Two main conclusions can be extracted. Within the theoretical framework of convention theory, it is possible to use quantitative methodology to investigate the quality claims of agrifood products. Second, such methodology permits the analysis of a wide range of cases and the possibility of obtaining a joint vision of the dominant orientations in terms of quality.
Quality conventions and quantitative methods – a new research line
The literature review in Section ‘Convention theory in the research on the agri-food system’ has shown that research on quality conventions in the agri-food system is based on qualitative methods and case studies. Quantitative methods are virtually absent, and information obtained from statistical sources is barely present. Semi-structured interviews and other qualitative methods are especially desirable in regard to identifying social and cultural attitudes, as well as to reveal the practices of agents and agreements established among them. This aspect is how their systematic use in the study of quality conventions is explained.
However, statistical data can also be useful as indicators of certain attitudes and relationships; for example, consider two wine producing areas, one of which uses a variety of native grapes and other of which uses global varieties, that is, the most popular varieties in the world market. It is logical to believe that the native grape wine producers seek quality that is local and traditional, that is, they are guided by domestic conventions; in contrast, the latter seeks quality by trying to adapt to the tastes of a majority and cosmopolitan consumer who demands wines from global varieties. It could be argued that the latter is guided by commercial conventions. Similarly, it is logical to think that an area in which many producers have obtained ecological crop labels can understand quality in the context of sustainability criteria that are characteristic of civic conventions. It is therefore possible to identify the existing categories of quality conventions from properly chosen statistical indicators.
Quantitative indicators do not allow the recognition of concrete agreements between producers, consumers and other agents involved in wine quality evaluation; qualitative methodologies are most likely irreplaceable for acquiring such knowledge. However, quantitative indicators themselves show the existence of certain generalised practices that could be seen as the result of agreements between producers to satisfy consumers’ expectations. In that sense, if a PDO region earns many awards in international wine competitions, for example, it means that wineries of that region include in their common practices the participation in competitions. This practice indicates that the wineries pay attention to the trends marked by the juries, to the work of other participants – their competitors – and to the expectations of consumers whose preferences are influenced by such competitions results, which implies an adherence to certain rules, explicit or implied, required to achieve the recognition of an award and obtain the consequent reputation that such an award provides the company. As a result, given that conventions are practices, routines, agreements and rules (Salais and Storper, 1992: 171), quantitative indicators have much to contribute to the knowledge of conventions.
In the third section we have developed a series of mathematical indicators that were constructed from statistical sources to enable the identification of existing types of conventions and to evaluate and measure the relative intensity of each type in large groups of companies in the wine subsector.
A single indicator may not be enough. For example, the fact that most of the wineries in an area are native points to domestic conventions because entrepreneurs have strong roots in the area and are best able to learn and develop traditional ways of making wine; however, it is possible that these entrepreneurs have removed themselves from tradition, adopting other conventions of quality. Therefore, a single, isolated indicator may not be useful. On the contrary, various indicators that are valued together will suffice. If an area has high values for several indicators in a convention category, its average score will be high in this category, indicating that the quality is defined by that category.
Two final remarks should be added. First, it must be emphasised that an alternative to qualitative methodologies of quality conventions research within the agri-food sector was not sought by this study. Instead, the purpose was to show the possibility of using some quantitative methodologies which have not been used to date. In the social sciences, both methodologies are complementary and their joint use reinforces research reliability.
Second, this work refers to a specific subsector, but could be applied to the whole agri-food industry. Some of the indicators may be directly transferable to all or many of the subsectors, while others only apply to wine. Examples of indicators that are easily transferable to other subsectors include 3, 7, 11, 15 and 18, replacing “wine” with the product and “winery” with “company”. In any case, each subsector of the agri-food industry has its peculiarities, and a table of specific indicators would have to be designed.