The World’s Largest Open Access Agricultural & Applied Economics Digital Library This document is discoverable and free to researchers across the globe due to the work of AgEcon Search. Help ensure our sustainability. Give to AgE con Search AgEcon Search http://ageconsearch.umn.edu [email protected] Papers downloaded from AgEcon Search may be used for non-commercial purposes and personal study only. No other use, including posting to another Internet site, is permitted without permission from the copyright owner (not AgEcon Search), or as allowed under the provisions of Fair Use, U.S. Copyright Act, Title 17 U.S.C. 1 Minimum Quality Standards and brand development in agrifood chains Giraud-Héraud, E.1, Grazia, C. 2 and Hammoudi, A. 3 1 Institut National de la Recherche Agronomique and Ecole Polytechnique, Paris, France 2 Agricultural Economics and Engineering Department, University of Bologna, Italy 3 Institut National de la Recherche Agronomique and ERMES/Université Paris II, France Abstract— This paper develops an original raw material, may be sufficient or, on the contrary, framework to better understand the interaction between insufficient to facilitate this strategy. Hence, firms the development of brands and the quality of raw might be lead to select only the most effective materials. We consider different levels of consumer trust producers or also to encourage their suppliers to for a brand and we examine the incentive for firms to upgrade upstream production conditions, through the improve the quality of a processed product by requiring creation of a private standard. This input’s that upstream suppliers adopt a private standard. In normalization strategy often corresponds to more or contrast to previous literature, the incentive for firms to less irreversible investments and procedures develop a more stringent private standard may increase (suppliers’ selection, contracts’ setting, norm’s with the level of the regulated minimum quality standard. Moreover, the creation of a private standard development, product’s certification, etc.). It also may can reduce the risk of consumer dissatisfaction while influence the firms’ short term decisions concerning increasing the marketed quantity. Unexpected positive quantity and price to adapt in fine to the evolution of effects of a reinforcement of the minimum quality demand and competition environment (see for standard may arise, in the sense that both market access example, Maurer and Drescher, 1996, Ponssard et al., for upstream producers and consumer surplus are 2005). improved and final price may decrease with respect to This paper shows how a medium-long term simply complying with the regulation. strategic choice about the mode of input procurement influences the short-term strategies, which may be Keywords— Minimum Quality Standard, brand, developed by the firm to provide the brand’s vertical relationship development. By considering different contexts of consumer trust in the brand, we thus illustrate the I. INTRODUCTION reasons why a firm would prefer the reinforcement of the upstream production condition and the conditions As in any industrial sector, the development of such that this strategy is implemented. Moreover, we brands by agrifood firms results from the intention to show that, unlike an accepted idea, this private meet consumer demand, while forming the basis of standard strategy is not necessarily due to a laxity of product differentiation from competitors. Moreover, the authorities in the definition of Minimum Quality the success of a brand depends both on a specific Standards (MQS). communication politics towards consumers and on the Two examples in the agrifood sector may illustrate consumer trust in firm statements about the brand (see how the choice of a brand development strategy is for example, the seminal works in the marketing strongly affected by both the level of MQS and the literature since Copeland, 1923). communication provided to final consumers: However, the brand success depends, above all, on i) The wine represents an emblematic example of the strategic manufacturing decisions, which are made brand development in the presence of upstream MQS. according to the technological possibilities offered to In this sector, there exists a great number of MQS that firms. Brand development is thus highly depending on – given the issues of sanitary safety or the respect of upstream raw materials production conditions, from the region of origin – mainly concern the vine which the final product results. Therefore, the public growers, which produce grapes or wine in bulk and regulation, which defines the standards concerning 12th Congress of the European Association of Agricultural Economists – EAAE 2008 2 sell it to downstream processing and/or retailing firms. ii) In the fresh products sector, a large development Within the European Union, an important part of of high premium labels by retailers has been observed production concerns Appellations of Origin and some in the last decade. With respect to the wine sector, of the well known regional ones (like Bordeaux in described above, one of the main interesting issues of France or Rioja in Spain) stand a lack of brand the supply chain management is given by the creation development to compete on the international market.1 of private standards, which reinforce the MQS. These The influence of the production criteria requested for private standards have been usually defined in the Appellations of Origin is often considered in response to increasing food safety concerns, namely in explaining this feature.2 One of the most frequent the meat sector (for example the “Filière Qualité arguments, which is based on the increasing trend of Carrefour”, the “Traditional Beef” of Sainsbury or the brands in the “New World” (for example, E&J Gallo “Selected Beef” by Mark and Spencer), but also for in US or Jakob’s Creek in Australia), is that too fruit and vegetables, fish and seafood or cheese constraining upstream production conditions are (Fearne, 1998). Specifically after the mad cow crisis, dissuasive for improving market strategies. That is the and despite the reinforcement of the MQS (such that reason why a French firm as Pernod Ricard prefers to the prohibition of using bone meal for livestock invest on a brand development strategy in Australia in feeding), the high premium labels in the meat sector order to avoid the too constraining regulations have been largely increased in the EU and have concerning grape production.3 Nevertheless, some involved an increasing number of upstream producers vineyard characterized by a high international participating in the brand creation5. Developed in notoriety (like Champagne, Porto or Chianti) have periods characterized by a crisis of consumers’ trust, been able to maintain a good reputation towards these strategies have reinforced the public regulation consumers. In these cases, brands are quite developed while surprisingly leading to an improvement of and a high intermediary price has allowed the upstream producer market access (see for example upstream producers to comply with relatively highly O’Brien and Diaz Rodriguez, 2004). demanding production conditions4. The objective of this paper is to illustrate some of these economic mechanisms associated to the brand development. We propose an economic formalization of the creation of a brand, in a context where the 1 According to Mora (2006), for several years now, Bordeaux's vineyards upstream production conditions are normalized. We have suffered from what would appear to be an interminable crisis. Some analysts view overproduction as the cause. Others blame the product thus refer to the specific case of the agricultural sector, Bordeaux puts out, decrying its lack of adaptation to new consumer where the upstream supply is fragmented if compared expectations. The author argues that Bordeaux producers do not tend to spontaneously adopt a market orientation. See also ViniPortugal, Monitor to the downstream processing and retailing sector. In Group (2003) for an analysis of the wine sector in Portugal and an this model, we consider a downstream firm with a illustration of strategies to improve competitiveness towards international monopolist position towards the final market and a markets. 2 The market access conditions for an Appellation of Origin are often monopsonist position towards the upstream atomized considered as MQS in the sense that the production of a wine outside the supply. Hence, the potential suppliers are numerous Appellation does not give access to the same markets and as far as an and price-taker in their decision whether to participate Appellation of Origin may represent a pertinent market. 3 As illustrated by Green et al. (2006), the French group Pernod Ricard has in the intermediary market. Upstream producers are largely invested on the international market, by developing wine brands as differentiated according to their equipments’ levels, Jacob’s Creek (Australia), Wyndham Estate (Australia), Etchard (Argentina), Río de la Plata (Argentina), Long Mountain (South Africa). which in turn determine the quality of their supply As illustrated by Pomarici et al. (2006) some of the leading Italian wine from the point of view of the consumers. Thus, the companies have invested abroad (expecially in US, Argentina and Central- East Europe), see for example the strategy of Antinori with brands like Antica Napa Valley (California), Col Solare (Columbia Valley), Albaclara 5 The Group Carrefour has launched the first FQC in 1992 (la “Boule and Albis (Cile), or developed partnerships with foreign companies (see for Bio”). Today, this strategy concerns 245 supply chains (in France) and 74 example the one between the Italian Frescobaldi and the Robert Mondavi products and involves 35.500 producers. About 40% of the products Corporation to create the brand “Luce”). concern the fruit and vegetable sector (Le Journal de Carrefour, 2005). 4 See Grazia (2006) for an illustration of the evolution of production With 200 suppliers in 1994, the production of the FQC fruit and vegetables conditions in the Chianti (namely, with the creation of the Appellation of has reached today a production of about 50.000 tonnes per year (Gaulet, Origin “Chianti Classico” in 1996) and a strong increase in intermediate 2000). See also Aragrande et al. (2005) for an analysis of the European price corresponding to the production conditions’ reinforcement. quality assurance schemes and implications on supply chain. 12th Congress of the European Association of Agricultural Economists – EAAE 2008 3 implementation of a MQS or a private standard might consumers may be better off, both in terms of quantity lead upstream producers to undertake investments in and final price. order to join the intermediary market. We thus provide an original contribution to the In this context, the downstream firm faces a quality- existing agricultural economics literature. A large quantity trade-off. That is, for a given level of quantity swathe of this literature examines the reasons for the supplied on the final market, an increase of the development of private quality and safety standards standard concerning the raw materials implies a and the effects of the level of MQS on the incentive decrease of the “risk”6 associated to the processed for firms to implement private standards. The main product, whereas, for a given level of standard, an idea is that firms will arguably have the greatest increase of quantity increases the risk for the incentive to implement private standards where there processed product. As a result, the implementation of are missing or inadequate public food safety and/or a private standard is likely to be necessary to avoid the quality standards; here private standards act as a negative effects of a high procurement quantity on the substitute for missing public institutions (Henson, risk. Therefore, firms may have different strategies for 2006; Henson and Reardon, 2005). In this spirit, brand development, which depends both on the level Giraud-Héraud, Rouached and Soler (2006) propose of MQS and on consumers trust in the brand. Namely, an original model of vertical relationship between if the trust is relatively high the firm has two options: producers and retailers which takes into account two i) choosing to select only some of the initially well- supply sources: i) a competitive spot market on which equipped producers, when the MQS is sufficiently low the retailers by a MQS product and ii) supply contracts (what we denote by a "Strict selective strategy"); ii) aimed at marketing higher quality private labels (PL). choosing to select the initially well-equipped The authors take into account the negotiation power- producers and also help some producers to upgrade sharing between downstream and upstream firms. It is their equipments to comply with the MQS, when this shown that if the MQS is relatively too high, then latter is higher (what we denote by "MQS adaptive retailer will not perceive any benefit in developing the strategy"). However, if the consumers trust is PL. Nevertheless, this literature recognizes that even if relatively low, and even if the MQS is relatively high, public standards are well-developed and afford a high we show the incentive for the firm to have a proactive level of food safety and/or quality, there may still be role and set a private standard more constraining than an incentive to implement private standards. Then, the the current MQS (what we denote by "MQS main reason to argument the coexistence of private reinforcing strategy"). Hence, we show that, it is not standards with highly demanding public regulation is when the MQS is relatively weak that the firms have given by the necessity for the firms to manage interest in substituting to the public authority and exposure to liability, limit exposure to potential implementing a private standard. Indeed, the regulatory action and/or anticipate future regulatory implementation of a private standard leads to a developments (Lutz et al., 2000). Despite, we show reduction of the risk of consumer dissatisfaction. how the incentive for firms to implement a private Hence, the processing firm can benefit from an standard when public regulation is relatively high may improvement of consumer willingness to pay and thus result from the strategic behaviour of firms in terms of increases the marketed quantity of the processed quality-quantity strategic choices in the context of a product. We thus show that when the downstream firm vertical relationship. Another set of contributions deals has interest in remunerating the upstream producer with the compliance process of firms to a process compliance process, market access may be improved standard and, more specifically, with the related issue through a reinforcement of the standard. Moreover, of producers’ capacity to comply with it. Thus, the compliance process represents a long term decision and results in more or less high adaptation costs for 6 In this paper we use the term “risk” to specify the non-compliance of the firms (Henson and Heasman, 1998). Hence, several processed product with respect to an expected quality. This terminology refers to the notion of “credence qualities” (Darby and Karni, 1973), which contributions examine the economic implications of is important in the agrifood sector, especially when the product standards using a cost and benefit analysis, which normalization concerns the aspects of certification of origin or food safety (see for example, Grunert, 2005 and Loureiro and Umberger, 2007). 12th Congress of the European Association of Agricultural Economists – EAAE 2008 4 attempts to measure the cost for firms of implementing A. Heterogeneity of producer equipments and risk of (food safety) regulations and compare it to the benefits product failure in terms of the reduced food borne illness (see for example Caswell and Kleinschmit, 1997; Antle, 1999; Following Giraud-Héraud, Hammoudi and Soler Viscusi, 2006). The main argument is that the more (2006), the upstream producers are differentiated the standard is constraining, the higher is the risk of according to their “equipment” level, which is firms’ exclusion from the market. Hence, it is shown represented by a one-dimensional parametere, for example, that the compliance with standards may assumed to be uniformly distributed on the pose a greater burden on small firms, due to the large interval[0,1], according to the density investments needed (Henson and Caswell, 1999, function f(e)≡1. Namely, the equipment level e Unnevehr and Jensen, 1999). Moreover, even if a represents the technical level of the farm before the standard is not mandatory in the legal sense, it could implementation of the standard. Thus, given the initial be de facto mandatory (Henson, 2006). Hence, when a equipmente, an upstream producer who wants to particular set of products or specifications gains supply the intermediary market must achieve at least market share such that it acquires authority or the level of equipmente , which corresponds to the influence, the set of specifications is then considered a s de facto standard (The Nature’s Choice standard of “process standard” in force in the market. We consider Tesco Stores PLC in the UK, that commands a market that the compliance with the standard, for a producer share of over 30 percent, is arguably an example). of typee, implies a fixed cost, which is assumed to Even if standards promulgated by private entities, take a linear formMax{0,e −e}. Namely, the cost of s unless referenced by regulations, cannot be legally compliance is given by (e −e) for a producer, whose s mandated, through market transactions such standards level of equipment is lower than the standard and zero may become involuntary in practice; firms have little otherwise. Hence, given the heterogeneity of upstream or no option but to comply if they wish to enter or supply, this cost function allows to explicitly take into remain within a particular market. However, the account the heterogeneity of the compliance costs7. strategic behaviour of the downstream processing or We interpret the risk of product failure on the final retailing firm, namely the quantity strategy in response market as the probability that the product does not to consumer demand, may be positive for producers, meet consumer expectations. The risk of product even if the standard is reinforced. failure is assumed to technically result from the upstream supply characteristics, whereas the II. THEORETICAL BACKGROUND downstream is assumed not to influence the level of risk8. Namely, the heterogeneity and the limited We consider a vertical relationship between J production capacity of suppliers implies that the upstream producers and a downstream firm. We probability of product success on the final market is assume that the downstream firm has a monopsonistic altered (and the image of the brand is compromised) position towards upstream producers and is a by the use of inputs, which do not meet the «ideal» monopolist on the final market. The firm buys x units production conditions expected by consumers (e=1). of input in order to market a quantity y of output. Hence, we consider that the risk associated with each Since each of the upstream producers is assumed to producer of typee, is affected by his level of offer one unit of the input on the intermediary market, equipment and is given byσ(e); where σ(.) is a then the firm has to source from different producers in decreasing function ofe. For the sake of simplicity, order to obtain the quantity x of input. 7 For an illustration of this heterogeneity in the empirical literature, see for example Kleinwechter and Grethe, 2006. 8 This assumption is crucial as regards the objectives of this paper. Namely, it makes it possible to isolate the influence that the downstream firm may have on the actual level of risk through its strategic behaviour (namely, short term quantity/price choice), regardless of the influence that the firm may have from a technical point of view. 12th Congress of the European Association of Agricultural Economists – EAAE 2008 5 we consider thatσ(e)=1−e. Hence, the individual C. Risk perception and trust in the brand on the final risk is maximal when the producer is characterized by market the minimum level of equipment; otherwise the risk is Consumers identify the firm’s product through the zero. Hence, each producer contributes with (1−e) to brand. The communication on the product’s attributes the risk of product failure on the final market. is provided either by the firm or by third parties B. Downstream processing stage (certifiers, consumer guides, etc.). We consider that the effects of this communication on consumer purchase decisions depend on the degree of consumer The processing stage may concern processing, trust, which in turn affects the level of perceived risk. preserving, conditioning or packing operations. The Let us detail these mechanisms. downstream firm converts the raw material into a Firstly, consumers are assumed to be imperfectly finished product according to a fixed-proportions informed about the product’s attributes, in the sense production function. Here, we consider that the that they react to the perceived probability of product downstream firm does not influence, through the failure rather than to the actual one. The concept of processing operations, either the risk or the number of risk perception includes all the risks associated with units sold. We denote by ɶe the threshold of equipment starting consumer choices at the point-of-purchase. Indeed, as highlighted by McCarthy and Henson (2005), risk from which producers are selected by the downstream perception concerns not only the health (for example firm. Hence, the firm always selects the producers characterized by equipment between ɶe and 1, that is, fat content) or safety (for example food poisoning) risks associated with the product, but also the chance the best level of equipment. The firm is assumed to that the product may not meet taste expectations, buy x units of input and convert them into y units of money is wasted, a poor meal is served to guest, etc. finished product, according to the fixed proportion Hence, this concept relates to the perception of both production functiony =T(x), where we simply the probability of product failure and the negative setT(x)= x. Hence, the threshold ɶe is given by: consequences of buying/using a product or service. Consumer perception of the risk may be influenced by x ɶe =1− (1) perceived product’s consistency, interest in cooking, J interest in the product, experience and confidence in Since we consider that each producer always purchase location (McCarthy and Henson, 2005), supplies the same quantity (one unit) of product (non- health loss, followed by psychological, financial, time elastic individual supply), the Benchmark situation and taste losses (Yeung and Yee, 2002). As a consequence, consumers may underestimate or (whene =0) is then defined by the following quantity s overestimate the risk of product failure, with respect to and risk of product failure: the actual level of risk. y= x Secondly, the perceived risk of product failure is assumed to be affected by the degree of trust in the 1 1 y (2) σ= ∫σ(e)f(e)de= ( )2 brand. Hence, as highlighted by Delgado-Ballester and ɶe 2 J Munuera Alemán (2000), trust in a brand can be Expression (2) represents the quantity bought and defined as “a feeling of security held by the consumer sold by the downstream firm and the associated risk of that the brand will meet his/her consumption product failure when no MQS is in force. When a expectations”. It is noteworthy that the process by MQS is implemented, the initial probability of product which an individual attributes a trust image to the failure given by (2) may change if at least one of the brand is based on his/her experience with that brand. producers upgrades his equipment. Thus, the density Hence, trust will be influenced by the consumer’s f(e) will shift to a density f'(e) and change the level evaluation of any direct (e.g. trial, usage, satisfaction in the consumption) and indirect contact (advertising, of σ with respect to (2). word of mouth, brand reputation) with the brand 12th Congress of the European Association of Agricultural Economists – EAAE 2008 6 (Keller, 1993; Krishnan, 1996). Moreover, trust is failureσ. The parameter λ is interpreted as a measure based on the two general dimensions of brand of the extent of consumer trust in the brand, reliability and brand intentions towards the individual, withλ∈[−1,1]. Hence, the aggregate inverse demand which involve the role of time. The first dimension is for the product, when the perceived risk is(1−λ)σ, is related to the assumption that the brand has the given by: required capacity to respond to the consumer needs, for example, by offering the new products that the p (α,l,σ,x)=α−(1−λ)σl−x (3) λ consumer may need or by a constant quality level in its offering (Deighton, 1992). The second dimension is Following (3), both the information about the concerned with the belief that the latter is not going to likelihood of product’s success and the consumer trust take opportunistic advantage of the consumer affect consumer willingness to pay, for a given level vulnerability (Michell et al., 1998). Given these of quantity. Namely, given the mechanisms illustrated premises, we focus on the perceived risk-reducing above, the lower the degree of trustλ, the higher the effect of the brand trust, this latter being interpreted as perceived risk of product failure and the stronger the an exogenous market (demand side) condition9. consumer reaction to a communicated decrease of the Finally, the level of perceived risk affects the extent risk. In equation (3), the parameter l represents the to which consumers react to a communication on the monetary loss for consumers for each unit of the product’s attributes. In a context of asymmetric product that fails11. We assume that α is sufficiently information, the main approach taken by consumers to high, namely α> J+2l (HP1)12. reduce the perceived risk experienced at the point-of- purchase consists in enhancing the probability of D. The game product success through the use of “risk relievers”, that is “a piece of information that increases the Given the MQS e set by the public authority in the 0 likelihood of product success” (McCarthy and Henson, long term, we consider the following game. 2005; Mitchell and McGoldrick, 1996)10. These Stage I. The firm chooses the level of private authors show that consumers characterized by the standard e >e ore =e . 1 0 1 0 highest level of perceived risk (“sceptic consumers”) Stage II. The firm decides the quantity x of inputs to tend to use more frequently extrinsic risk relievers to purchase (stage I.1). The firm then chooses N decrease the probability of product failure. When risk upstream producers (N ≤ J ) and proposes an relievers are given by the information provided either intermediary price ω in order to obtain the quantity x by the firm or by third parties and the perceived risk is (stage I.2). The N producers accept or reject this offer determined by the level of trust, ceteris paribus, the and upgrade their equipment if necessary (stage I.3). lower the trust in the brand, the higher the consumer Stage III. The firm converts the obtained inputs into reaction to a communicated decrease of the risk of a finished product and sells it to the end market. product failure, in terms of willingness to pay The game is solved using backward induction. We (marginal effect). firstly analyze the firm’s short term quantity/price Hence, following Polinsky and Rogerson (1983), we consider that in the end market consumers are choice, given a standarde . In this sense, we place the s identical and we denote by (1−λ)σ each consumer's analysis in the context of the traditional literature on perception of the actual level of risk of product MQS which aims at analyzing the effects of MQS on 11 According to McCarthy and Henson (2005), two dimensions of 9 If the level of trust would exclusively depend on the action of the firm, perceived risk can be distinguished, namely the perceived probability and then the firm would choose the highest level of trust, which corresponds to the importance of loss to the individual. the highest consumer willingness to pay for a given quantity. A different 12 This first assumption is obtained as follows. The final price given by (3) result may arise if the costs associated to the construction of brand trust are is positive, for any given level of quantity, if and only ifα>(1−λ)σl+x. considered. Moreover, a further contribution to this analysis may results from the assumption on a level of trust depending on the level of standard Given that i) x≤ J , ii) the risk varies from 0 to 1 and iii) the degree of in previous periods of time. trust is assumed to vary from -1 to 1, the final price is positive for any 10 See also Mitchell and Greatorex (1990) for an analysis of risk relievers given level of quantity and in any context of trust considered, if and only if in the UK food market. the parameter α is sufficiently high, that is: α> J +2l . 12th Congress of the European Association of Agricultural Economists – EAAE 2008 7 the firm’s strategic behaviour (see for example A. Producer compliance process with endogenous risk Ronnen, 1991; Crampes and Hollander, 1995; Scarpa, 1998) by considering that the MQS is exogenous, We denote by ˆx = J(1−e ) the quantity demanded s rather than explicitly consider the endogenous choice by the firm, whereby all the initially well-equipped of a MQS which maximizes social welfare13. Hence, producers are selected (eɶ =e ). Using (1), we verify we illustrate the effects of the standard on the strategic s behaviour of the firm in terms of quantity/price and that eɶ ≥e if and only ifx≤ xˆ. The quantity choice of s the related effects on the risk, on the number of the firm (that is, the relative position of the requested upstream producers selected and on consumer surplus. quantity x with respect toxˆ) thus determines the Turning to the first stage of the game, we then relative position of ɶe with respect to the standarde . examine the decision of the firm whether to implement s Given that, the firm’s quantity choice may result in the or not a private standard which reinforces the MQS set following two scenarios, according to whether the by the public authority. In this sense, we refer to the firm’s short term quantity strategy requires an upgrade literature dealing with the analysis of the incentive for of upstream production characteristics or not (we firms to implement private standards, according to the define more precisely these scenarios below). level of MQS (see for example, Henson, 2006; Henson On the one hand, if the quantity selected by the firm and Reardon, 2005). As only one product is sold on the market, only one standard can be operational. is relatively low, that is x≤ xˆ (eɶ ≥es), then the firm’s Hence, the standard e required on the intermediary quantity choice does not affect upstream production s market may be either a MQS (whene =e ) defined characteristics. Namely, if x< xˆ (eɶ >es), then the 1 0 firm selects only some of the initially well-equipped by the public authority or a private standard producers, while refusing some initially well-equipped implemented by the firm (whene >e ). 1 0 ones, namely those located between e andɶe. Hence, The paper is organized as follows. In section III, we s when x≤ xˆ no selected producer has to modify his provide an analysis of the firm’s quantity/price choice, equipment in order to supply the intermediary market. given the level of MQS. In section IV, we examine the As a consequence, the statistical distribution of decision of the firm whether to implement or not a producer equipment on the interval [ɶe,1] is private standard which reinforces the MQS set by the public authority. unchanged with respect to f(e)≡1. On the other hand, if the quantity selected by the firm is relatively high, that isx > xˆ, then the firm’s III. EFFECTS OF THE MQS ON THE STRATEGIC quantity choice affects upstream production CHOICE OF THE FIRM characteristics. Namely, the firm also involves some initially not well-equipped producers in order to obtain In this section, we analyze the firm’s quantity/price choice and the related effects on the level of risk, the quantity x (eɶ <e ). As a consequence, the s upstream producer participation in the market, final producers, who are initially located between ɶe and e s price and consumer surplus, given the MQS. have to upgrade their equipment in order to supply the intermediary market. The statistical distribution then 13 Even if a few contributions consider the endogenous choice of the MQS changes with respect to f(e) and is given by f'(e): (see for example, Ecchia and Lambertini, 1997), the choice of the criterion for determining the MQS is a very complex issue. Hence, there exist several criteria for the definition of a MQS, especially in the agricultural 0 if ɶe≤e<es  sector. In addition to the traditional criteria of maximization of social f'(e)=es−ɶe if e=es (4) welfare, other criteria could represent the public authority’s concerns, as  for example the minimization of the risk, especially in the case of product’s 1 if es<e≤1 safety, or the minimization of upstream producers’ exclusion. Following the main swathe of the economic literature on MQS, we thus examine the We now detail how the firm’s strategy influences effects of the level of MQS on the firm’s strategic behaviour, on the the risk, depending on whether it requires an average quality provided on the market and on the surplus of the other economic agents, without specifying the criterion of choice of the MQS. 12th Congress of the European Association of Agricultural Economists – EAAE 2008 8 upgrading of upstream production characteristics or evidence so that individual contracts rarely exist in the not. We denote by σ(e ,x) the risk for a given level of agrifood sector (see for example, Royer, 1998) and s intermediate price is usually negotiated between the standard e and for a quantityx. The equipment s retailer and the Producers Organizations and/or the distribution depends on the type of strategy chosen by cooperatives and rarely between the processing and/or the firm; we denote by h(e) this distribution, where retailing firm and each of the upstream farmers (see h(e)= f(e) if x≤ xˆ and h(e)= f'(e) ifx > xˆ. Using for example, Malorgio and Grazia, 2007, for an analysis of the role of Producers Organizations in the (1) and (4), we then obtain (see section 1 in the implementation of EurepGap by fruit and vegetables Appendix): farmers, Kleinwechter and Grethe, 2006).14 1(x)2 ifx≤xˆ First, we assume that the intermediary price is the σ(e,x)=1∫σ(e)h(e)de=2 J (5) same for all the producers, regardless of their initial s level of equipment. Hence, the downstream firm does ɶe  x 1 (1−e )[ − (1−e )] ifx>xˆ not have the possibility to discriminate between  s J 2 s upstream producers. Note that this assumption is As illustrated by expression (5), whenx≤ xˆ, since consistent with the absence of individual contracts since with different intermediary prices, each producer the firm does not have any influence on upstream would choose the highest price. Second, if the supply characteristics, the risk is not affected by the requested quantity is relatively low, the firm will only standarde . Conversely, whenx > xˆ, then the firm s select producers whose equipment is better than the procurement strategy determines an equipment standard (x≤ xˆ); otherwise – and given that the upgrading for the producers who are initially located production capacity of each producer is limited – the between ɶe ande . As a consequence, the level of the firm will be forced to also source from initially under- s standard e has an influence on the risk. In both cases equipped producers (x > xˆ). This assumption is also s consistent with the existence of an intermediary illustrated by (5), the risk is an increasing function of organization who can select the producers who want to the quantity. The reason is that an increase of the participate to the collective transaction. quantity requested on the intermediary market Thus, ifx≤ xˆ, the firm anticipates that all the implicitly leads to an increase of the number of selected producers enter the market without any cost producers involved and namely to the involvement of and can obtain the quantity with a zero intermediary more and more under-equipped producers. Hence, the price. Conversely, whenx > xˆ, the producers initially expression (5) illustrates the existence of a quantity- risk trade off in the following sense. Namely, the risk located between ɶe andeshave to invest in better increases in quantity, for a given level of standarde , equipment (eɶ <e ). In particular, the producer located s s whereas it decreases when the standard is reinforced, in ɶe is the last (less equipped) producer who for a given level of quantity. upgrades his equipment by investinge −eɶ. Hence, he s B. Intermediary price does not agree to participate in the market if the intermediary price is lower thane −eɶ. In order to s Since we consider that the downstream firm has a obtain the optimal quantity of input, the downstream monopsonist position towards upstream producers, firm proposes a price so that the less-equipped then it has complete negotiation power in the producer can participate in the market. Thus, using (1), definition of the intermediary priceω. The firm thus the intermediary price ω(e ,x) is given by: s sets the quantity x by anticipating the necessary price in order to obtain this quantity x (see Xia and Sexton, 14 We have voluntarily left out the explicit formalization of the 2004, for the original modelling of this decision intermediation assured by the Producers Organization, with which the process). The analysis is developed by the two downstream firm negotiates (as shown by empirical evidence). Indeed, taking into account this intermediary in the model would not change either following assumption, supported by the empirical the analysis or the qualitative results. 12th Congress of the European Association of Agricultural Economists – EAAE 2008 9 0 if x≤ xˆ Using (7), we then maximize the expected profit  ω(es,x)=x (6) πλ(es,x) with respect to the quantityx, given the J −(1−es ) if x> xˆ standardes. For every degree of trustλ, and given the standarde , we show that there exist two levels of In the first scenario, whereas all the producers s equipment, e ande, decreasing inλ, such that the located within the interval [e ,1] would agree to s * enter the intermediary market, the firm exerts at the optimal quantity x (e ) chosen by the firm is given λ s maximum level its monopsonist power by refusing the by (see section 2 in the Appendix for details): producers, whose equipment is lower thanɶe . Otherwise, ifx> xˆ then the firm chooses an J[1−e] if es≤e  intermediary priceω(es,x), so that the less equipped x*λ(es )=J[1−es] if e≤es≤e (8)  producer participates in the market. As a consequence, for a given quantity, the higher the standard, the higher JΨλ(es) if es≥e the compliance cost of the less equipped producer, the Setting: higher the intermediary price. Moreover, a direct consequence of the absence of price discrimination is 1 (1−λ)l(1−e )2 +2(α+1−e ) the existence of a positive externality for all the Ψλ(es )= 4[ (1−λ)l(1s−e )+(J +1) s ] (9) producers, whose equipment is higher thanɶe . s We can verify that Ψ (e)=1−e and thus the C. Standardization, optimal quantity and effect on the λ risk optimal quantity choice of the firm is continuous ine . s The two levels of equipment, e and e are two We now characterize the firm’s expected profit. For thresholds that identify the relative position of the a degree λ of consumer trust, the firm’s expected optimal quantity with respect toxˆ. In order to examine profit πλ(es,x) as a function of the standard es and the firm’s strategy in all the possible cases, we place the quantityx, is given by: the analysis in a context of the parameters whereby0≤ e<e <1, by assuming (see details in π (e ,x)=[p (α,l,σ(e ,x),x)−ω(e ,x)]x (7) λ s λ s s Appendix) that α≤2J (HP2). Furthermore, in order Where the risk σ(e ,x) is given by (5), the final to be consistent with (HP1), we pose J >2l (HP3), s which is also consistent with the assumption of price- price p (α,l,σ(e ,x),x) is obtained by substituting λ s taker upstream producers. (5) into (3) and the intermediary price is given by (6). Hence, the objective of the firm is to maximize the profit, given by (7), according to the quantityx. As IV. MINIMUM QUALITY STANDARDS AND BRAND DEVELOPMENT illustrated by (7), the quantity choice affects the expected profit in different ways. On the one hand, the The selection and remuneration of upstream lower is the quantity, the lower is the intermediary producers result from: (1) the exercising of both price, for a given level of standard. On the other hand, upstream and downstream market power by the the lower is the quantity, the higher is the final price. downstream firm within the vertical relationship; we This latter result is given both by a rarity effect (direct thus examine the influence of the public regulation on effect of quantity on price) and by the risk-reducing the downstream firm’s strategy concerning the (and WTP-increasing) effect of a quantity decrease. selection of upstream producers and on the The magnitude of this indirect effect of quantity on mechanisms governing the definition of the price depends both on the actual level of risk and on intermediary and final prices; (2) an imperfect consumer trust. consumer information about the actual level of health 12th Congress of the European Association of Agricultural Economists – EAAE 2008