Perceived Fairness and
Willingness To Pay for Sustainable
Products in Pay What You Want
Pricing
Tessa Simone de Leede | 10016341 | Words: 11,666
Supervisor: dr. A. el Haji Second reader: dr. J.J. Ebbers
June 21th, 2016 Final version
MSc. in Business Administration | Entrepreneurship and Management in the Creative Industries | ABS, UvA
This document is written by Student Tessa de Leede who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
2 Abstract
Ethical consumerism is a rising movement. Among consumers there is a growing trend of interest in ethical aspects associated with product consumption. As a result, today an increased demand to legitimate its practices to society is experienced in every industry. However, an intension-behavior gap still exists. Even though consumers want to buy in an ethical manner, they often do not practice what they preach. They are expected to pay premium prices while they experience a lack of information about these products. In previous literature Pay-What-You-Want (PWYW) was presented as a mechanism to overcome this problem. Some research argues that consumers perceive PWYW as more fair compared to fixed prices. It was found that although consumers have the option to pay nothing, most of them pay above zero. In this study, a field experiment was conducted at a Dutch sustainability platform. The manipulation was designed according to a 2x2 design. It was found that a perceived fairness of PWYW positively affected WTP. Contrary to what was expected, there was no effect of underscoring the fairness of PWYW on the perceived fairness of PWYW. An exception was the youngest group of participants, they were negatively influenced. Also, participants in the charity treatment actually had a lower WTP, compared to other treatments. This could partly be explained by the credibility of the product, brand, and pricing mechanism. As expected, brand credibility had a positive effect on WTP. Some additional effects were found and discussed.
3 Table of Contents
1. Introduction 4
2. Literature Review 6
2.1 Ethical consumerism 6
2.1.1 Corporate Social Responsibility
2.1.2 Ethical consumption intention-behavior gap
2.2 Pay what you want 2.2.1 Willingness to pay in PWYW 2.2.2 Perceived fairness of PWYW 2.2.3 Cause-related marketing 2.2.4 Brand credibility 6 7 8 9 10 11 13 3. Methodology 15 3.1 Data collection 14 3.2 Research design 16 3.3 Sample 17 4. Results 19 4.1 Descriptive statistics 4.2. Hypothesis testing
4.2.1 Hypothesis 1 – Perceived fairness 4.2.2 Hypothesis 2 – Underscoring fairness 4.2.3 Hypothesis 3 – Donation to charity 4.2.4 Hypothesis 4 – Brand credibility
4.3 Other effects
4.3.1 Controlling effects of perceived fairness of PWYW 4.3.2 Regression analysis for perceived fairness of PWYW 4.3.3 Interest and sustainability
4.3.4 Perceived fairness of price
4.3.5 Alternative explanations effect charity on WTP 4.3.5 Gender
4.3.6 Age
4.3.7 Regression analysis for WTP
19 21 21 23 24 25 26 26 28 30 31 32 33 34 35 5. Discussion 37 5.1 WTP in PWYW 5.2 Perceived fairness
5.3 WTP for charity in PWYW 5.4 Credibility in PWYW 5.5 Limitations and implications
37 38 39 40 41 References 45 Appendices 49
Appendix 1: newsletter Goodclub Appendix 2: survey flow
Appendix 3: treatments
49 52 59
4 1. Introduction
Interest in ethical products is becoming more prevalent. Nowadays, some consumers are not only seeking for quality and esthetics while purchasing a product. They explicitly look for products with social or environmental elements (Pelsmacker, Driesen & Rayp, 2006). Retailers are aware that customers increasingly evaluate them according to their socially responsible behavior and acknowledge their intermediary role in the marketing channel (Schramm, Morschett & Swoboda, 2015). At the same time, there is evidence that suggests that consumers are not willing to pay a premium price for these social or environmental elements (Pelsmacker et al., 2006). This seems paradoxical since for example the high prices for Fair Trade products are directly related to guaranteeing producers a fair wage (Araque-Padilla, Montero-Simó, & Aragón-Gutiérrez, 2015). According to Nicolls (2002) this unwillingness might partly be due to skepticism about the organization’s motives. To inform consumers, organizations should therefore communicate about their ethical elements in such a way that it is perceived as credible.
Moreover, Mendoza-Abarca & Mellema (2015) argue that organizations with a social aim need to implement mechanisms that simultaneously can create social value and an appropriate amount of economic value. They expect Pay What-You-Want (PWYW) to be a promising mechanism. With PWYW, consumers themselves decide what price to pay for a product (Kim, Natter & Spann, 2009). Because the consumer has full control over the price, PWYW could be perceived as more fair than fixed prices (Kim et al., 2009; Mendoza-Abarca & Mellema, 2015). However, the effect of the perceived fairness of PWYW on willingness to pay (WTP) has not been measured yet.
This study measures to what extent the perceived (relative) fairness of PWYW could be manipulated. Moreover, it examines the effect of these manipulations on WTP. Lastly, the
5 study measures the influence of brand credibility on WTP. First, participants are explicitly told why the organization thinks the pricing mechanism is fair. They might therefore feel socially pressured to pay a fair amount of money (Gneezy, Gneezy, Riener & Nelson, 2012). Second, participants are promised that a donation to charity is given. Consumers seem to be willing to pay more for products when the seller promises to donate to charity (Gneezy et al., 2012). However, it is not clear if this is still the case when the offered product itself has an ethical element. Implementing PWYW for ethical products or social organizations has received little attention in the literature. Gneezy et al. (2012) examined the effect of charity on WTP at PWYW. In this study a combination is examined, i.e. the effect of perceived fairness of the pricing mechanism and donating to charity for ethical products in PWYW. This leads to the following research question: What is the effect of donating to charity and underscoring the
fairness of Pay-What-You-Want Pricing on the willingness to pay for ethical products under Pay-What-You-Want Pricing?
By examining its potential in The Netherlands, the organization of this study – and potentially others as well - might get motivated to implement PWYW. It is examined to what extent consumers pay a higher price when a donation to charity is done and fairness is underscored. Moreover, this experiment shows whether the credibility of the seller influences the WTP. Therefore, this study provides implications in the PWYW field as well as in the field of business ethics. Results might give organizations that consider to invest in business ethics more insight about how to implement it.
This study is structured as follows. In the next section, previous research in ethical consumerism and PWYW is outlined. Based on this literature, three hypotheses are formulated. These hypotheses present potential conditions influencing the WTP in PWYW. Then, the method and data collection of the online field experiment are described. The results are outlined and discussed. Finally, a conclusion is presented.
6 2. Literature review
2.1 Ethical consumerism
Ethical consumerism is defined as “the conscious and deliberate choice to make certain consumption choices due to personal and moral beliefs” . More business oriented it means that individual consumers can promote ethical corporate practices through their daily purchase decisions (Auger & Devinney, 2007, p.362). With ethical consumerism both environmental, business practices, and social justice are incorporated (Chatzidakis, Kastanakis & Stathopoulou, 2016). Among consumers there is a growing trend of interest in ethical aspects associated with product acquisition and consumption (Araque-Padilla et al., 2015). These 'ethical consumers', are concerned for both the product itself and for the possible effects of its consumption. They feel responsible toward society and express these feelings with their purchasing behavior, i.e. by buying products for their positive qualities or boycotting products for their negative qualities (Pelsmacker et al., 2006). Shopping in alignment with their ethical consumption values provides these consumers a sense of connection with who they are and where they come from (Carrington, Neville and Whitwell, 2014).
Products with ethical attributes refer to those with explicit social elements (e.g. protecting human rights, questioning world trade rules, or the promotion of social enterprises) or environmental elements (e.g. reducing humans’ ecological impact, combating environmental pollution, objecting genetically modified food, fostering animal welfare or biodiversity). Two typical examples are Fair Trade and organic products (Araque-Padilla et al., 2015).
2.1.1 Corporate Social Responsibility
Sustainability had been defined as "meeting the needs of the present without compromising the ability of future generations to meet their needs" (WCED, 1987, p.43). Today there is virtually no industry that has not experienced increasing demands to legitimate its practices to
7 society at large. Companies have responded to this agenda by advocating Corporate Social Responsibility (CSR) (Crane et al., 2013), in which context-specific organizational actions and policies take into account stakeholders’ expectations and the triple bottom line performance (i.e. economic, social, and environmental) (Aguinis & Glaves, 2012).
Also retailers are aware that their customers are increasingly evaluating them according to their socially responsible behavior (Schramm-Klein et al., 2015). They are not only held responsible for their own CSR activities. They also have to guarantee the responsible behavior of all parties in the supply chain towards their consumers. Therefore, the retailer acts as an intermediary in the marketing channel (Schramm-Klein et al., 2015). Biong, Nygaard and Silkoset (2010) describe retailing as a ballot box for ethical decisions, in which consumers ‘vote’ with their shopping carts. Moreover, customers may rely on the retailers reputation to confirm that a product is ethical, since they do not have the resources to examine this themselves. Hence, the credibility of the retailer is important in the marketing of ethical products (Pearson & Henryks, 2008).
2.1.2 Ethical consumption intention-behavior gap
Despite the ethical consumerism movement, ethically-minded consumers rarely purchase ethically (Carrington et al., 2014). Consumers’ buying behavior is not consistent with their positive attitude toward ethical products (Pelsmacker et al., 2006). Carrington et al. (2014) call this the 'intention-behavior gap', a misalignment between ethical intentions and actual behavior. They identified the following obstacles: (1) alternative personal values, (2) extant habits, (3) inability to form plans, (4) unwillingness to sacrifice, (5) lack of available information and an unwillingness to conduct effortful searches for information, and (6) the distraction of the situational environment in effortful and spontaneous shopping modes.
8 According to the fourth reason, one of the major sacrifices consumers have to make is financially. They are required to pay a premium price to contribute to what is supposed to be a valuable social cause. It assumes that their purchase is more due to a desire to make socially responsible choices than for price reasons (Araque-Padilla et al., 2015). Pelsmacker et al. (2006) found in their experiment in Belgium that a mere 10% of participants is prepared to pay a price premium for Fair Trade coffee. In addition, Schmeltz, Bishop, and Gruen (2014) found in their experimental auction that 30% of participants bid significantly more for pro-environmental wine (i.e. high premium price group), compared to regular wine. More interestingly, 28% bid significantly less for the pro-environmental wine, compared to regular
wine. Those participants saw low social or emotional benefit in the pro-environmental
product. Araque-Padilla et al. (2015) found in their study that that the perceived value of a product decreased as the price increased. In order of importance, the perceived ethics, quality, and healthiness of the product attenuates this relationship.
This highlights the importance of informing consumers about ethical (as well as quality and health) aspects of the product, in line with the fifth reason of the intention-behavior gap. Araque-Padilla et al. (2015) argue that due to lack of information, skepticism and a lack of perceived utility of ethical products exists. Nicolls (2002) state that communication of the human element of sustainability should allow the individual consumer to give them the feeling that they make a difference. Consumers then make the commitment of regular purchase behavior and develop brand loyalty.
2.2 Pay what you want
Pay-What-You-Want (PWYW) simply means that the buyer decides on the price of an offered product. Regardless of the amount paid (even zero), the seller accepts the buyer's price and the transaction automatically proceeds (Kim et al., 2009). Long-existing examples where PWYW has been used are street music (Jang & Chu, 2012), museums and charity. Nowadays,
9 the pricing mechanism is being used in several other industries (i.e. new music albums online, lunches, theatre, and amusement parks) (Schmidt, Spann & Zeithammer, 2014).
Schmidt et al. (2014) name three main reasons why PWYW could be an attractive pricing strategy. First, PWYW is a means of endogenous price discrimination (i.e. different consumers pay different prices for the same products). Second, PWYW is suitable for organizations seeking to maximize unit sales (e.g. to sell a complementary product, enter new markets, realize experience curve gains, or achieve network effect). Lastly, PWYW can be used as a competitive strategy (e.g. undercutting competition using posted prices by not competing on price). Additionally, Kim et al. (2009) mention that because of its novelty PWYW may be used as a marketing instrument for new businesses. It may increase the chance of word of mouth and build a positive price image among consumers.
2.2.1 Willingness to pay in PWYW
Although consumers are allowed to pay nothing, recent studies on PWYW (in different industries) repeatedly show that the price paid is significantly higher than zero (Kim, Kaufmann & Stegemann, 2014). Gneezy et al. (2012) explain that people choose to pay because they feel it is the right thing to do. The payment depends on social norms. In face-to-face situations, paying in PWYW signals the morality of the person to others. Moreover, the payment serves to maintain their positive self-image. According to the self-signaling model, people feel uncertain about their true identity and use their actions to recall their beliefs. As a result, people would rather avoid a situation in which they violate the norm by choosing not to buy the product. If they do choose to purchase, they often choose to pay a fair price that does not have a negative effect on their self-image (Gneezy e al., 2012). In addition, Kunter (2015) found that in his empirical studies that guilt has a stronger effect on PWYW prices than shame.
10
Kunter (2015) was the first to study the impact of simple textual cues on PWYW
payments. He found that participants who saw statements about satisfying the seller and
wanting to be fair towards the seller paid the highest amount of money on average. However,
most textual cues did not change PWYW payments. Contrary, Pruckner and Sausgruber (2008) found that implementing the sign “Thank you for being honest” induces higher payments. Yet, it did not decrease the number of zero-payments. Kunter (2015) acknowledged
that the effects of other cues might have overwhelmed the textual cue, because the field
experiments took place in a zoo. In online field experiments, participants should be less
overwhelmed with cues. Besides, the effect of text cues on the perceived fairness of PWYW
has not been researched yet.
2.2.2 Perceived fairness of PWYW
Price fairness refers to the perceived fairness judgment of a buyer about a seller’s prices. It is part of a broader judgment of the overall merits from the transaction (Haws & Bearden, 2006). Consumers may prefer PWYW because of the level of control that is offered. When playing a role in the price-setting, consumers have greater fairness perceptions (Kim et al. 2009). This assumption is based on the finding that prices set through bidding were perceived as more fair than posted prices. However, bidding is quite different from PWYW, since you
do not have the guarantee that you will own the product after bidding. At PWYW the product
is yours for any price you offer. This indicates that PWYW would even be perceived as fair
when a consumer receives the product for a price below cost price.
For ethical products, PWYW could be perceived as more fair compared to fixed
prices, because of the high initial pricing of these products. Social organizations aim to realize their social mission by selling products. This makes them financially dependent. Pricing too low could therefore threaten the social organization’s ability to sustain its operations. Besides, high prices are largely related to guaranteeing producers a fair price for ethical products.
11 Nevertheless, most consumers are unwilling to pay a premium price for them (Araque-Padilla et al., 2015). A price that is perceived as too high may signal to customers that the organization is merely using the social mission as a device to maximize its own profits. Therefore, social organizations need to implement mechanisms that can create social value and an appropriate amount of economic value simultaneously. PWYW gives consumers the opportunity to decide for themselves how much they want to contribute to the social mission (Mendoza-Abarca & Mellema, 2015).
In this study it is examined whether underscoring the fairness of PWYW affects the
perceived fairness of the pricing mechanism, and whether this perceived fairness has an effect
on the WTP.
H1: Consumers are willing to pay more for ethical products in PWYW when they perceive the
pricing mechanism as more fair.
H2: Consumers perceive PWYW as a more fair pricing mechanism when this fairness is underscored, rather than when it is not underscored.
2.2.3 Cause-related marketing
Cause-related marketing implies that linking product sales to a charitable cause can serve as a marketing tool (Koppel & Schulze, 2013). Previous studies found that consumers who perceive products to be highly related to a social cause also extract more benefits from the transaction (Mendoza-Abarca & Mellema, 2015). Likewise, charity auctions in which a percentage of the price is donated to charity increase positive feelings from supporting the mission for altruistic consumers. Additionally, it made them willing to place higher bids (Elfenbein & McManus, 2010). According to Gneezy et al. (2012), this also seems to be the case for PWYW. They found that people’s WTP is higher when half of their payment goes to
12 charity. At the same time, consumers are more likely to make small donations linked to the purchase of a product rather than independently to a donation box (Koppel & Schulze, 2013). By purchasing a product that is linked with a social cause, the customer implicitly donates money to the cause. This is in line with ethical consumerism, where the more ethical products you buy, the more ‘good’ you do for planet, people and profit (Low & Davenport, 2006).
It has not yet been examined if a donation to charity still provides a higher WTP if the product has a social element as well. Purchasing an ethical product with an additional donation to charity might increase the positive feelings from supporting the mission. However, it could also be the case that consumer who buy an ethical product may already perceive its purchase as a way of supporting the mission, e.g. a fair wage for everyone or carbon free consumption. Moreover, an additional donation to charity may even signal to customers that the seller is using the social mission as a device to maximize its own profits.
Besides, in the study of Gneezy et al. (2012) fewerparticipants chose to buy the product when they promised a donation to charity. As Jang and Chu (2012) explain, consumers are willing to increase the fairness of a transaction with some additional cost (i.e. by paying an appropriate amount of money). However, they tend to give up when they perceive this cost as too high, as already explained by the average unwillingness to pay a premium price. Consumers would rather not receive the product than risk paying too little and damage their positive self-image.
Therefore, the perceived fairness seems to be even more important when consumers buy a PWYW product that is linked to a social cause. Some researchers examined the effect of cause-related marketing in PWYW (e.g. Gneezy et al., 2012). No studies have been found which focused on PWYW for ethical products. In this study the WTP for ethical products in
13 PWYW is measured. Moreover, it is examined to what extent a donation to charity affects this WTP.
H3: Consumers are willing to pay more for ethical products in PWYW when a donation to charity is made, rather than when it is not made.
2.2.4 Brand credibility
A broad definition of credibility is “the believability of an entity’s intentions at a particular time”. It consists of trustworthiness and expertise. Brand credibility refers to “the believability of the product position information contained in a brand”. Trustworthiness of a brand has a positive effect on the likeability of preferring the brand. (Erdem & Swait, 2004, p. 192).
This perceived trustworthiness makes organizations scared to communicate about CSR, due to a skeptical society. When consumers are skeptical about the credibility of the information, a negative effect may occur. When consumers have reasons to doubt the organization’s motives for CSR, they evaluate the organization more negatively than when it does not engage in CSR (Castaldo, Perrini, Misani, & Tencati, 2009). When consumers hear about CSR from the organization itself and perceive that the organization obtains high benefits from it, this backfire effect is most likely (Yoon, Gürhan-Canli & Schwarz, 2006).
At the same time, prior research suggests that there is still lack of information of ethical products. For organizations to differentiate themselves with social attributes, consumers must be fully aware of them. If consumers are not aware of the social attributes of a product, they might purchase a similar product without these aspects (McWilliams & Siegel, 2001). McWilliams and Siegel (2001) state that organizations not only should do good (i.e. actually implementing CSR), but also talk about doing good. Therefore, they state that advertising plays an important role for ethical consumers. The rise of the Internet gives
14 markets the chance to reduce the information asymmetry by using advertisement (Granados & Gupta, 2013). Eventually, it is the perceived trust that accounts for the success (or the failure) of a socially oriented company (Castaldo et al., 2009).
In Schmidt et al.’s (2014) experiment a significant fraction of PWYW participants paid for the strategic reason to keep the seller in business. They assume that an effortful, friendly seller may induce a buyer to pay more. Additionally, as mentioned above, ethical consumers want to support retailers who act in an ethical manner. It could therefore be expected that more participants want to keep the seller in business when they perceive them as ethical. This is in line with Kunter’s (2015) expectation that the effect of pro-social cues on WTP is moderated by the reputation of the seller. Therefore, in this study brand credibility is
taken into account. This leads to the following hypothesis.
H4: Consumers are willing to pay more for ethical products in PWYW when they perceive the brand as credible, rather than not credible.
15 3. Method
3.1 Data collection
In this study a field experiment was conducted to measure the WTP for an ethical product under PWYW. Contrary to laboratory experiments, in field experiments control is more challenging. However, it gives a researcher the opportunity to observe its effect in practice, and is therefore more realistic. Because of this realism, field experiments provide a lot of practical implications (Saunders, Lewis & Lewis, 2012).
The platform adopted for this study was Goodclub. It is a Dutch platform for people
with a sustainable mindset. Goodclub sells products with social as well as environmental elements. Although it would be relevant to find out the difference between them, in this study different measures are done for one specific ethical product with environmental elements. It was expected that due to the seller’s sustainable emphasis and participants’ interest in this topic unethical behavior (i.e. paying an unfair price) was less likely. The ethical product was a
sustainable water bottle from the brand ‘MiZu’. The data were collected through a Qualtrics survey, which was assigned for this experiment. It was conducted through the weekly newsletter of Goodclub. The newsletter was send to approximately 45.000 participants on March 4th, 2016 07:00 p.m. The water bottles were sold out after twenty hours, and therefore the survey was closed on March 5th, 2016 01:00 p.m.
The experiment proceeded as follows. First, after reading a short introduction about the experiment, members of the newsletter could participate in the experiment by clicking on the invitation link (Appendix 1). Second, participants received an explanation about the procedure of the experiment (Appendix 2) Third, participants were randomly allocated into one of the four treatments. They were shown the product and some additional information about the sustainable element of the product. Depending on which treatment they were in, participants also saw an announcement about a donation to charity or a text box that
16
underscored the fairness of PWYW according to Goodclub. This is further explained in the part below, as well as shown in Appendix 3. Fourth, participants were asked which color they preferred, what price (zero was possible) they thought was appropriate with shipping costs included, and if they would actually purchase the water bottle for that price. Fifth, seven or eight statements were shown, depending on the treatment. The statements contained the perceived interest in the product, sustainability of the product, fairness of the price, fairness of PWYW, relative fairness of PWYW compared to fixed prices, trustworthiness of Goodclub, and (if assigned to charity treatment) additional value of charity. Participants were asked to indicate to what extent they agreed with the statements. A five-point Likert Scale was used (i.e. strongly agree – strongly disagree). Sixth, those participants that wanted to purchase the product were asked to enter their personal information. They were also asked if they could name their motivation to buy the product. The survey ended for them with a thank you note and a confirmation that the product would be sent within ten working days. Thereafter, they were automatically sent to the webpage of Goodclub, i.e. www.goodclub.nl. Participants who did not want to purchase the product were only asked about their gender and age. The survey ended for them with a thank you note, after which they were automatically sent to the webpage of Goodclub. Thus, the survey slightly differed between participants that did (not) purchase the product. See Appendix 2 for the whole survey.
3.2 Research design
The field experiment had a 2 x 2 design containing four different treatments. The participants were randomly divided among the four treatments (Table 1, Appendix 3).
17 Underscoring fairness No Yes Charity No T1 T2 Yes T3 T4 Table 1: Experimental Design
In the charity condition the price paid by the subject went to charity, i.e. ‘Milieudefensie’. Since 1971 Milieudefensie has been working toward a more sustainable and fair country. The association now contains 84,000 members and donors. This price was paid in ‘Goodcoins’, an imaginary currency that can be converted into real money. Members receive a Goodcoin as a reward for ‘doing good’, in this case buying a product from one of Goodclub’s partners. Every Goodcoin equals one euro in the Goodclub shop. They can be saved, traded for new ‘good’ products, or donated to charity. Participants were familiar with Goodcoin, as Goodclub has been using this currency for years. For the underscored fairness, the content that was communicated was written by the researcher and approved by an employee the Goodclub.
3.3 Sample
The sample size (N) was 559. Some participants were filtered out. First, 213 participants had missing data on the first statement. Since the statements were mandatory to fill out, missing values meant incomplete surveys. There was no significant difference between the different treatments for the incomplete surveys (p= .309), i.e. treatment 1: 23.4%, treatment 2: 30.1%, treatment 3: 24.6%, and treatment 4: 23.1%. Second, from the participants who wanted to purchase the product, some did not fill out their personal information (N= 68).
18 Gender and age
The answers about gender from participants that purchased and did not purchase the product were merged together into one dummy variable. The sample consisted of 61.4% females. Participants were asked in which year they were born. The answers on this question from participants that purchased and did not purchase the product were also merged together. Since the experiment took place in the first week of March 2016, age was determined by subtracting the year of birth from 2015. Therefore, only participants born in January or February are shown as one year younger than in reality. The unlikely ages of 108, 114 and 115 and all unanswered values were categorized as missing variables (N= 5). Without these missing values, the mean age was 48 years with 13 years as youngest and 85 years as oldest participant (SD= 13.33). When ages are categorized (i.e. <20, 20-40, 40-65, 65-80, 80+), the category 40-65 consisted 61.9% of the sample (N= 346). As the youngest (N= 11) and the oldest (N= 7) categories consisted of too few people, ages are redistributed with an interval of 24 years, i.e. <37 (N= 132), 37- 61 (N= 331), 61+ (N=96).
19 4. Results
4.1 Descriptive statistics Willingness to Pay
In Table 2 the descriptive statistics are outlined. The average price participants indicated to be appropriate (WTP) for the water bottle was 11.35 euro, with a minimum of zero euros and a maximum of 35 euro (SD= 5.06). Only 1.6% of participants were willing to pay zero euro. When the price of zero euro was omitted, the average price was 11.53 euro (SD= 4.89) with a minimum of 1.00 euro. The average WTP in the four treatments ended between 10.81 and 11.67 euros.
Purchase
Only 37.9% of the participants wanted to purchase the product for the chosen price. Participants who purchased the water bottle, had a significant lower WTP than participants who did not purchase the water bottle. This effect size was moderate and significant (F (1,557) = 55.651, p < .000, η² = .09). The costs of the product for Goodclub were 14.52 euros (i.e. 8.57 euros for production and 5.95 euros for sending). Therefore, on average one product had a loss of 5.12 euros. As 81.6% of participants offered less than the costs of the product, this experiment resulted in a loss. Moreover, the product is usually sold for 26.98 euros, including sending. No purchasing participant offered this amount of money or more. For purchasing participants, the average WTP in the four treatment was between 8.63 and 10.41 euros.
20
Participants N % WTP (€) SD Minimum (€) Maximum (€)
All 559 100 11.35 5.06 0.00 35.00 No zero euro 550 98.4 11.53 4.89 1.00 35.00 Purchasing 212 37.9 9.40 5.05 0.00 25.00 Not purchasing 347 62.1 12.54 4.69 0.00 25.00 Base treatment 142 25.4 11.83 5.18 1.00 35.00 Charity treatment 133 23.8 11.05 4.99 0.00 25.00 Fairness treatment 143 25.6 11.67 5.00 0.00 29.00 Charity and fairness treatment 141 25.2 10.81 5.07 0.00 25.00 Table 2: descriptive statistics of WTP
Statements
Table 3 and Table 4 outline the results of the extent to which participants agreed with the seven statements (1: Strongly agree – 5: Strongly disagree).
Statement Mean SD Purchasing participants SD
Trustworthiness Goodclub 2.03 .71 1.94 .71
Fairness price 2.17 .79 2.33 .90
Sustainability water bottle 2.06 .78 1.89 .70
Fairness pricing mechanism 2.76 .86 2.68 .87
Relative fairness pricing mechanism 2.93 .91 2.88 1.00 Familiarity with pricing mechanism 3.29 1.14 3.40 1.34
Interest product 2.72 1.16 1.79 .62
Table 3: descriptive statistics of statements
Statement T1 SD T2 SD T3 SD T4 SD
Trustworthiness Goodclub 2.00 .74 2.05 .66 2.06 .75 2.02 .68 Fairness price 2.12 .73 2.21 .69 2.22 .84 2.15 .89 Sustainability water bottle 2.10 .85 2.13 .71 2.01 .73 1.99 .80 Fairness pricing mechanism 2.68 .87 2.74 .85 2.78 .88 2.84 .84 Relative fairness pricing mechanism 2.85 .95 2.93 .90 2.96 .86 2.99 .93 Familiarity with pricing mechanism 3.30 1.32 3.41 1.24 3.27 1.40 3.17 1.93 Interest product 2.84 1.25 2.73 1.10 2.69 1.16 2.62 1.11 Table 4: descriptive statistics of statements per treatment
21 4.2 Hypotheses testing
4.2.1 Hypothesis 1 – Perceived fairness
H1: Consumers are willing to pay more for ethical products in PWYW when they perceive the
pricing mechanism as more fair.
Perceived fairness PWYW
Participants were asked whether they thought the pricing system was fair (fairness PWYW) and more fair than fixed prices (relative fairness PWYW). The two statements have a strong, positive correlation (r=.63, N=559, p<.001). The alpha coefficient is .772 and corrected item – total correlation is .629. This suggests that the statements have an acceptable internal consistency. Scale variance increases for both variables if an item is deleted. The items are not taken together as a factor for the fairness of the system. Yet, the statements both seem to measure the perceived fairness of the system.
Nearly half (42.2%) of the participants neither agreed nor disagreed with fairness PWYW (M= 2.76; SD= .859). Likely, 45.4% of participants neither agreed nor disagreed with relative fairness PWYW (M= 2.93; SD=.907). When the third value (i.e. Neither agree nor disagree) was named as missing value, the sample size became 323. Of those participants 67.8% thought the pricing system was fair, 53.8% thought the pricing system was more fair than fixed prices.
Effect of perceived fairness PWYW on WTP
There was a weak significant effect of fairness PWYW on WTP (F(4,554) = 4.885, p= .001, η² = .03). Tukey post-hoc test revealed that this effect was only significant between the two highest and lowest prices (.004 < p < .025). For relative fairness PWYW this effect is weak and significant as well (F(4, 554) 5.203, p< .001, η² = .04), and was only significant between the lowest price and the three highest (.001 < p < .033). Participants that neither agreed nor
22 disagreed with relative fairness PWYW paid the highest amount (M=12.09; SD= 5.00) and participants that strongly disagreed the lowest (M= 7.44; SD= 7.75), see Figure 1.
Figure 1: perceived (relative) fairness PWYW and WTP
When only purchasing participants were taken into account, the size effect became moderate for both fairness PWYW (F(4, 207) 5.353, p< .001, η² = .09) and relative fairness PWYW (F(4, 207) 3.507, p= .009, η² = .06), see Figure 2. For both statements, the effect is only significant between participants that strongly disagreed, compared to the other levels (.002 < p < .030). Only for perceived fairness of PWYW, the difference between participants that disagreed and strongly disagreed was not significant for WTP either (p= .081). Thus, H1 is adopted.
Figure 2: perceived (relative) fairness PWYW and WTP for purchasing participants € 6.00 € 7.00 € 8.00 € 9.00 € 10.00 € 11.00 € 12.00
Fairness PWYW Relative fairness PWYW
Strongly agree Agree
Neither agree nor disagree Disagree Strongly disagree € 0 € 2 € 4 € 6 € 8 € 10 € 12
Fairness PWYW Relative fairness PWYW
Strongly agree Agree
Neither agree nor disagree Disagree
23 4.2.2 Hypothesis 2 – Underscoring fairness
H2: Consumers perceive PWYW as a more fair pricing mechanism when this fairness is underscored, rather than when it is not underscored.
Underscoring fairness
An additional text box which explained the fairness of PWYW (underscoring fairness) was implemented in two treatments, i.e. only underscoring fair (N= 143) and charity plus underscoring fairness (N= 141), see appendix 3. There is no effect of underscoring fairness on fairness PWYW (F(1, 557) 1.791, p=.181, η² = .00) nor relative fairness PWYW (F(1, 557) 1.213, p=.271, η² < .01). This effect is also no significant effect if only purchasing participants are taken into account (p= .115). Moreover, there is neither a direct effect of underscoring fairness on trustworthiness Goodclub (p= .181), nor fairness price (p= .771). Thus, H2 is rejected.
Interaction effects underscoring fairness on fairness PWYW
The interaction effect of gender on the effect of underscoring fairness on fairness PWYW was not significant (p= .343). The interaction effect of age on the effect of underscoring fairness on fairness PWYW was significant (b= -.013, t(550) -2.132, p=.034), but only for the youngest third of participants (b= .107, t(550) = 2.55, p= .011). One way ANOVA showed that participants younger than 42 years who were not in the underscoring fairness treatments significantly stronger agreed with fairness PWYW (M= 2.70; SD= .890) than who were in the underscoring fairness treatments (M= 3.03; SD= .835), see Figure 3.
Figure 3: amount of participants < 42 years with extend of agreement on fairness PWYW 10
20 30 40
(Strongly) agree Neither agree nor disagree (Strongly) disagree
No fairness claim Fairness claim
24 There were no interaction effects of interest (p= .465) nor sustainability (p= .078). Neither were there interaction effects of trustworthiness Goodclub (p= .778), fairness price (p= .771), nor interest (p= .184) for the effect of underscoring fairness on fairness PWYW. For relative fairness PWYW, no interaction effects were found for the above mentioned variables, i.e. gender (p= .111), age (p= .289), interest (p= .137), sustainability (p= .248), trustworthiness Goodclub (p= .778), and fairness price (p= .771).
There was no direct effect of underscoring fairness of PWYW on WTP (p= .617) neither for participants younger than 42 years (p= .487). There was also no interaction effect of the perceived (relative) fairness of PWYW on the effect of underscoring the fairness of PWYW on WTP (p= .069), neither for participants younger than 42 years (p= .111)
4.2.3 Hypothesis 3 – Donation to charity
H3: Consumers are willing to pay more for ethical products in PWYW when a donation to charity is made, rather than when it is not made.
Donation to charity
A donation to charity was implemented in two treatments, i.e. only charity (N= 133) and charity plus underscoring fairness (N= 141). Participants in the charity treatments paid less (M= 10.93; SD= 5.02) than participants in the treatments without charity (M= 11.75; SD=5.08). However, the effect of charity on the willingness to pay was only significant on a 90% confidence interval (F(1, 557) 3.713, p= .055, η² = .01). When purchasing participants alone are taken into account, there is a small effect for WTP (F(1,210) 4.738, p= .031, η² = .02). Purchasing participants in the charity treatment paid a significant lower price (M= 8.66; SD= 5.04), compared to the other treatments (M= 10.16; SD = 4.96), see Figure 4. Therefore, H3 is rejected. More importantly, the results show that there is evidence of an opposite effect. Consumers seem to pay a lower price for ethical products in PWYW when a promise to
25 donate to charity is added, rather than when it is not added. The results of H4 might explain this paradoxical effect.
Figure 4: WTP for treatments with (no) charity
There was no interaction effect on the effect of charity on WTP for age (p= .527), nor gender (p= .615), neither when only looking at purchasing participants (age: p= .570; gender:
p= .246).
4.2.4 Hypothesis 4 – Brand credibility
H4: Consumers are willing to pay more for ethical products in PWYW when they perceive the brand as credible, rather than not credible.
Brand credibility
Participants were asked whether they thought Goodclub is trustworthy. There is a weak significant effect of the perceived trustworthiness of Goodclub on WTP (F(4, 554) 5.605, p< .001, η² = .04). Participants that strongly agreed paid a higher price (M= 12.58; SD= 5.67), than participants that neither agreed nor disagreed (M=9.71; SD=5.06) and strongly disagreed (M= 10.50; SD= 2.12). Purchase had a significant interaction effect on the effect of trustworthiness Goodclub on WTP (b= -.127, t(555) -2.018, p= .044), see Figure 5. Thus, the effect of trustworthiness Goodclub on WTP was stronger for purchasing participants (F(2,209) 13.164, p < .001, η² = .11) than for participants that did not purchase the product
10.93 8.66 11.75 10.16 € 8.00 € 9.00 € 10.00 € 11.00 € 12.00
Participants Purchasing participants
WTP Charity WTP No charity
26 (F(4, 342) 2.345, p= .054, η² = .03). For participants that did not purchase the product, this effect was only significant on a 90% confidence interval. Thus, H4 is adopted.
Figure 5: difference in WTP between relative difference in trustworthiness Goodclub.
Purchasing participants that strongly agreed paid a higher amount (M= 11.30; SD= 5.65) than participants that neither agreed nor disagreed (M= 6.54; SD= 4.93). There were no purchasing participants that (strongly) disagreed with trustworthiness Goodclub. Therefore, Figure 5 only displays the relative high and low trust in Goodclub.
4.3 Other effects
Some additional effects were found and are discussed in the following section. 4.3.1 Controlling effects of perceived fairness of PWYW
Familiarity PWYW
Participants were asked whether they agreed that they were already familiar with the pricing system (familiarity PWYW). Of all participants, 36.5% (strongly) agreed they were already familiar with PWYW. There was a very weak statistically significant effect of familiarity PWYW on variable fairness PWYW (F(4,554) 3.532, p= .007, η² = .03). Tukey post-hoc tests revealed that the amount to which the participants agreed with fairness PWYW to be significantly stronger in the group that strongly agreed with familiarity PWYW (M=2.40; N=52; SD=1.07), compared to those who disagreed (M= 2.83; N= 151; SD= .80; p= .015) and strongly disagreed (M= 2.88; N= 133; SD= .90; p= .006).
13.36 12.59 11.83 10.85 9.19 7.53 € 7.00 € 8.00 € 9.00 € 10.00 € 11.00 € 12.00 € 13.00 € 14.00
High trust Average trust Low trust
No purchase Purchase
27 Trustworthiness of Goodclub
There were only four participants that (strongly) disagreed with trustworthiness Goodclub. Therefore, a dummy variable was made for participants that (strongly) agreed. The effect of trustworthiness Goodclub on fairness PWYW was significant (F(1,557) 9.170, p=.003, η² = .02). The amount to which participants agreed with fairness PWYW was significantly stronger in the group that (strongly) agreed with trustworthiness Goodclub (M= 2.70; SD= .83), compared to those who did not (M=2.96; SD= .92). For relative fairness PWYW, this effect was only significant on a 90% confidence interval (F(1,557) 3.029, p= .082, η² = .02).
When only purchasing participants were taken into account, there is a very weak significant effect for both fairness PWYW (F (1,210) 8.365, p= .004, η² = .02) and relative fairness PWYW (F (1, 210) 4.301, p= .038, η² = .02). From purchasing participants that (strongly) agreed that Goodclub was trustworthy (N= 424) 48.5% (strongly) agreed with fairness PWYW and 39.4% with relative fairness PWYW, compared to 27.7% and 29.8% of other purchasing participants.
Interest product
There is a very weak significant effect of the interest in the product on fairness PWYW (F(4,554) 2.775, p= .026, η² = .02) and relative fairness PWYW (4,554) 2.764, p=.027, η² = .02). Looking at the effects the other way around, relative fairness PWYW has a low significant effect on interest (F(4, 554) 5.257, p< .001, η² = .04). Participants that strongly agreed with interest product also stronger agreed with relative fairness PWYW (M= 2.18; SD= 1.40).
Fairness price
There is a moderate significant effect of the perceived fairness of the offered price (fairness price) on fairness PWYW (F(4, 554) 11.819, p < .001, η² = .08) and relative fairness PWYW
28 (F (4,554) 12.121, p < .001, η² = .08). Participants that (strongly) agreed also stronger agreed with fairness PWYW and relative fairness PWYW, and the other way around.
When only purchasing participants are taken into account, there is a high effect size for both fairness PWYW (F (4, 207) 8.718, p < .001, η² = .14) and relative fairness PWYW (F (4, 207) 12.452, p < .001, η² = .19). From purchasing participants that strongly agreed with fairness price 65.4% (strongly) agreed with fairness PWYW and 53.8% with relative fairness PWYW, compared to 14.3% and 0.0% of purchasing participants that strongly disagreed.
Sustainability product
There is a very weak significant effect of the perceived sustainability of the product (sustainability product) on fairness PWYW (F(4, 554) 4.330, p= .002, η² = .03) and relative fairness PWYW (F (4,554) 4.161, p= .002, η² = .03). From participants who strongly agreed with the sustainability of the product 44.0% (strongly) agreed with fairness PWYW and 29.6% with relative fairness PWYW, compared to 20.0% and 0.0% of participants who strongly disagreed.
4.3.2 Regression analysis for perceived fairness PWYW
The predictors in Model 1 explained for 12.3% the total variance in the perceived fairness of PWYW (p < .001). Only the statement about the trustworthiness of Goodclub did not have significant Beta value (p=.639). After entering gender and age, Model 2 was significant as well (p < .001) and explained 13.6% of total variance. However, gender did not have a significant Beta value (p=.146). Therefore, both trustworthiness Goodclub and gender were omitted from the model. The final model explained 13.1% of the total variance in fairness PWYW. All five predictor variables (i.e. familiarity PWYW, trustworthiness Goodclub, fairness price, interest product, and sustainability product) were statistically significant (<.001
29 < p < .031), see table 5. The more participants agreed with the statements, the more they agreed with the fairness of PWYW.
Model 1 Model 2 Model 3
Familiarity PWYW .076 (.026)*** .081 (.026)*** .086 (.026)*** Trustworthiness Goodclub .024 (.052) .063 (.053) . Fairness price .294 (.045)*** .267 (.045)*** .280 (.044)*** Interest product .086 (.031)*** .094 (.032)*** .099 (.031)*** Sustainability product .114 (.048)** .118 (.048)** .131 (.047)*** Gender . .104 (.071) . Age . -.006 (.003)** .006 (.003)** R2 .123*** .136*** .131*** dF 5 7 5 F 15.503 12.286 16.554 *p < 0.1, **p < 0.05, ***p < 0.01
Table 5: hierarchical multiple regression for fairness PWYW
Likewise, the same predictors were entered in a hierarchical multiple regression for relative fairness PWYW. Model 1 explained for 10.0% of the total variance and was statistically significant (p < .001). Both trustworthiness Goodclub (p= .331) and familiarity PWYW (p= .614) did not have significant Beta value. After entering gender and age, the model was still significant (p < .001) and explained 12.7% of total variance. Both trustworthiness Goodclub (p=.769) and awareness (p=.340) still did not have significant Beta value and were omitted from the model. The final model explained 12.6% of the total variance for the extent to which participants agreed that PWYW is more fair than fixed prices. Three out of four predictor variables (i.e. fairness price, interest product, age, and gender) were statistically significant. Only sustainability product was not significant (p=.105). The more participants agreed with the statements, the more they agreed with the fairness of PWYW. Moreover, of all males 34.7% (strongly) agreed with the relative fairness of PWYW, compared to 25.9% of females. Older participants stronger agreed as well. Tukey post-hoc
30 tests revealed the youngest group of participants significantly less strong agreed (M= 3.16; SD= .93), compared to the middle group (M= 2.89; SD= .818; p= .010), and the oldest group (M= 2.78; SD= 1.10; p= .005). See table 6.
Model 1 Model 2 Model 3
Familiarity PWYW .014 (.028) .026 (.028) . Trustworthiness Goodclub -.054 (.056) -.017 (.057) . Fairness price .328 (.048)*** .294 (.048)*** .295 (.047)*** Interest product .099 (.033)*** .121 (.034)*** .118 (.033)*** Sustainability product .087 (.052)*** .087 (.051)*** .080 (.049) Gender . -.140 (.076)*** -.148 (.074)** Age . .010 (.003)*** .009 (.003)*** R2 .100 .127 .126 dF 5 7 5 F 12.283 11.371 15.758 *p < 0.1, **p < 0.05, ***p < 0.01
Table 6: hierarchical multiple regression for relative fairness PWYW
4.3.3 Interest and sustainability
Of all participants 76.2% (strongly) agreed with that the product was sustainable. There was a moderate significant effect of sustainability on interest product (F(4, 554) 12.969, p < .001, η² = .09). The more participants agreed with sustainability the more they agreed to be interested in the product (strongly agree: M= 2.29; SD= 1.31, strongly disagree: M= 4.40; SD= .89). Not surprisingly, there was a weak significant effect of interest in the product on WTP (F(4, 554) 4.354, p= .002, η² = .03). Of purchasing participants 91.5% (strongly) agreed with interest in the product, compared to 23.4% of participants who did not purchase. Moreover, there was a significant weak effect of perceived sustainability on purchase (F(4, 554) 4.254, p= .002, η² = .03). Of purchasing participants, 84.4% (strongly) agreed with sustainability. The effect of the perceived sustainability of the product on WTP was weak and significant as well (F(4,554) 4.092, p= .003, η² = .03). Participants paid a higher price when they strongly agreed with
31 12.73 13.42 12.01 9.86 9.95 8.51 7.08 5 1.38 0 € 0 € 2 € 4 € 6 € 8 € 10 € 12 € 14
Participants Purchasing participants
Strongly Agree
Agree
Neither agree nor disagree
Disagree
Strongly disagree sustainability (M=12.34; SD=5.74), compared to those who strongly disagreed (M=6.60; SD=3.65). This effect was of moderate size for purchasing participant (F(4, 207) 3.593, p= .007, η² = .07).
4.3.4 Perceived fairness of price
Participants were asked whether they thought they had offered a fair price for the product (fairness price). WTP and the perceived fairness of the price were moderately correlated (r= -.322; N= 559; p< .001). The higher the WTP was for the product, the stronger they agreed to have offered a fair price. Moreover, there was a weak significant effect of the perceived price fairness on purchase (F(4,554) 4.471, p= .001, η² = .03). The more participants agreed to have paid a fair price, the less they actually bought the product.
Additionally, for purchasing participants, this correlation was stronger yet still moderate (r= -.460; N= 212; p< .001). Purchasing participants that strongly agreed with fairness price also had a higher WTP (M=13.42; SD= 6.04), compared to participants that strongly disagreed (M= 1.57; SD= 1.62). Nevertheless, 17.9% of participants that (strongly) agreed to have paid a fair price purchased the product for five euros or less. This price is far below cost price. For ten euros or less this is even 57.9%. This is illustrated in Figure 6.
32 4.3.5 Alternative explanations charity effect on WTP
It is remarkable that participants want to pay less in the charity conditions. When asked whether the charity component added value for them 78.2% of participants agreed or even strongly agreed (M= 2.05, SD=.928). Participants that strongly agreed had a higher WTP (M= 12.17; SD=5.03), compared to those that strongly disagreed (M= 8.40; SD=7.09). However, the effect of the perceived value of charity on WTP was not significant (F(4,128)= 1.194, p= .317). This might be due to the fact that there were not enough participants for the neutral (N= 21) or (strongly) disagree value (N= 8). For purchasing participants this effect was of moderate size, yet only significant on a 90% confidence interval (F(1,51) 2.989, p= .090, η² = .06). It was not possible to measure interaction effects for this variable due to the relatively large number of participants (strongly) agreed with the statement.
There was no interaction effect of sustainability on the effect of charity treatments on WTP (b= .08, t(555) .13, p= .895), neither for trustworthiness Goodclub (p= .179), fairness PWYW (p= .333), nor relative fairness PWYW (p= .403).
However, since the effect of charity was only significant for purchasing participants, the mediating effect of fairness PWYW was measured only for purchasing participants. First, the bias-corrected and accelerated confidence interval (BCA95= [-.804, -.021] excluded zero, indicating that fairness PWYW mediated the effect of charity on WTP. The total indirect effect of fairness PWYW on WTP was found to be significant (b= -1.127, p= .004) whereas charity did not statistically predict WTP (b= -1.225, p= .074). Thus, fairness PWYW mediated the effect of charity on WTP for purchasing participants. The direct effect of charity on fairness PWYW was weak and significant for purchasing participants (F(1, 210) 3.979, p= .047 η² = .02). Participants in the charity conditions less strongly agreed and more strongly disagreed with fairness PWYW (M= 2.80, SD= .925), than those in other conditions (M= 2.56, SD= .810). Consequently, participants paid a lower price. Moreover, the difference
33 between the treatments with and without charity was bigger when participants (strongly) agreed or disagreed, compared to when they neither agreed nor disagreed. This is shown in Figure 7.
Figure 7: Mediating effect of fairness PWYW on effect of charity on WTP for purchasing participants.
Nevertheless, for all values of fairness PWYW the WTP is lower in the charity treatments. There were no other interaction effects for age (p= .838), gender (p= .262), relative fairness PWYW (p= .357), sustainability (p= .590), nor trustworthiness Goodclub (p= .321).
4.3.6 Gender
It has already been mentioned that males were more likely to agree with the statement that PWYW is more fair than fixed prices, compared to females.
Moreover, there was a very weak effect of gender on purchase (F (1,557) = 8.383, p= .004, η² = .03). From all male participants 45.4% purchased the water bottle, against 33.2% of female participants. More interestingly, there was a significant interaction effect of gender on the effect of fairness PWYW on WTP (b= 1.0377, t(555) 2.049, p= .041). For males, there is
€11.62 €10.13 €10.28 €9.50 €1.00 €9.58 €9.30 €9.81 €6.36 €0.67 0 5 10 15 20 25 30 35 40 45 50
Strongly agree Agree Neiter agree nor disagree
Disagree Strongly disagree
No charity
34 a moderate effect of fairness PWYW on the WTP (F(4, 211) 7.052, p< .001, η² = .12), as shown in Figure 8.
Figure 8: difference fairness PWYW for gender
For males there was an inverted U-shape effect of relative fairness PWYW on WTP (F(4, 211)= 4.900, p= .001), as shown in Figure 9. For females, this effect is not significant (p= .445).
Figure 9: difference relative fairness PWYW for gender 4.5.7 Age
As mentioned earlier, the youngest group of participants (< 42 years old) less strongly agreed with the statement that PWYW is more fair than fixed prices. When the fairness of PWYW was underscored, this group agreed less with the statement that PWYW is fair.
11.27 10.17 9.08 12.24 12.03 11.83 € 8.50 € 9.00 € 9.50 € 10.00 € 10.50 € 11.00 € 11.50 € 12.00 € 12.50
High fairness Average fairness Low fairness
Male Female € 9.68 € 10.40 € 11.49 € 9.06 € 4.60 € 10.36 € 11.68 € 12.42 € 11.93 € 10.61 € 3 € 4 € 5 € 6 € 7 € 8 € 9 € 10 € 11 € 12 € 13
Strongly agree Agree Neither agree nor disagree
Disagree Strongly disagree
35 Moreover, there is a very weak effect of age on the purchase of the product (F (1,552) = 11.370, p < .001, η² = .02). Participants that purchased the product were statistically younger (M= 45.57; SD= 14.04) than participants that did not buy the product (M= 49.47; SD=12.68). Additionally, purchase had an interaction effect of moderate size on the effect of age on WTP (b= .156, t(550) 4.595, p< .001), see Figure 10. The difference in WTP for participants that purchased and did not purchase the product was significantly bigger for young participants, compared to old participants. All other measured (interaction) effects of age were not significant.
Figure 10: interaction effect of purchase on effect of age on WTP
4.5.8 Regression analysis for WTP
All statements had been reversed in value order, i.e. value 1 as “Strongly agree” was now value 5 and the other way around. This was important for the interpretation of the b-value. Thus, the higher the value of the statement the stronger the participant agreed with the statement.
The predictors in significant Model 1 explained for 22.2% the variance in WTP (p < .001). Statements about the interest in the product (p= .063) and the fairness of PWYW (p= .110) did not have a significant Beta Value. After entering gender and age, Model 2 was significant as well (p < .009) and explained 23.5% of total variance in WTP. Age did not have
8.44 9.65 10.86 13.51 12.65 11.78 8 9 10 11 12 13 14
Young Average Old
Purchase No purchase
36 a significant Beta value (p= .865), neither did interest in product (p= .074) and fairness PWYW (p= .073). As the statements for interest in product and fairness PWYW were significant on a 90% confidence interval, only age was omitted from the model. The final model explained 23.5% of the total variance in WTP (p < .001). Six of eight predictor variables were statistically significant on a 95% confidence interval, the other two on a 90% confidence interval, see table 7. The more participants agreed with the statements, the more they were willing to pay. The exception is the perceived relative fairness of PWYW. This statement has a negative effect on WTP. Participants that neither agreed nor disagreed had the higher WTP. Moreover, females had a higher WTP (M= 12.03; SD= 4.75) compared to males (M= 10.27; SD= 5.36). As mentioned before, participants that did not purchase the product had a higher WTP as well.
Model 1 Model 2 Model 3
Fairness PWYW .466 (.291) .524 (.291)* .486 (.289)*
Relative fairness PWYW -.592 (.274)** -.585 (.277)** -.546 (.272)**
Trustworthiness Goodclub .836 (.289)*** .636 (.298)** .705 (.290)** Fairness price 1.60 (.263)*** 1.668 (.265)*** 1.640 (.261)*** Interest product .408 (.219) .396 (.221)* .359 (.218)* Sustainability product .794 (.269)*** .793 (.269)*** .788 (.267)*** Purchase product -3.690 (.508)*** -3.486 (.514)*** -3.445 (.511)*** Gender . 1.235 (.403)*** 1.210 (.399)*** Age . .003 (.514) . R2 .222*** .235*** .235*** dF 7 9 8 F 22.502 18.534 21.134 *p < 0.1, **p < 0.05, ***p < 0.01
37 5. Discussion
Consumers seem to increasingly care about social and environmental issues correlated to the products they buy. However, for these ethical products a premium price is asked, which most consumers are not willing to pay. Mendoza-Abarca & Mellema (2015) argue that organizations with a social aim need to implement mechanisms that simultaneously create social value and an appropriate amount of economic value. Herefore, they expect Pay What-You-Want (PWYW) to be a promising mechanism.
5.1 WTP in PWYW
In line with previous literature on PWYW (e.g. Kim et al., 2009), prices paid in this experiment significantly differed from zero. Of all participants, 98.4% paid above zero euros. However, 77.8% of the participants that thought zero euros was an appropriate price, purchased the product. Contrary to previous literature (Schmidt et al., 2014), 81.6% of purchasing participants offered less than the costs of the product. Pelsmacker et al. (2006) outline that 10% was willing to pay a price premium. In this study, no participant was willing to purchase the product for the selling price. Although participants were not aware of these prices, 26.9% of purchasing participants paid five euros or less, which is not a realistic price. This assumes that these participants did not mind violating the norm, contrary to what Gneezy et al. (2012) found. Since the perceived fairness of the paid price was correlated with the WTP, purchasing participants that paid a very low price were fully aware of their unethical behavior. Potentially, these participants just wanted to get great bargains and therefore paid a very low price (Jang & Chu, 2012).
Nonetheless, the majority of participants did not purchase the product even though someof them were interested in the product. The mean WTP of participants that did not buy the product was significantly higher, compared to purchasing participants. This could be
38 explained with two theories from the literature. First, purchasing participants wanted to get great bargains and offered a low price (Jang & Chu, 2012). Second, participants might have considered what they thought was a fair price. Thereafter, they decided this price was too high for them and did not purchase it. Indeed, participants that agreed to have paid a fair price paid a higher price. The stronger participants agreed with the fairness of the price and the higher their WTP, the lower the number of participants who purchased the product. This is in line with Gneezy et al.’s (2012) argument that consumers give up when they perceive the costs of increasing fairness as too high. Nevertheless, some participants that agreed with fairness price and purchased the product, still paid prices far below cost price.
5.2 Perceived fairness
Because the consumer has full control over the price, PWYW could be perceived as more fair than fixed prices (Kim et al., 2009; Mendoza-Abarca & Mellema, 2015). As this assumption was based on an experiment in auctions, this study was the first to measure the perceived fairness of PWYW. It was found that nearly half of participants neither agreed nor disagreed with the statements that PWYW was fair and that PWYW was more fair than fixed prices.
The perceived (relative) fairness of PWYW had an effect on WTP. Surprisingly, for both fairness of PWYW and relative fairness of PWYW the highest WTP was for participants that neither agreed nor disagreed. Thus, when participants were not sure if they thought PWYW was (more) fair, they paid a higher amount of money.
Underscoring the fairness of PWYW did not have an effect on the perceived fairness of PWYW nor on the perceived relative fairness of PWYW. This is in line with Kunter (2015) who did not find effects on textual cues about fairness either. Only for purchasing participants younger than 42 years there was a significant negative effect of underscoring the fairness of PWYW on the perceived fairness of PWYW. This effect showed exactly the opposite as what
39 was expected in the hypothesis. However, the effect is in line with the negative effect of CSR, mentioned by Castaldo et al. (2009). It highlights the importance of the credibility of information about CSR. If the information is perceived as not credible, a backfire effect might occur.
Besides underscoring the fairness of PWYW, some interaction effects occurred for perceived fairness of PWYW. First, the effect of familiarity on the perceived fairness of PWYW highlights the importance of informing consumers about PWYW. At the same, since the familiarity did not have an effect on the perceived relative fairness of PWYW, making them aware does not make them perceive the pricing system as more fair than fixed prices. Second, the effect of the trustworthiness of the seller (i.e. Goodclub) is in line with Pearson and Henryks (2008). They argue that consumers rely on the retailer’s reputation to confirm ethics. It also assumes that for organizations of which customers have low trust it might not be good to implement PWYW. Third, for relative fairness of PWYW, the effect with interest was two-sided. Thus, participants who thought that PWYW was more fair than fixed prices were also more interested in the product. This highlights the marketing aspect of PWYW, named by Kim et al. (2009). Lastly, sustainability had a very weak effect on both statements as well. This assumes that PWYW is more suitable for sustainable products. However, since this experiment took place at a platform for sustainability, this is not remarkable.
5.3 WTP for charity in PWYW
Both Elfenbein and McManus (2010) and Gneezy et al. (2012) showed in their experiments that consumers are willing to pay more when a part of the revenue is donated to charity. Contrarily, the results of this study showed that participants in the charity treatments paid a lower price for the product, compared to those in the other treatments. This effect was only
