Hierarchization of social impact subcategories: towards a systematic approach for enhanced stakeholders’ representativeness

Abstract

Abstract Purpose Social life cycle assessment (S-LCA) aims to assess the potential social impacts related to stakeholders over the life cycle of a product or service. For legitimacy and meaningful results, direct consultation of stakeholders ranks among the most recommended approaches. This paper aims to provide the methodological basis for S-LCA to target potential impacts and to support decision-making using this kind of participatory approaches. In particular, the work aims to address some of the limitations of the systematization of stakeholders’ consultation. An approach to facilitate and speed up the access to stakeholders and the construction of respondent panels is proposed. Then, representativeness of the collected answers is verified using a statistical data treatment. The method is applied to hierarchize social impact subcategories in the offshore wind energy sector, a huge up-coming sector in France. This emerging sector raises a number of socio-economic issues that can be related to the development of a new industrial sector and its coexistence with local communities. Methodology Based on the participatory approach principle, the hierarchization of social impact subcategories is carried out by stakeholders. The developed methodology includes 5 steps. In step 1, the social impact subcategories from the UNEP in Guidelines for Social Life Cycle Assessment of Products and Organizations 2020 (2020) list are adapted to the sectoral context. In step 2, the hierarchization criteria are defined. Instead of using a ranking based on an importance criterion, hierarchization is based on two quantitative criteria to target impact subcategories that are both important and perceived as potentially problematic. In step 3, the stakeholders and a sampling approach are defined. Then, in step 4, an online survey consultation methodology is used and improved for the selection of qualitative variables. Finally, in step 5, the methodology specifies the data treatment protocol. The data treatment protocol in this fifth step aims at addressing the issue of the representativeness and relevance of the responses obtained from surveys. Indeed, hierarchization approaches based on consultations typically consider responses at the aggregated level of the stakeholder category. However, it is likely that different stakeholder profiles of respondents within a large heterogeneous stakeholder category influence the perception of social impact subcategories. To verify this point, it is necessary to look at a disaggregated scale of stakeholder sub-groups. This potential bias led to the need to adjust the survey responses. Results and discussion Large-scale sampling allowed us to collect 82 responses from value chain actors and 50 responses from local community with a respective response rate of 13% and 16%. Firstly, hierarchization of social impact subcategories was possible at the level of the whole aggregated stakeholder category. Then the disaggregated level was considered. To do so, qualitative data in the surveys allowed different profiles within a stakeholder group of the panel to be identified. Then, chi-squared tests on a representative variable were conducted and an adjustment of the responses and, therefore, on the resulting hierarchical order of social impact subcategories was applied. The study of the disaggregated responses led to the identification of a significant dispersion of the responses and the influence of certain variables of the respondents on their perception of social impacts. Conclusions Participatory approaches were found to be useful to legitimate the selection of impact subcategories when applying S-LCA. However, considering aggregated hierarchization results at the whole stakeholder category level may mask some polarized opinions within the same stakeholder category. An adjusted hierarchization can serve to enhance the representativeness of the consulted stakeholders’ perceptions. It would be good practice for the practitioner to highlight the limitations and possible biases. For this, one recommendation is to provide transparency on the dispersion of responses and disaggregated information on the stakeholder panels involved. With the proposed method, it was possible to both adjust the hierarchization results and express the residual uncertainty for the sake of transparency. The proposed method is designed to be transferable to any sector where stakeholders are assembled in sectoral clusters. We were able to access many stakeholders with different profiles. This broad sampling supports a holistic view of the social impact subcategories. The hierarchization results allow the practitioner to target a priority order to address the impacts subcategories for next S-LCA steps and to specify the chosen scope of the study.