Using misperceived social norms as a license: does pluralistic ignorance trigger complacency in the food environment?

ABSTRACT

The current food environment strongly communicates the normality of consuming unhealthy and unsustainable food products. However, it is unclear whether people truly support this unhealthy and unsustainable social norm, or that they follow the norm (reluctantly) because they believe that other people agree with it, a phenomenon that is generally known as pluralistic ignorance. While previous research has documented the existence of pluralistic ignorance in a variety of settings, it is unknown to what extent it directly influences behavior and which mechanism may account for this influence. The present study examines whether the perception that others seem to agree with unhealthy and unsustainable eating norms acts as a license to not change one’s eating behavior and leads to complacency. We assessed pluralistic ignorance by comparing self- and other-scores on the importance, frequency, normalcy, and intentions dimensions of consuming healthy and sustainable food in a large sample of Dutch participants (N = 415). To investigate the effect of pluralistic ignorance on self-licensing and complacency, we calculated healthy and sustainable ‘misperception scores’ per dimension. Healthy eating misperceptions only marginally predicted self-licensing, but healthy misperceived intentions did predict an increase in complacency. Sustainable eating misperceptions seem more influential because misperceptions on importance and frequency predicted an increase in self-licensing, and sustainable misperceived normalcy predicted a decrease in complacency and intentions predicted an increase in complacency. These findings suggest that pluralistic ignorance may be more influential in sustainable eating since people could be uncertain what appropriate sustainable food choices are. Prospects for future research and suggestions to address pluralistic ignorance to potentially increase healthy and sustainable food choices are discussed.

KEYWORDS:

• pluralistic ignorance
• social norms
• food choices
• food environment
• self-licensing, complacency

Introduction

Food choices are influenced by a multitude of factors, including intra-personal processes (e.g., intentions, habits, self-control; Will Crescioni et al., 2011), the food environment on a micro (e.g., people’s socio-economic position) and macro level (e.g., pricing of food; Lakerveld et al., 2018; Townshend & Lake, 2017), as well as the social context (Higgs, 2015). The social context communicates what other people in the food environment consume and consider appropriate to consume (De Ridder et al., 2013; Herman et al., 2003; Higgs, 2015). In public places, it is considered normal to consume unhealthy snacks or to purchase unhealthy food items, which may convey the appropriateness of unhealthy food consumption (Brownell, 2004). This development is particularly challenging for people holding the intention to eat more healthily and sustainably (Lakerveld et al., 2018; Swinburn et al., 2011; Townshend & Lake, 2017) since people base their food choices in part on the actions and preferences of others (Herman et al., 2003; Higgs, 2015). As the food environment on a micro- and macro-level degrades, with potential consequences for social norms about healthy and sustainable eating, people are more likely to make food choices that are not in line with their intentions or make choices they do not wholeheartedly support. Over time, these actions and perceived preferences of people may communicate a misperception that others do not want to consume healthily and sustainably like they do, and establish a misperceived social norm of unhealthy and unsustainable consumption, which we will refer to as undesirable consumption. Many people engage in the practice of undesirable consumption (Willett et al., 2019), but previous studies have revealed that a substantial number of people regard healthy and sustainable food as desirable (Van Loo et al., 2017), and also have intentions to increase desirable food consumption (De Ridder et al., 2017). From here on, we will use the term ‘desirable eating’ when referring to healthy and sustainable eating (Willett et al., 2019).

Social norms are implicit guidelines that consciously and subconsciously guide a person’s behavior, and are formed by the perceived behavior and attitudes of other people in groups they identify with (Bicchieri, 2005). There is a clear distinction between two types of social norms: descriptive norms (perceived behavior of other people) and injunctive norms (expected or endorsed by other people; Cialdini et al., 1990). By observing other people, social norms can help individuals by modeling their food choices after others (Higgs, 2015). For instance, when individuals see evidence that other people make the healthy choice over an unhealthy choice, they are more likely to make the healthy choice themselves (Prinsen et al., 2013). However, the Western food environment is an example where social norms can also be detrimental for people, as it communicates undesirable eating norms (Brownell, 2004). People with intentions for desirable eating may perceive others to agree with these undesirable social norms, even when they themselves do not necessarily endorse this norm. For instance, people who intend to purchase a desirable food item may abandon their intention and model their food choice in accordance with the perceived norm of undesirable consumption. When they see other people picking an undesirable food item, and especially if they identify themselves with the social group these other people belong to, they are more likely to follow this social norm (Bicchieri, 2005). Because of this undesirable example, people might misperceive that other people do not have desirable intentions like them and that they themselves differ from the social group. Since people do not want to differ from the group they identify with (Bicchieri, 2005), they may make a more undesirable choice than intended. This misperception is a social comparison error (Halbesleben & Buckley, 2004) and is known as ‘pluralistic ignorance’ (Miller & McFarland, 1991; Prentice & Miller, 1993; Toch & Klofas, 1984). It has been researched among a wide variety of contexts, including, for example, alcohol consumption, bullying behavior, drug use and environmental concerns (Sargent & Newman, 2021).

Pluralistic ignorance is a social psychological phenomenon that ‘occurs when a majority of individuals falsely assume that most of their peers behave or think differently from them when in fact their attitudes and/or behavior are similar’ (Berkowitz, 2004, p. 7). By definition, social norms are subjective impressions of what others do or should do. However, these subjective perceptions may vary in the extent that they are accurate (Sargent & Newman, 2021). Consequently, people adjust their behavior to be more in line with what they (mis)perceive the social norm to be. The very essence of pluralistic ignorance implies that people may overestimate undesired behavior and underestimate desired behavior performed by others, which may subsequently facilitate undesired behavior or inhibit desired behavior by themselves. Resulting in more undesired behavior and less desired behavior than people would exhibit without their flawed perceptions of what others do (Sargent & Newman, 2021).

Research on pluralistic ignorance in undesirable food consumption is scarce. For example, recent research shows that secondary and high school students reported consuming more undesirable foods when they perceived their peers to be more in favor of undesirable food consumption. In reality, however, their peers reported to be less in favor of undesirable consumption than the students perceived, indicating the presence of pluralistic ignorance (Calvert et al., 2021; Perkins et al., 2018). Previous research has also documented pluralistic ignorance in a neighborhood setting (Moojen et al., 2022). In contrast to previous pluralistic ignorance research, which mostly employs one or just a few items, Moojen et al. (2022) assessed pluralistic ignorance through four dimensions to better determine its scope in the food consumption context. Participants considered themselves to find desirable food more important and normal than others and to consume desirable food more often. However, they perceived other people to have higher intentions to increase desirable consumption than themselves (Moojen et al., 2022). Whereas these findings suggest that people indeed perceive others to hold unfavorable norms regarding eating, in line with the pluralistic ignorance phenomenon, they also suggest that participants may use perceived discrepancies on importance, frequency and normalcy as a license and be complacent and delay their own intentions to increase desirable food consumption. When people think that other do not engage in desirable food consumption, they may think that they do not need to do so themselves; presumably because they were already doing better than others according to their own perception (Sargent & Newman, 2021).

Self-licensing is a deliberate form of motivated reasoning, and has been defined as ‘the act of making excuses for one’s discrepant behavior before actual enactment, such that the prospective failure is made acceptable for oneself’ (De Witt Huberts et al., 2014, p. 121). Self-licensing can be used to resolve self-regulatory dilemmas as it enables someone to choose for immediate gratification without jeopardizing their long-term goal. Theoretically speaking, any reason can be used as a license to delay goal-consistent behavior. Previous research has documented that self-licensing encouraged more undesirable snacking in participants who were made to believe that they had to perform a boring task twice, as compared with participants who thought that they needed to do the task just once (De Witt Huberts et al., 2012). Furthermore, other studies show that people who encounter many food temptations are, as a result, more likely to give in to licensing behavior (Kivetz & Zheng, 2006; Okada, 2005). Despite abundant research examining how licensing may favor undesirable behavior, thus far only a few studies (Guarino et al., 1994; McFerran et al., 2010) have examined whether perceiving other people engaging in undesirable behavior more than oneself may act as a license to justify one’s own undesirable behavior. That is why, we investigate to what extent dimensions of pluralistic ignorance may serve as a license for engaging in undesirable behavior in the current study. In this case, the license to do so is the perceived undesirable eating behavior of other people in the food environment.

Next to examining the licensing properties of pluralistic ignorance, we will also examine whether thinking that others are doing worse than oneself is associated with complacency, defined as a psychological state in which people are satisfied with the status quo (Luciano et al., 2018). Complacent individuals are characterized by overestimating themselves and reporting high levels of self-satisfaction (Kawall, 2006). As a result, they are less likely to make efforts to change their behavior (Vancouver et al., 2008). In addition, the phenomenon of pluralistic ignorance is closely connected to over- or underestimation of the self in comparison to other group members people identify with (Geiger & Swim, 2016; Sandstrom et al., 2013). Those who compare themselves with people they do not feel very connected to, are less likely to conform to their perceived behavior or social norms (Geiger & Swim, 2016). Finally, there is some preliminary evidence that in the context of desirable eating, people may overestimate themselves regarding desirable eating behavior (Moojen et al., 2022).

Current study

The current study aims to investigate the effect of misperceptions on four dimensions of pluralistic ignorance that have been documented in previous research (Moojen et al., 2022): importance, frequency, normalcy and intentions on self-licensing and complacency. Observing a direct association of pluralistic ignorance with self-licensing and complacency may contribute to our understanding what the consequences of pluralistic ignorance can be in the food environment, and assist the development of interventions to alleviate pluralistic ignorance and increase desirable food consumption. Before examining the main question of how pluralistic ignorance affects licensing and complacency, we will first attempt to inspect the robustness of the pluralistic ignorance dimensions identified in a previous study (Moojen et al., 2022), and explore the relevance of connectedness with relevant comparison groups (i.e., ‘family & friends’ vs ‘neighbors’) that people likely encounter in the food environment.

Methods

Ethics statement and funding

This study was approved by the Ethics Review Board of the Faculty of Social and Behavioral Sciences and filed under number 22–0242. Data collection was supported by the Regio Deal Foodvalley (grant nr. 162135), a collaboration between the Dutch government and the Foodvalley region¹ to stimulate the transition to a more desirable food system for citizens.

Participants and design

For the current study, data were collected by a research panel (Flycatcher.eu). Panel members had given their permission prior to research participation. Data were collected online in the spring of 2022. Personal information on participants’ gender, age and education level was available from the panel. Education level was categorized according to the Dutch education system, comprising three levels: high, middle and low. Participants who were 18 years or older were randomly selected from the panel’s database and invited via e-mail. In total, 415 participants participated in the study with a median completion time of 4.37 min. Since people are influenced more by groups they closely identify with (Bicchieri, 2005), we will examine the effect of pluralistic ignorance on self-licensing and complacency within two different groups. The ‘close connection’ group compared themselves with their family and friends, whereas the ‘distant connection’ group compared themselves with people in their neighborhood. Before starting the questionnaire, participants were randomly assigned to the neighborhood condition (n = 203) or the ‘family & friends’ condition (n = 212). Next, participants filled out the informed consent and continued with a questionnaire that consisted of the following measures: dimensions of pluralistic ignorance, feeling of connectedness, self-licensing and complacency.

Measures

Connectedness – Feelings of connectedness with the comparison group (either neighborhood residents or family and friends) was used as a manipulation check, arguing that feelings of connectedness should be higher in the family and friends group. There are multiple comparison groups in the food environment, and these two groups are common modeling targets (Howland et al., 2012; Kaisari & Higgs, 2015; Salvy et al., 2007). The groups are expected to differ in feelings of connectedness (Moojen et al., 2022). Connectedness was assessed by using the ‘Inclusion of Community in the Self’ (ICS) scale (Mashek et al., 2007). The ICS consists of six pictures of two same-sized circles representing the self and the other group with varying levels of overlap. Participants were instructed to encircle the picture that best described their feelings of connectedness with their assigned comparison group.

Pluralistic ignorance measure – To assess pluralistic ignorance we used an adjusted scale developed in a previous study (Moojen et al., 2022). This scale assesses how people perceive themselves and others on four dimensions of pluralistic ignorance: the importance of desirable food, the frequency of consuming desirable food (reflecting descriptive norms), the normalcy of consuming desirable food (reflecting injunctive norms) and the intention to increase desirable food consumption, with others referring to neighborhood or ‘family and friends’, depending on condition. In a previous study (Moojen et al., 2022), desirable food was specified as specific food products (e.g., fruits & vegetables and meat), possibly creating a narrow scope of the concept of healthy and sustainable eating. To avoid this problem, we collapsed the specific food products into healthy and sustainable food concepts, which were explained to participants according to the guidelines of the Dutch Nutrition Centre.² Each of the four dimensions was assessed by four items: two items addressed healthy products and two sustainable; of the healthy items one addressed the self (e.g., I consider it normal to eat healthy products) and one their perceptions of the other (e.g., People in my neighborhood/My family and friends would like to eat more healthy products), likewise for the sustainable items. This resulted in a total of 16 items, all answered on a 7-point Likert-scale.

Operationalizing pluralistic ignorance for further analyses

In previous pluralistic ignorance research, researchers have employed different terms for the same concept, such as pluralistic ignorance score (Buzinski et al., 2018), norm misperception score (Sandstrom et al., 2013) or accuracy score (Sargent & Newman, 2021). In the present research, we will employ the term ‘misperception score’ by subtracting the participant’s perception of the other-score (the perceived norm) from the mean of the group’s self-score (the actual norm; cf. Buzinski et al., 2018; Sandstrom et al., 2013). We added seven to each score to avoid negative values and facilitate the interpretation of the ‘misperception scores’. As a result, a low ‘misperception score’ of lower than 7 indicated an underestimation of healthiness and sustainability of self on the respective dimensions; a ‘misperception score’ of 7 indicated an accurate estimation; a high ‘misperception score’ of higher than 7 indicated an overestimation of healthiness and sustainability of the self on the respective dimension.

Self-licensing – To assess if people license undesirable food consumption by comparing themselves with others, we constructed an adapted self-licensing scale by adjusting an existing self-licensing scale (Prinsen et al., 2019). Within the adapted scale, the license to eat undesirably was their perceived undesired behavior or attitudes of other people. The scale consisted of eight items on unhealthy and unsustainable food consumption (e.g., ‘I do not think unhealthy eating is a problem, because other people also eat unhealthy’), answered on 7-point Likert scales (does not apply to me at all – totally applies to me). The adapted self-licensing scale had a Cronbach’s Alpha of .70, indicating acceptable internal consistency.

Complacency – Due to the lack of a behavioral complacency scale, we constructed a new scale and based it on the Inertia scale by Polites and Karahanna (2012). Participants were asked whether they were complacent regarding changing their undesirable (unhealthy and unsustainable) food consumption with six items (e.g., ‘My normal diet is already healthy enough’) on 7-point Likert scales (fully disagree – fully agree). The complacency scale had a Cronbach’s Alpha of .71, indicating acceptable internal consistency.

All measures used in this study are provided in the supplementary materials.

Data analyses

To test the robustness of the pluralistic ignorance measure from the previous study (Moojen et al., 2022), we examined how participants viewed themselves and others regarding differences on the four dimensions of healthy and sustainable eating separately, using a repeated measure ANOVA analysis.

To assess pluralistic ignorance, we used paired t-tests to investigate the differences between perceptions of self-scores and their perceptions of others-scores on the pluralistic ignorance dimensions. A significantly lower or higher self-score than the other-score would mean a misperception on the corresponding dimension, indicating pluralistic ignorance.

For the manipulation check, we investigated differences between neighborhood vs ‘family and friends’ conditions by conducting a t-test of both connectedness scores. To inspect potential differences between conditions on the other-scores of the pluralistic ignorance dimensions, we conducted a MANOVA. Finally, we assessed levels of pluralistic ignorance among the two conditions.

For the main analysis, we investigated the effect of pluralistic ignorance by using ‘misperception scores’ as the independent variables on self-licensing and complacency (the dependent variables) with a series of multiple linear regression analyses. To encompass contextual factors of the food environment, we included comparison groups and socio-demographic information in the analyses. In total, four multiple linear regression analyses were conducted: two analyses examining the association of either healthy or sustainable ‘misperception scores’, comparison groups and socio-demographic information with self-licensing, and two examining the same associations with complacency. A significant relationship between ‘misperception scores’ and self-licensing and complacency would support our hypothesis that pluralistic ignorance may elicit licensing and complacency.

Results

Sample descriptives

The mean age of the participants (N = 415) was 50.89 (17.27 SD) years and 51.6% of the sample was male; 27.2% had a low educational level, 39.1% a middle educational level and 33.7% a high educational level.

Testing robustness of the pluralistic ignorance dimensions

To test the robustness of previous pluralistic ignorance results (Moojen et al., 2022), we first assessed the differences within the four dimensions of pluralistic ignorance. The repeated measures ANOVA, with Bonferroni correction for multiple testing, confirm previous findings: there were significant negative differences within the four dimensions in how participants perceived themselves and others on the healthy and sustainable items, showing that healthy intentions in particular negatively differed from the other dimensions, and sustainable frequency and intentions negatively differed from importance and normalcy. Full details of this analysis are provided in the supplementary materials.

Pluralistic ignorance was assessed by examining how participants compared their own perceptions with how they perceived other people’s behavior on the four dimensions with paired t-tests. The discrepancies between perceptions of oneself and of others on the dimension of importance, frequency and normalcy dimensions were similar to a previous study (Moojen et al., 2022). This indicates that people considered their own healthy and sustainable food choices to be more important and normal than those of others. They also indicated that they themselves consumed desirable food more frequently than others (all p’s < .05). However, the discrepancy between one’s own intentions to increase desirable consumption and the perception of other people’s intentions was different from results reported by Moojen et al. (2022). Rather than people perceiving others to have higher intentions to increase desirable consumption than they themselves had, participants indicated that they themselves had higher intentions to increase healthy t(414) = 8.78, p < .001, d = 0.43 and sustainable t(414) = 9.93, p < .001, d = 0.49, consumption than others (see Table 1; full results are reported in the supplementary materials). This indicates pluralistic ignorance across all four dimensions.

Table 1. Ratings of total own and other scores on importance, eating frequency, normalcy and intentions to eat more healthy and sustainable food.

	Healthy				Sustainable
	Self	Others	p	Cohen’s D	Self	Others	p	Cohen’s D
Importance	6.04 (.85 sd)	5.09 (1.07 sd)	***	0.81	5.16 (1.20 sd)	4.57 (1.15 sd)	***	0.52
Frequency	5.60 (1.03 sd)	4.76 (1.06 sd)	***	0.69	4.47 (1.15 sd)	4.16 (1.01 sd)	***	0.29
Normalcy	6.07 (.92 sd)	5.10 (1.15 sd)	***	0.81	5.01 (1.22 sd)	4.56 (1.10 sd)	***	0.37
Intentions	5.15 (1.49 sd)	4.61 (1.29 sd)	***	0.43	5.07 (1.43 sd)	4.47 (1.25 sd)	***	0.49

Note: N = 415, *p < .05, **p < .01, ***p < .001.

Manipulation check

Next, to assess whether the manipulation was successful, we examined whether participants who compared themselves with family and friends (N = 203) would report differently on connectedness, than those comparing themselves with neighborhood residents (N = 212). The neighborhood condition felt less connected (M = 2.65, SD = 1.30) to their respective group than the ‘family & friends’ condition (M = 4.15, SD = 1.23). This difference was statistically significant t(413) = 12.06, p < .001, d = 1.19, confirming our expectation that participants perceive a closer connection with family and friends. In addition, a multivariate analysis of variance using Wilks’ lambda showed a significant difference of the other-scores of the pluralistic ignorance dimensions depending on comparison group, Λ = 0.89, F(8, 406) = 6.45, p < .001. This indicates that people believe that family and friends perform more desirable eating behavior than people in their neighborhood. Separate univariate ANOVAs on the pluralistic ignorance dimensions revealed significant effects of condition on all dimensions, except healthy intentions, F(1, 413) = .88, p = .350 and sustainable intentions, F(1, 413) = .86, p = .354. This indicates that there was no difference in intentions between conditions depending on feeling connected with a particular comparison group. Considering these findings, we also examined the degree of pluralistic ignorance (the significant differences between self- and other-score) between the two conditions using paired t-tests (Table 2). These findings warrant taking comparison groups into account when examining the effects of pluralistic ignorance on desirable food choices.

Table 2. Ratings of own and neighborhood scores on importance, eating frequency, normalcy and intentions to eat more healthy and sustainable food.

	Healthy				Sustainable
	Self	Others	p	Cohen’s D	Self		Others		p	Cohen’s D
Importance	6.11 (.81 sd)	4.79 (1.09 sd)	***	1.13	5.23 (1.13 sd)		4.44 (1.06 sd)		***	0.65
Frequency	5.70 (1.05 sd)	4.48 (1.03 sd)	***	1.02	4.54 (1.16 sd)		4.04 (.92 sd)		***	0.44
Normalcy	6.11 (.90 sd)	4.78 (1.14 sd)	***	1.08	5.04 (1.19 sd)		4.40 (1.04 sd)		***	0.51
Intentions	5.13 (1.43 sd)	4.55 (1.18 sd)	***	0.44	5.08 (1.33 sd)		4.42 (1.13 sd)		***	0.51
Note: N = 203, *p < .05, **p < .01, ***p < .001.
Ratings of own and family & friends scores on importance, eating frequency, normalcy and intentions to eat more healthy and sustainable food
	Healthy					Sustainable
	Self	Others	p	Cohen’s D	Self			Others	p	Cohen’s D
Importance	5.98 (.88 sd)	5.38 (.97 sd)	***	0.56	5.09 (1.26 sd)			4.70 (1.22 sd)	***	0.40
Frequency	5.51 (1.01 sd)	5.03 (1.03 sd)	***	0.43	4.41 (1.13 sd)			4.27 (1.07 sd)		0.13
Normalcy	6.02 (.95 sd)	5.41 (1.09 sd)	***	0.58	4.97 (1.24 sd)			4.72 (1.14 sd)	***	0.43
Intentions	5.17 (1.54 sd)	4.67 (1.38 sd)	***	0.42	5.06 (1.52 sd)			4.53 (1.36 sd)	***	0.47

Note: N = 212, *p < .05, **p < .01, ***p < .001.

Main analyses

Relations between pluralistic ignorance, self-licensing and complacency

Before assessing the impact of pluralistic ignorance with the ‘misperception scores’ on self-licensing and complacency, we first examined the relationship between self-licensing and complacency. This correlation proved not to be significant (r = .03, p = .580), giving credit to our analytic approach to analyze effects of pluralistic ignorance on both dependent variables separately. A full overview of correlations between the ‘misperception scores’, self-licensing and complacency is presented in the supplementary materials.

Healthy and sustainable ‘misperception scores’

Overall ‘misperception scores’ per healthy and sustainable dimension are shown in Table 3. For all ‘misperception scores’, means were higher than 7, indicating that on all four dimensions of pluralistic ignorance, participants reported that they did better than others according to their own perception. For exploratory reasons, we investigated whether participants who overestimated themselves on the pluralistic ignorance dimensions would underestimate others more. The correlation analysis between self-score and the ‘misperception scores’ shows significant negative correlations. This implies that people who score high on the dimensions do not underestimate others more, but rather are more accurate or overestimate others more. An overview of correlations between ‘misperception scores’ and self-scores is shown in the supplementary materials.

Table 3. Range, means and standard deviations of the ‘misperception scores’.

Misperception scores	Min	Max	M	SD	Skewness
Importance Healthy	6.04	12.04	7.95	1.07	.68
Importance Sustainable	5.16	11.16	7.59	1.15	.53
Frequency Healthy	5.61	11.61	7.85	1.06	.24
Frequency Sustainable	4.48	10.48	7.32	1.01	.44
Normalcy Healthy	6.07	12.07	7.97	1.15	.57
Normalcy Sustainable	5.01	11.01	7.45	1.10	.41
Intentions Healthy	5.14	11.14	7.53	1.29	.26
Intentions Sustainable	5.05	11.05	7.57	1.25	.21

Note: N = 415.

Impact of pluralistic ignorance on self-licensing and complacency

To examine the effects of dimensions of pluralistic ignorance on self-licensing and complacency, we employed multiple linear regression models with the ‘misperception scores’, comparison groups and socio-demographic variables. Healthy and sustainable ‘misperception scores’ were entered in separate models. Self-licensing and complacency were entered separately as outcome variables. This resulted in four models that are shown in Tables 4, 5 , 6 and 7.

Table 4. Linear model of healthy ‘misperception scores’ on self-licensing.

			95% CI
Variable	Beta	SE	LL	UL	β	p
Condition	0.21	0.09	0.039	0.376	0.12	.016
Gender	−0.20	0.08	−0.360	−0.029	−0.11	.021
Age	−0.01	0.00	−0.016	−0.006	−0.23	<.001
Education	−0.18	0.06	−0.292	−0.075	−0.17	<.001
Importance	0.08	0.06	−0.027	0.191	0.10	.142
Frequency	0.11	0.06	0.005	0.223	0.14	.040
Normalcy	−0.03	0.06	−0.137	0.080	−0.04	.608
Intentions	0.01	0.04	−0.062	0.074	0.01	.859

Note: adjusted R² = .09. Condition = neighborhood compared to family & friends, gender = male vs female, education = low vs medium & high. N = 415.

Table 5. Linear model of sustainable ‘misperception scores’ on self-licensing.

			95% CI
Variable	Beta	SE	LL	UL	β	p
Condition	0.16	0.08	0.003	0.324	0.94	.046
Gender	−0.18	0.08	−0.351	−0.024	−0.11	.024
Age	−0.01	0.00	−0.015	−0.005	−0.20	<.001
Education	−0.18	0.06	−0.284	−0.068	−0.16	.001
Importance	0.13	0.05	0.029	0.224	0.17	.011
Frequency	0.13	0.05	0.022	0.227	0.14	.017
Normalcy	−0.03	0.05	−0.137	0.077	−0.04	.586
Intentions	−0.04	0.04	−0.114	0.036	−0.06	.304

Note: adjusted R2 = .12. Condition = neighborhood compared to family & friends, gender = male vs female, education = low vs medium & high. N = 415.

Table 6. Linear model of healthy ‘misperception scores’ on complacency.

			95% CI
Variable	Beta	SE	LL	UL	β	p
Condition	−0.06	0.09	−0.228	0.115	−0.03	.516
Gender	−0.22	0.09	−0.391	−0.054	−0.12	.010
Age	0.01	0.00	0.006	0.016	0.21	<.001
Education	−0.08	0.06	−0.193	−0.026	−0.07	.135
Importance	0.07	0.06	−0.040	0.183	0.08	.206
Frequency	−0.08	0.06	−0.191	0.030	−0.09	.155
Normalcy	−0.07	0.06	−0.179	0.041	−0.09	.217
Intentions	0.19	0.04	0.124	0.262	0.27	<.001

Note: adjusted R² = .17. Condition = neighborhood compared to family & friends, gender = male vs female, education = low vs medium & high. N = 415.

Table 7. Linear model of sustainable ‘misperception scores’ on complacency.

			95% CI
Variable	Beta	SE	LL	UL	β	p
Condition	−0.06	0.08	−0.223	0.103	−0.03	.469
Gender	−0.22	0.08	−0.381	−0.050	−0.12	.011
Age	0.01	0.00	0.004	0.015	0.18	<.001
Education	−0.07	0.06	−0.177	0.042	−0.06	.229
Importance	0.02	0.05	−0.075	0.123	0.03	.632
Frequency	−0.08	0.05	−0.182	0.026	−0.09	.143
Normalcy	−0.12	0.06	−0.230	−0.012	−0.15	.029
Intentions	0.25	0.04	0.171	0.323	0.34	<.001

Note: Adjusted R² = .18. Condition = neighborhood compared to family & friends, gender = male vs female, education = low vs medium & high. N = 415.

The model examining the effect of healthy eating misperceptions, comparison groups and socio-demographic information on self-licensing was significant (Table 4), F(8, 406) = 5.91, p < .001, R² = .11. Participants who perceived themselves to perform better than others in frequency of healthy eating significantly predicted an increase in self-licensing (B = 0.11, SE = .06, p = .040). The other healthy ‘misperception scores’ did not predict self-licensing among the participants. Condition was associated with an increase in self-licensing (B = 0.21, SE = .09, p = .016), indicating that comparing oneself with people to whom one feels less connected, is associated with more self-licensing. Gender, age and education level were associated with self-licensing as well. With female participants scoring lower on self-licensing (B = −0.20, SE = .08, p = .021) than male participants, older participants scoring lower on self-licensing (B = −0.01, SE = .00, p < .001) than younger participants, and higher educated participants scoring lower on self-licensing (B = −0.18, SE = .06, p < .001) than lower educated participants.

The model examining the effect of sustainable eating misperceptions, comparison groups and socio-demographic information on self-licensing was significant (Table 5), F(8, 406) = 7.01, p < .01, R² = .12. Participants who considered themselves to find sustainable eating more important as compared with other people exhibited higher licensing scores (B = 0.13, SE = .05, p = .011). Also, participants who considered themselves to engage in sustainable eating more frequently than others reported higher self-licensing scores (B = 0.13, SE = .05, p = .017). The other two sustainable ‘misperception scores’ were not associated with self-licensing. Condition was associated with self-licensing (B = 0.16, SE = .08, p = .046). This indicates that participants who compared themselves to others in their neighborhood scored higher on self-licensing than those comparing themselves with family and friends. Gender, age and education level were associated with self-licensing. With female participants scoring lower on self-licensing (B = −0.18, SE = .08, p = .024) than male participants, older participants scoring lower on self-licensing (B = −0.01, SE = .00, p < .001) than younger participants, and higher educated participants scoring lower on self-licensing (B = −0.18, SE = .06, p = .001) than lower educated participants.

The model examining the effect of healthy eating misperceptions, comparison groups and socio-demographic information on complacency was also significant (Table 6), F(8, 406) = 10.35, p < .01, R² = .17. Participants who considered themselves to have higher intentions to increase healthy food consumption than others reported higher scores on complacency (B = 0.19, SE = .04, p < .001). The other three ‘misperception scores’ were not associated with complacency. Gender was associated with an increase in complacency (B = −0.22, SE = .09, p = .010), which implies that female participants portray more complacency than male participants. Age was associated with a decrease in complacency (B = 0.01, SE = .00, p < .001), suggesting that older participants portray less complacency than younger participants.

The model examining sustainable eating misperceptions, comparison groups and socio-demographic information (Table 7) significantly predicted complacency, F(8, 406) = 11.45, p < .01, R² = .18. Participants who considered themselves to have higher intentions to increase sustainable food consumption in comparison to other people were associated with an increase in complacency (B = 0.26, SE = .04, p < .001). Participants who considered themselves to have found sustainable consumption more normal than others were associated with a decrease in complacency (B = −0.12, SE = .06, p = .029). The ‘misperception scores’ for importance and frequency were not associated with complacency. Gender was associated with an increase in complacency (B = −0.22, SE = .08, p = .011), indicating that female participants portray more complacency than male participants. Age was associated with a decrease in complacency (B = 0.01, SE = .00, p < .001), implying that older participants portray less complacency than younger participants.

Discussion

The aim of the present study was to investigate the effects of the social comparison error pluralistic ignorance regarding undesirable eating behavior on self-licensing and complacency. We hypothesized that participants may use their misperceptions of other people’s eating behavior as a license to not change their own behavior and as a reason to be complacent, implying an inhibiting effect of pluralistic ignorance (Sargent & Newman, 2021) on desirable eating behavior. Whereas previous research has documented a wide variety of licenses that may lead people to abandon their eating goals and procrastinate on their good intentions to change their eating behavior, thus far pluralistic ignorance has not been considered as a factor that may lead people astray. We provide preliminary evidence that comparing oneself with others, to the extent that someone thinks that they are doing better than others, may serve as a license and lead to complacency.

Our main results suggest that the effect of pluralistic ignorance regarding healthy eating seems to be less pronounced than pluralistic ignorance regarding sustainable eating. Misperceived frequency of healthy eating predicted self-licensing, whereas misperceived intentions to eat healthily predicted an increase in complacency. In contrast, misperceptions of importance and frequency of sustainable eating predicted an increase in self-licensing, and misperceptions of normalcy of sustainable eating predicted a decrease in complacency, and misperception of intentions to eat sustainably predicted an increase in complacency. These findings suggest that pluralistic ignorance effects may be larger for sustainable eating. The importance of healthy food consumption has been established for a long time, but the importance of sustainable consumption is a relatively new development and could possibly leave people in doubt about the appropriate course of action. Indeed, in case of doubt, people are more likely to observe others to assess what is normal or appropriate behavior (De Ridder et al., 2013; Higgs, 2015).

Interestingly, our findings suggest that different dimensions of pluralistic ignorance relate differently to self-licensing and complacency. Whereas misperceived importance, frequency and normalcy dimensions were positively related to self-licensing, the misperceived intention dimension was positively related with complacency. This potentially implies that people use misperceptions on the importance, frequency, and normality of behavior as input for motivated reasoning to abandon one’s goals. Misperceived information on intentions to change could be relevant when people do decide to take action themselves or stay complacent.

We examined robustness of pluralistic ignorance dimensions by using measures from a previous study (Moojen et al., 2022). We found similar results with participants overestimating themselves on the importance, frequency and normalcy dimensions (Moojen et al., 2022), except for the intentions dimension. In contrast to earlier findings, participants in the present study perceived themselves to have higher intentions to increase desirable consumption than others. It is difficult to explain these results, but it may be related to a change in the measures, which will be addressed in the limitations section.

In the current study, we also put an emphasis regarding investigating the differences in misperceptions between the neighborhood and family and friends settings. We opted for this approach since people are more inclined to follow a social norm that is supported by the social group they identify themselves with (Bicchieri, 2005), and there are numerous comparison groups in the food environment (Herman et al., 2003; Higgs, 2015). We found that misperceptions on importance, frequency and normalcy were greater among participants who compared themselves to others in their neighborhood, in comparison to participants who compared themselves with family and friends. This may be because people expect to be provided reliable eating behavior information from those they strongly identify with (Higgs, 2015), and thus possibly expect closely connected members to agree with their perceptions. For the intentions dimension, there was no difference in misperceptions between groups.

Similarly, as in Moojen et al. (2022), we found that participants considered desirable eating to be important, frequently consumed and normal to consume, and they believed other people also endorsed a similar desirable norm (albeit slightly less desirable than themselves). Still, there was a misperception present which predicted increases in self-licensing and complacency. Combined, these studies show initial evidence for the potential inhibiting effects of pluralistic ignorance on desirable food practices in the food environment. Therefore, it is important for future research to take the next step and address pluralistic ignorance to hinder self-licensing and complacency and possibly stimulate desirable food consumption.

Implications

Our results reveal preliminary indicators that an intervention addressing pluralistic ignorance, with a focus on sustainable eating, may be successful in reducing self-licensing and complacency and could help stimulate desirable food consumption. Further, the different relations between the pluralistic ignorance dimensions and self-licensing and complacency may suggest that if researchers want to decrease self-licensing or complacency within the food environment, it could be important to address the specific misperception dimensions in their intervention which are relevant for self-licensing or complacency. In addition, the lack of difference between groups among misperceptions in the intention dimension could suggest that researchers who wish to address misperceived importance, frequency or normalcy with an intervention, might possibly expect greater effects in the target group when they feel closely connected to the comparison group. When addressing misperceived intentions, researchers may expect similar outcomes in target groups regardless of feelings of connectedness to the comparison group. However, these implications are based only on initial findings, and additional research is required.

Next, previous pluralistic ignorance research has attempted to address the effects of pluralistic ignorance by correcting the social norm and giving people accurate information on the actual social norm (Buzinski et al., 2018). We recommend using a similar norm-correcting approach when addressing pluralistic ignorance in the food environment. This can be implemented in a nudge-based approach, via a social norm nudge, where the actual social norm is communicated in conjunction with making desirable food options easier to choose. For example, if there proves to be a misperception in the frequency of desirable food consumption, rather than a misperception of how important other people think it is, interventions should focus on the communication of what people actually eat and making the food option easier to choose, rather than stating that other people find desirable eating important. Finally, to measure the effect of these interventions, we also suggest adding a behavioral measure to measure the direct effect of pluralistic ignorance on eating behavior.

Strengths & limitations

Our study has several strengths and limitations. The strength of our study is the diverse sample of community residents with a variety of different education levels that were included. Another strength is the inclusion of a heterogeneous comparison group, namely the neighborhood group. In many previous pluralistic ignorance studies (e.g., Geiger & Swim, 2016; Halbesleben et al., 2007; Prentice & Miller, 1993; Sandstrom et al., 2013), participants had to compare themselves with homogenous comparison groups or salient in-groups. With this study, we aimed to investigate the presence of pluralistic ignorance in a broader context. We recommend continuing this effort by including more heterogeneous groups to determine the influence of pluralistic ignorance on food choices. One limitation of our study is the lack of a behavioral outcome variable, which should be added in future research. Furthermore, the correlational nature of the study design hinders the establishment of causality within our study. That is why we recommend implementing an experimental design in future research. Further, in a previous study (Moojen et al., 2022), specific food products were used and not specified as healthy or sustainable and thus had no potential element of social desirability. However, we have now opted to employ healthy and sustainable concepts to encompass more of the healthy and sustainable spectrum, instead of just a few specific products. This adjustment may explain why we found different results in the intentions dimension since participants could have been more susceptible to social desirable answers. However, as we hypothesized, the results show that misperceptions in intentions increased complacency. In addition, with the current concepts, we reduced the amount of items, and we also explicitly explained to the participants what was considered healthy and sustainable by experts.

Conclusion

The current study aimed to investigate the influence of pluralistic ignorance on self-licensing and complacency. We found that healthy eating misperceptions had little influence on self-licensing but were associated with an increase in complacency. Sustainable eating misperceptions were associated with both self-licensing and complacency. In addition, feelings of connectedness in comparison groups may influence factors which are related to healthy and sustainable food consumption. Finally, future interventions may consider utilizing the specific pluralistic ignorance dimensions to effectively alleviate the phenomenon. This study is another step toward understanding the effects of pluralistic ignorance in the food environment, and how it can be tackled in future interventions to increase healthy and sustainable food choices.

Author contributions

All authors contributed to the study design. RM oversaw the data-collection and carried out the data analysis. RM drafted the manuscript and DdR, and MG provided critical feedback. They collectively approved submitting the manuscript.

Ethical statement

The study was approved by the Ethics Review Board of the Faculty of Social and Behavioral Sciences and filed under number 22–0242.

Availability of data and materials

The research data and materials are publicly available on OSF via the following link: https://osf.io/kpxrh/?view_only=2c70d0c06ff24e4384b992bf2afc5564

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15534510.2023.2251643

Additional information

Funding

The data collection was funded by the Regio Deal Foodvalley (grant nr 162135). A collaboration between the Dutch government and the Foodvalley region to stimulate more healthy and sustainable food choices among Dutch citizens.

Notes

1. The Foodvalley region is a collaboration of several municipalities, educational institutions and local entrepreneurs within the Netherlands.

2. Healthy products were described as, ‘According to the food pyramid (Dutch: schijf van vijf): plenty of fruits and vegetables, mainly wholegrain products, less (red) meat, spreading and cooking fats and drinks (tea, tap water, coffee)’ Sustainable products were described as follows, ‘Mainly plant-based products and locally produced.’