Pro-Environmental Behaviour in Aviation

Consumer Decisions to Participate in Voluntary Carbon Offsetting

Maastricht University Faculty of Economics and Business Administration Maastricht, 12 August 2009 Behrens, P. ID number: i347949 MSc in International Business Supply Chain Management Master Thesis Supervisor: Prof. Dr. J. Semeijn 2nd Supervisor: Dr. M.J.H. van Birgelen

Abstract
It has now been widely acknowledged that aviation is a major contributor to global warming. In the absence of regulatory mechanisms, airlines and airports have started offering voluntary carbon offsetting schemes to mitigate the negative impact on the environment. Few studies have investigated an air travellers willingness to pay for the compensation of their CO2 emissions through these schemes. The current body of academic research lacks studies which discuss the influence of consumer-related factors on air travellers acceptance of voluntary carbon offsetting schemes. This thesis adds to the existing literature by analyzing the influence of these factors on the willingness to compensate for CO2 emissions from air travel and by analyzing the link between willingness and likelihood of compensation. An online survey among 128 respondents gives empirical support for all but one hypothesis. The findings suggest that perception of severity, self-perception and importance are positively related to willingness to compensate. Additionally, a positive link between willingness to compensate and likelihood of actual compensation was established. Empirical evidence for the influence of perceived behavioural control on willingness to compensate was not found. Perceived behavioural control may, however, directly determine likelihood to compensate.

Table of Contents
CHAPTER 1. INTRODUCTION............................................................................................ 7 CHAPTER 2. LITERATURE REVIEW & HYPOTHESES DEVELOPMENT ............. 11 2.1 2.2 2.2.1 2.2.2 2.2.3 2.3 THE IMPACT OF AVIATION ON CLIMATE CHANGE................................................. 11 CARBON OFFSETTING .............................................................................................. 12 What is a Carbon Offset?.................................................................................... 13 The Market for Carbon Offsets........................................................................... 13 Voluntary Carbon Offsetting in Aviation ........................................................... 14 WILLINGNESS TO PAY (WTP) FOR VOLUNTARY CARBON OFFSETTING IN AVIATION ................................................................................................................. 16 2.3.1. Factors that Influence Participation .................................................................. 17 2.3.2. Average Payment Amount................................................................................... 19 2.3.3. Reasons for the Lower than Expected Performance of Existing Schemes....... 19 2.4 THE INFLUENCE OF CONSUMER-RELATED FACTORS ON PRO-ENVIRONMENTAL BEHAVIOUR IN AVIATION......................................................................................... 20 2.4.1. Perception of Severity.......................................................................................... 21 2.4.2. Perceived Behavioural Control........................................................................... 22 2.4.3. Self-Perception .................................................................................................... 23 2.4.4. Importance........................................................................................................... 23 2.4.5. Likelihood to Compensate for CO2 Emissions from Air Travel ........................ 24 2.5 SUMMARY................................................................................................................. 25

CHAPTER 3. RESEARCH DESIGN ................................................................................... 26 3.1 3.2 SAMPLE CHOICE & DATA COLLECTION................................................................. 26 QUESTIONNAIRE DESIGN ......................................................................................... 27

3.2.1. Measurement Items for Independent Constructs............................................... 27 3.2.2. Measurement Items for Dependent Constructs.................................................. 27 3.2.3. Additional Measurement Items........................................................................... 28 3.3 3.4 PRE-TEST ................................................................................................................. 28 ANALYTICAL METHOD ............................................................................................ 29

3.4.1. Reliability ............................................................................................................. 29 3.4.2 Construct Validity................................................................................................ 30

CHAPTER 4. ANALYSIS & RESULTS .............................................................................. 31 4.1 RESULTS OF THE MEASUREMENT MODEL .............................................................. 31

4.1.1. Unidimensionality and Reliability ...................................................................... 31 4.1.2. Construct Validity................................................................................................ 32 4.2 4.3 4.3.1 4.3.2 4.3.3 4.4 RESULTS OF THE STRUCTURAL MODEL.................................................................. 33 ADDITIONAL ANALYSES & FINDINGS ..................................................................... 35 Demographics...................................................................................................... 35 Mediation ............................................................................................................. 36 Monetary Value ................................................................................................... 37 CONCLUSION ............................................................................................................ 38

CHAPTER 5. DISCUSSION ................................................................................................. 39 5.1 5.2 5.3 5.4 EMPIRICALLY VALIDATED MODEL ........................................................................ 39 DEMOGRAPHICS ....................................................................................................... 41 MEDIATION .............................................................................................................. 41 CLOSED VS. OPEN-ENDED QUESTIONS ................................................................... 42

CHAPTER 6. CONCLUSION............................................................................................... 43 6.1 6.2 6.3 SUMMARY OF THE RESEARCH ................................................................................. 43 THEORETICAL AND MANAGERIAL IMPLICATIONS ................................................. 44 LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH ................................... 45

REFERENCES ....................................................................................................................... 47 APPENDICES ........................................................................................................................ 52 APPENDIX A COMPARISON OF CO2 EMISSIONS FROM DIFFERENT TRANSPORT MODES 52 APPENDIX B APPENDIX C APPENDIX D DEMOGRAPHICS .......................................................................................... 53 QUESTIONNAIRE ......................................................................................... 56 OVERVIEW OF MEASUREMENT ITEMS ....................................................... 61

List of Tables
TABLE 1. COMPARISON OF CARBON OFFSET CALCULATORS AMS JFK (LONG-HAUL) ...... 15 TABLE 2. COMPARISON OF CARBON OFFSET CALCULATORS AMS TXL (SHORT-HAUL).... 15 TABLE 3. DESCRIPTIVE STATISTICS ON ITEM LEVEL............................................................. 31 TABLE 4. DESCRIPTIVE STATISTICS ON FACTOR LEVEL........................................................ 32 TABLE 5. RESULTS FOR STRUCTURAL MODEL ....................................................................... 33 TABLE 6. MEDIATION ANALYSIS ............................................................................................. 37

List of Figures
FIGURE 1. COMPARISON OF CO2 EMISSIONS FROM DIFFERENT TRANSPORT MODES ............ 11 FIGURE 2. CONCEPTUAL MODEL ............................................................................................ 25 FIGURE 3. EMPIRICALLY VALIDATED MODEL ....................................................................... 34 FIGURE 4. MEDIATED MODEL ................................................................................................. 36

Chapter 1.

Introduction

Climate change is hot. Natural disasters like Hurricane Katrina in the United States in 2005 and frequent news headlines about the Earths rising temperature have made climate change one of the most pressing issues worldwide. According to the latest assessment of the Intergovernmental Panel on Climate Change (IPCC), warming of our climate system is unequivocal (IPCC, 2007). Sea levels have increased, snow and ice extents have decreased and extreme weather events have changed in frequency and severity over the past 50 years. The major cause of these changes is the emission of greenhouse gases (GHG) that result from human activities. One of these human activities is air travel. Air travel has now been widely acknowledged to impact upon the global climate system through the emissions of GHG, including carbon dioxide (CO2). With the inclusion of aviation into the European Union Emissions Trading Scheme (EU ETS) as from 2012, this global transportation sector is about to feel the full blast of regulatory heat (Turner, 2009, p. 29). As a consequence, the aviation sector is urged to take action to mitigate the environmental impact of its activities. A fundamental and simultaneous change in technology and air travel behaviour will become a necessity to achieve the upcoming CO2 emissions reduction targets (Hooper, Daley, Preston & Thomas, 2008). Big steps have already been taken in improving an aircrafts technological configurations. Most of these configurations are focused on achieving a more efficient fuel use. With the introduction of the A380, Airbus has set a milestone for greener aviation technologies. The configuration of the A380 entails the use of lighter materials, improved aerodynamic properties and enhanced engine efficiency (kerman, 2005). Although aviation is making advances with improved and innovating technologies, the demand for air travel is growing too fast for environmental benefits to be realized from these advances (Chapman, 2007). Thus, technological improvement must be complemented with a change in air travel behaviour to effectively reduce CO2 emissions. However, whereas innovating technologies have been developing in a slow manner, there is no evidence for a change in air travel behaviour. Each and every one of us can make changes in the way we live our lives and become part of the solution [to climate change], Al Gore (2006), 45th Vice President of the United States, said in his movie 'An Inconvenient Truth'. This quote implies a shift from unsustainable to 7

pro-environmental behaviour in a variety of areas including air travel. Ideally, people should start flying less and rather use alternative transport modes that are less harmful to the environment. Nonetheless, people might not be able to forego flying as alternative transport modes cannot replace air travel on huge intercontinental distances. Moreover, previous research revealed that people are not willing to forego flying altogether for convenience reasons (Hooper et al., 2008; Gssling, Haglund, Kallgren, Revahl & Hultman, 2009). In fact, the demand for air travel is growing due to fast globalization implying movement of people around the world. This paradox entails that other options are required for inducing behavioural change among air travellers. One of these options is voluntary carbon offsetting. Some airlines or airports have recently started offering this service. It enables passengers to compensate their individual CO2 emissions by means of a payment scheme. Voluntary carbon offsetting has evolved as a strategy to assure stakeholders that the aviation sector is taking action to mitigate the environmental impact of its activities. However, these schemes are in their early years and critics question their efficiency and credibility in providing a long-term solution to the problem. In addition, little is known about the adoption of these schemes among air travellers (Hooper et al., 2008). Hence, this thesis focuses on the willingness and likelihood of air travellers to pay for CO2 compensation. In other words, it focuses on their decision to adopt pro-environmental behaviour in aviation. The literature on an air travellers willingness to pay for voluntary carbon offsetting in aviation is scarce. The latest research was conducted by Gssling et al. (2009) who found that the majority of air travellers does not seem to be aware of options to compensate for CO2 emissions from flying. Overall it has been found that the willingness to participate in voluntary carbon offsetting is positively influenced by an individuals level of environmental concern, its knowledge and awareness of aviations impact on the environment and the type of offsetting project. In contrast, air travellers generally do not see themselves responsible for providing a solution, which negatively impacts their willingness to compensate CO2 emissions (e.g. Gssling et al., 2009; Hooper et al., 2008). The existing literature thus far has disregarded factors from established theories in the field of consumer behaviour in predicting a consumers willingness to pay for voluntary carbon offsetting in aviation. Some of these factors, however, have successfully been used in 8

predicting and explaining pro-environmental behaviour in areas like beverage packaging, food consumption and recycling (Van Birgelen, Semeijn & Keicher, 2008; Vermeir & Verbeke, 2006; McCarty & Shrum, 1994). In particular, the Theory of Planned Behaviour (Ajzen, 1991) appears to be a powerful predictor for ecological behaviour in these areas. Therefore, this thesis aims at investigating the explanatory power of consumer-related factors for proenvironmental behaviour in aviation with a focus on passenger transport. The central problem statement for the research at hand is:

WHICH FACTORS RELATE TO A CONSUMERS DECISION TO OFFSET ITS INDIVIDUAL CO2 EMISSIONS FROM FLYING? To investigate this problem statement several sub questions have been formulated: What is voluntary carbon offsetting? Which factors relate to an air travellers willingness to compensate its CO2 emissions?

Does willingness to compensate affect likelihood to compensate?

This thesis will contribute to the current body of literature in the field of voluntary carbon offsetting in air travel by providing insights about the influence of consumer-related factors on consumers willingness to adopt these schemes. Furthermore, this study differentiates from earlier ones by not only predicting the willingness to compensate but also by analyzing the link between willingness to compensate and likelihood to compensate in real. From a practical perspective, the findings will serve the aviation sector in identifying potential users of voluntary carbon offsetting. These users can then be effectively targeted with the necessary marketing tools. Also, this thesis has the potential to raise awareness of the efforts taken by airline companies in tackling climate change. A better understanding of consumer behaviour will determine the success of any proactive approach towards environmental issues from which competitive advantage may arise. The remainder of this thesis is structured as follows: in chapter two, the reader of this thesis is provided with a thorough review on the existing academic literature about the topic at hand. This literature review will serve as a basis for the development of five hypotheses. These hypotheses will be visualized in a conceptual model which will be presented at the end of the 9

chapter. Chapter three will describe research design, questionnaire development and data collection. Also, the analytical method will be briefly introduced. The fourth chapter will present the results of the measurement model and the findings concerning the hypothesized relationships in the structural model. The implications of these results will be thoroughly discussed in chapter five. Chapter six will conclude the research study by discussing its managerial & theoretical implications, its limitations and by providing suggestions for future research.

10

Chapter 2.

Literature Review & Hypotheses Development

To answer the problem statement and research questions formulated in the introduction, this thesis draws on two streams of academic literature. The first stream consists of previous research on consumers willingness to pay (WTP) for carbon offsetting in aviation. The second stream comprises established theories about the understanding of consumer behaviour which discuss the relation between behavioural intentions and actual behaviour as well as consumer-related factors that influence behaviour. Subsequently, this chapter will review both literature streams in depth to find support for the research for this thesis and that serve to construct the conceptual framework. The in-depth review will be preceded by a clarification about the impact of aviation on climate change and the concept of carbon offsetting to provide a common basis for understanding.

2.1

The Impact of Aviation on Climate Change

Climate change is a hot topic that already fosters major counteracting efforts in a variety of industry sectors with the exception of the transport sector. Close attention has only recently been given to the contribution of the transport sector to global warming. It is now widely being recognized by industrialised countries that transportation activities are accountable for a substantial portion of CO2 emissions. It is further being recognized that, of all transport modes that carry passengers as well as freight, aviation is the most polluting. Figure 1 shows a comparison between CO2 emitted from a train, car or airplane carrying passengers on the track Amsterdam Central Station (NL) to Berlin Central Station (DE) with a distance of approximately 654 kilometres. A detailed overview of this comparison is displayed in Appendix A (ecopassenger.org, 2009).

Figure 1. Comparison of CO2 emissions from different transport modes (adapted from ecopassenger.org, 2009) 11

The comparison shows that the CO2 emission of the flight (including transport to and from the airport) is highest with 95.6 kilogram per passenger. The difference would even get higher if one takes the total global warming effect of aviation into account, which also stems from greenhouse gases other than CO2 (e.g. nitrogen oxides). Furthermore, CO2 emissions from air travel are especially serious in character since they are released in the higher layers of the atmosphere (Gssling et al., 2007) and are not absorbed on the ground (e.g. by trees). Consequently, CO2 emissions from aircrafts impact the atmospheric composition (IPCC, 1999). A change in this composition triggers, for instance, the formation of condensation trails which contribute to climate change. An additional concern comes from the prediction that air travel activity is about to grow by five percent annually (kerman, 2005; Chapman, 2007). Taking into account this expected growth rate, aviation is considered the most unsustainable mode of transport at present (Chapman, 2007). Until now, the aviation industry has mainly reacted to the climate change issue through technological improvements and flight route optimization. However, the slow pace of technological progress in the aviation sector is likely to be outpaced by the rapid growth in air travel (Gssling et al., 2007; Frew & Winter, 2008). In spite of the recognized detrimental impact of air travel on climate change, few regulatory mechanisms aiming at reducing this impact have been implemented. Previous attempts for regulation, such as the Dutch flight tax1, have been fruitless due to difficulties in reaching international agreements. In the absence of regulatory mechanisms, some airlines have started to address the growing concern about aviations impact on the environment by offering its passengers the possibility to offset their individual flights CO2 emissions by means of a voluntary payment scheme. This possibility is also referred to as voluntary carbon offsetting.

2.2

Carbon Offsetting

Prior to discussing voluntary carbon offsetting in aviation in more detail in this section, the reader of this thesis is provided with a general overview on carbon offsets and carbon markets.

The Dutch Government implemented a mandatory flight tax on July 1, 2008. The implementation resulted, amongst others, in a weakened competitive position of Dutch airlines. Strong resistance and opposition finally led to the abolishment of the flight tax as from July 1, 2009 (nieuwsbank.nl, 2009)

12

2.2.1 What is a Carbon Offset? According to Taiyab (2006, p. 3), a carbon offset negates or neutralises a ton of CO2e (carbon dioxide equivalent) emitted in one place by avoiding the release of a ton of CO2e elsewhere or absorbing/sequestering a ton of CO2e that would have otherwise remained in the atmosphere. The term equivalent is used to account for the fact that not only CO2, but also other greenhouse gases can be offset. The focus of the present study, however, is on the offset of CO2 emissions. The creation of a carbon offset follows from a variety of projects. Specific projects, such as reforestation or renewable energy, serve to generate carbon credits or emission reductions in the first place. Those carbon credits or emission reductions can then be purchased by individuals or companies that wish to offset their CO2 emissions to become carbon neutral. The price of a carbon credit varies according to different market- and/or project characteristics (Taiyab, 2005, 2006). One of the crucial aspects of carbon offsets is the principle of additionality (Taiyab, 2005, 2006; Gssling et al., 2007). It implies that purchasers of carbon offsets must ensure that emissions are truly and additionally reduced, which requires financial support through the sale of carbon credits. In other words, any emission reduction that would have occurred without financial support is not considered as being additional. In order to trade carbon offsets, several carbon markets have emerged. Carbon markets can be either regulatory or voluntary, as will be elaborated in the next section.

2.2.2 The Market for Carbon Offsets The market for trading carbon offsets has begun to emerge in the twentieth century as a direct consequence of the consciousness that our climate is changing. Taiyab (2005, p. 4) refers to carbon markets as the buying and selling of carbon credits and allowances. Most of todays carbon markets are regulatory, in which an authority determines and caps the amount of CO2 that a participant is allowed to emit during a certain period. The authority also issues tradable permits that enable a participant to exchange unused allowances of CO2 emissions with other participants in the market. The goal of such a trading mechanism is to eventually reduce CO2 emissions. Probably the most well-known regulatory trading mechanisms are the Clean Development Mechanism (CDM) used under the Kyoto Protocol of 1997 and the EU ETS.

13

Now, in recent years of the twenty first century, heightened awareness about the detrimental effects of climate change on future generations and the lack of regulatory mechanisms has led to a rapid expansion of voluntary carbon offset markets. Voluntary carbon offsetting allows individuals, organizations or companies to assume responsibility for any activity that brings harm to the environment (Strasdas, 2007; Taiyab, 2005). It gives them an opportunity to buy or sell carbon credits or emission reductions from retailers for reasons other than regulatory compliance. These credits are generally referred to as Verified Emission Reductions (VERs). VERs are generated through projects outside of the regulatory market and will be evaluated through an external auditing process (Taiyab, 2006). One of the main reasons for companies or organizations to participate in the voluntary carbon market is to show corporate social responsibility to consumers. Individuals, in contrast, participate to mitigate personal responsibility for environmental harm (Hooper et al., 2008).

2.2.3 Voluntary Carbon Offsetting in Aviation Voluntary carbon offsetting has only recently been considered with regard to the environmental impacts of air travel. Until today, only a few airline companies or airports have started to react to the detrimental impact of their business by offering air passengers the possibility to voluntarily offset their flights CO2 emissions. For example, the Dutch airline KLM2 recognized that air passengers do not only want to rely on the airlines efforts to lower the negative consequences of air travel, but would like to contribute to improvement themselves. Therefore, the airline launched an extra online service in March 2008, named CO2ZERO, for passengers who desire to fly carbon neutral. KLM states that CO2 compensation is the easiest, cheapest and most effective way to fly carbon neutral (klm.com, 2009). In a similar way, as from September 2008, the German airport Cologne-Bonn started supporting worldwide climate change projects of the independent Swiss organization Myclimate (Kln Bonn Airport, 2008). Myclimate is one of the worlds largest providers of CO2 compensation measures (myclimate.org, 2009). By visiting the airports website or by using specific electronic climate-counters which are situated in every terminal, air passengers can offset their individual CO2 emissions from any flight with any airline on a voluntary basis. Cologne-Bonn Airport is the first German airport that provides such a service

KLM = Koninklijke Luchtvaart Maatschappij www.klm.com

14

to its customers. Cologne-Bonn Airport, as well as KLM, tries to appeal to a growing share of green customers by providing the offsetting service. Todays efforts of airlines and airports with regard to CO2 emission compensation can be regarded as a precursor of future regulatory mechanisms for the aviation sector. Most importantly, the European Union has announced that the aviation sector will be included into the EU ETS as of 2012 (European Union, 2008). It is assumed that this inclusion will serve as a starting point for additional regulatory mechanisms in aviation on a global scale. Although voluntary carbon offsetting schemes in aviation look promising at a first glance, their success and efficiency remains questionable. Criticism comes from a variety of actors, such as NGOs. Gssling et al. (2007) investigated the efficiency and credibility of existing voluntary carbon offsetting schemes in aviation. By means of a website analysis of 50 organizations that provide voluntary offsetting schemes, the authors found substantial differences between the approaches used by those offset providers. The authors identified differences in compensation projects, price levels and accountability. Major differences also exist between the carbon calculators that an airline or airport uses to determine the CO2 emissions from a particular flight (Gssling et al., 2007; Hooper et al., 2008). A comparison between the calculators of KLM and Myclimate, mentioned earlier, for a one-way flight from Amsterdam Schiphol (AMS) to either New York (JFK) or Berlin Tegel (TXL) shows the following:

Provider KLM Myclimate Table 1. Provider KLM Myclimate Table 2.

Price per person in 2.96 28

CO2 emissions in kilograms 490 1.195

Source www.klm.com (2009) www.myclimate.org (2009)

Comparison of carbon offset calculators AMS JFK (long-haul) Price per person in 0.40 3.95 CO2 emissions in kilograms 66 158 www.klm.com (2009) www.myclimate.org (2009) Source

Comparison of carbon offset calculators AMS TXL (short-haul)

15

According to the German climate organization Atmosfair, such differences mainly stem from different perspectives of airlines and general climate organizations (atmosfair.de, 2009). More specifically, airlines only take the emissions of CO2 into account. Climate organizations such as Myclimate seem to account for the emission of other greenhouse gases than CO2 as well, although their calculator is called CO2 calculator (myclimate.org, 2009). The use of voluntary carbon offsetting schemes in aviation has also been criticized for being an insufficient way to address climate change (Gssling et al., 2007; Hooper et al., 2008; Strasdas, 2006; Borup, 2007). They rather provide an indirect solution to the problem than directly reducing CO2 emissions. It is further argued that such schemes will only encourage passengers to stick with their current flying behaviour, instead of reducing the demand for flights. The voluntary nature of the schemes gives no incentive to airlines to improve technology, as would be the case under a regulatory regime. Furthermore, critics are concerned about the verification of different projects, which is not handled in a uniform way. Inconsistent methods for verification will affect the credibility of voluntary carbon offsetting schemes (Strasdas, 2006; McKerron, Egerton, Gaskell, Parpia & Mourato, 2009). Although voluntary carbon offsetting schemes in aviation demonstrate a number of problems, they represent an important first step in dealing with the conflicting trends of growing demand for air travel on the one hand and growing concern for aviations impact on the environment on the other hand. Such schemes have an awareness-raising potential, since they enable passengers to see how much they contribute to climate change when flying. It is assumed that the knowledge about this contribution will encourage passengers to take corrective action. The key success factor of such schemes, however, is their acceptance by air travellers. In other words, air travellers must be willing to pay for the offset of CO2 emissions from flying. Since the emergence of voluntary carbon offsetting schemes in aviation, some studies have investigated factors that motivate air passengers to participate in such a scheme.

2.3

Willingness to Pay (WTP) for Voluntary Carbon Offsetting in Aviation

Few studies have been devoted so far to investigating an air passengers willingness to offset the CO2 that is emitted during a flight (MacKerron et al., 2009; Brouwer, Brander & Van Breukering, 2008; Gssling et al., 2009; Becken, 2004; Hooper et al., 2008). On the one hand, there seems to be a huge market potential for voluntary carbon offsetting in aviation. A 16

consumers WTP to offset CO2 seems to be much higher than is generally assumed. On the other hand, the share of passengers that has actually offset CO2 emissions through existing schemes has been found to be low. This section will discuss factors that influence participation in voluntary carbon offsetting, the average amount consumers are willing to pay for it and reasons for the disappointing performance of existing schemes.

2.3.1 Factors that Influence Participation The first factor which influences voluntary participation in carbon offsetting is an air travellers level of environmental concern. It has been found that there is a rather common understanding that climate change poses a genuine threat on current and future generations (Gssling et al., 2009; Hooper et al., 2008; Brouwer et al., 2008). The majority of respondents from previous research is either concerned or very concerned about the environment, while only a small share does not seem to worry at all. WTP for voluntary carbon offsetting seems to be highest for air travellers who report to be very concerned about the environment. Hooper et al. (2008) have even found WTP among those travellers to be higher than expected. More surprisingly is the paradox that there is a self-reported WTP among those travellers that do not at all worry about the environment (Gssling et al., 2009). A similar result has been obtained for the second factor of influence, which is a passengers knowledge and awareness about the relationship between air travel and climate change. Previous research has found that a large majority of people knows that air travel is environmentally harmful. Travellers who are knowledgeable and aware about the detrimental effect of aviation on the environment report a higher WTP than those that are not (Gssling et al., 2009; Hooper et al., 2008). Gssling et al. (2009), however, have found that the results on WTP between those groups do not differ substantially. According to these authors, such a paradox might be explained by a general uncertainty among air travellers about the way in which aviation contributes to climate change. The level of knowledge and awareness about the relationship between flying and climate change as well as the level of environmental concern seem to be primarily influenced by the nationality of an air traveller and its frequency of travel:

17

Nationality. European travellers, in particular from the Netherlands and the United Kingdom (UK), have been found to be significantly more aware and knowledgeable about aviations impact on climate change than any other nationality (Brouwer et al., 2008). As has been identified earlier, this high level of knowledge and awareness is positively related to these travellers WTP for carbon offsetting. The level of awareness and knowledge was also found to be high among North American passengers. Passengers from Asia seem to be least aware and knowledgeable about the relation between aviation and climate change. A reason for this low result might be that Asians, as well as Latin Americans and Africans, do not express concern about environmental issues in general. In contrast, UK travellers report the highest level of general environmental concern, closely followed by other European and North American travellers (Brouwer et al., 2008). Travel Frequency. Gssling et al. (2009) distinguish between frequent and less frequent travellers who make 1-30 or respectively more than 30 flights per year. According to them, the majority of air travellers falls into the category of less frequent travellers. Whereas only a small share of less frequent travellers seems to know about the possibility to offset CO2 emissions, about half of the frequent travellers is knowledgeable about the concept. Surprisingly, however, there are only slight differences between the stated general WTP for voluntary carbon offsetting of both categories. A difference was found, however, for the actual participation in voluntary offsets. In contrast to frequent travellers, none of the less frequent travellers has ever participated in voluntary carbon offsetting in aviation. Gssling et al. (2009) have thus stated that voluntary carbon offsetting is more common among frequent travellers due to a higher level of knowledge and awareness. Third, an air travellers WTP is influenced by the type of offset being offered by an airline or airport. There are various forms of offsets that can be purchased. For example, MacKerron et al. (2009) distinguish between certified and uncertified offsets, with or without specific cobenefits. Certified offsets are generated through projects in countries not participating in the Kyoto Protocol and are subject to registration by the UNFCCC3 (Gssling et al., 2009). That registration brings them a high level of credibility. Co-benefits of a voluntary carbon offsetting scheme include among others the support of human development, the promotion of biodiversity or technological development in the low-carbon sector (Taiyab, 2005, 2006). It
United Nations Framework Convention on Climate Change, an international treaty that sets general rules and goals for confronting climate change. 192 countries have joined this treaty (unfccc.int, 2009)
3

18

has been found that air travellers are willing to pay more for offsets that are certified. Air travellers are also willing to pay more for schemes that possess co-benefits. In this respect, air travellers did not differentiate between different types of co-benefits in stating their WTP (MacKerron et al., 2009). Finally, influence on WTP comes from air travellers perceptions of their responsibility in tackling climate change. Air travellers that feel a sense of moral obligation to pay for environmental protection have generally been found to be willing to pay for voluntary carbon offsetting in aviation (Brouwer et al., 2008). However, previous research has shown that only a small group of passengers feels responsible for compensating the emission of CO2 during a flight (Gssling et al., 2009; Hooper et al., 2008). The majority of passengers rather assume responsibility from government, airlines or aircraft producers.

2.3.2 Average Payment Amount Apart from analysing factors that stimulate participation in voluntary carbon offsetting schemes, previous studies have also sought to determine which price a consumer is willing to sacrifice for participation. It has been found that the average extra amount that air passengers are willing to pay for an offset exceeds the average price currently set by the supply side (Brouwer et al., 2008; MacKerron et al., 2009; Hooper et al., 2008). For example, Brouwer et al. (2008) have found that passengers are willing to pay on average 25 Euros per ton CO2e, while suppliers offer offsets at an average price of 12 Euros per ton CO2e. Thus, there is market potential since a consumers WTP seems to be much higher than is generally assumed. Becken (2004) even found that passengers (in particular tourists) view voluntary carbon offsetting as a rather cheap way of redeeming the guilty feeling when travelling by air. This view might explain the fact that passengers rather pay for offsets than to reduce or refrain from flying.

2.3.3

Reasons for the Lower than Expected Performance of Existing Schemes

Even though there seems to be evidence for a huge market potential for voluntary carbon offsetting schemes in aviation, the current performance of existing schemes is disappointing. Tour operators and airlines that currently offer schemes to offset CO2 report that customers show limited interest in these schemes (Gssling et al., 2009). An important reason seems to

19

be that a large amount of air travellers is not aware of the existence of such schemes. For example, Gssling et al. (2009) found that only 24 percent of all air travellers surveyed is aware about the concept. Awareness is lowest among less frequent travellers. This finding might pinpoint on a lack of communication towards air passengers. A second reason is that informed air travellers might have a negative attitude towards offsetting schemes due to a lack in credibility and efficiency. Third, air passengers that are willing to pay for voluntary carbon offsetting might be discouraged by so-called free riders, who do not see the need of paying as long as others do it. This problem could only be handled by the introduction of a mandatory payment scheme (Brouwer et al., 2008). Lastly, a low level of participation might stem from different perspectives of airlines and passengers about who is responsible for the reduction of CO2 emissions. As has been outlined before, airlines assume active participation from its passengers, while passengers see the government, aircraft producers or airlines in charge of the issue (Gssling et al., 2009).

2.4

The Influence of Consumer-Related Factors on Pro-Environmental Behaviour in Aviation

Pro-environmental consumption behaviour has been researched extensively (e.g. Van Birgelen et al., 2008; Laroche, Bergeron & Barbaro-Forleo, 2001; Vermeir & Verbeke, 2006; Cleveland, Kalamas & Laroche, 2005) using consumer-related factors from renowned theories for explaining consumer behaviour (e.g. Ajzen, 1991). However, previous research was mainly focused on predicting pro-environmental behaviour in the consumption of tangible products. For example, Vermeir and Verbeke (2006) centre their analysis on the purchase of organic dairy products. To the best of the researchers knowledge, consumerrelated factors have thus far been disregarded in previous studies on voluntary carbon offsetting in aviation. Therefore, in the present study, it is intended to analyze the influence of these factors on pro-environmental consumption behaviour for a service, which is of an intangible nature. One of the theories that have successfully been used in predicting pro-environmental consumption behaviour is the Theory of Planned Behaviour (TPB) (e.g. Van Birgelen et al., 2008; Oreg & Katz-Gerro, 2006). The TPB has been developed by Ajzen (1991) and entails

20

that a consumers actual behaviour is preceded by a consumers behavioural intention. In other words, a behavioural intention can adequately predict actual behaviour. A behavioural intention refers to a course of action that a person intends to follow. For example, before an air traveller will actually pay for offsetting his flights emissions, he must have the intention to do so. A consumers behavioural intention, in turn, can accurately be predicted by its attitude towards that behaviour, its subjective norms and its notion of perceived behavioural control. In the present study, the TPB serves as the basis for a conceptual model which will describe the influence of different constructs on a consumers pro-environmental behaviour in aviation. First, four constructs (perception of severity, perceived behavioural control, self-perception and importance) are examined for their predictive nature on a consumers intention to behave pro-environmentally when flying. This behavioural intention is referred to as a consumers willingness to compensate for CO2 emissions from air travel. Then, it will be analyzed whether a consumer that is willing to pay, is also likely to do so in a real-life context. More specifically, it will be investigated whether there is a gap between the behavioural intention and actual behaviour with regard to environmental-friendly flying. The actual behaviour is referred to as a consumers likelihood to compensate for CO2 emissions from air travel. In the following section, the hypotheses for the underlying research will be developed.

2.4.1 Perception of Severity A consumers attitude towards certain behaviour refers to the extent to which a consumer evaluates certain behaviour either favourably or unfavourably (Ajzen, 1991). Previous studies have found that attitudes are valid predictors for behavioural intentions (e.g. Minton & Rose, 1997). More specifically, consumers who show a favourable (unfavourable) attitude towards certain behaviour have a stronger (weaker) intention to perform the behaviour in question. For example, a person who demonstrates a positive attitude towards organic food products is more likely to purchase in organic supermarkets. Some studies in the field of environmentally friendly consumer behaviour use an individuals perception about the severity of ecological problems in order to denote a consumers attitude towards the environment (Laroche et al., 2001; Banerjee & McKeage, 1994). Laroche et al. (2001) found that consumers, who perceive ecological problems as having severe consequences for the security of the world, are willing

21

to pay more for the consumption of tangible ecological products. Based on this finding, it is worthwhile to explore whether a consumers perception about the severity of ecological problems will also make him/her pay more for an intangible product. More specifically, it will be investigated whether a consumers perception about the severity of ecological problems caused by CO2 emissions from air travel affects the willingness to compensate for those emissions. For this purpose, the following hypothesis is developed:

Hypothesis 1: Consumers, who perceive that CO2 emissions from air travel create a severe ecological problem, are more willing to compensate for these emissions.

2.4.2 Perceived Behavioural Control A behavioural intention can also be determined by a consumers perception about its ability to control its behaviour. According to Ajzen (1991), perceived behavioural control (PBC) refers to whether an individual perceives the performance of the behaviour in question as being easy or difficult. PBC can also be described in terms of perceived self-efficacy (Bandura 1977, 1982), which is concerned with judgments of how well one can execute courses of action required to deal with prospective situations (Bandura, 1982, p. 122). In other words, it denotes the extent to which a consumer believes that his individual efforts will contribute to the solution of a problem. In their study about sustainable food consumption, Vermeir and Verbeke (2006) refer to this concept as perceived consumer effectiveness (PCE). It was found that a high PCE or a high self-efficacy is necessary to evoke a consumer to transfer a positive attitude towards a certain issue into actual behaviour (Vermeir & Verbeke, 2006; Bandura, 1977, 1982). In the context of the present study, we refer to PBC as being the extent to which an air passenger believes that his or her individual payment for CO2 compensation will have an impact on the environment as a whole. In line with earlier studies, it is expected that a consumer who is convinced of the positive effect on the environment of his individual contribution, is more likely to be willing to compensate for CO2 emissions from air travel. Therefore, the following hypothesis is formulated:

22

Hypothesis 2: Consumers, who perceive their individual efforts to prevent or reduce CO2 emissions from air travel to have a positive effect on the environment as a whole, are more likely to compensate for these emissions.

2.4.3 Self-Perception In a study conducted by Van Birgelen et al. (2008), the authors use the self-perception theory (Bem, 1967) as a predictor for consumers pro-environmental behaviour in their purchaseand disposal decisions of beverage packages. The authors hypothesize that a consumer who behaves pro-environmentally in one area will also show that behaviour in other areas. As their hypothesis was supported, the present study will test the validity of this predictor in an aviation context. For example, it is examined whether a consumer that behaves ecologically in the way that he or she for instance collects and recycles paper is also willing to compensate for CO2 emissions from air travel. For this purpose, the following hypothesis is developed: Hypothesis 3: Consumers, who behave pro-environmentally in other areas than air travel, are more willing to compensate for CO2 emissions from air travel.

2.4.4 Importance McCarty and Shrum (1994, 2001) conducted a behavioural study in the context of recycling. They consider the importance of recycling and its inconvenience as antecedents of recycling behaviour. The importance of recycling refers to the degree to which a consumer expresses concern about environmental issues. The inconvenience relates to the ease or difficulty of performing recycling behaviour. Both concepts were also applied in a study by Laroche et al. (2001) in which the authors investigate the demographic, psychological and behavioural profile of consumers who are willing to pay more for environmentally friendly products. Both studies found evidence for the strong influence of the inconvenience of environmentally friendly behaviour on the performance of such behaviour. Although McCarty and Shrum (1994) found the influence of importance to be weak, the study of Laroche et al. (2001) revealed that consumers who are willing to pay more for environmentally friendly products think that it is important to behave in an ecological way.

23

As there is some controversy in the results on importance, the present study would like to investigate this particular item in the context of the topic at hand. In the present study, we consider consumers who find it important to behave in an ecologically favourably way to be more willing to compensate for CO2 emissions from air travel: Hypothesis 4: Consumers, who view pro-environmental behaviour as important to themselves or society as a whole, are more willing to compensate CO2 emission from air travel.

2.4.5 Likelihood to Compensate for CO2 Emissions from Air Travel As has been mentioned before, behavioural intentions have been found to adequately predict actual behaviour. Thus, we will use the results on a consumers willingness to compensate for CO2 emissions from air travel to analyze their influence on the actual likelihood of compensation. In line with Ajzen (1991), it is expected to find a positive relationship between those two constructs. A prerequisite for that positive relationship is a consumers volitional control about the behaviour in question (Ajzen, 1991). A person has volitional control if it can liberally decide to perform certain behaviour or not. Only in the presence of volitional control, will an individuals intention to perform certain behaviour be transferred into actual behaviour. Since voluntary carbon offsetting schemes allows for volitional control, the following hypothesis is added:

Hypothesis 5: A consumers willingness to compensate CO2 emission from air travel is positively related to its likelihood to compensate.

Figure 2 shows the conceptual model as a visual representation of the posited hypotheses.

24

H1

H2 H5 H3

H4

Figure 2.

Conceptual Model

2.5

Summary

To conclude this review, there is a growing concern about aviations impact on climate change. Nowadays, the recent implementation of voluntary carbon offsetting schemes by some airline companies or airports is heavily discussed in the media. On academic level, little research has been conducted on the acceptance of such schemes among consumers. The review of existing studies revealed that there is a need for further research about factors that will induce consumers to compensate CO2 emissions from air travel. It is assumed that those factors can be found in well-known theories about consumer behaviour, which will subsequently be analyzed in the research for this thesis. Only if consumers are prepared to adopt pro-environmental behaviour in aviation will voluntary carbon offsetting schemes be a successful way of mitigating aviations negative impact on climate change.

25

Chapter 3.

Research Design

This chapter will outline the design and structure of the research that has been conducted to validate the conceptual model. First, a description of the sample and data collection is provided. Then, the design of the questionnaire and its pre-test is discussed. Last, this chapter will introduce the analytical method used.

3.1

Sample Choice & Data Collection

For the purpose of the current research, cross-sectional data were gathered through an online survey among air travellers. An online survey is advantageous in terms of low cost, ease of creating, quick data gathering and fast online statistical analysis (Schmidt & Hollensen, 2006). The online survey has been enabled by special software, NetQuestionnaires, which access is provided by Maastricht University. The invitations to participate in the online survey were sent out by e-mail in June 2009. The e-mail contained a link to the website of the university that administers the survey. The sample consisted of adult persons (aged 18) under the assumption that every adult qualifies for being an air traveller. No further restriction on demographic characteristics of the sample has been made. First, the researcher targeted people with whom she is directly acquainted, such as friends or relatives. These acquaintances were then kindly requested to forward the invitation to as many persons as possible. This procedure is referred to as snowball-sampling (Schmidt & Hollensen, 2006). Second, the invitation was sent to the universitys online research panel, consisting of students and non-students who subscribed for regular participation in a wide variety of surveys. Third, the invitation link has been published on online community websites, such as Facebook. For the purpose of increasing participation a monetary incentive was provided. In order to limit social desirability bias, which is associated with environmental issues (MacKerron et al., 2009), anonymous participation was guaranteed to the respondents. About 250 responses were received. Of these responses, only 128 questionnaires were completed and hence qualified for the research. The sample consists of 45.5% male and 55.5% female respondents. Approximately 16.4% of the respondents are frequent business travellers. The remaining 83.6% usually fly for leisure. Most respondents are aged between

26

25-34 years and reside in the Netherlands. A detailed overview of demographic characteristics is presented in Appendix B.

3.2

Questionnaire Design

The questionnaire was developed in English as it is assumed that the targeted sample has sufficient knowledge of this language to understand the questions. The questionnaire is attached to this study in Appendix C. Apart from demographic questions the questionnaire contained 11 closed questions and 1 open-ended question related to the constructs under investigation. The closed questions contained 3 to 4 statements or items that measure a specific construct. Respondents were asked to indicate their level of agreement/disagreement with these items which were all scored on a 7-point Likert-item scale ranging from 1 = totally disagree to 7 = totally agree. All items were adapted from previous research on the topic at hand, which will be elaborated in the upcoming paragraphs. Negative wording was sometimes incorporated to check whether respondents answered in a consistent way. Negative items were recoded before analysis. A summary of the items used can be found in Appendix D.

3.2.1 Measurement Items for Independent Constructs The independent constructs included in the framework are perception of severity, perceived behavioural control, self-perception and importance. The independent constructs perception of severity and perceived behavioural control were each measured using 3 items that were adapted from Van Birgelen et al. (2008), who studied pro-environmental consumption behaviour with regard to beverage packaging. For measuring the construct of self-perception, 4 items were used to assess a respondents current environmental behaviour in areas other than aviation. These items were adapted from Kaiser, Wlfing and Fuhrer (1999) who established environmental attitude as a powerful predictor of pro-environmental behaviour. The influence of the importance of ecological behaviour on the dependent variables was measured using 2 items from McCarty and Shrum (1994, 2001). All items were customized for the research topic at hand.

3.2.2 Measurement Items for Dependent Constructs The dependent constructs are willingness to compensate and likelihood to compensate for CO2 emissions from air travel. Two questions were devoted to measuring the dependent 27

variable willingness to compensate. The first question uses 3 items to assess a respondents willingness to compensate. These items were adapted from Oreg and Katz-Gerro (2006) who investigated pro-environmental behaviour on a cross-national level. This question was then followed by an open question, in which respondents could state the maximum amount that they would be willing to pay for compensating CO2 emissions from flying. A distinction was made between short- and long-haul flights under the assumption that a passengers willingness to compensate is influenced by the flight distance. In order to assess likelihood to compensate, respondents were instructed to carefully read a scenario about an airlines offer to compensate for CO2 emissions from a flight from a European airport to the United States. This scenario was adapted from MacKerron et al. (2009). Respondents were explicitly asked to view the scenario as a real-life situation. The scenario was then followed by 4 items that asked about the respondents acceptance of the mentioned offer under differing conditions.

3.2.3 Additional Measurement Items In addition to the items used for measuring the main constructs of the conceptual framework, 3 items were used at the beginning of the questionnaire to test the general awareness among respondents about climate change and its consequences. These items were adapted from Van Birgelen et al. (2008). The questionnaire ended with demographic questions about gender, age, place of residence and the usual purpose of flying (business vs. leisure). Respondents that were interested in taking part in the prize lottery could enter their e-mail address before submitting the questionnaire.

3.3

Pre-Test

Before its implementation, a questionnaire should be pre-tested to identify and eliminate any problems or misinterpretations (Schmidt & Hollensen, 2006). Thus, to ensure that the questionnaire provides the information sought, a participating pre-test was conducted using a smaller sample of approximately 12 respondents. The goal of the research remained undisclosed to these respondents, as one could otherwise expect biased results. Respondents were asked to complete and evaluate the questionnaire with regard to e.g. item wording or question order. Overall, they had little difficulties in understanding the questions. Only minor adaptations in item wording were made. The answers to the open question about the maximum amount that a respondent would be willing to pay for CO2 compensation revealed

28

that the payment level in the real-life scenario was set too low. Therefore, this level was raised from initially 16 to 25, which more or less corresponds to the average amount reported by the test-respondents for a long-haul flight.

3.4

Analytical Method

The data that have been gathered for the present study were analyzed by means of partial least squares (PLS). The PLS algorithm is applied with the support of the statistical software tool SmartPLS (Ringle, Wende & Will, 2005). PLS is an analysis technique that enables simultaneous estimation of both the measurement and the structural model (Haenlein & Kaplan, 2004). The use of PLS is advantageous in different ways. First, the estimations are very robust against skewed data distributions and multicollinearity (Cassel, Hackl & Westlund, 2000). The kurtosis values in Table 3 show that the data set is skewed and consequently not normally distributed. Table 4 displays the correlations between the latent variables. This table shows that PBC and Importance are inherently intercorrelated (0.68) and thus multicollinearity exists. Second, PLS is suitable for estimations from small sample sizes. Third, PLS is able to deal with bias in the responses (Cassel et al., 2000). The analysis and interpretation of a PLS model involves two stages (Tenenhaus, Vinzi, Chatelin & Lauro, 2005). First, reliability and validity of the measurement or outer model is assessed. The measurement model relates manifest variables to their respective latent variable. A latent variable is an unobservable variable that is measured by indicators (also called manifest variables) (Haenlein & Kaplan, 2004). The second stage involves the assessment of the relation between the latent variables, also called the structural or inner model.

3.4.1 Reliability Reliable measures are a prerequisite for drawing valuable conclusions about inter-construct relationships. According to Dun, Seaker and Waller (1994) reliability refers to the internal consistency of the manifest variables or indicators that are used to measure a specific latent construct. Reliability can be assessed by examining the composite reliability values. Indicators are considered to be internally consistent if this value exceeds a boarder of 0.70 (Nunally, 1978). Before assessing reliability, however, the unidimensionality of the measurement scale should be guaranteed. A scale is unidimensional if all measurement items

29

indeed estimate the construct they are supposed to estimate (Dunn et al., 1994). Unidimensionality is assessed by examining the individual item loadings that result from confirmatory factor analysis (CFA). Item loadings that score lower than 0.50 must be excluded from the scale to ensure unidimensionality.

3.4.2 Construct Validity Once a unidimensional scale has been established, construct validity can be assessed. Construct validity refers to the extent to which the manifest variables measure the latent construct that they are assigned to (Dunn et al., 1994). Convergent and discriminant validity are most frequently used to support construct validity. Convergent validity is established if there is agreement between various attempts in measuring one specific construct by using different methods. It is assessed by examining each constructs average variance extracted (AVE), which must exceed a value of 0.50. Discriminant validity refers to the extent to which the items of a scale measure distinct constructs. It is assessed by comparing the square root of the AVE of an individual construct to the correlation of that constructs with remaining constructs of the model. Discriminant validity is established if the square root of the AVE exceeds the respective correlations. The findings on reliability and validity of the conceptual model will be presented in the following chapter.

30

Chapter 4.

Analysis & Results

This chapter will outline the results of the empirical research. The collected data will be used for testing the hypotheses that have been developed. First, the results of the measurement model containing reliability and validity will be presented. Second, a summary of the results for the structural model will be given. Then, the reader is provided with some additional findings. The chapter ends with a brief conclusion on the results.

4.1
4.1.1

Results of the Measurement Model

Unidimensionality and Reliability

As has been outlined in the previous chapter, a unidimensional scale should be established before assessing reliability. Therefore, a CFA has been performed in order to obtain information about the individual items factor loading. These factor loadings are presented in Table 3. Only one item (LikelyC) has a factor loading lower than 0.50. Consequently, this item is deleted from the model. As all remaining items load higher than 0.50 on their respective factor, unidimensionality of the scale is established. After establishing a unidimensional scale, the internal consistency of the measurement items can be assessed. The examination of the composite reliability values in Table 4 shows that all values are above the 0.70 border. Hence, all items measuring the different latent constructs can be considered internally consistent. The unidimensionality and internal consistency of the measurement items support their overall reliability.

Table 3.

Descriptive Statistics on Item Level Items Mean 5.13 4.17 4.79 4.41 3.89 3.84 5.96 5.11 4.87 Std. dev 1.24 1.27 1.29 1.38 1.42 1.29 1.29 1.53 1.56 Loading 0.89 0.85 0.91 0.85 0.92 0.81 0.69 0.53 0.73 t-value 29.88 23.50 53.53 12.57 69.75 12.48 4.64 2.89 5.78 Skewness Kurtosis -0.50 -0.07 -0.45 -0.69 -0.07 -0.17 -1.70 -0.67 -0.48 0.49 -0.07 0.61 0.63 -0.22 0.01 3.71 -0.09 -0.07 SevereA SevereB SevereC PBCA PBCC* SelfpercA SelfpercB SelfpercC

Perception of Severity

Perceived

Behavioural PBCB Control SelfPerception

31

SelfpercD Importance ImportA ImportB Willingness WTPA WTPB WTPC Likelihood LikelyA LikelyB* LikelyC** LikelyD Notes: * Item recoded

4.75 4.35 4.71 4.44 4.76 4.41 4.16 4.23 3.69 3.50

1.67 1.23 1.12 1.52 1.11 1.41 1.52 1.65 1.63 1.47

0.63 0.84 0.91 0.90 0.70 0.88 0.94 0.57 0.58

4.05 12.57 28.91 38.77 6.73 27.90 54.27 3.43 4.27

-0.44 -0.49 -0.35 -0.60 -0.69 -0.29 -0.42 -0.14 0.09 0.20

-0.63 0.54 0.75 0.07 2.31 -0.05 -0.09 -0.52 -0.38 0.06

** Item omitted due to factor loading < 0.5

4.1.2

Construct Validity

To find support for construct validity, convergent and discriminant validity have to be established first. Convergent validity is given if the AVE of all individual constructs is above 0.50. The AVE values in Table 4 show that this condition is satisfied for all but one construct (Self-Perception). Following the argument from Dunn et al. (1994), however, convergent validity still exists since the t-values in Table 3 show that factor loadings from CFA are all statistically significant from zero. These t-values were obtained through a bootstrap procedure in SmartPLS (Ringle et al., 2005). Table 4 also displays the correlation between the different latent constructs. The square root of the AVE of each construct is displayed on the diagonal of the correlation matrix. Since this square root value exceeds each constructs correlation with all other construct, discriminant validity is assured. The proof of convergent and discriminant validity supports the existence of construct validity.

32

Table 4.

Descriptive Statistics on Factor Level (1) (2) (3) (4) (5) (6)

(1) Importance (2) Likelihood (3) PBC (4) Perception of Severity (5) Self-Perception (6) Willingness

0.88 0.42 0.68 0.53 0.16 0.60 0.87 0.77 0.72 0.46 0.43 0.27 0.60 0.75 0.52 0.86 0.60 0.29 0.57 0.90 0.74 0.89 0.18 0.58 0.92 0.78 0.65 0.39 0.74 0.42 0.83 0.87 0.69

Composite Reliability Average variance extracted

4.2

Results of the Structural Model

The following section will examine the hypothesized relationships between the different constructs of the conceptual model. The empirical results for the structural model are displayed in Table 5. A bootstrapping procedure with construct level changes has been performed to obtain the t-values of the path coefficients. The procedure contained 500 runs.

Table 5. Relationship

Results for Structural Model Coefficient Willingness 0.31 0.08 0.26 0.34 0.60 t-value 3.38 0.74 4.03 3.42 7.56 p-value < 0.001 > 0.05 < 0.001 < 0.001 < 0.001 Conclusion H1 supported H2 rejected H3 supported H4 supported H5 supported

Perception of Severity PBC Willingness Self-Perception Importance Willingness Note:

Willingness Likelihood

Willingness

The hypothesis is rejected if p > 0.05 As can be seen in the table above, 4 out of 5 hypothesized relationships appear to be statistically significant. The extent to which a person views CO2 emissions from air travel as severe for the ecology seems to have a positive influence on the willingness to compensate for those emissions ( = 0.31; t = 3.38). This result supports H1. The influence of a persons

33

perceived behavioural control on its willingness to compensate does not appear to be significant ( = 0.08; t = 0.74). As a result of this low but statistically insignificant relationship, H2 is rejected. The third relationship, namely between self-perception and willingness to compensate, is shown to be positive and significant ( = 0.26; t = 4.03). As this positive relationship was hypothesized, H3 is supported. The importance of adopting proenvironmental behaviour seems to have a positive and significant impact on an individuals willingness to compensate for CO2 emissions from air travel ( = 0.34; t = 3.42). This result shows support for H4. Hence, only perceived behavioural control does not appear to have an influence on a persons willingness to compensate. The strongest influence on the willingness to compensate comes from the importance of pro-environmental behaviour ( = 0.34), followed by perception of severity ( = 0.31) and self-perception ( = 0.26). Very importantly, a strong and positive relationship was found between the willingness and the likelihood to compensate ( = 0.60; t = 7.56). This finding is in line with the hypothesized relationship. Thus, H5 is supported. A summary of the empirical validation of the model is depicted in Figure 3. Figure 3 also displays the R squares of the dependent latent variables. The R squared value of 0.53 shows that the independent latent variables explain nearly 53 percent of the variation in the dependent latent variable willingness to compensate. The construct of willingness to compensate, in turn, explains approximately 36 percent of the variation in the latent variable likelihood to compensate.

34

Figure 3.

Empirically Validated Model

4.3

Additional Analyses & Findings

In addition to the analyses pertaining to the conceptual model, an analysis was conducted to identify differences with respect to demographic data. Also, a mediation analysis was performed to assess if willingness to compensate acts as a mediator of direct relationships between the four independent variables and likelihood to compensate. Furthermore, the monetary value that a respondent is willing to sacrifice for CO2 compensation has been estimated. The results of these additional analyses will be discussed below.

4.3.1

Demographics

By means of using the statistical software tool SPSS it was investigated whether the respondents answers to the measurement items of the constructs were moderated by demographic characteristics obtained from the online survey. Either an independent sample ttest or a one-way ANOVA test was used to relate gender, age, purpose of flying or place of living to the answers. Significant results were found for the construct of likelihood to compensate. More specifically, the variables explaining this construct appear to be moderated

35

by the usual purpose of flying (business vs. leisure). Nevertheless, since the proportion of business travellers (n = 21) was much lower than the proportion of leisure travellers (n = 107) we consider this result to be unreliable. All other results appeared to be insignificant.

4.3.2 Mediation Mediation occurs when a direct causal relationship between an independent variable and a dependent variable is intervened by another variable (Baron & Kenny, 1986). In this studys conceptual model the construct of willingness to compensate might be a mediator of direct relationships between the four independent constructs and the dependent construct of likelihood to compensate. Thus far we did not analyze a possible mediation effect as no direct relationships between the consumer-related factors and likelihood to compensate were assumed. To perform a mediation analysis, the initial model was expanded to account for these direct relationships. The mediated model is depicted in Figure 4.

Figure 4.

Mediated Model

According to Baron and Kenny (1986), mediation can either be complete or partial. The authors suggest four steps in establishing mediation. Complete mediation occurs if all of these steps are met. The four steps have been applied in this study by means of a bootstrap procedure in SmartPLS. First, the total effect of the four independent constructs on likelihood to compensate was estimated. The corresponding coefficients are displayed in table 6. Since these coefficients are nonzero it is established that there is an effect that may be mediated. Hence, the first step is met. The second step involves the estimation of the effect of each of

36

the four independent constructs on the mediator. These effects are equal to the coefficients in table 5. Since these coefficients are all different from zero, step 2 has also been met. Third, the direct effect between each independent constructs and likelihood to compensate was estimated. The corresponding values are displayed in table 7. To establish that willingness to compensate completely mediates the above relationships in the fourth step, the values of the direct effect found in step 3 should equal zero. However, the values displayed in table 7 are only reduced in absolute size compared to the total effect rather than being equal to zero. Thus, the fourth step has not been met. Following the argument from Baron and Kenny (1986), complete mediation could not be established. Thus, it seems that the construct of willingness to compensate only partially mediates the relationship between the four independent variables and the final outcome. The amount of mediation, also called the indirect effect, is equal to the difference between the total effect and direct effect (see table 6).

Table 6. Relationship

Mediation Analysis Total Effect Direct effect Likelihood 0.16 0.23 0.17 0.23 0.05 0.20 0.07 0.10 Mediation 0.11 0.03 0.10 0.13

Perception of Severity PBC Likelihood

Self-Perception Importance

Likelihood

Likelihood

4.3.3

Monetary Value

In the questionnaire, respondents were asked to indicate the maximum amount (in Euro) that they would be willing to sacrifice for CO2 compensation in air travel (Laroche et al, 2001; Brouwer et al., 2008). In this respect, a difference was made between short-haul (<3 hours) and long-haul flights (>6 hours). In analyzing the answers in Microsoft Excel, missing values were regarded as non-payment. Approximately 16% of the respondents reported not to be willing to pay extra for CO2 compensation for both types of flight. Among those respondents that would be willing to pay, the average reported amount for a short-haul flight was approximately 24. More specifically, the most frequently reported amount for a short-haul flight was 10 by 17% of the respondents. Approximately 12% of the respondents would be willing to sacrifice 5, 15% reported an amount of 20 and 11% would even be willing to pay 50 on top of the ticket price. For long-haul flights, the most frequent answer was 50 which 37

was given by approximately 23% of the respondents. An amount of 20 has been reported by approximately 11% of the respondents, 9% would be willing to sacrifice 30 and 12% is willing to give an amount of 100 for CO2 compensation. On average, respondents were willing to sacrifice about 55. These results show that the majority of the respondents (74%) would be willing to pay for CO2 compensation. Furthermore, the reported amounts were somewhat higher for a long-haul than for a short-haul flight.

4.4

Conclusion

In conclusion, the establishment of the models reliability and validity ensures that the constructs were measured as intended. The test of the hypotheses revealed that four out of five hypotheses were supported. Perception of severity, self-perception and importance appear to be positively related to willingness to compensate. Willingness to compensate appears to be strongly related to likelihood to compensate. Additional analysis has established partial mediation between the four independent constructs and likelihood to compensate. Also, meaningful information on monetary sacrifices was obtained. The following chapter will further discuss the results and their implications.

38

Chapter 5.

Discussion

The aim of this study was to assess the influence of consumer-related factors on peoples willingness to compensate for their CO2 emissions from air travel and the effect of this willingness on their likelihood to compensate in reality. The findings gave support to all but one of the hypothesized relationships. This chapter will discuss and interpret these findings by means of relating them to those from existing literature. It will start with a discussion on the findings pertaining to the conceptual framework which will then be followed by a discussion on additional findings.

5.1

Empirically Validated Model

A persons perception of the severity of air travels contribution to climate change was shown to have a significant and positive influence on willingness to compensate. In other words, the more a consumer agrees on the environmental harmfulness of CO2 emissions from air travel the more he or she is willing to pay for compensation. This positive influence supports the first hypothesis and is in line with findings from previous studies (Gssling et al., 2009; Hooper et al., 2008; Brouwer et al., 2008). It must be noted, however, that the most frequently reported answer to all three measurement items was agree (= 5) which implies a rather reserved attitude towards the relationship between air travel and climate change. A possible explanation is found in Gssling et al. (2009). According to these authors, it might be unclear how air travellers perceive aviations impact on the environment compared to the impact of other human activities. A direct comparison with other harming activities might favourably change peoples attitude and hence strengthen the positive relationship. The findings for the second hypothesis show a positive relationship between perceived behavioural control and willingness to compensate. Whereas a significant relationship was expected, as suggested in the Theory of Planned Behaviour (TPB) (Ajzen, 1991), the effect appears to be insignificant. Apparently, although air travellers seem to know about the negative impacts of flying on the environment they doubt the positive effect of their individual contribution to mitigating these impacts. The finding is not in line with previous research by Hooper et al. (2008) who found that passengers who believed that their individual actions limit the negative impact of aviation on climate change were very much more likely to participate in voluntary carbon offsetting. However, with the use of the word action

39

(Hooper et al., 2008, p. 49) in the measurement item the authors do not account for voluntary carbon offsetting in particular which might explain the discrepancy with the result of the present study. Van Birgelen et al. (2008) also found only partial support for the TPB. A fundamental reason for these contradictory findings could be the belief that every person should assume responsibility for mitigating climate change rather than behaving as a free rider (Brouwer et al., 2008). The results for the third hypotheses are in line with expectations. A positive and significant relationship was found between self-perception and willingness to compensate. This result implies that air travellers who behave ecologically in areas other than aviation (e.g. recycling) seem to be willing to transfer this behaviour to the aviation sector as well. This finding further supports the self-perception theory (Bem, 1967). Earlier support for this theory had been found by Van Birgelen et al. (2008). The finding stands in contrast to the one from Pickett et al. (1993). This contradiction typifies the ongoing debate about whether ecological behaviour is specific to a situation or activity (Pickett, Kangun & Grove, 1993; Bratt, 1999) or general across domains (Thgersen & lander, 2006; Thgersen, 2003). Encouragingly, the latter and more recent studies have found evidence for a tendency towards general ecological behaviour. Thus, it seems as if peoples minds are greening. Probably, the extent to which the adoption of certain ecological behaviour (e.g. using solar energy) has been eased over time in terms of availability and costs causes this change in mindset. Support was also found for the fourth hypothesis. Peoples perceived importance of behaving ecologically was found to have a positive effect on willingness to compensate. This finding implies that the more important individuals believe pro-environmental behaviour to be, the more willing they are to compensate CO2 emissions from air travel. This finding is in line with Laroche et al. (2001). It does not fully support, however, the study of McCarty and Shrum (1994) who found a positive but insignificant relationship between importance and recycling behaviour. This insignificant effect, however, was associated with a strong perception of the inconvenience of recycling behaviour which might explain the contradictory finding. Apparently, no matter how important recycling is perceived to be, its inconvenience discourages individuals from adopting recycling behaviour which might not be an obstacle in voluntary carbon offsetting.

40

The most striking result was found for the effect of willingness to compensate on likelihood to compensate. This relationship was found to be strong, positive and significant. The finding is particularly noteworthy as previous studies have not investigated this link. It supports Ajzens (1991) suggestion of a positive relationship between a behavioural intention (willingness to compensate) and actual behaviour (likelihood). Interestingly, incentives seem to have no effect on likelihood to compensate. When incentives are offered, the average level of likelihood to compensate does not increase.

5.2

Demographics

The findings showed that the various demographic characteristics did not moderate the respondents opinions about the constructs under investigation. Previous studies that have related demographic characteristics to the outcome found that the awareness about aviations impact on the environment differs between nationalities (Brouwer et al., 2008) or that willingness to pay for voluntary carbon offsetting differs between age groups (Hooper et al., 2008). Thus, the current findings partially differ from earlier ones. Given an unequal number of observations for nationality or age group in either study, however, results have to be interpreted with the necessary care.

5.3

Mediation

The mediation analysis showed that the willingness or intention to compensate for CO2 emissions from air travel partially mediates the direct relationships between the four consumer-related factors and likelihood to compensate. This result is in line with the findings in the TPB (Ajzen, 1991). Interestingly, the low amount of mediation (0.03) in the direct relationship between perceived behavioural control and likelihood to compensate seems to indicate that perceived behavioural control can be used in a direct fashion to predict actual compensation behaviour. This finding further supports the TPB. Overall, perceived behavioural control seems to act as a better predictor of likelihood to compensate ( = 0.20, p < 0.1) than of willingness to compensate ( = 0.08, p > 0.5). Apparently, air travellers who do not doubt that their individual CO2 compensation has a positive influence on the environment are very likely to actually and directly pay for this compensation.

41

5.4

Closed vs. Open-Ended Questions

The findings on likelihood to compensate are somewhat contradictory to the findings from the open-ended question concerning a maximum amount that a consumer is willing to pay for a short-haul or respectively long-haul flight. The scenario on likelihood to compensate included a payment amount of 25 for a long-haul flight. Accordingly, one would have expected respondents who have indicated a maximum amount of > 25 in the subsequent open-ended question to at least agree (5) with all measurement items. However, the average answer to every statement lies beneath 5, which indicates inconsistency between answers. A possible explanation is social desirability bias which might have forced respondents to portray themselves favourably with regard to CO2 compensation by reporting an arbitrary amount rather than reporting 0.

42

Chapter 6.

Conclusion

The last chapter of this thesis provides the reader with a conclusion of the research. It will start with a brief summary of the research. Then, theoretical and practical implications of the findings will be discussed. Last, it will provide limitations of the study as well as suggestions for future research.

6.1

Summary of the Research

This research was meant to provide new insights into what motivates air travellers to behave in a pro-environmental way by means of offsetting their CO2 emissions. Ajzens (1991) TPB served as a basis for a conceptual model. The intention was to find consumer-related factors which influence an air travellers willingness to pay for voluntary carbon offsetting in the first place. Factors that were taken into consideration were perception of severity, perceived behavioural control, self-perception and importance. This study differentiates from earlier ones in this research area since the factors under investigation were adapted from often-used theories on consumer behaviour. Furthermore, it was intended to establish the link between an air travellers willingness to compensate CO2 emissions and its likelihood of paying for this compensation in reality. The investigation of this link differentiates this study from previous ones which were only focused on explaining the willingness to pay for compensation. To test the hypothesized relationships data had been gathered by means of an online survey among adult air travellers. The analysis of the obtained data set revealed that three out of four consumer-related factors had a significant and positive influence on willingness to compensate. The strongest influence came from importance, followed closely by perception of severity and self-perception. Contrary to expectations, perceived behavioural control showed no significant relationship with willingness to compensate. The relationship between willingness to compensate and likelihood to compensate was shown to be strongly positive and significant. In the mediated model, a direct relationship between the four consumerrelated factors and likelihood to compensate is partially mediated by willingness to compensate. No meaningful differences were found with respect to demographic characteristics.

43

6.2

Theoretical and Managerial Implications

The findings of this study have several theoretical and managerial implications. From a theoretical side this study contributes to current academic literature in several ways. First, it gives further insights into voluntary carbon offsetting in aviation. More specifically, the research identified three consumer-related antecedents of the willingness to voluntarily pay for carbon offsetting. Second, by defining and testing a conceptual model based on the Theory of Planned Behaviour (Ajzen, 1991) this study has been first in establishing the link between the willingness or intention to compensate for CO2 emissions from air travel and the likelihood of actual compensation. Third, the findings of the study add to previous research on pro-environmental behaviour in that they have shown that well-known consumer-related factors are not only useful in predicting this kind of behaviour in product-oriented industries but also in service-oriented ones. From a practical side this study has an awareness-raising potential, which is twofold. First, it has the potential to raise awareness among air travellers about the detrimental impact of flying on climate change. Second, it can potentially increase awareness about the existence of voluntary carbon offsetting schemes in aviation as previous research has found that only a small share of air travellers knows about their existence. To further increase awareness, the management of airlines or airports offering voluntary carbon offsetting is urged to market this extra service more effectively. Managers should focus their marketing efforts towards people who consider ecological behaviour as being important in the first place since the findings showed that these people are most willing to compensate their CO2 emissions. In this way, the management can please those air travellers that are happy to redeem their guilt about flying. Particularly noteworthy is the finding that perceived behavioural control rather has a direct influence on likelihood to compensate than a mediated influence via willingness to compensate. Airlines and airports can take huge advantages of this finding. It implies that an individual with perceived behavioural control does not doubt the effectiveness of compensating CO2 emissions in air travel and is ready to behave environmentally friendly. A prerequisite of performing this behaviour, however, is information about CO2 compensation and the availability of resources (Ajzen, 1991). Therefore, airports and airlines should take care of better communication and more transparency about the credibility of their schemes, in particular about their way of calculating CO2 emissions. Furthermore, the necessary resources

44

for easing the service encounter with respect to voluntary carbon offsetting have to be in place. Resources are e.g. climate-counters or specific service desks. Encouragingly, the findings support Brouwer et al. (2008) who concluded that there is market potential for offsetting schemes. In line with their findings, the average amount that an air traveller is willing to pay for CO2 compensation on a short or long-haul flight appears to be above the amount calculated in Chapter 2 of this study. Most potential seems to come from the tourist segment.

6.3

Limitations and Suggestions for Future Research

This study has several limitations of which the most important ones will be named hereafter. First, our study is limited in that it mainly concentrates on the compensation of emissions that pollute the earth, rather than truly reducing them. By focusing only on CO2 emissions, this study disregards other harmful emissions that aviation releases to the atmosphere (e.g. nitrogen oxides). Future research should therefore focus on carbon dioxide equivalents (CO2e) to include all kinds of emissions. Also, a focus on the compensation of CO2 emissions from flying is by far not enough to save our planet. The aviation sector should rather try to impede a further growth in demand to truly reduce emissions. Previous research has found that a tiny share of air travellers would be willing to reduce or even to forego flights. Future researchers are therefore encouraged to investigate by which consumer-related factors this share is motivated. Also, the questionnaire did not give respondents the opportunity to state their opinion on who should be responsible for providing a solution to the problem. This opportunity should be included in future research. Another limitation of this study is the scenario-based method used for examining the likelihood that respondents would compensate CO2 emissions in reality. Even though respondents were asked to think of the scenario as if it were real, the self-reported nature of this method still limits a reliable reflection of actual behaviour due to social desirability bias. Future research should rather make use of, for instance, observation techniques to gather reliable data on actual behaviour. Furthermore, the scope of the study is quite limited with a sample size of 128 respondents. Future research should use a larger sample size to further validate the findings of this study. A 45

more representative sample should also be used for making valuable comparisons between demographic characteristics; in particular business vs. leisure travellers. Lastly, this study had a focus on passenger transport. However, a major part of greenhouse gas emissions results from full freighter flights. An interesting task for future research is to investigate if consumers or businesses would be willing to pay a price premium for compensating CO2 emissions from freight carried by air.

46

References
Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes, Vol. 50, pp. 179-122. kerman, J. (2005). Sustainable air transporton track in 2050. Transportation Research Part D., Vol. 10, pp. 111-126. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, Vol. 84, pp. 191-215. Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, Vol. 37, pp. 122-147. Banerjee, B. McKeage, K. (1994). How Green is my Value: Exploring the Relationship Between Environmentalism and Materialism. Advances in Consumer Research, Vol. 21. pp. 147-152 Baron, R.M., Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, Vol. 51, pp. 1173-1182. Becken, S. (2004). How Tourists and Tourism Experts Perceive Climate Change and Carbonoffsetting Schemes. Journal of Sustainable Tourism, Vol. 12, No. 4, pp. 332-345. Bem, D.J. (1967). Self-Perception: An Alternative Interpretation of Cognitive Dissonance Phenomena. Psychological Review, Vol. 74, No. 3, pp. 183-200. Boon, B.H., Schroten, A., Kampman, B. (2006). Compensation Schemes for Air Transport. CE Delft. Borup, M. (2007). Carbon compensation scheme for air mobility in Norway. Proceedings: Cases in Sustainable Consumption and Production: Workshop of the Sustainable

47

Consumption Research Exchange (SCORE!) Network, supported by EUs 6th Framework Program, Paris. Bratt, C. (1999). Consumers Environmental Behavior: Generalized, Sector-Based, or Compensatory? Environment and Behaviour, Vol. 31, No. 1, pp. 28-44 Brouwer, L. Brander, P. Van Breukering, P. (2008). A convenient truth: air travel passengers willingness to pay to offset their CO emissions. Climatic Change, Vol. 90, pp. 299-313. Cassel, M.C., Hackl, P., Westlund, A.H. (2000). On measurement of intangible assets: a study of robustness of partial least squares. Total Quality Management, Vol. 11, No. 7, pp. S897S907. Chapman, L. (2007). Transport and climate change: a review. Journal of Transport Geography, Vol. 15, pp. 354-367. Cleveland, M., Kalamas, M., Laroche, M. (2005). Shades of Green: Linking Environmental Locus of Control and Pro-Environmental Behaviors. Journal of Consumer Marketing, Vol. 22, No. 4, pp. 198-212. Dunn, S.C., Seaker, R.F., Waller, M.A. (1994). Latent Variables in Business Logistics Research: Scale Development and Validation. Journal of Business Logistics, Vol. 15, No. 2, pp. 145-172. European Union (2008). Question and Answers on the Revised EU Emissions Trading System. MEMO/08/796, European Union, Brussels Frew, E., Winter, C. (2008). Purchasing carbon offset flights in Australia: an exploration of airline websites. CAUTHE 2008 Conference. Gore, Al. (2006). An Inconvenient Truth. Directed by Davis Guggenheim. DVD. Paramount Home Entertainment, 2006.

48

Gssling, S., Broderick, J., Upham, P., Ceron, J.P., Dubois, G., Peeters, P., Strasdas, W. (2007). Voluntary Carbon Offsetting Schemes for Aviation: Efficiency, Credibility and Sustainable Tourism. Journal of Sustainable Tourism, Vol. 15, No. 3, pp. 223-248. Gssling, S. Haglund, L.Kallgren, H., Revahl, M. Hultman, J. (2009). Swedish air travellers and voluntary carbon offsets: towards the co-creation of environmental value. Current Issues in Tourism, Vol. 12, No.1, pp. 1-19. Haenlein, M., Kaplan, A.M. (2004). A Beginners Guide to Partial Least Squares Analysis. Understanding Statistics, Vol. 4, No. 3, pp. 283-297. Hooper, P., Daley, B., Preston, H. Thomas, C. (2008). An Assessment of the Potential of Carbon Offset Schemes to Mitigate the Climate Change Implications of Future Growth of UK Aviation. Centre for Air Transport and the Environment, Manchester. Intergovernmental Panel on Climate Change (1999). Aviation and the Global Atmosphere: A Special Report of IPCC Working Groups I and III. Cambridge University Press, Cambridge. Intergovernmental Panel on Climate Change (2007). Climate Change 2007: Synthesis Report. Intergovernmental on Climate Change. Kaiser, F.G., Wlfing, S., Fuhrer, U. (1999). Environmental Attitude and Ecological Behaviour. Journal of Environmental Psychology, Vol. 19, pp. 1-19. Kln-Bonn Airport (2008). Medien Information: Klima-neutral fliegen ab Kln/Bonn. Released by press office on 03.09.2008, Kln-Bonn Airport. Laroche, M. Bergero, J., Barbaro-Forleo, G. (2001). Targeting consumers who are willing to pay for more environmentally friendly products. Journal of Consumer Marketing, Vol.18, No. 6, pp. 503-520. MacKerron, G.J., Egerton, C. Gaskell, C., Parpia, A., Mourato, S. (2009). Willingness to pay for carbon offset certification and co-benefits among (high-)flying young adults in the UK. Energy Policy, Vol. 37, pp. 1372-1381. 49

McCarty, J.A., Shrum, L.J. (1994). The Recycling of Solid Wastes: Personal Values, Value Orientations, and Attitudes about Recycling as Antecedents of Recycling Behavior. Journal of Business Research, Vol. 30, pp. 53-62. McCarty, J.A., Shrum, L.J. (2001). The Influence of Individualism, Collectivism, and Locus of Control on Environmental Beliefs and Behavior. Journal of Public Policy and Marketing, Vol. 20, No. 1, pp. 93-104. Minton, A.P., Rose, R.L. (1997). The effects of Environmental Concern on Environmentally Friendly Consumer Behavior: An Exploratory Study. Journal of Business Research, Vol. 40, pp. 37-48 Nunally, J.C. (1978). Psychometric Theory (2nd ed.). McGraw-Hill, New York. Oreg, S., Katz-Gerro, T. (2006). Predicting Environmental Behaviour Cross-Nationally. Environment and Behaviour, Vol. 38, No. 4, pp. 462-483. Pickett, G. M., Kangun, N., Grove, S. J. (1993). Is there a general conserving consumer? A public policy concern. Journal of Public Policy and Marketing, Vol.12, pp. 234-243. Schmidt, M.J., Hollensen, S. (2006). Marketing Research: An International Approach. Pearson Education Limited, Harlow. Strasdas, W. (2007). Voluntary Offsetting of Flight Emission: An effective way to mitigate the environmental impacts of long-haul tourism? University of Salzburg, Austria. Taiyab, N. (2005). The market for voluntary carbon offsets: a new tool for sustainable development? Gatekeeper Series 121, International Institute for Environment and Development, London. Taiyab, N. (2006). Exploring the market for voluntary carbon offsets. International Institute for Environment and Development, London.

50

Tenenhaus M., Vinzi, V.E., Chatelin, Y-M., Lauro, C. (2005). PLS path modelling. Computational Statistics & Data Analysis, Vol. 48, pp. 159 205. Thgersen, J. (2004). A cognitive dissonance interpretation of consistencies and inconsistencies in environmentally responsible behavior. Journal of Environmental Psychology, Vol. 24, pp. 93-103 Thgersen, J., lander, F. (2006). To What Degree are Environmentally Beneficial Choices Reflective of a General Conservation Stance? Environment and Behaviour, Vol. 38, pp. 550569 Turner, A. (2009). Finding the Formula. Flight International, Vol. 175, No. 5177, pp. 29-32. Van Birgelen, M., Semeijn, J., Keicher, M. (2009). Packaging and Proenvironmental Consumption Behavior: Investigating Purchase and Disposal Decision for Beverages. Environment and Behavior, Vol. 41, No.1, 125-146. Vermeir, I., Verbeke, W. (2006). Sustainable Food Consumption: Exploring the Consumer Attitude-Behavioural Intention Gap. Journal of Agricultural and Environmental Ethics, Vol. 19, pp. 169-194. Ringle, C.M., Wende, S., Will, S. (2005). SmartPLS 2.0 (M3) Beta. Hamburg http://www.smartpls.de.

Internet
www.myclimate.org www.klm.com www.atmosfair.de www.ecopassenger.org www.nieuwsbank.nl www.unfccc.int

51

Appendices
Appendix A Comparison of CO2 Emissions from Different Transport Modes

52

Appendix B
Gender

Demographics

Frequency Valid Male Female Total 57 71 128

Percent 44,5 55,5 100,0

Valid Percent 44,5 55,5 100,0

Cumulative Percent 44,5 100,0

What is your gender?

80

60

Frequency

40

20

0 Male Female

What is your gender?

Age
Frequency Valid 16-24 25-34 35-49 50-64 65+ Total 37 44 28 18 1 128 Percent 28,9 34,4 21,9 14,1 ,8 100,0 Valid Percent 28,9 34,4 21,9 14,1 ,8 100,0 Cumulative Percent 28,9 63,3 85,2 99,2 100,0

53

What is your age?

40

30

Percent

20

10

0 16-24 25-34 35-49 50-64 65+

What is your age?

Usual Purpose of Flying

Frequency Valid Business Leisure Total 21 107 128 Percent 16,4 83,6 100,0 Valid Percent 16,4 83,6 100,0 Cumulative Percent 16,4 100,0

What is your usual purpose of flying?

100

80

Percent

60

40

20

0 Business Leisure

What is your usual purpose of flying?

54

Place of Living
Frequency Valid The Netherlands Germany Belgium Other European Country Non-European Country Total 54 41 19 12 2 128 Percent 42,2 32,0 14,8 9,4 1,6 100,0 Valid Percent 42,2 32,0 14,8 9,4 1,6 100,0 Cumulative Percent 42,2 74,2 89,1 98,4 100,0

Where do you live?

50

40

Percent

30

20

10

0 The Netherlands Germany Belgium Other European Country Non-European Country

Where do you live?

55

Appendix C

Questionnaire

RESEARCH QUESTIONNAIRE Dear respondent, My name is Pia Behrens and I am a Master student at Maastricht University. In order to obtain my Masters degree in Supply Chain Management, I need your help! I am currently conducting research for my Master Thesis for which the following questionnaire was designed. I would highly appreciate your help in completing this questionnaire. In return, Ill give you the chance to win 30, 20 or 10! If you wish to take part in the lottery, please enter your e-mail address at the end of the questionnaire. Completing the questionnaire will take approximately 8 minutes. Participation is anonymous and all information will be treated confidentially. Thank you very much in advance!! Pia --------------------------------------------------------------------------------------------------------------Climate Change The research for my Master Thesis is related to the current hot topic of climate change. In particular, my research intends to provide insight into the behaviour of consumers in relation to climate change. 1. Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a. I am concerned about the environment. b. I am aware of current environmental problems. c. I consider myself to be well informed about environmental problems. 1234567 1234567 1234567

2. Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a. I collect and recycle used paper. 1234567 56

b. I usually buy drinks in returnable bottles. c. I prefer a paper bag over a plastic bag when shopping.

1234567 1234567

d. When possible in nearby areas, I use public transportation or ride a bike. 1 2 3 4 5 6 7

Carbon Dioxide (CO2) Emissions from Air Travel Air travel is expected to grow rapidly over the next 20 years. Recently, increased attention has been paid to the impact of air travel on climate change. An aircraft produces CO2 emissions during a flight. 3. Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a. In my opinion, CO2 emissions from air travel have a serious negative impact on the environment. b. One of the major causes of environmental harm is CO2 emissions from air travel. c. I believe that CO2 emissions from air travel is a very important environmental issue. 1234567 1234567 1234567

Voluntary Compensation Schemes There are schemes that allow people to voluntarily compensate for the emissions of Carbon Dioxide (CO2) when taking a flight by paying an extra amount on top of the ticket price. This extra money will then be invested in projects, such as protecting or planting forests, which prevent or reduce CO2 emissions. 4. Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a. Compensating CO2 emissions from air travel will reduce pollution. b. Compensating CO2 emissions from air travel is important to save natural resources. 1234567 1234567

57

5. Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a. When I compensate CO2 emissions from air travel, I feel that I do something positive for the environment. b. I believe that my decision to compensate CO2 emissions from air travel has a direct influence on the environment as a whole. c. My choice to compensate CO2 emissions from air travel has no direct impact on the environment. 1234567 1234567 1234567

6. Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a) I am willing to compensate for CO2 emission from air travel to protect the environment. b) I am willing to accept cuts in standards for living to protect the environment. c) I am willing to pay higher (ticket) prices to protect the environment. 1234567 1234567 1234567

7. What is the maximum extra amount (in Euro) that you would be willing to pay for compensating CO2 emissions from air travel? a. ______Euro for a one-way short-haul flight (< 3 hours) b. ______Euro for a one-way long-haul flight (> 6 hours)

8. Please read the following scenario: Imagine you are flying alone (either business or leisure) from a European Airport to the United States. You have paid an average ticket price for this trip, and you are given the following information by the cabin crew:

Extra CO2 in the atmosphere is causing climate change, which could have serious and damaging effects worldwide, on people and the environment. Our airline takes all steps to minimise our contribution to this problem through using modern technology. However, all aircrafts still produce some CO2.

58

We now offer you the chance to offset these CO2 emissions by investing in projects which absorb CO2 or reduce the amount produced elsewhere. Your flight today has emitted approximately 1.000 kilograms of CO2 to the atmosphere per passenger. This is more than an average citizen emits per month!

You can offset your contribution to these emissions for 25

If you wish to offset your emissions, please contact a member of our cabin crew!

PLEASE NOTE: When answering the upcoming question, please try to think as if the situation described above were real - in other words, as if you would actually have to pay the amount of money stated, and this would mean less money available to spend on other things.

Please indicate your level of agreement/disagreement with the following statements: (1= totally disagree, 7 = totally agree) a. I would take up this offer as a leisure traveller. b. I would only take up this offer if part or all of it is paid by my employer. c. I would take up this offer if I get extra air miles in return. d. I would take up this offer if the airline offers me an extra free drink. 9. What is your gender? o Male o Female 1234567 1234567 1234567 1234567

59

10. What is your age? o 16-24 o 25-34 o 35-49 o 50-64 o 65+ 11. What is your usual purpose of flying? o Business o Leisure 12. Where do you live? o The Netherlands o Germany o Belgium o Other European country o Non-European country 13. If you wish to take part in the lottery (not mandatory), please enter your e-mail address: _______________________________________

Thanks a lot for your participation and support!!!

60

Appendix D

Overview of Measurement Items

Construct

Measurement Items In my opinion, CO2 emissions from air travel have a serious negative impact on the environment. One of the major causes of environmental harm is CO2 emissions from air travel. I believe that CO2 emission from air travel is a very important environmental issue. When I compensate CO2 emissions from air travel, I feel that I do something positive for the environment. I believe that my decision to compensate CO2 emissions from air travel has a direct influence on the environment as a whole. My choice to compensate CO2 emissions from air travel has no direct impact on the environment. I collect and recycle used paper.

Perception of Severity (adapted from Van Birgelen et al., 2008)

Perceived Behavioural Control (adapted from Van Birgelen et al., 2008)

Self-Perception (adapted from Kaiser et al., 1999)

I usually buy drinks in returnable bottles. I prefer a paper bag over a plastic bag when shopping. When possible in nearby areas, I use public transportation or ride a bike. Compensating CO2 emissions from air travel will reduce pollution. Compensating CO2 emissions from air travel is important to save natural resources. I am willing to compensate for CO2 emission from air travel to protect the environment. I am willing to accept cuts in standards for living to protect the environment.

Importance (adapted from McCarty & Shrum, 1994, 2001)

Willingness to Compensate (adapted from Oreg & Katz-Gerro, 2006)

61

I am willing to pay higher (ticket) prices to protect the environment. Likelihood to Compensate (adapted from MacKerron et al., 2009) I would take up this offer as a leisure traveller. I would only take up this offer if part or all of it is paid by my employer. I would take up this offer if I get extra air miles in return. I would take up this offer if the airline offers me an extra free drink.

62