Nudging

This review focuses on interventions that aim to promote sustainable food choices and discourage less sustainable options. More specifically, we provide an overview of the results obtained by means of nudging. Nudging aims to change people's behaviour in a predictable way without forbidding any options or significantly changing their economic incentives^{(Reference Thaler and Sunstein17)}. Hence, nudging differs from initiatives that aim to make sustainable consumption more economically appealing, for instance subsidies for solar panels. Moreover, nudging initiatives tend to be relatively less costly than educational initiatives^{(Reference Benartzi, Beshears and Milkman18)}. Nudging also does not impose any restrictions as do enforced laws, such as the banning of plastic, but embraces the freedom of choice for the customer^{(Reference Sunstein19)}. This principle of libertarian paternalism is one of the two core principles of nudging^{(Reference Thaler20)}.

Indeed, nudging interventions are rather unobtrusive and trigger consumer responses without requiring much cognitive effort^{(Reference Thaler and Sunstein17)}. This difference in cognitive effort is also exactly what differentiates nudging from classic informational campaigns and education, which aim to convince consumers to change their attitudes with rational arguments^{(Reference Guthrie, Mancino and Lin21)}; a process that requires more cognitive effort^{(Reference Andrews, Netemeyer and Burton22)}. This approach might not be suitable in a food context, where a choice between options is often guided by fast, automatic and/or cognitively effortless responses to environmental stimuli. A thoughtful, slow and/or analytical processing of all available food options would be too time-consuming in many conditions, such as in making food decisions^{(Reference Wansink and Sobal23)}.

Indeed, many so-called dual process theories distinguish between two processing styles^{(Reference Evans24)}, of which Kahneman's System 1 (fast and effortless processing) and System 2 (slower and analytical processing) are the most prominent^{(Reference Kahneman, Frederick, Gilovich, Griffin and Kahneman25)}. Even if communication campaigns succeed, such that sustainable information is processed under high cognitive effort and positive attitudes have been created, behavioural changes are still not guaranteed. The attitude–behaviour gap for sustainable behaviour illustrates that consumers often have positive predispositions towards sustainable consumption, but these do not always translate into actual behavioural changes^{(Reference Vermeir and Verbeke26)}.

For this reason, nudging starts from the perspective of the consumer. Knowing how consumers make food decisions in a complex food environment provides insight into how food presentation may lead to sustainable food purchase choices by the consumer. This is called an optimisation of the choice architecture, which establishes the second core principle of nudging^{(Reference Thaler20)}. Many elements in a food environment will affect our choices, such as the way products are positioned^{(Reference Bucher, Collins and Rollo27)}, their visibility^{(Reference Wilson, Buckley and Buckley28)} or packaging^{(Reference Tijssen, Zandstra and de Graaf29)}. The results of these types of interventions at the point of purchase, such as in retail stores, cafeterias and restaurants, seem promising^{(Reference Vermeer, Steenhuis and Leeuwis30,Reference Brambila-Macias, Shankar and Capacci31)} , and consumers also believe that these interventions can help them more easily regulate their in-store consumption decisions^{(Reference Vermeer, Steenhuis and Seidell32)}.

Our review provides a non-exhaustive overview of the application of nudging for more sustainable food choices at the point of purchase. For some categories of nudging, research on sustainable food choices is still in its infancy, so we enriched our discussion with insights from general sustainable consumption, health research, policy making and consumer behaviour. In doing so, we also highlighted the domains where more research is needed. To the best of our knowledge, no framework on sustainable food choices has yet been published. Therefore, we built our review on the framework for the in-field health interventions of Cadario and Chandon^{(Reference Cadario and Chandon33)}. They categorised the nudges according to whether the nudge exerted an influence on consumers' cognition (i.e. consumer knowledge), effect (i.e. consumers' feelings) or behaviour (i.e. motor responses).

Descriptive labelling

Consumers' interest for specific product attributes, such as origin, ingredients and production process, has gradually increased over the years^{(Reference Costanigro, Kroll and Thilmany34)}. Consequently, labels focusing on food attributes have become increasingly popular, as demonstrated by the many studies on energy and nutrition labelling (see Cowburn and Stockley^{(Reference Cowburn and Stockley35)} for a systematic review). A similar trend can be detected for sustainable products, such that many products display sustainable labels on food packaging, menus in restaurants or next to the product in buffets^{(Reference Costanigro, Kroll and Thilmany34)}. The complexity of what can be considered sustainable consumption also implies that a wide variety of sustainable labels can be used^{(Reference Gruère36)}, ranging from where the product is coming from (i.e. ‘local’^{(Reference Feldmann and Hamm37,Reference Telligman, Worosz and Bratcher38)} ), to how it is produced (i.e. ‘organic’, ‘fair trade’, ‘recycled’^{(Reference Poelman, Mojet and Lyon39,Reference Songa, Slabbinck and Vermeir40)} ) or which non-sustainable ingredients are removed from the product (i.e. ‘free from’^{(Reference Hartmann, Hieke and Taper41)}).

Sustainable labels might function as eye-catchers when consumers are shopping in an environment where many products are competing for attention and cognitive resources are limited^{(Reference Milosavljevic, Navalpakkam and Koch42)}. Labels can also help consumers determine whether a product is sustainable^{(Reference Newell and Siikammki43,Reference Kallbekken, Saelen and Hermansen44)} . For example, labels can give information on the seasonality of products^{(Reference Feldmann and Hamm37,Reference Slabbinck, Vandenbroele and Van Kerckhove45)} , and whether the product is certified by a third party according to specific sustainability criteria (cf. Rainforest Alliance certificate⁽⁴⁶⁾) or human respect criteria (cf. Fair Trade certificate⁽⁴⁷⁾). Recent research also shows the positive effects of sustainable labels on product perceptions; eco-labels have positively affected taste judgements and willingness to pay^{(Reference Sörqvist, Haga and Langeborg48)}, wine labelled as organic has led to higher taste ratings^{(Reference Wiedmann, Hennigs and Behrens49)}, and products that are defined as ‘local’ have become more popular because of higher taste and quality associations^{(Reference Feldmann and Hamm37)}.

However, descriptive labels not only impact consumers' attention, information and associations with sustainable products but also have been shown to evoke emotional and behavioural responses. Products labelled as organic are believed to contribute to environmental protection (i.e. appealing to utilitarian attitudes), and as such, people experience emotional gratification from buying organic (i.e. appealing to hedonic attitudes^{(Reference Lee and Yun50)}). Even before buying, feelings of anticipated pride can predict pro-environmental choices later on^{(Reference Schneider, Zaval and Weber51)}. Nonetheless, a more negative feeling (i.e. guilt) has been demonstrated to make people feel more personally responsible for environmental damage and increase the likelihood that people buy products with sustainable labels^{(Reference Antonetti and Maklan52)} or support pro-environmental behaviour^{(Reference Rees, Klug and Bamberg53,Reference Lu and Schuldt54)} . Eco-labels that focus on omitting non-sustainable ingredients, such as ‘free from palm oil’, affect behavioural intentions such as the ‘willingness to pay a price premium’ among people who actively look for information and have a high preference for naturalness^{(Reference Hartmann, Hieke and Taper41)}.

Numerous studies have demonstrated the effects of sustainable labels on consumer response; however, the direction of the effects is ambiguous and largely determined by personality traits such as environmental concern^{(Reference Poelman, Mojet and Lyon39,Reference Sörqvist, Haga and Langeborg48,Reference Verain, Sijtsema and Antonides55–Reference Piqueras-Fiszman and Spence58)} . For example, whether consumers notice sustainable labels has been shown to depend largely on their implicit (i.e. neurophysiological emotional reactions) and explicit (i.e. self-reported feelings) positive predispositions towards sustainable behaviour^{(Reference Songa, Slabbinck and Vermeir40,Reference Van Loo, Caputo and Nayga59)} . Moreover, the positive effects of sustainability claims are more profoundly observed when consumers understand and trust the labels^{(Reference Vermeir and Verbeke26,Reference Janssen and Hamm60–Reference Samant and Seo62)} . Organic labels may also backfire, such that consumers who are less concerned with the environment perceive organic labelled products as less tasty^{(Reference Schuldt and Hannahan63,Reference Lee, Shimizu and Kniffin64)} . Buying organic labelled products might also trigger compensation behaviour (i.e. additional purchases of less sustainable products), so that consumers end up with a higher total environmental impact of their food purchases^{(Reference Gorissen and Weijters65)}. Moreover, the wide variety and multiple application of sustainable logos^{(Reference Pierrot, Giovannucci and Kasterine66)} can be confusing for consumers^{(Reference Gadema and Oglethorpe67)}, so that they might follow the heuristic that a labelled product is better than one without label, instead of actually deciphering the provided information on the sustainability label of the product^{(Reference Hartmann, Hieke and Taper41,Reference Janssen and Hamm60)} . To better align labels with consumers' preferences, more research is needed on the level at which labels should be created (e.g. Belgian v. European^{(Reference Van Loo, Caputo and Nayga68)}) and on which measures they should be based (e.g. food miles v. CO₂ emissions^{(Reference Van Loo, Caputo and Nayga68)}). Lastly, few studies have investigated how sustainable labels affect actual food choices; a 2013 study^{(Reference Grunert, Hieke and Wills69)} even found that sustainable logos did not affect the food choices of European consumers at the time.

Evaluative labelling

As consumers face a hard time deciphering different sustainable attributes, it is suggested to ease their understanding by providing evaluative information rather than descriptive information⁽⁷⁰⁾. Evaluative labels of sustainability can use ratings that provide an immediate indication of how sustainable the product is, such that food labels carry low–medium–high descriptions or a one-, two- or three-star rating^{(Reference Sutherland, Kaley and Fischer71,Reference Rosi, Zerbini and Pellegrini72)} . A field study on healthiness showed that significantly fewer ‘high-fat’ and ‘high-energy’ products were sold when labelled as such, while the sales of ‘low-fat’ and ‘low-energy’ products went up^{(Reference Cioffi, Levitsky and Pacanowski73)}. One study on sustainable evaluative labelling^{(Reference Cho and Baskin74)} also highlights the importance of healthiness perceptions. They found that a high sustainability label increases product preference only if the food product is perceived as being high in healthiness. Hence, a fit between the level of sustainability and healthiness was needed for the ‘high’ sustainable label to be successful, perhaps because the average consumer expects sustainable products to be more healthy, as demonstrated in another study^{(Reference Verain, Sijtsema and Antonides55)}. In a similar vein, ample research efforts have recently been put into creating labels that combine the impact of food on healthiness and sustainability^{(Reference Hallström, Davis and Woodhouse75,Reference Heller, Keoleian and Willett76)} . These labels provide an overall dietary quality score, although the approach on how to quantify this measure is not yet unified among researchers.

Colour coding can be another way to support consumers in making better decisions^{(Reference Vanclay, Shortiss and Aulsebrook77)}. For example, traffic light labelling builds on our socially learned associations that the colour green signals safety, while red signals danger, which leads to the assumption that red can function as a consumption-stopping cue^{(Reference Genschow, Reutner and Wänke78)}. Applying this idea to evaluative labelling, more sustainable products would be marked with a green label, while less sustainable products would be marked with a red label^{(Reference Thorndike, Sonnenberg and Riis79,Reference Olstad, Vermeer and McCargar80)} . Research showed that health decisions were improved by applying traffic light labelling in a cafeteria or university canteen^{(Reference Thorndike, Sonnenberg and Riis79–Reference Thorndike, Riis and Sonnenberg81)}, although an online experiment found mixed results depending on the country of origin (positive effect in Germany v. no significant difference in Poland)^{(Reference Aschemann-Witzel, Grunert and van Trijp82)}. In a grocery setting, the sales of black labelled products (i.e. high carbon emission) decreased by 6 %, while the sales of green labelled products (i.e. low carbon emission) increased by 4 %^{(Reference Vanclay, Shortiss and Aulsebrook77)}. Interestingly, the intention to avoid red labels is greater than the preference for green labels^{(Reference Rosi, Zerbini and Pellegrini72,Reference Scarborough, Matthews and Eyles83)} . This might be explained by consumers' perception that negative information (i.e. red logo) looms larger ^{(Reference Baumeister, Bratslavsky and Finkenauer84)} than positive information. However, the effectiveness of colour labelling does not generalise across products; the negative effect of a red surface on food selection appears to be more profoundly observed for unhealthy (v. healthy) products^{(Reference Reutner, Genschow and Wänke85)}. This highlights the importance of a fit between consumers' perceptions on what is ‘bad’ and the colour (e.g. unhealthy food and red) and that the effects are less clear when there is a mismatch (e.g. healthy food and red). This raises the question of whether colour coding would work for sustainable food choices, for which consumers' associations on what is ‘bad’ and ‘good’ might be less strong.

Note that not all studies find positive effects in evaluative labelling on food choices^{(Reference Aschemann-Witzel, Grunert and van Trijp82,Reference Hoefkens, Lachat and Kolsteren86)} , and some authors warn of compensatory consumption^{(Reference Chandon and Wansink87)}. Consumers might compensate for ‘low’ label purchases with less sustainable choices in side dishes or in their following food choices. Moreover, they may be exclusively focused on the characteristic that is ‘low’, so they overlook other negative characteristics of the food product (e.g. low-fat and high-sugar food^{(Reference Wansink and Chandon88)}). Future research should determine whether evaluative labelling on sustainable products also triggers such compensatory effects.

Visibility enhancements

Visual attention (i.e. the amount of fixation time) is an important proxy for product selection^{(Reference Armel, Beaumel and Rangel89)}; therefore, the positioning of sustainable products at the point of purchase should be carefully considered. For example, products that are presented at eye level^{(Reference Glanz, Karpyn and Wojtanowski90,Reference Gamburzew, Darcel and Gazan91)} have been shown to draw more attention and enter more easily the set of products considered by the consumer^{(Reference Keller, Markert and Bucher92)}. Another attention hotspot in retail stores is the shelf next to the cash register, where consumers have time to process the offerings while queuing to pay^{(Reference Kroese, Marchiori and de Ridder93)}. Even the order of products on a restaurant menu can impact food choices, such that products placed at the top of the menu benefit from a primacy effect and tend to be selected more often^{(Reference Dayan and Bar-Hillel94,Reference Policastro, Smith and Chapman95)} . A recent study^{(Reference Kurz96)} applied two visibility interventions on sustainable products for 9 months in a university canteen by changing the menu order (i.e. from the middle to the top position) and increasing the visibility of vegetarian products (i.e. more visible location). Combining both interventions increased the sales share of the vegetarian dishes significantly compared to all purchases with six percentage points. Moreover, a significant long-term effect was observed, so that after the intervention period, the share of vegetarian dishes was still four percentage points higher compared to the baseline period.

Increasing the availability of sustainable products in the assortment is also often related to greater visibility, and together, they have a strong effect on purchase intentions^{(Reference Wilson, Buckley and Buckley28)}. Although the assortment of more sustainable food choices, such as organic and vegetarian products, has been increasing over the years, its share is still relatively small compared to general products^{(Reference Sadler97,Reference Schösler, Boer and Boersema98)} . Studies on healthier food choices show promising results in that increasing the relative share of healthier foods (e.g. from 25 to 75 %) translates into more healthy purchases: the chances of buying a healthy product were multiplied by 2·9 times in an online study and 3·5 times in a canteen^{(Reference van Kleef, Otten and van Trijp99)}. Interestingly, regarding the principles of nudging^{(Reference Thaler and Sunstein17)}, consumers did not perceive the assortment to be more restrictive by the intervention. Another field experiment combines healthy labelling and increased availability so that students chose more healthy options in the vending machine^{(Reference Rosi, Zerbini and Pellegrini72)}. One study found mixed results: increasing the share of healthy options encouraged the sales of healthy dishes in only two out of six cafeterias^{(Reference Pechey, Cartwright and Pilling100)}.

Incorporating sustainable products in the existing assortment can be challenging for managers^{(Reference van Herpen and Bosmans101)}, as they can choose to mix them with the existing assortment (i.e. comparative positioning) or offer them separately (i.e. unique feature positioning). For example, meat substitutes, which are vegetarian products that are sensory similar to a specific meat product, are currently mostly shelved in a unique feature positioning^{(Reference Gravely and Fraser102)}. However, this separate shelf might be consistently skipped by non-vegetarians, so that meat substitutes do not enter the consideration set of these consumers. Non-users could be helped in recognizing meat substitutes as valuable alternatives by placing them next to the mimicked meat product^{(Reference Vandenbroele, Slabbinck and Van Kerckhove103)}. Suggestions are made that in this choice architecture, meat substitutes might appear as more familiar and less novel, as their sensory similarity with the mimicked product can be easily observed^{(Reference Adise, Gavdanovich and Zellner104–Reference Hoek, Pearson and James106)}.

This positioning question also pertains to when evaluative labels are used (e.g. one-, two- or three-star sustainability rating) and raises the question of where the ‘mildly’ sustainable products should be offered. Research suggests^{(Reference van Herpen, Fischer and van Trijp107)} that when there is a fit between the price setting and the product (i.e. intermediate price and intermediate star rating), sales increase if these products are positioned in a separate section. However, if there is no match and the two-star product is offered at a low price, mixing them with the low and high ratings (i.e. comparative positioning) has been found to boost the sales of the intermediate product. Positioning is also a critical challenge in online food retail, as the categorisation of the products highly affects purchases^{(Reference Chandon, Hutchinson and Young108)}. Moreover, environmental attitudes should again be considered when optimizing shelf management. That is, if the assortment is aligned with the consumers' goal, so that organic products are presented separately and the consumer has the intention to buy an organic product, organic purchases have been found to be higher compared to when they are comparatively positioned with other products that are not related to consumers' goal^{(Reference van Herpen, Fischer and van Trijp107)}.

Hedonic enhancements

Hedonic interventions trigger our senses so that products become more attractive by appealing to our taste, vision, olfaction (i.e. smell), audition and haptics (i.e. touch). Research on sensory marketing has gained importance over the past few years^{(Reference Krishna109)}. Therefore, we will extend the original category in the framework of Cadario and Chandon^{(Reference Cadario and Chandon33)} on hedonic enhancements by discussing research on each sense. We categorise on the main sensory characteristic of the nudge, although some nudges might trigger other senses and processes too, so the categorisation is not mutually exclusive.

Social influences

In addition to changing the food choice architecture, people can also be nudged by other consumers' behaviour^{(Reference Griskevicius, Tybur and Van den Bergh157)}. Because social desirability plays an important role in sustainable behaviour^{(Reference Cerri, Thøgersen and Testa158)}, we will extend the discussion of Cadario and Chandon^{(Reference Cadario and Chandon33)} on ‘healthy eating calls’ by incorporating insights into different types of social influences. First, the use of social norms is a frequently applied intervention to nudge consumers towards a desired behaviour; it accompanies rules and standards that are shared by the members of the same community^{(Reference Cialdini, Trost, Gilbert, Fiske and Lindzey159)}. Social norms can be either descriptive or injunctive, such that descriptive norms describe the behaviour of people (e.g. the average household energy consumption), while injunctive norms also evaluate this information in terms of consumers' approval (e.g. adding a happy or sad smiley to indicate whether people are under or above the average energy consumption^{(Reference Schultz, Nolan and Cialdini160)}). Authors warn that descriptive norms can backfire, for example, when people score better than the provided norm (and consequently increase their energy consumption^{(Reference Schultz, Nolan and Cialdini160)}) or when the message implies that it is acceptable to perform harmful behaviour because many people are doing so (e.g. the majority of the people throw cigarette butts on the street^{(Reference Cialdini161)}).

Nonetheless, these counterproductive effects can be eliminated by adding an injunctive norm to descriptive information (e.g. sad/happy smiley^{(Reference Schultz, Nolan and Cialdini160)}). Anthropomorphic cues, attributing human characteristics (e.g. sad faces) to non-human products, have been demonstrated to evoke feelings of sympathy towards the messenger that appear to reinforce normative behaviour, so that it eventually leads to increased sustainable behaviour to favour the messenger, but only when this behaviour does not come at an extra cost^{(Reference Ketron and Naletelich162)}. Similarly, anthropomorphism can be applied to advertise wonky foods (i.e. misshapen produce). Although generally perceived as less tasty, higher purchase intentions when wonky foods are anthropomorphised show the potential of anthropomorphism as an intervention for battling food waste^{(Reference Cooremans and Geuens163)}. Other symbols, such as the use of predetermined compartments on lunch trays for vegetables and fruit, can also evoke a social norm to increase the intake of targeted products.

Another recommendation is to frame the descriptive message in a positive way (e.g. the majority of the people throw cigarette butts in the bin^{(Reference Elgaaied-Gambier, Monnot and Reniou164)}). A recent study^{(Reference Cerri, Thøgersen and Testa158)} shows the effectiveness of using positive descriptive norms if consumers perceive the message as credible, even when the promoted behaviour is not sustainable (i.e. avoid overpackaging) and not yet adopted by the majority of the people. Conversely, descriptive (but not injunctive) norms are better for promoting behaviour than preventing it^{(Reference Melnyk, van Herpen and Fischer165)}. The persuasiveness of the message might also play a role, such that a field experiment shows that stronger (v. weaker) norms are more effective in enhancing eco-labelled purchases^{(Reference Demarque, Charalambides and Hilton166)}. However, some studies found no effects of positive descriptive messages^{(Reference Momsen and Stoerk167,Reference Abrahamse and Steg168)} . Overall, social norms and feedback seem to be promising nudges^{(Reference Rettie, Burchell and Riley169)}, as also identified by a meta-analysis that classifies these interventions as having one of the greater effect sizes among interventions in promoting sustainable behaviour^{(Reference Osbaldiston and Schott170)}.

Another promising intervention indicated by this meta-analysis^{(Reference Osbaldiston and Schott170)} is verbal prompting. A recent field study^{(Reference Kristensson, Wästlund and Söderlund171)} applied a diverse set of verbal prompts, such as questioning whether customers would buy non-environmental or environmentally friendly bananas (cf. question-behaviour effect^{(Reference Sprott, Spangenberg and Block172)}), and approaching customers with the assumption that they seem interested in buying eco-labelled products. All verbal prompts increased sustainable choices compared with the baseline sales. In a self-service restaurant, more healthy side dishes were bought when customers were verbally prompted, while customers indicated feeling no buying pressure^{(Reference van Kleef, van den Broek and van Trijp173)}. Signalling towards other people appears to be very important; sustainable purchases can be increased by verbal prompting in the presence of other store-employees^{(Reference Kristensson, Wästlund and Söderlund171)}. Indeed, pro-environmental behaviour shows some links with status signalling^{(Reference Puska, Kurki and Lähdesmäki174)}, such that a status motive has been demonstrated to lead to more sustainable purchases^{(Reference Griskevicius, Tybur and Van den Bergh157,Reference Puska, Kurki and Lähdesmäki174)} . In fact, some evidence^{(Reference Hartmann, Ruby and Schmidt175)} shows that the signal works in that vegetarian and meat alternative (e.g. insect-based burger) eaters are perceived as being more environmentally friendly and even more brave and interesting compared to meat eaters.

When eating in a social environment, people are also affected by the food choice and intake of others and adapt these according to the group, as demonstrated in a meta-analysis^{(Reference Robinson, Thomas and Aveyard176)}. Field data from a restaurant show that people tend to seek variety in their individual dishes but not in the menu category^{(Reference Ellison177)}. This is an interesting finding to nudge sustainable food choices; perhaps if someone orders vegetarian food, the others might follow with dishes within the vegetarian food category. Promoting vegetarian foods by framing them as the ‘recommendation of the chef’ is also a way to prompt vegetarian food choices among infrequent eaters of vegetarian food^{(Reference Bacon and Krpan110)}.

Social influences even go beyond written and verbal interactions with people. For example, consumers attach some social beliefs to the positioning of products in-store, such that products in-between two products of the same category are believed to be more popular because of the middle position, which functions as a positive cue leading to higher purchases^{(Reference Valenzuela and Raghubir178,Reference Missbach and König179)} . If consumers do not have strong preferences for products, the scarcity of products on shelves might also function as a popularity cue that affects consumer choices^{(Reference Parker and Lehmann180)}. Lastly, in view of the many ways in which social influences affect sustainable purchases, note that social desirability biases in the survey and experimental research on sustainability should always be taken into account^{(Reference Cerri, Thøgersen and Testa158)}.

Convenience enhancements

To make sustainable food choices more attractive, they should be convenient and easy to access, while the required effort for buying less sustainable products is preferably larger. Studies on choice architecture show that accessibility enhancements, for example, placing fruit and vegetables at the beginning of the buffet increase the self-serving portions of these target products^{(Reference Kongsbak, Skov and Nielsen181)}. Moreover, the increase in fruit and vegetables was not compensated for by other meal components; the increase substituted servings of other meal components. Another study in a cafeteria setting^{(Reference Rozin, Scott and Dingley182)} changed the order of foods so that energy-dense products were less accessible to favour the sales of low-energy foods. However, no effects were observed by placing wholegrain bread in a more convenient place^{(Reference de Wijk, Maaskant and Polet183)}. Placing products on a lower v. higher shelf also did not affect the sales of healthier snacks^{(Reference van Kleef, Otten and van Trijp99)}. Perhaps, the effectiveness of product accessibility as a nudge depends on the type of product to which the nudge is applied. For example, a field experiment was not able to generate greater sales of unfamiliar vegetables by increasing their accessibility^{(Reference Broers, Van den Broucke and Taverne184)}.

A concept that is closely related to accessibility is proximity. In a trade-off between healthy (i.e. apples) and non-healthy options (i.e. popcorn) that are manipulated in terms of distance (within-reach v. 2 m away), researchers found higher consumption of the food that is positioned closer^{(Reference Privitera and Zuraikat185)}. Putting a bowl of M&Ms at 20, 70 and 140 cm decreased intake with increasing distance^{(Reference Maas, de Ridder and de Vet186)}. At work, it is shown that fewer snacks are purchased if the vending machine is placed further from the office desks^{(Reference Baskin, Gorlin and Chance187)}. Even small required efforts, such as when potato chips are located in a distant snack bar v. at the cash register, have been found to discourage unhealthy choices^{(Reference Meiselman, Hedderley and Staddon188)}. Notably, no main effects were found on the proximity (20 v. 70 cm) or snack type when simultaneously presenting a bowl of chocolates and fruit^{(Reference Knowles, Brown and Aldrovandi189)}. Rather, a relative proximity effect was observed, such that higher consumption of fruit was only obtained when the chocolate bowl was placed further (v. closer).

A typically employed and effective nudging strategy is to present the consumer with a ‘default option’, in which consumers are presented with a pre-determined item as a first or more prominent choice, thus requiring less effortful analytical decision making on the part of the consumer^{(Reference Goldstein, Johnson and Herrmann190)}. Defaults can also be used in food research; research shows that parents select healthier breakfast options for their children when they are readily available (v. only available on request)^{(Reference Loeb, Radnitz and Keller191)}. Another study shows that healthier whole wheat (v. white) bread was more often selected when it was the default option for ordering a sandwich^{(Reference van Kleef, Seijdell and Vingerhoeds192)}. Further evidence can be found in a restaurant setting^{(Reference Bergeron, Doyon and Saulais193)}, where healthier choices were ordered more when presented as the default option on the menu. Overall, these studies suggest the potential to increase sustainable food choices by making them the default option. Note that nudging theory does not support choice restriction^{(Reference Lombardini and Lankoski194)}.

Convenience can also be related to the way in which food products are offered. For example, self-servings are found to be lower when customers have to take the food with a pair of tongs v. a spoon^{(Reference Rozin, Scott and Dingley182)}. Other research shows that servings from squeeze tubes (v. regular jars) are more easy to monitor by consumers^{(Reference Huyghe, Geuens and Vermeir195)}, and that people eat less from resealable (v. non-resealable) packages because they have to re-open them^{(Reference De Bondt, Van Kerckhove and Geuens196)}, which might be interesting findings in the case of less healthy content. Findings are less clear on preparing foods so that they are easier to eat; pre-sliced fruit was perceived as less attractive in terms of sensory and packaging characteristics^{(Reference von Germeten and Hirsch197)}. Supermarkets are also increasingly offering bundles of food (e.g. all ingredients for a dish in one box). A recent study shows that this might be a promising nudge to counter the effect that people choose less healthy options when they are cognitively loaded as a result of the mental process of planning and sourcing individual ingredients^{(Reference Carroll, Samek and Zepeda198)}. Food bundles simplify shopping choices so that customers under cognitive load have been found to more frequently choose food bundles, although only if the food bundles are non-discounted (v. discounted). Packages of sustainable food products might thus be an appealing nudge in making sustainability more convenient.

Size enhancements

Portion size interventions might be especially interesting in decreasing the sales volume for products of which its production has a high impact on the environment, such as meat^{(Reference Steenhuis, Leeuwis and Vermeer199)}. Research has extensively demonstrated that portion sizes and intake are also positively related (i.e. portion size effect^{(Reference Zlatevska, Dubelaar and Holden200,Reference Hollands, Shemilt and Marteau201)} ) so that smaller portion sizes also decrease consumption. Verbal prompting by asking customers in a restaurant whether they want to downsize their portion successfully increased the number of smaller portions by one-third, while no compensation behaviour (i.e. higher energetic side dishes or drinks) was observed^{(Reference Schwartz, Riis and Elbel202)}. Another way to encourage smaller portions is offering them next to default larger portion sizes, which significantly nudged customers in a cafeteria to switch from a larger to a smaller portion size^{(Reference Vermeer, Steenhuis and Leeuwis30)}. A field experiment^{(Reference Vandenbroele, Slabbinck and Van Kerckhove203)} found that adding two smaller portions of meat sausage units next to the default portion size unobtrusively encouraged customers to buy smaller units. During a 1-month intervention period, more than half the units sold (52 %) were smaller portions than the default larger portion size, which established a decrease in meat volume of 13 % if all units sold would have been the default size. Portion sizes can also be adapted to promote sales volume and intake, and increasing the portion sizes of vegetables and fruit led to a greater intake among children^{(Reference Miller, Reicks and Redden204)}. Nonetheless, the effects of changes in portion sizes on food choices and consumption should be carefully examined. Authors warn of the possible backfire effects that (multiple) smaller portions would eventually lead to higher consumption, and that smaller portions (compared to larger ones) are less likely to activate a self-control conflict that leads to careful monitoring of consumption^{(Reference do Vale, Pieters and Zeelenberg205)}.

Portion size judgement is also heavily influenced by our perception. For example, the same portion size has been found to be perceived as being larger when presented horizontal (v. vertical) on a plate^{(Reference Rowley and Spence206)}. Moreover, consumers perceive food portions placed centrally on a plate (v. slightly offset) as larger. In addition to the way food is presented on a plate, plate size also affects perceptions, so that consumers have been demonstrated to eat more from larger plates^{(Reference Hollands, Shemilt and Marteau201)}. Nonetheless, other studies found no effect of plate size on consumption^{(Reference Rolls, Roe and Halverson207)}, and larger plate sizes might also lead to a greater intake of vegetables^{(Reference Libotte, Siegrist and Bucher208)}. Considering consumers' preference for middle options^{(Reference Valenzuela and Raghubir178)}, size labels might trigger people to always go for the in-between-option, regardless of its portion size. Moreover, people report to feel less guilty when consuming a larger portion of hedonic food if it is labelled as ‘small’^{(Reference Aydinoğlu and Krishna209)}. Nonetheless, the authors also found that consumers are not ‘fooled’ by smaller items carrying a ‘large’ label.