Polina Burashnikova 10602925
Consumers willingness-to-pay for effort and rarity
Course: Thesis Proposal
Track: Entrepreneurship and Management in the Creative Industries Supervisor: B. Kuijken
Abstract
Quality signals are proved to have an influence on consumers willingness-to-pay. In this study two different quality signals such as rarity and effort and their effect on consumers purchase decisions are analyzed. Consumers are assumed to be more willing to pay for such products or services that required more effort from producers because people are willing to express their gratitude for the effort made by firms in order to satisfy consumers. On the other hand, rare objects or experiences are usually valued higher because of their unique nature.
The current study combines two concepts and adds to the knowledge about the quality signals. In order to investigate this issue, a set of experiments was run where effort and rarity were addressed to the consumers as quality signals. The empirical setting of this study is an online second-price sealed bid auction where items will be offered and sold. The results indicate the significant effects of both rarity and effort signals on consumers willingness-to-pay.
Table of contents
1. INTRODUCTION 5 2. LITERATURE REVIEW 7 2.1INFORMATION ASSYMETRY 7 2.2SIGNALING THEORY 8 2.3WILLINGNESS-TO-PAY 9 2.4THE CONCEPT OF EFFORT 9 2.5THE CONCEPT OF RARITY 11 2.6INFORMATION OVERLOAD 13 3. RESEARCH METHODOLOGY 14 3.1RESEARCH DESIGN 143.2THE PROCESS OF THE EXPERIMENT 14
3.3.PRODUCT TO BE AUCTIONED 15 3.4AUCTION TREATMENTS 16 3.4.1TREATMENT 1 17 3.4.2TREATMENT 2 17 3.4.3TREATMENT 3 18 3.4.4TREATMENT 4 18 3.5VARIABLES 18 3.5.1INDEPENDENT VARIABLES 18 3.5.2DEPENDENT VARIABLE 18 3.5.3CONTROL VARIABLES 19 3.6SAMPLE 19 4. RESULTS 21 4.1DESCRIPTIVE STATISTICS 21 4.1.1S 21
4.1.2DATA CHARACTERISTICS 22
4.2TESTING FOR NORMALITY 24
4.3HYPOTHESIS TESTING AND REGRESSION ANALYSIS 26
4.3.1HYPOTHESIS 1 27 4.3.2HYPOTHESIS 2 28 4.3.3HYPOTHESIS 3 29 5. DISCUSSION 31 5.1FINDINGS 31 5.2PRACTICAL IMPLICATIONS 31
5.3LIMITATIONS AND FURTHER RESEARCH 32
6. CONCLUSION 34
7. REFERENCES 35
8. APPENDICES 38
APPENDIX 1.QUESTIONS ASKED AT THE END OF THE AUCTION 38
APPENDIX 2.AMSTERDAMMERTJE ADVERTISEMENTS 39
1. Introduction
In many cases, when consumers are making purchase decisions they do not possess enough information about the product offered to them to thoroughly examine its quality. This information asymmetry leads to high uncertainty of the consumers about the product (Akerlof,1970). In order to facilitate consumers’ purchase decisions and make a product more appealing for consumers’, producers have to find strategies for its promotion. One of the ways of promoting a product is through providing consumers with the quality signals about it and therefore reducing uncertainty for potential buyers.
The focus of this thesis is on the analysis of how consumers react to particular information about the product and how these aspects influence their willingness-to-pay. The quality signals that are analyzed in this research are those of rarity of the product and effort required to produce it. There has already been research on the topic of rarity and effort, which can be found in the relevant literature. Kruger (2004) and Morales (2005) have studied consumers’ responses to the information about effort provided by the firms to produce a good or service. The results of their studies suggest that consumers positively react to this kind of information. Therefore, consumers show their gratitude by accepting higher prices for these products or services. The concept of rarity has been analyzed in the works of Brock (1968) and Aggarwal (2011). Similarly, the results of the studies regarding the concept of rarity show that signaling this quality to the consumers increases their willingness-to-pay. After looking at the influence of the chosen quality signals on consumers’ willingness-to-pay separately, it is analyzed whether signaling them at the same time will have the same effect on the WTP. The reason for this suggestion is the fact that consumers might feel “overloaded” with information and make a poor purchase decisions (Jacoby, 1974).
This research is expected to contribute to the current literature on the topic of willingness-to-pay in the following way. This analysis will be the first attempt to examine the influence of the combination of the signals of rarity and effort. Both of these concepts were studied separately but there is no research that combines the analysis of the use of these concepts. In order to make the product more appealing for consumers and increase their willingness-to-pay producers can provide the information about the rarity of the product. In most cases this logically implies that the effort for producing the good or the service is higher due to its scarcity. Therefore, it would be useful to investigate how the two concepts are related to each other. Thus, the following paper is expected to fill the research gap and contribute to the understanding of the topic.
According to the discussion, the thesis will be based on the following research question: What is the influence of providing information about the effort
needed to produce the product and the rarity of the product on consumers’ willingness-to-pay?
The thesis is structured in the following way. First, the literature review on the relevant concepts is presented. Second, the research design and the methods used are described. Third, the results of the analysis are defined and discussed. Finally, the conclusion, practical implications, limitations and propositions for further research are presented.
2. Literature review
The following section includes the discussion of the most relevant topics used in this research. First, the concept information asymmetry is mentioned and followed by the discussion on signaling theory. Second, the concept of willingness-to-pay is defined for a better explanation of the purpose of the following study. Third, the concepts of rarity and effort are described and analyzed. Does the fact that the product offered is rare makes it more attractive for the consumers to buy? What effect does the information about the effort required to produce a product or service has on consumers’ behavior? Finally, the concept of information overload is discussed.
2.1 Information asymmetry
One of the market failures in economics is known as information asymmetry. Extensive research has been done to investigate the sources of information asymmetry and ways to overcome this type of market failure. The most significant research on the topic has been done by Joseph Stiglitz (2001), George Akerlof (1970) and Michael Spence (1973). In 2001, Nobel laureate Joseph Stiglitz presented his study on information asymmetry. This concept takes place when some of the actors in the transaction possess more information than others about some aspects of the bargain (Stiglitz, 2001). George Akerlof (1970) analyzed the market for “lemons” where buyers are drawn away from the market due to the lack of reputation of the sellers and high chance of buying poor-quality products. This kind of situation can arise in different environments: between sellers and buyers, between insurance companies and individuals, between borrowers and lenders. In our case, sellers have more information about an offered object than potential buyers. There are different strategies that can be used by producers in order to reduce consumers’ uncertainty and increase their awareness about the product.
Therefore, in order to reduce the uncertainty cause by information asymmetry, producers need to send relevant quality signals to the consumers.
2.2 Signaling theory
One of the ways to overcome the problem of asymmetric information for producers is to provide consumers with some additional information about the products offered to them. This would decrease uncertainty for consumers and increase their desire to purchase a certain product. There are different types of advertising used by producers to convince people to choose for their products. In this case, a firm “communicates the level of some unobservable element in a transaction by providing an observable signal” (Kirmani, Rao, 20066). There has been significant research on the topic of the influence of different types of quality signals on consumers that suggest the positive relationship between signaling and consumers WTP (Kalra and Li 2008, Kirmani and Rao 2000, Soberman 2003, Moorthy and Srinivasan 1995, Spence 1973). The signals about the quality of the product or service enhance consumer perceived quality and increase their willingness-to-pay (Tsui, 2012). Perceived quality in this case can be defined as the overall subjective judgment of quality relative to the expectation of quality. In relation to the following research, the focus is on how quality signals can influence consumers’ willingness-to-bid (WTB) for objects offered on the market. The results of the study by Tsui (2012) show that advertising regarding different types of product quality has a high influence on consumers’ WTP. Therefore, this thesis will contribute to the current literature on the effect of sending different signals to the consumers in order to increase their desire to purchase the product. In comparison to the previous research on the same topic, the quality signals analyzed in this paper will be those of rarity of the product and effort required for its production. Li et al (2009) investigated the effects of quality signals on consumers in the setting of an online auction. The findings of the research suggests that “quality indicators that
directly reveal information on product quality and seller credibility encourage bidders not only to participate but also to shade bids”(Li, 2009).
2.3 Willingness-to-pay
The main conceptual issue of the paper – willingness-to-pay – can be defined as «the intention to pay a certain amount of money for engaging in a leisure activity or for attaining any other public good» (Ajzen, Peterson, 1988). In our case, consumers' willingness-to-pay is measured by the maximum amount of money that a person is ready to pay for a particular product. Therefore, the higher the bid of the participant of the auction – the higher his/her willingness-to-pay for the product.
2.4 The concept of effort
Even though the concept of effort is an important issue regarding the characteristics of the product, it has not been studied extensively in relation to the consumers willingness-to-pay. Most of the research on the following topic has been focused on consumers gratitude for the effort provided by sales advisers but not on the effort provided by firms in order to create the product. This explains the fact that the current research will contribute to the literature available on the topic of effort. However, some of the studies analyzed how the effort that firms put into the satisfaction of the consumer can influence consumers’ behavior towards the product. Buell and Norton (2011) use the term “labour illusion” to explain signaling of exerting effort by firms. According to the authors of this study, the appearance of effort in terms of labour illusion increases consumers’ perceptions of value (Buell, 2011).
Palmatier et al. (2009) analyze the concept of gratitude as a part of the Relationship marketing strategy. Authors of this study propose that providing extra effort by sellers is one of the relationship marketing investments that “generates
customer feelings of gratitude, which lead to gratitude-based reciprocal behaviors, resulting in enhanced seller performance (Palmatier, 2009).
The concepts of effort and gratitude can also be linked to the concept of reciprocity. The difference between gratitude and reciprocity is that the former concept means sincere acknowledgement and appreciation while the latter one takes place when there is a fit between what is given and what is returned that results in an evaluative match (Tevenar, 2006). Reciprocity was extensively analyzed by Lawrence Becker in his book (1986) on morality and moral theory. He defined the following maxims of reciprocity “that we should return good for good, in proportion t what we receive; that we should resist evil, but not do evil in return; that we should make reparation for the harm we do; and that we should be disposed to do those things as a matter of moral obligation” (Becker, 1986,p. 4).
One of the most extensive studies on the connection between effort and consumers’ willingness-to-pay has been done by Morales (2005). In his paper on consumer responses to high-effort firms Morales (2005) studied the reasons why consumers reward firms for their effort. The results showed that “consumers reward high-effort firms, even when they do not benefit directly from the effort” (Morales, 2005, p. 811). This can be explained by the universal principle, which states that people repay favors provided to them by others. In his study Morales also mentions the concept of reciprocity and distinguishes between “personal” and “general” reciprocity. In the first case, consumers reward firms for actions from which they don’t receive personal benefit while in the second case there is individual effort addressed to the consumer. In his first study Morales investigated whether people reward firms for extra effort how persuasion motives may prevent the rewarding of effort. The results of this study suggest that higher effort levels by firms increases consumers willingness-to-pay (Morales, 2005). Even when two products are perceived to have the same quality, consumers are willing to pay more for an extra effort by the producer.
Another research by Kruger et al (2004) presented an experiment, which was conducted in order to understand whether consumers value objects that required higher effort more. The results of the experiment showed that “the greater the perceived effort invested by the artist, the better they were assumed to be” (Kruger, 2004, p. 97). It is important to mention that this was proved to hold not only in terms of subjective liking of the work but also in terms of assessing the quality of the object.
Based on the following arguments from the current literature on the topic of effort and its influence on the consumers’ behavior, the first hypothesis to be tested was derived:
Hypothesis 1: The higher the effort needed to produce the product, the more
consumers are willing to pay.
2.5 The concept of rarity
Rarity is another aspect that can serve as a quality signal. It is believed that people seem to value rare objects higher due to different factors, such as distinguishing themselves from others, feeling special etc. (Brock, 1968, Aggarwal, 2011, Pocheptsova, 2010). There has been an extensive research in the current literature on the following topic that analyzed the reason why people evaluate rare objects and experiences higher than regular ones.
In their research, Pocheptsova et al. (2010) investigated how consumers evaluate products according to their preferences and choice. Authors of this study analyzed whether consumers prefer special-occasion products to everyday goods and how this choice can be influenced by metacognitive difficulty. This concept can be described as a situation when products feel unfamiliar to consumers, which results in a decrease of the attractiveness of the product.
One of the first theorists who analyzed the concept of rareness and its influence on consumer was Brock (1968). In his paper he discusses implications of commodity theory on value change. According to Brock (1968) people value scarce things higher because of the feeling of personal uniqueness, which is followed by the possession of a scarce good. “Constraining the opportunity to own or experience an object signifies a loss of freedom” (Aggarwal, 2011, p.19). Therefore, people have higher desirability for such objects that are unique in its nature. In their research, Aggarwal et al. (2011) analyze the effects of scarcity messages provided by the producer of the good to the consumers. They refer to this as a concept of Limited-quantity scarcity (LQS). In this case “the promotional offer is made available for a predefined quantity of the product” (Aggarwal et al, 2011,p.19). The results of the studies show that LQS messages about the product or service do have a positive effect on the consumers’ willingness-to-pay. Moreover, compared to the Limited-time scarcity (LTS) where the offer is made available for a predefined period, LQS was found to have a greater influence on consumer behavior.
The same result has been found and described by Lynn (1989) in his paper on scarcity effects on desirability. He studied the effect of scarcity on desirability and found a positive relationship between these two variables. Moreover, Lynn linked scarcity effects to the assumed expensiveness by consumers. An interesting result of his studies was that scarcity effects on desirability were found only in those cases when subjects were likely to think that the scarcity implied expensiveness (Lynn, 1989). Therefore, we can suggest that information provided by the producers about the scarcity of an object increases consumers willingness-to-pay.
As mentioned above, one of the reasons for a purchase of a rare good is the consumers’ desire to differentiate themselves from others by owning an object that others don’t own. Tian and McKenzie (2001) analyzed the aspects of counter conformity that consumers employ in order to differentiate themselves from others.
They define the concept of “consumers’ need for uniqueness” as “individuals' of differentness relative to others that is achieved through the acquisition, utilization and disposition of consumer goods for the purpose of developing and enhancing one’s personal and social identity”. (Tian, McKenzie, 2001, p. 172)
Applied to the case of the research, the following hypothesis regarding the concept of rarity is derived to be further examined in the research:
Hypothesis 2: The more rare the product is the more consumers are willing to pay.
2.6 Information overload
According to the current literature on quality signals, it is assumed that signaling rarity and effort will increase consumers’ willingness-to-pay for a product that meets both of these criteria. However, would the effect on consumers be the same if these two signals were addressed to them simultaneously in the information provided about the product? There has been extensive research on the amount of information that should be presented to the consumers’ about the product in order for it to be effective in terms of increasing their desire to purchase this product. Malhotra (1982) and Jacoby (1984) analyzed the concept of providing consumers’ with “too much” information, which is referred, to as information overload. The result of the study by Jacoby (1974, p.432) suggest that consumers "actually make poorer purchase decisions with more information". The same results were found by Scammon (1977), who proved in his analysis that more information provided to the consumers doesn’t facilitate consumers’ purchase decisions but vice versa. Another interesting finding on the same topic was made by Jacoby et al (1984) who found that consumers decision can be seen as a function of information load. This means that there is an optimal amount of information that can be processed by consumers, while higher or lower amounts of information both lead to a decrease in the quality of purchase decisions. Malhotra et al. (1982) have attempted to re-analyze the
findings of the previous researchers on the following topic of information overload. According to their analysis, it is questionable whether “too much information” actually leads to poorer purchase decisions. Therefore, the findings on the topic of information load and its influence on consumer behavior have been rather contradictory and no clear conclusions have been made. However, according to most of the research on the information load issue, the more information addressed to the consumers, the poorer the purchase decisions made. Following the discussion, the final hypothesis that will be tested in this research was derived:
Hypothesis 3. Signaling effort and rarity together decreases consumers willingness-to-pay.
3. Research methodology
3.1
Research design
The following research is based on the series of experiments. The
experiment type that is used in the research is the field experiment. This is the case
because the experiment will take place in the real world and not in the laboratory conditions. A chosen product will be auctioned online. By looking at the bids of the participants of the auction we will be able to understand what drives consumers willingness-to-pay. Veylinx online auction platform will be used for the auctioning of the objects. All of the auctions on Veylinx are second-price sealed bid auctions also know as Vickrey auctions. This means that participants of the auction do not know each others’ bids and the one bidding the highest price wins but pays the second highest price instead. The advantage of such kind of a demand-revealing technique is that individuals’ limit price can be measured directly and consumers’ willingness-to-pay can be revealed (Noussair, 2004).
3.2 The process of the experiment
Before being able to participate in one of the Veylinx auctions, the potential subscribers had to register with the Veylinx website and provide such information as name, year of birth, gender, email and home address. The process of the auction itself is the following:
First of all, people registered with Veylinx auctions platform receive an invitation to participate in an online auction to their email address. Then the subscribers might decide whether to participate in the auction or not. If a person decides to participate in the auction, he (she) is redirected to the page of the auctioned product.
Second, after being redirected to an ongoing auction the participant has a limited amount of time to place the bid online. During this time, the participant has to decide on the bid for the product. At this stage the participant can still leave the auction by closing the link. Depending on the amount of treatments for a particular auction, participants are redirected to the add corresponding to one of the treatments. For the auction run in this research, four ads were created corresponding to four different treatments. A detailed description of each of the treatments can be found in the next section. Therefore, there were four winners of the auction, corresponding to the amount of treatments.
Third, after bidding on the object the participant receives a notice about his participation in the auction.
Finally, the participant receives a list of miscellaneous multiple-choice questions. Some of the questions were asked to collect more general information about the participants and some were linked to the research question and the corresponding treatments. For this particular research, questions were chosen in such a way so that more information is collected about the participants of the treatment and more correlations are observed. All of the questions were
multiple-choice questions. The list of the questions in Dutch (with the English translation) is enclosed in the Appendix 1.
3.3 Product to be auctioned
The purpose of the current study is to understand how people react to the signals of rarity and effort about products and whether this has an influence on their purchasing decisions and willingness-to-pay. In order to fit the research question, the product to be offered at the online auction had to meet the following criteria. First of all, the product had to be rare and not easy to attain. Second of all, the product had to require effort to be produced and offered to the consumers. After a long process of searching for such a specific product to be auctioned, it was decided to auction an Amsterdammertje – a steel bollard that can be found in the streets of Amsterdam. This object is rare because of the fact that every year the amount of the bollards is being reduced due to their removal from Amsterdam streets by the government ( http://bollardsoflondon.co.uk/the-amsterdammertje-of-amsterdam). On the other hand, it is not easy to acquire and Amstedammertje from the process of digging it from the ground up to the shipment of the bollard to the house. Apart from the initial process of getting an Amsterdammertje out of the ground, it was also polished, packed and shipped by post to the winners of the auction.
3.4 Auction treatments
The experiment included 4 treatments for the product. Table 1 includes the 2x2 matrix that will serve as a base for the treatments of the experiment. Therefore, the experiment in the research will have a factorial design.
Table 1. Treatment matrix
NO SIGNALS RARITY SIGNAL
EFFORT SIGNAL RARITY/EFFORT SIGNAL
Each of the four treatments was presented with a special advertisement of an Amsterdammertje. Participants of the auction would randomly see one of the four ads of the product. All of the ads had exactly the same size (600x625), font and basic features and images. The purpose of this was to exclude the chance of the participants to bid differently according to the design of the ad. The only difference between the advertisements was the different signals that they addressed to the potential customers. In the next section, each of the treatments is discussed and their corresponding signals are described.
3.4.1 Treatment 1
Treatment 1 was supposed to be a baseline of all the treatments. It did not include any of the signals and only had an image of the Amsterdammertje and its dimensions on the advertisement. It was also mentioned on the advertisement of the baseline treatment that an Amsterdammertje can be placed in the garden to give consumers an idea of where to put a steel bollard. The purpose of including such a treatment to the research was to compare the results of bidding for a product when looking at the advertisement with no signals to the other treatments where signaling was used. The advertisement of Treatment 1 can be found in Appendix 2.1.
3.4.2 Treatment 2
Treatment 2 was used to test Hypothesis 1, which referred to the effect of effort signaling on consumers willingness-to-pay. Therefore, it was necessary to provide consumers with information about the effort that it takes to offer this particular product to them. In regards to this an Amsterdammertje, the most difficult part of the process of getting the steel bollard is digging it out of the ground. Therefore, in order to signal effort it was decided to mention on the advertisement that it takes a long time to get an Amsterdammertje out of the ground (see Appendix 2.2).
3.4.3 Treatment 3
Treatment 3 corresponded to signaling rarity of the product, namely Amsterdammertje. Even though there are still thousands of Amsterdammertjes in the streets of Amsterdam, they are currently being removed by the government. Moreover, it is not something that can be easily found in a special shop. Therefore, in order to signal rarity of Amsterdammertjes, it was decided to mention that they are being removed by the government and that it is the last chance to get it for yourself because soon they will disappear from the streets (see Appendix 2.3)
3.4.4 Treatment 4
Finally, Treatment 4 was used to test Hypothesis 3, which corresponded to the question of whether rarity and effort should be signaled together. Therefore, the advertisement of an Amsterdammertje for Treatment 4 combined the information that was used for Treatments 2 (effort) and Treatment 3 (rarity) (see Appendix 2.4).
3.5. Variables
3.5.1 Independent variables
The following research was based on two independent variables. The first independent variable is effort that is signaled to the consumers through the information provided in the product description. The second independent variable used in the research is the rarity of the product signaled to the consumers.
3.5.2 Dependent variable
The dependent variable in the following research is the consumers’ willingness-to-pay which will be measured by the price that people bid for a particular object. This is the best way to evaluate how much the information provided to the consumers about the product influences their purchasing behavior. In the rest of this thesis, variables representing bid amount will be labeled as willingness-to-pay.
3.5.3 Control variables
For the following research some of the control variables were used that were expected to predict the variance of the dependent variable represented by bid amounts of the participants of the auction. The first control variable used in the data analysis was the variable corresponding to the City of Residence of the participant. The collection of this data was done by looking at the answers of the participants about their city of residence. This variable could only take two values: 0 or 1 depending whether a participant lived in Amsterdam or in another city. The second control variable used in the analysis corresponded to the place of birth of the participants. In the same way as the first control variable, it could only take the values of 0 or 1.
The decision to use the first two control variables can be explained by the effect of the identity with the city of the participants. According to Bilkey (1982) there is a significance of the location of the production on the demand. Following
this reasoning, it is suggested that the fact that the product offered comes from a particular city can have an influence on consumers willingness-to-pay for it, especially for those bidders who have a connection with this city.
Finally, the last control variable added to the analysis was the education variable. This variable showed whether a participant has high education or not. The use of the following variable as a control can be explained by the fact that it could serve as a proxy for income of the participants of the auction.
3.6. Sample
Even though there are already more than 3200 subscribers to Veylinx, not all of them are active participants of the weekly auctions. In our case, 870 people participated in the auction of the Amsterdammertjes. Looking at the gender distribution of the sample, we can state that 44,2 % of the participants are female and 55,8% are male. If compared to the average gender distribution of the Dutch population (49,5% - male, 50,5% - female), we can see that the sample is overrepresented by men. Table 3 presents the comparison of the Gender distribution of the sample compared to the Dutch population. Regarding the age distribution, the highest percentage of the participants falls into the young age group (20-40 years old). Therefore, the sample is overrepresented with participants of age between 20 and 40 years old (48,3%) and between 40 and 65 years old (43,9%). The full comparison of the Sample and Dutch population age distribution can be found in Table 2.
Table 2: Age distribution (sample compared to the Dutch population)
Age Sample Dutch Population
< 20 years old 1 % 23,1 %
20 - 40 years old 48,3 % 24,6 %
40 – 65 years old 43,9 % 35,5 %
65 – 80 years old 5,9 % 12,6 %
> 80 years old 0,5 % 4,2 %
Table 3: Gender Distribution
Men Women Dutch Population 49,5% 50,5% Sample 55,8% 44,2%
4.Results
4.1. Descriptive statistics 4.1.1. Sample characteristicsAs stated before, each participant was randomly shown one of the four advertisements, corresponding to one of the treatments. Most of the people participated in the first treatment – 235 people. The rest of the treatments had 195, 219 and 222 participants correspondingly.
The age distribution of the participants was rather homogeneous. It is interesting to mention, that the highest percentage (23,2%) is corresponding to the youngest age group of the participants, which can be interpreted as a higher interest in the auctioned product by younger people.
Another interesting characteristic of the sample that is relevant for the following research is whether participants live in Amsterdam or outside of the
capital. In our case, only 202 participants were residents of Amsterdam while the rest – 668 participants lived in other cities. Table 4 show the percentage of the participants living in Amsterdam and in other cities.
Table 4: People living in Amsterdam
Frequency Percentage
Yes 202 23,2 %
No 668 76,8 %
Total 870 100 %
It was also interesting to see how many people from the sample were not only residents of Amsterdam but were actually born in this city. As mentioned earlier, after the auction participants were asked a few questions related to the auction. One of the questions was whether a person was born in Amsterdam or not. Here data had some missing variables because not all participants answered this question. Out of 870 people, 827 answered the question about the city of birth, of which 98 stated that they were born in Amsterdam and the majority – 729 participants were born in other places. The frequency and the percentage of people who were born in Amsterdam can be found in Table 5. These two variables are important for this research because it is quite interesting to understand whether people who were born in Amsterdam or are the residents of the city are willing to pay more for an Amsterdammertje which is considered to be a symbol of Amsterdam.
Table 5: People born in Amsterdam Frequency Percentage Yes 98 11,3 % No 729 83,8 % Total 827 95,1 % Missing 43 4,9 % Total 870 100 % 4.1.2. Data characteristics
After looking at the characteristics of the sample of the analysis, the next step was to look at the actual data and some important descriptive statistics. The mean bids (in euros) per treatments are shown on the graph 1. From this graph it is visible that the bid amounts increase from Treatment 1 to Treatment 4 but the degree of this increase reduces with each treatment.
Graph 1 Mean bids per treatment (euros)
0 1 2 3 4 5 6 7 8 9 10
Treatment 1 Treament 2 Treatment 3 Treatment 4
Mean bid amount
Moreover, it is important to look at the means, standard deviation, standard error and minimum and maximum bids for each of the four treatments.
From the Table 6 we can see that the lowest bid corresponds to the Treatment 1 (the baseline) and equals 35 euros and the highest bid corresponds to Treatment 4 and is 150 euros. Treatment 4 include both signaling of effort and rarity combined together. Therefore, there is an increase in the bid amount from Treatment 1 to Treatment 4.
Table 6: Descriptive statistics (treatments)
N Mean (€) Std. Dev. Std. Err. Min Max
Treatment 1 235 4,22 7,075 0,462 0 35
Treatment 2 195 7,88 13,219 0,947 0 75
Treatment 3 218 8,43 14,618 0,990 0 100
Treatment 4 222 9,50 17,328 1,163 0 150
Total 870 7,44 13,653 0,463 0 150
Because of the high percentage of the zero bids, it was decided to perform the same analysis including only the bids in the top 50%. The descriptive statistics after splitting the data set can be found in table 7. After the lowest 50% of the bids were excluded from the data, the mean for each of the treatments increased in total from 7,44 euros to 14,77 euros. In this case the minimum total bid increased to 10 euros. If we look at the minimum bids for separate treatments, it is visible that it is higher for treatments 2 and 4 compared to Treatments 1 and 3.
Table 7: Descriptive statistics (bids in the top 50%)
N Mean (€) Std. Dev. Std. Err. Min Max
Treatment 1 117 8,44 8,067 0,745 10 35
Treatment 2 97 15,69 14,179 1,541 20 75
Treatment 3 108 16,79 17,127 1,648 15 100
Treatment 4 110 18,70 20,938 1,996 50 150
Total 432 14,77 16,388 0,788 10 150
4.2 Testing for normality
The next step of the analysis was to test the normality of the distribution of the bids for the product. In order to check whether the data was normally distributed, Kolmogorov-Smirnov and Shapiro-Wilk tests were used. To visualize the distribution, the histograms of the variable were created and can be found on Appendix 3.
First, the test was used to check the distribution for the variable willingness-to-pay. Table 8 shows the results of the normality testing for this variable. From the histogram we can see that bid amount is not distributed normally and is strongly skewed to the right. Furthermore, from the Kolmogorov-Smirnov and Shapiro-Wilk tests it is visible that the variable is not distributed normally (p<0.05). The skewness value of 3.893 proves that the data is skewed. Moreover, the high value of kurtosis (23,179) suggests that the distribution is pointy and heavy-tailed.
Table 8: Normality tests
Kolmogorov-Smirnov Shapiro-Wilk
Statistic P-value Statistic P-value Skewness Kurtosis Willingness-to-pay
In order to normalize the distribution, two steps were taken regarding the transformation of the data. First, the data for the variable of willingness-to-pay represented by the bid amount was transformed using the natural logarithm into a new variable – ln_. Second, the data set was split into two parts – 50% of the highest bids and 50% of the lowest bids. The bids in top 50% were taken for the analysis. After transforming and splitting the data set, the distribution of the bid amount was normalized compared to the results before the transformation. The results of Kolmogorov-Smirnov and Shapiro-Wilk tests for the new variable ln_bid_amount are presented in Table 9.
Table 9: Normality test after the transformation
Kolmogorov-Smirnov Shapiro-Wilk
Statistic P-value Statistic P-value Skewness Kurtosis Willingness-to-pay
(ln_bid_amount)
.121 .000 .970 .000 -.128 .077
Even though Kolmogorov-Smirnov and Shapiro-Wilk tests results show significant results and still suggest that the distribution of the variable of the bid amount is not normal (p<0.05), some improvements of the distribution have been made. However, skewness and kurtosis values were reduced and are now close to zero values. Even though it could not be proved that the dataset is normally distributed, it can be due to the fact that the sample of the research was rather small.
4.3 Hypothesis testing and regression analysis
Finally, linear regression was conducted to explore what are the predictors for particular treatments that explain the results of the bids by consumers. In order to
conduct the regression, a few steps were taken in advance to the analysis. Some additional variables were created each corresponding to the specific treatments – the ones signaling effort or those with rarity signals. Variable “effort_treatments” only took into account the results of Treatments 2 and 4 when variable “rarity_treatments” referred to Treatments 3 and 4. Furthermore, variable “effort_treatments_X_rarity_treatments” referred only to Treatment 4 where rarity and effort was signaled together. Moreover, some of the control variables have been added to the regression analysis. First control variable is the city of residence which refers to whether the participant lives in Amsterdam or not. Finally, education was chosen as a control variable as a proxy for income of the participants. This variable could only get two values according to the level of education of the participants (high education or no high education). For the regression analysis the dataset was split into two groups according to whether the bid was in the top 50% of the bids or not.
For each of the Hypotheses we have first performed the regression analysis with only the control variables (Model 1). In Model 2, the corresponding independent variables were added.
4.3.1 Hypothesis 1
H1 stated: The higher the effort needed to produce the product, the more consumers are willing to pay.
From the results of the auction we can see that H1 is supported because the bids are higher for Treatment 2 (75 euros) compared to Treatment 1 (35 euros). Therefore, there is an increase of 46.6 % in the bid amount for the product.
Table 10: Regression model output (effort treatments)
Dependent variable: Willingness-to-pay
Model 1 Model 2 City of residence .395 (.130)** .384 (.128)** Place of birth .280 (.149)* .231 (.148) Education .182 (.122) .131(.121) Effort Treatments - .351 (.107)*** R2 0.076 0.109 dF 3 4 F 8.082 8.961
Note: standard errors in parentheses *p < 0.1, **p < 0.05, ***p < 0.01
First, the regression analysis was performed only with the three chosen control variables. In this case, Model 1 explains 7,6% of the variables (R2 = 0,76). In
Model 1 the city of residence has a significant influence on the bid amount of consumers (p<0.05). Another significant contribution to the explanation of the variance of the dependent variable comes from the Place of birth variable. The fact that a participant was born in Amsterdam leads to a 28% increase in the bid amount. Furthermore, in Model 2, effort treatments variable was added to the regression. This way, it was possible to see how high is the influence of the effort signaling on the amount consumers were willing to bid for the product. In Model 2, R2 increased to 0,109, which means that Model 2 account for 10,9% of the variance
of the dependent variable of willingness-to-pay. The city of residence still had a significant contribution to the variance of the dependent variable, but the Place of birth did not show significant results anymore. Most importantly, adding the variable of effort treatments to the regression did show significant results (p<0.05). The bids of the participants were 35% higher when there was effort signaled on the advertisement of the product.
4.3.2 Hypothesis 2
Hypothesis 2 stated that: The more rare the product is the more consumers are willing to pay.
In order to test this hypothesis, it was necessary to compare the results of
bidding for an Amsterdammertje from Treatment 1 to the results of Treatment 3 where rarity was signaled. The results suggest that the bids in Treatment 3 were higher than in Treatment 1 with the highest bids of 100 and 35 accordingly.
Table 11 presents the results of the regression analysis used to test Hypothesis 2. In this case, Model 2 included the same variables as Model 1 but also the variable corresponding to rarity treatments. In this case, after the addition of the new independent variable R2 increased to 16,4%. Rarity treatments variable shows a significant contribution (p<0.01) to the variance on the dependent variable represented by willingness-to-pay. The bids of the consumers increase by 56,5% when rarity is signaled to them.
Table 11: Regression model output (rarity treatments)
Dependent variable: Willingness-to-pay
Model 1 Model 2 City of residence .395 (.130)** .408 (.124)** Place of birth .280 (.149)* .273 (.142)* Education .182 (.122) .151. (.117) Rarity Treatments - .565 (.102)*** R2 0.076 0.164 Df 3 4 F 8.082 14.391
4.3.3 Hypothesis 3
Finally, Hypothesis 3 stated that: Signaling effort and rarity together
decreases consumers willingness-to-pay.
The last regression was used to analyze whether signaling effort and rarity together has a significant effect on consumers willingness-to-pay. In this case, in Model 2 the variable corresponding to the combination of rarity and effort treatments. From Table 11 it is visible that R2 shows an increase to 0.118 compared to Model 1 which means that the model accounts for 11,8% of the variance.
Moreover, the regression table shows us that signaling rarity and effort together leads to a 44% increase of the bids for the product. Adding the variable corresponding to the combination of effort and rarity treatments results in a significant contribution (p<0.01) to the variance of willingness-to-pay.
Table 12: Regression model output (effort and rarity treatments) Dependent variable: Willingness-to-pay
Model 1 Model 2
City of residence .395 (.130)** .396 (.120)**
Place of birth .280 (.149)* .237 (.147)
Education .182 (.122) .159 (.116)
Effort & Rarity Treatments - .441 (.118)***
R2 0.076 0.118
Df 3 6
F 8.082 12.396
Note: standard errors in parentheses *p < 0.1, **p < 0.05, ***p < 0.01
5. Discussion
The research presented in this thesis was intended to investigate how particular quality signals influence consumer behavior, specifically their willingness-to-pay for a product. This was tested with the help of an experiment including four different treatments of the same good.
5.1 Findings
The purpose of the following thesis was to understand whether rarity and effort serve as quality signals and increase consumers willingness-to-pay for products. This has been tested with the help of an experiment where auction participants could show their willingness-to-pay through placing bids for the auctioned product.
The results of the research show a high level of significance. It was proved that the two studied characteristics of the product do have an effect on consumers’ purchase decisions. Regarding Hypothesis 1 where the effect of signaling effort was analyzed, it was found that people are willing to pay a higher price when effort is signaled to them. Therefore, Hypothesis 1 was supported. This finding confirms the previous results of the research on the relevant topic. By receiving information about the effort required to bring the product to the market, consumers showed their gratitude and rewarded the firm with higher bids. Morales (2005, p. 811) states that this relationship changes “when firms are thought to be exerting effort with the intent to persuade”. However, this is not the case in the following research which means that the information about effort used in the advertisement did not have an aggressive convincing character and positively influenced consumers’ behavior.
Hypothesis 2 was derived to test the contribution of the signal of rarity on the increase in consumers WTP. In the same way as Hypothesis 1, this suggestion was supported. Moreover, it is interesting to mention that rarity has a higher influence on consumers. In the framework of the current research it was found that consumers are willing to pay more for the same product when the rarity signal is used rather
than effort signal. The findings of this research are in line with the results if the study by Aggarwal et al (2011) who found that signaling rarity by firms can be a important strategic technique. Bringing a “limit” to the quantity of the product provided to the consumers leads to a psychological perception of scarcity (Aggarwal et all, 2011). The high significance of the rarity signal in this case can be explained by the fact that the message combined both the Limited-Quantity scarcity (LQS) message as well as the Limited-Time scarcity message (LTS). This was achieved by signaling that the product is being removed in big amounts (LQS) and will not be on offer in the nearest future (LTS).
Finally, the last Hypothesis tested in this research was used to investigate whether it is efficient to signal such different signals as rarity and effort simultaneously. The regression analysis regarding Hypothesis 3 led to its rejection. Even though the most research on information overload showed that too much information leads to poorer purchase decisions made by consumers (Scammon, 1977), (Jacoby, 1984) in our case the results are the opposite. However, the results of the regression analysis of the current research can be explained in the following way. As stated in the work of Jacoby, Speller, Kohn (1984) consumers’ ability to make the best purchase decisions is poorer at the high levels of information as well as at the low levels. Nevertheless, the results regarding consumer decisions are the opposite at the intermediate levels of information provided to the consumers (Jacoby et all, 1984). In the case of the current research there were only two signals addressed to the consumers which can be seen as the intermediate level of information provided. Moreover, in his research Scammon (1977) states that the quality of consumers’ purchase decisions also depends on the simplicity and the usefulness of the information about the product. In our case, both the information provided about the effort and rarity of the product was simple and useful for the understanding of the purpose of the product and its utilization. Scammon (1977) also mentions that in many case it is better to use “some” information rather that no
information at all. All of the reasons mentioned above can be used as an explanation for the rejection of the third hypothesis of this research and prove that the strategy of using two signals while advertising a product can be an efficient marketing technique.
Moreover, the results of the regression analysis suggest that there are other predictors that can contribute to the understanding of the amounts consumers bids. For example, one of the findings of the research show that consumers who have a higher level of connection to the place where the product comes from. In our case, people who were born in Amsterdam or are citizens of the capital of the Netherlands are willing to pay a higher price for an Amsterdammertje which is a typical Amsterdam symbol than people who do not have this connection with the city. There has not been a lot of research regarding the influence of the connection to the city of birth or residence on the purchase behavior of the consumers who are being offered a product from the same city. However, there is a tendency towards the view that the connection with “home” is a part of the self-definition for people (Beck, 2011).
5.2 Practical implications
The following research has important theoretical and practical implications. First of all, the contribution of this paper to the current literature about quality signals can be explained by the fact that there was no research about the influence of the combination of rarity and effort signals in combination.
In terms of practical implications, the results of the following research an be used in marketing strategies when it comes to advertising the products. As we have seen, both signaling rarity and effort can add to the evaluations of the products by consumers. Moreover, both rarity and effort signals can be combined and signaled simultaneously in order to increase consumers willingness-to-pay.
5.3 Limitations and further research
Even though the analysis of this research suggests some significant results and findings, there are some limitations to it.
First of all, the dataset used to test the hypotheses of the current research only contains participants from the Netherlands which makes the results of the study country-specific. No proof can be found that the same results would be found in the setting of different countries.
Second, from the testing for the normality of the distribution we have found that the data used in the research was not distributed normally. In order to correct for this limitation of the analysis, it would be useful to achieve a higher level normality distribution through the expansion of the sample.
Third, in the further research it would be interesting to investigate whether consumers’ WTP decreases if more information is signaled to them. Therefore, it would be possible to analyze when the amount of information turns from being “sufficient” to “too high” by adding more quality signals to the advertisements.
Finally, the purpose of the following research was to understand how such quality signals as effort and rarity influence consumers’ willingness-to-pay. In our case, this has been tested for a particular product. However, it could be interesting to further investigate whether there is a difference in the degree of influence of these signals between products and services. If further research could be performed in the same setting, the bids for the products could be compared to the bids for the services. In order to perform such a research, there needs to be a service that meets the same criteria as the product – be rare and require effort to be offered to the consumers. An example of such a service could be a museum exhibition of rare objects that requires a long time to be collected and organized. Therefore, the tickets to such an exhibition could be auctioned in the same way as in the current
research and it could be analyzed whether the bids are higher when effort and rarity are signaled to the consumers.
Another possible further research on the same topic could be to investigate whether this kind of signals has an influence on consumers willingness-to-pay when the product offered is radically new innovation on the market. Moreover, it is interesting if the same holds for new products that a lot of information about it actually does not add to consumers WTP in the same way as in the current research.
6. Conclusion
The following study was intended to contribute to the current literature on the quality signals and their influence on consumers willingness-to-pay. The focus of this research is on two specific quality signals: rarity and effort. The hypotheses of the research tested whether these two concepts, both signaled separately and simultaneously contribute to an increase in consumers willingness-to-pay.
In the same vein as the results of the previous studies on the topic of signaling theory, the results of the current research suggest some significant findings regarding the effect of the signals on consumers willingness-to-pay which is represented by the bid amount. From the regression analysis performed in this research it was found that consumers are willing to bid higher amount for the same products when they are being signaled such qualities of the product as rarity and effort. However, the last hypothesis was rejected because the results of the regression analysis showed that signaling rarity and effort simultaneously can actually be effective in terms of increasing consumers willingness-to-pay.
Regarding the managerial implications of the findings of the following research it can be proposed that quality signals decrease consumers uncertainty about products and facilitate purchase decisions. Moreover, the use of the effort and rarity signals on the advertisement of the product can contribute to the amount
of money consumers are ready to pay for it. This is also the case for the combination of these two quality signals that does not lead to an information “overload”, but works as an effective marketing strategy.
7. References
Aggarwal, P., Jun, S.Y., Huh J.J. (2011). Scarcity messages: A Consumer Competition Perspective. Journal of Advertising, 40 (3), 19–30.
Ajzen, I., & Peterson, G. L. (1988). Contingent value measurement: The price of everything and the value of nothing? Amenity resource valuation:Integrating
economics with other disciplines, 65-76.
Akerlof, G. (1970). The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism. Quarterly Journal of Economics, 84 (3), 488–500.
Beck, J. (2011). The Psychology of Home: Why Where You Live Means So Much.
The Atlantic magazine, 2011.
Becker, Lawrence C. (1986). Reciprocity. New York: Routledge & Kegan Paul.
Brock, T.C. (1968). Implications of commodity theory for value change.
Psychological foundations of attitudes, 243-215.
Buell, R.W., Norton, M.I. (2011). The Labor Illusion: How Operational Transparency Increases Perceived Value. Management Science, 57(9), 1564–1579.
CBS (2013). Bevolking; kerncijfers. Accessed on 23 June, 2014.
(http://statline.cbs.nl/StatWeb/publication/?VW=T&DM=SLNL&PA=37296ned&D1 =a&D2=0,10,20,30,40,50,(l-1)-l&HD=120420-1611&HDR=G1&STB=T)
Jacoby, J. (1974). Consumer Reaction to Information Displays: Packaging and Advertising. American Marketing Association, 101-118.
Jacoby J. (1984). Perspectives on Information Overload. Journal of Consumer
Jacoby, J., Speller, D., Kohn, C. (1984). Brand Choice Behavior as a Function of Information Load: Replication and Extension.
Kalra, A. , Li S. (2008). Signaling Quality Through Specialization. Marketing Science, 27 (2), 168–84.
Kirmani A., Rao A.R. (2000). No Pain, No Gain: A Critical Review of the Literature on Signaling Unobservable Product Quality. Journal of Marketing, 64(2), 66-79.
Kruger, J. et al. (2004). The effort heuristic. Journal of Experimental Social
Psychology, 40(1), 91-98.
Li, S., Srinivasan, K., Sun, B. (2009). Internet auction features as quality signals.
Journal of Marketing. 73, 75–92.
Lynn, M. (1989). Scarcity effects on desirability: mediated by assumed expensiveness? Journal of Economic Psychology, 257-274.
Malhotra N. et al. (1982). The Information Overload Controversy: An Alternative Viewpoint. Journal of Marketing. 46(2), 27-37.
Morales, A.C. (2005). Giving Firms an “E” for Effort: Consumer Responses to High Effort Firms, Journal of Consumer Research, 31 (4), 806-12.
Moorthy, S., Srinivasan, K. (1995). Signaling Quality with a Money-Back Guarantee: The Role of Transaction Cost. Marketing Science, 14 (4), 442–66.
Noussair , C. et al. (2004). Revealing consumers’ willingness-to-pay: A comparison of the BDM mechanism and the Vickrey auction, Journal of Economic Psychology, 25, 725–741.
Palmatier, R. et al. (2009). The role of customer gratitude in relationship marketing.
Pocheptsova, A. et al. (2010). Making Products Feel Special: When Metacognitive Difficulty Enhances Evaluation. Journal of Marketing Research, 47, 1059-1069.
Scammon, D. (1977). “Information Load” and Consumers, Journal of Consumer
Research, 4 (December), 148-55.
Soberman, D. (2003). Simultaneous Signaling and Screening with Warranties.
Journal of Marketing Research, 40, 176–85.
Spence, M. (1973). Job Market Signaling. Quarterly Journal of Economics, 87 (3), 355–74.
Tevenar, G. (2006). Gratitude, Reciprocity, and Need. American Philosophical
Quarterly, 43(2), pp. 181-188.
Tian, K., McKenzie, K. (2001). The Long-Term Predictive Validity of the Consumers' Need for Uniqueness Scale. Journal of Consumer Psychology, 171-193.
Tsui (2012). Advertising, quality, and willingness-to-pay: Experimental examination of signaling theory. Journal of Economic Psychology. 33, 1193–1203.
8. Appendices
APPENDIX 1. Questions asked at the end of the auction:
1) Hoe vaak per jaar koop jij tweedehands spullen? (How often do you buy second hand things?)
• Nooit (never)
• een keer per jaar (once a year) • twee keer per jaar (twice a year)
• drie keer per jaar (three times per year)
• vaaker daan drie keer per jaar (more often than three times per year)
2) Heb jij een tuin? (Do you have a garden?)
• Ja (yes) • Nee (no)
3) Heb jij een verzameling? (Do you have a collection?)
• Ja (yes) • Nee (no)
4) Ben je in Amsterdam geboren? (Were you born in Amsterdam?)
• Ja (yes) • nee (no)
5) Zou je een Amsterdammertje voor jezelf uit de grond halen als de gemeente daar toestemming voor gaf? (Would you get an Amsterdammertje yourself if you were allowed to?)
• Ja (yes) • Nee (no)
Appendix 2. Amsterdammertjes advertisements 2.1. Treatment 1. Baseline. No signaling.
2.3. Treatment 3. Rarity signaling
Appendix 3. Normality histograms
3.1 Distribution normality histogram (bid_amount)
