1Scope and limitations for national food safety regimes within a globalisingworldPaper accepted at the Launch Conference “Learning about Risks” of the Social Contexts and Responses to Risk Network, 28-29 January 2005, CanterburyFIRST DRAFTAuthorBettina RudloffAssistant Professor at Department for Economic and Agricultural PolicyInstitute for Agricultural Policy, Market Research and Economic Sociology/ Universityof Bonn, GermanyContact AddressInstitut für Agrarpolitik, Marktforschung und Wirtschaftssoziologie,Nussalle 21D- 53115 BonnPhone: 0049 (0)228 – 73 2335E-mail: [email protected] national food safety regimes may conflict with respect to the appropriatesafety level when food is traded. The WTO defines general rules for balancing freetrade and national sovereignty of domestic policies. Key principles for food trade areharmonisation and scientific risk assessment:Harmonization restricts national sovereignty via a set of safety standards that aredeveloped by international standard-setting organisations. These standards deem toachieve an appropriate safety level and thereby can be required from imports.National sovereignty to enforce domestically higher safety levels by trade barrierscan be realised if scientific risk assessment justifies their necessity.An empirical survey on existing WTO-disputes on food issues will specify how theseprinciples are interpreted in concrete cases. Additionally, it identifies whether there isscope for non-scientific dimensions of risk - such as cultural and economic issues that may justify trade barriers in order to enforce sovereign food safety regimes.

2Structure1Introduction . 32Economic issues of links between food safety and trade. 42.1Domestically optimal food safety levels . 52.2Impact of national food policies on trade . 73 WTO’s key principles for food safety and potential scope for nationalsovereignty. 93.1Food safety measures as non-tariff barriers . 93.2WTO provisions and scope for national sovereignty. 123.2.1Dimensions of risk . 123.2.2The appropriate safety level or the acceptable risk . 163.2.3Least-trade distortion of food safety instruments . 184 Empirical survey on WTO food disputes: the actual acceptance ofsovereign national food safety policy . 225Summary and Conclusions . 256Bibliography . 27GraphsGraph 1: Diverging optimal food safety levels in different countries. 5Graph 2: Categories of food safety measures as non-tariff barriers. 11Graph 3: Process- and product standards . 14

31 IntroductionFood safety issues are increasingly at stake within the context of international trade:global food trade is raising due to the reduction of tariffs covered by the WTO’s“Agreement on Agriculture” (AoA). But this liberalisation process is affecting countrieswhich potentially have a different understanding of food safety and of what food risksto accept. This has led to supplementing the original objective of WTO of liberalisingtrade by the aim of ensuring domestic policy targets as food safety.1This two-fold objective, ruling both free trade and national sovereignty, is covered bythe “Agreement on the Implementation of Sanitary and Phytosanitary Measures”(SPS) adopted in 1994 that will be at the centre of the following analysis.2Chapter 2 starts with a brief introduction into the economic dimension of food safetyby first explaining what parameters may lead to different national safety levels.Thereafter, it scetches how different national safety levels influence trade to identifythe role of WTO. Chapter 3 analyses limitations and scope for national sovereignpolicies within the relevant WTO-agreements. This analysis takes into account (1) theconcept of non-tariff bariers and its’ application to food safety, (2) the risk dimensionsthat can build a basis for justifying trade barriers, (3) the scope within existingprovisions on the choice of safeyt levels and (4) the scope within the existingprovisions on the choice of instruments to avoid the undermining of the nationalsafety level. In Chapter 4 a survey on decided SPS disputes identifies whether thepreviously identified scope for national sovereignty has been interpreted in favour oragainst national flexibility. Finally, a summary is given and conclusions for thenational design of food polices are drawn in chapter 5.1BHAGWATI discussed as “Law of Constant Protection” whether the reduction of tariffs is automaticallycombined with an increase of other border measures (BHAGWATI 1988).2Another relevant agreement will be Agreement on Technical Barriers to Trade, adopted in 1995 aswell but with a precursor existing already since 1979 as Standard Codex that has not been adoptedby all countries that were WTO members at that time.

42 Economic issues of links between food safety and tradeMarket forces can fail to achieve food safety and therefore political action can be justified. Such market failures can be caused by information asymmetries (AKERLOF1970) and spillovers (COASE 1969): The consumer can be faced with the problem ofinformation asymmetries in terms of not being able to differentiate between differentproduct safety qualities e.g. caused by different drug residua. Thereby they cannotconsume according to their preferences on product safety. Spillovers may appearwhen producers do not bear all costs caused by drug residua. Those costs arespilled over to the society that has to handle potential health effects. In both examples the societies’ welfare cannot be maximized and the respective goods are notoffered in optimal quantity.Both reasons can justify public action to solve information asymmetries and spilloversand thereby to increase welfare.The optimal level of public goods like “food safety” is addressed by different economic theories, such as the “public choice theory” or the “theory on public goods”.These approaches analyse the optimal policies when the market fails (BUCHANAN1968).Applying market failures to international trade is a famous and well-established research area in resource and environmental economics (DASGUPTA et al 1978). Incontrast, the economic analysis of food quality and safety issues is still a young research area especially if related to trade.3The following explanation of parameters determining different optimal national foodsafety levels will be based on assumptions similar to those analysed in the researcharea of environmental policy.3One exception is CASWELL who covers this topic already since years. CASWELL 1991.

52.1 Domestically optimal food safetylevelsThe optimal quantity of the public good “food safety”4depends on different parame-ter that can be separated into those influencing production and those influencing demand on food safety as public good.The following graph shows a very simplified model for exemplary different conditionsin two countries.Graph 1: Diverging optimal food safety levels in different countries5“price”for food safetyFood safe ty demand income risk relevance, and perceptionFood safe ty supply technical/ human resources opportunity costsEquilibrium“prices”degree of riskoptimal level of food safety acceptable riskCountry 2quantity of food s afety safety levelCountry 1Based on models for optimizing the level of public goods. See for example MANKIW, G. (2002): Principles of Economics.The parameter will be described in a broad and principal to deduce the optimal levelin general:4Different quantities can be understood as different degrees of safety, i.e. higher quantities correspond with higher safety.5There exist several studies on the question whether food safety can be interpreted as public good orbeing linked to spillovers (see for an overview RUDLOFF 2003). Without analysing this principalquestion these approaches are used in order to facilitate the general problem of optimal levels.Related to the approach of spillovers the model reflects the production or demand of the privategood causing safety risks. Thereby the level of food safety is not directly addressed but the optimal quantity of the commodity causing the spillovers. This would require that the social costs andbenefits are already reflected in the demand and supply functions. If safety is explicitly expressedas public good – an approach that may be used equivalently to spillovers (MANKIW 2000) – a necessary assumption is that the demand represents the one for public goods. Hereby the damagecosts resulting from safety risks are covered. In this model the supply reflects the abatementcosts. A general assumption for both public goods and spillovers is that all relevant information isknown and quantifiable. For a more detailed analysis see OECD (1997): Process and productionmethods (PPMs): conceptual framework and considerations on use of PPM-based trade measures, OECD / GD (97)137, Paris.

6On the production side the available technical resources and the necessaryknow-how to produce the public good “food safety” determine the supply.6 Thesecosts of resources such as for labour and capital can differ due to production capacities and production technologies that require a different resource input. Furthermore,natural conditions may influence the costs for realising safety, e.g. salmonella infections are more expensive to be avoided in countries characterised by a hot climatecompared to those facing a cold climate. Demand is characterised as benefit thatconsumer gain out of consuming a certain quantity of goods.7 Different benefits canarise from different income levels as the available income influences the choice between different products such as the public good “safety” and others like education orenvironmental protection. It is often assumed that these kinds of goods can be understood as “luxury” goods for which demand increases with increasing income(MANKIV 2000). As a result developing countries may have a smaller demand for foodsafety compared to developed countries. Other essential parameters are directly related to different risk dimensions, such as the traditional risk concept covering probability and damage amount or broader concepts including risk perception. Related tothe traditional quantitative risk understanding both probability and damage amountcan differ due to natural conditions: with respect to the example of salmonella infections the probability of emerging infections may be higher in countries facing a tropical climate. The damage amount of actually emerging salmonella infections can behigher in less developed countries, as a generally weak human constitution leads toa more problematic course of disease. The resulting higher degree of risk caused byeither higher probabilities or higher damage amounts increases the benefit of reducing it.8 Another risk component influencing demand is risk perception. The perception6Economical more precisely the supply is characterised as marginal costs necessary for an additionalunit of the respective product, i.e. in this case an additional safety level. See VARIAN 2000. Theconcept of opportunity costs describes the supply as determined by waiving of alternative production options as production resources are assumed to be limited. This limitation leads to the necessary choice of what alternative to be produced and related costs are higher the higher already thelevel of production is what means that just low capaicities over for the alternatives. (HENDERSON2004.7Economic precisely the demand describes the marginal utility of an additional unit of the consumedquantity. Due to the Law of decreasing marginal utility the additional utility is reduced as higher thealready consumed level is. The problem of aggregating individual benefits to a public one is excluded here to focus on the simplified relations. See for that MANKIW 2000.8In the concept of understanding demand as damage costs increasing risks lead to increasing damage costs.

7of risks defines how damage amounts are evaluated and thereby again influencesthe damage amount. Such an evaluation can differ due to cultural and religious reasons, e.g. the evaluation of animal health risks for cattle can be higher in hinduisticsocieties than in others due the high valuation of cattle. Several studies deduce theexistence of cultural differences in risk perception.9As a consequence of the explored parameter national safety levels may differ between countries. In graph 1 the demand for food safety is assumed to be lower incountry 2 compared to country 1 resulting in a lower safety level that can be understood synonymously as lower acceptable risk. Both national safety levels are set optimally, i.e. to maximize national welfare.The deduced safety levels can be achieved by using different instruments like settingstandards in terms of requiring certain production methods, imposing taxes on methods leading to unsafe products or by paying subsidies for safe products. The instrumental aspects of handling food safety will be discussed under chapter 3.1 and 3.2.3.As far as different safety levels are not affecting trade and thereby have no negativeimpact on trading partners food policy is of no concern to the WTO.2.2 Impact of national food policies on tradeAccording to welfare theory liberal trade without any barrier increases the welfare forall trading countries due to international division of labour. Hereby countries specialise in production areas with existing advantages compared to other countries (GANDOLFO1998).Diverging national food safety levels can have consequences for trading partnerswhat defines the reason for WTO action:10As direct effects the following can be noticed:1. Defining comparative advantages. Countries with national lower safety levelsmay have cost advantages in producing the respective good. If the resultingcommodity price is relatively low compared to countries where the production is9See for different national risk approaches RENN 2003, HENSON, S., AND TRAILL, W B. (2000) Relatedto GMO see MARRIS, C.; WYNNE, B.; SIMMONS, P.; W ELDON, S. 2001 and . Regarding beef and tomato see 10A quantification of such effects in trade is tried in MASKUSEN AND W ILSON 2001, KRISSOF, BOHMANNand CASWELL 2002.

8more expensive due to a higher safetylevel,theexportopportunitiesofcountries with lower safety levels are higher. Although this represents the usualdetermining factor for international trade the competing countries are tempted tostart strategic actions to influence the comparative advantage.2. Undermining of domestic safety level. High-safety countries like country 1 inthe graph face the problem that imported food from low-safety countries may reduce their safety level targeted by safer products (JOSLING, ROBERTSANDORDEN2004, p. 35f). This is related to the mentioned problem of asymmetric informationif the different qualities of domestic and imported products cannot be differentiated by consumers who therefore cannot demand according to their preferences.The high-safety country may wish to impose instruments on imports to ensure theown domestic safety level.These trade effects reveal the potential for strategic options to improve one’s owntrade position. Such strategic incentives are well-analysed related to environmentalpolicy (HUANG AND LABY 2001):1. Dumping. Dumping is related to the strategic underestimation of the in fact optimal policy levels. In the environmental area this effect is discussed as ecodumping. A strategic lax safety policy may reduce prices and improve the exportopportunities. Several food policy parameters cannot be proofed externally, suchas culturally determined risk perceptions. This bears the risk that it hardly can bedecided whether a lax policy is reflecting the national conditions or the result of adumping strategy.2. Hidden protectionism. Strategic overestimation of national safety goals mayserve as rationale for protecting the domestic market by import barriers to avoidthe undermining effect. Another objective can be the protection of the productionsector by paying subsidies to farmers or food processors justified by higher production costs to be fulfilled. This effect is currently discussed as political issue ashidden or disguised protection in the ongoing WTO negotiation on Agriculture(ANDERSON AND POHLSEN 2000).These implications of divergent food safety levels identify the role for WTO whoseprimary objective always has been to ensure liberal trade without any hindering barrier. Within the 50 years of existence the WTO framework has been constantly developed and extended. Hereby the initial target of free trade has been supplemented

9by integrating mechanisms to ensure na- tionalsovereigntyinsomeinternalpolicies (BAGWELL AND STAIGER).3 WTO’s key principles for food safety and potential scope for national sovereignty3.1 Food safety measures as non-tariff barriersThe WTO is addressing national policies only when trading partners are negativelyaffected. Such effects can be differentiated depending on the underlying specific national instruments a country has chosen to achieve food safety: (1) Subsidies to bepaid domestically for fulfilling certain safety levels may increase export opportunitiesand thereby lead to losses for other WTO members. (2) Tariffs can be imposed in acountry in order to avoid the undermining of national strict safety level by cheaperimports fulfilling lower standards. Tariffs reduce directly market access for tradingpartners who consequently face losses. (3) Instead of tariffs other types of bordermeasures can be established aiming at enforcing the domestic safety level.Whereas the first two categories, food-related subsidies and tariffs, are covered bythe “Agreement on Agriculture” (AoA)11 the third category of other measures is covered by the “Agreement on the Implementation of Sanitary and Phytosanitary Measures (SPS).12Those measures different from tariffs are covered by the concept of non-tariff barriers(NTBs) (NTBs) (JOSLING, ROBERTS AND ORDEN 2004):The definition of GANDOLFO implies a negative trade effect caused by types of barriers different from tariffs. This trade impact justifies the action of the WTO.No explicit mentioning of NTBs can be found neither in the SPS nor in other agreements. The SPS defines its’ scope of coverage by defining as SPS measures (Annex1):“. all laws, decrees, regulations, requirements and procedures, including, interalia, end product criteria; processes and production methods; testing; inspec-11Including the exceptional role related to general subsidies handled by the Agreement on Subsidiesand Countervailing measures.12Other NTBs are covered by the Agreement on Technical Barriers to Trade, which has its origin already early in 1979.

10tion, certification and approvalprocedures,quarantinetreatmentsincluding relevant requirements associated with the transport of animals orplants, or with the materials necessary for their survival during transport; provisions on relevant statistical methods, sampling procedures and methods ofrisk assessment; and packaging and labelling requirements directly related tofood safety.”In the followi