Unhealthy diets and associated adverse health conditions including overweight and obesity are a seemingly intractable global challenge related to contemporary global food systems(Reference Afshin, Sur and Fay1,Reference Peeters2) . It is estimated that 11 million deaths globally were attributable to dietary risk factors in 2017, with the most important risk factors being high intakes of Na and low intakes of whole grains and fruits(Reference Afshin, Sur and Fay1).
Within its Global Action Plan to prevent and control non-communicable diseases (NCDs), the WHO encourages its Member States to develop and implement a range of measures to promote healthier diets, including actions that address the food environment(3). With respect to the food retail environment, where food is sold to, and purchased by consumers, WHO specifically recommends ‘policy measures that engage food retailers and caterers to improve the availability, affordability and acceptability of healthier food products’(3). Recommended strategies to prevent diet-related conditions increasingly include measures which seek to regulate the food environment to decrease the health and economic burden of NCD(Reference Mah, Luongo and Hasdell4–Reference Brimblecombe, McMahon and Ferguson6). Public health research describes a range of regulatory interventions that seek to enable healthy food purchases by consumers by targeting the food environment(Reference Mah, Luongo and Hasdell4,Reference Mah, Cook and Rideout7,Reference Brimblecombe, Ferguson and McMahon8) .
Food retail regulatory interventions can take a variety of forms and involve both government and non-government stakeholders. For the purposes of this review, we differentiate between forms of regulation developed by government, also called ‘public regulation’ (e.g. reformulation programmes, front of pack labelling, sugar taxes, zoning/bylaws limiting the opening of new unhealthy food retail outlets)(Reference Mah, Cook and Rideout7,Reference Colchero, Rivera-Dommarco and Popkin9–Reference Black11) , and forms of ‘private regulation’ developed by non-government actors, such as arrangements between organisations and retailers or food retail organisations themselves (e.g. policies or contracts specifying the type, labelling, amount or placement of healthy food or beverages in food retail and vending)(Reference Boelsen-Robinson, Blake and Backholer12–Reference von Philipsborn, Stratil and Heise14).
While there has been significant focus in the academic literature and international policy recommendations on public regulation, different forms of private regulation are increasing at both national and global levels, including in the regulation of food retail environments(Reference Lane, Naylor and Tomlin13). Private or multi-stakeholder forms of regulation are increasingly used to address issues such as fair food trading, food safety and environmental sustainability in food retailing (as with fair trading certification schemes developed by non-government organisations and business actors)(Reference Havinga, Hammoudi, Grazia, Surry and Traversac15–Reference Muradian and Pelupessy17).
Private regulation can be voluntary or mandatory in nature, that is, enforceable. Voluntary private regulation relies on the agreement of the regulated entity (the food retailer in the case of this review) to implement, and there are no enforceable consequences for non-compliance. Enforceable private regulation includes contractual obligations often found in vending contracts to provide a certain percentage of healthier food options, accompanied by mandatory sanctions for non-compliance including dismissal of the vendor(18).
Available evidence suggests challenges in implementing effective private regulation to support healthy food retail environments(Reference Blake, Boelsen-Robinson and Hanna19–Reference Gupta, Alston and Needham22). Where they have been attempted, such interventions are often externally driven and maintained by health sector actors, with variable interest from food retailers themselves(Reference Mah, Cook and Rideout7,Reference Huse, Orellana and Ferguson23) . Various barriers (lack of customer demand, lack of retailer interest in menu labelling and lack of standardised recipes) and enablers (improved business image, consumer interest and competitive advantage) have been identified(Reference Thomas24,Reference Kerins, McSharry and Hayes25) . However, the provision and promotion of healthy food in food retail settings remains difficult to implement and sustain(Reference Gupta, Alston and Needham22). Existing studies indicate that food retailers can perceive interventions like menu labelling as a potential threat to profit and without specific intervention from the public health community, retailers currently have little incentive to independently label, promote and sell healthier food items(Reference Thomas24,Reference Middel, Schuitmaker-Warnaar and Mackenbach26) .
Further, research from the fields of regulation and public health law show that in order to be effective, all regulatory measures must be accompanied by adequate processes for monitoring, enforcement and review(Reference Jones, Neal and Reeve27,Reference Magnusson and Reeve28) . The inclusion of monitoring processes allows for an evaluation of the regulatory measure’s performance in achieving its objectives and enables enforcement action (for mandatory schemes)(Reference Reeve29). Likewise, processes of review and enforcement are important for enabling continuous improvement, deterring non-compliance and enhancing the credibility of private regulation. Ideally, monitoring, enforcement and review processes should be undertaken by external, independent actors, although this is relatively rare in private regulatory systems(Reference Reeve29). This review, therefore, places a novel focus on the use of private regulation in food retail settings that has the aim of improving diet-related health with a specific focus on the processes used to implement this form of private regulation. Drawing on insights from public health law and regulatory theory, this review examines the types of private regulatory measures used to create healthy food retail environments, how these measures were implemented, monitored, reviewed and enforced, and the barriers to and enablers of these effective regulatory governance processes. In doing so, this review contributes to the emerging area of healthy food retail research in public health nutrition.
Methods
Protocol and registration
We undertook a scoping review informed by the Johanna Briggs Institute guidelines for scoping reviews(Reference Peters, McInerney and Munn30) and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis extension for Scoping Reviews (PRISMA-ScR)(Reference Tricco, Lillie and Zarin31). The review protocol was developed by our team of public health nutrition and public health law researchers prior to registration with Open Science Framework https://osf.io/7th83.
Definitions and eligibility criteria
For the purposes of this research, we defined ‘food retail’ as any physical location that sells food for consumer consumption where the consumer has a choice in regard to what they will purchase(Reference Spencer and Kneebone32–35). We included take-away food outlets, supermarkets, restaurants, cafes, vending machines and hospital cafes and excluded online food environments and institutionalised food service (where food is provided free of charge and with no or limited consumer choice) found in settings such as aged care, defence, hospitals and correctional services settings.
Our definition of private regulation includes regulatory measures developed by private actors to implement guidelines or policies developed by public (government) actors. For example, a national or state government may produce a healthy eating framework that they encourage organisations to implement within their own settings(36,37) . Where such frameworks are locally implemented by an organisational policy or contract, we include this as an example of private regulation that falls within the scope of this review.
To describe the types of private regulatory measures, we used a framework developed by Mozaffarian that classifies policy interventions by level, target, domain and mechanism(Reference Mozaffarian, Angell and Lang38). Originally developed to analyse features of government-led (public) regulation, we adapted this tool to suit our focus on private regulation and used it to extract information on the domain and mechanism for each regulatory measure (Table 1). For our purposes ‘domain’ refers to the broad type of action or intervention used and includes instore point of purchase information, fiscal policies (e.g. pricing strategies), food quality standards (percentage of healthy items offered for sale) and built environment changes (e.g. changing the physical environment to favour the selection of healthier foods). ‘Mechanism’ refers to the modification the intervention is attempting to achieve and includes attempts to alter consumer preference or choice, altering the composition of food sold so it is healthier (prepared or pre-packaged products with less salt/sugar/fat), and altering the availability and accessibility of healthier food options in the food retail setting.
Table 1 Study Characteristics
To ensure we captured all articles pertaining to our definition of ‘private regulation’, we used broad search terms for regulation in our initial searches and then excluded forms of ‘public regulation’ at the stage of full-text screening. Searches were limited to articles published in English. We aimed to capture articles that described the use of private regulation that had been embedded in the organisation (i.e. was not a research trial) to create healthy food retail environments. We initially included school food settings in our search; however, this proved problematic. The decision to exclude these articles at the stage of full-text screening was made due to the difficulty of interpreting the results of these articles which reported a combination of both user-pays and institutionalised food service provision. Results were not reported separately according to the different means of food service provision, and therefore these articles were deemed to not meet our inclusion criteria. Our inclusion and exclusion criteria are presented in Table 2.
Table 2 Final eligibility criteria
* Added to the exclusion criteria at full-text review.
Search
A detailed search strategy was developed with the aid of a university librarian and the example for Ovid MEDLINE is presented in see online supplementary material, Supplementary Table S1.
Information sources
Six databases (Ovid MEDLINE, PsycINFO, Embase, CINAHL Plus, Business Source Complete and Scopus) covering the fields of public health, nutrition, business and law were included in our search strategy to maximise our chances of capturing existing literature. Articles identified from searches conducted by JD on 8–9 October 2020 and repeated on 1 and 6 September 2023 by EvB were downloaded from each of six databases to EndNote and screened for duplicates. Covidence was used to identify and exclude further duplicates and to manage the screening, review and extraction process.
Selection of sources of evidence
JD and MF independently screened an initial 700 (5 %) titles and abstracts using the inclusion and exclusion criteria and variance was forty-six articles (6·6 %). After discussion, refinements to the criteria were made, and an additional 700 (5 %) of articles were independently screened by both reviewers and variance was thirteen articles (1·9 %). Title and abstract screening on the remaining articles was then conducted by JD with reference to MF for clarification, if required.
Data charting process
Data were extracted from each article by JD in Covidence using templates designed by the research team. Ten percent of articles were cross checked by a second author to ensure consistency. Data extracted in Covidence were then exported into Microsoft Excel (2018), and the key data were transferred to Microsoft Word (2018) (Tables 1 and 3) and edited for clarity.
Table 3 Responsibility for regulatory processes and voluntary/mandatory nature
NVivo (2020) software was used to support our qualitative analysis of reported barriers and enablers. JD developed codes in NVivo for barriers to and enablers of implementation, monitoring, review and enforcement of the regulation described and analysed each article. JD and JB compared analyses for 10 % of articles to ensure consistency. A case classification summary report containing all the identified barriers to and enablers of implementation, monitoring, enforcement and review for all thirty-five articles was exported from NVivo to Microsoft Word. Evaluation of the reported barriers to and enablers of regulatory governance processes drew on a framework developed by Reeve and further adapted by other scholars for evaluating and strengthening the performance of public health law and regulation(Reference Magnusson and Reeve28,Reference Reeve29,Reference Reeve and Magnusson71) . This framework evaluates the dimensions of regulatory content and the processes established by regulation, including administration/implementation, monitoring, enforcement and review.
Data items
Data extraction templates were designed to collect data on article demographics, type of regulation including domain and mechanism, voluntary or mandatory nature of the regulation; regulatory governance processes regarding implementation, monitoring, enforcement and review and who had responsibility for them and the barriers to and enablers of these regulatory processes as described by the authors(Reference Mozaffarian, Angell and Lang38). Voluntary regulations were defined by an acceptance from the organisation, institution or food retailer to implement, but with no enforceable consequences for not implementing the regulation. Mandatory regulations defined by an expectation of implementation whether legally binding or organisationally endorsed.
Synthesis of results
Article demographics, type of regulation, including domain and mechanism, and compulsory nature were descriptively analysed (Tables 1 and 3). Deductive analysis using Reeve and Magnusson’s framework was used to identify barriers and enablers related to monitoring, enforcement and review (Table 4). Inductive thematic analysis was then used to group the large number of documented barriers to and enablers of the implementation process(Reference Reeve and Magnusson71,Reference Boyatzis72) .
Table 4. Qualitative analysis of the barriers to and enablers of effective regulatory governance processes
Results
Selection of sources of evidence
The final set of thirty-five articles were identified from 586 full-text articles assessed for eligibility from an initial screening of 17 694 articles. Reasons for exclusion of full-text articles are reported in our PRISMA flow diagram in Fig. 1.
Fig. 1 PRISMA
Characteristics of sources of evidence
The thirty-five articles identified were published between 2011 and 2023, with twenty-seven articles (77 %) published from 2015 onwards. Article characteristics are described in Tables 1 and 3.
Results of individual sources of evidence and synthesis of results
Types of private regulatory measures used
The thirty-five articles identified reported on twenty-six unique private regulatory initiatives (hereafter ‘initiatives’) used to create healthier food retail environments, as some articles evaluated the same initiative but from a different perspective. Five articles reported on the Canadian Alberta Nutrition Guidelines for Children and Youth in Recreation Centres(Reference Kirk, Olstad and McIsaac56,Reference Olstad, Lieffers and Raine62,Reference Olstad, Raine and McCargar64,Reference Olstad, Downs and Raine73,Reference Olstad, Raine and McCargar74) . One of the articles that reported these Guidelines also reported on similar guidelines in Nova Scotia and British Columbia(Reference Kirk, Olstad and McIsaac56). Four articles(Reference Lam, Ejlerskov and White57,Reference Ejlerskov, Sharp and Stead75,Reference Ejlerskov, Stead and Adamson76) reported on United Kingdom (UK) voluntary supermarket checkout food policies, and three articles(Reference Mueller, Anzman-Frasca and Blakeley59,Reference Anzman-Frasca, Mueller and Lynskey77,Reference Anzman-Frasca, Mueller and Sliwa78) reported on the United States (US) National Restaurant Association’s Kids LiveWell program in the regional restaurant chain, Silver Diner. The remaining twenty-three articles reported on initiatives used in hospital and health service food retail outlets (n 6)(Reference Boelsen-Robinson, Blake and Backholer12,Reference Bell, Pond and Davies43,Reference Fandetti, Dahl and Webster52,Reference Moran, Krepp and Johnson Curtis58,Reference Stead, Eadie and McKell68,Reference Walker, Littlewood and Rogany79) , vending machines (n 4)(Reference von Philipsborn, Stratil and Heise14,Reference Blake, Peeters and Livaditis44,Reference Pharis, Colby and Wagner65,Reference Wickramasekaran, Robles and Dewey70) , fast food outlets (n 3)(Reference Choi, Hyary and Fleming-Milici47,Reference Robinson, Burton and Gough80,Reference Harpainter, Hewawitharana and Lee81) , supermarkets (n 3)(Reference Havinga, Hammoudi, Grazia, Surry and Traversac15,Reference An and Sturm39,Reference Fildes, Lally and Morris54) , independently owned food retail outlets (n 2)(Reference Bagwell42,Reference Bogart, Castro and Cohen45) , remote and regional community stores (n 2)(Reference Butler, Tapsell and Lyons-Wall46,Reference Ferguson, O’Dea and Holden53) , universities (n 2)(Reference Fandetti, Dahl and Webster52,Reference Rickrode-Fernandez, Kao and Lesser66) and sports and aquatic centres (n 1)(Reference Naughton, Romaniuk and Peeters60). Table 1 provides a summary of the characteristics of the included studies.
Of the twenty-six initiatives, ten were implemented in the USA(Reference Bogart, Castro and Cohen45,Reference Choi, Hyary and Fleming-Milici47,Reference Eneli, Oza-Frank and Grover51,Reference Fandetti, Dahl and Webster52,Reference Moran, Krepp and Johnson Curtis58,Reference Pharis, Colby and Wagner65,Reference Rickrode-Fernandez, Kao and Lesser66,Reference Wickramasekaran, Robles and Dewey70,Reference Anzman-Frasca, Mueller and Sliwa78,Reference Harpainter, Hewawitharana and Lee81) , seven in Australia(Reference Boelsen-Robinson, Blake and Backholer12,Reference Bell, Pond and Davies43,Reference Blake, Peeters and Livaditis44,Reference Butler, Tapsell and Lyons-Wall46,Reference Ferguson, O’Dea and Holden53,Reference Naughton, Romaniuk and Peeters60,Reference Walker, Littlewood and Rogany79) , five in the UK(Reference Bagwell42,Reference Fildes, Lally and Morris54,Reference Stead, Eadie and McKell68,Reference Ejlerskov, Stead and Adamson76,Reference Robinson, Burton and Gough80) , two in Canada(Reference Lane, Naylor and Tomlin13,Reference Kirk, Olstad and McIsaac56,Reference Olstad, Lieffers and Raine62) , one in Germany(Reference von Philipsborn, Stratil and Heise14) and one in South Africa(Reference An and Sturm39).
Sixteen (62 %) initiatives were voluntary(Reference Lane, Naylor and Tomlin13,Reference von Philipsborn, Stratil and Heise14,Reference An and Sturm39,Reference Bagwell42,Reference Bogart, Castro and Cohen45,Reference Choi, Hyary and Fleming-Milici47,Reference Ferguson, O’Dea and Holden53,Reference Moran, Krepp and Johnson Curtis58,Reference Naughton, Romaniuk and Peeters60,Reference Olstad, Raine and McCargar64,Reference Rickrode-Fernandez, Kao and Lesser66,Reference Ejlerskov, Sharp and Stead75,Reference Anzman-Frasca, Mueller and Sliwa78–Reference Harpainter, Hewawitharana and Lee81) , six (23 %) mandatory(Reference Butler, Tapsell and Lyons-Wall46,Reference Eneli, Oza-Frank and Grover51,Reference Fandetti, Dahl and Webster52,Reference Pharis, Colby and Wagner65,Reference Stead, Eadie and McKell68,Reference Wickramasekaran, Robles and Dewey70) , two (7·5 %) used both voluntary and mandatory approaches(Reference Boelsen-Robinson, Blake and Backholer12,Reference Bell, Pond and Davies43) and two initiatives (7·5 %) did not describe the status or provide enough information to determine the status (Table 3)(Reference Blake, Peeters and Livaditis44,Reference Fildes, Lally and Morris54) .
Of the twenty-six initiatives, forty (54 %) were developed and implemented by the organisation and twelve (46 %) were created based on healthy food retail frameworks or programs developed by governments, with implementation occurring at the organisational level via some form of private regulation.
Mozaffarian’s classification of policy interventions: domain and mechanism(Reference Mozaffarian, Angell and Lang38)
Domain
As noted in Table 1, of the twenty-six initiatives described, fourteen initiatives operated within one of the domains(Reference von Philipsborn, Stratil and Heise14,Reference An and Sturm39,Reference Choi, Hyary and Fleming-Milici47,Reference Ejlerskov, Stead and Adamson50,Reference Fandetti, Dahl and Webster52–Reference Fildes, Lally and Morris54,Reference Lam, Ejlerskov and White57–Reference Mueller, Anzman-Frasca and Blakeley59,Reference Rickrode-Fernandez, Kao and Lesser66,Reference Stead, Eadie and McKell68,Reference Wickramasekaran, Robles and Dewey70,Reference Robinson, Burton and Gough80,Reference Harpainter, Hewawitharana and Lee81) and twelve operated across multiple domains(Reference Boelsen-Robinson, Blake and Backholer12,Reference Bagwell42–Reference Bogart, Castro and Cohen45,Reference Choi, Hyary and Fleming-Milici47,Reference Fildes, Lally and Morris54,Reference Naughton, Romaniuk and Peeters60,Reference Olstad, Lieffers and Raine62,Reference Pharis, Colby and Wagner65,Reference Anzman-Frasca, Mueller and Sliwa78,Reference Walker, Littlewood and Rogany79) . Thirteen operated in the domain of point of purchase information;(Reference Boelsen-Robinson, Blake and Backholer12,Reference Bagwell42–Reference Bogart, Castro and Cohen45,Reference Choi, Hyary and Fleming-Milici47,Reference Naughton, Romaniuk and Peeters60,Reference Olstad, Downs and Raine61,Reference Pharis, Colby and Wagner65,Reference Robinson, Burton and Gough67–Reference Walker, Littlewood and Rogany69,Reference Anzman-Frasca, Mueller and Sliwa78) sixteen took the form of food quality standards;(Reference Boelsen-Robinson, Blake and Backholer12,Reference von Philipsborn, Stratil and Heise14,33,Reference Bagwell42–Reference Blake, Peeters and Livaditis44,Reference Choi, Hyary and Fleming-Milici47,Reference Fandetti, Dahl and Webster52,Reference Fildes, Lally and Morris54,Reference Kirk, Olstad and McIsaac56,Reference Moran, Krepp and Johnson Curtis58,Reference Naughton, Romaniuk and Peeters60,Reference Olstad, Downs and Raine61,Reference Pharis, Colby and Wagner65,Reference Rickrode-Fernandez, Kao and Lesser66,Reference Walker, Littlewood and Rogany69,Reference Wickramasekaran, Robles and Dewey70) six were in built environment changes;(Reference Lane, Naylor and Tomlin13,Reference Butler, Tapsell and Lyons-Wall46,Reference Eneli, Oza-Frank and Grover51,Reference Fildes, Lally and Morris54,Reference Harpainter, Hewawitharana and Lee55,Reference Naughton, Romaniuk and Peeters60) three were in population education(Reference An and Sturm39,Reference Ferguson, O’Dea and Holden53,Reference Pharis, Colby and Wagner65) and three were in the fiscal policy domain(Reference An and Sturm39,Reference Ferguson, O’Dea and Holden53,Reference Pharis, Colby and Wagner65) .
Mechanism
Of the twenty-six initiatives, thirteen used one mechanism and thirteen used multiple mechanisms. Within the initiatives described: twenty-four targeted altering food availability or accessibility,(Reference Boelsen-Robinson, Blake and Backholer12–Reference von Philipsborn, Stratil and Heise14,Reference An and Sturm39,Reference Anzman-Frasca, Mueller and Lynskey41–Reference Blake, Peeters and Livaditis44,Reference Butler, Tapsell and Lyons-Wall46,Reference Choi, Hyary and Fleming-Milici47,Reference Eneli, Oza-Frank and Grover51–Reference Kirk, Olstad and McIsaac56,Reference Moran, Krepp and Johnson Curtis58,Reference Naughton, Romaniuk and Peeters60,Reference Olstad, Downs and Raine61,Reference Rickrode-Fernandez, Kao and Lesser66,Reference Stead, Eadie and McKell68–Reference Wickramasekaran, Robles and Dewey70) nine targeted altering consumer preference or choice(Reference Boelsen-Robinson, Blake and Backholer12,Reference von Philipsborn, Stratil and Heise14,Reference Anzman-Frasca, Mueller and Lynskey41,Reference Bogart, Castro and Cohen45,Reference Ferguson, O’Dea and Holden53,Reference Pharis, Colby and Wagner65,Reference Robinson, Burton and Gough67–Reference Walker, Littlewood and Rogany69) and ten targeted altering food formulation(Reference Boelsen-Robinson, Blake and Backholer12,Reference von Philipsborn, Stratil and Heise14,Reference Anzman-Frasca, Mueller and Lynskey41–Reference Bell, Pond and Davies43,Reference Choi, Hyary and Fleming-Milici47,Reference Fandetti, Dahl and Webster52,Reference Kirk, Olstad and McIsaac56,Reference Naughton, Romaniuk and Peeters60,Reference Rickrode-Fernandez, Kao and Lesser66) .
Reporting of and responsibility for regulatory governance processes
In terms of the regulatory governance processes established by the initiatives, 34 articles reported on some aspect of implementation, 15 (43 %) articles reported a form of monitoring,(Reference Anzman-Frasca, Mueller and Sliwa40,Reference Bell, Pond and Davies43,Reference Blake, Peeters and Livaditis44,Reference Eneli, Oza-Frank and Grover51,Reference Fandetti, Dahl and Webster52,Reference Fildes, Lally and Morris54,Reference Kirk, Olstad and McIsaac56,Reference Moran, Krepp and Johnson Curtis58,Reference Naughton, Romaniuk and Peeters60,Reference Olstad, Raine and McCargar64–Reference Rickrode-Fernandez, Kao and Lesser66,Reference Stead, Eadie and McKell68–Reference Wickramasekaran, Robles and Dewey70) two (11 %) articles reported on enforcement(Reference Blake, Peeters and Livaditis44,Reference Stead, Eadie and McKell68) and six (17 %) articles reported on a review process(Reference An and Sturm39,Reference Anzman-Frasca, Mueller and Lynskey41,Reference Bell, Pond and Davies43,Reference Fandetti, Dahl and Webster52,Reference Rickrode-Fernandez, Kao and Lesser66,Reference Stead, Eadie and McKell68) . Table 3 also describes the entity that had responsibility for the regulatory governance processes.
Two articles reported on implementation, monitoring, enforcement and review(Reference Rickrode-Fernandez, Kao and Lesser66,Reference Stead, Eadie and McKell68) . The article by Stead and colleagues, described the Healthcare Retail Standard (HRS), a regulatory scheme developed by the Scottish Government that applied to all food retail outlets in the Scottish National Health Service and aimed to increase healthy food options and limit the promotion of unhealthy food(Reference Stead, Eadie and McKell68). The HRS was a mandatory inclusion in any contract negotiated with a commercial retail outlet, which provided a process for enforcement, although the specific details of how the enforcement took place were not described. Non-commercial (National Health Service run) food retail outlets were also required to comply with the HRS, but this was not incorporated into their contracts, so no enforcement process was apparent. Monitoring of the HRS was managed by an external partner, the Scottish Grocer’s Federation, which is the trade association for the retail convenience sector in Scotland. The Scottish Grocer’s Federation conducted initial inspections and provided guidance to retailers on how to meet the HRS. It conducted biennial quality assurance inspections thereafter. The authors noted two examples of the HRS being reviewed and then modified: (1) the inclusion of lower fat baked potato crisps/chips in meal deals after the observation of an increase in full fat crisp/chip sales and (2) a revision allowing packaged snack items with the price marked prominently on their packaging, which were initially banned, after feedback from retailers that no alternative could be sourced. Whether the review process was regular, or reactive, was not described.
Barriers and enablers to effective regulatory governance processes
Table 4 lists the barriers to and enablers of effective regulatory governance processes, as described by the authors of the included studies. Barriers to and enablers of implementation were frequently identified in the literature but the barriers to and enablers of monitoring, review and enforcement were reported less often. The use of voluntary private regulatory measures was noted in some articles to be a barrier to both implementation(Reference Olstad, Raine and McCargar63,Reference Olstad, Raine and McCargar64,Reference Robinson, Burton and Gough67) and enforcement(Reference Bogart, Castro and Cohen45). Studies reported a perception from retailers and managers that mandatory policies enabled implementation because they ‘levelled the playing field’(Reference Olstad, Downs and Raine61,Reference Stead, Eadie and McKell68) . Bogart and colleague’s article evaluating the American Beverage Association’s voluntary Better Calories initiative also noted concern within the public health community regarding compliance (and therefore effectiveness) of voluntary industry self-regulation given that industry’s primary aim is beverage sales (including unhealthy options)(Reference Bogart, Castro and Cohen45,Reference Simon82,Reference Ronit and Jensen83) .
Nine dominant themes emerged as either barriers to or enablers of implementation, including the regulatory substantive content, including the specific goals, terms, definitions, and conditions included in the regulation(Reference Magnusson and Reeve28); retailer issues, customer issues and operational issues – factors of concern related to retailers or customers, or practical/logistical issues related to operating a food retail outlet; financial issues related to financial cost/profit/loss associated with implementing the initiative; communication issues related to stakeholders, retailers and consumers being informed of initiatives; choice issues related to the perception of ‘free choice’ by consumers in selecting products; relationship management related to relationships between individuals and/or organisations within and/or outside their organisation such as internal stakeholders or food and drink supplier relationships and leadership – including organisational leadership and support.
The substantive content of a contract was identified as both an enabler and a barrier to implementation. Where a contract was due for renewal, this created an opportunity for change to occur, however, where an existing contract still had a significant time before renewal, this created a barrier to change(Reference Kirk, Olstad and McIsaac56,Reference Rickrode-Fernandez, Kao and Lesser66) . One article noted that the very nature of contracts or leases created a defined period of time that may be too short for effectiveness to be demonstrated(Reference Lane, Naylor and Tomlin13).
Enablers to monitoring included audit processes, provision of expert feedback to vending contractors on compliance with policy, monitoring of sales data to determine policy impact and an expectation of compliance by a defined date. Barriers included lack of time and staff resources to conduct monitoring, poorly defined targets and specific nutrition standards being left out of contractual obligations.
The enablers of a review process included proper monitoring that enabled the unintended consequences of the regulatory measure to be identified and modified. In this way, the monitoring data fed into the review process so that modifications could be made. One article noted that there was a lack of independent evaluation(Reference von Philipsborn, Stratil and Heise14), but otherwise the articles did not comment on the absence of any review or evaluation processes.
The enablers of enforcement were the inclusion of obligations and enforcement measures in contractual arrangements, the education of stakeholders regarding the policy and the presence of a specific policy compliance procedure.
Discussion
This study identified a range of private regulatory measures that aimed to create a healthier food retail environment. Our review found that private regulation was used under the auspices of programmes, standards, schemes, interventions, initiatives, policies, pledges charter, strategies, guidelines and contracts. The majority of initiatives described were voluntary despite recognition of the limitations of this format, particularly where commercial profit motives may be in conflict with the objectives of the initiative.
In the articles identified in this review, priority was given to reporting on implementation with less attention paid to other regulatory governance processes such as monitoring, review and enforcement. Accordingly, it was unclear from these studies whether many of the private regulatory measures described had established these important regulatory governance processes. Given that many of these articles were not focussed on regulatory governance, we do not discount the possibility that these processes may have been in place, but not reported on. In a recent review of healthy food retail interventions, Gupta and colleagues noted that the majority of published reviews also focused on implementation, with fewer focusing on programme sustainability and scale up(Reference Gupta, Alston and Needham22).
To enable improvement of healthy food retail initiatives, there needs to be greater reporting in the literature on the processes of monitoring, review and enforcement, along with evaluations of the barriers to and enablers of these regulatory governance processes. As discussed in the introduction section, these regulatory governance processes are key to the effective implementation of regulation, and effective regulatory implementation is more likely to result in improvements to the healthiness of the food retail environment, which the regulations under review in this study hope to achieve. Such reporting will also help to identify best regulatory practice design measures that facilitate the creation and sustainment of healthy food retail environments. The literature would benefit from the use of a robust, standardised framework that examines the entire regulatory process so that a comprehensive evaluation of the use of private regulation in healthy food retail environments can be made.
The barriers to and enablers of implementation reported in our study largely reflect those identified in two recent systematic reviews of healthy food retail interventions(Reference Gupta, Alston and Needham22,Reference Middel, Schuitmaker-Warnaar and Mackenbach84) . Retailer nutrition knowledge and beliefs, retailer concern over consumer demand or acceptance of healthier foods, profitability concerns and poor communication are reported as barriers to implementation across all three studies(Reference Gupta, Alston and Needham22,Reference Middel, Schuitmaker-Warnaar and Mackenbach84) . Similar enablers reported by all three studies were ease of intervention/implementation, no cost or profitable for retailer, consumer acceptance of changes, strong relationships/partnerships with all stakeholders and clear communication(Reference Gupta, Alston and Needham22,Reference Middel, Schuitmaker-Warnaar and Mackenbach84) . These barriers and enablers focus on the factors influencing implementation rather than the effectiveness of the implementation strategies themselves and/or the implementation strategies needed to bring about ongoing change. Our review brings attention to the need for researchers to go beyond reporting implementation and provide critical examination of the regulatory governance processes which in turn are important for effective implementation of healthy food retail initiatives(Reference Gupta, Alston and Needham22,Reference Middel, Schuitmaker-Warnaar and Mackenbach84) .
In an age of ‘big data’, we note that data, and access to it, was mentioned in only four articles as an enabler to monitoring(Reference Blake, Peeters and Livaditis44,Reference Ferguson, O’Dea and Holden53,Reference Naughton, Romaniuk and Peeters60,Reference Wickramasekaran, Robles and Dewey70) . Contractual obligations to electronically submit sales and nutrition data were noted as an enabler in the article by Wickramasekaran and colleagues evaluating a County-based healthy vending policy(Reference Wickramasekaran, Robles and Dewey70). However, they also noted that data were missing for some months, indicating that despite contractual obligations, sales data can still be difficult to access and/or problematic for monitoring purposes(Reference Wickramasekaran, Robles and Dewey70). Conversely, lack of detail or lack of data were identified as barriers to monitoring(Reference von Philipsborn, Stratil and Heise14,Reference Wickramasekaran, Robles and Dewey70) . While the article by Stead and colleagues was the only one to include details of implementation, monitoring, review and enforcement, it did not detail the specific monitoring processes employed. The authors noted that future research could focus on retailer financial viability, and that longer-term monitoring is required for this purpose. This suggests that sales data were perhaps not monitored in their study and/or they were not privy to data on profit or other business metrics. Point of sale data is a rich source of information for monitoring the outcomes of regulation in the food retail environment, including profit and/or loss(Reference Brimblecombe, McMahon and Ferguson6,Reference Brimblecombe, Liddle and O’Dea85) . It is also worth noting that the monitoring conducted in the Stead article was managed by an external partner, the Scottish Grocer’s Federation, which is the trade association for the retail convenience sector in Scotland. Whilst independent monitoring is seen as best practice, in this example, the monitoring is independent of the retailer itself, but conducted by an industry trade association which may introduce a conflict of interest(Reference Reeve29). This also points to the importance of private regulation being accompanied by transparency and accountability processes and for further research evaluating the presence and operation of these processes.
In Australia, as in many industrialised economies who have pursued a ‘deregulation’ agenda, there has been little government appetite to pursue public regulation to create a healthier food retail environment(Reference Baker, Gill and Friel86). This has created an opportunity for private regulation to fill the gap and diffuse throughout society, as various entities seek to create healthier food retail environments(Reference Black87). This diffusion of regulation away from government comes with risks and opportunities which need careful attention to enable equitable health outcomes(Reference Lacy-Nichols, Jones and Buse88). In democratic societies governments have responsibilities to their citizens in a way that profit-driven companies do not, thereby enabling checks and balances on governments that are not otherwise applied to companies(Reference Lacy-Nichols, Jones and Buse88). Commercial actors can act in ways beneficial to health; however, the literature notes the negative impact that powerful industries, such as the ultra-processed food industry, can have on health(Reference Lacy-Nichols, Jones and Buse88,89) .
Our finding that most of the articles were published relatively recently could reflect either (a) an increase in the use of private regulatory measures or (b) an increasing academic interest in reporting regulatory approaches to health-enabling food retail. This research may provide support to private actors involved in, or interested in implementing private regulatory measures, and empower them to include effective quality processes for monitoring, review and enforcement when drafting measures designed to create healthy food retail environments.
Limitations
Due to their nature as agreements between private parties (and therefore often commercially sensitive and treated as confidential), there may be examples of private regulation being used to create healthy food retail environments that have not been subject to academic investigation and are therefore not captured by our search. However, this does not weaken the key finding that reporting on regulatory governance, specifically monitoring, review and enforcement processes, appears to be overlooked.
The large number of articles identified in the searches created a significant burden of articles to screen. JD and MF both have experience in the field of healthy food retail environments, and therefore, it was agreed that if we could decrease our inter-observer variability to < 5 % then JD could continue the screening alone (inter-observer variability reduced to 1·9 %). To minimise reviewer bias, two reviewers should screen all articles; however, the decision to review the articles by one researcher was made to ensure timely completion of the research.
Conclusions
To be effective, private regulatory measures must be accompanied by effective processes for implementation, monitoring, review and enforcement(Reference Jones, Neal and Reeve27,Reference Magnusson and Reeve28) . Our research demonstrates that there is inadequate reporting in the peer reviewed literature on the processes for monitoring, review and enforcement, making it difficult to evaluate the presence or effectiveness of the regulatory processes established by each initiative. Strengthening reporting on the governance processes beyond implementation will improve the evidence base for forms of private regulation that aim to create a healthier food retail environment and enable the identification of design features that are more likely to lead to the creation of sustained healthier food retail environments.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors. J.D. is supported by an Australian Government Research Training Program Scholarship and Monash University.
Conflict of interest
There are no conflicts of interest.
Authorship
J.D., B.R., A.J. and J.B. conceived and designed the analysis.
J.D. and E.V. collected the data.
J.D., B.R., A.J. and J.B. contributed analysis tools.
J.D., M.F., E.V. and J.B. performed the analysis.
J.D., B.R., A.J. and J.B. wrote the paper.
Ethics of human subject participation
Not applicable.
Supplementary material
For supplementary material accompanying this paper visit https://doi.org/10.1017/S136898002400065X.
Open access
