Co-designing food waste services in the catering sector


Purpose
The purpose of this paper is to present results from the action research project, where sustainability professionals, local businesses and academic researchers collaborated on exploring barriers for food waste recycling in SMEs food outlets in order to inform local policy and business practices in Bristol, UK.


Design/methodology/approach
The researchers conducted face-to-face, qualitative surveys of 79 catering businesses in three diverse areas of the city. The action research methodology was applied, where a range of co-researchers contributed towards study design and review.


Findings
The research reveals the main barriers to recycling and how such perceptions differ depending on whether the respondents do or do not recycle, with “convenience” and “cost” being the main issue according to the already recycling participants. On the other hand, participants who do not recycle state that their main reason is “not enough waste” and “lack of space”.


Practical implications
Participants recommended a range of measures, which could improve the current food waste services in Bristol. For example, they suggest that business engagement should address the barriers voiced by the participants applying the framings used by them, rather than assuming restaurants and cafes are not aware of the issue. By inviting a variety of non-academic stakeholders into the process of research design and analysis, the project addressed the imbalances in knowledge production and policy design.


Originality/value
Despite the local and qualitative focus of this paper, the results and research methodology could act as a useful guide for conducting food waste action research in the policy context.


professionals to reach the above conclusions. 116 Thus, the solutions proposed reflect the composition of the participants' pool, i.e. 117 managers, academic experts, and policymakers. They suggest interventions at high-level 118 decision-making, e.g. "a multi-stakeholder dialogue" (Goebel et al., 2015;Priefer et al., 2016), 119 "improving data availability and measurements by agreeing on the definitions of "food 120 waste/surplus food" or "mandatory collection of food waste" (Priefer et al., 2016). predominantly recycling measures as an answer to food waste. She pointed out that this practice is against the widely accepted hierarchy of waste, which seeks to prevent, then redistribute and 139 then recycle waste (Papargyropoulou et al., 2014). As a result, surplus food turns into a waste 140 commodity (Mourad, 2016). 141 However, even after reducing food production and redistributing surplus to people in 142 need, there will be "unavoidable waste" left, e.g. cores, egg shells or bones. It is estimated that 143 a quarter of food waste in catering is "unavoidable", a category defined by WRAP (2017)  scale optimisation of the existing supply chains, leaving the current modes of over-production 160 and over-consumption unchallenged. In other words, they are underpinned by the "economic" 161 discourse understood as "resource efficiency" rather than "protest against capitalism". In turn, describe a societal shift from arguing for "wartime resourcefulness" to contemporary concerns 166 about "feeding global population with limited resources" (ibid). Furthermore, they argue that 167 at the catering level, food waste is constructed as a moral issue and a matter of incompetency 168 in business management and food handling (ibid).

169
In summary, the academic literature provides comprehensive reasons for food waste policy and managerial recommendations. The following section will elucidate why the 177 approach presented in this paper, action research, is suitable for closing the gap in the literature.  The central point of the primary data collection was the qualitative survey, designed in 192 collaboration with the practice-based co-researcher (Author 2). Following the data collection 193 and preliminary analysis stage, co-researchers also contributed to the scrutiny of the results.

194
Action research is used in this study as it focuses on practical and applied knowledge, 195 and it strives to break down the hierarchies and imbalances in knowledge production (Hawkins,196 2015). It acts as a conduit between practitioners, policymakers, and researchers. Moreover,   202 The qualitative design was applied in this study to derive diversity and "richness" of 203 answers and participants rather than statistical analysis of results (Jansen, 2010). Therefore, the 204 results do not aim to represent the whole catering sector, but they act as an evidence for co-   209 The researchers carried out 79 face-to-face surveys in January 2018. Businesses were 210 purposively selected, so each business type and research area (see Table 1 for area 211 characteristics) was adequately represented. Furthermore, the areas selected reflect the 212 diversity of Bristol's high streets. The sample size was determined so that the dataset achieves  The majority of the interviews lasted between 5 and 10 minutes, however, in 8 cases 217 they lasted 15-25 minutes (including 1 waste facilities tour). Answers were recorded in writing 218 on a survey sheet. Two respondents opted for sending email responses instead of participating 219 in a face-to-face survey. The interviews were conducted with the staff at the front of the house 220 unless they specifically requested another staff member to contribute (e.g. an off-duty manager 221 or a chef). Since the level of seniority was not a requirement for participation, the survey 222 allowed to capture a more diverse range of experiences and opinions. Furthermore, the concise 223 survey design contributed to a high response rate as the day-to-day work wasn't disturbed, nor 224 was a separate meeting was required as the willing participants were recruited using the door-225 knocking technique.

226
When distributing the survey, the researchers avoided prompting. They also took care   236 The researchers coded participants' answers and analysed them using thematic-   243 The analysis of survey data should not be statistically relied upon since the sample size 244 is not representative of the whole city. Seventy-nine participants and three neighbourhoods 245 cannot reflect the participation rate for some 1000 catering outlets located across all 34 wards 246 in the city. However, the nature of action research does not require results to be generalisable 247 as the focus of the survey is the themes and discourses derived from the qualitative data.

248
Similarly, the recycling participation figure might be an overestimation, as participants who do 249 not recycle could refuse taking part in the surveys or do not reveal its practices truthfully.

250
However, a high response rate and a range of honest and detailed responses from non-recycling 251 businesses encourage trust in the data.

252
The researchers encountered a language barrier in a few cases, which affected the 253 "richness" of the dataset, particularly in Easton. The researcher used plain language and 254 repetitions to encourage complete answers. For the future, the researchers recommend working 255 with interpreters.
The length of the questionnaire (5 open-ended questions) could potentially affect the 257 "richness" of data. However, a variety of answers, high response rate and the presence of 258 forward-looking insights suggest that the data achieved saturation. The researchers decided to 259 conduct a short survey, as this was more appropriate in busy, customer-facing environments.

261
In total, 79 out of a population of 95 approached businesses responded to the survey 262 (83% response rate). Table 2 outlines the demographic characteristics of survey respondents.

263
The participating businesses were located in the following areas: city centre (39.2%),

264
Gloucester Road (40.5%) and Easton (20.3%). The smaller sample size in Easton reflects the 265 size of the area. They characterised themselves as the following: restaurants (29.1%), pubs 266 (12.7%), cafes (30.4%), fast food takeaways (22.8%) and bakeries (5%).   274 Out of 79 respondents, 42 (53%) confirmed that they already use food waste collection 275 services. Table 3 outlines the response by area and business type. The recycling rate is not 276 evenly distributed across the areas and business types, with Easton having much lower 277 participation rate than other areas. While restaurants achieved high recycling participation rate 278 (78%), takeaways and bakeries recycled the least (respectively 33% and 0% participation in 279 recycling services). Although the results are not statistically significant, they indicate that participation in recycling services may depend on the type of the business and the location of 281 the catering business. As such, improved waste services could target its recipients according to 282 businesses in needs and potential priority areas.   314 Thirteen participants indicated that food waste recycling should be a legal requirement, 315 e.g. "It should be done by the council, not waste companies" takeaway/ city centre. Notably, 316 12 out of 13 answers came from participants, who already recycle. This result should not be 317 used as an extrapolation for the acceptance of compulsory food waste recycling policy. The 318 survey did not explicitly ask: "are you in favour of compulsory food waste management?".

319
Instead, the question was the following: "how could waste collection services be improved?".

345
Participants emphasised that the quality of the communication, rather than the quantity is the 346 key. In extreme cases, a lack of communication is the issue. For example, a manager of an 347 Easton restaurant recalls "we've never even been offered recycling, only general waste!". 348 Participants believe that business engagement should be meaningful and offer more than    Co-researchers agreed that sharing stories and discourses ought to help uptake. Traders 416 groups could act as knowledge sharing spaces; areas lacking such way of self-organising should 417 get help from the local authority with setting up such business community. They also agreed 418 that lack of space is the major issue for small businesses. However, a group deal and discount 419 could offer frequent collection, which would reduce the need for storage. • Area targeted for street cleaning (BCC, 2017) • 88% residents concerned about climate change (BCC, 2016) • 91% residents think litter is a problem (BCC, 2016) • No Business Improvement District present 4 • Most common sociodemographic ACORN 5 categories: