weighting based on import values,perhaps the most commonly used weighting scheme,may bias the average downward, because items withthe highest tariffs will receive virtually no weightbecause little or no trade will take place under suchtariffs. Weighting based on shares of domestic value ofproduction would assure that highly protected com-modities produced in large amounts get appropriatelylarge weights, but this method can result in an upwardbias, because many factors other than tariffs affectagricultural production levels. In addition, productiondata at the tariff-line level are rarely available. Theshare of the domestic value of consumption is anotheralternative, but biased to the extent that high tariffsreduce consumption. Similar to production, consump-tion data are generally not available at the tariff-linelevel. One alternative is to calculate a simple(unweighted) average aggregated to a level where dataon appropriate production weights are available (the 4-or 6-digit HS level), as was done by the Organizationfor Economic Cooperation and Development (OECD)in a recent analysis (OECD, 1999). Ultimately, there is
4 A detailed specification of commodity groupings is availablefrom the authors ([email protected]).
3 As of November 30, 2000, WTO membership totaled 140 coun-tries or customs territories. Of this number, 16 are accounted for bythe European Union; one each for the EU Commission and the 15member states.
no ideal weighting scheme and the transparency ofunweighted aggregations has some advantages.
Statistics To Characterize Tariff Profiles
A critical component of this study is to determineappropriate statistical measures to characterize thelevel of tariff protection in each country or commoditysector. The two most commonly used measures are thearithmetic mean (or average) to capture the overalllevel of tariffs and the standard deviation to measurethe spread or distance of most observations from themean. While each is the most efficient measure fornormal or bell-shaped distributions, arithmetic meanand standard deviation are not the most appropriatemeasure for highly skewed distributions.5
Tariff schedules sometimes have distributions that arehighly skewed to the right, meaning that the tariffs
continue much farther to the right of the mean than tothe left. For these distributions, the mean may overesti-mate the central tendency of the data. The most com-mon alternative measure is the tariff median, whichmeasures the midpoint of the tariff schedule’s distribu-tion. If a country’s tariff schedule is normally distrib-uted, then the mean and median tariffs would be veryclose, and there would be no need to report more thanone. But, when the tariff schedule is highly skewed,both the mean and median give useful information,although the median tariff might be considered a more“representative” measure for comparing the overallheight of each country’s regime, since it is less sensi-tive to a few extremely high rates.6
This report uses means and medians as the two statis-tics to characterize tariff distributions. While the rela-tionship between the mean and median represents acontinuum, four benchmark combinations are identi-fied with associated economic interpretations.
High mean/high median: High levels of protection fora country or commodity sector found across most tar-iff-lines.
High mean/low median: Extremely high levels of pro-tection for a few specific commodities result in highmean, although most tariff-lines are low. This suggeststhe need for more detailed analysis that breaks outcountries and/or disaggregates commodities to under-stand nature of protection.
Low mean/high median: Extremely low levels of pro-tection for a few specific commodities result in lowmean, although most tariff-lines are high. This sug-gests the need for more detailed analysis that breaksout countries and/or disaggregates commodities tounderstand the nature of protection.
Low mean/low median: Low levels of protection for a country or commodity sector found across most tariff-lines.
Before applying these benchmarks to the data, defi-nitions of high and low are required. The dividinglines are the global mean agricultural tariff equal to 62 percent and the global median tariff equal to 40percent. In parts of the analysis, tariffs for a specific
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 7
5 When the word “mean” is used without a modifier, it refers to thearithmetic mean, or simple average.
6 This report does not provide a direct measure of the spread or dis-persion of the data such as the standard deviation. A comparison ofthe mean and median provides some information about the disper-sion and also indicates the influence of megatariffs.
A number of sources provide the bound andapplied data used in this report. The primarysource of bound tariffs is the Agricultural MarketAccess Database (AMAD). The AMAD is themost comprehensive collection of available publicdata on WTO market access, containing detaileddata on WTO tariff and TRQ schedules, importdata, applied tariffs, production, consumption,and trade, among other information. The AMADcontains data on about 40 WTO members, includ-ing all major agricultural trading members.AMAD data can be accessed through its website,www.amad.org. Tariff bindings in this report forcountries not included in the AMAD are from tar-iff bindings of the WTO Secretariat. These bind-ings are reproduced on the CD-ROM “Results ofthe Uruguay Round,” WTO Secretariat. Additionaldata on applied tariffs is from the United NationsConference on Trade and Development (UNC-TAD) Trade Analysis and Information System(TRAINS) database. UNCTAD TRAINS containsa comprehensive collection of applied tariff data.Applied tariff data for developing countries for theyears 1995-99 included in this analysis, as well asin the AMAD, are from the UNCTAD TRAINSdatabase.
Tariff Data
8 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
commodity and country are also compared withregional or commodity mean and median tariffs.
This study also identifies markets subject to extremelyhigh tariffs. This is important because these are mar-kets where tariffs could be significantly reduced with-out actually improving market access. No internation-ally accepted definition exists to categorize these“megatariffs.” In this report, tariffs equal to or above
100 percent qualify as megatariffs. Another term formegatariffs used in this study is “international tariffpeaks,” or those tariff-lines that exceed some commonyardstick. 7
The levels of tariff protection profiled in this reportrefer to Most Favored Nation (MFN) bound tariffs.Tariff rates on trade under regional or preferential tradeprovisions, such as North American Free Trade Agree-ment (NAFTA) or Generalized System of Preferences(GSP), are not considered.
Bound MFN tariffs are tariff commitments sched-uled by WTO members and are generally consideredthe maximum allowable tariffs that a member maylevy on imports. The establishment of bound tariffrates on agricultural trade among WTO memberswas a major accomplishment of the Uruguay Round.Under WTO rules, application of tariffs above boundrates generally requires that compensation be offeredto trading partners adversely affected by an increasein tariffs above bound levels. Bound MFN tariffs arethe rate against which regional tariff preferences orother import reductions are referenced. Bound rateshave typically been the rate used as the basis for tar-iff reductions in multilateral trade negotiations. Thetariff schedules of most WTO members reflect thetariff rates established by the Uruguay Round. Tariffschedules of members who joined the WTO since1995 were developed through accession negotiations.In general, these bindings reflect the rate effectivefor 2000 and beyond for developed countries and2004 and beyond for developing countries, althoughall ceiling bindings took effect in 1995.
Some tariffs take the form of tariff-rate quotas(TRQs). TRQs specify that a limited quantity of a
good may be imported at a low tariff, the “in-quota”tariff. Once the quota level has been reached, unlimitedimports of the same good may be imported at a higher“over-quota” tariff. Prior to the Uruguay Round, manyWTO members applied a wide range of nontariff barri-ers (NTBs) on imports of agricultural products. TheUruguay Round replaced NTBs with tariff-rate quotas;a process also known as tariffication. The tarifficationprocess provided for two types of TRQs: minimumaccess and current access. The minimum access levelis the quantity allowed to be imported at the lower tar-iff. It was set at 3 percent of consumption in 1986-88in the base period, to be increased to 5 percent of baseconsumption by 2000 (2004 for developing countries).Current access was to be provided for products subjectto tariffication with imports exceeding 5 percent ofdomestic consumption in the base period.
Although all WTO members established bound tariffs in the Uruguay Round, the actual applied tariffthat a country imposes may be lower than the tariffbinding. Unlike bound tariffs, applied tariffs may be raised above published levels (up to bound rates)without notice or compensation to affected trade partners. A comprehensive database of applied tariffdata across WTO members is not readily available.However, a subset of applied tariff data for severaldeveloping countries from the AMAD and UNCTADTRAINS databases was collected for this report. Thedata are used to illustrate the differences that may be observed between bound and applied rates in somecountries.
Bound Tariffs, TRQ Tariffs, and Applied Tariffs: What’s the Difference?
7 The WTO often uses the term “international tariff peaks” to referto tariffs above 15 percent. This definition has generally been usedwhen examining tariffs on imports of manufactures. For agricul-tural tariffs, however, defining international peaks as tariffs equal toor above 100 percent has more meaning.
Tariff Profiles by Region andCommodity
This section compares bound tariffs (see box, BoundTariffs, TRQ Tariffs, and Applied Tariffs: What’s theDifference? for information on different types of tar-iffs) on agricultural commodities across regional andcommodity groupings. The tariff means and medianspresented in this report are useful measures to comparethe potential levels of protection built into countries’tariff schedules. However, differences between boundand applied tariffs, market conditions, and other policies also influence the actual barrier to marketaccess.8
Tariffs by Region
Against a high global average tariff rate of 62 percent,considerable variation exists in tariff levels acrossregions. Average tariffs for WTO members by regionrange from an ad valorem tariff equivalent of 25 per-cent to 113 percent (figure 2). By region, average tariffs in agriculture are over 100 percent in two
regions: South Asia (113 percent) and the non-EUcountries of Western Europe (104 percent). At 25 per-cent, North America registered the lowest regional tar-iff average.
Table 2 displays mean and median tariffs, by regionalgrouping, for each of the commodity aggregates. For most of the developed country groupings, theregional tariff aggregates are among the lowestregional averages. The main exceptions to this trendare the non-EU countries of Western Europe, whichinclude Norway, Switzerland, and Iceland. Each ofthese has relatively high average tariffs, at 142, 120,and 113 percent, respectively. Like North America, theEU-15 also registers a relatively low average regionaltariff, at 30 percent.
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 9
8 The issue of “water in the tariff,” or the condition where a coun-try’s domestic price is below the import price plus the tariff, is anexample of how market conditions help determine the actual levelof protection. State trading enterprises and sanitary and phytosani-tary (SPS) measures are examples of policies that can raise theactual level of import protection.
World agricultural tariff averages, by regionFigure 2
0
20
40
60
80
100
120
South
Asia
Non-E
U Wes
t Eur
ope
Caribb
ean I
sland
s
Sub-S
ahar
an A
frica
North
Africa
Centra
l Amer
ica
Middle
East
Easter
n Eur
ope
South
America
Southe
rn A
frica
Asia-P
acific
Europ
ean U
nion-
15
North
America
Percent
Average agricultural tariff (62 percent)
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
10✺
������������������
������������������
������/ A
ER
-796E
conomic R
esearch Service/USD
A
����������������������������� ������������������������������
��������� .����� �*���� *��������� .����*�� .� �.� �*���� &����*���� 3���������
���� ������ ���� ������ ���� ������ ���� ������ ���� ������ ���� ������
*�������������� 45 46��� 47 '8 %%4��� %$$��� 68��� 9$��� 6%��� 47��� 79��� 48���
"�� � 46 44 :$ '8� %$4��� %$$��� 68��� 9$��� 97��� 47��� 7$��� %8���"�� ��������� 87 7% 87 '8 %%6��� %$$��� 68��� 9$��� 9'��� 47��� 78��� 48���;��� 74 74 '' %8 %%$��� %$$��� 68��� 9$��� 4'��� 47��� 47��� %8���.����� 5 5 :7 4$ %%4��� %$$��� 68��� 9$��� 84��� 47��� 7%��� 47���<������� 47 7$ 44 '8 %%$��� %$$��� 68��� 9$��� 84��� :5��� 4:��� '4���<���-� 44� 44 '' %$ %%'��� %$$��� 68��� 9$��� 69��� :5��� 45��� '8���=���� ������� 9%� 9% '7 '4 %47��� %$$��� 68��� 9$��� %$:��� 47��� 49��� '9���;���>����� '7� '$ '4 %: %'9��� %$$��� 68��� 9$��� :9��� 47��� 4:��� ':���?�#�� ���� $� $� 4$ '% %%%��� %$$��� 68��� 9$��� 54��� 47��� 74��� %:���3��/�����!�������@� ��������� 68 68 '6 '8 %%%��� %$$��� 68��� 9$��� %%%��� 9$��� 8$��� 48���3��/������ ���!����-!����������� 5� 6: 4'��� 4$��� %%4��� %$$��� 68��� 9$��� 57��� :9��� :'��� 48���3��/����@� � ���!����-!����������� :9 :5��� 4%��� 4$��� %%$��� %$$��� 68��� 9$��� 98��� :$��� :'��� 48���3��/�������� 77��� 46��� 48��� '8��� %''��� %$$��� 68��� 9$��� 89��� 47��� :6��� 48���.-� ��>����� '$��� '$��� '$��� %$��� %$%��� %$$��� 68��� 9$��� 8$��� 47��� 45��� %8���2��� 46��� ''��� 64��� 4$��� %$7��� %$$��� 68��� 9$��� 67��� 96��� :8��� 48������� %5��� %5��� '9��� '6��� %%5��� %$$��� 68��� 9$��� 4:��� 47��� 84��� 48���;����/����� ''��� 4$��� 4$��� 4$��� %$9��� %$$��� 68��� 9$��� 4:��� 47��� :8��� 48���;����/����@� ':��� ':��� 4$��� '8��� %'$��� %$$��� 68��� 9$��� 47��� 47��� 87��� 84���;����/�������>����������� ���!������>������ 7��� $��� ':��� '9��� %%'��� %$$��� 68��� 9$��� 47��� 47��� 6$��� 48���
;����/������������ �� '4��� '4��� '8��� '8��� %$6��� %$$��� 68��� 9$��� 47��� 47��� %$%��� 48���;����/��������� � 46��� 46��� '9��� '9��� %%6��� %$$��� 68��� 9$��� 47��� 47��� 8$��� 48���;�����A���� ':��� ':��� '9��� '7��� %$6��� %$$��� 68��� 9$��� 47��� 47��� 85��� 48���=���� ���/����� 4%��� 44��� 4%��� '8��� %%$��� %$$��� 68��� 9$��� 4$��� 47��� :$��� 48���=���� ���/����@� 46��� 46��� '7��� %5��� %%%��� %$$��� 68��� 9$��� 47��� 47��� 6$��� 76���=���� ���/����@� ������������������ 47 46 49��� '6��� %%7��� %$$��� 68��� 9$��� 48��� 47��� 88��� 48���=���� ���/�������>������������>��� ��� 6��� $��� 67��� '8��� %%8��� %$$��� 68��� 9$��� 4%��� 47��� 77��� %5���=���� ���/������ 4:��� 48��� 87��� '8��� %$9��� %$$��� 68��� 9$��� 8'��� 47��� 8:��� 48���=���� ���/��������� � 74��� 74��� '9��� '7��� %$5��� %$$��� 68��� 9$��� 48��� 47��� 85��� 45���=���� ���A����/������ ':��� ':��� 4'��� 4$��� %$6��� %$$��� 68��� 9$��� 47��� 47��� 59��� :9���&��� %9��� $��� 4%��� '8��� %%4��� %$$��� 68��� 9$��� 89��� 47��� 76��� 48���&����>������/������!�����!�>�������� 49��� 46��� 4$��� '8��� %%6��� %$$��� 68��� 9$��� 47��� 47��� 88��� 49���B�����������/���#� $��� $��� '4��� '8��� 5%��� %$$��� 68��� 9$��� 47��� 47��� 8%��� '$���B�����������/�������������>������� 75��� 75��� '5��� 4$��� %'$��� %$$��� 68��� 9$��� 47��� 47��� ::��� 48���.���� ��� ':��� ':��� ''��� %9��� %%$��� %$$��� 68��� 9$��� 44��� 47��� 49��� '4���.����� � ':��� ':��� '7��� '8��� %%$��� %$$��� 68��� 9$��� 4'��� 4'��� 49��� '4���.����� ��� ''��� ''��� 49��� 4$��� %'%��� %$$��� 68��� 9$��� %74��� %:9��� 7'��� %8���(� ���/�� � �������� 77��� 77��� '$:��� 7$��� %%$��� %$$��� 68��� 9$��� 59��� 88��� 89��� 48���(� ���/��������� 87��� 87��� 4'��� 4$��� %%5��� %$$��� 68��� 9$��� 85��� 48��� 65��� %$$���;� �� %4��� 5��� '%��� %$��� %$7��� %$$��� 68��� 9$��� 74��� 47��� 46��� %8���;������������ � 84��� 46��� 44��� '4��� %%6��� %$$��� 68��� 9$��� :4��� 47��� 8$��� 48��������� %%5��� %%5��� '5��� 4$��� %%9��� %$$��� 68��� 9$��� 4$��� 47��� 44��� %8���������/����� 69��� 69��� 4'��� '9��� %%9��� %$$��� 68��� 9$��� 44��� 47��� 45��� 48���(��>����������� 8%��� 8% 7%��� 4$��� %4$��� %8$��� 68��� 9$��� 48��� 47��� 48��� %8�������� � ��>��������� %$��� %6 '8��� '8��� %%8��� %$$��� 68��� 9$��� 47��� 47��� 46��� %6���.����� 6��� $ '7��� '$��� %%7��� %$$��� 68��� 9$��� 47��� 47��� 46��� '%������� �������� ''��� ''� ''��� %:��� %'$��� %$$��� 68��� 9$��� %$$��� %$$��� 77��� 48���.������� ������� ������ ��
���������
Econom
ic Research Service/U
SDA
������������������
������������������
������/ A
ER
-796✺
11
����������������������������� ���������������������������������������������� ����� ������� &� ��1������������ ������ �1 �� �%8 .����*����� �� ����*����� ��� � ���� �� &����*�����
���� ������ ���� ������ ���� ������ ���� ������ ���� ������ ���� ������ ���� ������
*�������������� 75 '$��� %$7��� 78��� 4$��� %4��� 45��� 48��� 87��� 78��� 9:��� %$$��� '8��� :���
"�� � 76��� '$ %$$��� 8$��� 84��� :4��� 7:��� 48��� 88��� 78��� 9:��� %$$��� '8��� 4���"�� ��������� :8 48��� %''��� 84��� 79��� 78��� 7$��� 48��� 8$��� 7$��� 9:��� %$$��� %5��� 8���;��� %6 '��� %4%��� 8$��� 76��� %%��� 45��� 48��� 8$��� 7$��� 9:��� %$$��� '4��� 4���.����� 75��� 48 54��� 78��� '7��� '$��� 49��� 48��� 84��� 7$��� 9:��� %$$��� %7��� 6���<������� %7��� ' 5$��� 8$��� $��� $��� 46��� 48��� 7'��� 7$��� 9:��� %$$��� %9��� $���<���-� 5��� $ 9%��� 78��� 4��� $��� 7$��� 48��� 78��� 48��� 9:��� %$$��� %4��� 4���=���� ������� 47��� 5 58��� 99��� %4��� :��� 45��� 48��� 6'��� :$��� 65��� %$$��� %6��� 9���;���>����� 44��� %4 98��� 8$��� %$��� 4��� 49��� 48��� 84��� 74��� 9%��� %$$��� '9��� 9���?�#�� ���� :8��� '5 '44��� 6$��� 4$��� ''��� 47��� 4$��� 79��� 7$��� 99��� %$$��� '%��� '���3��/�����!�������@� ��������� :5 %6��� '$:��� %66��� 6$��� 67��� 49��� 48��� 84��� 7$��� 9:��� %$$��� %$��� %���3��/������ ���!����-!����������� 5$ 45��� '67��� 8$��� 7%��� '6��� 74��� 48��� :9��� :$��� 5%��� %$$��� 75��� ':���3��/����@� � ���!����-!����������� 9' 74��� 4$5��� 4%$��� ::��� 49��� 74��� 48��� 9:��� :$��� 5$��� %$$��� 9$��� '6���3��/�������� 67 45��� '9'��� '%5��� 74��� ':��� 7%��� 48��� 88��� 78��� 5$��� %$$��� 7%��� :���.-� ��>����� 78 %��� ''��� %$��� $��� $��� 46��� 48��� 85��� 8$��� 9:��� %$$��� :��� $���2��� 98 8%��� '4$��� 54��� 96��� 6$��� 74��� 48��� :9��� :$��� 96��� %$$��� 98��� 78������� 75 ':��� %95��� :%��� ''��� '7��� 49��� 48��� 8%��� 7$��� 9:��� %$$��� :$��� 49���;����/����� 45� %8 8%��� '$��� '%��� %'��� 7$��� 48��� 8'��� 7$��� 9:��� %$$��� %$��� '���;����/����@� 45 %6 47��� 7$��� '$��� '%��� 7$��� 48��� 8'��� 7$��� 9:��� %$$��� %6��� %$���;����/�������>������
����� ���!������>������ %7 %9 '%��� '$��� 7��� :��� 7%��� 48��� 88��� 78��� 9:��� %$$��� %8��� 8���;����/������������ �� %: ' %5��� %$��� '��� '��� 49��� 48��� 75��� 7$��� 9:��� %$$��� 6��� '���;����/��������� � 75 '$ 79��� 78��� '%��� '%��� 45��� 48��� 8'��� 7$��� 9:��� %$$��� %'��� 6���;�����A���� :: '$ 75 ''��� 46��� ''��� 46��� 48��� 79��� 78��� 9:��� %$$��� %'��� 4���=���� ���/����� '9 %4 %68��� 8$��� %:��� %$��� 7%��� 48��� 87��� 78��� 9:��� %$$��� %%��� 4���=���� ���/����@� 76 %5 %7:��� 8%��� %7��� %8��� 45��� 48��� 78��� 7$��� 9:��� %$$��� %6��� %$���=���� ���/����@� �
����������������� '4 5 %$4��� 8$��� %9��� %'��� 7$��� 48��� 8$��� 7$��� 9:��� %$$��� %4��� %$���=���� ���/�������>�
�����������>��� ��� 4'��� 5��� 6$��� 78��� 49��� %:��� 45��� 48��� 7:��� 7$��� 96��� %$$��� %%��� 4���=���� ���/������ ''��� %9��� 76��� '6��� '��� $��� 4:��� 4$��� 87��� 78��� 9:��� %$$��� %%��� 4���=���� ���/��������� � 76��� %5��� %'4��� 8$��� '%��� %7��� 49��� 48��� 8%��� 7$��� 9:��� %$$��� %'��� 9���=���� ���A����/������ 99��� ':��� ':��� ''��� %:��� %:��� 45��� 48��� 8'��� 8$��� 9:��� %$$��� '8��� '8���&��� %6��� 9��� 4%��� %$��� 8��� 7��� 49��� 48��� 8'��� 7$��� 9:��� %$$��� %9��� 7���&����>������/������!�����!�>�������� 46��� %$��� 75��� 4'��� %:��� %6��� 49��� 48��� 8%��� 7$��� 9:��� %$$��� %%��� :���B�����������/���#� 9��� '��� :6��� %6��� 8��� 6��� 44��� 4$��� 76��� 7$��� 9:��� %$$��� %��� %���B�����������/�������������>������� 47��� %6��� 5%��� 78��� 8��� 4��� 4:��� 48��� 78��� 7$��� 9:��� %$$��� %4��� 6���.���� ��� 75��� 4$��� %77��� 79��� 475��� 475��� 49��� 48��� 78��� 49��� 9:��� %$$��� %'��� $���.����� � 47��� %8��� 55��� 78��� 8:��� 8:��� 49��� 48��� 78��� 49��� 9:��� %$$��� %'��� %���.����� ��� 64��� :$��� 9'��� 8$��� 85��� 86��� 45��� 48��� :8��� :$��� 9:��� %$$��� 8$��� %$���(� ���/�� � �������� 7'��� 8%��� '9��� 4$��� %7��� %%��� 49��� 48��� :7��� 67��� 9:��� %$$��� '9��� 8���(� ���/��������� :7��� 88��� '5��� '4��� 49��� 47��� 49��� 48��� ::��� 67��� 9:��� %$$��� %%'��� %4���;� �� 7$��� 4��� '4��� 4��� $��� $��� 46��� 48��� 85��� 78��� 9:��� %$$��� %'��� 4���;������������ � 79��� 4$��� %$8��� 76��� %8��� %4��� 4:��� 48��� 79��� 7$��� 98��� %$$��� 4$��� %%��������� ''��� %4��� '$��� 5��� :��� 9��� 49��� 48��� 8$��� 7$��� 9:��� %$$��� %9��� $���������/����� '8��� %9��� 46��� ''��� %$��� %'��� 49��� 48��� 86��� 78��� 9:��� %$$��� %5��� %���(��>����������� %5��� '8��� '4��� 8��� '��� $��� 49��� 48��� 87��� 78��� 9:��� %$$��� %7��� $�������� � ��>��������� :%��� %4��� 97��� 7%��� %6��� %8��� 4:��� 48��� 88��� 78��� 9:��� %$$��� '9��� %9���.����� %7��� 5��� ':��� 7$��� '��� $��� 49��� 48��� 8$��� 7$��� 9:��� %$$��� 9��� '������� �������� '7��� 5��� '4��� 4��� 4��� 4��� 4%��� '$��� 85��� :$��� 9:��� %$$��� %7��� 7���% (��������� �� ��3;&������ ����� ��� ��10** ������� ���� )�������.�����/����� �����0������.��#���!�1.2*)
12 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
The average tariff of the Asia-Pacific region of 34 per-cent conceals a wide range of country averages. Aver-age tariffs of 10 percent or less are registered for Sin-gapore, Hong Kong, Macau, Australia, and NewZealand. However, tariffs for Japan and Korea average58 percent and 66 percent, respectively. Most of theremaining countries in the region maintain tariffs of 25to 45 percent.
With the exception of the high tariff region of non-EUWestern Europe, the regional groupings with the high-est average tariffs are regions that comprise non-OECD countries. The developing country regions ofAfrica, the Caribbean, and South Asia, with averagesranging from 71 to 113 percent, are all above theglobal average rate of 62 percent. The regions of SouthAmerica, the Middle East, and Central America allhave tariff averages ranging from 39 to 54 percent.
Within most regional groups, the tariff means acrosscommodity groupings tend to show a high degree ofvariation. In the three European regional groupings, aswell as in North America, North Africa, and the Asia-Pacific region, there is a high dispersion rate acrosscommodity means. In particular, tariffs greater than theoverall average tariff on agriculture of 62 percent arefound in the meat, dairy, sugar, and sweetener cate-gories. In addition, in some regions, comparativelyhigh tariffs are recorded for tobacco, oils, and severalcategories of prepared vegetables.
In a few regions, tariffs tend to be rather uniform, withaverage rates across commodity groups varying littlefrom the regional averages. The average rate of 75 per-cent found across all commodities in Sub-SaharanAfrica, for example, reflects the fact that each countryin the region set a uniform tariff rate across the
World agricultural tariff averages, by commodityFigure 3
Toba
cco:
unm
anuf
actu
red
Dai
ry
Mea
t: fro
zen
beef
, por
k, o
r pou
ltry
Mea
t: fre
sh b
eef,
pork
, or p
oultr
yM
eat:
Pre
pare
d
Mea
t: fre
sh, o
r fro
zen
othe
r mea
t
Sug
ar b
eet
Veg
etab
les:
fres
h
Gra
in p
rodu
cts
Sw
eete
ners
Live
ani
mal
sS
tarc
hes
Egg
s
Gra
ins
Toba
cco:
pro
duct
s
Veg
etab
les:
drie
d &
fres
h ro
ots
& tu
bers
Veg
etab
les:
froz
enV
eget
able
juic
e: to
mat
oV
eget
able
oils
Sug
ar c
ane
Veg
etab
les:
froz
en o
r pre
pare
d (o
ther
)
Veg
etab
les:
pre
para
tions
Food
pre
para
tions
Coc
oa b
eans
& p
rodu
cts
Veg
etab
les:
drie
d
Oilc
ake
Oils
eeds
Feed
Tea
& te
a ex
tract
sFr
uit:
fresh
Ski
ns &
hid
es
Nut
sFa
ts &
oils
Frui
t jui
ce
Hor
ticul
ture
: cu
t flo
wer
s &
folia
geC
offe
e: o
ther
Cof
fee
Frui
t: fro
zen
Frui
t: dr
ied
& fr
esh
(coc
onut
s, d
ates
& fi
gs)
Ess
entia
l oils
Fibe
rFr
uit:
prep
arat
ions
Frui
t: dr
ied
(rai
sins
)S
pice
s
Nut
s &
frui
t: dr
ied,
fres
h, &
pre
pare
dH
ortic
ultu
re: l
ive
Percent
Average agricultural tariff (62 percent)
0
10
20
30
40
50
60
70
80
90
100
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 13
entirety of its agricultural tariff schedule. While theseuniform rates differ by country, when averaged acrosscommodity groupings, the mean tariffs are identicalfor each. In South Asia, Central America, and South America, mean tariffs across the commoditygroupings are also relatively close to the overallregional averages.
Tariffs by Commodity
Figure 3 compares average tariffs by commodity groupfor all WTO members reviewed in this report. Of the46 commodity aggregates listed, average tariffs on 18of the groups are above the global agricultural tariffrate of 62 percent. These commodity groups are madeup of tobacco, dairy, meats, sugar, sweeteners, severalcategories of vegetables, grains, grain products, andbreeding animals. Tariffs on the remaining 28 com-modity groups are at or below the 62-percent average
tariff. At rates of more than 50 percent, commoditygroups with the lowest tariffs (coffee, fiber, severalfruit categories, spices, and live horticulture), are nev-ertheless all relatively high. These high global tariffrates across all commodity groupings reflect the hightariffs found in many developing countries’ WTOschedules.
Table 2 also reveals regional patterns in tariff protec-tion by commodity. For North America and the EU,tariffs on most of the commodity groups are beloweach region’s respective average tariff. Commoditieswith tariffs below regional means in both the EU andNorth America include live breeding animals, coffee,fiber, fruit, nuts, oilseeds and oilcake, skins and hides,spices, tea, and vegetables. Unmanufactured tobacco inthe Asia-Pacific region and sweeteners in North Africahave the highest tariffs for any commodity category inthe respective regions.
High protection evident in global mean, median pairs
0
20
40
60
80
100
120
140
160
180
200
Figure 4
0 20 40 60 80 100 120 140 160 180 200
Mean
Median
Global mean (62%)
Global median (40%)
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
14 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Means and Medians across Commodities and Regions
Several patterns emerge from comparing means andmedians across the regional and commodity averages.Figure 4 plots all the entries in table 2 where the meanfalls within 200 percent (this excludes only eightregional-commodity points). The mean is on the hori-zontal axis and the median is on the vertical axis. Theglobal mean and median are marked with solid lines.They divide the figure into four quadrants that corre-spond to the categories presented in the methodologysection, e.g., low mean/low median in the lower left-hand corner, high mean/high median in the upper
right-hand corner. The dashed line at a 45-degree angleshows where the mean equals the median.
Several observations from figure 4 help characterizethe global pattern of protection. First, most tariffs liebelow the 45-degree line, indicating that high tariffs ona few specific lines are one cause of high overall aver-ages. This is also reflected in a global median of 40percent that lies below the global mean. The distancebelow the 45-degree line is an indicator of the magni-tude of the influence of megatariffs on the mean value.For example, the point with a mean of 87 and amedian of 70 represents EU dairy and shows that tar-iffs are high across most specific products. In contrast,the point 88, 26 represents Eastern Europe vegetable
Low tariffs in North America
Commodity and Country SnapshotsFigure 5a
Median
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100 120 140 160 180 200
Mean
Global mean (62%)
Global median (40%)
High tariffs protect traditional sectors in EUFigure 5b
Median
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100 120 140 160 180 200
Mean
Global mean (62%)
Global median (40%)
Megatariffs' importance shown in dairy tariffsFigure 5c
Median
Global median (40%)
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100 120 140 160 180 200
Mean
Global mean (62%)
High tariffs are the exception for fruitFigure 5d
Median
Global median (40%)
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100 120 140 160 180 200Mean
Global mean (62%)
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 15
juice (tomato) and indicates that a few, very high tar-iffs distort the average.
The large number of tariffs in the low mean/lowmedian category indicate that the majority of commod-ity averages for these regions lie below the globalaverage. A comparison with the global average manu-facturing tariff of 5 percent (Hertel and Martin) makesthe point that all but a very few regional, commoditypairs lie above this average. Although relativelysmaller in number, the scatter plot shows that a signifi-cant number of commodity groups across regions havehigh and likely prohibitive tariffs. Thus, the problemof high tariffs is not isolated in a small number ofcountries and commodities.
Figures 5a-5d show how scatter plots of means andmedians can help uncover patterns of protection forboth countries and commodities. Figures 5a and 5bdisplay the commodity mean and median combina-tions for North America and the EU. Both regionshave a small number of commodities with tariffs abovethe global average: meats, dairy, sugar (EU only), andtobacco products (North America only). EU tariffs onfresh meat and dairy have a higher median value indi-cating that a larger percentage of the tariff lines havehigh tariffs. Comparison of the two charts also showsthat both countries have a large number of commoditycategories with average tariffs of 20 percent or less,although North America has the largest number.Although no charts are included for the Asia-Pacificregion, examination of data for this region shows asimilar pattern to North America and the EU with the
highest levels of protection applied selectively, withina commodity grouping. A different pattern emerges for Central and South America, where tariff means areclustered in a smaller range. Many countries in theCentral and South America region also show tariffmedians that are at levels comparable to the means,suggesting that tariffs are not skewed by a large num-ber of megatariffs. In non-EU Western Europe, and inin developing countries, high levels of protection tend to be specified more broadly across commoditygroups.
Differences in the profile of tariffs across countries fordairy compared to fruits are shown in figures 5c and5d. All but two average tariffs for dairy (South Amer-ica and Southern Africa) are above the global mean. Inaddition, the median for each region is close to themean, indicating that high tariffs exist across mostlines. The fruits chart (figure 5d) encompasses sixcommodity categories and contains more observationsthan dairy (figure 5c), which is a single commoditycategory. The means and medians show a different pat-tern of protection where most tariffs are below theglobal average and median values lie close to themean. However, the scatter plot detects the presence ofhigh tariffs, including points in the high mean/lowmedian quadrant that indicate megatariffs that likelyisolate domestic industries from international competi-tion—for example, juice in Eastern Europe, and freshand dried fruit in the Middle East. The high tariffs inthe high mean/high median quadrant represent ceilingbindings for Sub-Saharan Africa and South Asia.
16 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Tariff Rate Quotas
On the surface, TRQs pose a paradox in that they coveronly 6 percent or a small set of tariff-lines and are usedby only 35 of 113 countries in this study, but are per-ceived to play an important role in agricultural protec-tion. TRQs began as an instrument to provide limitedmarket access for sensitive commodities because coun-tries were worried that tariffication in the URAA wouldlead to extremely high tariffs. The use of TRQs in mostregions makes them a factor in trade around the globe.TRQs were scheduled by countries in all regionsreviewed in this report except for South Asia and Sub-Saharan Africa. Table 3 lists the countries that useTRQs, and these countries include the largest agricul-tural importing members of the WTO. Appendix tables1 and 2 contain average tariffs by region and by chapter
of the harmonized system and show that TRQs existacross all commodity groups.
Some notable differences across TRQs show that whilethe problem touches most regions and commodities,the role of TRQs varies significantly. Some regions,such as Eastern Europe, scheduled TRQs for productsin most commodity groups (see appendix tables 1 and2). A more common practice was to schedule TRQsfor a subset of specific, narrowly defined commoditiesor sub-commodities. Looking across commodities, inall regions with TRQs, at least one country scheduledTRQs for meats; dairy; cereals; and preparations ofvegetables, fruits, nuts, or other parts of plants. Theprevalence of TRQs in the sensitive sectors of meats,dairy, and cereals provides at least a partial explana-tion for their importance in trade policy discussions.
�������� � ����� ���� ��������������������!���� ���"# ���������$������������ �<#���C��� (������� �� (������� �� (0D��� ����
��� ��� (���� � �C��������� ����� (������� �� ����� �C��� �����#��� ����� ��������� ��� �(0D��� ��� �(0D��� ��� �������(0D� ������ C���������� �������
��������������������������� ��������������������������������� ������������������������������������������������������������� ������
*#���������� �� 75 :4 %'9 �� �� �� '4(��������� �� �� �� �� '6!776 7!554 7!56' ��
*������ 7 %$ '8 698 %% 5 'E� ��� %$' %7% %%7 :$6 46 %$5 %5E���� 45 '$ :5 '5: 6' %69 7:E�@�� 46 6 7' %7%8 7 7 %� � '4 4 %45 %$'$ %7% %8% ''����� � 96 %4' %46 '47 68 :6 49�����0�� 7' 77 :9 ::5 64 9: %5�@���0��� ��� %' '9 79 %6'9 7: '7: %7������ ': 4$ 74 996 %9 '' 7���.�#��� 7% '8 68 :6$ 46 95 %:������ �1 �� 4$ %6 69 %854 444 '97 '9"����� 75 4' %%9 :55 4% %$: %:B� ��� '5 ': 7$ 758 9: 7%: 8$���� � %%4 75 %9% 6%6 8'7 7%6 86� �� ��� 79 :8 %65 %!4%9 ' %4 %����� 68 65 %8% 966 '4 46 :F� 89 '' 7'' %%9% %99 %'' '%G���!�0��� ������ :: %5 4%7 %%47 %58 %64 '83���� '8 %$: '79 %'49 64 6% %$3����� 74 79 %79 99' :5 :9 %43������ :8 %79 %%8 %'$8 4$ '65 '$&���H�� � 6 $ 6 565 7 7 %&����� :% 77 6% :6% %6 '5 :&���� %7' ':' '$4 6'' 4:9 8$' 88� � 74 %8 94 %9% 86 8: 49������� �� 47 7$ 4: :9$ %8 :6 %%��� � 79 4% 85 64 %:5 '89 98.��#-�0��� ��� %4 4$ 7' 659 7$ %66 '%.��#� � 8% %6 6% %:$ 44 44 '5.����*���� 45 '$ :5 '5: 6' %69 7:.���@��� � %'$ 68 '%$ 987 '4: 458 7'(�� � 48 '6 5% :94 48 87 %'(� ��� %%$ ': %$5 '9: %7 74 %61 �����.���� %' %$ 8' %%59 %5$ %9' '7=� �@��� 8' 46 %$% '%: :' :4 46
���I�&�������� ��)%(��������� �� ��3;&������ ����� ��� ��10** ������� ���� ����.�����/����� �����0������.��#���!�1.2*)
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 17
High Tariffs Characterize TRQs
Many of the high over-quota tariffs for TRQs appear toreflect countries’ objectives to protect sensitive agri-cultural sectors. The average over-quota tariff for TRQlines is 128 percent (table 3). The average over-quotatariff for 25 of the countries is higher than the 62-per-cent average for all tariffs. Of this group, the averageover-quota tariff of 8 countries is between 100 and 150percent. Over-quota rates of 6 countries are between150 and 250 percent, and 2 countries, Japan andKorea, schedule over-quota rates that average above300 percent.
At 63 percent, the average in-quota tariff equalsapproximately the overall average tariff of 62 percent.In general, WTO members scheduled in-quota tariffrates at less than 50 percent. However, eight WTOmembers set in-quota tariffs over the global averagetariff of 62 percent: Norway, Morocco, Barbados,Colombia, Malaysia, Israel, Switzerland, and Indone-sia. Of this group, most scheduled in-quota tariffsbetween 65-150 percent. Although in-quota tariffswere designed to provide market access for a limitedquantity of imports at relatively low tariffs, table 3
shows that, in practice, in-quota tariffs were alsoscheduled at very high levels.9
The ratio of the average tariff for all tariff-lines com-pared with the average for only the over-quota TRQlines supports the expectation that TRQs generally pro-tect sensitive sectors. Figure 6 shows the ratio of theaverage tariff for all lines to the average tariff for over-quota TRQ lines. In 14 countries, the average tariff forthe TRQ lines is at least twice that for all lines. Threecountries stand out with rates more than six times thatof the average for all lines. Australia, with one of theoverall lowest tariff averages, has a small number ofTRQs that protect the dairy sector. Canada’s TRQs pro-tect mainly the dairy and poultry sectors and have anaverage over-quota tariff of 139 percent, although verylow in-quota tariffs. Not surprising, with the highestaverage over-quota rate at 388 percent, Japan’s over-quota rate is seven times higher than its overall aver-age. While potentially posing a barrier to its markets,Japan’s in-quota average of 22 percent represents asmall fraction of the over-quota rate.
9 See analysis in Burfisher et al., for analysis of administration ofTRQs and the possible role of in-quota tariffs.
TRQs are associated with high over-quota tariffsFigure 6
Aus
tralia
Bar
bado
sB
otsw
ana
Bra
zil
Can
ada
Col
ombi
aC
osta
Ric
a C
zech
Rep
ublic
Ecu
ador
El S
alva
dor
Eur
opea
n U
nion
Gua
tem
ala
Hun
gary
Icel
and
Indo
nesi
aIs
rael
Japa
nR
epub
lic o
f Kor
eaM
alay
sia
Mex
ico
Mor
occo
New
Zea
land
Nic
arag
uaN
orw
ayP
anam
aP
hilip
pine
sP
olan
dS
lova
k R
epub
licS
love
nia
Sou
th A
frica
Sw
itzer
land
Thai
land
Tuni
sia
Uni
ted
Sta
tes
Ven
ezue
la
Ratio
Over-quota average/Country average
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
6
0
12
10
8
4
2
18 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Summing Up Tariff Protection:OECD vs. Non-OECD
Indicators of tariff protection for OECD and non-OECD countries in table 4 complement the characteri-zation of tariffs using means and medians. The firstfour columns show averages for the commodity cate-gories in absolute terms (table 4, columns 1 and 2) and as a percentage of the global mean tariff (table 4,columns 3 and 4). The data reinforce the finding thatprotection in OECD countries is concentrated in a few sectors: grains, dairy, livestock, sugar, and sweet-eners. Non-OECD countries have overall high rates of protection with less variation across commoditygroupings. They have high protection on the samecommodities as OECD countries, but tobacco standsout with the highest average tariff for non-OECDcountries.
The number of countries with tariffs higher than theglobal average shows the prevalence of high tariffsacross countries. As in other cuts of the data, dairystands out with the highest mean in OECD countries.Dairy also has the largest number of OECD and thirdlargest number of non-OECD countries with highmeans. Sweeteners and frozen meat also have highmeans across a large number of countries.
OECD countries use megatariffs in a limited numberof commodity groups, but have TRQs in all but two
commodity groups. The concentration of megatariffsamong the familiar sensitive sectors is another mani-festation of high protection for a few (albeit large) sec-tors. The number of TRQs notified by OECD countriesis also concentrated in a few sectors. However, TRQsare found in at least one region for most commoditygroups, indicating that sensitive products exist acrossthe agricultural sector.
Non-OECD countries rely on megatariffs for protec-tion along with more selective application of TRQs.Non-OECD, or developing countries, often apply tar-iffs far below these high, bound rates. The followingsection examines the use of applied tariffs in develop-ing countries.
Overall, different patterns of protection betweenOECD and non-OECD countries emerge. OECD coun-tries have higher rates on “traditional” agricultural sec-tors, such as dairy, livestock, and sugar, while non-OECD countries have high tariffs across most com-modities. Both OECD and non-OECD countries pro-vide extremely high protection to a few commodities.However, as a result of tariffication, OECD countriesapply more TRQs than non-OECD countries. Non-OECD countries use megatariffs more than OECDcountries. Many of the megatariffs associated withdeveloping countries were not subject to reductionunder the Uruguay Round because they were estab-lished as ceiling bindings.
Econom
ic Research Service/U
SDA
������������������
������������������
������/ A
ER
-796✺
19
������%�&���������� �������������� ������������� ����'() �����!'() �
(���� (����J��� ��#���������� ��� �����������#��������K�:' ��� ������������������� ��� ����������(0D�
��������� <��2 &� �<��2 <��2 &� �<��2 <��2 &� �<��2 <��2 &� �<��2 E�� <��2 &� �<��2 E��
��� ��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
(� ���/�� � �������� '' %%$ $)7 %)9 $ 7' $ 4$ 4$ 4 8 92��� %%: 67 %)5 %)' 5 77 8 '5 47 %' %: '93��/����@� � ���!����-!����������� %$: 68 %)6 %)' : 76 7 4% 48 %$ %6 '63��/������ ���!����-!����������� 5: 64 %)8 %)' 7 78 4 4$ 44 %$ %: ':3��/�������� 5' :9 %)8 %)% 8 49 4 '5 4' 9 6 %83��/�����!�������@� ��������� 9' :5 %)4 %)% 7 7' 4 '8 '9 : %$ %:.���� ��� %$7 :7 %)6 %)$ 7 48 4 '9 4% ' $ '=���� ���/����� 96 :7 %)7 %)$ 7 46 4 '9 4% 5 9 %6"�� ��������� 98 :6 %)7 %)% 8 4: 4 '9 4% 5 5 %9.����� ��� :7 6$ %)$ %)% : 7% ' 4% 44 9 %' '$?�#�� ���� 9' :: %)4 %)% 7 49 7 4$ 47 6 : %4.����� 97 :7 %)4 %)$ 7 4: 4 '9 4% 6 % 9���� 67 :: %)' %)% 7 4: 7 '6 4% 9 8 %4"�� � 69 :: %)4 %)% 8 7$ 4 '9 4% %$ %: ':(� ���/��������� 8% :5 $)9 %)% 4 7% ' '5 4% % 4 7=���� ���/�������>������������>��� ��� 68 :' %)' %)$ 4 4: ' '6 '5 7 % 8=���� ���/����@� :% :7 %)$ %)$ ' 46 % '5 4$ 7 ' :=���� ���A����/������ '% 6% $)4 %)' $ 4: $ '9 '9 % $ %<���-� 4% :7 $)8 %)$ 4 48 % '6 '9 4 : 5.����� � 8' :7 $)9 %)$ ' 48 % '9 '5 ' $ '=���� ���/����@� ������������������ 8' :4 $)9 %)$ 7 4: 4 '9 4% 6 ' 5=���� ���/��������� � 76 :7 $)9 %)$ 7 49 ' '9 4$ : : %';������������ � 84 :' $)9 %)$ 4 49 ' '6 '5 %$ 6 %6����� � ��>��������� 7% :7 $)6 %)$ 4 48 % '9 '5 6 $ 6=���� ���/������ 76 :' $)9 %)$ 4 46 4 '9 4% 6 6 %7<������� 7: :' $)6 %)$ 8 48 ' '9 4$ : : %';��� 79 :% $)9 %)$ ' 47 ' '6 '5 : $ :(��>����������� 4$ :4 $)8 %)$ % 48 % '6 '9 8 % :;����/�;��� '8 :8 $)7 %)% % 46 $ '6 '6 9 8 %4.-� ��>����� 7 :8 $)% %)$ $ 48 $ '6 '6 $ $ $&��� '% :4 $)4 %)$ % 4: % '9 '5 8 7 5;���>����� 4: :' $): %)$ 4 4: % ': '6 6 8 %';�����A���� '8 :9 $)7 %)% % 4: $ '9 '9 : ' 9B�����������/�������������
>������� 44 :4 $)8 %)$ ' 49 % '9 '5 ' $ '������/����� '5 :4 $)8 %)$ % 4: $ '6 '6 7 ' :������ %4 :4 $)' %)$ $ 48 $ '6 '6 % 4 7;����/�;��@� %9 :7 $)4 %)$ % 46 $ '6 '6 % $ %;����/�������>������
����� ���!������>������ %% :' $)' %)$ $ 4: $ '9 '9 % % '���� �������� 5 :7 $)% %)$ $ 48 $ '9 '9 $ $ $;� �� 9 :4 $)% %)$ $ 4: $ ': ': 4 4 :;����/��������� � %5 :: $)4 %)% % 46 $ '9 '9 ' 4 8;����/������������ �� 6 :% $)% %)$ $ 4: $ '9 '9 % 4 7.����� %$ :% $)' %)$ % 47 $ '6 '6 ' % 4&����>�������������!�����!�>��������� '' :' $)7 %)$ 4 47 % '9 '5 7 7 9B�����������/���#� 4% 85 $)8 %)$ 4 46 ' '6 '5 4 $ 4=���� ������� 45 :9 $): %)% 4 7% ' '9 '9 7 %$ %7
&���/��*#��������������������� ���������������#��������������� ����� ���������������� )���������.�����/���� �����0������.��#���!�1.2*)
20 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Comparison of Bound andApplied Tariffs
As shown in the preceding sections, many WTO mem-bers maintain high bound tariffs in their WTO marketaccess schedules. In practice, however, not all coun-tries apply tariffs at the bound rate (see box page 8,Bound Tariffs, TRQ Tariffs, and Applied Tariffs: What’sthe Difference?). Latin American countries present agood example of this, partly because of data availabil-ity. Figure 7 compares final WTO bound tariffs with1998 applied tariffs for 15 countries, 12 of which arein Latin America. The final bound tariff for thesecountries is the tariff binding to be effective no laterthan 2004. In all cases, the average 1998 applied tariffis considerably lower than the final 2004 WTO boundrate. The average bound tariff for the 12 countries is45 percent, while the average applied tariff in 1998was 13 percent, or less than one-third the level of theaverage bound tariff. Not only do they tend to belower, there is also less dispersion across applied tar-iffs than corresponding bound rates. While the averagebound tariffs of these countries range rather widely,
from 26 to 110 percent, the average applied tariff in1998 fell within a much lower and narrower range of10 to 43 percent.
For developing countries in other regions, a more lim-ited set of applied tariff data for one or more of theyears 1995-99 was available. Table 5 presents thesetariff averages. For 7 of the countries listed, appliedtariffs for the various years reported were at levels thataveraged from about one-quarter to about three-quar-ters of the bound rates. India, Pakistan, and Tunisia allscheduled final bound tariffs which average over 100percent, while the applied tariffs for the years shownwere considerably lower, at between 30 and 43 per-cent. Korea and Morocco, however, set applied tariffsfor the years listed at about 75 percent of bound rates.
Although the countries listed above apply tariffs belowbound rates, many developing countries and mostdeveloped countries tend to apply tariffs at the boundrate. Some countries, such as Thailand and Turkey,appear to be violating their Uruguay Round commit-ments by applying tariffs at rates higher than theirbound rates, but this is explained by the fact that many
The difference between bound and applied tariffs in selected developing countriesFigure 7
Arg
entin
a
Bra
zil
Col
ombi
a
Ecu
ador
El S
alva
dor
Gua
tem
ala
Mex
ico
Nic
arag
ua
Pak
ista
n
Pan
ama
Par
agua
y
Phi
lippi
nes
Tuni
sia
Uru
guay
Ven
ezue
la
Average tariff
0
20
40
60
80
100
120
BoundApplied 1998
1
Bound tariffs are MFN rates based on final URAA implementation, and applied tariffs represent annual average.
Source: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 21
of their bound tariffs are not scheduled to becomeeffective until 2004. The applied rates, on the otherhand, reflect the tariff schedule published for Thailandin 1995 and Turkey in 1997, well before full imple-mentation of all tariff reductions.
The differences between bound and applied ratesreflect a conflicting set of interests of importers andgovernments. The lower tariffs provided by appliedrates in the examples shown in figure 7 may be prefer-able for importers seeking to import and sell foreigngoods. However, given the ability of governments toraise applied rates without penalty, the tariff applied ona shipment when it clears customs may not be the tariff
published in the country’s applied tariff schedule. Thisuncertainty can have a dampening effect on the level ofadditional trade one might expect to occur at the lowerapplied rate. On the other hand, from the government’sperspective, the lower applied rates give the country theability to raise tariffs quickly in order to insulate itsdomestic market from fluctuations in world prices andthus minimize harm to the national economy. Unfortu-nately, when countries utilize high bound tariffs as anumbrella under which they vary their applied tariffs,they can eliminate much of the advantage that stable,bound tariffs have over nontariff barriers and can con-tribute toward greater instability in world prices.
������*�+$����� ������������������� ����� ���������������� ��������� �
��� ��� E�� � *�������%558 *�������%55: *�������%556 *�������%559 *�������%555
*��� �� 48 %$ %4
E�@�� 46 %$ %$ %$ %4 %4
����� � 96 %8 %7 %7 %8 ''
�����0�� 7' %7 %6
������ ': %8 %7 %7 %:
���.�#��� 7% %7 %4 %4
"����� 75 %7 %%
� �� %%7 4$
� �� ��� 79 %: %:
G���!�0��� ������ :: %6 8$
3����� 74 %7 '$ '$
3������ :8 75
&����� :% %7 %%
�-��� %$% 74
� � 74 %'
����� 48 %% %$
������� �� 47 ': %5
(�� � 48 46
(�� ����>�(� �� %$$
(� ��� %%$ 47 48
(��-�� 45 78
1����� 4' %$ %4
=� �@��� 8' %8 %8 %8 %8 �I�&���#�� ��)���%E�� ������������3;&������ ����� ��� ��10** ������� ���� !� ����������������������� �� ���#����)���������.�����/����� �����0������.��#���!�1.2*)
22 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Tariff Escalation
Tariff escalation refers to the situation where tariffs arezero or low on primary products and then increase, orescalate, as the product undergoes additional process-ing. Further, when tariffs on products escalate with thestage of processing, the effective rate of protection, orthe tariff expressed as fractions of value-added afterdeducting intermediate inputs from product value, alsoincreases. Thus, tariff escalation potentially signalshigh rates of protection for value-added or processedproducts, and can inhibit international trade in thesegoods. For a few countries, however, the opposite mayoccur, with higher tariffs on bulk commodities raisingraw material costs, thus placing a country’s processedexports at a competitive disadvantage to other coun-tries, a situation known as tariff de-escalation.
The commodity breakouts presented in table 2 identifya number of primary and processed commodity stages,albeit at a somewhat aggregate level. To give someindication of the extent to which tariffs escalate in theagricultural sector, table 6 shows various processingstages for a number of commodity groupings and givesthe mean tariff by region.
A number of important points emerge from table 6.First, although there is evidence of tariff escalation ina number of commodities across both developed anddeveloping regions, there are also many regions andcommodities in which tariff escalation does not appearto be a problem. In 7 of the 13 regions, tariffs onprocessed products exceed those on the raw material inmore than half of the cited examples. Tariff escalationis most evident in the schedules of Eastern Europe andthe Middle East, followed by North America, SouthAsia, and the EU. In Eastern Europe, tariffs tend toescalate by at least 10 percentage points in all butthree processing chains. The largest example of escala-tion, however, is for sweeteners in North Africa, wherethe mean tariff increases by over 100 percentage pointsover those on sugar beets and sugarcane.
Processed products in which escalation is most pro-nounced include meats, sweeteners, and vegetable oils.Tariffs increase with processing in 10 regions withinthe meats and sweeteners sectors and in 9 regionswithin the vegetable oils sector. In some cases (meat inSouthern Africa and Other Western Europe and veg-etable oils and sweeteners in North Africa), the aver-age spread between primary and processed commoditytariffs is over 50 percentage points. Other examples ofspreads exceeding 50 percentage points include veg-etable juice in Eastern Europe and tobacco products inNorth America.
Tariffs in some processing chains do not increase andmay even decline with additional processing. Thecountries of Sub-Saharan Africa and the Caribbeantend to have uniform tariffs across all agriculturalproducts and account for the bulk of the cases whereno change occurs across the processing chain. Thehides and skins sector provides the best example oftariff de-escalation, or tariff protection declining withprocessing. In 9 of the 13 regions, the average tariff onhides and skins declines compared with the average onlive animals. Other studies of tariff escalation suggestthat tariff de-escalation is particularly common in thecase of multiple outputs (Lindland). Thus, while tariffson hides and skins are lower than those on live ani-mals, the tariff on meat, the main output in this multi-ple processing relationship, tends to be much higher. Apattern of tariff de-escalation can also result when theprocessed import is at the first stage of processing. Inthis case, tariffs on the finished product (in our exam-ple, leather goods) would then escalate. In agriculture,a pattern of tariff de-escalation might be tied to thelevel of support provided by farm programs, which, tobe effective, might require high border protection onprimary products. In some of these cases, however,products at a higher level of processing may receiveprotection in forms other than tariffs, such as highertransport costs or the ability of domestic firms to exer-cise monopoly power (Yeats).
Econom
ic Research Service/U
SDA
������������������
������������������
������/ A
ER
-796✺
23
������,�+$������������ �����������������$����������� ����
&��� �� ��� ��� � .��� �1�%8 <�����) ����� 3����� &��� .� �.� .����� *��� .��� *��*����� *����� ��� �� *����� ������ ������ ��� *���� *���� *���� ������ *�� ����� �
"�� � '8 88 9: 7: 84 %$$ 76 7$ 97 68 46 :$ %$4 :5"�� ��������� %5 8$ 9: 7$ 79 %'' :8 78 9' 68 87 87 %%6 6$
<�������� %9 7' 9: 46 $ 5$ %7 4: 84 68 47 44 %%$ 85<���-� %4 78 9: 7$ 4 9% 5 45 69 68 44 '' %%' 86=���� ������� %6 6' 65 45 %4 58 47 49 %$: 68 9% '7 %47 :'
?�#�� ���� '% 79 99 47 4$ '44 :8 74 54 68 $ 4$ %%% :5B�����>��-� � : 85 9: 46 $ '' 78 45 8$ 68 '$ '$ %$% 86;����>����@� ���� :8 66 5$ 74 87 '5% 9: :' 95 68 9' 4' %%% 67����������� 7% 88 5$ 7% 74 '9' 67 :6 89 68 77 48 %'' 67
;����!����� %$ 8' 9: 7$ '% 8% 45 :8 4: 68 '' 4$ %$9 89;������������� � %' 8' 9: 45 '% 79 75 8$ 47 68 46 '9 %%6 88;�����A���� %' 79 9: 46 46 75 :: 85 47 68 ': '9 %$6 86
=���� ���/����� %% 87 9: 7% %: %68 '9 :$ 4$ 68 4% 4% %%$ :5=���� ����������� � %' 8% 9: 49 '% %'4 76 85 48 68 74 '9 %$5 :$=���� ���A���� '8 8' 9: 45 %: ': 99 59 47 68 ': 4' %$6 :4
.���� ��� %' 78 9: 49 475 %77 75 49 44 68 ': '' %%$ 6$
.����� � %' 78 9: 49 8: 55 47 49 4' 68 ': '7 %%$ :'
.����� ��� 8$ :8 9: 45 85 9' 64 7' %74 68 '' 49 %'% :5
(� ������ � ��������� '9 :7 9: 49 %7 '9 7' 89 59 68 77 '$: %%$ 95(� �������������� %%' :: 9: 49 49 '5 :7 65 85 68 87 4' %%5 ::
%(��������� �� ��3;&������ ����� ��� ��10** ������� ���� ).�����/����� �����0������.��#���!�1.2*)
24 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Agricultural Tariff Schedules ofthe United States, European
Union, and Japan
This section presents a more detailed examination ofthe agricultural tariff schedules for three of the world’slargest importers of agricultural goods, the UnitedStates, the EU, and Japan. These schedules are amongthe most complex in the world, consisting of a mix oftariffs and TRQs, with a combination of ad valoremand non-ad valorem rates. Some of the rates vary overthe course of the year. Some are set on the basis of acomplex technical relationship, while others are acombination of ad valorem and specific rates, set up sothat either component can be binding. This sectionprovides a detailed picture, on a commodity basis, ofwhere bound tariffs in each country’s schedule remainhigh and where they are already low or zero.
Measuring the Impact of High Tariffs
Figure 8 consists of three histograms containing theproportion of each country’s tariff-lines falling in 6categories ranging from zero (duty-free) tariffs to tar-iffs greater than 100 percent. This breakout illustratesthat there are both widespread differences in the distri-bution of agricultural tariffs across the three countriesand that none of the countries’ tariff schedules are dis-
tributed symmetrically around the tariff mean. Distrib-utions such as these are described as being highlyskewed to the right, meaning that the tariffs continuemuch farther to the right of the mean than to the left.This is somewhat obscured by the fact that, in figure 8,all tariffs above 100 percent are lumped into one inter-val on the far right of each distribution. About 2 per-cent (24 tariff-lines) of the U.S. schedule consists oftariffs above 100 percent, with the highest rate equal-ing 350 percent. For the EU the figures are 8 percent(141 lines) with a high rate of over 500 percent, while11 percent (142 lines) of Japan’s schedule is made up of megatariffs, with the highest rate exceeding2,000 percent.
As shown in figure 8, the means for each of thesecountries are clearly inflated by the presence of a rela-tively small number of very high rates. As discussedpreviously, for skewed distributions, the mean alone isnot sufficient to characterize the overall level of tariffs.Medians provide a useful complement since they arerobust to outliers. In each case, the tariff medians areconsiderably lower than the tariff means. In contrast tothe median, which defines the center of the distributionin each country’s tariff schedule, only 12 percent ofJapan’s agricultural tariffs are larger than its tariffmean. Only 21 percent of U.S. tariffs are greater thanthe mean, while in the EU’s schedule only 28 percentof all tariffs exceed the mean.
Relative frequency distributions of agricultural tariffs for the United States, EU, and JapanFigure 8
Percent
mean11.8
mean30.1
mean58.5
median2.7
median12.8
median10.0
0
10
20
30
40
50
United States EU Japan
Duty-free>0-10>10-20>20-30>30-100>100
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 25
The mean, median, and frequency of distribution givea more complete picture of each country’s agriculturaltariff schedule as well as additional information usefulwhen comparing tariff schedules across countries.Judging from these measures, the overall level of tariffprotection in the EU and Japan is considerably higherthan in the United States. But the picture is not as
clear when comparing the EU and Japan, since the rel-ative size of their tariff means and medians differs,with the EU having a lower mean but a higher median.What is clear, though, is the extent to which eachcountry’s tariff mean is inflated by the presence ofmegatariffs in each schedule.
������-������������������������������������ ��������.������/���� ������(.������0�����1 �����.���� �1 F�
3� 3��� 3�������� 3� 3��� 3�������� 3� 3��� 3��������
*�������������� %' 4 '7 4$ %4 %7% 89 %$ %7'
"�� � ' % �� 84 :4 ' %5% 4 6"�� ��������� 9 ' �� 79 78 ' %:' '7 ':;��� %8 $ ' 76 %% 5 5 $ ��.����� % % �� '7 '$ �� %': 84 4<������� %6 $ ' $ $ �� 6' $ '<���-� ' ' �� 4 $ �� % $ ��=���� ������� 7 ' �� %4 : % %$ 5 ��;���>����� 4 ' �� %$ 4 % 7 7 ��?�#�� ���� % $ �� 4$ '' �� %$6 $ 83��/�����!�������@� ��������� % $ �� 6$ 67 '5 45 $ '3��/������ ���!����-!����������� %' % �� 7% '6 : 78 6 43��/����@� � ���!����-!����������� 5 8 �� :: 49 '7 49 5 43��/�������� ' ' �� 74 ': 6 65 '$ 6.-� ��>����� $ $ �� $ $ �� % $ ��2��� 74 49 6 96 6$ 7% 4'' ''6 79���� 5 9 �� '' '7 �� %9 '% ��;����/�;��� 7 % �� '% %' % %$ : ��;����/�;��@� 9 5 �� '$ '% �� 5 %$ ��;����/�������>����������� ���!�
�����>������ 9 7 �� 7 : �� 4 4 ��;����/������������ �� ' ' �� ' ' �� % % ��;����/��������� � : 7 �� '% '% �� %9 %6 ��;�����A���� $ $ �� 46 '' 4 '' '4 ��=���� ���/������ 6 7 �� %: %$ ' 4 4 ��=���� ���/����@� 5 9 �� %7 %8 �� %$ %$ ��=���� ���/����@� ������������
������ : 8 �� %9 %' % %%$ 5 %=���� ���/�������>������������
>��� ��� : 8 �� 49 %: �� 6 6 ��=���� ���/������ 4 ' �� ' $ �� %56 : :=���� ���/��������� � : 8 �� '% %7 ' %4 %4 ��=���� ���A����/������� �� �� �� %: %: �� ': ': ��&��� %6 4 4 8 7 �� 9 9 ��&����>������/������!�����!�
>�������� : 7 �� %: %6 �� %4 %' ��B�����������/���#� %$ % �� 8 6 �� $ $ ��B�����������/�������������>������� 7 7 �� 8 4 �� ' 4 ��.���� ��� $ $ �� 475 475 ' $ $ ��.����� � % % �� 8: 8: �� �� �� ��.����� ��� 7: 8% 8 85 86 9 9' 88 %4(� ���/�� � �������� 94 8 4 %7 %% �� $ $ ��(� ���/��������� %$' 5 % 49 47 �� 5 4 ��;� �� 4 $ �� $ $ �� '% $ ';������������ � %6 %$ �� %8 %4 �� 8' '% %$������ $ $ �� : 9 �� : : ��������/����� 8 % �� %$ %' �� 46 %6 '(��>����������� 6 4 �� ' $ �� 86 %6 '����� � ��>��������� %9 %9 �� %6 %8 �� %: %7 ��.����� % $ �� ' $ �� ' $ ������ �������� % $ �� 4 4 �� ' ' �����I� �������� ��)�% (��������� �� ��3;&������ ����� ��� ��10** ������� ���� �).�����/����� �����0������.��#���!�1.2*)
26 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Finding Megatariffs by Commodity Groupings
This section focuses on how agricultural tariffs differacross commodity sectors and identifies which productsare subject to high tariffs that might preclude profitabletrade from taking place. Figures 9-11 display the tariffmeans for the same 46 commodity groupings used in theprevious section, comparing these means with the overalltariff mean for each country.10 The individual commod-ity means exceed the country’s overall mean in onlybetween 10 (U.S.) and 14 (EU) of the 46 product cate-gories in each country. In seven of the commodity
sectors in Japan and in one each in the United States and the EU the means are greater than or equal to100 percent.
In addition to containing the means found in figures 9-11, table 7 gives the tariff medians for these commod-ity groups and identifies the extent to which megatar-iffs are being applied in each group. Large differencesbetween the mean and median tariffs indicate that afew, extremely high rates distort the mean. Megatariffsare found in between 7 (U.S.) and 17 (EU and Japan)of the 46 product categories in each country. It is interesting to note where the differences and similari-ties lie in the levels of tariff protection each countryaccords its agricultural and agri-food producers and how these compare with the overall level of tariff protection.
10 Recall that these groupings represent a subset of all the agricul-ture tariffs in the countries’ schedules. Some of the missing linesrepresent sectors where tariff equivalents cannot be calculated,e.g., alcoholic beverages.
United States averages by commodity groupFigure 9
Gra
ins
Gra
in p
rodu
cts
Feed
Sta
rche
sO
ilsee
dsO
ilcak
eV
eget
able
oils
Fats
& o
ilsLi
ve a
nim
als
Mea
t: fre
sh, o
r fro
zen
othe
r mea
t
Mea
t: fre
sh b
eef,
pork
, or p
oultr
y
Mea
t: fro
zen
beef
, por
k, o
r pou
ltry
Mea
t: pr
epar
edS
kins
& h
ides
Dai
ryE
ggs
Frui
t: fre
shFr
uit:
froze
nFr
uit:
drie
d (r
aisi
ns)
Frui
t: pr
epar
atio
nsFr
uit:
juic
eV
eget
able
s: fr
esh
Veg
etab
le ju
ice:
tom
ato
Nut
sH
ortic
ultu
re: l
ive
Sug
ar b
eet
Sug
ar c
ane
Sw
eete
ners
Toba
cco:
pro
duct
sFi
ber
Food
pre
para
tions
Cof
fee:
oth
erTe
a &
tea
extra
cts
Coc
oa b
eans
& p
rodu
cts
Spi
ces
Ess
entia
l oils
Cof
fee
Toba
cco:
unm
anuf
actu
red
Hor
ticul
ture
: cut
flow
ers
& fo
liage
Nut
s &
frui
t: dr
ied,
fres
h, &
pre
pare
d
Veg
etab
les:
drie
dV
eget
able
s: p
repa
ratio
ns
Veg
etab
les:
froz
en
Veg
etab
les:
drie
d &
fres
h ro
ots
& tu
bers
Veg
etab
les:
froz
en o
r pre
pare
d (o
ther
)
Frui
t: dr
ied
& fr
esh
(coc
onut
s, d
ates
& fi
gs)
Percent
0
10
20
30
40
50
60
70
80
90
100
110
U.S. average agricultural tariff (12 percent)
1
Tariffs are bound MFN rates based on final URAA implementation.
So rce: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 27
The highest mean tariffs in the United States are theresult of some very high duties levied on imports oftobacco products and unmanufactured tobacco. Eventhough most of the tariffs in these categories are below10 percent, the means are inflated by the presence ofseven megatariffs (all equal to 350 percent), each ofwhich is an over-quota rate in a TRQ. The mean tariffon products in the sweeteners category is also high.While it contains only two megatariffs (on glucose andfructose imports), a large proportion of the duties inthis category exceed 50 percent. All of these high tar-iffs form the over-quota rates of a TRQ. A high meanand median, as well as the largest number of megatar-iffs, are found in the dairy sector. The seven mega-tariffs in this category apply to the imports of dairy
products other than cheese or butter (including milkand cream, yogurt, and sour cream). Similar to sweet-eners, the high median indicates that most of the otherrates in the dairy group are also fairly high (over 65percent of all dairy tariffs are above 30 percent). All ofthese high dairy tariffs are the over-quota rates of aTRQ. Other commodity groups with means above theoverall average include cocoa beans and products,feeds (oilmeals, pellets, and other feeding residues),food preparations (including sauces, soups, and condi-ments), oilseeds, and tree nuts. The oilseeds categorycontains two of the highest tariffs in the U.S. schedule,on shelled and unshelled peanuts, but generally lowtariffs across all other oilseeds, and thus has a mean ofonly 17 percent. All 24 of the megatariffs in the U.S.
EU averages, by commodity groupFigure 10
Gra
ins
Gra
in p
rodu
cts
Feed
Sta
rche
sO
ilsee
dsO
ilcak
eV
eget
able
oils
Fats
& o
ilsLi
ve a
nim
als
Mea
t: fre
sh, o
r fro
zen
othe
r mea
t
Mea
t: fre
sh b
eef,
pork
, or p
oultr
y
Mea
t: fro
zen
beef
, por
k, o
r pou
ltry
Mea
t: pr
epar
edS
kins
& h
ides
Dai
ryE
ggs
Frui
t: fre
shFr
uit:
froze
nFr
uit:
drie
d (r
aisi
ns)
Frui
t: pr
epar
atio
nsFr
uit:
juic
eV
eget
able
s: fr
esh
Veg
etab
le ju
ice:
tom
ato
Nut
sH
ortic
ultu
re: l
ive
Sug
ar b
eet
Sug
ar c
ane
Sw
eete
ners
Toba
cco:
pro
duct
sFi
ber
Food
pre
para
tions
Cof
fee:
oth
erTe
a &
tea
extra
cts
Coc
oa b
eans
& p
rodu
cts
Spi
ces
Ess
entia
l oils
Cof
fee
Toba
cco:
unm
anuf
actu
red
Hor
ticul
ture
: cut
flow
ers
& fo
liage
Nut
s &
frui
t: dr
ied,
fres
h, &
pre
pare
d
Veg
etab
les:
drie
dV
eget
able
s: p
repa
ratio
ns
Veg
etab
les:
froz
en
Veg
etab
les:
drie
d &
fres
h ro
ots
& tu
bers
Veg
etab
les:
froz
en o
r pre
pare
d (o
ther
)
Frui
t: d
ried
& fr
esh
(coc
onut
s, d
ates
& fi
gs)
Percent
0
50
100
150
200
250
300
350
400
EU average agricultural tariff (30 percent)
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
28 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
schedule form the over-quota tariff in a TRQ, so somemarket access is being provided at the lower in-quota rates.
The EU’s highest tariff rates affect mainly products inthe dairy and meat sectors. Of the EU’s 141 megatar-iffs, 70 percent are found in these product categories.In the dairy sector, megatariffs are applied on almostall items with the exception of cheeses, while most ofthe meat megatariffs apply to the imports of beef,lamb/mutton, and goat meat. While the means aresomewhat inflated by the presence of these high rates,overall the tariffs in both sectors tend to be high, with78 percent of dairy tariffs and 57 percent of meat tariffs bound above 30 percent. Most of the meatmegatariffs form the over-quota portion of a tariff-rate
quota (TRQ), so there might be some possibility ofmarket access at the lower in-quota rate. However,with the exception of some TRQs for butter, the highdairy tariffs are not associated with a TRQ, thus thesetariffs would apply on all imports. Other commoditysectors with high mean tariffs include sugar beet, sugarcane, sweeteners, grains, grain products, and preparedfeeds. Most of these categories also have high mediantariffs, since a large proportion of the tariffs in thesecategories are quite high. The maximum EU tariff is540 percent, applied to imports of dried or powderedsugar beets (which contributes to the high average forsugar beets in figure 10). Some other product linesaffected by megatariffs include grape juice, preparedor preserved mushrooms, and bananas.
Japan averages, by commodity groupFigure 11
Gra
ins
Gra
in p
rodu
cts
Feed
Sta
rche
sO
ilsee
dsO
ilcak
eV
eget
able
oils
Fats
& o
ilsLi
ve a
nim
als
Mea
t: fre
sh, o
r fro
zen
othe
r mea
t
Mea
t: fre
sh b
eef,
pork
, or p
oultr
y
Mea
t: fro
zen
beef
, por
k, o
r pou
ltry
Mea
t: pr
epar
edS
kins
& h
ides
Dai
ryE
ggs
Frui
t: fre
shFr
uit:
froze
nFr
uit:
drie
d (r
aisi
ns)
Frui
t: pr
epar
atio
nsFr
uit:
juic
eV
eget
able
s: fr
esh
Veg
etab
le ju
ice:
tom
ato
Nut
sH
ortic
ultu
re: l
ive
Sug
ar b
eet
Sug
ar c
ane
Sw
eete
ners
Toba
cco:
pro
duct
sFi
ber
Food
pre
para
tions
Cof
fee:
oth
erTe
a &
tea
extra
cts
Coc
oa b
eans
& p
rodu
cts
Spi
ces
Ess
entia
l oils
Cof
fee
Toba
cco:
unm
anuf
actu
red
Hor
ticul
ture
: cut
flow
ers
& fo
liage
Nut
s &
frui
t: dr
ied,
fres
h, &
pre
pare
d
Veg
etab
les:
drie
dV
eget
able
s: p
repa
ratio
ns
Veg
etab
les:
froz
en
Veg
etab
les:
drie
d &
fres
h ro
ots
& tu
bers
Veg
etab
les:
fro
zen
or p
repa
red
(oth
er)
Frui
t: dr
ied
& fr
esh
(coc
onut
s, d
ates
& fi
gs)
Percent
0
50
100
150
200
250
300
350
Japan's average agricultural tariff (58 percent)
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 29
Among Japan’s 142 tariff-lines subject to rates inexcess of 100 percent are 49 of the 50 highest boundtariffs found within the three countries. The highestcommodity tariff mean is that for dairy products, withan average of 322 percent. Megatariffs account for 63percent of all tariff-lines in the dairy sector, with 20 ofthese rates exceeding 500 percent. The median tariff of227 indicates how high the bulk of tariffs are in thissector. As with the EU, all dairy imports, with theexception of cheese, are protected by megatariffs.Unlike the EU, however, most of these rates form theover-quota tariff of a TRQ. Imports of dried legumesare also subject to TRQs with very high over-quotarates and are the reason why the mean on dried vegeta-bles of 197 is so high. The means on the grains, 191,and grain products, 162, are also very high, largely aresult of Japan’s having recently tariffied its protectionon the imports of rice and rice products. Tariffs onindividual tariff lines in these three groups include 43megatariffs, nine of which range from 710 percent to1,364 percent on various categories of rice. The pro-duction of starches is also a highly protected industryin Japan, with tariffs averaging 126 percent. The liveanimals category has a very high average tariff, but azero median tariff. Imports of certain breeds of horses,buffalo, and swine are subject to megatariffs, whileimports of all other animals in this category are per-mitted duty-free entry. A large number of megatariffsare also applied on imports of meats and sweeteners.The highest Japanese tariff, of over 2,000 percent, isapplied to imports of konnyaku (konjac) tubers, aproduct found in the other vegetables category.
Existing Low or Zero Tariffs
While high tariff rates affect several products in theUnited States, EU, and Japan, some product groupsface zero or very low tariffs. In particular, skins andhides, certain fibers (cotton, wool, flax, and hemp), awide range of horticultural products, dried fruit, cof-fee, tea, and essential oils tend to enter each countryduty-free or at a very low duty.
If low tariff rates are defined as those below 10 per-cent (single digits), then the corresponding proportionof low agricultural tariff-lines is equal to 76 percent inthe U.S. schedule, 50 percent in Japan, and 43 percentin the EU. Thirty-four of the 46 commodity groupingsin the U.S. tariff schedule have average tariffs at or
below 10 percent, while 18 in Japan and 14 in the EUfall into this category. In many cases, these low tariffsare applied to raw materials, with the correspondingprocessed products subject to higher rates. Grains andoilseeds are generally subject to lower tariff rates thantheir products in the United States and Japan; the tar-iffs on live animals are less than those on meats in theUnited States and the EU, and raw tobacco faces lowertariffs than tobacco products in all three countries.This suggests that there are a number of incidences oftariff escalation in these countries, although the evi-dence should be interpreted with caution, given theaggregate level of the analysis.
Summing Up United States, EU, and Japan Comparisons
Prohibitive tariffs block trade in many agriculturalproducts, particularly in Japan and the EU. The exis-tence of triple-digit tariffs alongside zero tariffs illus-trates the extremes that characterize the distribution.The analysis identifies product categories withmegatariffs that could block trade and highlights dif-ferences between means and medians that indicatewhere a few, highly protected products have a distort-ing effect on the average rate of protection.
Across commodity groupings, broad similarities existin the level and distribution of tariff protection withincountries. The results demonstrate that, while the tar-iffs most critical for protection of the domestic agricul-tural sector might differ somewhat by country, theygenerally are only a subset of the country’s total agri-cultural tariff schedule. Dairy and sugar products arehighly protected in all three countries, while hides andskins and fibers are almost free of protection. On theother hand, levels of protection vary greatly amongsome commodities in all three countries for variousreasons. Japan applies high tariffs on raw silk and silkcocoon imports, while they enter the United States andEU at zero or minimal duties. Because of its proximityto neighboring sugar beet producing countries, the EUapplies a high tariff on sugar beet imports, while theUnited States and Japan allow sugar beets duty-freeentry, relying instead on high transport costs to pro-vide protection to producers. The United Statesimposes its highest tariffs on tobacco and tobaccoproducts, which Japan imports duty-free and the EU atrelatively low duties.
30 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Tariffs on Commodities ofExport Interest to the
United States
In 1999, U.S. agricultural exports totalled almost $53billion, spread across more than 130 countries. Theexistence of import tariffs in these countries was oneof several factors affecting the size of this trade. Tariffsalter the relative prices of imported and domesticallyproduced goods and thus alter the volume of imports.How much greater would U.S. agricultural exports beif global agricultural tariffs were eliminated or sub-stantially reduced? This is a question not easilyanswered, as it is subject to a host of factors, includingproducer and consumer responses to price changes,market structures, and time lags in the adjustmentprocess. While the answer is beyond the scope of thisstudy, some insight can be gained by identifying thosemarkets in which U.S. agricultural exports continue toface high tariffs.
Main Agricultural Products Exported by the United States
The top 30 categories of U.S. agricultural exports areshown in table 8. For the countries reviewed in thisreport, these items earned $32.7 billion, or about 62percent of total U.S. agricultural export revenue in1999. Of these, the top 10 each accounted for at least$1 billion in revenue and include the traditional bulkcommodities: corn, soybeans, wheat, and tobacco, aswell as intermediate goods such as beef (fresh/chilledand frozen), frozen chicken cuts, and soymeal. Alsoincluded in the top 10 are two consumer-oriented cate-gories: cigarettes and miscellaneous food preparations.
The top 30 destinations for these U.S. agriculturalexports are also shown in table 8. The countries listedare a subset of the countries reviewed in this report,which accounted for 86 percent of the $32.7 billionU.S. exports attributed to these 30 categories.11 Thetop 30 countries alone accounted for $26.4 billion, or81 percent. Japan was by far the most important desti-nation for the U.S. commodities making up these 30categories, with imports of over $7.6 billion. The EU,Mexico, Korea, and Canada represented billion dollarmarkets for these commodities. In terms of both com-modities and countries, there is a high degree of con-
centration at the top. The top ten commodity group-ings account for 71 percent of the $32.7 billion subto-tal, while the top ten destinations for this trade accountfor 68 percent.
Also contained in table 8 are the top 30 markets forthe top 30 U.S. agricultural exports. In 1999, theUnited States registered exports worth $14.8 billion tothese markets. A large share of the markets for theseU.S. exports is found in Japan and the EU. The singlemost lucrative export destination for U.S. agriculture isassociated with import demand for cigarettes by Japan.Other billion dollar markets for U.S. exporters in 1999resulted from import demand for corn in Japan andsoybeans in the EU. Rounding out the top five werethe Japanese markets for soybeans and fresh andchilled beef. The sixth largest market for U.S.exporters (soybeans to Mexico) was one of eightNAFTA markets listed in table 8. U.S. exports of soy-beans, wheat, corn, sorghum, fresh and chilled beef,and cotton to Mexico and bread, pastries, etc., and miscellaneous food preparations to Canada wereamong our top 30 export markets in 1999 (for the top30 categories).
It is informative to compare the level of tariffs in thosemarkets that imported U.S. products with those thatdid not. While most U.S. exports to Mexico andCanada would have been subject to preferential, and in some cases, zero tariffs, U.S. exports to some othermarkets were constrained by very high tariffs, in some cases high enough to preclude any trade fromtaking place.
Exports Subject to Megatariffs
Figure 12 displays the mean and upper bound tariffsfacing U.S. exporters, for each of the top 30 U.S. agri-cultural exports.12 To better illustrate the means, theupper bounds have been cut off at 500 percent. Thesimple means range from 47 percent for mixed feedsto 98 percent for frozen beef. Also shown is the globaltariff mean of 62 percent. As might be expected, thesemeans are inflated by a few very high tariffs in somecountries. Note, in particular, that ten of the categories(corn, sorghum, rice, tobacco, frozen beef, frozenpotatoes, apples, wine, whiskey, and miscellaneousfood preparations) are subject to at least one tariff inexcess of 500 percent. This section focuses on thosemarkets where U.S. exports continue to face tariff
11 Of the remaining trade, two-thirds went to just four of the coun-tries not currently WTO members, and therefore not reviewed inthis report: Taiwan, China, Saudi Arabia, and Russia.
12 Consistent with previous sections, the means are simple averagesand do not include the in-quota rates of TRQ’s.
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 31
peaks, defined here as being synonymous withmegatariffs, or tariffs equal to or greater than 100 percent.
Table 9 summarizes, for each of the top U.S. exportcategories, selected characteristics of the marketswhere this trade faces megatariffs.13 In these 30 com-modity categories, 47 different countries have at leastone tariff bound at 100 percent or above. Twenty-fiveof these countries have bound their entire agriculturalschedules at rates equal to or above 100 percent. Forthe remainder, megatariffs are found in between 1
(India, Malaysia, Morocco, and Thailand) and 17(Norway) of the 30 commodity categories. Across cat-egories, the two beef groupings, frozen and fresh/chilled boneless beef, top the list, with U.S. exports of these products subject to megatariffs in 36 and 37countries, respectively.
Eleven of these markets (wine and whiskey exports toEgypt; unprocessed tobacco to Malaysia; frozen beefto Iceland, Norway, and Switzerland; milled rice toJapan; apples to Israel; and corn, sorghum, and miscel-laneous food preparations to South Korea) are subjectto at least one tariff above 500 percent. In eight ofthese cases, however, the tariff is the over-quota rate ofa TRQ, so there is some opportunity for exports at thelower in-quota rate. In most cases, the within-quotatariff is significantly below the over-quota megatariff,
13 Appendix table 3 lists these markets, the tariffs faced by U.S.agricultural exports, and the value of U.S. exports. Not includedin this list are those countries that bound tariffs at 100 precent orabove, but where available data indicated that they were applyingrates at below 100 percent.
������1������2�.3/3���������������4���� �����5������� ������������������������� ���������� ��������5��
(���4$�1).)������������������� (���4$������ ��� � (���4$���-���������� 1).)��������� ��� ���� 1).)��������� ������� ��� ��� 1).)��������
���� ���� ����
��� 7!564!5%6� F� 6!:'4!695� ��������� F� %!6%5!'':�.�� � � 7!887!58$� �1�%8 7!'54!475� ��� F� %!7':!7$8���� 4!49:!8:6� 3����� 4!4'%!979� .�� � � �1 %!$75!497���������� 4!'4%!8$7� G��� %!569!55'� .�� � � F� 698!798�;������������ !�&�. %!8'7!:$9� � � %!5:'!9:4� E���!� � �����!�E���!� � �����! ����J������ F� 6$8!8'$�
����J������ %!'6:!45$� ����� 95'!8::� .�� � � 3����� ::'!6%:�������������!����@� %!%%%!5$'� B� ��G� � :48!%:5� (� ���!�� ��������� �1 :%6!$'$�(� ���!�� ��������� %!$9'!45:� ������� �� 867!645� E���!� � �����!����@� F� 865!565�.����� %!$:5!:7'� � �� ��� 7:$!7'$� ��� G��� 867!54:�E���!� � �����!����@� %!$'7!$':� (��-�� 74$!%::� ��� 3����� 847!9:9������ 5:9!''$� ����� 7$6!66:� ��� ����� 765!%%8������������>��-� � 965!''6� .���@��� � 496!5$8� ��� F� 78'!66%�2��� ��������� :4%!649� ����� � 477!$$%� 0������������� ��) �1 7%$!$%:�0������������� ��) 86%!588� (�� � 445!77:� ��������� �1 7$9!$7:�.����� 888!4$9� =� �@��� '97!%74� .����� 3����� 498!$57�0���!������� 888!'88� 2��� �� �0��� ��� '9$!86:� *��� ��!�����J���!������� �1 449!'%9�*��� ��!����J���!������� 87$!589� ���� '7:!$%8� E���!� � �����!�����J��L� 3����� 4':!4:8�3����������!����) 7::!%%%� 3���� '4:!897� ;������������ !�&�. � � 4%7!9%:������ ��� ������#��#�� 7:%!574� *������ %6%!995� ���-!�����J������ F� 4$$!'58��� � 77$!'97� 3������ %:8!96'� �����������>��-� � G��� '5:!:$4�E���!�����!������� 466!4'4� "����� %:'!%66� ����� 3����� '98!48'�E���!������!����) 46'!7'%� &���� %:$!9$5� ��� ����� '9'!:'8����-!�����J������ 487!:9:� .� ����� %84!68%� ����� ��� ������#��#�� .���@��� � ':8!$87��������!����@� 484!87%� ���.�#��� %49!$7%� �� � �1 '8:!5$:�*�����!����� 476!:84� .����*���� %46!668� ;���� F� '8'!6'5�3 ����������� ��� 44$!74$� ���� %'%!6%8� E���!� � �����!����@� G��� '7$!9%%����-��� 4':!559� �����0�� %%7!:4'� ��� ������� �� '47!:88�.����� 4'$!$85� B� ���� %%%!$9:� .�� � � G��� ''8!'4'�"����!����� 4$9!85:� � � %$5!%9%� ��� 3����� '%7!:'8�;���� '58!4%8� 1 �����*� �������� %$9!7$7� E���!������!����) � � '%%!'4$�
.� ����� 4'!:54!5''� .� ����� ':!488!:9%� .� ����� %7!94:!$56�<���� '$!'$4!%::� <���� :!449!'7%� <���� %6!986!9'8������ *��16-�211� ����� ���,6��6��� ����� ���,6��6���
&���/�����������������������������:�������B.���#��)����������������.�����/���������������������������������������1).)�2������ �������������)
32 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
and U.S. product is being imported (see appendix table 3).
The value of U.S. exports to markets where megatar-iffs exist totalled $3.8 billion in 1999, an average ofabout $4.4 million per market. This compares with anaverage trade flow of $11.2 million per destination toall other markets in this report for these 30 commodi-ties. The difference between those markets wheresome access was offered via a TRQ versus thosewhere no TRQ was in effect was dramatic. U.S.exports to TRQ markets totalled $2.2 billion, an aver-age of $35.6 million per market. When one excludesmarkets where a TRQ exists, average U.S. exportsdrop to under $2 million per market. This suggeststhat, in those markets subject to megatariffs, TRQs areoffering some market access for U.S. imports,
although one must also keep in mind that most of theTRQs tend to be in the wealthier OECD countries.
Japan, the EU, and Korea represent the three mostimportant non-NAFTA destinations for these 30 U.S.commodities. In 1999, U.S. exports to Japan of thefour commodities (wheat, rice, fresh and chilled pork,and miscellaneous food preparations) where megatar-iffs were levied, averaged $244 million, versus averageexports of $256 million to the 26 other markets. U.S.exports to the four EU markets subject to megatariffs(frozen boneless beef, rice, mixed feeds, and residuesof starch manufacture) averaged $133 million versus$145 million to the others. Korea applies megatariffsin five of these markets (corn, sorghum, soybeans, for-age, and miscellaneous food preparation). U.S. exportsaveraged $173 million to these markets versus $45million to the other 25. For these three countries, at
Average, maximum, and minimum tariffs faced by top 30 U.S. exports Figure 12
Cor
nS
oybe
ans
Whe
atC
igar
ette
s
Toba
cco:
unp
roce
ssed
Pou
ltry
cuts
: fro
zen
Soy
mea
l
Cot
ton
Res
idua
l sta
rch
man
ufac
turin
g
Dog
& c
at fo
od
Sor
ghum
Ric
e: m
illed
Mix
ed fe
eds,
etc
.
Win
eB
eef,
shee
p, g
oat f
atB
read
, pas
try, e
tc.
Por
k: f
resh
or c
hille
dP
otat
oes:
froz
enA
pple
s: fr
esh
Whi
skie
sS
oyoi
lG
rape
s: fr
esh
Fora
ge
Man
ufac
ture
d to
bacc
o
Pep
tone
s &
der
ivat
ives
Alm
onds
: fre
sh o
r dry
she
lled
Cat
tle h
ides
& s
kins
Bee
f: bo
nele
ss, f
roze
n
Bee
f: bo
ned,
fres
h or
chi
lled
Food
pre
para
tion,
NE
S
Percent
0
100
200
300
400
500
630 754 1,203 877
Line represents range for tariff with the top beingthe maximum and bottom being the minimum.
Note: All minimums are 0.
Global average
Product average
779 756 3,000 553554 3,000
1
Tariffs are bound MFN rates based on final URAA implementation.
Source: Economic Research Service, USDA
1
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 33
least, the presence of megatariffs in a market did notresult in U.S. exports being significantly less than inmarkets where megatariffs were not being applied.There are several explanations for this situation. Inmost of the markets where megatariffs are found inthese countries, we also find TRQs being applied.With the exception of the Japanese rice TRQ, all havefairly low in-quota rates, and the minimum accessamounts in most of these markets are being filled orclose to being filled.
Another explanation has to do with the fact that theseexports are for all products within these 6-digit cate-gories. In many cases, megatariffs might be applied onsome of the sub-categories of these products whileother sub-categories are subject to zero or very lowtariffs. One example might be a low tariff on corn usedas seed, but a high tariff on corn destined for use asfood or feed. In the case of some perishable products,tariffs vary over the course of the year, with high tar-iffs when the product is in season and low ones duringthe rest of the year. The value of imports may be veryhigh during the time the tariff is low and drop to zerowhen the megatariffs are in effect. The result is that itcan be difficult to have a clear vision of the effect thathigh tariffs are having on trade, particularly if tariffsand trade are not compared at the same HS level. Onething that is evident, however, is that the wide range intariffs levied on individual commodities within a num-ber of these 6-digit commodity markets (see appendixtable 3) indicates the extent to which countries havestrategically tailored their tariff schedules to provideprotection for very specific products.
������6������2�.3/3���������������4���� ������������������������������ ��������� ������� �
��� ����� 3��������%
�������������� 4$ 47.�� � � '5 4$��� 4% 48��������� '9 '9;������������ !�&�. 4% 49E���!� � ��!�����J������ 4: 49������������!����@� ': ':(� ���!�� ��������� '9 '5.����� '6 '6E���!� � �����!����@� 46 87����� '8 '8�����������>��-� � ': ':2��� ��������� ': ':0������������� �������� '9 '9.����� '5 4%0���!������� 4% 76*��� ��!�����J���!������� ': ':3����������!����) '5 7'����� ��� ������#��#�� '8 '8�� � '9 7%E���!�����!������� '5 44E���!������!����) '9 46���-!�����J������ 4$ 48�������!����@� '9 47*�����!����� '5 7%3 ����������� ��� '5 4'���-��� 47 7%.����� 4% 44"����!����� '9 4';���� '5 4%
&���/��;����������� ��-���!�����*��� ����� ���4)%��� ���������#���C���� �� � �(0D����������� ����� � �� �!3;&��������������� ��10** ������� ��� ).�����/���� �����0������.��#���!�1.2*)
34 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
Conclusions
While this report has focused on relative relationshipsamong tariffs across countries and commodities, ineach slice of the data a story of high tariffs emerged.With the global average tariff estimated at 62 percent,it is not surprising that high tariffs characterize mostcountries’ agricultural schedules. Only in a very fewcases was a country’s agricultural tariff average closeto the industrial country 5-percent average for tariffson imports of manufactures (e.g., the average agricul-tural tariff for Australia). Given the high level of protection that high tariffs allow, current (Burfisher etal.) analysis shows that tariffs contribute the largestshare of the cost of current agricultural protection and,thus, should be a priority for the next round of tradenegotiations.
Across regions and countries, a few stand out withhigh average levels of protection. Mean agriculturaltariffs are over 100 percent in South Asia (113 percent)and the non-EU countries of Western Europe (104 per-cent). In Africa, average tariffs for the Sub-Saharanand northern regions range from 71 to 75 percent. Theaverage rate in Central America is about 54 percent,followed by Eastern Europe, where the average tariff is49 percent. Tariffs in the EU, Asia-Pacific, and SouthAmerica range between 30 and 39 percent. At 25 per-cent, North America registered the lowest regional tar-iff average. The large differences in average tariffsacross countries indicate the potential for farmers inone country to benefit from protection while reducingprices and incomes of farmers in other countries.
Across commodities, tobacco, meat, dairy, sugar, andsweetener products generally have the highest tariffs.For other commodities, high protection may exist inselective countries. For example, in some regions, suchas Asia-Pacific, Europe, and the Middle East, compar-atively high tariffs within each respective region arerecorded for several categories of prepared vegetables.For these commodities, the global profile of tariffsindicates that producers in some countries benefit fromhigh levels of protection at the expense of producers ofthose commodities in other countries.
Megatariffs contribute to, but do not explain, the highoverall tariff averages. Comparisons of means andmedians across countries and commodities uncovercases where megatariffs largely explain the high meanor average tariffs for a specific commodity. However,in many cases similar values of means and medians
indicate uniformly high tariff levels. The overall pic-ture is one of high tariffs across a large number ofregions, countries, and commodities.
As might be expected from their relationship withproducts previously protected by nontariff barriers,TRQs are associated with high tariffs and sensitivesectors. The average over-quota tariff of 128 percent isslightly more than two times the overall average. Thisis a product of the Uruguay Round tarifficationprocess, which accommodated the conversion of somebase period NTBs into very high tariff equivalents.These new tariffs were set at such high levels that noimports, other than those provided by the minimumaccess amount, are likely to enter. Both surprising andcontrary to the principle that TRQs should providemarket access is the estimated average in-quota tariffof 63 percent—slightly above the average of 62 per-cent for all other tariffs. A number of countries havebound their in-quota rates at extremely high levels,even though the process of tariffication called for min-imum access to be provided “on the basis of a tariffquota at a low or minimal rate.” While it is true thatno numerical rule defined “low or minimal,” theserates would seem to contradict the spirit of the agreement.
Both developing and developed countries have highaverage tariffs, but tariffs for developed countries showmore variation across commodities. Developed coun-tries’ high tariffs are concentrated in dairy, meats,sugar, and sweeteners while developing countries pro-vide more uniform tariffs across commodities. Themethod of providing extremely high protection variesas megatariffs in developed countries often form theover-quota tariff in a TRQ, while those in developingcountries do not. This suggests that in developed coun-tries, at least, some market access may be provided atthe generally lower in-quota tariff in those marketsaffected by megatariffs. At the same time, we foundthat many developing countries levy applied tariffs thatare considerably below the bound rates.
The role of developing countries in future WTO nego-tiations is likely to increase significantly. A major con-tribution of this study is the breadth of developingcountry coverage, a feature that has generally beenlacking in previous studies of agricultural marketaccess. While the variation in tariff protection acrossdeveloping countries is considerable, our results indi-cate that bound agricultural tariffs in developing coun-
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 35
tries are considerably higher, on average, than indeveloped countries. This is, in part, a reflection of thespecial and differential treatment provided to thesecountries, particularly the flexibility provided on ceil-ing bindings and the lower reduction commitments.
These results are important for several reasons. First,they contradict assertions that the Uruguay Round didnot provide reciprocity for developing countries;specifically, that the mutual concessions agreed to inagriculture benefitted developed countries at theexpense of developing countries. This assertion isoften accompanied by claims that Uruguay Roundmarket access concessions have damaged the agricul-tural sectors in these countries. This view also does notappear to be supported by the evidence. On the con-trary, the comparisons between bound and appliedrates, where data on applied tariffs is available, suggestthat much of the market access being provided foragricultural imports by developing countries is takingplace at rates that are well below their WTO bindings.
Second, while developing countries continue to facetariff peaks and tariff escalation in developed countrymarkets, they also face these problems in tradebetween themselves, even perhaps to a greater extent.As results for the United States, the EU, and Japanindicate, one-quarter of all tariff-lines in these coun-tries are duty-free, involving a large number of prod-ucts of export interest to developing countries. In addi-tion, the actual tariff rates these countries apply toimports from individual developing countries are oftenlower than the MFN rate would indicate, due to theexistence of the Generalized System of Preferenceswhich provides for lower rates for selected countriesand commodities, and to other concessions affordedthrough various preferential trading arrangements.
No matter how one views tariff data across countriesand commodities, high average tariffs create barriers to
markets for U.S. and other farmers. The height of theaverage tariff signals the need for large cuts to expandmarket access broadly in agriculture. In addition, thepresence of megatariffs, particularly those that formthe in-quota rate of a TRQ, points to the need toaggressively cut tariffs in some sectors if any addi-tional market access is to be provided.
Finally, our findings have uncovered a number of othermarket access issues, beyond simply the high level oftariffs, which deserve consideration. As already men-tioned, the complexity of many country tariff sched-ules makes it very difficult to compare tariffs acrosscountries and commodities. In particular, matching tar-iffs and imports is a laborious and cumbersomeprocess. If tariffs and imports were matched, it wouldbe easier to approximate ad valorem equivalents fornon-ad valorem tariffs as well as use import weights tocalculate mean tariffs. The former is especially impor-tant because of the lack of transparency associatedwith non-ad valorem tariffs. As already noted, the non-ad valorem equivalents of these tariffs tend to behigher than their ad valorem cousins. One of the rea-sons for this almost certainly derives from their lack oftransparency, which serves to hide the actual level ofprotection being provided. The difficulties associatedwith deciphering TRQs also bears mentioning. Somecountries scheduled TRQs in ways that require carefulinterpretation of each line, and in some cases TRQsappear to cover a number of overlapping tariff lines.While these are problems that confront researchersattempting to unravel the accomplishments of theUruguay Round, they must surely have also hindered negotiators attempting to assess or quantifythe extent or importance of other countries’ tariff con-cessions. As such, there would appear to be consider-able merit in establishing certain rules for imposingconsistency and transparency across tariff schedules.
36 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
References
Bureau, J.C., L. Salvatici, and L. Fulponi. “ComparingEU and US Trade Liberalisation under the UruguayAgreement on Agriculture,” European Review ofAgricultural Economics 27(3): 259-280, 2000.
Burfisher, M. et al. Options for Agricultural PolicyReform in the WTO Negotiations. EconomicResearch Service, USDA, Agricultural EconomicReport Number 797, January 2001.
Corden, W.M. Trade Policy and Economic Welfare.Oxford: Oxford University Press, 1974.
Finger, J. M. and A. Olechowski, ed. The UruguayRound: A Handbook on the Multilateral TradeNegotiations. World Bank, Washington, DC, 1987.
Hertel, T.W. and W. Martin. “Liberalizing Agricultureand Manufactures in a Millenium Round: Implica-tions for Developing Countries,” World Economy23:455-470, 2000.
Houck, J.P. Elements of Agricultural Trade Policies.New York: Macmillan Publishing Co., 1986.
Irwin, Douglas A. “Changes in U.S. Tariffs: The Roleof Import Prices and Commercial Policies.” Ameri-can Economic Review: 88(4), Sep. 1998.
Laird, S. “Multilateral Approach to Market AccessNegotiations,” WTO Staff Working Paper TPRD-98-02, May, 1998.
Lindland, J. “The Impact of the Uruguay Round onTariff Escalation in Agricultural Products,” FoodPolicy 22(6): 487-500, 1997.
OECD. Review of Tariffs Synthesis Report. TradeDirectorate, Trade Committee, TD/TC(99)7/Final,July 1999.
Wainio, J., P. Gibson and D. Whitley. “Implemen-tation of Uruguay Round Tariff Reductions.”Agricultural Outlook, Economic Research Ser-vice, USDA, November 1999.
World Trade Organization. Ad Valorem, Specific, andOther Tariffs: Background Paper by the Secretariat.WTO, AIE/S5, February 6, 1998.
Yeats, A.J. “The Escalation of Tariff Barriers,” in Fin-ger, J. M. and A. Olechowski, ed. The UruguayRound: A Handbook on the Multilateral TradeNegotiations. World Bank, Washington, DC, 1987.
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 37
In their WTO schedules, members used a variety offormats and levels of precision to specify individualtariff commitments. In most cases, members specifiedtariffs in ad valorem terms, as a simple percentage ofthe value of the imported product. However, somecountries elected to specify some or all tariffs in spe-cific or other non-ad valorem terms. Tariff FormatsConceal High Levels of Protection (page 5), presentstwo examples of common forms of non-ad valoremtariffs. Before comparing tariff protection across coun-tries and commodities, detailed tariff bindings must beconverted into a common format. Calculating ad val-orem equivalents (AVEs) of specific or other non-advalorem tariffs allows aggregation of tariffs across thewidest group of commodities and countries.
Commodity coverage in this report is based on the def-inition of agriculture as specified in Annex 1 of theWTO Agreement on Agriculture (see appendix box,Product Coverage of the WTO Agreement on Agricul-ture). This definition includes all items from chapters1-24 of the Harmonized System (HS), minus chapter 3(fish and crustaceans). Also included are selected agri-cultural products from other chapters, such as selectedchemicals, fibers, and other substances. The HS pro-vides a nomenclature for classifying internationallytraded goods. Each of the chapters listed in Appendixtable 1 is classified at a 2-digit level in the HS. Suc-cessive levels of disaggregation, found at the 4-, 6-, 8-or 10-digit levels, define products in narrower and nar-rower terms, or levels of specificity.
North America:Canada, Mexico, United States
Central America:Belize, Costa Rica, El Salvador, Guatemala, Hon-duras, Nicaragua, Panama
South America:Argentina, Bolivia, Brazil, Chile, Colombia,Ecuador, Guyana, Paraguay, Peru, Suriname,Uruguay, Venezuela
Caribbean Islands:Antigua & Barbuda, Barbados, Cuba, Dominica,Dominican Republic, Grenada, Haiti, Jamaica, SaintKitts & Nevis, Saint Lucia, Saint Vincent & TheGrenadines, Trinidad & Tobago
EU-15:European Union
Non-EU West Europe:Cyprus, Iceland, Malta, Norway, Switzerland
Eastern Europe:Czech Republic, Hungary, Poland, Romania, SlovakRepublic, Slovenia
Middle East:Bahrain, Israel, Kuwait, Qatar, Turkey, United ArabEmirates
North Africa:Egypt, Morocco, Tunisia
Sub-Saharan Africa:Angola, Benin, Burkina Faso, Burundi, Cameroon,Central African Republic, Chad, Congo, Coted’Ivoire, Djibouti, Gabon, Gambia, Ghana, Guinea,Guinea-Bissau, Kenya, Madagascar, Malawi, Mali,Mauritania, Mauritius, Mozambique, Niger, Nigeria,Rwanda, Senegal, Sierra Leone, Tanzania, Togo,Uganda, Zaire, Zambia, Zimbabwe
South Asia:Bangladesh, Bhutan, India, Pakistan, Sri Lanka
Asia-Pacific:Australia, Brunei, Fiji, Hong Kong, Indonesia, Japan,Korea, Macau, Malaysia, Maldives, New Zealand,Papua New Guinea, Philippines, Singapore, SolomonIslands, Thailand
Southern Africa:Botswana, Lesotho, Namibia, South Africa, Swazi-land
Countries in Dataset and Regional Groupings
Appendix: Technical Details of AVE Calculations
38 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
For example, at the 2-digit, or the HS chapter level, wefind the aggregate category, Meat and Edible MeatOffal (chapter 2). Chapter 2 is disaggregated into 10categories at the 4-digit level, ranging from 0201,Meat of Bovine Animals, Fresh or Chilled, to 0210,Meat and Edible Meat Offal, Salted, In Brine, Dried orSmoked; Edible Flours and Meals of Meat or MeatOffal. Within each 4-digit grouping, a further level ofdisaggregation would be at the 6-digit level, for exam-ple, 020110, Carcasses and Half-carcasses, and020120, Other Cuts with Bone-in. Up to the 6-digitlevel, tariff schedules across countries use the samecategories for breaking out successive commodity dis-aggregation. Therefore, a 6-digit commodity definitionfor a given commodity in one country that uses the HSwould correspond to the same items in another countryusing the HS. The definitions of HS commodity group-ings up to the 6-digit level are established regularly bythe World Customs Organization.
Tariff schedules of WTO members were specified atvarious levels of commodity detail. Thus, some coun-tries’ schedules have as little as a few hundred individ-ual tariff-lines, with some of these specified at the 2-or 4- digit level.1 Other countries specified tariffs at a10-digit level, which resulted in schedules containingnearly 2,000 tariff-lines. At 8-digit and higher levels ofdisaggregation, commodity definitions vary fromcountry to country, therefore specific comparisonsacross countries are increasingly difficult at that levelof detail.
When a country uses non-ad valorem tariffs, the abilityto compare levels of protection across countries andcommodities is further complicated. A recent paper bythe WTO secretariat, Ad valorem, Specific, and OtherTariffs, discusses issues raised in the calculation ofAVEs of non-ad valorem tariffs. In order to calculateAVEs, it is necessary to divide the specific tariff by animport price. Given the lack of detailed data availableon import prices at the HS level, AVEs were calculatedusing world import unit values as a proxy for country-specific import prices. The world import unit valueswere defined at the 6-digit HS level, which, as notedabove, is the most disaggregate level at which tariffnomenclatures are internationally comparable. Importunit values were calculated for available world importsfrom all sources (minus EU-intra-trade), in value and
volume terms, using global trade data from the UnitedNations Trade Data System (COMTRADE). Theimport unit values used were for the period 1995-97,the most recent period available, and were obtainedfrom the Agricultural Market Access Database. Theworld import unit values expressed, where available,the unit value in U.S. dollars for each 6-digit category,in kilograms or pieces. For countries that did notschedule their tariff bindings in U.S. dollars, a finalstep prior to calculating the AVEs was to convert theimport unit values, for each year, into national curren-cies, and then calculate average import unit values for1995-97, in national currencies.
Tariff schedules of 129 WTO members were reviewedin this report, yielding a total of about 91,000 individ-ual tariff lines. Calculations of AVEs were needed forabout 5,600 non-ad valorem tariff lines. Of this total,AVEs could not be calculated for 387 of the tariff-
1 Some developing countries’ schedules contain a single uniformtariff rate (such as 100 percent) across all commodities. In suchcases, a 6-digit tariff schedule for the country was constructedusing the uniform rate across all tariff lines.
Product Coverage of the WTO Agreement onAgriculture (HS96)HS Chapters 1 to 24 less fish and fish products, plus1--
HS Code 2905.43 (mannitol)7
HS Code 2905.44 (sorbitol)
HS Code 2905.45 (glycarol-othre than crude)
HS Heading 33.01 (essential oils)
HS Code (ex) Ex 3302.10 (preparations based on
odoriferous substances, of
a kind used in the manu-
facture of beverage)
HS Headings 35.01 to 35.05 (albuminoidal substances,
modified starches, glues)
HS Code 3809.10 (fishing agents)
HS Heading 3823 (oleochemicals)
HS Code 3824.60 (sorbitol n.e.p.)
HS Headings 41.01 to 41.03 (hides and skins)
HS Heading 43.01 (raw furskins)
HS Headings 50.01 to 50.03 (raw silk and silk waste)
HS Headings 51.01 to 51.03 (wool and animal hair)
HS Headings 52.01 to 52.03 (raw cotton, waste and
cotton carded or combed)
HS Heading 53.01 (raw flax)
HS Heading 53.02 (raw hemp)
%The product descriptions in round brackets are not necessarily exhaustive.Source: WTO Agreement on Agriculture, annex 1.
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 39
lines. The majority (198) of the tariff lines for whichAVEs were not calculated were in chapter 22 of theHS, which covers beverages, spirits, and vinegar.Duties on many items in this chapter are specified interms of the percentage of alcohol. In these cases, thetariffs could not be matched to the world import unitvalues and, thus, AVEs were not calculated. The nextlargest number (95) of the remaining items for whichAVE calculations were not possible pertained to manyof the complex tariffs scheduled by Malaysia on prod-ucts outside of chapter 22.
AVEs were calculated at the tariff-line level, whether itbe the 2-, 4-, 6-, 8-, or 10-digit level, but using importunit values at the 6-digit level. When a tariff wasscheduled at a more disaggregated level, using theworld unit import value could have led to underesti-mating the AVE for some of these tariffs while overes-timating it for others. When a tariff was scheduled atthe more aggregate 4-digit level, the price used was asimple average of all 6-digit import unit values withineach given 4-digit tariff.
Tariffs used throughout this report, including the AVEcalculations, are the final bound MFN tariffs scheduledby WTO members. The final tariff bindings reflect therate effective after phased implementation of Uruguay
Round tariff cuts. As a general rule, developed coun-tries phased in their tariff reductions during the period1995 to 2000. Developing countries began phasing intheir tariff reductions in 1995 as well; however, theyhave until 2004 to complete implementation. In caseswhere developing countries applied tariffs that wereunbound, they had the flexibility to offer ceiling bind-ings on these products. These ceiling bindings wereexempt from the reduction commitments; therefore,the final bound tariff would take effect in 1995.
Tariff averages are calculated to reflect average MFNbound tariff rates. These averages are calculated in oneof two ways, depending on whether or not the countryscheduled TRQs. If the country did not scheduleTRQs, all tariff lines in the schedule were used to cal-culate average tariffs. If the country did scheduleTRQs, the over-quota TRQ rates and all non-TRQrates were used to calculate average tariffs. An alterna-tive in the second case would have been to first com-pute tariff averages for each TRQ by simply averagingthe in-quota and over-quota rates. The in-quota ratesare not included in this report’s tariff averages since,as some have argued, using the over-quota rate alone ismore appropriate, because it represents the marginal,binding constraint on additional trade (Laird).
40✺
������������������
������������������
������/ A
ER
-796E
conomic R
esearch Service/USD
A
+������4���������+$�������$��!7���������������� ����������������&/��������.����� *�� &��� 3����� ����� &� ��1 �1� .��� �� ��� ��� � &���*���� ������ *���� ��� ������ ����������� %8 *����� *����� ��� �� *�����
$% ?�#�� ���� �� 4' %'% 9: 7: 7%: 68 �� �� %97 '%$$' 3��� ����� ����������� 64 :' %89 %:5 8$ 468 95 %%' 59 %97 %:7$7 2�����������M� ���������M� ������ ��M�
��� ��������������� ��������� !� ���� 68 '67 %': %:' 84 '6: 67 %$6 66 %7$ %'%$8 ������������ ��������� !� ���� �� '' �� �� %5 99 �� �� �� �� ��$: ?�#�������� ��������� ��M� �� �!������� ����
��-�M������������� ���� �� ��������� �� %9 �� �� 48 '%6 �� �� �� �� ��$6 ��� ���#���� ���� ������� ������� ���� ��� 48 467 :8 79 '$% 8: %69 68 %7' %98$9 ��� ��������� �� ���M����������������������������� � 7' '77 :$ 5' 46 94 7' '5 5: %76 ��$5 ������!���!����� �������� %78 %8% 47 �� �� 6'%$ ������ 45 4'% %%4 %'9 76 %'6 6' %$$ :6 �� 9$%% �������������������� ��� ������M����M�������M�� ��� M
��������� :8 744 �� �� 88 %84 88 %$8 6' %$4 %8%' <��������� ������� ����������M�������� �����
��� �!������� ��������M�� ����������������� ���� ��M������ �������� 78 798 75 �� 48 %6$ �� %'7 84 �� %79
%4 ?�M�����!����� �� �������#���� ������� ��������� �� 687 �� �� 9: �� �� �� �� �� ��%7 =���� �������� ���������M�#���� ������������ ���� �� �� �� �� �� �� �� �� �� ��%8 * �������#���� ������� ������� ����������#���
�������������������� ������M� �������#���� ������� �� %69 %%6 98 48 56 �� %$9 %%6 %89 %:$%: �������� ��������!��������������������� �!���������
��������C������ #���� ���� �� �� �� %6$ :5 454 88 �� 57 %97 '$8%6 .����� ��������� ������ ��� 9$ :% %87 �� :8 %$4 %%7 5' 97 %'' %$5%9 ����� ��������������� � �� �� �� �� '5 58 �� �� �� �� 77%5 �������� �����������!������!������������-M� -���L����� 9% '$9 �� �� 44 '44 �� %%: �� �� 77'$ �������� �����#���� ���!������!� ���������������������� �� 4: 8: %$$ %'4 48 '75 %$8 %46 9% %45 %4''% 3������ �������� ����������� � 75 '5' �� �� 7' %48 �� 5$ 58 �� 6$'' E�#�����!��������� ��#� ��� %:5 96 �� �� 86 %5% % �� �� %'4 %%6'4 0�������� ��������������������� ��������M���������
�������� 4: 75 :4 �� 4% '$8 4% %$: 7$ �� 55'7 (� ���� ��� ����������� ������ �������� 77 %!$46 �� �� 94 �� �� �� 67 �� ��44 ���� ��������� ������ ����M����������!��������������������
�������� �� �� �� �� %5 �� �� �� �� �� ��48 *� ��� ������� �� ���M����������������M������M�� @���� �� 4$6 �� �� 46 �� '7 %4' �� �� 9749 3������ �������������������� �� �� �� �� '6 �� �� �� �� �� ��7% 0������� ���-� ��������� �����-� ��� ������� �� �� �� �� '' �� �� �� �� �� ��8$ .��- �� %8: �� �� �� �� �� �� �� �� ��8% ����!��� ����������� ������M���������� � ��
��#� �� ��� �� �� �� �� %5 �� �� �� �� �� ��8' ����� :$ �� �� �� �� �� �� 55 �� �� %884 <����#���� ������������� ���M�������� � ����#� �� ����
���������� �� �� �� �� ': �� �� �� �� �� ���
���I� �������� ��).�����/���� �����0������.��#���!�1.2*)
Econom
ic Research Service/U
SDA
������������������
������������������
������/ A
ER
-796✺
41
+������4���������+$��������!7���������������� ����������������&/��������.����� *�� &��� 3����� ����� &� ��1 �1� .��� �� ��� ��� � &���*���� ������ *���� ��� ������ ����������� %8 *����� *����� ��� �� *�����
$% ?�#�� ���� �� %9 '6 44 '% %$8 %7 �� �� %97 %$' 3��� ����� ����������� '$ :$ 8: 99 '' %%$ %7 %%4 47 %97 %6$7 2�����������M� ���������M� ������ ��M
��� ��������������� ��������� !� ���� '$ '5 77 %68 46 '4' 4$ 6: '9 %45 %%$8 ������������ ��������� !� ���� �� 6 �� �� %8 $ �� �� �� �� ��$: ?�#�������� ��������� ��M� �� �!������� �����
��-�!������������� ���� �� ��������� �� 9 �� �� '9 49 �� �� �� �� ��$6 ��� ���#���� ���� ������� ������� ���� ��� '$ '7 '8 '9 %$$ 8 %'5 4' %77 8$$9 ��� ��������� �� ���M����������������������������� � '$ 48 74 %: '8 :4 : %7 44 %'8 ��$5 ������!���!����� �������� '$ 44 �� �� '$ �� �� �� �� �� 8$%$ ������ '$ %' 56 98 %8 86' %: 89 4$ �� '8%% �������������������� ��� ������M����M�������M�� ��� M�
��������� '$ %: �� �� 47 %67 7: :8 '$ %$4 4%' <��������� ������� ����������M�������� �������� �!�������
��������M�� ����������������� ���� ��M������ �������� '$ %5 %%: �� %6 %56 �� 9: 4$ �� %$%4 ?�M�����!����� �� �������#���� ������� ��������� �� '$ �� �� 78 �� �� �� �� �� ��%7 =���� �������� ���������M�#���� ������������ ���� �� �� �� �� �� �� �� �� �� ��%8 * �������#���� ������� ������� ����������#��
�������������������� ������M� �������#���� ������� '$ '4 '%8 %' '% 96 �� 99 '6 $ ':%: �������� ��������!��������������������� �!�������������
�����C������ #���� ���� �� �� %'$ 47 6% '9 �� 4: %97 4%6 .����� ��������� ������ ��� '$ '6 5' �� :: '7 : :8 7: %'' '9%9 ����� ��������������� � �� %4 �� �� 4$ '% �� �� �� �� %4%5 �������� �����������!������!������������-M� -���L����� '$ '% �� �� '4 %89 �� 7$ �� �� 9'$ �������� �����#���� ���!������!� ���������������������� �� '$ 79 74 '8 4$ 486 '9 %46 %8 %49 4'% 3������ �������� ����������� � '$ '4 �� �� 4$ 84 �� 7$ 88 �� %5'' E�#�����!��������� ��#� ��� '$ '8 �� �� 8$ 8% �� �� �� %'4 4%'4 0�������� ��������������������� ��������M���������
�������� '$ 6 :4 �� %8 4%% '$ 6% %$ �� 4'7 (� ���� ��� ����������� ������ �������� '$ 469 �� �� 69 �� �� �� 46 �� ��'5 <�� ���������� �� �� �� �� %8 �� �� �� �� �� ��44 ���� ��������� ������ ����M����������!��������������������
�������� �� �� �� �� %8 �� �� �� �� �� ��48 *� ��� ������� �� ���M����������������M������M�� @���� �� 9 �� �� 4$ : %' 5' �� �� 849 3������ �������������������� �� �� �� �� '' �� �� �� �� ��7% 0������� ���-� ��������� �����-� ��� ������� �� �� �� �� '4 �� �� �� �� ��74 ;���-� �� ��������������M�� ��������������� �� �� �� �� �� �� �� �� �� ��8$ .��- �� %$ �� �� �� �� �� �� �� ��8% ����!��� ����������� ������M���������� � ��
��#� �� ��� �� �� �� �� 4$ �� �� �� �� �� ��8' ����� '$ �� �� �� �� �� 5$ �� �� '84 <����#���� ������������� ���M�������� � ����#� �� ����
���������� �� �� �� �� %$ �� �� �� 78 �� ��
���I� �������� ��)��������.�����/���� �����0������.��#���!�1.2*)
42 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
+������4��������������2�.3/3���������������4���� ������������������������������� 8��!�"#�����$��!
)������ ��������������� �!7����������� 7������������ .3/3��4����
��� G���!�0����� %)9����4 4'9����:4$ 867!54:���� � %9$ $����%68 '!7%50�� � �� '7$ %!464&���� %:: $����474 974.���@��� � �� $����''$ 84:
.�� � � G���!�0����� 8 796 ''8!'4'.���@��� � �� %4����%'9 '!8'6���� � $ $����%68 $&���� '$6 $����'$6 $
��� F� $����'$ '8: 78'!66%����� 98 %'9 94!'':.���@��� � %46 %%4����'56 %!6%'&���� 476 476 %!:7$���� � %9$ $����%68 9:'0�� � �� 78����'7$ $
��������� ����� �� %'8 55!:98��� � 5$ %64 %!:'50�� � �� '4' %!857
;������������ !�&�. � � 8)7����6)7 '6����'68 4%7!9%:F� %'����'8 %8����4$6 %96!$48G���!�0����� '$ 687 76!:78����� �� %8����%'$ %'!$59&���� �� %����745 :!846���� � �� $����%'8 89%
E���!� � ��!�����J������ .���@��� � 48 46: :!$%:��� � 4$ %86 58����� %'$ %5$ '$.����*���� '$ %:$ %5E���� '$ %:$ $���� � 4' 749 $3������ 94 '45 $&�� � �� %:$ $&���� �� 45' $0�� � �� '99 $.�@�� � �� %:$ $
������������!����@� E� ��� %97 %97 5%8�-��� �� %$$ $
(� ���!�� ���������) 3���� 7$7 %'$4 45!%96����� �� %'8 '!68'��� � 88 %$8 $
.����� ���� � %9$ $����%68 6%'&���� %6' $����%6' $
E���!� � �����!����@� ������ �1 �� '$ %4'����%66 7!'%8&���� '5: 698 4!94$.���@��� � 6% :59 %!5%7��� � 4$ '$: 848����� %'$ %'9 %87E���� '$ %:$ $���� � 4' 966 $&�� � �� %:$ $0�� � �� 4%8 $.����*���� '$ %:$ $.�@�� � �� %:$ $(� ��� '6 %$$ $
�����������>��-� � 0�� � �� %68 $2��� ��������� ���� � %9$ '74 8740������)������ ��) ������ �1 �� �� $����%9: 7%$!$%:
&���� '%% $����'%% 4!:54���� � �� $����%68 $
.����� G���!�0����� 4 665 '%%.���@��� � %6$ %����%97 46���� � �� $����%68 $
continued--.������� �������� ������ ��)
Economic Research Service/USDA ������������������ ������������������������ / AER-796 ✺ 43
+������4��������������2�.3/3���������������4���� ������������������������������� ��������8��!�"#�����$��!
)������ ��������)������ �!7����������� 7������������ .3/3��4����
0���!������� ������ �1 �� $����4%)%4 %'4 ::!48:&���� �� %����4%9 %!%85���� � $ $����%68 '%$��� � %8 %86 8:0�� � �� %'$ $F� :79 68: 4:!%4:
*��� ��!�����J���!������ ����� �� 8%����%$' 6!4563����������!����) � � $����')' $����'$: %'7!:97
������ �1 �� �� 7����'4% 75!7$5&���� �� $����%8: 9'6���� � %9$ %68 '8
�� � ����� �� %79 ':6����� �� 4$$$ %970�� � �� 4%8 '4
E���!�����!������� ���� � 7$ %%5 $&���� %6$ )4����%6$ $0�� � �� %'$ $.���@��� � �� %'8����'65 $
E���!������!����) ����� �� 74����%%' %!9$9&���� �� 55����'98 5$'0�� � �� '$$ $
���-!�����J������ F� �� $����%$% 4$$!'58E� ��� %97 %97 4%���� � 4' 786 $&���� �� 4:4 $0�� � �� 444 $(� ��� �� %'$ $
�������!����@� ����� �� %6$ 58&���� 76: 76: $0�� � �� %'9 $
*�����/����� ����� �� 884 7!947&���� % )6����%99 '680�� � �� '$$ $.���@��� � 7)6%����9)'7 7)6����%9$ $
3 ���������� ��� (��-�� �� %4$����%7: ::!:''��� � %'$ '4$ 48!%'4����� �� %'8 $0�� � �� %'$ $
���-��� .����*���� '$ :6����%'% '!7$'����� �� %79 6''��� � %$8 ':9 :78����� �� 4$$$ %7'0�� � �� ''$ %%8� �� �� %8$ :6E���� '$ :6����%'% $&�� � �� :6����%'% $.�@�� � �� :6����%'% $
.����� ���� � 6 %$6 $&���� %6: %)4����%6: $�-��� �� %$$ $0�� � �� %:$ $.���@��� � �� 54����%99 $(�� � '$ %7: $
"����!����� ����� �� 475 $0�� � �� '$$ $.���@��� � �� :����%:' $
;���� G���!�0����� 8 %$% %9!'57���� � �� $����%68 $&���� $����'99 '8'����'99 $
*��������������' E� ��� �� %$$ '7!'%5E ����� �� '$$ 58!$8'?����� �� '$$ $
continued--.������� �������� ������ ��)
44 ✺ ������������������ ������������������������ / AER-796 Economic Research Service/USDA
+������4��������������2�.3/3���������������4���� ������������������������������� ��������8��!�"#�����$��
)������ ��������)������ �!7����������� 7������������ .3/3��4����
*��������������' ������ &����� �� %8$ %:$!9$5H�� �� �� %8$ 6!'97.�)�=� �� ��>�"�� � �� %4$ 5!6%93��� �� %'8 '5H� � �� %'8 :%3������� �� %'' 4%6( @ � �� %'$ %:!78'"� � �� %%$ '!$%5* �����>�E� ���) �� %$$ 6!88$E���@� �� %$$ 5!79$E��-� �;�� �� %$$ 5$$E��� �� �� %$$ $2��� �� �� %$$ 7!7%:"�� � �� %$$ :!867"�� �� %$$ %8!677F��� �� %$$ %$9!%98G� � �� %$$ %:!89:G���� �� %$$ ::!8:73�@� �C�� �� %$$ %$!7:5.�)�G�����>�&�#�� �� %$$ '!%:8.�)�?��� �� %$$ 7!99:(�� ����>(� ��) �� %$$ :4!$$8(��� 4!9%$!$%7
&���/�(������ ������������������ � �� �������� ���������������� ����������������� �����������������)���I�&�������� ���%(�������� ���������� �����#���C������������������ ������� ������� ���!�������������������������� ����� A��������(0D)'(����������������� �� ������������������������N��������������N������������� � �� ������'8��(<���� ������� �� �������� ���������������������������C��������� �#��%$$������ �)��(�����(<���� ���������� ���������#����� ������������#���������� �����).�����/����� �����0������.��#���!�1.2*)