Background
Worldwide, overweight and obesity have become a major public health problem as they constitute a considerable risk for many chronic diseases such as certain cancers, type 2 diabetes and coronary heart diseases [
1,
2]. Consequently, effective public health strategies are urgently needed to prevent overweight and obesity and reduce the burden of disease [
3]. A high intake of energy dense foods has been associated with excess weight gain [
4]. Over the past decades, the number of snacking occasions per day and energy density of snacks have increased substantially [
5].
As empirical evidence for the strong environmental influence on overeating and excess weight gain is growing [
6], environmental approaches are increasingly being advocated as offering important potential for improving eating habits. Consumers make their decisions in an environment in which many cues may influence their purchase, although they are often not aware of this influence. Increasingly, efforts are being done to reshape the environment in which consumers make their food decisions [
7]. Recently, the concept of nudging has become very popular in this respect [
8]. Nudging works from the setting in which a choice is presented and aims to make beneficial choices (from a consumers’ perspective) more appealing [
9]. They can be seen as relatively simple, easy to implement and inexpensive interventions. A key characteristic of nudging is that consumers maintain their liberty of choice [
10]. This implies that interventions do not ban ‘forbidden’ products or enforce consumers to make a particular choice.
The present study examines the possibility of nudging consumers towards healthier snack choices near the checkout counter by altering the accessibility and availability of both healthy and unhealthy snacks. Both availability and easy accessibility of energy dense foods have been identified as risk factors for overeating [
11]. In food studies, availability and accessibility has been conceptualized in various ways, such as spatial accessibility and means of transportation to stores [
12]. Both of them have been recognized as promising starting points to design effective intervention tools. Interventions based on these features have typically been studied separately or combined with other interventions such as price alterations or labelling. For example, results from an intervention at a large hospital cafeteria suggest that a color-coded labelling intervention in combination with a rearrangement of healthy items (displaying them at eye level) can contribute to increased sales of these items [
13]. In another study, accessibility of (un)healthy foods and the provision of caloric information were manipulated on paper menus distributed at a fast food sandwich chain. It was found that participants were more likely to choose low-calorie options when these were put front-page rather than presented at the back page of the menu [
14]. Hanks and colleagues [
15] showed that adding a convenience line displaying only healthy foods in a school lunch room increased sales of healthier food by 18% and decreased sales of less healthy foods by nearly 28%. Maas and colleagues [
12] manipulated accessibility by changing distance to unhealthy snacks in a lab study. They found that putting snacks further away reduced the likelihood and amount of snack intake. Similarly, making a food slightly more difficult to reach reduced intake of salad bar food by 8-16% [
16]. Not all studies find such strong effects, probably because of strong existing preferences for particular products [
17]. For example, Meyers and colleagues [
18] manipulated the accessibility of high- and low-calorie desserts in a hospital cafeteria. Making low-calorie desserts less accessible decreased the likelihood of their selection and resulted in fewer desserts taken at all. However, when high-calorie desserts were made less accessible and visually salient, people accepted the inconvenience of reaching for them rather than eating a low calorie dessert. In sum, the few experimental studies that have explored accessibility provide promising but sometimes inconclusive results, probably due to different type of manipulations and settings (i.e. lab environment, school, hospital canteen, fast food chain).
Besides manipulating accessibility or convenience, studies manipulating availability typically origin in the marketing field [
19,
20]. For example, although this study was conducted almost four decades ago, Curhan [
20] showed that doubling shelf space for hard fruit in a supermarket increased sales by 44%. In a more recent study on vending machines in bus garages, the number of available healthy items was increased while at the same time prices of these items were lowered. Although the effect of lowering prices cannot be separated from the effect of increasing availability, the interventions resulted in 10-42% higher sales of healthy items [
21]. All in all, previous studies particularly focused on the effects of accessibility, while availability has received less research attention in the field of public health promotion. In this study, accessibility is defined as the convenience or closeness of physically obtaining a product in a shelf space. We define availability as the presence of snacks ready for immediate choice by consumers.
No research to date combined both the effects of availability and accessibility to better understand their potential as a simple strategy to help people make healthier food choices in a worksite canteen. To inform new policy making in public health, the contribution of this paper is that we address the effect of availability on healthy food choices, both as a main effect and in combination with accessibility that has been studied more widely. Moreover, a lab study alone is not robust enough to provide reliable predictions about how people behave in a real-life context. Therefore, we address these effects both at the fundamental level (controlled lab experiment) as well as at the public health implication level (a realistic real life setting).
Research in the field of behavioural economics has shown that consumer decisions are often irrational and prone to biases [
10]. Visitors of a self-service restaurant or canteen are often in a hurry and tend to be hungry, which makes them more prone to these biases. The interventions in our study aim to take advantage of biases related to easy accessibility and availability. Altering the ease of access to snacks basically means that it requires less effort to obtain them. People furthermore have the tendency to go for the default option as this typically requires the least effort due to habits or even ‘laziness’ [
22]. Availability can influence consumer choices in various ways. More healthy snacks present for purchase increases the likelihood that consumers find a snack that fits their need. Higher availability of healthy snacks leads to a larger assortment which tend to raise consumer expectations and satisfy consumers with a high need for variety [
19,
23]. By enlarging the available assortment of healthy snacks, we made these snacks the implicit default. This may function as a cue that implies a consumption norm. Typical checkout counter assortments including a variety of chocolate bars, candy and savoury snacks speak to the desire of immediate pleasure rather than uncertain rewards such as achieving a healthy body weight. Food products and snack displays are often designed to stimulate the automatic affective system of human behaviour, which is being driven by triggers in the environment. Therefore, our intervention aims to make taking a healthy snack slightly more convenient and attractive while preserving freedom of choice.
Recently, ethical concerns have been raised that nudges could be too intrusive and restrict (perceived) freedom of choice [
24,
25]. For example, strategies that focus on offering people a default option are problematic if people tend to overlook the possibility to make alternative choices [
26]. Moreover, according to reactance theory [
27], eliminating choice options or putting pressure on people to eat healthy may lead to reactance in that some consumers will do exactly the opposite (e.g. go for the unhealthy option). Therefore, an additional aim of this study is that we also examined participants’ perception of their freedom in making choices and choice satisfaction.
The first study reported provides a lab experiment into whether shelf arrangement (location of healthy snacks on top versus bottom of shelf display) and assortment structure (assortment includes either 75% or 25% healthy snacks) influences the type of choices consumers make. Pictures of shelf displays were created to simulate snack food displays that can typically be found near the checkout counter of self-service restaurants or canteens. Because this study was conducted with tightly controlled stimulus material in an imaginative choice situation, the question is whether the results would generalize to a ‘real-world’ snack choice situation. Therefore, the second study is a field experiment in which the same manipulations were applied to an actual snack shelf display in a hospital canteen. Worksite restaurants are important intervention sites because of trend towards increasing out-of-home consumptions [
28].
To summarize, changes in a snack assortment in terms of effort required to actually grab a product and the size of an assortment may make healthier snacks more salient, attractive, normative and convenient. For both studies, it was hypothesized that the individual and combined effect of increasing the availability of healthy snacks (while at the same time not banning unhealthy snacks) and positioning them at the top of the shelf display would increase sales of these healthy options. We manipulate the number of facings of healthy snacks and the vertical position of 16 snacks, while keeping total shelf space constant and retaining the availability of unhealthy options.
In sum, the objective of this article is to examine the effects and interplay between shelf arrangement and assortment structure on consumer choices for healthy and unhealthy snack products.
Discussions and conclusions
In this paper, we presented two experiments, one controlled lab study and one field study in a hospital canteen that examine the combined effects of the availability and shelf arrangement of healthy snacks on snack decisions and consumer perceptions of assortment and choice. The results of both studies show that changing the snack assortment can change consumers’ choice in a more healthful direction when the majority of the assortment consists of healthy options while allowing for more unhealthy choices. Within the setting that we investigated, more shelf space for healthy snacks increased the likelihood that people choose for a healthy snack (lab study) and led to higher sales of healthy snacks (field study).
There are several possible explanations for these findings. A probable mechanism which could explain our findings is that people try to minimize the energy spent in obtaining food [
16]. More healthy foods in shelves probably enhanced their desirability and visibility. Normally, the small number of fruit and other healthy snacks hardly attract consumers’ attention. Measurement of shelf spaces in various types of stores showed that stores all devote more shelf space to unhealthy than healthy items. Smaller shops such as convenience shops typically assign a disproportionally high amount of shelf space to more unhealthy food such as energy rich snacks foods and beverages [
31]. As such, our findings resemble findings of marketing studies showing that the number of shelf facings has strong impact on how consumers pay attention and evaluate products and brands [
32,
33]. A larger assortment also makes it more likely that consumers find a product that fits their needs. Recall that across studies, we manipulated both the proportion of available healthy snacks and the position of these snacks at shelves. As a result, not only the number of available products, but also the product variety within type of snack (i.e. healthy or unhealthy) differed across conditions. Previous research suggests that more variety increases satisfaction [
34]. Unfortunately, our design did not allow for disentangling this combined effect of number and variety of options. Future research could address this issue.
Putting healthy snacks at the most convenient (top) shelf space did not impact consumer choices and sales, as both the lab study and the field study showed. In other words, making unhealthy snacks less accessible did not discourage their selection. Although a marginal effect was obtained for shelf arrangement in that snacks on top shelves were more often selected, this effect did not reach statistical significance. This absence of an effect of accessibility was unexpected. It may be that the accessibility manipulation was not strong enough to actually change convenience or the amount of data was too limited. The simulated choice task in the lab study probably did not change the nearness of item selection. In the field study, visitors had to bend at their knees to obtain a snack from the lower shelves. Furthermore, a limitation of adding healthy alternatives to choice sets can be that this increases the consumption of indulgent food items [
35]. In that case, the presence of healthy items provides a license to indulge. We could not find such an effect.
Nudging is based on the premise that it is justifiable to encourage consumers to make better choices, by gently pushing them in the right direction. Although the majority of respondents of the field study survey did not observe any changes in the assortment, for a group of employees the special and new shelf display captured their attention. Many choices in settings such as canteens and grocery stores are relatively low involvement choices; consumers do not actively process available information about choice alternatives. However, the unexpected display in the canteen may have led to more effortful attention of employees, who might otherwise neglect it after a longer time period. Interestingly, although our manipulations did not alter perceptions of freedom of choice in the lab study, surveyed hospital employees experienced more freedom in snack choice in situations were healthy snacks were visibly at top shelves. This may be understandable given their strong preference for a relatively large healthy snack assortment. Overall, about 31% of all sold snacks in the hospital were unhealthy while for students in the labs study about 71% of all snacks choices were unhealthy. This could be the result of the different samples studied, each with its own snack preferences. It could also be related to expectations that people have of the type of snacks that are appropriate to eat within the context that they buy them (hospital versus university canteen). Additionally, the assortments with 75% unhealthy snacks were seen as more realistic than assortments with 25% unhealthy snacks. The increased healthy choices may be due to the demand effects of the research context which may have hinted at health and food choices.
Despite the clear results, some limitations must be noted about both studies. One is the artificial nature of the choice task in the first lab study and the involvement of student participants. Although the assortments in the lab study including 75% healthy snacks were seen as less realistic, results showed that our manipulation did not alter choice satisfaction. Perhaps the assortment did not look like a typical snack assortment at a university canteen or the instruction being given (select favourite snack) influenced choice. Another limitation refers to the relative small number of snacks sold. In the field study, fruit had a lower price which may have given these products an additional benefit. Moreover, it is unclear whether our intervention could produce sustained changes in snack choice behaviour. It may be possible that eating a healthy snack is compensated later in the day by eating or snacking more. Despite such limitations, our experiments demonstrated that increasing the availability of healthy snacks may be effective in a canteen setting. Caution, however, should be exercised in generalizing these findings to other settings such as grocery stores or restaurants. Seymour and colleagues [
36] reviewed the effectiveness of nutritional interventions in worksite and university cafeterias involving the availability, access to, pricing or and information about fruit and vegetables. They concluded that these interventions are more likely to be successful than those in restaurant and grocery stores. This is because these ‘limited access’ sites have fewer other options available. Nevertheless, the work environment can have a large impact on food choices as people tend to consume a large part of their daily energy intake during their work day. Moreover, previous research has shown that a healthy worksite food environment can have productivity-enhancing effects [
37].
Our results show that a relative large assortment of healthy snacks is able to influence consumers’ healthy snack purchases. With the present studies, we extend our understanding of the effects of accessibility and availability of snack foods on consumer choices. Our research examined two factors that might play a role in healthy snack choices. An interesting topic for further research would involve the identification of other important factors that gently nudge consumers towards healthier snack choices and the effectiveness of these factors in various contexts.
Competing interests
The authors declare that they have not competing interests.
Authors’ contributions
All authors (EvK, KO, and HvT) contributed to the design of the study. EvK and KO conducted the statistical analysis and drafted the manuscript. All authors participated in the interpretation of the data and critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.