Time value: differences in opportunity cost of time per service and situation

(1)

Time value: differences in opportunity

cost of time per service and situation

(2)

2

Time value: differences in opportunity

cost of time per service and situation

Arnoud van Huffelen s1508520

Siegfriedpad 6

3194VW Hoogvliet Rotterdam 06-48831384

a.w.van.huffelen@student.rug.nl

Faculty of Economics and Business – University of Groningen Department of Marketing

Master thesis MScBA Marketing Management & Marketing Research April 2012

(3)

3

Management summary

In today’s economy, people’s time is a scarce resource, maybe even more scarce than money. Therefore, questions about the value of time (or opportunity cost of time) are very relevant. The literature is divided on the issue whether the adage “time is money” is true, but the synthesis “time is worth money” appears to be acceptable, certainly in consumer decision making. The conventional notion is that the opportunity cost or value of time is always the wage rate. This thesis examines whether there are factors that influence the time value of one consumer over time, and thus whether time value may not be constant. The four factors considered in this thesis are: service complexity, service client presence, time pressure and social context.

Four time-saving services are used to test the influence of the two service factors: lawn mowing (simple, low extent of required client presence), tax return advice (complex, low extent of required client presence), home hair dresser (simple, high extent of required client presence) and career advice (complex, high extent of required client presence). It may be that consumers are willing to pay higher prices for some services than for others. When each service saves an equal amount of time but has a different willingness to pay (WTP), this means that consumers indirectly value their time differently, depending on which service is hired. This notion is tested with self-stated WTP and the Price Sensitivity Measurer method, using data from a questionnaire (106 respondents). Although there are some differences in WTP when the time saved is not taken into account, these differences disappear when time saved is corrected for. Therefore it is concluded that there is no difference in time value for services based on complexity and required client presence.

(4)

4 that time value would not be constant. The effect of time pressure is tested using observational data from 172 visitors in a French drugstore.

Examining time pressure in isolation, it appears that its effect on product price and A-brand purchase is not linear (higher time pressure means higher opportunity cost of time) but U-shaped: for both high and low time pressure, product prices and A-brand purchases are high, while these values are low for intermediate levels of time pressure. It may be that low time pressure (which means longer time in the store) means very long deliberation which can result in A-brand purchase. For the second part (intermediate through high time pressure) the hypothesized pattern remains. This U-shaped effect is incorporated in the models by adding a quadratic term. A model is estimated for each of the four product categories: Body & bath, Perfumes, Make-up and Luxury (expensive perfumes and body-care for men and women). In the sales model (with the dependent variable product price) of the Luxury category, time pressure has a significant effect. This means that, going from an intermediate level of time pressure, higher time pressure causes purchase of higher priced goods. Following the reasoning stated above, this means that in the Luxury category, higher time pressure causes a higher opportunity cost or value of time.

Presence and importance of social context are also examined. When a consumer is with another person, the value he places on time may be higher. This may be even more prominent when the other person is important, for example a superior instead of a colleague. This is tested in the same questionnaire as for the service variables. Respondents are given a text which describes a situation they are in (a service environment) and specifies a social context (no context, context or important context). In a restaurant situation respondents had to state a desired compensation for an inconvenience during the service process. Correcting for compensation tendency, consumers ask more money in compensation for a lost experience when with another person than alone. When consumers see the compensation as a compensation for the entire group, more money is asked when another person is present, so the time is considered more valuable in the presence of social context. There is not enough evidence to find an effect of importance of social context on the value of time.

(5)

5

Preface

Inspired by the book ‘Freakonomics’ by Steven D. Levitt and Stephen J. Dubner, I decided not to go into traditional marketing topics for my thesis, but instead choose a more unconventional topic. The field of human motivation and choice provided many interesting avenues for research. Time value proved to be a topic that is a perfect combination of my Bachelor Business Economics and Master Marketing: choices under scarcity (economics) that lead to business outcomes for goods and service companies (business and marketing). In times where every consumer feels time pressure, the subject of time value and research in this field are very relevant. Diving into previous research, I found that Gary S. Becker wrote one of the most influential articles in the field. That same Gary Becker is said by Steven Levitt to be the biggest influence for his research, completing the circle.

This thesis is the final chapter of my time studying in Groningen. Therefore I would like to thank some people who were involved in the process of writing this thesis, but also some people who helped me during my entire time in Groningen.

First of all I would like to thank my supervisor, Dr. Jaap Wieringa, for his excellent supervision, knowledge, and guidance in the tougher parts of the process. Second, I would like to thank my second supervisor Niels Holtrop for his comments in the final stage of the thesis process. Third, I would like to thank Dr. Julien Schmitt, who was willing to share his observational data collected in the drugstore. This data provided a rich source of information, without which I could not have completed this thesis. The discussion on the data also helped me organizing my analysis and report. Fourth, I would like to thank my sister Martine van Huffelen who proved to be the best editor one could possibly imagine. Not only her comments on concept versions of the thesis, but also the discussions on the subject of time value were of great help. I would also like to thank Jiska Rutteman for her comments on the literature review and Malou Nijboer for her marketing insights on the recommendations section.

For their unceasing support, in more than one way, I would like to thank my parents Henk and Ilse van Huffelen. Finally I would like to thank my girlfriend Marchien de Vries for all her patience and support.

(6)

6

Chapter 1 – Introduction

1.1 Motivation

“Unusual services that save you time” by Anthony (1988) describes companies that facilitate time saving. “These entrepreneurs take over routine chores so you can focus on more important things”. These services range from car care, pet sitters, handymen, moving services (to the point of folding clothes and putting them in closets), personal shoppers for clothing, but also for birthday gifts to even a company that picks up to do lists at the train station in the morning, completed when a customer returns from work.

An article in the Wall Street Journal (Spencer 2003) also addresses time-saving services and products. It discusses the value of time in the economy and the affordability of time-saving services, among others the aforementioned personal shoppers.

The reason that these services exist is the current life environment. In today’s economy, people’s time is a scarce resource, maybe even more scarce than money. With so many possible work and non-work activities, people do not have the time to do everything they want to do. As a consequence, people are looking for ways to save time. One way is to let other people do things, mostly routine activities or chores, but also complex activities such as accounting, for you. But how to decide whether it is worth to outsource the activities? For that, the value of the time saved has to be compared to the cost of paying for the service.

But, what is the value of time, or opportunity cost of time? It may be the wage rate, since an hour saved may be used to work for an hour, which is paid the hourly wage. This is an opportunity cost, because the alternative use of time of the activity is working. For time-saving services, someone is better off if a service is provided for a price lower than his or her wage that could be earned in the time that the service is provided. Spencer (2003) also states that more expensive services should only be used when earning a higher income. Combining the time saved and the cost per hour of the service gives the threshold income above which to use the service. For example, buying chopped garlic should be done from an annual income of $10,000, hiring a professional desk organizer should be done from an annual income of $570,000.

However, people pay very different prices for different services, so it may be that the opportunity cost of time differs per situation. This may also be true for situations in which there is time pressure. When time pressure is high, the value of time may be higher.

(9)

9 pay is explained to stress the importance of a benchmark to which purchases should be compared: is the purchase worth the hours of work? And is the time at work worth the money (thesimpledollar.com)? In an entrepreneurial context, the concept of opportunity cost of time is discussed. The saying “Don't pay someone $25 an hour to do $7 an hour work” (entrepreneurs.about.com) is used to indicate the opportunity cost of outsourcing: when a person’s wage is $25 an hour, he should not do a job that takes an hour and can be done by someone else for $7. The best alternative here is to work an hour, and hire a service employee to do the job. Opportunity cost of time is also discussed in the context of leisure: ‘the story of the riding lawnmower’ tells that someone considered buying a riding lawnmower but was unsure whether it was worth the expense (cashmoneylife.com). When he saw the time savings it would create he bought it and used the saved time to play outside with his children. In considering to use household services, again the wage and opportunity cost of time are considered (money.cnn.com).

The value of time is an important issue in social science research. There are many scientific publications on time value in (among others) economics, psychology and marketing journals. A large share of them relates time value to consumer decision making.

However, the role of time value is not totally clear yet. A call for more research in this area is already made in 1976, by Jacoby, Szybillo and Berning, who state in their review of time literature that time is only treated as a secondary variable, but should become the focus of research: “Perhaps (…) investigators will be stimulated to engage in much-needed conceptual and empirical considerations of the relationship between time and consumer behavior”. Moreover, in the service literature there is also a call to examine the structural positioning of (time-saving) services to increase the influence of marketing in the service sector (Shostack 1987). A more recent call comes from Berry, Seiders and Grewal (2002): “In service economies that include so many time- and energy impoverished consumers, learning more about service convenience should be a priority”.

(10)

10 may change willingness to pay. Second, under time pressure, consumers may value their time differently and therefore may change their product choice (e.g. make faster decisions, choose convenience products).

1.2 Research questions

The nature and role of opportunity cost of time are not clear yet. Therefore, the problem statement that will be discussed in this thesis is as follows:

“Are there intrapersonal differences in opportunity cost of time and what is the role of opportunity cost of time in consumer decision processes?”

The research questions that are used to answer the problem statement are:

1 - Does the opportunity cost of time differ per service?

1.1 - What are relevant characteristics of services concerning opportunity cost of time? 1.2 - What is the willingness to pay for several services based on the dimensions of 1.1?

2 - What is the influence of contextual variables on opportunity cost of time?

2.1 - Is there a difference in opportunity cost of time in situations with different time pressure? 2.2- Is there a difference in opportunity cost of time in situations with different social context?

While research question 1 mainly addresses the question of intrapersonal differences in opportunity cost of time, research question 2 addresses both the intrapersonal differences and the role of opportunity cost of time in consumer decision processes. The entire problem statement is therefore researched, and the topic of the value of time is discussed from different angles.

1.3 Position of this thesis in the literature

(11)

11 for one person in different situations or over time. Several literature streams, as well as theoretical and many empirical studies focus on this type of opportunity cost of time.

The literature on intrapersonal differences in opportunity cost of time is less developed. There are less major topics concerning this subject, and they mainly focus on the differences and similarities of time and money. The papers that are written on the subject are either theoretical, or (in the case of empirical articles) spend only a small part of the study on the value of time. This thesis is positioned in that under-researched area (table 1.1). It focuses solely on the value of time in a comprehensive way.

Table 1.1 will be discussed more elaborately in the closing section of chapter 2, after the relevant articles are discussed in the first part of chapter 2.

Differences in opportunity cost of time

Interpersonal Intrapersonal

Literature Literature Price discrimination Mental accounting

stream Yield management

Travel time

Theoretical Becker 1965 Howard and Sheth 1969

Cesario 1976

Okada and Hoch 2004

Empirical Prochaska and Schrimper 1973 Leclerc, Schmitt and Dubé 1995

Crafton 1979 Rajagopal and Rha 2009

Bellante and Foster 1984

Carlin and Sandy 1990 This thesis

Casey, Vukina and Danielson 1995 Table 1.1: Place of this thesis in the literature.

1.4 Structure of the report

(12)

12

Chapter 2 – Theory

2.1 Introduction to chapter 2

Chapter 2 will provide the theoretic framework for this thesis. Sections 2.2 through 2.4 discuss the basis of opportunity cost of time, its traditional determination and the problems surrounding this determination. Sections 2.5 and 2.6 provide challenges to the traditional determination, addressing some of the problems of section 2.4. In these sections the hypotheses that test the research questions are formulated. Section 2.7 presents the conceptual model which summarizes the hypotheses and sets the stage for the empirical part (chapters 3 and 4). The final section of this chapter discusses the contribution of this thesis to the literature.

2.2 Opportunity cost of time

2.2.1 Opportunity cost

Opportunity cost is the cost associated with the opportunities that are forgone by not using resources for an alternative action (Pindyck and Rubinfeld 2001, p. 204). The concept of opportunity cost dates back to at least the 19th century’s works of John Stuart Mill (Stigler 1955), who wrote that

“…when land capable of yielding rent in agriculture is applied to some other purpose, the rent which it would have yielded is an element of the cost of production of the commodity which it is employed to produce.”

(13)

13 2.2.2 Time is money

There are situations in which “time is money” is true. In deterministic situations, people make decisions about time loss in the same manner as decisions about money (Leclerc, Schmitt and Dubé 1995).

In the domain of money, people maintain mental accounts for different sources and uses of money and treat each account differently, which violates the notion of perfect substitutability or fungibility of money (e.g. money won in a lottery is likely to be spent very differently than money obtained from regular income, even if the amount and the timing of receipt from these two sources were identical). People create the same mental accounts for time, and thus treat time the same as money (Rajagopal and Rha 2009).

The same result is found in another study (DeVoe and Pfeffer 2007). Hourly payment leads individuals to mentally account for their time more like money.

2.2.3 Time is not money

There are, however, also situations in which “time is money” is not true. Although more money can be obtained to a certain extent (e.g. working overtime, having two jobs), time is finite per definition (Berry 1979). In other words, “even the wealthy are limited to 24 hours per day” (Leclerc, Schmitt and Dubé 1995).

Furthermore, decision making in the domain of time is not exactly the same as financial decision making, as Leclerc, Schmitt and Dubé (1995) find. These authors conclude that, under conditions of risk, consumers are significantly more risk averse for time decisions than for monetary decisions This is because time is less fungible, meaning less transferable or substitutable over time. Losses or savings in time cannot as easily be transferred (i.e. recouped or applied) as losses or savings in money (Leclerc, Schmitt and Dubé 1995), because time is harder to store and gain. Therefore planning is important, and uncertainty reduces the capability of consumers to plan. Consequently, consumers do not like uncertainty and are more risk averse for time decisions.

(14)

14 that they expended in the acquisition. This kind of flexible valuation is harder to do with money. Driving force between these differences is the ambiguousness of the concept of time value. The opportunity cost of time is harder to imagine than the opportunity cost of money: money is storable and liquid, while time is not as liquid and highly perishable (Okada and Hoch 2004). Thus, while money can be saved perfectly and used later on, saving and storing time is harder or impossible. Moreover, consumers have less experience with exchanging time than with exchanging money (Okada and Hoch 2004). The smaller amount of knowledge about the opportunity cost of time leads to a less clear image of its value.

Saini and Monga (2008) also find a difference in decisions concerning time and money. For decisions regarding time, consumers use more heuristics than for decisions regarding money. The authors attribute this finding, as Okada and Hoch (2004) do, to the psychological difference between time and money: time is harder to account for. When participants are primed to account for their expenditures, there is no longer a difference in the use of heuristics for decisions regarding time and money.

2.2.4 Time is worth money

For decision making, time is only partly similar to money. Thus, time and money may not be dealt with similarly from a decision making standpoint (Rajagopal and Rha 2009). But this does not mean that time has no value. Many people feel pressed for time (DeVoe and Pfeffer 2007), and for many time is not just a scarce resource, it is the scarce resource (Leclerc, Schmitt and Dubé 1995). As society gets more affluent and products and services become less scarce, time becomes increasingly important to people because “it is scarcity which creates value”. Thus, people attempt to balance their time across work and non-work activities (Rajagopal and Rha 2009). They trade off between working time and leisure time and want to spend less time on work (DeVoe and Pfeffer 2007). There is even a tradeoff between several activities inside leisure time (Casey, Vukina and Danielson 1995). So, time does have value. This has been recognized in the domain of goods: convenience goods (e.g. frozen dinners) are goods that save people time (Berry, Seiders and Grewal 2002). Consumers desire products that conserve time (Jacoby, Szybillo and Berning 1976), and generally pay a premium for convenience (Okada and Hoch 2004), so they ‘buy’ time with money.

(15)

15

2.3 Determining the value of the opportunity cost of time

2.3.1 Wage rate as opportunity cost of time

An economic rationale for the inclusion of time in the opportunity cost of an action is provided by Stigler (1961). Stigler reasoned that if nonprice factors were not relevant to the amount of search, a complete search would be conducted and the lowest price would be found by everyone. As a result, there would soon be no price dispersion. Since this is not the case, it becomes imperative to include the time factor in any such analysis. When a value is placed on time, a consumer will continue to search until the expected savings from the search are less than the cost of time (Jacoby, Szybillo and Berning 1976).

But the question remains what the cost or value of time is. Time has no explicit market value (Cesario 1976). A central work in the early stage of research on this subject is the article of Becker (1965). He studies the allocation of time to various activities using a rational-choice paradigm, which assumes rational behaviors (i.e. utility maximization) on the part of time consumers (Rajagopal and Rha 2009). He states that:

“Most economists have now fully grasped the importance of forgone earnings […] and criticise educationalists and others for neglecting them. In the light of this it is perhaps surprising that economists have not been equally sophisticated about other non-working uses of time. For example, the cost of a service like the theatre or a good like meat is generally simply said to equal their market prices, yet everyone would agree that the theatre and even dining take time, […] time that often could have been used productively. If so, the full costs of these activities would equal the sum of market prices and the forgone value of the time used up.”

Thus, the opportunity cost of time should be included in the full cost of an action. This is supported by the finding that estimates of the income elasticity of demand for different goods are biased when the cost of time is ignored (Becker 1965). For instance, the income elasticity of demand for goods (e.g. going to the theatre, as in the quote above) is positive, but when time costs are included the price of the good is higher, so the additional demand will be lower (meaning a lower income elasticity).

(16)

16 lead to this valuation of time: the assumption that time is fungible (transferable) and the assumption that people optimize their utility with time. This is still an economic approach to the opportunity cost of time, and it is also a normative approach which focuses on how people ought to behave rather than how people actually behave (Rajagopal and Rha 2009). For the opportunity cost of time to be equal to the wage rate, two conditions must hold: i) time is transferable, i.e. consumers can flexibly exchange work hours (for which they are paid per hour) for leisure hours and vice versa and ii) the opportunity cost of time is fixed per person. An example of the use of condition ii) can be found in Carlin and Sandy (1990). They state that parents with high wage rates are less likely to properly use car seats for their children because their time is more valuable. Implicitly they assume that time is always worth the same.

While the wage rate is a reasonable starting point for the value of time, there are serious problems with both conditions under which it should hold. These are addressed in section 2.4.

2.3.2 The use of travel time to estimate opportunity cost of time

An alternative approach to estimate the value of time is used in recreation benefit studies. To estimate the primary benefits of outdoor recreation sites (which are public, i.e. not marketed) are calculated with a proxy: the cost of travel with the inclusion of time costs (Cesario 1976). Because recreation takes place in the “nonwork” time, the consumer is not trading off work time to nonwork time, so the wage rate is an inappropriate estimate of time value. Rather, consumers are trading off one type of nonwork time to another, thus, the opportunity cost of time is estimated by the amount one is willing to pay to save time spent travelling (Cesario 1976).

A similar notion is raised by Crafton (1979). He also includes time costs in the full cost of a product. Hence, when two stores are located at a similar distance from home and have the same prices, the one with the quicker checkout should be preferred. Conversely, a store closer by can ask a higher price, because of the time gain.

2.4 Problems with the traditional determination of the opportunity cost of time

2.4.1 Problems with wage rate as opportunity cost of time

(17)

17 prompted (Hoskin 1983, Okada and Hoch 2004). Indeed, several studies report that the value of time (with respect to nonwork travel/commuting) is between 25% and 50% of the wage rate (Cesario 1976). Thus, the use of wage rate as a measure for opportunity cost of time is not necessarily appropriate. An explanation for this is, as is stated above, time is less fungible (transferable) than money, it has no market value, and that the concept of time is ambiguous (Okada and Hoch 2004).

Second, there are two specific problems with wage as the opportunity cost of time. These relate to the conditions mentioned above. The problem with condition i) (time is transferable) is that many people can not easily exchange work hours for leisure, because they have a fixed amount of hours in their contract. The problem with condition ii) (the opportunity cost of time is fixed) is that the opportunity cost of time may vary widely under different circumstances. This is discussed in more detail in section 2.4.2.

2.4.2 Problems with fixed opportunity cost of time

As already mentioned, people create mental accounts for time, and for each account the value of time differs (Rajagopal and Rha 2009). For example, a delay at the airport is imagined to be more upsetting for a business trip than for a vacation. Thus, the value of time is different for work and non-work situations.

This notion is supported by Okada and Hoch (2004), who state that the opportunity cost of time is flexible, and dependent on context. The valuation placed on travel time is highly subjective, varying from individual to individual and from situation to situation (Cesario 1976). Situational variables in the travel time valuation study are for example the purpose of the trip, trip length, time of day and other variables.

So, context has an influence on the value of time. Contextual variables that are likely to have an influence on the value of time are time pressure and social context. Two contextual variables are discussed: time pressure and social context (section 2.6). The difference of time value may also be visible in the price people are willing to pay for different types of services. This will be discussed in the next section (2.5).

2.5 Services and the opportunity cost of time

2.5.1 Time-saving services

(18)

18 cooking, etc.), clothing care (drycleaning and laundry sent out), personal care (barbershop and beauty parlor services) (Bellante and Foster 1984), car care, pet sitters, handymen, moving services (Anthony 1988) and legal services (Dubé-Rioux, Regan and Schmitt 1990).

As stated before, a service should be purchased when the opportunity cost of time is lower than the price of the service. A higher opportunity cost of time would then lead to a higher usage of time-saving services. This is confirmed in the literature: opportunity cost of time has a positive effect on out of home food consumption (Prochaska and Schrimper 1973) and the extent to which wives are working positively influences usage of time-saving services (Bellante and Foster 1984).

The ‘do or let do’ decision is analogous to the ‘make or buy’ decision made in outsourcing. Many business processes are outsourced, i.e. handed over to another company (‘external service provider’). An example of this is human resource management outsourcing (Greer, Youngblood and Gray 1999). A major reason for outsourcing in this sector (but also generally) is money savings. Money can be saved because the external service provider has expertise (lower costs) and can therefore ask a lower price. Money savings are important in all economic circumstances, but are deemed important especially in bad economic times. Saving time is another reason for outsourcing (Payroll Manager’s Report 2007). When a business process is outsourced, employees save time which can be used for more important or value-adding processes, thus focusing on the core competencies of the company. This is especially important in good economic times, because the work load is higher, and thus there is more time pressure. Time is too precious to spend it on low-value activities, so more services (outsourcing) are used. Time pressure is discussed in more detail below.

Thus, consumers trade off money for the services, which means that they trade off money for time. Via this tradeoff the value of time can be derived: time is saved, and money is spent. It must be noted, however that the time saved for the consumer of the service is not necessarily equal to the time that is spent (and billed) by the service provider. It is likely that the service provider has developed expertise, and thus spends less time in the process.

2.5.2 Opportunity cost of time and services

(19)

19

characteristics: intangibility, heterogeneity, inseparability, and perishability (Lovelock and Gummesson 2004). Davis, Guiltinan and Jones (1979) discuss the same four characteristics and add that there is no transfer of ownership in service transactions, and that marketing is different for services.

After this phase of research in the difference between goods and services, scholars proposed categorizations of services (Li, Yang and Wu 2009). An influential article in this field is that of Lovelock (1983), who summarizes previously proposed schemes for classifying services and adds a new set of characteristics to describe services. Each characteristic is examined on two dimensions, because combining classification schemes in a matrix yields better marketing insights than classifying service organizations on one variable at a time (Lovelock 1983). The resulting five characteristics are the nature of the service act (directed to a person or possession, tangibility of the act), relationship with a customer (transactional or continuous, membership), degree of customization (customization of the service, employee judgment/latitude), demand and supply (demand fluctuations, supply constraints), and service delivery (amount of service outlets, electronic communication or customer goes to service provider or vice versa).

Lovelock (1983) also summarizes previously proposed schemes for classifying services. Four major characteristics appear in the list of items from previous articles: the object to which the service is provided (rented good/owned good/nongood or person), the portion of physical and intangible elements, the supply method (equipment/people based, and for people based the required skill level of the employee) and the extent of required client presence.

Dubé-Rioux, Regan and Schmitt (1990) distinguish concreteness, with for example car repair being concrete and legal services being abstract, and specificity, with car repair being generic and Fix-it Muffler/Brake Shop being specific. Shostack (1987) distinguishes complexity, the number and intricacy of steps required to perform the service, and divergence, the executional latitude or variability in the performance. In this sense, concreteness and complexity are rather similar constructs, because more abstract services generally require more steps to be performed. Moreover, skill level (from Lovelock’s (1983) supply method) is likely to be correlated with concreteness and complexity, because more difficult service processes generally require more highly educated employees.

(20)

20 actual service benefits). Shostack (1987) states that barber shops changed their positioning from merely “hair cutting” to “hair styling”, by adding beauty salon elements. In this way, they tapped or created a new market segment of men willing to pay substantially higher prices for a more elaborative process. Partly due to additional services, and partly due to repositioning, a higher price for the service could be asked.

While practices such as price differentiation and yield management acknowledge that there are interpersonal differences in opportunity cost of time, they do not account for intrapersonal differences.

Different prices that are paid due to different characteristics or positioning of services may point to the existence of differences in opportunity cost of time. This offers opportunities for positioning (Shostack 1987). In conclusion, service characteristics are likely to influence willingness to pay and opportunity cost of time.

2.5.3 Complexity

Two characteristics are particularly relevant regarding their influence on opportunity cost of time. The first is complexity (in the definition of Shostack (1987)): consumers are likely to pay more for a more complex service. As pointed out by Shostack (1987), this may be solely due to the presence of that characteristic, and not the added benefits. Hence, the presence of the characteristic enhances willingness to pay and opportunity cost of time. This is, however, not tested empirically. To test the effect of complexity on opportunity cost of time, the following hypothesis is formulated:

H1: Degree of complexity has a positive effect on the opportunity cost of time.

As stated above, complexity is similar to the concreteness characteristic of Dubé-Rioux, Regan and Schmitt (1990), and the skill level of Lovelock (1983), and is therefore likely to capture some of the variation of these characteristics.

2.5.4 Extent of required client presence in the service process

(21)

21 time is saved. Therefore, time in the service process is analogous to wait time. This wait time is a waste of time, which is a loss of time that could be used more productively (Leclerc, Schmitt and Dubé 1995), and is thus an opportunity cost. In determining how much time is saved by the service, the time present in the service process would then be discounted implicitly by consumers. Therefore, when a consumer does not need to be present when the service is performed, this consumer saves more time. This time has value, because it can be used for other purposes.

When the effect of saved time is taken into account, the presence or absence of the consumer in the service process is likely to still have an effect, whether through miscalculation of the time spent in the service process or through the mere need to be present during the service. Therefore, it is likely that the willingness to pay and opportunity cost of time are higher when a consumer does not need to be in the service process. To test this, the following hypothesis is formulated:

H2: Extent of required client presence has a negative effect on the opportunity cost of time.

There are some characteristics that are similar to presence in the service process. First, Davis, Guiltinan and Jones (1979) discuss delivery characteristic: impersonal or personal delivery. Second, the object to which the service is provided, from Lovelock (1983), is also similar. Goods are usually serviced without the consumer in the process (e.g. car repair), while nongoods services usually require presence of the consumer (e.g. health services or leisure facilities). Thus, ‘extent of required client presence’ covers a broad range of similar variables (in the same way that complexity covers a range of similar variables).

By assessing two service characteristics, the finding of Lovelock (1983), that combining classification schemes in a matrix yields better marketing insights than classifying service organizations on one variable at a time, is applied.

2.6 Contextual variables and the opportunity cost of time

2.6.1 Time pressure and consumer behavior

(22)

22 of consumer behaviors. The decision where to shop is influenced by time pressure: consumers combine shopping purposes and locations to save time, termed multipurpose shopping, multistop shopping or one stop shopping (Dellaert et al. 1998, Jacoby, Szybillo and Berning 1976). Moreover, consumers may decide to do their shopping at home, via telephone (Jacoby, Szybillo and Berning 1976) or internet (Verhoef and Langerak 2001).

The purchasing process itself is also influenced by time pressure, which is supported by a large body of literature. Information search is influenced negatively by time pressure almost by definition, since only a finite amount can be processed in a given time period (Iyer 1989). This leads to lower exposure to the offering, lower recognition of needs and lower retrieval of information from memory, and thus to higher failure to make intended purchases and lower unplanned buying (Iyer 1989, Park, Iyer and Smith 1989).

Howard and Sheth (1969) also find that time pressure influences the value of time. Time pressure is an exogenous variable and can be different for each decision situation. When a buyer feels pressed for time, because of any of several environmental influences, he must allocate his time among alternative uses. In this process a reallocation unfavorable to purchasing activity can occur. Time pressure will create inhibition and will also unfavorably affect the search for information (Howard and Sheth 1969). Consumers under time pressure examine information faster, focus on more important and negative attributes, and use simpler (noncompensatory) decision rules (Dhar and Nowlis 1999, Nowlis 1995). Thus, an increase in time constraints from low time pressure results in a reduction in the extent of systematic information processing (Suri and Monroe 2003).

Consumers under time pressure are less likely to adopt a new brand. The consumer is less likely to learn about new product offerings and less likely to try one even if he does learn about it (Berry, Seiders and Grewal 2002, Jacoby, Szybillo and Berning 1976). Stated differently, brand loyalty is higher under time pressure (Jacoby, Szybillo and Berning 1976, Roselius 1971). Thus, time pressure leads to more routine, habitual behavior: less information processing, lower adoption of new brands and higher repeat purchase.

(23)

23 2.6.2 Rationale for differences in opportunity cost of time

Although the literature above explains why consumer behavior changes (namely, time pressure), it does not explain why time pressure has an effect. It is likely that this process is mediated by the value of time: time pressure makes the (little) time more valuable, and therefore consumers cannot afford to spend much of it, so they want to ‘save time’, and thus change their behavior. Time saving occurs in the form of reduction in shopping time, information processing, risk, and an increase in routine, habitual behavior (see above). The amount of time pressure is mostly determined before the shopping trip, so the amount of time saving a consumer applies is mostly the same within one shopping trip but can be different for another shopping trip. The presence of alternative consumer behavior during different shopping trips therefore points to a change in the opportunity cost of time.

2.6.3 Trading off time

The subject of time pressure and increased purchase of high quality brands or national brands is particularly interesting. There is a tradeoff between time and information search (Berry, Seiders and Grewal 2002), and therefore a tradeoff between time and decision quality. As mentioned above: with unlimited time, every consumer reaches the optimal decision (Stigler 1961). However, under time pressure, the purchase decision is suboptimal (Nowlis 1995). When there are only two alternatives, one with high price and high quality, one with low price and low quality, consumers may overpay when they buy the expensive alternative, or may purchase an unsatisfactory product. The latter alternative is unlikely, because consumers under time pressure have a tendency to buy high quality or major, well-known brands (Nowlis 1995, Roselius 1971).

Thus, consumers are likely to overpay when they feel pressed for time. Stated differently, to reduce time spent in the decision making process, consumers accept a suboptimal decision, and willingly incur extra costs (this is also true if there is a predisposition for low quality products, in that case costs are the above mentioned time, convenience and effort wasted to adjust, repair or replace a failed product). The extra costs that consumers will trade for time point to an increase in the value of time. Thus, time pressure is likely to increase the opportunity cost of time.

(24)

24 These emotions do not exist when a consumer is not pressed for time, i.e. does not place much value on his time. To test the effect of time pressure on the opportunity cost of time, the following hypothesis is formulated:

H3: Time pressure has a positive effect on the opportunity cost of time.

2.6.4 Social context

After the first contextual variable, time pressure, the second contextual variable, social context, is discussed. Social context is proven to have an effect on consumer behavior. The presence of others increases embarrassment of consumers while buying embarrassing products such as condoms (Dahl, Manchanda and Argo 2001). In a group context wherein group members voice their choices sequentially, there is more variety seeking than when consumers make decisions individually (Ariely and Levav 2000). Puntoni and Tavassoli (2005) indicate that there are three levels of social context: interaction with others, presence of others (but no interaction), and imagined social context. While most of this research treats social context as a dichotomous variable (absence or presence of others), Luo (2005) specifies presence of others, in the context of impulse purchasing. Two groups are discussed, peers and family, and they have different effect on consumer behavior: the presence of peers increases the urge to purchase, and the presence of family members decreases it (Luo 2005). Thus, the type of social context is also important.

Wakefield and Inman (2003) find that consumers are less price sensitive in a social context. This is attributed to self-presentation, the process of conveying an image of themselves to others (Ariely and Levav 2000), of consumers. However, this is tested in the (food) product domain, and may not be dominant for (time-saving) services.

Because the amount of social context has a negative influence on price sensitivity, consumers are expected to spend more in the presence of others, which may indicate a higher opportunity cost of time. Therefore:

H4a: The presence of social context has a positive effect on opportunity cost of time.

(25)

25 time and money than when the same product is to be used by the buyer (Belk 1982). This again can be extended by specifying the ‘other’ in degrees of importance. Some people or relationships are more important for a person than others (e.g., Zimmer-Gembeck 1999). While purchases for others may take more expenditures, even more expenditures may be taken when the other is more important. This may indicate a higher opportunity cost of time when the other is more important. A more intuitive reasoning may be that the more important the other is with whom time is spent, the more value that time has. These findings lead to the following hypothesis:

H4b: The importance of social context has a positive effect on opportunity cost of time.

2.7 Conceptual model

The research questions from section 1.2 and the literature from the previous sections lead to the conceptual model in figure 2.1. The number of the hypotheses that are outlined in sections 2.2 through 2.6 are given next to the arrows of the effects, alongside the expected sign of the effect.

The service characteristics from section 2.5 are on the left side. Complexity is expected to have a positive influence on opportunity cost of time, while client presence is expected to have a negative influence on the focal variable. The contextual variables time pressure and social context (from section 2.6) are on the right side. All the contextual variables are expected to have a positive influence on the opportunity cost of time. If one (or more) of the hypotheses can be confirmed, there is proof against the traditional notion of a fixed opportunity cost of time.

2.8 Contribution to literature

Most research only recognizes interpersonal differences in opportunity cost of time (table 2.1). Price discrimination and yield management claim that different prices can be asked to different consumers, because they have different price sensitivities and values of time. Becker (1965) argues that the value of time is equal to the wage rate. Since income differs per consumer, the value of time can differ per consumer. The travel time literature incorporates time (as the wage rate) in the total cost of a product.

(26)

(Prochaska and Schrimper 1973), working wives and usage of time

seat usage (Carlin and Sandy 1990) and l

As stated before, this literature does not recognize that the opportunity cost of time can be different from moment to moment

incorporates intrapersonal differences in opportunity cost of time is smaller than the ‘interpersonal differences’ stream (table 2.1). Although some articles state that

time varies per situation (Howard and Sheth 1969, Cesario 1976, Okada and Hoch 2004) only state possible situations, but do not test them empirically.

The only literature stream that researches intrapersonal differences of time is that of mental accounting. However, these articles mainly focus on the differences and similarities of time and money. For example, Leclerc, Schmitt and Dubé (1995) execute eight stud

only two are solely about time value. Furthermore, these authors only manipulate the time and money variables itself: the value of a travel time reduction for a short trip vs. a long trip, and the value of wait time for a cheap ticket vs. an

the decision are manipulated. Rajagopal and R

differences, and the one experiment that is solely about time has a small sample. In Figure 2.1: Conceptual model.

Prochaska and Schrimper 1973), convenience store pricing or travel time ( working wives and usage of time-saving services (Bellante and Foster 1984), c

Carlin and Sandy 1990) and leisure activities (Casey, Vukina and Danielson 1995). fore, this literature does not recognize that the opportunity cost of time can be different from moment to moment for the same person. The body of literature which incorporates intrapersonal differences in opportunity cost of time is smaller than the

rpersonal differences’ stream (table 2.1). Although some articles state that

(Howard and Sheth 1969, Cesario 1976, Okada and Hoch 2004) only state possible situations, but do not test them empirically.

y literature stream that researches intrapersonal differences of time is that of mental accounting. However, these articles mainly focus on the differences and similarities of time and money. For example, Leclerc, Schmitt and Dubé (1995) execute eight stud

only two are solely about time value. Furthermore, these authors only manipulate the time and : the value of a travel time reduction for a short trip vs. a long trip, and the value of wait time for a cheap ticket vs. an expensive ticket. Thus, no variables external to the decision are manipulated. Rajagopal and Rha (2009) also focus on time and money differences, and the one experiment that is solely about time has a small sample. In

26 convenience store pricing or travel time (Crafton 1979), Bellante and Foster 1984), child car safety Casey, Vukina and Danielson 1995). fore, this literature does not recognize that the opportunity cost of time can be

. The body of literature which incorporates intrapersonal differences in opportunity cost of time is smaller than the rpersonal differences’ stream (table 2.1). Although some articles state that the value of (Howard and Sheth 1969, Cesario 1976, Okada and Hoch 2004), they

(27)

27 conclusion, these articles only touch upon the value of time, and do not provide a comprehensive overview of it.

This thesis is positioned in that under-researched area (table 2.1). It focuses solely on the value of time in a comprehensive way. The subject is studied from different theoretical angles, for both services and products, and with the use of various different research methods. It will try to falsify the traditional notion that the value of time is always equal to the wage rate, by studying consumer decisions when purchasing time-saving services, when under time pressure, and with various social contexts.

Differences in opportunity cost of time

Interpersonal Intrapersonal

Literature Literature Price discrimination Mental accounting

stream Yield management

Travel time

Theoretical Becker 1965 Howard and Sheth 1969

Cesario 1976

Okada and Hoch 2004

Empirical Prochaska and Schrimper 1973 Leclerc, Schmitt and Dubé 1995

Crafton 1979 Rajagopal and Rha 2009

Bellante and Foster 1984

Carlin and Sandy 1990 This thesis

(28)

28

Chapter 3 – Research design

3.1 Introduction to chapter 3

The hypotheses of chapter 2 will be tested with the analyses outlined in this chapter. The analyses are organized in three themes. The first set of analyses will test hypothesis 1 and 2 with different price measurement methods for several services. The second set of analyses concern time pressure. They are used to test hypothesis 3, with purchase models. The third and final set of analyses is in the area of social context. They test hypotheses 4a and 4b by the outcomes of consumer decisions in imagined situations.

3.2 Theme 1: Willingness to pay for services using open-ended contingent valuation

and price measurer.

In the first set of analyses, four services based on two characteristics will be tested with two price measurement methods. The service characteristics are, as stated above, complexity (Shostack 1987) and the extent of required client presence (Lovelock 1983). This 2x2 matrix gives: lawn mowing (simple, low extent of required client presence), tax return advice (complex, low extent of required client presence), home hair dresser (simple, high extent of required client presence) and career advice (complex, high extent of required client presence), see table 3.1.

Complexity

Simple Complex

Extent of required client presence Low Lawn mowing (2.0, 1.5) Tax return advice (4.6, 2.7)

High Home hair dresser (2.6, 7.0)

Career advice (4.9, 6.6)

Table 3.1: Selected services positioned on the two characteristics, with mean scores (complexity and required presence, respectively) in parentheses.

This selection of services is arrived at via a pretest. The pretest was held among a convenience sample of 21 respondents. For a more detailed description of the pretest, see appendix 1.

(29)

29 should be high. For example, the low complexity – high client presence service should score closest to 1 for complexity and 7 for client presence (deviations for practical reasons are discussed below). Their mean scores (on a 7-point scale from low to high) for complexity and required client presence are reported in table 3.1. Only the mean scores for tax return advice are less than optimal (which would be 7, 1), so the introductory text of this service is slightly adapted in the final version of the questionnaire to better represent the focal characteristics. Career advice is chosen above mortgage advice, because consumers may consider a mortgage advice as a free service, and the tradeoff in values on the two dimensions is about even (0.4 points more on required presence and the same amount less on complexity).

There are several methods to measure consumers’ willingness to pay. De Pelsmacker and Van Kenhove (2006, p. 391) discuss a series of methods in the context of new product development: buying intentions, Price Sensitivity Meter, Gabor Granger (monadic, noncomparative), Brand Price Trade Off and conjoint analysis (comparative). In this set of analyses, a monadic method will be used.

For theme 1, more than one method will be use to increase validity. The first method is the open-ended contingent valuation (self-stated reservation price) method, in which respondents state their WTP for a product directly (Backhaus et al. 2005, Kalish and Nelson 1991). The open-ended method is preferred to the closed-ended method (in which respondents indicate intention to buy for a given price level) because closed-ended contingent valuation approaches are not always statistically efficient or robust and the open-ended data-collection process is more time and cost efficient because there is only one question (Backhaus et al. 2005). Moreover, the closed-ended method would be too similar to the second method. The second price measurement method is the Price Sensitivity Meter (PSM) of van Westendorp (De Pelsmacker and Van Kenhove 2006, pp. 394-396, Breidert, Hahsler and Reutterer 2006). In this method, the respondents have to indicate per product the price levels of which they think the product is too cheap (to be confident in its quality), cheap, expensive and too expensive (to pay the price for the product). Plotting the cumulative percentages of the four categories (with the ‘too cheap’ and ‘cheap’ lines inverted) gives the intersections which mark the boundaries of the acceptable price range, or ‘normal’ price. With two lines sloping upwards and two sloping downwards, a diamond shape in the middle of the graph marks these boundaries. Four price points are distinguished (De Pelsmacker and Van Kenhove 2006, pp. 395-396):

(30)

30 2. Point of marginal expensiveness (intersection of cheap and too expensive): upper

boundary of acceptable prices.

3. Optimal price point (intersection of too cheap and too expensive): point of lowest resistance to the price.

4. Indifferent price point (intersection of cheap and expensive): the ‘normal’ price. The pretest included questions about the price levels. Acceptable prices ranged between €0 and €100, which will be the range of possible answers in the PSM questions. The average acceptable prices (across all respondents) for the four selected services ranged between €10 and €50. However, for some services, the acceptable price range seemed to be between about €10 and €25. To ensure that respondents can indicate the price levels of the PSM with enough detail, the price intervals are small for low values and increase for higher values. For example, respondents can choose €5, €7.5, €10, but later only €80, €90, €100. For more details on the price ranges in the pretest, see appendix 1.

Breidert, Hahsler and Reutterer (2006) state that the direct approach to ask consumers’ willingness to pay has several flaws. The most relevant ones to this thesis are listed here, accompanied with a reason why their influence is mitigated (quotes from Breidert, Hahsler and Reutterer 2006). First, “by directly asking the customers for a price, there is an unnatural focus on price” over other attributes. Because there is no real product or service involved here, this does not matter. Moreover, to estimate time value, a precise estimate of prices is necessary. Second, “customers do not necessarily have an incentive to reveal their true WTP”. They may overstate it fore prestige reasons (which is mitigated by the anonymous nature of the questionnaire) or understate it to keep the price of the focal products low. This is probably less of a factor since this is a scientific study, which is not firm-related. Third, “directly asking for WTPs especially for complex and unfamiliar goods is a cognitively challenging task for respondents”. It may be hard to put a single price on a complex service. Therefore, it may still be that the retrieved numbers are not the true WTP’s. This is the reason to use the PSM method, which has more questions about price and can aid the respondent in stating prices.

3.3 Theme 2: Influence of time pressure on opportunity cost of time and thus

consumer decisions.

(31)

31 consumers feel pressed for time, they have a tendency to buy high quality or major, well-known brands (Nowlis 1995, Roselius 1971). They incur extra costs (lower decision quality) to save time, which is precious under time pressure. Alternatively, when a consumer has sufficient time, he can examine products, and examine whether the lower quality of a lesser brand is accompanied by a sufficiently lower price. Thus, when a consumer purchases A-brands under time pressure and lower-quality A-brands when there is less or no time pressure, this indicates that the value of time differs under different levels of time pressure.

Observational data from 172 customers in a French drugstore is used to test this. Information is available on position in the store and actions in front of the shelf (such as searching, grabbing and investigating) for each moment in time. Also purchases are recorded along with their prices, and self-reported time pressure (on a 5-point scale). Finally, data is collected on gender, age and social company.

The data is collected by a researcher, who is in the store with a PDA. On the screen is a map of the store to indicate the position of the customer. When he or she stops in front of a shelf, a second screen is used to record the actions of the customer, both global actions (look, search, ask salespeople, check shopping list) and product actions (grab, investigate, buy, put back, try). Each entry of position or action is recorded as a line in the dataset, with the accompanying time and position (which is linked to a category of products) in the store. There are 16,145 lines in the database. In the third screen, the gender, age and social company of the customer are recorded. The actual purchases are recorded separately by using the receipt. There are 10 zones in the store, which relate to four product categories: Body & bath, Perfumes, Make-up and Luxury (expensive perfumes and body-care for men and women). The effect of time pressure will be assessed in isolation and in conjunction with other possibly relevant variables in the form of two models. These models will use as explanatory variables not only time pressure, but also moment (time and situation) variables and behavioral variables to make a more complete, realistic model (Leeflang et al. 2000, pp. 41-45).

(32)

32 Figure 3.1: Database construction.

The form of the sales model is a multiplicative model. Its advantages over an additive model are that it automatically accommodates interaction effects between the independent variables and that it is more robust (Leeflang et al. 2000, p. 74). However, the data have to be transformed to make the model linear.

Because the data for the models is panel data and thus has multiple observations for one respondent, it is likely that the regressors are correlated to the individual effects. Therefore, Mundlak’s approach is used to accommodate this issue (Mundlak 1978). The means of each variable for a respondent are inserted in the model next to the individual variables (e.g. direct examination of products before the purchase, and the average amount of direct examinations per session during the entire shopping trip). The mean variables accommodate the correlation between the regressors and the individual effects.

To check the results of the sales model, a second model will be estimated. A dummy variable for A-brands will be constructed that divides above-median and below-median price products in a category. The second model is a binary logistic model, with the A-brand (yes or no) dummy as the dependent variable. Including a dummy for each respondent to account for respondents’ tendencies or within-cross-section homogeneity leads to highly insignificant models and parameter estimates because of the low observations-to-variables ratio (smaller than 2:1). Therefore, one purchase per respondent is selected randomly, resulting in a database with 104 cases. This database is labeled as the ‘A-brand database’ (see figure 3.1). A

A-brand database

Sales database

Raw data

(16,145 rows)

Each purchase

(297 rows)

1 purchase per

respondent

(104 rows)

(33)

33 similar set of independent variables as in the first model will be used (the action values, aggregated per respondent) along with time pressure, to examine the effects.

3.4 Theme 3: Influence of social context on opportunity cost of time and thus

consumer decisions.

In theme 3, hypotheses 4a and 4b, the effects of presence (among others, Ariely and Levav 2000) and importance (Zimmer-Gembeck 1999) of social context on opportunity cost of time are examined. In other words, is time more valuable when spent with others?

Respondents will be indicating which type of service they would use in the given social context. To rule out self-presentation as much as possible, time saving services are used. While a higher priced product may be used to ‘show off’, a higher priced and more time saving service is likely to be used primarily to save time. Thus, choosing a more expensive alternative indicates a higher need to save time, and thus a higher value of time.

Furthermore, situations are set in which respondents are asked to report an amount of money that would be appropriate to compensate for a lost experience (Casey, Vukina and Danielson 1995). Since the amount of money relates to the wage for a certain amount of work hours, this measure is comparable to the value of time. Moreover, the tendency to ask compensation is corrected for by the amount that is asked in an additional situation.

Social context will be manipulated by imagined social context (Puntoni and Tavassoli 2005), i.e. letting respondents think about a certain other person. To ensure that the social context is thoroughly entered in the respondent’s mind, the method of Fitzsimons and Bargh (2003) is used. They let respondents fill out questions on another person which are easy to answer but require some visualization or deliberative thought. For example, the initials of the person are asked, along with questions on the appearance, age and hobbies of that person. This method of ‘setting the stage’ is more thorough than a mere description of the social context.

Each situation, the choice of service (transportation type) and amount of compensation (in a restaurant) will be examined on two different levels of social context. The presence of social context will be examined in the restaurant situation, with an ‘alone’ and ‘with a friend’ condition. In this situation, compensation amount is the dependent variable. To control for other variables besides social context, another compensation situation (in a hotel) will be given to the respondents prior to the restaurant situation (see figure 3.2).

Time value: differences in opportunity cost of time per service and situation

Time value: differences in opportunity

cost of time per service and situation

Time value: differences in opportunity

cost of time per service and situation

Management summary

Preface

Contents

Chapter 1 – Introduction

1.1 Motivation

1.2 Research questions

1.3 Position of this thesis in the literature

1.4 Structure of the report

Chapter 2 – Theory

2.1 Introduction to chapter 2

2.2 Opportunity cost of time

2.3 Determining the value of the opportunity cost of time

2.4 Problems with the traditional determination of the opportunity cost of time

2.5 Services and the opportunity cost of time

2.6 Contextual variables and the opportunity cost of time

2.7 Conceptual model

2.8 Contribution to literature

Chapter 3 – Research design

3.1 Introduction to chapter 3

3.2 Theme 1: Willingness to pay for services using open-ended contingent valuation

and price measurer.

3.3 Theme 2: Influence of time pressure on opportunity cost of time and thus

consumer decisions.

A-brand database

Sales database

Raw data

(16,145 rows)

Each purchase

(297 rows)

1 purchase per

respondent

(104 rows)

3.4 Theme 3: Influence of social context on opportunity cost of time and thus

consumer decisions.