When Payments Go Social: The Use of Person-to-Person Payment Methods Attenuates the Endowment Effect

Consumer-to-consumer e-commerce platforms allow people to buy and sell goods directly with other people. This market segment has grown rapidly in recent years, led by well-known brands such as Facebook Marketplace, eBay, Kickstarter, and Etsy. Craigslist alone had 277 billion monthly visits and generated $660 million in revenue in 2021 (AIM Group 2022). Early research shows that trust in the platform, and in the potential transaction partner, are key drivers of consumers’ intention to transact in this type of market (Leonard 2012; Strader and Ramaswami 2002). The current research extends these findings by examining how elements of the transaction, such as the payment method used, may also affect consumers’ pricing decisions in these contexts.

Most traditional large retailers (e.g., Amazon, Target) use a fixed pricing mechanism, in which the firm sets a price and consumers can choose to buy or not buy at that price (Ellickson and Misra 2008; Lee and Staelin 1997; Shankar and Bolton 2004). In contrast, many consumer-to-consumer e-commerce platforms rely on participative pricing mechanisms (Chandran and Morwitz 2005; Chernev 2006), which require consumers (as sellers) to decide the price at which they are willing to sell the product, and consumers (as buyers) to name the price they are willing to pay. This type of pricing decision is believed to evoke markedly different psychological processes compared with those that occur when consumers respond to a fixed price (Chandran and Morwitz 2005; Haws and Bearden 2006; Kim, Natter, and Spann 2009; Lee, Kim, and Fairhurst 2009), in some instances leading consumers to be more responsive to social contextual factors (Saccardo et al. 2021) such as the potential for prosocial benefits (Gneezy et al. 2010).

Understanding the unique considerations that affect these types of participative pricing decisions is of central interest to managers, as revenue for the consumer-to-consumer e-commerce segment is largely driven by the value of successful transactions on the platform, which in turn requires that buyers and sellers converge on a mutually agreeable price. However, decades of research show an asymmetry in pricing decisions known as the endowment effect, such that people typically demand more to sell a good than they are willing to pay to acquire the same good (Kahneman, Knetsch, and Thaler 1990; see Morewedge and Giblin [2015] for a review). In the marketplace, if the endowment effect obtains and buyers and sellers cannot converge on a mutually agreeable price, transactions are less likely to occur (Kahneman, Knetsch, and Thaler 1990). For example, Kahneman, Knetsch, and Thaler (1990) estimated that the endowment effect can decrease the transaction volume of a product by as much as 60%. Given the significance of participative pricing for both managers and theorists, the current research investigates distinct considerations related to this type of decision. We find that one novel influence on participative pricing decisions is the type of payment method used in the transaction.

Specifically, we examine consumers’ participative pricing decisions when they use P2P payments, compared with their decisions when they use more traditional payment methods such as cash, credit cards, and debit cards. P2P payments are electronic money transfers traditionally made from one person to another through a technological intermediary, such as Zelle, Venmo, or WeChat. P2P is one of the fastest-growing payment methods, with U.S. P2P mobile payment transaction value projected to reach $983.85 billion in 2022 (Broadbent 2022). P2P payments have been heavily marketed as a social method of paying peers (Alfiky 2019) and are commonly used in social contexts, such as roommates splitting rent or friends sharing the cost of a party (Stern 2014). Moreover, consumers can now use P2P payment methods to buy goods from traditional businesses (CenturyLink 2021) and are increasingly likely to use P2P methods when transacting with a service provider (Huang et al. 2020).

In the current research, we told participants that they would be buying or selling an item with a stranger and that the transaction would use either a traditional payment method or P2P payment. We then asked the participants to decide on an acceptable price at which they would be willing to buy or sell the item. We compare the discrepancy between buying and selling prices when consumers anticipate using traditional payments and when they anticipate using P2P payment. Across eight studies, the classic endowment effect emerges when consumers use traditional payment methods, such as cash, credit cards, or debit cards, with sellers demanding more to relinquish the good than buyers are willing to pay for the same good. However, when consumers anticipate using P2P payment for transactions, we find that sellers are willing to accept slightly lower prices, and buyers are willing to pay a bit more; as a result, the endowment effect is eliminated.

To illustrate the potential practical importance of these findings for managers, we conducted two market simulations using the data from our studies: one using pricing offers from the P2P conditions and one using pricing offers from the traditional payment method conditions. For each round we randomly drew a seller price and a buyer price, repeated 5,000 times. If the seller’s minimum selling price (i.e., willingness to accept, or WTA) was lower than or equal to the buyer's maximum amount they would pay for the item (i.e., willingness to pay, or WTP), it was counted as a mutually acceptable match, and a successful “trade” was recorded (see Web Appendix A). On average, 9% more of the potential transactions were successful in the P2P market than in the traditional payment method markets. Thus, one straightforward practical implication of this research is that by encouraging the use of P2P payments, managers of consumer-to-consumer e-commerce platforms may be able to increase transaction volume, potentially by as much as 9%, improving market efficiency and revenue.

To explain this pattern, we draw on social norm theory. The exchange of goods for payment can range from traditional purely economic transactions to more social transactions, such as buying or selling in a community market or with a personal connection (Mandel 2006). We propose that different behavioral norms have come to be associated with making participative pricing decisions in purely economic versus more social transactions. For example, in purely economic transactions people tend to make strongly self-interested pricing decisions (Miller 2001), such as a seller pricing a product as high as possible to extract maximum surplus, or a buyer making a lowball offer that may lead to disagreements with the seller (Mandel 2006). In contrast, the social transaction norm for pricing decisions is to be somewhat less selfish and more cooperative, such that sellers accept a bit less and buyers pay a bit more (Saccardo et al. 2021).

Building on a long tradition of research on social and behavioral norms, we suggest that payment methods can be a subtle cue about whether a given transaction is purely economic or more of a social transaction, which in turn can evoke associated norms for pricing behavior. Specifically, our central proposition is that compared with the use of a traditional payment method, paying with P2P subtly cues a social transaction context, activating the associated normative behavior of making somewhat more cooperative participative pricing offers. This occurs even though consumers are explicitly aware that they are trading with a person whom they do not know. As a result, when paying with P2P, sellers are willing to accept a bit less and buyers are willing to pay a bit more, attenuating the endowment effect.

In addition to these useful practical insights for managers, our findings advance theory in several domains. First, we add depth and nuance to the understanding of how consumers make pricing decisions, by showing that payment methods can influence participative pricing decisions. Moreover, by documenting that an orthogonal contextual factor such as the payment method can reduce the gap between selling and buying prices, we add nuance to the understanding of the endowment effect, a ubiquitous and important phenomenon.

Second, consumer researchers interested in payment methods have provided valuable insights into how payment methods (e.g., cash vs. credit cards) can affect consumer behavior, often by evoking different levels of “pain of paying.” However, to our knowledge, no research has specifically documented the unique psychological consequences of using P2P payment, the fastest-growing payment method in the marketplace (Greene and Stavins 2018). The current research begins to address this gap by identifying a new dimension on which payment methods can vary: the degree to which the payment method activates different social norms in a transaction.

Finally, prior research has provided mixed results on how communal relationship norms and actual interpersonal relationships affect the endowment effect (Aggarwal and Zhang 2006; Halpern 1997; Mandel 2006; McGraw, Tetlock, and Kristel 2003). The current findings offer an additional data point in this discussion by documenting that P2P payment leads to more cooperative pricing decisions and thus attenuates the endowment effect in a one-shot (and not communal) market exchange, when consumers explicitly report that they are transacting with strangers.

Conceptual Development

Traditional Versus Participative Pricing Decisions

In most traditional retail transactions, the retailer sets a fixed price, and consumers can choose to accept or reject the set price to buy or not buy the product. Thus, most pricing research focuses on how firms can determine the optimal price for a product and how consumers process, evaluate, and respond to a set price (Ellickson and Misra 2008; Lee and Staelin 1997; Shankar and Bolton 2004). However, with the rise of consumer-to-consumer e-commerce platforms such as Craigslist, eBay, and Facebook Marketplace, and with more retailers offering a “pay what you want” pricing scheme (Jung et al. 2014; Kim, Natter, and Spann 2009), the participative pricing mechanism has become increasingly common. This type of decision is defined as consumer participation in price determination, whereby consumers as sellers and buyers make decisions on the price at which they are willing to sell or buy a product (Chandran and Morwitz 2005; Kim, Natter, and Spann 2009).

Consumers making participative pricing decisions go through different psychological processes and respond differently to contextual information than when they face a fixed price. From a cognitive perspective, a participative pricing mechanism increases consumers’ perceived price fairness (Haws and Bearden 2006) and their implementation rather than evaluation mindset (Chandran and Morwitz 2005), increasing transaction intention. More important to this research, from a social perspective, consumers are more sensitive to social cues when engaging in participative pricing. For example, compared with a fixed-price offer, consumers engaged in participative pricing decisions (e.g., pay what you want) are willing to pay significantly more when a purchase has a prosocial component (Gneezy et al. 2010).

A different stream of research on the endowment effect has examined consumers’ pricing decisions extensively. The well-known finding is that consumers often exhibit a behavioral asymmetry such that sellers tend to demand more to relinquish a good than buyers would be willing to pay to acquire the same good (Kahneman, Knetsch, and Thaler 1990). For example, in the valuation experimental paradigm (Kahneman, Knetsch, and Thaler 1990; Morewedge and Giblin 2015), half of the participants are endowed with a good, such as a coffee mug, and told that they now own the good but have the chance to sell it. These “sellers” then indicate the lowest price they would accept to relinquish the item (i.e., WTA). The other half of the participants are shown the same good and told that they can purchase it. These “buyers” then indicate the highest amount they would pay for the item (i.e., WTP). Typically, sellers demand a higher price to relinquish the good than buyers are willing to pay to acquire it, a pattern that has been demonstrated in a myriad of ways, across contexts and goods. Proposed explanations include loss aversion (Kahneman and Tversky 1979), an asymmetry in the self–product association of sellers and buyers (Kim and Johnson 2010), biased information processing (Carmon and Ariely 2000), and concerns about transaction disutility (Weaver and Frederick 2012). However, to our knowledge, relatively little work has investigated whether this pattern may be moderated by the type of payment method used (but see Novemsky and Kahneman 2005; Svirsky 2014). In the current research, we suggest that the use of P2P payments can lead to more cooperative pricing decisions by both sellers and buyers, attenuating the endowment effect and improving the efficiency of economic exchange markets. We explain this pattern by drawing on literature related to social norm theories.

Social Norms and Human Behavior

Social norms are defined as a shared understanding of the behaviors that are deemed “normal” and thus expected in a given situation (Bicchieri 2005; Cialdini and Trost 1998; Interis 2011). For example, a social norm in a library is to keep quiet and speak in a whisper (Aarts and Dijksterhuis 2003), and a social norm at a formal meeting is to be attentive (Eriksson, Strimling, and Coultas 2015). Social norms provide guidance to human decisions by dictating what behaviors are considered acceptable and appropriate in a specific context (Gergen and Gergen 1981). Social norms are a powerful predictor of human action, even over and beyond reported attitude (Schwartz and Tessler 1972; Zuckerman and Reis 1978), and they serve many social functions, such as sustaining the smooth operation of groups and communities (Van Kleef, Gelfand, and Jetten 2019), increasing social efficiency (Huck, Kübler, and Weibull 2012), and improving social welfare (Azar 2005).

People learn social norms by associating normative behavior with its corresponding context (Aarts and Dijksterhuis 2003; Cialdini and Trost 1998). Such learning can come from direct experience or indirect observation (Lieberman 2000; Sperber 1990). For example, people learn to associate being quiet with the library by repeatedly being quiet and/or observing others being quiet in that environment (Aarts and Dijksterhuis 2003). Different contexts can be associated with different social norms: just as people come to associate being quiet with a library, they learn to associate talking loudly and joking with a bar. Over time, associations between the contexts and normative behaviors will become established and stored in one's cognitive structure (Harvey and Enzle 1981).

A key feature of social norm theory is that once a behavior is associated with a specific context, subtle reminders of that context can directly evoke the associated behavior. This is consistent with findings in the mental association literature that show that the activation of one node in a mental association will automatically activate its associated concepts (Carver et al. 1983). Importantly, this automatic activation of behavior can happen even when the motivation that originally led to the behavior is no longer present (Burger et al. 2009; Hoffman, McCabe, and Smith 1996; Stout 2006). Using the library example, the social norm to be quiet in the library may have originally been motivated by the sight of others being quiet, a desire to not disturb others, or an attempt to avoid the ire of the librarian. However, once the library–quiet association is established, cuing the context by presenting a picture of the library is sufficient to cause participants to speak with a quieter voice in an unrelated pronunciation task (Aarts and Dijksterhuis 2003), even though the original motive to not disturb others or to abide by the library rules is no longer relevant. Similarly, a classic example of a social norm is the norm of reciprocity, which is when people learn that the expected behavior when one receives a gift or favor is to return that gift or favor (Gouldner 1960). Researchers suggest this norm may have initially been established to satisfy the self-presentation motivation (Cialdini 2001). However, once this behavioral pattern is established, receiving a favor can automatically activate the norm of reciprocity and prompt consumers to reciprocate, even without the presence of social others (i.e., when self-presentation is not possible; Burger et al. 2009). In summary, a long tradition of research on social norms has found that people come to associate certain behaviors with specific contexts. Later, subtle reminders of that context can directly evoke the associated behavior, without activating the motivations that may have originally prompted the behavior.

Building on this theory, we argue that P2P payment can subtly cue a social transaction context, activating social transaction norms in which consumers make less selfish and more cooperative pricing decisions. In the next section, we define social transaction norms, compared with traditional economic transaction norms, and discuss their influence in guiding consumer pricing decisions.

Economic Versus Social Transaction Contexts and Pricing Norms

A basic unit of commerce is the exchange of a good or service for payment. Historically, researchers usually characterize such transactions as economic exchanges or money markets (Heyman and Ariely 2004; Shore et al. 2006), to the point that one accepted definition of economic exchange is “an interaction whereby goods are given in return for money” (Bagozzi 1975). According to multiple theories, consumers engaged in economic transactions tend to emphasize the financial and tangible aspects of the exchange (Shore et al. 2006) and rarely take the other party into account (Podolny 1994). Patterns of behavior typically associated with economic transaction contexts include the norm of self-interest (Mandel 2006) and the notion that buyers and sellers should maximize value for themselves, often by extracting surplus from the other party (Miller 2001). In other words, it is acceptable and expected that buyers and sellers primarily pursue incremental benefits for themselves (Kohn 2008; Miller 2001; Wallach and Wallach 1983). Thus, when it comes to pricing decisions, it is normative for consumers to set prices and make offers based primarily on self-focused considerations, such as getting the best “deal” for themselves. As a result, sellers tend to price a product as high as possible, and buyers may make lowball offers, often resulting in disagreements between buyers and sellers (Mandel 2006).

However, not all exchanges of goods for payment are purely economic transactions (Costello and Reczek 2020; Saccardo et al. 2021; Santana and Morwitz 2021). Instead, the exchange of goods for payment can range from purely economic transactions, such as buying or selling with authorized agents, to transactions that are more social in nature, such as buying or selling at a local community market or with a close personal connection (Mandel 2006). In contrast to pure economic transactions, when making pricing decisions in social transaction contexts consumers still consider the value of the good and their personal profit, but they also take into account additional social considerations, such as facilitating social agreement (Mandel 2006), enhancing self-image (Andreoni and Bernheim 2009), preserving interpersonal relationships (Webster 1992), and receiving socioemotional benefits (Shore et al. 2006). As a result, in such situations, it has become normative for consumers to make pricing decisions that balance the desire for personal profit and other social considerations. For example, compared with a purely economic transaction, in a social transaction consumers may be less selfish and more cooperative: sellers may accept somewhat lower prices (Mandel 2006; Webster 1992), and consumers are willing to pay more (Costello and Reczek 2020; Saccardo et al. 2021).

Importantly, several recent studies show that a given transaction, even when it involves payment for a good or service, can be made to seem more or less like an economic versus social transaction through subtle contextual cues like framing and wording (Costello and Reczek 2020; Saccardo et al. 2021). When these subtle contextual cues make a given transaction seem more like an economic transaction, people act according to traditionally understood “economic transaction norms” and make more self-interested pricing decisions. However, when subtle cues suggest a given exchange is more like a social transaction, consumers display the associated normative behavior of making more cooperative pricing decisions. For example, Saccardo et al. (2021) find that a subtle wording change (e.g., “pay what you can” vs. “pay what you want”) during a marketing appeal makes the transaction feel more like a social exchange, resulting in higher willingness to pay for a cup of coffee. When a ride-share company frames a transaction as provider (vs. platform) focused, it highlights the social orientation of the transaction, increasing consumers’ willingness to pay for a ride (Costello and Reczek 2020). Interesting new work also shows that consumer characteristics such as gender can drive the chronic tendency to take social considerations into account when making pricing decisions, resulting in more cooperative pricing offers (Santana and Morwitz 2021).

Research in economic games provides consistent evidence as well. For example, one well-known and somewhat puzzling finding is that participants in a one-time dictator game often give a portion of their allocation to an anonymous stranger. Hoffman, McCabe, and Smith (1996) explain this pattern as participants having developed a set of beliefs and norms about interacting with others in a social context based on past experiences, which they inadvertently apply during the dictator game exchange. Specifically, “people have unconscious, preprogrammed rules of social exchange behavior that suit them well in the repeated game of life's interaction with other people. These patterns are imported into the laboratory. Thus, when they encounter a dictator game for the ‘division’ of $10 ‘provisionally allocated’ to them and an anonymous counterpart, not many act in their strict self-interest because the situation seems similar to the day-to-day sharing” (Hoffman, McCabe, and Smith 1996, p. 659). In simpler words, Hoffman, McCabe, and Smith argue that subtle cues in how the dictator game experiment was conducted evoked a more social transaction context, which led participants to apply the transaction norms they had previously developed in everyday life to the novel dictator game transaction context.

In summary, even during traditional one-shot market exchanges with anonymous strangers, subtle cues can evoke social transaction norms, in which consumers balance self-focused motives and social considerations, in turn prompting less selfish and more cooperative pricing decisions.

P2P Payment Methods and Social Transaction Context

In the current research, we examine the effect of different payment methods on consumers’ participative pricing decisions. Specifically, we contrast P2P payments with traditional payment methods such as cash, credit cards, and debit cards. A primary dimension on which P2P payments differ from traditional payment methods is that they tend to be considered more social in nature. This is because they have been positioned that way by managers (Alfiky 2019; see Figure 1), and because P2P payments are often used in social and interpersonal contexts, such as paying rent, splitting utilities with roommates, paying others back for shared meals, and sending gifts (Alfiky 2019; Stern 2014). In a pretest, we validated the mental association between P2P payment and social transactions (see Web Appendix B). While consumers using P2P payments are equally likely to consider their personal profit as consumers using cash (p = .98), they are significantly more likely to take social considerations into account in their decisions (p < .001), the key feature of a social transaction. Therefore, because of managers’ positioning and consumers’ past experiences using P2P payment in social contexts, we expect that anticipating the use of P2P payment (compared with a traditional payment method) can subtly cue the social transaction context and associated pricing norms.

OPEN IN VIEWER

Figure 1.

Social Market Positioning of P2P Payment Platforms.

The Current Research

Bringing together theorizing on social norms, mental associations, and payment methods, our main proposition is that P2P payment subtly cues a social transaction context, evoking the social transaction norm of making somewhat less selfish and more cooperative pricing decisions. This occurs even when the transaction is with a stranger. As a result, in an endowment effect paradigm, the use of P2P payment will cause buyers to offer somewhat higher prices and sellers to be willing to accept somewhat lower prices, attenuating the endowment effect. Formally,

H₁: Compared with the use of other payment methods (cash, credit cards, debit cards), anticipating the use of P2P payments in a novel transaction attenuates the endowment effect.

H₂: This occurs because anticipating the use of P2P payment cues a social transaction context, activating the associated social transaction pricing norm in which consumers make more cooperative pricing decisions.

Eight studies examine these hypotheses and provide support for the proposed conceptual framework, using hypothetical and consequential choices, hedonic and utilitarian products, tangible and virtual goods, and products of lower and higher value (H₁, Studies 1–6). We rule out alternative explanations related to perceived anonymity during transactions and digital representation of money (Study 3). Studies 4 to 6 use moderation (Spencer, Zanna, and Fong 2005) and mediated moderation to test H₂, and provide evidence consistent with our conceptual framework.

Study 1

Study 1 tests the proposition that the endowment effect will attenuate when consumers expect the transaction to occur through P2P payment, compared with when they expect to use cash (H₁). Students were shown a picture of a smoothie blender and imagined themselves selling or buying the blender. They were further told that the transaction would occur using either cash or a P2P payment. When the transaction will occur using cash, we expect a significant difference between buyers’ WTP and sellers’ WTA, replicating the endowment effect. However, we expect this difference to attenuate when the transaction will occur using P2P payment.

Method

Two hundred six students (33% female, M_age = 20.0 years) participated in the study for partial course credit with a 2 (payment method: cash vs. P2P) × 2 (role: seller vs. buyer) between-subjects design. We showed all participants an image of a smoothie blender. Sellers were told to imagine they owned the blender and had the opportunity to sell it at an acceptable price. Buyers saw the same blender and were told they had the opportunity to buy the blender. Orthogonally, half of the participants were told that the transaction would be made using cash, and half were told that the transaction would be made using a P2P payment (i.e., they would make/receive a payment through a payment app, such as Venmo or PayPal). All participants were then presented with a list of prices ranging from $.00 to $100.00 in $5 increments (Kahneman, Knetsch, and Thaler 1990), and indicated whether they would be willing to make the transaction at each price. We recorded the WTA to be the lowest price at which sellers agreed to sell the blender and WTP to be the highest price at which buyers agreed to buy the blender (Kahneman, Knetsch, and Thaler 1990). We excluded participants who did not finish the study from the analysis. See Web Appendix C for exclusion criteria and Web Appendixes D through K for the full stimuli for this and subsequent studies.

Results and Discussion

A two-way analysis of variance (ANOVA) with payment method (cash vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ chosen price as the dependent variable revealed a significant interaction (F(1, 202) = 4.41, p = .037, partial eta² = .021). Specifically, there was a significant WTA–WTP discrepancy in the cash conditions (M_{cash_sell} = $43.33, SD = 21.42 vs. M_{cash_buy} = $30.70, SD = 18.24; F(1, 202) = 10.86, p = .001, Cohen's d = .63), but in the P2P conditions the WTA–WTP discrepancy was eliminated (M_{P2P_sell} = $36.92, SD = 18.37 vs. M_{P2P_buy} = $35.57, SD = 18.88; F(1, 202) = .13, p = .72).

We conceptually replicated this finding using a coffee mug, a different price elicitation method, and an adult population (see Web Appendix L). The results revealed the predicted interaction (F(1, 198) = 4.64, p = .032) and pattern of results: in the cash conditions, the classic endowment effect replicated (F(1, 198) = 14.28, p < .001), but in the P2P conditions there was no difference between WTA and WTP (F(1, 198) = .502, p = .48). A second replication with a calculator revealed the same interaction (F(1, 186) = 5.301, p = .02) and pattern of results (see Web Appendix M).

The studies thus far test H₁ with student and Amazon Mechanical Turk (MTurk) samples, with a lower-priced coffee mug (average price $5), a midpriced calculator (average price $23), and a relatively more expensive smoothie blender (average price $37). Across these studies, when consumers anticipated using P2P payment compared with cash, sellers and buyers made more convergent pricing decisions. In the next two studies (2a and 2b), we examine H₁ using an incentive-compatible, consequential choice.

Study 2a

Study 2a has multiple objectives. First, we test H₁ using a consequential dependent measure, when all participants had a physical product to evaluate (a university-branded portable device charger; see Figure 2), and a subset of participants were randomly selected to have their choice executed for real. Second, to ensure participants’ pricing decisions are incentive aligned and not driven by strategic motives, we employed the Becker–DeGroot–Marschak (BDM) procedure (Becker, DeGroot, and Marschak 1964), in which participants were told that the computer would randomly pick one price from the given price list, and participants’ choice at that price (accept or not accept) would be executed. This method is believed to better elicit true pricing preferences (Kahneman, Knetsch, and Thaler 1990; Plott and Zeiler 2005).

OPEN IN VIEWER

Figure 2.

University-Branded Portable Device Charger.

Third, we propose that P2P payment activates the social transaction norm, nudging consumers to make less selfish pricing decisions. However, an alternative explanation is that participants mistakenly believe that because they are using P2P payment, they are selling to or buying from a person they know. As previous research demonstrates the attenuation of endowment effect among real friends (Halpern 1997; Mandel 2006), it was important to rule out this alternative. Thus, we included a measure designed to capture participants’ explicit beliefs about their actual relationship with the potential transaction partner. If P2P payment affects people's relationship beliefs, we would expect a main effect of P2P payment on relationship perception; however, if P2P payment activates a social transaction pricing norm as we argue, we expect no difference on the perceived relationship measure as a function of payment method.

Method

Two hundred sixteen undergraduate students (52% female, M_age = 21.1 years) participated in the study for partial course credit with a 2 (payment method: cash vs. P2P) × 2 (role: seller vs. buyer) between-subjects design. Participants entered the behavioral lab and were presented with a university-branded mobile device charger (a portable, external battery, which can be used to charge a phone or device on the go), with a full description of product attributes (see Web Appendix E). Sellers were told that the product was given to them, and they could take it home at the end of the study, but they had the opportunity to sell it at an acceptable price. Buyers were told that they had the chance to purchase the product. Orthogonally, half of the participants were told that the transaction would be made using cash, and half were told that the transaction would be made using a P2P payment (i.e., they would make/receive payment through a payment app, such as Venmo or PayPal).

All participants were shown a list of prices ranging from $.00 to $50.00 in $2.50 increments. They were told that a subset of participants would be randomly selected to have their choice be executed for real, making the pricing decision incentive compatible and consequential. The participants were further informed that the execution of their choice would follow the BDM procedure (Becker, DeGroot, and Marschak 1964). Specifically, the computer would randomly select one of the prices from the list, and their choice for that price (i.e., accept/not accept) would be honored. All participants chose their acceptable prices from the price list, as in Study 1. Next, we measured their perceived actual relationship with the transaction partner on a seven-point scale (“How well do you think you know the seller/buyer?”; 1 = “Not at all,” and 7 = “Very well”). Participants also indicated if they had a P2P account. At the end of the study, a subset of participants were randomly selected and took the product or money based on the choice they made during the study.

Results

Pricing decision

Eight participants who did not have a P2P account and thus would not be able to execute the transaction were excluded. Including them does not change the results. With the remaining 208 participants (52% female, M_age = 21.0 years), a two-way ANOVA with payment method (cash vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ chosen price as the dependent variable revealed a significant interaction (F(1, 204) = 4.39, p = .037, partial eta² = .021). Specifically, there was a significant WTA–WTP discrepancy in the cash conditions (M_{cash_sell} = $21.54, SD = 13.53 vs. M_{cash_buy} = $14.81, SD = 10.24; F(1, 204) = 9.45, p = .002, Cohen's d = .56), but this discrepancy attenuated in the P2P conditions (M_{P2P_sell} = $15.82, SD = 8.12 vs. M_{P2P_buy} = $15.58, SD = 12.04; F(1, 204) = .012, p = .91).

Perceived relationship with transaction partner

A two-way ANOVA with payment method (cash vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ perceived actual relationship with the transaction partner as the dependent variable revealed no main effect of payment methods (M_cash = 1.86, SD = .87 vs. M_p2p = 1.90, SD = .83; F(1, 196) = .143, p = .71), no main effect of transaction role (F(1, 196) = 1.66, p = .20), and no interaction (F(1, 196) = .41, p = .52). Overall, participants do not consider their transaction partner to be someone they know (M = 1.88 compared with the midpoint of the scale; t(199) = −35.35, p < .001).

Discussion

Study 2a provides important evidence in support of our main hypothesis, when prices are elicited using a BDM method to minimize strategic considerations, when the choice is consequential, and when participants are physically in the presence of the product. Further, our findings extend the research on friendship and the endowment effect by showing that P2P payment can lead to more cooperative pricing decisions even when consumers are explicitly aware that they are trading with a stranger.

In Study 2a, only a subset of participants were randomly selected to execute their choices for real. Although prior research demonstrates that using this probabilistic selection method does not influence the perceived realism of the study, nor change consumers’ incentives in decision making (Charness, Gneezy, and Halladay 2016; Engel 2011), in Study 2b we replicated Study 2a with all participants’ choices being consequential. Specifically, we adopted the classic endowment effect design whereby every seller is endowed with a product to sell and every buyer is shown the same product to buy.

Study 2b

Study 2b conceptually replicates the previous studies using a university-branded water cup (see Figure 3). We endowed every seller with a water cup and showed the buyers the same cup. We reminded all participants that their decisions would be implemented for real; that is, according to their choice, they would end up with either the cup or money at the end of the study. We expect the classic endowment effect to replicate in the traditional payment conditions but to attenuate when people anticipate transacting with P2P payments. The study was preregistered (https://aspredicted.org/av8fv.pdf).

OPEN IN VIEWER

Figure 3.

University-Branded Water Cup.

Method

Five hundred eight students (31% female, M_age = 21.0 years) participated in the study for partial course credit with a 2 (payment method: cash vs. P2P) × 2 (role: seller vs. buyer) between-subjects design. Sellers were endowed with a university-branded water cup. They were told that they owned the cup and had the opportunity to sell it at an acceptable price. Buyers were told they had the opportunity to buy the cup. Orthogonally, half of the participants were told that the transaction would be made using cash, and half were told that the transaction would be made using a P2P payment (i.e., they would make/receive a payment through a payment app, such as Venmo or PayPal). Participants were shown a list of prices ranging from $.25 to $5.00 in $.25 increments and were informed that the execution of their choice would follow the BDM procedure, as in Study 2a (Becker, DeGroot, and Marschak 1964). We included two attention check questions asking if participants were clear about their role as sellers and buyers and the payment method used for the transaction (see Web Appendix F). Participants also indicated whether they had a P2P account. At the end of the study, participants took the cup or money based on their choice.

Results and Discussion

Following the preregistered exclusion criteria, participants who (1) did not finish the study, (2) failed attention checks, or (3) did not report having a P2P account were excluded from the final analyses. The two-way ANOVA with payment method (cash vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ chosen price as the dependent variable revealed a marginal interaction (F(1, 328) = 2.80, p = .095, partial eta² = .008). Specifically, there was a significant WTA–WTP discrepancy in the cash conditions (M_{cash_sell} = $2.47, SD = 1.29 vs. M_{cash_buy} = $1.83, SD = 1.13; F(1, 328) = 10.09, p = .002, Cohen's d = .53), but in the P2P conditions the WTA–WTP discrepancy was eliminated (M_{P2P_sell} = $2.31, SD = 1.22 vs. M_{P2P_buy} = $2.12, SD = 1.33; F(1, 328) = 1.03, p = .31), again supporting H₁.

So far, our studies compare P2P payment with cash. Although cash is commonly used in the endowment effect research (Kahneman, Knetsch, and Thaler 1990; Morewedge and Giblin 2015), other payment methods such as debit and credit cards are available in the marketplace. Debit and credit cards are especially interesting to compare with P2P payment, as they share some features with it, such as the digital representation of money (Soman 2003) and the electronic transfer of funds from one account to another in the absence of physical money (Runnemark, Hedman, and Xiao 2015). However, relevant to our theorizing, these electronic payment methods differ from P2P payment with respect to their association with social transaction contexts and, thus, their ability to cue a social transaction norm. Therefore, as a further test of our theoretical framework, in the next study we compare P2P payments with credit cards. This helps provide evidence that the activation of a social transaction norm by P2P payment drives the attenuation of the endowment effect, and casts doubt on several alternative explanations.

Study 3

Study 3 extends the investigation in two ways. First, to better specify the activation of a social transaction norm as the key driver of our effect, we compared P2P payments with credit card payments. Credit cards and P2P payments share some technical similarities (Runnemark, Hedman, and Xiao 2015; Soman 2003), but credit cards are not associated with social transaction contexts. Therefore, if the endowment effect attenuates for P2P payments but not for credit cards, we can be more confident that it is the activation of a social transaction norm, but not other factors such as digital format, that drives our findings.

Second, Study 3 examines an alternative explanation related to perceived anonymity between transaction partners. Because P2P apps sometimes display one's username and picture to the other person in a transaction, consumers may worry that when they use P2P payments, they are less anonymous than when they use non-P2P payments, which could prompt less selfish behavior. Although in Study 2a we measured the degree to which participants perceive knowing each other and found no difference between cash and P2P payments, to further investigate this potential mechanism, in Study 3, we experimentally equate the anonymity between transaction partners across payment methods, and we also measure perceived anonymity between transaction partners. We expect the endowment effect to replicate with credit card payment but to attenuate with P2P payment, while the perceived anonymity between the two conditions remains constant. This study was also preregistered (https://aspredicted.org/5vi8w.pdf).

Method

Four hundred two people recruited from MTurk (48.5% female, M_age = 40.2 years) participated in the study for a small payment. The study used a 2 (payment method: credit card vs. P2P) × 2 (role: seller vs. buyer) between-subjects design. All participants were first shown an image of a smoothie blender. Sellers were told to imagine they owned the blender and had the opportunity to sell it at an acceptable price. Buyers saw an image of the same blender and were told they had the opportunity to buy it. Orthogonally, half of the participants were told that the transaction would be made using Square (a common technology used for credit card transactions), and half were told that the transaction would be made using P2P payment (i.e., they would make/receive a payment through a payment app, such as Venmo or PayPal). To ensure that the participants’ relationship with the transaction partner and their anonymity was constant across conditions, all participants (in both the credit card and P2P payment conditions) were told that they would be trading with someone they did not know and that their names and pictures would be displayed to the transaction partner during the transaction. All participants were then presented with a list of prices ranging from $.00 to $100.00 in $5 increments, and indicated whether they would sell or buy the blender at each price. Participants also indicated the extent to which they felt anonymous to their transaction partner (1 = “Not anonymous at all,” and 7 = “Fully anonymous”) and the extent to which they knew the person (1 = “Not at all,” and 7 = “Very well”). Finally, we included three attention checks to test whether participants were clear about their role as sellers or buyers, knew the payment method used, and knew that their name and picture would be displayed to the partner (see Web Appendix G).

Results

Pricing decision

Following the preregistered criteria, participants who did not finish the study or failed attention checks were excluded from the final analyses. A two-way ANOVA with payment method (Square vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ chosen price as the dependent variable revealed a significant interaction (F(1, 277) = 4.10, p = .044, partial eta² = .015). Specifically, there was a significant WTA–WTP discrepancy in the credit card conditions (M_{credit_sell} = $44.79, SD = 21.79 vs. M_{credit_buy} = $33.75, SD = 16.63; F(1, 277) = 11.29, p = .001, Cohen's d = .57), but the WTA–WTP discrepancy attenuated in the P2P conditions (M_{P2P_sell} = $39.30, SD = 19.08 vs. M_{P2P_buy} = $37.73, SD = 20.37; F(1, 277) = .22, p = .64).

Perceived anonymity

A two-way ANOVA with payment method (credit card vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ perceived anonymity to transaction partner as the dependent variable revealed no main effect of payment methods (M_credit = 2.11, SD = 1.40 vs. M_p2p = 2.11, SD = 1.32; F(1, 277) = .001, p = .97), no main effect of transaction role (F(1, 277) = .26, p = .61), and no interaction (F(1, 277) = .028, p = .87), making it unlikely that differences in perceived anonymity are a central driver of our findings.

Perceived relationship with transaction partner

A two-way ANOVA with payment method (credit card vs. P2P) and role (seller vs. buyer) as the independent variables and participants’ perceived relationship with the transaction partner as the dependent variable revealed no main effect of payment methods (M_credit = 1.86, SD = .58 vs. M_p2p = 1.85, SD = .50; F(1, 277) = .023, p = .88), no main effect of transaction role (F(1, 277) = .96, p = .33), and no interaction (F(1, 277) = .69, p = .41). As in Study 2a, participants did not consider their transaction partner to be someone they know (M = 1.85 compared with the midpoint of the scale, t(280) = −66.87, p < .001).

Discussion

The results of Study 3 indicate that it is unlikely that differences in anonymity or the mistaken belief of transacting with known others when using P2P payment can explain our findings. Study 3 also showed that the attenuation of the endowment effect does not occur for a traditional electronic payment method like a credit card, which functions in the same manner as P2P payments (i.e., with the digital transfer of money) but is not associated with social transaction contexts.

We conceptually replicated this finding in an additional study (Web Appendix N), which used specific, named P2P payment platforms (Venmo and PayPal) instead of describing P2P in a more general way as in previous studies. As in Study 3, the endowment effect emerged in the credit card condition (F(1, 297) = 9.14, p = .003) but attenuated for each of the specific P2P payment platforms (Venmo: F(1, 297) = .08, p = .78; PayPal F(1, 297) = 1.31, p = .25), yielding a significant interaction (F(2,297) = 4.52, p = .012). This finding also indicates that the effect generalizes across specific, named P2P platforms.

In a second conceptual replication (Web Appendix O), we compared P2P payment with debit cards, another traditional electronic payment method. Using a consequential study design (as in Study 2a) and a digital product (streaming access to a popular movie), we observed a significant WTA–WTP discrepancy in the debit card conditions (F(1, 245) = 12.44, p = .001). However, this discrepancy was eliminated when P2P payment was used (F(1, 245) = .71, p = .40). We also ruled out pain of payment or pleasure of receiving payment as alternative explanations in the study. Together, these data increase our confidence in the proposal that it is the cuing of a social transaction norm, but not other features of P2P, such as digital transaction format, that attenuates the endowment effect.

Thus far, our studies test H₁ with student and online samples, using consequential and hypothetical choices, hedonic and utilitarian products, tangible and virtual goods, and products with a lower price (e.g., water cup, average price $2), middle price (e.g., mobile device charger, average price $17), and relatively higher price (e.g., smoothie blender, average price $37). We further show that this effect is not driven by knowing the transaction partner, by concerns about perceived anonymity, or by the digital format of the payment transaction. Instead, we argue that it occurs because P2P activates the social transaction norm and prompts consumers to make less selfish pricing decisions. The remaining studies provide evidence of this underlying process.

Study 4

Prior research shows that the extent to which a contextual cue influences behavior depends on their strength of association (Webb and Sheeran 2007). Thus, as an important test of our conceptual framework, we propose that P2P payment should only cue a social transaction pricing norm and lead to more cooperative pricing decisions when consumers associate P2P payment with social transaction contexts. Specifically, we expect that for consumers who strongly associate P2P payment with social transaction contexts, P2P payment should activate the social transaction norm and lead to more cooperative pricing decisions, attenuating the endowment effect. By comparison, we expect that this will not occur among consumers who do not associate P2P payment with social transaction contexts and thus for whom P2P payment should not activate the social transaction norm.

Method

One hundred ninety-eight undergraduate students (32% female, M_age = 20.4 years) who reported having a P2P payment app participated in the study with a between-subjects design (role: seller vs. buyer [manipulated] × measured P2P–social transaction association). We showed all participants an image of a university-branded coffee mug. Sellers imagined that they were endowed with the coffee mug and had a chance to sell it at an acceptable price. Buyers imagined having an opportunity to buy the mug. We emphasized to all participants that they would be trading with someone they did not know. All participants read that the transaction would be made using a P2P payment. To engage participants in the scenario and increase the accessibility of their associations of P2P payment with different types of transaction contexts (e.g., economic transactions), we asked everyone to write a few sentences about how P2P payment can be used for business transactions. Next, participants chose their acceptable prices from a list of prices ranging from $.00 to $10.00 in $.50 increments. We measured the strength of the P2P–social transaction context association by asking participants to answer the question “In your opinion, mobile payment is typically used to …” on a slider scale (0 = “Transfer money with friends and family,” and 10 = “Pay for merchandise and services”). In the analysis, we reverse coded the value for ease of interpretation, such that a higher score indicates stronger P2P–social transaction context association. Finally, as an attention check, participants indicated whether they were assigned to the seller or buyer role (see Web Appendix H). Also, they reported the extent to which they perceived their transaction partner to be a friend or a stranger (0 = “Friend,” and 10 = “Stranger”).

Results and Discussion

Twenty-five participants who failed to correctly identify their role as seller or buyer in the study were excluded from subsequent analyses. To test our moderation hypothesis, we conducted a regression analysis (using SPSS PROCESS Model 1) with role (seller vs. buyer), strength of the P2P–social context association, and the interaction term as predictors and WTP or WTA as the dependent measure. The predicted interaction emerged (β = .53, t = −3.17, p = .002; see Figure 4). We applied the Johnson–Neyman procedure to identify regions of significance of the effect of seller or buyer role across different levels of P2P–social transaction association (Spiller et al. 2013). As expected, among consumers with weak associations between P2P payment and social transaction contexts (at or below 3.61), the social transaction norm was not activated, and the endowment effect emerged, mirroring the pattern we observed in previous studies with traditional payment methods (see Figure 4). Also mirroring our previous findings, among participants who more strongly associated P2P with social transaction contexts (i.e., above 3.61), the gap between WTP and WTA closed, and the endowment effect was attenuated. Interestingly, when participants reported a strong association between P2P and social transactions (i.e., association strength at or above 7.73), the endowment effect was reversed, and buyers were willing to pay significantly more than sellers would charge. Although we did not predict this reversal, it is consistent with our theory that when P2P payment is strongly associated with social transaction contexts, and thus when the social transaction norm is especially salient, consumers become especially other-regarding, even with strangers.

OPEN IN VIEWER

Figure 4.

Study 4 Results: Estimated WTP or WTA Simple Effect with Confidence Bands.

Finally, examining each group independently, an increase in P2P–social transaction context association led to a decrease in sellers’ WTA (β = −.27, t = 2.59, p = .01) and an increase in buyers’ WTP (β = .26, t = −1.99, p = .048), again consistent with our proposition that the more strongly P2P payment activates the social transaction norm, the more consumers will make cooperative pricing decisions. Participants’ role as a seller or buyer did not affect the strength of the P2P–social transaction association (M_sell = 6.28, SD = 2.73 vs. M_buy = 6.71, SD = 2.36; t(171) = 1.11, p = .27). Further, as in previous studies, participants did not believe that they were transacting with someone they knew (M = 6.72 compared with the midpoint of the scale, t(172) = 8.14, p < .001).

Study 4 provides initial evidence for H₂, consistent with our theoretical framework, by showing that when there is little association between P2P payment and social transaction contexts, and thus when the use of P2P payment does not effectively activate the social transaction norm, the traditional endowment effect emerges. However, when the same materials and payment method are used and there is a strong association between P2P payment and social transaction contexts, and thus the use of P2P can activate the social transaction norm, consumers make more cooperative pricing decisions and the gap between WTP and WTA attenuates. As in previous studies, this occurs even though people do not believe they know their transaction partner.

In the next study, we provide additional evidence in support of the theoretical framework. We propose that P2P activates the social transaction pricing norm by subtly cuing the social transaction context, which prompts consumers to make more cooperative pricing decisions. If our theory is correct, then directly cuing the social transaction context, independent from payment methods, should lead to similar pricing decisions as when P2P payment is used. In the next study we investigate this possibility by adapting a previously validated manipulation from Santana and Morwitz (2021) to directly cue the social transaction context. We expect that directly cuing the social transaction context via experimental intervention when a traditional payment method is used will have a similar effect on pricing decisions (i.e., attenuating the endowment effect) as telling people that they will be using P2P payment.

Study 5

In Study 5 we use an established manipulation to directly cue the social transaction context. Santana and Morwitz (2021) find that telling consumers that a transaction will occur on a community-oriented platform increased participants’ social payment motives, specifically leading them to become less self-focused. We adapted their manipulation to directly cue the social transaction context, by telling some participants that their transaction would occur on a community-oriented platform. When the exchange occurs on a community platform, we expect that it will directly cue the social transaction context and the endowment effect will attenuate, even though consumers are using a traditional payment method for the transaction. This study was preregistered (https://aspredicted.org/ar68t.pdf).

Method

Six hundred one people recruited from MTurk (51.0% female, M_age = 40.6 years) participated in the study for a small payment. The study used a 3 (credit card vs. P2P vs. credit card–social) × 2 (role: seller vs. buyer) between-subjects design. Participants imagined either selling or buying a smoothie blender. In the credit card and P2P conditions, the payment method was manipulated using the language from Study 3. In the credit card–social condition, the payment method was described with identical language as in the credit card condition, but we further told participants that the blender would be listed on a community-oriented marketplace that is known for its warm and caring environment where consumers can feel at home and be treated like family (adapted from Santana and Morwitz 2021, Study 4). In all conditions, participants were reminded that they would be trading with someone they did not know and that their names and pictures would be displayed during the transaction, using language described in Study 3. All participants were then presented with a list of prices ranging from $.00 to $100.00 in $5 increments, and indicated whether they would agree to sell or buy the blender at each price. Finally, we included three attention checks to test if participants were clear about their role as sellers or buyers, knew the payment method used, and knew that their name and picture would be displayed to the partner (see Web Appendix I).

Results and Discussion

Following the preregistered exclusion criteria, participants who did not finish the study or failed attention checks were excluded from the final analyses. A two-way ANOVA with the remaining participants revealed a significant interaction (F(2, 435) = 3.68, p = .026, partial eta² = .017). Replicating previous studies, there was a significant WTA–WTP discrepancy in the credit card control conditions (M_{credit_sell} = $42.27, SD = 19.56 vs. M_{credit_buy} = $35.37, SD = 17.59; F(1, 435) = 4.99, p = .026, Cohen's d = .37). In contrast, the WTA–WTP discrepancy was attenuated in the P2P condition (M_{P2P_sell} = $40.32, SD = 19.63 vs. M_{P2P_buy} = $38.01, SD = 17.62; F(1, 435) = .22, p = .42) and in the credit card–social condition (M_{P2P_sell} = $30.61, SD = 16.51 vs. M_{P2P_buy} = $35.07, SD = 15.78; F(1, 435) = 2.37, p = .13).

These results support the notion that P2P influences consumers’ WTA or WTP by activating social transaction norms. One strength of this study is that it cued the social transaction context using a managerially relevant marketplace platform manipulation that prior research has validated. Through two moderation studies (Spencer, Zanna, and Fong 2005), we provide evidence consistent with the proposed conceptual framework: that P2P payment leads to more cooperative pricing decisions because it subtly cues a social transaction context that in turn activates the social transaction pricing norm. The next two studies provide additional evidence in support of the process using a measured mediation approach.

Study 6

Our central proposal is that the payment method can be a subtle cue about whether a given transaction is purely economic versus more of a social transaction, which in turn evokes the associated norm for pricing behavior. Specifically, we have robustly demonstrated that paying with P2P can activate social transaction pricing norms and lead to more cooperative pricing decisions. In Study 6 we use a mediation design to demonstrate that this occurs because compared with traditional payment methods, P2P payment subtly cues a social transaction context, which in turn affects pricing decisions and attenuates the endowment effect. Specifically, Study 6a measures how the payment method affects perception of the transaction context, and validates that even when all other aspects of the transaction are identical, the payment method can be a subtle cue about the economic versus social nature of the transaction context. This study was preregistered (https://aspredicted.org/th2pd.pdf).

Finally, Study 6b shows that this subtle difference in economic versus social transaction context activation can account for the previously observed effects on pricing behavior, using a mediated moderation design with the same measure. Study 6b was also preregistered (https://aspredicted.org/vi49y.pdf).

Study 6a

Method

Four hundred people recruited from Prolific (62% female, M_age = 36 years) participated in the study for a small payment. The study used a 2 (payment method: credit card vs. P2P) × 2 (role: seller vs. buyer) between-subjects design. All participants were first shown an image of a smoothie blender. Sellers were told to imagine they owned the blender and had the opportunity to sell it at an acceptable price. Buyers saw an image of the same blender and were told they had the opportunity to buy it. Orthogonally, half of the participants were told that the transaction would be made using a credit card, and half were told that the transaction would be made using P2P payment. To ensure that participants’ relationship with the transaction partner was constant across conditions, all participants (in both the credit card and P2P payment conditions) were told that they would be trading with someone they did not know.

We measured the effect of payment method on activation of economic versus social transaction context by asking participants their perception of the transaction context on a seven-point scale (1 = “Purely social transaction,” and 7 = “Purely economic transaction”). Specifically, all participants read the following description:

Market transactions can be described as occurring along a continuum, from purely economic transactions to more social transactions. Purely economic transactions are usually asocial, and consumers tend to be self-interested and to focus primarily on the value of the good and getting the best deal for themselves. In purely economic transactions, consumers rarely take the other party into account. An example of an economic transaction is when people buy from an authorized dealer. In contrast, in a social transaction, while consumers still focus on the value of the good and personal profit, they also take into account other factors, such as the relationship with the other party or other social considerations. An example of a social transaction is when people buy from a community market. To what extent do you think this transaction is an economic vs social transaction?

Finally, we included two attention checks to test whether participants were clear about the payment method used and the fact that they did not know their transaction partner before the transaction (see Web Appendix J).

Results and discussion

Following the preregistered exclusion criteria, participants who did not finish the study or failed attention checks were excluded from the final analyses. A two-way ANOVA with the remaining participants revealed a significant main effect of payment methods (M_{credit_econ} = 5.94, SD = 1.46 vs. M_{P2P_econ} = 5.45, SD = 1.61; F(1, 376) = 9.84, p = .002, Cohen's d = .32), confirming our proposition that using P2P payment for transactions subtly cues a social transaction context. There was no main effect of seller/buyer role on perceived transaction type (M_sell = 5.74, SD = 1.47 vs. M_buy = 5.65, SD = 1.63; F(1, 376) = .22, p = .64). Although not predicted a priori, the results also revealed a significant interaction (F(1, 376) = 5.80, p = .017). The unexpected interaction is driven by a difference in the P2P condition such that buyers saw the transaction as significantly more social than sellers did. We speculate that there may be a stronger association between P2P payment and social transaction context for buyers in part because it is more common for people to use P2P payment to buy rather than to sell.

The results of Study 6a validate that for otherwise identical transactions, and when the transaction is with an anonymous stranger, the mere use of P2P payment is sufficient to cue a social transaction context. A related possibility is that P2P payment might affect expectations of physical proximity or time spent with the transaction partner, such that consumers using P2P payment believe they would be physically closer or spend more time with the transaction partner, compared with the use of traditional payment methods. To test this possibility, we replicated Study 6a and added several new measures of perceived proximity and time spent with the transaction partner. As in Study 6a, participants in the P2P conditions perceived the exchange as more of a social transaction. However, they did not believe they would be physically closer to their transaction partner, and they even expected to spend marginally less time with the transaction partner (see Web Appendix P). The data suggest it is unlikely that our observed effect is driven by P2P payment changing consumers' concrete expectations about the transaction, such as physical proximity and time spent with partners, and thus increase our confidence that P2P payment influences pricing decisions by evoking social transaction pricing norms. In Study 6b we extend these findings and show that this subtle evoking of social transaction contexts mediates the effect of P2P payment on consumer pricing decisions.

Study 6b

Method

Eight hundred one people recruited from Prolific (54.3% female, M_age = 39.8 years) participated in the study for a small payment. The study used a 2 (payment method: credit card vs. P2P) × 2 (role: seller vs. buyer) between-subjects design. The first part of the procedure used methods and materials identical to Study 5 (but without the credit social condition). After participants chose their highest WTP or lowest WTA, we measured the evoking of social transaction context using a measure identical to that described in Study 6a. Finally, participants responded to three attention checks about their role as sellers or buyers, the payment method used, and whether they knew their transaction partner before the transaction (see Web Appendix K).

Results and discussion

Following the preregistered exclusion criteria, participants who did not finish the study or failed attention checks were excluded from the final analyses. A two-way ANOVA with the remaining participants revealed a significant interaction in the pricing decision (F(1, 729) = 4.72, p = .03, partial eta² = .006). Specifically, the WTA–WTP discrepancy emerged in the credit card condition (M_{credit_sell} = $45.35, SD = 21.06 vs. M_{credit_buy} = $36.64, SD = 17.93; F(1, 729) = 19.15, p < .001, Cohen's d = .45) but was attenuated in the P2P conditions (M_{P2P_sell} = $39.30, SD = 17.85 vs. M_{P2P_buy} = $36.69, SD = 19.05; F(1, 729) = 1.74, p = .19).

Next, we performed a conditional indirect effects analysis. We predicted that anticipating the use of P2P payment cues a social transaction context, activating the social transaction pricing norm that leads to more cooperative pricing decisions. Replicating Study 6a, we observe a main effect of payment type on social transaction context activation, such that participants perceive the transaction to be significantly more like a social transaction context when P2P is used (M_{credit_econ} = 5.89, SD = 1.48 vs. M_{P2P_econ} = 5.21, SD = 1.79; F(1, 729) = 30.79, p < .001, Cohen's d = .41).

We further predicted that when P2P payment activates the social transaction context and thus the social transaction pricing norms, sellers would accept a somewhat lower price (i.e., a decrease in WTA) and buyers would pay a somewhat higher price (i.e., an increase in WTP), resulting in a mediated moderation. To test this model, we used SPSS PROCESS Model 15 (Hayes 2022), entering the payment method as the independent variable, perceived social transaction context as the mediator, the role in the transaction as the moderator, and participants’ WTA or WTP as the dependent variable. This analysis yielded a mediated moderation (β = 1.63, 95% confidence interval: [.43, 3.21]). Specifically, after controlling for payment method, we found an interaction between perceived social transaction context and exchange role (seller vs. buyer) on participants’ WTA or WTP (β = −2.39, t = −2.78, p = .006), indicating that evoking a social transaction context increases buyers’ WTP and decreases sellers’ WTA. This result is consistent with the proposed mechanism that P2P payment influences consumers’ participative pricing decisions in a market transaction by activating a social transaction context and the associated pricing norm of making more cooperative pricing decisions.

General Discussion

Successful transactions, and efficient markets, require a buyer and seller to converge on a mutually acceptable price. Yet prior research regularly finds an endowment effect, such that sellers tend to ask for a higher price than buyers are willing to pay, which inhibits transactions from occurring and makes markets less efficient (Morewedge and Giblin 2015). Such asymmetry in pricing decisions between sellers and buyers has the potential to decrease the transaction volume of a product by as much as 60% (Kahneman, Knetsch, and Thaler 1990). This type of market inefficiency can impede revenue and growth for consumer-to-consumer e-commerce platforms such as Etsy, Facebook Marketplace, and eBay, where participative pricing decisions are common and revenue is determined by the volume of successful transactions. Across eight studies, we demonstrate that using P2P payment methods, compared with traditional payment methods such as cash, debit cards, and credit cards, can increase the likelihood that buyers and sellers will converge on a mutually agreeable price, attenuating the endowment effect. Through a market simulation (see Web Appendix A), we show that by encouraging the use of P2P payments, managers of consumer-to-consumer e-commerce platforms may be able to increase transaction volume, potentially by as much as 9% on average, which can translate into a sizable increase in revenue.

A natural question that follows this market implication is whether P2P payment affects sellers, buyers, or both. In short, the answer is that, compared with non-P2P payments such as cash and credit or debit cards, P2P payment both decreases WTA for sellers and increases WTP for buyers. To examine this question, we pooled the data from all studies designed to test H₁ and calculated the standardized effect size that P2P payment has on sellers’ WTA and buyers' WTP separately (see Web Appendix Q). The results show that across studies, P2P decreases sellers’ WTA by .267 units (p < .001) and increases buyers’ WTP by .122 units (p = .048). Further, there is no difference between the size of the two effects (p > .1). Thus, the use of P2P payments significantly influences both sellers’ and buyers’ pricing decisions with approximately equivalent effect sizes, consistent with our theorization that P2P payment activates social transaction norms for both transaction parties.

Beyond the important substantive implications of this work, our findings contribute to several theoretical domains. First, participative pricing decisions are known to be affected by consumer mindset (Chandran and Morwitz 2005), social norms (Kim, Natter, and Spann 2009), and environmental cues (Gneezy et al. 2010), but no previous research examines how the payment method can affect participative pricing decisions. We show that this external factor can alter sellers’ WTA and buyers’ WTP, even when the valuation of the product is unchanged (as measured in a conceptual replication of Study 1; Web Appendix M).

Second, previous work has shown that payment methods, such as cash and credit cards, can affect consumer judgment and decisions (Prelec and Simester 2001; Raghubir and Srivastava 2008; Soman 2003). To the best of our knowledge, the current research is the first to systematically investigate the effect of P2P payment methods on consumer behavior. Whereas prior work mainly focuses on how payment methods can induce different levels of pain of payment (Prelec and Simester 2001; Raghubir and Srivastava 2008; Soman 2003), we extend this stream of literature and identify a unique psychological process that can be evoked by payment methods, specifically the activation of social transaction norms, which influence consumers’ participative pricing decisions.

Third, we add to a growing body of research that examines how relationship norms and actual interpersonal relationships interact with consumer pricing decisions in the context of the endowment effect, and in which scholars have documented seemingly divergent empirical patterns. Whereas one stream of research shows that making communal relationship norms accessible increases the price that sellers demand and amplifies the endowment effect (Aggarwal and Zhang 2006; McGraw, Tetlock, and Kristel 2003), other studies show that transacting with actual close friends can attenuate or even reverse the endowment effect (Halpern 1997; Mandel 2006). We add to this investigation by examining how the use of P2P payment can subtly activate the social transaction norm and prompt consumers to not only consider maximizing the value for themselves but also take social considerations into account in their pricing decisions, and in doing so we document a third empirical pattern: P2P payment increases the price buyers are willing to pay and decreases the price sellers are willing to accept, even when consumers report being explicitly aware that they are transacting with true strangers. Thus, our data provide an additional, novel perspective on the complex interplay between social considerations and participative pricing.

Limitations and Future Research

Our empirical package provides consistent evidence that P2P payment prompts less selfish and more cooperative pricing decisions among both sellers and buyers, attenuating the endowment effect. Some remaining questions that emerged deserve more investigation. One unexpected pattern that emerged in our data was that for people who strongly associate P2P with social transaction contexts, buyers tend to offer a higher price than sellers are willing to accept (Study 4). We speculate that the observed reversal may happen when the contextual cue makes the social transaction norm especially salient and when consumers take the social considerations into account to a greater extent, even when they are transacting with a stranger. As a result, sellers are willing to offer a much lower price and buyers are willing to pay even more. Future research could examine the specific boundary conditions under which social transaction norm activation is strong enough to prompt a reversal in the endowment effect.

An open question is whether and how findings related to the effect of social transaction norms within market exchanges relate to the notion of “communal norms,” the set of behaviors and motivations developed to govern an entirely different class of interactions (e.g., helping friends move, motivated by genuine care for one another). Although prompting social considerations does appear to make consumers somewhat less selfish, which is generally consistent with communal transactions, both in our work and in recent studies, we believe the scope of social transaction norm activation is likely to be different from communal norms in many important ways. For instance, although activating social transaction norms during a pricing decision may make people somewhat less selfish in their WTP or WTA, we believe it is unlikely to prompt genuine concern or caring for the other party in the transaction. In our context, although paying with P2P may encourage consumers to take social considerations into account, we believe it is unlikely that consumers develop a communal relationship with the transaction partner just because of the use of P2P payment, nor would they consider it their responsibility to take care of the partner or be willing to incur costs to meet that person's needs, all of which are hallmarks of communal norms. Future research might directly investigate and further elucidate the empirical relationship between social transaction norm activation within an exchange context (i.e., our theoretical framework) and other nominally related but empirically distinct notions, such as communal relationship norms.

Whereas our study focuses on consumers’ pricing decisions in the context of the endowment effect, future research could also examine consumer decisions in different contexts, such as willingness to donate, where social considerations are more salient. One possibility is that P2P payment may amplify prosocial behaviors in a donation context because of the congruency between the activated social transaction norm and the donation context. However, there may be a ceiling effect because the donation context is already dominated by social transaction norms, and thus people will act in a more other-regarding manner regardless of the payment method used. Future research can investigate how transaction contexts can amplify or attenuate the effect of P2P payment on other-regarding behaviors.

In a traditional view, one-shot transactions with a stranger are often characterized as economic exchanges and are expected to be governed primarily by economic transaction norms, that is, focusing on maximizing benefits for oneself (Clark and Mills 1993; Heyman and Ariely 2004). However, researchers are increasingly showing that even within these prototypical market transactions, there is considerable variation in the degree to which social considerations come into play (Costello and Reczek 2020; Saccardo et al. 2021; Santana and Morwitz 2021). For example, Costello and Reczek (2020) show that using provider-focused (vs. platform-focused) marketing communication can cue the behaviors of perspective taking and empathy, which in turn affect pricing decisions. Although we do not directly examine the domains discussed by Costello and Reczek, we believe that there may be potential synergies between our findings. For example, platform-focused marketing communications seem to prompt behaviors similar to those prompted by traditional payment methods, whereas provider-focused marketing communications seem to prompt behaviors more similar to those prompted by P2P payment. We speculate that beyond affecting pricing decisions, additional behaviors that are normative in social transaction contexts may prompt consumers to take the perspective of the other party, as in Costello and Reczek's work. A potentially fruitful area of future research may be to bring together these various findings related to social transactions to develop a synergistic, overarching framework (MacInnis 2011) for the nomological network of consumer behaviors that may emerge in social transaction contexts.

More generally, we believe our work advances thinking on the breadth of motivations and considerations that come into play when consumers engage in market transactions. Although most traditional research characterizes market-based exchanges with strangers of goods for payment as primarily “economic” transactions that gravitate toward self-interest and personal value maximization (Bagozzi 1975; Miller 2001), it may be time to broaden our treatment of such markets. Sharing economies, consumer-to-consumer platforms, and crowdfunding are all powerful market examples of social transaction marketplaces in which consumers are increasingly likely not only to consider and seek to maximize their own needs, but also to take into account social considerations. Academics would be well served to expand our conceptualization of traditional markets to include these perspectives. Finally, more than ever, issues like environmental impact, political ideology, responsible sourcing, and labor policies are all increasingly likely to be considered and to influence pricing and purchase intent. On many levels, this is good news: factors that increase people's tendency to take a broad social perspective in their consumption choices will ultimately improve long-term societal welfare. Future research could continue to seek out small nudges, beyond the method of payment, that can make consumers less self-focused and more likely to consider broader social perspectives.

Conclusion

The past decade has witnessed the rise of P2P payments and explosive growth of consumer-to-consumer e-commerce platforms, such as Etsy, Facebook Marketplace, and eBay, which largely rely on a participative pricing mechanism. This research investigates the effect of P2P payment on consumers’ participative pricing decisions. In doing so we extend theory in payment methods, consumer participative pricing decision making, and the role of contextual cues for understanding consumer behavior, and we offer insights relevant to managers about two large and important trends in the marketplace: the rise of P2P payments and the explosive growth of digital consumer-to-consumer markets.

Supplemental Material

sj-pdf-1-mrj-10.1177_00222437221128255 - Supplemental material for When Payments Go Social: The Use of Person-to-Person Payment Methods Attenuates the Endowment Effect

Supplemental material, sj-pdf-1-mrj-10.1177_00222437221128255 for When Payments Go Social: The Use of Person-to-Person Payment Methods Attenuates the Endowment Effect by Liang Huang and Jennifer Savary in Journal of Marketing Research