Preference for imperfect produce: The influence of political ideology and openness to experience

Food waste is a critical global issue, posing substantial environmental, social, and financial challenges, even as millions continue to face uncertainty regarding their next meal source. The seriousness of this problem necessitates a closer look into ways to reduce food waste. In this investigation, we identify a key factor which contributes to food waste — imperfect fruits and vegetables (FaVs), defined as foods that are non-uniform in shape, color, or texture. Because of their unusual and sometimes ugly appearance, people are often averse to buying them, making it important to understand factors that influence people ’ s acceptance or rejection of imperfect FaVs. Across four online studies (including an exploratory study) with nearly 1400 U.S. respondents and using varied designs, we test political ideology as a moderator of this aversion to imperfect FaVs. Our findings indicate that politically conservative people are less likely to purchase imperfect FaVs (vs. perfect FaVs) compared to politically liberal people. The last study also uncovers the psychological mechanism underlying this greater aversion to FaVs by conservatives: lower openness to experience explains why this segment of population may be less willing to purchase imperfect FaVs. Implications and future research directions are


Introduction
Each year, close to 1 billion tons of edible food go to waste globally, with around 61% of this waste occurring at the household level (United Nations Environment Programme [UNEP], 2021).Meanwhile, worldwide, the number of people experiencing hunger and malnutrition reached 828 million (United Nations, 2022).In the United States, approximately 40% of all the food produced is wasted, even as close to 50 million Americans continue to reside in households that qualify as food insecure (Gundersen & Ziliak, 2015;Vines, 2022).The implications of these findings are simple, dramatic, and clearfood waste has non-trivial and negative impacts on the environment, society, and economy (Morone et al., 2019;UNEP, 2021).In response to this pressing issue, the U.N. Sustainable Development Goal (SDG 12.3) aims to reduce food waste by 50% at the retail and consumer level by 2030 (U.N. General Assembly, 2015).Reducing food waste not only enhances current and future food security but also mitigates adverse environmental and financial impacts.Considering the gravity of the problem, recommending ways to reduce food waste assumes great significance, and to do so, we first need to understand the factors that drive food waste.
In this investigation, we identify one such factorimperfect produce, defined as foods that are non-uniform in shape, color, or texture (De Hooge et al., 2017;Loebnitz et al., 2015).Food waste is often predicated on consumer rejection of ugly or imperfect produce (Helmert et al., 2017), and some estimates suggest that in the U.S., as much as 20% of fruits and vegetables are thrown away because of cosmetic imperfections (Berkenkamp, 2015).We examine how political ideology might influence consumers' acceptance/rejection of aesthetically imperfect fruits and vegetables (henceforth FaVs).One of the reasons why we invoke political ideology as our moderator is because of a link observed between food insecurity and political beliefs -as food insecurity rises worldwide (World Bank, 2023), so do political divisions (Silver, 2022).In addition, political ideology in and of itself exerts profound influence not only on how people interpret social events but also the choices they make in their everyday lives (Adaval & Wyer, 2022).Some view political ideology as motivated cognition and one of a constellation of system justifying beliefs with behavioral and neurological implications and associations (Jost & Amodio, 2012).Thus, besides its association with food insecurity, political ideology represents a ubiquitous facet of human existence with far-reaching psychological as well as policy implications.
We demonstrate that conservatives are less likely to purchase imperfect FaVs than liberals and that this effect is driven by the former's lower openness to new experiences.This link between political ideology and imperfect food consumption is the first to theoretically bridge the literatures on advancing the understanding of how political ideology can affect food choices.Practically, marketers might target the liberal segment or foster a mindset that embraces novel experiences to enhance the purchase of aesthetically imperfect food which is otherwise healthy, delicious, and of good quality.Stemming from this insight, our investigation offers one way to help advance the goal of minimizing food waste and to help the planet at large.

Imperfect food consumption
Imperfect produce refers to food that is non-uniform in shape, color, or texture, but is free of any damage that might pose health threat (De Hooge et al., 2017;Loebnitz et al., 2015).Consumers tend to devalue and are less inclined to purchase aesthetically imperfect, deformed, scarred, and/or discolored FaVs (Campos et al., 2022;Grewal et al., 2019;Xu et al., 2021).Recent reviews (e.g., Hartmann et al., 2021;Varese et al., 2023) suggest that the negative effect of food appearance abnormality on consumers' willingness to buy FaVs is consistent across countries in North America, Europe, and Asia.As a consequence, retailers are often hesitant to stock them leading to greater food waste as well as lower income for many producers.To help overcome this problem, some farmers have tried to sell imperfect FaVs on their own, and companies such as Misfits Market (MM; misfitsmarket.com) have appeared to shine a light on the problem of food waste related to imperfect FaVs.The content on the company's website explains this as follows: "Because we work closely with farmers and food makers, we know just how much food is going to waste every single day.And yet, there are millions of food-insecure households in the U.S." and "We take in high-quality organic produce that sometimes looks a little different.Fruits and veggies that are too big, too small, or just sort of funny-looking."The emergence of such ventures is a welcome occurrence; however, much work remains.In particular, MM appears to be attempting to overcome the resistance to imperfect FaVs by claiming that the produce it acquires is "high-quality" and "organic", hoping that these features may compensate for the negative consequences of visual imperfection.In more mundane settings like a regular grocery store, imperfect FaVs may not always be organic, and consumers may experience ambiguity about their quality.Yet another company (imperfectfoods.com;henceforth IF) which works directly with farmers in sourcing imperfect produce, claims that "Together, the Imperfect movement of farmers, suppliers, customers, and community partners has saved 172.5 million pounds of food from lesser outcomes" (bold in original) making a more direct food waste reduction pitch.While MM's utilization of compensatory product features and IF's appeal of minimizing food waste are two ways to help address the food waste issue, research unraveling consumer-related factors that can provide deeper insights into the phenomenon is nascent and scarce (see Lagerkvist et al., 2023 for an exception), and can help in more precise strategic targeting.
Resistance to imperfect FaVs appears to be driven by a variety of reasons.Consumers use visual cues to judge the risk perceived vis-à-vis the food they are considering consuming (Castagna et al., 2021) being more accepting of aesthetic imperfections related to color as opposed to shape or other physical aspects (Lagerkvist et al., 2023).In particular, the degree of imperfection of FaVs guides shoppers' perceptions of their quality (Lombart et al., 2019), the degree of ripeness influences purchase intentions (Symmank et al., 2018), and oddly shaped vegetables evoke perceptions of unnaturalness, lowering purchase intentions (Loebnitz & Grunert, 2018).Food is also intrinsically enmeshed with consumer identities (Durif-Bruckert, 2017) making it less appealing when (imperfect) food is inconsistent with one's identity.
Recent studies suggest that well-designed pricing and communication strategies can help lower buyer resistance to imperfect produce (Hartmann et al., 2021;Varese et al., 2023).For example, research in a French setting has found that identity focused interventions may be necessary to reduce tensions in identity related imperfect FaV consumption (Debucquet, Lombart & Labbe-Pinlon, 2021).Similarly, interventions targeting the "landed/rooted" consumer type may help increase imperfect FaV uptake (Debucquet & Lombart, 2017).In addition, discounted pricing to compensate for imperfections helps enhance quality perceptions and purchase intentions (De Hooge et al., 2017;Giménez et al., 2021;Helmert et al., 2017).Researchers have also found that communicating the quality of imperfect foods (Aschemann-Witzel et al., 2018), boosting consumers' self-perception through in-store messaging (Grewal et al., 2019), focusing on sustainability and authenticity (De Hooge et al., 2022;Louis & Lombart, 2018), labeling foods as 'ugly' (Mookerjee et al., 2021), and anthropomorphizing imperfect food and imbuing it with human qualities in advertising (Shao et al., 2020) can help promote sales of imperfect foods.These findings are encouraging but recent, and our understanding of processing and judgments associated with imperfect FaVs is consequently fairly impoverished.In this intellectual environment, our contribution lies in examining a consumer-related moderatorpolitical ideologyand its influence on the purchase intention of imperfect FaVs.Notably, we also explore the psychological process underlying this political ideology driven moderation.We propose that political ideology impacts consumers' openness to experience, which in turn drives the outcome variable of purchase intention for imperfect FaVs.

Political ideology, openness to experience, and imperfect FaVs decisions
Political ideology has been defined as a shared "set of beliefs about the proper order of society and how it can be achieved" (Erikson & Tedin, 2003, p. 64), being construed as enabling an interpretation of the social world, and a normative specification of "good and proper ways of addressing life's problems" (Jost et al., 2007, p. 307).Based on these lenses, there is widespread agreement, and empirical evidence bears it out, that a bipartite structure of political ideology as left-right (aka liberal-conservative) helps capture its complexities parsimoniously (e. g., Erikson & Tedin, 2003;Iyengar & Westwood, 2015;McClosky & Zaller, 1984;Rathbun, 2007).Comporting with this dichotomy, in an important investigation, Jost (2006) showed that from 1972 to 2004, where respondents identified themselves ideologically as left or right successfully predicted voting intention in American National Election Studies (see also Abramowitz & Saunders, 2008;Layman & Carsey, 2002).
That political ideology is consequential in many other ways than only in its ability to predict voting intention has been documented in innumerable settings.Conservatives and liberals differ in their worldviews, cognitive processing styles, and motivational concerns (Jost, 2017).Compared to liberals, conservatives tend to be more status quo oriented and more resistant to change, associate with greater rigidity, accord greater weight to maintaining societal norms, believe more in A. Aydinli et al. free will, and trust their own decisions more (Buechner et al., 2021;Fernandes et al., 2022;Jost, 2006;Ordabayeva & Fernandes, 2018).Conservatism is also linked to a greater desire to overcome fear and uncertainty and mitigate threats (Jost et al., 2003, p. 383).
In the consumer space, research has generally converged onto the viewpoint that political ideology strongly affects consumption and behavior (Adaval & Wyer, 2022;Jung & Mittal, 2021;Shavitt, 2017).In particular, Rogers and Jost (2022) find that liberals tend to have a more diverse set of consumption-related activities than conservatives.They also prefer global experiences (Buechner et al., 2021) and, more generally, have distinct processing styles (Jost & Krochik, 2014), motivations (Jost & Banaji, 1994), and brand preferences (Shepherd et al., 2015).Recent work has found that conservatives respond to micro-level stimuli like rounded surge pricing more negatively than liberals (Cui et al., 2022), show greater donation intent in response to COVID-19 messages featuring identifiable victims (van Esch et al., 2021), and increase their risk-taking as their perceived self-efficacy increases (Han et al., 2019).
In the food domain, political values shape people's perceptions of health risks associated with food (Boeuf, 2019) and influence their food choices and consumption (Lindeman & Väänänen, 2000).Compared to liberals, conservatives are more reluctant to eat unfamiliar, novel food (Guidetti et al., 2022), and show less interest in local food (Witzling & Shaw, 2019), organic food (Onyango et al., 2007), and plant-based meat (Yule & Cummings, 2023).Conservatives also tend to stigmatize vegetarianism (Rosenfeld & Tomiyama, 2019) and express a preference for an omnivore diet over a vegetarian diet (Milfont et al., 2021).Furthermore, conservatives and liberals respond differently when faced with the trade-off between overconsumption and food waste.When food portions are large and food waste becomes likely, compared to liberals who are more concerned about environmental consequences of waste and thus overconsume to prevent waste, conservatives are more concerned about personal responsibility and thus waste food to reduce overconsumption (Mas et al., 2022).
A robust aspect of the liberal-conservative belief typology is that liberals prefer social change, equality, progress, and flexibility, while conservatives anchor on inequality, tradition, and stability (e.g., Anderson & Singer, 2008;Jost, Banaji, & Nosek, 2004, 2008;Nosek et al., 2009).Jost et al. (2007, p. 990) summarize these differences eloquently as follows: "The idea is that there is an especially good fit between the need to reduce uncertainty and threat, on the one hand, and resistance to change and acceptance of inequality, on the other, insofar as preserving the [inegalitarian] status quo allows one to maintain what is familiar and known while rejecting the risky, uncertain prospect of social change" (see also Jost, Fitzsimons, & Kay, 2004, pp. 271-272).Congenial with this marked contrast between liberals and conservatives when it comes to preference for the status quo, openness to experience has been found to correlate negatively with conservatism quite robustly (Osborne & Sibley, 2020).Indeed, conservatives prefer familiar products, reject novel products, and favor national brands which presumably represent tradition and stability (Glasgow et al., 1985;Khan et al., 2013).Also, creativity, as represented by patents, is positively related to openness to experience but negatively related to conservatism (McCann, 2011).Further evidence of conservatives' preference for tradition over change in the context of food is offered by Wrenn (2017) and Hodson and Earle (2018), who found that conservatives are generally less open to vegan or vegetarian experiences.
Specific to imperfect produce, over time, consumers have come to expect produce to be aesthetically perfect (or at a minimum, not unusual), and this expectation has become the norm.A deviation from this norm (aka imperfect FaVs) represents a) change from the baseline expectation, and b) uncertainty regarding its more downstream associations -taste, quality, etc.To be accepting of such a normative deviation requires an openness which is more likely a liberal than a conservative response.This reasoning leads us to the following formal hypotheses.
Hypothesis 1.As compared to liberals, conservatives will disfavor imperfect FaVs more.
Hypothesis 2. This preference is driven by the lower openness to experience of conservatives.

Empirical overview
We conducted four online studies to examine the above hypotheses.In the main paper, we report three studies (Studies 1A, 1B, and 2), including one pre-registered study (Study 1B) while the fourth (exploratory) study is reported briefly in the "Key findings" section (Section 5.1).Details of each of the four studies (including materials, procedure, measures, analysis, and results) are reported in the Web Appendix.
For all studies, we recruited adult (18+) participants from the United States using an online data collection platform, Prolific, in exchange for a small monetary incentive.All hypotheses, sample sizes, and analytic plans were specified before data collection.Specifically, we determined the desired number of participants across all studies before data collection, targeting 80 or 100 per condition, which exceeds the recommendation of Simmons et al. (2011) for an appropriate sample size to provide adequate statistical power.Furthermore, we conducted a post-hoc power analysis with G*Power (Faul et al., 2007), which is reported for each study.We also ensured that participants paid close attention by using attention check questions, and participants who failed attention checks were excluded from data analyses.Across all studies, we tested our hypotheses using validated scales from prior literature and adapted the stimuli from previous studies that specifically examined imperfect produce.
We utilized both within-subject and between-subject designs to assess participants' purchase likelihood of perfect versus imperfect FaVs.Study 1A featured a within-subject design where participants were exposed to both perfect and imperfect FaVs simultaneously, while in the other three studies (Studies 1B, 2, and Exploratory Study), participants either saw perfect FaVs or imperfect FaVs, depending on the conditions they were assigned to (between-subject design).In addition, we used the single-item measure to tap into respondents' political orientation (Jost, 2006;Studies 1A, 1B, 2, and Exploratory Study), as well as the multi-item scale to assess participants' political ideology (Kidwell et al., 2013;Study 1A).Moreover, in Study 2, we examined openness to new experiences as an underlying psychological process and perceived risk as an alternative explanation for political value congruence (Boeuf, 2019).We also controlled for price perception of the presented food box, participants' price consciousness, and purchase frequency of imperfect FaVs.
We tested Hypothesis 1 (H1) and Hypothesis 2 (H2) in four studies.H1, the foundational hypothesis, was examined in Study 1A and replicated in subsequent studies.We tested H2, pertaining to the mechanism in Study 2. We find consistent results across studies by utilizing different study designs and operationalizations of the focal measures.Specifically, the results show that conservatives disfavor imperfect FaVs to a greater extent than liberals.Furthermore, we find that differences in openness to experience between conservatives and liberals underlie the effect of political ideology on preference for imperfect versus perfect FaVs.Fig. 1 provides an overview of the conceptual model and the operationalization of the focal constructs.

Participants and Design. A total of 160 participants from Prolific
A. Aydinli et al. responded to our probes, in exchange for a small monetary incentive 1 in a mixed design, in which type of food (perfect-FaVs vs. imperfect-FaVs) served as a within-subject factor and individual political ideology was measured.Using a post-hoc power analysis with G*Power (Faul et al., 2007), our current design had an 88% chance of detecting a medium effect size of 0.25.All participants passed the attention check and were included in the analysis (M age = 37.27, Age range = 19 to 75; 61.9% female).
Procedure.Participants were shown two boxes of FaVs sold by a trusted online grocer that met USDA safety standards and indicated their preference.The stimuli (including images, text descriptions, and prices) of aesthetically unattractive and attractive FaVs were adopted from Mookerjee et al. (2021), who examined the impact of "ugly" labeling on the purchase of unattractive produce.Specifically, Box 1 featured relatively perfect oranges, apples, carrots, and cucumbers for a special price of $20 (regular price: $35) while Box 2 carried imperfect versions of the same produce for a special price of $15 (regular price: $25) (see Fig. 2).To represent the marketplace reality, where imperfect produce is typically priced lower than its regular counterparts in grocery stores (such as Trader Joe's; Alex, 2018), the price of Box 2 was lower than that of Box 1.As such, following Mookerjee et al. (2021) we maintained different prices for perfect and imperfect FaVs and kept the discount rate comparable across the two boxes.Moreover, by holding the USDA prime constant across conditions, we presumably minimized the negative quality association with imperfect produce to some extent.In concert with the price differential favoring the imperfect box of produce, the design served to provide a more stringent test of the hypothesis.
In order to maintain realism, we used the terminology "imperfect" to describe the box of imperfect FaVs in this study (as well as in subsequent studies).This choice of terminology is consistent with messaging used by companies like Misfit Markets and Imperfect Foods, which typically use similar verbiage to promote fruits and vegetables that are not perfect.Participants indicated which box they would purchase, on a 5-point scale ("Which box would you purchase?";anchors: 1 = "definitely Box 1" to 5 = "definitely Box 2," M = 2.55, SD = 1.53), with a higher score indicating a greater preference for imperfect FaVs (key dependent variable).
Participants then responded to two measures of political ideology.The single-item measure tapped into respondents' political orientation on a 7-point scale ("Please locate yourself on the following scale of political orientation"; anchors: 1 = "extremely liberal"; 7 = "extremely conservative"; M = 3.79; SD = 2.05; Jost, 2006).The multi-item scale had participants indicate their opinions (1 = "strongly against" to 7 = "strongly favor," M = 3.65, SD = 1.83) on seven issues, including capital punishment, abortion*, gun control*, socialized healthcare*, same-sex marriage*, illegal immigration*, and democrats*, where * indicates reverse-coding (α = 0.92; Kidwell et al., 2013).Responses were averaged and scored so that a higher score represented closer alignment with a conservative ideology.The two measures of political ideology were randomly presented and were correlated highly (r = 0.89).
Next, participants responded to demographic probes (age, gender, income).To ensure that participants were taking the survey seriously, we inserted an attention check question in the questionnaire.Participants were instructed to drag a slider anchored by 0 and 100 to the number 47 ("This question is to check whether you read the instructions carefully.Please drag the slider to the number 47").Any participant who made a mistake was excluded from the analysis (Oppenheimer et al., 2009; please see Appendix A for detailed measures).

Data analysis
We performed a regression analysis using SPSS® to test Hypothesis 1-whether political ideology predicted preference for imperfect (vs.perfect) FaVs, and considered p < .05 as statistically significant.

Results and discussion
A regression of relative preference for imperfect FaVs on the singleitem measure of political ideology indicated a significant negative effect of political ideology (b = − 0.19, S.E.= 0.06, t (158) = − 3.34, p = .001),confirming that conservatives were less likely to purchase the imperfect FaVs over perfect FaVs than were liberals.The results were identical when the multi-item political ideology scale was used as the predictor (b = − 0.19, S.E.= 0.07, t (158) = − 2.96, p = .004).This effect of political ideology remained significant even when age, gender, and income were added as covariates to the analyses.Overall, Study 1A provides initial support for the hypothesized link between political ideology and preference for imperfect FaVs, supporting H1.

Study 1B
Study 1B (pre-registered: https://aspredicted.org/2CN_6QT) sought to replicate the observed effect of political ideology using a betweensubjects manipulation of the type of FaVs (imperfect vs. perfect).

Method: participants, design, and procedure
Participants and Design.Four hundred respondents from Prolific participated in the study in exchange for monetary compensation. 2Using a post-hoc power analysis with G*Power (Faul et al., 2007), our sample size was sufficiently powered to detect a medium effect (0.25) 99% of the time.Twelve participants who failed the attention check were eliminated (with no change in the results), leaving a sample of 388 (M age = 39.24;Age range = 18 to 82; 50.8% female).The study employed a mixed design, where political ideology was measured and type of food (perfect-FaVs vs. imperfect-FaVs) was between-subjects.
Procedure.As in Study 1A, participants were instructed to imagine that they were going to purchase a box of fruits & vegetables sold by a trusted online grocer whose products met USDA safety standards, and that they came across a box of fruits & vegetables that were sold with a special deal.We used a different set of aesthetically unattractive and attractive FaVs for generalizability.In the perfect-FaVs condition, the box featured relatively perfect tomatoes, strawberries, cherries, kiwis, cucumbers, potatoes, carrots, paprika, eggplants, and apples.In the imperfect-FaVs condition, the box featured the imperfect versions of the same produce.To eliminate concerns about different price levels across the two food boxes driving results in Study 1A, in Study 1B, the price was kept constant across conditions, with both boxes offered at a reduced price of $25 (regular price: $40) (see Fig. 3).
After seeing the food box, participants indicated their purchase likelihood using the probe: "How likely are you to purchase this box of fruits & veggies?" (1 = "not at all likely" to 7 = "very likely") and responded to the same single-item political ideology scale ("Please locate yourself on the following scale of political orientation"; anchors: 1 = "extremely liberal" to 7 = "extremely conservative"; Jost, 2006; M = 3.79, SD = 2.03) and demographic questions (age, gender, and income) as in Study 1A.
We used two attention checks for this study.As the first attention check, participants were instructed to drag a 0-100 slider to the number 47.In addition, participants were presented with both the perfect and imperfect boxes of fruits and veggies and asked to recall the box that they had seen earlier in the study (Which fruits and veggies box did you see earlier in this study?Please select what you saw; Box 1/Box 2).Any participant who failed the slider attention check or incorrectly recalled the box they had seen was excluded from the analysis (Oppenheimer et al., 2009; please see Appendix B for detailed measures).

Data analysis
We ran PROCESS Model 1 (Hayes, 2017) in SPSS® to test the interactive effect of political ideology (continuous variable) and food type (1 = perfect, 2 = imperfect) on purchase likelihood.We also performed a spotlight analysis (Spiller, Fitzsimons, Lynch, & Mcclelland, 2013) to provide an estimate and statistical test of the simple effect of type of food at different values of the continuous measure of political ideology.We mean-centered political ideology and considered 95% CI and p < .05 as statistically significant in the analyses.

Study 2
Study 2 tested the proposed psychological process -whether conservatives had a weaker preference for imperfect FaVs than perfect FaVs because they are less open to new experiences.We also explored perceived risk as an alternative mechanism based on the finding that both, liberals and conservatives perceive risk associated with food uniquely when it is pitched as being consistent or inconsistent with their values (Boeuf, 2019).

Method: participants, design, and procedure
Participants and Design.We recruited 400 respondents from Prolific who participated in this study in exchange for monetary compensation. 3A post-hoc power analysis using G*Power (Faul et al., 2007) indicated that this study was sufficiently powered to detect a medium effect (f2 = 3 Because the Prolific respondent pool is skewed towards Liberals, we recruited 200 Republicans and 200 Democrats in order to achieve a balanced sample. A. Aydinli et al. 0.25) 99% of the times.Eight participants failed the attention check and were excluded from the analysis (with no change in results), resulting in a final sample of 392 (M age = 39.4;Age range = 18 to 79; 51.0% female).The study design was the same as that of Study 1B-political ideology (measured) x 2 type of food (perfect-FaVs vs. imperfect-FaVs; between subjects).
Procedure.We used the same set of perfect and imperfect produce as in Study 1B (see Fig. 3).Participants were instructed to imagine buying a box of fruits and vegetables from a reputable, USDA-certified online grocer, and that a box of fruit and vegetables currently was offered on a special deal.Respondents next viewed either a box of relatively perfect fruit and vegetables (the perfect-FaVs condition) or a box of imperfect versions of the same produce (the imperfect-FaVs condition), the prices of which were kept constant (reduced price of $25).After seeing the food box, participants reported their purchase likelihood ("How likely are you to purchase this box of fruits & veggies?"; anchors: 1 = "not at all likely" to 7 = "very likely").
Participants then responded to the same single-item political ideology scale as in Study 1B ("Please locate yourself on the following scale of political orientation," Jost, 2006, 1 = "extremely liberal" to 7 = "extremely conservative," M = 3.72, SD = 2.06).We examined the proposed underlying mechanism of openness to experience by using a well-established scale tapping into respondents' self-perceptions (John & Srivastava, 1999).In particular, participants were asked to indicate how each of ten characteristics may or may not apply to them (sample items: "inventive," "curious about many different things"; 1 = "disagree strongly" to 5 = "agree strongly," M = 3.59, SD = 0.72, α = 0.86; see Appendix C).We also explored perceived risk associated with imperfect FaVs as an alternative account for our effect (Castagna et al., 2021), considering the possibility that conservatives perceive greater risk associated with imperfect FaVs, leading to their lower likelihood of purchase.To test this account, participants responded to the query "How would you rate the box of fruits & veggies you have seen on the following dimensions?" using three items: risk, safety, and harm (1 = "very low" to 7 = "very high"; Castagna et al., 2021; M = 2.47, SD = 1.16, α = 0.60).
Next, as control variables, participants reported price perception of the presented food box ("In your opinion, the box of fruits & veggies you have seen is:"; 1 = "very cheap" to 7 = "very expensive," M = 3.64, SD = 1.27), their price consciousness using a five-item scale (Lichtenstein et al., 1993; sample item: "I will shop at more than one store to take advantage of low prices; " 1 = "strongly disagree" to 7 = "strongly agree, " M = 5.34, SD = 1.32, α = 0.88), and purchase frequency ("How often do you buy imperfect produce?") of imperfect FaVs (defined as "Imperfect produce is that which has a significant natural aesthetic deviation in shape and/or color from prototypical produce, but has no damage or disease that could affect safety, taste, or nutrition"; 1 = "never," 2 = "occasionally," 3 = "once a month," 4 = "two to three times a month," 5 = "at least once a week"; M = 1.95,SD = 0.86).These covariates were included to examine the possibility that the advertised price, respondents' general price sensitivity, and their purchase pattern for imperfect FaVs may influence the predicted findings.Finally,  A. Aydinli et al. participants reported their demographics (age, gender, and income) and responded to the same two attention checks as Study 1B.Please see Appendix C for detailed measures.

Data analysis
To test Hypothesis 1, we employed PROCESS Model 1 (Hayes, 2017) with SPSS to test the interactive effect of political ideology (continuous variable) and food type (1 = perfect, 2 = imperfect) on purchase likelihood.We used PROCESS Model 8 (Hayes, 2017) to test the moderated mediation effect (Hypothesis 2), with political ideology as the predictor, food type as the moderator, openness to experience as the mediator, and purchase likelihood as the dependent variable.The continuous variable was mean-centered and 95% CI and p < .05were relied on in the analyses.
Effects of control variables.Purchase likelihood was significantly correlated with price sensitivity (r = 0.12, p = .01),price perception (r = − 0.34, p < .01),and purchase frequency of imperfect food (r = 0.25, p < .01).To further examine the influence of these covariates, we ran another regression analysis which showed that the predicted interaction effect between political ideology and food type remained significant when controlling for price perception, price consciousness, and purchase frequency of imperfect produce (β = − 0.21, S.E.= 0.09, t (385) = − 2.45, p = .01).

Discussion
Consistent with Studies 1A and 1B, the results of Study 2 supported Hypothesis 1 by showing that conservatives exhibited lower relative purchase likelihood for imperfect FaVs than liberals.The effect remained significant after controlling for individual differences in price perception, price sensitivity, and purchase frequency of imperfect FaVs.Importantly, Study 2 provided evidence supporting Hypothesis 2: lower openness to experience explains why conservatives are less willing to purchase imperfect FaVs.Finally, perceived risk was ruled out as an alternative mediator.

Key findings
This investigation demonstrates that conservatives are less likely to purchase imperfect FaVs than liberals, and this effect is driven by conservatives' less openness to new experiences.In Studies 1A and 1B, we established the basic effect and in Study 2, we garnered process evidence showing that openness to experience drives the effect of political ideology on the purchase likelihood for imperfect FaVs.In the same study, we also ruled out perceived risk as an alternate process account.There may yet be other processes at play and that are worth investigating.For instance, conservatives and liberals vary in their cognitive flexibility (Angle et al., 2017), which may influence their preference for imperfect FaVs.In addition, political ideology may be associated with perfectionism, explaining why conservatives and liberals exhibit differential attitudes toward imperfect FaVs.To test these accounts, we ran an exploratory study using the same design as in Studies 1B and 2 (see Web Appendix: Exploratory Study).A departure from the earlier studies was that we also asked respondents to indicate their political party affiliation as Democrats or Republicans while measuring their cognitive flexibility (Haran et al., 2013) and trait perfectionism (Feher et al., 2020).Overall, findings comported with the earlier studies.Republicans reported lower A. Aydinli et al. purchase likelihood for imperfect FaVs than perfect FaVs while Democrats reported comparable purchase likelihood.The interaction effect was marginally significant, driven largely by Republicans showing differential purchase likelihood for imperfect vs. perfect FaVs.Although Democrats had higher cognitive flexibility than Republicans, political ideology was not associated with perfectionism.Furthermore, none of the indices of moderated mediation (PROCESS Model 8) for the two alternate processes was significant.

Contribution
Our findings have significant implications for theory and practice.With respect to the literature, we contribute to the burgeoning research on the influence of political ideology on food choices.Prior research showed that compared to liberals, conservatives exhibit higher reluctance to try novel foods (Guidetti et al., 2022), and are less interested in local food (Witzling & Shaw, 2019), organic food (Onyango et al., 2007), and plant-based meat (Yule & Cummings, 2023) and vegetarian options (Milfont et al., 2021).Extending this line of research, to the best of our knowledge, we provide the first empirical evidence regarding the impact of political affiliation on preference for imperfect produce.Notably, our research uncovers how conservatives' lower openness to new experiences explains their aversion to FaVs that are abnormal in shape and color.
In addition, we contribute to the nascent literature on imperfect foods by identifying a consumer-related factor that influences the likelihood of purchasing such products.Prior studies have focused on how the visual imperfections of FaVs influence consumers' inferences related to risk, quality, and naturalness, subsequently explaining their purchase intentions (Castagna et al., 2021;Loebnitz & Grunert, 2018;Lombart et al., 2019).Our research identifies a consumer characteristic that serves as an indicator of a higher or lower preference for imperfect produce and provides insights into the underlying mechanism.
Also, our study underscores the idea that consumption of imperfect foods may be related to a specific mindset.In particular, though we do not focus on this, it may be the case that part of the reason that conservatives do not want to buy imperfect FaVs is that they perceive imperfect FaVs as something that the "other," liberals, in this case, does.As such, the polarization of the two groups contributes to a singularization of each group within itself such that each group views the other as wholly distinct.Thus, the consumption, or not, of imperfect FaVs may become relevant to promulgating that mindset and identity.Future research should further investigate this.
Our findings are also relevant for retailers and food marketers.Sellers of imperfect foods may use this information to identify relevant customer segments and more specifically target their products to those people who are more likely to buy them thus avoiding wasting potential marketing dollars on consumers who are less likely to purchase them.In addition, as openness to experience is the mechanism through which purchase behavior is influenced, marketers as well as N.G.O.s looking to minimize food waste may consider ways to enhance buyers' openness to experience.

Limitations and future research directions
Our investigation uncovers the effect of political ideology on preference for imperfect producea finding that can nudge food producers and marketers to reduce food waste.However, much work remains to be done towards accomplishing this important goal and this research can be foundational for future inquiries.First, a key implication is that imperfect produce may face less resistance in blue states within the U.S. Thus, producers' efforts directed at blue states might be more productive as well as efficient.In addition, a different strategy may be more appropriate for conservatives, who are less open to new experiences.For example, future research should explore whether an appeal of tradition ("most produce was imperfect decades ago!") may have greater traction than an appeal of novelty or an appeal towards decreasing food waste.This idea is similar to the notion of normalizing imperfect or ugly produce, first introduced to literature by Makhal et al. (2020).Perhaps, product feature-based appeals (high-quality, organic, etc.) like those used by MM and imperfectfoods.commay be successful in reaching conservatives as well through enhancing food literacy with this consumer segment.
Second, anthropomorphizing has been shown to move the needle on increasing purchase intention for imperfect food (Shao et al., 2020).However, we do not yet know whether the type of anthropomorphic image matters.Indeed, research has shown differential persuasive effects depending on whether a persuasion object is anthropomorphized as a partner or a servant (Kim & Kramer, 2015).This observation raises the intriguing questionmight conservatives respond positively to a specific type of anthropomorphic representation of imperfect FaVs?Third, our model is ripe for testing in different countries, particularly A. Aydinli et al. because the liberal-conservative dichotomy exists in many countries.However, in addition to the left-right belief systems, countries vary in many other ways, e.g., in their individualism-collectivism, power distance belief, economic strength, administrative structure, degree of political polarization, food culture, etc.Therefore, even though we do not have specific predictions, it would be useful to test our predictions in other countries since it will be meaningful both, theoretically as well as from a policy perspective.For example, culturally specific food identities may override the political preference moderator or have an interactive effect.Fourth, our research is purely quantitative and does not delve deeply into the singularization of political parties or imperfect FaVs.Thus, there remains a great deal of work to be done to understand how risk is constructed for FaVs as well as the general sensemaking process that occurs for both liberals and conservatives in this area.
Fifth, we used the word "imperfect," to describe fruits and vegetables that were not perfect.This choice mirrors real-world marketing language used by companies such as Imperfect Foods.However, the use of such verbiage may have inadvertently predisposed participants more negatively towards such produce which we tried to counter by supplying images.Future research should seek to understand whether our findings replicate using varied terminologies and descriptions.Lastly, our findings which are based exclusively on online primary data collection help set the stage for large scale field testing using behavioral measures.

Conclusion
Considering the vast swaths of human beings going under-nourished every day, our planet urgently needs mechanisms to address the gargantuan and related problem of food waste.One of the manifestations of such waste is the large proportion of produce which gets disposed without being consumed, primarily because it is aesthetically imperfect.Our inquiry can serve as a precursor to more systematic search for productive solutions that can alleviate human suffering while enhancing sustainability of the agricultural industry.

Fig. 1 .
Fig. 1.Conceptual model and operationalization of key constructs across studies.