4.1. Abstract

1This chapter proposes a model of the welfare effects of the protection in a country (Home) of a geographical indication (GI) of another country (Foreign), when said GI had previously held generic status at Home. In international trade jargon, the procedure is known as the claw-back of a GI. The setting includes two countries (called, for convenience, Home and Foreign), three varieties (GI-original goods produced in Foreign, and GI-variety and generic goods produced at Home), and a continuum of heterogeneous consumers. Two regimes – protection and no protection – are analysed in two scenarios regarding Foreign firms, perfect and oligopolistic competition. Only the equilibrium at Home is analysed. Although a loss in global welfare is always expected when fewer varieties are available in a market, results suggest that industrialized Home countries, with sophisticated consumers and higher relative costs, tend to lose less in protecting Foreign GIs than developing Home countries, where the opposite is true. When oligopolistic competition is assumed, and Foreign GIs are protected, GI firms become from differentiated from their closest competitor (i.e. generics), furthering the competitive distortion. Industrialized countries’ consumers stand to lose more from this situation than under perfect competition, while consumers in less-developed countries, which value the variety but are unwilling or unable to pay for origin become the biggest losers. Interestingly, contrary to conventional wisdom, oligopolistic competition by Foreign firms leads to less stringent competition conditions for Home GI-varieties, and does not affect generics. In effect, if, after protection, Home GI-varieties can successfully differentiate themselves from Foreign GI-original goods without the (unlawful) use of the GI label (either through the development of their own GI or through proper branding) and remain competitive, oligopolistic competition from Foreign firms is more favourable to the development of Home firms than perfect competition.

4.2. The non-misleading trade of non-original GI-labelled goods

2In 1996, based on a survey conducted over 12,800 nationals of the European Union42, the European Commission concluded that the name “Feta” evoked Greek origin among European consumers, and thus registered the name “Feta” as a protected geographical indication43 to cover Feta cheese produced in Greece. Denmark, Germany, and France contested the decision on the basis that Feta cheese had long been produced and legally marketed under the name “Feta” in their countries. The European Court of Justice (ECJ) concluded that the Commission had played down the actual situation in the Member States, and annulled the regulation on the basis that Feta had become the generic name of a particular type of cheese. After the European Commission submitted new evidence, the ECJ overturned its previous ruling in October 2005, removing the right of any non-Greek EU producers to use the name Feta for cheese.44

3The Feta case is illustrative of the interests at stake in the identification of goods through names originating in a particular locality or territory, wherein the legal protection of the name as a GI is sought on the grounds that a given quality, reputation, or other characteristic of the good is linked, precisely, to that particular geographical origin. First, the trend towards agricultural product differentiation makes it imperative for users of indications of geographical origin to add niche market value to their products by obtaining GI protection. Second, the act of protection represents a potential loss to producers that currently label their products with the said indication as either a generic term (Feta outside Greece) or a registered trademark (such as Budweiser beer in the US, or Parma in Mexico and Canada). The third constituent aspect of this process is the consumers’ right not to be misled as to the true origin of the goods they buy.

4The parallel with trademarks is informative. There are cases in which terms that began as trademarks eventually became the generic name for the general product, such as “PC”, “Rollerblades” and “Thermos”. There is also, of course, the converse case of descriptive terms eventually acquiring secondary meaning as brands, such as “All Bran” and “Holiday Inn” (Landes and Posner 2003). GIs are confronted with similar issues, usually of the first type (a GI becoming generic), as geographical names are generally not descriptive. The final determination of generic versus secondary meaning status is the problem of the courts, which adjudicate such matters on a case-by-case basis.

5Under trademarks law, generic words cannot be trademarked. In fact, “if a trademark becomes a generic name, trademark protection immediately ceases” (Landes and Posner 2003, p. 291). Although the same standard with respect to trademarks usually applies to GIs (a generic GI cannot be trademarked), the countries that are GI demandeurs would like to impede the cessation of GI protection in territories in which GIs have become generic. The standard of protection solicited by these countries is thus greater than that currently granted to trademarks.45

6In this chapter, I follow the approach adopted by Grossman and Shapiro (1988b) in their paper on foreign counterfeiting of status goods, and develop a model for the non-misleading international trade of non-original GI-labelled goods, with heterogeneous consumers. The setting includes three varieties within the same product group in two countries. I analyse the impact of the protection of GIs, and of the enforcement of such a protection on Home and Foreign firms and consumers, in the case of both perfect and oligopolistic competition.

4.3. Literature review

4.3.1. The economics of information

7A branch of the economics of information scholarship analyses and models the market failures and distortions stemming from information asymmetries between buyers and sellers, the tools available to correct these distortions, and the policy implications thereof.

8In his pioneering 1961 piece, Stigler shows the role of advertising and reputation in economizing on search costs and solving information asymmetries. He analyses the “ascertainment of market price”, stressing that since price dispersion is “ubiquitous, even for homogenous goods”, it must be a “manifestation of ignorance in the market”. Stigler notes that the gap between the information of buyers and sellers may be bridged through advertising, but concludes that “reputation is a word which denotes persistence of quality, and reputation commands a price (or exacts a penalty) because it economizes on search costs” (Stigler 1961, p. 187). In this way, reputation, due to its connotation of reliability, may be a more effective means of reducing consumer search costs.

9Nelson, relatedly, develops the concepts of search and experience goods. Similar lines of inquiry led him and, in another study, Darby and Karni, to classify goods into search, experience, and credence goods, on the basis of how consumers obtain or are conveyed information about product characteristics, which are assumed to be known to the producer (Nelson 1970 and Darby and Karni 1973). Akerlof demonstrates that information asymmetries between sellers and buyers regarding quality can cause an entire market to collapse or, alternatively contract into an adverse selection of low-quality products. He looks at the effect of consumer uncertainty with respect to quality in a market for used cars. He proves that high-quality cars would be driven out of the market by “lemons”, leading to an overall reduction in the average quality of the goods offered, the price of the goods, and the size of the market, an overall Pareto inefficient outcome (Akerlof 1970). Along this line of reasoning, Howse and Neven argue that the most favourable case for trademark protection is that “repeat purchases help alleviate problems of moral hazard and adverse selection” (Howse and Neven 2005, p. 184).

10Landes and Posner build upon Stigler’s concept of search costs to develop a formal model of the economics of trademarks. They model the effect of trademarks protection on pricing, output, and quality, and discuss the economic impact of “property rights in words” or “exclusive rights to words” (Landes and Posner 1987, p. 273, 276). Trademarks help reduce search costs for consumers and provide an incentive to increase quality (Landes and Posner 1987 and 2003). Grossman and Shapiro model the non-deceptive trade of counterfeits, unbundling the status and quality aspects of trademarked goods, in an analysis that may be adapted to the so-called “misleading requirement” regarding GI protection (Grossman and Shapiro 1988b). Crespi and Marette analyse whether generic advertising lowers the differentiation among competing brands of the same good. They find that high-quality producers do not benefit from generic promotion when the benefits of the increased demand generated by generic advertising are outweighed by the costs of lower product differentiation (Crespi and Marette 2002; Crespi 2007).

11Zago and Pick (2004) consider the welfare impact of the adoption of labelling policies for agricultural commodities. They use a model of vertical differentiation, and derive the effects of the introduction of the regulation on equilibrium and welfare. The emergence of two differentiated competitive markets leaves consumers and high-quality producers better off, while low-quality producers are worse off. They also indicate that with high costs and low-quality differences, the total welfare impact of the regulation can be negative. Moreover, in cases where high-quality producers have market power, regulation would have a negative effect on consumers.

12Moschini, Menapace, and Pick (2008) assess the economics of geographical indications (GIs) within a vertical product differentiation framework. It is assumed that certification costs are required for GIs to serve as credible (collective) quality certification devices, and that the production of high-quality products is endogenously determined. They find that GIs can support a competitive provision of quality and lead to clear welfare gains, though they fall short of delivering the (constrained) first best. The main beneficiaries are consumers. Producers may also accrue some benefit if the production of the high-quality products draws on scarce factors that they own.

4.3.2. Empirical studies

13Most empirical studies focus on disentangling the layers of value embedded in a GI good price. Rosen’s 1974 paper series, for instance, estimates so-called hedonic price functions for specialty agricultural products, differentiating implicit prices for quality, varieties, and regional as well as brand reputations.

14To provide additional examples, a path-breaking study of wine in Australia and New Zealand between 1992 and 2000, for instance, suggests that for Australia, regional reputations in general have become increasingly significant through time, while price premia based on brand reputation are also significant (Schamel and Anderson 2003). A similar study of the U.S. wine industry, after correcting for variety, regional origin, and age, demonstrates that wine price premia in relation to quality signals may negate regional effects, a fact that may also bias estimated brand values (Schamel 2003). A study for Portugal provides empirical support for the hypothesis that the region of origin matters to consumers in the wine, olive oil, and cheese sectors (Santos and Ribeiro 2005). Finally, two recent papers focus on Bordeaux wine qualities and prices (Gergaud and Ginsburgh 2008 and Ashenfelter 2008). Recent compilations of case studies on this subject include Giovanucci et al. (2009), Vandecandelaere et al. (2009), El Benni and Reviron (2009) and Reviron et al. (2009).

4.4. Model set-up: International trade and GI protection

15This section develops a model for the non-misleading international trade of non-original GI-labelled goods with heterogeneous consumers. The setting includes two countries (Home and Foreign), three varieties (GI-original, GI-variety, generic), two regimes (GI taken as generic, GI protected as IPR), and two scenarios for Foreign firms (fringe of competitors versus monopolistic competition). Only the equilibrium at Home is considered.

4.4.1. Supply side

16This model assumes the existence of two countries, Home and Foreign, and one product group with high substitutability and cross-price elasticities of goods within that group. Consistency in quality over time is ensured through repeat purchases. There are three types of goods; within each type, goods are perfect substitutes:

1. GI-original goods are high-quality goods produced by Foreign firms and effectively protected in the Foreign territory as GIs.

2. GI-variety goods are high-quality goods produced by Home firms. They closely follow the product specifications of the GI-original good.

3. Generics are the low-quality generic version of a similar product, also produced by (different) Home firms.

17For Foreign firms, two scenarios are explored. First, it is assumed that Foreign firms operate under perfect competition, with price equal to marginal cost and average cost (zero-profits). In the alternative scenario, Foreign firms are N identical firms, N fixed, that operate under oligopolistic competition with Cournot Nash-equilibrium strategic pricing. Each firm takes its rivals’ output levels as given, and chooses its own output level to maximize its profit. Foreign firms produce at a marginal cost of production assumed constant at c.

18At Home, Foreign firms face a competitive fringe of Home firms with marginal cost pricing and zero profits (elastic supply and marginal cost equal average cost). GI-variety goods are produced at a constant marginal cost of b, with b < c. Generics are produced at a marginal cost of a, with a < b < c.

4.4.2. Regimes

19At Home, the Foreign GI is protected according to the TRIPS Agreement. Two regimes are envisaged: 
Under regime (0), Home only complies with the “non-misleading requirement”, the Foreign GI is not protected as such. This implies that, following the genericity exemption (or the grandfather of the genericity clause for wines and spirits), GI-variety producers are allowed to market their products with the Foreign GI label, although no single Home firm can appropriate the GI as a trademark. In addition, it is assumed that consumers are not misled as to the true origin of the goods they buy (i.e. the consumer knows that French Feta is made in France).

20Under regime (1), Home protects the Foreign GI as such, with the highest level of protection granted (i.e. that presently accorded only to wines and spirits). Correspondingly, Home firms must stop using the GI label for their GI-variety goods. To differentiate themselves from low-quality generics, GI-variety firms incur an extra cost. The cost of the enforcement of protection is assumed to be completely financed by the State, and therefore does not impact the investment or production decisions of Foreign or Home firms.

4.4.3. Demand side

21It is assumed that the GI label has an intrinsic value to consumers in so far as it lowers their search costs in finding the particular variety they prefer. In effect, the label is assumed to reflect certain product specifications (or “standard of identity” under U.S. law46) that add value to GI-labelled goods (both GI-original and GI-variety goods) over generics.

22Home consumers also place a value on the origin of the good, as the imported “real thing” is perceived to be worth more than domestic “copies”. For domestic (Home) consumers, the “real thing”, reflects a given quality, reputation, or other characteristic of the good which is essentially attributable to its geographical origin, the terroir, and that is not necessarily replicable outside the GI region. Therefore, if Home firms are permitted to use the Foreign GI label, they achieve two things: they unbundle the variety and origin aspects of the goods, and diminish the market power of GI-original firms by competing in the same variety segment.

23There are D heterogeneous consumers indexed from 0 to D to represent their level of sophistication in consumption. Utility is assumed to be additive. This assumption simplifies the analysis and allows cross-price substitution effects to be embedded in the utility function, reflecting the capacity of the non-misleading GI labels to unbundle origin from variety, and variety from genericity.47 The more sophisticated the consumer (and the lower his index), the higher is his utility from (and thus his willingness to pay for) consumption of any good, variety, and origin. U[d] is the utility function for the consumption of any of the three product types, V[d] is the additional utility derived from consumption of the GI variety, and G[d] is the additional utility derived from consumption of the original GI product. In addition, each of these three functions is assumed to be monotonically decreasing, implying their inverse functions exist and individuals’ demand curves do not cross.

24At Home, a Foreign firm finds its market among consumers with low indexes d, while consumers with intermediate indexes will purchase Home GI-variety products and those with high indexes will opt for generics. It is further assumed that each consumer consumes one unit of the good, or none, but no more. Total sales of GI-original goods amount to X; total sales of GI-labelled goods amount to Y (sum of GI-original and GI-variety goods); and total sales of the three types amount to Z, with Z ≤ D. Total sales of GI-variety goods, therefore, are Y – X and total sales of generics are Z – Y.

25To facilitate comparisons under different regimes and scenarios, a dashed square corresponding to GI-variety sales under no protection/perfect competition (area B in panel A of Figure 4.5.a) is included in all panels of Figure 4.5.a, Figure 4.6.a and Figure 4.6.b.

4.5. Perfect competition equilibrium

4.5.1. Foreign GI taken as generic (not protected as a GI)

26The first legal regime explored is one in which Home GI-variety goods are lawfully labelled with a Foreign GI considered to have acquired generic status at Home (regime (0)). Such a regime, for example, exists under Section 5388(c) of Title 26 of the United States Code, which establishes special rules regarding the use of GIs for wines deemed semi-generic. In general, these may be used to designate wines of an origin other than that indicated by the GI if the true place of origin is disclosed, and if the wine conforms to any standards of identity for such wine. The Section goes on to list the relevant semi-generic names: Angelica, Burgundy, Claret, Chablis, Champagne, Chianti, Malaga, Marsala, Madeira, Moselle, Port, Rhine Wine or Hock, Sauterne, Haut Sauterne, Sherry, and Tokay.

27Under perfect competition, there is marginal cost pricing. In equilibrium, the marginal consumer of a generic must be indifferent between consuming the good and buying it at all; the price of a generic, therefore, must be equal to the valuation of consumption by the marginal consumer of a generic. Similarly, the indifference condition for GI-varieties is such that the premium of the generic must equal the marginal consumer valuation of variety. Similarly, the indifference condition for GI-original goods is such that the premium of the GI-variety must equal the marginal consumer valuation of origin:

28It will be shown that, unless generics are not competitive, the marginal consumer of generics is always the same, therefore Z₀* = Z₁* = Z*. The competition condition is necessary (and sufficient) for generic, GI-variety, and GI-original sales to occur. Due to additive utilities, the price of the GI-variety does not amount to the valuations of variety and genericity of the marginal consumer (b < U[Y₀*] + V[Y₀*]). The price of the generic, a, pins down the demand for the GI-variety. The slope of the demand curve for the GI-variety with positive sales of generics is lower that its slope without; it exhibits a kink at Z*, so that demand to the right of Z* is given by U + V, but demand to the left is given by a + V. For the same reason, the demand curve for GI-original goods exhibits kinks at both Y₀* and Z* with flatter slopes for lower indexes; from right to left, the segments are given by U + V + G, a + V + G, and  b + G. Panel A of Figure 4.5.a shows the demand curves for the three types, with utility functions assumed to be linear for simplicity. Total sales (price times quantity) of generics, GI-variety, and GI-original goods correspond to areas A, B and C respectively.

Figure 4.5.a: Equilibria under perfect competition

29Panel B of Figure 4.5.a shows a scenario in which generics are not competitive because the marginal cost of the GI-variety is below the level that corresponds to the kink in demand for the GI-variety (which is itself a function of the marginal cost of generics). Generics are therefore not consumed. For b_low < b, consumers are better off consuming GI-varieties than generics. A lower b also impacts negatively the sales of GI-original goods; area C in panel B is smaller than in panel A. The higher the valuation of origin G over that of the variety V, and the less differentiated the GI-variety from the generic good, the lower the impact of such competition.

4.5.2. Foreign GI protected

30Landes and Posner (2003) discuss the economic implications of a “language monopoly” over words with primary (generic, descriptive) meaning (p. 191). If the producer of a brand or a GI could appropriate the generic name of the product, he would earn rents due to the added cost of periphrasis to his rivals. A regular PC (a former brand that today identifies any computer that is not an Apple), for instance, would be described as “a programmable electronic device for storing, manipulating and retrieving data” (Landes and Posner 2003, p. 191), while a Dijon mustard could be described as a “pungent and spicy mustard with a smooth, creamy texture that comes up to the nose (qui monte au nez)”48. The appropriation of generic names reduces the number of words available to competitors to describe their products, shifting the industry supply curve up.

31On the other hand, giving generic status to a GI may increase the search costs of consumers who believe there are quality differences between the old products bearing the GI and the new ones. These costs must be balanced against the costs to those who are forbidden usage of the generic term. As Landes and Posner point out, “terminating legal protection when a trademark (or a GI) achieves generic status might be criticized as imposing a dichotomous solution to a continuous problem” (Landes and Posner 2003, p. 195), since generic status is achieved gradually. But economic efficiency suggests that protection should cease when the costs of continued protection (higher search costs, periphrasis costs, deadweight loss) exceed the benefits (minimizing consumer confusion/search costs and maximizing the incentive of firms to maintain consistent product quality).

32If the Foreign GI is protected (regime (1)), GI-variety sales of Home firms can no longer use the GI label. If such firms want to compete, they must incur a differentiation cost,49 which may be financed at the industry level (for example through the development of their own GI, hinging on the issue of replication) or at the firm level (through brand development). For simplicity, a fixed cost per GI-variety firm will be assumed, such that the marginal cost of production of GI-varieties increases to b' > b.

a) GI-variety sales effectively deterred

33In the case of exorbitantly high differentiation costs, GI-variety firms with a marginal cost b_high are not competitive; only GI-original and generics sales occur. In that case, the indifference condition for GI-original goods is such that the premium over the generic (c – a) must equal the valuation of origin and variety by the marginal consumer. The equilibrium is characterized by:

34In that case, the demand curve for GI-original sales exhibits a single kink at Z*. To the left of the kink, the slope of the demand curve faced by Foreign firms is steeper than that under no protection, as it includes the valuation of both variety and origin. Panel C of Figure 4.5.a indicates that after protection, when

35sales of GI-original products and of generics both fill the void left by GI-varieties. The maximum level of marginal cost at which GI-varieties are competitive is a function of the marginal cost of GI-original goods

36The fact that more consumers buy the original GI good still implies a loss of welfare, as these consumers are paying a high premium solely for origin, when in reality at least part of what they value is the variety. Such consumers would rather pay less for a GI-variety than more for a GI-original good, but they no longer have the option. Similarly, there is a loss of welfare for the consumers that had to switch to generics. Welfare analysis is formally derived in Section 4.7.

b) GI-variety sales still competitive

37Attempts at the replication of successful GIs are commonplace. For example, though any reference to the use of the “méthode champenoise” in the production of sparkling wines other than Champagne is now unlawful in Europe (and in many other countries following TRIPS special protection of wines and spirits), it is a well-known fact that wines all over the world use the method, most notably several European Protected Appellations of Origin, such as Crémant (France), Cava (Spain), Espumante Bairrada (Portugal) and Trento (Italy).

38If GI-variety firms remain competitive

39either because they effectively replicate the GI under a different name or because of branding, the equilibrium is the same as when the Foreign GI was not protected, but with b' > b instead of b, implying lower sales of the GI-variety. Panel D of Figure 4.5.a shows that sales of generics and GI-original goods increase in volume, while sales of GI-variety goods decrease. Foreign and GI-variety sales are determined residually. If GI-variety sales are not deterred, the demand curve for GI-original goods is not steeper, rather, it is shifted upwards, as the marginal cost that pins down the valuation for variety is higher, commanding a higher level of sales under marginal cost pricing. This is due to Foreign firms appropriating the surplus as a result of the valuation of variety by the marginal consumer of GI-variety goods.

4.6. Oligopoly equilibrium

40In this section, the same analysis is performed, this time under the assumption of oligopolistic competition by Foreign firms in the market for GI-original products. The assumption is that N identical firms compete in that sector with Cournot pricing. Competition conditions, with the exception of the competitiveness of GI-variety goods once the Foreign GI is protected, are not essential to the analysis and are therefore omitted.

4.6.1. Foreign GI taken as generic (not protected as a GI)

41Under a regime of no-protection of the Foreign GI (regime (0)), the inverse demand curves faced by Home and Foreign firms are the same as under perfect competition, as are the equilibrium quantities Z* and Y₀*, corresponding to total sales and to sales of GI-labelled goods. In this case, however, Foreign firms maximize profits by setting marginal revenue, rather than price, as equal to marginal cost, following Cournot pricing. The segment of the demand curve relevant to those purposes is b + G:

42where G' is the partial derivative of G with respect to each firm’s quantity (G' < 0). Since the demand curve faced by Foreign firms is kinked at Y₀*, the marginal revenue curve is discontinuous at that point; the segment to the left of Y₀*, is placed above the segment to its right because demand (average revenue) is flatter. The market-clearing price and premium for GI-original goods are determined by the demand function. Panel A of Figure 4.6.a illustrates the demand curves (which have not changed) and the marginal revenue curve for Foreign firms (thick dashed lines).

Figure 4.6.a: Equilibria under oligopolistic competition by Foreign firms

43Oligopolistic competition leads to less restrictive competition conditions for GI-varieties, and does not affect generics – only the closest competitor is affected. This distortion is entirely due to the fact that Foreign firms have some market power, but exercise it exclusively over the origin segment of demand. GI-varieties, which do not compete in that segment of the market, see an expansion of output.50 Consumers of GI-original goods under marginal cost pricing are worse off; they either pay a higher price for the same GI-original good, or they switch to GI-varieties as their valuation of origin is not worth the price increase under oligopolistic competition.

44Additionally, the inter-linkage between the three types of goods implies that Foreign firms have room to expand output and threaten the equilibrium for GI-varieties. That equilibrium is now determined residually from the right by the cost of generics (which pins down demand), and from the left by Foreign firms which can easily exercise their market power to reduce sales of GI-varieties. If the market power of Foreign firms were to disappear, the situation observed under perfect competition, with GI-variety sales reduced to area B, would recur.

45Panel B of Figure 4.6.a shows a scenario in which GI-varieties are not competitive because GI-original goods are low cost. This would be the effect, for example, of a productivity shock or a State subsidy. In that case, GI-varieties are out of the market with or without protection. Protection has no effect, as the upper segment of the demand curve is never reached. Sales of generics are also affected, and are lowered.

4.6.2. Foreign GI protected

a) GI-variety sales effectively deterred

46Assuming that GI-variety goods are not competitive after protection (regime (1)), but that generics are, the inverse demand curve faced by Foreign firms is the same as that in the scenarios of protection/perfect competition (steeper than before protection). This time, however, with Cournot pricing, output is defined by the first-order condition for profit maximization – that marginal revenue equal marginal cost:

47where V' and G' are the partial derivatives of V and G with respect to each firm’s quantity (negative). The new closest competitor, generics, benefits from the distortion created by the oligopoly, with higher sales at the same price a than under protection. This because for a few consumers, it is not worth switching to the expensive GI-original, meaning they are better off with the generic. As before, consumers are worse off, a result always obtained when there are fewer varieties in a market.

48The equilibrium price for Foreign products under protection is unambiguously higher than when protection is not granted, as the demand curve is steeper. However, with Cournot pricing, there are two possible results for Foreign firms regarding volumes. Panel C of Figure 4.6.a shows the typical case, wherein the marginal revenue curve for a regime of protection is to the right of the marginal revenue curve under no protection, such that GI-original sales under protection are higher than under no protection.

49Figure 4.6.b presents, in panel B an example of a marginal revenue curve under protection that is steeper and to the left of the marginal revenue curve under no protection (panel A), which indicates that it is in the interest of Foreign firms after protection to lower sales in order to increase profits. This is the reverse of the result found under perfect competition, wherein sales under protection are higher than under no protection with marginal cost pricing. This distortion is due to oligopolistic competition with a more differentiated product.

50An intermediary situation, represented by panels C and D of Figure 4.6.b, occurs when some productivity shock or subsidy lowers the marginal cost for GI-varieties to a level c' that crosses the horizontal segment of the broken marginal revenue curve Y* at  in panel C. GI-variety goods are out of the market, but sales and pricing by Foreign firms differ under each regime. If the GI is not protected, sales are capped at X* in volume, a limit-pricing equilibrium due to the discontinuity of the marginal revenue curve (panel C). If the GI is protected, at equilibrium sales will be lower in volume, and at a higher price, to the detriment of consumers. This occurs despite the fact that generics make up the decrease in volume of GI-original goods (panel D).

Figure 4.6.b: Reduction of sales due to oligopolistic competition by Foreign firms

b) GI-variety sales still competitive

51If GI-variety firms are still competitive

52the equilibrium is the same as the case in which the Foreign GI is taken as a generic but with b' > b instead of b. In this iteration, the effect is ambiguous. Sales of generics will increase as Y* moves to the left, but the shape of the new demand curve for Foreign (at b' + G), and of its corresponding marginal revenue curve, may be such that it is in the interest of Foreign firms to lower output in order to increase profits. Panel D of Figure 4.6.a illustrates the case in which Foreign firms expand output.

4.7. Welfare analysis

53Welfare analysis includes consumer and producer surplus (profits). Refer to Appendix C for derivations.

54Consumer surplus at Home is given by:

55Note that consumer surplus is not given exclusively by the areas between the inverse demand functions and the equilibrium price. Because utilities are additive, the existence of positive sales of GI-varieties and generics pin down the demand functions for their closest competitor up-scale, but consumers of GI-original goods add their surplus over the three utility functions, G, V and U; those of GI-varieties over V and U and those of generics over U.

56It is assumed that aggregate operating profits (producer surplus) of Foreign firms are based on a mark-up over marginal cost, so that ∏ = sX. Profits are zero under perfect competition (s = 0) and positive under oligopolistic competition (s > 0). In the case of GI protection, when GI-variety sales are deterred, consumer surplus becomes:

57Aggregate welfare is the sum of Foreign profits and Home consumer surplus: W = CS + ∏.

58There are two welfare effects omitted from this analysis: the welfare effect of government spending on enforcement, which implies a monetary disbursement ultimately paid by taxpayers, but which is assumed not to affect the producers and consumers of this market; and the welfare effect of the casual consumer mistaking a GI-variety for a GI-original.

4.7.1. Perfect competition

59Under perfect competition, sales of generic and GI-original goods both increase after protection, while Foreign firms have zero profits. The net welfare effect of protection is given by Home consumer surplus (derivation in Appendix C):

60Asterisks were omitted for simplicity. Whether GI-variety sales are deterred after protection or not, the effect on Home consumer surplus (equal to global welfare in this case in absence of profits for Foreign) is an unambiguous loss, as all terms are negative.

61In case GI-variety sales are deterred (Y₁ = 0), the first term in the welfare function corresponds to the welfare loss of consumers that previously consumed the GI-variety and switched to GI-original goods. On those units, there is a gain from valuation of origin (G) to which the price premium on those units is deducted (c – b), a net loss (in panel C of Figure 4.5.a, it is visualized as the area between demand G + b and price/cost c). Demand before protection was pinned down by the price of the GI-variety; after protection it is pinned down by the price of generics. The second term is the loss of welfare to consumers that switched from the GI-variety to generics; the loss from valuation of variety (V) minus the savings on those units (b – a), a net loss (in panel C of Figure 4.5.a, it is the area between demand V + a and price/cost b).

62In a case where GI-variety producers remain competitive (Y₁ > 0), the same losses appear, in addition to a third term that corresponds to the extra cost spent on GI-varieties by those who do not switch, a net loss (panel D of Figure 4.5.a).

63Although a loss in global welfare is always expected when fewer varieties are available in the market, some situations arise:

1. If the valuation of variety V is relatively flat at a low level, implying that only the “real thing” is of some value to consumers (and G is either steep or flat at a high level), sales of GI-varieties would be low with or without protection. As such, the relative efficiency of Foreign firms is what matters in terms of limiting losses once protection is granted.

2. In contrast, if the valuation of origin G is relatively flat at a low level, and valuation of variety is high (V is steep or flat at a high level), sales by Foreign firms would be low. It would then be expected that under protection generic sales would increase disproportionately. As above, the relative efficiency of Foreign firms is the salient feature.

3. Taken altogether, the above analysis suggests that industrialized (Home) countries, with sophisticated consumers (high G, low V) and higher relative costs (Home b and a close to Foreign c), tend to lose less by protecting foreign GIs than developing (Home) countries, which have less sophisticated consumers and lower relative costs of production.

4.7.2. Oligopoly

64In the case of oligopoly, four possible situations arise after protection: GI-variety sales are competitive or not; and GI-original sales increase or decrease. In all instances, Foreign firms have positive profits and the sale of generics increases (derivations in Appendix D). To make economic sense, it is assumed that net profits increase in all scenarios.51 This constitutes a net transfer from Home to Foreign (the decrease in sales exacerbating the oligopoly distortion):

65In the case where GI-original sales increase, net global welfare is formulated in identical terms to that under perfect competition, but the equilibrium quantities, and thereby the range of the integrals, differ:

66The first term in both equations starts positive (at X₀, G + b – c = s₀> 0) and may or may not become negative, indicating that, depending on the parameters of the model, consumers who switch to GI-original goods might be better off. Those who switch to generics after protection (second term in both equations) and those who stick to GI-varieties (third term in the case Y₁ > 0) have a net and unambiguous loss. This case is illustrated in panels C and D of Figure 4.6.a. Global welfare is ambiguous.

67If Foreign sales decrease in volume, net global welfare is given by:

68The effect both on Home consumer surplus and net global welfare is an unambiguous loss (proof in Appendix D).

69In case GI-variety sales are deterred (Y₁ = 0, drawn in panel B of Figure 4.6.b), the first term in the welfare function corresponds to the welfare loss of consumers that previously consumed the GI-original good and now switched to generics. On those units, there is a savings on those units (c – a), but it does not compensate for the loss from valuation of origin and variety (G + V). The second term is the net loss of welfare to consumers that switched from the GI-variety to generics; the loss from valuation of variety (V) minus the savings on those units (b – a). Net welfare is unambiguously negative, as both terms are always negative.

70In a case where GI-variety producers remain competitive (Y₁ > 0), the first term is the valuation of origin (G) minus the cost paid on each additional unit of GI-original good (c – b'). The second term is the loss due to the valuation of variety (V) of those who consumed GI-varieties and must switch to generics, minus the monetary saving on those units (b – a), a net loss). The third term represents the net loss to those who stick to GI-varieties.

71Many of the conclusions of this analysis are in sharp contrast with those found under perfect competition:

1. After protection, GI-original goods are more differentiated from their closest competitors (generics) than before protection (GI-varieties). This will allow Foreign firms to potentially obtain additional rents through a decrease in sales after protection, intensifying the competitive distortion. Industrialized countries, with sophisticated consumers (high G), stand to lose more from this situation than developing countries. Net welfare always sees a loss when sales of GI-original goods decrease in volume.

2. If sales of GI-original goods increase with protection, the possibility of a global welfare gain exists if the valuation for variety is low, and the valuation of origin is high. Industrialized countries will benefit from this situation. This is because the term

1. may compensate for the valuation losses of those who stick to GI-varieties or of those who switch to generics.

1. Relative efficiency in production, under these conditions, is less critical than under perfect competition.

4.7.3. Enforcement

72The Chile-EU agreements on wines and spirits include in their appendices the list of Chilean registered trademarks conflicting with European GIs, which were eventually protected, following a phase-out period of five years for exports and of 12 years in the internal market. These brands include: Asti, Baden, Borgoño, Burdeos, Carmen Margaux, Cava Vergara, Champagne Grandier, Champaña Valdivieso, Champenoise Rabat, La Rioja, Viña Manquehue Oporto (wines), and Cognac Juanico, Coña Col, and Grappa San Remo (spirits).52 This agreements, negotiated at the margins of a Free Trade Agreement (FTA), exemplifies the necessary equilibrium between redressing “usurped” GIs and compensating for the “claw-back” of GIs: Chile agreed to protect European GIs through a combination of trade-offs within the FTA and an extended implementation period.

73If Home protects the GI, but enforcement is deficient, Foreign firms will face some competition from GI-variety firms. Producers of GI-varieties internalize the risk of confiscation and eventual fines, and thus operate at a higher marginal cost, which, by assumption, is such that the conditions for Home GI-varieties to compete are satisfied. The better the enforcement, the higher the risk of confiscation and fines, the higher the marginal cost b, and the higher the equilibrium price. It is assumed that confiscated goods are destroyed. As long as the competition condition is satisfied, and enforcement is strengthened, only legally marketed products will prevail. Only the oligopoly case is considered (for the perfect competition case, sis set to 0).

74Alternatively, it may be assumed that Foreign firms producing the GI-variety abide by the law, and, as they are forbidden the shorthand use of the GI-label to signal their type, must incur a cost of periphrasis in labelling, marketing, and advertising their products. The effect, again, is an increase in marginal cost. Such a situation could arise, for instance, from the extension of the Article 23 special protection of wines and spirits to all products – thus preventing the use of the GI even when the true origin is indicated or when used in translation or accompanied by expressions such as “kind”, “type”, “style”, “imitation”, and the like. This situation is analogous to the impact of protection when GI-variety sales are not deterred.

75These results correspond to panel D of Figure 4.6.a:

76For Foreign firms, a marginal increase in enforcement implies an increase in b, a decrease in GI-labelled goods Y, and a shift upward of the demand curve for GI-original goods. Provided that the initial oligopoly equilibrium is stable, ∂X/∂b is positive. An increase in sales implies a lower valuation of origin by the marginal consumer of the additional units of Foreign goods, leading to an increase in the final price of less than the change in b. The first term has the sign of ∂s/∂X; it represents the loss/gain from the pricing of all original units at a new mark-up due to the increase in b. The second term is always positive and reflects the profits from the additional units sold.

77For Home consumers, the first term is the negative of the first term in the change in Foreign profits; it constitutes a net transfer from Home to Foreign. The main loss at the margins stems from the second term, as stricter enforcement raises the price of GI-variety goods above their marginal cost of production (increase in b), an increase that applies to all units of GI-variety goods sold. The destruction of confiscated items, and/or the cost of periphrasis, is a waste transferred to consumers in the form of a higher price.

78The effect of enforcement on global welfare is ambiguous. The first term captures the welfare gain owed to the increase in Foreign sales. In this context, enforcement is pro-competitive (offsetting part of the oligopoly distortion), as the supply of original GI products increases as Foreign firms respond to the upward shift in the supply curves of their competitors. The second term, however, is a social loss resulting from stricter enforcement, equal to the elevation of GI-variety sales costs (Y – X) above their marginal cost of production.

4.8. Concluding remarks

79This chapter aims at evaluating the potential economic impact of the protection of a geographical indication that had acquired generic status in a given territory (a procedure, in trade jargon, known as the claw-back of GIs). The focus is on the non-misleading international trade of GI-labelled products. Special attention is paid to the competitive and welfare implications of GI protection by countries importing the original product. As this book speaks directly to current negotiations occurring at both the multilateral and bilateral levels regarding stronger standards of protection for particular geographical indications, this work has immediate policy implications.

80It is demonstrated that a level of protection sufficiently strong to drive national firms competing with the Foreign GI out of the market would lead to a loss in global welfare – a situation that is always expected when fewer varieties are available. This finding, however, is nuanced under different scenarios. Under the assumption of perfect competition, industrialized Home countries, with sophisticated consumers and higher relative costs, tend to lose less in protecting Foreign GIs than developing Home countries, where the opposite is true. When oligopolistic competition is assumed for Foreign firms, moreover, GI-original firms become further differentiated from their closest competitor (now generics), exacerbating the competitive distortion. In this scenario industrialized Home countries’ consumers stand lose more than under perfect competition. Developing Home countries, however, with a low willingness to pay for origin but a high degree of valuation for variety, will be the greater losers.

81A case characterized by the failure to deter the sale of domestic copies of Foreign GIs, either due to branding, successful GI replication, or deficient enforcement, is also analysed. Interestingly, it is demonstrated that oligopolistic competition among Foreign GI developers is accompanied by ambiguous welfare effects and is actually more favourable to the development of Home copies than the scenario of perfect competition.