Abstract
Fairness is an important component of all marketing exchange. While previous literature has focused on companies’ fair actions toward consumers, this article examines fair actions of consumers toward companies. Through a series of experiments, the authors investigate consumer fairness in the context of a pay-what-you-want pricing scheme. Results show that some consumers act fairly toward companies, even if they have no obligation to do so. For such consumers, there seems to be a self-signaling motive, in that they want to signal to themselves that they are fair. The authors also show that the distribution of price paid has a similar pattern to that of the dictator game in behavioral economics. Finally, the authors show that consumers can be influenced into taking fair actions by providing cues about “socially correct” actions others are taking. The reason for this is that many consumers act unfairly not because of their inherent propensity but because they believe others are doing the same. Implications for distributive justice in marketing exchange are discussed.
Introduction
Dynamic economies are born out of healthy competition, and it is the role of government agencies to ensure that fair competition takes place. Fair competition is often legislated through fair trade and antitrust laws, and other times enforced by monitoring agencies. However, laws and monitoring can only go so far, and there will be areas where unfair practices may still be within legal boundaries. In such instances, it is important to understand what makes companies act fairly, especially with regard to how companies exercise their bargaining power to extract surplus from the exchange (i.e., the issue of distributive justice).
Studies have shown that unfair exchange arises from differences in bargaining power and information asymmetry between transactors (Adkins and Jae 2010; Alwitt 1995; Mascarenhas, Kesavan, and Bernacchi 2008). Namely, the party with greater bargaining power and more information will force a solution in which it will obtain a larger proportion of surplus.
However, another area that is much less studied is the means to induce consumers, not firms, to behave fairly in their consumption practices. As consumers gain more knowledge, consumer power and knowledge increase, as do their tendency to take advantage of the companies (Rezabakhsh et al. 2006). For instance, some consumers free ride on the sales service of offline retailers by visiting them to obtain information about the product, with no intention of buying, and ordering the same product through an e-commerce sellers who may offer lower prices because they do not incur the cost of maintaining a store (Shin 2007). Therefore, if we require firms to behave fairly, so should consumers, since they are the reciprocal side of a transaction. The principle of reciprocity demands that fair action be mutual: if one party reneges on the goodwill, the ensuing equilibrium is for both to act unfairly. Restricting companies from unfair actions toward consumers would work in the beginning, but it will only be sustained if consumers reciprocate. Indeed, there are many instances where companies have changed their “generous” policies because of unethical consumers.
Wirtz (1998) and Chu, Gerstner, and Hess (1998) state that customer cheating and fraudulent invocation of guarantees are causes of firms’ reluctance to adopt unconditional guarantees and generous return policies. Also, imagine a pharmaceutical company deciding to sell its nonprescription medicines at cheaper prices to low-income families, while charging full price to others. This decision would contribute to distributive justice in the community in that it gives low-income families access to inexpensive medicine. However, if middle- or higher-income consumers gain access to lower priced pharmaceuticals either through false representation of their income or by rebuying from consumers who have bought them cheaply (i.e., arbitrage), then the pharmaceutical company may be forced to discontinue the price discrimination policy and charge full price to all consumers. Because of such reasons, many pharmaceutical companies provide medicine at lower prices only to consumers in underdeveloped countries, not to low-income consumers within the same country. This example shows that fair action on the part of consumers may be a key variable in the achievement of distributive justice in society. This article aims to shed light on the conditions under which consumers will act fairly toward companies, as well as the ways to encourage such actions.
The basic tenet of microeconomics, and especially agency theory, is that economic agents will act opportunistically when given the chance (i.e., moral hazard; Hölmstrom 1979). However, experimental game theory research showed that this may not always be true. Experiments conducted in behavioral game theory often show an outcome wherein the more powerful party leaves something on the table for the counterpart (i.e., see the outcome of dictator games in Camerer and Thaler 1995; Fehr and Gächter 2000; Rabin 1993). Further evidence of fair behavior is shown in the business strategy literature of interfirm trust, where it is shown that some original equipment manufacturers in a favorable bargaining position will not act opportunistically toward suppliers, even if given the chance to do so (Dyer and Chu 2000, 2003). These studies of individual–individual or firm–firm interaction show that economic agents may have fairness motive too. We delve a step further in this stream of research to see if fairness can be achieved in an individual–firm relationship.
An interesting marketing experiment by a British rock band, Radiohead, in 2007 allows us to observe how consumers act in reality when they are given power over sellers. The band released its new album online and let consumers download it by paying whatever amount they wished, a pay-what-you-want (PWYW) pricing scheme. Although observers worried about the feasibility of this pricing scheme, the idea was quite successful: approximately 1 million people downloaded In Rainbows in the first month of its release. While 62 percent of these consumers paid nothing at all, 38 percent paid more than zero, and 12 percent paid the full price of $8–$12. If consumers always take full advantage of their opportunities, this type of pricing scheme can never be successful.
Also, various types of donations (e.g., street musicians, the Metropolitan Museum of Art in New York, university alumni fund drives) represent forms of PWYW pricing, since donation amount cannot be forced. Panera Bread’s café in suburban St. Louis, Clayton restaurant, for instance, combined PWYW with donation, where some revenue goes to charity, and the store has been operating successfully. 1 Gneezy et al. (2010) demonstrate that the profitability of PWYW is higher when it is combined with a charity component than without it. In their field experiments with PWYW buffet restaurants, theater tickets, and warm beverages, Kim, Natter, and Spann (2009) found that of 1,452 people questioned, everyone paid something. These behavioral observations have sparked academic interest in PWYW (e.g., Chen, Koenigsberg, and Zhang 2010; Elberse and Bergsman 2008; Gautier and van der Klaauw 2012; Gneezy et al. 2010; Kim, Natter, and Spann 2009; Regner and Barria 2009).
Despite the success of PWYW, the theoretical understanding of the underlying mechanism is still limited. Why PWYW works is an increasingly significant issue, not only because it has been successful but also because it helps all stakeholders in the market understand consumers’ fair behavior. Why would consumers pay more than what they have to pay? In this article, we investigate why consumers behave fairly toward companies and how to induce fair behavior in consumers. Theoretical investigation of the PWYW pricing scheme would increase the understanding of the fairness of consumers. Greater understanding about consumer fairness would change the thinking and actions of firms, governments, and civic groups, and these changes would hopefully result in higher levels of fairness and distributive justice in society.
Methods of Establishing Distributive Justice
In order for marketing exchange to develop into a relationship, distributive justice and fairness need to be in the exchange. Therefore, these two concepts are central to a long-term marketing relationship, and many articles have been written on them. 2 Prior marketing research on establishing distributive justice and fairness can be divided into three camps: (a) why an unfair exchange takes place, (b) whether parties act opportunistically when given the chance, and (c) how one party responds to the unfairness of the other. Some have explored the causes of imbalances between parties and have recommended corrective measures (Adkins and Jae 2010; Alwitt 1995; Mascarenhas, Kesavan, and Bernacchi 2008). Mascarenhas, Kesavan, and Bernacchi (2008) state that buyer–seller information asymmetry creates structured injustices in the marketing system and suggest corrective action to rectify the injustices. Alwitt (1995) proposes that the poverty of consumers causes an inequity of exchange favoring marketers and suggests that buyers organize into groups and buy in large quantities. Marketplace imbalance may also result from language difficulties of consumers (Adkins and Jae 2010).
Others have conducted research on how to prevent parties from taking advantage of their greater power and have encouraged them to make fair decisions (Crul and Zinkhan 2008; Ferrell and Ferrell 2008; Nevin, Hunt, and Levas 1981). Crul and Zinkhan (2008) examine stakeholder theory as the most appropriate firm perspective for the fair treatment of all stakeholders as a foundation of distributive justice. Ferrell and Ferrell (2008) suggest that improvement in distributive justice can be achieved by adopting a stakeholder orientation within the company. Nevin, Hunt, and Levas (1981) even propose legislation preventing unfair practices in the franchising context. Finally, some researchers have investigated how weaker parties respond to the unfair actions of powerful parties. In marketing research, these aspects were actively explored in the contexts of supply chain management, service failure, and price fairness, as we show below.
Companies’ Unfair Behaviors and Consumers’ Responses
Starting with the seminal work by Kahneman, Knetsch, and Thaler (1986), economists and marketing scholars have explored fairness as an economic concept in the exchange between consumers and companies (Bolton, Warlop, and Alba 2003; Camerer and Thaler 1995, Campbell 1999, 2007; Rabin 1993; Xia, Monroe, and Cox 2004). One of the conclusions of this research is that, in the long run, companies benefit by acting fairly toward consumers. Thus, companies cannot increase prices to the maximum extent, regardless of overdemand during peak season, because of the perceived unfairness of such actions (Kahneman, Knetsch, and Thaler 1986). Consumers are less likely to do business with unethical sellers (Whalen, Pitts, and Wong 1991). This propensity leads consumers to play an active role in fair trade even though the results of an exchange between weak suppliers and large corporations are not related directly to consumer welfare. In fact, consumers have forced companies to behave fairly by preferring to buy fairly traded products, sometimes even paying a premium (Loureiroa and Lotade 2005).
Most prior research on fairness in marketing focuses on corporate decision making. It is usually assumed that companies are the causes of unfair exchanges, while consumers are powerless victims. Research on consumers’ fairness perception focuses mainly on how consumers perceive fairness of firms and react to it. For example, Goodwin and Ross (1990) studied how consumers perceive inequity and make complaints to management when there is a service failure. Moreover, research on how to enhance fairness in a marketing system suggested firm’s adaptation of the stakeholder orientation as a remedy for the firms’ unfair behavior (Ferrell and Ferrell 2008).
Consumers’ Unfair Behavior
Only few articles have dealt with fairness aspects of consumers’ behavior by observing shoplifting and questionable product returns. Alas, the results are not so favorable to the consumer. Several research suggest that one-third of consumers provided fraudulent product returns after using the product (King and Dennis 2003; Piron and Young 2000; Reynolds and Harris 2005) and a significant proportion of consumers have shoplifted at least once (Cox, Cox, and Moschis 1990; Klemke 1982; Kraut 1976).
The fact that exchanges are made in the company–consumer relationship and not in the individual–individual relationship can cause consumers to act less fairly. Fiske’s (1992) identification of different forms of social relations might be illustrative. For example, in an equally matching relationship, each party is entitled to an equal amount, whereas in a market-price relationship, allocations are based on free market principles (McGraw, Tetlock, and Kristel 2003). It has also been found that people tend to rationalize their dishonest behavior by citing companies’ immoral and deceptive practices, so much so that the companies “deserve” unfair treatment (Wilkes 1978). Further, people also feel less guilt when they steal from large and impersonal organizations (Cox, Cox, and Moschis 1990; Smigel 1956). Kahneman, Knetsch, and Thaler (1986) confirm that people tend to apply different concepts of fairness, depending on whether they transact with friends or strangers: with friends, they adopt motives affiliated with not-making-a-profit, whereas with strangers, the market price motive applies. Thus, people appear more likely to engage in unfair behavior when they transact with strangers or companies. These empirical observations seem to support the prediction that consumers will not act fairly toward companies.
From this perspective, it seems reasonable that many economists and others in the market were suspicious of the feasibility of the PWYW pricing scheme when it was introduced by Radiohead (Porter 2007). However, consumers were not always selfish under this pricing scheme. Thus, we see that they have a duality, and the interesting question is under what conditions the benevolent (or dishonest) side will emerge. Fairness research in behavioral economics might help us to understand these conditions.
Consumers and Fairness
Research in behavioral economics has explored the fairness of people mostly in individual to individual exchange context. Individuals in behavioral games seem to act fairy. Experimental evidence from behavioral economics regarding agents’ other-regarding preferences—such as fairness, altruism, and reciprocity—in the context of dictator games (Camerer and Thaler 1995; Fehr and Gächter 2000; Rabin 1993) show that players will donate 20–30 percent of an endowment to the anonymous recipient (Camerer 1997). Even when participants in the experiment were given titles as seller and buyer, instead of dictator and recipient, only 20 percent of sellers gave zero to the buyer, and about 10 percent of sellers gave half of their endowment to buyers (Hoffman et al. 1994).
The research on dictator games gives more insight into our PWYW research, given that dictator games and PWYW are similar. 3 In both PWYW and dictator games, the consumer/dictator does not have to leave any surplus for the company/recipient. In this sense, PWYW seems like a dictator game couched in a market transaction context.
As in the dictator game, where the dictator makes an offer, in the PWYW experiment, the consumer chooses her price. Also, in either the dictator or PWYW experiment, one that makes the offer is revealing her propensity for fairness. By contrast in a normal pricing situation, the company makes the offer, which the consumer can accept or refuse. In essence, companies play the role of pie cutter, and consumers the pie chooser. Pie cutters reveal their propensity for fairness through their division of the pie, while pie choosers’ decision is binomial: accept or reject. In this case, it is easier to comment on the fairness of the pie cutter, while it is difficult to say anything about the fairness of the pie chooser. Now in PWYW, consumers who used to be pie choosers become pie cutters, so the issue of consumers’ fair action becomes relevant. In general negotiation conditions, even though a party makes an offer that appears unfair, it will not be the final outcome because the other party has the power to make a counter offer. Therefore, an apparently unfair offer cannot be justified as a strategic tactic in a PWYW since the game ends right there. Some dictators choose not to play the dictator game even though such a decision leads to a lower payoff (Dana, Cain, and Dawes 2006), implying that the need to appear fair to others is a burden that they want to avoid.
We hypothesize that the aforementioned feature of PWYW leads consumers to pay certain amounts of money, which is motivated by their need to appear fair. Indeed, many studies of dictator games demonstrate that dictators pay more than zero because of their fairness motives. In some cases, the motive to signal their fairness to others will induce players to give up as much as half of their entitlement (Andreoni and Bernheim 2009). Other studies have also shown that some consumers want to pay higher prices for prosocial products since it signals greater social responsibility (Gneezy et al. 2010). Such costly prosocial behaviors serve as a signal of prosocial identity to the self, leading to an increase in subsequent prosocial behaviors (Gneezy et al. 2012). Therefore, the motive to signal fairness will lead to higher payments in PWYW. We also note that PWYW worked in Radiohead’s case where consumers’ payments were not observable, which implies that observability is not a necessary condition for the feasibility of PWYW. Greenberg (1983) argues that allocators’ adherence to distributive justice is due to their concern for the impressions they make on themselves (self-image) and on others (impression management). Bodner and Prelec (1995) go even further, suggesting that people might not even know their own true selves and might therefore engage in charitable behavior to prove to themselves that they are benevolent (i.e., self-signaling). Based on these prior studies and on the Radiohead case, we can assume that the self-signaling motive to appear fair is one of the underlying mechanisms of PWYW.
PWYW and Social Norm Conformity
Since markets are embedded in a larger social system, social norms may determine consumer behavior (Varman and Costa 2008). Cialdini, Reno, and Kallgren (1990) postulate that two social norms are at play: descriptive norms (i.e., what most people do) and injunctive norms (i.e., what ought to be done). The inferred social norm in the PWYW context is the descriptive norm. People initially act according to the injunctive norm, but if they find that others have a lower standard for fairness, these well-intentioned consumers will regress along the slippery slope of unfair behavior (Cialdini, Reno, and Kallgren 1990). Bicchieri and Xiao (2009) also show that when the two norms are in conflict, the descriptive norm usually exerts the greater influence. The title of Bicchieri and Xiao’s (2009) article “Do the Right Thing: But Only if Others Do So” explains this phenomenon succinctly. Goldstein, Cialdini, and Griskevicius (2008) show that providing descriptive norms (e.g., the majority of guests reuse their towels) to customers of a hotel results in a significantly higher rate of towel reuse in field tests.
Overview of Experiments
We examined the two main hypotheses concerning the fairness of consumers toward companies in experiments 1–5. Experiments 1, 2a, and 2b were conducted to investigate whether there is a group of consumers who offer more than zero price and to see if this price offer was dependent on the consumers’ awareness of the sellers’ cost in making the products. Experiments 3–5 were conducted to test the effects of social norms on consumer fairness. All of the experiments were conducted to ensure participants’ anonymity, which would exclude the effects of impression management motives. Even though PWYW had already been adopted by some restaurants, Radiohead’s success was surprising because consumers could pay in private, whereas customers of restaurants pay in public. Ensuring anonymity in our experiments helped us to understand the underlying mechanism of the PWYW scheme.
Experiment 1 was designed for two purposes. The first goal of experiment 1 was to compare outcomes of the dictator game and PWYW to determine whether the two generate similar patterns. Second, we aimed to observe how fairly consumers act toward companies. We knew from the Radiohead case that consumers were not fully selfish under PWYW, but we did not know to what extent. In a dictator game, it is easy to measure the dictator’s fairness because the allocation of the endowment is apparent (e.g., 50–50), but in a PWYW, the consumer’s valuation for the product is private information, so we do not know if a consumer who paid $10 is being greedy or generous. Thus, we asked participants about their willingness to pay (WTP) and used a ratio of the price paid to the WTP as a measure of fairness. 4 We conducted a PWYW experiment with four different products (recording album, mobile phone, cake, and DVD title) to exclude the explanation that Radiohead’s success was because of its fervent fans, not the characteristics of PWYW.
In experiment 2a, we examined whether consumers pay more when they are given cost information, which would induce consumers to think about fairness from the sellers’ perspective. We took note of the fact that cost is a central concept in fairness evaluations and cueing the seller’s costs reduces consumers’ perceptions of price unfairness (Bolton, Warlop, and Alba 2003). In equity theory, equity is defined as the ratio of outcome to the input of all parties (Cook and Hegtvedt 1983). In a consumer–company exchange, the price is the consumer’s input and the company’s outcome, the product/service is the consumer’s outcome, and the cost is the company’s input. (Searching time and effort can also be the consumer’s input, but price is the most common element in all cases.) Price and product/service are relatively salient compared to cost. In many instances, consumers accept a company’s price increase with less resistance if that increase reflects a cost increase, rather than the company taking advantage of increased demand (Kahneman, Knetsch, and Thaler 1986). People also take others’ losses into consideration (Gneezy 2005) when making a decision. In a dictator game, dictators who believed that recipients were entitled to receive a positive amount allocated more positive sum to recipients (Hoffman et al. 1994; List 2007). In a PWYW setting, we assume that a fair consumer is more inclined to pay at least what it costs the seller to manufacture the product, such that the price paid/WTP ratio should increase. Just providing cost information can make consumers more aware of the existence of the other party. The salience of the counterpart might lead them to think more about the effects of their own behaviors on others’ gain and loss. Thus, in experiment 2a, we hypothesize the following.
Hypothesis 1: Compared with a control group, a group that receives cost information has a higher price paid/WTP ratio.
In experiment 2b, we tested whether increase in payment is motivated by the signaling motive or intrinsic fairness motive. We did this by varying the amount of cost and observing subsequent price paid. We tested how consumers responded differently to cost information when the information was higher or lower than they expected. Cost information would act as a criterion for buyers’ monetary burden as well as criterion for companies’ profit. If the behavior is driven by the consumer’s desire not to want to cause companies to lose money, consumers would pay more or at least the amount equivalent to the cost. In this case, payments of consumers would increase monotonically as costs increase. In contrast, if the signaling motive is the main driver, consumers would be willing to pay more when costs are lower because it is easier to look fair with only a small payment. Consumers would withdraw their payments at a certain moment when they judge that the payments are not effective in contributing to their self-image. Thus, in experiment 2b, we predict that the signaling motive drives the payment more and hypothesize the following.
Hypothesis 2A: The group with cost information that is higher than their expectation has a lower price paid/WTP ratio than the group that receives lower cost information.
Hypothesis 2B: The group with cost information that is equal to or is lower than their expectation has a higher price paid/WTP ratio than the group that receives no cost information.
In experiments 3–5, we investigated the effect of descriptive norm on the payment of consumers. If consumers are influenced by how many others act fairly, their payments would vary depending on what others are doing. Experiment 3 examined how consumers react to the information of the descriptive norm. In experiment 3, we predict the following.
Hypothesis 3: The group with information about the cost and the unfair descriptive norm (i.e., majority pay nothing) has a lower price paid/WTP ratio than the group that receives only cost information.
Experiment 4 also tests the effect of the descriptive norm. We replace cost information with “fair” price and test the effect of two different descriptive norms, one describing fair behavior and the other unfair behavior. We hypothesize the following.
Hypothesis 4A: Consumers who observe the “fair” price and are informed that most people pay nothing have a lower price paid/WTP ratio than consumers who only observe the “fair” price.
Hypothesis 4B: Consumers who observe the “fair” price and are informed that most people pay the fair price have a higher price paid/WTP ratio than consumers who only observe the “fair” price.
Finally, in experiment 5, we expand our experiments to a field setting. We tested the effect of descriptive norm on consumers’ fairness in a real transaction setting, selling coffee to adults in a park.
Experiment 1: Measuring the Degree of Fairness
Procedure
We conducted a PWYW experiment of four products (recording album, mobile phone, cake, and DVD title) with seventy undergraduate students at a Korean university. We included mobile phones to see whether PWYW works for expensive products as well, and we included cakes because restaurants are adopting PWYW more actively than other businesses. The order of four products was counterbalanced. The total number of observations was 280 (4 products × 70 students), but 6 observations with zero WTP were excluded from the analysis.
We were interested in gauging the value of the products to the students and how much they would pay. Therefore, we asked them to indicate their WTP, a common metric used to measure the valuation of the product. Next, we told participants about the PWYW scheme and asked them what they would actually pay.
Results and discussion
To measure the proportion of the value of the product that each consumer would pay, we used a ratio of the price paid to their valuation (i.e., price paid/WTP). The results across all respondents, as we show in Figure 1, indicate that approximately 5–15 percent of respondents would pay the full value, even for mobile phones, which were quite expensive.
Ratio of price paid/willingness to pay (WTP; observation = 274).
Asking for payment under a PWYW scheme after asking about WTP may cause a halo effect. Since participants indicated their WTP first, they are likely to state higher prices than the honest prices in their minds. Thus, we conducted an additional experiment to investigate whether any halo effect influenced the results. We compared payment when asking WTP first and payment without asking WTP. If the halo effect has influence, prices paid in two conditions will be different. Sixty graduate students participated in this posttest for the return of the chocolate bar. Participants were assigned to one of the two conditions (asking WTP vs. not asking WTP). We used coffee as the product that participants had to pay for, which was used in experiment 2a and 2b as well. The results demonstrate that there is no significant difference between the “asking WTP” condition (M = $23.00) and “not asking WTP” condition, M = $22.16; t(58) = 0.255, p = .8. It was shown that asking WTP first does not significantly increase the price paid.
From Figure 1, we can draw several conclusions. First, separate distributions for the four products indicate a consistent pattern. Second, the 100 percent price paid/WTP ratio for some participants indicates that they were proactive in signaling their fairness. Third, the curve decreases continuously until the 80 percent ratio, reaches zero at 90 percent, and sharply increases at 100 percent, which is a typical outcome of a dictator game. In the dictator game, the bimodal points are 0 percent and 50 percent, representing complete greed and complete fairness, respectively. In PWYW, the corresponding points are 0 percent and 100 percent. Further, distribution in the dictator game shows very few respondents at 40 percent and then a sharp increase at 50 percent (Andreoni and Bernheim 2009; Forsythe et al. 1994).
Andreoni and Bernheim (2009) argue that the agglomeration of offers at 50 percent indicates players’ penchant to signal their fairness to others. Even those who want to offer less than 50 percent, if not for the signaling effect, persuade themselves to offer 50 percent, resulting in a pooling equilibrium. Similarly, consumers who would prefer to pay 90 percent of the value of the product in the PWYW setting instead pay the extra 10 percent to signal to others that they are fair. As a result, we find no respondents at 90 percent and a peak at 100 percent. Our results suggest that such signaling behavior extends to consumer–company relationships.
We tested whether distribution of PWYW and dictator game are different statistically. Since only data from Forsythe et al. (1994) and Hoffman et al. (1994) 5 (N = 102) are reported in their articles, we used them as dictator game data to compare with our PWYW data. The results of the comparison demonstrate that the two distributions are not different (Mann–Whitney = .849). This finding suggests that the two are similar in their distribution (see Figure 2), though the finding cannot be generalized because a specific PWYW and the dictator game data were compared.
Distribution of price paid/willingness to pay (WTP) in pay-what-you-want (PWYW) and offer/endowment in dictator game.
Experiment 2a: Do Consumers Care About Companies Operating at a Loss?
As an extension, we investigate if consumers actually care about whether the selling company makes a profit from the transaction. If consumers know that their behavior will hurt the company, do they act differently? Take the example of a fire sale. Most people are happy about the great bargains that they get, but some might feel sad about profiting from someone else’s misfortune and choose not to participate. To examine this issue, we tested whether informing consumers of the cost of production increases their PWYW price.
Procedure
Sixty students participated, with thirty in the control and thirty in the experimental group. The former group viewed a recording album and a mobile phone, which were counterbalanced, and indicated their PWYW price (i.e., control group); the latter received the same stimulus but also had information about the cost of the products (i.e., cost provision group).
Results and discussion
The pooled results for both products, in Figure 3, indicate that the price paid/WTP (M = 60) in the cost provision group is higher than that in the control group, M = 24; t(116) = 5.34, p < .001. When we compare the numbers of zero-price buyers and 100 percent buyers, we find that the cost provision group exhibits its highest peak at 100 percent, whereas the control group peaks at zero. These differences are statistically significant,λχ 2 (1, N = 59) = 16.28, p < .001. Consumers who do not want to hurt the company pay much more, and cost information encourages them to exhibit more fairness to companies.
Distribution of price-paid/willingness to pay (WTP) ratio: Effect of cost information.
Experiment 2b: Do Consumers Pay to Look Fair?
The goal of experiment 2b was to test the effect of signaling motive on the price paid. We investigate how consumers responded to cost information when the information was higher or lower than they expected. If consumers regard paying under PWYW as means of looking fair, they might pay more when they know that the costs are lower than when the costs are higher than they expected. We collected cost estimation on mobile phones in prestudy and provided different cost information: lower, equal, and higher cost information than the average estimated cost drawn from the prestudy. Then the payments from three different conditions were compared.
Procedure
A prestudy was conducted to measure cost estimation. Ninety students participated and were asked to indicate their cost estimation of the Samsung Galaxy S2. They were instructed to include administrative costs and sales costs as well as the manufacturing cost. The average estimated cost was $472, and its standard deviation was about $212. Thus, $470 was used as the “equal” cost information, one standard deviation lower ($260) as “lower” information and one standard deviation higher cost ($680) as “higher” information in this experiment 2b.
Experiment 2b was designed the same as experiment 2a, except the use of the three categories of cost information provided. One hundred and twenty students participated and were assigned randomly to one of the four conditions (lower cost, equal cost, higher cost, and control conditions). The participants in the fourth control condition were not given any cost information. All participants were asked to indicate their WTP of the mobile phone and what they would actually pay under the PWYW scheme.
Results and discussion
Figure 4 presents all the price paid/WTP ratios of four conditions. As expected, the “lower” condition group (M = 73.18) is higher than the “higher” group, M = 48.48, t(55) = 3.010, p = .004. Also, the “equal” condition group, M = 72.00, t(58) = 3.702, p < .001, and the “lower” condition group, M = 73.18, t(56) = 4.480, p < .001, as well as are higher than the “control” group (M = 40.23). However, no significant difference is observed between the “higher” condition (M = 48.48) and the “control” condition, M = 40.23, t(57) = 0.896, p = .374, groups.
Mean ratio of price-paid/willingness to pay (WTP): Effect of varied cost information.
This finding suggests that consumers’ decisions are influenced by their signaling motives. They pay more only when they believe the additional payment will contribute to their appearance of fairness. If their payment decisions are only made by considering the company’s losses, they would have paid more when informed that the company’s costs were higher than they expected. However, the signaling motives exceed the caring motives. When appearing fair requires a significant investment, consumers pay less, even if doing so could hurt the company’s financial situation.
Experiment 3: Influence of Descriptive Norm (with Cost Information)
Experiment 3 tested the effect of the descriptive norm on consumer payment. We compare three groups with respect to their price paid/WTP ratio: a control group, a group informed about the cost, and a group that received information about both the cost and that the majority of people would pay nothing.
Procedure
Ninety students participated, in groups of thirty. The students considered a smart phone developed by Samsung and indicated how much they would pay in a PWYW scheme. The control group received no information, the second group received information on cost (i.e., cost provision group); the third group received information on cost and the following paragraph (i.e., norm-inconsistent group):
The Galaxy S phone costs $180 to Samsung (including manufacturing cost, selling cost, and administrative cost). Assume that recently Samsung implemented a pricing scheme where the consumer pays what he or she wants to pay. Under such a scheme, how much would you pay? Research shows that 72 percent of people intended to pay zero.
Results and discussion
In Figure 5, compared with the control group (M = 40), the cost provision group (M = 60) expressed a higher price paid/WTP ratio, t(58) = 2.31, p = .024, similar to our finding in experiment 2a. The norm-inconsistent group (M = 34) were influenced by the descriptive norm and therefore expressed a lower price paid/WTP than the cost provision group, M = 60; t(58) = 3.21, p = .002. Finally, we find an insignificant difference between the control (M = 40) and norm-inconsistent (M = 34) groups, t(58) = .60, p = .551, which implies that the increase in the price paid/WTP resulting from the provision of cost information is equivalent to the decrease in this ratio resulting from the unfair descriptive norm.
Distribution of price-paid/willingness to pay (WTP) ratio: Effect of descriptive norms with cost provision.
Distribution of price-paid/willingness to pay (WTP) ratio: Effect of descriptive norms with fair price stimulus.
Mean Amount of Price Paid: Effect of Descriptive Norms with Cost Provision in the Field
Our data thus suggest that, in a consumer–company context, the descriptive norm provides a better predictor of behavior than does the injunctive norm when the two are in conflict. We infer that consumers in the norm-inconsistent group initially were induced to price fairly, in response to the provision of cost information, but the subsequent information about unfair descriptive norm induced them to act more unfairly. Despite the prevalence of this kind of compromising behavior, there were still a few diehards in our sample who would insist on paying 100 percent.
Experiment 4: Influence of Descriptive Norm (Fair Price Presented)
In this experiment, we investigate whether we can induce people to pay a certain price by presenting it as “fair.” The procedure was the same as that of experiment 3 with two exceptions. First, we used “fair” price instead of cost. Second, we included a norm-consistent condition.
Procedure
Again, three groups of thirty college students are the respondents. In this study, we used Starbucks coffee, and the “fair” price that was presented to the respondents was $4∼$4.50. The first control group saw the coffee and the fair price and provided their PWYW price. The second group received the same stimulus, as well as the information that most people (92 percent) pay nothing (i.e., norm-inconsistent group). Finally, for the norm-consistent group, we provided the same stimulus as in the first group but indicated that most people (92%) pay an amount in the fair price range.
Results and discussion
In figure 6, compared with the control group (M = 72), the norm-inconsistent group (M = 46) expressed a lower price paid/WTP, t(58) = 3.28, p = .002, which indicated that people are affected by the descriptive norm. As we expected, the norm-consistent group (M = 70) expressed higher price paid/WTP than the norm-inconsistent group, M = 46; t(58) = 3.15, p = .003. Finally, the results for the control group were not significantly different from those for the norm-consistent group, M = 72 versus 70; t(58) = 0.25, p = .805. Therefore, the initial influence of an injunctive norm is not enhanced by a consistent descriptive norm; the descriptive norm is useful only for maintaining a consistent initial injunctive norm. When information on one type of norm (injunctive or descriptive) appears, expectations about both norms are affected (Bicchieri and Xiao 2009). This result indicates that it is critical for marketers or policy makers to set an appropriate initial standard and ensure that the descriptive norm does not go against it.
Experiment 5: Influence of Descriptive Norm (Field Setting)
Experiment 5 was designed to overcome some limitations of experiments 1–4 and enhance the external validity of the findings. In experiment 5, we conducted a field study in which adult consumers had to pay money in a real-world setting. The theoretical goal of this experiment was to examine the effect of social norms, as in experiments 3 and 4. We used a canned coffee drink as the product to be sold, as it is familiar to buyers, easy to sell outside, and one of the products used in previous experiments.
Procedure
Two hundred and eighty-nine adults bought the products. The field experiment was conducted for three days in a large park located in the center of the city where many ordinary people passed by. We sold the canned coffee in a market stall and explained to customers the PWYW pricing scheme and provided them with purchasing cost information, which was displayed on a board. To manipulate the social norm information, the boards shown to customers in both the inconsistent and the consistent condition featured the following paragraph, which also included the social norm information compared to the control condition group: The canned coffee drinks were purchased at $0.25. We did business here last week under the same pricing scheme. At that time, 72% of people [paid zero (inconsistent condition)/paid above the purchase cost (consistent condition)].
Customers were instructed to put the amount of money they wanted to pay into an envelope and put the envelope into the box located behind the sellers. This was done to ensure that customers would believe that the seller would not see their payment decision. We did not ask customers to indicate their WTP to enhance the realism of the field study.
Results and Discussion
As in experiment 4, the inconsistent group (M = $0.30) paid less than the norm-consistent group, M = $0.42, t(189) = 3.47, p < .001, and the control group, M = $0.37, t(191) = 2.05, p = .042, paid more (see figure 7). The norm-consistent group was not significantly different from the control group in its payments, M = $0.42 versus $0.37; t(192) = 1.21, p = .228. These patterns of results are similar to those in experiment 4. Our findings suggest that social norms have an effect on payment under PWYW in an actual transaction setting.
Discussion
This study attempted to examine consumer fairness through six experiments. The first three experiments explored the self-signaling motives of consumers. In experiment 1, we confirmed that PWYW and dictator game have similarities in their distributions and that consumers have motives to send signals of their fairness to themselves. Also, we measured consumer fairness and demonstrated the presence of consumers who are willing to act fairly toward companies by showing that some consumers would pay 100 percent of the value of the product. We gave consumers anonymity and perfect power in negotiation and assessed how they acted. Even in situations where they could exploit the other party without being observed by others, some consumers acted fairly. This result provides an alternative to the more cynical view of consumers put forth in prior research. In experiment 2a, we provided support for the existence of the fairness motive in consumers. Consumers paid more when they were provided with company’s cost information. The results of experiment 2b concluded that increases in the payment were driven more by signaling motives than by fairness motives. When consumers believe that they can increase fairness with only a small additional cost, they are willing to pay that additional cost. However, when they realize that the cost of appearing fair is higher, they tend to give up. Namely, consumers did not maintain the level of payment and dropped their payment. This behavior reveals that consumers paid more because they cared about their self-image more rather than the loss to the companies.
The last three experiments focused on consumers’ conformation to the descriptive norm. Experiments 3 and 4 demonstrated that consumers are influenced by the descriptive norm in the context of PWYW. Consumers paid more when they recognized that most of the other consumers responded fairly and they paid less when the opposite information was provided. Experiment 5 was a field experiment, which extended experiments 3 and 4 to a real purchase setting with real payment. Consistent results in payment were also observed in the field test.
Conclusion
In this research, we have attempted to understand the fairness motives of individuals when dealing with companies. By investigating PWYW, we found that self-signaling and norm conformity are the two principal underlying mechanisms of PWYW. Providing moderate cost information and a fair descriptive norm lead to more fair behavior by consumers. Our study also identified a proportion of fair-minded consumers who could be induced into fair action by appropriate guidelines.
The provision of cost information was used to trigger the injunctive norm by making the risk of the other party more salient. What we found was that the effect of cost provision was weakened when costs were too high or when an unfair descriptive norm was provided. That is, consumers wanted to pay more to not want to make the seller lose money, but when this cost was too high or when they found others paying a low price, consumers dropped their price: a case where the consumer may have started with good intentions but then regressed into unfair behavior by circumstances, as also noted by Cialdini, Reno, and Kallgren (1990). Thus, for a company to implement PWYW, it needs to heed careful attention to circumstances surrounding the consumer, even to the point of managing correct norm behavior.
Our research contributes to distributive justice by expanding the scope of research to the consumer side. Prior distributive justice research focuses on direct ways to attain fair practices of firms. However, such efforts are bound to fail if consumers do not reciprocate by fair behavior of their own. Thus, inducing distributive justice is really a joint responsibility of consumers and firms. Without this reciprocity, one party is bound to retreat into a defensive position. Thus, parallel improvement in stakeholders’ fairness can expedite achievement of distributive justice. In this sense, research on consumers’ fair behavior can be a critical component of increasing distributive justice in the total market system.
Our findings also have clear implications for companies that remain wary of providing fair offers to consumers such as generous refund or guarantee policies for fear of consumers’ opportunistic behavior. New understanding that consumers are less selfish than the companies perceive may allow companies to design new marketing policies, some as innovative as PWYW. If companies can utilize consumers’ signaling motives and induce the conformity of consumers to the descriptive norm, giving power to consumers can be more effective than controlling them through a strict marketing policy.
Finally, this research contributes to the macromarketing literature by building a bridge between behavioral economics and distributive justice research in macromarketing. Behavioral economics have been exploring individual’s fair behavior through behavioral games. They have found several conditions in which players behave more fairly or less fairly. Even though some contexts differ between the two fields, the findings from the literature on individuals’ fairness provide insights into understanding consumers’ fairness and distributive justice.
Footnotes
Acknowledgments
The authors thank Terrence Witkowski (editor), Alexander Nill (associate editor), and three anonymous reviewers for their insightful comments and guidance.
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 received no financial support for the research, authorship, and/or publication of this article.