Consumers’ perceived corporate social responsibility evaluation and support: The moderating role of consumer information

This study analyzes how consumers’ evaluations of various dimensions of corporate social responsibility (CSR) affect their support of it and how consumers’ searches for CSR information influence such evaluations. The empirical analysis relies on data on CSR from a wide representative sample of 3543 Spanish hotel consumers. We use hierarchical multiple regressions to test the relationships and use factorial analysis to test the validity of the different CSR dimensions. The proposed positive effects of legal, ethical, economic, philanthropic, and environmental dimensions of CSR on consumers’ support for corporate reputation, the selection of an establishment, and future purchase intention are corroborated, although they are only partially corroborated in the case of the economic dimension. These relationships are moderated in some cases by consumers’ search for information about hotels’ CSR practices. Managerial and economic implications are derived from the results.


Introduction
Corporate social responsibility (CSR) involves the implementation of business practices that contribute positively to society by supporting sustainable economic, social, and environmental development (Du et al., 2010). CSR also contributes to firms' competitive advantage and performance (Carroll and Shabana, 2010;Lee and Park, 2009). The business case for CSR involves arguments and empirical justifications for the benefits of businesses adopting CSR practices (Carroll and Shabana, 2010) and includes the long-term economic advantages. Prior studies have found positive effects of CSR on financial performance (Orlitzky et al., 2003). In an experimental study, León and Araña (2014) also found that the costs of not engaging in CSR practices can be high. To enhance the positive impact of CSR practices on society and on firms' competitiveness, firms should understand how CSR is evaluated by different stakeholders (Ö berseder et al., 2013) and specifically by consumers. This field of research requires greater development (Ö berseder et al., 2014). We contribute to this stream of literature by analyzing what CSR practices consumers recognize and intend to reward.
The hotel sector has high visibility in society, particularly with regard to its impact on the communities and the environments where hotels are located. Therefore, the analysis of and decisions about CSR practices in this sector may have strong effects on the behavior of a wide range of agents, such as government agencies, travelers, and local communities, and specifically on the behavior of hotel consumers. The perceived value of CSR differs depending on which firm stakeholder is evaluating the actions (Campbell, 2007;Jamali, 2008;Jones et al., 2017). In this sense, consumers are one of the most important company stakeholders (Jones et al., 2017). An examination of consumers' perceptions of the relevant CSR actions and how these perceptions affect consumers' intended future behavior and evaluations can help firms understand the importance of introducing CSR practices. Moreover, appropriate choices of CSR activities can contribute to firms' competitiveness and to the overall impact of CSR on society if stakeholders support these actions and legitimate these companies' behavior.
The literature still lacks consensus regarding the dimensions that consumers think CSR should include (Xiao et al., 2017). The present research adopts Carroll's (1979Carroll's ( , 1991 theory and the theory of sustainable development (Panapanaan et al., 2003;Panwar et al., 2006) in proposing different CSR dimensions. Using the two approaches, we extend the analysis of previous literature. The valuable study by Xiao et al. (2017) analyzes the effect of consumers' perceptions of the four dimensions proposed by Carroll (1979Carroll ( , 1991 on the overall consumers' support. Likewise, Xiao et al. (2017) analyze whether there are differences among the different hotel categories. Our study tries to extend some of their results by considering two theories on the proposition of the CSR dimensions and other aspects that could moderate the relationship.
The development of a CSR measure specifically for the hotel sector and from consumers' perspective is very recent, and only Xiao et al. (2017) have analyzed the separate effect of each of the dimensions of CSR on the hospitality sector. The importance of analyzing the separate effects lies in disentangling the value consumers attribute to different CSR actions. According to Green and Peloza (2011), consumers may have different responses depending on the type of CSR activity analyzed. Loureiro et al. (2012) stated that some dimensions do not influence consumer responses. The recent economic crisis calls for a new evaluation of what consumers perceive to be important actions with regard to CSR. Moreover, previous findings show differences between consumers' evaluation of the dimensions (e.g. Currás- Pérez et al., 2018;Loureiro et al., 2012), especially in the case of the economic dimension and specifically within the tourism sector (e.g. Xiao et al., 2017). Therefore, the specific dimensions that consumers value remain unclear (Xiao et al., 2017). Examining the different dimensions separately rather than using an overall measure of consumers' perception of CSR also allows the interaction of these dimensions with other variables to be analyzed. In this study, we propose that consumers' search for information about hotel CSR practices has a moderating effect on the proposed relationships between the CSR dimensions and consumers' support. Prior studies find that customers may reward CSR involvement if they are aware of what firms are doing with regard to responsible actions of the firm (McWilliams and Siegel, 2001;Sen and Bhattacharya, 2001). In this vein, the degree of consumer information about CSR practices influences consumers' perceptions. Consumers who are aware of CSR practices have more information; consequently, they may perceive different values for different CSR dimensions. In addition, Servaes and Tamayo (2013) establish consumer awareness of hotel CSR practices as a necessary condition to modify consumer behavior.
Following this logic, the aim of this research is to analyze the relationship between the perceived value consumers attribute to the different dimensions of CSR and consumers' support for hotels while also considering the moderating effect of consumer searches for information regarding hotels' CSR practices.

Hypotheses development
To approximate a conceptualization and measure of CSR, scholars have frequently used Carroll's pyramid framework (Carroll, 1979(Carroll, , 1991. This model ranks firms' responsibilities from economic to legal, to ethical and to philanthropic, situating the most basic responsibilities at the bottom of the pyramid and those that contribute to avoiding harm and to improving the quality of life at the top. Carroll's framework thus requires a multidimensional concept to adequately reflect all the CSR components (Alvarado-Herrera et al., 2017;Aupperle et al., 1985;Maignan, 2001;Xiao et al., 2017).
The theory of sustainable development (Kakabadse et al., 2005;Panapanaan et al., 2003;Panwar et al., 2006) and the related concept of corporate sustainability contribute to the comprehension of the construct and facilitate the translation of the model into companies' operating objectives (Panwar et al., 2006;Van Marrewijk, 2003). There is confusion in the literature regarding the use of similar terms from different theoretical approaches. Specifically, Montiel (2008) reviews the concepts of CSR and corporate sustainability and encourages collaboration between the two fields due to their similarities and the opportunities to further academic progress because both concepts share the vision of balancing economic, social, and environmental issues. Several reviews have been conducted on the relationship between the two concepts (e.g. Chang et al. 2017;Linnenluecke et al. 2009;Lozano, 2012;Montiel, 2008;Schwartz and Carroll, 2008;Steurer et al., 2005;Van Marrewijk, 2003). In a more recent paper, Ashrafi et al. (2018) also claim to integrate the two fields. The concept of sustainable development originates more at the macro level of analysis, while the term CSR originates more at the organizational level; however, sustainable development also requires contributions from organizations to achieve sustainability. The Brundtland Report (WCED, 1987) is attributed to be the starting point in defining sustainable development. The concept then began to be analyzed in the management literature as well (e.g. Bansal, 2005;Gladwin et al., 1995). This concept emphasizes the long term and originally focused only on environmental aspects before later evolving to include the social, environmental, and economic pillars due to the need to take into account all of a business's processes that influence sustainable development. This approach gained more popularity when Elkington (1998) linked the idea of the "triple bottom line" to the concept of sustainability (Carroll, 2015), and firms began to present social reports through the prism of the three aspects of their economic, social, and environmental performance.
Environmental sustainability is understood as the efficient use and protection of natural resources and the environment, and social sustainability denotes the social well-being of all the society and of the stakeholders of the firm. Therefore, the sustainable development approach includes environmental issues as a separate dimension of the responsibilities in the CSR concept, even though they are not specifically included in Carroll's framework (Clarkson, 1995;Maignan and Ferrell, 2004;Mitchell et al., 1997).
Taking the two approaches into consideration, CSR can be better represented by economic, legal, environmental, ethical, and philanthropic aspects. The economic aspect is present in both approaches. Economic sustainability refers to economic profitability, employment, and other aspects that allow firms to obtain utilities by producing and selling quality products and services. The distinctions among the legal, ethical, and philanthropic dimensions of Carroll's model are also relevant, especially given that the former considers firms as corporate citizens who are subject to legal norms. Nevertheless, a proactive vision of CSR cannot consider only legal requirements. Moreover, legal rules are not representative of all of the concerns of stakeholders and are sometimes obsolete in the context of rapid social and technological changes in developed countries. Therefore, the ethical dimension has special relevance because it refers to the appropriate conduct that goes beyond the established law. These dimensions are also different from the philanthropic dimension, which refers to charitable activities to help humanity (Carroll, 1998). The social dimension of the sustainable development approach is embodied in other aspects of Carroll's dimensions, as it refers to the general well-being in society. Additionally, according to a more recent review of research on hotels (Serra-Cantallops et al., 2018), social dimensions are mainly reflected in philanthropic actions.
Carroll himself (Schwartz and Carroll, 2003) recognized that some of the practices included in each category can overlap, especially in the case of the ethical dimension. Ethical issues permeate the other categories in the pyramid (Carroll, 2016). Nevertheless, he maintains the categorization with a separate ethical dimension to reflect practices that, for example, are different from the legal dimension because they go beyond the minimum required by law (Carroll, 2016). Moreover, as in the case of the philanthropic dimension, in which activities can be motivated by ethical considerations (Carroll, 2015), the environmental dimension can also be motivated by ethical or legal considerations. Nevertheless, the importance of the environmental dimension as a pillar of the concept of corporate sustainability and the importance of environmental issues in resource constraints and in their contribution to future generations is sufficient to consider it as a separate dimension. Furthermore, empirical studies that apply Carroll's model sometimes lack items that specifically represent environmental aspects (e.g. Maignan, 2001;Xiao et al., 2017).
The inclusion of the environmental dimension has great relevance in studies regarding the tourism industry, given the impact of this dimension on companies' surroundings . The tourism sector has substantial impacts on the physical environment (Jamrozy, 2007;Martínez et al., 2013;Timur and Getz, 2009), and the importance of the physical environment's degradation makes the environmental responsibilities of tourism companies even more relevant. In addition, the location of most hotels in resource-constrained surroundings (Bohdanowicz et al., 2011) makes the tourism sector one of the most important with regard to environmental issues. Therefore, especially in this sector, including a distinct environmental dimension can contribute to a better understanding of hotels' responsibilities. Studies focused on the CSR concept are now including environmental responsibilities in their analyses (e.g. Ö berseder et al., 2014;Pérez et al., 2013). Moreover, the importance of environmental aspects to consumer perceptions and evaluations has also been examined in other studies (e.g. Chen, 2010;Pérez et al., 2013). Some scholars have also added the environmental dimension to other theoretical schemes (e.g. Liu et al., 2014) to examine consumers' perception of its value. Accordingly, we propose that consumers are likely to differentiate among these five dimensions. Therefore, this study proposes analyzing the economic, legal, environmental, ethical, and philanthropic dimensions.
The analysis of the separate effects of the CSR dimensions is relevant because it remains unclear what specific dimensions consumers value (Xiao et al., 2017). Some authors propose that not all of the dimensions affect consumer responses (e.g. Loureiro et al., 2012).
With regard to the economic dimension, mixed results have been found. Some studies show a positive relationship between the economic dimension and consumers' support for CSR (e.g. Brown andDacin, 1997: Walsh andBeatty, 2007;Wang et al., 2003), while others have found a nonsignificant relationship in this regard (e.g. Podnar and Golob, 2007). Furthermore, Aupperle et al. (1985) report a negative relationship between this dimension and the other CSR components. The most recent study by Xiao et al. (2017) finds a negative effect of the economic dimension on consumers' support for CSR. However, the authors collected data in 2011, and they note that consumer perceptions may change over time and may vary in different countries because of differing cultural contexts. Other authors argue in favor of including economic responsibilities when analyzing consumers' perceptions of CSR because the recent economic crisis made consumers more concerned about economic sustainability and avoiding unemployment, insecurity, and financial risks (Choi and Ng, 2011). According to the underlying theories developed by Carroll and in the sustainable development literature and considering that the present research is conducted in a different context and period from Xiao et al.'s (2017) study, we argue that the economic dimension has a positive effect on consumers' support. To develop an image of a firm, consumers are expected to value firms' compliance with the aspects included in the traditional economic function. Firms create value by efficiently providing quality products that consumers demand or desire (Torugsa et al., 2013), and consumers are one of the main stakeholders directly affected by a firm's capacity to produce and sell quality products and services.
Some studies also state that the recent economic crisis and numerous firm scandals likely have had an effect on consumers' perceptions and sensitivity regarding the importance of legal and ethical firm responsibilities (Xiao et al., 2017). A firm's responsibilities must be carried out within the context of legal requirements. A firm's reputation and sales will be affected if companies do not fulfill their legal requirements (Park et al., 2014). In addition, today's "litigious society" increases the importance of this dimension for consumers (Carroll, 2016: 6).
Perceptions of firms' ethical behaviors are also rewarded by consumers' support (Bendixen and Abratt, 2007;Creyer and Ross, 1997). In a recent study, Ferrell et al. (2019) conclude that ethical responsibilities are critical to consumers' attitudes toward firms.
A firm's philanthropic actions are also expected to have a positive influence on consumers' attitudes, and empirical evidence confirms this relationship (Brown and Dacin, 1997;Sen and Bhattacharya, 2001;Yoon et al., 2006). These responsibilities reflect society's "desire to see businesses involved in the betterment of society" (Maignan and Ferrell, 2001: 39), and as a consequence, they improve consumers' perceptions.
Furthermore, environmental CSR aspects are easily understandable by the general public and consumers (Liu et al., 2014;Rahbar and Whahid, 2011). Global climate change and specific related aspects such as habitat destruction and rising sea levels are increasingly stimulating more environmental consciousness among consumers (Choi and Ng, 2011). Scholars have tested the effects of ecological issues on public attitudes toward the image of companies in both empirical studies and through experimental approaches (e.g. Montoro-Ríos et al., 2006).
This study analyzes the relationships between the different CSR dimensions and consumers' support for CSR. Consumer's support can include attitudes, intentions or behaviors. We use the term "consumer's support" specifically to refer to perceived firm reputation, firm selection and future purchase intention in the same sense as other authors employ "overall CSR support" to refer to behavioral intentions (e.g. Xiao et al., 2017) or "consumer behavior" and employ "consumer responses" to refer to a wide range of cognitive, affective, and behavioral outcomes among consumers (e.g. Golob and Podnar, 2018;Sen et al., 2006). In fact, some studies have analyzed the effects of CSR on purchase intentions (e.g. Murray and Vogel, 1997), on preferences (e.g. Liu et al., 2014) and on corporate reputation (e.g. Su et al., 2015). After a review of the literature, Aguinis and Glavas (2012) outline increases in the reputation of the firm, evaluations of products and the company, choice of the company/product, and customer loyalty as the main outcomes of CSR. Overall, the literature notes that CSR positively affects firm reputation, firm image, brand value, and consumer attitudes toward the firm and reinforces future buying intentions (Bhattacharya and Sen, 2004;Brammer and Pavelin, 2006;Brown and Dacin, 1997;David et al., 2005;García de los Salmones et al., 2005;Sen and Bhattacharya, 2001;Smith and Higgins, 2000;Varadarajan and Menon, 1988;Yoo and Lee, 2018). Therefore, consumer attitudes and intentions are among the variables that are most affected by CSR (Bhattacharya and Sen, 2004). Accordingly, in this study, we focus on analyzing the effect of CSR on firms' perceived reputation and on behavioral intentions, specifically purchase intention and establishment selection. If a consumer attributes importance to the different dimensions of CSR, then CSR likely also influences the formation of the consumer's attitude toward a firm and their behavioral intentions, their company preferences, and their perception of the firm's brand image, thereby potentially increasing the consumer's overall positive impression of the firm (Murray and Vogel, 1997) and the possibility that the consumer will select that firm (Arora and Henderson, 2007). Firms engaging in CSR practices are more likely to meet stakeholders' expectations and specifically consumers' expectations and, therefore, those firms will have more opportunities to increase their reputation (Jalilvand et al., 2017;Lamberti and Lettieri, 2009).
Accordingly, the first hypothesis states as follows: H1: The economic, legal, environmental, ethical, and philanthropic dimensions of CSR have a positive association with (a) firms' perceived reputation, (b) the possibility of firm selection, and (c) future purchase intention.
Furthermore, the proposed relationships may even be stronger if the consumers are aware of hotels' CSR practices. The importance of CSR awareness has been analyzed as one of the relevant factors in the ability of CSR to generate a positive impact on firm value (Rhou et al., 2016). This relevance is consistent with information processing theory (Miller, 1956). This theory is grounded on the idea that human behavior is influenced by the information processed, and it analyzes the stages in which a person processes information. According to Ricks (2005), this theory would predict that a positive perception of CSR activity promotes a positive view of the firm carrying out the activity and subsequently the attitudes and behaviors toward the firm. Based on this theory, Tian et al. (2011) state that the first phase in the information processing stages regarding CSR corresponds to consumer information or consumers' awareness of the CSR practices. Their study demonstrates that higher levels of consumers' awareness of CSR are related to more positive responses to CSR. Fatma and Rahman (2016) also state that consumers' positive attitude toward the firm and behavioral intentions is stronger if consumers are aware of the firm's CSR practices.
Ö berseder et al. (2011) note that most studies of consumers and CSR share the problem of assuming that consumers are aware of CSR practices, when this is not always the case. Consumers' lack of awareness of CSR initiatives is a major limiting factor in those consumers' ability to respond to these initiatives (Bhattacharya and Sen, 2004;Schuler and Cording, 2006;Pomering and Dolnicar, 2009). In contrast, if consumers are aware of CSR activities, their intentions toward the firm may be more positive (Barmer and Gray, 2000;David et al., 2005;Tian et al., 2011;Xiao et al., 2017) and their response to CSR initiatives could improve (Bhattacharya and Sen, 2004;Schuler and Cording, 2006). McWilliams and Siegel (2001) even state that potential customers should be fully aware of a firm's CSR activities. Some studies conclude that a positive relationship exists between the information consumers have about CSR and their purchasing behavior in particular (Brown and Dacin, 1997;Mohr and Webb, 2005;Sen et al., 2006). In fact, a poor understanding of firms' CSR practices has been shown to explain the disconnect between attitudes and behaviors (Bhattacharya and Sen, 2004;Mohr et al., 2001;Pomering and Dolnicar, 2009).
Specifically, this gap has been analyzed from the perspective of the theory of planned behavior (Ajzen, 1985(Ajzen, , 1991. This approach is one of the most relevant theories used to explain consumer behavior with regard to CSR issues in particular (e.g. Chen and Tung, 2014;Han and Kim, 2010;Kang et al., 2012). The framework analyzes intentions as the main antecedents of behaviors and considers that intentions are determined by attitudes, subjective norms, and perceived behavioral control. Some studies focus on analyzing specific aspects of this theory, such as attitudes and perceptions, as antecedents of consumer intentions or behavior (De Pelsmacker and Janssens, 2007). Moreover, additional variables can also be added to this model (Ajzen, 1991;Ozcaglar-Toulouse et al., 2006). Along this line of research, information has been proposed as an important variable in the model (e.g. Pérez and García de los Salmones, 2018). Bamberg et al. (2003) analyze how the predictors of behavioral intentions proposed in the theory of planned behavior are contingent on new information.
Given these ideas, we propose that CSR dimensions and their relationship with consumers' support may be strengthened if consumers are aware of hotels' CSR practices. In the context of buyer-supplier relationships, Homburg et al. (2013) also propose the importance of CSR awareness as a moderating factor. Specifically, our hypothesis states that consumers' search for information about hotels' CSR practices strengthens the positive relationship between CSR dimensions and consumer' intention toward the firm and its perceived reputation. Consumers who are more aware of hotels' CSR activities are hypothesized here to attribute more importance to CSR dimensions in the formation of their attitudes and intentions toward the firm. In the same sense and following the theory of planned behavior, Schuler and Cording (2006) integrate the importance of information for the theory. A lack of this information about a specific company limits consumers' responses to CSR activities (Bhattacharya and Sen, 2004;Schuler and Cording, 2006). Hildebrand et al. (2011) also propose the relationship between CSR and firm reputation as being contingent on the awareness of a firm's actions.
The rationale of the hypotheses we propose is that CSR has a greater influence on consumers' support (reputation, selection of the establishment, and purchase intention) if the consumers are also aware of the hotel's CSR practices (specifically, by searching for information about the hotel's practices). The hypothesis does not distinguish among information searches for each dimension of CSR practices. Craig Smith et al. (2010) examine how consumers' awareness of one CSR dimension affects their positive perceptions of the other dimensions, even without any information about the firm's practices with regard to those dimensions. Such awareness could, therefore, enhance the positive influence of the different CSR dimensions on the consumer's support (Rhou et al., 2016).
A consumer's beliefs about whether a company is socially responsible are greater when the consumer is aware of the firm's CSR activities (Du et al., 2007), and an information search for a hotel's CSR practices could lead to greater levels of awareness. Together with the importance that consumers attribute to CSR in hotels, a specific information search about a hotel's CSR practices could, therefore, strengthen consumers' support, given that CSR practices lead to stronger positive attitudes and intentions (e.g. Pomering and Dolnicar, 2009;Sen et al., 2006). Accordingly, our second hypothesis states as follows: H2: The positive association between the economic, legal, environmental, ethical, and philanthropic dimensions of CSR and (a) the firm's perceived reputation, (b) the possibility of firm selection, and (c) future purchase intention is strengthened when the consumer searches for information about the hotel's CSR activities.

Methodology Data
The data were collected through an online questionnaire addressed to Spanish consumers. Before the beginning of the survey questions, a letter was included in which we introduced the researchers, described the aim of the study, ensured the complete anonymity of the responses and the aggregated treatment of the data, and explained the voluntary character of the survey and the observation that there are no right or wrong answers (Podsakoff et al., 2003). For the fieldwork, firms and regional and local organizations affiliated with the tourism sector signed collaboration agreements to distribute the questionnaire on their respective websites and social networks (Valencian Tourism Agency via Invat.tur; the Catalan Tourism Agency, Segittur; Barcelona Tourism; Minube, Escapada Rural; Chic Travelling). The data were gathered between July and September 2015.
Given the impossibility of arriving at a probabilistic sample of hotel customers in Spain, we used a weighting approach to balance the sample and make it representative of the population (Brick and Kalton, 1996;Elliot, 1991;Hinkin, 1995). This approach required taking into account the ideal population. To do so, demographic data from the Residents' Travel Survey from the Spanish Statistical Office (https://www.ine.es/dyngs/INEbase/en/operacion.htm?c¼Estadistica_ C&cid¼1254736176990&menu¼ultiDatos&idp¼1254735576863) were used to parameterize the sample by age, gender, and employment situation. Through this method, we obtained a weighted representative sample with an efficiency value of 36% and an error of 2.50%.
The final sample included 3585 valid answers. The demographic information is presented in Table 1.

Measurement of variables
Dependent variables. The three dependent variables that were included in each of the regressions to represent consumer support were perceived corporate reputation, selection of the establishment, and future purchase intention (with values ranging from 1 ¼ Totally disagree to 5 ¼ Totally agree). Single-item measures have been employed in the marketing literature to measure single perceptions (e.g. Bergkvist and Rossiter, 2007;Du et al., 2007;Yoo and Lee, 2018) as well as corporate image (e.g. Singh et al., 2008). We follow this approach to measure the three dependent variables in this study. Perceived corporate reputation was evaluated with the following item: "CSR improves the hotel's reputation," which is adapted from Sen and Bhattacharya's ( 2001) study. In the case of establishment selection ("I'll choose a responsible hotel"), the item was also adapted from other studies (Liu et al., 2014;Tian et al., 2011). Following other scholars (David et al., 2005;De Pelsmacker et al., 2006;Green et al., 2018;Sen and Bhattacharya, 2001), purchase intention was also measured with a single item ("I'll purchase services from responsible hotels").
Control variables. Given that demographic characteristics may influence consumer perceptions, specifically in the hotel sector (Heo and Lee, 2011), three control variables were introduced.
Educational level was measured in the questionnaire through a question with four possible answers. Thus, we created four dummy variables: Studies_1 is equal to 1 if an individual has primary or secondary education and 0 otherwise; Studies_2 is equal to 1 if an individual has a degree in job training or a lower degree at university and 0 otherwise; Studies_3 is equal to 1 if the individual has a degree in undergraduate studies and 0 otherwise; and Studies_4 if the individual has a master's degree or PhD and 0 otherwise. In the regression, we exclude Studies_1 since it is our baseline category.
Gender is coded as 0 for male and 1 for female. Finally, the level of personal monthly income is asked on the questionnaire as a question with four categories. Accordingly, we created four dummy variables: Incomes_1 is equal to 1 if an individual reported to have a monthly income level lower than 1000 Euros and 0 otherwise; Incomes_2 is equal to 1 if the individual's monthly income is between 1000 Euros and 2000 Euros and 0 otherwise; Incomes_3 is equal to 1 if the individual's monthly income is between 2001 Euros and 3500 Euros and 0 otherwise; and Incomes_4 is equal to 1 if the individual's monthly income is higher than 3500 Euros and 0 otherwise. Then, we introduced these dummy variables in the regression, considering Incomes_1 as a baseline category. Moderating variable. The moderating variable is consumers' CSR information, which is measured as consumers' search for information about the hotel's overall CSR practices, and it ranges from 1 (I never search for information about a hotel's CSR practices) to 5 (I always search for information about a hotel's CSR practices). Following other studies (e.g. Homburg et al., 2013), the variable measures whether consumers search for information about a hotel's overall CSR practices because we wanted to know whether consumers are aware of any of the hotel's CSR activities. Authors such as Du et al. (2007) have also measured CSR awareness with a single-item measure.
We also measure the correlation between the moderating variables and the other indicator in the questionnaire about consumers' previous behavior. This was a yes/no question that asked consumers whether they had previously purchased a stay in a responsible hotel. The correlation is 0.341 (p < 0.01), indicating that there is a relationship between more informed consumers and consumers who know that some of their previous stays were in a responsible hotel, a fact which provides consistency to the measurement.
CSR dimensions. To develop a measure of CSR dimensions, we follow the traditional guidelines proposed in the literature for construct specification, item selection, purification, and scale validation (Churchill, 1979;DeVellis, 2003). After an extensive review of the published CSR scales from the consumer perspective, we performed an assessment conducted by a panel of experts in the tourism sector. The experts were selected the following guidelines that set out by Skjong and Wentworth (2001) and Ouellet (2007), which stated that the experts should have (1) experience in making judgments and taking decisions based on evidence or knowledge; (2) a reputation in the academic community, a research profile, knowledge of the scientific method, and the subject (CSR in this case); (3) a willingness and availability to participate; and (4) impartiality and inherent qualities such as self-confidence and adaptability.
Following these guidelines, we finally obtain the collaboration of 10 academics that have experience in the tourism sector. We presented the initial set of indicators to the experts. As the literature recommends (DeVellis, 2003;Ouellet, 2007), the experts provided recommendations about the importance and adjustment of the indicators, the indicators' adequacy to measure the different company responsibilities, the necessity of retaining or dropping the items, or the necessity of including other dimensions or indicators. The experts were then asked to rate the importance degree of the resulting indicators, and only those who were clearly representative for 7 of the 10 experts (more than 66.7% of the experts) and representative at least to some extent for the other 3 experts were retained (Ouellet, 2007). Finally, after these recommendations and the authors' assessment, a battery of 44 items was proposed to measure the different categories in CSR.
The answers reflect the extent to which respondents agree with the importance of each statement regarding CSR in hotels and were categorized using a five-point Likert-type scale ranging from 1 ¼ Totally disagree to 5 ¼ Totally agree. Although a wide sample of studies was reviewed, the 21 items included in the study were adopted from proposed indicators used in the following  Martínez et al. (2013), and Zahra and La Tour (1987).
A series of exploratory factor analysis were performed, and items below the cutoff of 0.40 were dropped, resulting in 33 items. In the final exploratory factor analysis, all the items loaded comfortably above the 0.50 threshold (Tabachnick and Fidell, 2001) and on five factors (see Table 2), as proposed in the literature review. The 33 items represented 63.63% of the variance explained (18.29% legal dimension, 15.69% ethical dimension, 12.40% economic dimension,  8.80% philanthropic dimension, and 8.55% environmental dimension), with eigenvalues ranging from 1.34 to 13.02. The internal consistency of each variable is shown by the item-to-total correlations exceeding the threshold of 0.3 in all cases. The economic factor included 5 items (a ¼ 0.815); the legal factor included 4 items (a ¼ 0.855); the environmental factor included 10 items (a ¼ 0.929); the ethical factor included 6 items (a ¼ 0.891); and the philanthropic factor included 8 items (a ¼ 0.916). Therefore, the items show internal consistency, with as supporting initial content validity (Hinkin, 1998).
A confirmatory factor analysis (CFA) was then performed. The results are presented in Table 3. A total of 21 indicators were ultimately retained: 3 in the economic dimension, 3 in the legal dimension, 7 in the environmental dimension, 5 in the ethical dimension, and 3 in the philanthropic dimension. All of the indicator loadings in the proposed dimensions were significant (Table 3) and ranged from 0.633 to 0.868. Thus, they were above the threshold of 0.6 and demonstrated convergent validity (Gerbing and Anderson, 1988).
All the dimensions showed an average variance extracted (AVE) higher than 0.5 (Table 3; the lowest was 0.564 in the economic category, and the highest was 0.638 in the legal category), suggesting convergent validity (Hair et al., 1999), which was further assessed with composite reliability values (Hair et al., 1999) that ranged from 0.793 (economic) to 0.920 (environmental; Table 3). Regarding discriminant validity, all the correlations were significant and below the cutoff of 0.70 (Pallant, 2007), and the AVE for all the latent variables was higher than the squared correlations were (see Table 4). Thus, discriminant validity was also demonstrated (Fornell and Larcker, 1981). Discriminant validity was also tested using the confidence interval of the correlations (Anderson and Gerbing, 1988;Bagozzi, 1994), which included the value of 1 in each case, providing an additional evaluation of discriminant validity. Moreover, the goodness of fit indexes of the measurement model showed appropriate values, indicating that the model adequately fits the data: confirmatory fit index ¼ 0.969, incremental fit index ¼ 0.969, normed fit index ¼ 0.961, nonnormed fit index ¼ 0.964, and root mean square error of approximation index ¼ 0.031.

Analytical techniques
The hypotheses were tested using a hierarchical regression analysis. To introduce the variables in the hierarchical regression, each CSR dimension was calculated as the mean score from the items included, thus generating composite scores for all the dimensions. Table 5 presents the descriptive statistics and correlations of all the variables included in the series of models.
Although the analysis of the variance inflation factors shows that multicollinearity is not a problem (with the highest factor at 2.150, comfortably below the cutoff level of 10) (Cohen et al., 2003), to reduce multicollinearity, the main variables were mean centered to introduce them into the test of the moderating effects (Aiken and West, 1991;Cohen et al., 2003).
Three series of models were estimated, one for each of the dependent variables: perceived firm reputation, possibility of establishment selection, and future purchase intention. Each model was estimated in three steps by introducing the control variables first, then the main variables, and finally the interaction terms. Then, the improvement in each step was analyzed with the significance of the change in R 2 through an F-test.

Results
Regarding the first hypothesis, all the CSR dimensions show a positive relationship with the three dependent variables, except in the case of the ethical and economic dimensions, which present a positive and significant relationship only with perceived corporate reputation (see Table 6). Therefore, hypothesis 1 is partially supported.      In each model, the addition of the interaction terms produces significant increases in R 2 , confirming some of the proposed moderating effects. Specifically, when the dependent variable is perceived firm reputation, the introduction of the interaction terms increases the variance explained by 3.1% (dF ¼ 22.207, p < 0.01). Regarding the series of models with the importance of establishment selection, model 6 with the interaction terms results in an increase in R 2 of 14.6% (dF ¼ 119.557, p < 0.001). In the case of the dependent variable future purchase intention, model 9 increases the variance explained by 11.6% (dF ¼ 90.540, p < 0.001) over model 8. Therefore, the increases are significant, thus confirming the following moderating effects: a positive moderating effect of consumers' search for information in the relationship between the ethical dimension and the evaluation of firm reputation; an unexpected negative moderating effect in the relationship between the philanthropic dimension and all three dependent variables; and a significantly positive moderating effect in the relationships between the economic and ethical dimensions and the importance of establishment selection. Therefore, we find only partial support for hypothesis 2.
Considering that the increases in the R 2 values due to the interaction term appear to be moderate, we have performed some additional robustness checks on the moderator effects. We thus carried out a Chow test (Chow, 1960), comparing the differences between less informed and more informed consumers. The sample was sorted in ascending order of the moderating variable (information search). The top and the bottom 35% of cases were selected, and the remaining 30% of the cases were omitted, in accordance with standard econometric procedures (Golfeld and Quant, 1965;Gounaris and Venetis, 2002;Kohli, 1989). A series of regressions were then estimated for each set of dimensions with the interactions that were significant in the previous analyses. The differences in the sum of the squared residuals from the new sample (n ¼ 2510) and the regressions on each subsample (low-and highinformed consumers) were then introduced to calculate the Chow test. The results show significant differences in the regression coefficients of both groups. In Table 7, we summarize the main statistics, confirming the previously described moderating effects, which therefore provide additional support for hypothesis 2 with respect to the economic, ethical, and philanthropic dimensions.

Discussion
The results of this study confirm the importance of separately analyzing the CSR dimensions and their effect on different aspects of perceived consumer support as well as the effect of consumer information in these relationships. The positive impact on both consumer perceptions of Carroll's traditional dimensions and the environmental dimension proposed by the theory of sustainable development is confirmed, although the impact was only partially confirmed in the case of the economic dimension. In contrast to the findings of Xiao et al. (2017), which indicated a negative effect of the economic dimension, our results show a positive relationship with perceived consumer support, although only in the case of firms' perceived reputation. A nonsignificant relationship is found in the case of establishment selection and future purchase intention. The strong effect of the economic crisis on perceived consumer support for traditional economic aspects of firms is noted in Xiao et al.'s (2017) study. Nevertheless, the temporal period in which the present research was conducted is farther away from the worst years of the economic crisis, and consumers' evaluation of the economic dimension may have changed. Moreover, cross-cultural differences could also be a reason for the different perceptions obtained in our sample compared with the perceptions of the US consumers analyzed by Xiao et al. (2017). According to the macroeconomic data, the effects of the economic crisis appear to have lasted longer in Spain. Although we cannot test these arguments because in our study, we neither compare two different periods nor compare consumers in different countries, it appears that consumers' perceptions of this dimension change over time.
The environmental dimension was also found to be strongly perceived by consumers. It is the dimension with the highest coefficient in the first two models' series. Therefore, consumers distinguish between all the CSR dimensions proposed in this study. The exploratory and confirmatory analyses and the validity of the different dimensions also support the necessity of including all the proposed dimensions.
The importance of consumers' information about hotel CSR practices was also partially tested. This analysis supported the positive impact of the ethical dimension and the economic dimension on some aspects of consumers' support and, contrary to our hypothesis, weakened the positive relationship of the philanthropic dimension. These results lead us to different conclusions. The first involves the importance of consumers' information about hotel CSR practices in the configuration of consumers' image of a company and in influencing their intentions with regard to the economic, ethical, and philanthropic dimensions. Specifically, for more informed consumers, the ethical and economic dimensions have more importance than they do for less informed consumers, whereas in the case of the legal and environmental dimensions, we did not find differences between consumers. Moreover, the philanthropic dimension has less importance for the intentions of more informed consumers. A possible explanation for this unexpected result is that the philanthropic dimension may be perceived by more informed consumers as not reflecting the core of the company. In terms of Porter and Kramer's (2006) distinction between responsive and strategic CSR, informed consumers seem to attribute more value to strategic actions because they are more closely connected to the firm's operational context, such as in the case of the economic dimension, whereas responsive CSR is indirectly related to the main activity of the company, such as in the philanthropic dimension.

Conclusion
The results of this study have economic implications for firms. Understanding what consumers perceive as the practices that CSR includes is relevant because it encourages hotels to develop and prioritize their CSR activities. Previous research has focused on the valuable analysis of the relationship between CSR and financial performance and has provided arguments about firms' economic incentives to engage in CSR practices. CSR can produce economic rewards for firms (Chrisman and Carroll, 1984). Kurucz et al. (2008) outline that the business case for CSR includes its positive effects on competitive advantage, costs, the level of risk, and reputation. Our study furthers the comprehension of these positive effects that have economic implications since consumers are one of the main stakeholders of companies and legitimize the companies' actions through their buying behavior. An analysis of consumer intentions may be able to predict future consumer behavior. As a starting point, hotel companies can obtain valuable information from these results and based on the CSR practices most valued by consumers, decide the type of CSR on which their efforts should focus. Specifically, our study confirms that consumers distinguish between the economic, legal, environmental, ethical, and philanthropic CSR dimensions and affirms the value of considering the sustainable approach and Carroll's model with regard to theory. The environmental dimension is relevant for all of the analyzed aspects of consumers' support. The economic dimension has a weaker impact on some aspects of consumers' support, in line with the results of other studies, but this dimension still has a role in CSR since consumers recognize this dimension when evaluating a firm's reputation. The time and the country in which this research was carried out are different from the context of Xiao's study (Xiao et al., 2017), which also analyzes the separate impacts of different CSR dimensions on consumers' support. Their study shows a negative impact of this dimension, but they also suggest that consumers' evaluation of the importance of the different dimensions could evolve in time. Moreover, crosscultural differences between countries have also been examined with regard to CSR perceptions (e.g. Singh et al., 2008). Specifically, the general public awareness in Spain about sustainability issues developed later (Singh et al., 2008) than it did in other countries such as the United States, which was the context for Xiao et al. (2017). This consideration could be an explanation for the differences between the results of the two studies. The negative effects of the recent economic crisis also lasted longer in Spain, which could also be a reason why Spanish consumers attribute more importance to basic economic responsibilities. Moreover, for consumers that show more interest in searching for hotels' CSR information, the economic dimension of CSR is also important to selecting the establishment, which provides more support for the argument that consumers consider this dimension to be relevant.
Furthermore, hotel managers can also use these results to develop guidelines on how to differentiate their advertising with regard to the type of consumer (more or less active in searching for information). This last point is relevant due to the high costs of publicity and the consequent pressure for hospitality companies to make effective advertising investments (Assaf et al., 2017). For example, one of the larger Spanish international hotel chains devoted 12.1 million Euros in 2015 to publicity through different channels (Statista.com, 2018). To increase consumer behavior toward firms applying CSR practices, managers should facilitate CSR awareness (Ö berseder et al., 2011). According to our results, especially when consumers are choosing between different brand alternatives, the ethical and economic aspects of CSR seem to be more important for more informed customers. Therefore, to differentiate a hotel brand, managers should strengthen the information and advertising the firm provides regarding their ethical and economic responsibilities. To do so, managers should identify when this step takes place during the buying process in this segment of consumers. In contrast, the perceptions of the philanthropic dimension are less valued for more informed consumers, which is likely because more informed consumers attribute more value to the fact that CSR should be part of the core of the company and should not only be reflected in philanthropic activities that can be outside the main activity of the company. General advertising campaigns should emphasize philanthropic aspects because consumers distinguish and attribute importance to this dimension, but for the specific segment of more informed consumers, no emphasis should be given to this aspect.
From a macro perspective, sustainability helps to conserve and strengthen the value of tourism resources that attract visitors. From a micro perspective, achievements in this area can enable hotels to promote their sustainable activities, which can add considerable value given its potential influence on a consumer's decision to purchase services. CSR practices also provide an opportunity for consumers to act responsibly (Font and McCabe, 2017). Improving the understanding of consumers' perceptions can help companies better understand the value of adopting responsible practices, thus contributing to destination sustainability.
This study also has some limitations. The cross-sectional data do not allow us to derive conclusions about the causality of the proposed relationships. Panel data, by maintaining a sample of consumers over time, could improve the analysis in future research. Moreover, the study only asked consumers about their intentions and perceptions, whereas their actions were not tested (e.g. regarding their final purchasing behavior). In addition, in this study, we did not control for firm effects. This approach is similar to other CSR studies (e.g. Xiao et al., 2017). Nevertheless, consumers' perceptions could differ across different brands. Future research should address these questions by asking customers about the CSR practices of some specific hotels whose actual practices are known to the researchers. Moreover, this study only considered consumers' general searches for information about CSR practices. Future research should link consumers' awareness of each CSR activity or dimension with the relationships proposed in this article by gathering data on the information consumers have about a specific brand with regard to each of the dimensions proposed.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received financial support by the "Conselleria de Economía Sostenible, Sectores Productivos, Comercio y Trabajo de la Comunidad Valenciana (España) (EBCORG/2018/4)."

ORCID iD
Montserrat Boronat-Navarro https://orcid.org/0000-0002-9062-4161 José Antonio Pérez-Aranda is director and teacher at the University School of Hospitality and Tourism CETT-University of Barcelona. His research focuses on the development of quality management systems, the environment, and corporate social responsibility. His thesis has received the Award for the Best Doctoral Thesis on Corporate Social Responsibility and Sustainability (2017 call) awarded by the Network of Santander Chairs of CSR (University of Málaga and Banco Santander). Likewise, the research carried out was recognized by the Generalitat de Catalunya, being a finalist in the "Guardons de Turisme" awards for the best tourism research project in 2017.