Study title

Brazil, China, South Africa

Individual

Organization

Household

In Phase One, in each country we met with 30 food system actors including people working in government, civil society, wholesale, retail, hospitality, public procurement, NGO’s, production and campaign groups, influencers and celebrities. Each interview explored the background and role of the interviewee in the organisation; their understandings and criteria of sustainable and ethical food; their understandings of middle class demographics and their consumption of sustainable and ethical food; organisation policies and practices as they pertain to sustainable and ethical food; and key influences and trends that they saw as influencing their organisation and wider publics. In Phase Two, we extended our research to food consumer/citizens. In each of our cities we interviewed 30 people from a range of middle-class backgrounds and followed up a smaller sub-set of 10 households from each case study city for ethnographic work. Interviews addressed household food consumption practices, judgments about ‘good’ food and influences on notions of ethicality and environmental and social values. Respondents were also asked to reflect on (and demonstrate) their online practices as consumer-citizens, including how they collect information, shop and/or review online. Ethnographic research involved ‘go-along’ (Kusenbach 2012), accompanied shopping interviews and co-cooking sessions to capture the complexities and nuances of food choices, judgments, engagement with government and corporate ethical initiatives, and the ordinary ethics of food purchase and use. As part of this phase, digital ethnography explored the role of social media and online tools as devices implicated in ethical judgement and the creation of markets for sustainable foods. In Phase 3, to evaluate the cultural influences, trends and technologies affecting ethical consumption of food in and across the three country case studies we undertook research focusing specifically on the digital sphere. Phase 2 produced culturally specific understandings of which platforms, apps and tools are relevant to consumers. Participants were asked: a) which top 3 search terms they use to look for information on consumption choices online; b) which top 3 key accounts (e.g. of campaign organisations) they follow on social media, and; c) whether they follow celebrities who might influence their buying decisions. These terms and account names guided text mining of social media in Brazil and South Africa to examine the traction of key values, actors and lines of influence. For a period of 12 months, cloud-based Discovertext software was used to mine the public Facebook and Twitter accounts of the campaign organisations and celebrity accounts named by participants (in English in South Africa and in Portuguese in Brazil). Government restrictions and technical differences inhibit access to social media for text-mining purposes in China. To widen the reach of the research, and produce material with traction with policy and commercial actors, in Phase 4 we built on the in-depth analysis of phases 1 and 2 to gather quantitative data to assess how values and behaviours identified in each country apply across a larger sample of consumers and identify values and behaviours transcending local context and scale. Quantitative data was gathered through a web-based survey of the drivers of sustainable food consumption and the behavioural intentions arising from these (e.g. intentions to purchase sustainable food). Comparable samples of the middle-class demographic in each of the three case study countries were selected to obtain a total of 1500 responses (500 per country). The questionnaire was developed in English (the most commonly-spoken second language and the lingua franca of the public domain) for use in South Africa and professionally translated into Portuguese, Cantonese and Mandarin using double-back translation procedures for use in Brazil and China respectively. For quality and cost effectiveness, survey participants in the three countries were accessed using Qualtrics.