The impact of perceived CSR on corporate reputation and purchase intention

Abstract

Purpose

The purpose of this paper is to examine the influence of consumers’ perceived corporate social responsibility (CSR). The aim is to provide insight into the effect of perceived CSR on purchase intention (short-term effect) and corporate reputation (long-term effect), whilst considering the role of brand image, satisfaction (affective and cognitive) and brand loyalty.

Design/methodology/approach

The sample comprised 429 consumers selected using non-probabilistic sampling with age and gender quotas. Confirmatory factor analysis was used to validate the measurement model. Structural equation modelling was used to validate the research hypotheses.

Findings

All direct and mediated influences in the model were significant, except for the effects of perceived CSR on affective satisfaction. Thus, the proposed causal chain is valuable to understand how perceptions of CSR influence purchase intention and perceived reputation.

Research limitations/implications

Perception is considered a dual phenomenon (cognitive and affective). It would be advisable to consider both dimensions in the future. The same is true of affective satisfaction.

Originality/value

Direct and mediated relationships that have previously been studied separately are considered together in a single model. This approach provides a better understanding of how perceived CSR influences purchase intention and reputation.

Citation

Bianchi, E., Bruno, J.M. and Sarabia-Sanchez, F.J. (2019), "The impact of perceived CSR on corporate reputation and purchase intention", European Journal of Management and Business Economics, Vol. 28 No. 3, pp. 206-221. https://doi.org/10.1108/EJMBE-12-2017-0068

1. Introduction

Corporate social responsibility (CSR) is crucial under the approach of sustainable economics because it encourages firms to actively improve their social, economic and environmental context, thereby creating value for consumers (Green and Peloza, 2011). CSR can be considered both a strategy and a management system for developing competitive advantages (Motilewa and Worlu, 2015), making social and environmental issues a core rather than peripheral part of the business (Porter and Kramer, 2011).

The literature provides evidence of the positive influence of consumers’ perceived CSR on purchase intentions (Aksak et al., 2016; Bigné et al., 2005) and perceived corporate reputation (Hur et al., 2014; Park et al., 2014). However, scholars have tended to develop simple models to study the influence of perceived CSR on important yet highly specific aspects such as brand image (Alvarado and Schlesinger, 2008), affective and cognitive satisfaction (Bigné et al., 2011) and loyalty (de los Salmones et al., 2007). These models have the advantage of rigorously exploring the internal issues associated with the target phenomena, but they ignore the interrelations between these phenomena. Thus, they fail to provide a broad overview that aids our understanding of how perceived CSR influences purchase intention (short-term influence) and perceived reputation (long-term influence).

To mitigate this shortcoming, our aim is to bring together in a single model the aforementioned phenomena as classical relational variables (Alvarado-Herrera et al., 2015). The goal is to improve our understanding of the causal chain via which perceived CSR influences purchase intention and perceived reputation. To achieve this goal, we have divided the remainder of this paper into four sections. The next section presents a literature review of the full range of theoretical constructs that support the proposed conceptual model describing CSR. The following section describes the method, sample and variables used in this study. The penultimate section presents the quantitative findings, reporting the model’s fit and the results of the hypothesis testing. The final section provides the academic and managerial implications of this research.

2. Literature review

2.1 CSR and consumer response variables

According to stakeholder theory, firms must accept certain social responsibilities that transcend purely economic considerations (Bigné et al., 2005), with an emphasis on handling ethical concerns. Almost in parallel with the emergence of this theory, the “sustainability paradigm” echoes this idea, stressing the goal of meeting today’s needs without compromising the development of future generations. CSR refers to the voluntary initiatives that the firm undertakes to champion social and environmental causes and communicates to stakeholders to garner their support (Taghian et al., 2015) through transparent, ethical processes.

These initiatives can contribute to forging relationships with different stakeholders (Lai et al., 2010), including the consumer, a key stakeholder who acts as both recipient and judge of these initiatives when making purchase decisions (Mohr et al., 2001; Aksak et al., 2016). In fact, consumers tend to identify more with firms that undertake CSR initiatives because doing so boosts self-esteem and increases satisfaction (Chung et al., 2015).

Corporate management seeks not only to pursue short-term survival or improvement but also to develop strategies that create enduring competitive advantages. In the case of the former, marketing managers have traditionally used purchase intention as a proxy for future sales (Morwitz et al., 2007) and satisfaction and loyalty as predictors of future business development (Eskildsen and Kristensen, 2008). In the case of long-term competitive advantages, reputation is another indicator that measures accumulated prestige, enabling firms to create customer loyalty whilst reducing stakeholder-related risks (Siano et al., 2010).

Intention, satisfaction, loyalty and reputation all reflect the consumer’s predisposition towards the brand and the company, be it positive or negative. All largely depend on the consumer’s perceptions. Ellen et al. (2006) found that CSR affects purchase intention as a function of the motives that the consumer ascribes to these initiatives. According to the aforementioned study, purchase intention is enhanced by motives that are values driven or strategic and is undermined by selfish motives (i.e. for the benefit of the firm), whilst there is no significant effect if the motive is stakeholder driven.

Perceived reputation refers to an assessment of the general image that consumers have of the firm (Fombrun and van Riel, 2004). It is a representation of the firm’s record and perceptions of the firm’s future. It is a cumulative and collective judgement that combines the appeal or scorn ascribed by the market for either best practices adopted by the firm (e.g. Lego and Google) or harm that consumers feel the firm has inflicted (e.g. Goldman Sachs and Monsanto). Under this approach, reputation is vulnerable to risk derived from specific crises and the social and environmental impact of the firm’s activities.

2.2 Consumers’ perceived CSR

The way in which consumers perceive CSR can affect purchase intentions and corporate reputation. Mohr et al. (2001) showed that the evaluation of products, companies and purchase intentions depends on the amount and nature of the CSR information that is shared. Lee and Shin (2010) found a positive relationship between perceived CSR and purchase intention. Gatti et al. (2012) reported that CSR, together with perceived quality, can provide a competitive advantage by influencing purchase intentions. These authors reported a positive effect of CSR on purchase intention via mediators such as attitude towards corporate experience and values. Likewise, Lee and Lee (2015) suggested that beliefs about CSR positively affect purchase intention through consumer ethics. In other words, perceived CSR can affect purchase intention.

Perceived CSR and perceived reputation are two sides of the same coin. A minimum degree of credibility is necessary to undertake CSR initiatives because consumers can be sceptical towards this type of initiative when undertaken by companies with corporate image issues (Liebl, 2011). Therefore, reputation is an intangible asset that is scarce and difficult to imitate (Rodríguez, 2002) and that is effective at promoting consumer purchase intentions (Pirsch et al., 2007; Aksak et al., 2016) and corporate reputation (Park et al., 2014)[1].

2.3 Perceived CSR as an antecedent to brand image

Brand image is a subjective, perceptual phenomenon that is formed through rational and emotional interpretation (Alwi and Kitchen, 2014). Brand image transcends the technical, functional or physical properties of the product because it is linked to the personality of the subject and is developed through the interactions between the brand and its stakeholders (Bigné and Currás, 2008).

The literature indicates that perceived CSR affects brand image cognitively and affectively (He and Li, 2011) because it provides information on the values of the company (Martínez et al., 2014), although the valence of this influence may vary by stakeholder (Popoli, 2011). Wu and Wang (2014) also affirmed that this relationship exists, although they noted that different dimensions of perceived CSR exert different degrees of influence. Nevertheless, there is a consensus that brand image is strengthened by CSR if the firm knows how to elicit trust, build credibility and develop a solid reputation in the eyes of others (Maldonado et al., 2017).

From a strategic perspective, Werther and Chandler (2005) argued that the perception of CSR initiatives may offer a form of insurance for the firm by aiding its positioning. However, Aldás et al. (2013) argued that positioning occurs whenever CSR is an intrinsic component of the global strategy of the business. The impact of perceived CSR on brand image also varies depending on the consumer’s evaluation processes, and it has positive valence (Martínez et al., 2014; Sen and Bhattacharya, 2001). Based on these arguments, we propose the following hypothesis:

H1.

Perceived CSR has a direct positive effect on brand image.

2.4 Perceived CSR as an antecedent to satisfaction

Customer satisfaction is a key concept in marketing. It can be interpreted as an overall evaluation of a product, service or firm based on an emotional or cognitive response and associated with a specific event that occurs at some specific time (Espejel et al., 2008). Customer satisfaction has been examined extensively in the literature, principally from the perspective of Oliver’s (2014) model of expectation disconfirmation theory. Satisfaction can be understood as the gap between what consumers perceive they have received and their expectations prior to consumption. Under this approach, disconfirmation mediates the relationship between the perceived outcome and satisfaction, although there is also a direct relationship between the actual outcome and satisfaction, which is one of the standard metrics used by marketing managers (Farris et al., 2010).

Satisfaction has a certain duality. It can be conceived as a cognitive–affective state resulting from rational evaluations (e.g. assessment of functionality and compliance with attributes and behaviours) and feelings (Matilla and Wirtz, 2000), which separately or jointly create a state of consumer activation and encourage subsequent behaviour.

Customer satisfaction has been cited as a consequence of CSR (Luo and Bhattacharya, 2006). Numerous authors have reported that perceived CSR directly influences satisfaction (Rivera et al., 2016; Saeidi et al., 2015). Bigné et al. (2011) also showed that CSR influences satisfaction, but conceptualised satisfaction as a two-dimensional construct consisting of affective and cognitive dimensions. We therefore propose the following hypotheses:

H2a.

Perceived CSR has a direct positive effect on cognitive satisfaction.

H2b.

Perceived CSR has a direct positive effect on affective satisfaction.

2.5 The relationship between brand image and satisfaction

A direct positive relationship between brand image and satisfaction has been well established, having been studied for a wide range of sectors. The relationship has also been studied in the context of CSR, and scholars have reported a relationship between CSR and customer satisfaction (Martínez and Rodríguez, 2013). Bigné and Andreu (2004) illustrated the cognitive–affective nature of satisfaction, combining the idea of disconfirmation with affective elements from environmental psychology.

However, we have found few studies in the area of CSR that have examined the relationship between brand image and customer satisfaction. Rivera et al. (2016) recently addressed the role of attitude instead of brand image, reporting the direct influence of CSR on customer satisfaction. Only Wu and Wang (2014) have studied the relationship between the dimensions of CSR (customer, employee, environment, economy and community) and customer satisfaction, considering brand image (functional, symbolic and experiential) and brand attitude as mediating variables. They observed significant relationships between perceived CSR and brand image and between image and satisfaction mediated by brand attitude. We therefore propose the following hypotheses:

H3a.

Brand image has a direct positive effect on cognitive satisfaction.

H3b.

Brand image has a direct positive effect on affective satisfaction.

2.6 Satisfaction as an antecedent to loyalty

Loyalty can be understood as trust in or a desire to preserve a relationship with a brand or firm (Delgado-Ballester and Munuera-Alemán, 2001). It entails rewards for both the firm (by enabling better marketing management and easier strategic planning) and the customer (by offering service and price benefits and reducing risks associated with switching brands) (Magatef and Tomalieh, 2015). The influence of satisfaction on loyalty has been confirmed for a range of sectors. Indeed, satisfaction is considered a necessary element to develop loyalty, although it has also been suggested that satisfaction alone is not sufficient to explain loyalty (Bielen and Demoulin, 2007).

Matute-Vallejo et al. (2011) and Chung et al. (2015) have reported a direct link between perceived CSR and satisfaction and an indirect link (via satisfaction) between CSR and loyalty. Martínez and Rodríguez (2013) studied the relationship between CSR and loyalty in the hotel sector, where loyalty is indirectly influenced by CSR through the mediating roles of trust, consumer identification and satisfaction. The aforementioned authors observed that despite theoretical (Sen and Bhattacharya, 2001) and empirical (Marin et al., 2009) evidence of the relationship between loyalty and CSR, our understanding of the ways in which this relationship arises is limited. However, Martínez et al. (2014) found a direct positive relationship between perceived CSR and loyalty.

Pérez and Rodríguez (2015) reported an indirect effect of perceived CSR on loyalty mediated by identification, emotions and satisfaction. They identified two paths through which CSR creates loyalty and satisfaction. The first is through the beliefs and emotions elicited by the firm at the institutional level, whereas the second is through the emotions, feelings and attitudes generated by the product. We therefore propose the following hypotheses:

H4a.

Cognitive satisfaction has a direct positive effect on brand loyalty.

H4b.

Affective satisfaction has a direct positive effect on brand loyalty.

2.7 Loyalty as an antecedent to purchase intention and corporate reputation

Purchase intention is a key indicator for firms. It helps them to predict the likelihood that a consumer makes a purchase within a given period and acts as a proxy for actual behaviour (Farris et al., 2010). The literature defines purchase intention from two angles: as a preference to re-purchase a given product and as the probability that the consumer chooses a particular product. The first case refers to the post-purchase behaviour of existing customers, whereas the second case refers to an indicator of general consumer preferences (whether or not the consumer is an existing customer) in relation to whether the consumer perceives that the product or brand meets her or his expectations, needs and desires.

This preference may be modulated by a wide range of factors such as experience, cognitions, emotions and so forth. One of these factors is the actual behaviour of the firm. The literature provides solid evidence that CSR behaviour is an important factor behind consumers’ purchase decision processes (Creyer and Ross, 1997). Scholars have also reported that perceived CSR strengthens consumer loyalty, which increases the likelihood of choosing the firm’s products (Chen et al., 2015).

The literature shows that loyalty plays an important role in purchase intention (Hong and Cho, 2011). We therefore propose the following hypothesis:

H5a.

Brand loyalty has a direct positive effect on purchase intention.

It is widely accepted that corporate reputation is an antecedent to brand loyalty (Walsh et al., 2009), playing an important role in the management of the firm’s reputation. Thus, a good reputation potentially boosts loyalty. However, CSR is not an instrument to handle crises or an insurance policy against crises. It is a relational strategic approach that the firm adopts with respect to society and its environment. Because reputation can be perceived as a cumulative and collective judgement, it can also be based on individual’s loyalty, which may or may not strengthen the firm’s reputation. We can therefore establish the following hypothesis, which posits the influence of loyalty on the consumer’s opinion or idea of the firm:

H5b.

Brand loyalty has a direct positive effect on corporate reputation.

3. Method

3.1 Participants

All individuals contacted for this study were at least 18 years old. They were selected using non-probabilistic sampling based on gender and age quotas. The sample comprised 456 individuals. The elimination of outliers reduced the sample to 429 valid cases. In terms of gender, 56.6 per cent of the individuals in the sample were women. Regarding age, 25.7 per cent were aged under 30 years, 31.3 per cent were aged 31–45 years, 24.3 per cent were aged 46–65 years and 18.7 per cent were aged over 65 years. Finally, we established four groups by level of education: incomplete secondary studies (7.7 per cent), complete secondary studies (13.3 per cent), incomplete university studies (36.6 per cent) and complete university studies (42.4 per cent).

The sample had a slightly higher proportion of women than men (the population of Argentina consists of 51 per cent women). There was a high presence of individuals aged 31–45 years (the mean age in Argentina is 34.4 years). There was a bias towards people with higher education, so the sample did not reflect the mean for Argentina (9 per cent according to the National Institute of Statistics and Censuses, Instituto Nacional de Estadística y Censos, 2018). We accepted this bias because the questionnaire contained certain complex items that required a minimum level of education to respond.

3.2 Measures

Perceived CSR

We considered the five basic dimensions of CSR: economic, social, philanthropic, environmental and ethical/legal (Dahlsrud, 2008). We selected items related to each of these dimensions. We focused particularly on the social dimension because it is considered the most difficult dimension to measure (Beske-Janssen et al., 2015).

Brand image

Because of the sheer range of constructs covered by brand image, the literature presents both unidimensional and multidimensional scales. We used Villarejo-Ramos’s (2002) scale because it is unidimensional and does not refer to any specific product.

Satisfaction (cognitive and affective)

We adopted Oliver’s (2014) definition, distinguishing between cognitive and affective satisfaction. We borrowed the measure from Bigné et al. (2011). This measure has three cognitive items and five affective items. We removed the item “I enjoy staying in […]” because it was unrelated to our study. The aforementioned authors reported high reliability (0.82–0.93).

Brand loyalty

We used the scaled developed by Wong (2004) because of its high reliability (0.90–0.92). This scale was designed to measure loyalty towards a shop. However, we adapted it for use in relation to the respondent’s selected brand using four items related to saying positive things about the brand, recommending the brand, encouraging others to buy the brand’s products and considering the brand as a first choice.

Purchase intention

We conceptualised purchase intention as an initial preference that leads to a future desire to purchase a given product. We used Putrevu and Lord’s (1994) three-item scale. These items relate to intentions regarding the next purchase, purchasing in the long term and the surety of making a purchase. This scale has been used by Bigné and Currás (2008) and Chen et al. (2015) with high reliability (0.81–0.91).

Perceived reputation (REP)

We adopted the definition used by Fombrun and van Riel (2004). Our three-item measure was based on research by Ahearne et al. (2005) (for the item “company X is well respected”) and Schwaiger (2004) (for the items “[…] is one of the best in the sector” and “[…] is well established”). These last two items were used by Riley et al. (2014) for the reputation dimension of the brand image construct. There are no available data on the reliability of this scale.

All items were based on seven-point Likert-type scales (1=completely disagree; 7=completely agree). Table I lists the scales, their dimensions, the items and the sources.

3.3 Study design

To test the model, we conducted a quantitative study. The first phase consisted of an exploratory study of the most prominent consumer categories in Argentina with firms that were recognisable for consumers. The categories were food and drink, electrical appliances, cosmetics, clothing, school items, children’s toys, tourism and technology. The second phase consisted of developing the scales to measure the variables in the model. During this phase, a structured pilot questionnaire was developed and administered through a personal survey at respondents’ homes.

Because some variables were constructed using items from different sources and because we had some concerns over respondents’ comprehension, the third phase consisted of a pre-test. The pre-test was carried out with undergraduate university students, whose collaboration helped to create the items and ensure comprehension (Hinkin, 1995). The aim of the pre-test was to check the structural features of the questionnaire, formulation of the items, understanding of questionnaire content and ease of response. We also confirmed that the questionnaire could be adapted to the eight product categories mentioned earlier. Following the pre-test, we eliminated two items for perceived CSR (pCSR1 and pCSR9). Respondents indicated that they were unable to respond to these items because they felt that these items depended more on knowledge than on perceptions. We changed the way certain items were phrased to enhance respondents’ understanding. We also removed two brand image items (IMA1 and IMA6) because they could not be understood and one loyalty item (LEAL4) because “priority” could be interpreted in several ways. Finally, we decided that the respondents should focus their responses on the brand of the last purchase they made after identifying what they recognised as the most socially responsible brand in the chosen product category.

The last phase consisted of designing the procedure for the pen-and-paper and electronic data collection. Data were collected in September and October 2015 in Córdoba (Argentina). Córdoba, Buenos Aires and Rosario are home to 70 per cent of the population of Argentina. This status makes Córdoba representative of Argentina.

4. Results

We performed the analyses in two stages. First, we conducted confirmatory factor analysis to validate the measurement model (reliability and validity). Next, we used structural equation modelling to test our hypotheses.

4.1 Validation of the measurement model

We performed confirmatory factor analysis of covariance structures using the robust maximum likelihood method because multivariate normality did not hold (Mardia test=231.11). Using all items (except those eliminated following the pre-test) yielded good values for the fit indicators (BBNFI=0.90; IFI=0.94, CFI=0.94), with an RMSEA value of 0.05. The value for the robust fit statistic χ Satorra – Bentler 2 of 573.82 (df = 278, p = 0.00) was significant because of the effect of the large sample size. After we eliminated this effect, the value for χ²–normed of 2.06 fell within the acceptable range recommended in the literature (Schreiber et al., 2006).

However, we observed problems with perceived CSR because the factor loadings for pCSR6 and pCSR8 were low and their respective R² values were less than 0.4. Given the reflective nature of these items, we deleted both from the perceived CSR scale and recalculated the goodness of fit for the measurement model. The new configuration had a better fit (BBNFI=0.91, IFI=0.96, CFI=0.96), with an RMSEA value of 0.05. The value for the robust fit statistic χ Satorra – Bentler 2 of 445.50 (df=231, p=0.00) was still significant, and the value for the χ²–normed of 1.92 was lower than the previous value and was acceptable.

We measured reliability using Cronbach’s α and the composite reliability (CR). Values for these measures were greater than 0.8 for all variables, thereby meeting the requirement of min(α, CR) = 0.7. We assessed convergent validity using two criteria: factor loadings of all items should be greater than 0.5 and significant, and the average variance extracted (AVE) should be greater than 0.5. In both cases, the minimum requirements for reliability and convergent validity were met (see Table II).

Discriminant validity was also checked using two criteria: confidence intervals and the test of AVE. The confidence intervals for the correlations between variables should not contain the value 1. No confidence interval for each pair of variables contained a correlation equal to 1. For the second criterion, we compared the AVE for each factor with the squared correlation between each pair of factors. Discriminant validity holds if the AVE values for two factors are both greater than the squared value of the correlation between those factors. This condition held for all values of AVE and all correlations.

4.2 Structural model results

We performed regression analysis of the structural model. Thus, we accept H1 because perceived CSR influenced brand image (β=0.31; p<0.01). Perceived CSR also influenced satisfaction, but only for the cognitive dimension (H2a: β=0.11; p<0.05) and not for the affective dimension (H2b: β=0.06; p>0.05). We accept H3a and H3b because the influence of brand image on satisfaction (cognitive and affective) was significant, although the influence on the affective component (β=0.70; p<0.01) was greater than the influence on the cognitive component (β=0.53; p<0.01).

The two dimensions of satisfaction exerted a direct positive influence on brand loyalty. The influence of cognitive satisfaction (H4a: β=0.39; p<0.01) was weaker than the influence of affective satisfaction (H4b: β=0.53; p<0.01). It was therefore possible to deduce that the effect of brand image on loyalty was strong (β_{Total effect}=0.60, p<0.01), indicating that brand image is closely linked to loyalty through cognitive and affective satisfaction. The indirect effect of brand image on loyalty through cognitive satisfaction was 0.23 (brand image–cognitive satisfaction=0.59 × cognitive satisfaction–loyalty=0.39) and the effect through affective satisfaction was 0.37 (brand image–affective satisfaction=0.70 × affective satisfaction–loyalty=0.53). Thus, the mediating effect of affective satisfaction was greater than the mediating effect of cognitive satisfaction on the relationship between brand image and loyalty. Finally, the influence of loyalty on purchase intention (H5a: β=0.71; p<0.01) and the influence of loyalty on reputation (H5b: β=0.62; p<0.01) were confirmed. Table III provides a summary of these results.

Because the model showed the mediation of the endogenous variables, it was advisable to check whether the effects of mediation were significant. Table IV shows that all were highly significant (all t-values>4.00, p<0.001). Notably, all indirect effects of perceived CSR on the other dependent variables in the model were significant. The same occurred with the other endogenous mediating variables.

5. Conclusions

We observed that the proposed relationships were consistent with those described in the literature in terms of both direct relationships and indirect effects. We also observed high coefficients of determination, which implies that the model is capable of explaining purchase intention and corporate reputation in the context of perceived CSR. This conclusion supports the idea that the proposed causal chain captures important relationships and indicates that perceived CSR has a significant indirect effect on purchase intention and perceived reputation.

However, we did not observe the expected results regarding the influence of perceived CSR on the affective component of satisfaction. Our findings differed in this regard from those reported by Bigné et al. (2011). More specifically, we did not observe a direct effect, although we did observe an indirect effect when brand image mediated the relationship. This finding has two possible explanations. First, the measure that we used to capture consumers’ perceptions focused on the rational aspects of CSR. A review of the content of the items showed that no item was related to affect, feelings or emotions. Thus, not accepting H2b is reasonable. For example, it is possible for two different consumers (one in favour of social causes and another against them) to give the same scores to the item “brand X helps social organisations in the community” because the firm objectively helps in this way. However, if this item had been worded to capture emotional aspects (e.g. “I feel that brand X helps as companies should through CSR”), the scores given by each respondent might have differed. The second explanation is that the perception of cognitive aspects of CSR contributes to satisfaction through brand image because brand image is based on what consumers think and feel about the company and brand (Keller, 1993).

Despite being measured in terms of rational considerations, perceived CSR had a significant indirect effect on reputation, which is built over time, and purchase intention, which is a short-term response. This finding reflects the importance that firms ensure that consumers are aware of CSR initiatives. This finding is consistent with those reported by Andreu et al. (2015), who showed that rational messages help raise consumers’ awareness of CSR.

Our findings have three direct implications for firms. First, if consumers do not perceive CSR initiatives, firms will struggle to view CSR as a strategic investment, as already affirmed by Andreu et al. (2015). Second, considering that consumers have little awareness of CSR initiatives, companies should develop communication strategies that promote these initiatives as being positive for society and the environment as well as for the firm. Finally, perceived CSR is an important antecedent to promote brand image and customer satisfaction and loyalty. CSR has the ability to have a short-term impact, by encouraging purchase intentions, and a long-term effect, by helping to develop the firm’s reputational capital, which can provide a competitive advantage.

Like all research works, this study has limitations related to the sample. More specifically, the study used a non-random sample of consumers from one city in Argentina. The study also has contextual limitations. The study was performed in a moment of political turmoil in a politically and economically polarised country. It was therefore difficult to elicit consumer sentiment regarding consumers’ perceptions of firms’ social, economic and environmental responsibility given the social and economic situation facing these consumers. Furthermore, the essentially cognitive measure of perceived CSR meant that the influence on affect was not captured. The fact that perception can be considered a dual construct (cognitive and affective) reflects the interest in studying both dimensions. Because this study focused on the measurement of the cognitive dimension, we advocate considering the affective component when measuring perceived CSR.

Table I

Measurement scales

Items	Dimension	Research
Perceived CSR: the brand/company…
pCSR1: …treats employees very wella	Economic	Bigné et al. (2005), Bigné and Currás (2008), Dean Lichtenstein et al. (2004), Maignan and Ferrell (2001), Menon and Kahn (2003), Van Herpen et al. (2003)
pCSR2: …is socially responsible	Social
pCSR3: …helps civil society organisations in the community	Philanthropic
pCSR4: …is committed to ecological issues	Environmental
pCSR5: …returns some of what it has received to society	Social
pCSR6: …act thinking about society	Social
pCSR7: …integrates philanthropic contributions in their business activities	Philanthropic
pCSR8: …behaves honestly with their customers	Ethics
pCSR9: …respect the legal regulationsa	Legal
Brand image: the brand/company X…
IMA1: …has a strong personalitya		Villarejo-Ramos (2002)
IMA2: …has a powerful image
IMA3: …is very good
IMA4: …is a very nice brand/company
IMA5: …is a very attractive brand/company
IMA6: …is a hugely sympathetic brand/companya
Satisfaction
CS1: buying the brand/company was an intelligent decision for me	Cognitive	Bigné et al. (2011)
CS2: I think I did the right thing by buying this brand/company
CS3: the brand offers exactly what I needed/expected from it
AS1: the brand/company has made a positive impression on me	Affective
AS2: I like to buy that brand/products of that company
AS3: buy that (brand/ use the products) of that company I find it great.
Loyalty
Leal1: I say positive things about brand/company to other people		Wong (2004)
Leal2: I recommend to others this brand/company
Leal3: I encourage friends and relatives to buy that brand/company
Leal4: I consider that brand/company my first choicea
Purchase intention
INT1: I will purchase brand/company the next time I need a product		Putrevu and Lord (1994), Bigné and Currás (2008), Chen et al. (2015)
INT2: it is very likely that I will buy that brand/company
INT3: definitely, I will buy some product of that brand/company
Perceived reputation: the brand/company…
REP1: …is one of the best in the sector		Ahearne et al. (2005), Schwaiger (2004), Riley et al. (2014)
REP2: …is well established
REP3: …is well respected

Note: ^aItems eliminated in the pre-test phase

Table II

Reliability and convergent validity

Construct	Item	M(SD)	Li	R2	α	CR	AVE
Perceived CSR (pCSR)	pCSR2	4.96 (1.26)	0.65	0.42	0.84	0.85	0.53
	pCSR3	4.12 (1.24)	0.74	0.55
	pCSR4	4.39 (1.50)	0.80	0.65
	pCSR5	4.13 (1.39)	0.76	0.58
	pCSR7	4.06 (1.26)	0.65	0.43
Brand image (IMA)	IMA2	5.91 (1.24)	0.74	0.54	0.87	0.87	0.63
	IMA3	6.12 (1.06)	0.89	0.78
	IMA4	5.92 (1.13)	0.77	0.59
	IMA5	5.86 (1.23)	0.78	0.60
Cognitive satisfaction (CS)	CS1	5.81 (1.11)	0.78	0.61	0.83	0.87	0.65
	CS2	5.93 (1.02)	0.89	0.80
	CS3	5.92 (1.06)	0.74	0.54
Affective satisfaction (AS)	AS1	6.02 (1.03)	0.86	0.74	0.89	0.89	0.73
	AS2	5.98 (1.07)	0.87	0.76
	AS3	5.61 (1.21)	0.83	0.69
Loyalty (LEAL)	LEAL1	5.69 (1.20)	0.83	0.69	0.87	0.88	0.71
	LEAL2	5.59 (1.26)	0.88	0.78
	LEAL3	5.22 (1.49)	0.81	0.66
Purchase intention (INT)	INT1	5.49 (1.32)	0.90	0.81	0.91	0.91	0.76
	INT2	5.72 (1.24)	0.89	0.80
	INT3	5.47 (1.43)	0.83	0.69
Perceived reputation (REP)	REP1	5.48 (1.34)	0.80	0.68	0.87	0.88	0.71
	REP2	5.83 (1.18)	0.88	0.77
	REP3	5.60 (1.24)	0.85	0.72

Notes: M, mean; Li, factor loading; α, Cronbach’s α; CR, composite reliability; AVE, average variance extracted

Table III

Structural equation model results

Efectos	β	Hypothesis	R2
H1: pCSR → IMA	0.309**	H1 (accepted)	0.095
H2a: pCSR → CS	0.105*	H2a (accepted)	0.400
H3a: IMA → CS	0.594**	H3a (accepted)
H2b: pCSR → AS	0.034ns	H2b (not accepted)	0.521
H3b: IMA → AS	0.711**	H3b (accepted)
H4a: CS → LEAL	0.396**	H4a (accepted)	0.622
H4b: AS → LEAL	0.522**	H4b (accepted)
H5a: LEAL → INT	0.708**	H5a (accepted)	0.500
H5b: LEAL → REP	0.622**	H5b (accepted)	0.390

Notes: Standardized estimates. *p <0.05; **p<0.01

Table IV

Analysis of effects of mediation (standardized solution)

Indirect effect	β (t-value)
Through brand image
pCSR → Cognitive satisfaction (CS)	0.17 (4.40)
pCSR → Satisfaction affective (AS)	0.20 (4.41)
Through satisfaction (CS and AS)
pCSR → Loyalty	0.25 (5.10)
Brand image → Loyalty	0.68 (6.79)
Through loyalty
CS → Purchase intention	0.36 (7.53)
CS → Perceived reputation	0.29 (7.53)
AS → Purchase intention	0.50 (9.23)
AS → Perceived reputation	0.40 (7.79)
pCSR to dependent construct
pCSR → Purchase intention	0.21 (4.44)
pCSR → Perceived reputation	0.17 (4.77)

Note: All t-value> 4.00 (p<0.001)

Note

1. A summary of contributions from 2004 to 2016 is available from the corresponding author upon request.