3.1. The Study Area
The Salento peninsula is an area located in the southeast of Italy in Apulia Region (Figure 2
). It covers the provinces of Lecce, Taranto (the eastern part) and Brindisi (central-southern part), with a population of approximately 1.5 million inhabitants. Salento spans 5329 km2
and has more than 300 km of coastline along the Ionic and Adriatic Seas.
In recent decades, Salento became known at both national and international levels due to its beautiful coasts and numerous events and entertainments proposed along with its well-known historical and artistic heritage. The economy, once purely based on agriculture and artisanal fisheries, has experienced a significant increase in the secondary and tertiary sectors, making this area one of the richest in Southern Italy. One of the most important economic sectors is tourism. Tourist arrivals registered an 80% increase between 2002 and 2009, followed by a more moderate 10% increase between 2009 and 2015 (ISTAT 2016
). The tourism industry is especially intensive in MSW generation compared to other economic sectors, such as manufacturing or agriculture, and more prone to produce other kinds of polluting outputs (Mateu-Sbert et al. 2013
). Therefore, the relationship between tourism growth and MSW generation in Salento and related management strategies are worth studying for at least three reasons: (i) the development of the tourism industry has resulted in an increase in waste generation (UNEP/GPA 2006
); (ii) inadequate MSW management can bring negative effects on the attractiveness of the touristic area, reducing tourism inflows (Arbulú et al. 2015
); (iii) valorizing available waste materials, in circular economy models, allows for closing the loop not only material-wise but also energy-wise (Pan et al. 2018
). Due to these premises, the tourism industry in Salento represents an interesting case of investigation for understanding the socio-political dynamics based on experts’ insights and awareness in order to support a radical form of sustainability transition. Italy has recently taken steps in this direction with two second-generation biorefineries, namely Gela and Porto Marghera. However, they still represent a small industrial niche, facing strong socio-economic challenges (Imbert et al. 2017
). From this perspective, the tourism industry in Salento could represent an open-air laboratory for the application of the most advanced environmental and renewable technologies and become a frontrunner not only for Italy but also for the whole EU.
With the aim of determining the quantitative values of SWOT factors, the Analytic Hierarchy Process (AHP) or the ANP are the most suitable techniques (Saaty 1996
). The AHP is generally employed to determine the quantitative values for SWOT analysis, since it works on the idea that elements function independently of one another in a hierarchical configuration (Catron et al. 2013
; Saaty 2005
). This represents a severe assumption to meet, especially when the considered attributes become interdependent owing to a complex situation. Such a degree of complexity makes the ANP appropriate to study factors’ dependencies (Starr et al. 2019
). Assessing the conditions able to promote socio-institutional changes for a sustainable energy transition in Salento (e.g., second-generation biorefinery) includes a number of complexities and interdependencies involving different stakeholders. For example, urbanization, demographic trends and related socio-cultural changes will likely impact the waste management practices in the area. Accordingly, SWOT-ANP is the appropriate method for this analysis.
The methodological approach can be divided in two distinct phases:
Identification and selection of relevant SWOT factors by means of a literature review and expert interviews;
Prioritization of the internal and external factors identified through a survey administered to a variety of knowledgeable stakeholders.
In the first phase, a literature review was carried out by looking at two main databases of scientific literature, i.e., Scopus and Web of Science, to ascertain a list of relevant factors to be used in our investigation. A broad keyword search was conducted in order to retrieve relevant papers within the publication timeframe of 2015–2019. We paired some anchor keywords (i.e., “bio*,” “circular*,” and “sustainab*”) with search strings (i.e., “tourism”, “energy”, “transition”, “refinery”). Our in-depth literature review uncovered more than 150 papers engaging with the sustainability of the tourism sector and more than 20 regarding waste-to-energy transitions. With the aim of selecting the most relevant factors, we refined this pool of articles by carefully examining the text of each article in order to ascertain the presence of a well-defined idea or value judgement with regard to the area of investigation. In doing so, we employed the QDA Miner 5.0 software package (Provalis Research 2015
), which allowed us to perform a qualitative assessment of the context in which relevant keywords appeared in the selected documents. Table 1
reports some descriptive statistics about the documents analyzed and the relative keywords found.
We accessed 151 documents for a total of 1,583,418 words, with an average number of words per document equal to 10,486. The majority of documents refer to the sustainability and development of the tourism sector. In this framework, we found 847 keywords corresponding to an average of 5.6 per each accessed document. The word map below (Figure 3
) allows the visualization of the most relevant words employed in literature to characterize the sustainability of the tourism industry.
From the total number of words used in the 151 articles, the map includes only those terms which appear at least 5 times in the analyzed corpus of the single article. The bigger the letter size, the more frequent the word. It is important to mention that the term energy is in the middle of the map, and it is connected to all the focal points of the current research (biorefinery, development, tourism, circular), but also to other important aspects, such as: transition, policy, jobs, etc. In a further stage, with the aim of choosing the most relevant factors and summarizing them by way of a 2 × 2 matrix (internal factors: strengths and weaknesses; vs external factors: opportunities and threats), we conducted two interviews with two academicians (i.e., an agricultural economist and a commodity scientist) with long-term involvement (i.e., more than a decade) in the field under investigation. This allowed the labelling of the factors retrieved by means of literature review as internal and external to the tourism industry.
In the second phase, and building on the protocol followed in Starr et al.
), a survey was developed and administered to a group of knowledgeable local stakeholders. A larger group of experts was identified, starting from a preliminary list of actors derived from the Italian Association of Tourism Professionals and Cultural Operators (AIPTOC). The association has more than 300 effective members covering different categories (e.g., managers, researchers, evaluators of management systems, consultancy companies, institutions and trade associations, etc.). Successively, considering the information collected by means of websites, technical reports and blogs, we refined the list by focusing only on actors with long-term involvement (i.e., more than a decade) in the field under investigation for the selected study area. Interviewees were selected with the intention of representing a wide range of actors involved in the tourism industry. In particular, the group of experts taking part in the survey were: two tourism industry professionals, a trade association, two representatives of a consumer association and environmental associations, a local policy maker, and two researchers. The eight interviews were conducted by telephone over the period of February to April 2019, and lasted approximately one hour. Respondents were asked to make several pairwise comparisons between the identified SWOT factors using a scale suggested by Saaty
). The scale ranges from equal importance (participant assigns a numerical value of 1) to extreme importance (participant assigns a numerical value of 9) of one element over another. After comparisons between each factor within the SWOT categories were made, comparisons between each category were made by employing the same protocol. Therefore, two matrices were administered to respondents: (i) pairwise comparisons per group (Table 2
); and (ii) pairwise comparisons of the groups (i.e., strengths, weaknesses, opportunities and threats) (Table 3
In Table 2
and Table 3
, F1,..., F5 are the identified factors, G1,…, G4 are the SWOT groups, VF2F1
represents the value of factor F2 with respect to factor F1 (the same logic is applied to all factors), VG2G1
is the value of group G2 with respect to group G1 (the same logic is applied to all groups), SCF1
are the sum of values regarding the columns of group F1 and G1, respectively (the same logic is applied to all groups)1
The local factor priority is obtained evaluating the average values of the expert comparisons among the factors in the same SWOT group. Meanwhile, the group priority is based on the average of the expert comparison among all groups. The global factor priority of all SWOT factors is calculated as the product of the local factor priority and the respective group priority.
The two aforementioned phases are common in both AHP and ANP procedures. However, to appraisal the interdependence between SWOT factors, an additional analysis is needed. Table 4
was employed to weight the interdependence of each category. For example, respondents were asked to consider how strengths may be used to mitigate weaknesses or enhance opportunities (Catron et al. 2013
; Starr et al. 2019
In Table 4
represent the factors’ interdependence. They measure the relative importance of weaknesses, opportunities, and threats in enhancing strengths, respectively, and IG1G2
represent the relative importance of strengths, opportunities, and threats relative to mitigating weaknesses and so forth.
The global priority value based on factor interdependence, for individual SWOT factors, can then be calculated as: global priority of factor Gij = priority value of factor Gij * (interdependent scaling value of SWOT category).