A market for time : fairness considerations in queues

(1)

A Market for Time: Fairness

Considerations in Queues

A Market for Time

Liang Rui (10697357)

University of Amsterdam

Faculty of Economics and Business Business Administration

Specialization: Marketing Academic year: 2015-2016 Supervisor: Anouar El Haji

(2)

Statement of Originality

This document is written by Liang Rui who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

This study aims to investigate the time market - the trading of time with money. By learning about people’s perceptions of fairness about this trade, this study examines the influence of fairness considerations on people’s market exchange norms. An experiment consisting of 360 participants was conducted; the results suggest that people believe that it is fair to trade time with money. Furthermore, the willingness to pay for time is positively predicted by fairness considerations. However, the moderating effects of expected waiting time and recipient were found to be insignificant.

(4)

List of tables and figures

Tables

Table 1 Experimental Manipulation 15

Table 2 Changes of Expected Waiting Time 15

Table 3 Statistic Summary 19

Table 4 Statistic Summary by Positions 20

Table 5 Statistic Summary by Recipient 20

Table 6 Means, Standard Deviations and Correlations 21

Table 7 Result of Direct Effect Analysis 22

Table 8 Summary of one-way ANOVA 23

Table 9 Comparison between treatments 23

Table 10 Result of the moderation effect of Expected Waiting Time 27

Table 11 Result of the moderation effect of Recipient 27

Figures

Figure 1 Conceptual model 12

(6)

1

1. Introduction

In recent years, customer waiting management has become an increasingly important marketing issue. Many scholars have shown growing interests in this area of study. The literature on the subject regards waiting time as an obstacle which constrains the improvement of customer satisfaction (Hui, Thakor & Gill 1998). Katz, Larsen and Balire (1991) conducted an experiment in a grocery store to investigate the relationship between waiting time and customer satisfaction. The study result provides strong evidence that the amount of time customers spend waiting in checkout lines is significantly related to their final satisfaction. Chebat and Filiatrault (1993) confirm the same correlation between waiting time and customer satisfaction. According to Jones and Dent (1993), over 70 % of customers are clearly concerned about waiting times. Extensive waiting time causes negative responses such as anxiety, tension, and helplessness (Carmon, Shanthikumar & Caron 1995). Those negative responses could affect the final evaluation of customers towards services or goods, thus, negatively affect the overall customer satisfaction. In addition, because of improvements in productivity and the increase in customers’ income, customers’ opportunity cost of waiting has increased significantly (Kostecki, 1996). As a result, customers tend to increase their perceptions of the value of time: the more value the customer places on wasted time, the more likely that customer will generate a negative evaluation of the service.

However, in most places, consumer’s waiting time remains an unresolved issue.

Considering the limited capacity of the market, it is hardly enough for service or goods providers to address the issue by simply increasing supplies. In other words, minimizing the actual waiting time is difficult and sometimes impossible. Therefore, it becomes important for firms to adopt a number of coping strategies to deal with this concern. Various research have demonstrated that customer satisfaction is affected not only by waiting time, but also by factors such as customers’ initial expectations, the waiting environment, information transparency and queuing forms (Bitner, 1990; Taylor, 1994; Tom & Lucey, 1995). The research findings provide

(7)

2

evidence that it is possible to address the waiting issue through changing customers’ perception of waiting. By doing so, firms can design operations to provide pleasant waiting environments, keep customers well-informed about the waiting process, and provide fair queuing order. Improving such operations can generate promising results in customers’ waiting evaluation. If customers understand that they are obliged to wait, they feel more pleasant if they receive a better service while waiting in a fair and transparent environment. Therefore, in terms of coping with waiting, this operational management becomes essential for firms.

Among coping strategies, queuing is most widely used. Queues are viewed as social systems created to coordinate the delivery of goods and services, and to reduce the costs of service delivery (Hall, 1991; Saaty, 1961; Schwartz, 1975). They have been viewed as the fairest way to allocate goods and services when market demands exceed supplies. Rafaeli, Barron and Haber (2002) point out that different queuing orders and structures can affect customers’ perceptions of waiting which in turn generate different evaluations for the services customers are waiting for. Research on social norms suggests that, besides the market exchange norm, people are also under the control of social norms (Fiske, 1992; Mcgraw et al, 2003; McGraw & Tetlock, 2005). The most important social norm in queues is fairness, also called social justice. Customers perceive that whether waiting behaviors are fair or unfair will determine their perceptions toward the service. In other words, when customers feel that they are subjected to unfair treatment, they generate more negative perceptions about the service. As suggested by Larson (1987), customers may become infuriated when they experience social injustice, which was defined as a violation of the first-in, first-out norm (FIFO norm). This violation results in negative attitudes, and ultimately negatively impacts a firm’s market share. In other words, fairness consideration does have an effect on the customer’s evaluation of the service received. However, using queues to allocate the delivery of goods and services may not be the most efficient, or at least the most economic way, as queuing positions cannot reflect the real opportunity costs of each customer in the queue (Oberholzer-Gee, 2006). For instance, people

(8)

3

standing at the front of the queue may not be in a hurry, while those at the back may require the service urgently. Therefore, the concept of a market for time has been raised. Perhaps queuing can be more efficient if customers are allowed to trade positions. Oberholzer-Gee (2006) studied this question. In his experiment, he assessed people’s reactions after letting participants ahead of the queue by offering monetary compensation to those in front. As suggested by Oberholzer-Gee, the market for time can hardly exist as the disruptive behavior caused increased negative externality to others, and people do not feel that they have the right to do so. Indeed, such behavior causes a great deal of negative externality not only to those standing in line, but also to the service provider; this makes queuing uncontrollable. An interesting finding in Oberholzer-Gee’s study is that most participants expressed the desire to trade places for money at the beginning. While, due to fairness considerations, this was not acceptable in the end.

Apart from Oberholzer-Gee’s research, in previous literature, limited attention has been paid to the role of social norms in waiting time market. In this paper, the waiting time market was studied to investigate how fairness considerations affect people’s perception of queues.

(9)

4

2. Literature review

This chapter covers the most relevant research from the existing literature related to fairness considerations and the willingness to pay. First, a general description about the waiting issue in the market will be provided. Second, key concepts of this paper such as queuing and social norms will be presented. This chapter then discusses the link between social norms and willingness to pay, in other words, how social norms affect people’s willingness to pay. Finally, research model with related hypotheses will be demonstrated.

2.1 Waiting time and customer perceived value

Waiting, as a universal phenomenon, is considered a negative experience for customers (Scotland, 1991). A significant amount of consumers’ time is spent waiting, thus it creates high marketing costs, which managers should know how to deal with (Kostecki, 1996). As suggested by Kostecki (1996), consumer satisfaction is an increasingly important determinant of business success, as extensive waiting time can greatly harm customers’ evaluation of services received. Furthermore, customers’ perception of the value of time is greatly increasing due to the increase in income and improvements in productivity. The importance of waiting time management not only affects consumer satisfaction, but also is influenced by the idea that time has become a major factor of competitiveness (Kostecki, 1996). However, in most cases, waiting is still an unresolved marketing issue.

According to Hornik (1984), there is a distinction between objective waiting time (actual waiting time) and perceived waiting time (expected waiting time). Based on a field study, Gail and Scott (1997) found that customers’ satisfaction is mainly determined by the perceived waiting time, and not by the objective waiting time. In general, customers tend to experience less satisfaction when they spend a significant amount of time waiting. Hence, service satisfaction is negatively related to the perceived waiting time (Scotland, 1991).

(10)

5

In the field of marketing, managers usually adopt two methods to address waiting time: operation management and perception management. Considering the limited capacity of service or goods providers, perception management becomes particularly important in dealing with waiting (Yu, Kang and Lien, 2014). As suggested by Homik (1984), customers tend to overstate the amount of time they spend waiting in lines. Factors such as anxiety, uncertainty, unfairness, and perceived value of goods or services can add to consumer’s perceived waiting time (Scotland, 1996). Therefore, managers must recognize the specific effect of waiting environments and checker speed on perceived waiting time and customer satisfaction.

It is interesting to note that Kostecki (1996) points out that waiting is not always a negative experience. Consumers tend to overestimate the value of goods or services due to long waiting periods. As suggested by Kostecki (1996), long waiting periods intensify attractiveness to consumers and, in certain cases, increases overall satisfaction, as consumers can be more prepared to be served. Thus, it is logical that some managers use long waiting period as a tool to stimulate consumers’ desirability of goods or services. However, using waiting period as a sign of attractiveness is inconclusive, tricky, and may increase negative experience.

2.2 Waiting time and market efficiency

Various companies use customer waiting time as a marketing tool to stimulate demand for specific products. Nevertheless, it is obvious that waiting time in most cases brings negative externality to customers. In the study by Lin and Chang (2011), factors such as queuing, delays, or a customer’s early arrival could add to the waiting time. Customers evaluate the value of waiting by their desire for a specific product or service by comparing the actual waiting time with the expected range of waiting time. According to Lu, Musalem and Olivares (2013), purchase incidence is considered to be affected more by the length of the waiting line rather than the speed of service, and customer’s sensitivity to the queue length is negatively correlated with price sensitivity.

(11)

6

The experiment result in Lin and Chang’s study (2011) suggests that it is important for managers to provide pleasant waiting experiences and shorten the expected range of waiting time. In a later study, Osmehkin and Deleze (2013) researched waiting time distribution and market efficiency in pairs trading. The result suggests that waiting time results in two different prices: inefficient price and efficient price. Waiting time distribution illustrates that inefficient price states outside optimal barrier levels rapidly converge back to efficient price states within the optimal boundaries. A later study by Bin Li and Ozgua (2013) pointed out that waiting costs of customers may reduce the positive desire for a product which in turn reduces market efficiency. The longer people wait, the more likely the business or organization will lose a potential or optimal transaction. Another study by Kulshreshtha (2015) provides an interesting view on waiting trade. The study analyzes the relationship between opportunity costs of waiting and bribery in rationing by waiting. The result suggests that illegal behavior to reduce waiting time actually benefits both parties, and can help to achieve maximum profit. This could show that a short waiting time does have a positive benefit to market efficiency.

2.3 Queue and structures

Almost every organization uses queues to distribute the delivery of goods and services. It is generally agreed that queues hold three major features: first, it regulates and coordinates the order in which people receive goods or services; second, queuing must assume a spatial form; third, the basic order of queues relies on people’s shared knowledge of standards (Milgram & Toch, 1969). The literature concerning queuing dates back to Mann (1969). He regarded queuing as a social system with its own set of norms, and suggested that people waiting in queues follow a basic principle of the first-in, first-out (FIFO) priority. The FIFO norm can be regarded as a form of social justice.

According to Hall (1991), three major types of queues exist in the market: Multiple Queues, Single Queues and Numbered Queues. The level of people’s considerations for fairness

(12)

7

varies according to each type of queue. In the Multiple Queue, several waiting lines are formed at the same place. Individuals decide which line to enter based on a short observation. A typical example of multiple queuing is demonstrated in a supermarket where multiple lines exist while customers wait for checker services. However, as the processing speed of each queue cannot be guaranteed at the same level due to employees’ skills or motivation, this multiple queue cannot always serve the FIFO norm, and even suggests a potential lack of fairness (Anat et.l, 2013). In the Single Queue, all customers enter the same line in which they wait for service. In the Numbered Queue, each customer is given a number according to their orders of arrival. Instead of standing in line, customers are free to stand or sit anywhere until their numbers are announced. In both Single Queues and Numbered Queues, the process of queuing is performed under the regulating FIFO norm. Thus, these queuing structures hold an implicit promise of fairness (Miller, Kahn & Luce, 2006). Therefore, I will use the Single Queues as my research type in a later section of this paper.

Another factor which has a major effect on people’s considerations of fairness is the actual waiting condition. As shown in the study by Miller, Kahn and Luce (2006), factors such as environment, processing time, and number of people can affect the overall waiting experience. The environmental factor has become particularly important to firms in terms of waiting time management. By creating pleasant waiting conditions, service providers can affect customers’ perceptions of waiting, thus improve the overall evaluation of customers about the service.

2.4 Social norms and fairness considerations

As mentioned above by Mann (1969), queuing has its own set of norms. The FIFO norm has been regarded as the fundamental principle of queuing. People consider queuing behavior as fair or unfair depending on whether that behavior transgresses queuing norm, in thiscase, the FIFO norm. Schwartz (1975) studied social norms of queuing and how those norms regulate queuing. His study further confirmed the finding by Mann (1969) that the principle of queuing

(13)

8

is sociological: “competitive allocation cannot operate without institutionalization of norms defining the limits of legitimate action, particularly in this case with regard to the legitimacy of means of attaining the goal.” (p. 93).

However, queuing behavior does not always follow the FIFO norm. When perceived waiting time or risk of disappointment increases, people try to “jump in queue”. As a matter of fairness, such behavior is strictly banned. Those intending to jump in queue usually do so as a matter of urgency (Gad & Eran, 2012). As this behavior often increases the waiting time of others, thus raising the negative externality to other people, such excuse is unacceptable. Milgram (1986) conducted an experiment to determine how people react toward intrusions in queues. He regarded queuing behavior as typical of the manner in which individuals build a social system. According to Milgram, what regulates an individual’s behavior in queuing is a commonly accepted belief, also called distributing justice. In his experiment, people who wanted tojump in queue by providing an urgent excuse were rejected. Milgram’s (1986) study illustrates that when those waiting in a queue have the same waiting cost, the FIFO becomes the only priority. Schmitt (1992) studied a waiting line from an individual perspective. He assumed that an individual’s behavior can be better explained by personal interests rather than social norms. Nevertheless, his study presents strong evidence that queuing forms a social system with its own set of norms and an individual’s queuing behavior is greatly regulated by those norms.

2.5 The effect of fairness norms

With respect to queuing theory, how does fairness consideration matter? Many studies on this issue have been conducted. Larson (1987) studied the effect of fairness norms on individuals’ attitudes. His study suggested that features such as waiting environment, social injustice, and service delay can affect customers’ attitudes. He defined social injustice in queuing as deviations from the FIFO norm. However, he does not provide further explanation for this definition. Allon and Hanany (2012) extended Larson’s study by proving that the FIFO norm is

(14)

9

indeed the only priority in queuing. LoMonaco (2008) conducted a research in relation to an overnight queue for tickets for a music event. The results illustrate that satisfaction is affected by whether queuing behaviors can be regulated by distributing justice.

Furthermore, there is a growing interest in the study of various factors, apart from economic, that could affect decision-making. Numerous laboratory studies prove that people do not always make decisions for the sake of maximizing pay-offs. This could be explained by social consideration. In fact, in some cases the norm of social exchange affects decision-making over the norm of market exchange (Fiske 1992, McGraw et al. 2003, McGraw and Tetlock 2005). The study by Heyman and Ariely (2004) suggests that people are willing to recognize more effort from a social contract than from a small or moderate amount of money.

According to Konow (2003), individuals are likely to make sacrifices to change the pay-offs of other players as a matter of fairness considerations. As suggested by Charness and Rabin (2002), individuals also make sacrifices in order to help others. In certain cases, people are willing to wait longer when the increased overall welfare of doing this is large (Oberholzer-Gee, 2006). Although the element of fairness can be regarded as a constraint on profit-seeking activities in the market (Kahneman & Thaler, 1986) – it prevents individuals from accepting monetary compensation – customers are still willing to accept certain negative externality if their actions can help more people (Oberholzer-Gee (2006). In other words, people feel that they do not have the right to exploit monetary benefits when market demand exceeds supply, but they feel that they have the obligation to take social responsibility when they have to make sacrifices

2.6 Economic reasons undermine the FIFO

Although the FIFO has been viewed as the fundamental principle of queuing, it is undermined quite often. People ignore the FIFO norm and cut in line, thus raising negative externalities to others. From the service provider’s perspective, if the expected waiting time increases due to

(15)

10

delay or limited capacity, some customers tend to jump in the queue, even though it is generally unacceptable by most people. However, economic reasons can also explain this jumping behavior. From the customer’s point of view, previous literature pointed out the conflict between FIFO norm and economic reasoning, thus provides an explanation to behaviors that undermine the FIFO norm. Strategic customers tend to maximize the efficiency of their market exchanges. When customers wait in queue, positions do not reflect an individual’s opportunity cost (Oberholzer-Gee, 2006). As suggested by Allon and Hanany (2012), people do not follow the FIFO norm just for the sake of it. People overserve the FIFO norm as it serves their best interests. Therefore, when economic reasons are taken into account, the FIFO is hardly the priority for customers. However, Allon and Hanany criticized the economic conflict by pointing out that the FIFO norm may be value-enhancing for both customers and service managers in the long run. Therefore the FIFO norm is consistent with strategic customers’ value maximization. Nevertheless, most queues are short-lived, and the conflict between social norm and economic norm can hardly be avoided. Oberholzer-Gee conducted a research about people jumping in line with monetary compensation. The result suggests that most participants were willing to trade positions with monetary compensation. However, as a matter of fairness considerations, participants who initially expressed willingness to trade places ultimately did not do so, as they did not feel that they have the right to trade positions. According to Oberholzer (2006), there are two factors which people consider beyond social norms, before making decisions: transaction costs and negative externality created by position-trading strategies. In his experiment, trade of places increases waiting time, thus raising negative externality to others. Other experiments conducted by Kahneman, Knetsch and Thaler (1986), and Frey and Pommerehne (1993) suggest that people are not willing to let others jump in line because of considerations of fairness. However, in Oberholzer’s experiment, participants who did not agree to trade places were afraid of possible punishment by others. This finding suggests that the element of fairness may not be the major reason that prevents people from trading places, as the

(16)

11

fear of negative consequences might be the real issue here. Knoow (2000) conducted a study about fair shares in allocation suggesting a similar implication – that customers do not feel they have the right to adjust position allocation, which provides an explanation for the above experimental results.

2.7 Willingness to pay for time

As stated before, strategic customers tend to experience the tension between economic norms and social norms. In other words, between social considerations and value or utility maximization, customers can hardly map the trade-off. The time market exists in reality; the re-sale of tickets is a typical time market. Customers who do not want to spend time waiting or have access to tickets are likely to go for re-sale tickets. In this case, strategic customers seek value maximization to trade time with money. However, due to legal considerations, this market is not accepted by most institutions. Another example of time market is the U.S. medical market. A field study by Anderson et.al (2015) illustrates that people are willing to pay for short waiting periods in hospitals. In addition, people believe that it is fair for serious patients to pay to receive quick treatment. In fact, as suggested by Anderson et.al (2015), customers are willing to do this due to the increase of waiting time in the American medical service market.

With respect to the willingness-to-pay theory, it is hard to measure customers’ willingness for time trade, as individuals’ market exchange considerations are quite different. The time customers willing to spend on waiting based on the desired size for a particular good or service.This means that it is possible to measure customers’ willingness to pay for shorter waiting time by desired set size. Previous theory highlights that customers’ initial desire for purchase has a direct effect on their willingness to pay. In the study by Hafenbrad, Hoffrage and White (2013), the desired size has a linear effect on willingness-to-pay and the non-monetary costs of waiting may reduce customers’ willingness to pay.

(17)

12

2.8 Research gap

As the literature review above demonstrates past studies focus on the relationship among fairness perception in queues, waiting time and willingness to pay method. Unfortunately, to my best knowledge, no further study has focused on the effect of fairness considerations on willingness to pay in terms of the time market, particularly when people are allowed to trade time for money, how fairness considerations play a role. Therefore, this paper will present a potential mechanism for time and position trading in queues. The research question of this paper is “To what extent considerations of fairness affect people’s willingness to pay in the market of time”.

2.9 Conceptual Framework

This study is aimed to investigate how fairness considerations affect willingness to pay in terms of time trading. The conceptual framework is generated to illustrate the expected relationship between fairness consideration and willingness to pay. Therefore, in the conceptual model, fairness consideration is the predictor variable for willingness to pay (WTP). Furthermore, expected waiting time and payment recipient are expected to moderate the direct relationship between fairness consideration and willingness to pay.

(18)

13 Fairness consideration

Fairness consideration is defined as the extent to which people consider whether one’s behavior or exchange is fair or unfair. Fairness consideration has been viewed as a constraint on profit seeking activity (Kahneman, Knetsch & Thaler, 1986). When people perceive a market exchange as unfair, they will be less willing to accept the trade; or provide negative evaluations if they have to accept it. In Oberholzer-Gee’s (2006) experiment, participants who perceived more fairness shown more willingness to let people cut in line with monetary compensation. Thus, I hypothesize:

H1. Fairness consideration is positively related to willingness-to-pay.

Recipient

Individuals are likely to make sacrifices if their behaviors are particularly helpful to social welfare. (Andreoni & Miller, 2002; Charness & Grosskopf, 2001; Charess & Rabin, 2002). As suggested by Oberholzer-Gee (2006), the size of externality can explain an individual’s desire to accept monetary compensation. When the exchange is made through a donation, there is more positive externality when compared to when the exchange is made through direct trade. Therefore, I hypothesize:

H2: The relationship between fairness consideration and WTP is negatively moderated by

the payment recipient. When people exchange places, the direct effect of fairness consideration on WTP is low when compared with exchanging places by donation.

Expected waiting time

(19)

14

different levels of fairness consideration. The expected waiting times for those at the front is shorter than people stand at back places. Therefore, people at the front tend to care less about whether queuing is fair. Thus, I hypothesize:

H3: The positive relationship between fairness consideration and WTP is positively moderated by the individual’s expected waiting time.

(20)

15

3. Method

This chapter represents the start of the empirical stage of my study. First, a description of the experiment design is given. Then, the most relevant characteristics of the collected data are shown. After that, all variables in the surveys are discussed.

3.1 Experimental design

In order to conduct my research, an experiment was designed to investigate fairness consideration and how it affects people’s willingness to pay. The experiment was performed based on the follow scenario. Three people are standing in a queue in an airport waiting for check-in service, as illustrated in Figure 2. The person in the front (A) must wait for 5 minutes to check in, the person in the middle (B) must wait for 20 minutes, and the person in the back (C) must wait for 35 minutes. The experiment consists of six treatments according to various experimental manipulations as demonstrated in table 1. By completing an online survey, participants entered 6 treatments randomly as being one person in that queue (A, B, or C).

(21)

16 Table 1

Experimental manipulations

Manipulator Front position (A) Middle position (B) Back position (C)

Trade for money(M) MA MB MC

Trade for Donation (D) DA DB DC

Table 2

Change of expected waiting time (minutes)

Waiting time before trade Waiting time after trade Waiting time changes

A 5 30 + 25

B 20 20 0

(22)

17

3.2 Sample

In order to conduct my experiment, I approached 385 individuals online. On 27th December 2015, I distributed my survey on the online marketing research platform Amazon Mechanical Turk (www.mturk.com). The survey is compensated. Each participant who successfully submits a survey could receive 0.15 U.S. dollar on the platform. Among the 385 participants that have started completing the survey, 360 have fully completed it (response rate 93.5%), and each treatment received 60 responses separately. On 8th January 2016, the survey was closed.

The collected sample consists of 172 males and 128 females (Mage = 33.9, SDage = 30, age-rage: 18-76 years). 72.15 % of participants completed at least a bachelor education (Bachelor = 46.7%, Master = 18.6%). In addition, 91.8% have had work experience (Mwork = 12.8, SDwork = 10, work-rage: 0-48 years).

3.3 Measurement of variables

Fairness consideration

A scale was used to measure fairness consideration (Kahneman, Knetsch and Thaler, 1986). The survey asked respondents to what extent he or she considers trading places as fair. The variable was used on a 5-point scale (completely unfair, unfair, neither unfair nor fair, acceptable, completely fair).

Willingness-to-Pay

In this measurement, participants who entered the survey as being people at the front place were asked to state the amount which they would like to receive; participants who entered the survey as being people at the back place were asked to state the amount which they would like to pay; participants who entered the survey as being people at the middle place were asked to state the

(23)

18

amount which they consider should be paid for the trade. No maximum or minimum amount was set.

Recipient

As a recipient acts as a manipulator in the experiment, the measurement is fixed. In treatment MA, MB, and MC, the recipients are people standing in the queue, in treatment DA, DB and C, the recipient is a charity organization.

Expected waiting time

The experiment consisted of six treatments. The variable of expected waiting time depends on the random role of the participant. Therefore, each participant’s expected waiting time is fixed. For example, in treatment MA, the participant plays the role of A, thus the expected waiting time for A is 5 minutes.

Control Variables

Four control variables were used: Gender, Age, Education level, and Work tenure. The last section of the survey includes those items.

(24)

19

4. Results

This chapter presents the results of my study. First, the process of data preparation will be presented. Then, a correlation matrix (Table 6) of all variables will be outlined. After that, linear regressions for direct effect analyses will be shown. Finally, the moderating effect of expected waiting time will be discussed.

4.1 Data preparation

Before performing the data analysis, I examined the frequency table for errors in the raw data, no errors were found. Also, there were no missing values in the compulsory questions of the survey. In addition, as no counter-indicative items were used in the survey, there was no need to recode items. Furthermore, I performed dummy coding on the control variables of age, education and work tenure. For age, 0=young age (≤30 years), 1= old age (>30 years); for education, 0= low education (highest received education is high school), 1=high education (at least received a bachelor education); for work tenure, 0=short work experience (≤10 years), 1=long work experience (>10years).

4.2 Statistic summary

Table 3 lists the descriptive statistics of study variables. As shown in Table 4, participants’ fairness considerations were significantly higher when they were at the front position than at the back position. Accordingly, participants at the front have a higher willingness-to-pay, while those at the back have lower willingness-to-pay. Furthermore, as shown in Table 5, participants under the donation condition have a higher level of fairness consideration than those under the trade condition. More interesting, participants were willing to pay more for donation than trade.

(25)

20

Note: N=360, for age, 0=young age (≤30 years), 1=old age (>30 years); for education, 0= low education (the highest received education is high

school), 1=high education (at least received bachelor education); for work tenure, 0=short work experience (≤10 years), 1=long work experience (>10years). Table 3 Statistic summary Treatment Variable MA MB MC DA DB DC M SD M SD _M SD M SD M SD M SD Fairness consideration 3.68 1.07 3.37 1.15 3.35 1.18 3.11 1.12 3.28 1.18 3.30 1.24 WTP 17.63 14.96 11.67 10.43 6.95 6.57 17.90 18.43 14.71 15.30 8.23 7.16

Age (0=young age, 1=old age) .56 .52 .51 .50 .49 .48 .50 .49 .53 .51 .52 .51

Gender (0=female, 1=male) .49 .47 .55 .53 .51 .50 .51 .50 .52 .50 .50 .49

Education (0=low edu, 1=high edu) .23 .42 .26 .45 .21 .40 .22 .41 .23 .43 .25 .44

(26)

21 Table 4

Statistic summary by position (expected waiting time)

Position Fairness Consideration WTP

M SD M SD Front (A) 3.40 1.13 17.76 16.71 Middle (B) 3.24 1.14 13.19 13.13 Back (C) 3.02 1.20 7.59 6.87 Note: N=360 Table 5

Statistic summary by recipient

Recipient Fairness Consideration WTP

M SD M SD

Trade 3.03 1.14 11.08 11.96

Donation 3.57 1.18 13.61 14.91

Note: N=360

(27)

22

4.3 Correlation analysis

Table 6 provides an overview of the correlation between each two variables. First, work experience is found to be positively related to age(r=.90, p<.01) and negatively related to education (r= -.11, p<.05). In terms of willingness-to-pay, fairness consideration is highly positively related to it (r=.38, p<.01). Furthermore, fairness consideration is found to have different correlations with demographic variables: it is negatively related to age (r=-.11, p<.05) and work experience (r=-.14, p<.01).

Note: N=360

*_{Correlation is significant at the .05 level (one-tailed).} **_{Correlation is significant at the .01 level (one-tailed).}

Table 6

Means, Standard Deviations and Correlations

Variables M SD 1 2 3 4 5

1. Age (young age=0, old age=1) .52 .50 -

2. Gender (female=1, male=0) .52 .50 .00 -

3. Education (Low edu=0, high edu=1) .23 .43 .02 .02 -

4. Work experience (Short work=0, long work=1) _.44 _.50 _.90** .01 -.11* -

5. Fairness Consideration 3.35 1.16 -.11*_-.09 .04 -.14** -

(28)

23

4.4 One way ANOVA

A one-way ANOVA analysis was performed to compare means of WTP between treatments. As shown in Table 10, there was a statistically significant difference between treatments as determined by one-way ANOVA (F (5, 354) = 7.18, p<.001). Table 11 illustrates that WTP was statistically significantly lower in treatment MC (M= 7.02, p<.001) than in treatment MA (M=17.63), and lower in treatment DC (M=8.23, p<.05) than in treatment DA (M=16.21). Furthermore, WTP was significantly higher (p<.05) in treatment DA (M=16.21, p<.05) than in treatment MA (M=15.63), higher in treatment DB (M=15.90, p<.05) than in treatment MB (M=12.73), and higher in treatment DC (M=8.23, p<.05) than in treatment MC (M=7.02).

Table 7

Summary of one-way ANOVA analysis

Sum of Squares DF Mean Square F Sg.

Between treatments 5920.55 5 1184.11 7.18 .000

Within treatments 58377.51 354 164.91

Total 64298.06 359

Table 8

Comparison of WTP between treatments

Treatment Mean SD N MA 15.63 1.93 60 MB 12.73 1.32 60 MC 7.02 .85 60 DA 16.21 2.43 60 DB 15.90 1.88 60 DC 8.23 .92 60

(29)

24

Table 9

Result of direct effect analysis

Willingness-to-Pay

Variable B SE B β

Step 1

Age (0=young age, 1=old age) .46 2.01 .17*

Gender (0=female, 1=male) .09 1.41 .00

Education(0=low edu, 1=high edu) _-.06 _1.68 _-.00

Work experience(0=short work, 1=long work) -.39 2.02 -.15*

Step2

Age (0=young age, 1=old age) 3.66 1.87 .14*

Gender (0=female, 1=male) _1.09 _1.31 _.04

Education (0=low edu, 1=high edu) -.52 1.56 -.02

Work experience (0=short work, 1=long work) -3.60 1.88 -.13*

Fairness Consideration 4.33 .60 .38**

Note: N = 360.

*_{significant at the .05 level (one-tailed).} **_{significant at the .01 level (one-tailed).}

(30)

25

4.5 Direct effect

In order to test the direct effect of fairness consideration on people’s willingness-to-pay, I performed hierarchical regression analyses. In step 1, I entered control variables of age, gender, education and work experience as independent variables and WTP as the dependent variable. As presented in Table 9, a direct relationship was found between two control variables (age and work experience) and WTP: age is positively related to WTP (β= .46, p<.05) and work experience is negatively related to WTP (β= -.39, p<.05). However, the control variables of gender (β= .09, p>.05) and education (β= -.06, p> .05) cannot significantly predict WTP. In step 2, I entered control variables and fairness consideration as independent variables and WTP as a dependent variable. The model was statistically significant F (3, 356) = 21.32; p<.01. The introduction of fairness consideration explained the additional 13.7% variance in WTP, after controlling for age, gender, education and work experience (R2_{change= .137; F (1, 136) = 57.5).}

In the final model, among the five predictor variables, three were statistically significant, with fairness consideration recoding a higher Beta value (β=.38, p<.01) than age (β=.14, p<.05) and work experience (β=-.13, p<.05).

Overall, the regression results suggest that fairness consideration is significantly related to willingness to pay. Thus, hypothesis 1 was supported. In addition, willingness to pay was found to be significantly predicted by control variables age and work experience.

(31)

26

4.6 Moderation effect

In chapter 2, hypotheses 2 and 3 proposed that there are two moderation effects of expected waiting time (cited as position) and recipient on the direct relationship between fairness consideration and willingness-to-pay. In other words, people standing at different positions (A, B, C) and facing different recipients have different levels of fairness consideration and willingness-to-pay. In order to test hypotheses 2 and 3, I performed moderation analyses in SPSS. Before starting the test, I constructed two interaction terms to my regression model by adding standardized multiplications of fairness consideration position (ZF*ZP) and fairness consideration * recipient (ZF*ZR).

Table 10 presents the result of expected waiting time as a moderator of the relationship between fairness consideration and WTP. As shown in Table 8, fairness consideration has a significant effect on WTP (p<.01). However, there was no significant interaction between fairness consideration and expected waiting time (p>.05), which means that there was no moderation effect of expected waiting time on the direct relationship between fairness consideration and willingness-to-pay. Therefore, further steps cannot be followed to analyze the moderated effect; hypothesis 2 was rejected.

In terms of the moderating role of recipients, Table 11 presents results of the statistics. The predictor variable, fairness consideration, was found to have a significant effect on WTP (p<.01), while the interaction term (ZF*ZR) had no significant effect on WTP (p>.05). Thus, hypothesis 3 was rejected.

As neither expected waiting time nor recipient has a significant moderating effect on the direct relationship between fairness consideration and WTP, both hypotheses 2 and 3 were rejected.

(32)

27 Table 10

Result of the Moderation Effect of Expected Waiting Time

Coefficient SE t Sig. Constant .06 5.19 .31 .69 Fairness consideration 6.46 1.46 4.44 .00** Position -.79 2.35 -.33 .53 ZF*ZP -1.08 .66 -1.63 .10 Note: N=360

**_{Correlation is significant at the .01 level (one-tailed).}

Table 11

Result of the Moderation Effect of Expected Waiting Time

Coefficient SE t Sig. Constant -4.05 2.98 -1.36 .17 Fairness consideration 4.76 .82 5.83 .00** Recipient 4.16 4.02 1.03 .30 ZF*ZR -.64 1.13 -.56 .57 Note: N=360

(33)

28

5. Discussion and conclusion

This chapter presents the most relevant findings of my study, together with theoretical and practical implications. Furthermore, limitations and a number of suggestions for further research will be given.

5.1 Discussion

In my experiment, participants were allowed to exchange positions by paying or making donations. The average value (M=3.35, SD= 1.16) of fairness consideration of all participants demonstrates that trading places is perceived as fair. This finding is not consistent with Oberholzer-Gee’s (2006) study, which suggests that people consider trading places as unfair. One possible explanation could be the difference in externalities. In Oberholzer-Gee’s experiment, participants were offered money to let strangers cut in line; this increases the waiting time for others, thus increases negative externality. However, in my study, the position exchange is between a person at the front of the queue with a person at the back. Therefore, no negative externality was created for others. Moreover, in treatment DA, DB, and DC, participants were told that the trade will be made through a donation to charity. Participants may feel fairer when they realize that their behaviors are helpful to social welfare. This might explain why the overall fairness consideration in my study is significantly higher than that of Oberholzer-Gee’s study. The result of manipulation analysis confirms such explanation. In my experiment, the fairness consideration in treatment DA, DB, and DC (M=3.57, SD=1.18) is higher than the fairness consideration in treatment MA, MB, and MC (M=3.03, SD=1.14). It indicates that participants considered it fairer to exchange places through donation than through direct trade. The other experimental manipulator – expected waiting time – shows that participants at back of the queue have less fairness consideration (M=3.02, SD=3.40) than those at the front (M=3.40, SD=3.02). However, as suggested by Rafaeli, Barron and Haber (2002), people with less expected waiting time tend to care less about whether the queuing is fair. In

(34)

29

this study, it is those with more expected waiting time that suggests trading places. They were expected to give monetary compensations or social donation. Either way, they were about to pay for a short waiting time. As shown in the one-way ANOVA analysis, WTP was significantly higher in treatment MA (M= 15.63, p<.05) than in treatment MC (M=7.02), and higher in treatment DA than in treatment DC, which suggests than people expect to receive more than pay. I found that WTP of participants at the back of the queue (M=7.59, SD=6.87) is significantly lower than those at the front (M= 17.76, SD=16.71). The WTP difference between people at the front and back is above 10 dollars, this may suggest that, if time can be viewed as a product, people have more desire to sell it than buy it.

As for the direct effect of fairness consideration on WTP, the relationship is significant. (β=.37, R2=.15, p<.01). A closer examination of the direct effect according to different position has also confirmed this relationship, as shown in Table 7. This finding supports my hypothesis 1. People with more fairness consideration have a higher WTP. This finding is consistent with the study by Kahneman, Knetsch and Thaler (1986). Fairness does constrain profit-seeking activity. When people consider a market exchange as unfair, they will have less desire for the product. Moreover, the regression results have illustrated that control variables, such as age and work experience, also have effects on WTP. More specifically, age has a positive effect on the relationship between fairness consideration and WTP (β=.33, R2=.02, p<.01) and work experience is negatively related to the direct relation (β=-.29, R2=.02, p<.01). This result suggests that young people have more desire for the trade in time, and people with long work experience (over 10 years) do not prefer such trade.

In terms of the moderating roles of recipient and expected waiting time, both effects were non-significant (p>.05). The result suggests that the direct effect of fairness consideration on WTP is not moderated by recipient nor by expected waiting time. However, as shown in Table 4 and Table 5, fairness consideration and WTP is significantly different by recipient and the expected waiting time. One possible explanation could be that recipient and expected

(35)

30 waiting time can act more efficiently as factors that affect WTP rather than moderators.

Overall, as the queue experiment demonstrates, most individuals do indeed consider it fair to trade time. The data provides an interesting result demonstrating that when payment of trade goes to the charitable organization, participants exhibit higher fairness consideration and WTP. This provides evidence that individuals are ready to make sacrifices when a market exchange is considered socially beneficial (Charness & Rabin, 2002).

5.2 Theoretical and practical implications

My study contributes to the current literature of fairness consideration and WTP in the following ways. First, it confirms the theory that fairness consideration affects decision-making which is part of a growing study (Bolton and Ockenfels 2000, Charness and Robin 2002, Fehr and Schimidt 1999). People’s desire for a market exchange is affected not only by market norms, but also by social norms. In the market for time, fairness consideration could play an essential role in purchasing decisions. Because time is hardly measured by monetary items, the central norm which dominates people’s purchasing desire would be social norms, not the market exchange norm.

Second, my study links the research gap between fairness norms and willingness-to-pay. More specifically, fairness consideration has a positive relationship with willingness-to-pay. In my experiment, the results have confirmed the expected direct effect of fairness consideration. Furthermore, two demographic variables, age and work experience, have demonstrated direct effects on WTP. People at an older age tend to have higher WTP and people with longer work experience tend to have less. The direct relationship between fairness consideration and WTP suggests that fairness norms can be used as a good predictor of willingness-to-pay. In the study by Oberholzer, the author explained why the market for time cannot exist. He suggested that fairness consideration constrains people’s willingness to trade. However, my study shows that fairness consideration can hardly be the reason, or least be the major reason, to explain why

(36)

31 people do not trade places often.

In terms of practical implications, my study made the follow contributions. With respect to customer waiting time management, my study provides evidence that it is possible to adjust queues from the internal perspective. By letting customers allocate places voluntarily, service providers can make the queues more efficient. As a result, an after-trade queue can better reflect customers’ real opportunity costs. When customers are allocated to positions consistent with their opportunity costs, they will provide better evaluation of the service received.

5.3 Limitations

The result of my study is subject to a number of limitations. First, the experiment was performed through an online surveying. Because of that, I could not observe and control participants freely, and it cannot be guaranteed that there was no self-enhancement or bias in the data. Second, the market for time I created in the experiment is very limited. Factors such as the actual waiting environment and purchasing price were not taken into consideration. Therefore, it can be hardly stated that the experiment condition is practical. More important, the difference of individuals’ hurry was not under considered in my study. It makes the fairness considerations participants included unlikely to reach real social consideration. Lastly, the expected waiting time and recipient did not have the expected moderating effects. One reason could be that the expected waiting time and recipient should be considered as other factors that affects willingness-to-pay rather than moderators.

(37)

32

5.4 Further research

The findings of this study suggest a number of interesting directions for further research. First, the market for time is complicated and consists of a number of constraints. Future researchers could study the limits of time market such as transaction cost and externality. Second, the perception of fairness consideration affects people’s decision-making in various ways; an interesting study area in this regard is which factors could influence fairness consideration according to different market environments. In terms of willingness to pay in the time market, my study is a simple investigation. Future studies could pay more attention to the measurement of willingness to pay in the time market. Last, my study suggests some demographic factors, such as age and work experience, could affect the perception of fairness considerations and willingness to pay. More work could have been done on demographic factors that predict fairness perception and willingness to pay.

(38)

33

6. References

Anat Rafaeli, Greg Barron and Keren Haber (2002). The Effects of Queue Structure on Attitudes. Journal of Service Research, Volume 5, No.2

Anat Rafaeli, Efrat Kedmi, Dana Vashdi and Greg Barron (2003). Queues and Fairness: A Multiple Study Experimental Investigation. Faculty of Industrial Engineering and Management Technion-Israel Institute of Technology

Andreoni, James and John Miller (2002). Giving According toGARP: An Experimental Test of the Consistency of Preferences for Altruis. Econometrica. 70(2): 737–753.

Bin Li and Ozgua (2013). Exploring the tradeoff between waiting time and service cost in non-asymptotic operating regimes. IEEE Conference Publications

Bitner, M. J. (1990). Evaluating service encounters: The effect of physical surroundings and employee responses. Journal of Marketing, 54, 69-82.

Chebat, J., & Filiatrault, P. (1993). The impact of waiting in line on consumers. International Journal of Bank Marketing, I 35-40.

Carmon, Ziv, J. George Shanthikumar, and Tali F. Carmon (1995). A Psychological Perspective on Service Segmentation Models: The Significance of Accounting for Consumers’

Perceptions of Waiting and Service. Management Science 41 (11), 1806–15.

Charness, Gary and Brit Grosskopf (2001). Relative Payoffs and Happiness: An Experimental Study. Journal of Economic Behavior & Organization. 45(3): 301–328.

Charness, Gary and Matthew Rabin (2002). Understanding Social Preferences with Simple Tests. The Quarterly Journal of Economics. 117(3): 817–869.

Daniel Kahneman, Jack L. Knetsch and Richard Thaler (1986). Fairness as a constraint on profit seeking: entitlements in the market. The American Economic Review. Vol. 76, No.4 (Sep. 1986). Pp. 728-741

Elizabeth Gelfand Miller, Barbara E. Kahn and Mary Frances Luce (2007). Consumer Wait Management Strategies for Negative Service Events: A Coping Approach. Journal of Consumer Research. Vol. 126, No.9 (Aug. 2007). Pp. 163-181

Frey, Bruno S. and Werner W. Pommerehne (1993). On the Fairness of Pricing – an Empirical Survey among the General Population. Journal of Economic Behavior and Organization. 20: 295–307

Felix Oberholzer-Gee (2006). A Market for Time Fairness and Efficiency in Waiting Lines. The American Economic Review. Vol. 59 – 2006 – No. 3, 427–440

Fiske, A. P. (1992). The four elementary forms of sociality: Framework for a unified theory of social relations. Psych. Rev. 99 689-723.

(39)

34 Gad Allon & Eran Hanany (2012). Cutting in Line: Social Norms in Queues. Management

Science. Vol.50, No.3 March 2012, pp.493-503.

Gad Allon, Eran Hanany (2012), Cutting in Line: Social Norms in Queues, Management Science. Vol.186 Nov 2012, pp.126-143

Gerard Anderson, Charlyn Black, Elaine Dunn, Jordi Alonso, Jens Christian-Norregard, Tavs Folmer-Anderson, and Peter Bernth-Peterson (2015). Willingness To Pay To Shorten waiting time For Cataract Surgery, content.healthaffairs.org by Health Affairs on June 15, 2015

Huang-Yuan Lin and Ting-Yuen Chang (2011). The Customer’s Perspective On Waiting time in electronic Marketing, Social Behavior And Personality. Vol. 59 – 2006 – No. 3, 427–440 Hall, R.W. (1991). Queuing Methods for Service and Manufacturing. Englewood Cliffs, NJ:

Prentice Hall.

Hui, Michael K., Mrugank V. Thakor, and Ravi Gill (1998). The Effect of Delay Type and Service Stage on Consumers’ Re-actions to Waiting. Journal of Consumer Research, 24 (4), 469–79.

Jones, P. and Dent, M. (1993). Improving service: managing response time in hotel and restaurant operations. Operations Management Association. pp. 331-7.

James Knoow (2000). Fair Shares: Accountability and Cognitive Dissonance in Allocation Decisions. The American Economic Review. Vol. 59 – 2006 – No. 3, 427–440

Kahneman, Daniel, Jack L. Knetsch and Richard Thaler (1986). Fairness as a Constraint on Profit Seeking: Entitlements in the Market, American Economic Review. 76(September): 728–741.

Katz, K., Larsen, L., Blaire, M., & Larsen, R. C. (1991). Prescription for waiting-in-line blues: Entertain, enlighten, and engage. Sloan Management Review, 32, 44-53.

Leon Mann (1969). The Waiting Line as a Social System, American Journal of Sociology. Vol. 75. No3, pp. 340-354

Moshe Haviv, Ya'acov Ritov (1998). Externalities, Tangible Externalities, and Queue Disciplines. Management Science. Vol. 126, No.9 (Aug. 1998). Pp. 163-181

Michel Kostecki (1996). Waiting Lines as a Marketing Issue European Management Journal. Vol. 14, No. 3, pp.295-303.

Rongrong Zhou and Dilip Soman (2003). Exploring the Psychology of Queuing and the effect of the number of people behind, Journal of Consumer Research, Vol. 29, No. 4

Sebastian Hafenbrad, Ulrich Hoffrage, Chris M.White (2013). The impact of effect on willingness-to-pay and desired-set-size, Journal of Consumer Research, 24 (4), 469–79. Saaty, T.L (1961), Elements of Queuing Theory. New York: McGraw-Hill.

Schwartz, B. (1975). Queuing andWaiting: Studies in the Social Organization of Access and Delay. Management Science. Vol.186 Nov 1972, pp.126-143

(40)

35 Taylor, S. (1994). Waiting for service: The relationship between delays and evaluations of

service. Journal of Marketing, Vol 298 Apr 1994, pp. 56-69

Tom, G., & Lucey, S. (1995). Waiting time delays and customer satisfaction in supermarkets. Journal of Services Marketing, Vol. 76, No.4 Aug 1995. Pp 283-296.

Praveen Kulshreshtha (2015), Rationing by Waiting, Opportunity Costs of Waiting and Bribery, Indian Economic Review, New Series, Vol. 38, No. 1 (January-June 2003), pp. 59-75 Yina Lu, Andrés Musalem, Marcelo Olivares and Ariel Schilkrut (2013), Measuring the Effect

(41)

36

Acknowledgement

Foremost, I would like to express my sincere gratitude to my supervisor Anouar El Haji for the continuous support of my master thesis, for his patience, motivation and immense knowledge. His guidance helped me in all the time of research and writing of this thesis.

Besides my supervisor, I would like to express my gratitude to all my teachers who educated me in the past study year.

Last not the least, I would like to thank my family, especially my grandfather, for supporting me spiritually throughout my life.

(42)

37

Appendix: Questionnaire

A Survey for Fairness Considerations in Time Market

Welcome

Welcome to this survey. I appreciate your effort in completing it. By participating in my survey, you will help me complete the research that I am currently conducting for my master’s thesis, the MSc programme at the University of Amsterdam (UvA).

Instruction

Please read the instructions and questions carefully. There are no correct or incorrect answers. The survey consists of two parts. In the first part you will be asked to answer questions concerning with given scenario. In the second part, you will be asked to provide some background information.

Duration

The duration of the survey is approximately 2-3 minutes.

Confidentiality

The collected information is for research use only. The published results of my research will not identify any individual, or the choice he or she has made in any way. Nor will I give the information to any third parties.

Questions

(43)

38

Treatment 1

In an airport, three people are standing in a queue waiting for check-in service, as shown in above figure. The person in the first position (A) needs to wait 5 minutes to check in, the person in the middle position(B) needs to wait 20 minutes, and the person in the last position(C) needs to wait 35 minutes.

Imaging that you are person A. If people A, B and C are allowed to voluntarily trade positions for money. For example, if C pays 5 dollars to A, then C goes to the first position, which means that A goes to the last position.

Question 1: To what extent do you consider that it is fair that C trade positions with you for money?  Very fair

 Unfair

 Neither unfair nor fair  Acceptable

 Completely fair

(44)

39

Treatment 2

Imaging that you are person B. If people A, B and C are allowed to voluntarily trade positions for money. For example, if C pays 5 dollars to A, then C goes to the first position, which means that A goes to the last position.

Question 1: To what extent do you consider that it is fair that C trade positions with A for money?  Very fair

 Unfair

 Completely fair

(45)

40

Treatment 3

Imaging that you are person C. If people A, B and C are allowed to voluntarily trade positions for money. For example, if C pays 5 dollars to A, then C goes to the first position, which means that A goes to the last position.

Question 1: To what extent do you consider that it is fair that you trade positions with A for money?  Very fair

 Unfair

 Completely fair

(46)

41

Treatment 4

Imaging that you are person A. If people A, B and C are allowed to voluntarily trade positions by donating money to a charity organization. For example, if C donates 5 dollars to A accepts the deal, then C goes to the first position which means that A goes to the last position.

Question 1: To what extent do you consider that it is fair that C trade positions with you by

donating money to a charity organization?

 Very fair  Unfair

 Completely fair

(47)

42

Treatment 5

Imaging that you are person B. If people A, B and C are allowed to voluntarily trade positions by donating money to a charity organization. For example, if C donates 5 dollars to A accepts the deal, then C goes to the first position which means that A goes to the last position.

Question 1: To what extent do you consider that it is fair that C trade positions with A by donating

money to a charity organization?

 Completely fair

(48)

43

Treatment 6

Imaging that you are person C. If people A, B and C are allowed to voluntarily trade positions by donating money to a charity organization. For example, if C donates 5 dollars to A accepts the deal, then C goes to the first position which means that A goes to the last position.

Question 1: To what extent do you consider that it is fair that you trade positions with A by

donating money to a charity organization?

 Completely fair

(49)

44

Background information

Part Ⅱ

Now, please provide with some background information

Question 1: What is your gender?  Male

 Female

Question 2: What is your current age?

Question 3: What is your highest completed education?  Middle school

 High school  Bachelor’s degree  Master’s degree

 Advanced graduate work or Ph. D  Not sure

Question 4: What is the length of your work experience in years?

Ending

This is the end of this survey, please click continue button to finish. Thank you for participating. Your input is greatly appreciated.

Best regards, Liang Rui

rui.liang@student.uva.nl

A market for time : fairness considerations in queues