Elsevier

Food Policy

Volume 43, December 2013, Pages 100-107
Food Policy

Effects of the Guiding Stars Program on purchases of ready-to-eat cereals with different nutritional attributes

https://doi.org/10.1016/j.foodpol.2013.08.013Get rights and content

Highlights

  • We study the effect of nutritional rating display on demand for ready-to-eat cereals.

  • We estimate Rotterdam demand model using store scanner data.

  • We use treatment-control empirical approach by matching control and treatment stores.

  • We find that sales of low-ranked cereals decreased, while high-ranked ones increased.

  • Nutritional ranking reduced substitutability of low-ranked and high-ranked cereals.

Abstract

Over the past decade, the food industry has increased its use of front-of-package and shelf-tag nutrition labeling designed to present key nutritional aspects and characteristics of food products. One such system is the Guiding Stars Program™ (GSP), which uses an algorithm to score the nutritional values of food products from one to three stars, where more stars mean more nutritious. We studied how the introduction of the GSP in one supermarket chain affected the demand for ready-to-eat cereals. We estimated the demand for cereals and measured the effect using a treatment–control approach. We found that the GSP significantly increased the demand for cereals that GSP considers more nutritious at the expense of cereals that GSP considers less nutritious.

Introduction

Poor diets are a significant cause of obesity, heart disease, stroke, cancer, diabetes, osteoarthritis, and other health conditions that impose an economic burden on individuals and society overall (USDHHS/USDA, 2011, USDHHS, 2010). Medical costs associated with obesity were estimated as high as $147 billion, or 10% of all medical costs, in 2008 (Finkelstein et al., 2009, O’Grady and Capretta, 2012, Tsai et al., 2011). Growing recognition of the obesity epidemic and the prevalence of diet-related chronic diseases have led to an array of efforts aimed at increasing physical activity and promoting nutritious eating, including changes in the formulation, packaging, labeling, and marketing of food and beverage products (IOM, 2010).

The 2009-2010 National Health and Nutrition Examination Survey found that 22.5% of Americans do not use or rarely use the standard Nutrition Facts label when they shop (CDC, 2013). Other consumers struggle to understand the Nutritional Facts label (Rothman et al., 2006). Over the past decade, the food industry has increased its use of front-of-package (FOP) and shelf-tag nutrition labeling designed to present or summarize key nutritional aspects and characteristics of food products. These initiatives include programs such as the Guiding Stars Program™, NuVal™, Traffic Light and Facts Up Front (see IOM (2010) for descriptions of various FOP nutrition labeling systems).

The Institute of Medicine (IOM) of the National Academies recently reviewed observational and experimental studies on the effectiveness of FOP nutrition labeling systems in influencing food purchases under applied and experimental settings (IOM, 2012).1 In evaluating the strengths and weak nesses of past studies, the IOM committee assigned the largest weight to peer-reviewed results from field or natural experiments. After reviewing four field experiments available in the literature, the IOM concluded that, as currently developed, the FOP systems alone do not show consistent evidence of dramatically influencing consumer choice. However, some evidence suggests that easy-to-understand FOP systems encourage consumers to choose more nutritious foods, especially in cases where consumers make quick purchase decisions.

One of the studies examined changes in sales for all products before and after the Guiding Stars Program™ (GSP) was implemented at the Hannaford supermarket chain in September 2006 (Sutherland et al., 2010). The GSP was designed to “help consumers readily choose foods” that are consistent with dietary recommendations such as those put forth in the Dietary Guidelines for Americans (Fischer et al., 2011). GSP gives foods one to three stars ratings based on the nutritional value of the food (three stars is the best); the products that are not qualified for a star rating remain unstarred. After adjusting for seasonality, Sutherland et al. (2010) found a slight increase in the sales of products labeled with stars after the introduction of the GSP. Sutherland et al. (2010), however, did not incorporate a control group. Thus, their findings may not represent sales effects attributable to the GSP at Hannaford and may instead capture the national trend in cereal sales. Furthermore, their model does not include sale promotions, changes in product variety, and increased awareness of importance in nutrition due to the release of Dietary Guidelines for Americans 2005 (USDHHS and USDA, 2005). Not accounting for these factors and ignoring the effect of price changes on food demand could bias the estimated effectiveness of the GSP.

The primary objective of this study was to determine if the GSP increased sales at Hannaford stores of ready-to-eat (RTE) cereals that the GSP considers more nutritious at the expense of RTE cereals that the GSP considers less nutritious. We used two proprietary data sets to estimate cereal demand for Hannaford (treatment) and Hannaford-like (control) stores. We estimated changes in cereal demand pre- and post-GSP for both Hannaford and the control stores and then statistically tested the pre- and post-GSP differences between Hannaford and control stores.

Section snippets

Data

The data for this study came from two proprietary sources—the Guiding Stars database provided by the Guiding Stars Licensing Company and Scantrack StoreView data of RTE cereal sales in 13,175 stores across the continental United States (US) provided by Nielsen. The GSP is implemented at Hannaford stores, which represent the treatment group. The selection of Hannaford-like stores that constitute the control group is explained later in this study.

Identification of control stores

Field experiments offer the best method for evaluating a program such as GSP (IOM, 2012). A key element in designing a strong field experiment is the inclusion of a control group, which helps separate program effects from non-program effects. These non-program effects may come from a variety of sources: macroeconomic conditions; changing consumer attitudes and knowledge about nutrition; as well as other supply shifters such as cost changes, advertisements, and product reformulations (Shum, 2004

Rotterdam demand model

We utilized the Rotterdam demand system (Barten, 1964, Theil, 1980) to examine purchases of cereals, which were aggregated into four GSP star rating categories – unstarred, 1-star, 2-star, and 3-star. The original Rotterdam model can be specified aswidlnqi=μiDQ+Σjπijdlnpj,i=1,2,,n,where wi = piqi/m is the budget share for good i; dln qi and dln pj are log changes in quantity (qi) and price (pj), respectively; μi = pi(∂qi/∂m) is the marginal propensity to consume; DQ = Σiwidln qi is the Divisia volume

Empirical model

The primary objective of this study was to estimate the effect of the GSP on sales of four star categories of cereals. Our data span from September 2005 to August 2006 (1 year prior to the implementation of the GSP in September 2006) and September 2006 to April 2008 (20 months under the GSP). We hypothesized that the GSP effect is cumulative and can be captured by a time trend variable (taking the value of 0 before September 2006 and the value of 1 for the first week of September 2006 and a

Descriptive statistics

We examined weekly sales data for four cereal categories at 134 Hannaford stores and 134 control stores over 140 weeks—the pre-GSP period of September 2005 to August 2006 and the post-GSP period of September 2006 to April 2008. The descriptive statistics of the variables included in the Rotterdam model are reported in Table 2. The cereal market is dominated – in terms of volume sales, dollar sales, and number of SKUs – by less nutritious cereals. Note that cereal prices are not correlated with

Results of demand estimation

The GSP coefficients indicate that the market shares of starred cereals (i.e., 1-, 2-, and 3-star) rose following implementation of the GSP at the expense of unstarred cereals, as shown in both Hannaford and control results (Table 3). These GSP coefficients are significant at a 5% probability level for both the Hannaford and control stores.

Empirical results suggest that cereal demand by nutritional attribute follows a socioeconomic gradient at Hannaford. The demand for unstarred cereals falls

Conclusions

The Institute of Medicine (IOM) recently completed a comprehensive review of the rich literature on the front-of-package (FOP) nutrition labeling system. IOM noted several methodological weaknesses in the literature, including the lack of control conditions to establish unconfounded effects of FOP labeling on consumers’ food choices. Given IOM’s findings, the primary objective of this study was to rigorously investigate, as a case study, the effect of one of the FOP systems, the Guiding Stars

References (33)

  • Culver, S.F., Cape, E., Fischer, McBride, W., Sutherland, L.A., Vitagliano, J.A., 2011. U.S. Patent No. 208,151. U.S....
  • E. Finkelstein et al.

    Annual medical spending attributable to obesity: payer-and service-specific estimates

    Health Affairs

    (2009)
  • L.M. Fischer et al.

    Development and implementation of the guiding stars nutrition guidance program

    American Journal of Health Promotion

    (2011)
  • X.S. Gu et al.

    Comparison of multivariate matching methods: structures, distances, and algorithms

    Journal of Computational and Graphical Statistics

    (1993)
  • S. Ichimura

    A critical note on the definition of related goods

    Review of Economic Studies

    (1950)
  • Institute of Medicine (IOM)

    Examination of Front-of-Package Nutrition Rating. Systems and Symbols: Phase I Report

    (2010)
  • Cited by (58)

    View all citing articles on Scopus

    The views expressed in the study are those of the authors and cannot be attributed to the U.S. Department of Agriculture or the U.S Department of Health and Human Services.

    View full text