Public perceptions of the UK food system: public understanding and engagement, and the impact of crises and scares

Lead Research Organisation: National Centre for Social Research
Department Name: Research Department

Abstract

We propose addressing each of these research aims through a three stage sequential mixed method design. Each element has its own aims but also feeds into the broader research design.

1. Social media research

Cardiff University has been collecting a random 1% of all UK tweets since 2012. We will analyse the tweets collected during the period of the horsemeat scandal and review them to identify key search terms that can be used to identify other tweets that are likely to relate to the horsemeat scandal. We will then use this 'lexicon' to search all Twitter traffic from the time of the scandal and purchase all of the relevant tweets for one month following the start of the scandal.

We will use this data to map levels of activity and changing sentiment against a timeline of the key events. This will identify:

- Key announcements, information and media coverage which had the greatest impact on social media.
- Key messages, opinions or attitudes that developed at least partly because of social media.
- How sentiment towards the scandal and particular actors changed over time.

Secondly we will develop a network map showing how information flowed through Twitter, focussing on the producers and proliferators of key messages. This will include modelling the likelihood that particular opinions or information will be proliferated allowing us to:

- Understand how and why particular information (and misinformation) on the horsemeat scandal proliferated through Twitter.
- Identify key actors in bridging information between different groups represented on social media.

2. Qualitative workshops

We propose conducting six half-day workshops with a total of up to 300 participants. A purposive sampling strategy will be adopted, in order to capture the views of a diverse range of people in terms of five characteristics which are known to be linked to food behaviours: socio-economic status, age, household composition, diet and ethnicity. We will conduct them across the UK to achieve further demographic and geographic spread.

Each workshop will host 40 to 50 participants, include a mix of plenary sessions and smaller discussion groups and will be based around two themes:

i) Food supply chains, trust, responsibility and values including:

- How people understand food supply chains.
- Whether they consider food supply chains when buying food, what they consider, why and how they obtain the information required to do this.
- Who they trust for information about supply chains and food supply.
- Who they see as responsible for food information and safety.

ii) Managing food "scares" including:

- How and why public perceptions of food supply chains were affected the recent horsemeat scandal.
- How perceptions were affected by the actions of organisations involved in food supply chains, the government response, media coverage of the incident, and wider public debate.
- Who people see as responsible for managing food "scares" and who they trust in these situations, and what they expect from government.

3. Quantitative survey

We propose adding a module of 20 questions to the British Social Attitudes survey developed from the data collected during the other stages of the research. The exact content will be agreed once the results from these first two stages are known but may include:

- Levels of engagement with issues related to supply chains
- Key issues of concern regarding food chain management
- Who the public trust to manage food production and food supply chains
- What values should be applied to the management and regulation of food chains
- Views on food scandals and scares and responses to them

Data will be analysed against the full range of socio-demographic variables as well as other attitudinal data from other parts of the survey using a range of statistical methods, including segmentation.

Planned Impact

A range of organisations who would benefit from this research, and the specific ways in which they may use the data have been identified:

The Food Standards Authority (FSA):
This research will garner new insights into how people engage with and understand food supply chains; where there may be gaps in knowledge, information or understanding; and how this varies between socio-demographic groups. This feeds directly into the FSAs strategic objective to ensure the consumers have the the information and understanding required to make informed choices about what they eat. Further to this the interrogation of the impact and dynamics of scares or 'crises' in UK food supply chains will provide invaluable insight into how such events can be best managed in the future.

Government and other policy makers:
The benefits to the FSA will also be shared with a range of other government departments and policy makers, including departments like DEFRA, the Department of Health and Public Health England whose interests overlap with those of the FSA. The insights gained from this research will help departments such as these better understand patterns of consumption as well as how the public respond to food scares, which often have health dimensions. Furthermore it will show, with great depth and detail, how this varies across different groups allowing future interventions to be targetted for maximum efficiency and effectiveness.

NGOs and charities:
Our scoping work and wider discussions with NGOs and charities have demonstrated that the proposed research would be valuable to the work of organisations in a range of different areas. These include organisations working on areas related to sustainable production and consumption, food security, consumer rights and information as well as a number of others. This research will provide new insights into how the public understand and engage with the food supply chains including identifying the current extent of understanding and engagement, key knowledge or awareness gaps, where more information is required, how this information should best be presented and communicated (both generally and during particular 'crises') and how this varies across different socio-economic groups. This will allow NGOs and charities to develop updated approaches to engaging with these issue and to better target campaigns, information or interventions where they are most needed.

Private sector:
The private sector and particularly large food producers, retailers and supermarkets play a key role in design, maintenance and management of UK food supply chains. Among many producers, retailers and supermarkets there is an increased emphasis on where food has come from, whether it be related to 'buying British', supporting local producers, reducing food miles, or simply knowing where food has come from (which often links into debates about animal welfare and the use of chemicals in agriculture). The insights from this research will help these private sector actors understand how consumers engage with the issues central to these initiatives which in turn will allow them to better tailor, market and manage them in the future.

Advisory group:
As detailed in the Pathways to Impact attachment we will set up a stakeholder advisory group which will in part be used to ensure that the outputs from the study are relevant and can be used in a way that will ensure impact on policy and practice.
 
Description Opportunity to view themes across different methods

This project had three distinct components: Twitter analysis, qualitative workshops and a representative quantitative survey. This offered a unique opportunity to explore meta-themes or threads that cut across the findings from our different methods. The qualitative work was the richest data source, but this was an opportunity to see if social media (specifically Twitter) had the potential to provide a 'parallel epidemiology' of public concerns about particular food incidents.

In the end we looked at three threads in detail that related to our project aims:
1) Public understandings and concerns
In the qualitative workshops, in the absence of a food scare or other trigger, the food supply chain (FSC) was not a high priority for participants and their understanding was often vague. Furthermore, the survey data showed that people were most likely to say that healthiness matters to them when buying food and less likely to rate concerns around the FSC as important. Women and older people were more likely to be concerned about issues around the FSC.
2) Trust and confidence
Workshops participants saw responsibility for FSCs as shared by all, but with different actors having differing levels of trustworthiness. The survey data showed that only a third of consumers trusted the government to make sure food is safe to eat and that people trust food from Britain more than food from abroad.
3) Public concerns and perceptions of risk in the event of a scandal/scare
Analysis of the Twitter data from the 'horsemeat scare' of 2013 showed that tweets did capture the distinctive "risk signature" of horsemeat i.e. the transgressive humour and that the main concern was about the breaching of a cultural taboo about eating horses. Twitter did not pick up a constant theme (that had emerged from the workshops and survey) which was the break in trust between consumers and suppliers/supermarkets. This loss of trust was described by many as having been profound at the time and added to their sense of ongoing mistrust in provenance, labelling and the general integrity of the FSC.

How Twitter could be used in future food scares

Statistical modelling of Twitter data allowed us to explore what made a tweet more likely to be retweeted during the 'horsemeat scare'. Who tweets is important i.e tweeting from the FSA @foodgov account significantly increased the retweet likelihood, as did tweeting from a mainstream media account compared to everyone else. Tweets produced at night survived longer than tweets produced at any other time of day, which may be explained by people having the time to check their Twitterfeed and retweet content when they wake up.
We also asked people in the survey if they had a Twitter account. This is the first time researchers have been able to start estimating which groups can be reached via this media. UK Twitter users are not representative of the wider UK population: they are more likely to be young and male and from higher social classes.
Exploitation Route There are key themes from the findings that speak to a number of the Food Standards Agency and other Government agencies' interests and priorities around different consumer concerns and behaviours, engagement on issues of the food system, importance of trust, response to an incident and longevity of effect, lessons in using social media tools and lessons in pulling different strands of evidence together. There are already plans to incorporate relevant findings from the project into the Food Crime Annual Strategic Assessment and key initiatives like Our Food Future.

We now have a much better idea of who uses Twitter following the survey questions and some of the characteristics that increase retweeting or survivability are not intuitive and may have a significant impact on social media communication strategies for organisations like the FSA in the event of future scares. Links to practice could also be explored.

From our conversations with the media, there is particular interest from retailers around how they can build up trust with customers and what to do in the event of a food scare. There are lessons that can be learned from our findings around labelling, being seen to put consumer interests first and knowing their shopper.
Sectors Agriculture, Food and Drink,Digital/Communication/Information Technologies (including Software),Retail
URL http://natcen.ac.uk/our-research/research/understanding-the-challenges-of-the-food-system/?_ga=1.49524006.1471327549.1435248982
 
Description BSA media 
Form Of Engagement Activity A press release, press conference or response to a media enquiry/interview
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Media (as a channel to the public)
Results and Impact Press release and subsequent media coverage following chapter published as part of British Social Attitudes. This included interviews with FoodQualityNews.com and Retail Express. We sent e-flyers to civil servants, trade bodies and senior people at supermarkets and we sent out some infographics on Twitter (@natcen).?
Year(s) Of Engagement Activity 2016
 
Description ESRC DISSEMINATION EVENT Belfast 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Postgraduate students
Results and Impact As part of the dissemination of their project, Queen's University Belfast held a dissemination event. We presented under two sessions: 1) how to design and get data on consumer perceptions and behaviour and 2) results from our project. The audience was made up of students learning about data collection methods and representatives from the public sector.
Year(s) Of Engagement Activity 2017
 
Description Ethics in social media workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact The workshop was held at the University of Aberdeen called "Social Media Privacy and Risk: Towards More Ethical Research Methodologies", which is part a grant from the Economic and Social Research Council "Social Media - Developing Understanding, Infrastructure & Engagement (Social Media Enhancement)" Aim was to co-produce a set of ethical guidelines around the use of social media data for social research. The workshop was predominantly attended by academics including our collaborator Dr Luke Soan from Cardiff and one of our project researchers. Our project came up a few times as an example of using social media data to understand public attitudes and was discusssed in relation to issues such as sharing data (Twitter T&C's of data sharing), anonymization (or the fact that you're not allowed to anonymise Twitter data when reporting individual tweets) and the right to withdraw (once Twitter data is collected via purchaser there is no mechanism to allow users to delete their data).
Year(s) Of Engagement Activity 2016
 
Description FSA Comms 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact This was a presentation to the Food Standard's Agency's Communications team of findings from Twitter analysis. The purpose was to show results from modelling Twitter data from the horsemeat scare and how this might inform the way in which FSA could manage future food scares e.g. time of day to tweet. The findings were extremely well received and we understand that they may be applied in a future crisis.
Year(s) Of Engagement Activity 2016
 
Description Royal Statistical Society International Conference 2016 (Manchester) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Invited presentation for panel titled "Social Statistics: Paradigms in non-random sampling". Presented talk titled: "Who Tweets? First Steps in Understanding and Evaluating Representation on Twitter" drawing upon BSA data on Twitter use. Reference to the horsemeat study to make the point that the people who are most likely to have been affected by the scare are those with lower incomes, who are less likely to be on Twitter. Audience was primarily those interested in survey methodology and alternative sources of data.
Year(s) Of Engagement Activity 2016
 
Description Twitter for Academics (post-graduate led conference at Cardiff University) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Postgraduate students
Results and Impact Interdisciplinary postgraduate-led conference exploring how to engage with and utilise Twitter effectively for teaching, research and impact. Dr Luke Sloan gave keynote presentation titled "Who Tweets? First Steps in Understanding and Evaluating Representation on Twitter" drawing on BSA data to demonstrate the cutting edge work being done on estimating the Twitter population. Horsemeat project discussed as case study.
Year(s) Of Engagement Activity 2016
 
Description Understanding the Challenges of the Food System Final Event 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact This event brought together the five research teams under the ESRC-FSA co-funded research programme Understanding the Challenges of the Food System. This was an opportunity to present all the final results to a mixed audience of policymakers and academics. Particularly of value to the FSA in terms of future engagement and communications and the ESRC for linking to their latest call around resilience of the UK food system.
Our project very much complemented the work of some of the other groups in looking at consumers views.
Year(s) Of Engagement Activity 2017