Leasing green : the willingness to pay for sustainable and leasable fashion

(1)

Leasing green

The willingness to pay for sustainable and leasable fashion

Olaf Schouws 6023088

MSc Business studies, EMCI Track Master Thesis

Supervised by B. Kuijken MSc 30 June 2014

(2)

Abstract

The aim of this study is to determine consumers‟ willingness to pay (WTP) for

sustainable versus conventionally produced products, and to explore the role of leasing sustainable products. In addition, the moderating effects of perceived quality and fashionableness are investigated. Contrary to previous research, this study incorporates the effect of both sustainability and leasing by using an experimental method. The

empirical setting is the jeans industry. An experimental auction with 809 participants was used to estimate WTP, along with a follow-up survey to collect information about the perceived quality and fashionableness of sustainable products. Although sustainability and leasing had some positive effect on WTP, (a price premium of 12%) a clear

significant result has not been found. The results did indicate significant effects of both age and education on WTP. Younger, higher educated consumers have a higher WTP than older, lower educated consumers. This study contributes to the literature by further

investigating the effect of sustainability on WTP and by making a first attempt to

incorporate the effects of leasing on WTP. This topic could be further researched by using other product types and a more representative sample.

(3)

1.0 Introduction

The last couple of years the concept of sustainability is appearing in every industry and fashion is no exception to this (Casadesus-Masanell et al, 2009; Dong, Richards, & Feng, 2013; Kuik, 2005). With several scandals covered by the media about abuses and even the collapsing of factories in Bangladesh, the consumer now knows about the social and environmental problems concerned with the fashion industry (Kazmin & Allchin, 2013; Kuik, 2005). Surprisingly, still many fashion brands are reluctant of incorporating a more sustainable policy because of the costs involved. This seems strange because according to previous research the consumer is willing to pay a premium for sustainable fashion (Casadesus-Masanell et al, 2009; Ellis et al., 2012; Hustveld & Bernard, 2008; Lee, 2011). Because much of these results stem from research using a survey method, it is difficult to measure the actual willingness to pay (WTP) at the point of purchase. This is supported by the fact that over 50% of Europeans claim they are willing to pay a premium for sustainable fashion but the actual market share is only 1% (De Pelsmacker et al., 2005). Thus there seems to be a gap between what people say they do and what they actually do when it comes to buying sustainable fashion.

In this paper, the WTP for sustainable and leasable fashion is investigated through an experimental second price auction. To our knowledge the effects of both sustainability and leasing on WTP have never been investigated simultaneously. Previous research has mentioned the contradictory relationship between sustainability and leasing (Agrawal et al., 2012; Fishbein, McGarry, & Dillon, 2010). It has been argued whether leasing is the more sustainable option compared to purchasing. Ellis et al. (2012) studied the WTP for sustainable t-shirts through an experimental auction. They found that their respondents were willing to pay a premium for sustainable fashion. However, they argue that a more representative sample and the inclusion of other product categories are necessary to

(5)

generalize the relationships found in their study. The empirical setting of this study is the fashion industry, more specific, the jeans industry. This is relevant because jeans are an iconic piece of fashion (Harris, 2010; Kuik, 2004). More importantly, the concept of sustainability is very relevant to the jeans industry because the production of jeans has a large environmental impact. When producing jeans a lot of water and chemicals (e.g. bleach for „distressing‟) are needed. If jeans would be produced in a more sustainable manner it would be positive for socio-environmental circumstances. However, more insight into consumers‟ WTP for a pair of sustainable jeans is needed to see whether it is also economically viable to do this.

This study has several theoretical and managerial implications. First, with respect to existing theories, this study aims at further clarifying the influence of sustainability and leasing on the WTP. Because of contradictions in the literature about sustainability and leasing and the lack of generalization of previous research about the actual WTP at the point of purchase for sustainable products, more insight into this particular topic is needed (Agrawal et al., 2012; Ellis et al. 2012). Second, with respect to managerial implications, this study informs fashion brands about the importance of sustainability on an economical and ecological level.

2.0 Literature review

2.1 Sustainable products

Although existing literature about sustainable products presents many definitions of the term, it is still unclear what constitutes a sustainable product (Dangelico & Pujari, 2010). Chen (2001) provides a very short definition stating simply that sustainable product development addresses environmental issues through product design and innovation. The environment plays a central role in most of the definitions about sustainable products. Lin

(6)

and Chang (2012, p. 125) state in their research about the role of environmental

consciousness in green product usage that “compared with regular products, sustainable products tend to use biodegradable, nontoxic ingredients and are packaged in bottles that can be recycled”. In their definition they are not only focusing on the content of the product, but on the packaging as well. This implies that the whole product is sustainable. It is important to note this because if only one component of a product is sustainable, can the whole product than be referred to as a sustainable product? Beside the product itself, it is also important that the manufacturing process is sustainable.

In this study the research setting is the fashion industry. Therefore previous literature about sustainable fashion is reviewed. Before a clear definition of sustainable fashion can be given, it is important to point out that there are different nomenclatures to describe the concept: green fashion, ethical fashion and sustainable fashion are used interchangeably (Dong, Richards, & Feng, 2013). Casadesus-Masanell et al. (2009, p. 204) define green goods in the fashion industry as “goods that are supposed to have lower adverse environmental impacts either in production or in use”. This definition succeeds to mention the environment as an important stakeholder when producing fashion but fails to mention the fair-trade and ethical components that include the people who make the clothes. Dong, Richards, and Feng (2013, p. 35) provide in addition a much more elaborate definition of sustainable fashion. Stating that sustainable fashion is “clothing that incorporate fair trade principles with sweatshop-free labour conditions; that does not harm the environment or workers by using biodegradable and organic cotton; that is designed for a longer lifetime use; that is produced in an ethical production system, perhaps even locally; and which causes little or no environmental impact and makes use of eco-labelled or recycled materials”. This definition mentions both the environment and

(7)

the people who make the clothes as important stakeholders. Therefore this definition of sustainable fashion will be used in this paper.

2.2 Leasing

Previous research has defined leasing as a contract by which the owner of the asset (the lessor) grants the right to use the asset for a given term to another party (the lessee) in return for a periodic payment of rent (Anderson & Lazer, 1978). This definition of leasing provided by Anderson and Lazer (1978) will be used in this paper. However, it is

important to note that from an economic perspective, the critical distinction concerned with leasing is between “short-term leases” and “long-term leases” (Miller & Upton, 1976). A short-term lease is a commitment to acquire and pay for the services of a

specific capital good for the shortest practicable interval of time. A long-term lease is one in which the user has a commitment extending over more than a single period. In this paper, the term leasing will be used from a “short-term lease” perspective because in some literature, long-term leasing (for more than twenty years) is considered the same as buying or selling (Shih & Chou, 2011).

Lease contracts are not solely for the business world anymore but are now becoming more common in the consumer market as well (Dasgupta, Siddarth, & Silva-Risso, 2007). Durable products that used to be purchased outright, such as cars,

computers, furniture and other household appliances are now leased over extended periods. Because this paper is about leasing sustainable products, it is important to review the literature about what is known about this certain topic. Leasing and sustainability were long considered to be going hand in hand. Because leasing is considered as being the more sustainable option as opposed to buying or selling (Fishbein et al. 2000; Mont 2002 in Agrawal et al. 2012). This notion was build upon the argument that because firms are responsible for their lease-products post-use, the incentive for firms to design their

(8)

products more durable became higher. If there are more durable products than the demand for new products would go down and this reduces the environmental impact of

manufacturing and disposal. As a consequence, environmental groups recommend firms to offer leasing and take-back options.

However, literature has appeared that contradicts the sustainable lease perspective. With their research, Agrawal et al. (2012) provide guidance on when firms can justifiably promote leasing as the greener strategy. They show that the two main drivers of whether leasing can be both a more profitable and a greener strategy are the product durability and whether it has a high use impact. The latter depends on the product type. The conclusions they draw from their research are that leasing is greener for products that have a higher use impact (compared to their production and disposal impacts) and low durability (e.g. printers). In contrast, leasing is environmentally worse for products that have a higher production and disposal impact (e.g. laptops). What can be concluded from the research done by Agrawal, et al. (2012) is that whether leasing is the more sustainable option is different for each product type and it has to be investigated if it is more sustainable to lease it or to sell it.

2.3 Willingness to pay

Consumers‟ willingness to pay (WTP) a certain price for a product can be valuable

information to firms. It provides guidance for pricing a product in the market (Wertenbroch & Skiera, 2002). Especially when introducing new products, this information can be very

valuable. Wertenbroch and Skiera (2002, p. 228) define WTP, as “the denotation of the maximum price a buyer is willing to pay for a given quantity of a good”. It is a ratio-scaled measure of the subjective value the buyer assigns to that quantity. He or she buys that item from a set of alternatives, for which his or her WTP exceeds purchase price the most”. They further state that knowing a consumers‟ WTP is important in estimating demand and

(9)

designing pricing schedules. However, Wertenbroch and Skiera (2002) argue whether existing market research techniques (e.g. surveys) provide an incentive to consumers‟ to reveal their true WTP and in whether they simulate actual point-of-purchase contexts. Auger and Devinney (2007) also note that simple survey questions are too “noisy” to provide

operationally meaningful information and overestimate intentions to a considerable extent. As an alternative to survey questions, Wertenbroch and Skiera (2002) discuss the use of

experimental Vickrey (1961) auctions. A Vickrey auction ensures incentive compatibility because the auction bid is sealed off, meaning that the purchase price is determined solely by the other participants‟ bids. Also, the participant that wins the auction must purchase the good in a real transaction. This provides an incentive for bidders in a Vickrey auctions to reveal their WTP truthfully.

2.4 The effects of sustainability on willingness to pay

Elaborate research has been done on the effects of sustainability on the WTP for products and services (Chen, 2010; De Pelsmacker et al. 2005; Hetterich et al. 2012; Jansen et al. 2011; and Lin & Chang, 2012). In most of the research it is hypothesized that

sustainability has a positive effect on the WTP, resulting in higher prices for sustainable products. This effect is proven by the research of Hetterich et al. (2012) who investigate the willingness to pay a premium for car interiors made of renewable resources. Results show that more than 66% of the respondents would accept green car components for a moderate price increase. However, what exactly constitutes a moderate price increase? This can be a 10% or 30% price premium; it is not made clear in the research. Also, in the research done by Hetterich et al. (2012) respondents‟ only say they would pay a price premium for a more sustainable product. It is not certain if they would actually do it. Lin and Chang (2012) provide more specific results in their research stating that consumers are willing to pay a 20-25% price premium for sustainable household products. All these

(10)

results seem very promising for the market of sustainable products. However, De Pelsmacker et al. (2005) state the contradictory fact that 50% of European consumers claim that they are willing to pay a higher price for sustainable products, but the final market share is less than 1%. This is called the attitude-behavior gap; consumers‟ positive attitudes are not always translated in the final action (Boulstridge & Carrigan, 2000).

The above-mentioned literature provides some knowledge about the WTP for sustainable products from all sorts of categories. In addition to this, literature has also elaborated on the WTP for sustainable fashion (Casadesus-Masanell et al., 2009; Ellis et al., 2012; Hustveld & Bernard, 2008; Lee, 2011). Previous research indicates that

consumers are also willing to pay a premium for sustainable fashion. Casadesus-Masanell et al. (2009) found that consumers were willing to pay significant premiums for organic cotton garments of the Patagonia brand. Even though the organic cotton provided no demonstrable private incremental benefits for the consumer. Ellis et al. (2012) found using an experimental auction that consumers were willing to pay a 25% premium for an organic cotton t-shirt over the visibly similar t-shirt made from conventionally produced cotton. Previous research provides enough support for the notion that consumers are willing to pay a premium for sustainable products and that sustainable fashion is no exception.

Based on the above, sustainability can be expected to have a positive effect on WTP. However this effect may be different depending on the demographic characteristics of the consumer. Previous studies have investigated the relationship between demographic characteristics and the WTP for sustainable products (Anderson & Cunningham, 1972; De Pelsmacker et al., 2005; Dickson, 2001; MORI, 2000; Roberts, 1996). For example, Anderson and Cunningham (1972) found that younger consumers were more willing to buy sustainable or ethical products. However, the effect of the level of education of the

(11)

„typical sustainable consumer‟ was not clear. Roberts (1996) did found that the typical ethical or sustainable consumer was a person with a relatively high income, education and social status. In contrast to this result, Dickson (2001) found no differences in age and income between ethical consumers that think sustainability is important and consumers who do not. In the same study it was found that women more often than men buy apparel with a sustainable label. Nevertheless, most studies concluded that gender does not influence the buying behaviour or WTP for sustainable fashion (e.g. MORI, 2000). The above leads to the following hypotheses:

H1a: Consumers are willing to pay more for a sustainable product than for a conventionally produced product.

H1b: Younger consumers are willing to pay more for a sustainable product than older consumers.

H1c: Consumers with a higher education are willing to pay more for a sustainable product than consumers with a lower education.

H1d: Women are willing to pay more for a sustainable product than men.

2.5 The effects of leasing on willingness to pay

The main question concerned with leasing in previous research often had to do with whether it is better for a firm to lease or sell its assets (Dasgupta, Siddarth, & Silva-Risso, 2007; Miller & Upton, 1976). This question was mainly set in an economical context: “Can a firm gain more profit by selling or leasing its assets? According to Miller and Upton (1976) the decision whether to lease or to buy is neither a matter of indifference for the typical firm or one for which any general presumption can be established a priori. Each case must be examined on its merits.

In other words, the answers to the question if it is economically better to lease or sell a product depends on the product characteristics and the situation surrounding the

(12)

product. To illustrate this, Dasgrupta, Siddarth, and Silva-Risso (2007) show how financial contracts require consumers to make trade-offs among financial elements, such as interest rates, monthly payments and contract length, and non-price components, such as maintenance costs and operating costs. They found that consumers often prefer leases with lower up-front costs and monthly payments, even if the net prices of these contracts are higher than short-term contracts or purchasing the product. In order to give more insight into this phenomenon, it is argued in this paper that consumers are willing to pay more for a lease-product compared to a product that needs to be purchased. Based on the above, the following is hypothesized:

H2: Consumers are willing to pay more for a lease-product than for a product that needs to be purchased.

2.6 The effects of leasing and sustainability on willingness to pay

Previous research has shown that sustainability has an increasing effect on the WTP (Casadesus-Masanell et al., 2009; Ellis et al., 2012; Hustveld & Bernard, 2008; Lee, 2011). The same effect was found for leasing on WTP (Shih & Chou, 2011). Shih and Chou (2011) look at the effect of uncertainty in the lease-or-buying decision process. They conducted a survey of some uncertain factors that may affect the adoption of an emerging type of technology (e.g. solar power systems). The results showed that consumers with more concerns about uncertainty are willing to pay more for a shorter lease time as opposed to purchasing (leasing for over twenty years). What can be drawn from this conclusion is that it is presumably more appealing for consumers to lease an innovative product than to purchase one at once, because this reduces the risk and uncertainty.

In this paper, the WTP for sustainable fashion is investigated. Fashion cannot be considered as a new, innovative product. However, sustainable fashion as it exists today,

(13)

not the „sustainable fashion‟ of the 1960s hippie movement, is a relatively new product to the marketplace (Winge, 2008). Although previous research has shown that consumers are willing to pay a premium for sustainable products and 50% of the Europeans claim that they do this, it remains unclear why the final market share of sustainable products is less than 1%, the so-called attitude-behaviour gap (De Pelsmacker et al., 2005). In this paper, it is argued that sustainability combined with leasing creates a synergy effect on the WTP for products. Perhaps, if a sustainable product is offered to the consumer in the form of a lease contract, this will decrease uncertainty about the innovative product and result in a higher WTP compared to a conventionally produced product, just a lease-product or just a sustainable product. Based on the above, the following hypothesis is formulated:

H3: Consumers have a higher WTP in the treatment with a sustainable lease-product compared to all the other treatments.

2.7 The effect of perceived quality and fashionableness

In their study, Lin and Chang (2012) not only found that consumers are willing to pay a price premium for sustainable household products but also tend to use more of it. This effect occurs because consumers perceive sustainable products as being less effective than conventional products and therefore try to compensate by using more of it. As a

consequence, resources are unnecessarily wasted, which has a negative impact on the environment and surpasses the point of buying sustainable products in the first place. The sustainable products investigated in the research by Lin and Chang (2012) are limited to those in product categories in which strength is a major determinant of product choice and in which there is a possibility of overuse without the consumer realizing it. Because in this research the empirical setting is the fashion (jeans) industry, strength can be considered as a determinant for product choice as well. In other words, the sustainable jeans should not fall apart very easily. Therefore the present research suggests that

(14)

consumers hold similar stereotypical views of sustainable and conventional products when it comes to the quality and strength of the product. According to Luchs et al. (2010) consumers infer a positive association between ethicality and gentleness and a negative association between ethicality and strength. This result shows that sustainable and conventional products are perceived differently in terms of product effectiveness, with sustainable products being associated with lower levels of quality. Moreover, the

perceived quality of a sustainable product has a changing effect on the buying behaviour of consumers (Lin & Chang, 2012). The above leads to the following hypotheses:

H4a: Sustainable products are perceived as being of lesser quality than conventional products.

H4b: The effect of sustainability on the willingness to pay for fashion is positively moderated by perceived quality.

Besides the perceived quality there is another dilemma that sustainable fashion brands face today (Beard, 2008). „Ecofashion‟ brands face a challenge as they seek to position and brand themselves not only as ethically worthy, but also increasingly as "Fashionable". Not just to a niche audience, but to everyone. In the past, sustainable or eco-conscious fashion was associated with the Hippie subculture of the 1960s (Winge, 2008). At that time the sustainable fashion design mostly consisted out of the stereotypical images of eco-dress like rope sandals, tie dye T-shirt, and hemp cargo pants. Wearing sustainable fashion was more used to make a statement: “It represented a form of deliberate and cultivated anti-fashion, and empowered the wearer through visual support of

sociopolitical ideals and values associated with the protection of environmental, human, and animal rights.” (Winge, 2008 p. 514). Aside this, sustainable fashion was also associated with being drab, shapeless and as mentioned before: „anti-fashion‟.

(15)

only adopted by a few. However, according to Winge (2008) this is about to change. Because celebrities are wearing sustainable fashion and magazines like Vanity Fair start claiming that “green is the new black”, it set in motion the idea of sustainable being the latest fashion trend. Therefore, in this paper it is investigated whether sustainable fashion has lost its negative stereotype as not being fashionable and if this affects the buying behavior of sustainable fashion. This discussion leads to the following hypotheses: H4c: Sustainable products are perceived as being more fashionable than conventional products.

H4d: The effect of sustainability on the willingness to pay is positively moderated by perceived fashionableness.

3.0 Methodology

3.1 Method

In this research an experiment was conducted to measure the WTP for sustainable and leasable fashion. An experiment method was chosen to be most suitable for two reasons. First, testing the hypotheses involves the manipulation of an independent variable (e.g. sustainability or leasing) and measuring its effects on a dependent variable: the WTP. According to Saunders et al. (2007) experiments allow for examining such a manipulation in a controlled context. Therefore, any effect on the dependent variable can be attributed to the stimulus in the experiment. Second, an experiment can test for effects and causal explanations in a controlled context, which means that its internal validity is usually high. However, because of the controlled nature of an experiment, the external validity, or the extent to which findings can be generalized across the entire population, is more

(16)

To elicit a consumer‟s WTP for a product various methods exist. A perfect method does not (yet) exist; the appropriateness of the method depends on the research approach, resources available and on the type of good (Ellis et al., 2012). However, experimental auctions are emerging as alternatives for eliciting WTP to more traditional data collection methods such as surveys (Auger & Devinney, 2005; Lusk & Hudson, 2004 in Ellis et al., 2012). Surveys allow people to hypothetically indicate their WTP, yet previous research shows that people tend to overestimate their WTP when no real money is involved in transactions (List & Gallet, 2001 in Ellis et al., 2012). In an experimental auction however, real exchange of money is applied to provide a more realistic condition of markets. List & Gallet (2001) found that using real money in an experimental auction instead of hypothetically indicating the WTP in a survey reduces bias. This finding is also supported by the research done by Auger and Devinney (2005). They investigated the WTP for ethical consumerism by comparing traditional survey instruments and results obtained from experimental methods (e.g. auctions). According to their results there is a lack of relationship between the results of the traditional survey and the experimental method.

Based on the abovementioned literature a field experiment methodology was chosen to determine consumers‟ WTP for sustainable and leasable fashion. The

experiment was conducted by using the online auction platform called „Veylinx‟ (Kuijken & El-Haji, n. d.). This platform has been set up in cooperation with the University of Amsterdam and has around 3000 panel members. During the experiment a sustainable pair of jeans was auctioned online via a Vickrey auction or second-price sealed-bid auction (Vickrey, 1961; Wertenbroch & Skiera, 2002). As discussed earlier, a Vickrey auction takes place in a sealed environment. Participants place a hidden bid and the highest bid wins. However, the winner of the auction pays a price that is equal to the

(17)

second highest bid. This prevents participants from bidding strategically so there is no incentive to misrepresent their true WTP.

3.2 Research setting

The empirical setting of this study is a Dutch jeans brand called Mud jeans

(www.mudjeans.eu). This brand is specialized in sustainable fashion and inventor of the „lease-a-jeans‟ concept. The lease concept allows customers to lease jeans for a year and pay per month. After a year, the customer can decide whether to keep the jeans, send them back for recycling or trade them for a new pair of jeans and extend the lease contract. The retail price for the lease contract is a one-time payment of 20 euro followed by a monthly payment of 5,95 euro for a period of twelve months. Mud jeans can also be purchased without a lease contract. The retail price for purchasing is 109,95 euro. The International Working Group on Global Organic Textile Standard (GOTS, 2014) certifies mud jeans as being a sustainable fashion brand. Jeans were the chosen product because they

traditionally have a large environmental impact (Kuik, 2005). During the production of jeans a lot of water and chemicals is used. Beside this, jeans are often produced in countries with low labour costs (e.g. China and Bangladesh). As a result the retail prices for jeans vary a lot. All these facts about jeans make it an interesting product to determine the WTP.

3.3 Data collection & research design 3.3.1 Data collection

This study is a quantitative research. Data was collected through the online auction platform Veylinx. At 10 am on the day of the experiment, every panel member received e-mail with an invitation to participate in the auction. Panel members participated in the auction by clicking on the link provided in the e-mail. Before the actual auction the

(18)

participants were informed about the rules of Veylinx. The following steps describe the data collection procedure:

Step 1: A detailed explanation was given of how the auction works.

Step 2: The participants were informed that they have the opportunity to bid accordingly to what they think the product is worth. Each bid was binding.

Step 3: The participants were shown an advertisement of the pair of jeans and were given six minutes to fill in their bid. The other bidders couldn‟t see their bid.

Step 4: After they placed their bid the participants were given a short survey. The survey questions were about brand recognition (Do you know the brand shown in the

advertisement? Yes/no), statements about the perceived quality of the jeans and if they perceive sustainable jeans as fashionable (5 point Likert scale: “totally agree” vs. “totally don’t agree”). Participants were also asked how long they usually wear a pair of jeans before buying a new one (0-6, 6-12, 12-18 months etc.). Consult appendix 2.0 for all of the survey questions. All panel members had till 10 pm on the day of the experiment to fill in their bids. When the auction closed all bids were gathered and ranked. The highest bidder was notified via e-mail and paid an amount equal to the second highest bid.

3.3.2 Research design

The field experiment had a 2 x 2 design containing four different treatments (table 1). The participants were randomly divided among the four treatments. In each treatment,

everything was kept consistent except for the independent variables sustainability and leasing. The first treatment (T1) was the baseline treatment. Only a picture of the pair of jeans was shown without any logos or statements about sustainability or leasing (appendix 1.1 & 1.5). In the second treatment (T2) a picture of the jeans was shown with sustainable logos and a statement saying it is a sustainable pair of jeans (appendix 1.2 & 1.6). In the third treatment (T3) the jeans were sold with a lease contract and participants placed a bid

(19)

per month instead of purchasing the jeans at once (appendix 1.3 & 1.7). In the last treatment (T4) the picture of the jeans was shown accompanied by sustainable logos, a statement and a lease contract allowing the participants to bid per month (appendix 1.4 & 1.8).

Table 1: Research design: 2 x 2 factorial design

Leasing

No Yes

Sustainable logo No T1 T3

Yes T2 T4

3.4 Sample

Table 2 presents some interesting descriptive statistics of the sample. The total number of the sample (N) was 811. However, two participants were excluded because they reported an age of 103 and 114, which is very unlikely to be real and therefore considered as non-useful. The final sample size is N = 809. It contains 357 females (44%) and 452 men (56%). Compared to the Dutch population (see appendix 3), men in the sample are slightly overrepresented and older whereas women are slightly underrepresented and younger (CBS, 2013). The age differed from 16 to 93 with an average of 41.55 (M = 41.55, SD = 16.01). The sample is skewed because the age group 20-40 years (24.6%) is in the sample almost twice the size compared to the total Dutch population (47.5%). Most participants in the sample were highly educated (36.8%). This result is also skewed because in the total Dutch population most people are educated at a vocational (MBO) level (40.1%). This difference between the sample and the Dutch population can be explained by the fact that most participants were recruited within the students‟ direct networks resulting in mostly other students and family.

The table also shows that the average bid was € 12.87 (SD = 18.96) with a maximum of € 119.40. Finally, it‟s interesting to look at brand recognition, 10,1% stated

(20)

that they knew the brand shown in the advertisement (Mud jeans). Because this is a fairly low percentage, it was not considered to be a problem during further analyses.

Table 2: descriptive statistics total sample

N Minimum Maximum M SD Gender 809 0 1 .44 .50 Age 809 16 93 41.55 16.01 Education 550 0 5 3.78 1.24 WTP 809 0 119.40 12.87 18.96 Brand recognition 759 0 1 .10 .30

3.5 Variables and measurement 3.5.1 Dependent variables

The dependent variable was the maximum WTP for a pair of jeans at the point of

purchase. According to Wertenbroch and Skiera (2002) the WTP is a ratio-scaled measure of the subjective value the participant will assign to a specific quantity. The participants filled in the maximum amount of euro they were willing to spend on the pair of jeans shown in the advertisement. The euro was the chosen currency because all the auction participants were Dutch and Mud jeans is a Dutch company located in the euro region.

Beside the variable WTP were two more dependent variables: „perceived quality‟ and „perceived fashionableness‟. These variables were used to measure whether

sustainability has lost the stereotype of poor quality and not being fashionable (Luchs et al., 2010; Winge, 2008). Lin and Chang (2012) investigated the product effectiveness of sustainable products by measuring the attitude towards the product. They conducted a survey containing questions about the quality of the product on a 5-point Likert scale (“bad quality” vs. “good quality”). Based on the questions by Lin and Chang (2012), two statements were made to measure the participants‟ perceived quality of the jeans and if they perceive the jeans as being fashionable (appendix 2.0)

(21)

3.5.2 Independent variables

Sustainability and leasing were the two independent variables used in this study. First, sustainability communicated to the participant that the pair of jeans in the advertisement was manufactured in a sustainable manner (appendix 1.2). Displaying a „Max Havelaar fair-trade‟ logo and a „Soil and More – Climate Neutral‟ logo in the advertisement of the jeans operationalized the sustainable variable. The Max Havelaar fair-trade organization is a Dutch non-profit organization that focuses on improving the working conditions and supply chain of small companies in third world countries. Soil and More focuses on preventing the loss of soil fertility by reducing the emission of carbon dioxide. The following statement accompanied the logos: “Mud jeans produces sustainable fashion with respect for people and the environment”. The advertisement also mentioned that the pair of jeans contains 98% biological cotton and 2% elasthan.

Second, leasing communicated to the participant that the pair of jeans was to be sold with a lease contract (appendix 1.3). Instead of bidding to purchase the jeans at once the participant was told to bid per month. The following sentence operationalized the leasing treatments: “You are leasing the jeans for one year after which you have to give it back. Your bid is your monthly payment which you have to pay for 12 months.”

3.5.3 Moderating variables

Other than a dependent variable, perceived quality was also used as a moderating variable because it is thought to moderate the relationship between sustainability and WTP (H4b) (Lin & Chang, 2012). The variable perceived fashionableness was also used as a

moderator variable for H4d. Consumers that perceived sustainable fashion as being more fashionable than conventionally produced fashion were expected to have a higher WTP. Perceived quality and perceived fashionableness were operationalized in the same manner as described under dependent variables.

(22)

3.5.4 Control variables

Demographic characteristics were expected to have an effect on the relationship between sustainability and WTP (Anderson & Cunningham, 1972; De Pelsmacker et al. 2005; Dickson, 2001; MORI, 2000; Roberts, 1996). Therefore, the first hypothesis, stating that consumers have a higher WTP for sustainable products compared to conventional products was controlled for the variables: age, gender and education. Participants that signed up for the online auction platform Veylinx were asked to state their age, gender and education. Participants with an age of 30 or more were operationalized as being „older consumers‟. Participants with an education at HBO (University of applied sciences) or University level operationalized the highly educated variable.

3.6 Data analysis

This section provides an overview of the tests that were used to analyze the data. First, descriptive statistics will be analyzed focusing on the assumptions that are needed before testing the hypothesis (Field, 2009). These assumptions include testing for normality and homogeneity of variance. Finally two parametric tests: ANOVA and regression analysis will be performed to test the hypotheses. In order to test the hypotheses several dummy variables were created. The first dummy variable: „sustainable‟, only contained treatments 2 and 4, because these treatments contained the advertisement with the sustainable logo‟s and statements. The second dummy variable: „leasing‟, contained treatments 3 and 4, because in these treatments, the jeans were sold via a lease contract.

Finally, two other dummy variables were created for age and education. The dummy variable for age had a value of „0‟ for age under 30 and a value of „1‟ for ages of 30 or higher. Education had two different values for levels under HBO (University of applied sciences) and levels for HBO or higher. Another precaution that was made before testing the hypotheses was multiplying the bid amounts in the lease treatments (T3 and T4) by a

(23)

factor of 12. This was done in order to compare the lease amounts to the purchase amounts during the hypothesis testing.

For the dependent variable WTP no missing values were found because

participants were not able to participate in the auction without entering a bid amount. The missing values of the independent variables were below ten percent; therefore „hotdeck‟ imputation (Myers, 2011) could be possible. However, it was not possible to find proper deck variables for imputation because all the other variables had more missing data than the independent variables. Therefore, the missing values were handled by using „listwise deletion‟ (Field, 2009). SPSS 21 was used for all statistical analyses.

4.0 Results

4.1 Descriptive statistics

First, the assumption of normality will be checked visually by looking at the histograms that show the distribution of bids (Field, 2009). Then, applying a Kolomogorov-Smirnov test will quantify normality.Although the sample size of each treatment in this study is larger than N = 30 and therefore can be approximated by a normal distribution, it is always useful to test for normality (Field, 2009). Finally, the assumption of homogeneity of variance will be explored.

4.1.1 Willingness to pay.

Table 3a shows the mean WTP for treatments 1, 2, 3, and 4 in euro. For treatment 1, M = 11.86 (SD = 15.89), for treatment 2 M = 12.75 (SD = 16.74), for treatment 3, M = 12.49 (SD = 19.53), and for treatment 4, M = 12.87 (SD = 23.54). The table also shows that the highest bid was placed in treatment 4, containing the sustainable and leasing manipulation (maximum = 119.40). The lowest bid was placed in treatment 2, containing only the sustainable manipulation (maximum = 80.00).

(24)

In order to check visually whether the variable WTP is normally distributed the histogram showing the frequency distribution of bids was consulted (Field, 2009). The distribution of the histogram including bids of zero didn‟t show a nice bell shaped curve. This means that the samples are highly positively skewed. This was also the case when the bids of zero were excluded and for the bids in the top 50 percent. The results of the Kolmogorov-Smirnov test for normality shown in appendix 4.1 confirm that the treatments seem to violate the normality assumption. The test was performed for both including and excluding zero bid values. The table shows that the WTP in all treatments was significantly non-normally distributed (all ps <.001). A Shapiro-Wilk test similarly confirmed this finding (all ps <.001).

Next, Levene‟s F test will explore the assumption of homogeneity of variance (Field, 2009). Table 3b shows that the differences in variance between each treatment were significant (F (3, 805) = 9.70, p < .001), indicating that equal variances cannot be assumed. However, SPSS offers an alternative version of the F-ratio: Welch‟s F (Field, 2009; Welch, 1951). Welch‟s F test shows the differences in variance were not significant (F (3, 434.25) = .61, p = .609), indicating that equal variances can be assumed.

In summary, the assumptions of normality are violated both visually and

quantified by the Kolmogorov-Smirnov test. Additionally, Levene‟s and Welch‟s F test show that equal variances cannot be assumed, except for Welch‟s F test including zero bids. As a consequence performing parametric tests like a one-way ANOVA and

regression analysis can lead to inaccurate results. According to Field (2009), the data must be corrected for these errors before testing the hypotheses.

Table 3a: descriptive statistics willingness to pay (N = 809)

Treatments N M SD Std. Error Min. Max.

1. Baseline 208 11.86 15.89 1.10 0 86.35

2. Sustainable 212 12.75 16.74 1.15 0 80.00

3. Leasing 210 12.49 19.53 1.35 0 84.00

(25)

Table 3b: homogeneity of variances willingness to pay Variable Levene statistic Df1 Df2 Sig. Welch statistic Df1 Df2 Sig. Willingness to pay 9.701 3 805 .000 .610 3 434.25 .609

4.1.2 Perceived quality and perceived fashionableness

In this section the descriptive statistics of the variable perceived quality and perceived fashionableness are discussed. Table 4a and b show the mean of perceived quality and perceived fashionableness in treatment 1, 2, 3 and 4.

First, normality was checked visually by looking at the histograms. Similar to the willingness to pay data the distribution of both perceived quality and perceived

fashionableness was not nicely bell shaped indicating a non-normal distribution. This finding was supported by the results of the Kolmogorov-Smirnov test, which can be seen by the significant results in appendix 4.2 (all ps < .001).

Secondly, homogeneity of variance was tested. Table 4c shows the results of Levene‟s homogeneity of variance test. Both perceived quality and fashionableness were non significant, respectively F (3, 737) = .25, p = .862 and F (3, 742) = .06, p = .980. This means that equal variances can be assumed.

Table 4a: Descriptive statistics of perceived quality (N=741) in different treatments

Treatments N Mean SD Std. Error

1. Baseline 192 2.76 .834 .060

2. Sustainable 197 2.78 .775 .055

3. Leasing 190 2.80 .818 .059

4. Sustainable & leasing 162 2.76 .787 .062

Table 4b: Descriptive statistics of perceived fashionableness (N=746) in different treatments

Treatments N Mean SD Std. Error

1. Baseline 193 2.57 .950 .068

2. Sustainable 198 2.63 .946 .067

3. Leasing 191 2.74 .980 .071

(26)

Table 4c: Homogeneity of variance perceived quality and fashionableness Levene’s statistic Df1 Df2 Sig. Perceived quality .249 3 737 .862 Perceived fashionableness .061 3 742 .980

In summary, the data of the WTP, perceived quality and fashionableness were non-normally distributed. Therefore, to correct for this error a „log transformation‟ was

performed in SPSS (Field, 2009). Although the histogram of the transformed data showed a more bell shaped curve it was still not normally distributed. The Kolmogorov-Smirnov test confirmed this (all ps < .001). However, as noted before, when a sample exceeds N = 30 it can be approached as being normally distributed (Field, 2009). Therefore, the untransformed dataset was used in further analysis. Also, the dataset including the bids of zero was used in further analysis because when a participant placed a bid of zero it presumably said more than that they „just did not want the jeans‟. This assumption is based on the fact that the percentage of participants that placed a bid of zero in the lease treatments was 56%. In contrast, the percentage of participants that placed a bid of zero in the purchase treatments was much lower, 34,5%. Perhaps the participants in the lease treatments liked the jeans but did not want to be tied into a contract and therefore placed a bid of zero.

4.2 Hypothesis testing

4.2.1 Willingness to pay for sustainable products

The first hypothesis (H1a) stated that consumers are willing to pay more for a sustainable product compared to one that is produced conventionally. In order to test this hypothesis the dummy variable containing all the sustainable treatments was compared to the other treatments. Although tests of normality state that the data was non-normally distributed a one-way ANOVA analysis was performed.

(27)

Table 6a shows that, although the average WTP was higher for the group in the sustainable treatment (M = 13.62; SD = 20.14) than for the group that was in the baseline treatment (M = 12.17; SD = 17.79) there was no significant effect of sustainability on the WTP, F (1, 807) = 1.17, p = .28). The one-way ANOVA was also performed for only the top 50% of the participants. Although there still wasn‟t a significant effect of

sustainability on WTP, F (1, 396) = 2.45, p = .12, the value of p did become a lot lower and approached the borderline of significance. Also, the average WTP of participants in the sustainable treatment was again higher (M = 27.24; SD = 21.31) than the WTP of participants in the baseline treatments (M = 24.07; SD = 19.10). It is therefore safe to state that there is a trend of an increasing WTP in the sustainable treatments. This finding is confirmed by the fact that there was a significant effect of sustainability on WTP when the bids of zero were excluded, F (1, 444) = 4.89, p < .05. However, as discussed earlier, only the WTP including bids of zero were used. Therefore, H1a was rejected.

Table 6a: One-way ANOVA between conventional and sustainable treatments.

Variables Treatments N Mean SD F Sig.

Bid amount Conventional 418 12.17 17.79 1.172 .279 Sustainable 391 13.62 20.14

Bid amount in top 50% Conventional 205 24.07 19.10 2.45 .118 Sustainable 193 27.24 21.31

The effect of sustainability on WTP was controlled for the demographic characteristics age, education and gender (H1b, H1c, and H1d). There was a significant effect of age on the WTP for sustainable products, F (1, 389) = 7.044, p < .01. Younger consumers (age < 30) had on average a higher WTP (M = 17.25, SD = 20.65) than older consumers (M = 11.64, SD = 19.61). This means H1b was accepted.

H1c stated that consumers with a higher education have a higher WTP for

sustainable fashion than consumers with a lower education. A one-way ANOVA analysis shows that there was a significant effect of education on the WTP for sustainable

(28)

products, F (1, 264) = 6.676, p < .05. Consumers with a higher education (M = 16.38, SD = 21.06) had on average a higher WTP than consumers with a lower education (M = 9.64, SD = 18.77). This means H1c was accepted.

Finally, H1d stated that female consumers have a higher WTP than male consumers. However, there was no significant effect of gender on WTP, F (1, 389) = .112, p = .738. This means H1d was rejected.

4.2.2 Willingness to pay for leasable products

In order to test H2, consumers have a higher WTP for lease products compared to products that need to be purchased, the dummy variable containing all the lease treatments was compared to the other treatments using a one-way ANOVA. Table 7a shows there was no significant effect of leasing on WTP, F (1, 807) = .782, p = .377. A similar result was found for the bid amounts in the top 50%, F (1, 396) = 3.109, p = .079. Again, the value of p became lower; this indicates a trend toward significance. The WTP was higher in the group with the lease product, M = 27.46 (SD = 23.56) than in the group with the conventional product, M = 23.89 (SD = 16.47). This means that the second hypothesis was partially accepted.

Table 7a: One-way ANOVA between conventional and leasing treatments.

Bid amount Conventional 420 12.31 16.31 .782 .377

Lease 389 13.49 21.47

Bid amount in top 50% Conventional 207 23.89 16.47 3.109 .079

Lease 191 27.46 23.56

4.2.3 Willingness to pay for sustainable, leasable products

The third hypothesis states that the participants have a higher WTP for the product in the „sustainable-lease‟ treatment compared to the other treatments. In order to test this hypothesis the treatment where the sustainable jeans is offered with a lease contract was compared to all the other treatments. There was no significant effect of the different

(29)

treatments on WTP, F (3, 805) = .756, p = .519. However, when the one-way ANOVA analysis was performed for only the top 50 percent of the WTP the effect did become more significant, F (3, 394) = 2.024, p = .110. This result indicates that there is a mild trend towards significance. Because of this low p value Fisher‟s least significant

difference (LSD) post-hoc test was consulted to determine which treatment affected WTP the most (Field, 2009). Table 8b shows that the sustainable and leasing treatment (M = 29.81, SD = 26.02) had a significant effect on the baseline treatment (M = 22.66, SD = 16.76) p < .05. Participants in the sustainable and leasing treatment placed an amount that was € 7.15 higher than in the baseline treatment. There was no significant effect between the sustainable and leasing treatment and the other treatments. This means H3 was rejected. The WTP of the sustainable-lease product was only higher compared to the product in the baseline treatment.

Table 8a: One-way ANOVA between all treatments

Willingness to pay Baseline 208 11.86 15.89 .756 .519 Sustainable 212 12.75 16.74 Leasing 210 12.49 19.53 Sustainable & leasing 179 14.65 23.54 Willingness to pay in top 50% Baseline 102 22.66 16.76 2.024 .110 Sustainable 105 25.09 16.18 Leasing 103 25.46 21.16 Sustainable & leasing 88 29.81 26.02

Table 8b: LSD Post-Hoc test for WTP in top 50%

(I) Treatment (J) Treatment Mean

Difference

Std. Error Sig.

Sustainable & leasing Baseline 7.15 2.93 .015

Sustainable 4.71 2.91 .107

Leasing 4.34 2.93 .139

4.2.4 perceived quality and fashionableness

Hypotheses 4a en 4c are about the perceived quality and fashionableness of sustainable products. H4a states that sustainable products are perceived as being of lesser quality than

(30)

conventional products. In order to test this hypothesis the dummy variables containing all the sustainable treatments was compared to the other treatments using a one-way

ANOVA analysis. There was no significant effect of sustainability on the perceived quality of the product, F (1, 739) = .02, p = .89. The participants in the sustainable

treatments (M = 2.77, SD = .78) perceived on average the quality of the jeans as being the same as the participants in the other treatments (M = 2.78, SD = .83). Therefore H4a was rejected.

Testing H4c lead to a similar result. This hypothesis stated that sustainable products are perceived as being more fashionable than conventional products. There was no significant difference between the participants in the sustainable treatment (M = 2.66, SD = .94) and the participants in the other treatments (M = 2.66, SD = .97), F (1, 744) = .00, p = .99. In fact, the participants perceived the jeans in the sustainable treatment as being almost as fashionable as the jeans in the other, baseline treatments. This means H4c was rejected.

Hypothesis 4b states that perceived quality positively moderates the relationship between sustainability and the WTP. In order to test this hypothesis the PROCESS add-on for regression analysis in SPSS was used (Hayes, 2012). There was no significant

moderating interaction effect of perceived quality, F (1, 737) = 2.836, p = .093. Hypotheses 4d states that there is a moderating effect by perceived

fashionableness on the relationship between sustainability and the WTP for sustainable fashion. Again, there was no significant moderating interaction effect by perceived fashionableness, F (1, 742) = 1.275, p = .259. Both H4b and H4d were rejected.

4.3 Robustness checks

In addition to the aforementioned analyses several robustness checks were performed. This seemed necessary because all of the analyses were done using the variable that

(31)

contained the bid amounts of the lease treatments that were multiplied by 12 (months) to resemble one year leasing a pair of jeans or „buying it‟. On the other hand, people can choose to wear jeans longer than a period of one year. As a consequence, all the results of the aforementioned analyses would be impure and don‟t predict WTP for sustainable, leasable fashion correctly. Based on this reasoning the participants were asked to state how many months they wear a pair of jeans before they discard them (appendix 2.0). It seems participants wear a pair of jeans between 18 and 24 months on average (M = 4.69, SD = 1.65). This is twice as long as expected and been accounted for when constructing the dependent variable „WTP‟ that was used in the analyses.

Based on the above, a new, more fine-grained dependent variable was constructed. This was done by dividing the bid amounts of the treatments in which the jeans were purchased by the number of months the corresponding bidder wears the jeans. This resulted in a new variable „robust WTP‟ that contained the WTP for jeans per month. The hypothesis testing was performed again by using „robust WTP‟ as a dependent variable. Concerning the first hypothesis, although the mean WTP per month was slightly higher for the group in the sustainable treatment (M = 1.10, SD = 1.68) compared to those in the baseline treatment (M = 1.00, SD = 1.55) there was no significant effect between

sustainability and WTP, F (1, 789) = .671, p = .413). By using only the top 50% of bidders the effect did become more significant but not enough, F (1, 382) = 1.367, p = .243. The effect of sustainability on „robust WTP‟ was controlled for age, gender and education. The results indicated that both age (p <.01) and education (p <.05) had again a significant effect on WTP for sustainable products. This result is consistent with the earlier findings.

The second hypothesis stated that consumers have a higher WTP for lease

(32)

effect of leasing on WTP, F (1, 789) = 1.632, p = .202. However, when the analysis was performed for only the top 50% of bidders the value of p did become much lower, F (1, 382) = 3.540, p = .061. This result was borderline significant. The group in the lease treatments had on average a higher WTP per month (M = 2.29, SD = 1.96) compared to the group in the baseline treatment (M = 1.95, SD = 1.54).

To test the third hypothesis, all the different treatments were compared with each other using the „robust WTP‟ variable. There was no significant effect of one of the treatments on WTP for the complete sample, F (3, 787) = .954, p = .414. For only the top 50% of bidders the value of p became again lower, F (3, 380) = 1.897, p = .130. The table with the LSD post-hoc analysis even showed a significant effect between the sustainable and leasing treatment and the baseline treatment (p < .05). The participants in the

sustainable and leasing treatment were willing to spend € 0,58 more on a pair of jeans than the participants in the baseline treatment. This result is similar to the analysis that was done with the untransformed WTP variable.

In order to test hypothesis 4b and 4d a regression analysis was performed with „robust WTP‟ as the dependent variable. There was no significant moderating effect by perceived quality on the relationship between sustainability and WTP, F (1, 737) = .958, p = .328. Perceived fashionableness did not have a significant effect either, F (1, 742) = .308, p = .579. The results of the regression were consistent with the earlier findings.

5.0 Discussion

5.1 General discussion

This study explored the effects of sustainability and leasing on the WTP for fashion by using an experimental auction method. Additionally, the perceived quality and

(33)

(Casadesus-Masanell et al, 2009; Ellis et al., 2012; Hustveld & Bernard, 2008; Lee, 2011) extensively researched the effects of sustainability on the WTP for fashion. However, to our knowledge, the concepts of leasing and sustainability have never been researched simultaneously using an experimental method. Based on the aforementioned literature, it was expected that sustainability and leasing have a positive effect on the WTP.

Furthermore, it was expected that perceived quality and perceived fashionableness have a moderating effect on the relationship between sustainability and the WTP. Although the ANOVA analyses indeed indicate a positive effect of sustainability and leasing on the WTP, a clear significant result has not been found. The only significant results that were found were those of age and education on the WTP for sustainable fashion. Some possible explanations for these results are discussed.

The first hypothesis suggested that consumers have a higher WTP for sustainable products compared to conventional products. The results indicated that the participants in the sustainable treatments had a higher WTP on average than those in the baseline

treatments. Although this difference was not significant, it showed a decreasing trend toward significance. The additional robustness check showed a similar result. Earlier studies indicated that consumers are willing to spend a 20-25% price premium for sustainable products (Ellis et al, 2012; Lin & Chang, 2012). The findings of this study indicate a much lower price premium of around 12% for sustainable products. There may be several reasons for this finding. The first explanation is a familiarization with

sustainable products in general. Sustainability was an innovative concept, often associated with expensive products a few years ago; now the tables might have turned (Lin & Chang, 2012). Consumers don‟t perceive sustainable products as an innovation anymore and therefore value them less. The attitude-behaviour gap described by De Pelsmacker et al. (2005) could provide another explanation (Boulstridge & Carrigan, 2000). This approach

(34)

portrays a gap between what people say they do and what they actually do. De

Pelsmacker et al. (2005) refer to this approach to explain why 50% of the consumers in their study say they are willing to pay a price premium for sustainable fashion but the actual market share is only 1%. This effect could be caused by the fact that a lot of

previous research used a survey method to determine the WTP for sustainable fashion. As a consequence, consumers might overestimate their WTP for sustainable fashion (List & Gallet, 2001 in Ellis et al., 2012). This study provided an incentive for participants to state their WTP more precisely. As a result, the average WTP is lower than those of studies using survey methods. Finally, there are some more straightforward explanations for the difference in results. It could be that the manipulation failed. If the logo‟s indicating the sustainable product didn‟t stand out enough, then it is possible that they did not have any effect. Second, the choice of product might have an effect on the smaller difference between the WTP for sustainable and for conventional products.

Contrary to the insignificant results found for the effect of sustainability on WTP, there was a significant result for both education and age on WTP. The results indicated that the younger participants were willing to spend more on sustainable products than the elder participants. This result corresponds with the earlier findings by Anderson and Cunningham (1972) but conflicts with those by Dickson (2001). A possible explanation could again be the choice of product. Perhaps the jeans appealed more to the younger consumer and therefore led to a higher WTP. Another significant relationship was found between education and WTP. This result is consistent with the findings done by Roberts (1996). This result could be explained by the fact that the sample of this study was skewed towards higher educated people.

Findings concerning the effect of leasing were insignificant as well, although a decreasing trend toward significance could be detected. This was also the case for the

(35)

result of the robustness checks. The participants on average had a higher WTP for the lease product than for to the purchase product. The results resemble the findings done by Dasgrupta, Siddarth and Silva-Risso (2007). This could be explained by the fact that consumers tend to prefer leases with lower up-front costs and monthly payments, even if the net prices of these contracts are higher than purchasing the product.

The level of perceived quality was not different for sustainable products compared to conventionally produced products. In fact, the sustainable and conventional products were perceived as being of comparable quality. This finding is conflicting with the study by Luchs et al. (2010). They suggested that consumers infer a positive association between ethicality and gentleness and a negative association between ethicality and strength, with sustainable products being associated with lower levels of quality. This finding supports the notion of the aforementioned familiarization with sustainable products. Consumers might have lost their (negative) prejudices toward sustainable products and now consider them to be of equal quality with conventional products. The same could be said of the perceived fashionableness of sustainable products. Winge (2008) argued that sustainable fashion is about to lose its negative stereotype of being not fashionable. The results of this study didn‟t indicate that sustainable fashion is perceived as being more fashionable opposed to conventionally produced fashion. However, it wasn‟t perceived as being less fashionable either. Therefore, a turning point might have been reached wherein sustainable products aren‟t a point of difference anymore but becoming a point of parity.

5.2 Implications

5.2.1 Theoretical implications

This study extends the understanding of the effect sustainability has on WTP for products. Furthermore, this study took a first step looking at the effect of both sustainability and

(36)

leasing on the WTP. Additionally, it looked at the role perceived quality and

fashionableness plays in that relationship. The results indicated that although there were positive effects of both sustainability and leasing on WTP, most of these effects were not significant. Possible explanations could be of methodological nature (e.g. sample size too small, manipulation not prominent). Or have a more practical nature, the aforementioned familiarization with sustainable products or the consumer-behaviour gap (Boulstridge & Carrigan, 2000).

5.2.2 Managerial implications

The results of this study have certain implications for practitioners. Although no clear significant difference was found between the WTP for sustainable products and

conventional products, the results do indicate a positive effect of sustainability on WTP. This is also the case for leasable products. Therefore, the fashion industry would be wise to invest in the development of a more sustainable policy. The current study also support the fact that sustainable products are beginning to lose their negative stereotype

concerning the level of quality and that it is perceived as being less fashionable. This could indicate that the average consumer is getting more familiar with the concept of sustainable products and that the barriers of adopting the innovation are disappearing.

The concept of leasing also has several implications for practitioners. Agrawal, et al. (2012) argued that it depends on the type of product whether it is more sustainable to lease it or buy it. They conclude that leasing is more sustainable for products that have a high use impact and low durability. Fashion fits this description. However, results from this study indicate that the consumer might not be ready to lease their clothes yet. Because the number of participants that placed a bid of zero was higher in the lease treatments as opposed to the purchase treatments. Therefore, precautions should be taken when offering leasable fashion to the consumers.

(37)

5.3 Limitations and future research

This study has several limitations; the most interesting ones are discussed here. Probably the most significant limitation was the problem of comparing the WTP of the lease-treatments to the purchase treatment during the analyses. In other words, the lease treatment contained bids per month (for a period of 12 months) and the purchase

treatment just one bid to have complete ownership of the jeans at once. This problem was handled by multiplying the bids in the lease treatment by twelve (months). However, this method is not representative for the real life situation because someone can choose to wear his or her jeans longer than one year. As a consequence, the price per month would go down for the people that purchased their jeans. Therefore, additional robustness checks were performed in which the dependent variable was constructed by dividing the purchase price with the average amount of months a person wears a jeans. The results of the

analysis with the more robust dependent variable were more or less similar to the other analysis. Nonetheless, future studies could find other ways to compare leasing to purchasing.

Furthermore the options to manipulate the independent variables were limited because the Veylinx platform did not allow for elaborate advertisements. Therefore the participants couldn‟t clearly see the object that was auctioned and there was a limited space for the manipulations and follow-up questions. Moreover, this study failed to implement a manipulation check. Future studies could find other ways to manipulate the independent variables and check whether the manipulations have succeeded.

Besides the more practical limitations there also were a few methodological limitations. First, the sample was highly skewed. The sample was prominently highly educated and had an age between and 20-40. This does not represent the Dutch population (CBS, 2013). This skewness could have impacted the results; future studies should

(38)

provide a more representative sample. Second, most of the data was non-normally

distributed. The main reason behind this was the large amount of people that placed a bid of zero. Also, the homogeneity of variance tests was most of the times significant.

Therefore, equal variances could not always be assumed. This violates the assumptions of doing the parametric tests (Field, 2009). Future studies could correct their sample for these errors.

6.0 Conclusion

The aim of this study was to investigate the effects of sustainability and leasing on the WTP for products using an experimental method. Based on previous studies, it was expected that both sustainability and leasing would have positive effects on sales price. In addition, it was expected that the relationship between sustainability and WTP would be positively moderated by perceived quality.

Although sustainability and leasing had some positive effect on WTP, a clear significant result has not been found. The results did indicate significant effects of both age and education on WTP. Younger, higher educated consumers have a higher WTP than older, lower educated consumers.

This study contributes to the literature by further investigating the effect of sustainability on WTP and by making a first attempt to incorporate the effects of leasing on WTP. Using other product types and a more representative sample could be applied in the further research on this topic.

(39)

7.0 Appendix

1.0 Advertisements

Ad 1.1: Female baseline treatment

(40)

Ad 1.3: Female lease treatment

(41)

Ad 1.5: Men baseline treatment

(42)

Ad 1.7: Men lease treatment