Abstract
Introduction
The consumer press is a key player in today's markets. 1 On the one hand, it contributes to shaping the demand side through recommendations and advice given to the public concerning the quality of various products and services. On the other hand, it tends directly to influence the supply side through a series of operations targeted at firms or public authorities, which are visible enough in the public sphere to generate decisions that carry economic consequences. These operations include the denunciation of dishonest manufacturers, alerts concerning dangerous products, court cases to trigger changes in legal decisions on consumption, etc. As the commercial success of periodicals and magazines specializing in various domains of consumption and the multiplication of ‘consumer’ columns in national dailies show, this approach is not the hallmark of consumerist magazines. Nowadays, offering information on new products and criticizing the innovations circulating in the market are relatively common exercises. Yet consumerist magazines in the strict sense of the term, like Que choisir and 60 millions de consommateurs in France, retain a particular position in the press as a whole. This position is related as much to the specific nature of the point of view they wish to represent, as to their goal of promoting a rational conception of consumer choice. 2 In practice, this ambition of consumerists is based largely on a particular exercise: the comparative testing of products and services.
Social scientists have paid little attention to the way in which this type of consumerist information is produced. For researchers interested in the consumer press, the production of comparative information on goods and services is conceived as a painstaking activity at best, and at worst, as a kind of craze with limited effects on customer behaviours. Scholars usually suggest that comparative testing constitutes the reasonable side of a press characterized, above all, by protest journalism (Pinto, 1990); one that most often reports on unrealistic comparisons never encountered in the consumers’ real life (Marcus-Steiff, 1977); or that does not influence their opinions in any substantial way (Aldridge, 1997). Whatever the empirical evidence provided to support this kind of criticism, the practical processes of producing comparative information on product performance have been relatively ignored. Yet it seems to me that these processes are worthy of investigation because they reveal many interesting aspects of the micro-politics of value (Appadurai, 1986) that underpin the circulation of commodities in contemporary markets. Through an experimental approach, consumerist testing elaborates the nature and properties of economic goods that usually stand in opposition to those promoted by market suppliers. As with any deconstruction, this invites us to reflect on the logic of construction itself and, more generally, to question the way in which various actors compete for the qualification of economic goods in the market.
I propose to investigate in detail the processes that give birth to comparative tests, in order to show the originality of the consumerist approach to quality evaluation. In the first section, I will briefly review the resources available in the realm of the social sciences to characterize consumerist information, and specify a frame of analysis rooted in an economic sociology perspective. In the second part I will study the construction of comparative tests from this frame. I will deliberately ignore the reception of test results by readers, focussing instead on the work carried out by the professionals who perform comparative testing in order to illuminate the practical operations they bring into play and the troubles they encounter. I will suggest that this practice – a relatively rich and fairly original one – aims to define a particular position for the consumer, one in which his or her ability to choose is subject to a specific form of what I call ‘disengagement’ from the market.
A perspective for understanding comparative testing
The sociology and economics of consumerist information
Even though consumption has been widely studied by sociologists, their approach leaves little room for discussing the role consumerist information plays in consumption practices. Baudrillard (1998), for instance, argues that goods are no longer characterized by their value or their properties but by the signs they represent. This conception undermines the importance of information obtained through comparative tests as goods are defined, to a large extent, in relation to their performance and not to the semiology of which they partake. Similarly, Bourdieu's emphasis on the social logic of distinction tends to restrict the question of evaluating goods to the exercise of taste (Bourdieu, 1984). The rationality of consumer choices – in so far as this expression can be used to denote processes that are as unconscious as they appear in the theory of the habitus – is embedded in a social context largely untouched by public information on goods and services. As a whole this perspective is of little help in understanding the impact of performance tests on consumers’ opinions.
In contrast, economists have shown a great deal of interest in the information necessary for accomplishing an economic exchange. Yet the way in which consumer theory has taken consumerist information into consideration remains complex. Indeed, economics has spent a lot of energy to delineate the relationship between information on prices and quantities but it usually encounters difficulties when dealing with information concerning quality – which is precisely the information produced by comparative testing. The failure of economic theories to integrate product quality as a relevant variable has been criticized for a long time and solutions have been proposed to improve the neo-classical model in this respect (Eymard-Duvernay, 1989). For example, Lancaster (1966) has proposed to split goods up into their constitutive characteristics: each good is then seen as a combination, as a ‘basket’, of different characteristics that can be qualified (for instance, through the use of consumerist tests). Although Lancaster notes that these characteristics – the only relevant variables for informing the consumer's choice – are objective properties, he says nothing about the devices and practices used to reveal them or about their modes of appropriation and interpretation by economic actors.
Consumerist actors as prescribers
In contrast, cultural theory of consumption tends to consider that appraising the performance of products and services falls mainly under the realm of arbitrary judgement disconnected from the goods themselves, while economics rely on an objectivist definition of performance. The former places little value in investigating the modes of production and appropriation of consumerist information (why bother since that's not where the consumer's choice is made?), while the latter raises the question without providing the means to answer it (the analysis of consumerist know-how remains outside the economic model). I suggest shifting the perspective slightly by focusing on the role of prescribers (Hatchuel, 1995).
An analysis in terms of prescription stresses the action of mediators that ‘fit’ the space separating suppliers and customers and participate in the structuring of their relations. This perspective leads us to take into account the existence of a plurality of prescribers in the market, as well as the mechanisms of competition between them. Consumer magazines are by no means the only actors to propose information on the quality of goods and services, and it is important to identify what differentiates these materials from other categories of prescribers: advertisers, public authorities, consultants, distributors, etc. This question runs through Alan Aldridge's research (Aldridge, 1994; Aldridge, 1997) on British consumer unions. Following a British sociological tradition strongly influenced by Marxist critique (Keat et al., 1994) 3 , he considers the issue of consumer power and authority in relation to the other actors in the market.
In a study devoted to the magazine Which, the British equivalent of the French magazine Que choisir, Aldridge investigates competition in the market for information on products and services. He compares consumerist magazines and specialized magazines that also publish advice and information on goods. In comparison, consumerist publications are characterized by their independence from manufacturers, especially in financial terms. Magazines affiliated to independent consumer unions (eg, Que choisir) or to public institutions (eg, 60 millions de consommateurs, created by the Institut National de la Consommation) do usually not appeal to advertisement in order to remain autonomous from manufacturers.
This independence is not the only difference between these two types of prescribers. Aldridge (1994) identifies several other points: first, consumer organizations relate to a form of depersonalized authority rather than to the charismatic authority usually mobilized by specialized magazines; second, their product evaluation is based on scientific analysis rather than on connoisseur-ship; and, lastly, they portray consumption in an austere light, as opposed to specialized magazines’ implicit emphasis on consumption as an hedonistic activity. More generally, Aldridge contrasts the consumerists’ ascetic and critical discourse with the proliferation of ‘promotional discourse’ accompanying goods and encouraging consumption. With regard to financial services and investment advice, his analysis (Aldridge, 1998) plays down the real effectiveness of consumerist methods in the empowerment of the consumer. This powerlessness shows the difficulty of holding a key position in a context marked by the influence of more pre-eminent forms of prescription (especially advice by financial experts and investment clubs).
Consumerist prescribers in the process of qualifying goods
Alan Aldridge's research is suggestive of the reasons why the notion of prescription for the study of consumerist activity might be interesting, since it emphasizes the capacity to produce and transmit information on the goods traded, in a specific way. I propose to take this research question further and to consider the task performed by consumerist prescribers from the point of view of qualifying goods. As Callon et al. (2002) show, the process of qualifying commodities that lies at the heart of competition in modern markets operates through the attribution of particular positions and competencies to economic actors. From this perspective, the distinctive feature of prescribers is not that they provide information that would be more objective than that disseminated by producers. This would merely boil down to the traditional mise en scène of the transaction provided by economic theory, resorting to notions like imperfect information or bounded rationality in order to explain the mediating role of prescribers. It is rather that prescribers offer alternative modes of engagement between consumers and goods – and therefore, between consumers and suppliers. Prescription supports the ability to calculate – in the extensive sense defined by Callon and Muniesa (2005) – in the making of market choices, an ability which is not necessarily concentrated on the individual actor and can be distributed among several actors and material settings.
From this perspective, let me briefly examine the performance test published in a consumerist journal. Two interesting features immediately appear, which provide a preliminary understanding of the way the consumerist prescription alters the consumer's engagement in the market. First of all, the act of reading a performance test in a consumer magazine is designed to shape a situation of consumption choice prior to the economic transaction as such. Yet this type of distilled situation is actually rare in the realm of consumption, for the constitution of information intended to guide choices is most often rooted in the interaction with market actors and networks. Consumerist activists like to recall that the consumer's capacity to choose from within traditional trade institutions is more restricted than market professionals usually claim: even when they are confronted by countless products sold in the supermarket, consumers are already in the shop where everything has been done to organize their route and lead them to the purchasing act, from the packaging of products to the salesperson's smile, through to the arrangements of the shelves or the choice of background music. 4 Consumers are already engaged in the commercial act. In contrast, using the terms of a distributed cognition specialist (Hutchins, 1995), I would describe the consumerist magazine device as carrying out a ‘pre-computation’ of the purchasing act: choose first, then (maybe) buy. It aims to separate the emergence of choice and the purchase and to distribute them to two different time-frames and situations.
The second striking feature relates to the rhetoric of testing itself. The very idea of setting up a sort of experiment, with all of the quasi-scientific characteristics this entails, tells us much about the way consumerists consider the correct attribution of qualities to economic goods. Social scientists like to say that one of the first steps in the development of a scientific approach is to fight against ‘prenotions’. Similarly, consumerists want the consumer to get rid of any prenotion that blurs a correct understanding of the commodities, and in particular the prenotions transmitted by other economic actors. Of course, as the sociology of scientific knowledge has extensively shown, there is no scientific project without a related political device. As we will see, the consumerist testing apparatus is not bare of political significance.
As a whole, we can consider the consumerist's goal as precisely opposite to that of the actors on the supply side: it aims to disengage consumers from the market infrastructure 5 – to the point where it may even advise them simply to cancel the planned transaction and to not buy. The market abounds with equipment enabling consumers to calculate so that they can shape their preferences (Canu and Mallard, 2006) and escape the paradox of too much choice (Cochoy, 2002). Consumerism would like to replace this equipment by the apparatus of comparative tests, that is, a quasi-scientific setting of prescription designed to allow the consumer to make a choice without engaging emotionally in market devices – and possibly without even buying. Yet this ambition entails various costs and requires a specific kind of work, with regard to the identification of the economic goods to qualify, to the construction of a representation of the supply and to the defence of a discourse of autonomy vis-à-vis the market. I will focus on these points throughout my examination of the practical achievement of performance testing.
Investigating consumerist work
Let me now outline the empirical material I will mobilize for this purpose. It is based on a survey of performance tests concerning telecommunication products and services that were published in 60 millions de consommateurs and Que choisir in the end of the 1990s. 6 I examined the comparative tests run on telecom products (faxes, mobile phones, Internet subscription, phone cards, cordless phones, etc.) in 1997, 1998 and 1999 and tried to determine the work that lay behind those tests. Like many other activities, comparative testing is standardized. The main ethical principles (eg, ‘Tested products have to be bought in conditions comparable to those of the average consumer’) are found in the French standard NF X 50–005. These rules form a framework of legitimacy for consumerist magazines that distinguishes them from rival prescribers. 7 Testing, however, is of course more complex in practice than the rules depicted in the standard, and only a dialogue with professionals enables a thorough understanding of this work.
Performing a comparative test in a consumerist magazine requires the collaboration of two different types of actors: first, the engineers of the test centre working with the magazine, in charge of designing and monitoring the performance tests usually sub-contracted to test laboratories; and, second, the journalists or market researchers who play only a small part in the design of the technical testing protocol but who are key in drafting the final reports and writing up the results in articles. I met the managers of test centres and the engineers participating in the tests concerned. 8 The data concerning the more journalistic part of this activity are partly drawn from this survey and partly from a former survey undertaken in 1997, on the role of consumer organizations in controversies over water quality (Mallard, 2001).
All in all I was able to collect general information on testing, supported mainly, but not exclusively, by examples from the telecommunication sector. In the end of the 1990s, the liberalization of this sector extended the scope of competition to a wider range of goods, thus lending a new dimension to the question of choice. Yet all situations of consumption are obviously not identical with regards to choice. It is worth noting that much of the information provided here applies, on the whole, to household appliances. 9 The examples I use in the analysis below pertain largely to telecommunications products and services but also, here and there, to other consumer goods.
Comparative tests on telecommunication products and services: elements of a practice
As historians and sociologists of science have shown, the significance of an experiment never lies only in the technical achievement of the test itself. For instance in their seminal study of Robert Boyle's experiments on the properties of air, Shapin and Schaffer (1985) pay as much attention to the technical design of the air pump, as to the various social, political, literary practices that were mobilized to constitute an audience for these experiments, and to disclose their results far beyond the laboratory. Similarly, I think that in order to take the measure of performance testing, it is necessary to examine the entire process from the elaboration of the experiment to the publication of results in the consumerist journal. I will focus on five successive categories of operation comprising a comparative test: the preliminary market analysis; the identification of the goods that takes place through the sample's construction; the formulation of criteria for the tests; the formatting of the results; and the presentation of results to the manufacturers concerned.
Analysing the market and informing the consumer: a matter of qualification and critique
Let me begin with an obvious fact: to inform the consumer, consumerist workers have to be informed themselves. Like scientists investigating a new field, the consumerist workers have to constitute their expertise for each new subject, and a comparative test usually starts with a data collection phase. Engineers and journalists have to update their know-how regularly by reading the specialized literature and by reviewing a subject each time the magazines plans an investigation. The engineer in charge of the test, possibly in collaboration with the journalist or market researcher concerned, carries out a preliminary study and sometimes even questions professional experts in the field – who will cooperate in the exercise more or less willingly.
This phase is extremely important because it initiates the mechanism of deconstruction and reconstruction of the economic environment supporting the design of a test protocol that will ultimately prove relevant and legitimate for the magazine's readers. In order to provide their own representation of what the market is made up of, consumerists have to gather various pieces of information, discuss their relevance and rearrange them in a new combination that they will pass onto the readers. Indeed, consumer magazines aspire to train readers, not just to inform them. As Hatchuel (1995) would put it, they are not content to supply information, the utility of which is known to readers in advance (a prescription of facts). Rather, they define criteria that readers did not know about and were unable to formulate (a prescription of judgement). Prior to the presentation of tests themselves, this ambition requires the elaboration of products and services as well as their contextualization. Thus, reading articles presenting performance tests is an effective way of seeing the task of market analysis that consumerist workers perform at the beginning of each new test. In these articles, it appears that economic goods are (re-)qualified at three different levels: at the level of the technology, of the possible uses of the technology, and of the market actors that are involved in its circulation.
On the technical side, articles explain the way objects work or the way they are made, list the services they offer, distinguish between useful functions and gadgets, and so on. In a test on phone-faxes, for instance, the difference between thermal transfer and laser or ink jet technology is explained. With regards to use, the consumerist magazines describe practices (eg, ‘What does surfing the web mean?’) and answer a number of questions that users are likely to ask (‘Should you insure your mobile phone?’). 10 Finally, a significant proportion of information provided relates to market actors’ behaviours and exchange mechanisms: the strategies and range of actions of the main actors (manufacturers, distributors, service-providers, etc.) are identified, the consumer is warned about usual traps to avoid, about possible bargains and bad deals, price trends and market share are qualified, etc. For instance, in the telecom market the ambiguous role played by service providers working as intermediaries between operators and customers is regularly discussed in articles on mobile phones. Indeed, consumerist magazines consider these actors – the intermediaries – to be parasites in the market relationship, as providing no real service and as needing to be treated with caution.
An important aspect involved in this multifaceted analysis is the identification of the traded good itself. The telecom sector is interesting in this respect since many goods are combinations of products and services. Depending on the good's ‘perimeter’, its performance and associated costs will vary. The case of mobile phones, for example, is generally meticulously investigated in consumerist magazines. The slightest comparison is an opportunity to take into account not only the performance of the telephone itself but also of its battery, subscription rates, call rates, and the various options that combine to make a whole. Here, consumerists prompt consumers to question what is really involved in the act of buying and using a good, against the definitions provided by the suppliers’ promotional information.
This work of constituting and providing information to the reader, a necessary preliminary step in the construction of understandable and legitimate comparisons, has a polemical and critical objective. It sometimes makes it possible to suggest that what the supplier wants to sell to consumers does not match what the consumers will eventually have to pay. More generally, it proceeds from a critique of the market, the flip-side of providing information. One can even suggest that the specific nature of the information produced by consumerism stems from a dual construction of the critical stance: consumerist magazines aim both to reveal the hidden properties of consumer goods and to denounce injustices in the market. The first part of the work is performed by the engineers who construct the comparative tests, while the second part is carried out mainly by the journalists. They contextualize the tests in the critical context of consumerist discourse and sometimes do field studies resembling investigative journalism applied to the market. Even if it generates multiple tensions, the link between these two activities of qualification and critique (and the subsequently sometimes complicated relationships between the journalists and the engineers) distinguishes the originality of consumerists’ ambition: to challenge not only the quality of goods but also the trust that one can have in the different categories of prescribers and in their modes of engagement in the market.
Identifying and collecting consumer goods
I shall now examine the practical constitution of the sample of products that will be submitted to the test. At first sight, this step seems mundane, but in fact, it conceals many tensions and difficulties that are emblematic of the episteme of performance testing. In the previous section, I alluded to the fact that consumerists discuss the perimeter of the goods that have to be tested. The delicate question of the identification of the supply lies at the heart of the operation. In order to highlight the significance of this operation, let me compare this task with the work of an entomologist interested in testing the reaction of insects to a new insecticide designed to kill a species of harmful butterflies. The entomologist has to catch the insects in the natural environment and bring them to the laboratory, where he can perform the test. Before he departs to the fields with his net, the entomologist needs a way to recognize the butterflies belonging to the particular species concerned. He will use a classification, which makes possible the identification of the butterflies that are relevant for the test, and he will discard the ones that are not concerned. The consumerists’ job is quite analogous: he has to identify a particular category of goods, catch them ‘in their natural environment’ and bring them to the laboratory in order to test their performance.
In the domain of consumption, the equivalent of the natural classification used by the entomologist is segmentation. As economists know, segmentation is a tool that helps to solve the first problem of the consumer theory: identifying the goods that are substitutable for one another, whose quantities can be added up to calculate the total amount available in the market. Commodities may be dissimilar but comparable in the sense that they fulfil analogous functions for the consumer. These will be put in a given segment. Yet as anthropology has shown – see for instance Kopytoff (1986) – there is nothing straightforward or neutral about assembling separate goods into a category. Therefore, the consumerists usually scrutinize the segmentations proposed by suppliers or inscribed in technical standards. My survey shows that segments devised by marketing professionals can be at least partially ignored during the testing.
If, for instance, the market analysis has revealed that many consumers in the general public buy products that suppliers target at professionals, these products may be included in the test sample: the opposition between ‘products for professionals’ and ‘products for the general public’ will not be taken as such in the constitution of the sample. Segmentations based on technical criteria are likewise critically examined. The test managers try to see which segmentation is developed from the point of view of use. For example, it might be possible to segment the refrigerator market according to the number of compressors each fridge comprises, but this is of little relevance to the consumer whose interest when buying a fridge rather refers to its capacity. In other cases, a technical segmentation can match a segmentation that is meaningful for the user. A case in point is washing machines, where the spinning speed corresponds to the price variable:
‘It may seem strange that we segment the market by saying ‘Que Choisir tests washing machines that spin at over one thousand revs per minute’. However, there's an almost perfect correlation between the price of a product and its spinning speed. So, in the end, as far as I'm concerned it amounts to testing products with the same spinning speed.’ (Interview with the testing manager at Que choisir.)
The next step is selecting the goods. The influence of this phase in the overall structure of a test can vary from case to case, depending primarily on the degree of competition in the market under consideration and on the speed at which products evolve. Two phenomena are particularly interesting here: the pressure of time, and the dilemma of variety reduction.
Let me come back to the comparison with the entomologist. For his experiment to be valid, he has to bring the butterflies to the laboratory, alive. No test of the insecticide's efficiency will be possible with dead butterflies. It is the same with the performance test: all the goods tested have to be ‘alive’, in that they must still be being sold when the magazine article is published. The demand for up-to-date products is central to the spirit of comparative testing, since this experiment does not bear on the respective performance of different technical objects but rather on the assessment of different purchasing options. Test managers have to select products that are already available when the test is scheduled and that will still be available – and preferably in line with current tastes – when the results are published. Due to the time needed to perform the tests, however, a few months may separate these two points in time, during which products are withdrawn from the market and innovations are introduced. A telephone handset that has not been available in the market for over two months may be a ‘product’, but it can no longer stand as a ‘good’ in a comparative test. The market is a shifting entity, constantly evolving, whereas the comparative test needs to ‘freeze it’ in order to produce a photograph that can be valid at two different times.
Controlling variety also has important effects on the test. What is at stake is to determine which good can be included in the comparison and which do not belong. It sometimes proves necessary to reduce the commercial offer in order to reach an acceptable size for the sample. The offer may be so vast and complex that for practical reasons or for cost considerations it is unthinkable to test all possible products. In 1998 the magazines 60 millions de consommateurs and Que choisir published test results at the same time (in November and December, respectively). In total, 27 telephone models were tested: 21 by Que choisir and 10 by 60 millions de consommateurs but only 4 of them were featured in both samples. Under such circumstances, it is not surprising that the final ranking of the products differs significantly from one magazine to the next – even when both achieve good quality assessments. In some cases the magazine articles even point out that the reader has to take this factor into account in appraising the test results.
Many of the tensions outlined here come from the very nature of the comparative test as a device with ambitions to represent. Using Pinch's (1993) terms, there must be a similarity, a possible projection, between the situation of choice as it is simulated in the test report and the one that is really experienced by consumers when they are shopping. An analogous requirement is present in the entomologist's practice: the act of killing a series of butterflies in the laboratory has no particular relevance in itself, and makes sense only because these butterflies represent other butterflies present in the natural environment, that will be eradicated if the insecticide is disseminated outside the laboratory. It is interesting to note that the requirement of similarity and projection, which is so important for consumerists, does not concern other market actors as much. Retailers do not have to carry out a representational process: the available offer, that is, their offer, is precisely the one the buyer will find in the shop's shelf space. In a sense, specialized magazines leave it up to manufacturers and distributors to fill their advertising space with lists of products and prices, thus constituting a representation of the offer available in various places. Mail order selling must put together an offer, contextualize it in a representative device and keep it available, but the rules and process mobilized to do so probably do not face the constraint of representativeness that consumerists aim for. Comparative testing set up a fictitious point of view on the supply which is available nowhere as an actual juxtaposition of rival goods but which has to be realistic enough to include what buyers may possibly find if they enter into the market. Organizing the ‘pre-computing’ of buying is, in the final analysis, a complex and somewhat strange task, especially when one is not engaged in the business of commercial distribution.
Devising criteria and trials
Once the goods have been identified and purchased in shops or supermarkets (taken from their ‘natural environment’), they are ready to be submitted to the tests. But devising the criteria and trials themselves proves to be a tricky task in comparative testing. What comes to mind are the words of Lancaster who, with regard to relevant characteristics, wondered how many people would choose their car depending on the colour of the engine. This suggests, among other things, that it is very difficult to propose a general definition of the relevance of a criterion of quality. I will limit my comment here to a few pragmatic points and resources mobilized in devising criteria and tests.
Tests of solidity and security probably constitute the basic components of a large array of comparative tests. They point to traditional background know-how, rooted in a long history of consumerist practice. Blacklisting products that fail to withstand shocks during use is a traditional feature of performance tests, one probably inherited from the industrial practice of quality control where these practices originate (Gabriel and Lang, 1996; Winward, 1993). Attention to product safety has also become a systematic concern, since the Corvair affair that made consumerism ‘à la Nader’ 11 popular. The engineers that I met consider that, whatever their importance, criteria of this kind are neither the most interesting nor the most distinctive of the consumerist approach. According to them, the most specific dimension of their work today relates to the way they take use into account.
The predominance of use in identifying relevant criteria for the consumerist evaluation calls for comment. In reflecting about the nature of the commodity in the capitalist economy, one may think, firstly, of the Marxian distinction between use value and exchange value. Whereas traditional economics (be it neo-classical or Marxist) has been very absorbed with determining the exchange value of goods, consumerist actors consider investigating use value to be of main concern. Secondly, one may notice that considering use implicitly or explicitly means taking into account the user's social identity. The choice of representing one or another category of user has a political significance. Hence, consumerist tests usually testify on behalf of particular categories of users who may statistically represent minorities but who still deserve consideration. In a test of telephone handsets, they will for example scrutinize the size of buttons, in order to qualify the product's appropriateness for the elderly and the partially sighted.
Even if the design of criteria and trials for testing use is undoubtledly culturally and politically laden, one should not conclude that it is arbitrary. Consumerist actors claim to have a privileged access to the consumer's irreducible point of view concerning use. Hence, although they share this know-how with other specialists – manufacturers, standardization engineers – they downplay the abilities of these other actors to interpret and assess the consumers’ view correctly. One may ask, then, what kind of approach supports the correct qualification of use. Indeed, the notion of usage is at least as general and as cumbersome as that of ‘quality’. The measurement of object properties and the identification of scenarios of predictable interactions with users will not be enough to qualify use (Thévenot, 1993). It appears that a peculiarity of the consumerist approach may lie in a comprehensive approach to what constitutes the context of use. It rests on a sound knowledge of the entire trajectory of the user-product relationship:
‘We put ourselves virtually in a user's position and imagine the context or contexts of use of the product. We imagine the trajectory of the product, that also includes the phase of the intention to buy, the purchase itself, the installation, the putting into service, the use, with different modes every time, and then even the scrapping of the product or the failure scenario.’ (Interview with the test manager of Que choisir.)
Consumerists follow the biography of the good after it is bought and decommoditized (Kopytoff, 1986) in order to identify relevant trials to which it will be submitted. The approach assumes certain kind of projective attitude into the user's position, the legitimacy of which is largely based on the fact that everyone can naturally accede to that position (after all, are we not all users?). Knowledge mobilized to construct the tests, however, goes far beyond this intimate knowledge of users’ practices. Here again, the information collected during the market analysis phase plays an important part because this makes it possible to identify relevant non-trivial criteria or, on the contrary, useless criteria. As the test manager at Que choisir noted, ‘when you know that do-it-yourself tools are used for a few minutes a year only, you don't expect a drill to work for several hours continuously in the endurance tests’. It is through this type of investigative process that the test protocols emerge. The most interesting evaluation criteria and associated tests will be determined at each stage. 12 Tests for evaluating criteria may vary fairly widely and the whole paraphernalia of trials in quality evaluation can be mobilized: standardized tests, determining the physical properties by means of a measurement tool, appealing to individual experts (eg, for ergonomy), collective expert or to user panels (eg, for the sound quality of a mobile phone), and so on.
Let me note, finally, that the consumerist tradition has a longstanding experience in the field of services – insofar as the service can be reduced to the set of contracts and rates linking the consumer to the service-provider. Years of studying service regulations and struggling against unfair conditions have helped shape a set of skills that can be applied to widely diverse sectors (banks, insurance companies, public services, etc.) and are one of the main drivers of court action to defend consumers. The most useful tool here is the double-entry comparative table that summarizes, for each supplier, the proposed services along with their costs and a qualification of the related contractual commitments (eg, minimum length of subscription). The limit of this approach is that it addresses more directly the promises of services rather than their quality as such, as can be seen in tests devoted to mobile telephony or Internet access plans. The comparative tests that Que choisir and 60 millions have run on these subjects in the period under scrutiny have focused mainly on product evaluation, with a secondary, qualitative or exploratory analysis of related services (hotline, on-line services, etc.) or, on an even more complex subject, the quality of the telecommunication network strictly speaking.
Indicating the choice or choices
Consumerist performance testing is commonly seen as an exercise aimed at selecting the best product on the market. Yet carefully reading the consumerist press suggests a less clear-cut reality. A performance test does not look like a demonstration that is able to reveal the best of all products and to calculate the optimal quality/price ratio. In fact many articles show a relative indecision, or tension, between two ways of accomplishing rational choice from the consumerist standpoint: as a) the discovery of the best product or service (‘given a series of criteria, there is an intrinsically best good that wins’), or as b) the identification of the plurality of possible answers to the consumers’ needs and desires. The magazine articles therefore often propose contrasting approaches to the issue of choice. It is interesting to look carefully at the ‘literary devices’ that these articles mobilize in order to outline the test results, in the same way as Shapin and Schaffer (1985) examined the experimental accounts written by Robert Boyle to communicate the facts discovered within his laboratory to the external world.
Let me examine how the best choice is presented. The most usual device is a ranking of products and the attribution of grades to them, for instance on a scale from 0 to 20. Such a system indicates an unambiguous order of preference. In the two magazines considered here, the grades are calculated on the basis of a linear weighting, with the use of restrictive criteria: a bad score for one of the criteria – eg, a safety flaw – brings down the overall grade. In this type of approach, setting the respective weight of each criterion has significant consequences for the results. Grades and ranking may be altered substantially simply by changing the balance between the different parameters. How are the underlying trade-offs made? It seems that for each of the two magazines, principles of weighting are adopted prior to the tests and their relevance is re-examined and adjusted ex-post. Depending on the grades and ranking obtained for the different products, the engineers may discuss their validity and perform computer simulations to test their sensitivity to the weighting coefficients. This pragmatic approach, which relies on the testing professionals’ experience and know-how, allows them to qualify the robustness of weighting functions and to discard arbitrary results.
Other literary devices bring into play more circumstantial situations of choice. Test managers confess they would like to have a more flexible medium than the magazine, allowing them to take into account the variability of consumer preferences. They imagine interactive terminals in public places or internet websites where consumers could themselves express their preferences and obtain a ‘personalized best choice’ in return. A very traditional medium of communication, the magazine nevertheless affords certain opportunities to take this variability into account. Two different kinds of literary devices are used, one being centred on the product and the other on the consumer. In the former case, the magazine indicates the product(s) that, notwithstanding the grade obtained for the test, get(s) the best evaluation according to a particular criterion. The diversity of possible choices relates here to the diversity of criteria that can be conceived of to qualify the product. In the latter case, the readers are given a choice of user profiles, each of which is associated with a suitable product. For instance a section devoted to choosing the best mobile phone rate proposes several use scenarios similar to the following one: ‘On average, you phone twice a day, essentially for family reasons. And you like to be contactable all the time’. For each scenario, the reader is prompted to choose the corresponding rate. The idea here is to offer the reader a figure for projecting his or her own identity or needs so that questions on the criteria themselves no longer need to be answered.
What conclusion can we draw from the fact that the magazine resorts to various literary settings in the publication of the consumerist experiment's output? For test managers, this indecision is a response to the readers’ various expectations. There is, on the one hand, the feeling that ‘consumers like clear, straightforward advice’ and, on the other that, ‘depending on the consumer, certain criteria may be more important than others and that one should not single out a given product’. What is important to me is that test managers and above all journalists seem to have a deep understanding of the cognitive processes that underlie the consumer's making of choice. They know that readers have unequal interpretative capabilities and that information must be transmitted through distinctive formats. Moreover, they rightly feel that the proliferation of mediations is an efficient way to shape the judgement concerning a good's quality: a persuasive prescription is not always the one that says ‘Buy this product!’ Instead, they prefer juxtaposing and illustrating various forms of embodiment of the relation between the consumers and the goods. In some sense, this plurality ratifies the idea that consumption involves different ways of filling in the gap between evaluation (product qualification) and action (buying).
Keeping manufacturers at a distance
Consumerist testing has developed an attitude of keeping distance from manufacturers. Although consumerist magazines need the cooperation of firms to obtain information on products and services, they are very careful not to be taken over by the commercial point of view. When results are received, they follow strict protocols in order to involve manufacturers without letting them have the last word. Before the final publication, each manufacturer will receive the results of the tests concerning its product – but it will not by privy to those of competitors. If ever a dispute arises concerning the results, possible explanations are sought together with the causes of divergence. Whenever manufacturers’ arguments are acceptable the magazine may – although rarely does – publish a comment that in light of these arguments, puts certain results into a different perspective.
The aim of this preliminary procedure is, among other things, to avoid unpleasant surprises at the moment of publication. Situations of disagreement may persist, however, and may give rise to varying degrees of conflict: verbal dispute by phone, mail, registered letters, requests for court action to suspend the publication, or even outright prosecution. The latter occurs only rarely, which highlights the importance of a dialogue with manufacturers in the preliminary information phase. 60 millions de consommateurs has been sued ten times in its 20-year existence and has lost only once. Another area of tension between consumerist magazines and industry relates to the appropriation of comparative test results by manufacturers for promotional purposes. In the consumerist world two policies exist in this respect, found in the two French magazines considered here.
Que choisir (like the British magazine Which?) does not want its results to be used by manufacturers and refuses to grant authorization to be quoted in this respect. According to the test manager at Que choisir, this deontological rule helps avoid having consumerist information misused in a context over which the magazine has no control:
‘We always remind our readers that the test results are valid only when they are published and during the period that just follows: their lifetime is short precisely because in the market products largely evolve with time.’ (Interview with the test manager of Que choisir.)
What is at stake once again, is the capacity of the goods tested in the magazine to represent accurately the goods that the consumer will actually find in the shops. Here the consumerists wish to distinguish themselves very clearly from a certification process, which usually has more resources to check up on the compliance of products with standards over time. I shall add that Que choisir's policy on the question of quotation probably also originates in the necessity of preserving their image as observers untouched by the vicissitudes of commerce. Such purification clearly contributes to the credibility of consumerism as an institution that is uncontaminated by the interests and strategies of economic actors. 13
In contrast, 60 millions de consommateurs (like the German magazine Test, associated with the Warentest Foundation) does allow its results to be quoted by producers. This position relates to a more comprehensive policy implemented since 1987 by the INC (Institut National de la Consommation), a policy which, according to the test manager, had not really proved successful and was scheduled for amendment at the time of the survey. The original purpose of this policy was to reinforce the authority of the INC's as a test centre that would be also respected by manufacturers through a more formal testing approach. The model used was German, where the periodical Test had a very wide readership and where test results were quoted extensively by manufacturers. The system, however, was not very successful in France. In a spirit of control over the information supplied, the way to proceed in order to quote test results had been codified very precisely in a standard of the AFNOR (Association Française de NORmalisation, the French standardization organization). The manufacturer was allowed to mention only an overall qualitative assessment based on the comparative test result, without any indication of grade or rank and without any reference to rival products. This system proved unattractive to manufacturers who were unable to incorporate it in their promotion:
‘If we wanted to subvert the system it would be easy. We receive requests daily, to quote results, all the time. But the manufacturers are limited by the little normative information sheet, which isn't useful for advertising, it's not promotional enough’ (Interview with the test manager at the INC.)
Moreover, certain forms of competition with certification institutions (which were being reorganized in France at about the same time) probably contributed to weakening this device. Consumerist testing was confronted here with quality assessment as it is viewed in the certification system – a system in which manufacturers are more directly involved.
Conclusion
The sociology of the consumerist work outlined in this chapter is based on an analysis of one of its most emblematic activities: comparative testing. The chapter aims not so much to describe in itself the social and organizational context in which the defenders of consumers act, as to demonstrate the processes through which a very particular representation of consumption choices is produced. It is by reconstructing the challenges and practical constraints governing the fabrication of consumerist information that the specific character of this information becomes visible. The hypothesis followed throughout this chapter is that in performance testing there is as a dynamic of disengaging the consumer from the market. I will briefly sum up the different efforts that consumerist actors have to make to produce a situation of disengaged choice.
As the survey presented shows, consumerists perform work similar to that of the marketing specialists, but turned the other way around. To be in a situation to reveal what can be done with products and the cost one should pay for them, it is first necessary to take the place of the usual spokespersons of the modern market. Consumerist actors accuse those who manage supply of being much too engaged in the market to be considered as reliable prescribers of the best choices. In order to make this denunciation legitimate they need in turn to constitute their own credibility. Deposing of the established spokespersons is a laborious task – it is the alpha and the omega of journalists’ critical work. But instating themselves as credible spokespersons is equally difficult – and the engineers’ contribution is essential. This is why it is important to recall that the goal of reflecting the specific point of view of the consumer is far more than a mere claim or a formal justification of consumerist discourse. The inquiry on the consumer's point of view is embodied in a series of procedures and practices, and supported by a multitude of operations and know-how constituting the trade of performance testing specialists.
The investigation suggests that the task of testing is a somewhat exhaustive one. Firstly, it implies a process of collecting and centralizing data about the market which is not trivial for a number of reasons: because there is, a priori, no pre-existing standardization procedure nor place for the concentration of data relative to the market; because the manufacturers cooperate to a greater or lesser degree in the transmission of important data; because it is necessary to check the veracity of producer's statements through field inspections at retail shops. Drawing on this information, consumerist actors re-order and re-describe the sometimes messy sphere of supply – that is, they animate it with categories of their own, which are not systematically borrowed from the definitions that producers give.
Secondly, performance testing involves the constitution of a link of representativeness, of drawing similarities and the possibility of projection between this sphere and a particular place of experiment. This requires identifying and collecting entities that can prove fragile, unstable and sometimes short lived: that is, the economic goods. At this stage of testing, the definition of what constitutes the goods will be critically examined and the way in which consumers segment the market through their consumption habits will be taken into account.
Thirdly, consumerist actors carry out the attribution of qualities to the goods through an experimental process which is quite sophisticated and which intertwines various points of view. As opposed to technical definitions of objects and the anticipation of buyers’ assumed needs, consumerist actors seek the actual use of goods, considered as a point of access to their ultimate truth. They do not systematically restrict points of view on the question of quality and can be forced instead to multiply them, perhaps to the extent that uses and consumers are multiple. As a whole, the consumerist evaluation of products and services is not simply a trial of objects, in charge of distinguishing their intrinsic characteristics. It aims to construct differences between goods, through a relatively detailed exploration of their worlds, a testing of their qualities, a redefinition of their shapes and extension, and a composition of criteria of judgement.
In so doing, consumerists end up representing the issue of choice in a way that is quite distinctive, when compared against other representations provided by market professionals. As a representational device the comparative test is at first sight very similar to many other tools used in the organization of trade and in the mise en scène of competition (catalogues, prospectuses, supermarket shelves, sales arguments, specialized press benchmarks, etc.). As our study shows, however, its elaboration incorporates a whole series of situated hypotheses, decisions and values. In this sense, performance testing offers us a striking example of a socio-technical process that makes possible the attribution of particular qualities to consumer goods: even if these qualities and the logic of attribution are, as always, highly negotiable, they are not contingent in the sense that they embody a specific politics of consumption characterized by the figure of a consumer disengaged from the market.
This approach naturally raises questions about the impact of tests, from the point of view of their reception by the general public, by political decision-makers, and by the suppliers themselves. It seems likely that, apart from a more or less hypothetical impact on the general public, comparative tests have ‘indirect’ effects of prescription on various actors of supply, especially distributors and manufacturers. For instance, the survey enables us to assume that, as with the mobile phone, telecom operators, anticipating a potential effect on buyers, are attentive to the results of consumerist tests when they have to select the models around which they plan their packages. More in-depth investigations should be undertaken fully to answer the question of impact. To conclude, it should be noted that the research I have presented points to another direction for research, stemming from the notion of disengagement that I have placed at the heart of my reflection. There is no operation of disengagement that does not imply other forms of engagement. It seems particularly interesting to investigate how the move from disengagement vis-à-vis the market to modes of more political engagement typical of consumerist associations takes place, for example towards defending the interests of consumers or of the environment. An analysis of forms of criticism and denunciation of inequalities or imperfections of markets – which, I noted earlier, constitutes the work of consumerist journalism – seems to be an interesting starting point to address this issue.
Footnotes
1
An earlier version of this chapter has been published in French (Mallard, 2000). The present version has largely benefited from the editors’ comments.
2
In France, three magazines are concerned: Que choisir, 60 millions de consommateurs and Test Achat (the latter is available only to subscribers). Throughout this chapter I use the term ‘consumer magazine’ to refer to these particular magazines and not to the vast range of magazines and newspapers that publish reports or special sections on consumer products.
3
The persistence of this frame of analysis and its relative sturdiness probably owes a lot to the deregulation of markets and public services in the UK in the 1980s. This trend might have resulted in patterns of relations between buyers and sellers more strongly characterized by the power struggle than is the case in other national contexts.
4
The theme of the consumer's uncontrollable desire and temptation generated by the market environment is recurrent in the consumerist literature. Aldridge refers to a set of practical advice given by the magazine Which? in an article bordering on self-parody, to help consumers resist the sirens of consumption and to not allow themselves to lose control over their own actions: keep strictly to your shopping list, wear earplugs so that you don't hear the coaxing music, walk to the supermarket and take a basket rather than a trolley physically to limit your buying power, avoid taking along children who will push you to buy, do take along a stick to help you to reach the cheaper products placed higher up in the shelf space, etc.
5
It is of interest that based on a similar uncoupling of situations of choice and consumption, we find here the exact opposite of the difficulty experienced by promoters of e-commerce trying to find ways to engage the virtual consumer. In an e-commerce situation the consumer can leave the transaction framework at the click of a button. With the current upsurge of e-commerce, the role of various intermediaries in consumers’ engagements in market transactions is being re-examined.
6
Throughout this chapter, I will emphasize the many convergences that exist between the two magazines investigated (Que choisir and 60 millions de consommateurs), without examining in detail significant points of divergence that would need further work.
7
For instance, a controversy was started in the magazine Que choisir on tests performed by the FNAC, one of the most powerful retail companies for technology products in France. The controversy revolved around the question of the compliance of the FNAC testing protocols with the methods recommended in the norm.
8
When the survey was made, Que choisir employed seven engineers and 60 millions de consommateurs six, specialized in various areas of consumption (agri-food, chemicals, household appliances, electronics, etc.). The five individuals that I met, who all worked on telecom products, were qualified engineers and had 5 to 12 years of experience in this field.
9
It also stands to reason that the socio-demographic profile of the readers of comparative tests plays an important part in their impact. As noted, I excluded the issue of tests results’ reception from this study. It is nevertheless possible to clarify some of the conditions. According to the journalists, most Que choisir subscribers are middle-aged or retired, with average or substantial buying power. There is a large percentage of teachers and middle and senior managers, and more home owners than tenants. Most live in the provinces in average-sized towns, as opposed to Paris and other cities.
10
Use-related aspects receive more attention when dealing with new products. Note that the propensity to represent or promote new uses (found far less in consumerist magazines than in specialized ones) differs in the two magazines concerned. At Que choisir the journalists’ mission seems to be less to provide information about new products (they consider that suppliers make enough of a fuss on their own) than to alert the readers of fraud and abuse hidden behind novelties. In 60 millions de consommateurs, in contrast, articles fairly often gently promote new products and innovations.
11
In this affair, one of the most sensational in the history of US consumerism, Nader, a lawyer, accused car manufacturers of having a criminal attitude since they sold vehicles that had not been designed to guarantee their drivers a minimum of safety.
12
Other requirements also play a role in designing tests for a collective of magazines. This is the case for ICRT (International Consumer Research and Testing), a consortium of European organizations to which the magazine Que choisir belongs. The issue of the cultural relativity of the definition of quality might be raised, as this definition is embodied in the diversity of national behaviours and habits. Although this applies in any branch of activity, it is obviously highly sensitive in such cases as food technologies (eg, the French, Italians and Swiss easily agree on criteria for testing coffee percolators, but agreement on the subject is more difficult to reach with other European partners). In a test on mobile phones, for instance, the importance of the ‘low battery’ alarm was a subject of disagreement between ICRT participants.
13
In the same way, Aldridge stresses that the magazines refuse to endorse the manufacturer's sales strategies, due to a form of pure aversion to the market. He also suggests a complementary, more critical analysis: consumerist magazines have to ensure themselves of a monopoly on the distribution of their own information, to guarantee their independence based essentially on income from sales.