**Associations between Governmental Policies to Improve the Nutritional Quality of Supermarket Purchases and Individual, Retailer, and Community Health Outcomes: An Integrative Review**

**Alyssa J. Moran 1,\*, Yuxuan Gu 2, Sasha Clynes 2, Attia Goheer 1, Christina A. Roberto <sup>3</sup> and Anne Palmer 4,5**


Received: 7 September 2020; Accepted: 30 September 2020; Published: 15 October 2020

**Abstract:** Supermarkets are natural and important settings for implementing environmental interventions to improve healthy eating, and governmental policies could help improve the nutritional quality of purchases in this setting. This review aimed to: (1) identify governmental policies in the United States (U.S.), including regulatory and legislative actions of federal, tribal, state, and local governments, designed to promote healthy choices in supermarkets; and (2) synthesize evidence of these policies' effects on retailers, consumers, and community health. We searched five policy databases and developed a list of seven policy actions that meet our inclusion criteria: calorie labeling of prepared foods in supermarkets; increasing U.S. Department of Agriculture (USDA) Supplemental Nutrition Assistance Program (SNAP) benefits; financial incentives for the purchase of fruit and vegetables; sweetened beverage taxes; revisions to the USDA Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) food package; financial assistance for supermarkets to open in underserved areas; and allowing online purchases with SNAP. We searched PubMed, Econlit, PsycINFO, Web of Science, and Business Source Ultimate to identify peer-reviewed, academic, English-language literature published at any time until January 2020; 147 studies were included in the review. Sweetened beverage taxes, revisions to the WIC food package, and financial incentives for fruits and vegetables were associated with improvements in dietary behaviors (food purchases and/or consumption). Providing financial incentives to supermarkets to open in underserved areas and increases in SNAP benefits were not associated with changes in food purchasing or diet quality but may improve food security. More research is needed to understand the effects of calorie labeling in supermarkets and online SNAP purchasing.

**Keywords:** food purchase; policy; retail food environment; food and beverage; grocery; federal nutrition assistance programs; beverage tax; menu labeling; financial incentives; health disparities

#### **1. Introduction**

Poor diet is widely considered a public health crisis, contributing to many of the leading causes of morbidity and mortality in the United States (U.S.) and globally [1,2]. There is growing recognition that dietary behaviors are shaped by the environments in which people live, learn, work, and play, and public health interventions increasingly target these settings [3]. Compared to nutrition interventions aimed at individuals or groups, upstream interventions designed to alter the environments in which people make food and beverage choices may be more effective for improving health, and are less costly to implement in the long-term [4,5]. For example, environmental interventions to treat obesity, such as sugary drink taxes and reductions in child-directed television advertising, are shown to be more cost-effective than commonly reimbursed clinical interventions, such as nutrition counseling or bariatric surgery [5].

In the U.S., supermarkets are natural and important settings for implementing environmental interventions to improve healthy eating. These stores, which generate more than \$2 million annually in sales volume, are the primary retail store choice for the vast majority of U.S. households [6]. According to data from the U.S. Department of Agriculture (USDA) Food Acquisition and Purchase Survey, in 2012–2013, 89% of households did their primary shopping at supermarkets or other large grocers, with only 5% doing their primary shopping at other stores (e.g., convenience or dollar stores) [7]. During this time, supermarket purchases made up the majority of calories purchased by U.S. households (65%) and accounted for between 56% (households participating in the USDA Supplemental Nutrition Assistance Program [SNAP]) and 64% (higher income non-participating households) of household food expenditures [6]. Restaurant closures necessitated by the COVID-19 pandemic have likely increased reliance on supermarkets as a primary food source for many households.

The in-store environment is a well-recognized and powerful driver of dietary behaviors in supermarkets. Prior work has documented the important role of in-store food and beverage marketing, including availability, affordability, prominence, and promotion, in shaping consumer choices [8]. While these strategies hold promise for promoting healthy choices, they are often used to increase purchases of ultra-processed, nutritionally-poor products. A study of nearly 70 food retailers in three states found that sugary drinks were the most commonly promoted beverage, displayed in an average of 25 locations throughout the store [9]. National survey data show that the nutritional quality of purchases from supermarkets is generally poor, with diet quality scores (measured using the Healthy Eating Index 2010) closely mirroring those from national surveys of dietary intake [6,10].

Prior research has assessed the effectiveness of in-store promotions for healthy foods, finding that changes to product pricing, availability, packaging, display, signage, and labels are associated with consumer purchasing in the short-term [8]. Implementing these interventions long-term and scaling them across the nation's more than 30,000 supermarkets, however, has proven challenging [11]. Grocery stores operate at low margins and rely on trade fees and discounts from food and beverage companies for revenue [12]. These fees often favor the largest manufacturers and distributors, allowing them to control which products are stocked and how items are promoted in the store. It is estimated that supermarkets collect more than \$50 billion a year in trade fees and discounts, with fees accounting for a large proportion of total grocery revenue relative to sales (although fees vary greatly by product, manufacturer, and store type) [11,12]. These exorbitant financial incentives make voluntary interventions to promote healthy purchases difficult to implement in supermarkets without food and beverage company buy-in.

In the absence of widespread and sustained voluntary action, governmental policies could help increase healthy purchases and decrease unhealthy purchases in the supermarket setting. To this end, several policies have been implemented across the U.S., and many studies have been conducted to evaluate their effects. This integrative review aims to synthesize the academic literature on this topic by: (1) identifying U.S. governmental policies, including regulatory and legislative actions of federal, tribal, state, and local governments, designed to promote healthy choices in supermarkets; and (2) summarizing the available evidence related to these policies' effects on retailers, consumers, and community health via changes to the supermarket environment. The objective of this review is to provide researchers and policymakers with information on existing policy options, their relative effectiveness in improving dietary behaviors, co-benefits or unintended consequences (e.g., impacts on retailer revenues or community economic development), and areas in need of further research. Although previous reviews have examined the effects of some of these policies, in isolation, on individual dietary or health outcomes, this will be the first, to our knowledge, to compare a wide range of outcomes across multiple policy approaches.

#### **2. Methods**

This review sought to answer two research questions: (1) which governmental policies in the U.S. aim to promote healthy choices in supermarkets; and (2) what is known about the effects of these policies on retailers, consumers, and communities? To answer these questions, we conducted searches of the peer-reviewed, academic, English-language literature published until January 2020. We searched PubMed, Econlit, PsycINFO, Web of Science, and Business Source Ultimate to identify papers spanning the economics, public health, marketing, consumer behavior, and business literature. Methods used to select and analyze results were consistent with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Figure 1) [13].

**Figure 1.** Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Flow Diagram.

#### *2.1. Policy Identification*

To answer the first research question, we reviewed five nutrition policy databases: (1) World Cancer Research Fund International's NOURISHING Database [14]; (2) The World Health Organization Global Database on the Implementation of Nutrition Action [15]; (3) Healthy Food Access Portal [16]; (4) Growing Food Connections Local Government Policy Database [17]; and (5) Healthy Food Policy Project [18]. From each database, two authors (Y.G. and S.C.) recorded the name, brief summary, date of enactment, and locale of government regulatory or legislative actions, meeting the following inclusion criteria: (1) implemented in a supermarket setting; (2) enacted in the U.S. as of 26 September 2019; and (3) intended to promote healthy food purchases (with policy intent inferred by the authors based on subject matter expertise). Policies that had been proposed but not enacted; policies that may influence food or beverage choices without changes to retailer practices (e.g., food formulation or front-of-package labeling policies); policies affecting only small stores or other non-supermarket settings (e.g., a healthy staple food ordinance); or policies that may affect food and beverage purchases, but were not designed with such intent (e.g., small sales taxes on soda or federal mandates requiring states to issue Supplemental Nutrition Assistance Program [SNAP] benefits no more than once monthly) were not included.

#### *2.2. Search Strategy and Inclusion Criteria*

Policies identified in this initial search were used to define our search terms, which were developed in collaboration with an informationist with expertise in public health policy (Appendix A). Papers were included if they met all of the following criteria: (1) peer-reviewed, original research; (2) conducted in the U.S.; (3) written in English; (4) evaluated a governmental policy, as defined above; and (5) assessed outcomes related to retailers (e.g., supermarket or manufacturer sales, revenue, or employment), consumers (e.g., dietary intake, food purchases, food security, body mass index), or communities (e.g., healthcare costs). In addition to quantitative evaluations, implementation, mixed methods, and qualitative research were included. Studies were excluded if they: (1) only assessed outcomes in small stores or other non-supermarket settings (e.g., availability of healthful foods in convenience stores); (2) described policy development but did not evaluate policy effects; or (3) described public comments or public opinion prior to policy enactment. After each search, duplicates were removed and titles, abstracts, and full texts were independently screened for inclusion by two authors (Y.G. and S.C.) using Covidence, a software for evidence synthesis (Covidence, Melbourne, Australia). Backward reference searching of included articles and reviews was conducted to identify additional papers. A third author (A.J.M.) was available to resolve disagreements (Figure 1).

#### *2.3. Evidence Synthesis*

Research was catalogued in alignment with the NOURISHING framework, which classifies policies into specific actions (e.g., posting calories on menu boards) within ten broad approaches (e.g., nutrition labeling). Three authors (Y.G., S.C., and A.J.M.) read each article and abstracted data on study setting (U.S. census region and urbanicity); design (controlled experimental, controlled quasi-experimental, descriptive (quantitative, including uncontrolled interventions, microsimulations, and modeling studies); descriptive (qualitative), or mixed/multiple methods); and population (adults, children, households, or other (e.g., stores, prices, benefit redemptions)). For each study, one study design was selected, but multiple settings and populations could be selected. Given the large and varied amount of research reviewed, additional quantitative or meta-analysis was not possible. Instead, a summary of findings, including outcomes, approaches, and research gaps, was generated using thematic analysis and narrative synthesis. Results are presented by policy action and approach.

#### **3. Results**

We identified 147 peer-reviewed academic research studies for inclusion in this review. These studies evaluated seven policy actions within three policy approaches (Table 1). The majority of studies used a descriptive, quantitative (52%) or controlled, quasi-experimental design (37%); were conducted among adults (48%); and used national data (44%) or were set in the Northeast (30%) (Table 2). Few studies used experimental (7%), qualitative (4%), or mixed/multiple methods designs (3%); were conducted among children (24%); or were set in the south (6%), Midwest (7%), or rural areas (6%).


**Table 1.** Description of policy actions included in the review.


**Table1.***Cont.*

Department of Agriculture).


**2.** Study design features of included articles, by policy action

 area.

**Table** 

*IJERPH* **2020**, *17*, 7493

#### *3.1. Nutrition Labeling*

Require Calorie Labeling of Prepared Food in Supermarkets

Few (*n* = 3) studies examined outcomes related to calorie labeling of prepared foods in supermarkets [35–37] and only one estimated the effects of calorie labels on food choices in a real-world supermarket setting [35]. Bachman et al. studied 393 women before and after calorie labeling in nine locations of a regional supermarket using a quasi-experimental design. Only 16% of study participants exposed to calorie labeling reported noticing the labels, and calorie labels did not influence food choices, although the sample size was small. In two studies, people trying or wanting to lose weight were more likely to rate calorie labels as important than people who were satisfied with their current weight [35,37]. Both studies used self-reported measures of consumer perceptions to assess the impact of calorie labeling; no studies have measured outcomes using validated dietary assessment surveys or objective food purchase data.

#### *3.2. Economic Tools to Address Food A*ff*ordability and Purchase Incentives*

#### 3.2.1. Increase Supplemental Nutrition Assistance Program (SNAP) Benefits

Thirteen studies assessed the effect of increased SNAP benefits (see Table 1 for description of policy) on household expenditures [38–41], food security [24,40,42,43], dietary behaviors [24,40,43–47], obesity [25], and healthcare utilization [48], with the majority using experimental (*n* = 2) or quasi-experimental (*n* = 8) designs. Studies indicate that increasing SNAP during the American Recovery and Reinvestment Act (ARRA) increased food-at-home expenditures but not food-away-from home expenditures [39,41,47], increased the share of benefits spent at superstores versus small stores [38], and increased spending on other necessary goods and services, including housing (mortgage, rent, utilities), transportation, and educational tuition [41]. Studies consistently demonstrated improvements in food security resulting from the ARRA benefit increase and Summer Electronic Benefits Transfer (EBT), as well as decreases in food security when the ARRA benefit increase ended [24,40,42,43]. Two studies found that benefit increases resulting from ARRA significantly reduced, but did not eliminate, declines in energy intake at the end of the benefit month [44,45], and one study found that ARRA was associated with a 65% reduction in outstanding medication needs due to cost among SNAP-eligible children [48].

The evidence for improving dietary behaviors and obesity is mixed. Most studies have found null or limited effects of a SNAP benefit increase on adult dietary quality [43,44,46,47]. A microsimulation study that directly compared the effects of an increase in SNAP benefits with a targeted subsidy on fruits, vegetables, and milk found that for the cost, targeted subsidies were more than ten times as effective in reducing deficiencies of recommended food groups [47]. One study of Summer EBT observed a small increase in children's fruit and vegetable, whole grain, and dairy intake, but no change in consumption of unhealthful foods and beverages [24,40]. One study showed a reduction in BMI among adults; however, that study assessed the impact of an indirect increase in SNAP benefits resulting from children's enrollment in school (and thus, participation in school meal programs), and may have been confounded by other changes affecting weight that correspond with school enrollment, such as changes in childcare expenses [25].

#### 3.2.2. Provide Financial Incentives for Fruits and Vegetables to Low-Income Households

Nineteen studies examined the impact of supermarket fruit and vegetable subsidies, incentives, vouchers, or prescriptions targeted towards low-income households or individuals [47,49–66]. Results from randomized trials and natural experiments consistently demonstrate increases in household fruit and vegetable purchases or adult fruit and vegetable intake when incentives are targeted towards SNAP participants [50–58]; yet, few studies have been conducted with children [52]. Studies assessing substitution found little evidence that fruit and vegetable incentives changed unhealthful food intake or expenditures [51–53,59]. Intervention effects, however, may not be sustained after

the financial incentive ends [54,57]. Although incentive programs in supermarkets would have high start-up costs if implemented nationally, they are expected to be cost-saving in the long-term, largely due to reductions in type 2 diabetes, heart disease, and stroke [60–63]. Compared to research on SNAP incentives, there are limited data on produce prescription programs, which, to date, have been most frequently implemented in farmers' markets or other limited-service food retail settings [49].

When the design and delivery of incentive programs are considered, the impacts on purchasing and consumption may increase with the size of the incentive. For example, the Healthy Incentives Pilot, which provided a 30% incentive on fruits and vegetables, saw a 26% increase (equivalent to approximately <sup>1</sup> <sup>4</sup> serving per day) in consumption among adults participating in SNAP, while the Shop Five for ME study, which provided a 50% incentive, found a 54% increase in fruit and vegetable purchases among SNAP households [52,53]. Additionally, incentives delivered as same-day discounts versus future rebates, administered electronically versus as paper coupons or vouchers, and offered without a minimum purchase requirement may increase uptake by lower-income households [52,55–57]. Frequent engagement with participating households and store staff about how the incentives work, which items qualify, and where they can be used may also be important for increasing utilization [49,52,56,58]. Complementary interventions focused on changing policies or environments to reduce unhealthy food purchases appear more effective in improving total diet quality at the population level than nutrition or cooking education, which tend to have low participation [51,52,59,61,63].

#### 3.2.3. Tax Sweetened Beverages

Forty-eight studies evaluated the impact of sweetened beverage excise taxes on a variety of behavioral, economic, and health outcomes in the U.S [5,60,63–65,67–109]. Evidence from real-world natural experiments shows that these taxes increase retail prices, although pass-through (the proportion of the tax that is passed on to the consumer) varies by city, store type, and beverage type and size [70,71,73,77,98,100]. Excise taxes reduce sales of taxed beverages, but the magnitude of the reduction is highly variable across cities [68,98,100,101]. For example, a \$0.01 per ounce tax on calorically sweetened beverages in Berkeley, California was associated with a 9.6% decline in sugary drink sales volume in Berkeley supermarkets after one year [100]. In a natural experiment, a \$0.015 per ounce tax on calorically and non-calorically sweetened beverages in Philadelphia, Pennsylvania was associated with a 38% reduction in taxed beverage volume sales after accounting for people who avoided the tax by purchasing sweetened beverages outside city limits [98]. Differences across studies may be due to baseline purchasing habits or income of the population, size of the tax, types of beverages included in the tax, the proportion of the population able to easily avoid the tax (i.e., by shopping in a bordering city), or tax salience.

There are less consistent data on changes in beverage consumption, total diet, or health outcomes, particularly among children. Evidence on dietary intake is mixed, possibly due to measurement error in dietary assessment tools and inadequate sample sizes [69,78,90,100,108]. For example, one study found a statistically significant reduction in sales of sugary drinks and increase in sales of untaxed beverages, but no statistically significant change in adult beverage intake one year after the Berkeley tax [100]. Another study one year after a Philadelphia tax found no changes in children's beverage intake overall, but significant reductions in sugary drink intake among children who were high consumers prior to the tax [71]. There are limited quantitative data on the long-term (>1 year) effects of sweetened beverage taxes; one study of the earliest tax in Berkeley found an average reduction in sugary drink consumption (−0.55 times/day) and increased water consumption (+1.02 times/day) among adults 3 years after the tax [90]. Modeling studies with varying assumptions consistently show taxes improve long-term health outcomes related to obesity, cardiovascular disease, and diabetes among adults, and reduce childhood obesity [5,60,63–65,84,87,91,93,94,96,99,103,104]. Several of these studies suggest a tax on calories or sugar in beverages may better target health harms and encourage industry reformulation, but these taxation strategies have not yet been implemented in the U.S [85,86,106,109].

Economic research shows taxes are highly cost-saving from a public health and societal perspective, but may be costly to industry [104], with some studies documenting reduced supermarket combined sales [98], reduced sugar producer revenues [76,104], and increased cross-border shopping in cities with a tax (which may not be detrimental to supermarkets if shoppers visit the same chain in a different city) [69,70,98,100]. Although the food and beverage industries frequently voice concerns over job loss resulting from taxation, no job loss in these industries within the first year of a tax has been documented [89,97]. Strong and consistent evidence shows that beverage taxes raise revenue for city programs, such as parks and early childhood education, and these investments may affect social determinants of health [76,88]. For example, a simulation of Philadelphia's tax found that investments in quality pre-kindergarten would further reduce sugary beverage consumption among young children by 8% [88].

3.2.4. Revise Composition and Quantities of Foods Provided through the USDA Special Supplemental Nutrition Program for Women, Infants and Children (WIC)

Forty-four studies assessed the association between the 2009 WIC food package revisions and availability of foods and beverages in supermarkets; purchases, redemptions, or dietary intake among WIC participants; obesity in early childhood; perinatal and birth outcomes; or outcomes related to breastfeeding [110–153]. There is consistent evidence of an association between the WIC food package revisions and improvements in household food purchases and dietary intake among both adults and children [111,114–116,119,121,135–140,145,147–149,151]. Specifically, revisions to the food package are associated with improvements in total diet quality, increases in fruit, vegetables, whole grains, dietary fiber, and low-fat dairy, and reductions in full-fat dairy, saturated fat, and juice (with no evidence of complete substitution to other sugary drinks). The cash-value voucher, in particular, increased the perceived value of the program for many participants, although voucher redemption varied across communities and may be limited in some areas by poor access to fresh fruits and vegetables or negative store experiences [112,113,117,123,128,131,132,146]. Impacts of the revisions on breastfeeding are mixed, with some studies showing increases in breastfeeding initiation [129,153], others showing no effect [118], and none finding a relationship with breastfeeding at six months [129,153]. Recent research using interrupted time series or controlled quasi-experimental designs show improvements in maternal and child health outcomes resulting from the food package changes, including reductions in infant and young child obesity [125–127,130], improvements in infant birth weight outcomes (low birth weight, small for gestational age, and large for gestational age) [120], and reductions in maternal weight gain and preeclampsia [120].

With regard to the retail food environment, several studies have documented changes in WIC food availability, variety, quality, or pricing after implementation of the food package revisions and minimum stocking requirements [133,134,141–144,152]. While outcomes have generally been positive for small and medium-sized stores (i.e., greater availability and lower prices of healthful foods), results in supermarkets and mass merchandisers are mixed, likely due to the wide variety of food options offered in these stores at baseline. It is important to note that, in addition to the federal requirements, states have the authority to establish stronger stocking requirements for authorized retailers and there is substantial variation in regulatory guidance across states [142]. Variation in minimum stocking standards, as well as other flexibilities in how WIC programs are administered at the state and local levels, may partially explain observed differences in program impacts across localities; however, this has not been well studied [130].

#### *3.3. Incentives and Rules to Create a Healthy Retail Environment*

3.3.1. Provide Financial Assistance to Supermarkets to Locate in Underserved Areas

Twenty-two studies assessed the impacts of new supermarkets locating in underserved areas. Though many community residents support the introduction of a new supermarket [154], studies,

including many using controlled, quasi-experimental designs, show low adoption of the new supermarket [155–157] and little or no improvement in body mass indices [155,158–162], household food purchases [156,157,163,164], or dietary intake [155–160,165] attributable to the new store. Similarly, modeling studies show these interventions are less cost-effective for supporting the introduction of new stores and increasing shopping at supermarkets than policies that increase SNAP benefits or coverage [38,166]. However, distance to the store, health and economic characteristics of the community, and baseline shopping habits within the population, may be important effect modifiers [167,168]. Additionally, supermarkets may positively impact health independent of effects on diet. One longitudinal natural experiment showed no improvement in dietary intake, but reductions in SNAP participation, food insecurity, and diagnoses of high cholesterol and arthritis one year after the opening of a new store [169]; however, the mechanism through which these positive health effects occurred is unclear. Several studies have suggested that investments in healthy retail may positively impact health by improving economic opportunity, social cohesion, or safety, but these mechanisms have not been studied [168,170,171].

#### 3.3.2. Allow Payment for Online Grocery Purchases with SNAP

No research has studied the effects of online grocery shopping on the diets of SNAP participants, but several recent studies provide insight on the availability and uptake of online SNAP purchases [172–175]. Three studies have shown that, while online grocery can address transportation barriers and food availability, perceptions related to higher food costs online, lack of control over food quality, and distrust of the online shopping process may prevent SNAP participants from utilizing these services [173–175]. One attempted trial was unable to recruit enough SNAP participants to make online grocery purchases, mainly due to participants' perceived lack of control over the quality of food selected [174]. Additionally, a recent study found that online grocery delivery services disproportionately serve urban areas; services are rarely available in rural areas [172]. It will be important to continue to monitor equitable access to online grocery shopping and delivery over the course of the COVID-19 pandemic, particularly in communities and sub-populations at the highest risk of infection.

#### **4. Discussion**

This integrative review aimed to identify governmental policies enacted in the U.S. to promote healthy choices in supermarkets and to synthesize the academic literature on these policies' effects. We identified 147 papers in seven policy areas: calorie labeling, SNAP benefit increases, financial incentives to purchase fruit and vegetables, sweetened beverage taxes, revisions to the WIC food package, financial assistance for supermarkets to locate in underserved areas, and allowing online purchases with SNAP. The majority of identified papers were related to sweetened beverage taxes (33%), followed by revisions to the WIC food package (30%), and financial incentives for supermarkets (15%); few studies assessed calorie labeling of prepared foods in supermarkets (2%) or online SNAP (3%). Most studies leveraged natural experiments to evaluate policy effects, utilizing controlled, quasi-experimental study designs, microsimulation or agent-based modeling, longitudinal approaches, and interrupted time series methods; very few studies employed experimental, qualitative, or mixed methods approaches. With regard to population, many studies were conducted among adults, except in the case of WIC food package revisions, in which studies of young children were more common. Studies frequently used national data or data collected in Northeastern or Western U.S. cities.; far fewer studies were set in the Southern or Midwestern U.S. or in rural areas.

When effects were compared across policy action, we found consistent evidence, including from real-world randomized trials and natural experiments, of an association between economic tools to address food affordability and dietary behaviors. Specifically, sweetened beverage taxes were associated with increased prices and decreased purchases of taxed beverages, revisions to the WIC food package were associated with improvements in total diet quality and maternal/child health outcomes, and fruit and vegetable incentives increased purchases and consumption of discounted foods. In modeling studies, all three policies reduced the incidence of cardiometabolic diseases and were cost-effective in the long-term, but those restricting or discouraging consumption of unhealthful foods (i.e., taxes, WIC revisions) showed greater gains than those solely encouraging consumption of healthful foods (i.e., incentives). When incentives were paired with restrictions or taxes on unhealthful purchases, however, their combined effects on dietary behaviors were greater than those of any single policy action. This highlights the importance of multiple, synergistic policy interventions delivered together.

In contrast to the economic levers described above, financial assistance for supermarkets to open in underserved neighborhoods and increases in the SNAP benefit amount had little effect on diet, but reduced food insecurity. Food insecurity is associated with a wide range of negative outcomes, including increased risk of obesity and cardiometabolic diseases [176], poor mental health [177], and poor early childhood development [178]. Thus, these policy interventions could improve long-term health by reducing food insecurity, but these mechanisms have not yet been studied. Longitudinal natural experiments are needed to understand the role of supermarkets in neighborhood revitalization and the complex relationship between economic development strategies and improved health of neighborhood residents.

This review exposed several gaps in the literature that could be addressed in future research. First, research on calorie labeling of prepared foods in supermarkets and online SNAP is nascent and could be examined using natural experiments or interrupted time series designs. Second, very little research has been conducted in rural areas or in the Southern or Midwestern U.S.—regions with a disproportionately high prevalence of obesity and related health conditions [179]. Similarly, few studies assessed policy impacts on racial or socioeconomic disparities. While some policies may not substantially improve average dietary intake, they may contribute to improving equity. Third, although this review sought to include a wide range of outcomes, most studies evaluated policy effects on food security, household purchases, dietary intake, or obesity. Other important health-related outcomes, such as changes to social norms, parental feeding practices, and modeling of healthful behaviors, are needed and could be assessed using qualitative or mixed methods. Fourth, very few studies examined outcomes of importance to retailers, such as customer loyalty or sales revenue, which could help foster retail partnerships and industry buy-in. Fifth, policy implementation was rarely addressed. Implementation of federal policies often varies at the state level, and state and local policies with similar goals often differ in scope. This variation in implementation could explain variation in outcomes across settings, but it has not been quantitatively or qualitatively assessed. Similarly, process outcomes, such as policy adoption, acceptability, or fidelity, were infrequently measured and could help explain null effects. Lastly, many quasi-experimental studies were limited by small sample sizes and crude dietary assessment tools, which may have limited investigators' abilities to detect small policy effects. Investigators should carefully consider required sample sizes and appropriate dietary assessment methods to avoid false null findings, which can be detrimental to policy and advocacy efforts.

#### *Strengths and Limitations*

This review has several limitations and strengths that should be considered. In line with the project aims, a large amount of literature was reviewed and, thus, strength and quality of evidence was not quantitatively assessed but rather qualitatively synthesized. As a means to limit included studies to the highest quality papers, non-academic, non-peer reviewed sources were not included. This decision likely led to exclusion of some important evidence, such as reports commissioned by the USDA or other agencies or organizations. It also excludes industry reports, which may be more likely to assess outcomes relevant to retailers. We did not include food formulation, front-of-pack, or back-of-pack labeling policies in this review, although such policies could theoretically influence the types of products stocked or how they are priced or promoted within the store. Strengths of the study include extraction of relevant policies from five policy databases, comprehensive search strings and database searches across the psychology, economics, business, marketing, policy, and public

health literature, inclusion of a wide range of effects on individuals, retailers, and community health, and narrative comparison of effects across seven distinct policy actions.

#### **5. Conclusions**

Governmental policies, particularly sweetened beverage taxes, revisions to the WIC food package, and financial incentives for fruits and vegetables, are associated with improvements in dietary behaviors. Providing financial incentives to supermarkets to open in underserved areas and increases in SNAP benefits are not associated with changes in food purchasing or diet quality but may improve food security. More research is needed to understand the effects of calorie labeling in supermarkets and allowing online purchases with SNAP.

**Author Contributions:** Conceptualization, A.J.M., A.P.; Methodology, A.J.M., A.G., C.A.R.; Formal analysis, A.J.M., Y.G., S.C.; Investigation, S.C. and Y.G.; Data curation, S.C. and Y.G.; Writing—original draft preparation, A.J.M.; Writing—review and editing, A.J.M., Y.G., S.C., A.G., A.P., C.A.R.; Visualization, A.J.M.; Supervision, A.J.M., A.P., A.G., C.A.R.; Project administration, A.J.M.; Funding acquisition, A.P., A.J.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the Johns Hopkins Center for a Livable Future. Publication fees were supported by Healthy Eating Research, a national program of the Robert Wood Johnson Foundation.

**Acknowledgments:** A version of this work was presented at the January 2020 Healthy Retail Research Convening in Washington, D.C. The authors would like to thank the meeting attendees for their constructive feedback, and Donna Hesson for her assistance in developing the search strategy for this paper.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **Appendix A. Search Terms**

#### **Search #1**

*Terms related to supermarket* (grocer\* OR supermarket OR store OR Retail\* OR outlet OR e-commerce OR mercado) AND *Terms related to food and beverage*

food OR foods OR beverage\* OR fruit OR fruits OR vegetable\* OR snack\* OR drink OR drinks OR lunch OR dinner\* OR breakfast OR meat OR poultry OR beef OR chicken OR fish OR milk OR cheese OR yogurt OR juice OR soda OR grain OR grains OR meal OR bean OR beans OR nut OR nuts OR candy OR sweets OR cookies OR chips OR "ice cream" OR sugar OR salt OR sugar-sweetened OR "sugar sweetened" OR sweet OR nutrition\* OR calorie OR calories

#### AND

#### *Terms related to policy*

policy OR policies OR Law OR laws OR regulat\* OR ordinance\* OR statute\* OR tax OR taxes OR taxation OR incentive\* OR "healthy food financing" OR subsid\* OR "menu labeling" OR "calorie labeling" OR WIC OR "supplemental nutrition" OR "food stamps" OR access OR rule OR rules OR "retail expansion" OR "community development" OR "food trust" OR "food desert" OR loan OR loans OR "healthy food business" OR "healthy neighborhood" OR "federal nutrition program" OR "fresh food fund" OR "grocery financing" OR zoning OR "minimum stocking" OR "staple food" OR "excess food" OR "food waste" OR "grocery store development program" OR "closer to my grocer" OR license OR licensing OR permit OR permitting OR "Baton Rouge" OR frameworks OR "grocery access" OR "fresh food financing" OR "supermarket access" OR "healthy food center" OR "neighborhood development" OR "healthy families" OR "fresh food retailer" **Search #2**

#### *Terms related to beverage taxes*

("sweetened beverage\*" OR "sugary drink\*" OR "sugary beverage\*") AND (tax OR taxes OR taxation)

#### **Search #3**

*Terms related to SNAP benefit increases*

(ARRA OR (benefit AND increase)) AND "Supplemental Nutrition Assistance Program"

#### **Search #4**

*Terms related to WIC food package revisions*

((WIC AND (fruit\* OR vegetable\*) AND voucher)

#### OR

(((WIC OR "Special Supplemental Nutrition Program for Women, Infants, and Children") AND ("food package OR revisions))) AND ("2014/04"[Date–Publication]: "3000"[Date-Publication])) **Search #5**

*Terms related to the online purchasing pilot*

"online grocery" OR "online purchas\* pilot" OR (online AND shopping AND "Supplemental Nutrition Assistance Program")

#### **Search #6**

*Terms related to produce prescription programs*

((fruit OR vegetable\* OR produce) AND (prescription OR rx)) AND (supermarket OR grocer\*)

#### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

International Journal of *Environmental Research and Public Health*

## *Review* **Influence of Food and Beverage Companies on Retailer Marketing Strategies and Consumer Behavior**

**Amelie A. Hecht 1,\*, Crystal L. Perez <sup>1</sup> , Michele Polacsek <sup>2</sup> , Anne N. Thorndike <sup>3</sup> , Rebecca L. Franckle <sup>4</sup> and Alyssa J. Moran <sup>1</sup>**


Received: 11 September 2020; Accepted: 3 October 2020; Published: 10 October 2020

**Abstract:** The retail food environment plays an important role in shaping dietary habits that contribute to obesity and other chronic diseases. Food and beverage manufacturers use trade promotion—incentives paid to retailers—to influence how products are placed, priced, and promoted in stores. This review aims to: (1) catalogue trade promotion practices that manufacturers use to influence retailer marketing strategies, and (2) describe how these retailer marketing strategies affect consumer purchasing behavior and attitudes. Researchers searched five databases, Academic Search Ultimate, Business Source Ultimate, PsycINFO, PubMed, and Web of Science, to identify literature from industry and academic sources published in English through November 2019. Twenty articles describing manufacturer trade promotion practices were synthesized and provided insight into four types of trade promotion practices: category management, slotting allowances, price discounts, and cooperative advertising. Fifty-four articles describing the impact of retailer marketing on consumers were synthesized and graded for quality of evidence. While comparison across studies is challenging, findings suggest that retailer marketing strategies, such as price promotions and prominent placement, lead to increased sales. Results can guide efforts by policymakers, public health practitioners, and food retailers to design retail environments that improve healthy eating while maintaining retailer financial interests. Additional research should measure the impact of retailer marketing strategies on consumer diet quality and retailer outcomes (e.g., return-on-investment).

**Keywords:** trade promotion; price; promotion; placement; food and beverage; food retailer; grocery; consumer behavior; marketing; chronic disease; choice architecture

#### **1. Introduction**

The retail food environment plays a critical role in shaping dietary habits and is an important setting for interventions to improve diet quality and prevent diet-related chronic diseases, including diabetes, obesity, and cardiovascular disease [1]. Evidence suggests that marketing of unhealthy foods and beverages may be more common and effective at driving sales compared to marketing of healthy foods and beverages [2–9]. Low-income and racial and ethnic minority populations are disproportionately targeted by unhealthy food marketing, which may exacerbate disparities in diet quality and diet-related chronic disease [10]. For example, advertisements for low-cost, high-calorie, and low-nutrition foods and beverages appear more often in media watched by African Americans [11]; and retailers increase

marketing of sugar-sweetened beverages when Supplemental Nutrition Assistance Program (SNAP) benefits are issued each month [12].

Retail food stores, which include both online and brick-and-mortar retailers (see Appendix A for a list of retail formats), are the primary source of food for many populations in both developed and developing economies [13]. In the US, consumers acquire the majority of their calories from supermarkets and superstores [14]. Considering that consumers make an estimated three-quarters of their purchasing decisions while shopping [15], in-store marketing techniques may play an important role in shaping purchase attitudes and decisions [9,16].

Food and beverage manufacturers use trade promotion practices (TPP), or incentives to retailers, to shape in-store marketing [17]. This paper focuses on how TPP influence three out of the "4Ps" of marketing: price, place (both the channels through which products are sold and where products are placed in stores), and promotion (efforts to engage consumers and communicate product features, such as signs) [18]. The fourth "P" of marketing, "product," is less frequently shaped by TPP, but rather by manufacturers in-house, through efforts such as packaging and product formulation. Similarly, TPP more commonly shapes where items are placed in stores and on shelves (i.e., product placement) rather than the channels through which products are sold. Food and beverage manufacturers allocate about \$1 trillion annually to TPP—between 50 and 70% of their marketing budgets and nearly 20% of their total revenue [17,19].

There is growing interest among policymakers, researchers, advocates, and retailers in creating policies and corporate practices that promote healthy food retail. To inform efforts to improve the food retail environment, it is important to understand (1) the types of TPP currently used by food and beverage manufacturers to influence retailer marketing strategies, and (2) how retailer marketing strategies, in turn, affect consumers. The first part of this research question—which types of TPP are used to influence retailers—is understudied, particularly in the public health literature. A 2016 investigative report commissioned by the Center for Science in the Public Interest, which describes TPP but did not use a systematic approach to gather data or survey the literature, served as a launching point for this aim [17].

The second part of this research question—how retailer marketing strategies impact consumers—has been only partially explored in previous reviews. Specifically, three previous reviews have focused on price promotions' impact on consumers; all three concluded that price promotions were associated with consumer behavior [3,9,20]. In a 2012 integrative review, Glanz et al. synthesized literature on the impact of price, placement, and promotion on consumer behavior but limited their search to literature focused on brick-and-mortar grocery stores. They found that all three marketing strategies were associated with increased product liking and purchasing, with some variation in degree of impact by strategy [21]. This review serves as an update to and expansion of the Glanz et al. review, synthesizing literature since 2011 and including other nontraditional retail settings such as online retailers and convenience stores. This review focuses on identifying, where possible, whether and how outcomes differ when healthy versus unhealthy products are marketed. Findings from this study can inform efforts by advocates, policymakers, public health practitioners, and food retailers to design food retail environments that promote healthy eating while maintaining retailer financial interests. This study will also identify gaps in the literature and provide directions for future research.

#### **2. Methods**

Two research questions were identified: (1) how do food and beverage manufacturers use TPP to influence retailer marketing strategies; and (2) how do retailer marketing strategies impact consumer purchasing behavior and attitudes? Searches were conducted for peer-reviewed and grey literature (e.g., conference abstracts and proceedings, reports, dissertations) in English. To identify publications from diverse disciplines including public health, business, economics, marketing, and social sciences, the following databases were searched: Academic Search Ultimate, Business Source Ultimate, PsycINFO, PubMed, and Web of Science. Search terms for each research question were developed by the study authors in consultation with industry and academic experts and a research librarian (Appendix B). The selection and analysis of the results were carried out under the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [22].

#### *2.1. Research Question 1: Search Strategy and Inclusion Criteria*

To answer the first research question, a narrative review was conducted to identify and catalogue types of trade promotion practices used by food and beverage manufacturers to influence retailer marketing strategies (Figure 1). Articles published through November 2019 were included. Article titles and abstracts were independently screened by two authors (AH and CP) for inclusion. Full-text review was completed by the first author (AH). Any questions about study inclusion were resolved through discussion with the second author (CP).

**Figure 1.** Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Diagram for Research Question 1.

#### *2.2. Research Question 2: Search Strategy and Inclusion Criteria*

To answer the second research question, a systematic review was conducted to understand the impact of retailer marketing strategies on consumer behavior and attitudes (Figure 2). Inclusion criteria were that the article must (1) be published between January 2011 and November 2019 (to capture studies published since the Glanz et al. review) and (2) measure the impact of retailer marketing strategies influenced by TPP on consumer purchasing behavior or attitudes. Studies were excluded if they assessed (1) an investigator-driven healthy retail intervention (review by Karpyn et al., forthcoming); (2) retailer or manufacturer practices unrelated to TPP (e.g., product labeling); (3) restaurants, vending machines, cafeterias, or schools, or (4) were not original research (e.g., literature reviews). The excluded literature reviews were incorporated into the background and discussion section of this review. Two authors (AH and CP) independently reviewed titles, abstracts, and full texts for inclusion and met to reconcile differences. Reference lists of included articles were also scanned, and relevant articles included.

**Figure 2.** PRISMA Diagram for Research Question 2.

#### *2.3. Quality of Studies*

The quality of included studies for the second research question was assessed using the Newcastle–Ottawa quality assessment scale, adapted for cross-sectional studies [23] (Appendix C). The Newcastle–Ottawa scale assigns studies composite quality scores by awarding a certain number of stars out of a total of nine possible stars. Similar to an approach used by Bennett et al., amendments were made to the scale; for articles using aggregate sales data, a "not applicable" option was allowed for categories of "non-respondents" and "controlling for confounding variables" [3]. The denominator (total number of possible stars) was reduced appropriately. Two authors (AH and CP) independently graded the included studies and met to reconcile differences. As described by Takehashi and Hashizume [24], studies that earned fewer than a third of the possible stars were classified as low-quality studies.

#### **3. Results**

#### *3.1. Narrative Review of Trade Promotion Practices*

Twenty articles were identified that described TPP used by manufacturers to influence retailer marketing strategies [25–44]. Of these, 13 articles were published in the peer-reviewed literature or through conference proceedings and seven articles were published in trade publications. Of peer-reviewed publications, two were in public health or public policy journals and the remainder were in journals focused on retail, economics, or marketing. Thirteen articles focused on the US, six focused on other countries, including Brazil, the United Kingdom (UK), New Zealand, Sweden, Finland, Italy, and Portugal, and one used a global perspective.

Results indicate that manufacturers use four types of TPP to shape retailer marketing strategies: (1) category management, (2) slotting allowances, (3) price discounts, and (4) cooperative advertising (Table 1). These terms may differ across retailers, manufacturers, and countries; for example, in Europe, slotting allowances are also referred to as listing charges [25]. Certain types of TPP may be used more often for some product categories and in some retail formats. For example, slotting allowances are more often used in highly concentrated, processed product categories such as beverages, snacks, and candy [32]. In smaller stores, such as convenience and corner stores, more informal incentive-based agreements between suppliers and retailers are common [36].


**Table 1.** Definitions of trade promotion practices.

Note: some articles discussed multiple trade promotion practices, so ns sum to greater than the total number of included articles.

#### 3.1.1. Category Management

Eleven articles focused on category management [27,28,31,34,35,37–39,41,43,44]. Category management is the collaboration between manufacturers and retailers to make decisions regarding product assortment, space allocation, pricing, and in-store promotion for entire product categories. Categories (e.g., ice cream, yogurt) are treated as strategic business units to ensure maximum efficiency and boost sales for the whole category, rather than for individual brands [27]. Category management typically uses a shopper-centric and research-based approach to promote consumer satisfaction and loyalty [39,41,44].

A leading manufacturer in a category often serves as the "category captain," overseeing category management and customizing plans on a store-by-store basis. Such an arrangement is often considered beneficial for both retailers and manufacturers: it allows retailers to concentrate on other aspects of their business, and manufacturers to focus on increasing category market share and profitability [28]. While some retailers have safeguards in place to ensure category captains are not unfairly advantaged, critics contend that because category captains have influence over which brands and products within a category are stocked and promoted, category captains may be able to exclude competitors [28,43].

#### 3.1.2. Slotting Allowances

Seven articles focused on slotting allowances, or lump-sum fees paid by manufacturers to retailers in exchange for access to the consumer market (i.e., shelf space) [25,29,32,33,36,40,42]. These include slotting fees to introduce a new product onto shelves, pay-to-stay fees to maintain shelf position, floor fees to make sales presentations and offer in-store samples, and display fees, which may cover premium placement, display materials (e.g., wire racks, prefabricated displays), and promotional signage. Theoretical explanations for why slotting allowances have become widely used include a market power explanation (i.e., slotting allowances reflect growing power among retailers who control access to the market) and an efficient market explanation (i.e., slotting allowances enable efficient allocation of scarce shelf space) [25]. According to the efficient market rationale, slotting allowances help retailers defray the costs and risks associated with new product introductions in light of an estimated 70% failure rate for new products [45]. Evidence suggests that slotting allowances in the US alone total between \$6 billion and \$18 billion per year [25,46]. Nationwide introduction of a new product in the US can cost up to \$1–2 million in slotting fees [45]. In countries with more independent retailers, slotting allowances are less common.

#### 3.1.3. Price Discounts

Four articles discussed price discounts that manufacturers provide to retailers to incentivize retailers to stock, display, or provide promotional discounts for their products [26,33,36,47]. Manufacturer discounts may be fixed or performance-based [47]. Fixed discounts are price reductions offered to the retailer on a per unit or per case basis, often at the time of billing, for a limited period of time. Performance-based discounts are tied to a measure of retailer performance such as number of units sold, displayed, or offered on price promotion. Discounts may be passed on to the consumer in the form of temporary price reductions (TPR) or coupons, affecting final sale prices [33,47]. Manufacturers may also provide retailers products for free to encourage retailers to stock new products or provide customer discounts, giveaways, or in-store samples [26,36].

#### 3.1.4. Cooperative Advertising

One article focused on cooperative advertising. Cooperative advertising is the collaboration between manufacturers and retailers to create and distribute local promotional materials such as newspaper inserts or direct mail flyers [42]. A cooperative advertising agreement may be initiated by either a retailer or manufacturer. Typically, the manufacturer will design the promotional materials, providing product images and templates, and the manufacturer and retailer will share the cost of printing and distribution.

#### *3.2. Literature Review of Impacts of Retailer Marketing Strategies on Consumers*

Fifty-four articles that describe the impact of retailer marketing strategies on consumer behavior or attitudes were identified (Table 2). These included peer-reviewed literature (*n* = 44), dissertations (*n* = 4), conference proceedings (*n* = 3), reports from government or industry (*n* = 2), and trade publications (*n* = 1). Studies occurred in the US (*n* = 17), UK (*n* = 11), other European countries (*n* = 8), Asian and Middle Eastern countries (*n* = 8), Australia or New Zealand (*n* = 6), Canada (*n* = 1), and Egypt (*n* = 1); two articles did not specify location. Articles focused on a range of retail formats, including supermarket/grocery stores (*n* = 43); convenience/corner stores (*n* = 9); online retailers (*n* = 4); dollar stores (*n* = 1); other (e.g., organic markets, liquor stores, pharmacies, *n* = 9); and four articles did not specify the retail format assessed. Ten articles evaluated multiple retail formats. Data sources used varied widely; scanner or panel data was the most commonly used data source (e.g., Kantar Worldpanel data) (*n* = 26), followed by customer survey (*n* = 21), direct observation (*n* = 9), customer interviews or focus groups (*n* = 8), marketing data from the manufacturer or retailer (*n* = 5), retailer loyalty card data (*n* = 4), and other data sources (e.g., customer diaries, eye scanner, store audits, bag checks, *n* = 6); one article did not specify the data source used. Nearly one third (*n* = 17) used multiple data sources. No articles declared conflicts of interest.

TPP influence three categories of retailer marketing strategies: how products are priced, placed, and promoted. Results below are organized according to these three domains. Notably, comparison across studies is challenging given they focus on different products, use different study designs, and employ different outcome measures. The two final sections of the results describe findings from studies that compare outcomes across two or more retailer marketing strategies and compare the impact of marketing of healthy versus unhealthy products.

#### 3.2.1. Pricing

Retailers employ a variety of price promotion strategies, including coupons, bundle deals (e.g., buy-one-get-one, 2-for-1), and TPR (also called rollbacks). In the US and the UK, an estimated 40% and 34% of all purchases are price promoted, respectively [6,48]. Estimates indicate that between 24% [49] and 67% [4] of unhealthy foods and beverages are purchased while price promoted, though prevalence of promotions differ across retailer formats and neighborhood [4]. A review of price promotions among Scottish retailers found that TPR are the most prevalent form of price promotion, accounting for 74% of promotions, followed by bundle deals, which represent 23% of promotions [9]. Price promotions are offered more frequently for unhealthy compared to healthy products [2,3,5,7–9,20,50,51].

Thirty-two articles focused on price promotions. Eight presented results separately for coupons, seven presented results separately for TPR, and the remainder did not specify the type of price promotion assessed or assessed multiple types of price promotion and did not present results separately.

#### Coupons

Coupons may be distributed by retailers or manufacturers. In 2017, 302 billion coupons for consumer packaged goods were distributed in the US [52]. Six studies evaluated coupons and reported coupons were associated with increases in overall purchase volume, impulse purchase volume, brand choice, and product trialing (first-time purchase), but not brand loyalty [53–58]. Two studies assessed customized coupons, which target consumer groups based on demographic characteristics or past shopping behavior, and found they were associated with increased purchasing of targeted products [54,56]. Coupons in some product categories may be more impactful than others: one study found that coupons led to greater product trialing when promoting leading brands and categories that were popular, easy to store, had fewer products in the category (easier for customers to process less options), and were frequently on sale [55]. Another study found that while customized coupons led to increased purchases for both healthy and unhealthy products, they were more effective for unhealthy products [54].

#### Temporary Price Reductions

All eight studies that evaluated TPR detected associations with one or more consumer shopping behaviors, including purchase volume, impulse purchase volume, brand choice, and brand market share [58–65]. TPR may have a stronger impact on some outcomes compared to others: one study that assessed wine purchases in the UK found that TPR strongly influenced brand selection, somewhat influenced purchase volume, but did not influence purchase initiation [61].

Three articles assessed the impact of TPR in online retail [63–65]. Two out of three studies found that online price promotions were associated with increased purchases [63,64]; the third found no association [65]. One of the two studies that detected an association reported that because online purchases were delivered, barriers to stockpiling were eliminated, resulting in increased purchase volume compared to in traditional brick-and-mortar retail outlets [64]. The other reported that when a retailer with both online and brick-and-mortar retail outlets offered price promotions online, online sales increased, but sales in the brick-and-mortar stores decreased [63]. That study also found that high frequency of online promotions led to diminished effects over time [63].

#### Other Price Promotions

Thirteen articles on price promotions did not specify the type of price promotion studied or examined several types of price promotions together [6,8,50,66–75]. Many studies using panel data were unable to distinguish between types of price promotion used by customers. All studies identified positive associations between price promotions and one or more outcomes, including purchase volume, stockpiling purchase volume, purchase initiation, product trialing, and store choice. Within some studies, however, price promotions were positively associated with some outcomes and not others. For example, one study assessing Japanese market trends over time found that manufacturer expenditure on sales promotion was associated with an increase in total purchase volume but a decrease in manufacturer profits [66]. Another study found that price promotions led to short-term sales increases, but in more than half of cases, did not increase category revenue due to brand-switching (substitution) effects within the category [70].

Quantitative estimates on the impact of price promotions are difficult to compare because researchers used different outcome measures. Three studies, all using data from the Kantar Worldpanel, illustrate this challenge [6,8,69]. Nakamura et al. estimated that a 1% increase in price discount led to a

sales uplift of 1.44% within a given category [6]. Smithson et al. found that approximately one-fifth of foods and beverages bought on price promotion were purchased in addition to what would be expected absent a price promotion, leading to an overall increase in food and drink purchase volume [8]. Revoredo-Giha et al. found that the presence of a price promotion increased spending between 2% and 10%, depending on the product category [69].

The effect of price promotions may differ across product categories and consumer characteristics. For example, one study found that while, price promotions did not, on average, affect beef sales, they did influence sales for certain cuts of meat and consumer groups (e.g., young families versus older adults) [68]. Another study found that price promotions were associated with increased soda sales across all levels of consumer education and retail formats, but the effect was weaker in neighborhoods with a higher proportion of residents with at least a post-secondary certificate or diploma [71].

Three studies compared differences in the impact of price promotion on healthy and unhealthy products [67,69,74]. Two of these studies found that purchase volume increased as price decreased for unhealthy foods but not for healthy foods [67,74]. Another, however, found that price promotions led to increases in total spending and spending by category for both healthy and unhealthy foods, though the effect was greater for less healthy foods [69]. Specifically, they found greater increases in spending for unhealthy categories such as confectionery (10%) and beverages (9%) and smaller increases for healthier categories such as fruits and vegetables (5%), grains (3%), and dairy (2%).

#### Perceived Importance of Price Promotions

Eight articles assessed consumer perceptions regarding the importance of price promotions in shaping their purchasing decisions [67,73,76–82]. Though the populations and contexts assessed varied across articles, all studies found that shoppers considered price promotions to be an important factor influencing their shopping behavior. Three of these studies assessed perceived importance of price promotions within specific cultural and religious contexts. In one study, Egyptian Muslim shoppers reported that TPR and bundled deals led them to engage in more stockpiling and spending, but other discount promotions considered not compliant with Shari'ah law, such as sweepstake draws and scratch-and-win promotions, did not shape their behavior [79]. In a study of Pakistani Muslim shoppers, participants reported that their intentions to purchase Halal products were shaped by price promotions [80]. Through interviews with "ethnic" shoppers in the UK, a final study found that participants reported diverse responses to price promotions, ranging from responsive to hostile, depending on the perceived "net worth" of the promotion [81].

#### 3.2.2. Placement

Sixteen studies focused on how products were placed within stores, measuring visual attention, purchase volume, or spending as the primary outcomes [15,46,58,62,83–94]. Through slotting allowances and category management, manufacturers are able to secure placement in premium store locations, including on the endcap (i.e., end-of-aisle displays free from direct aisle-based competition), in the checkout aisle, and on freestanding displays. In 2012, an estimated 60% of products in stores were cross-promoted, meaning they are were displayed in secondary locations away from their "home" aisle [15]. Displays may be located anywhere in the store: approximately 42% of displays are located on the endcap, 28% in the aisle, 23% on the perimeter of the store, and 7% at the front of the store [15]. In an evaluation measuring shoppers' visual attention, 13% of all eye-fixations were drawn to in-store displays; of these, 44% were to endcaps, 34% to floor stands, 12% to in-line displays (i.e., gondola, or freestanding wire or metal shelving), and 10% to power wings (i.e., sidekick displays, or cardboard displays that attach to shelving) [15].

#### Endcaps

Five studies focused on placement in endcaps; all found significant positive effects on purchasing [46,83,85–87]. In a study of UK stores, endcap displays led to increased purchase volume for beer by 23.2%, for wine by 33.6%, for spirits by 46.1%, and for carbonated drinks by 51.7%; sales uplift was even greater for tea and coffee [85]. Two studies found that endcaps located at the rear of the store are more impactful than those at the front of the store [47,91]. In an experimental study in Australian grocery stores, placement of unhealthy products on rear endcap displays generated a 416% uplift in sales, while placement on front endcap displays generated a 346% uplift in sales [46]. Findings also suggest that endcaps are most impactful when located away from in-store sampling [87] and in stores without middle, perpendicular aisles [86].

#### Shelf Placement and Space

Category management and display fees can also affect where categories are placed within a store, and where individual products are placed on shelves (e.g., at eye-level for adult shoppers). Three experimental studies suggest that placement at the front of the store, in central aisles, at eye-level, and away from other popular categories can have positive effects on sales [88,91,92]. In one study, moving fruits and vegetables to a prominent location at the front of a grocery store led to an increase in sales volume and spending on fruits and vegetables [92]. In another, placement of dairy products in a central aisle was associated with increased product sales and purchase incidence, while placement next to popular categories had an "attention stealing" affect, leading to decreased sales [88]. In a laboratory-based study of college students, junk food items placed at eye-level received more visual attention than those on higher or lower shelves [91]. However, an observational study in New York City bodegas found no association between unhealthy beverage purchases and the placement of healthy products in prominent locations (i.e., water at eye-level and produce in at the front of the store) [89].

Total amount of dedicated shelf and display space (measured in feet) was associated with increased sales in two studies [84,90]. In one study, Minneapolis stores with more shelf space dedicated to fruits and vegetables had healthier purchases (i.e., more fruits and vegetables, more whole grains, and higher healthy eating index scores) [84]. Similarly, in a study of Hispanic shoppers in San Diego tiendas, each additional square foot of display space for fruits and vegetables was associated with a \$0.02 increase in weekly amount spent on fruits and vegetables [90].

#### Other Placement Strategies

Four additional studies evaluated the impact of placement but did not specify how or where evaluated products were displayed [58,62,93,94]. All four studies found that presence of displays was positively associated with impulse purchase volume, spending, or brand choice. One of these studies was an industry report that assessed a multifaceted marketing campaign, however, and it is unclear what proportion of the sales uplift was attributed to placement [93].

#### 3.2.3. Promotion

Sixteen articles focused on promotion [53,73,76,80,82,84,87,89,90,95–101]. Manufacturers use cooperative advertising and display fees to secure promotional signage, in-store sampling (i.e., taste tests), loudspeaker announcements, games, and other giveaways.

#### Signs

All three studies that measured the relationship between signs on shelf facings (called shelf-talkers or aisle violators) and purchase behavior focused on promoting healthy products; none detected a significant association [84,89,90]. In tiendas in San Diego, the number of signs promoting fruits and vegetables was not associated with fruit and vegetable purchases among Hispanic consumers [90]. In Minnesota stores, healthy advertising inside stores was not associated with purchasing, and, in fact, healthy advertising outside stores was associated with less healthy purchases [84]. In New York City bodegas, neither signs advertising water nor signs advertising sugar-sweetened beverages were associated with sugar-sweetened beverage purchases [89].

One study assessed "feature advertising" in two competing grocery stores, but did not describe components of "feature advertising" [73]. This study found that feature advertising led customers to choose to shop at the store with featured advertising over another store.

#### In-Store Sampling

In-store sampling was found to be associated with greater brand loyalty and purchase volume in three studies [53,87,99]. Several factors may moderate the impact of in-store sampling on purchases: studies suggest that benefits are maximized when the product being offered on sample matches the product displayed on the closest endcap [87,99]. One study also found a sales increase when in-store samples were offered close to the weekend compared to earlier in the week, when store personnel were present to offer the sample (24.3% increase compared to without store personnel present), when there was a sign promoting the product (90.8% increase compared to no sign), and when a commercial for the product is played on an in-store TV (36.3% increase compared to no commercial) [99].

#### Games, Giveaways and Limited-Time Offers

Findings on the impact of games, giveaways, and limited-time offers differed across studies [53,82,100]. In one study, customers reported that in-store games and lotteries led to greater customer loyalty and stronger relationships with promoted brands [53]. In another study, giveaways of collectible items increased the probability of brand choice and category purchase incidence, particularly when paired with a price discount, but did not change the purchase volume decision [100]. In a final study, both limited-time and membership deals were found to increase purchase incidence in an organic market [82].

#### Perceived Importance of Promotions

Seven articles assessed consumer perceptions regarding the importance of promotional activities in shaping their purchasing decisions [76,80,95–98,101]. Studies investigated different types of promotions and used different methods to assess customer perceptions, and found varying levels of perceived importance. Five studies found that consumers reported high levels of perceived importance of marketing on their attitudes toward purchasing [80,96–98,101]. Two studies, however, found promotional offers to be less persuasive: in a survey of Australian shoppers, 41% said they were influenced by promotional offers, but, in focus groups and interviews, many said that while promotional offers engaged them initially, trust and emotional connection to the brand was the primary driver of their purchase decisions [95]. In a survey of Vietnamese urban shoppers, participants described merchandise display and promotion as the least important factor from a list of seven factors influencing impulse purchase behavior [76].

#### 3.2.4. Comparison of Marketing Healthy versus Unhealthy Products

As previously described, a small number of studies compared marketing of healthy versus unhealthy products [54,67,69,74,84,89,90]. Of these, four focused on price, three on placement, and three on promotion. Half of price-focused studies found that price promotions led to increased purchasing of unhealthy but not healthy products, [67,74] whereas the other half of studies found that while the effect was stronger for unhealthy products, price promotions led to increased purchasing of both healthy and unhealthy products. One of the three studies focused on placement found no association between prominent placement of healthy products and purchasing [89]. The other two studies, however, found that stores with more shelf and display space dedicated to fruits and vegetables had healthier sales [84,90]. Notably, both of these studies were cross-sectional and thus were unable to determine causality. Finally, none of the three studies focused on signs promoting healthy products detected a relationship with purchasing [84,89,90].

#### 3.2.5. Comparison across Marketing Strategies

A small number of articles directly compared one retailer marketing strategy to another. Four of these asked participants to rank factors that shape their purchasing; in all four, participants reported that price promotions were the most or one of the most influential factors shaping their attitudes toward purchasing. Vietnamese shoppers reported that price promotions influenced their spontaneous purchase tendencies more than displays [76]. Taiwanese organic market shoppers reported that discounts and free giveaways impacted their shopping behavior more than membership or limited-time offers [82]. Two other studies focused on Muslim shoppers: in one [80], shoppers reported being equally influenced by Halal marketing promotions and pricing, while in the other [79], shoppers reported price discounts influenced their purchase intention more than giveaways, games, and in-store samples.

One study compared different types of price promotions, finding that sensitivity to coupons was greater than sensitivity to TPR [58]. The remaining studies quantitatively compared price promotions to either promotion or placement; results largely indicated that price promotions are more impactful than other types of marketing strategies [53,59,73,85]. Specifically, one study found that price was a stronger driver of stockpiling purchases than feature and display promotion [73]. Another found that a 20% TPR increased fair trade coffee sales more than providing information or a moral appeal [59]. Another study found that the effect size for endcap placement was equivalent to a price decrease for alcohol categories of between 4% and 9% per volume, and a price decrease for non-alcohol categories of between 22% and 62% per volume [85]. One study, however, found that price promotion and in-store sampling produced different benefits: in-store sampling helped nurture consumer loyalty more than coupons, but coupons resulted in more purchases [53].

#### 3.2.6. Quality of Evidence Grading

On average, included studies received 65% of total possible stars (Appendix D). Only three of the 54 studies included in this review were of low-quality, having earned less than a third of all possible stars. The two categories in which studies most often earned zero stars were sample size (*n* = 27) and non-respondents (*n* = 28). Nearly half of the included articles omitted sample size calculations or justification; this was particularly common among studies using questionnaires or published in non-peer-reviewed sources. Only one study compared respondents and non-respondents or reported their response rate, though for 25 articles, this information was considered not applicable due to use of panel data. More than half of all studies earned the maximum number of stars in the assessment of outcome category by linking records or using an independent blind assessment to determine the outcome.


**Table 2.** Study design, marketing strategy, retailer format, country, study duration, data source, objectives, outcomes, and key findings for studies

included

in





















#### **4. Discussion**

This review is the first to synthesize literature from academic and industry sources on the approaches that manufacturers use to shape retailer marketing strategies, and, in turn, consumer behavior and attitudes. More than half of the included studies focused on pricing; fewer articles assessed placement or promotion and many of these articles focused on purchase attitudes rather than behavior.

Findings suggest that all types of price promotions, including coupons, multi-buys, and TPR, shape purchasing behavior. Placement in premium store locations, such as on endcaps, and in-store samples are also effective drivers of sales. Other promotion activities, such as giveaways, games, and signs, may be less impactful. Notably, findings suggest that retailer marketing strategies may be less effective at driving sales for healthy foods and beverages [54,67,69,74,84,89,90]. Of the small number of studies that specifically considered sales of healthy products, the majority found that retailer marketing strategies, including signs and price promotions, were not associated with increased sales of healthy products. Two studies did find increases in healthy purchases, but effect sizes were smaller than for unhealthy products [54,69]. Previous reviews have similarly found that promotions of unhealthy products are more impactful than those for healthy products [4,9]. Studies of retailer- or investigator-driven interventions to specifically promote healthy purchases, however, were outside the scope of this study (these interventions are reviewed by Karpyn et al. in a paper published as part of this special issue) and may have identified retailer marketing strategies that effectively increase healthy purchases.

Findings regarding the impact of price promotions and product placement on consumer behavior are consistent with findings of previous reviews [3,9,20,21]. To the authors' knowledge, the review by Glanz et al. is the only study to also explore promotion; however, they did not identify any studies related to signage, and only one related to in-store sampling [21]. Several previous reviews, including Glanz et al., excluded studies in nontraditional retail settings, such as convenience and dollar stores and online retail. Findings from included studies of nontraditional retail formats suggest that retailer marketing strategies have similar effects across retail settings. One notable difference, however, is that consumers may be more likely take advantage of promotions in online retail by stockpiling, as they are not required to transport their purchases home themselves [64].

While only three studies were rated as "low quality," analytic rigor and rigor of data sources in included articles varied widely. Many articles from industry publications did not describe their analytic methods; thus, it was challenging to assess the quality of evidence of these articles. Additionally, while no studies listed conflicts of interest and many did not disclose funding sources, several were written by industry representatives and published in trade publications and may, therefore, have been more inclined to include findings that portrayed the companies favorably. A growing number of studies used store scanner and loyalty card data; these data sources, which provide large sample sizes and detailed sales information, should be used widely, particularly when paired with information on customer demographics. Several of the included studies occurred in controlled, laboratory settings; strategies that proved impactful in these settings, such as placement of products at eye-level, should be adapted and tested in real-world retail environments.

Study findings point to strategies that policymakers, public health practitioners, and retailers can use to ensure that retail environments promote healthy eating. Results suggest that policies and corporate practices that limit promotion of unhealthy products, rather than interventions to promote healthy products, may be needed to improve diet quality. Policies and practices can target each of the four TPPs identified in this study to curb promotion of unhealthy products. For example, policies could prohibit category captains from excluding competitors, or create healthy checkout aisles by prohibiting retailers from accepting stocking fees to display ultra-processed foods in checkout aisles. Considering SNAP is an important revenue stream for many US retailers, restrictions on promotion of unhealthy products could also be integrated into requirements for SNAP-authorized retailers [102].

#### *4.1. Future Research Directions*

Findings from this study highlight directions for future study. Research is needed to evaluate:


#### *4.2. Limitations*

It is possible that different search terms or databases might have identified further studies. The quality of evidence grading tool was adapted from the Newcastle–Ottawa quality assessment

scale, which was initially designed for case-control and cohort studies. Thus, it was challenging to assess quality of evidence for qualitative and observational studies, as well as industry reports and news articles, which provide few details on methods. Additionally, while study inclusion criteria were designed to capture studies from any country, only studies published in English were included. The majority of identified studies focused on the US and UK, and many countries were not represented in these findings. As a result, results may not be generalizable across countries and cultures.

#### **5. Conclusions**

This review finds evidence that by influencing retailer marketing strategies through TPP, manufacturers can shape consumer behavior and, ultimately, diets. The 74 studies included in this review suggest that TPP have a considerable effect on product placement, pricing, and promotion, and, in turn, on a range of customer outcomes, including purchase volume, spending, and attitudes. Findings point to a particularly strong relationship between price promotions and consumer behavior and differential impacts by product type and consumer characteristics. This review builds on previous work by synthesizing findings from recent studies and studies focused on non-traditional retail formats. Study findings provides valuable insight that can guide efforts by policymakers, public health practitioners, and food retailers to design retail environments that promote healthy eating. Public health practitioners and policymakers could consider policies that regulate promotion of unhealthy products by targeting each of the four TPPs identified in this study. Further investigation is warranted to determine the impact of retailer marketing on dietary outcomes and outcomes of importance to retailers. Further research is also needed in online and nontraditional retail settings.

**Author Contributions:** A.A.H. and C.L.P. were involved in conceptualizing this study, defining the methodology, conducting the literature search, analyzing findings, drafting and editing the manuscript, and funding acquisition. M.P., A.N.T., R.L.F. and A.J.M. were involved in conceptualizing the study, defining the methodology, and reviewing and editing the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by Healthy Eating Research, a national program of the Robert Wood Johnson Foundation. Training support for A.A.H. was provided through the Johns Hopkins Center for a Livable Future-Lerner Fellowship.

**Acknowledgments:** The authors would like to thank attendees of the January 2020 Healthy Retail Research Convening and the 2020 Healthy Eating Research Annual Meeting for their constructive feedback on research findings and conclusions. The authors would like to thank Megan Lott, Kirsten Arm and Mary Story from Healthy Eating Research for their project support.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders commissioned research on this topic but had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **Appendix A**

#### *Food and Beverage Retail Formats \*:*

• **Supermarkets:** A food retailer with greater than 9000 square feet of selling space and at least \$2 million in annual sales. Non-food sales must account for 15% or less of total store sales (e.g., Kroger, Safeway).

• **Drug stores:** Drug stores feature prescription-based pharmacies but generate at least 20% of their total sales from other categories, including general merchandise and food (e.g., Rite-Aid, CVS).

• **Mass merchandisers:** Large department stores that sell primarily general merchandise and nonperishables but also carry a limited assortment of grocery products (e.g., K-Mart, Target).

• **Supercenters:** Also known as hypermarkets and superstores, supercenters are hybrid stores that combine mass merchandisers with full supermarkets (e.g., Walmart Supercenter, Fred Meyer).

• **Convenience and corner stores:** Convenience stores typically sell gasoline, general merchandise, alcohol and tobacco, and a limited selection of staple and ready-to-eat, prepared foods (e.g., 7-Eleven, Exxon).

• **Dollar stores:** Dollar stores typically emphasize low prices (many products cost \$1) and offer little in the way of customer service. Traditionally, they focused on staple consumer goods and other nonfood items but, have increasingly offered food (e.g., Dollar Tree, Dollar General).

• **Club stores:** Also referred to as warehouse or volume stores, these are large-format outlets that specialize in selling food and selected general merchandise in bulk for relatively low prices, per unit. Consumers need paid memberships to shop at them (e.g., Costco, Sam's Club).

• **Online retailers:** Retailers with an ecommerce presence, providing grocery click-and-collect or delivery services. These retailers can take any brick-and-mortar format (i.e., convenience stores, supermarkets, and mass merchandisers can all be online retailers).

• **Other:** Includes military commissaries, hospitals, and other food service providers, as well as direct-to-consumer food outlets such as farmers markets and community-supported agriculture.

\* Definitions adapted from Volpe R, Kuhns A, and Jaenicke T. 2017. Store Formats and Patterns in Household Grocery Purchases, EIB-167, U.S. Department of Agriculture, Economic Research Service.

#### **Appendix B**

#### *Search Terms*

Research Question 1 Search String

(("Food" OR "Beverage")

AND ("Manufacturer \*" OR "Distributor \*" OR "Supplier \*" OR "Processor \*" OR "grower \*" OR "trade association \*" OR "producer \*" OR "industry" OR "company")

AND ("Category captain \*" OR "Category management" OR "Cooperative advertising" OR "Cooperative marketing" OR "dealer aid \*" OR "fee \*" OR "In-kind remuneration" OR "Pay-to-stay" OR "pouring right \*" OR "Promotional allowance \*" OR "Slotting" OR "Trade promotion" OR "Trade spend")

AND ("Grocer \*" OR "Grocery store \*" OR "Supermarket \*" OR "Food store \*" OR "Online grocer \*" OR "Superstore \*" OR "Warehouse \*" OR "Club store \*" OR "Convenience store \*" OR "Food retailer \*"

OR "Food outlet \*" OR "pharmac \*" OR "corner store \*" OR "discount store \*" OR "dollar store \*") NOT "pharmaceutical \*"

Research Question 2 Search String

(("Food" OR "Beverage")

AND ("Manufacturer \*" OR "Distributor \*" OR "Supplier \*" OR "Processor \*" OR "grower \*" OR "trade association \*" OR "producer \*" OR "industry" OR "company") OR ("Grocer \*" OR "Grocery store \*" OR "Supermarket \*" OR "Food store \*" OR "Online grocer \* " OR "Superstore \*" OR "Warehouse \*" OR "Club store \*" OR "Convenience store \*" OR "Food retailer \*" OR "Food outlet \*" OR "pharmac \*" OR "corner store \*" OR "discount store \*" OR "dollar store \*"))

AND ("Category captain \*" OR "Category management" OR "Cooperative advertising" OR "Cooperative marketing" OR "dealer aid \*" OR "fee \*" OR "In-kind remuneration" OR "Pay-to-stay" OR "pouring right \*" OR "Promotional allowance \*" OR "Slotting" OR "Trade promotion" OR "Trade spend" OR "Buy-one-get-one" OR "buy one get one" OR "Circular \*" OR "Coupon \*" OR "Cross Promotion \*" OR "Discount \*" OR "Display \*" OR "endcap" OR "Feature and display" OR "Incentive \*" OR "In-store sampl" OR "loss leader \*" OR "Placement" OR "POINT-of-sale advertis \*" OR "Price Promotion \*" OR "Rollback \*" OR "Sales Promotion \*" OR "Sign \*" OR "Shipper \*" OR "Shelf talker \*" OR "Temporary price reduction \*" OR "Two-for-one" OR "two for one")

AND ("Customer" OR "Shopper" OR "Consumer")

AND ("Behavior" OR "Preference" OR "Demand" OR "Consumption" OR "Choice" OR "decision" OR "purchas \*" OR "attitude \*" OR "willingness to pay")

NOT ("pharmaceutical \*")

#### **Appendix C**


<sup>1</sup> "Target population" defined based on authors' definition of their "target population." The Newcastle–Ottawa quality scale assigns studies composite quality scores by awarding up to nine stars. A study can be awarded a maximum of one star (\*) in the categories of: representativeness of the sample, sample size, non-respondents, and statistical test. A maximum of two stars (\*\*) can be awarded in the categories of: ascertainment of risk factor, comparability, and assessment of outcome.



**Table A2.** Quality assessment of the included studies using the Newcastle–Ottawa quality scale (*<sup>n</sup>*

= 54).

Mortimer and Weeks (2011)

 1

 1

 0

 2

 1

 1

 1

 7/10



#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*International Journal of Environmental Research and Public Health* Editorial Office E-mail: ijerph@mdpi.com www.mdpi.com/journal/ijerph

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18