Aboriginal and Torres Strait Islander people living in remote areas generally experience the poorest health outcomes and hold the worst economic position in Australia [1
]. Aboriginal and Torres Strait Islander people experience unemployment at 4.2 times, and have an average disposable income 70% of, non-Indigenous Australians [3
]. Poverty is greatest for Aboriginal and Torres Strait Islander people living in very remote areas and is growing [2
]. The life expectancy of Aboriginal and Torres Strait Islander people is approximately 10 years less than non-Indigenous Australians. The majority of this gap is due to chronic disease, especially cardiovascular disease and cancer, and injury for the 35–74 years age group [4
]. The gap is largest in remote areas where Aboriginal and Torres Strait Islander people experience a burden 2.4 times that of non-Indigenous people [5
]. Dietary intake is a key risk factor contributing to this gap [4
Nutrient-rich traditional, non-market food continues to contribute to dietary intake [6
], though the rapid nutrition transition resulting from colonization has led to a population diet high in sugar, salt and fat and low intakes of vegetables, fruit and other nutrient-rich foods [7
]. In remote Aboriginal and Torres Strait Islander communities, Western foods are predominantly purchased from the single retail food outlet, referred to as the store, operating in a challenging, remote environment, which contributes to the high cost of food. Many stores are community-owned, providing a unique opportunity for local policy development [8
The remote store landscape has undergone considerable change in the last decade, particularly in policy and services. In 2008, a Close the Gap statement of intent was agreed to by a number of Aboriginal and Torres Strait Islander people and organizations and the Australian Government [9
]. In the same year, the Council of Australian Governments released the Closing the Gap Strategy that aimed to achieve health equity within 25 years [10
] and in 2009 developed the National Strategy for Food Security in Remote Indigenous Communities which linked food security (i.e., the ability to acquire appropriate and nutritious food in a regular and socially acceptable manner) and nutrition with the national Closing the Gap targets [11
]. Two years prior to this in the Northern Territory (NT) of Australia, the Northern Territory Emergency Response was implemented and included a number of measures indirectly related to food ‘security’. One of these was for the compulsory income management of welfare recipients [12
] (i.e., restriction of available cash and purchase of specific products), which has since been extended beyond the NT [13
]. A second measure was the introduction of a regulatory framework for the operation of remote stores, including minimum standards relating to food security; this remains effective today [14
]. The Australian National Audit Office reports however, that government policies have made minimal contribution to addressing food insecurity in remote communities [15
]. Reports on the Closing the Gap targets show mixed outcomes, though importantly that the target to close the life expectancy gap is not on track [10
] and that outcomes are worse in remote than non-remote areas [2
]. The Productivity Commission highlights the importance of developing an evaluation culture in Aboriginal and Torres Strait Islander policy where policy evaluation informs future policy [16
During this time of policy change there has also been a growth in organizations that provide retail management services to remote community storeowners, alongside an increasing recognition of the role that the stores play in the health of the communities [17
]. The historical tension between economic and health outcomes may be giving way as organizations publicly demonstrate valuing health outcomes as an objective of sustainable business [19
]. In remote Aboriginal and Torres Strait Islander community stores, there are examples of local policies (i.e., the rules of operation determined by the governing body [26
]), which aim to promote health outcomes within a sustainable business model [24
]. There is significant opportunity in this dynamic remote retail context to work with storeowners and the systems they operate within to influence local store food policy to create health-promoting environments.
Food pricing is considered one of the more effective practices to influence consumer purchasing patterns [28
]. Health-promoting food pricing policies exist in remote stores, but there is little understanding of the decision-making process informing their design development including the magnitude of the price increase or decrease and promotion of the policy [29
]. Policy analysis can help understand the process of design development and thereby identify opportunities to strengthen design and improve health outcomes through the store [30
]. Policy development models have evolved to consider trade-offs between multiple and often conflicting objectives [32
]; they may have utility in understanding efforts in the remote retail context where governing bodies deal with the dual and potentially conflicting objectives of consumer health outcomes alongside commercial viability of stores. Decision-making which incorporates evidence will hopefully lead to consideration of a greater range of policy options and result in more effective outcomes [33
This paper describes health-promoting food pricing policies including their alignment with evidence, and the decision-making processes in their development in very remote Aboriginal and Torres Strait Islander community stores in Australia. We specifically refer to ‘food policy’ as the local-level food policy implemented in stores aimed at modifying the price of food/beverages in order to promote health.
Approximately 175 stores supply food in some of the 1187 discrete Indigenous communities in remote locations across Australia [8
]. A total of 92,960 Aboriginal and Torres Strait Islander people and a small number of non-Indigenous people reside in these communities. Seventeen communities have a population greater than 1000 and almost 75% are located in very remote locations [34
]. Our study included very remote communities only [35
]. These are located largely in the NT, Queensland (Qld), South Australia (SA) and Western Australia (WA). Some stores are owned by the government or are privately owned, though the most common model is of incorporated community ownership where Aboriginal and Torres Strait Islander residents comprise the membership. These stores function as either not-for-profit or business enterprises and are often responsive to community priorities. The owners of community-owned stores employ a store manager/s or engage the services of a retail organization to manage the store’s operation, with the latter model accounting for approximately 55% of stores in remote Aboriginal and Torres Strait Islander communities in Australia [17
]. In addition to operating an effective retail operation, a number of stores and retail organizations aim to employ local Aboriginal and Torres Strait Islander people and promote positive nutrition outcomes and healthy lifestyles [8
A qualitative study was conducted that applied a methodology informed by Thow’s framework used in the Pacific Region. This framework was informed by policy theories related to lesson drawing to understand the form of food policies and how to engage with policy-makers [30
]. It was successfully used to describe the common elements of policy processes across the diversity of policy processes identified in different countries in the Pacific Region. Our methodology was informed by this framework as we similarly anticipated a diversity of policy processes across different remote communities, states and territories and governance models. We first focused on determining the range of pricing policies in place in remote stores and secondly on an understanding the stages of the process [32
], the people involved [30
], identification of objectives [32
], consideration of assessment criteria applied [32
] including a list of pre-determined criteria previously used in food policy assessment (i.e., feasibility, sustainability, acceptability, importance, effectiveness, unintended consequences [36
]), and the evidence considered.
Purposive sampling was employed, informed by the snowball method, to maximize coverage of the types of policies implemented. Participants were: (i) retailers, who were the store managers employed by the owners of a community store or store managers and retail management staff employed by a retail organization, and (ii) health professionals, including public health nutritionists (hereafter, nutritionists) and others working in roles with stores employed by a retail organization, government or non-government organization. Participants were required to identify that they had knowledge of health-promoting food pricing policy in remote Aboriginal and Torres Strait Islander community stores. At least one retailer and where applicable, the nutritionist from each of five retail organizations representing the majority of these entities, and all nutritionists in service provision and food supply policy known to Megan Ferguson and Julie Brimblecombe operating in remote NT, Qld, SA and WA, were invited to participate. Participants were invited by email from the lead researcher or by a potential participant in the study. This study did not seek to quantify policy implementation by store, store governance model or state/territory.
A semi-structured interview guide was used in all interviews. It focused on two sets of data. The first was the health-promoting food pricing policies in stores. We included price increases and subsidies in the form of price discounts, rewards, vouchers and free product give-away. We excluded takeaway food outlets as a setting and government policy instruments that might impact on food purchases such as income management. The second set of data focused on the decision-making process for one of the policies reported. Interviews lasted on average 50 min, and were conducted by Megan Ferguson, a nutritionist who has worked in both the remote health and retail sectors. This background was important in terms of understanding the context and relating to participants’ experiences. Interviews were conducted in English, in person or by phone. In one case, responses to the interview questions were e-mailed by a participant. Participants provided consent and all interviews were audio-recorded, transcribed verbatim and returned to participants for checking. Documents describing food pricing policies were sourced or provided by participants and used to complement interview data. Data were uploaded and managed in NVivo (QSR International Pty Ltd. Version 11, Melbourne, Victoria, Australia, 2012). Ethical approval for this study was provided by Human Research Ethics Committees in the NT (HREC NTDHMSHR 2012–1711; CAHREC HREC-12-13; CDU HREC H12096), Qld (FNQ HREC HREC/16/QCH/35-1041) and WA (WACHS HREC 2016/13; WAAHEC 715; KAHPF 2016-006). Informed consent was obtained from all participants.
The dataset was reviewed independently by two researchers, Megan Ferguson and Julie Brimblecombe, who have extensive research, policy and practice experience in the remote retail and health sectors. This strengthened the analysis by ensuring research quality and relevance. The authors discussed and agreed on the coding framework. The data were coded by Megan Ferguson and the findings reviewed with Julie Brimblecombe.
Firstly, a data content analysis relating to the types and design of food pricing policies was conducted, with allowance for additional codes. The coding framework included the following: Under the three pre-determined codes, subsidy, price increase, subsidy/price increase combination; the sub-codes relating to each code of targeted food or beverage, magnitude of price change, duration, administration, complementary strategies, other design elements; and, a fourth emergent code, business fundamentals. Secondly, a deductive, thematic analysis of the decision-making process was conducted to identify why, how and who was involved.
Health-promoting food pricing policies implemented in very remote Aboriginal and Torres Strait Islander community stores in Australia were dominated by subsidies and subsidy/price increase combinations. These had a small to moderate impact on food prices of fruit, vegetables, bottled water, artificially sweetened and sugar sweetened carbonated beverages, and broadly used ‘healthy/essential’ and ‘unhealthy’ food classifications. Decision-making was a deliberative process, which evaluated policy feasibility and sustainability, though generally lacked incorporation of research-informed evidence.
4.1. Designing Health-Promoting Food Pricing Policy
The dominance of subsidies and subsidy/price increases reported in this study is in line with recommendations to support healthier choices in low socioeconomic populations with the subsidy/price increase combination possibly mitigating concerns about the potential regressive nature of taxes [40
]. The range of products targeted only partially align with the current evidence. The lack of criteria applied to the ‘healthy’ category for example, results in a misalignment with guidelines for good health and a lost opportunity to promote a healthy diet. Targeting artificially sweetened carbonated beverages may not support positive health outcomes as reducing the price of these is unlikely to decrease the consumption of sugar sweetened carbonated beverages [42
]. Additionally, there are calls for a greater focus on policy targeting discretionary foods [43
]. Magnitude of price changes were at best in line with recommendations for modifying purchasing [47
]. Equity was the objective of decision-making in some cases, and the magnitude of the price changes went some way to achieving this [48
]. The ongoing nature of most policies which are not routinely advertised to customers prevented the use of price as a signal to customers; this was described by participants and supported by others as a significant missed opportunity [49
Food pricing policies in this context which aim to improve health would be more aligned with research evidence if there was: (i) further targeting of products (e.g., specify healthy foods, foods likely to have a greater response to price changes [43
]); (ii) increased magnitude of price change [47
]; (iii) use of price and price promotion to send a signal to customers, such as through a price increase alone or dynamic, rotating subsidies and promoting the change in price to customers [29
]. Policies need to be assessed within the local context and may require new avenues for funding, such as by manufacturers, suppliers and wholesalers, by government or through evaluation of current food pricing policy or funds dispersal.
4.2. Enhancing Policy Development Processes
This analysis indicates that the process of decision-making was deliberative [32
]. Improved health, and to a lesser extent equity, were key objectives in the decision-making process. These objectives of health and equity inform policy development differently, including the sources of evidence required. Whilst assessment of effectiveness was considered a priority, participant response and the design of current policies, indicates limited use of research-informed evidence. Although consideration of unintended consequences was not universally viewed as important to the process, research-informed evidence would go some way to inform the assessment of this criterion whether it was explicitly included or not. Acceptability and importance were not well-considered criteria, although they were regarded a priority and likely to be best addressed through further engagement with Aboriginal and Torres Strait Islander storeowners and others they elect to involve. Given articulating and communicating problems is a crucial stage in decision-making [32
], the processes reported in this context are likely to be improved with further assessment of the criteria, acceptability, importance, effectiveness and unintended consequences of potential policies. The processes were generally focused on a single policy rather than evaluation of a suite of options. They were based on analysis of retail data, informed by an assessment of cost in terms of retail impact though not cost-effectiveness, nor health impact, and limited in terms of robust monitoring and evaluation. Greater incorporation of research-informed evidence into the design of food pricing policies which have an objective of dietary or health improvement, is likely to result in more effective policy, and was called for by study participants [33
Complex policy with multiple and potentially conflicting objectives, is likely to create tension [32
]. There appears to be a shift in the well-documented tension between commercial profit and health outcomes in remote stores [22
]. Opportunities exist for well-designed health-promoting food pricing policies to be considered within the suite of business practices by storeowners, and precedent has been set for this as described in our study. Currently, retailers are front and center of the decision-making process in remote stores, hence the reliance on retail-focused evidence and criteria in the decision-making process. Current processes offer opportunities to further progress health-promoting policy, such as using the role of benchmarking against other stores and organizations as a potential mechanism for dissemination of good practice. Mechanisms to support decision-makers to access research-informed evidence and to assess acceptability, importance and unintended consequences of policies for the local context could lead to more effective health-promoting policies. This might involve a greater role for Aboriginal and Torres Strait Islander storeowners and nutritionists in decision-making.
4.3. Strengths and Limitations
This study has captured the views and experiences of retailers and health professionals across remote Aboriginal and Torres Strait Islander communities in Australia. Effort was made to ensure retailers operating in independent stores were included, though without a census of all stores, this is a more challenging cohort to identify and locate. The resources for this study did not allow for the conduct of interviews in remote communities with Aboriginal and Torres Strait Islander store committee/board members. Interviewing those persons known to work closely with storeowners provided insight into the roles and processes which could be further explored. Participants were invited to contribute where health-promoting food pricing policies were implemented and as such, this is likely to represent the best-case scenario rather than the situation in all remote Aboriginal and Torres Strait Islander community stores. The case considered was food pricing policies, and the process of policy development may be different to that of other health-promoting food policies in stores.