Does corporate social responsibility result in better hotel guest attitudinal and behavioral loyalty?

Abstract

This study aims to determine whether and how investing in corporate social responsibility (CSR) practices results in better attitudinal and behavioral loyalty. It seeks to comprehend the path linking CSR and customer loyalty through customer-company identification (C–CI), trust, and satisfaction as mediators. Partial least squares (PLS)-based structural equation modeling was applied to examine the theoretical model using SMART PLS. The data were collected through online questionnaires from 385 customers of Morocco's three-, four- and five-star hotels. The results conclude that CSR practices significantly contribute to customer loyalty through the mediating components of C–CI and trust. Surprisingly, unlike previous studies, the influence of CSR on customer satisfaction is insignificant. CSR influences customer satisfaction only through C–CI and trust. Moreover, the relationships between mediators, rarely explored by previous researchers, revealed that C–CI contributes to trustworthiness and customer satisfaction. Trust can also generate customer satisfaction.

1. Introduction

Managers find corporate social responsibility (CSR) to be an emerging marketing tool that generates competitive value. Similarly, customers, one of the most critical stakeholders, expect socially responsible behavior from companies [1]. One extremely valuable Key Performance Indicator (KPI) for an organization's financial and reputational performance is the retention rate of customer loyalty. Defining what makes a business profitable [2,3]. Scholars have been committed to discovering what contributes to customer loyalty to a brand [4] in all industries, particularly the hospitality sector. While investigating the impact of customer experience on loyalty, Cetin and Dinçer (2013) found out that a loyal customer contributes to a sustainable competitive advantage for hotels [5]. Similarly, several researchers [6,7] in the developed world have confirmed customers' willingness to pay more for hotels opting for CSR practices. Given the increasing importance of CSR and its influence on customers, hotels need to determine whether the investment in CSR activities is worth it. Does investing in CSR practices result in better attitudinal and behavioral loyalty? This study aims to comprehensively comprehend the path that links CSR and customer loyalty in Morocco's most significant sector: the hotel sector.

On the one hand, the intended contribution of this study is to position the study within one of the two distinct streams, namely Freeman's theory or Friedman's theory. Hence, if CSR promotes positive word of mouth, recommendations, and repurchase services, we have to position our research within the stream of Freeman's stakeholder theory, which supports socially responsible practices by satisfying the expectations of various stakeholders [8]. In this case, businesses are responsible for different stakeholders in society; thus, they cannot focus solely on increasing profits. Nevertheless, if CSR does not contribute to customer loyalty, the research positions itself within Friedman's utilitarian view that companies should fulfill their economic responsibilities by serving only their shareholders [9]. Throughout the study, we prove that either CSR is a must or that companies should focus only on maximizing their profits.

On the other hand, this study shows the social elements of consumer behavior that are impacted by CSR and lead to customer loyalty. Thus, companies will know which variables (C–CI, trust, and satisfaction) are affected by socially responsible activities and affect their attitudinal and behavioral loyalty. Hotel industry managers will benefit from this study because it provides them with the necessary tools to design their CSR practices while planning their strategies. Recommendations and implications from the research findings will help them acquire a competitive advantage.

This exploratory research suggests a strong theoretical framework that considers the gaps in the literature. Over the past few decades, many theories and approaches have been identified to explain why customers respond positively to CSR activities. Scholars have applied various theories, including the stakeholder theory [10,11], means-end value model theory [12], social exchange theory [11,[13], [14], [15]], and social identity theory [2,16,17].

Study utilizes a combination of social exchange and social identity theories as a foundation to explain the path that links socially responsible activities to customer loyalty. Social exchange theory refers to when one party favors the other because it expects a benefit in return [18]. According to this theory, companies will inevitably benefit from increased customer satisfaction and positive customer evaluation, which, in turn, generate retention and revenue for the company. In the same context, the social identity theory refers to how individuals’ perceptions of personality are based on their belonging to social groups [19]. Therefore, once the customer is a member of a hotel (has stayed at least once) and is highly engaged in CSR practices, positive attitudes and behaviors are generated.

Thus, this study combines two theories, social identity (C–CI) and social exchange (trust and satisfaction), which remain restricted [2]. To our best knowledge, except from the empirical research of Martinez and del Bosque (2013) previous hospitality research has not incorporated the role of C–CI as a social identity variable along with trust and satisfaction as a social identity variable on explaining the path from CSR to hotel customer loyalty.

Prior research stresses the importance of customer-company identification in linking socially responsible activities with customer loyalty but does not study it as a mediator along with trust and satisfaction impacting both attitudinal and behavioral loyalty. The process by which C–CI not only serves as a bridge linking CSR to loyalty but also as a precursor to trust and satisfaction represents an important advance in the literature within this field. This broader vision of the customer loyalty antecedents allows to have mediators from two different theories to explore within a single study.

Another research gap is that most previous studies used either the attitudinal or behavioral approach, while others combined them as one construct. To remedy this gap, the two dimensions of loyalty were considered separately in this study for a thorough understanding.

Moreover, several researchers have chosen the dimensions of CSR based on stakeholder theory; hence, the majority have considered the economic dimension of Caroll's pyramid. Little attention has been paid to companies' desired responsibilities. Hence, this study considers only the desired dimensions of Carroll's pyramid. Finally, the research on CSR has mainly been conducted in the developed world, with little guidance in the emerging context, specifically in the African continent.

From this perspective, our research aims to answer the following questions:

• -What is the path that links CSR to customer loyalty?
• -What is the relationship between CSR and the three customer behavior variables (trust, identification with the company, and satisfaction)?
• -What is the relationship between customer loyalty (attitudinal and behavioral) and its three antecedents (trust, identification with the company, and satisfaction)?
• -Is there a positive relationship among mediators?

2. Literature review and hypothesis development

2.1. CSR and its impact on mediators

Carroll [20] categorizes the four responsibilities (economic, legal, ethical, and philanthropic) of a business under new classifications (required, expected, or desired). The economic and legal responsibilities fall under the “required” category, whereas ethical and philanthropic responsibilities fall under the expected or desired category. However, some researchers [21,22] have modified Carroll's CSR pyramid to meet industry attributes. In the same way, this research has adapted the “desired” category of Caroll's pyramid [20] to match the attributes of the hospitality sector [22]. The philanthropic dimension of Carlo's pyramid was used to measure the impact of CSR on customer loyalty. Being a corporate citizen is a compulsory attribute of a socially responsible company. It is voluntary but highly desired by society. However, the environmental and societal dimensions were added to the desired CSR category. Environmental issues are among the most popular topics among countries, environmentalists, media, and communities [23]. The hospitality industry's involvement in environmental crises, such as global warming and water depletion, is not trivial [24]. The environmental dimension of CSR has been revealed as the most prominent component in the literature on CSR in the tourism industry [25].

Therefore, consumers have become increasingly sensitive to environmental issues. Hence, hospitality scholars [24,26] that analyzed CSR practices have primarily focused on environmental practices. However, the community component is also critical, and businesses should devote considerable attention to the well-being of society [27].

Accounting scholars [4,12,[28], [29], [30]] have considered the societal dimension in studying the impact of CSR on customer loyalty. Accordingly, this study considered three dimensions (philanthropic, societal, and environmental) to portray CSR practices.

Trust is considered the first achievement of a company's socially responsible practices [31]. The ethical and responsible principles of companies increase the trust of all stakeholders, including customers [32]. The company provides information on its CSR activities [27]. This information is used to improve trust in the company. Consequently, firms are expected to trust each other [33] Various scholars [3,34] used customer trust as a critical mediator to link the path between CSR and customer loyalty. Therefore, this study posits that CSR influences customer trust.

H1

CSR has a positive influence on customer trust.

Customer-company identification intensifies with CSR activities. Positive attitudes and behaviors are generated when a customer is a member of a hotel that is highly engaged in CSR practices. Communicating socially responsible practices is mandatory because customers identify with companies that know about their CSR practices [35]. CSR initiatives reveal a company's personality and create strong connections and identification with customers by reflecting unselfish values [36]. Furthermore, Pérez and Rodríguez del Bosque [28] and Li et al. [37] reveal a direct positive connection between CSR images and C–CI. Customers are eager to identify with companies that opt for CSR strategies to improve their self-esteem. Accordingly, this study proposes that CSR positively influences C–CI.

H2

CSR has a positive influence on customer-company identification.

Customers are satisfied with socially responsible companies [38]. In addition, CSR has been proven to positively influence customer satisfaction [39,40]. Socially responsible practices are considered market value drivers and are oriented towards making customers more satisfied with products and services [41]. Thus, the literature has shown that CSR impacts customer satisfaction [2]. Thus, customer satisfaction is a mediating variable that clarifies the relationship between PCSR and customer loyalty. Consistent with previous research findings that postulate the crucial role of customer satisfaction based on the socially responsible activities of companies, this study suggests that CSR positively influences customer satisfaction.

H3

CSR has a positive influence on customer satisfaction.

2.2. Relationship between mediators

To properly analyze the effect of the three constructs (trust, C–CI, and satisfaction) on customer loyalty, we explicitly considered the relationships among them. One of the most significant long-term social exchanges is trust [42]. This is an antecedent of the identified interactions [35]. Trust arises when a customer identifies with a company that communicates good value [43,44]. However, customers' self-definitions and self-esteem are communicated similarly [45]. Martínez and del Bosque [2] claimed the existence of trust in C-CI, but their findings did not support it. Accordingly, researchers have argued that C–CI may be a precursor of trust rather than an outcome [11]. Therefore, trust is based on the personal perception of a specific group's attributes and values [46]. In this vein, Glaveli [11] found that identification was driving customer trust in his research about CSR and customer loyalty.

Based on the previous discussion, this study postulated that C–CI positively influences trust.

H4

C–CI has a positive influence on customer trust.

The second precursor of customer satisfaction considered in this study is C–CI. Customers identify with companies that increase their satisfaction [47]. Therefore, a company's overall assessment is based on a high level of C–CI, which positively affects customer satisfaction [48]. However, they identify with the company before becoming customers; thus, C–CI precedes satisfaction [35,47]. Furthermore, the affective attachment linked with identification tends to enhance or decrease the company's level [48]. Therefore, based on this discussion, this study assumes that C–CI positively influences customer satisfaction.

H5

C–CI has a positive influence on customer satisfaction.

Customer trust is a precursor to customer satisfaction because it creates positive attitudes and emotions. Researchers have found that trust precedes customer satisfaction [49]. Indeed, trust is based on a pre-purchase experience, whereas satisfaction is based on a post-purchase experience [42]. In the hospitality sector, researchers [50,51] have proven a positive influence of trust on satisfaction.

When a hotel is believed to be trustworthy, satisfaction is created systematically. Therefore, trust is a precursor to satisfaction.

H6

Customer trust has a positive influence on customer satisfaction.

2.3. Customer loyalty and its antecedents

Oliver [52] defined customer loyalty as a profound dedication to regularly rebuying a favorite product or service in the future, resulting in recurring same-brand purchasing and word-of-mouth without marketing efforts. This concept is essential in the tourism industry and, more significantly, in the hospitality sector. Previous studies have shown that loyal customers contribute to a sustainable competitive advantage for hotels [5]. In addition, a rise of five percent in customer loyalty can increase profits by 25–80% in profit [53]. However, scholars have primarily examined one of the three main components of loyalty: attitudinal, behavioral, or composite. The attitudinal dimension refers to the emotional commitment to a given service or product, whereas the behavioral component of loyalty refers to the repetition of using or purchasing a particular service or product over time [53]. Finally, the composite dimension refers to choosing and recommending products or services [54]. Söderlund [55] criticized how the authors examined customer loyalty by considering only one component, which may have biased the results. Nevertheless, most authors [14,16,34] have assessed customer loyalty as a single dimension.

The comprehensive vision of loyalty supported in this study treats each dimension separately for a thorough understanding of how socially responsible activities impact customer loyalty. This study produces a conceptual framework for analyzing loyalty through three antecedents (trust, C–CI, and satisfaction), which are considered the main variables.

Trust has been studied in various disciplines, explaining the extensive literature and the multiplicity of definitions [56]. Organizational trust is the customer's belief that an organization will behave to protect its interests and maintain what the company promises [48]. The company can count on serving the long-term interests of its customers [57]. Overall, it is conceptualized based on credibility and benevolence [58]. Credibility is a company's capacity to maintain its promises and fulfill consumers' expectations. In contrast, benevolence is the consumer perception that a company cares beyond its economic profit. Trust is crucial for building relationships in the hospitality industry, and researchers such as Chaudhuri and Holbrook [48] have found a positive relationship between loyalty and trust. Accordingly, this study proposes that trust positively influences both CL dimensions of customer loyalty.

H7.a

Trust has a positive influence on customer behavioral loyalty.

H7.b

Trust has a positive influence on customer attitudinal loyalty.

The identification with a company also affects customer loyalty [35]. Identifying customers in certain businesses is understandable in today's social context, where many individuals need to identify with other groups, organizations, and companies to fully express their identities. C–CI refers to the cognitive state of customer closeness to a firm [59]. Various scholars [2,28,41] have revealed that customer–company identification positively influences loyalty. Therefore, this study suggests that C–CI positively influences customer behavioral and attitudinal loyalty.

H8.a

C–CI has a positive influence on customer behavioral loyalty.

H8.b

C–CI has a positive influence on customer attitudinal loyalty.

Finally, customer satisfaction is critical to improving a company's competitive advantage and retaining customers. This concept has been studied extensively in consumer behavior and marketing literature. A satisfied customer experiences a product or service that is equal to or greater than the expected experience. This refers to the customer's overall product performance assessment [60], which includes subjective reactions. It has also been viewed as a difference between perceptions and expectations [61].

Along with this definition, satisfaction represents a good or bad feeling arising from the distinction between pre- and post-consumption perceptions. Various scholars [2,16,29] have confirmed the positive influence of satisfaction on loyalty. Therefore, this study suggests that customer satisfaction positively influences customer behavioral and attitudinal loyalty.

H9.a

Customer satisfaction has a positive influence on customer behavioral loyalty.

H9.b

Customer satisfaction has a positive influence on customer attitudinal loyalty.

Based on the above hypotheses, this study proposes the research model shown in Fig. 1. The questionnaire items are displayed in Table I.

Fig. 1.

Proposed research model.

Table 1.

Questionnaire items.

Variable	Items code/items	References
Philanthropic (PHI) –CSR	PHI1. The hotel directs part of its budget to donations to social causes. PHI2. This hotel supports the culture and art activities of the local community. PHI3. Managers and/or employees of the hotel participate in voluntary and charitable activities. PHI4. The hotel provides free services/products to needy people. PHI5. The hotel assists voluntarily projects that enhance a community's quality of life.	Baden (2016)
Society (SOC) –CSR	SOC1. The hotel is committed to improving the welfare of the communities in which it operates. SOC2. The hotel actively participates in social and cultural events (music, sports, etc.). SOC3. The hotel plays a role in society that goes beyond mere profit generation. SOC4. The hotel provides a fair treatment of employees (without discrimination and abuse, regardless of gender, race, origin, or religion). SOC5. The hotel provides training and promotion opportunities for employees. SOC6. The hotel helps to solve social problems.	Martínez et al. (2013)
Environmental (ENV) –CSR	ENV1. The hotel protects the environment. ENV2. The hotel reduces its consumption of natural resources. ENV3. The hotel recycles its waste. ENV4. The hotel communicates to its customers its environmental practices. ENV5. The hotel exploits renewable energy in a productive process compatible with the environment. ENV6. The hotel conducts annual environmental audits. ENV7. The hotel participates in environmental certifications.	Martínez et al. (2013), Gürlek et al. (2017)
Behavioral loyalty (BL)	BL1. I would be loyal to the hotel. BL2. I will remain a customer of the hotel. BL3. The hotel opting for CSR activities will be my first choice compared to other hotel brands.	Zeithaml et al. (1996), Chang and Yeh (2017)
Attitudinal loyalty (AL)	AL1. I am likely to say positive things about the hotel. AL2. I will recommend the hotel to my family and friends. AL3. I would recommend the hotel if somebody asked my advice.
Customer-Company Identification (C–CI)	C–CI1. When someone compliments the hotel, it feels like a personal compliment. C–CI2. I tell others that I am proud to be a customer of this hotel. C–CI3. The hotel fits my personality. C–CI4. I feel closely linked to the hotel. C–CI5. I feel good to be a customer of the hotel. C–CI6. I'm very interested about what others think about the hotel.	Mael and Ashforth (1992); Kang et al., 2015 and Perez and del Bosque, 2015); Akbari et al. (2020)
Satisfaction (SAT)	SAT1. Overall, I am satisfied with this hotel service. SAT2. My decision to choose this hotel was a wise one. SAT3. I think that I did the right thing when I purchased this hotel service. SAT4. I feel happy about my decision to choose this hotel. SAT5. This hotel is exactly what I need for my accommodation.	Cronin et al., 2000; Martinez and del Bosque, 2013; Kang et al., 2015; Akbari et al., 2020
Trust (TRU)	TRU1. I trust in the quality of this hotel. TRU2. The hotel is safe to patronize. TRU3. The services of this hotel make me feel a sense of security. TRU4. This hotel is interested in its customers. TRU5. This hotel is honest with its customers. TRU6. Hiring services of this hotel is quality assurance.	Morgan and Hunt (1994) and Sirdeshmukh et al. (2002)<

3. Methodology

Our research population included local tourists visiting three-, four- and five-star hotels in Morocco. One of the country's priorities is sustainable development. Various conventions, protocols, and initiatives have been developed, showing Morocco's strong commitment to sustainable development. Therefore, the Moroccan government highly and clearly expressed his encouragement and support to hotels opting for a CSR strategy showing the importance of analyzing the customer behavior in relation to those socially responsible initiatives.

Regarding the hotels type, the choice of star hotels is justified because these categories are more involved in socially responsible activities. Thus, the total population in 2019 for the three hotel categories was approximately 2.318.467 local tourists [62]. Furthermore, this year was selected because of the COVID-19 pandemic and the massive lockdown that followed. Therefore, because our population included customers of three-, four-, and five-star hotels, we opted for probabilistic sampling using a stratified random sample. Of the 385 valid responses, 114 (29.6%) were customers of five-star hotels, 151 (39.2%), and 120 (31.2%) were customers of three-stars hotels. Regarding the filter questions, only customers who chose the hotel for their vacation and not for work or other reasons were considered.

Ethical concerns, including consent, confidentiality, and transparency, were respected throughout the research process. Hence, all respondents in the survey were informed about its purpose and objectives and received a cover letter informing them about the research aim and use of the questionnaire.

The questionnaire items were validated in previous studies. All variables were measured using a five-point Likert scale ranging from strongly agree (5) to strongly disagree (1) to ascertain the participants’ level of agreement [2].

Tick-box responses were used to facilitate the participants’ answers and provide an organized view of the questionnaire. The author fine-tuned the questionnaire after the pretest and pilot study. This study was conducted in June 2022.

Structural equation modeling (SEM) is a statistical model applied to examine theoretical models’ validity and is considered a second-generation tool [63]. Two SEM methods were used to analyze the data: covariance-based (CB–SEM) and variance-based (PLS-SEM) [64]. PLS-SEM uses total variance, while CB-SEM uses the covariance matrix to estimate the parameters [65]. In our study, we chose PLS-SEM using SMARTPLS software for various reasons.

• o It has proven to be superior in evaluating complex and multiple mediation analyses [66]. In our study, we used three mediators, two dependent variables, and 40 items.

• o It achieved greater statistical power for all sample sizes [67].

• o This research aims to predict if a relationship exists and if the mediators explain the path. Thus, the recommended method is PLS-SEM when the research goal is exploratory, theory development, or prediction-oriented [67], meaning that the model can detect relationships as they are present in the population.

• o While CB-SEM assumes the normality of data rarely found in social sciences, PLS works well with non-normal distributions [67]. To determine which method was suitable for our research, we tested data normality and found that the data were not normally distributed. Hence, the use of the PLS-SEM is justifiable.

• o Because we have higher- and lower-order constructs in our model, to test the reliability and validity of the analysis following the two-stage approach, the scores of the lower-order constructs are used as observed variables of CSR. Accordingly, Hair et al. [67] recommend using PLS-SEM when latent variable scores are required for subsequent analysis.

4. Results

4.1. Descriptive analysis

Among all participants, 230 (59.7%) had previously stayed in the same hotel, showing behavioral loyalty toward these hotels. The three prominent locations of the hotels last visited are Marrakech (39.5%), Agadir (15.1%), and Tangier (13%). Regarding sex, 196 subjects (50.9%) were male, and 189 (49.1%) were female. The sample was slightly skewed toward young adults, with 125 (32.5%) aged 35–44 and those aged 25–34 (23.1%). The respondents were mainly managers (41.82%), and in terms of educational level, our sample has a relatively high level of education. Indeed, 54.3% of respondents completed a university degree (Table 2).

Table 2.

Main study – sample demographics.

Item	Variable	N	%
Stay in the same hotel (Behavioral loyalty)	Yes	230	59.70%
	No	155	40.30%
	Total	385
Gender	Male	196	50.90%
	Female	189	49.10%
	Total	385
Age	From 18 to 24 years	24	6.20%
	From 25 to 34 years	89	23.10%
	From 35 to 44 years	125	32.50%
	From 45 to 54 years	75	19.50%
	From 55 to 64 years	68	17.70%
	Over 65 years	4	1.00%
	Total	385
Occupation	Manager	161	41.82%
	Worker	87	22.60%
	Entrepreneur/Self-employed	62	16.10%
	Retired/pensioner	29	7.53%
	Student	26	6.75%
	Housewife	20	5.19%
	Total	385
Cities	Marrakech	152	39.50%
	Agadir	58	15.10%
	Tangier	50	13.00%
	Tetouan	20	5.20%
	Fez	16	4.20%
	Casablanca	14	3.60%
	Others	77	20%
	Total	385
Education	University	209	54.29%
	Postgraduate	154	40%
	High school	22	5.71%
	Total	385

4.2. Assessment of measurement model

Before examining our conceptual model, the first step in assessing PLS-SEM findings requires testing the scale's psychometric properties, named the “measurement model.” The purpose of evaluating the measurement model was to ensure that it fits the data correctly. This study includes CSR as a higher-(second-) order construct; therefore, it is necessary to evaluate the reliability and validity of higher (first) lower-order constructs separately. In their critical review of the assessment of second-order constructs, Sarstedt et al. [66] reported that researchers often ignore the assessment of second-order constructs' reliability and convergent and discriminant validity when it comes to validation. The two most common methods used to assess second-order constructs in the literature are the repeated indicators approach and the two-stage approach, which comprises the embedded and disjoint approaches [68]. Nevertheless, Becker et al. [69] found that the two-stage approach better explained the relationship between the main constructs and their dimensions. With this in mind, we adopted a disjointed two-stage approach.

4.3. Lower order constructs

In terms of reliability, Cronbach alpha (α) and composite reliability (CR) were used. Cronbach's alpha is concerned with how closely associated a set of items are as a group, and composite reliability is defined as the “indicator of the shared variance among the observed variables used as an indicator of a latent construct” [70]. The values for α and CR should be equal to or above 0.7 [65]. Table 3 indicates that the reliability indicators are within acceptable ranges.

Table 3.

Lower-order constructs’ reliability and validity.

Constructs	Cronbach alpha	CR
AL	0.944	0.964
BL	0.878	0.926
ENV	0.943	0.954
PHI	0.926	0.945
SOC	0.93	0.945
TRU	0.953	0.964
SAT	0.928	0.945
C–CI	0.929	0.944

For convergent validity, Average Variance Extracted (AVE) and factor loadings were used [64]. The AVE is a measure of the variance explained by a construct regarding its indicators. Factor loadings are concerned with the degree to which the observed variables correlate with a common latent variable. This study confirms that all lower-order constructs converge to explain the variance of their items, as all the factor coefficients are above the agreed range. Fig. 2 shows the convergent validity values of all constructs. The AVE values are shown in the main constructs, whereas the factor loadings are shown as the values going out from the constructs toward their indicators and other variables.

Fig. 2.

Lower order constructs' reliability and validity.

After assessing reliability and convergent validity, the discriminant validity of the lower-order constructs was evaluated. This is related to the extent to which a construct differs from others [65].

The method developed by Fornell and Larcker [70] requires a larger square root for the AVE of any construct than for its association with all other constructs. Table 4 shows that all values in bold and diagonal representing AVE's square root of the AVE are larger than the inter-construct correlation, indicating discriminant validity. The second method, “the heterotrait-monotrait ratio of correlation” (HTMT), differs from the first in its interpretation. This approach, introduced by Henseler et al. [71], shows that when the values are close to one, there is a problem with discriminant validity. Henseler et al. [71] and Hair et al. [67] suggested a value below 0.9. As shown in Table 6, all values were below the recommended score, which confirms that discriminant validity was successfully validated using both methods.

Table 4.

Lower order constructs discriminant validity (Fornell-Larcker criterion).

Constructs	AL	BL	C–CI	ENV	PHI	SAT	SOC	TRU
AL	0.949
BL	0.753	0.898
C–CI	0.559	0.603	0.86
ENV	0.313	0.376	0.458	0.864
PHI	0.283	0.32	0.386	0.521	0.879
SAT	0.643	0.594	0.645	0.407	0.268	0.881
SOC	0.379	0.444	0.479	0.655	0.77	0.388	0.86
TRU	0.623	0.684	0.581	0.357	0.341	0.61	0.438	0.918

Table 6.

Higher order construct's reliability and convergent validity.

Constructs	Items	Item loadings	AVE	Cronbach alpha	CR
CSR	PHI	0.840	0.766	0.847	0.908
	SOC	0.854
	ENV	0.930

4.4. Higher-order constructs

Because we opted for the disjoint two-stage approach, the next procedure was to assess the second-order constructs. The measurement model of our second-order constructs is reflective formative because CSR has three different dimensions. Following this approach, lower-order components are considered items of higher-order constructs. In the first stage, the construct scores (ENV, SOC, and PHI) were saved. In the second phase, we use them to measure CSR. Higher-order construct reliability and convergent validity results are provided in Table 5, where the model's internal consistency through α, AVE, and CR are confirmed. In addition, Table 7, Table 8 show that discriminant validity was established through the Fornell-Larcker and HTMT criteria.

Table 5.

Lower order constructs discriminant validity (HTMT criterion).

Constructs	AL	BL	C–CI	ENV	PHI	SAT	SOC	TRU
AL
BL	0.821
C–CI	0.594	0.665
ENV	0.332	0.412	0.485
PHI	0.301	0.358	0.414	0.557
SAT	0.683	0.654	0.69	0.434	0.289
SOC	0.405	0.492	0.513	0.7	0.828	0.417
TRU	0.653	0.742	0.615	0.377	0.362	0.641	0.465

Table 7.

Higher order construct's discriminant validity (Fornell-Larcker criterion).

Constructs	AL	BL	C–CI	CSR	SAT	TRU
AL	0.949
BL	0.753	0.898
C–CI	0.559	0.603	0.86
CSR	0.375	0.439	0.507	0.875
SAT	0.643	0.594	0.645	0.411	0.881
TRU	0.623	0.684	0.581	0.436	0.61	0.918

Table 8.

Higher order construct's discriminant validity (HTMT criterion).

Constructs	AL	BL	C–CI	CSR	SAT	TRU
AL
BL	0.821
C–CI	0.594	0.665
CSR	0.393	0.478	0.536
SAT	0.683	0.654	0.69	0.438
TRU	0.653	0.742	0.615	0.455	0.641

4.5. Common method bias test

Before evaluating the structural model, collinearity was examined to ensure no bias in the regression findings [65]. This test is known as the Common Method Bias (CMB) and reduces the probability of collinearity-related problems. The VIF (Variance Inflation Factor) is the measure of the amount of multicollinearity of multiple regression variables. When the VIF ratio of a variable is high, the variable is highly collinear with the other variables in the model. Becker et al. [69] and Mason and Perreault [72] accepted values close to five and lower. Preferably, the VIF values should be close to 3.3 and lower [73,74]. In our study, the lowest VIF value was 1, and the highest was 1.998. As no value exceeded the recommended threshold of 3.3, no collinearity issues were reported.

4.6. Structural model assessment

After obtaining satisfactory results for the measurement model, we tested it [65]. For this purpose, we analyzed the coefficient of determination (R2) and Stone–Geisser test (Q2).

The model's goodness of fit is determined by R2 [74], which provides the variance percentage of each endogenous construct supported by all the constructs that explain it. Some researchers followed the argument of Falk and Miller [75], who stated that the value of R2 should be greater than or equal to 0.1. Whereas other researchers interpret their findings according to Chin [76], who classifies the value into three main categories: strong (R2≥0.67), moderate (0.33≤R2<0.67), weak (0.19≤R2<0.33), and unacceptable (R2≤0.19). Table 9 provides the R2 values for all the endogenous constructs. Applying the method of Falk and Miller [75], all of our values were greater than 0.1.

Table 9.

Structural model indicators.

	R2	Q2
AL	0.509	0.453
BL	0.546	0.436
C–CI	0.257	0.188
SAT	0.501	0.383
TRU	0.364	0.304

In contrast, in the method of Chin [76], all R2 values fall under the moderate level, except for customer-company identification, which falls under the weak category but is still acceptable. Hence, our predictive capabilities were established by following both methods. The next step is to test the Stone–Geisser (Q2). This metric is critical in evaluating the predictive validity of complex models [[76], [77], [78]]. Q2 establishes the predictive relevance of the endogenous constructs and is based on the blindfolding procedure, a sample-reuse technique that removes single points from the data and then calculates the model parameters [65]. A Q2 value above 0 indicates that the model has predictive relevance [65]. As a rule of thumb, the Q2 values are either small (0≤ Q2<0.25), medium (0.25≤Q2<0.50), or large (Q2 ≥ 0.50). Table 9 reveals that all Q2 values are positive and above zero, which proves that the model used in our study can accurately predict.

4.7. Hypothesis testing

We test our hypotheses using bootstrapping. This technique is based on a resampling method, whereby large numbers of subsamples of the same size are repeatedly drawn randomly from the original sample and replaced to provide data for an empirical examination of the variability of parameter estimates and indices of fit [79]. The path significance of each hypothesized relationship in our research model, through the bootstrapping method, allowed us to accept or reject a direct path between the two constructs. In our study, the significance level was 0.05. However, to accept p-values, they should be below our significance error level threshold. Table 10 shows the positive and significant relationships among all the hypothesized relationships, except for the impact of CSR on customer satisfaction.

Table 10.

Hypothesis testing.

Relationship	Hypothesis		Standardized β	t-values	p-values	Decision
Independent variable and Mediator	H1	CSR - > TRU	0.19	3.594	.000	Supported
	H2	CSR - > C–CI	0.507	11.559	.000	Supported
	H3	CSR - > SAT	0.047	1.123	.262	Not supported
Mediator and Mediator	H4	C–CI - > TRU	0.484	8.797	.000	Supported
	H5	C–CI - > SAT	0.421	7.242	.000	Supported
	H6	TRU - > SAT	0.345	5.815	.000	Supported
Mediator and Dependent variable	H7.a	TRU - > BL	0.443	6.729	.000	Supported
	H7. b	TRU - > AL	0.325	5.908	.000	Supported
	H8. a	C–CI - > BL	0.234	3.77	.000	Supported
	H8. b	C–CI - > AL	0.143	2.55	.011	Supported
	H9. a	SAT - > BL	0.173	2.465	.014	Supported
	H9. b	SAT - > AL	0.353	5.552	.000	Supported

On the one hand, the total effect of the independent variable, as shown in Table 11, has an impact on both the dependent variable's customer attitudinal and behavioral loyalty with the existence of the three mediating variables (CSR on BL: β = 0.383, t-values = 9.468, p-values = 0.000; CSR on AL: β = 0.359, t-values = 8.894, p-values = 0.000). The total effect is informed by the direct effect between CSR and customer loyalty and the indirect effect through mediators. In contrast, the direct effect represents the impact of CSR on behavioral and attitudinal loyalty without the presence of any mediator. The path becomes non-significant in the absence of mediators in the model (CSR on BL: β = 0.08, t-value = 1.792, p = 0.074; CSR on AL: β = 0.023, t-value = 0.532, p = 0.595). This assumes that mediators are important constructs that tie the path between CSR and customer behavioral and attitudinal loyalty.

Table 11.

The total and direct effect of the dependent variable on the independent variable.

	Total effect			Direct effect
	Coefficient	t-value	p-value	Coefficient	t-value	p-value
CSR - > BL	0.383	9.468	0.000	0.08	1.792	0.074
CSR - > AL	0.359	8.894	0.000	0.023	0.532	0.595

5. Discussion

5.1. Impact of CSR on trust, C–CI, and satisfaction

This study supports the proposed theoretical model, excluding the path linking CSR to customer satisfaction.

A literature review of CSR and customer loyalty reveals that the relationship between both constructs is mainly indirectly supported through mediators. According to the research findings, the positive impact of CSR on customer trust was well-established. Similarly, in previous studies for the hotel industry, authors like Palacios-Florencio et al. [3] and Choi and La [34] have confirmed the positive impact of CSR on customer trust. In the Spanish hotel sector, Martínez and Rodríguez del Bosque [2] found that customers confirm that socially responsible companies contribute to their trustworthiness toward these companies. Similarly, the path linking CSR to C–CI was validated and is consistent with the findings of previous studies [28,37]. This means that once the customer is a member of a hotel highly engaged in CSR practices, positive attitudes and attachments are generated in response to those practices. Our research confirms the findings of [80], who state that customers are generally keen to identify with socially responsible companies to improve their self-esteem and embody a better social image.

However, the final hypothesized path linking CSR to customer satisfaction has not yet been confirmed. Indeed, even though satisfaction is considered a significant predictor of loyalty in all marketing literature, the path was inconclusive, as the relationship between CSR and satisfaction was very weak, whereas the majority of researchers [28,29,81] found a direct relationship between CSR and customer satisfaction, which surprisingly contradicts those of previous studies. Moroccan hotels’ customers do not seem to be directly satisfied with CSR.

However, this result is similar to that of a recent study by Mariño-Romero et al. [82], who stated that this was explained by the fact that their study's sample was composed of perceptions of executives, who may have a restricted idea about their customer satisfaction. However, our explanation is not the same because our sample comprises customers. The discrepancy between our research and previous studies is that, unlike most studies conducted in the developed world, this research proposes a new avenue for a different context regarding customer behavior. However, for being satisfied other factors, such as price, location, and services offered, may interfere with this relationship.

5.2. Relationship between trust, C–CI, and satisfaction

Unlike the traditional model framework, which explains the relationship between CSR and customer loyalty through mediators, this study thoroughly examined the underlying mechanisms of the mediators. The literature generally overlooks the immediate effects of customer-company identification, trust, and satisfaction [83]. Based on the findings, the immediate impact of C–CI on trust and satisfaction was validated. The direct relationship between C–CI and trust established in this study follows Glaveli [11]. Trust is a crucial aspect of long-term social exchange. This has been found to be an antecedent of the identified relationships [35]. Customers are more likely to identify with trustworthy companies that convey the image of caring and honesty. However, customer self-definition and self-esteem are communicated similarly [45]. Identification with a trustworthy hotel is a significant component of creating long-term relationships. Martínez and Rodríguez del Bosque [2], who first studied the direct impact among mediators, theoretically argued for the existence of trust in C–CI. However, their empirical evidence did not follow. This gap between theory and empirical evidence directed our research to consider the possibility that C–CI may not be an outcome of trust but a precursor [11]. Thus, our findings support those of Lewicki and Bunker [46], who claimed that the foundation of trust is in response to the personal perception of the attributes and values of a specific group. For the impact of C–CI on satisfaction, our research is in accordance with previous studies [2,47,84]. Our findings support the statement of He and Li [41] that the higher the identification, the more satisfied the customer is with the company. The last direct impact among the mediators was established regarding the effect of trust on satisfaction. This finding concurs with studies in the hospitality industry [50,51] confirming the relationship between trust and satisfaction. These findings are of interest for further research and analysis. These represent new avenues for explaining the relationship between CSR and customer satisfaction. However, the direct link between C–CI and satisfaction, and between trust and satisfaction, highlights the importance of the mediating role of C–CI and trust in linking CSR to customer satisfaction and, thus, to customer loyalty.

5.3. Relationship between customer loyalty and its antecedents

These research findings show a positive correlation between customer trust and customer loyalty, consistent with various empirical studies [3,11,34,48]. The relationship between trust and customer loyalty is the strongest among the two antecedents. Indeed, the path linking trust to customer behavior and attitudinal loyalty was significant. This concurs with a study by Chaudhuri and Holbrook [48], which confirms the correlation of trust with both behavioral and attitudinal loyalty. Building trust is a prerequisite to long-term relationships. This result is consistent with Reichheld and Schefter [85], who assert that companies must start with a vital component of customer behavior: trust.

The second path in our conceptual framework, linking C–CI to customer loyalty, was also confirmed. While social identity theory is gaining increasing attention in explaining long-term relationships [47], this study took the opportunity to take a closer look at it in relation to customer loyalty. According to the research findings, C–CI influences customer behavioral and attitudinal loyalty. Various studies support this impact [2,11,28,41].

Nevertheless, the last path linking satisfaction to customer loyalty has been confirmed; although, as shown previously, CSR does not positively impact customer satisfaction, it remains an antecedent of loyalty. Based on these research findings, customer satisfaction positively impacts customer loyalty, which is consistent with various studies [28,81].

6. Theoretical implications

First, the impact of CSR on customer loyalty in the context of an emerging world, particularly in African countries, is a confirmed relationship and a real contribution to the literature. Where most of the literature has been dedicated to developed market economies, a gap in the literature has been filled by this contribution [86].

Second, the concept of perceived membership in a particular company that guides customers’ behavior has shown a powerful impact on the business world [87]. Indeed, based on the results, it is clear that customers identify with hotels to fully express their identity. Thus, the social identity theory was tested in the hospitality industry of an emerging country. The findings show that the theory was tested empirically and yielded fruitful results in contributing to the desired outcome.

Third, unlike previous studies [39,40], this research confirms that in Morocco's hospitality industry, CSR does not directly generate customer satisfaction. In other words, CSR customers are more likely to identify with hotels [2,28,41] and develop trustworthiness [88] as a precursor of customer satisfaction. This contradictory result may be because for hotel customers in Morocco, for their satisfaction to be directly linked to CSR, other factors may interfere, such as the price, location, and services offered; therefore, the critical insight of this research brings to the literature a new avenue for research that hotels opting for CSR activities should consider the emphasis on their identity and trustworthiness of their customers as a bridge that connects to customer satisfaction. Considering CSR and customer satisfaction models separately is a new theoretical implication that shows the role of identification and exchange constructs as mediators.

Finally, although there is a considerable difference between the two loyalty measures [89], this study reveals that both dimensions are correlated in response to CSR practices. The literature on customer loyalty asserts that attitudinal loyalty may not systematically lead to attitudinal loyalty and that neither construct is correlated. In other words, a positive attitude toward a product or service does not lead to its repurchase. Studies have revealed that both dimensions are weakly correlated [90] and must be examined independently [2]. Therefore, we found that attitudinal and behavioral loyalty go hand-in-hand in response to CSR activities. Although the literature has shown the inadequacy of assembling both dimensions of customer loyalty, this research contributes to the body of knowledge by proving that both constructs are correlated with socially responsible activities.

7. Practical implications

This study indicates that hotels contribute to building individual identities. Hotels are valued because they reaffirm individual principles and beliefs. This result is critical for managers to achieve improved financial performance. The results shed light on the significant impact of CSR on C–CI [91] and suggest that hotels should invest in corporate identity management. A crystalline identity increases the impact of a company or brand. In addition, it strengthens customers' feelings of belonging, fostering positive attitudinal loyalty through word-of-mouth and positive recommendations that may lead to effective behavioral loyalty through repeated purchases of the service. Therefore, the next step for hotels is investing in CSR practices in the real world, not just as a PR window. The fact that CSR is worth investing in also requires the precise formulation of CSR initiatives. By investing in CSR and making their investments known, companies can gain customer trust as the perceived knowledge of those socially responsible activities positively influences attitude and trust [92] and thus create better financial results [93]. The next vital step is to communicate these practices to the benefit of customers. Customers want to make CSR dimensions visible because they are interested in what others think of a hotel. Therefore, doing well and maintaining a low profile will not achieve the desired outcomes. Similarly, not doing good and communicating about CSR will lead companies to face image risk by being suspected of ‘greenwashing’ if their initiatives remain outdated [94]. Therefore, companies must constantly invest in and communicate these actions.

8. Limitations and future guidelines

This study has certain limitations. The study focused only on the hotel industry of the three selected star rankings in the Moroccan context. Therefore, the research framework may not be entirely applicable to other star hotels. Similarly, there may be differences between dependent and independent hotels. Also, the research was conducted on the online platform “Google Forms” and was based on a voluntary self-administered questionnaire. Although it has the advantages of high efficiency and cost savings, there is a lack of interaction with respondents to clarify misunderstandings during the process.

To generalize the results, future research should examine the model in other contexts, other subsectors of the tourism industry, other industries, and other emerging countries, particularly African countries, as research on the developing world is still restricted to CSR in general and with regard to customer behavior. Future scholars should consider other theories, such as sustainability, stakeholder, and means-end value theories. Accordingly, other mediators, such as brand image, emotions, and commitment, are included in the research framework. This will open new avenues and provide insights into the impact of CSR on customer loyalty. This study adopted Caroll's pyramid to define CSR associations by considering only the desired CSR. Future researchers could use the required CSR (economic and legal) to determine whether there are any differences between them. Other CSR dimensions, such as ethics and stakeholders, could also be used to provide a broader model. There is a tremendous lack of moderators to clarify the link between CSR and customer loyalty [42]. Some researchers have used moderators related to company type, such as commercial versus savings banks. Future researchers should consider moderating variables related to sociodemographic factors (age, educational level, and gender). In addition, because research varies depending on context and culture, other moderators, such as cultural background, generate data for comparison purposes. Moreover, using religion as a moderator may provide new insights into CSR.

Production notes

Author contribution statement

Fatima Ezzahra Jiddi, Ph.D.: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

The data that has been used is confidential.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.