Collaborative Consumers Can Be Ethical Consumers: Adapting the Defining Issues Test to Understand Ethical Reasoning in Collaborative Consumption Markets

Abstract

Collaborative consumption activities like saving food and buying used clothes are an important and rapidly growing part of sustainable consumer behavior. Many political and commercial campaigns promote collaborative consumption practices by highlighting subsets of normative motives, such as sustainable, social, and ecological effects. Whether or not consumers can comprehend these claims and incorporate them into their decision-making process is, however, unclear. This article introduces a new experimental study design to ethical consumer research—an adapted version of the Defining Issues Test—that enables to assess consumers’ ability to cognitively and intuitively understand moral arguments in collaborative consumption contexts. The experiment was conducted with N = 435 collaborative consumers from Germany. The findings indicate that in different collaborative consumption contexts, consumers consider different moral arguments as important. Furthermore, the findings also indicate that consumers can incorporate the provider’s point of view in their moral judgment.

Keywords

collaborative consumption consumer social responsibility Defining Issues Test ethical decision-making Kohlberg experiment

Saving food, buying clothes secondhand, and using car-sharing services appear to be the spikes of modern sustainable consumption. People who engage in such collaborative consumption (CC) activities and who we will therefore call collaborative consumers, receive and provide market goods and services in cooperation with other consumers, either peer to peer or with the support of mediator organizations. To support sustainable consumption styles and to push new business ideas, many political and commercial campaigns promote CC styles by highlighting a variety of normative motives. Companies like the local commuter app Faxi in Leeds, United Kingdom, use the same argument as the German federal agency of environment to promote car sharing (i.e., “sharing cars reduces pollution and saves money”; Umweltbundesamt, 2022; University of Leeds, n.d.). On their website, the recommerce giant eBay regularly posts stories like “Recommerce: Fueling Economic Opportunity and Sustainability” (Iannone, 2021). The global hospitality exchange service Couchsurfing, which enables its users to offer and request free accommodation, promotes a culture of sustainable social exchange on their website (Couchsurfing, 2021). Apart from that, there are numerous political actors, researchers, and media articles who embrace CC as an important cornerstone of sustainable economies (Buczynski, 2013; Walsh, 2011).

These are just a few examples of the wide range of moral arguments that are being used to promote the heterogeneous realm that are CC markets. Studies have shown that consumers are receptive to normative marketing communications in many contexts, that they can comprehend, accept, and also reject such claims and that they can even adapt their purchasing habits with regard to such claims (d’Astous & Legendre, 2009; Orazi & Chan, 2020; Schlaile et al., 2018). CC markets differ, however, from other consumer environments in certain important respects. People who engage in sharing, redistributing, and gift-giving activities do not only receive goods and services—they also are the ones who can provide those very services and goods. Collaborative consumers may therefore also act as prosumers, co-producers, or providers, and as such they may create value (Dellaert, 2019). That is the case for people who sometimes book accommodations on AirBnB and sometimes provide their own flat on the platform, or for people who sign up for carpooling on Carpoolworld. All these characteristics deviate considerably from established consumer roles (Belk, 2014). For this reason, a number of scholars claim that people in CC markets either switch between the consumer and the provider role (Scaraboto, 2015) or that they take on the new social role of the collaborative consumer (Ertz et al., 2019).¹ Regardless, collaborative consumers know the provider’s point of view firsthand or they can easily imagine themselves being in that position. Therefore, it is reasonable to argue that consumers inside CC markets might be more sensitive to moral arguments that highlight economic fairness and social benefits than consumers outside these markets. However, this thesis presupposes that collaborative consumers are able to cognitively or intuitively understand the moral arguments that are used to promote complex motives of fairness and social benefits, and that they can incorporate those arguments into their decision-making process (Septianto & Soegianto, 2017).

Contrary to this thesis and the presupposition that comes along with it, many extant studies cast doubt on whether moral statements enter the consumer decision-making process. For example, consumers may favor organic over conventional food not for ethical but solely for taste reasons (Delmas et al., 2016) and they may even hold on to non-ethical reasons when companies promote environmental or social benefits only (Apaolaza et al., 2018). Beside this egoistic perspective, certain research suggests that, in many contexts, consumers are simply not able to comprehend all ethical dimensions of their choices (Caruana, 2007; Kingston, 2021). This finding is widely accepted and goes hand in hand with the theory of bounded rationality (Sent, 2018; Simon, 1997). The theory argues that consumers as human beings have only limited receptivity and operate in markets characterized by great information and power asymmetries, limited market access, scarce time, and material resources. The number and complexity of arguments that consumers can cognitively consider in their moral deliberations are thus limited (Tanner & Carlson, 2009). This line of argumentation concludes that even if consumers are able to cognitively or intuitively understand moral arguments in favor of CC, it remains unclear if and how these arguments can influence their decision-making process (Carrington et al., 2010; Orazi & Chan, 2020).

Overall, the controversial research situation leads to many voids in the literature, which the present article aims to fill: It is not known which argumentative moral statements in favor of CC choices consumers can cognitively or intuitively understand, nor whether such statements affect their decision-making processes, nor whether collaborative consumers are sensitive to the consumer-provider role intersection and the moral points of view of each role.

We argue that addressing these issues requires a new methodological approach. Conventional study designs in consumer ethics research either focus on empirically observable purchase behavior without investigating the moral arguments that might have motivated the very individuals whose behavior is observed (Balderjahn et al., 2018) or, conversely, they collect different indicators to approximate consumer intentions without recognizing actual purchasing behavior (Beldad & Hegner, 2018). Both methods usually result in the well-known intention-behavior gap (Andorfer & Liebe, 2012). Alternative approaches survey consumers’ moral values and their proclaimed intentions in qualitative interview studies (Davies & Gutsche, 2016). Here, social desirability biases are difficult to avoid (Egloff & Schmukle, 2002). Other studies try to predict consumers’ attitudes and future behavior with the help of preference analysis (Auger et al., 2007). This method does not determine whether consumers understand and support the moral arguments promoted. Moreover, as Sunstein (1996) noted, future intentions and actions cannot be predicted from elicited preferences because those preferences may change over time and within new contexts.

To fill this methodological void, we have developed a new experimental study design that adapts and extends the Defining Issues Test (DIT2; Rest, Narvaez, Thoma, & Bebeau, 1999). By using short stories featuring moral dilemmas, the DIT2 places test subjects in a morally complex situation in which they are asked to argue how the story’s protagonists should act and why they should act so. In the following reflection process, the subjects have to, first, rate the relevance of different complex moral arguments that give good reason either for or against the protagonists’ action and, second, rank what in their opinion are the four best arguments. Based on the rating and ranking of those moral statements, the test subjects match one out of three levels of moral development. The classification allows a quantified evaluation of the subjects’ ability to comprehend moral arguments. Although scholars have already proven the validity of the DIT2 extensively in business ethics (Doyle et al., 2009; Mudrack & Mason, 2019) and while some have even adopted the original Kohlberg stories to every day ethical dilemmas (Weber, 1990), they have primarily assessed the moral reasoning of business students and managers. Adapting and applying the DIT2 in ethical consumer research is a novelty.

To fit the CC context, we embedded the DIT2 in an experiment and made the following methodological developments. First, we modified the original DIT2 short stories to reflect real-life CC dilemma situations. Second, to consider the social role intersection between the consumer and the provider, which is characteristic of CC markets, half of the stories were written from the consumer’s point of view and the other half from the provider’s point of view. Third, we developed a priming procedure, which we applied after the first part of the experiment to analyze the stability of moral judgments. Finally, we ensured the validity of the adjusted DIT2 with the help of well-established scales in empirical consumer ethics research, which cover functional, social and moral value, and sustainability attitudes. In the face of this methodological contribution, we need to answer the research question:

Research Question 1 (RQ1): Is the adjusted DIT2 a valid method to analyze consumers’ ability to cognitively and intuitively understand moral arguments in different CC contexts?

For the additional theoretical contribution, we also strive to answer the following research questions:

Research Question 2 (RQ2): Does consumers’ understanding of moral arguments vary with the CC context?

Research Question 3 (RQ3): Does consumers’ sensitivity to the provider’s point of view vary across different CC contexts?

Research Question 4 (RQ4): Are collaborative consumers able to incorporate moral arguments of various complexity in their decision-making process?

In this article we outline the theory of moral development and a number of characteristics of CC contexts, we describe the novel experimental design, report the analysis, and discuss the results. Table 1 presents the methodological steps in which the research questions are examined.

Table 1.

Methodological Approach.

Research question	Method	Main finding
RQ1	Comparing scores of the adjusted DIT2 (P-score, N2-score, U-score) with scores from standardized DIT2. Verifying results with survey items established in consumer research.	The adjusted DIT2 is a valid method to analyze consumers’ ability to understand moral arguments in different CC contexts. It can be utilized in future research.
RQ2	Comparing P- and N2-scores in four different CC contexts.	Varying between CC contexts, consumers consider moral arguments from different stages of moral development when they face a consumption dilemma.
RQ3	Comparing P- and N2-scores between the consumer and the provider point of view.	Consumers are sensitive to the providers’ points of view in CC contexts.
RQ4	Analyzing and interpreting U-scores	As long as no financial incentives are involved in moral decision-making, collaborative consumer’s motives and intentions align best.

Note. DIT = Defining Issues Test; CC = collaborative consumption.

Conceptual and Theoretical Background

Moral Development—Theory and Practices

Moral development theories propose that human beings develop their ability of moral thinking, reasoning, and intuition as they grow up (Kohlberg, 1958). Due to practical experience and education, people gain a better understanding of their social world and the norms that do and should apply to it, and they can better assess the consequences and side effects of different courses of action (Wellman & Liu, 2004). Practical experiences also improve people’s moral intuition, which helps them make moral judgments with less cognitive effort (Haidt, 2001; Lemerise & Arsenio, 2000).

What all moral development theories have in common, however, is the assumption that people’s ability to understand moral arguments is crucial to the process of moral judgment, and that this ability varies at different stages in life and in different social contexts. Ethical internalism bridges the gap between people’s moral judgment and their actions even further (Williams, 1981) by proposing that human beings are motivated by sincere moral judgments to perform corresponding actions. Why is that plausible? Imagine that the two scientists, Hannah and Nils, discuss the issue of producing gender-sensitive articles. Nils thinks that such norms are just a trend and by no means justified. Hannah thinks otherwise and convinces Nils with her strong arguments. If Nils changes his mind and sincerely believes gender sensitivity to be of importance in scientific articles, it would appear odd if he would not feel motivated to actually use gender-sensitive language in his next article. Proceeding from the internalist approach and the thesis that moral judgment can develop with experience, people in CC markets must be able to grasp normative arguments cognitively and intuitively to be motivated to act upon them.

Moral Judgment Interviews

Lawrence Kohlberg confronted children and youngsters in the moral judgment interviews (MJI) with a series of fictional short stories, all of which turned out to be moral dilemmas (Kohlberg, 1958). At the open end of each short story, the interviewees were asked to explain how the stories’ protagonist should act and why she or he should act in that manner. Probably the most famous of those stories is the Heinz dilemma: The protagonist, Heinz, faces the dilemma of whether or not to steal a drug to save his terminally ill wife. Based on the young interviewees’ judgments, Kohlberg created a three-level model of moral development that distinguishes between six stages of moral reasoning (Table 2). The first and second stage is summarized as the pre-conventional level in which a focus on physical consequences determines the idea of moral right and wrong. People in stage one want the stories’ protagonists to avoid punishment (Heinz will get in trouble if he lets his wife die), while in Stage 2, they highlight personal benefits and risks (Heinz may not get much of a jail term but his wife will probably die before he gets out). The third and fourth stages are both about social conventions. At this level, people want the protagonists to act in conformity with their peer groups (everyone will think Heinz is a criminal) or in accordance with general social norms like the law (Heinz will always feel guilty, if he don’t do his duty to his wife). On the conventional level, people start to incorporate intersubjective perspectives in their moral judgment and they learn to collaborate with each other. Finally, the post-conventional level represents universal moral norms. These norms are either utilitarian or contractual in stage five (The state should uphold democratic legitimized law, even though Heinz thinks it is wrong) or Kantian in stage six (If Heinz don’t steal the drug he would have lived up to the outside rule of the law but he wouldn’t have lived up to his own standards of conscience) (Kohlberg, 1971: 67 and 91).

Table 2.

Stages of Moral Development (in Accordance With Kohlberg, 1971; Rest, 1979).

Stage	MJI	DIT	Attitude	Category
6	Post-conventional		Act such that your actions and the effects of your actions can be desired by all humans and that everyone could potentially act in the same manner	Deontological
5	Post-conventional		Contract structures are recognized as socially desirable. Act on behalf of a social consensus that is desirable to all parties involved	Contractual and utilitarian
4	Conventional	Maintaining norms	Abstract normative structures can be recognized. Subjects recognize laws and authorities as important. Do your duty	Authoritarian
3	Conventional	Personal interest	Personal and other intentions can be rated as good/bad. Subjects want to meet the expectations of their peer group. Do what is best for the group	Social consensus/ conformity
2	Pre-conventional	Personal interest	Subjects recognize other people as bearers of their own goals and desires. Fulfilling or ignoring other people’s goals and desires serves the interests of the subject. An eye for an eye	Strategic self-interest
1	Pre-conventional		Subjects can only recognize their own desires. Action is conditioned through sanctions and rewards.As long as there is no punishment, it is ok	Blind egoism

Note. MJI = moral judgment interviews; DIT = Defining Issues Test.

In his empirical studies, Kohlberg identified many relational characteristics between the stages. A person who enters a given stage can understand all arguments that follow the logic of that very stage. As people improve their cognitive moral abilities, they can understand more complex arguments and enter higher stages. Thus, they become more skilled in three manners. First, starting from a self-interested perspective, people become better at adapting intersubjective arguments and recognizing other people’s interests. Second, they become more self-determined. Although, in an early stage, they accept norms without question, they can later consider and evaluate reasons for given norms, and at a certain point they are even able to develop moral rules by themselves. Finally, people become more skilled in reasoning. Despite these developments, people in a given stage are still able to assess a situation from a lower stage perspective (Elm & Weber, 1994). Kohlberg believed that all human beings in all cultures and at all times climb from stage to stage in their moral development, one stage at a time without skipping stages (Kohlberg, 1981). The pace of their development and the stage they will eventually end up in may differ. A high stage, therefore, does not equal maturity. Furthermore, neither genetic preconditions nor a particular level of socialization ensures that everyone will end up in Stage 6 (Kohlberg, 1971).

Although the MJI have been successfully used and tested for universal cross-cultural use, the method is too time consuming to be easily used in large-scale consumer studies (Kurtines & Greif, 1974). A simpler quantitative alternative that we believe to be most attractive to ethical consumer research is James Rest’s (1979) DIT.

Defining Issues Test

The DIT translates Kohlberg’s MJI into a paper-and-pencil version with six phases. For each phase, test subjects read a moral dilemma, they decide how the protagonist should act, they rate 12 argumentative moral statements according to their relevance and then they rank the four statements they believe to be most important for the protagonist’s decision. The 12 statements reflect Kohlberg’s moral Stages 2 to 6. Notably, Rest and colleagues found Stage 1 not to be of relevance since they did not consider participants younger than ninth grade and after that age nearly no test subject was reported to prefer Stage 1 arguments (Thoma, 2006). Omitting the first stage, they regrouped the five remaining stages, with Stage 2 and Stage 3 representing the pre-conventional or personal interest level, Stage 4 the conventional, maintaining norms level and Stage 5 and Stage 6 the post-conventional level (Table 2). That is not unique to the DIT. Weber and McGivern (2010), for example, collapsed both pre-conventional and both post-conventional stages in an adaptation of the MJI because these stages have been proven to be very similar statistically. With the revised version of the DIT2, Rest, Narváez, Bebeau, and Thoma (1999) improved the statistical assessment of the rating and ranking phase. Later, Xu and colleagues (2007) developed an online version.

When a subject in the DIT2 rates and ranks a statement (e.g., explications of moral principles), it can be assumed that she or he can understand the argumentative statement and, thus, prefers one over another that is ranked and rated lower. Hence, if a subject gives a low rating, or no rating, to a statement, she or he is either unable to understand the argument or considers it as unimportant (Rest, Narváez, Bebeau, & Thoma, 1999). In both cases, the test can reveal which moral arguments the test subjects cognitively and intuitively understand. A possible problem with this test design is that subjects can be biased by the statements they are presented with. Subjects might rank statements randomly or they might try to appear morally mature by giving a high ranking and rating to statements that sound pretentious (Elm & Weber, 1994). To eliminate this problem, the DIT2 includes a few non-sense statements that sound like Stage 5 or Stage 6 statements (Thoma, 2006). If a subject rates a number of non-sense statements as important, she or he is excluded from the test. Besides that, the DIT2 also bypasses the social desirability bias, which is a test subject’s tendency to give an answer to morally tense questions that she or he believes to be socially acceptable (Schwarz, 1999). The reason the DIT2 bypasses the social desirability bias is that there are no socially desirable solutions to the dilemma scenarios presented.

The two main differences between the MJI and the DIT concern the object they focus and the results they produce. In the MJI, subjects develop and articulate unique moral arguments. The method can, therefore, verify a subject’s cognitive and linguistic ability to create and understand moral reasons. In contrast, the DIT2 cannot recognize self-made arguments. Instead, it confronts subjects with linguistic representations of possible moral arguments that they have to rate and rank. Accordingly, the DIT2 can only assess the subjects’ ability to understand moral arguments (Elm & Weber, 1994; Thoma, 2006). Thus, the results of both methods are not comparable. The MJI is a “hard stage concept” since it uses content analysis to determine the moral judgment stage of a subject. The DIT is a “soft stage concept” that takes all weights that subjects put on representative arguments into account and calculates a score (Elm & Weber, 1994). These differences have various effects on the average scores (see scoring).

For consumer research in general and the CC research in particular, the idea of stages of moral development that are sensitive to people’s willingness to collaborate, and their ability to conceive moral arguments is rather appealing. Introducing the DIT to consumer research seems promising, too, because it has been tested and validated extensively in the business context (Doyle et al., 2009; Mudrack & Mason, 2019; Weber, 2018), and because it captures the moral reasoning of consumers that are confronted with normative arguments in a given context.

CC Contexts

The role² of the consumer is characterized by the ability to interact with other market actors, the ability to obtain goods and services and—from a neoclassical perspective—the motive to optimize personal benefits (Frank, 2017). In real-life circumstances, however, this role evolves in its complexity by recognizing additional role motives, role-specific social pressures and role responsibilities (Müller, 2022). People who perform the consumer role thus have to meet different expectations in different CC contexts. In certain contexts, they are expected to buy secondhand clothes, in others they are expected to share their car via an online app and in other contexts they are expected to take food for free. For these different practices, very different ecological, political, social and ethical motives are important (Bendapudi & Leone, 2018; Eckhardt & Bardhi, 2016). Since the different CC contexts appear to have a major influence on collaborative consumers’ course of action and the normative motives they usually entertain, it is paramount to highlight one key characteristic that all CC contexts have in common.

In all CC contexts, consumers are able to use various forms of compensation to either access or redistribute goods and resources (Akbar & Hoffmann, 2022; Mair & Reischauer, 2017; Puschmann & Alt, 2016). Collaborative consumers engage in redistribution contexts when they use non-profit services like Foodsharing, a platform that enables companies and private persons to donate food or leftovers to a digitally organized community. They engage in access-based consumption formats when they use services like Couchsurfing, which mediates private accommodations without charging a fee. Of course, there are also redistribution and access-based CC contexts in which consumers have to spend money (Böcker & Meelen, 2017; Frenken & Schor, 2017). AirBnB, for example, is the for-profit counterpart to Couchsurfing. AirBnB provides access to privately owned accommodations via a website and they profit from a commission. Perhaps the best-known example of a for-profit redistribution service is the online marketplace eBay, where consumers can either buy or bid on products or services that are in most cases owned and provided by private persons.

This article focuses exclusively on CC contexts that are characterized by the same individual being able to take the provider and the consumer role. All contexts that recognize for-profit sharing of company-owned goods are excluded (Belk, 2014). That consumers can become acquainted with the provider’s point of view firsthand is particularly interesting for consumer ethics research. To act in line with personal normative values, collaborative consumers can understand the perspective of other stakeholders in a collaborative transaction and empathize with them simply because they have taken their role at another time or because they have a good insight into what it means to take that other role.

The Experiment on Consumer’s Moral Judgment in CC Contexts

Method

To answer RQ1 to RQ4, we developed an experimental study design consisting of a short questionnaire that collected socio-demographics and CC habits (A), an adjusted version of the DIT2 (B), a priming procedure in the form of normative text cues (C) and a complementary survey part that asked test subjects about their opinions on CC markets and their individual values associated with sharing (D). The structure of the experiment is represented in Figure 1.

Figure 1.

Structure of the Experiment.

A: Socio-Demographics and CC Habits

At the beginning of the experiment, participants were asked to provide demographic data (sex, age, level of education, income, housing situation, religious and political beliefs). They had to select one out of five political labels that conforms best to their personal beliefs. The labels represent the current political landscape in Germany (conservative, liberal, ecological, patriotic, or social; see Table 5). Participants then had to indicate what CC practices they have participated in formerly and how frequently they had practiced them. As we argued before, we had to ensure that the respondents had prior experience with CC activities, as we assume that only those individuals can truly form beliefs and opinions about certain CC contexts they know firsthand.

B: The Adjusted DIT2

After having completed the questionnaire, subjects were randomly assigned to one out of four groups: the consumer perspective and its control group, and the provider perspective and its control group. Experimental group 1 received dilemma scenarios from the consumer’s point of view and Experimental group 2 received dilemmas from the provider’s point of view. The two control groups received the very same scenarios, but they were not primed in the middle of the experiment. All test subjects were confronted with a total of six dilemma stories (Rest, Narvaez, Thoma, & Bebeau, 1999). In each consumer dilemma story, the participants were confronted with a scenario, in which choosing the CC option is economically either advantageous or neutral, but morally complex. To determine potential nuances of moral judgments in different CC contexts and also to detect possible priming effects, test subjects did three iterative phases featuring dilemma stories of three different CC contexts. Then, all participants in the experimental groups were primed. Afterward, the experimental and the control groups continued with the second DIT2 part that featured three different dilemma stories from the same three CC contexts. To prevent possible spillover effects of the sequence of the stories, we randomized the order of the stories. Table 3 shows how the groups were assigned to the stories and how the randomization of the different stories was conducted.

Table 3.

Group-Story Allocation.

	DIT2 part 1			Priming	DIT2 part 2
Groups	Story 1	Story 2	Story 3	3 moral statements	Story 4	Story 5	Story 6
Experimental Group IConsumer perspective	1a	1b	1c	Yes	2a	2b	2c
	1b	1c	1d		2b	2c	2d
	1c	1d	1a		2c	2d	2a
	1d	1a	1b		2d	2a	2b
Experimental Group IIProvider perspective	1e	1f	1g	Yes	2e	2f	2g
	1f	1g	1h		2f	2g	2h
	1g	1h	1e		2g	2h	2e
	1h	1e	1f		2h	2e	2f
Control Group IConsumer perspective	1a	1b	1c	No	2a	2b	2c
	1b	1c	1d		2b	2c	2d
	1c	1d	1a		2c	2d	2a
	1d	1a	1b		2d	2a	2b
Control Group IIProvider perspective	1e	1f	1g	No	2e	2f	2g
	1f	1g	1h		2f	2g	2h
	1g	1h	1e		2g	2h	2e
	1h	1e	1f		2h	2e	2f

Note. a-d = stories from the consumer’s point of view; e-h = stories from the provider’s point. Part 1: first half of the DIT2 before priming. Part 2: second half of the DIT2 after priming. DIT = Defining Issues Test.

Each of the six iteration phases consisted of three steps that are characteristic for the DIT2. First, the test subjects had to read a short story (one of story 1.a-d/e-h) and decide how they thought the protagonist of the story should act (“Do it,” “Don’t do it,” “I can’t decide”; Xu et al., 2007). Second, the subjects had to rate 12 distinct moral statements by their relevance for the dilemma on a five-point scale (great relevance/much relevance/somewhat relevance/little relevance/no relevance). In a third step, participants were asked to rank the four most important statements (Xu et al., 2007).

Thus far, the iteration phases conform to the well-known standard. The adjustments on the DIT2 consisted of three parts. First, the hypothetical life-or-death dilemma stories that are characteristic of Rest were replaced by everyday CC dilemma scenarios. The adjustment addresses the criticism that hypothetical stories make it difficult for subjects to engage emotionally (Gilligan, 1982; Skoe et al., 2002), and it acknowledges theoretical and empirical findings according to which the stage of moral reasoning varies with the context of the moral dilemma. Jones (1991) argues that because human beings are cognitively limited they employ simplification techniques to judge familiar moral issues. They cannot employ those techniques, however, when an issue gets to intense. That is why the stage of moral reasoning surveyed depends on the intensity of the moral issue at play. If the moral stakes in a given dilemma story are high, so is the measured stage of moral reasoning. Several studies by Weber and colleagues support Jones thesis but also emphasize the importance of other contextual factors. They conducted adapted MJI, in which managers and business students were confronted with Kohlberg’s Heinz dilemma followed by two other business-related dilemmas. The results show that the average stage of moral reasoning was different for all three dilemma stories. The score was high in the Heinz dilemma, confirming Jones thesis, and varied in the other two stories (Weber, 1990; Weber & McGivern, 2010; Weber & Wasieleski, 2001). Similar dilemma adaptations are common for the DIT (Doyle et al., 2009).

For this experiment, two representative dilemma story for each type of CC context (access-redistribution/profit or not for profit) were developed (2×4). Each story features a different moral issue that involved juridical, social, ecological, and economic components and that varies in intensity (Table 4). Just like Rest’s stories their adapted versions do not only put two values in conflict but multiple. Every course of action in every story affects the protagonist’s future economic and social well-being, each one has moral consequences and each one is either in line with the protagonist’s beliefs or not. In addition and to increase moral intensity, some of the protagonist’s options are unlawful or bad for the environment. The adapted moral statements in the rating phase reflect this multitude of conflict. However, test subjects are always able justify their recommendations with personal economic benefits or egoistic reasons.

Table 4.

DIT2 Dilemma Stories for the Collaborative Consumption (CC) Contexts.

Dilemma stories	Redistribution		Access-based
Dilemma stories	For-profit	Non-profit	For-profit	Non-profit
Consumer perspective	Story 1.bShould Hannah (Dominik) fake a good eBay review in return for an urgently needed discount?	Story 1.cShould Amelie (Sebastian) illegally go dumpster diving to help out people in need?	Story 1.aShould Chiara (Fabian) illegally book an AirBnB flat that she needs to attend a short-term internship?	Story 1.dShould Julia (Alex) give way to a penniless student in a double booking on Couchsurfing?
Consumer perspective	Story 2.bShould Valerie (Mike) give a provider on Kalaydo a bad review for not fully assembling an ordered kitchen as they agreed to?	Story 2.cShould Caroline (Paul) pay money for trees that were actually advertised as free on the platform FreeYourStuff, when she can use those trees to help out her neighbors?	Story 2.aShould Sophia (Hermann) accept a Getaround deal, if the car is in much worse condition than described online?	Story 2.dShould Tinka (Nils) take the help of a racist pensioner to repair her bike?
Provider perspective	Story 1.fShould Hannah (Dominik) give a discount on an eBay deal to avoid a negative review?	Story 1.gShould Amelie (Sebastian) give away vegetables on foodsharing and risk that a view of her customers might start sharing food instead of buying it?	Story 1.eShould Chiara (Fabian) accept a private and generous accommodation request on AirBnB, even if the offer exceeds the legal limit?	Story 1.hShould Julia (Alex) prefer a penniless student over a financially independent tourist on Couchsurfing?
Provider perspective	Story 2.fShould Valerie (Mike) deliver a kitchen that she sold on Kalaydo a second time, if no one was there at the appointed time?	Story 2.gShould Caroline (Paul) give a negative review to a FreeYourStuff user who let donated trees go to waste?	Story 2.eShould Sophia (Hermann) rent their car to someone on Getaround, if they can only pay 2/3 of the price?	Story 2.hShould Tinka (Nils) help a migrant fix his bike in her/his garage if the migrant is very rude?

The second adjustment of the DIT2 concerns social roles as important contextual elements. The effect of different social role components on moral reasoning, like a subject’s profession and the rules of conduct in its line of work, has been tested in business case dilemmas (Weber & Siniora, 2022; Weber & Wasieleski, 2001). It is plausible to expect similar effects in CC contexts. For her family holiday, a person can swap her apartment with someone else’s on Couchsurfing and she can do so in her social role as collaborative consumer, accommodation provider, mother, and spouse. When the same person is promoted to CEO, she might be convinced that using Couchsurfing is no longer appropriate. It is not feasible to include a large number of social roles into this study. However, since CC contexts are characterized by the fact that the same people know the provider and the consumer role firsthand, those two role perspectives were introduced to the stories (2×8).

When dealing with moral dilemma stories, some scholars argue that the gender of the protagonists might affect the test subjects reasoning as well (Cretser et al., 1982; Timmers et al., 2003). For Kohlberg’s original stories, scholars found no evidence of such an affect (Donleavy, 2008). However, for dilemma stories adapted to the business context, there is weak evidence that in some contexts test subjects score higher if their gender matches the gender of the stories’ protagonist (Weber & Siniora, 2022). To test this possible effect in CC contexts, test subjects who identify as female received stories in which the protagonist was female and vice versa (2×16). In total, there were 32 short stories. All stories had been pre-tested. The normative issues of the stories are represented in Table 4. Full consumer point of view stories are in the supplementary file.

Third, corresponding rating arguments, the rationale of which matched the logic of Kohlberg’s six stages, were developed. The AirBnB story, for example, had argumentative statements like: “The landlord of the apartment asks for a price that is not legal. It feels wrong to rent from such a person” (pre-conventional stage), “It is important that Chiara finds a place to live and start her internship. Otherwise, her parents will be very disappointed” (conventional stage), and “Heterogeneous living spaces are important for a politically open and just culture. People cannot want this diversity to be systematically destroyed by high housing prices” (post-conventional stage). Each stage of moral development was represented by two statements: one in favor of the protagonists’ action and one that argues against the proposed action. This equals 12 statements unique to each story. The reason why we did not exclude stage one arguments for this research like Rest does in the traditional DIT2 is simple: Neoclassic economics pictures consumers as homines oeconomici who are interested in nothing but their personal benefit. Soper summarizes this view quite aptly: “Consumers are without question departing from the most egoistic forms of self-interest in the love and care they bring to shopping or socializing with family and friends” (Soper, 2007: 207). By considering the moral stage of blind egoism, test subjects in the experiment were able to confirm to this view. We will show that this decision does not have a distorting effect on the results, and that real-life collaborative consumers are thus no homines oeconomici.

C: Priming Procedure

After the first DIT2 part, test subjects had to pass a second DIT2 part with new dilemma scenarios (2.a-d/e-h). Although the new stories featured new protagonists and different platforms and consumer–provider relations, they represented the same CC contexts as the stories from the first part. Before they started the second part, test subjects in the Experimental group 1 and 2 were confronted with well-known sayings like “Tit for tat” or “Do one good deed a day.” Each of those sayings can be associated with one of the three levels of moral development (personal interest, maintaining norms, post-conventional). Test subjects had to read three statements of the same level of moral development and each time they had to describe in writing a situation in which they complied with it. If the moral judgment of a test subject were to be stable during the experiment, it can be expected that this priming phase does not show any effect. To test whether the priming had any effect at all, two control groups underwent the experiment without being primed.

D: Measurement

After the main experiment, the test subjects were asked to fill in a questionnaire not only on how they perceived the functional (Hamari et al., 2016; Lamberton & Rose, 2012), moral (Bucher et al., 2016) and social value (Lamberton & Rose, 2012; Tussyadiah & Pesonen, 2016) of sharing but also on their attitude toward sustainability (Möhlmann, 2015). All these constructs are well established in the literature and it has been shown that they influence consumers’ sharing behavior (see Appendix). We added this section to the experimental design to assess the validity of the adjusted DIT2 in CC contexts. Particularly, we checked for potential correlations between the well-established concepts from the survey and the normative arguments from the rating phase.

Reliability

All stories, the rating statements and the priming text cues were developed by all four authors and had been tested in two pre-tests with a total of N = 82 business students. In the DIT2, there are several reliability checks to identify unreliable test subjects and render their answers invalid. When subjects fail to rate statements, this may be because they lack motivation. For that reason, if more than eight ratings for any two stories are left out, the test is invalid. The same is true for the ranking phase. If a subject skips more than three ranking items on any two stories, the test is also invalid. Another important issue relates to Stage 5 and 6 statements inevitably standing out due to a more complex sentence structure and choice of words. To also assure that test subjects do not complete the rating based on an optical pattern, the statements were presented in a randomized order. Moreover, certain participants may try to score high on the tests by giving a high rating to the statements that sound the most complex without understanding them. To test whether participants only gave high ratings to the statements that they understood, 12 complex sounding non-sense statements had been put into the rating phase of all 16 stories. This means that a zero to one rating statement (Stage 5 or 6) of each story just sounded pretentious but lacked meaning (Enden et al., 2019; Rest et al., 1997). Another reliability check was to exclude participants who rated 11 statements the same for more than one story and thus failed to discriminate their answers (Rest, Narvaez, Thoma, & Bebeau, 1999). Furthermore, we excluded participants from the analysis who were unreasonably quick in their response time and completed the whole experiment in <12 min (two SDs below the mean time of 30 min). A total of 29 participants failed at least one of the reliability checks and were excluded from further analysis. This led to a final sample of 435 participants.

Sample

For the acquisition procedure, social media ads, an email distribution list for the experimental lab of a German university and flyers that were spread all over the campus were used. Since the experiment was all about understanding moral statements in CC contexts, the ads asked only those people to participate who had firsthand experience with CC practices. Participants were also asked about their CC experience before they were randomly assigned to the experimental groups (see A socio-demographics and CC habits). This sampling ensures internal validity (Spilski et al., 2018).

In total, 435 collaborative consumers from various social groups and occupations were tested, 230 of whom were female (52.9%). The test subjects ranged in age from 18 to 71 years (mean age 26.8, median age 25, SD = 7.89). Due to the acquisition strategy, it turned out that most subjects lived in rental or shared apartments (90.7%), their education was above high school level (94.8%) and their monthly income was below €1,000 (66.2%) (Table 5). As another measure to ensure the stability of the findings, we conducted the experiment both in the lab and online. About half of the participants (46.7%) completed the experiment in the German university’s computer lab, the other 53.3% did it via an online link at a personal computer. We tested for comparability between the respondents who participated in the lab and those who participated at home by comparing the average time each respondent took for the whole experiment. Furthermore, we included the place of the experiment (laboratory vs. at home) in the analysis to test for the robustness of our analyses. None of the results showed any significant differences between test subjects who completed the experiment in the lab vs. those who did it at home. Each participant received €10 in cash as compensation.

Table 5.

Final Sample After Reliability Checks.

Criterion	N	% of total sample
Gender
Female	230	46.9
Male	204	52.9
Other	1	0.2
Education
Below high school	23	5.2
Above high school	412	94.8
Income (in euro, after housing costs)
≤1,000	288	66.2
>1,000	147	33,8
Housing situation
Shared apartment	134	30.8
Rented apartment	218	50.1
Own house	25	5.7
Own apartment	8	1.8
Rented house	34	7.8
Other	16	3.8
Religious
Yes	219	50.3
No	214	49.2
No answer	2	0.5
Political views (multiple choices possible)
Conservative	81	18.6
Liberal	167	38.4
Ecological	211	48.5
Patriotic	17	3.9
Social	251	57.7
Total	435

Scoring

Both Kohlberg and Rest suggest a scoring system that displays a subject’s moral development with a numerical index. The most frequently used scores are the P-score and the N2-score.

P-Score

The P-score represents the proportion of statements that appeal to post-conventional statements. In long-term studies it also displays individual moral development. The P-score is calculated by dividing all ranking points that were given to post-conventional statements [p] by the number of ranking points in total (10n), where (n) is the number of stories used in the DIT2. If a subject does not rank a statement (a) for a given story (k), the score is recalculated using the basis of (10n) points minus the points that have been left out. P(k) is the number of ranking points for post-conventional statements in story (k).

P - Score = \frac{\sum_{k = 1}^{n} P (k)}{(10 n) - a}

N2-Score

However, the P-score falls short of recognizing any other interesting data but the post-conventional ranking. Rest therefore developed the N2-score. The N2-score is used to put the P-score (without adjusting left-out rankings) in relation to the degree to which statements that appeal to personal interests (Stage 2 and 3) are given the lowest rating (Rest et al., 1997). In contrast to the P-score that recognizes post-conventional ranking information only, the N2-score also recognizes rating information as well as ranking information concerning conventional and pre-conventional stages. Since our experiment includes stage one statements as well, we calculated the degree of personal interest with statements from stage one to stage three. The N2-score is calculated in two parts. The first is the non-adjusted P*-score. The second part is the difference between the average ratings (R) in post-conventional stages and the personal interest stages. This rating is standardized by dividing it by the standard derivation. The standardized rating is weighted by three because it has an empirical standard derivation of $\frac{1}{3}$ from the P* score. In this manner, the N2-score and the P*-score are comparable. If subjects are able to discriminate personal interest statements and post-conventional ones, the N2-score is higher than the P-score. Likewise, subjects score lower if they fail to discriminate properly.

\begin{array}{l} N2 - Score = \sum_{k = 1}^{n} \frac{p (k)}{10 n} \\ + (\frac{m e a n (\sum_{k = 1}^{n} R_{k} (s t a g e 5 + 6)) - m e a n (\sum_{k = 1}^{n} R_{k} (s t a g e 1 + 2 + 3))}{S D (R_{k} (s t a g e 1 + 2 + 3 + 4 + 5 + 6))}) * 3 \end{array}

The P-score and the N2-score are always between 0 and 96, since six iterative phases include a total of four to six non-sense statements. The N2-score is, therefore, a little lower than the P-score on average because it uses more data that can potentially be invalid (Rest et al., 1997). In this experiment, we used both the P-score and the N2-score to indicate the test subject’s understanding of complex moral statements. If both scores are low for all stories, the test subject failed to intuitively or cognitively understand post-conventional arguments that promote or criticize CC. Both scores were also used to verify the priming effect as well as the intrapersonal deviations between consumers and providers. For the priming effect, both scores were compared story-wise for the non-primed and the primed cases. For example, if a subject obtained a high P-score in story 1a and after the priming phase she scored significantly lower in story 2a, it can be assumed that the priming was effective. To ensure that the differences are not due to the specifics in the stories, we compared the outcome with the control group. For the intrapersonal deviation between the consumer role and the provider role, the P-score and the N2-score between both groups were compared story-wise. Again, this calculation was compared with a control group to exclude random effects.

Researchers have noticed that the P-score tend to be higher in the DIT2 than the score stages in Kohlberg’s MJI. This effect is explained with the task of selecting items to be cognitively less demanding than the task of developing one’s own normative arguments (Weber, 2018). Conversely, it has also been argued that the relatively low P-score in the MJI are due to linguistic biases (Gibbs et al., 1992). However, for the present experiment, this exciting issue is secondary, because we study the ability of consumers to understand moral statements in CC contexts.

U-Score

In contrast to the P-score and N2-score, the U-score does not refer to the subjects’ stage of moral reasoning. Instead, it shows if subjects rated statements in accordance with their decision-making. As long as the subjects did not choose the “can’t decide” option, the U-score checks if the decision that is made for each story fits the decisions in the rating phase (Cohen, 1988: 147; Thoma et al., 1991). The U-score is calculated like this:

U s c o r e = \frac{\sum_{k = 1}^{n} \sum_{i = 1}^{n * 2} T_{i k} (3 - R_{i k}) E_{i k}}{\sum_{k = 1}^{n} \sum_{i = 1}^{n * 2} [2 | E_{i k} |]} with E_{i k} = \frac{(M P_{i k} - M C_{i k})}{S_{i k}}

T_ik is a dichotomous variable (it is either −1 or +1), and is used to maintain a positive numerator for responses consistent with the selected action choice. R is the rating of a statement (i) in story (k). E is the effective size of the difference between how all subjects who choose the “do it” option rated a statement (i) in story (k) and those ratings of the subjects who choose the “don’t do it” option. E, thus, standardizes average rating values with regard to all participants. Accordingly, MP_ik is the average rating of statement (i) in story (k) for all participants who choose “do it” and MC_ik is the rating for the same item computed for the participants who choose “don’t do it..” S_ik is then the standard derivation of all choices made. The score range is between −1 and 1. A low U-score indicates that a subject is not able to properly select items that support her or his choice of action. For the analysis, it is important to keep in mind that not all pro-action accounts are pro-CC approaches. Whether an action can be considered to be in favor of CC depends on the story design. The DIT2 scores and their meaning are displayed in Table 6.

Table 6.

DIT2 Scores.

Description	P-Score	N2-Score	U-Score
Meaning	Correlation between the ranking points of post-conventional statements and others	Consistency between post-conventional ranking and rating behavior	Consistency between choice of action and rating behavior
Highest value	96 = post-conventional statements ranked highest for every story	96 = post-conventional statements rank highest for every story AND statements that refer to self-interest were not rated	1 = rating behavior equals the average rating behavior of all others that also chose “do it”/”don’t do it”
Lowest value	0 = no post-conventional statements were ranked	0 = no post-conventional statements were ranked	−1 = rating behavior is opposed to the average

Results

This section reports all experimental results pertaining to the research questions: First of all, to answer whether the adapted DIT2 is a valid method for analyzing consumers’ ability to cognitively and intuitively understand moral arguments in CC contexts (RQ1), the results of the DIT2 (i.e., P-score, N2-score, and U-score) and the corresponding validity checks are presented and validated with the help of the supplementary survey. Furthermore, to answer RQ2 and RQ3 (i.e., whether collaborative consumers believe different moral arguments are of relevance in different CC contexts), and whether collaborative consumers are sensitive to arguments from the provider’s point of view—the second part focuses on a CC-context-specific and role-specific analysis of the scores. Finally, the third part analyses the potential effects of the priming phase and the U-scores and, thus, answers RQ4 (i.e., whether consumers can potentially incorporate such moral arguments in their decision-making process).

Validity of the Adapted DIT2 Method

Table 7 shows the mean values for the P-score, the N2-score and the U-score and for their standard deviations across the whole sample. As described above, the different scores tackle different aspects of the moral reasoning process. They cannot be compared among each other. However, they can be interpreted and compared with scores from past DIT2 research.

Table 7.

P-score, N2-Score, U-Score in the CC Contexts.

Sample			N2-Score		U-Score
Sample	M	SD	M	SD	M	SD
N = 435	42.02	15.81	47.04	18.33	0.59	0.12

The mean values of the P-score (42.02) and N2-score (47.04) are comparable to other DIT2 studies, albeit ours show higher standard deviations and thus a higher variability. For example, Rest, Narvaez, Thoma, and Bebeau (1999) reported P-values among adults between 33.7 and 53.7, whereas the N2-score for the same individuals varied between 40.4 and 53.3, respectively. The slightly higher variation and the slightly higher sub-value in both scores might be an indicator of the very specific nature of the social roles individuals take in CC markets. Besides that, the P-score and the subjects’ age were positively correlated (r(433) = .110, p ≤ 0.05), as was the case for the N2-score and age (r(433) = .105, p ≤ 0.05). Furthermore, there were no statistically significant correlations between the scores and religious beliefs, income, or gender. These scoring results are in line with Rests and colleagues’ findings and are an indicator for the validity of the adjusted DIT2.

To verify the results of the DIT2, the scores were further compared with the survey results in a correlation analysis. In the whole sample, the P and N2-values correlate positively with the test subject’s sustainability attitude and also with the test subject’s perceived social and moral value of sharing. To account for variations between the consumer’s and the provider’s point of view, a correlation analysis for each sub-sample was conducted (Tables 8 and 9). In both sub-samples, the sustainability attitude significantly correlated with high N2-scores, and the functional value that test subjects recognized in CC practices correlated with low N2-scores. The indication here is that people who value sustainability also gave high ratings to items that were in favor of sustainable practices—and those items are stage four and above. Vice versa, people who value the economic benefits of CC also gave high ratings to items that promoted those benefits—and these items were stage three and below. Furthermore, the analysis revealed a significant positive effect of the moral value on the N2-score, indicating that individuals who engaged in CC contexts for moral reasons also ranked the corresponding items high.

Table 8.

Correlation of the Main Constructs Among the Consumers’ Point of View.

Construct	P-Score	N2-Score	Sustainability attitude	Functional value	Social value	Moral value
N2-Score	.992**
Sustainability attitude	.317**	.332**
Functional value	−.015	−.000	.351**
Social value	.040	.030	.305**	.209**
Moral value	.223**	.238**	.706**	.233**	.497**
Age	.104	.105	.108	−.193**	−.066	.097

Note. Pearson correlations; level of significance: * p ≤ .05. ** p ≤ .01 (two-tailed).

Table 9.

Correlation of the Main Constructs Among the Providers’ Point of View.

Construct	P-Score	N2-Score	Sustainability attitude	Functional value	Social value	Moral value
N2-Score	.992**
Sustainability attitude	.290**	.298**
Functional value	−.106	−.119	189**
Social value	.296***	.293**	.113	.166**
Moral value	.330**	.352**	.203**	.219**	.492**
Age	.116	.116	.054	−.171*	.056	.094

Note. Pearson correlations; level of significance: *p ≤ .05. ** p ≤ .01 (two-tailed).

To test whether the test subject’s financial situation might blind them for the protagonist’s dilemma, we compared their moral reasoning behavior with their income. We compared three income groups: (A) up to 500 € per month (n = 143), (B) 501-1,000 € (n = 143), C) <1,000 € (n = 146). Applying ANOVAs, we found no statistically significant difference of the P-Scores (P_A = 41.89, P_B = 41.78, P_C = 42.25, F(2,429) =.034, p = .967) and the N2-scores (N2_A = 46.51, N2_B = 46.89, N2_C = 47.55, F(2,429) = .119, p = .888).

The experiment’s U-scores deviated from the known standard. According to Thoma and colleagues (1991), a common U-score ranges between −.41 and .77. In this experiment, however, the average U-score was .59 (SD = .18). Furthermore, no subject showed a rating behavior that did not fit the choices she or he made to solve the dilemma stories and, therefore, there were no negative U-scores.

CC-Context-Specific and Role-Specific Effects

As Figure 2 shows, the N2 scores across all four CC context types are contingent on the social roles. A one-way ANOVA revealed that individuals who read the access-based/for-profit stories from the consumer’s point of view scored significantly lower (M_Consumer = 33.86) than individuals who were confronted with the provider’s point of view (M_Provider = 41.84, F(1, 343) = 10.216, p ≤ 0.01). Similarly, subjects in the consumer role had significantly lower N2-scores in dilemma stories featuring access-based / non-profit contexts (M_Consumer = 38.37) than those in the provider role (M_Provider = 50.11, F(1, 348) = 30.261, p ≤ 0.001). The inverse pattern emerges when analyzing redistribution/non-profit contexts. There, subjects who read the stories from a consumer’s point of view scored significantly higher (M_Consumer = 47.88) than the subjects who read the stories from the provider’s point of view (M_Provider = 42.93, F(1, 348) = 4.710, p ≤ 0.05). Finally, subjects who read the dilemma stories featuring redistribution/for-profit contexts from the consumer’s point of view scored significantly higher (M_Consumer = 47.77) than those who had the provider’s viewpoint (M_Provider = 41.98, F(1, 349) = 5.776, p ≤ 0.05).

Figure 2.

Role-Specific and Context-Specific N2-Score.

To verify these findings, we compared the N2-score with the survey results. For test subjects who took the provider’s point of view, there is evidence for a significant and positive correlation between perceived social value and both the P-score and N2-score. This indicates that people who value the social benefits of CC also rated conventional and post-conventional statements high. For test subjects who took the consumer’s point of view, there were no significant correlations between social values and the P-scores and N2-scores. This finding indicates that people recognize different moral norm sets as valid for different social roles and that those norms can be congruent in certain CC contexts.

The findings that are represented in Figure 2 and the survey data make two insights clearly visible. People recognize different moral norm sets to be of importance for different social roles and in different CC contexts. In access-based CC markets, the N2-scores of the provider stories are significantly higher than the scores of the consumer’s perspective. The pattern turns around in redistribution markets where the N2-scores of the consumer’s perspective are significantly higher than those from the provider’s point of view. This fact implies that subjects who read consumer point of view dilemma stories find abstract moral and social arguments to be very relevant in redistribution CC contexts, while they prefer arguments in favor of the protagonists’ personal interests in access-based contexts. Slightly less striking, but still noticeable, is the difference between for-profit and non-profit contexts. The highest N2-scores from the consumer’s and the provider’s point of view were reached here. The reason for that might be that people have no personal (e.g., financial) but only social incentives to engage in non-profit CC markets.

Stability of Moral Choices

The priming behavior as well as the U-scores are indications of the stability of moral convictions in the decision-making process. A repeated measure ANOVA with the four different priming manipulations (no prime vs. personal interest prime vs. maintaining norms prime vs. post-conventional prime) and the comparison of the pre-prime N2-scores and the post-prime N2-scores as dependent variables showed no significant effects. The priming statements, thus, had no significant effect on the N2-scores, irrespective of the stories’ point of view. A 2 (consumer vs. provider) × 4 (for-profit and non-profit access-based CC contexts vs. for-profit and non-profit redistribution CC contexts) ANOVA was used to test the validity of the U-score more carefully. This analysis revealed a significant effect of the CC contexts on the U-score (F[3, 2,728] = 6.932, p ≤ .001), with specifically high U-scores for stories from the non-profit contexts (Table 10). Although the effect is rather weak (R² = .014), this result supports the thesis that subjects in non-profit CC contexts are driven by motives that correspond to the higher moral development Stages 4 to 6.

Table 10.

Results of the ANOVA.

	Dependent variable: U-Score
Factor	df	Mean squares	F	p
Corrected model	7	.173	5.337	***
Constant	1	958.633	29,625.404	***
Role	1	.325	10.058	**
CC context	3	.224	6.932	***
Interaction
Role × CC context	3	.057	1.775	n.s.
Error	2,728	.032
R²	.014
R²_adjusted	.011

Note. CC = collaborative consumption.

Two-way ANOVA; dependent variable = U-score. n.s. p > .10; ^†p ≤ .10. * p ≤ .05. ** p ≤ .01. *** p ≤ .001.

Discussion

The DIT2’s adaptation for CC markets produces valid P-scores, N2-scores, and U-scores, the results are consistent with the values of other DIT2 studies and they are also confirmed by the supplementary survey that featured variables established in CC research. The only deviations noticeable are a slightly higher variation of the average P-score and N2-score. As known from moral reasoning research in various business contexts, this deviation is due to the real-life nature of the dilemma stories, with people relating easier to those stories than to Rest’s hypothetical life-or-death scenarios and due to only subjects doing the test that take the same social role as the stories protagonists (Doyle et al., 2009; Weber, 2018; Weber & Wasieleski, 2001). The results verify the adapted DIT2 as a sound method for analyzing consumers’ ability to cognitively and intuitively understand moral arguments in CC contexts (RQ1).

The adapted DIT2 reveals new insights into the entanglements of CC markets, social roles and moral arguments. Until now, it is known that consumers in non-profit contexts tend to recognize social motives as more important than motives of personal interest. The adapted DIT2 confirms these findings and, moreover, it shows that for every CC context consumers consider moral arguments of different stages of moral development to be relevant (RQ2). As Table 11 summarizes, collaborative consumers in redistribution contexts are more sensitive to maintaining norms and post-conventional arguments than consumers in access-based contexts. In contrast, consumers in access-based contexts give high ratings and rankings to arguments that advocate a maintaining norm (or even a self-interest) rationale and they are not sensitive to post-conventional motives. These results indicate that, first, consumers engage in access-based contexts to pursue individual rather than societal or ecological benefits and, second, that consumers who are engaged in redistribution contexts are driven by higher developed moral motives. In line with this, collaborative consumers have proven to be sensitive to the provider’s point of view (RQ3). All these role-related and context-related findings are new to the consumer ethics discourse.

Table 11.

Stages of Moral Development that are Considered to be Relevant in CC Contexts.

Social role	Redistribution		Access-based CC
Social role	For-profit	Non-profit	For-profit	Non-profit
Stage of moral developmentConsumer	Maintaining norms and post-conventional	Maintaining norms and post-conventional	Personal interest	Personal interest and maintaining norms
Stage of moral developmentProvider	Personal interest and maintaining norms	Maintaining norms and post-conventional	Personal interest and maintaining norms	Post-conventional

Note. CC = collaborative consumption.

Although the experiment cannot be used to model or predict the actual behavior of collaborative consumers, the analysis of the priming test and the U-score have shown that the collaborative consumer’s decision on how the stories’ protagonists should act as well as their rating decisions are stable (RQ4). The highest U-scores were reached in non-profit CC contexts, which indicate that people align their motives and intentions more easily when no financial incentives are involved.

Implications

As with no method before, the adapted DIT2 makes it possible to reveal exactly which line of normative argumentation people in different CC contexts consider to be relevant. This methodological asset is important for at least two reasons. First, policy makers and companies can use the results in Table 11 to promote CC practices. This can be achieved by using normative arguments that their target audience can understand. For example, in access-based and for-profit CC contexts where consumers are highly driven by motives of personal interest, it appears pointless to advertise the social impact of sharing products on the local community. Most people’s reason to use AirBnB is that it provides accommodations cheap and easy and not because they hope to support locals. Conversely, it makes sense to outline the ecologic and social benefits when marketing Couchsurfing. The results can also help successfully promote alternative behaviors. For example, messages that entertain motives of personal interest could convince some AirBnB customers to use Couchsurfing.

Second, the study changes how research will recognize the normative motives of collaborative consumers. Collaborative consumers are neither driven exclusively by hedonistic nor by altruistic motives. Instead, they put complex moral arguments in perspective to their social reality. Compared with other social contexts, it is easy for people in CC contexts to adopt the social perspective and the social role of other collaborative agents and, therefore, to consider their motives and needs. This explains why consumers in intense collaborative contexts like foodsharing are more sensitive to the provider’s point of view than people who use AirBnB.

Limitations

Even though a large number of subjects participated in the experiment, the sample is not representative. Particularly, the average age was rather low and the number of students was high. The sample also does not provide any information on possible cultural differences in the collaborative consumer’s moral reasoning. Cultural variations are likely (Hofstede et al., 2010) and should be addressed in future research. In particular, the moral reasoning of consumers from individualistic societies might differ from those of collective societies regarding CC dilemma situations. Furthermore, the U-score only considers the subject’s recommended action for various everyday consumption contexts; we did not observe de facto moral behavior of collaborative consumers in this study. Consequently, no conclusions with regard to the currency of ethical purchasing behavior can be drawn from this experiment.

Conclusion

Consumption dilemmas are neither a rare phenomenon nor unique to CC contexts. In everyday consumption, people understand and support different and often incompatible moral arguments. Allowing consumers to comment on dilemma scenarios can uncover the moral values and arguments that intuitively or deliberately are most important to them in a given context, and it also indicates which social role might be of importance to the decision-making process. With the adapted DIT2, researchers have the opportunity to analyze the moral rationale and intuition of people in common consumption contexts in which they take the social role of a consumer and many other roles as well.

Supplemental Material

sj-docx-1-bas-10.1177_00076503231157696 – Supplemental material for Collaborative Consumers Can Be Ethical Consumers: Adapting the Defining Issues Test to Understand Ethical Reasoning in Collaborative Consumption Markets

Supplemental material, sj-docx-1-bas-10.1177_00076503231157696 for Collaborative Consumers Can Be Ethical Consumers: Adapting the Defining Issues Test to Understand Ethical Reasoning in Collaborative Consumption Markets by Sebastian Müller, Nils Christian Hoffmann, Ludger Heidbrink and Stefan Hoffmann in Business & Society

Footnotes

Appendix

Appendix.

Measures and Scales of the Complementary Survey.

Construct	Items	M	SD	λ	Source(s)
Functional value (α = .636)	By using sharing, I usually get a good deal.	5.79	1.03	.805	Adapted from Lamberton & Rose (2012); Hamari et al. (2016)
	I use sharing to save money.	5.94	1.16	.819
	Sharing has many personal benefits for me, e.g., easier access to the products and services offered.	5.38	1.27	.667
Social value (α = .877)	Sharing is a great way to meet new people.	3.97	1.69	.906	Adapted from Lamberton & Rose (2012); Tussyadiah & Pesonen (2016)
	Sharing allows me to meet like-minded people.	4.01	1.59	.942
	Sharing makes me feel part of a group of like-minded people.	3.52	1.63	.841
Moral value (α = .824)	I use sharing to help others.	4.18	1.39	.726	Bucher et al. (2016)
	I use sharing because it is good for the environment	5.39	1.43	.851
	I use sharing because it feels right.	5.23	1.35	.857
	Sharing allows me to do something meaningful.	3.99	1.45	.801
Sustainability attitude (α = .904)	By using sharing services, I can reduce the consumption of energy and other resources	5.70	1.21	.907	Adapted from Lamberton & Rose (2012); Möhlmann (2015); Hamari et al. (2016)
	By using sharing services, I can be a more socially responsible consumer.	5.49	1.22	.913
	By using sharing services, I can reduce the negative impact of consumption on the environment.	5.64	1.27	.928
	It is important that others like the sharing services and platforms I use.	2.60	1.45	.853

Acknowledgements

We thank the anonymous reviewers as well as the editor Prof. Manika for their great support.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was funded by the German Federal Ministry of Justice and Consumer Protection (Funding Number: 28V12007).

ORCID iD

Sebastian Müller

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Sebastian Müller (PhD, Kiel University) is a senior researcher at the Center for Life Ethics at Bonn University, Germany. His research focuses on consumer ethics, social ontology, philosophy of economics, and public health ethics. His articles have appeared in such journals as BMC Medical Ethics, Business and Society Review, and Public Health Ethics.

Nils Christian Hoffmann (PhD, Kiel University) was a postdoc at the department of marketing at Kiel University, Germany, and at the Kiel Centre for Globalization (KCG). His research focuses on the ethics of consumption, ethical and sustainable consumer behavior as well as cross-cultural marketing. His work appeared in academic journals, such as Journal of Business Ethics, Journal of Marketing Management, and Management International Review.

Ludger Heidbrink (PhD, University of Hamburg) is a professor of Practical Philosophy at Kiel University (CAU), Germany. His special research areas are consumer ethics, business ethics, and philosophy of economics. He has published more than 15 books and articles in academic journals, such as Journal of Business Ethics, Business and Society Review or Public Health Ethics.

Stefan Hoffmann (PhD, Technical University of Dresden) is a professor of Marketing at Kiel University (CAU), Germany. His research interests focus on consumer behavior, sustainable and ethical consumption, and digital marketing. He has published in several leading business and marketing journals, including the Journal of the Academy of Marketing Science, Journal of Business Ethics, Journal of Consumer Psychology, and Journal of Service Research.

References

Akbar

Hoffmann

(2022). Collaborative spaces: Framework for collaborative consumption and the sharing economy. Journal of Service Marketing. Advance online publication. https://doi.org/10.1108/JSM-03-2021-0078

Andorfer

V. A.

Liebe

(2012). Research on fair trade consumption—A review. Journal of Business Ethics, 106, 415–435.

Apaolaza

Hartman

D’Souza

López

C. M.

(2018). Eat organic—Feel good? The relationship between organic food consumption, health concern and subjective wellbeing. Food Quality and Preferences, 63, 51–62.

Auger

Devinney

T. M.

Louviere

J. J.

(2007). Using best–worst scaling methodology to investigate consumer ethical beliefs across countries. Journal of Business Ethics, 70, 299–326.

Balderjahn

Peyer

Seegebarth

Wiedmann

K.-P.

Weber

(2018). The many faces of sustainability-conscious consumers: A category-independent typology. Journal of Business Research, 91, 83–93.

Beldad

Hegner

(2018). Determinants of fair trade product purchase intention of Dutch consumers according to the extended theory of planned behaviour. Journal of Consumer Policy, 41, 191–210.

Belk

(2014). You are what you can access: Sharing and collaborative consumption online. Journal of Business Research, 67(8), 1595–1600.

Bendapudi

Leone

R. P.

(2018). Psychological implications of customer participation in co-production. Journal of Marketing, 67(1), 14–28.

Biddle

B. J.

(1986). Recent developments in role theory. Annual Review of Sociology, 12(1), 67–92.

10.

Böcker

Meelen

(2017). Sharing for people, planet or profit? Analysing motivations for intended sharing economy participation. Environmental Innovation and Societal Transitions, 23, 28–39.

11.

Bucher

Fieseler

Lutz

(2016). What’s mine is yours (for a nominal fee)—Exploring the spectrum of utilitarian to altruistic motives for Internet-mediated sharing. Computer in Human Behavior, 62, 316–326.

12.

Buczynski

(2013). Sharing is good. How to save money, time and resources through collaborative consumption. New Society Publishers.

13.

Carrington

M. J.

Neville

B. A.

Whitwell

G. J.

(2010). Why ethical consumers don’t walk their talk. Towards a framework for understanding the gap between the ethical purchase intentions and actual buying behaviour of ethically minded consumers. Journal of Business Ethics, 97, 139–158.

14.

Caruana

(2007). A sociological perspective of consumption morality. Journal of Consumer Behaviour, 6, 287–304.

15.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum.

16.

Couchsurfing. (2021). About us. Couchsurfing. https://about.couchsurfing.com/about/about-us/

17.

Cretser

G. A.

Lombardo

W. K.

Lombardo

Mathis

(1982). Reactions to men and women who cry: A study of sex differences in perceived societal attitudes versus personal attitudes. Perceptual and Motor Skills, 55, 479–486.

18.

d’Astous

Legendre

(2009). Understanding consumers’ ethical justifications: A scale for appraising consumers’ reasons for not behaving ethically. Journal of Business Ethics, 87, 255–268.

19.

Davies

I. A.

Gutsche

(2016). Consumer motivations for mainstream „ethical“ consumption. European Journal of Marketing, 50, 1326–1347.

20.

Dellaert

B. G. C.

(2019). The consumer production journey: Marketing to consumers as co-producers in the sharing economy. Journal of the Academy of Marketing Science, 47, 238–254.

21.

Delmas

M. A.

Gergaud

Lim

(2016). Does organic wine taste better? An analysis of experts’ ratings. Journal of Wine Economics, 11(3), 329–354.

22.

Donleavy

G. D.

(2008). No man’s land: Exploring the space between Gilligan and Kohlberg. Journal of Business Ethics, 80, 807–822.

23.

Doyle

Frecknall-Hughes

Summer

(2009). Research methods in taxation ethics: Developing the defining issues test (DIT) for a tax-specific scenario. Journal of Business Ethics, 88, 35–52.

24.

Eckhardt

G. M.

Bardhi

(2016). The relationship between access practices and economic systems. Journal of the Association for Consumer Research, 1, 210–225.

25.

Egloff

Schmukle

S. C.

(2002). Predictive validity of an implicit association test for assessing anxiety. Journal of Personality and Social Psychology, 83(6), 1441–1455.

26.

Elm

D. R.

Weber

(1994). Measuring moral judgment: The moral judgment interview or the defining issues test? Journal of Business Ethics, 13, 341–355.

27.

Enden

Boom

Brugman

Thoma

(2019). Stages of moral judgement development: Applying item response theory to Defining Issues Test data. Journal of Moral Education, 48(4), 423–438.

28.

Ertz

Durif

Arcand

(2019). A conceptual perspective on collaborative consumption. AMS Review, 9, 27–41.

29.

Frank

(2017). Me, my family or the public good? Do inter-role conflicts of consumer-citizens reduce their ethical consumption behavior? International Journal of Consumer Studies, 42(3), 306–315.

30.

Frenken

Schor

(2017). Putting the sharing economy into perspective. Environmental Innovation and Societal Transitions, 23, 3–10.

31.

Gibbs

J. C.

Basinger

K. S.

Fuller

(1992). Moral maturity: Measuring the development of sociomoral reflection. Lawrence Erlbaum.

32.

Gilligan

(1982). New maps of development: New visions of maturity. American Journal of Orthopsychiatry, 52(2), 199–212.

33.

Haidt

(2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108(4), 814–834.

34.

Hamari

Sjöklint

Ukkonen

(2016). The sharing economy: Why people participate in collaborative consumption. Journal of the Association for Information Science and Technology, 67, 2047–2059.

35.

Hofstede

G. J.

Minkov

(2010). Culture and organizations. Software of the mind (3rd ed.). McGraw-Hill.

36.

Iannone

(2021, March 16). Recommerce: Fueling economic opportunity and sustainability. eBay Press Room. https://www.ebayinc.com/stories/news/recommerce-fueling-economic-opportunity-and-sustainability/

37.

Jones

T. M.

(1991). Ethical decision making by individuals in organisations: An issue-contingent model. The Academy of Management Review, 16(2), 366–395.

38.

Kingston

(2021). Shopping with a conscience? The epistemic case for relinquishment over conscientious consumption. Business Ethics Quarterly, 31(2), 242–274.

39.

Kohlberg

(1958). The development of modes of moral thinking and choice in the years 10 to 16. University Press.

40.

Kohlberg

(1971). Stages of moral development as a basis for moral education. In Beck

C. M.

Crittenden

B. S.

Sullivan

E. V.

(Eds.), Moral education: Interdisciplinary approaches (pp. 23–92). University of Toronto Press.

41.

Kohlberg

(1981). Essays on moral development volume 1. The philosophy of moral development. Harper & Row.

42.

Kurtines

Greif

E. B.

(1974). The development of moral thought: Review and evaluation of Kohlberg’s approach. Psychological Bulletin, 81(8), 453–470.

43.

Lamberton

C. P.

Rose

R. L.

(2012). When is ours better than mine? A framework for understanding and altering participation in commercial sharing systems. Journal of Marketing, 76, 109–125.

44.

Lemerise

E. A.

Arsenio

W. F.

(2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 71(1), 107–118.

45.

Mair

Reischauer

(2017). Capturing the dynamics of the sharing economy: Institutional research on the plural forms and practices of sharing economy organizations. Technological Forecasting and Social Change, 125, 11–20.

46.

Möhlmann

(2015). Collaborative consumption: Determinants of satisfaction and the likelihood of using a sharing economy option again. Journal of Consumer Behaviour, 14, 193–207.

47.

Mudrack

P. E.

Mason

S. E.

(2019). Moral reasoning and its connections with Machiavellianism and authoritarianism: The critical roles of index choice and utilization. Business & Society, 58(4), 779–812.

48.

Müller

(2022). Die Grenzen des Konsums. Eine Verantwortungstheorie der Konsumentenrolle [The limits of consumption. A theory of consumer role responsibility]. Campus.

49.

Orazi

D. C.

Chan

E. Y.

(2020). “They did not walk the green talk!”: How information specificity influences consumer evaluations of disconfirmed environmental claims. Journal of Business Ethics, 163, 107–123.

50.

Puschmann

Alt

(2016). Sharing economy. Business & Information Systems Engineering, 58, 93–99.

51.

Rest

J. R.

(1979). Development in judging moral issues. University of Minnesota Press.

52.

Rest

J. R.

Narváez

Bebeau

M. J.

Thoma

S. J.

(1999). Postconventional moral thinking: A neo-Kohlbergian approach. Lawrence Erlbaum.

53.

Rest

J. R.

Narvaez

Thoma

S. J.

Bebeau

M. J.

(1999). DIT2: Devising and testing a revised instrument of moral judgment. Journal of Educational Psychology, 91(4), 644–659.

54.

Rest

J. R.

Thoma

S. J.

Narvaez

Bebeau

M. J.

(1997). Alchemy and beyond: Indexing the Defining Issues Test. Journal of Educational Psychology, 89(3), 498–507.

55.

Scaraboto

(2015). Selling, sharing, and everything in between: The hybrid economies of collaborative networks. Journal of Consumer Research, 42(1), 152–176.

56.

Schlaile

M. P.

Klein

Böck

(2018). From bounded morality to consumer social responsibility: A transdisciplinary approach to socially responsible consumption and its obstacles. Journal of Business Ethics, 149, 561–588.

57.

Schwarz

(1999). Self-reports: How the questions shape the answers. American Psychologist, 54(2), 93–105.

58.

Sent

E. M.

(2018). Rationality and bounded rationality: You can’t have one without the other. The European Journal of the History of Economic Thought, 25(6), 1370–1386.

59.

Septianto

Soegianto

(2017). Being moral and doing good to others. Re-examining the role of emotion, judgement, and identity on prosocial behavior. Marketing Intelligence and Planning, 35(2), 180–191.

60.

Simon

H. A.

(1997). Models of bounded rationality vol. 3: Empirically grounded economic reason. MIT Press.

61.

Skoe

E. E. A.

Eisenberg

Cumberland

(2002). The role of reported emotion in real-life and hypothetical moral dilemmas. Personality and Social Psychology Bulletin, 28(7), 962–973.

62.

Soper

(2007). Re-thinking the ‘Good Life’: The citizenship dimension of consumer disaffection with consumerism. Journal of Consumer Culture, 7(2), 205–229.

63.

Spilski

Gröppel-Klein

Gierl

(2018). Avoiding pitfalls in experimental research in marketing. Marketing ZFP, 40(2), 58–90.

64.

Sunstein

C. R.

(1996). Social norms and social roles. Columbia Law Review, 96, 903–968.

65.

Tanner

R. J.

Carlson

K. A.

(2009). Unrealistically optimistic consumers: A selective hypothesis testing account for optimism in predictions of future behavior. Journal of Consumer Research, 35(5), 810–822.

66.

Thoma

S. J.

(2006). Research on the defining issues test. In Killen

Smetana

J. G.

(Eds.), Handbook of moral development (pp. 67–909). Lawrence Erlbaum.

67.

Thoma

S. J.

Rest

Davison

(1991). Describing and testing a moderator of the moral judgment and action relationship. Journal of Personality and Social Psychology, 61(4), 659–669.

68.

Timmers

Fischer

Manstead

(2003). Ability versus vulnerability: Beliefs about men’s and women’s emotional behaviour. Cognition and Emotion, 17(1), 41–63.

69.

Tussyadiah

I. P.

Pesonen

(2016). Impacts of peer-to-peer accommodation use on travel patterns. Journal of Travel Research, 55(8), 1022–1040.

70.

Umweltbundesamt. (2022, March 18). Car-sharing. https://www.umweltbundesamt.de/themen/verkehr-laerm/nachhaltige-mobilitaet/car-sharing

71.

University of Leeds. (n.d.). Carpooling (car share) and car hire—Sustainability service. https://sustainability.leeds.ac.uk/sustainable-transport/car-share-2/

72.

Walsh

(2011, March 17). Today’s smart choice: Don’t own. Share. Time (online). http://content.time.com/time/specials/packages/article/0,28804,2059521_2059717_2059710,00.html

73.

Weber

(1990). Managers’ moral reasoning: Assessing their responses to three moral dilemmas. Human Relations, 43(7), 687–702.

74.

Weber

(2018). Does it matter how one assesses moral reasoning? Differences (Biases) in the recognition versus formulation tasks. Business & Society, 57(7), 1440–1464.

75.

Weber

McGivern

(2010). A new methodological approach for studying moral reasoning among managers in business settings. Journal of Business Ethics, 92, 149–166.

76.

Weber

Siniora

D. N.

(2022). Investigating cognitive moral reasoning: The effect of dilemma context and gender agreement between subject and the dilemma actor. Business and Society Review, 126(4), 455–478.

77.

Weber

Wasieleski

(2001). Investigating influences on managers’ moral reasoning. Business & Society, 40(1), 79–110.

78.

Wellman

H. M.

Liu

(2004). Scaling of theory-of-mind tasks. Child Development, 75(2), 523–541.

79.

Williams

(1981). Internal and external reasons. In Williams

(Ed.), Moral luck: Philosophical papers 1973-1980 (1st ed., pp. 101–113). Cambridge University Press.

80.

Iran-Nejad

Thoma

(2007). Administering Defining Issues Test online: Do response modes matter? Journal of Interactive Online Learning, 6(1), 10–26.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB