Differentiated implementation and European integration: the development of EU food quality labelling

Abstract The article examines how differentiated national implementation of EU policies affects later European integration – whether it leads to renewed unified vertical integration, differentiated integration, the status quo or renationalisation. It examines the case of geographical indications (GIs), an EU labelling system for food and drink based on place of origin and processes of production. Despite features such as diverse national interests and gastronationalism, differentiated implementation has resulted in renewed unified vertical integration. The article identifies three processes: EU legislative requirements encouraging the establishment of producer groups; European Court of Justice decisions which gave priority to quality over cross-border trade; the use of free trade negotiations to alter EU rules. While the integration outcomes are those expected by neo-functionalist analyses, the processes are more institutionally-based. Differences in implementation may aid rather than hinder further integration due to institutional features of EU legislation and decision making.

fewer.A GI provides a status that enhances prices and the market for food and wine with GIs is expanding.Moreover, GIs are linked to local, regional or national identities and 'gastronationalism' .As the Prošek case illustrates, differences in implementation of EU policies can lead to diverse interests and conflicts among EU member states.
Our question here is: what effects do such differences in implementing EU policies have on further integration?This is important in understanding the evolution of the EU given that many EU policies have seen considerable differences in implementation across member states.Indeed, academic attention is now turning to differentiated implementation, as this special issue illustrates.Most studies have looked at why differences in implementation occur, especially in legal transposition.But here, we look at the next stage in the 'integration cycle' by considering the feedback effects of differential implementation of EU policies in practice on later integration.
We set out four possible effects: renewed unified vertical integration; differentiated integration; maintenance of the status quo; renationalisation of policy making.For each we identify possible processes for differentiated implementation linked to wider theoretical literatures, notably neo-functionalism, work on differentiated integration and post-functionalist and disintegration studies.We might expect differentiated implementation to increase pressures for differentiated integration, the status quo or renationalisation, because it would augment divergence of member state interests, especially for 'identity' issues that are open to politicisation.Indeed, in domains such as genetically-modified foods (GMOs) and even banking regulation, when faced with divergent state policies, structures and interests as well as politicisation, member states have gained opt outs, or else failed to engage in further integration.
Yet we argue that for GIs, the opposite has happened: differentiated implementation has created pressures for unified vertical integration.The scope of the GI scheme has expanded and become more detailed with rules applying across all member states, and the pro-GI coalition has broadened as previously sceptical member states have become supportive.Such an outcome is in line with expectations derived from neo-functionalism.But when we examine developments in detail, we identify processes that are linked to institutional features and not just cross-border trade in itself.In particular, we find that the provisions of the initial legislation created strong incentives to establish producer groups, aiding the creation of a pro-GI coalition that also included member states with many GIs and the European Commission.Then to resolve differences in the scope of GIs across member states, cases were taken to the CJEU which made rulings that furthered unified vertical integration based on protecting quality rather than removing barriers to trade.Free trade agreements also provided the pro-GI coalition with opportunities to strengthen the GI scheme within the EU.Our overall argument is that through a set of developing and cumulative institutionally-rooted feedback processes, differences in implementation contributed to further unified vertical integration.Differentiated implementation can aid rather than hinder EU integration.
The article begins with a discussion of differentiated implementation and possible integration outcomes and processes.It then turns to the methodology and sources used, and then the nature of the case.Thereafter it looks at differences in implementation and their effects on integration and analyses why these occurred.The conclusion returns to broader issues of differentiated implementation and integration.

Analysing the effects of differentiated implementation on European integration
The literature on differentiation in the EU has focussed on differentiated integration defined as a diversity of legal requirements. 1Many studies have argued that this form of integration is a response to obstacles to 'unified integration' and that the EU has become more differentiated over time (Holzinger and Schimmelfennig 2012; Leruth et al. 2019; Schimmelfennig  and Winzen 2019).Such studies have formed part of wider debates about whether EU institutions have become less able to overcome diverse member state positions due to 'politicisation' and 'politics traps' or whether there are ways out of the EU's 'polycrisis' that lead to deeper integration (for overviews, see for example, Hooghe and Marks 2019; Zeitlin et al. 2019;  Raudla and Spendzharova 2022).
More recent work on differentiation, including the current Special Issue, has turned to differentiation in implementation of EU law and policies, which the editors define as 'diversity in the existence and use of discretion during legal and practical policy implementation in the EU, both in terms of implementation processes and outcomes' (Zhelyazkova  et al. 2023).Most of the literature has focussed on legal implementation within member states, notably in terms of transposition of EU directives, for instance, examining their 'customisation' and whether member states apply different legal provisions to comply with EU requirements (e.g.Thomann 2015, 2019; Thomann and Zhelyazkova 2017).
However, the effects of differentiated policy implementation on integration are much less studied, especially going beyond legal compliance.The present article seeks to do so, in order to identify possible outcomes and processes.As policy cycle analysts underline, implementation is part of wider movements that affect other stages, even if these do not neatly follow one after another (Hill and Varone 2017; Knill and Tosun 2020).We would expect diverse implementation, over time, to affect further European integration.In terms of differentiated implementation, we go beyond legal compliance to include contrasts in how member states interpret and use EU policies in practice.In terms of effects, Raudla and  Spendzharova (2022) usefully distinguish three paths for the development of the Single Market: renationalisation, maintenance of the status quo ('resilience') and renewed integration.We develop this schema by also following Schimmelfennig et al. (2015) in distinguishing two dimensions of integration: vertical integration; differentiation.We therefore split renewed integration to offer four outcomes -renewed unified vertical integration, greater differentiated integration, the status quo, and renationalisation.For each outcome, many pathways for feedback can be envisaged (cf.Pierson 1993) and indeed the European policy making and integration literatures offer a host of possibilities depending on which theory of integration is applied (see Leuffen et al. 2022).Given space constraints, for each possible outcome, we select a strong theoretical literature that identifies possible processes and roles for key actors.
One possibility is that differentiated policy implementation leads to greater unified vertical integration -i.e. the centralisation of rules and laws at the EU level that apply to all EU member states.Neo-functionalists offer highly developed analyses of feedback mechanisms for such an outcome (see e.g.Sandholtz and Stone Sweet 1998, 2012).They suggest that as European integration proceeds, cross-border trade and firms expand.But cross-national differences in implementation increase the costs for these firms, who then press for EU harmonisation in alliance with the Commission and European Court, who treat implementation differences and gaps as obstacles to their aims of European integration and also opportunities to expand their powers (cf.Stone Sweet 2004).Eventually the expansion of cross-border firms alters the position of national governments, as the size and balance of domestic interests have altered and governments find that centralisation of governance offers them functional advantages (cf.Mathieu 2020).
A second possibility is that differentiated implementation leads to greater differentiated integration, which itself can divided into horizontal differentiation (the extent to which EU rules apply to all member state), vertical (differences in integration across policy areas) and external (EU rules applying to non-EU member states).Here we focus on the first form.Recent work argues that there is increasingly a trade-off between vertical integration and unified integration and therefore greater centralisation is accompanied by increasing differentiation.It identifies key factors that affect different types of integration (Leuffen et al. 2022;  Leruth and Lord 2015; Schimmelfennig et al. 2015; Schimmelfennig and  Winzen 2020).Differentiated implementation can affect three important factors, namely member state preferences, politicisation and member state capacities to cooperate.If it leads to greater heterogeneity of member state interests and increase the costs of further integration (institutional and policy 'misfit') differentiated integration becomes more likely (cf.Schimmelfennig and Winzen 2020: 20-46).If it contributes to politicisation, for instance, if policy makers in some member states attack contrasting forms of implementation in others, again differentiation becomes more likely.Member states bargain to overcome opposition through opt outs and other forms of differentiated integration; indeed, differentiated integration analyses sometimes explicitly link their analyses to liberal intergovernmentalism (e.g.Schimmelfennig and Winzen 2020).Differentiation can occur in the single market, notably for 'politicised issues such as regulation of GMO foods, but even elsewhere as a strategy to defend home markets for instance, in the banking union (Eckert 2022;  Howarth and Quaglia 2020; Weimer 2019).
A third possible outcome is maintenance of the status quo, which Raudla and Spendzharova (2022) term 'resilience' but admit can also be seen as statis.Many of the same factors that limit unified integration in analyses of differentiated integration can be applied, but the outcome differs because member states are unable to agree on further integration.Thus, as differences in national implementation lead to greater heterogeneity in preferences, so it becomes impossible to find a sufficient majority of member states for new legislation.But equally differentiation through further integration with opt outs is rejected, for reasons such as fears of loss of competitive advantage or national political strategies of preventing further integration.According to some 'postfunctionalist' work, politicisation increases blockages by creating a 'constraining dissensus' (Hooghe  and Marks 2009, 2019) as greater disagreements limit policy makers.
A final possibility is that 'renationalisation' or 'disintegration' occurs (cf.Leruth et al. 2019; Raudla and Spendzharova 2022; Vollaard 2014), in which powers and increased scope for discretion return to member states.Differentiated implementation can create pressures for renationalisation if it worsens features such as heterogenous preferences and politicisation.It may also mean that national actors do not see the benefits of the single market, as for instance cross-border transaction costs remain high.
The four stylised views of the effects of differentiated implementation offer not only contrasting outcomes but also different views of interests and processes.In particular, neo-functionalist analyses underline the economic gains of integration in terms of cross-border trade.They suggest that cross-border firms in coalition with the Commission and through European court decisions successfully press for reductions of national differences and that over time opposition from national governments reduces as European integration augments the gains from cross-border EU trade.In contrast, the other three outcomes arise if greater weight is given to conflicting state preferences that are not altered by integration and to 'politicisation' as a barrier to integration.Insofar as differentiated implementation increases these two factors, it reduces the scope for unified vertical integration.Supporters of such outcomes underline the power of national governments, although they differ over whether they are able to reach agreements through differentiated integration or instead face stalemate or engage in reversal of integration.

Research design and methodology
Using the case of GIs, we examine the effects of differentiated implementation, for which we follow the definition in the Special Issue that includes implementation processes and outcomes, in line with the increasing literature on EU implementation that calls for going beyond legal transposition (e.g.Treib 2014; Zhelyazkova et al. 2016).Hence we analyse national organisations and how they used the GI scheme in practice.We look both at the integration outcomes after differentiated implementation and processes that link them by creating pressures for certain forms of integration.Hence we adopt an 'effects of causes' approach in which we look at the effects of differentiated implementation rather than a complete explanation of the outcomes as many other factors can affect these (cf.George and Bennett 2005).
The dependent variable is operationalised in terms of vertical and differentiated legal integration -whether additional powers are given to the EU and whether such powers are differentiated, notably in terms of horizontal differentiation (legal exceptions to EU rules).We define 'integration' in legal terms, including greater powers for EU institutions and tighter EU rules including harmonisation, whether through new legislation or Court decisions.
We use a process tracing methodology based on rational actors, in line with 'actor-centred institutionalism' (Scharpf 1997), looking at how those actors seek to develop integration over time.We analyse outcomes and processes that led to them in the light of the four schematic views set out above, which, while they cannot offer a complete list of possible processes, do highlight the most theoretically and empirically applicable ones.In particular, we examine the different processes in terms of key actors and their coalitions and roles.We look at the timing and sequence of events and change over time.Thus, for example, we examine whether transnational firms in alliance with the Commission and Court lead integration, overcoming opposition from national governments who gradually alter their position, or whether national governments bargain with each other, agreeing differentiation to overcome diverse interests or failing to reach agreement and therefore accepting the status quo or even re-evaluating benefits of market integration to go towards renationalisation of policy.We look at whether pressures for integration arose after changes in cross-order trade or at other times.However, we are not confined to the processes that each theoretical framework offers but use the case to see whether other processes operate.The wider theoretical aim is to offer an explanation of how and why differentiated implementation affects integration outcomes and the conditions for certain effects.
Our evidence base consists of official documents by the EU and national responses to consultations, CJEU cases concerning GIs, newspaper and specialised articles and semi-structured anonymized interviews with senior policy makers in 2022 (see Online Appendices).We examined all EU official documents concerning GIs (listed in Appendix 1).In terms of legal cases, we undertook a comprehensive search of cases involving GIs before the CJEU using the CJEU database with key words (the methodology is described in Appendix 2).We interviewed both national officials and EU ones and asked a series of questions about their relationships with each other, producer groups and the role of CJEU cases.We selected interviewees who had direct experience of GI policies at both EU and national levels, taking both countries that had supported the scheme early on (e.g.France, Italy and Spain), and then others who had initially been much more resistant (e.g.Sweden and the Netherlands).We asked national policy makers about how GIs were implemented in their country, relationships with producers, relations with other member states and judicial cases as well as the role of EU institutions; questions were adapted for EU officials.The interviewees and the questions asked are set out in Appendix 3.

The case of geographical indications in the EU
The EU's GIs scheme is used as a case study.Labelling products such as wine, oil, or cheese with the name of their place of production and associated techniques used for cultivation and transformation signals a particular type of quality.As we will see, member states had very different starting points in terms of institutions for national labelling.GIs are proposed by national producers and governments, with very different usages of scheme and diverse national interests.Moreover, given that labelling food and drink is often linked to identity -be this national, regional or local-and to 'gastro-nationalism' (cf.Huysmans 2022), it offers an example of market regulation that is particularly open to politicisation.There are strong links with 'territories' and as one interviewee put it 'these products are not like others' . 2 In addition, as the GI scheme is based on a Regulation, member states have fewer opportunities for legal customisation compared to a directive, so unified vertical integration is likely to visibly reduce the scope for national choice without being hidden behind issues of transposition of EU directives.Due to all these factors, GIs offer a relatively 'hard case' within the wider class of cases of market regulation for differentiated implementation leading to renewed unified vertical integration.
In Europe, France, Italy, Spain, Greece, and Portugal have had systems for registering and protecting appellations of origin dating back to the beginning of the 20th century. 3But in other EU member states this policy barely existed and instead producers used general systems of individual trademarks and patents (Kireeva 2011).The Cassis de Dijon judgement in 1979 that products legally produced and marketed in one Member State could be marketed in others created pressures for a EU-wide GI scheme.EU policy makers argued that it was necessary to preserve the coherence of the internal market and enhance knowledge and credibility of certain products for consumers.In addition, the EU's GI scheme developed in the context of the Uruguay Trade Round, which sought to reduce protectionist measures and public intervention in agriculture, and also in the shifts in the EU Common agriculture policy moved from providing direct market support to ensuring producer viability, food quality and environmental protection.Instead of the long and costly process of obtaining a trademark, the GI scheme was expected to offer opportunities for groups of small-scale producers to use these quality labels to promote their products.
In 1992, the EU passed a Regulation for the protection of GI foods and drinks in the internal market, with a common registry and labelling system. 4GI protection fulfils three main functions.First, as a trademark, it protects the interest of producers as third parties cannot use a the same label if their product does not conform to the applicable standards.Second, it provides information to consumers about a product and help GI holders to protect their reputation as makers of a product.Finally, GIs are defended as a tool for the protection of traditional knowledge and are associated with gastronationalism (Huysmans 2022).
The EU GIs scheme has three main labels, depending on the type of product and production that cover four categories: • Protected Designations of Origin (PDO) labels products produced, processed and prepared in a specific geographical area, using the recognised know-how of local producers and ingredients from the region concerned; • Protected Geographical Indication (PGI) are foods, agricultural products, and wines with a quality, reputation, or other characteristics essentially attributable to geographical origin; • Geographical Indication (GI) labels spirits or aromatised wine whose particular quality, reputation or other characteristic is essentially attributable to geographical origin that also use the PGI label; • Traditional Speciality Guaranteed (TSG) highlights traditional aspects of production without linkage to a specific geographical area, but does not confer intellectual property rights.
The procedure to register a GI label for a product involves producers, national governments and the Commission.Producers initiate the procedure (article 48).But such producers must be 'recognised organisations' .Thus, the Regulation creates strong incentives for individual producers to maintain or create an organisation to represent their GI product.The organisation must provide information on the product, the geographical area of origin, the production process and link between these features and the product's quality, reputation or characteristics (article 7).It submits the GI application to the national authority where it is based.The member state sends the application to the Commission, which checks whether it meets the conditions of the Regulation and if so, then publishes it.Other member states or third countries can object to the proposed registration within 6 months by sending a substantiated statement to the Commission (article 10).After this, the Commission takes a final decision and publishes the registered names (article 50), with further challenges only through the CJEU.

Differentiated implementation and pressures for renewed unified vertical integration
From its creation in 1992, implementation of the GI scheme differed across member states and led to pressures for renewed unified vertical integration.We trace the development of differentiated implementation and then three processes for its effects on later integration: the establishment of GI producer organisations that formed a pro-GI coalition with their national governments and the Commission; CJEU decisions; using free trade negotiations to alter internal EU rules and procedures.These three sets of pressures began at different times and have been 'layered' on top of each other.Finally, we find extension in the interpretations of the CJEU and a broadening of the pro-GI coalition.

Differential implementation
The starting points for implementing the 1992 Regulation differed greatly among member states.The initial phases of the GI scheme saw two types of countries: those that used the scheme and took advantage of it, essentially France and Italy, but then also Spain, Greece, and Portugal; other EU countries, which made little use of the GI scheme, and thus gained less from it or even lost from it as their producers could no longer use certain labels (London Economics 2008).
As countries with a with long GI traditions, France and Italy provide high support and guidance to GI label applicants.In France, the Institut national de l'origine et de la qualité (INAO) supports national committees with specialists that assist producers in preparing their application (e.g.delimiting the geographical area, defining the characteristics of the product and economic analysis of the product's value chain) and ensures close cooperation between producers and the state. 5In Italy, the national and the regional government have central roles in regional authorities that aid applicants.They convene public assessment meetings for each product application, where representatives of all administrations and professional and trade stakeholders analyse the characteristics of the application and examine potential difficulties and problems.On the other hand, among the countries that provided moderate support are Sweden and Denmark, where the national authorities in charge of GI products manage GI registrations along with a larger remit (Veterinary and Food Administration in Denmark, the Food Agency in Sweden), offer less intensive and tailored advice, and have enjoyed many fewer staff for GIs than countries such as France and Italy, sometimes only one or two full-time equivalents 6 (Table 1).
The 1992 Regulation established that member States had to inform the Commission which of their existing legally-protected names they wished to register, resulting in dissimilar starting points in terms of numbers.Six months after the adoption of the regulation, France had registered half out the total 663 products of the EU GI scheme.Italy, Spain, Germany, Greece and Portugal registered most of the other half of the total number of GIs.In stark contrast, Belgium registered 3 GIs, Luxemburg, Sweden and the UK registered only 1 each and the Netherlands, Ireland and Denmark and Sweden had none (Figure 2).
The EU GI scheme became consolidated progressively in the following years.The total number of GIs grew steadily, from 663 in 1992, to 2150 in 2007 and then further increases.Some of this growth resulted from the enlargement of the EU in 2004 (Figure 3).
Most sales of GI products have been within domestic markets, especially for food products.Although overall data are difficult to obtain (especially for pre 2005), studies suggest that in 2005, 62% of total GI products were sold in their national market, with the figure of 83% for agricultural products and food.Nor did these percentages change much -by 2017 they were 58% and 75%.Indeed, one study found that GI producers believed that the scheme had helped increase their share of national markets, and often exports to other member states were small and remained so -the share of exports within the EU actually fell from 20.8% of the total to 19.5% (London Economics 2008; Table 2).Academic  economic studies underline that it is far from clear whether GI labels aid exports.However GIs can offer higher prices and profit margins, as well as very valuable marketing tool, offering in turn higher prices, attracting demand and differentiating the product from similar ones. 7lthough the economic size of GI products has risen, it has been concentrated in a few countries (see Table 3).In 2005, 58% of the total sales of GI products were produced in France and Italy.In 2017, this percentage was 55%.

GI producer organisations and the formation of a pro-GI coalition
GI producer organisations soon became major policy participants at national and EU levels. 8Establishing a recognised producers' organisation is a precondition for identifying and registering a new GI protected product.Their numbers have grown sharply since the early 1990s (see Table 4).However, the fastest percentage rise in their numbers occurred in the 1990s, especially given that figures from 2004 are boosted by new member states.Table 4 offers an estimate of their numbers. 9ut GIs and hence producer organisation were unevenly spread, especially in the early 1990s.This continued, but with some rebalancing, with most in France, Germany, Spain and Italy, followed by Poland, Greece and Portugal.One reason is that some member states such as Sweden had few GI producer groups and instead relied on trade marks.It is only since 2007 that other countries have seen growth in the number of GIs (Figure 4).Producers created wider European networks to demand more effective protection of PDOs and PGIs, leading to cross-national alliances. 11AREPO, oriGIn, EFOW are networks of regions and association of producers of GI products established in 2003 and 2004. 12A high proportion of their members have come from certain member states (notably France, Italy, Spain, Portugal and Greece).The networks have developed an increasingly visible profile in debates and consultations on EU food and drink policy.They have lobbied at both national and EU levels. 13rom the start, GI producers have considered monitoring and enforcement variations to be one of the most important deficiencies in the GI scheme and underlined the need to harmonise controls in member states and strengthen protection and enforcement against third parties.They have highlighted that the GI scheme does not establish requirements and processes for member states to identify unlawful practices.This, they argue, creates difficulties in ensuring the protection of GI rights when GI products are sold in a different EU member state to where the GI producers are based, as GI producers are forced to litigate according to the different national laws of the member state where the unlawful practice occurs, which is difficult and costly (AREPO 2021).
Therefore producer groups 14 have requested greater consistency in the system for obtaining, controlling and protecting GIs, including uniform and centralised processing by the European Commission, and the standardisation of procedures at the level of national institutions (in member states as well as in third countries).Food fraud, and in particular the use of the Internet for these illegal practices, is a major concern for producers (COPA-COGECA) and some have demanded the creation of a European agency dedicated to the theme of quality agricultural production (AREPO 2008).
In member states with many GIs, the numerous producer groups cooperated with and lobbied national agriculture ministries.They soon acted to seek development of the GI scheme, especially harmonisation across the EU.They were joined by other countries, both 'Southern European ones' such as Spain but also 'Northern' ones such as Germany and the UK which developed their own GIs and systems for processing GIs. 15 The pro-GI countries and producer groups often acted in cooperation with the European Commission.Their actions included court cases and using negotiations over free trade agreements.

Differentiated implementation and court cases
Differing national implementations of the GI scheme rapidly created conflicts between EU member states, as well as with the Commission, regarding the interpretation of the GI Regulation.However, national governments failed to respond with agreements to resolve these.Hence they were dealt with through court cases both directly going to the CJEU and also as national courts requested preliminary rulings from the CJEU.Often court cases saw the Commission and 'Southern' member states with many GIs opposed to others with few GIs.In its decisions, the CJEU tended to favour interpretations of the law that limited the degree of national discretion and generally benefitted the interests of GI label producers and countries, although less boldly than in recent cases (discussed below).Thus, cases arising from differential implementation led to court decisions that increased EU legal harmonisation of GIs.
This can be seen in a key issue for GI regulation -which products should be registered as GIs and which excluded as 'generic terms' .The latter are names of products that originally referred to places where the product was produced, but have become common names to refer to products with widespread characteristics.The principle behind the exclusion of generic terms from the GI register is that such products have lost their distinctiveness and so should not receive protection.
The 1992 legislation, however, does not include a list of generic names.An attempt by the Commission to establish such a list in 1996 failed as it did not receive sufficient support from national governments in the Council (Gangjee 2016).Thus, bargaining between member states did not produce further integration.In the absence of an agreement, conflicts soon arose when a product had a name associated with a particular quality and geographical provenance in one member state (which therefore sought to protect it with a GI) but not in another.It is important because the legislation establishes that once a label is GI protected through EU registration, it cannot become generic and lose its GI protection.EU member states took different sides, together with GI versus non-GI producers.
The Parmesan case and the Feta cheese cases are good examples.In the former (Case C-132/05), Italian producers complained that Germany allowed its producers to use the name 'Parmesan' while the German government argued that the word 'Parmesan' in Germany referred simply to hard cheeses that could be grated and thus should not be restricted for use.But the Court decided that the name 'Parmesan' in Europe only refers to Parmigiano Reggiano and cannot be used for any similar type of cheese.Similarly, in the Feta cheese case (Cases C-289/96, C-465/02 and C-466/02), Greece sought to register GI protection but producers of feta-type of cheese from other member states claimed that Feta was in fact a generic term, and therefore didn't qualify under the GI scheme.After an initial decision that 'feta' was generic and hence the Feta PDO should be deleted from the EU register, the Commission then ordered an examination of the uses of the term in the different EU member states to elucidate to what extent consumers linked Feta to a particular product quality and place and sought to protect it again.In 2005 Court ruled that Feta cheese was in fact not generic, and so the PDO registration was reinstated.
Other cases illustrate that the CJEU has supported the GI scheme, even explicitly accepting that it can limit cross-border trade.The Rioja wine case (C-388/95) saw Belgium, with the support of Denmark, the Netherlands, Finland and the United Kingdom, opposing Spain, which was supported by Italy, Portugal and the Commission.It concerned whether the maintenance of the quality and reputation of the Rioja GI label meant that it must be bottled in the region of production in Spain, as established by Spanish law.Belgium considered that the Spanish rules were detrimental to the EU market principle of free movement of goods and defended the right to keep the GI label while exporting the wine in bulk from Spain and bottling it elsewhere.The Spanish Government contended that its rules were justified by the specificity and quality of the product and the need to protect the Rioja controlled designation of origin.This is important because GI producers argue that they must follow strict rules to ensure the quality of the product which, however, disappear once their products are sold to other manufacturers who then benefit from referring to the GI label on their product ('made with Parma ham' , 'Parmesan' , etc.) without being subject to similar controls.
The Court concluded that the Spanish rules protecting the GI status of Rioja were justified and proportionate, even though they limited cross-border trade, as the Court gave priority to upholding quality through the GI scheme.It explicitly stated that the Spanish rule had 'an effect equivalent to quantitative restrictions on exports' (Belgium vs Spain, 2000, paragraph 42) which are usually prohibited, but 'must be regarded as being in conformity with Community law despite its restrictive effects on trade, since it constitutes a necessary and proportionate means of attaining the objective pursued [to preserve the considerable reputation of the wine bearing the designation of origin]' (Belgium vs Spain, 2000, P I-03123).The CJEU made similar rulings and explicit statements in other cases, recognising that the application of the GI may impose 'measures having equivalent effect to a quantitative restriction on exports' but are justified to preserve quality. 16hus by the mid-2000s, differences in national implementation of the 1992 GI Regulation had sparked a number of court cases.The CJEU had become increasingly involved in developing the GI scheme through its decisions.Its logic concerned protection of quality as defined in the 1992 Regulation rather than enhancing cross-border trade.

Free trade negotiations
Supporters of GIs also used opportunities arising from external trade negotiations not just to promote rules for GIs externally (i.e.external differentiated integration) but also to strengthen the GI scheme internally within the EU.Negotiating trade agreements is an exclusive EU competence but ratification is by all national parliaments.These institutional arrangements have given opportunities for the pro-GI coalition to press for new provisions for GIs.
Indeed, since the mid-1990s, ensuring the protection of GI labels has become a priority for certain member states and the Commission (Huysmans 2020).Countries such as France feared the loss of state protection of GIs and instead an international private law system based on trademarks, in which countries such as the US would benefit and pressed the Commission hard. 17Thus, for instance, Italy, France, Spain, Greece, have made the protection of their GI labels a condition for ratification, for both economic and cultural reasons.In 2018 Italy and Greece warned that they would not ratify CETA due to its insufficient protection of their respective geographical indications (cf.Huysmans 2020).For its part, the Commission has made agreements conditional on the inclusion of safeguarding measures for GIs.In fact, the EU's free trade agreements with countries such as Canada, Mexico, Chile and South Africa have all included lists of GIs that each party has agreed to protect (Huysmans 2020; O'Connor and Richardson 2012).
One set of reasons for protecting GIs related to exports outside the EU which have grown rapidly.However, these were and remain a rather small share of the total GI market (17% in 2005 and 22% in 2017 -see Table 2).Another important reason cited by the Commission was the defence of European cultural heritage, traditional methods of production and natural resources (Cernat 2019: 28-29).Such a defence involved not only exporting GIs outside the EU but also strengthening the GI scheme internally.Indeed, the development of a common external position on GIs in trade agreements offered additional pressures to make the GI scheme more centralised.One reason is that member states cannot opt out of the EU trade position on an issue by-issue basis and so differentiated integration is made less likely.
But another reason was that in response to WTO and overseas state demands, the pro-GI coalition found opportunities to modify the GI scheme to increase vertical integration.In 1999, the United States and Australia formally complained about the GI Regulation to the WTO, arguing that its application and enforcement processes discriminated against non-EU GIs.In 2005, a WTO panel concluded that the 1992 GI Regulation was inconsistent with TRIPS Agreement, as it discriminated against non-EU products (World Trade Organisation  2006).In response, the EU passed Regulation 510/2006.The 2006 Regulation established the same application procedures for EU and non-EU products and required EU member states to put in place national monitoring and enforcement structures for GIs, in line with what the EU demanded from third countries (London Economics 2008).However, it went further in response to difficulties in the existing system in terms of being slow, difficult to implement and burdensome, 18 by including a provision that member states would have to adopt a system of official controls, in line with those established for EU feed and food, animal health and animal welfare rules (Articles 10 and 11).While under the 1992 Regulation, member states only had to ensure that inspection structures were in place, the 2006 Regulation established more specific and concrete obligations, and linked the monitoring of the GI scheme to the wider initiatives in EU food and animal health policy.Hence the 2006 changes considerably tightened EU controls over member states.

Increasing pressures for further integration -bolder courts and wider coalitions
In recent years, the CJEU has become bolder in interpreting the scope of GI scheme.The legislation establishes that trademarks and products with the same or homonymous names as products already entered in the register cannot be used nor achieve GI designation.The clause has potentially far-reaching implications as it could mean that even long-standing names of non-GI products in a country are prohibited if there is GI product in another country with the same or similar name.
Legal cases have arisen due to variations in national interpretations of GI terms in different member states.Initially, the CJEU took a limited view of the Regulation's provisions.Thus, in the 2014 Tokaj decision (Case C-31/13 P), Hungary disputed the use of the name for wines made in a region in Slovakia, but the Court established that Tokaj label could be used by both Hungarian and Slovakian producers.In the Teran wine case (Case T-626/17), Slovenia initiated an action for annulment of the designation 'Teran' on the label of some wines produced in Croatia.The Court dismissed the Slovenian arguments and established that, under certain strict conditions, the Croatian producers could continue to produce wines indicating that Teran grape variety had been employed.
However, the Court has now adopted a more expansive interpretation of the GI scheme.Thus, the Champanillo case (Case C-783/19) saw Champagne producers opposing Champanillo, a Spanish tapas bar chain with five restaurants in the Catalonian region.The plaintiff argued that the term Champanillo -which means 'little champagne' in Spanish -as well as the use of two champagne glasses in the restaurant's logo, infringed Champagne's geographical protection.The CJEU ruled that registered GI names are protected not only against imitation, but also against evocation -which includes images and representations commonly associated with the protected product.Significantly, the court ruled that the protection is applicable to services, such as restaurants, and not only to food and drink.The ruling may be influential in the present dispute between the producers and governments of Italy and Croatia.For Italy, the name Prošek was too similar to Prosecco, and it argued that giving Prošek a PDO label would confuse consumers and make producers victims of counterfeiting -or 'agro-piracy' . 19Croatian producers countered that their traditional wine has always been called Prošek, and since the two products are very different, no confusion is possible.Judgement is currently awaited.
Moreover, the Court has now ruled that the GI scheme applies to exports of GI products from EU member states.In particular, in a recent (2022) decision it ruled upheld Commission action against Denmark for the latter failing to prevent or stop the use by Danish producers of the registered name 'Feta' for cheese intended for export to third countries (Case 159/20).Thus, the prohibition on EU producers on selling GI protected products applies even in non-EU markets where the GI label is not recognised.
While the CJEU has taken a broader view of the legislation, the pro-GI coalition has become wider.The Commission has encouraged support for GIs through funding and also to alterations in the Common Agricultural Policy towards such as 'sustainability' and a 'farm to fork' approach to encourage local production. 20Hence it has sought to link GIs to wider issues of protecting the environment.Equally, the European Parliament and the European Court of Auditors have supported greater harmonisation and stronger enforcement of the GI scheme.In 2010, the Parliament adopted a Resolution demanding that thorough routine protection of GIs labels to become an obligation in all Member States.The resolution argued for a better definition of the control procedures applicable at all stages of the marketing of products both before and after they are placed on the market (European Parliament 2011).This measure would introduce verification controls to detect misuse.The Court of Auditors has also argued that lack of regular checks across the EU increase the risks of disallowed practices not being found.This, they argued, weakens the objective of ensuring fair competition between GI producers, and reduces the credibility of the GI scheme (European Court of Auditors 2011).
Perhaps the most significant change in the pro-GI coalition has been the position of national actors.Until the 2010s, support for greater EU powers to harmonise the regulation GIs came from governments in member states that had a large number of GIs and GI producer groups such as France, Italy and Spain.They saw opportunities for economic growth not only from sales of products but also from tourism, as well as wider advantages in terms of promoting national or subnational cultures and identities. 21Germany became an important part of the pro-GI coalition together with several Eastern European countries, such as Hungary, Croatia and Poland. 22But over time, even member states such as Denmark and the Netherlands with fewer GIs have gradually altered their position away from opposition and had accepted the scheme by 2022. 23ember state positions have evolved as GI producer organisations have been set up and they and their national policy makers see new opportunities for the sector. 24Indeed, the number of GIs registered rose to 3287 in 2021.Although France, Italy, Spain, Greece and Portugal accounted for over 70% of all the GIs registered in 2021, numbers rose even in countries that traditionally had few, notably in Eastern Europe, which then supported the GI scheme. 25Even countries such as Denmark and the Netherlands began to have more GIs, and their agriculture and producers saw the advantages of the scheme, especially for high value products. 26As numbers of GIs have risen, so too has the size of GI markets.Thus, for instance, there was a 42% increase between 2010 and 2017 to reach an estimated sales value of 77.15 billion euros and the share of the GI/TSG products was 7.0% of the total sales value of the European food and drink sector in EU28 in 2017.France and Italy remained ahead, reaching over 26 and 16 billion respectively in 2017.But the values for all GIs in Germany and Spain were between 5 billion and 10 billion euros during the period.In Portugal, the Netherlands, and Greece, the values were between and 2 billion.Although none of the GI products in any of other 19 EU member states MSs had a value of over 1 billion euros, the expanding markets offered growth prospects.However, the markets remained for GI products remained highly domestic rather than cross-border within the EU (see Table 2).
The effects of these pressures can be seen in legislative changes and proposals that increase EU powers over the GI scheme.Regulation 1151/2012, currently in force, considerably widened EU requirements by obliging member states to establish official controls to verify that a product complies with the corresponding product specification, and to monitor of the use of registered names of products.Member states must designate a competent authority responsible for controls to detect and suppress misuse, imitation or evocation of a protected name or other practices that mislead the consumer (Articles 36 and 13) Member states must take appropriate administrative and judicial steps to prevent or stop the unlawful use of GI labels in their territories.Hence member states now have the obligation to monitor and enforce GI rights and producers no longer have the burden of taking the initiative in the case of unlawful practices (Gragnani 2013).
Since 2019, the GI scheme has been under review, and was due for revision in 2022.The European Parliament has stressed the need to reform national legislation in order to combat counterfeiting ensure the enforcement of EU food standards with effective measures such as a European force against food fraud (European Parliament 2021).The Council and the Commission have also supported increased integration and harmonisation, and linked it to environmental sustainability and rural development. 27The Programme for the French Presidency of the Council of the European Union (first semester 2022) set as one of the priorities an improved EU-wide system for the protection of GIs (French Presidency 2022).The Council has agreed to review the GI scheme and in October 2020 invited the Commission to strengthen the legislative framework on GIs.In October 2021, the governments of 17 of the 27 member states presented a common declaration welcoming 'the opportunity to strengthen the quality schemes for agricultural products, foodstuffs, wines and spirit drinks in the forthcoming legislative review' , and considering 'that controls and effective enforcement of GIs are essential to the proper implementation of European quality policy and must be strengthened' (Council of the European Union 2021).The Commission has argued that GI rules should be made more precise and its role and that of member states in the registration process should be better identified.It has claimed that expanding the GI model to other non-agrifoods could help to protect an important part of local identity, attract tourism, and contribute to job creation (European Commission 2020).However, the outbreak of war in Ukraine has stalled progress as attention has been given to food security. 28

Conclusion
Implementation of the EU's GI scheme has been highly differentiated across member states.Some have made much use of the scheme, building on previous arrangements, notably Italy, France and Spain, which have by far the highest number of GIs.In contrast, other member states, notably in Scandinavia and Eastern Europe have fewer GIs and sought to defend the scope for their producers to apply generic labels that are outside the GI scheme.
Diverse national interests and the links between food and drink and identity could lead us to expect obstacles to integration.Yet such differential implementation has led to pressures for greater vertical unified integration.We identify three processes that grew and were combined over time.Producers established new organisations at both national and EU levels and mobilised to support GIs, in alliance with their national governments and the European Commission.When agreement on new legislation was blocked, there were legal cases before the CJEU that laid down more detailed rules.International trade negotiations provided both pressures and opportunities for member states with high numbers of GIs and the Commission to press for greater EU powers and enforcement.Over time, the CJE has become bolder and member state preferences have become more favourable to extending the GI scheme.Legal decisions and new legislation increased EU powers and rules that apply across all the EU.
The outcome of greater unified integration and changes in member state preferences are in line with those expected by neo-functionalist theories which underline the role of transnational exchange and firms as well as the Court and Commission.However, not all the processes are necessarily closely linked to cross-border trade -indeed, most GI markets remained dominated by domestic sales, with exports both to other EU members states and outside the EU remaining a low proportion (albeit of an expanding market).Instead, the analysis offers a set of institutionally-based processes that developed over time and cumulatively.This increase in the number of national producer groups was stimulated by the 1992 Regulation which requires a recognised producer group for each GI.Court cases in reaction to the failure of the member states to agree greater specification of the regulation and before the GI market really expanded in the 2000s.They were based on recognition of GIs rather than cross-border trade -indeed, the CJEU upheld GIs despite recognising them as restrictions on such trade.In addition, the pro-GI coalition was able to use free trade agreements to give priority to GIs and alter the implementation of the GI regime.Its ability to do so was rooted in institutional rules under which external trade is an EU competence but agreements must be ratified by all member states, giving 'Southern' pro-GI states a powerful lever.These institutionally-based processes appear somewhat independent of cross-border trade.
GIs are only one case and the outcomes and processes found could be usefully researched in further cases.In particular, it would be useful to explore the importance and relationship of the institutionally-based processes identified above with cross-border trade.Thus, for instance, it would be valuable to seek cases where EU regulation encourages or requires producers to organise but the scope for cross-border trade is very low.Equally, a case lacking the external dimension could help analysis of the importance of free trade agreements.
Analysis of GIs presents a dynamic and institutional view of integration by suggesting that thanks to feedback processes, over time, differentiated implementation can become a force for vertical unified integration.Those processes are not just market based but also institutional.More broadly, by underlining processes linked to EU institutional arrangements, it suggests that the EU need not turn to differentiated integration or face statis or disintegration, and more broadly, can find responses to 'polycrisis' and 'politics traps' .The case of GIs suggests that at least within market making and regulation, differentiation in implementation of EU policies may lead to renewed unified vertical integration.

Figure 3 .
Figure 3. total number of Gis in the eu 1992 to 2021.source: eambrosia, european commission.

Figure 4 .
Figure 4. number of Gi protected products by member state, 1992, 2007 and 2021.note: colour online only.source: eambrosia, european commission.

Table 1 .
level of guidance available to pDo/pGi applicants.

Table 3 .
sales value of Gi products (total sales and sales in selected countries) in 2005, 2010 and 2017.

Table 4 .
agriculture and food Gi products, by number and percentage increase, in selected years.
Notes 1. E.g.Schimmelfennig and Winzen define DI as occurring 'if EU rules and policies are not legally valid in all member states-or not exclusively valid in member states' (2019: 1172).2. Interview 6. 3. E.g.France in 1905, the Appellation d'Origine Contrôlée, Italy in 1954 for diary products and in 1963 for the wine sector, or in Spain, 1932 for wine.4. Council Regulation (EEC) No 2081/92 of 14 July 1992 on the Protection of Geographical Indications and Designations of Origin for Agricultural Products and Foodstuffs. 5. Interviews 6 and 8.