Motivating Sustainable Change in Tourism Behavior: The First- and Third-Person Effects of Hard and Soft Messages

Educating and changing consumers ́ attitudes towards sustainable and more environmentally friendly holiday choices is often seen as a key challenge for the tourism industry. The primary objective of this study is, therefore, to increase our understanding of psychological mechanisms underlying consumers’ responses to communication that aims to alter their holiday behavior in a more sustainable direction. Drawing on reactance theory, as well as firstand third-person perception effects, we present an experimental study designed to test how different levels of message assertiveness (i.e., hard versus soft pressure) affect consumers’ intentions to change their traveling behavior. The results suggest that when respondents are presented with a socially desirable message, their individual intentions to change one’s holiday plans are affected to a greater extent compared to their perception of how others would react to such cuing. Furthermore, this first-person effect is most prominent under lower levels of message assertiveness, and when conveyed messages address socially desirable behavior in line with one’s current values. Hard pressure messages loaded with highly assertive prompts, on the other hand, are likely to evoke motivational reactance, especially when a consumer holds a weaker attitude towards sustainability and environmental issues. Practical and theoretical implications of the provided findings as well as avenues for future research are discussed.


Introduction
The consequences of tourism growth have been called a "double-edged" sword, as it has both positive and negative effects on the environment [1]. The attention on the negative effects of tourism has increased since the 1990s, primarily due to the apparent contribution of resource depletion and ecosystem degradation [2]. More recently, the effect of excessive tourism and traveling on climate change and global warming is also placed center stage in discussions on sustainable consumption [3]. In the tourism sector, like in any commercial market, an alignment takes place between the value a service produces and the consumers´higher-order values [4], and responsible for this alignment is the so-called value regime [5]. According to Gollnhofer, Weijo and Schouten [6], the value-creating operations of a value regime remains hidden as long as the value it produces is consistent with the higher-order values of consumers, but changes when value outcomes violate the higher-order values of important consumer segments or groups.
Transferred to the context of tourism, the growing awareness and concern for the negative environmental impacts of different sides of tourism and travel, are largely due to an increasing mismatch between the value outcomes of tourism and higher-order consumer values related to

Theoretical Background
Whenever a value regime is shaking in its foundations, the tone of the public debate and the persuasive messages communicated by different interest groups varies from the mild and rationally argumentative, to the harder and more emotionally laden. The 2-minute film "I am Nature", produced by Conservation International with Julia Roberts as the narrator, is an example of rational, though thought-provoking, messaging that explains the role of nature through history, nature's ability to adapt to changes, and that nature is ready to evolve even if humans are not. Oftentimes, the messages that receive the most media attention are the more controversial and emotional ones, and well known examples are the animal welfare organization, PETA´s, campaigns portraying naked celebrities under the slogan, "I´d rather go naked than wear fur", or other celebrities holding up skinned and bloody animal corpses accompanied by a message saying, "Here´s the rest of your fur coat". Furthermore, prior to the annual convention of the National Rifle Association in 2019, the organization Everytown for Gun Safety posted huge advertisements around Indiana where the convention took place, with controversial pictures of an automatic rifle and the statement "Shoot a school kid, only $29". Within the context of sustainability, a 2008 video from the Portuguese conservation group Quercus that showed a monkey, a polar bear and a kangaroo committing suicide because their natural habitats are ruined by global warming, might serve as yet another example of rather controversial and emotionally laden messaging. However, Quercus is just one of many examples, and in their study on the effect of environmental messages, Kronrod, Grinstein and Wathieu [7] found that while only 19% of slogans for consumer goods were assertive, 57% of the environmental slogans were assertive, the equivalent to what we call hard pressure messages. This is interesting, as previous research in consumer research has shown that the effectiveness of persuasive communication depends, amongst others, on the assertiveness of the message. In general, the harder the pressure put on people, the less interested they are in complying with the requests or suggestions put forward. The next paragraphs describe why.

Hard and Soft Messages and the Role of Reactance
The intention behind mass communication is to influence the attitudes and behaviors of the receiver in a specific direction. Within the field of communication research, the theory of reactance suggests that when communication is perceived as threatening to the freedom of independent choice, psychological reactance is motivated, and a rejection of the message a likely outcome [8,9]. Rejecting the message, or choosing the opposite of what is being encouraged, becomes a way to restore a consumer's freedom of choice. Research on reactance has focused heavily on explaining why certain types of communication campaigns have been unsuccessful in achieving the desired attitude change [10], and typically, reactance is evoked when the receiver perceives that a behavioral change is demanded of them, rather than an opportunity offered to them.
The assertive language used in a variety of messages communicating the need for sustainable change, should, according to reactance theory, have little effect in changing consumer attitudes and behavior [10]. However, in cases where the message is perceived to be important, Kronrod et al. [7] find that assertive language is more effective and leads to higher intentions to comply with the message than the same message using less assertive language. In their study, the messages were that participants should generally use more public transport, or be aware of water usage when doing dishes. However, to reach the goals of the UN and the Paris agreement, the UN itself admits that "we have to make unprecedented action immediately" [11]. What the UN says is that more dramatic changes in behavior are necessary to deal with the environmental challenges we are currently faced with, and these, we will add, are more far-reaching than water usage while dish washing. In this perspective, it is less certain whether the environmental behaviors applied in Kronrod et al. [7] can be perceived as very restrictive to free choice or limiting behavior that is important to consumers, which is a key pre-requisite for reactance [8]. Following from this, it is also unclear whether the perceived importance of the message will result in respondents being willing to accept the assertive language as suggested by Kronrod at al. [7], when the important behavioral change is of the "unprecedented" kind.
In the present study, we test messages encouraging consumers to make environmentally sustainable choices when going on holiday, using two different levels of assertiveness. The first offers general suggestions, the other demands behavioral change from the reader. Both have the same message content (see Appendix A for full text), and only the tone of the message is different. However, in contrast to the messages used in the study by Kronrod et al. [7], the changes suggested/demanded of the participants in our study are likely to be perceived as having far-reaching impacts on their lives. What we expect to see then, is that among the participants, the harder and more assertive message will motivate sustainable change less than the softer, less assertive message. However, we will also argue that respondents will assume that others are more influenced by the harder messages than they are themselves. In other words, we find reason to believe that hard and soft messages will affect first and third persons differently. The next section discusses this further.

The First-and Third-Person Effect
The Third-Person Perception (TPP) holds that "a person exposed to persuasive communication in the mass media sees this as having a greater effect on others than on himself or herself" [12] (p. 1). The TPP effect has been an important strand of study in the field of advertising and communication research [13][14][15][16][17]. Biased optimism is the most frequently used framework to explain the TPP effect [18,19], and holds that others are more likely to experience negative consequences compared to oneself. When transferring this concept to the field of communication and the research conducted on TPP, messages that are perceived negative to be influenced by will evoke a TPP [13][14][15][16][17]. In contrast, messages considered positive to be influenced by will in some cases evoke a First-Person Perception (FPP) [20,21]. The FPP is a development of the TPP and was identified later [12,22]. Contrary to the third-person effect, the first-person effect suggests that when exposed to socially desirable media messages, consumers believe they are just as prone to be influenced by the message as others, if not more.
Initial research on TPP focused mainly on testing the effect of negative influences such as ads, pornography and violence, which generally resulted in a TPP. However, Gunther and Mundy [18] tested a range of different message types, which varied in their degree of "smartness to be influenced by", and they found a clear link between this measure and the TPP. Topics that were considered less smart to be influenced by resulted in a significant TPP, whereas they saw no difference in first-and third-person effect for messages which were seen as "smart to be influenced by". Inspired by the work by Gunther and Mundy [18], Hoorens and Ruiter [22], found a significant relationship both for undesirable messages and TPP as well as for desirable messages and FPP. As proposed by Golan and Day [23], more FPP research is important to avoid the false perception that media communication is undesirable, and that persuasion is always seen as a negative by receivers.
Similar to reactance research, research on TPP has shown that persuasive intent is important for the effect to occur [8,9,18,20]. Scherr and Müller [24] directly tested the link between the two theories and found that perceived persuasive intent is positively linked to reactance and an important predictor for TPP. Due to its importance to TPP they suggest that reactance should always be included in future studies of TPP. Directly related to our context, Hoorens and Ruiter [22] showed that messages concerning environmental issues were generally considered socially desirable, and led to an FPP. As far as we are aware, no other study has explored the "optimal impact phenomenon" or TPP/FPP in the context of sustainability and environmental issues. Considering that awareness and knowledge are often the first steps towards behavioral change, gaining a better understanding of how to most effectively communicate the message of sustainability to consumers is, therefore, a key insight. In this study, we therefore test how the hard and soft messaging affects both FPP and TPP. The dominant and lasting societal focus on sustainability is typically a value regime in change, but the world population is, nevertheless, rather heterogeneous in their interest and involvement in pursuing a more sustainable behavior. Even though pursuing the UN Sustainability Goals are socially accepted, behavioral change strategies targeting "unprecedented change" might still be perceived as being too pushy, too offensive, and not motivating enough to get people to jump the bandwagon. Partly in line with Kronrod et al.'s [7] focus on importance, we will argue that consumers' higher-order values, which translates to the strength of their general attitudes towards sustainability, will influence how hard and soft messages influence sustainable change motivation. The next paragraphs outline this suggestion in more detail.

Attitude Strength
Brehm and Bryant [25] have shown that messages in line with or neutral to one's own convictions are less likely to evoke reactance, whereas for messages that are contrary to one's own beliefs, reactance is likely to be much stronger. Similarly, Vested Interest Theory (VIT) describes how a consumer's interest in an object depends on how it corresponds with his/her value orientation [26]. Although messages of sustainability and conscious consumption are largely considered to be in accordance with the current value regime shift, it is fair to say that the degree of focus and interest in the subject varies. Studies have shown that attitude strength increases behavioral intentions that are consistent with the attitude [27], which is in line with reactance theory [8]. According to reactance theory, messages that are consistent with currently held attitudes will rarely evoke reactance, as they are not urging consumers to behave differently than they already intended to. Hence, the message is not influencing their freedom to choose, as their choice is already made and thus, potentially only strengthened by the message. Translated to our context, this implies that even though a message is designed in a way that would otherwise be considered a hard, or assertive message, it is less likely to evoke reactance if it is consistent with the consumers higher-order values, i.e., their currently held altruistic and biospheric attitudes [28]. Moreover, drawing on vested interests being related to the consistency between attitude and behavior, we suggest that consumer reactions to advertising messages should differ, corresponding to differing values and attitudes [26]. We therefore argue, in line with the findings by Pham and Mandel [29], that consumer responses to persuasive sustainability messages depend on their preexisting attitudes towards sustainability and environmental issues. This is also in line with the expectations of how consumers' actions are consistent with their value-relevant involvement [30], and we therefore expect consumers with a stronger attitude towards environmental issues to be less concerned about the tone of the message, as the content of the message parallels their personal views. Consumers with a weaker attitude towards environmental issues will be more sensitive to message style, which results in reactance when the message becomes harder and more assertive. This further implies, that for the FPP-TPP distinction, we assume that when the message is aligned with one's own stronger held attitudes, the FPP will dominate also for hard messages. In contrast, when the existing attitude strength is low, harder message design will result in a TPP dominance.
In summary, this leaves us with the following hypotheses to be tested.

Hypothesis 1 (H1).
For soft messages, (a) FPP will be more prevalent than TPP, while (b) the opposite will result from hard messages.

Hypothesis 2 (H2).
Hard messages will result in significantly more reactance, measured as anger and perceived threat to consumers' freedom, than soft messages.

Hypothesis 3 (H3).
For consumers with strongly held attitudes towards sustainability, the FPP will dominate for both hard and soft messages.

Hypothesis 4 (H4).
For consumers with weaker attitudes towards sustainability, (a) FPP will dominate for soft messages, while (b) TPP will dominate for hard messages.

Procedure
To test our hypotheses, 137 undergraduate students at a publicly funded university were recruited to participate in a mixed design study, where message pressure Hard vs. Soft was experimentally manipulated, while attitude strength (High and Low) was based on a two-group split on a measured scale. The experimental session began with acquiring an informed consent, after which respondents were given a booklet containing the experimental material. The versions of the booklet, containing the hard/soft message variations about environmental impact of tourism, were randomly distributed to the participants to ensure random allocation to experimental conditions. Opening the booklet, subjects first answered a survey measuring their attitudes towards environmental issues. They then turned to the soft or hard message text. These were identical in content, but not in style. The texts were also of equal length (250 words) to ensure equal exposure to the treatment. Of importance, the soft-pressure message was framed as general suggestions that travelers be more environmentally responsible when going on holiday. The hard-pressure message focused on each individual's responsibility, and used directive language urging the individual reader to travel more responsibly when going on holiday (full texts are found in the Appendix A). It suggests limitations on how often, for how long and where to go on holiday. Based on an argument of fair allocation of resources, our message also suggests that higher prices should be paid for the goods and services used when on holiday. In that respect, the message conveyed addressed both altruistic (fair resource distribution) and biospheric (environmental concern) issues. Next, respondents completed item scales on perceived threat to freedom, and anger, which measured their reactance in the light of the message. Finally, and in line with procedures tested by Gunther [21] and Price and Tewksbury [31], respondents indicated to what degree the text they had read was likely to influence them personally when planning their next holiday, and to what degree they expected the text to influence people in general when planning their next holiday. Drawing on the experience from previous research, which shows no order effect nor any artificial contrast effect when presenting the first-and third-person perceived effect questions sequentially [21,31], we adopted this particular approach.

Measures and Scale Validation
All measures used were drawn from previous research. Five measures for attitudes towards environmental issues were adapted from Petty and Krosnick [32] and Pham and Mandel [29]. Example items are: "How much do you personally care about issues related to the environment?", and "How much do the environmental concerns directly affect you?". Anchors for the responses ranged from 1 (Not at all) to 5 (Very much). To measure reactance, we used the perceived threat to freedom and anger scales developed by Dillard and Shen [10]. The four items on threat to freedom had a five point scale anchored completely disagree (1) and completely agree (5), while the four anger items were anchored Not at all (1) and To a large degree (5).
In line with Gunther [21] and Price and Tewksbury [31], we measured the TPP and FPP by inviting respondents to answer the following questions: "To what extent do you believe that the information in the text you just read will influence you when you plan your next holiday?" and subsequently, "To what extent do you believe that the information in the text you just read will influence people in general when they plan their next holiday?". Anchors for both queries ranged from (1) Not at all to (5) To a great extent.
All scales were subject to a face validity check, as they were exposed to critical review by members of a research group working with various sustainability-related subjects. Questions that initially seemed confusing or ambiguous were identified and refined before data collection.
Prior to hypotheses testing, the multi-item scales were subject to an internal consistency check. The attitude scale saw a Cronbach alpha equal to 0.83, comfortably exceeding the oft-cited threshold of 0.70 [33], and indicating a high inter-item reliability. Also very satisfying, the four items for perceived threat to freedom yielded a Cronbach alpha of 0.89, while the four-item anger scale received an alpha score of 0.90.

First-Person and Third-Person Effects for Soft and Hard Pressure Messages
Consistent with past research [18,22], we find that in the entire sample (n = 137) a significant first-person effect occurs when respondents are presented with a socially desirable message (i.e., regardless of the level of pressure imposed). A paired sample t-test was run to evaluate the difference in perceived effect of the message on self as opposed to others. The results imply a significant difference, with M FPP = 2.65 and M TPP = 2.46 (p = 0.032).
However, the focus of this study was to explore whether the different levels of pressure used in the socially desirable message play a role in activating the first-person effect, so we further examined the impact of the two manipulations, i.e., hard-and soft-pressure messages.
The results (see Figure 1) revealed that respondents exposed to the soft-pressure message found themselves more likely to be influenced by the message as opposed to others. There was a significant difference in the scores, M FPP = 2.83 versus M TPP = 2.51 (p = 0.006). This result is in line with our Hypothesis 1a. However, the respondents exposed to the hard-pressure message report no thirdperson dominance, as this group believed there to be no difference in the perceived influence of the message on self and others (M FPP = 2.46, M TPP = 2.41, p = 0.708). We hypothesized that the TPP should dominate this condition, and our Hypothesis 1b is thus not supported.

Reactance: The Role of Perceived Threat to Freedom and Anger
A natural conclusion following from Figure 1 is that the predicted FPP is significantly reduced when the message pressure increases. An independent sample t-test was conducted to compare the effect on self in the two experimental conditions. Respondents exposed to the hard-pressure message rated the likelihood of the message to influence their next holiday significantly lower than respondents who were exposed to the soft-pressure communication (Mhard = 2.46, Msoft = 2.83, p = 0.034). However, the TPP was not influenced by the message manipulation (p = 0.476). As we hypothesized that this change in FPP but not TPP was due to motivational reactance, we ascertained this reasoning in the next step.
Results depicted in Figure 2 reveal a significantly higher perceived threat to freedom in the hardpressure communication condition than in the soft-pressure one, Mhard = 2.89 versus Msoft = 1.81 (p = 0.000). Additionally, a significant greater feeling of anger was identified in the hard-pressure condition than the soft-pressure communication, with Mhard = 1.96 and Msoft = 1.44 (p = 0.000).

Reactance: The Role of Perceived Threat to Freedom and Anger
A natural conclusion following from Figure 1 is that the predicted FPP is significantly reduced when the message pressure increases. An independent sample t-test was conducted to compare the effect on self in the two experimental conditions. Respondents exposed to the hard-pressure message rated the likelihood of the message to influence their next holiday significantly lower than respondents who were exposed to the soft-pressure communication (M hard = 2.46, M soft = 2.83, p = 0.034). However, the TPP was not influenced by the message manipulation (p = 0.476). As we hypothesized that this change in FPP but not TPP was due to motivational reactance, we ascertained this reasoning in the next step.
Results depicted in Figure 2 reveal a significantly higher perceived threat to freedom in the hard-pressure communication condition than in the soft-pressure one, M hard = 2.89 versus M soft = 1.81 (p = 0.000). Additionally, a significant greater feeling of anger was identified in the hard-pressure condition than the soft-pressure communication, with M hard = 1.96 and M soft = 1.44 (p = 0.000).

Reactance: The Role of Perceived Threat to Freedom and Anger
A natural conclusion following from Figure 1 is that the predicted FPP is significantly reduced when the message pressure increases. An independent sample t-test was conducted to compare the effect on self in the two experimental conditions. Respondents exposed to the hard-pressure message rated the likelihood of the message to influence their next holiday significantly lower than respondents who were exposed to the soft-pressure communication (Mhard = 2.46, Msoft = 2.83, p = 0.034). However, the TPP was not influenced by the message manipulation (p = 0.476). As we hypothesized that this change in FPP but not TPP was due to motivational reactance, we ascertained this reasoning in the next step.
Results depicted in Figure 2 reveal a significantly higher perceived threat to freedom in the hardpressure communication condition than in the soft-pressure one, Mhard = 2.89 versus Msoft = 1.81 (p = 0.000). Additionally, a significant greater feeling of anger was identified in the hard-pressure condition than the soft-pressure communication, with Mhard = 1.96 and Msoft = 1.44 (p = 0.000).    This result indicates that the hard message resulted in respondents experiencing motivational reactance, causing them to reject the message. This is in line with reactance theory where research shows that people will behave contrary to what is being suggested in order to regain the feeling of free choice [9], and in support of our Hypothesis 2.

The Role of Attitudes Towards Environmental Issues
In order to test how the effect of hard and soft messages influenced FPP and TPP at different attitude strengths, we split the sample in two based on a mean split of the attitude scale. A paired sample t-test was used to assess FPP versus TPP effects for both the hard and soft pressure type of message (see Figure 3 for individuals with high attitude strength, and Figure 4 for those with low attitude strength). Firstly, for consumers with strong attitudes, the soft pressure condition results in a dominating and significant FPP effect (M FPP = 3.38 versus M TPP = 2.79, p = 0.003). Moreover, also for the hard-pressure message there is a significant FPP effect for this specific group, with M FPP = 2.98 and M TPP = 2.48 (p = 0.011). Our Hypothesis 3 is thus supported. This result indicates that the hard message resulted in respondents experiencing motivational reactance, causing them to reject the message. This is in line with reactance theory where research shows that people will behave contrary to what is being suggested in order to regain the feeling of free choice [9], and in support of our Hypothesis 2.

The Role of Attitudes Towards Environmental Issues
In order to test how the effect of hard and soft messages influenced FPP and TPP at different attitude strengths, we split the sample in two based on a mean split of the attitude scale. A paired sample t-test was used to assess FPP versus TPP effects for both the hard and soft pressure type of message (see Figure 3 for individuals with high attitude strength, and Figure 4 for those with low attitude strength). Firstly, for consumers with strong attitudes, the soft pressure condition results in a dominating and significant FPP effect (MFPP = 3.38 versus MTPP = 2.79, p = 0.003). Moreover, also for the hard-pressure message there is a significant FPP effect for this specific group, with MFPP = 2.98 and MTPP = 2.48 (p = 0.011). Our Hypothesis 3 is thus supported.

Discussion of Results and Scientific Contribution
According to Steg et al. [34], people across the world, both in developed and less developed countries, endorse biospheric values related to pro-environmental beliefs, attitudes, norms and actions. Hence, a message that motivates sustainable tourism behavior is likely to be considered socially desirable in general.
Related to this, the first important result of our study is that for messages addressing a higherorder goal of more sustainable tourism behavior, an FPP effect is more prominent than a TPP effect. This implies that, in general, a message that motivates socially desirable behavior in line with current values, will have an effect on the individual that receives it, and not only on his/her perception of how others will react to it. This is an important starting point for further discussion, as it confirms that communicative messages of the "smart to be influenced by" kind increase the receiver's motivation to comply.
Furthermore, the results of this study show that when communicating the need for more sustainable tourism behavior, a soft message will render a dominating FPP effect, while for harder messages the FPP dominance disappears. The reason, we suggest, is that hard pressure messages motivate a higher level of reactance, which we also find in our study. Based on this, we will argue that the overwhelming focus on assertive and hard-push message appeals that seem to be dominating the communication practices within the realm of sustainability [7], are less effective than an alternative, softer approach. There are several reasons for this, all highlighted by our results. One is that, in line with VIT, the behaviors most attractive to tourists are the ones most closely aligned with their higher-order values and attitudes [26]. If these are dominated by egoistic and hedonic values, rather than altruistic or biospheric ones, a hard pressure message will cue the reduction in freedom of choice resulting from the message, and motivate reactance. Freedom of choice is an issue related to individual wellbeing, and thus closely related to both egoistic and hedonic values. Previous research suggests that cuing such values might turn the focus away from what benefits the environment, towards what the consequences are for the individual [34]. This then renders the message as something "negative to be influenced by", and TPP is then a more probable outcome [13][14][15][16][17]. In this context, how you think others might change their tourism behavior does not imply that you change yourself, which reduces the motivational effectiveness of the hard pressure message.

Discussion of Results and Scientific Contribution
According to Steg et al. [34], people across the world, both in developed and less developed countries, endorse biospheric values related to pro-environmental beliefs, attitudes, norms and actions. Hence, a message that motivates sustainable tourism behavior is likely to be considered socially desirable in general.
Related to this, the first important result of our study is that for messages addressing a higherorder goal of more sustainable tourism behavior, an FPP effect is more prominent than a TPP effect. This implies that, in general, a message that motivates socially desirable behavior in line with current values, will have an effect on the individual that receives it, and not only on his/her perception of how others will react to it. This is an important starting point for further discussion, as it confirms that communicative messages of the "smart to be influenced by" kind increase the receiver's motivation to comply.
Furthermore, the results of this study show that when communicating the need for more sustainable tourism behavior, a soft message will render a dominating FPP effect, while for harder messages the FPP dominance disappears. The reason, we suggest, is that hard pressure messages motivate a higher level of reactance, which we also find in our study. Based on this, we will argue that the overwhelming focus on assertive and hard-push message appeals that seem to be dominating the communication practices within the realm of sustainability [7], are less effective than an alternative, softer approach. There are several reasons for this, all highlighted by our results. One is that, in line with VIT, the behaviors most attractive to tourists are the ones most closely aligned with their higher-order values and attitudes [26]. If these are dominated by egoistic and hedonic values, rather than altruistic or biospheric ones, a hard pressure message will cue the reduction in freedom of choice resulting from the message, and motivate reactance. Freedom of choice is an issue related to individual wellbeing, and thus closely related to both egoistic and hedonic values. Previous research suggests that cuing such values might turn the focus away from what benefits the environment, towards what the consequences are for the individual [34]. This then renders the message as something "negative to be influenced by", and TPP is then a more probable outcome [13][14][15][16][17]. In this context, how you think others might change their tourism behavior does not imply that you change yourself, which reduces the motivational effectiveness of the hard pressure message. While Osburg et al. [26] suggest that individuals should be particularly involved in decisions that are consistent with their value orientation, Steg et al. [34] argue that values reflect which overarching goals are generally most important to people, and that normative goals drive people towards actions they believe are appropriate in a given situation. Related to our Hypotheses 3 and 4, where we in more detail scrutinize how soft and hard messages result in FPP and TPP across consumer attitudes towards environmental issues, it is interesting to see that our findings align well with the suggestions in previous research [26,34]. When consumers have developed strong pro-environmental attitudes, a message that corresponds to their attitude leads to an FPP dominance regardless of whether it is soft or hard. This would imply that the message is compatible with their value orientation, and complying with it is considered a smart thing to do. An interesting detail here is that when looking at TPP for this experimental group, the scores are also rather high. While bordering on speculations, this might be an indication of a false consensus effect, where respondents overestimate the degree to which others would agree with them and adopt a sustainable behavior. In contrast, in the absence of such value-behavior compatibility, under low attitude strength, a hard message instigates a TPP dominance, while a soft pressure message results in equal levels of FPP and TPP. This, we will argue, is the likely result of consumers with weak pro-environmental attitudes being less influenced by a socially desirable message than their strong attitude companions, and that increased pressure results in reactance and rejection of the message.

Conclusion and Practical Contribution
For tourism marketers, communicators, public policy makers and advertisers aiming to motivate more sustainable tourism behavior, our study has some clear practical implications.
First, a message that is socially desirable and consistent with general consumer values will motivate behavioral change.
Second, as the motivational effect on one's own behavior (FPP) is higher when the message includes a soft kind of pressure, there is no need to bring in the cavalry. While hard pressure messages are effective for consumers with strong pro-environmental attitudes, these will probably engage in pro-sustainable behavior regardless of the message, simply because this kind of behavior is consistent with their altruistic and biospheric values.
For the tourism industry, and public policy makers, a more fruitful way to motivate future sustainable tourism behavior will be to address the normative goals [34], and the altruistic and biospheric values [26] of those tourists who now fall in the low attitude strength group. While biospheric values are endorsed across the entire globe, the strengthening of these, relative to egoistic and hedonistic values, is a strategy that will enable harder pressure, and thus more effect, as these values develop. At present, we argue, this is most efficiently achieved with softer messaging than what we usually see within the realm of sustainability.

Limitations and Future Research
We believe it is important to acknowledge some limitations related to this study, which also offers the opportunity for further research.
Firstly, we have only tested one sustainable message that targets a hedonistically motivated behavior with a relatively infrequent occurrence. Kronrod et al. [7] focused on sustainable behaviors of a more instrumental kind, related to everyday routines. Between these two extremes, there are a whole range of behaviors that can be performed in a more sustainable way. Along this continuum, future research should aim to understand at what point assertive messages become less effective in favor of softer messages.
Secondly, the study was only conducted in one high-income country, which is currently not severely affected by the impacts of tourism. Consumers in countries and areas more directly affected by some of the challenges may react very differently to the issues raised by the communication. It would, therefore, also be fruitful to replicate this study using a more general, cross-national sample.
Finally, it would be of value to understand how long the effect of a message of this kind influences the consumer. In order to ensure lasting behavioral change it is important to understand what kind of message actually has a long lasting effect. Is there a possibility that softer messages are more acceptable at the point of exposure, while the effect over time is greater for more assertive messages due to them being more memorable? We suggest that all of these questions offer fruitful pathways for future research.

Conflicts of Interest:
The authors declare no conflict of interest.

Appendix A
Manipulation text translated from Norwegian. In Norwegian, the two texts contained the exact same number of words.

Appendix A.1 Hard Pressure Manipulation Text
Travel Responsible! You need to do this now to make your holiday more environmentally friendly. We travel more than we used to, but everybody knows that tourism has an impact on the environment. It is entirely possible for you to travel responsibly and safeguard the environment and nature, promote local culture and traditions, and contribute to local economic growth. It is your responsibility as a tourist to make sure this happens.
The "all-inclusive" packages offered by large travel operators are perhaps a comfortable option, but as any reasonable person knows, these have few positive impacts on the local environment. In the worst case, this type of tourism can cause locals to lose land and access to important resources such as water and electricity. Destruction of natural and cultural heritage is a consequence of your irresponsible holiday. You must take responsibility and minimize the negative consequences of your holidays.
Flights are polluting. Stop traveling on these fleeting weekend trips now! If you absolutely must fly, then travel less frequently and for a longer time. It contributes to a better understanding of the place you visit, and is more environmentally friendly.
As a responsible tourist, you must visit workshops where local crafts are produced based on local traditions and buy souvenirs that have real value. You must always buy goods and services at a fair price, and it is your responsibility to contribute to good working conditions for the locals. You must always eat organic food and local produce. Conservation of natural and cultural heritage, and thus preservation of the world's biological and cultural diversity, is something that you, as a tourist, must take responsibility for.

Appendix A.2 Soft Pressure Manipulation Text
How to Travel Responsible?
Here are some tips on how to make our holidays more environmentally friendly. We travel more than before, but tourism can also have an impact on the environment. At the same time, it is entirely possible to travel responsibly and safeguard the environment and nature, promote local culture and traditions, and contribute to local economic growth. As a responsible tourist, one can allow oneself unique experiences with a good conscience.
The "all-inclusive" packages offered by large travel operators are perhaps a comfortable option, but these have few positive impacts on the local environment. In the worst case, this type of tourism can cause locals to lose land and access to important resources such as water and electricity. Destruction of natural and cultural heritage is a consequence of this type of irresponsible tourism. Responsible tourism is therefore about minimizing the negative consequences of our holidays.
Flights are polluting. Traveling over longer periods contributes to a better understanding of the place you visit, and is more environmentally friendly. A trip where you are away for a longer time is usually more memorable than the fleeting weekend trips.
As a responsible tourist, you can visit workshops where local crafts are produced based on local traditions, and buy souvenirs that have real value. Please be aware if goods and services are sold at a fair price, as this can contribute to good working conditions for the locals. You can also eat organic food and local produce. As a responsible tourist, you have the opportunity to contribute to the conservation of natural and cultural heritage, and thus help preserve the world's biological and cultural diversity.