Introduction

In many urban areas facing food insecurity, “the problem … isn’t a shortage of food, but a shortage of healthy food” [1]. Inequities in food retail contribute to and reflect this problem. In the USA, half of high-income zip codes have at least one supermarket, compared with just one in six low-income zip codes [2]. Consequently, residents of low-income neighborhoods dubbed “food swamps” may rely on small stores for food, which compared to supermarkets, devote dramatically more shelf space to highly processed foods (e.g. four times the space to carbonated beverages) and less space for healthy foods (e.g. 64% less for fresh fruits) [3].

In neighborhoods like San Francisco’s low-income Tenderloin District, the problem of lack of healthy food intersects with an overabundance of advertising, display, and availability of tobacco and alcohol in its many small, typically family-owned corner stores [4]. Unhealthy diet, smoking, and heavy drinking contribute to the neighborhood’s disproportionately high morbidity and mortality [5].

Following a brief overview of the neighborhood and its Healthy Corner Store Coalition (the Coalition), which emerged to address this problem, we present the conceptual framework and methods used in this Community-Based Participatory Research (CBPR) case study. We explore the Coalition’s evolution and its CBPR processes, findings, contributions to policy change through a healthy retail ordinance, and the “ripple effect” in the neighborhood. The challenges faced, including likely contributions to gentrification and displacement, are discussed, along with lessons for other policy-focused CBPR for urban health equity.

The Tenderloin Neighborhood and Healthy Corner Store Coalition

The Tenderloin is one of San Francisco’s poorest neighborhoods, with 32% of its 28,000 residents living below the federal poverty line, compared to 13% citywide [6]. With no full-service supermarket and roughly 60 corner stores primarily stocking processed foods, tobacco, and alcohol, lack of access to healthy retail is a significant problem. The Tenderloin has by far the highest tobacco and alcohol outlet density in the city and correspondingly elevated rates of tobacco use and alcoholism [5]. Together with deep social inequities and the chronic stressors associated with life in poor neighborhoods, these forces compound health risks for residents, who have some of San Francisco’s highest chronic disease rates [5].

The catalyst for Tenderloin Healthy Corner Store Coalition was a 2011 youth-driven Google™ map of the neighborhood’s corner stores which illustrated the plethora of tobacco saturation and poor access to healthy food. The map resonated with several community-based organizations (CBOs) and agencies prompting follow-up meetings and the formation of the Coalition [7]. Initially led by the Vietnamese Youth Development Center (VYDC) and the Tenderloin Neighborhood Development Corporation (TNDC), the Coalition included local residents, CBOs, staff from the local health department (DPH), and academic partners from the University of California.

Crucial to the Coalition’s mission was the hiring and training of five local residents as “Food Justice Leaders” (FJLs) based in part on the “Food Guardian” model in the Bayview neighborhood across town. The FJLs work as participatory reserachers, community organizers, and advocates. They conduct detailed store assessments, resident surveys, and merchant and resident education, using their findings to work for municipal policy change. Together with the larger Coalition and its allies, the FJLs helped secure passage and implementation of the Healthy Food Retailer Incentive Program Ordinance (dubbed Healthy Retail SF [HRSF]) in Fall 2013, the city’s program for incentivizing select stores in food swamps to become healthy retailers.

Conceptual Framework

To examine the processes through which the Coalition and its partners helped effect policy and neighborhood change and evaluated implementation outcomes, we employed Kingdon’s model of the policymaking process [8]. Although policymaking is non-linear and embedded within socio-historical contexts [9,10,11], several steps “shape the content, course, pace, and development of policy” [12]: problem definition or identification of an issue, generating awareness and getting on the policymakers’ agenda, deciding which policy option to pursue, negotiating for a policy win, policy implementation, and monitoring and evaluating policy outcomes.

Kingdon’s widely used model summarized these steps in three streams [8]: The problem stream, convincing decision makers a problem exists and building awareness; the politics stream, proposing feasible, politically attractive solutions, and the policy stream, negotiating to get approval of a proposed measure. When positive developments occur in all three streams, a policy “window of opportunity” opens, increasing the likelihood of success.

Methods

To study the Coalition and its CBPR and policy work, Yin’s case study method [13] was used, with its accent on “empirical investigation of a contemporary phenomenon within its real life context [using] multiple sources of evidence.” In-depth, semi-structured interviews were conducted with eight Coalition members, with shorter interviews of 17 corner store owners or managers, and five with policymakers and other stakeholders. Four focus groups, two each with FJLs and merchants, archival review of internal documents, media coverage, and other sources additionally were conducted from 2014 to 2017. Two researchers coded each transcript independently, later reconciling discrepancies. Notes from participant observations and archival data provided context and additional perspectives.

To monitor potential impacts of the HRSF program on the retail environment, quantitative data also were collected using a 54-item, modified version of the Retail Standards for Health and Sustainability tool, developed by the DPH and a Bayview neighborhood coalition, Southeast Food Access in Action (SEFA) [14] and used in the Bayview and other neighborhoods well beyond California [15]. The instrument, expanded by the Coalition to include seven additional tobacco items, assesses availability and promotion of alcohol, tobacco, and healthy and unhealthy food. Following training by Coalition staff and DPH and pilot testing, the FJLs built on their trusting relationships with merchants to obtain permission to assess stores with this tool in 2013–2015 and 2017. Assessments allowed us to study our hypothesis that HRSF would result in a ripple effect, wherein improvements would be seen not only in the five stores participating in the program but also in many of those not participating. Monthly point-of-sale (POS) data from the first four of the five Tenderloin HRSF stores also were collected and analyzed to detect changes in produce and tobacco sales through month 12 (later analysis forthcoming).

Role of the Coalition in Linking CBPR and Policymaking

The Problem Stream and Creation of a Grassroots Coalition

While substantial literature exists on the associations among tobacco marketing, lack of healthy foods, and poor health [3, 16, 17], to create awareness of the problem, local data that “hits home” is particularly useful. In the Tenderloin, the “apple map” created by the VYDC with support from DPH was exemplary. Based on data they collected from 35 corner stores, the youths’ map depicted store quality using GoogleMaps™ and images of a good apple, bad apple, and most often, a rotten apple core. Of the stores, 42% had no fresh produce, 85% lacked required “no smoking” signs, and 76% had over a third of their storefronts covered with advertising—mostly for alcohol, tobacco, and sugary drinks. Only 19% were “good apples,” with 66% rated “rotten apples” [18]. Together with earlier CBPR studies on unhealthy retail in the Bayview neighborhood [12, 19], this local evidence attracted attention to help effect change. On the neighborhood level, DPH encouraged the youth researchers to distribute the mandatory no smoking signs to all stores not in compliance, boosting compliance by 82% [18].

Effective data sharing also increased policymaker attention to the problem. After seeing the apple map, a local supervisor visited the Tenderloin and commented, “A lot of stores are covered in cigarette and alcohol ads, or junk food and drink ads…. I’ve really come to see food access as a civil rights issue. Many people don’t have access to affordable, good quality food at a fair price, and corner stores are a key part of this.”

In CBPR, as Cacari-Stone et al. note [11], partnership dynamics impact the roles of evidence in civic engagement and political participation. To involve more community residents and organizations in refining the problem and identifying potential remedies, a community meeting was hosted by TNDC and VYDC, for which the apple map served as a catalyst. Attended by about 60 residents and representatives of DPH and CBOs, the meeting generated considerable interest. As a convener commented, “we had various topics for the community to give input on, but converting the corner stores from something negative into having a positive influence on the neighborhood had the greatest support” [7]. This desire for ownership, coupled with the shared goal of creating broader awareness of the problem to effect change, culminated in the formalization of the Coalition.

The Coalition’s members met monthly, sometimes joined by an architect interested in healthy retail and members of a local mosque. Described by an early co-leader as a “very, very, very diverse,” group, the Coalition sometimes experienced tensions. Yet, it quickly emerged as a highly effective organization, due to strong inclusive leadership and its members’ unifying belief that “food is a health equity issue.” Commenting on the collaborative, community-driven process that characterized the Coalition’s work, another participant noted that at its monthly meetings, members “drop [personal or organizational] agendas at the door” and focus on activities in support of their collective goal.

“The Politics Stream” and the Coalition’s Research and Advocacy Contributions

To make the case for action promoting healthy retail, FJLs led data collection on the needs of local residents that could be shared locally and with policymakers. A 2012 multi-language survey of 640 residents revealed that most shopped outside the neighborhood for staples (e.g. produce, meat/poultry, and grains), representing close to 50% of their total grocery expenditures. TNDC staff estimated from this figure that the neighborhood lost $11 million each year [20]. This reality and the fact that almost 80% of respondents reported that they would buy healthy food locally if it were available and affordable were seen by Coalition members as underscoring the need for healthier, comprehensive food options locally, which could strengthen the local economy.

The VYDC and the Coalition held a joint press conference in December 2012, to share survey findings and the earlier apple map to leverage support for change. Coverage included a local radio segment and an article in a district supervisor’s newsletter [21]. As another policymaker remarked, “The fact that local people provided actual numbers and facts from work on-the-ground made a difference,” since proposed policy measures “have to have support from the community” [4].

Support also needed to come from local merchants, who could be resistant to a policy discouraging the sale of what many perceived as their most profitable items: tobacco and alcohol. The Coalition, DPH, and an architectural firm specializing in store conversions educated merchants about the strong profit margins on dairy, bread, protein, and fresh produce (25–50%, and sometimes > 100% for pre-cut fruit and salads compared to 15–25% on tobacco and alcohol [22]) and emphasized the voluntary nature of an incentive program.

To further increase buy-in from key stakeholders, a legislative aide remarked that “we sent different iterations of the [proposed] legislation to the Arab American Grocers’ Association (AAGA), the Coalition, and others to talk through the language of the measure.” This inclusive approach proved critical with a particularly important player: the AAGA, representing 450 stores in the city, which eventually endorsed the ordinance.

Merchants, DPH, the Coalition, the Office of Economic and Workforce Development (OEWD), and other partners worked out the ordinance details. The final product, HRSF, provided technical assistance with redesigns and other benefits for selected stores which, in turn, committed to changes to meet the legislation’s definition of a healthy retailer: ≥ 35% of selling space to healthy foods and ≤ 20% to alcohol and tobacco combined, while removing specified amounts of cigarette and alcohol ads and paying minimum wage (www.healthyretailSF.org).

To ensure political feasibility, the roles of different players were clearly specified, including only a small initial investment from the city. The OEWD would house the program and contribute $60,000 annually for operations, technical assistance, and equipment, working closely with DPH in running the program. The DPH, in turn, would contribute part time staff and community engagement resources. The fact that the Coalition had already brought in a foundation grant for a pilot store conversion further underscored the commitment of stakeholders in assisting the city in this endeavor.

The legislation described detailed methods for ensuring accountability, monitoring, and evaluation, including through POS data and bi-weekly, FJL-conducted “report cards” from each participating store, and the continued use of the aforementioned Corner Store Standards for Health and Sustainability Tool used by the FJLs to assess most of the Tenderloin’s 60+ corner stores. The FJLs’ strong relationships with local stores facilitated participation. Findings were disseminated through the Tenderloin Neighborhood Healthy Shopping Guide which included a picture, rating of one to four stars, and summary of “healthy highlights,” e.g., no tobacco, for each store (www.healthyTL.org). The baseline finding that most stores received just one to two stars demonstrated the need for the ordinance.

The importance of sustaining trusting relationships with participating stores and of community engagement, while building the evidence base for policy change [11], remained evident in this second “politics” phase. The FJLs, for example, shared copies of the Shopping Guide first with merchants, as part of individualized feedback packets and one-on-one education. To enhance community engagement, the first (2013) Shopping Guide was distributed to residents at a community forum attended by ~ 150 residents with tabling by 12 CBOs, a nutritious meal and recipes, and speeches by a city supervisor and others about the proposed ordinance.

The “Policy Stream” and a Window of Opportunity

In the final stream of the policymaking process, the FJLs and other Coalition members, together with their Bayview counterparts, spoke with policymakers in person and at hearings before the board of supervisors to advocate for the healthy retail legislation. Reflecting on the importance of their testimony at a hearing shortly before a vote of the full board, a policymaker commented on how Coalition members’ words resonated with some board members who had not previously paid attention to the issue: “It also brought up things not apparent [to them]. So much publicity about [the city’s] great restaurants, the food culture, but hearing from people who couldn’t get healthy food in their own neighborhood was something else.”

A Supervisor co-sponsor of the legislation reflected, “The Coalition was extremely influential in drafting, refining, and then passing the healthy retailer ordinance,” in part because it “brought members in to educate the legislators [and] had very clear ideas in working with our staff on what the measure should look like.” Their work paid off, and HRSF was unanimously passed by the Board on October 9, 2013.

In retrospect, a policy window of opportunity [8] may have facilitated passage. As a policymaker noted, City Hall-based priorities of redevelopment, reducing community violence, and “doing something” about growing inequalities and hunger in this affluent city provided an ideal environment for getting healthy retail in the city’s poorest neighborhoods on the agenda. In the Politics stream, the measure’s low cost, engagement of multiple community stakeholders, public-private partnerships, and grounding in scientific evidence also proved favorable. Finally, in the Policy Stream, a progressive board of supervisors, a mayor concerned about both the poor and the needs of small business, and effective and timely media advocacy helped secure victory.

Beyond Policy Passage: Implementation, Monitoring, and Evaluation

Passing legislation, particularly municipal ordinances which may “lack teeth,” must be followed by timely implementation, including detailed measures for monitoring, evaluating, and course corrections where needed [9, 12]. Soon after passage, the HRSF infrastructure, including a refined model and five implementation steps, was established, along with staffing and creation of a Centralized Resource Center. A HRSF Advisory Committee was established to review progress and offer guidance in program decision-making. Importantly, three representatives of the Coalition were invited to join the Advisory Committee, meeting twice a year at City Hall.

HRSF implementation involved the conversion of nine stores by 2016, five in the Tenderloin. Successfully translating the ordinance into practice further included engaging > 4000 community residents in nutrition education and healthy retail efforts, food advocate training and workforce development, the strengthening of healthy retailer skills and collaborations, and the development of new local partnerships with other demand-side projects, e.g., a free fruit and vegetable voucher program, EatSF (www.EatSFvoucher.org). Finally, implementation also involved sharing promising practices and preliminary findings through sponsorship of a Bay Area convening, media advocacy including at least 14 press events in 2014–2015 alone, and Coalition member presentations at state and national professional meetings and the 2016 meeting of the Society for International Urban Health.

To monitor and evaluate progress and outcomes at the corner store, Coalition, and neighborhood and municipal policy (HRSF) levels, several approaches were used. These included repeated observational assessments in two-thirds of the Tenderloin’s corner stores in 2014, 2015, and 2017, enabling analysis of changes from baseline (2013). The addition of new items on tobacco and e-cigarette advertising, availability, and display enabled additional data collection pertinent to newly implemented tobacco control legislation [4].

As noted above, a major hypothesis of this case study was that there would be a ripple effect among Tenderloin corner stores, such that improvements would extend beyond the five Tenderloin stores participating in HRSF. This ripple effect was demonstrated; from 2013 to 2015, the percentage of stores with a poor composite score (one to two stars) decreased from 77 to 52%, and in 2017, dropped again, albeit more modestly, to 49%. The number of stores with four stars quadrupled from one to four over this period.

Interviews in 2016 with the owners or managers of 17 non-participating stores further supported a ripple effect, with most reporting and many showing us that they had made some healthy changes (e.g. offering some fresh fruit or reducing cigarette advertising). Reasons given included wanting to stay competitive, hoping to get into HRSF, and because “my customers deserve it” [23].

Analysis of interim POS data from the first four Tenderloin HRSF stores through month 12 of the program shows promise. The average monthly units of produce sold increased from 5299 at the start of POS data collection to 7174, a 35% increase. One store that did not sell any produce at baseline increased average monthly units sold to 1418 from months 6 to 12. Three stores decreased units of tobacco sold by 35%, while in a fourth store, tobacco as a percent of sales remained the same despite an increase in total units sold.

Analysis of data from all five Tenderloin HRSF stores is currently being conducted to examine overall program impact and sales fluctuations due to factors like seasonality, store launches, community events, and enforcement of recent tobacco legislation (e.g. e-cigarette advertising and display restrictions, excise taxes, and increasing minimum age to 21). Additionally, as a DPH partner commented, these evaluation tools have also been critical to program improvement in other ways, “[enabling] FJLs to help merchants receive feedback and problem-solve...”

Analysis of data on the Coalition and its functioning also are yielding useful findings. Although many of the facilitating factors were discussed above, detailed analysis of these and key challenges faced are still underway. Some of the striking obstacles that surfaced should be underscored here including, however, severe limitations in fiscal and human resources that constrain what can be accomplished by the Coalition and HRSF. Although the conversion of five Tenderloin stores from 2014 to 2016 is impressive, against the background of need, it remains a drop in the bucket. Additionally, as some merchants noted, practices like the well-intentioned distribution of free fruits and vegetables by local volunteer programs may result in customers not purchasing produce locally, with stores in turn seeing fresh produce go bad. Some merchants also shared their belief that by failing to require that “food stamps” be spent primarily on healthy foods, this critical government program was thwarting the sale of healthier fare.

Finally, and with respect to the need to increase the support of policymakers and other stakeholders for such programs, Coalition leaders and DPH staff both facilitated capacity building in areas like testifying and media advocacy, and noted the major challenges faced. As several Coalition members commented, while being able to testify at hearings was “huge,” in reality, conflicting family and other obligations, mobility limitations, and the frustrating tendency for hearings to be canceled or delayed without notice were major barriers to participation. With policymakers and other stakeholders reaching out to Coalition members for testimony and support on other issues related to healthy retail (e.g. a soda tax, tobacco control measures), the importance of addressing such barriers to increase their civic participation is underscored.

Discussion

As Butler and colleagues [16] have noted, essential to improving health and equity on the community and policy level is “endorsing a paradigm shift in how to look at equity. For example, supporting and building communities’ ability to engage in reducing inequities at the state and local level; identifying creative ways to eliminate inequities; and measuring equity differently…”

We examined how such a shift in thinking was demonstrated by a neighborhood Coalition and its partners through their focus on CBPR and policy change to address food justice and inequities in exposure to tobacco and other unhealthy products in a low-income urban neighborhood. We further illustrated how the Coalition and other actors helped impact each stream of the policymaking process, as well as the implementation and evaluation aspects of the new HRSF legislation.

The power of youth and community voices and participation in CBPR as a catalyst for change in low-income communities has been well documented [15, 24,25,26,27]. The Tenderloin case study underscored this reality, as well as the role of the Coalition’s partnership dynamics in strengthening both the evidence base and civic engagement for change [11]. Respectful, interactive training of the Coalition’s FJLs in research methods and related areas, co-led by a DPH partner and later supplemented by additional trainings with UC researchers, emphasized the bi-directional nature of learning, data collection, interpretation, and effective dissemination.

This study had several limitations. First, the focus on a particular neighborhood and city, while allowing some transferability of findings, precluded generalizability. Research designs involving multiple sites and cross-case comparisons may partially alleviate this issue [28].

Second, although store assessments in most of the Tenderloin’s corner stores in each of 4 years and monthly POS data in all five Tenderloin HRSF stores provided useful quantitative findings for assessing HRSF, both methods faced challenges. Personnel turnover, competing priorities for time and resources, and in some HRSF stores, a longer-than-anticipated lag time in beginning POS data collection and POS data gaps due to unforeseen problems, were among the obstacles faced. As Black and colleagues [29] suggest, moreover, “More nuanced measures of the food environment, including multidimensional and individualized approaches, would enhance the state of the evidence and help inform future interventions.”

Third, the fact that policy change takes place over a long period of time and “involves multiple players ‘hitting’ numerous leverage points” increases the difficulty of teasing apart the contribution of any one actor or effort [30]. Further, having multiple programs and strategies taking aim at different but related parts of a broad problem like food insecurity may enhance effectiveness [26] and prove more realistic in today’s complex world.

Finally, an often-overlooked question in the assessment of municipal interventions involves the possible unintended contributions of a new program or policy toward a problematic outcome.

Of great concern in HRSF and related programs [31] are fears of unwittingly contributing to the already rapid gentrification of affected urban neighborhoods. While positive in increasing access to healthy foods among low-income residents and bringing new revenue and customers into stores stocking healthier fare, such changes may contribute to squeezing out low-income residents and small businesses, as their rents continue to soar [31]. Although the “three-legged stool” of the HRSF business model and particularly its community engagement emphasis have tried to help buffer against the displacement of local residents, more attention to the implications of gentrification and how it might equitably be addressed is needed. Academic and other partners who work with communities on corner store interventions are well positioned to partner, as well, on studying and helping address problems like gentrification in these communities.

Kingdon’s three streams of the policymaking process [8] proved a useful framework for studying the Coalition and the HRSF program it helped bring to fruition. As this case study has suggested, multiple methods of evaluative data collection “can together improve a partnership’s assessment of its contributions to changing a policy or the policy environment” [32].

Another takeaway was the imperative of community-led data gathering and high-level involvement in all project phases. FJLs had much greater access to merchants than the outside academics, enabling detailed and repeated observational assessments in two-thirds of stores.

Local members of the Coalition also knew what to ask, how to ask it, and which organizations to go through in getting hundreds of residents to take part in a survey of their shopping habits and concerns. Their sharing of findings proved critical in reaching the media, public, and policy makers. Further, and whether educating merchants, leading forums, or speaking at municipal hearings and advising the city on implementation and sustainability, community leadership is the lifeblood of efforts like this one.

The imperative of “doing your homework” before pushing for a policy measure that requires broad-based support from diverse constituents also was demonstrated. For merchants who saw alcohol and tobacco as important sources of income, the FJLs’ ability to discuss the lower profit margins on these products compared to fresh produce was critical. Similarly, the Coalition, DPH, and other partners’ advance work to develop means for monitoring implementation and measuring progress and outcomes underscored for policymakers that this was a well-thought out proposal to which the Coalition and its partners would continue to contribute beyond the implementation phase. Particularly in uncertain economic times, the importance of accenting for government the potential funders and supporters for future sustainability of a new project is critical. Ongoing collaboration between DPH and the Coalition, and later with academic partners, helped raise foundation and other support for HRSF.

Local leadership development, community capacity building, and increased visibility helped keep the Coalition and the city’s HRSF program in focus, with the Coalition and its partners widely seen as viable and thriving contributors to the fight against inequities in healthy food access and neighborhood saturation with tobacco, alcohol, and junk food. “Food insecurity is a civil rights issue,” just as tobacco control is a health equity issue, if poor neighborhoods are no longer to bear the brunt of disproportionate exposure to the advertising and availability of cigarettes and the like. Policy-focused CBPR can play an important role in studying and addressing such injustices, while building community capacity and visibility as key players in helping make policy and community level change happen.