Screening social life cycle assessment (S-LCA) can be applied to identify the social hotspots associated with a production activity or supply chain. The objective of this paper is to explore how the quality of the results of a screening S-LCA can be improved, illustrated by a case study of sugarcane production in Brazil, an activity which has been criticized for its records on social sustainability due to issues such as poor working conditions for field workers and treatment of migrant workers.

Cradle-to-gate production of sugarcane in Brazil has been modeled using input-output analysis. The associated social impacts have been modeled using the framework of the Social Hotspots Database (SHDB), which is one of the first databases providing information on social risks along supply chains. The results from the SHDB were complemented with results from a systematic analysis of relevant literature. Content analysis was applied to 38 publications in English relevant to the social impacts of sugarcane production in Brazil, including peer-reviewed articles, “gray literature,” non-governmental organization reports, and conference presentations. Qualitative data analysis software NVivo 8 was used to facilitate the analysis of the publications. A deductive category system was established based on the subcategories recommended in the UNEP/SETAC social life cycle assessment guidelines. Social impacts were further aggregated and analyzed by social themes and impact categories.

The social impacts of the sugarcane life cycle in Brazil arise almost exclusively within the Brazilian sugarcane sector itself. Fifteen social themes are identified as hotspots in the SHDB, and nine of them are also identified by content analysis. Health and safety and labor rights and decent work are the impact categories with the highest risks. Besides negative impacts, content analysis is capable of identifying several positive impacts related to sugarcane production. Due to the use of aggregated country-level data, social impacts of manual and mechanical harvesting of sugarcane cannot be differentiated in SHDB; however, this can be achieved by content analysis.

SHDB is effective for identifying social impacts at the country level, but the data are inevitably aggregated and only show averages across different technologies and geographical areas; therefore, the database is of limited value in distinguishing between alternative operations and locations. Content analysis can facilitate foreground data collection by differentiating operations and identifying both negative and positive impacts at the level of individual activities. We recommend that S-LCA databases can be integrated with results of content analysis to improve the richness—representativity and specificity—of results from a screening S-LCA, to differentiate between alternative production routes and processes.