Can I still eat this? Using implicit and explicit measures to explore consumer behavior toward food products with date labels

This study investigates implicit and explicit attitudes toward products before and beyond the best-before date (BBD) using an Implicit Association Test and an online questionnaire. Moreover, we test whether consumer perception of and behavior toward products beyond the BBD can be manipulated using a priming task. We use a three-group between-subjects design where respondents had to recall either a frugal, a wasteful, or an unrelated behavior. Results show that consumers have negative implicit associations with products beyond the BBD. Reduced health and safety perceptions, consumers ’ strategies to determine edibility, and general risk perception of products beyond the BBD predict consumption of these products. While recalling a frugal behavior does not have significant effects, recalling a wasteful behavior prior to evaluating products beyond the BBD leads to a decrease in the perceived safety and healthfulness of these products.


Introduction
The UNEP (2021) estimates that around 17% of all food produced was wasted in 2019 (~931 million tons).Private households were found to be responsible for more than half (61%) of said food waste.Therefore, minimizing food waste at the household level remains highly relevant.One cause of (unnecessary) household food waste can be expiration date labeling (Newsome et al., 2014).Date labels (e.g., use-by and best-before dates (BBD)) are universally used by food manufacturers and are strictly regulated in the European Union.While use-by dates generally are concerned with food safety and products past this date should no longer be consumed, the BBD does not indicate food safety but is a mere quality index (European Commission, Directorate-General for Health and Food Safety, 2018) 1 Still, many consumers are not aware of this fact.In a recent study by Patra, Henley, Benefo, Pradhan, and Shirmohammadi (2022), the majority of respondents report being confused about date labeling and believe that products beyond the BBD are not safe to consume.This is confirmed by Kavanaugh and Quinlan (2020), who show that consumers tend to throw out products beyond the BBD, even if they appear to be still edible.Underlying reasons for this unnecessary waste can be concerns regarding product safety (Newsome et al., 2014) or freshness and healthfulness (Wansink & Wright, 2006).The European Commission, Directorate-General for Health and Food Safety (2018) estimates that 10% of the 88 million tons of annual food wasted in the EU can be traced back to date marking.Therefore, it considers revising EU rules on date marking as part of its Farm to Fork Policy (European Commission, 2020).Previous research from WRAP (2011) in the UK even suggests that the date label plays a role in 30% of disposal decisions.In view of food scarcity, inflation of food prices, and related timely issues, this unnecessary food waste is a big concern.Therefore, to further explore the reasons and mechanisms behind the behavior of throwing out food as soon as the date label has passed is highly relevant.

The role of implicit and explicit attitudes in consumer food waste behavior
Previous studies investigating consumer behavior toward food with date labels where the date has passed primarily focused on the effect of different date labeling (e.g., use-by/sell-by/best-by) on consumers' perception and behavior (e.g., Gong, Su, Zhang, Chen, & Wang, 2022;Wilson, Miao, & Weis, 2018), date label knowledge (e.g., Kavanaugh & Quinlan, 2020;Neff et al., 2019;Thompson, Toma, Barnes, & Revoredo-Giha, 2018), and safety and quality concerns as mediators of the relationship between date labels and willingness to consume food products where the labeled date has passed (e.g., Gong et al., 2022;Thompson et al., 2018).Other studies have used extended Theory of Planned Behavior models to explain consumer behavior toward such foods (e.g., Schmidt, 2019;Thompson, Toma, Barnes, & Revoredo-Giha, 2020).
However, the aforementioned studies tend to be concerned with conscious constructs, such as explicit attitudes.Still, previous research suggests that consumer decision-making is guided by unconscious processes (Dijksterhuis, Smith, van Baaren, & Wigboldus, 2005) and implicit attitudes, that is, unconscious associations or biases, which can contribute to explaining the attitude-behavior gap in ethical consumption choices (Govind, Singh, Garg, & D'Silva, 2019).While individuals are usually aware of their explicit attitudes and can self-report them in questionnaires, implicit attitudes are automatic evaluations without volitional control (Greenwald, McGhee, & Schwartz, 1998).A person can hold implicit and explicit attitudes toward the same object, which may also deviate from one another (Wilson, Lindsey, & Schooler, 2000).
Generally, in the field of suboptimal foods, 2 only a few studies consider implicit attitudes.One example is Bolos, Lagerkvist, and Kulesz (2019), who examined the effect of explicit and implicit attitudes on the purchase likelihood of suboptimal apples.They found that while explicit measures better predicted the purchase of apples, the implicit measure was superior in explaining the rejection of apples.Another recent example is provided by Niu, You, Ran, and Liu (2023), whoin several experimentsinvestigated the choice of suboptimal fruits and found that an incongruence between implicit and explicit beliefs regarding the relationship between unattractiveness and naturalness of produce can foster choice of imperfect fruits because it results in elaborative thinking.However, to the best of our knowledge, implicit attitudes remain underexplored when it comes to food items that are not suboptimal because of their visual appearance but because of being beyond the BBD.Therefore, our study combines implicit and explicit measures to investigate how unconscious and conscious processes determine behavior toward food items beyond the BBD.

The role of priming in consumer food waste behavior
Another question that, to the best of our knowledge, has received little attention so far is how to motivate consumers to re-evaluate date labels and consume products that might be beyond the BBD but still safe to eat.Previous research shows that sustainability priming can be a means to activate desired goals, which in turn can motivate consumers to behave in a more environmentally friendly way (e.g., Panzone, Ulph, Hilton, Gortemaker, & Tajudeen, 2021).In the context of sustainable consumption behavior, priming can, for instance, entail using an informational message to activate an environmental-protection goal, which was shown to affect subsequent pro-environmental consumption behavior (Tate, Stewart, & Daly, 2014).Generally, priming refers to the presentation of "some stimulus with the aim of activating a particular idea, category, or feeling and then measuring the effects of the prime on performance in some other task" (Cameron, Brown-Iannuzzi, & Payne, 2012, p. 330).The idea is that by activating an idea, category, or feeling, corresponding associations are activated, which may then translate into explicit attitudes and behavior.
While studies using priming in the context of food waste are still rare (Septianto, Kemper, & Northey, 2020), recent findings confirm the usefulness of priming when it comes to food waste reduction.For instance, Petit, Lunardo, and Rickard (2020) show that priming the food waste problem by providing an informational text decreases respondents' preference for larger package sizes, even if these are promoted.Another example is provided by Visser-Amundson (2022), who used priming as one of several nudging strategies to reduce food waste in the hospitality industry.In their field experiment, primes consisted of sad and happy smiley stickers and an informational poster in restaurant kitchens.Results show that priming (in combination with other nudges) was an effective tool to reduce food waste.Next to informational messages, another way to prime for sustainability can be through reminding people of their environmentally friendly behavior in the past.For instance, Panzone, Ulph, Zizzo, Hilton, and Clear (2021) show that "being required to recall past environmentally-friendly behaviour before shopping led consumers to purchase more sustainable food baskets" (p.1).Reminding people of their environmentally friendly behavior in the past is especially effective when the respective behavior strengthens environmental self-identity (Van der Werff, Steg, & Keizer, 2014).
Building on the previous research in the field, our study explores whether using a prime asking to recall past behavior can change consumers' perception of foods beyond the BBD and increase their willingness to consume them.Contrary to existing studies, we not only test whether recalling past pro-environmental behavior is effective.Next to recalling a sustainable past behavior (i.e., behaving frugally and not wasting), we also test the effect of recalling an unsustainable past behavior (i.e., behaving wastefully).

Objectives and conceptual framework
In summary, this study combines implicit and explicit attitudes toward food products before (from here on 'fresh' products) and beyond the BBD and tests whether consumers' health and safety perception of and behavior toward products beyond the BBD can be manipulated using a priming task.In addition, other factors possibly influencing consumer behavior toward products beyond the BBD will be examined.These include consumers' strategies to determine the edibility of food items (i.e., do they follow objectification strategies such as relying on external cues like date labels, or do they follow internalization strategies such as using their senses) (Neubig, Roosen, Karg, & Moser, 2022), consumers' perceived health risk (Visschers, Wickli, & Siegrist, 2016), and their food disgust sensitivity (Ammann, Hartmann, & Siegrist, 2018).As a use case, we choose dairy products (i.e., milk, yogurt, cream cheese, and curd) since most dairy products carry BBD in the investigated context, meaning they are expected to be safe to consume several days beyond that date.The study aims to answer the following three research questions: (1) How do unconscious and conscious processes determine behavior toward dairy products beyond the BBD? (2) How can a prime asking to recall past wasteful versus past frugal behavior change consumers' perception of dairy products beyond the BBD and increase their willingness to consume them?(3) What other factors contribute to predicting willingness to consume dairy products beyond the BBD?
Fig. 1 shows the conceptual model of the study.In line with previous research, we expect that priming past frugal behavior will increase willingness to consume (WTC) products beyond the BBD while priming wasteful behavior will decrease WTC.This effect is expected to be (partially) mediated by the explicit perception of products beyond the BBD regarding safety and healthfulness.Moreover, we expect implicit product associations to influence explicit perceptions of products beyond the BBD and WTC.We expect lowered explicit safety and health perception and lower WTC for consumers showing strong negative implicit associations toward products beyond the BBD.Finally, we expect that a tendency to use objectification of edibility strategies, higher food disgust sensitivity, and higher risk perception of dairy beyond the BBD 2 Suboptimal food refers to food that is often rejected by consumers because it deviates from the standard, for instance due to its visual appearance, but also products beyond the BBD are often considered to be suboptimal foods (Aschemann-Witzel, de Hooge, Amani, Bech-Larsen, & Oostindjer, 2015).
C.M. Neubig and J. Roosen lead to more negative explicit product perceptions and a lower WTC these products.In turn, a tendency to use internalization of edibility strategies is expected to lead to less negative perceptions and a higher WTC products beyond the BBD.

Procedure and sample
Data was collected in December 2021 in Germany through an online survey including an implicit association test (IAT) (Greenwald et al., 1998) and an experimental manipulation, i.e., a priming task, with a three-group between-subjects design.The study was approved by the ethics committee of the Technical University of Munich (approval number: 678/21 S).Respondents received information about the study objectives, tasks, data protection, and anonymity.They could withdraw from the survey at any time without giving a reason.All respondents provided informed consent before taking part in the study.
Respondents (N = 1135) were recruited via a market research firm and received financial compensation for participating.Inclusion criteria included being over 18 years old, living in Germany for at least five years, speaking German fluently, and being at least partly responsible for grocery shopping and preparing meals in their household.We also screened for consumption of the investigated dairy products.Respondents were required to consume milk and yogurt regularly (at least once a month).For cream cheese and curd, there were no exclusion criteria.Even so, only a minority of the sample stated never consuming cream cheese (2.2%) and curd (3.7%).The sample was quota-sampled to ensure representativeness for the German population regarding age, gender, and education.The minimum sample size was calculated using the number of free parameters to be estimated in the structural model.The number of parameters to be estimated in the specified model was equal to 77, 3 and a common rule of thumb states to use a minimum of 10 respondents for each parameter, making the minimum sample size for this study 770 respondents (Bentler & Chou, 1987).
Table 1 shows the socio-demographic characteristics of the sample in total and by treatment condition.The sample was 49% female, 50% male.Corresponding to the German population, older age categories were more represented.Most respondents had a university entrance qualification (35%) or an intermediate secondary education certificate (35%).Most respondents lived in a two-person household (46%).The demographic composition of the obtained groups was reasonably comparable (see Table 1).There were no statistically significant differences regarding gender, age, education, household size, and income between the three groups.

Questionnaire and experimental design
After agreeing to the informed consent and passing the screening criteria, respondents provided their socio-demographic information.Next, they completed the IAT to assess implicit associations toward dairy products beyond the BBD.After the IAT, they were asked questions regarding their determination of edibility, food disgust, and risk perception.Next, the priming task was carried out.Finally, respondents' explicit product perceptions and WTC dairy products beyond the BBD were assessed.Fig. A1 in the appendix provides an overview of the survey setup.

Implicit association test
In the IAT, respondents were asked to assign four types of stimuli to different category pairings as quickly as possible.The stimuli used were Notes.Additionally, we control for the time that has passed since the BBD, i.e., how the number of days a product is beyond the date label influences the explicit product perceptions, implicit product associations, and WTC these foods.(1) pictures of dairy products beyond the BBD, (2) pictures of fresh dairy products (date label not yet passed), (3) negative words (e.g., moldy, disgusting), (4) positive words (e.g., fresh, tasty).The word stimuli were adapted from Bolos et al. (2019) and translated into German (see Table 2).For the picture stimuli (see Fig. 2), we created pictures of four different dairy products (milk, yogurt, cream cheese, and curd) that only contained the following front-of-pack information: product name, graphical elements such as pictures of the content, and a BBD.Half of the products carried a date that had just passed (depending on when the respondents answered the survey invitation, the products had passed the indicated date between one and 12 days ago).In contrast, the other half of the products carried a date in the near future (depending on when the respondents answered the survey invitation, the products were between two and four weeks before their BBD).There were two sets of category pairings, one of which was expected to be congruent to respondents' implicit associations and the other to be incongruent.For the congruent condition, respondents had to sort stimuli into the categories 'dairy products beyond the BBD or negative' versus 'fresh dairy products or positive'.For the incongruent condition, respondents had to sort stimuli into the categories 'fresh dairy products or negative' versus 'dairy products beyond the BBD or positive'.If respondents associate dairy products beyond the BBD with negative words, this would be reflected in faster responses in the congruent condition.The IAT consisted of seven blocks.Three blocks were single sorting tasks (either picture or word stimuli) for practicing; the remaining four blocks that were considered for the analysis were double sorting tasks (combination of picture and word stimuli).We randomly assigned respondents to one of four permutations to reduce left-right biases and counterbalance the starting positions of the categories (Nosek, Greenwald, & Banaji, 2005).Table A1 in the appendix shows a schematic diagram of the IAT setup, including all permutations.

Determination of edibility and additional influencing factors
After completing the IAT, respondents answered questions referring to their strategies to determine the edibility of food items (adapted from Neubig et al., 2022).They were asked whether they generally follow objectification strategies, such as relying on date labels, or internalization strategies, such as using their senses (e.g., smelling, tasting).Moreover, respondents answered scales regarding their risk perception of dairy products beyond the BBD (adapted from Visschers et al., 2016) and food disgust sensitivity (adapted from Ammann, Hartmann, & Siegrist, 2018).An overview of the used scales is provided in the Appendix, Table A2.

Priming experiment
After answering these scales, respondents were randomly assigned to one of three treatments (frugal, wasteful, control), each receiving a different priming task.During the task, respondents were asked to recall and describe a specific situation in a text box in as detailed a way as possible.Respondents in the frugal condition were asked to recall and describe a situation in which they wanted to throw something away but then decided to use, recycle or donate it.Respondents in the wasteful condition were asked to recall and describe a situation in which they had thrown something away that was still good or could have been (re-)used.
Respondents in the control condition were asked to recall a typical (work) day and describe the first thing they do when they get up in the morning.Previous research has used similar tasks (e.g., Shin et al., 2021) to help consumers recall past behavior.

Explicit product perceptions and willingness to consume dairy products beyond the BBD
Subsequently, explicit product perceptions, i.e., health and safety perceptions, were measured by showing respondents the product pictures of milk and yogurt that they had also seen in the IAT and asking them to rate the fresh milk and yogurt 4 and the ones beyond the BBD regarding expected healthfulness and safety using 5-point Likert scales (from very unhealthy/unsafe to very healthy/safe).Lastly, WTC food beyond the BBD was assessed.Respondents were subsequently shown pictures of dairy products (milk, yogurt, cream cheese, curd) beyond the BBD and asked how likely they were to consume each of the four products using a 5-point Likert scale (from very unlikely to very likely).

Data analysis
Data was analyzed using IBM SPSS Statistics and AMOS 28.Results from the IAT were analyzed by calculating the D-score using the improved scoring algorithm by Greenwald, Nosek, and Banaji (2003).Accordingly, data from all four double-sorting tasks were considered, and trials with latencies smaller than 10,000 ms were eliminated. 5The D-score reflects the difference between the average response latencies in the incongruent minus the congruent condition in the IAT, divided by the pooled standard deviation of the associated trials. 6A D-score of zero would indicate that respondents were equally fast in assigning stimuli to categories in congruent and incongruent conditions and, therefore, were not implicitly biased in either direction.A positive D-score would indicate that respondents were faster in the congruent condition and, therefore, biased toward negative associations with dairy products beyond the BBD.A negative D-score would, in turn, indicate that respondents were faster in the incongruent condition and, therefore, biased toward positive associations with dairy products beyond the BBD.
For further analyses, the explicit product perceptions (i.e., health and safety perceptions) and WTC were not examined individually for each product category.Instead, the products were combined into one product category for the products beyond the BBD and one fresh product category due to high correlations of perceptions and WTC between the four investigated dairy products.We then performed dependent sample ttests to investigate differences between the explicit perception of dairy products beyond the BBD and fresh dairy products.
In a final step, drawing on our conceptual model (see Fig. 1), we performed confirmatory factor analysis (CFA) and structural equation modeling (SEM) with maximum likelihood estimation to investigate the interdependencies of implicit and explicit product attitudes, WTC dairy products beyond the BBD as well as the treatment effects and additional influencing factors.cream cheese, curd), for the explicit product perceptions (i.e., safety and health perceptions), we only asked respondents to evaluate milk and yogurt in order to reduce cognitive load and shorten the questionnaire.We expected milk and yoghurt to be the most relevant in terms of consumption frequencies and therefore chose to use these two products for a more in-depth evaluation. 5In addition, respondents with a latency below 300 ms in more than 10% of trials are to be deleted according to Greenwald et al. (2003); however this was not the case for any of the respondents in this study. 6Since four blocks were considered for the D-score calculation, two separate difference scores were created and then averaged to obtain the final D-score (see Greenwald et al., 2003).

Implicit attitude and explicit product perceptions
First, implicit attitudes, i.e., the results from the IAT, were analyzed.Results show that the D-score (M = 0.38; SD = 0.23) is positive and significantly different from zero, t (1134) = 54.954;p <. 001, indicating that on average, respondents were significantly faster in their responses when 'dairy products beyond the BBD' was paired with 'negative'.This suggests that respondents have a stronger implicit association between 'dairy products beyond the BBD' and 'negative' than 'positive'.
Moving to the explicit product perceptions, i.e., health and safety perception measures, results from dependent sample t-tests show that health and safety perceptions of fresh dairy products differ significantly from health and safety perceptions of dairy products beyond the BBD (see Table 3).A correlation analysis shows that health and safety perceptions of dairy products beyond the BBD were significantly correlated (r = 0.76, p < 0.001).Likewise, health and safety perceptions of fresh dairy products were significantly correlated (r = 0.63, p < 0.001).Therefore, safety and health perceptions were not only examined as separate constructs but also merged into the combined construct of Product Perception (once for fresh dairy products and once for dairy products beyond the BBD).

Structural equation model
Since the previous analyses showed that safety and health perceptions are highly correlated, safety and health perceptions remained combined for further analyses.However, in the SEM, a difference score between perceptions of fresh dairy products and perceptions of dairy products beyond the BBD was used to reflect the significant change in explicit product perceptions after the labeled date (see Table 3).Therefore, we subtracted the health and safety ratings of products beyond the BBD from the health and safety ratings of fresh products and used the resulting variables to combine into the latent construct 'reduction in safety and health perception after BBD passed' (RSHP).
Next, a CFA was carried out.The CFA included the latent constructs RSHP, objectification of edibility, internalization of edibility, disgust sensitivity, risk perception of dairy products beyond the BBD, and WTC dairy products beyond the BBD.Two items from the disgust scale were excluded due to low factor loadings (below 0.4).The final CFA model fulfilled the requirements for good model fit (RMSEA = 0.052; NFI = 0.940; TLI = 0.946; CFI = 0.954).
For the SEM, the observed variables of implicit associations (i.e., the D-score) and dummy variables for the priming treatments (with the control group as a reference) were added.Moreover, we included the time since the labeled date (i.e., the number of days the evaluated products were beyond the BBD at the time respondents answered the survey) as a control.We investigated the causal relationships proposed in our conceptual model (see Fig. 1).Table 4 shows the estimates, standard errors, standardized estimates, and significance of estimates.The final model fulfilled the requirements for good model fit (selected goodness-of-fit parameters are reported in the table notes of Table 4).
Results show that while the frugal priming did neither affect RSHP nor WTC dairy products beyond the BBD, the wasteful priming significantly increased RSHP but did not have a direct effect on WTC dairy   products beyond the BBD.WTC dairy products beyond the BBD was most strongly predicted by the explicit measure of RSHP.Moreover, using objectification of edibility strategies and having a higher risk perception of dairy beyond the BBD decreased WTC dairy products beyond the BBD, while using internalization strategies increased WTC dairy products beyond the BBD.Disgust sensitivity and implicit association had no direct effect on WTC but were mediated by RSHP.Stronger implicit associations between dairy products beyond the BBD and negative and higher disgust sensitivity led to a higher RSHP.In addition, objectification of edibility and risk perception of dairy beyond the BBD significantly increased RSHP.Finally, the time passed since the products' date label significantly decreased WTC dairy products beyond the BBD and increased the explicit measure of RSHP.However, interestingly, time passed since the labeled BBD did not influence the implicit measure.

Discussion
This study adds to the emerging literature on the role of implicit and explicit attitudes in consumer decision-making regarding suboptimal food products (e.g., Bolos et al., 2019;Niu et al., 2023) and priming in the field of sustainable food consumption behavior (e.g., Panzone, Ulph, Zizzo, et al., 2021).To the best of our knowledge, it is the first study to focus on implicit associations with food products beyond the BBD, which is a relevant category to consider when aiming to reduce avoidable food waste (European Commission, Directorate-General for Health and Food Safety, 2018).

The role of implicit attitudes and explicit product perceptions
Looking at the implicit measure, the results of this study confirm that consumers have negative associations with dairy products beyond the BBD, and they implicitly devalue them.This finding aligns with previous research on different types of suboptimal foods, i.e., visually suboptimal apples (Bolos et al., 2019).Looking at the explicit measure, we also find a significant difference in health and safety perception between fresh dairy products and products beyond the BBD, which aligns with previous research (e.g., Schmidt, 2019).While not surprising, this is potentially problematic since most dairy products are still edible many days after the BBD.Nevertheless, our results suggest that consumers perceive these products negatively (implicitly and explicitly) and, as a result, are less willing to consume them, leading to unnecessary food waste.
Interestingly, the time passed since the BBD does not influence reaction times in the IAT.This leads to the question of whether consumers think of BBD labels as a dichotomy (i.e., before versus beyond the BBD) or a continuum (i.e., one day past the date is less problematic than five days past the date).In the case of implicit associations, the IAT results point to the former.When it comes to explicit product perceptions, however, our results show that the RSHP is significantly influenced by the time passed since the BBD, as is the WTC products beyond the BBD: The longer the labeled date has passed, the less healthy and safe the product is perceived, and the less willing are consumers to eat or drink it.The fact that implicit associations were not influenced by the time passed since the labeled date could result from the IAT setup which forces consumers to think in dichotomies (since they have to sort products into two categories).However, more research is needed to investigate the role of the time that has passed since a product's date label in consumers' implicit associations with such food products.In addition, this study used product category-specific word stimuli in the IAT following Bolos et al. (2018).The size of the implicit association may depend on the choice of word stimuli.However, in other studies using the IAT on food items, word stimuli were also product categoryspecific (e.g., Raghunathan, Naylor, & Hoyer, 2006).
Considering the prediction of WTC foods beyond the BBD, our results show that the explicit measure, i.e., the RSHP, is the most influential factor.The implicit attitude, however, does not significantly influence WTC foods beyond the BBD.This is contrary to the findings of Bolos et al. (2019) and Niu et al. (2023) since these studies found that both explicit and implicit attitudes predicted the choice of suboptimal produce.One reason for the lack of direct effect of implicit attitude on WTC products beyond the BBD could be that the choice was hypothetical and might have been more cognitive and deliberate than a real-life consumption decision.Real-life consumption decisions are more likely influenced by unconscious processes such as implicit attitudes, which might explain why the implicit measure did not affect our dependent variable.However, in the present study, the implicit attitude plays a role in predicting RSHP, meaning that respondents with negative associations with dairy products beyond the BBD have a higher RSHP, i.e., explicit attitude.The relationship between implicit and explicit attitudes has received attention in previous literature already, with findings showing that explicit and implicit attitudes can be correlated (Hofmann, Gawronski, Gschwendner, Le, & Schmitt, 2005) or influence each other (Whitfield & Jordan, 2009).
Considering that explicit attitudes, i.e., safety and health perceptions, are the most relevant influencers in WTC, one crucial lever to reduce waste of foods beyond the BBD that are still safe to eat is to increase their perceived healthfulness and safety.Strategies could include using education campaigns or providing information about the meaning of date labels on the product packaging.Moreover, policies that change the meaning of date labels, for instance, to reflect safety instead of quality, might effectively reduce waste of still safe-to-eat food products since they would allow setting the date label to a later point in time.
In addition, changing implicit attitudes could play a part in increasing explicit attitudes, as suggested by the results of our study.Implicit associations, however, are more stable and more challenging to change (e.g., Govind et al., 2019).One possibility suggested by related research on suboptimal food (Bolos et al., 2019) is to constantly pair the negatively perceived productin this case, foods beyond the BBDwith positive stimuli in marketing campaigns.Moreover, providing more information about the BBD on the product packaging might also be beneficial since it can lead to more elaborative decision-making and reduce the influence of implicit attitudes.Future research should further examine potential measures to override the negative implicit associations toward food beyond the BBD.

The effect of priming past behavior
This study finds no effect of recalling previous pro-environmental behaviorin this case, acting frugal by saving something intended for discardon pro-environmental behavior in the presentin this case, WTC food beyond the BBD.This finding is opposed to that of Panzone, Ulph, Zizzo, et al. (2021), who found that recalling past pro-environmental behavior can influence current pro-environmental behavior.They argue that recalling previous behavior leads to consistency effects in the current behavior.It could be that respondents in the present study did not directly grasp a link between the recalled behavior and the WTC food beyond the BBD since they might have thought about something else in the recall task, e.g., something they threw out because of a broken package.In addition, it is unclear whether the memories recalled by the respondents in this study are positive or negative.For instance, someone might have recalled saving food beyond the BBD and feeling good about it.In contrast, someone else might have recalled the same instance but felt bad because consuming food beyond the BBD led to lower taste experiences or upset their stomach.Future research could, therefore, use different recall scenarios that are more targeted to the outcome measure.In addition, it is unclear how much respondents were engaged in the writing task.Some may only have thought about the task briefly and written a couple of words, while others may have thought about the task in detail before writing about it.Future research could investigate engagement levels in writing tasks and compare the effectiveness of writing task primes to other priming methods, such as informational primes.
In contrast, we find a significant effect of recalling past wasteful behavior.While recalling past wasteful behavior does not directly affect WTC foods beyond the BBD, it significantly decreases the safety and health perceptions of these products, which has a significant effect on WTC.This suggests that making consumers aware of previous waste situations or emphasizing that they have thrown away food previously is not advisable.This finding is thought-provoking, considering information campaigns that aim to reduce consumer food waste.Current campaigns tend to start by creating awareness around the food waste problem, for instance, by explaining to consumers that households create the most food waste along the supply chain and, therefore, trigger recall of previous food waste.The findings of the present study call for caution when using such campaigns.Critical reflection on how to communicate the issue of food waste to consumers is necessary to avoid potential backfiring effects.Future research should explore the functioning of backfiring effects in the food waste domain in more detail.For instance, focus could be given to the role of social norm activation in backfiring effects since providing information about the high occurrence of household food waste might lead to normalizing food waste behavior (Cialdini, 2003).

The influence of how consumers determine edibility
Further results of this study regarding consumers' strategies to determine the edibility of food items support that how consumers use date labels can influence WTC food beyond the BBD.We show that consumers who are less likely to focus on date labels (objectification) and those who increasingly trust their senses (internalization) are more willing to consume products beyond the BBD.This finding aligns with Neubig et al. (2022); however, they only find an effect of objectification of edibility strategies on the WTC aging produce.In the present study, we show that both the use of objectification strategies and internalization strategies affect WTC products beyond the BBD.While objectification of edibility reduces the safety and health perceptions of products beyond the BBD and the WTC them, internalization of edibility has no significant effect on safety and health perceptions of these products; however, it significantly increases the WTC them.
Therefore, the use of internalization strategies should be encouraged.To do so, consumers should not only be educated about how to use their senses to determine edibility, as done with "Look, Smell, Taste" pictograms on food products.Since we show that relying on date labels increases RSHP and can increase unnecessary food waste, providing information about how to reduce reliance on date labels and correctly interpret them is critical to rectifying the risk perception of dairy products beyond the BBD.

Conclusion
In conclusion, we show that consumers have unconscious negative associations with dairy products beyond the BBD.While the implicit associations lead to a significant RSHP, they do not predict WTC products beyond the BBD directly.Looking at the explicit measure, we find a significant difference in safety and health perceptions between fresh products and products beyond the BBD and that the RSHP is a direct predictor of WTC products beyond the BBD.Regarding the priming treatments, contrary to our expectations, recalling past frugal behavior does not have any effect in this study.However, recalling past wasteful behavior significantly reduces safety and health perceptions, thereby decreasing WTC dairy products beyond the BBD.This study calls for measures to improve the perceived safety and healthfulness of dairy products beyond the BBD, re-evaluate the meaning of date labels, and increase internalization of edibility strategies.Notes.* Reverse coded items.+ Two pictures were later excluded from the analysis due to low factor loadings.

Fig. 1 .
Fig. 1.Conceptual ModelNotes.Additionally, we control for the time that has passed since the BBD, i.e., how the number of days a product is beyond the date label influences the explicit product perceptions, implicit product associations, and WTC these foods.

Fig. 2 .
Fig. 2. Examples of Picture Stimuli used in the IAT Notes.The examples show the fresh version of the products.The same pictures but with a different date printed on were used as beyond the BBD versions of the products.

Table 1
Socio-demographic characteristics of the sample.

Table 2
Word Stimuli used in the IAT.

Table 3
Measures of explicit product perception.Means (M), standard deviations (SD) and significance levels of mean differences (p).

Table 4
Results from SEM.

Table A2
Overview of Survey Questions: Items, Scale, and Sources.When it comes to judging whether a food is still safe to consume, I stick to the stated best-before date.2.If the best-before date on a product has passed, I will no longer eat it.When it comes to judging whether a food is still safe to consume, I trust my own senses (sight, smell, taste).2.I think of the best-before date as a non-binding recommendation.believethat the risk of becoming ill as a result of eating expired dairy products is high.2.I am not worried that eating expired dairy products results in health damage.*3.I think that consuming expired dairy products is harmless.*4.I think that one can perfectly safely eat dairy products whose best-before dates expired a few days ago.*Respondents were asked the following question for eight + pictures: Closely look at this picture.Imagine, you were asked to consume this food item.Please indicate how disgusting you perceive this item to be.Please answer intuitively, there are no right or wrong answers.