Investigating the impacts of air travellers’ environmental knowledge on attitudes toward carbon offsetting and willingness to mitigate the environmental impacts of aviation

Transportation Research Part D 59 (2018) 96–107

Contents lists available at ScienceDirect

Transportation Research Part D journal homepage: www.elsevier.com/locate/trd

Investigating the impacts of air travellers’ environmental knowledge on attitudes toward carbon offsetting and willingness to mitigate the environmental impacts of aviation

T

⁎

Jin-Long Lu , Chiu-Yi Wang Department of Shipping and Transportation Management, National Kaohsiung Marine University, 142 Haijhuang Road, Nantzu District, Kaohsiung City 811, Taiwan, ROC

AR TI CLE I NF O

AB S T R A CT

Keywords: Air travellers Environmental knowledge Carbon offsetting Communication media

Several previous studies have explored air travellers’ willingness to compensate for their aviation carbon emissions or their willingness to pay for voluntary carbon offsetting programs. These studies have concluded that it is important to educate air travellers about the negative impacts of aviation on the environment. This study proposes two types of communication media and measures their effects on passengers’ knowledge about the environmental impacts of aviation. Then, travellers’ attitudes toward carbon offsetting and their willingness to offset their flights or change their travel behaviour are measured. Finally, a path model is estimated to analyse the effects of the various sources of knowledge on travellers’ knowledge about aviation and the environment, their attitudes toward the environmental impacts of aviation and their willingness to mitigate these impacts. The outcomes suggest that appropriate media could be adopted to enhance passengers’ knowledge of aviation impacts and the benefits of carbon offset programs, which could help passengers develop positive attitudes toward carbon offsetting and be more willing to offset their flights and change their travel behaviour. Not only does this study verify the relationships among knowledge, attitude, and behaviour (intention), it also has implications beyond academic research: it demonstrates that environmental education for air travellers should be prioritized.

1. Introduction Many airlines around the world have launched efforts to mitigate aviation emissions, including renovating their fleets to be more fuel efficient, optimizing route planning, improving operating procedures, reducing aircraft weight by using lightweight containers and cabin articles, and introducing market-based measures such as carbon offsetting. However, those environmental efforts initiated by airlines are not usually visible to passengers (Hagmann et al., 2015; Budianschi et al., 2012). In addition, passengers receive very little specific information on the environmental impacts of aviation, the concept of carbon offsetting, and the potential impact of travel behaviour change on reductions in carbon emissions. Dodds et al. (2008) found that no more than 16% of Canadian air travellers were aware of the concept of carbon offsetting. Another study by Lu and Shon (2012) indicated that only 6% of Taiwanese passengers knew the details of carbon offsetting schemes, and only approximately 5% believed that such schemes were effective in mitigating carbon emissions. A study by Cheung et al. (2015) investigating Australian travellers’ awareness of carbon offsetting found that less than 30% of respondents knew what carbon offsets were. Hagmann et al. (2015) revealed that approximately 32% of EU

⁎

Corresponding author. E-mail address: [email protected] (J.-L. Lu).

https://doi.org/10.1016/j.trd.2017.12.024

1361-9209/ © 2018 Elsevier Ltd. All rights reserved.

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

passengers, 84% of whom were German, had heard of carbon offsetting schemes but only 8% had ever used such schemes. Clearly, there is an information gap between the airline industry and air passengers regarding the environmental impacts of air travel and the actions that can be taken to mitigate these impacts. Therefore, several studies have suggested that such information should be provided to passengers and that it should efficiently improve air passengers’ awareness. Then, air passengers will be more likely to pay more for the environmental costs of their travel and/or to travel in a more environmentally friendly manner. For example, Brouwer et al. (2008) indicated that Asian travellers’ low willingness to pay for carbon offsets was due to their lack of understanding of aviation impacts on the environment. Lu and Shon (2012) also concluded that if passengers knew nothing about carbon offset schemes, their willingness to pay for such offsets would be significantly reduced. Furthermore, Kim et al. (2014) implied that air passengers’ environmental knowledge positively and significantly influenced their attitudes toward voluntary carbon offsetting. Jou and Chen (2015) concluded that passengers’ willingness to pay for carbon offsets can be enhanced by explaining the purpose and content of carbon offset policies. An investigation by Cliff (2014) also found that less than 9% of UK passengers had ever carbon-offset their flights, and one reason for such low engagement was a lack of information on – and understanding of – the concept of carbon offsetting. However, the studies cited above did not fully provide air passengers with information on aviation emissions, its impacts on the environment, and the concept of carbon offsetting prior to measuring or exploring their willingness to pay for carbon offsets. Those studies might neglect to offer the information or just want to investigate the original opinions of passengers; anyhow, the analyses were conducted under the condition of low knowledge levels and therefore, resulted in lower concerns and lower willingness to act on the problem. Hence, would air passengers’ attitudes toward carbon offsetting and their willingness to participate in an offset programme – or even change their travel behaviour – be further enhanced if they were well informed regarding aviation impacts on the environment and the idea behind carbon offsetting? The main purpose of this study is to develop media tools that can be used to provide passengers with basic knowledge of aviation impacts (on the environment) as well as the idea of carbon offsetting. This study also aims to measure the attitudes and intentions of air passengers once they know more about the issues (or after they have been educated). Influenced by existing programmes regarding cabin safety education, this study develops two types of media, a card and a video, for briefing passengers on the environmental impacts of aviation and corresponding strategies for carbon mitigation. Then, we investigate air passengers’ attitudes toward the carbon offset programmes and their willingness to participate in carbon offsetting or to change their travel behaviour to mitigate carbon emissions. Finally, a cause-effect model analysed by path analysis is used to explore the relationships among the knowledge, attitudes, and willingness of the participants. Taiwanese travellers who have ever travelled outbound at least once in the past two years are chosen as the population for our case study. 2. Research design 2.1. Developing the communication media Briefing cards and videos are two media commonly used for airline cabin safety education. This study adopts the same strategy of designing a card and a video to introduce passengers to the environmental impacts of air transportation and potential strategies for carbon mitigation. The structure of the two media are basically the same: first, they introduce the growing trend of air transportation and its possible environmental impacts in terms of carbon emissions; then, they explain the theory and benefits of voluntary carbon offsetting programmes; finally, the media suggest an alternative way to reduce aviation carbon emissions, namely, by changing travel behaviours. The sources of the content of these educational media are primarily the official websites of the International Air Transport Association (IATA) (IATA, 2015), the International Civil Aviation Organization (ICAO) (ICAO, 2015), and several research papers (e.g., Lu and Shon, 2012; Jou and Chen, 2015; Kim et al., 2014; Araghi et al., 2014). Before being released to interviewees, the content of the two media was reviewed by three researchers in Taiwan and one researcher in Australia who are familiar with issues related to aviation and the environment and the concept of carbon offsetting. Furthermore, the two media were sent to approximately 15 people via the authors’ personal connections, and those individuals’ perceptions of the content of the two media were collected for further revision. Hence, the content validity of the two media can be ensured. Fig. 1 shows the content of the printed card. The full content of the other media – a video with both a traditional Chinese version and an English version – can be watched on YouTube.com.1 (Note that both the card and video have traditional Chinese and English versions, but only the traditional Chinese version was used for the survey in Taiwan.) 2.2. Measurements and pre-test survey 21 measurements are defined for empirical analysis. The details are listed in Table 1. The first six measurements (M1-M6) are used to measure passengers’ knowledge of aviation impacts on the environment and the concept of carbon offsetting. The seventh through 12th measurements (M7-M12) are associated with passengers’ attitudes toward voluntary carbon offset schemes. The 13th through 16th measurements (M13-M16) are related to passengers’ willingness to voluntarily offset their flights. The last five (M17-M21) measurements correspond to passengers' willingness to change their travel behaviours. To measure passengers’ knowledge levels, a 1 Traditional Chinese version: http://youtu.be/-B2wgztowMs; English version: http://youtu.be/lkXIoHg4fQU; or search for “Voluntary carbon offset programme” on Youtube.com.

97

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Fig. 1. Information printed card.

five-point Likert scale is adopted and ranges from “I know/can explain the issue with less than 10% confidence (=1)”, “I know/can explain the issue with approximately 30% confidence (=2)”, “I know/can explain the issue with 50–50 confidence (=3)”, “I know/ can explain the issue with approximately 70% confidence (=4)”, and “I know/can explain the issue with more than 90% confidence (=5)”. For the other measurements, the five-point Likert scale is set from “Strongly disagree with the statement (=1)” to “Strongly agree with the statement (=5)”. Prior to the formal survey, a pre-test survey was conducted in the departure lounge of the Kaohsiung International Airport (KHH) in southern Taiwan. This pre-test survey primarily checked the wording and semantics of the measurements and hence, did not provide any information about the environmental impacts of air transportation, nor did it introduce the voluntary carbon offset programme before the survey. In other words, the pre-test survey was based only on what the passengers may have already known, 98

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Table 1 Measurements. Measurements

Referred sources

Related to knowledge about the environmental impacts of aviation and the concept of the voluntary carbon offset programmes M1: I can explain the environmental impacts caused by air transportation Kim et al. (2014); IATA (2008), van Birgelen et al. (2011), M2: I know what channel can be accessed to calculate carbon emissions from air travel Davison et al. (2014) M3: I know that the class of seat is an important factor to calculate the carbon emissions M4: I can explain the voluntary carbon offset programme and its benefits M5: I know what channel can be used to purchase carbon offsets M6: I can explain where the voluntary carbon offset funds go and how the voluntary carbon offset funds will be used for Related to attitudes toward carbon offsetting programme M7: Purchasing offsets to support the voluntary carbon offset programme benefits to humankind M8: Purchasing offsets to support the voluntary carbon offset programme is smart M9: Purchasing offsets to support the voluntary carbon offset programme is valuable M10: Purchasing offsets to mitigate the carbon emissions is the least we could do for our environment M11: Carbon offsetting air travel is important to our environment M12: Carbon offsetting air travel has a positive impact on our environment Related to the willingness to participate in the voluntary carbon offset programme M13: I am willing to pay the offsets to support the voluntary carbon offset programme when travelling by air M14: I am willing to encourage people to support the voluntary carbon offset programme when travelling by air M15: I am willing to support the carbon tax policy on passengers’ duty M16: I will recommend airlines to implement the voluntary carbon offset programme through customer service centre Related to the willingness to change travel behaviour M17: I am willing to use telephone meeting or videoconferencing to replace business air travel when possible M18: If travelling by air is necessary, I am willing to reduce my baggage weight to reduce carbon emissions M19: I am willing to fly less frequently to reduce environmental impacts M20: I am willing to pay extra fees to take a low-carbon-emission flight M21: I will plan my vacation to local places to avoid travelling by air

Kim et al. (2014), Chen (2013), van Birgelen et al. (2011)

Chen (2013), Davison et al. (2014), van Birgelen et al. (2011)

Davison et al. (2014), van Birgelen et al. (2011), Araghi et al. (2014)

vaguely known, or did not know about the environmental impacts of aviation and the concept of carbon offsetting. The total number of interviewees in the pre-test survey sample was 59. In this sample, the ratio of males to females was approximately 4–6; up to 80% of the pre-test sample were between 20 and 40 years old; over 60% of the pre-test sample work in business-related sectors; approximately 50% of the pre-test sample had a monthly income between NT $25,000 (New Taiwan Dollar) and NT $55,000 (NT $1 is approximately equal to US $0.33); and 80% of the pre-test sample travel for leisure. Finally, approximately 53% of the pre-test sample had never heard of the voluntary carbon offset programme; 34% of the sample had vaguely known of the programme; and 13% of the sample had known about the programme. After the preliminary analysis, two of the measurements were found to be inadequate due to their low correlations with the other measurements and a failure to form a meaningful construct; therefore, they were removed from the formal survey. These include the tenth measurement, ‘Purchasing offsets to mitigate carbon emissions is the least we could do for our environment’ (M10), and the 20th, ‘I am willing to pay extra fees to take a low-carbon-emission flight’ (M20). Finally, 19 measurements were used for the formal survey. 2.3. Questionnaire and survey In addition to the measurements mentioned above, respondents’ experiences of their most recent outbound air travel, including destination, travel purpose, and airlines used, as well as personal characteristics such as gender, age, monthly income, and frequency of travel abroad were also incorporated into the survey questionnaire. The formal survey was conducted via the Internet with the professional assistance of an online survey company in Taiwan, Pollster.com. Pollster.com is a member of the All Access Marketing Solution Group and has more than 150,000 registered members in the Taiwan area. According to Pollster.com, approximately 60% of the registered members are female; 38% are in the age range of 21–30 years old; 39% are 31–40 years old; 14% are 41–50 years old; and finally, approximately 80% of the registered members live in Taiwan’s five major cities. The primary reason for using an Internet survey is that the two media developed in this study can be easily and effectively embedded in the web interface, and respondents can easily review the questionnaire – including reading the card or watching the video – step by step on their own devices (PCs, pads, or mobile phones), then they can instantly respond to the survey. One limitation of an online survey is that the respondents were not actual air passengers at the airports but they did have flying experience in the past two years before the survey. We believe that the opinions of these respondents should be somehow representative for the analysis. The survey starts with questions regarding the respondents’ latest air travel experiences. The next question asks how much the 99

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

respondents know about voluntary carbon offset programmes. Then, one (and only one) of the two communication media was randomly assigned to the respondent. Respondents assigned to the card were required to read the information shown on the card for at least 20 s. This is to ensure that they paid attention to the information disclosed in the card. The respondents presented with the video could not proceed to the next part of the survey until the video ended (approximately 2 min and 35 s). After reading the information on the card or watching the video, six statements were used to measure the respondents’ degree of knowledge regarding the environmental impacts of aviation and the concept of carbon offsetting. This was followed by measurements regarding attitudes toward voluntary carbon offset (VCO) programmes and the willingness to carbon offset or change travel behaviours. The online survey questionnaire was sent to approximately 1200 registered members randomly selected from the database, screened by the condition that the members have had at least once outbound travel experience in the last two years (more than 10,000 members met this requirement). The survey began on April 15, 2015, and 627 samples were received after two weeks. The sampling error was ± 3.9% with a confidence level of 95%. However, some participants were found to have experience only on domestic flights; therefore, only 552 samples were retained for analysis. Among the sample, 269 received the card briefing while 283 received the video briefing. The time that each respondent spent on the survey was also recorded. Respondents who received the card briefing spent, on average, 6.94 min on the survey; the minimum time was 4.2 min, while the maximum time was 22.5 min. However, more than 75% of the card samples spent 4 to 7 min on the survey; approximately 99% of the respondents spent less than 16 min on the survey. Only one of the card samples used more than 22 min on the survey. As for the respondents who received the video briefing, the minimum time for the survey was 4.7 min and the maximum time was 19.2 min. The mean time spent on the survey was 6.90 min. Moreover, over 70% of the video sample spent 4–7 min on the survey; 95% of the respondents spent less than 14 min on the survey. Only two of the video samples spent more than 18 min on the survey. Hence, it can be said that both groups of respondents spent almost the same time on their surveys even though they received two different types of media, and each medium had a different time requirement for the briefing. In other words, the time that the respondents spent on the survey might not have moderate effects on the conveyance of knowledge. Table 2 further the profile of the respondents and their latest experiences of outbound travel. Differences between the proportions of the card and video briefing groups in the sample are tested in each category. According to Table 2, between the two media groups, only the proportion of male (or female) respondents, the proportion of respondents between 31 and 40 years old, and the proportion of respondents with a monthly income below NT $25,000 are significantly different at an α-level of 0.05. The distributions of the two media samples across various personal backgrounds and travel experiences are somewhat similar. The results of the respondents’ Table 2 Profile of respondents and their last outbound travel by media (%). Variable

Card (N = 269)

Video (N = 283)

Significance of the difference

Gender

Male Female

38.3 61.7

47.3 52.7

0.035* 0.035*

Age

21–30 31–40 41–50 51–60 > 61

39.4 36.4 16.7 4.8 2.6

32.9 48.1 13.1 4.6 1.4

0.127 0.006* 0.398 0.963 0.778

Monthly income (NT $1000)

≤25 25–40 40–55 55–70 70–85 85–100 > 100

29.0 37.5 17.8 10.8 1.5 1.5 1.9

19.8 37.5 23.7 11.3 3.9 2.8 1.1

0.031* 1.000 0.166 0.907 0.573 0.760 0.851

Outbound travel frequency (times per year)

<1 1–3 3–5 >5

60.3 32.7 5.0 2.0

53.7 37.8 6.0 2.5

0.122 0.981 0.814 0.907

Destination of the trip

Mainland China Hong Kong or Macau South East Asia Korea and Japan Outside Asia

9.7 16.4 11.5 53.9 8.6

13.4 15.2 14.5 48.4 8.5

0.385 0.778 0.481 0.196 0.981

Purpose of travel

Business Non-business

21.9 78.1

27.6 72.5

0.181 0.189

Used Airlines

Taiwanese airlines Hong Kong or Macau based airlines Mainland China airlines Low-fare airlines Others

71.7 10.4 3.3 6.3 8.2

69.6 10.2 3.2 9.2 7.8

0.622 0.963 0.981 0.496 0.925

* p < .05.

100

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

prior knowledge about the voluntary carbon offset (VCO) scheme revealed that less than 5% and 3% of the two media groups (card and video, respectively) clearly knew about VCO programmes; approximately 40% of the respondents vaguely knew about VCO programmes; and 55% of the card sample and 57% of the video sample knew nothing about VCO programmes. 2.4. Knowledge-attitude-behaviour (KAB) theory The relationships among knowledge, attitude, and behaviour have been verified in various fields. Gofin et al. (1990) performed one of the early studies discussing how mothers’ knowledge level of car safety affected their attitudes toward car seat belt use and the intentions to use a car seat for newborns. Their study concluded that positive and significant relationships existed between knowledge, attitude, and behaviour intention. Nilsson and Küller (2000) found that people’s knowledge about environmental impacts was positively correlated with an attitude of environmental concern, and both knowledge and attitudes were positively correlated with pro-environmental travel behaviour (i.e., less driving distances, less use of private transport, and the acceptance of traffic restrictions). However, attitudes toward environmental concerns were more potent than factual knowledge for promoting pro-environmental travel behaviour. Flamm (2009) also used the knowledge-attitude-behaviour (KAB) model and verified that households with pro-environmental attitudes had higher levels of knowledge of the environmental impacts of vehicle ownership and use and owned fewer and more fuelefficient vehicles, drove them less often, and consequently consumed less fuel. Another study by O’Connor et al. (2002) also found that cognitive components (i.e., understanding the causes of climate change and bad consequences from climate change) were the most powerful antecedent factors to support reducing greenhouse gas emissions. In the study of Walton et al. (2004), significant and positive correlations among environmental concerns, knowledge of emissions, and contribution to an environmental organization (behaviour) were found, but they were small associations. Moreover, the results from Lera-López et al. (2014) indicate that younger, better educated, and more environmentally-aware citizens were willing to pay more to reduce air and noise pollution caused by road transportation. Therefore, these authors suggested that developing effective communication policies to educate the population and stimulate their concerns over environmental problems was important. Regarding educating or conveying knowledge to the public, Liao (2014) offered a cabin safety education programme to elementary school students and applied the KAB model to the evaluation of the knowledge, attitudes, and behavioural intentions regarding cabin safety. She found that the students’ knowledge, attitudes and behavioural intentions regarding cabin safety were significantly improved after receiving the safety education. To clarify, increasing knowledge can be considered as a primary factor to influence an individual’s attitude and behaviour (intention). Waygood and Avineri (2016) provided four different formats for information regarding CO2 emissions to determine how these different formats would influence an individual’s car choice: mass, treeequivalent, earth-equivalent, and s carbon budget. They implied that the carbon budget information was the most efficient format for communicating CO2 emissions and motivating a behavioural change (i.e., reduce emissions). Furthermore, the results from Waygood and Avineri (2016) indicated that gender did not have a significant impact on an individual’s behavioural response associated with receiving different formats of information regarding CO2 emissions. However, women were more likely to report having engaged in an action to reduce climate change impact. KAB theory has been widely applied in various fields and the relationships among the three main constructs have been successfully identified. Several studies suggested that if useful information regarding CO2 emissions, the environmental impact of transportation, or transport safety was presented using an efficient means, it would improve an individual’s knowledge and therefore, result in a change in attitude and transport behaviour. We utilise the KAB model as the main framework for our analysis. Because the actual behaviour of offsetting flights cannot be observed in Taiwan, behavioural intention, an indicator of future behaviour, is used in the model analysis. 3. Findings 3.1. Effects of knowledge briefing Table 3 shows a comparison of the three groups: the pre-test survey sample, the card briefing sample, and the video briefing sample. Regarding the measurements related to knowledge, the scores on the first six measurements obtained from the pre-test survey sample were below 2.5, with the exception of M1 (i.e., I can explain the environmental impacts caused by air transportation); while the scores corresponding to the same measurements were between 2.9 and 3.4 and between 3.2 and 3.8 for the card and video samples, respectively. Hence, using a video as a tool for knowledge briefing could effectively improve air passengers’ knowledge of the environmental impacts of aviation and the concept of carbon offsetting by at least 10% compared with a card briefing and by at least 50% compared with the sample that received no knowledge briefing. Moreover, with the exception of M1, the mean differences of knowledge between any two of the three groups were all significant at the α-level of 0.01. This clearly indicates that people’s knowledge about aviation impacts on the environment and the concept of carbon offsetting would be enhanced by explaining the consequences of the rapid growth of air travel and possible carbon mitigation strategies. However, respondents in the card briefing sample reported attitudes toward carbon offsetting (M7 to M12) that were not significantly different from the pre-test sample. For the video sample, only M7 (i.e., Purchasing offsets to support the voluntary carbon offset programme benefits humankind), M11 (i.e., Carbon offsetting air travel is important to our environment), and M12 (i.e., Carbon offsetting air travel has a positive impact on our environment) differed from the pre-test sample. Similarly, for both the card and video samples, only M13 (i.e., I am willing to pay offsets to support the voluntary carbon offset programme when travelling by 101

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Table 3 Descriptive statistics and difference test among the three samples. Item

Pre-test sample μp

Card sample μc

Video sample μv

Significance of the diff. H0: μp = μc

Significance of the diff. H0: μp = μv

Significance of the diff. H0: μc = μv

M1 M2 M3 M4 M5 M6 M7 M8 M9 M11 M12 M13 M14 M15 M16 M17 M18 M19 M21

3.373 2.170 2.170 2.356 2.170 2.203 3.644 3.695 3.746 3.458 3.559 3.322 3.271 3.237 3.153 3.424 3.441 3.136 2.848

3.428 2.996 2.900 3.164 3.033 2.996 3.729 3.766 3.903 3.636 3.680 3.387 3.513 3.372 3.227 3.818 4.004 3.680 3.383

3.873 3.413 3.194 3.463 3.385 3.410 3.859 3.820 3.901 3.693 3.788 3.562 3.597 3.424 3.371 3.859 3.986 3.682 3.435

0.708 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.475 0.558 0.188 0.200 0.359 0.673 0.067* 0.395 0.616 0.003*** 0.000*** 0.000*** 0.001***

0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.074* 0.298 0.179 0.076* 0.068* 0.094* 0.010** 0.232 0.128 0.001*** 0.000*** 0.000*** 0.000***

0.000*** 0.000*** 0.003*** 0.001*** 0.000*** 0.000*** 0.077* 0.473 0.975 0.490 0.171 0.054* 0.290 0.586 0.103 0.577 0.809 0.984 0.571

Note: M10 and M20 were removed from the final survey. * p < .1. ** p < .05. *** p < .01.

air) and M14 (i.e., I am willing to encourage people to support the voluntary carbon offset programme when travelling by air) were associated with passengers’ willingness to participate in the VCO programme and had significant differences from the pre-test sample. With respect to the measurements related to people’s willingness to change their travel behaviour, the mean scores of M17 (i.e., I am willing to use telephone meeting or videoconferencing to replace business air travel when possible), M18 (If travelling by air is necessary, I am willing to reduce my baggage weight to reduce carbon emissions), M19 (i.e., I am willing to fly less to reduce environmental impacts), and M21 (i.e., I will plan my vacation to local places to avoid travelling by air) significantly differed between the pre-test and card samples as well as between the pre-test and video samples. 3.2. Factor analysis Although the measurements had been confirmed by several previous studies and further identified through a preliminary analysis using pre-test data in this study, the latent factors determining these measurements could be a bit inconsistent with the proposed structure. Hence, factor analysis was used to abstract meaningful factors. The determination of the number of factors is based on whether the eigenvalue of each factor is greater than one. Table 4 suggests a factor analysis solution after applying the orthogonal varimax rotation technique. M11 (i.e., Carbon offsetting air travel is important to our environment), M14 (i.e., I am willing to encourage people to support the voluntary carbon offset programme when travelling by air), and M21 (i.e., I will plan my vacation to local places to avoid travelling by air) were eliminated due to their low factor loadings. Finally, four factors were abstracted and the accumulated explained variance of these factors reached 76% of the total variance. According to the meanings of the variables (measurements) related to each factor, the first factor refers to respondents’ knowledge about the aviation impact on the environment and the carbon offset scheme (i.e. what it is, how it work, and how to join it, etc.); hence, the first factor was named knowledge about aviation impact and carbon offset schemes (KAO). The second factor implies respondents’ attitude toward participating in the voluntary carbon offset programme (i.e. is it smart, valuable, beneficial, and so on). It was named attitudes toward the VCO programme (ATO). Regarding the third factor, the meanings of the associated measurement indicate respondents’ willingness to voluntarily buy the offsets, to support the VCO scheme, and to recommend the VCO scheme to airlines; therefore, it was named willingness to offset flights (WTO). The last factor presents respondents’ willingness to use alternative ways except for air travel, to adjust travel frequency of air travel, and to reduce baggage weight if taking air travel. This factor was named willingness to change behaviour (WTC). Regarding the reliability coefficients, the Cronbach’s α of each factor exceeded 0.7, indicating good reliability of all four factors. 3.3. Model analysis and results Next, we used these four factors as the primary variables for the model analysis. The structure of the model, illustrated in Fig. 2, is primarily based on the KAB theory (behavioural intention) and the conclusions reported in relevant studies, e.g., Flamm (2009), Kim et al. (2014), Liao (2014), Nilsson and Küller (2000), and Waygood and Avineri (2016). In the model, WTO and WTC represent behavioural intentions and are the dependent variables. KAO and ATO are the explanatory variables for both WTO and WTC. The level of KAO is also presumed to have an impact on ATO. It is further assumed that the error terms of the two dependent variables 102

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Table 4 Results of factor analysis. Factor for interpretation (% variance explained)

Loading

Measurements

Reliability (Cronbach’s α)

Knowledge about aviation impact and carbon offset schemes (KAO) (27.34%)

0.757

M1: I can explain the environmental impacts caused by air transportation M2: I know what channel can be accessed to calculate carbon emissions from air travel M3: I know that the class of seat is an important factor to calculate the carbon emissions M4: I can explain the voluntary carbon offset programme and its benefits M5: I know what channel can be used to purchase carbon offsets M6: I can explain where the voluntary carbon offset funds go and how the voluntary carbon offset funds will be used for

0.920

M7: Purchasing offsets to support the voluntary carbon offset programme benefits to humankind M8: Purchasing offsets to support the voluntary carbon offset programme is smart M9: Purchasing offsets to support the voluntary carbon offset programme is valuable M12: Carbon offsetting air travel has a positive impact on our environment

0.915

M13: I am willing to buy the offsets to support the voluntary carbon offset programme when travelling by air M15: I am willing to support the carbon tax policy on passengers’ duty M16: I will recommend airlines to implement the voluntary carbon offset programme through customer service centre

0.875

M17: I am willing to use telephone meeting or videoconferencing to replace business air travel when possible M18: If travelling by air is necessary, I am willing to reduce my baggage weight to reduce carbon emissions M19: I am willing to fly less to reduce environmental impacts

0.775

0.843 0.806 0.860 0.846 0.833 Attitude toward the VCO programme (ATO) (22.42%)

0.824 0.832 0.840 0.788

Willingness to offset flights (WTO) (13.63%)

0.612 0.739 0.744

Willingness to change behaviour (WTC) (12.13%)

0.780 0.774 0.785

Fig. 2. Proposed model framework.

103

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Table 5 Definitions of variables (n = 611). Variables

Mean

Std.

KAO = average of the summation scores of the associated measurements (i.e. M1 to M6) ATO = average of the summation scores of the associated measurements (i.e. M7 to M9, M12) WTO = average of the summation scores of the associated measurements (i.e. M13, M15, M16) WTC = average of the summation scores of the associated measurements (i.e. M17 to M19) Card briefing = 1 if the respondent received card briefing; = 0 otherwise Video briefing = 1 if the respondent received video briefing; = 0 otherwise Business = 1 if the respondent travels for business purpose; = 0 otherwise FREQ1 = 1 if the respondent travels by air less 1 time per year; = 0 otherwise FREQ2 = 1 if the respondent travels by air 1–3 times per year; = 0 otherwise FREQ3 = 1 if the respondent travels by air 3–5 times per year; = 0 otherwise FREQ4 = 1 if the respondent travels by air over 5 times per year; = 0 otherwise AGE1 = 1 if the respondent ages 21–30 years old; = 0 otherwise AGE2 = 1 if the respondent ages 31–40 years old; = 0 otherwise AGE3 = 1 if the respondent ages 41–50 years old; = 0 otherwise AGE4 = 1 if the respondent ages more than 50 years old; = 0 otherwise INC1 = 1 if the respondent earns monthly income no more than NT $40,000; = 0 otherwise INC2 = 1 if the respondent earns monthly income in between NT $40,001 and NT $75,000; = 0 otherwise INC3 = 1 if the respondent earns monthly income more than NT $75,000; = 0 otherwise E-Meeting = 1 if the respondent has ever used tele- or video-conferencing; = 0 otherwise PRIOR = 1 if the respondent has ever heard or already known the concept of carbon offsetting; = 0 otherwise

3.192 3.793 3.377 3.789

0.965 0.767 0.945 0.767

0.440

–

0.463

–

0.242

–

0.522

–

0.379

–

0.067

–

0.098

–

0.378

–

0.416

–

0.146

–

0.061

–

0.619

–

0.314

–

0.067

–

0.201

–

0.457

–

might have a positive correlation. This hypothesis is partially motivated by the results from Hinnen et al. (2015), which suggests that travellers who reported a high willingness to pay for carbon offsets (and other green products during air travel) also valued the consumption of fair-trade products or products from the regions where they live. In addition, Mair (2011) indicated that purchasers of voluntary carbon offsets (VCO) had ecocentric attitudes and their socioeconomic profiles differed from others engaging in proenvironmental behaviour. Therefore, passengers who are willing to offset their flights would also be willing to change their travel behaviour, such as by reducing their baggage weight or reducing their frequency of air travel, and vice versa. Moreover, based on the conclusions of Section 3.1, using different media for the briefings is hypothesised to have various impacts on the level of KAO. Other variables, such as trip purpose, gender, age, travel experience, are also considered as potential explanatory variables in the model estimation. Table 5 lists the variables used in the model estimation and their definitions. The values of the four factors were defined by calculating the average of the summation scores of all the associated measurements (i.e., mean score values). Other variables were set as dummies. Here, all the samples were pooled together for model estimation. A path analysis model was estimated using the 104

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Table 6 Estimated results of the model. Dep. Var.

KAO

Variables

Coeff.

Constant KAO ATO Card briefing Video briefing Business AGE1 AGE2 AGE4 FREQ1 FREQ3 INC3 E-Meeting PRIOR Covariance R2 Overall R2 χ2 (df) RMSEA CFI TLI SRMR

ATO t-stat.

***

8.391

0.680*** 1.065***

5.340 8.393

2.203

0.430

***

0.154 0.272 27.98 (26) 0.011 0.998 0.996 0.017

5.958

WTO

Coeff.

WTC

t-stat.

Coeff.

t-stat.

2.760 0.307***

28.292 10.206

−0.153 0.097*** 0.818***

−1.133 3.336 22.607

−0.104

−1.575 0.046 0.194*

0.889 1.787

−0.144 −0.146* 0.281* 0.121*** (7.175)

−1.487 −2.253 5.327

***

0.205

*

0.145

***

1.738

2.518

0.182

0.570

Coeff. ***

t-stat.

1.720 0.109*** 0.405***

11.152 3.656 10.851

−0.108 0.083 0.189 0.191*** −0.079

−1.416 1.082 1.508 3.538 −0.899

0.212*** 0.099*

3.164 1.776

0.308

Note: RMSEA = Root mean square error of approximation (< 0.05); GFI = Comparative fit index (> 0.9); TLI = Tucker-Lewis index (> 0.9); SRMR = Standardized root mean square residual (< 0.08). * t(0.10) = 1.645. ** t(0.05) = 1.96. *** t(0.01) = 2.575.

statistics software, Stata 12.0. Table 6 demonstrates the results of the model estimation. The model fit is acceptable in terms of the values of the model fit indices. However, the R-square values of the first and the second equations, KAO and ATO, are lower than 0.2. According to the estimated result of the covariate between the two equations, WTO and WTC, there is a significant and positive correlation between the error terms in both equations suggesting a parallel direction for the two intentions. This implies that passengers who are willing to participate in carbon offset programme are also willing to change their travel behaviours, and vice versa. KAO is only explained significantly by two sources of knowledge, communication media and the passengers’ prior knowledge regarding the environmental impacts of aviation and the concept of carbon offsetting. Between the two mediums, using a video as the briefing medium has a strong and significant impact on enhancing passengers’ levels of environmental knowledge about aviation and the concept of carbon offsetting. Moreover, PRIOR could represent a cumulative effect, indicating that passengers have ever heard of or known about the carbon offset scheme, which can positively and significantly enhance their knowledge. This suggests that air travel authorities and airlines should continuously invest more resources in educating passengers about aviation’s impact on the environment and the idea of carbon offsetting, particularly by using an appropriate communication media. In the second equation model, ATO, KAO is the most significant explanatory variable that positively affects passengers’ attitudes toward carbon offsetting. PRIOR is also a key variable that influences ATO. Hence, the level of knowledge (combined with cumulative knowledge) about aviation impacts and carbon offsetting can significantly determine passengers’ attitudes toward carbon offsetting. With respect to passengers’ personal and trip characteristics, passengers over the age of 50 have significant and positive attitudes toward carbon offsetting, while business passengers have negative but insignificant attitudes toward carbon offsetting. In the WTO equation model, ATO is the most significant variable, implying that it is the most basic element triggering passengers' willingness to voluntarily offset their flights. Passengers' prior knowledge and KAO also play important roles in determining WTO. However, if passengers have ever used tele- or video-conferencing or have a monthly income over NT$75,000, they have a low willingness to voluntarily offset their flights (i.e., negative impacts on WTO). Finally, passengers over the age of 50 have a significant and high willingness to offset their flights. In the final equation model, WTC, ATO is a very important factor affecting passengers' willingness to change their travel behaviour. Passengers who have used tele- or video-conferencing show a higher willingness to change their travel behaviour. However, this differs from the associated results in the WTO equation model. A possible explanation is that if passengers have experience with tele- or video-conferencing, they know that air travel can sometimes be replaced by these technologies; therefore, they might feel that they have already changed their travel behaviour and may have a lower willingness to offset their flights. Moreover, if passengers travel by air only once per year, they show a significant and positive willingness to change their travel behaviour. In addition, young passengers (aged between 21 and 30 years old) have low willingness to change their air travel behaviours. Based on these model results, the direct, indirect, and total effects of the two media (i.e., card and video); passengers’ 105

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

Table 7 Estimated effects. WTO

Card briefing Video briefing KAO ATO

WTC

Direct

Indirect

Total

– – 0.097 (3.333) 0.818 (22.601)

0.236 (4.660) 0.370 (6.307) 0.251 (10.214) –

0.236 0.370 0.348 0.818

(4.660) (6.307) (9.147) (22.601)

Direct

Indirect

Total

– – 0.109 (3.654) 0.405 (10.849)

0.236 (4.660) 0.249 (6.307) 0.124 (10.210) –

0.236 0.249 0.233 0.405

(4.660) (6.307) (7.223) (10.849)

The values in parentheses are t-values.

environmental knowledge about aviation as well as carbon offset (KAO); and passengers’ attitudes toward VCO programme (ATO) on WTO and WTC were estimated and are listed in Table 7. Passengers’ attitudes toward the VCO programme play a key role in their willingness to offset flights (WTO) and their willingness to change their travel behaviour (WTC). This means that convincing air passengers of the positive benefits of the voluntary carbon offset scheme should be a top priority for airline management if they want to increase passengers' willingness to carbon offset and change their travel behaviours to mitigate carbon emissions. Although the effects of KAO on WTO and WTC are not as high as those of ATO, KAO has the most significant effects on ATO (cf. Table 6), particularly through the use of the communication medium of video to deliver information about the environmental impacts of aviation. Therefore, appropriate media tools should be considered for educating air passengers about the negative impacts of air travel and the actions that can be taken to mitigate those impacts. In this study, the use of video briefing was examined and shown to be superior to using card briefing for enhancing passengers' knowledge about the impacts of aviation and the concept of carbon offsetting. 4. Discussions and conclusions Several previous studies have revealed that passengers’ low engagement in carbon offsetting is likely due to a lack of knowledge about aviation’s impact on the environment and a lack of information about the benefits of carbon offsets (e.g., Brouwer et al., 2008; Lu and Shon, 2012; Cliff, 2014). In this study, no more than 5% of respondents fully understood the content and benefits of the voluntary carbon offset scheme. It is impractical to expect passengers to participate in such voluntary carbon offset programmes if they do not know – or vaguely know – about the programme and the environmental impacts of aviation emissions. To improve the environmental knowledge level of air passengers, this study proposed two different types of media but demonstrating the same information about the impacts of aviation on the environment, the benefits of carbon offsetting, and strategies to change travel behaviour, and we verified that both can deliver such information effectively. Particularly, using video media can further enhance respondents’ environmental knowledge of aviation and the idea of voluntary carbon offset scheme by more than 10% compared to the use of card briefing. This is because the way of presentation of the video looks vivid and attractive in comparison with the printed card and therefore, helps the respondents easily acquire the knowledge. The findings of this study align with the study of Molesworth (2014), which indicates that offering a pre-flight safety video that is vivid, humorous, or includes a celebrity can effectively attract passengers’ attention to safety information. However, the form of printed card does not mean useless. Practically, a well-designed, A4-sized, and two-sided printed card can also offer sufficient information regarding the environmental impact of aviation, theory of carbon offsetting, and the mitigating strategies (this is similar to the safety instruction card in all flights). Accordingly, even though the form of video briefing can easily catch the eyes of passengers, synergy effects of evoking air passengers’ awareness could be expected when printed card briefing is considered along with the adoption of video (Dijkstra et al., 2005). Furthermore, positive links among knowledge, attitude, and behavioural intention are successfully verified in this study, which are the same as the conclusions in Nilsson and Küller (2000), Flamm (2009), Kim et al. (2014) and Lera-López et al. (2014). We found that knowledge has both direct and indirect (i.e., mediated by attitude) effects on the impacts of the participants’ willingness to offset flights (WTO) and change their travel behaviour (WTC). In addition, this study indicates that enhancing respondents’ knowledge about the impacts of aviation and the carbon offset scheme could increase their concern and lead to actions to address the impacts. However, attitude (ATO) is the key to motivate actions (i.e., intentions). In Flamm (2009), environmental knowledge and attitudes were revealed to have significant effects on households’ ownership of fuel-efficient vehicles; however, households with pro-environmental attitudes can also affect driving behaviour. In addition, Nilsson and Küller (2000) argued that people’s attitudes toward environmental concerns were more effective than factual knowledge for promoting pro-environmental travel behaviour. To clarify, although enhancing people’s knowledge is the first step to raise the awareness of the environmental impacts of aviation and know what can be done to mitigate these impacts, it is much more important to convince people to realise the benefits of carbon offsetting and subsequently, to take part in the offset scheme and travel less harm to the environment. Nevertheless, there is always a gap between attitude and behaviour (intention) (Cohen et al., 2013; Kroesen, 2013; Dickson et al., 2013). Kroesen (2013) suggested that a better understanding of air travellers’ viewpoints toward aviation and climate change would resolve the inconsistency between their concerns regarding the environmental impacts and air travel behaviour, particularly among the various age groups of people. Our study revealed that older respondents significantly have positive attitudes toward carbon offsetting and have high willingness to offset flights. In addition, respondents travelling less than one time per year significantly have high willingness to change travel behaviour in terms of reducing baggage weight. Hence, these two segments can be targeted to 106

Transportation Research Part D 59 (2018) 96–107

J.-L. Lu, C.-Y. Wang

promote pro-environmental travel behaviour. In addition, Matasci et al. (2014) recommended that barriers for people to address climate change can be overcome by offering better information on the possible consequences of climate change and feasible mitigation measures. Waygood and Avineri (2016) found that people would change travel behaviour and reduce their negative impacts on the environment if useful information on climate change were provided. This also highlights the importance of educating air passengers regarding environmental issues. Furthermore, it is important to transform passengers’ willingness (intention) into actual action. In addition to the internal barriers (i.e., knowledge) of air passengers, external barriers, such as government policy, are critical for passengers to take action (Becken, 2007). Accordingly, the government should set policies to encourage airlines to invest more resources to meet international standards for aircraft emissions and work with airlines to better educate passengers about aviation emissions and the benefits of carbon offsetting. Especially, if airlines spontaneously disclose the amount of emissions of each flight and suggest various strategies for passengers simultaneously with the efforts of airlines themselves to reduce the emissions, the green image of airlines would be fulfilled further. Finally, the outcomes of this study are primarily based on respondents’ stated preferences and intentions. Future studies could use the media proposed in this study and analyse whether passengers actually purchase offsets after receiving environmental briefings. Acknowledgements This study was supported by the Ministry of Science and Technology of TAIWAN, R.O.C. (Grant number: MOST103-2410-H-022005). We thank all the anonymous reviewers for their valuable comments. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.trd.2017.12. 024. References Araghi, Y., Kroesen, M., Molin, E., Wee, B.V., 2014. Do social norms regarding carbon offsetting affect individual preferences towards this policy? Results from a stated choice experiment. Transp. Res. Part D 26, 42–46. Becken, S., 2007. Tourists’ perception of international air travel’s impact on the global climate and potential climate change policies. J. Sustain. Tourism 15 (4), 351–368. Brouwer, R., Brander, L., Van Beukering, P., 2008. “A convenient truth”: air travel passengers’ willingness to pay to offset their CO2 emissions. Clim. Change 90 (3), 299–313. Budianschi, C., Ekeroth, F., Marija, M., 2012. Eco-friendly flights? A consumer’s perspective. Bachelor’s Thesis within Business Administration, Jönköping International Business School, Jönköping University. Chen, F.-Y., 2013. The intention and determining factors for airline passengers’ participation in carbon offset schemes. J. Air Transp. Manag. 29, 17–22. Cheung, J., Kragt, M., Burton, M., 2015. The awareness and willingness of air travelers to pay for voluntary carbon offsets and their co-benefits. Working Paper 1505, School of Agricultural and Resource Economics, University of Western Australia, Crawley, Australia. Cohen, S.A., Higham, J.E.S., Reis, A.C., 2013. Sociological barriers to developing sustainable discretionary air travel behaviour. J. Sustain. Tourism 21 (7), 982–998. Cliff, A.D., 2014. An Investigation of UK Passenger Attitudes Towards the Carbon Offsetting of Both Flight and Airport Emissions. (Unpublished master’s thesis). University of Chester, United Kingdom. Available at: < http://www.hdl.handle.net/10034/346175 > . Davison, L., Littleford, C., Ryley, T., 2014. Air travel attitudes and behaviours: the development of environment-based segments. J. Air Transport Manage. 36, 13–22. Dickson, J.E., Robbins, D., Filimonau, V., Hares, A., Mika, M., 2013. Awareness of tourism impacts on climate change and the implications for travel practice: A Polish perspective. J. Travel Res. 52 (4), 506–519. Dijkstra, M., Buijtels, H.E.J.J.M., van Raaij, W.F., 2005. Separate and joint effects of medium type on consumer responses: a comparison of television, print, and the Internet. J. Bus. Res. 58, 377–386. Dodds, R., Leung, M., Smith, W., 2008. Assessing awareness of carbon offsetting by travellers and travel agents. Anatolia: Int. J. Tourism Hospitality Res. 19 (1), 135–148. Flamm, B., 2009. The impacts of environmental knowledge and attitudes on vehicle ownership and use. Transp. Res. Part D 14 (4), 272–279. Gofin, R., Palti, H., Adler, B., 1990. The use of car restraints by newborns and mothers: knowledge, attitude, and practices. Isr. J. Med. Sci. 26, 261–266. Hagmann, C., Semeijn, J., Vellenga, D.B., 2015. Exploring the green image of airlines: passenger perceptions and airline choice. J. Air Transport Manage. 43, 37–45. Hinnen, G., Hille, S.L., Wittmer, A., 2015. Willingness to pay for green products in air travel: ready for take-off? Bus. Strat. Environ. 26 (2), 197–208. IATA: International Air Transport Association, 2008. Aviation Carbon Offset Programmes: IATA Guidelines and Toolkit. IATA: International Air Transport Association, 2015. IATA Carbon Offset Program. Accessed: 01/23/2015. ICAO: International Civil Aviation Organization, 2015. Carbon Emissions Calculator. Accessed: 01/23/2015. Jou, R.-C., Chen, T.-Y., 2015. Willingness to pay of air passengers for carbon-offset. Sustainability 7, 3071–3085. Kim, Y., Yun, S., Lee, J., 2014. Can companies induce sustainable consumption? The impact of knowledge and social embeddedness on airline sustainability programs in the U.S. Sustainability 6, 3338–3356. Kroesen, M., 2013. Exploring people's viewpoints on air travel and climate change: understanding inconsistencies. J. Sustain. Tourism 21 (2), 271–290. Lera-López, F., Sánchez, M., Faulin, J., Cacciolatti, L., 2014. Rural environment stakeholders and policy making: willingness to pay to reduce road transportation pollution impact in the Western Pyrenees. Transp. Res. Part D 32, 129–142. Liao, M.Y., 2014. An evaluation of an airline cabin safety education program for elementary school children. Eval. Program Plann. 43, 27–37. Lu, J.-L., Shon, Z.Y., 2012. Exploring airline passengers’ willingness to pay for carbon offsets. Transp. Res. Part D 17, 124–128. Mair, J., 2011. Exploring air travellers' voluntary carbon-offsetting behaviour. J. Sustain. Tourism 19 (2), 215–230. Matasci, C., Kruse, S., Barawid, N., Thalmann, P., 2014. Exploring barriers to climate change adaptation in the Swiss tourism sector. Mitig. Adapt. Strat. Glob. Change 19, 1239–1254. Molesworth, B.R.C., 2014. Examining the effectiveness of pre-flight cabin safety announcements in commercial aviation. Int. J. Aviat. Psychol. 24 (4), 300–314. Nilsson, M., Küller, R., 2000. Travel behaviour and environmental concern. Transp. Res. Part D 5, 211–234. O’Connor, R.E., Bord, R.J., Yarnal, B., Wiefek, N., 2002. Who wants to reduce greenhouse gas emissions? Soc. Sci. Quart. 83 (1), 1–17. Van Birgelen, M., Semeijn, J., Behrens, P., 2011. Explaining pro-environment consumer behavior in air travel. J. Air Transport Manage. 17 (2), 125–128. Walton, D., Thomas, J.A., Deavitzki, V., 2004. Commuters’ concern for the environment and knowledge of the effects of vehicle emissions. Transp. Res. Part D 9, 335–340. Waygood, E.O.D., Avineri, E., 2016. Communicating transportation carbon dioxide emissions information: does gender impact behavioral response? Transp. Res. Part D 48, 187–202.

107