Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Ameliorating Food Desert Conditions Most of the second day of the workshop focused on interventions to change food deserts. Some of these interventions were designed as research intervention trials and these were discussed in session 4. Ses- sion 5 addressed several promising, although less formally evaluated, programs and policies that are currently under way to improve the food environment. These interventions range from incentives for grocery stores and supermarkets to locate in underserved areas, to city-wide programs to encourage healthier eating, and extend to support for small, corner- type stores and neighborhood-based farmers markets. RESEARCH INTERVENTIONS Researchers have been evaluating different interventions to amelio- rate food desert conditions. These include efforts aimed at changing the food environment in many different ways. Overview of Efforts to Change the Food Environment Joel Gittelsohn, of Johns Hopkins University, presented an overview of efforts to change the food environment and reminded the group that food outletsâincluding supermarkets, small food stores, restaurants, and school and worksite cafeteriasâare all part of the larger community nutrition environment. Policy, environmental, and individual variables 45
46 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS Policy Variables Environmental Variables Individual Variables Behavior Community Nutrition Organizational Environments Nutrition Environments â¢ Type & Location of Food Socio-demographics Home School Outlets (stores, restaurants) â¢ Accessibility (hours of Work Other Government operation, drive-thru) Psychosocial Eating and Factors Patterns Industry Policy Consumer Nutrition Environment â¢Available healthy options Perceived â¢Price, promotion, placement Nutrition â¢Nutrition Information Environments Information Environment (such as media and advertising) FIGURE 5-1â Model of community nutrition environments. Figure 5-1.eps SOURCE: Glanz et al., 2005. Reprinted with permission from the American ÂJournal of Health Promotion. combine to affect eating patterns on an individual and collective basis (see Figure 5-1). Given this caveat, changing the food environment has many poten- tial benefits. Among these benefits, such changes can limit or expand the range of choices available to consumers, increase access to healthy foods, complement individual behavioral change programs, reach large numbers of people, and provide long-term sustainability if efforts are institutional- ized. It is a practical way, perhaps the only practical way, to address the obesity epidemic. In addition to altering access, the food environment can also be changed within stores, within neighborhoods, and in other settings through provision of information and promotions to consumers. In all cases, the link between supply and demand is key to determining whether changes in the environment will be linked with healthier eating. As Gittelsohn termed it, the âtrifectaâ is to increase availability, reduce price, and promote healthier choices. Availability in Large and Small Stores Seymour et al. (2004) reviewed 11 supermarket intervention studies, 8 of which provided information about healthy foods to consumers and 3 of which combined information with changes in access, availability, and incentives. Six of the studies (four with information only) showed increases in sales of healthier foods, while five did not show a change. Of
AMELIORATING FOOD DESERT CONDITIONS 47 the three studies that also examined dietary data, one showed increased consumption of healthier foods and two did not register impacts. Based on this review, Gittelsohn concluded that informational shelf labeling seems to work, while incentives, in the form of coupons, had little impact. However, a longer duration of up to two years may be needed to show any significant change. Since 2004-2005, researchers have conducted a number of intervention trials in small stores, which are often the main source of retail food pur- chases among low-income, ethnic, minority populations. Gittelsohn and a colleague are now in the process of reviewing 14 such studies: 5 studies confined to the stores and 9 that combined in-store interventions with community social marketing. These studies indicate a potential for suc- cess, as measured by reported improvements in fruit and vegetable sales, consumer psychosocial behaviors, healthy food purchasing patterns, and consumer diet. Challenges to increasing fruit and vegetable availability in small stores include convincing store owners to stock healthier foods, especially fresh fruits and vegetables that are perishable and require spe- cial handling. He suggested first trying to convince small store owners to stock less risky (e.g., nonperishable) healthy foods, such as low-sugar or high-fiber cereals. In many cases, store layouts pose a barrier; some are so enclosed that customers cannot touch a food item until they purchase it. These closed settings also severely limit social engagement between the customer and the clerk and therefore create barriers for nutrition educa- tion opportunities. Price Manipulation Types of price manipulation include lowering prices of healthy foods, offering coupons and other incentives, and increasing prices of unhealthy foods to subsidize lowering the costs of healthy foods. The CHIPS (Chang- ing Individualsâ Purchase of Snacks) study (French et al., 2001) showed that modifying the prices of low-fat snacks in vending machines increased sales and did not decrease profits. Other studies have looked at price subsidies in school cafeterias and showed that healthier food intake con- tinued even when the subsidies stopped. Gittelsohn stated that research on price manipulation in stores as a public health intervention is needed, but one difficulty in setting up a price trial in a food store is that retailers are reluctant to share their pricing strategies or to give up control over this key aspect of their business. Other questions to resolve in changing the food environment include how to build and sustain community support, the role of locally produced foods, and the optimal combination of institutions to involve. Certain aspects of the food environment have been commonly measured, while
48 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS others still need examination. For example, sales of selected foods, such as fruits and vegetables, are commonly measured, while total sales, which represent total intake, are not; it is difficult to measure impacts without knowing the numerator and denominator. Psychosocial considerations that influence store manager decisions, which could affect their expecta- tions for stocking healthy foods, are rarely measured. On the consumer side, the impact on behavior and on health, in terms of actual food prepa- ration and diet and in terms of BMI, is rarely measured. Studying the Introduction of a New Supermarket in a Food Desert Neil Wrigley, of the University of Southampton, United Kingdom, reported on a ânatural experimentâ first mentioned on day 1 of the work- shop: the opening of a new supermarket in a food desertâin this case, a supermarket regenerated in an urban underserved area of Leeds, England. In the late 1990s, he said, the metaphor of a âfood desertâ captured Brit- ish policy makersâ attention. Reports and inquiries linked trends in retail development in which food stores were moving outside of urban areas and toward the edge of town to the development of food deserts and to public health consequences. However, empirical evidence on key aspects of these linkages was limited. The Leeds Urban Regeneration Supermarket Intervention Study (Wrigley et al., 2003) was set up to link the policy debate to an evidence base and to assess the impact of a non-healthcare intervention, specifically a retail provision intervention, on food consumption patterns. The study was developed rapidly in response to an opportunistic possibility to con- duct a ânatural experimentâ when one of the UKâs first urban regenera- tion partnership stores was being constructed. Although possibly overly ambitious, Wrigley said the study established important benchmarks for subsequent retail provision intervention studies and was characterized by high-quality social survey data collection. The focus was Seacroft, an area of about 15,000 households in one of the most deprived wards of England. By the 1990s, it had a crumbling shopping center with poor levels of food retail provision. Buying food entailed either leaving the area or using a limited range of smaller stores. In partnership with the city, a labor union, and a government agency, a large Tesco supermarket plus 10 smaller shops and other facilities opened in 2000, amidst much fanfare including a visit from then-Prime Minister Tony Blair. The intent was to improve food access along with increasing employment and revitalizing the local economy. The Leeds study involved a two-wave household panel survey: the first in summer 2000, five months before the supermarket opened, with 1,009 respondents; the second in summer 2001, seven months afterward, involv-
AMELIORATING FOOD DESERT CONDITIONS 49 ing 615 of the original group. A separate repeatability survey and focus groups were also carried out. Of the respondents, 45 percent switched to using the new store as their main food source, and 31 percent (nearly three times more than before) reported that they walked to the store to food- shop rather than relying on vehicles (often either taxis or borrowed cars) to travel to places further away. Small but significant increases in fruit and vegetable consumption were found among users of the new store. Qualita- tive evidence from focus groups found that people appreciated the benefits relating to ease of access, affordability, quality, and safety, although some were worried about temptations to overspend and were concerned about more affluent shoppers coming from outside areas. Wrigley reviewed the Leeds study, as well as the Glasgow study described by Steven Cummins (see Chapter 3), to draw conclusions about supermarket intervention studies. He recommended that future studies should take sample size-statistical power and endogeneity-simultaneity issues more seriously; attempt to assess unintended consequences of the intervention; try to separate the impacts of physical access, economic access, and choice on food consumption; and appreciate the linkages between existing intervention studies and the dynamics of the food envi- ronment. Natural experiments, he said, change reality. They do not take place in a scientific vacuum and can fundamentally change the public discourse. Retailers in the United Kingdom are sensitized to the issue of food deserts to further their âenlightened self-interest.â There are now 35 urban regeneration partnership stores in the United Kingdom, and Tesco is opening up stores in south-central Los Angeles under its Fresh & Easy brand. Wrigley concluded by noting that some academics in the United Kingdom are more comfortable with alternative food network solutions, rather than supermarkets, in addressing food deserts. While solutions such as farmersâ markets have a role to play, he disagreed with his UK colleagues and questioned the extent to which they can penetrate socially excluded areas, at least in the United Kingdom, and have an impact on public health problems. Working with Small Stores to Promote Healthy Eating Guadalupe Ayala, of San Diego State University, described her research with tiendas, which she described as small Latino-Hispanic gro- cery stores with at least 50 percent of store shelf space devoted to food products, including fruits and vegetables, ready-to-eat foods, and meat. Also called bodegas by some Spanish speakers, these stores are very plenti- ful in Latino communities. They play important social and economic roles in both new and established immigrant-receiving communities, and for
50 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS new immigrants they often serve as a gateway into U.S. communities. In studies funded by the USDA and the National Cancer Institute, Ayala found that households shop at these types of stores an average of eight times per month, and they represent 33 percent of a familyâs total food basket and 84 percent of a familyâs total produce purchases, with much of the rest purchased at supercenters. Ayala said her research shows that working with tiendas and other small grocery stores may be an effective method to address the prob- lem of food deserts. The study, Vida Sana Hoy y MaÃ±ana (Healthy Life Today and Tomorrow), examined the efficacy of a food marketing and environmental change intervention to promote sales and consumption of fruits and vegetables. Although Latinos eat more fruits and vegetables than other demographic groups, acculturation has a negative impact: the longer people have been in the United States, the lower is their fruit and vegetable intake (Duffey et al., 2008). The primary outcome measured in the study was the number of daily servings of fruits and vegetables. It secondarily measured their total variety, behavioral strategies to increase fiber, and psychosocial factors, such as perceived self-efficacy to purchase more produce. The intervention was a randomized controlled trial in four North Carolina tiendas. It included employee and manager training, structural changes in the stores, and store-centered food marketing campaigns. The training enabled store personnel to become âfruit and vegetable special- ists,â as well as strengthen their selling and marketing strategies. The stores received $1,000 each to prepare and display packages of fruits and vegetables called âPronto Paks.â The food marketing campaign included recipes, point-of-purchase materials, and a radio program. Consumer fruit and vegetable intake increased with this intervention by about one additional serving per day (see Figure 5-2). Self-efficacy in terms of purchasing and using fruits and vegetables declined, possibly because respondents felt less capable as more was learned and awareness heightened, especially in the short run. Ayala noted some challenges in small-store interventions. The own- ers may be reluctant to participate in government programs. For exam- ple, researchers first suggested tapping into a program sponsored by the North Carolina Department of Agriculture to link food retail businesses with local farmers, but the tienda owners did not want to get involved. They have no mechanism for electronically tracking sales data, which makes it hard to know what is sold. Follow-up is especially difficult with new immigrant populations, with only about two-thirds located for follow-up 10 months after the intervention. Ayala said that in terms of identifying what foods to target in future interventions, Latino stores tend to stock far less low-fat dairy (and at higher relative prices) and more
AMELIORATING FOOD DESERT CONDITIONS 51 Increase of 1 daily serving of fruits and vegetables. 2.81 2.51 2.51 1.87 p â¤.01 FIGURE 5-2â Effect of tienda intervention on consumption. SOURCE: G. Ayala, 2009. Figure 5-2.eps bitmap image with some vector type sugar-sweetened beverages and sweet and savory snacks, compared to non-Latino stores. Farmers Markets in Low-Income Communities Farmers markets, although a tiny percentage of the overall food envi- ronment, are expanding (see Box 5-1). Research from East Austin, Texas, shows some of the elements to consider in food deserts. Andrew Smiley, of the Sustainable Food Center (SFC), discussed its experiences in establishing farmers markets in East Austin, Texas. A well- intentioned effort to establish a central market did not succeed in the long term. Instead, SFC has found that smaller markets near Women, Infants, and Children (WIC) clinics are meeting the communityâs needs. A 1996 study of food access in East Austin found food desert condi- tions (SFC, 1995): two supermarkets, other retail food outlets with poor variety and generally higher prices, and transportation to outlets outside the neighborhood expensive and difficult to arrange. The targeted area is a predominantly minority low-income community, with about 25 percent of the 24,000 residents under age 12. Of the 38 convenience stores in the neighborhood, fewer than half stocked milk and only 5 stocked all of the ingredients for a well-balanced meal. One solution, still in operation, was a bus route, specifically designed to reach supermarkets, that has not been evaluated for its impact.
52 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS BOX 5-1 Farmers Markets: Small But Growing Market Segment Farmers markets occupy a small, but growing share of the U.S. food environ- ment. According to workshop participant Debra Tropp, of the USDA Agricultural Marketing Service, farmers markets accounted for about $1.2 billion in sales in 2007, up from $812 million in 2002, or a 30 percent increase after adjusting for inflation. In 2008, there were 4,685 markets operating, up from 1,755 in 1994, with more than 135,000 farms involved. Consumers are also using farmers markets as a source of fresh fruits and vegetables. Workshop participant Heidi Blanck, of the Centers for Disease Control and Prevention (CDC), mentioned a recent consumer survey conducted by the CDC and the National Cancer Institute on the use of farm-to-consumer venues and food attitudes and behavior. The survey showed that in 2007, one in five adultsÂ self-identified as primary food shoppers reported shopping at farmers mar- kets (20.1 percent). When asked how often in the summer they purchased fruits and vegetables from a farm-to-consumer venue (i.e., farmers market, farm stand, pick your own farm, community-supported agriculture), 56.1 percent of primary food shoppers reported at least monthly use and 27.1 percent reported weekly use. Weekly use was higher among middle-aged and older adults and lower in the South and Northeast compared to the West. Weekly use did not differ by sex, race or ethnicity, education, or income. SOURCES: H. Blanck, 2009; D. Tropp, 2009; USDA NASS, 2009. Another proposed solution was a centrally located farmers market. SFC already operated a large market in downtown Austin and several smaller ones at WIC clinics in other neighborhoods. To open a large market in East Austinâs Saltillo Plaza in 2003, SFC recruited vendors, pro- moted the market, took out insurance, managed operations, and Âcarried out a host of other tasks. Despite outreach, special events, and other efforts, evaluation indicated that the market lacked sufficient support for the resources expended. The majority of sales were due to WIC Farmersâ Market Nutrition Program (FMNP) vouchers, but these went only to fruit and vegetable farmers. Surveys and focus groups indicated that word of mouth generated awareness about the market, but people felt it was inconvenient for regular shopping. Those who did come especially valued the fruits and vegetables and, across income levels, the idea of supporting local farmers. However, there just were not enough customers to justify SFC or farmersâ costs, and SFC decided to close the market in 2005. Produce sellers were still interested in additional sales outlets, and customers with FMNP vouchers still wanted to purchase their products.
AMELIORATING FOOD DESERT CONDITIONS 53 In an alternative and still successful approach, SFC decided to support very small (one or two) farmers markets located next to WIC clinics. WIC clinic markets are not as staff-intensive, and WIC staff members are good partners in outreach. Six of these clinics are now operating. Discussion: Research Interventions Terry Huang, of the National Institute of Child Health and Human Development at NIH, moderated the session on food desert intervention research. Main points included the need for multicomponent interven- tions, formative research, and more robust price manipulation trials. Making Farmers Markets Viable Huang related questions about setting up farmers markets, espe- cially near WIC clinics. Smiley said that the state FMNP program made it easier for individual clinics to participate. SFC has liability insurance for the markets it coordinates, which greatly eased liability concerns. Some neighborhood convenience stores saw the markets as a threat. This required some outreach to explain that the small markets take FMNP vouchers almost exclusively and did not pose any competition. In answer to a question about locating markets so that they simultaneously straddle more and less affluent neighborhoods, Smiley noted that determining new sites is both a physical and a psychological issue. In the case of Aus- tin, for example, location vis-Ã -vis a highway that runs through the city makes a big difference psychologically and logistically. Incentives Gittelsohn noted that incentives can go to shoppers or to store own- ers. In addition to coupons, he has been involved with trials in which, for example, a shopper gets one free (healthy) item for every four purchased. However, these have not been successful in small stores, in part because owners worry about consumers abusing the incentives. Ayala said she was involved in an experiment in which a customer would get one free pound of produce for every 10 pounds purchased. It became too much of a burden for the store owners and was not sustainable. In contrast, incentives to small-store owners, such as $25-50 gift cards they can use with wholesalers, have been promising, Gittelsohn said. Ayala said interviews with tienda owners in California revealed that they try to make every square foot of the store as profitable as possible. They are willing to try something for a month to see if it will bring in more customers or profit.
54 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS The cost for small stores to equip themselves to sell fresh produce varies. In California, store owners told Ayala it would cost between $5,000 and $10,000 for refrigeration units to stock fruits and vegetables. How- ever, even with small additions that cost $1,500 to set up, changes were seen in customersâ dietary intake. In addition, states and counties may have different policies and regulations, which have an impact on imple- mentation costs for store owners. Defraying theses costs would help store owners that usually operate on thin profit margins. The Role of Price As noted throughout the workshop, price matters. Gittelsohn said literature is scarce on food store price manipulations for public health interventions, making it difficult to arrive at definitive conclusions about what does and does not work. That is, although it is widely known that lowering price is used to increase sales volume, there is little documen- tation or evaluation of the use of prices to improve the healthfulness of diets. Intuitively, increased availability and point-of-purchase promotion encourage people to try new foods, but they have to cost less or at least about the same price as less healthy alternatives. Wrigley observed that improved access and improved price frequently go together, so it is hard to distinguish one from the other. He suggested that urban regeneration stores could perhaps carry out pricing experiments. Smiley noted that testing perceptions is important, too. For example, a survey found some people do not shop at farmers markets because they perceive them to be more expensive, even when prices are comparable. Ayala said that low-fat milk is consistently more expensive than whole milk across Latino stores, which is not the case in other communities. Latino households drink more whole milk, and the price differential may be because of market demand. Another factor that may discourage sales of lower-fat products relates to packaging: low-fat products sometimes do not have âVitamin Dâ on their labels. While there are many barriers to changing learned behaviors and preferences, the price difference would have to be resolved first before educational campaigns are considered. Popkin noted that food processors continually manipulate price as part of their business practices. For example, food processors have found prepackaged vegetable servings to be successful, but these items are also priced well and highly promoted, which makes the separate elements difficult to evaluate. Wrigley reiterated findings from focus groups about worries from exposure to the temptations of full-scale retailers. The issue of being tempted to overspend fixed budgets is a big issue, and obviously a price-related issue.
AMELIORATING FOOD DESERT CONDITIONS 55 Multicomponent Approaches Multi-institutional, multicomponent approaches are a natural exten- sion of current research looking at single parts of the food environment. Formative research that involves community participatory processes would help plan interventions. The type of foods to focus on may vary by location and culture, but sweetened beverages seem to be a key prob- lem across many low-income settings. Huang asked whether beverages should be addressed from both the demand and the supply sides: dis- couraging their purchase but also working with companies to reformu- late the product. Gittelsohn has done some preliminary work with local manufacturers and distributors in terms of changing the mix offered to stores. Customers claim they do not eat healthy foods because the foods are not available, cost too much, or are of poor quality; on the other hand, store owners claim they do not stock the healthy foods because nobody buys them. Almost as a mediator, public health and other specialists can concurrently convince store owners to increase supply and work with consumers to increase demand. Another need, according to Ayala, is to better enumerate stock inven- tory. The type and number of products in tiendas is far different than those in convenience stores, and Ayala asserts that food desert conditions may not exist in predominantly Latino Hispanic communities. It would be interesting to understand whether the food environment is one of the factors that explain the âHispanic paradox,â which suggests that first- generation Latinos have better health outcomes than their acculturated counterparts despite greater poverty and lower socioeconomic status. E-commerce The question was raised about using the telephone or Internet to place orders to lower costs and increase access. Smiley said the SFC operates a Farm-to-Work program, a subscription program for employees at partner work sites. In another project in Austin, WIC families pre-order their full mix of groceries for delivery to centralized locations. Gittelsohn said cell phones are so widespread that using them in this way could offer an inter- esting possibility. Wrigley said the origins of e-commerce in the United Kingdom were to reach underserved markets. An experiment in New- castle from 1980 to 1982 involved computer access from public libraries in underserved areas of the city to place grocery orders. Ayala observed that the digital divide is not as pronounced as it was a few years ago, but interpersonal relationships are too strong to make e-commerce for food a viable option, at least in the Latino community. Diez Roux warned about creating new problems while supposedly fixing others. For example, hav- ing a store or other destination to travel to promotes physical activity and
56 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS social interactions. The much bigger problem is the way in which environ- ments affect a variety of health behaviors at the same time. Huang asked workshop attendees for further observations. Sarah Trehaft, of PolicyLink, urged listening to community voices in discussing food desert measurements, policies, and interventions. Popkin agreed that community participation and feedback are essential to successful approaches and outcomes. POLICY INTERVENTIONS Research interventions to modify the food environment are attractive because they are fundable and measurable, however, research interven- tions are merely one solution to solving food deserts in communities. The policies and programs discussed by the next session also aimed at improving access to healthy food in food deserts through supermarkets, corner stores, farmers markets, and other outlets. While these interven- tions were not set up as research experiments, they are, nevertheless, an interesting mix of initiatives launched by government agencies and grassroots efforts that began at the community level and became more widespread (see Box 5-2). Most have or are planning some types of forma- tive research and evaluation. Determining Sites for New Supermarkets In developing strategies to increase the number of supermarkets in food deserts, it is important to understand the prevailing business model of U.S. supermarket chains. According to William Drake, a former super- market executive now with Cornell University, most food retailers are aware of the issues of food security and urban food deserts. However, a combination of internal capabilities, external trade area characteris- tics, economic realities, and intolerance for risk raises difficult barriers to overcome and hence the difficulty for supermarkets to site new stores in food deserts. Drake described the prototype for most supermarkets in the United States today: large (48,000 square feet), with plentiful parking (250 spaces), and sales volume of $400,000 per week, all difficult to achieve in most core urban areas. About two-thirds of the nationâs 34,000 supermarkets belong to a chain, defined as stores with at least 11 units. The profit margin is thin, in the range of 1.5 to 1.75 percent of sales. To maximize profit, the most successful chains are âfinely tuned machinesâ that know their target consumers and operate in ways to attract them to buy. To diverge from their model, he said, is very inefficient and more likely to fail. Super- markets in the past 20 years have tended to locate in middle- and upper-
AMELIORATING FOOD DESERT CONDITIONS 57 BOX 5-2 Top-Down and Bottom-Up Approaches Top-down and bottom-up approaches have been used to launch and coor- dinate efforts to improve food access. âTop downâ refers to initiatives emanating from a government agency or other institution; âbottom upâ refers to initiatives that begin in neighborhoods or in community-based organizations and become larger programs or policies. For example, in 2002 the New York City Department of Health initiated a con- certed effort to focus health intervention in the South Bronx, Central and South Brooklyn, and East and Central Harlem. These were areas with high rates of poor health outcomes, including obesity and diabetes. Survey data also show that these areas have the lowest rates of fruit and vegetable consumption in the city. These District Public Health Offices have been the central point for four food access projects: (1) improving food choices in corner stores, (2) increasing the number of farmers markets, (3) increasing or at least maintaining the number of supermarkets, and (4) encouraging fruit and vegetable vendors. Although involving the community, these initiatives began at âthe top.â In Philadelphia in 1992, the nonprofit Food Trust began a community-level m Â ission to âensure that everyone has access to affordable, nutritious food.â The organization helped set up farmersâ markets in low-income neighborhoods to i Âmprove access to affordable nutritious food and now manages 30 markets in the greater Philadelphia area. income suburban locations where they have a good chance of meeting their targets. Chains also tend to site new stores relatively close to their existing outlets for a number of economic and logistic advantages. Once a general geographic area is identified, the chain makes site- specific decisions. Its business analyses rely on variables such as projected sales, occupancy costs, and labor expenses, among others. These models work less well in urban settings because the underlying data are often underestimated or misrepresented. For example, it is difficult to gauge sales when the current competitors are small corner stores rather than other supermarkets. Without good knowledge of an area, Drake said, sit- ing decisions are more prone to fail. State and local governments can assist retailers in entering urban mar- kets by providing real estate or establishing public transportation stops to commercial locations and food stores, particularly in inner cities. Yet the fact remains that locations in urban food deserts do not fit the positioning strategy of most large chain supermarket operations. As an alternative, he suggested working with voluntary and cooperative food wholesalers, the segment of the industry that serves independent retailers. Independent
58 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS retailers are better able to customize a positioning strategy and adapt to local conditions than the larger chains. Another idea is to target specific retail stores that have a business model with a better chance of success in urban inner-city markets. Known as limited assortment hard disÂcounters, these stores offer healthy foods at competitive prices but with a more limited assortment. Policies to Encourage Supermarket Entry As reported by Drake, the average supermarket in the United States is about 48,000 square feet and is set in a suburban location with plenti- ful parking. However, several presenters reported on variations that fit better in urban environments: smaller stores, perhaps 12,000 to 15,000 square feet, with more limited parking and convenient public transport or shuttles to help shoppers take their food home. John Weidman, of the Philadelphia-based the Food Trust, described how the nonprofit built on its many years of community food work to help develop the Fresh Food Marketing Initiative in Pennsylvania. As in Chicago, New York City, and other places, mapping in Philadelphia showed the coincidence of a lack of supermarkets with a high incidence of diet-related diseases. The mapping study sparked the interest of the City Council, which requested that the Food Trust convene a group of public health, economic development, government, and supermarket industry representatives to understand why stores did not locate in these commu- nities and what policies could fix the problem. The State of Pennsylvania also held hearings, and in 2004, this work culminated in the Fresh Food Financing Initiative (FFFI), the nationâs first publicâprivate funding initia- tive set aside for retailers to open and update stores in underserved food deserts. The FFFI is a $120 million initiative that funds food retail projects in underserved areas. It provides grants of up to $250,000 per store and loans of up to $2.5 million per store. Since 2004, it has funded 58 stores of various sizes that have provided almost 3,500 jobs. Most of the larger stores are independent or small, locally based chains. Spatial analysis confirms that these stores have gone into many areas with the greatest need. The Food Trust is now working on evaluating the health outcomes and expanding its efforts in other states, including New York, Louisiana, Illinois, and New Jersey. Cathy Nonas, of New York Cityâs Department of Health & Mental Hygiene, described how the city is trying to increase, or at least maintain, the number of supermarkets operating in high-need neighborhoods. A city planning standard was set to aim for a store of 15,000 square feet to serve on average 10,000 people living in a five-block radius. A superÂ
AMELIORATING FOOD DESERT CONDITIONS 59 market commission was established, with assistance from the Food Trust, to look at zoning regulations and tax incentives. Some city-owned spaces have been identified as potential sites for new stores. At the same time, community-based organizations and unions are working with city and state officials to stop the closing of supermarkets, and they have had some success in Harlem. SMALL STORES A point brought up throughout the workshop is that improving the food offerings of existing stores in a community can be a feasible solution to accomplish the goal of making healthy foods convenient and afford- able. For example, on the first day of the workshop, Joseph Sharkey suggested focusing on where people currently shop when he presented an overview of the rural Brazos Valley. Gittelsohn reported that research into small-store interventions has greatly increased since 2004, and Ayala shared her findings from small stores in Latino communities. In this ses- sion, Nonas and Weidman explained several small-store programs in New York and Philadelphia. Healthy Bodegas in New York City The New York City Department of Health & Mental Hygiene has targeted areas in three parts of the city where it is trying to improve food access, as previously reported in Box 5-2. Nonas explained that through the Healthy Bodega Initiative, the department is encouraging existing stores in these areas to improve their offerings of healthy foods. In a first phase, three district public health officers worked with about 350 bode- gas each (more than 1,000 total) to increase availability and purchases of low-fat milk. Extensive consumer education accompanied outreach to the bodegas (Figure 5-3). In the next phase targeting fruits and vegetables, more than 450 bodegas participated; the smaller number was chosen based on the storeâs interest in selling and increasing its quantity of fresh produce. Depending on the storeâs characteristics, the department helped bodega owners increase quantity, improve quality, provide prepackaged items, market healthy foods better, or obtain the appropriate permits to sell processed produce and produce in front of the store or on the stoop. The two campaign phases saw large increases in sales of low-fat milk and fruits and vegetables, although Nonas acknowledged that it is hard to evaluate the effect apart from other factors, such as offering WIC par- ticipants coupons for low-fat milk. She listed challenges in sustaining the initiative: not enough staff to reach out to so many stores and visit them sufficiently, the need to balance outreach efforts between community
60 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS FIGURE 5-3â Consumer materials used in the Healthy Bodega Initiative. SOURCE: C. Nonas, 2009. Figure 5-3.eps bitmap images buy-in and bodegas, limited infrastructure in many bodegas (such as refrigeration and storage), haphazard distribution systems, and the need for micro-loans and education to make the necessary improvements. The department is now working more closely with fewer bodegas to make more sustainable and substantive changes and to increase healthy food options in those stores. Each district public health officer works with 20 bodegas, chosen to ensure that each resident is within walking distance of at least one healthier bodega, which they can visit at least twice a month. As Nonas said, if three bodegas are located on a block, maybe only one needs to carry fruits and vegetables. Nonas noted that the department also works with these stores to decrease tobacco ads and identify healthier items in the store with promotional materials, and works with other city agencies and organizations (such as milk distributors, produce distribu- tion sites, micro-loaners, and permitting centers) to make it easier for city bodegas to stock and sell healthier items. Making Healthy Food at Corner Stores Kid Friendly John Weidman reported that the Food Trust works with Philadelphia corner stores where children stop for snacks. After starting a school nutri- tion program, the Food Trust did research on the role of corner stores in childrenâs nutritional status. They found that children consumed about 600 calories in snacks, almost all of them unhealthy, spending about $2 per day. In a pilot program in five stores in North Philadelphia, refriger- ated coolers were set up with fresh fruit. The attention-grabbing coolers and attractive packages appealed to kids, and sales were brisk. A private operator has taken on and expanded distribution to 50 stores, which solves the problem of a sustainable distribution network. A plan to create kid-pleasing water bottles is next.
AMELIORATING FOOD DESERT CONDITIONS 61 Initial evaluations show kids purchasing items with less fat and fewer calories. The owners are purchasing more healthy items, and the childrenâs knowledge of healthy eating is improving. The Food Trust is currently midway through a more rigorous randomized study to track a group of children for three years and look at BMI and calorie reduction. FARMERS MARKETS AND OTHER ALTERNATIVES IN LOW-INCOME COMMUNITIES As described above, farmers markets are a small but growing part of the food environment in the United States. Across income levels, consum- ers have shown they are willing to frequent these markets and, in many cases, they prefer the social interaction and direct contact with the grow- ers. At the same time, as Smiley described, they need to be convenient and worthwhile from the point of view of both the buyer and the seller. Panelists shared their experiences from markets in several cities, as well as more cross-cutting lessons about community buy-in and the vital con- nection with government programs. Health Bucks for Fresh Fruits and Vegetables Nonas described how the New York City Department of Health & Mental Hygiene has looked for ways to support low-income residents in purchasing fresh produce at farmers markets, in part by increasing the number of farmers markets in the target areas and enabling them to accept Electronic Benefit Transfers (EBT) from participants in the USDAâs Supplemental Nutrition Assistance Program (SNAP, formerly know as the Food Stamp Program). As an incentive to shop at neighborhood farmers markets, the city dis- tributes âHealth Bucks,â which are $2 coupons for the purchase of fresh fruits and vegetables. In 2005, the program began by distributing these coupons to community-based organizations for their constituents. The original program was so successful that in 2007, additional Health Bucks were distributed in the form of a $2 bonus for every $5 spent using EBTs at a farmers market. In addition, the department encouraged farmers to sell different types of healthy produce (such as peaches and plantains), and sales have been good. Health Bucks have been a major source of income for farmers selling at these markets and have ensured that markets profit in these low-income areas. Nonas mentioned that EBT sales have also skyrocketed since their introduction. Nonas said it takes a multiagency effort to make Health Bucks suc- cessful and, more generally, to keep farmers markets in low-income com- munities thriving. An ongoing challenge is how to pay for EBT machines
62 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS at the markets, as well as a market manager to oversee the machine and coordinate reimbursement to farmers. Also, she noted the need to con- tinue access to fruits and vegetables when the farmers market season is over because most markets in these areas are not year-round. Lessons from Philadelphia The Food Trustâs work with farmers markets has been successful in neighborhoods at all income levels, mentioned Weidman. In annual surveys, customers report significant vegetable intake and more frequent snacking with fruits and vegetables. He identified key components in making farmers markets successful in low-income communities: â¢ Strong management â¢ ite selection to locate a market in a place that works for the S c Â ommunity and the vendors â¢ Strong community partnerships â¢ Connections with government programs â¢ Marketing A pilot program in which each vendor stand had its own EBT machine resulted in increased usage. The machines processed credit cards too, so the farmers and other shoppers also benefited. Weidman suggested that this is maybe the future of vendor transactions as the economy becomes more âelectronic.â Involving Communities in California Andrew Fisher, of the Food Security Coalition, described the 250- member coalition as the connective tissue between advocacy and tech- nical assistance. He discussed three projects in greater Los Angeles to provide insights about the role of the community in food desert projects: the Villa Parke Center Farmersâ Market in Pasadena, Market-to-School project in Santa Monica, and urban agriculture project in Watts. The Villa Parke Center Farmersâ Market has operated in a low-income part of Pasadena for more than two decades. A benefit of this and other successful markets is that they build social capital as people interact with vendors and each other more so than in a supermarket. Many farmers hire helpers from within the community. Accepting government subsidy benefits is important to success in low-income communities: For this par- ticular market, farmers agree to participate so they can also sell at a more lucrative market in downtown Pasadena. Price is an issue at farmers markets in low-income communities for
AMELIORATING FOOD DESERT CONDITIONS 63 both shoppers and farmers. Shoppers are price-sensitive, but farmers also need to see a return on their labor. Product mix is crucial. Language and culture can create barriers between vendors and shoppers and affect sales, which is one reason that a community-organizing approach and local Âhiring are important. The market-to-school project in Santa Monica also succeeded through community support. Although Santa Monica is generally a well-off com- munity with no food deserts, Fisher noted that a large percentage of the school population is low income. The project began in 1997 when a parent worked with a local farmers market to improve the offerings of his childâs cafeteria salad bar. It has since spread to all the schools in the district, and one-third of children now regularly eat salad for lunch. The project was successful because it had the support of the school district and food service personnel. The logistics have been time-consuming but not insurmountable. Less successful, but a valuable lesson nonetheless in terms of the role of community support, was an urban agriculture project to provide vegetable gardens on a 2-acre plot near a housing project in Watts. The goal was to provide low- to no-cost healthy foods that could be grown year-round in the temperate Los Angeles climate and foster community interaction. Fisher said the project did not succeed because it was too top-down. The plan was presented to gardeners, some of whom were already tending plots, rather than involving them in planning how it would work. The gardeners were supposed to grow tomatoes and c Â arrots when they knew growing herbs was more lucrative. In addition to poor design, friction between the staff and the gardeners worsened the situation. Fisher noted that planners often see food as a private good, rather than a public benefit, or if so, under the purview of the federal govern- ment. As he commented, âThere are no [local] departments of food in the country.â Some policy solutions that could increase access on a commu- nity level include transportation changes, such as buses or shuttles, or a ban on lease restrictions that prevent new supermarkets from replacing those that close down. Publicâprivate councils, now operating in about 100 communities around the country, also bring diverse stakeholders to the table to find ways to enhance food access. Tapping into Public Funding The benefits provided through SNAP and WIC, as well as Supple- mental Security Income (SSI) and other federal and state assistance pro- grams, are critical elements in making farmers markets a going concern in low-income communities.
64 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS August Schumacher, Jr., a former USDA undersecretary of farm and foreign agriculture services and now consultant with the Kellogg Founda- tion, elaborated on how changes in the SNAP and WIC Programs could further reduce rural and urban food deserts. SNAP retailers (such as cor- ner stores and farmers markets) do not have to represent the bleak picture described in Detroit by Mari Gallagher (see Chapter 2). Federal and state nutrition funding is evolving. Starting in 2009, more than 6 million WIC mothers and children nationwide will receive vouch- ers to buy fruits and vegetables in supermarkets and farmers markets. WIC mothers will receive $8 per month, and WIC-eligible children will receive $5 per month for fruit and vegetable purchases. He expressed hope that the amount increases. Schumacher noted that Congress will be reauthorizing the vital Child Nutrition legislation later in 2009 and 2010. He believes that increasing the provisions for healthy food purchases in this legislationâparticularly, increasing the monthly funding for WIC mothers and children to purchase fruits and vegetablesâwould be excep- tionally helpful in partially alleviating the growing incidence of obesity and diabetes among Americaâs vulnerable children, many who live in food deserts. Schumacher also reviewed the status of new funding to promote healthy food incentives in the 2008 Farm Bill. The provision within Sec- tion 4141 is for a $20 million pilot program under SNAP to explore how the immense program can improve the dietary and health status of eli- gible households. Regulations are being developed within USDA to set up pilot programs. To support these evolving federal nutrition improvement programs, Schumacher cited a combination of foundations, states, and cities that fund healthy food incentivesâsuch as those in Holyoke and Boston, Massachusetts, and San Diego, Californiaâthat allow SNAP, WIC, and SSI clients to receive âdouble vouchersâ to increase their purchases of fruits and vegetables using SNAP EBT cards or WIC vouchers. He also noted that EBT cards could be an effective way to track fruit and vegetable purchases. Foundations such as the Wholesome Wave Foundation and the Humpty Dumpty Foundation provided funds in four states (California, Massachusetts, New York, and Connecticut) in 2008 and, with additional foundation partnerships, are considering expansion to eight more states in 2009. Schumacher said the diversity of the American food system, in terms of farmers and shoppers, will lead to growth in the sector. Data from the USDAâs Agricultural Census (USDA NASS, 2009) indicated 100,000 new farmers in the United States, with strong growth of Hispanic, Asian, Native American, and African-American farm operators. He is also opti- mistic about new demand drivers for fruits and vegetables, including an
AMELIORATING FOOD DESERT CONDITIONS 65 emphasis on child nutrition, healthier lunches and breakfasts at school, and expansion of SNAP and WIC use. He urged that new WIC vouch- ers add roadside stands to their permitted retail stores because many WIC mothers live in rural areas where farmers markets and full-service supermarkets are less prevalent. In addition, Schumacher suggested that funding for the Section 4141 Healthy Food Incentive program increase from $20 million to $100 million. He also said making it easier for farmers markets to get waivers to accept federal benefits would help them expand in low-income areas. DISCUSSION: POLICY INTERVENTIONS Robin McKinnon, of the National Cancer Institute at NIH, served as moderator for this panel discussion. The need for a connection between communities and food outlets ran through the questions and comments. Supermarket Incentives Drake said incentives at the state and local levels are especially impor- tant, such as helping to assemble real estate, job assistance, and training. Stores are only as good as their employees, and preventing turnover, which can be in excess of 100 percent per year in some inner-city stores, is a huge task. Transportation for shoppers and employees helps. ÂWeidman suggested a federal-level equivalent to the FFFI for publicâprivate fund- ing initiatives. Farmers Markets One workshop participant expressed concern about a relationship between farmers markets and gentrification. Weidman said this has not occurred in Philadelphia, but rather there is a revitalization factor that benefits the community. Schumacher said that sellers or employees from within the community help make markets successful. Fisher noted that certified farmers market in California must work with the community. Successful markets take about a year to get organized, including a process of amassing community support. The cost of setting up a farmers market varies. The Food Trust oper- ates a network of markets, which helps consolidate the workload. Nonas pointed out that it makes a difference whether the market accepts EBT, which adds costs. Often middle-sized farmers have the hardest time dedi- cating the resources to sell at markets, and they need to make enough money for this to be worthwhile. Weidman said that there was some con- cern about whether the Philadelphia area had enough farmers to set up
66 THE PUBLIC HEALTH EFFECTS OF FOOD DESERTS at all the markets. He said recent experience has shown there are plenty, including many younger farmers under the age of 30, an encouraging sign for the future of farming. Community Outreach As noted above, community acceptance is important to the success of farmers markets. Instilling ownership and buy-in is also important with supermarkets. Fisher said joint ventures have been helpful, such as a Pathmark in Newark, New Jersey, that partnered with the nonprofit New Community Corporation. Weidman noted that the bigger chains have not participated much in the FFFI, for reasons explained by Drake and others, but smaller independents have. Those that pay attention to com- munity needs find it pays off. He cited the example of the Brown Family Shoprite, which heeded community requests to offer halal meats, now a best-selling item. McKinnon asked about farmers markets in more dispersed rural areas. Schumacher said markets have been growing in rural areas, as he has seen in his work in Alabama and Mississippi. Weidman said that rural counties have also accessed the FFFI program.