Case Study: The Spanish Wine Industry

The definition of the sample universe is done through the combination of the following databases: Spain's existing public records in different protected designations of origin (DOP) and the information available in the database Iberian Balance Sheet Analysis System (SABI), registered in year 2015 with CNAE (National Classification of Economic Activities) code 11.02 "Winery Companies".

The number of independent entities resulting from these databases was 3286. Following Spanos and Lioukas, the authors dropped from the sampling frame firms with missing data, and the datum that resulted from a duplicate company with different location or brands without a formal structure. The total sample was then reduced to 2413 independent entities.

The questionnaire was conducted after an extensive review of the literature, and scales validated in previous studies were used. It focuses on the resources and capabilities that have been collected by theoretical studies and on the analysis of the competitive environment, business strategy and business performance. The questionnaire was also tested by previously sending it to various entities linked to the wine sector in Spain, associations, experts, as well as managers of wineries. The pretest was carried out in two phases. First a pre-validation was made with four winery managers and four sector experts, and then, a second validation phase took place with nine managers and three sector experts. As a result, some issues were modified in the questionnaire and the explanation of the different sections was expanded. The objective was to ensure that the questionnaire was understandable and that it reflected the peculiarities of the industry.

The process for collecting the data started in December 2015 and finished in May 2016. After sending the questionnaire via email to the manager of the different firms, the authors allowed for one month to receive an answer and if during that period it was not provided, a phone reminder was made. The final sample was made up of 339 valid responses, which meant a 14.0% response rate, similar to the amount reported by Baruch and Holtom, for the industrial sector, so that we considered it appropriate for our study.

The sample characteristics are given in , . Table 2 reports the distribution of the response percentages among the firm's type of ownership. With regard to the number of employees, its distribution percentages and the comparison with the mean of the sector is shown in Table 3. Therefore, non-response bias does not seem to be a major concern because of the data similarity between responses and sector.

Table 2. Response percentages in reference to the type of ownership of the winery.