Corporate Social Responsibility Drivers and Barriers According to Managers’ Perception; Evidence from Spanish Firms

Abstract

Corporate Social Responsibility (CSR) is becoming a dominant issue in both research and practice of management. However, the underlying processes in the relationship between the degree of development of CSR in companies and the drivers/barriers that determine this development are still at the center of an intense debate. The purpose of this empirical study is to examine these relationships. We investigate a sample of 416 Spanish firms; based on a multifactorial framework, our study considers both the subjective and objective drivers/barriers, analyzing their joint effect on the final degree of sustainability. A structural equation model is established and a Bayesian approach is used, enabling exact inferences about the model’s parameters and handling missing data with random imputation, thus increasing the study’s reliability. The results show that this degree is related to what managers believe CSR to be (subjective drivers/barriers) and what managers expect it to accomplish or outcomes (objective drivers/barriers).

1. Introduction

As a form of self-regulation, Corporate Social Responsibility (CSR) refers to voluntary business, social and environmental actions to correctly interact with the natural environment and stakeholders [1]. Ethics, environmental care, good corporate citizenship, social cohesion, transparency and stakeholder satisfaction are some of the qualities commonly associated and sometimes equated with CSR. In recent years, businesses around the world have been promoting economic, social and environmental initiatives trying to incorporate these qualities as they realize the importance of CSR in facing the global competition [2] This phenomenon has aroused the interest of academics, who have devoted considerable effort to defining the CSR concept, to delimiting its multiple dimensions, and to proposing operational ways to overcome difficulties inherent in its measurement and quantification. The measurement and quantification of CSR in companies is probably one of the aspects in which the scientific community can still advance. This interest in quantifying CSR is also growing in corporations [3]. Another area where academics can still make relevant contributions is related to the factors that determine the level of CSR achieved in the company. The study of these factors, drivers or barriers in the escalation towards advanced levels of CSR requires an adequate quantification of development of CSR in the company. This difficulty has hindered the empirical research in this topic.

Despite the long history of CSR, no consensus has been reached among the industry participants, academics or other interested parties on its precise definition, while they also recognize the complexity of the drivers and barriers that affect CSR [4]. One reason for this might be that a very diverse and multidimensional set of issues, ranging from climate change and environmental conservation to human rights and labor practices, is discussed under the category of CSR [5]. It is in this context, that a debate has arisen about the diversity of CSR drivers/barriers. Previous literature considers both the drivers of and the barriers to CSR [6]. In addition, previous literature distinguishes two types of factors that affect CSR: the moral driven (holistic interpretation) and the profit driven (instrumental interpretation) [6,7]. In this paper, we consider both perspectives to propose a classification into four categories of factors that determine the level of CSR in the company. We use the subjective/objective qualifiers to distinguish between drivers or barriers that may be conditioned by the moral beliefs of the manager and the drivers or barriers whose recognition is not conditioned by such beliefs.

On the one hand, from a subjective perspective, CSR is an umbrella term that summarizes how business-society relationships should be developed [8]. From this perspective, CSR responds to a manager’s subjective perception of what this relationship should be like, considering ethical, moral or philanthropic issues. On the other hand, from an objective perspective, CSR has an instrumental use. From this perspective, CSR gets outcomes and represents a mechanism with which to respond to the institutional pressures perceived by managers, to secure long-term benefits, and to obtain competitive advantages which result in supernormal profit [9,10], among other aspects.

Although there is abundant theoretical and empirical literature analyzing the factors that determine CSR, scholars agree that a better understanding of effective drivers/barriers is required [11]. Our objective is to contribute to this line of research by evaluating the impact of subjective and objective drivers/barriers perceived by managers on the achievement of an advanced level of development of CSR in their companies. We specifically investigate CSR drivers/barriers through two research questions. Firstly, from a subjective perspective, we try to answer this question: How does managers’ subjective interpretation of what CSR means affect the degree of CSR development achieved in firms? Secondly, from an outcome perspective, we try to answer the following question: How does managers’ perception of the benefits (objective drivers) and costs (objective barriers) of CSR affect the degree of CSR development achieved in firms?

We believe that this information should be integrated into a broad and well-articulated theoretical framework that, based on empirical data from managers’ perceptions, enables an analysis of the relationships between CSR initiatives and their effective drivers/barriers. As Tarabella and Burchi [11] claim, the use of perceptual data is appropriate for CSR assessment because managers’ opinions influence both the implementation and results of CSR strategies.

There are two distinctive elements of this research with respect to previous literature. First, we offer an analysis of the factors that determine the level of CSR with greater detail than that generally presented in the previous literature, revealing the role of the interpretation of managers by distinguishing between four types of factors: subjective drivers, subjective barriers, objective drivers and objective barriers. Secondly, we offer a careful and comprehensive measurement of CSR that thoroughly addresses the multiple dimensions of the concept.

To measure the development of CSR practices in firms in this paper we use an indicator of CSR, with different multidimensional aspects of CSR and their relative importance, or specific weight. To our knowledge, the integration of multiple CSR indicators into a single variable with a suitable procedure has not been previously approached in the literature. The effort devoted to this aspect is an essential contribution of this paper, which proposes a flexible methodology based on the use of structural equation models as useful tools for measuring constructs and analyzing the existence of interrelationships between different items and constructs. There is little empirical research on CSR drivers/barriers in Spanish companies, where studies have mainly focused on a single dimension of CSR; indeed, the drivers/barriers of CSR in Continental Europe are still relatively unknown and most of the literature is based on Anglo-American countries (US and UK) [12]. This study aims to fill some of these gaps by developing a tool that enables us to explain the socially responsible behavior of companies, considering CSR in relation to its drivers/barriers. This is achieved by analyzing the subjective (what managers believe CSR to be) and objective factors (what managers expect it to accomplish or outcomes) of CSR and studying the relationship between the variables involved, using a rich body of empirical evidence gathered through a survey conducted on a sample of 416 Spanish firms. The results confirm that a firm’s CSR depends on both the subjective drivers/barriers of CSR perceived by managers and their expectations about the objective drivers/barriers or outcomes.

The paper is organized as follows. First, in the next section, we briefly review the literature about the subjective and objective drivers of and barriers to CSR as perceived by managers. We then present the data, methods and results of our empirical research conducted on a sample of Spanish firms. Finally, we present a discussion and our main conclusions about how to reconfigure institutional structures to promote greater commitment to sustainability.

2. Theoretical Background, Materials and Methods

2.1. Theoretical Background and Theories about Drivers/Barriers

The various types of factors influencing the implementation of CSR measures can be understood by reviewing the four types of responsibilities considered in the Carroll’s CSR pyramid: economic responsibility (“be profitable”), legal responsibility (“obey the law”), ethical responsibility (“be ethical”) and philanthropic responsibility (“be a good corporate citizen”) [13]. In line with these responsibilities, Arevalo and Aravind [14] or Govindasamy and Suresh [6] consider four models of CSR according to the types of factors that determine them. The ethical model, determined by voluntary commitment to public welfare; the statist model, determined by legal requirements; the stakeholder model, determined by stakeholder pressure; and the liberal model, determined by economic drivers and barriers.

The literature highlights all these types of determinants. According to this research, some of the drivers that favor CSR development are related to ethical theories. Managers’ morality, charity, honesty and moral beliefs are some of the drivers that justify this interpretation of CSR. From this perspective, firms are under the obligation to be socially responsible because it is morally correct, even if it comes at a cost [15]. The managers of firms who see CSR in this way would be acting according to principles based on the achievement of a better society, the promotion of universal rights, sustainable development, or the “common good” approach [16]. These managers consider socioemotional wealth as a prosocial and positive stimulus to CSR implementation [8].

Likewise, the growing literature linking CSR and institutional theory can help us to explain institutional drivers to engage in CSR [17]. We follow Matten and Moon [8], who define “institutions” as “…not only the formal organisation of government and corporations but also norms, incentives, and rules…” and “…institutional theory brings interdependencies between and interactions among stakeholders into the analysis, which is vital to understanding CSR, given its societal orientation”. Matten and Moon [8] suggest that national institutional frameworks consisting of specific combinations of political, financial, educational, and cultural systems affect CSR behaviors. From this perspective, the legal requirements and the demands of the various stakeholder groups and society acquire importance as drivers for CSR.

In addition, in the framework of institutional theory, Campbell [18] argues that economic conditions affect the degree to which corporations act in socially responsible ways but that this relationship is mediated by a variety of institutional factors that constrain and enable these behaviors. Furthermore, it seems that corporate social responsibility differs across sectors [19]. One possible explanation is the presence of contextual factors, as indicated by Zhao et al. [20].

Other factors that favor the development of CSR are based on instrumental stakeholder theory [21]. In this group of theories, CSR is seen as merely a strategic tool for the achievement of economic objectives through social activities, strategies for competitive advantages and cause-related marketing [16]. In this respect, different authors mention the possibility of obtaining long-term benefits through CSR [22] or cost savings [23]. Specifically, CSR could be used to improve the relationships with stakeholders [18,24,25], improving company reputation [6,26], obtaining greater community solidarity [27], preventing incident-induced reputation risks [28], promoting customer loyalty [29], etc. On the other hand, based on instrumental stakeholder theory, most of the barriers that hinder the development of CSR in firms are related to property rights and the maximization of shareholder value [14,21]. This approach, adopted by some academics, together with the narrow conceptual interpretation of CSR by some firm managers, corresponds to Friedman’s classic capitalism [30], which posits that the primary function of any corporation is to maximize shareholder value.

Some academics and managers also consider the availability of the resources (financial, human, and time-related resources) associated with CSR development [31]. In this way, the drivers of and barriers to company CSR development would be explained based on resource theory and, as O’Connor and Spangenberg [32] point out, firms can only devote limited resources to CSR. Larger firms have more resources for the development of CSR, so firm size is a key factor and a predictor of CSR [33]. Some research projects such as Gamerschlag et al. [34] show size-related CSR behavior differences. From this resource-based perspective, CSR is seen as providing internal benefits by helping a firm to develop new resources and capabilities related to know-how and lowering costs [35] or improving product quality and product innovation [36].

2.1.1. Two Perspectives of CSR Drivers/Barriers

Based on our analyses of the academic literature on CSR drivers/barriers, we identified two types of drivers/barriers: subjective and objective. With respect to the former, it is known that managers’ conceptions of CSR can play a key role in its development [37] and the trouble in determining the subjective drivers/barriers comes from the difficulty of defining the concept, as it is “essentially contested”, “appraisive” (or considered as valued) and “internally complex” [38]; it could even be associated with the personal values of individual managers [39]. Secondly, the objective drivers/barriers of CSR come from an outcome-oriented position (profit or costs). These outcomes are commonly associated with the organization’s strategic interest [27] and the difficulty in determining the objective drivers/barriers comes from the fact that CSR is a dynamic phenomenon [40] with an intertwining of multiple considerations and relatively open rules of application [38].

In this study, we will extend this idea by proposing that both subjective and objective drivers/barriers are interconnected with the degree of CSR. Firms not only perceive barriers that generate net costs associated with the implementation of CSR, but also opportunities that generate net benefits, but all this is conditioned by the managers’ subjective interpretation of CSR. In the next section, we will propose two (extreme) perspectives on how CSR and its development depends on how the subjective drivers/barriers are perceived by managers and the disequilibrium between the two types of objective drivers/barriers in each organization.

2.1.2. Drivers of and Barriers to CSR Strategies: Content Analysis

To identify relevant studies on CSR drivers/barriers, we selected a set of primary keywords (CSR, sustainability, drivers/barriers, outcomes, concept and perception) for a computer search of the Science, Scopus and ScienceDirect databases. Only papers published in peer-review journals from 2001 on were included in the search. Subsequently, from this comprehensive but not exhaustive search, the authors manually scanned abstracts in order to remove papers not concerning the study’s objective. A complete review was subsequently made of the remaining papers in order to detect drivers/barriers that influence the development of CSR and after this last review, only 106 papers were considered. The main groups of CSR drivers/barriers were then classified by origin and grouped under general terms in the authors’ opinion. Finally, the authors classified the CSR drivers/barriers in the reviewed papers into two groups. The first includes CSR drivers/barriers based on considerations related to business culture or managerial values. This type of driver is found in 54 papers and will now be referred to as “subjective drivers/barriers”, which can be either “favorable/available” or “drivers”, and “unfavorable/not available” or “barriers”, similar to “wide” or “narrow” following the terminology used by Quazi and O’Brien [41]. An overview of the concept is shown in

Table 1. Overview of concept of responsibility based on literature review.

	Frequency (%)
Integrating ethics a	31.9%
Sustainable development b	18.7%
Organizational commitment to transparency c	13.2%
Philanthropy d	11.0%
Public relations exercise e	8.8%
Fashion following f	8.8%
A skeptical view only for large corporations g	4.4%
A utopic value h	3.3%
Total	100%

Sources: ^a Baron [64], Cacioppe et al. [49], Carter and Rogers [58], Ciliberti et al. [45], Cramer et al. [65], Deniz Deniz and Cabrera Suarez [47], Erwin [42], Graafland et al. [66], Harjoto and Jo [56], Hartmann [67], Hermingway [39], Ibrahim et al. [68], Idowu and Papasolomou [69], Lattemann et al. [10], López Gamero et al. [31], Lozano [70], Maignan and Ralston [24], Manner [71], Margolis et al. [28], Matten and Moon [8], Muller and Kolk [33], Pater and Van Lierop [72], Perrini and Minoja [73], Peterson and Jun [74], Sajjad, et al. [46], Salam [75], Tengblad and Ohlsson [76], Waldman et al. [77], Wu et al. [43]; ^b Bansal and Hunter [78], Benabou and Tirole [54], Cacioppe et al. [49], Carter and Rogers [58], Clement-Jones [52], Cheah et al. [79], Chan and Cheung [51], Cramer et al. [80], Dobers and Springett [48], Godkin [53], Hemingway [39], Hopkins [81], Jamali et al. [82], Maignan and Ralston [24], Pedersen [83], Peterson and Jun [74], Siltaoja Marjo Elisa [84], Strand et al. [50]; ^c Cacioppe et al. [49], Carter and Rogers [58], Clement-Jones [52], Ducassy [55], Harjoto and Jo [56], Hemingway [39], Lattemann et al. [10], López Gamero et al. [31], Margolis et al. [28], Perrini and Minoja [73], Sjostrom [57], Tschopp and Nastanski [59]; ^d Benabou and Tirole [54], Ciliberti et al. [45], Clement-Jones [52], Giannarakis and Litinas [62], Hemingway [39], Jamali et al. [82], Manner [71], Margolis et al. [28], Qian et al. [61], Sigurthorsson [60]; ^e Ciliberti et al. [45], Clement-Jones [52], Frankental [44], Greenfield [85], Hemingway and Maclagan [27], Hemingway [39], Jamali et al. [82], Manner [71]; ^f Chan and Cheung [51], Ciliberti et al. [45], Clement-Jones [52], Hemingway [39], Jamali et al. [82], Lewicka-Strzalecka [86], Manner [71], Rotter [63]; ^g Hemingway [39], Jamali et al. [82], Lewicka-Strzalecka [86], Manner [71]; ^h Hemingway [39], Jamali et al. [82], Manner [71].

. The second group covers studies that basically consider CSR drivers/barriers determined from an outcome-oriented position. These drivers/barriers are found in 79 papers and will now be referred to as “objective drivers” that can either be “favorable/available” or “drivers”, and “unfavorable/not available” or “barriers” to CSR. Essentially, they are based on the demands and general preferences of stakeholders, the availability of resources and difficulties in the firm itself. Objective drivers/barriers that extrinsically promote or hinder CSR are shown in

Table 2. Overview of the objective drivers of and the barriers to the adoption of CSR that are based on a review of the literature.

	Frequency (%)
Stakeholder pressure a	25.8%
Institutional framework b	15.9%
Reputation management c	13.6%
Availability of financial resources d	9.1%
Availability of temporary resources e	7.6%
Leading corporations f	6.8%
Sectorial trends g	6.1%
Availability of human resources h	6.1%
Difficulties involved in interpreting CSR i	5.3%
Low institutional interest j	3.8%
Total	100%

Sources: ^a Cacioppe et al. [49], Chih et al. [91], Clement-Jones [52], Den Hond and De Bakker [103], Dobele et al. [104], Doh et al. [105], Erwin [42], Fabrizi et al. [4] , Gamerschlag et al. [34], Golob and Podnar [89], Graafland and Zhang [102], Graafland et al. [66], Hartmann [67], Henderson [106], Hemingway and Maclagan [27], Huang and Kung [107], Hung [88], Jo and Harjoto [108], Kronenberg and Bergier [90], Lewicka-Strzalecka [86], Lozano [70], Maignan and Ralston [24], Margolis et al. [28], Marquis et al. [109], Moon [98], Morsing and Schultz [110], Pater and Van Lierop [72], Pedersen [83], Perrini et al. [111], Post et al. [112], Preuss [113], Sajjad, et al. [48], Tengblad and Ohlsson [76], Weber [114]; ^b Bondy et al. [17], Campbell [25], Campbell [18], Chih et al. [91], Gond et al. [93], Greenfield [85], Hartmann [67], Henderson [106], Husted and Allen [9], Jamali and Neville [94], Kronenberg and Bergier [90], Kolk and Perego [92], Lattemann et al. [10], Lewicka-Strzalecka [86], Lozano [70], Margolis et al. [28], Matten and Moon [8], Moon [38], Pedersen et al. [115], Salam [75], Tschopp and Nastanski [59]; ^c Block and Wagner [36], Cacioppe et al. [49], Clement-Jones [52], Eberle et al. [96], Harjoto and Jo [3], Hermingway [39], Hur et al. [97], Idowu and Papasolomou [69], Lozano [70], Martínez-Ferrero et al. [95], McWilliams and Siegel [35], Moon [38], Morsing and Schultz [110], Orlitzky et al. [116], Pérez and Rodríguez del Bosque [29], Reverte [12], Selvi et al. [26], Weber [114]; ^d Baron [64], Campbell [18], Deniz Deniz and Cabrera Suarez [47], Graafland and Zhang [102], Henderson [106], Jamali et al. [82], Jenkins [117], Lewicka-Strzalecka [86], López Gamero et al. [31], Lund-Thomsen et al. [100], Van der Heijden et al. [37], Welford and Frost [118]; ^e Baron [64], Deniz Deniz and Cabrera Suarez [47], Frankental [44], Graafland et al. [66], Jamali et al. [82], Jenkins [117], López Gamero et al. [31], Van der Heijden et al. [37], Waldman et al. [77], Welford and Frost [117]; ^f Bondy et al. [17], Cheung et al. [99], Idowu and Papasolomou [69], Kronenberg and Bergier [90], Lozano and Huisingh [119], Lozano [70], Marimon et al. [120], Schadewitz and Niskala [121], Zhao et al. [20]; ^g Campbell [25], Chih et al. [91], Idowu and Papasolomou [69], Jackson and Apostolakou [122], Lewicka-Strzalecka [86], Lund-Thomsen et al. [100], Marimon et al. [119], Pater and Van Lierop [72]; ^h Baron [64], Deniz Deniz and Cabrera Suarez [47], Jamali et al. [82], Jenkins [117], López Gamero et al. [31], Lund-Thomsen et al. [100], Van der Heijden et al. [37], Welford and Frost [117]; ⁱ Campbell [25], Frankental [44], Malik [123], Morrow and Mowatt, S. [124], Tschopp and Nastanski [59], Van der Heijden et al. [37], Welford [125]; ^j Gond et al. [93], Graafland and Zhang [102], Lund-Thomsen et al. [100], Sajjad, et al. [46], Van der Heijden et al. [37].

, with no apparent relationship with the first group of drivers/barriers. The steps followed in this content analysis are described in the sequence diagram in Figure 1.

A review of the papers from the first group shows that the most cited subjective CSR drivers/barriers by frequency of appearance are:

(1)

Integrating ethics: The managerial idea that CSR implies a new management idea [41], which integrates ethical standards [42] into the firm’s management and organizational culture [33,43] and corporate governance procedures [44] is the most commonly cited in 31.9% of the papers. The ethical values and preferences of top management [45,46], culture beliefs [47] and CEO characteristics [39], would thus be aspects with an impact on the degree of CSR development in an organization.

(2)

Sustainable development: CSR may be seen as sharing normative goals with the concept of sustainable development [48]. The second most cited concept, with a frequency of 18.7%, is that managers believe that, in general, CSR integrates aspects related to social–environmental–economic performance or sustainability [49,50,51]. This would include strategic approaches to CSR and commitment to earning profits or the need for a better bottom-line [52], the generation of reliable sustainable CSR [53], the adoption of a more long-term perspective or a realistic view of CSR [54], among other aspects.

(3)

Organizational commitment to transparency (stakeholder confidence): The managerial idea that CSR implies a new organizational commitment to transparency and stakeholder confidence is found with a frequency of 13.2% in the reviewed literature. Specific aspects such as risk management [55], conflict-resolution [56], responsibilities towards stakeholder activism [57] and transparency [58,59] could be contemplated here.

(4)

Philanthropy: The fact that managers believe that CSR is charitable behavior and solidarity-based is cited in 11.0% of the reviewed literature. Some of the aspects considered are philanthropy [60,61], altruism, social activities [62], generalized reciprocity and guilt [28].

(5)

Public relations exercise: The fact that managers believe that CSR is an image strategy rather than a true strategic conviction of the firm is found in 8.8% of the papers, with specific references to aspects such as an exercise in public relations and media campaigns [52].

(6)

Fashion-following: Considering CSR to be an ideal [63], the new buzzword or a passing fashion after corporate scandals [51,52] or corruption [45] is mentioned in 8.8% of the papers.

Managers’ belief that CSR is especially for large corporations or a utopic value that cannot be integrated into or managed by firms, is also mentioned, respectively, in 4.4% and 3.3% of the literature. Table 1 shows an overview of the subjective CSR drivers/barriers based on the literature review.

It is common in the literature to present aspects of CSR that are based on two extremes. However, these two extreme positions represent a theoretical constant because, in practice, companies could take positions between the two extremes and then shift from one perspective to another. The review of the literature shows that the subjective CSR drivers/barriers perceived by managers are based on two extremes associated with the personal values of individual managers. One manager might adopt a reactive stance towards society and deny social responsibility while another might adopt a more proactive approach and make his organization exceed societal expectations [72] or shift from a defensive to a more constructive attitude, similar to the “apathetic entrepreneur” and the “active corporate social entrepreneur” management predispositions towards CSR defined by Hemingway [39].

On the one hand, there are managers who have what academics refer to as a “narrow concept of social responsibility” [41]. The narrow conception of corporate responsibility is closely associated with the classical perspective, suggesting that the main function of business is to provide goods and services that lead to the maximization of profit within the framework of legal requirements [41]. This conception is widespread in the business world [46,82] and sees CSR as a utopia, a passing fashion, an image-oriented strategy, or as pure philanthropy, and focuses its attention on the organization’s institutional function, on property rights and maximizing shareholder value from instrumental stakeholder theory [30]. For them, CSR activities would only take place if they made a net contribution to profits [22], as they believe that they have just two obligations: to earn money for the owners and to comply with legislation [85]. In general, these managers have a narrow conception of corporate responsibility and they are skeptical about CSR as they consider it to be the opposite of good firm practice because it dilutes the creation of wealth [51]. It might be expected that firms whose managers hold such views on CSR in relation to the organization’s institutional function and maximizing shareholder value from instrumental stakeholder theory, would tend to have less commitment to CSR and, therefore, less CSR.

However, there are managers who have what academics refer to as a “wide concept of social responsibility”. The wide conception of CSR entails a wider range of economic, legal, ethical, moral, and philanthropic responsibilities [41,46]. This broad conception is basically associated with managers who support ethical management, transparency in relation to stakeholders and sustainable development (economic, social and environmental) in general. Managers belonging to this group see firms as a part of society with responsibilities over a wide range of issues, focusing on the common good and who suggest that their firms, as institutions that control many of society’s resources, should include social aspects in their strategic decision-making processes [87]. According to Hung [88], firms in which the managers show greater concern for issues affecting stakeholders appear to be more likely to see the need for effective CSR. Hung [88] concludes that managers who have an integral, responsible and effective view of CSR tend to develop it in their firms. One would expect that firms whose managers hold views on CSR stemming from ethical, political, integrative and instrumental stakeholder theories (rather than on the need to maximize shareholder value) would tend to have a greater commitment to CSR and, therefore, more CSR.

Based on all the above, we propose the following hypotheses:

Hypothesis 1A.

The narrower the management conception of CSR (increased presence of subjective barriers), the lower the firm’s CSR.

Hypothesis 1B.

The wider the management conception of CSR (increased presence of subjective drivers), the greater the firm’s CSR.

As with the subjective CSR drivers/barriers, we now analyze the objective CSR drivers/barriers found in 61 different papers. The most cited objective CSR drivers/barriers, by frequency of appearance, are:

(1)

Stakeholder pressure: Some kind of pressure from the relationship with stakeholders [4,34] and from stakeholder dialogue [89], from one or several stakeholders including NGOs [90], appears to be the most commonly mentioned aspect in favor of CSR in the organization, cited in 25.8% of the papers. The objective drivers cited include social demands [48] and stakeholders’ expectations and pressures [45].

(2)

Institutional framework: There is increasing top-down pressure impacting CSR [52]. Institutional issues like public and private regulations, rules regarding corporate behavior and associative behavior among corporations themselves [25,91] are the second most commonly cited objective CSR drivers, cited in 15.9% of our sample. Highlighted aspects include the country-level institutional factors [92], the role that governments can play [93], institutional pressure [10,94] and developing an institutional framework [8].

(3)

Reputation management: CSR has a positive effect on corporate reputation [36], and lowering the cost of capital [95], on loyalty [29] with the integration of the organization into its host community [27] on improving firms’ reputations in relation to their stakeholders [96,97] and their performance in the eyes of governments [98]. This is also one of the most commonly mentioned aspects as a CSR driver, cited in 13.6% of the papers.

(4)

The impact of leading corporations [17] based on aspects such as visibility, the publication of sustainability reports [90], foreign partners [86], international diversification [99], etc., is cited in 6.8%, and aspects in favor of CSR related to sectorial trends and private sector-led initiatives [100] are cited in 6.1%.

(5)

Availability of resources (financial, time-related and human resources): Economic factors such as lack of economic resources [100], a difficult economic situation [86], a situation in which it is difficult to show any significant positive correlation between CSR and the “bottom line” [51], etc., are most commonly cited as barriers, with 9.1% frequency. They are followed by lack of structure and human resources [31] with 7.6%, and lack of time-related resources [37], with 6.1%.

(6)

Difficulties involved in interpreting CSR: The concept and the scope of CSR are difficult to define [101]. Difficulty in determining core CSR targets, roles and responsibilities [51] and how to put them into practice it [37] are cited in 5.3% of the papers as barriers.

(7)

Low institutional interest: The lack of governmental willingness [100] and the lack of official grants [102] are cited in 3.8% of the papers.

The same thing that appears to occur with subjective factors of CSR also seems to occur with managers’ objective factors of CSR. However, as we have already mentioned, managers could take various positions in the middle and then shift from one perspective to another. In this line, following resource-based theory, McWilliams and Siegel [35] assert that managers conduct a cost-benefit analysis to determine the level of resources to devote to CSR activities or attributes, meaning that they assess the demand for CSR and also evaluate the cost of satisfying this demand. Thus, we propose that firms can manage their CSR drivers/barriers in two (rather extreme) ways: a “favorable outcomes—expected benefits” and an “unfavorable outcomes—estimated costs” perspective. Table 2 shows an overview of the drivers of and barriers to the adoption of CSR based on our literature review.

Firms whose managers find outcomes that are favorable to CSR development would be expected to present a greater commitment to CSR, and hence greater CSR. These motivational factors would be specifically related to ethical, political, integrating and instrumental stakeholder theories (instead of maximizing shareholder value) and, above all, resource-based theory. Likewise, firms whose managers find outcomes that are unfavorable to CSR will have less commitment to CSR and hence a lower degree of CSR. In this case, the motivational factors unfavorable to CSR would be related particularly to the organization’s institutional function or to ownership rights and the maximization of shareholder value.

This discussion, together with the reframing of current literature on the phenomenon under investigation, leads us to the following hypotheses:

Hypothesis 2A.

The greater the managers’ perception of profits and benefits derived from CSR (increased presence of favorable outcomes), the greater the firm’s CSR.

Hypothesis 2B.

The greater the managers’ perception of economic or organizational obstacles derived from CSR (increased presence of unfavorable outcomes), the lower the firm’s CSR.

This study’s primary objective is to test the validity of these four hypotheses.

2.2. Data and Methods

This section includes the design, application and results of an empirical analysis that enables us to test our hypotheses based on the information related to a sample of 416 Spanish firms. After a brief reference to the analytical methodology, we describe the sample and the variables. This is followed by the model’s specification and the results obtained.

2.2.1. Methodology

The study uses structural equation modeling (SEM) to identify how different drivers/barriers contribute to the CSR that an organization achieves. Given the abstract nature of these constructs, SEM is an appropriate tool (see, for instance, Bollen [126]). We specify a structural regression model that takes factors determining CSR initiatives as independent variables and CSR as a dependent variable and enables us to evaluate the impact of factors determining CSR initiatives. As there are missing data for some of the study variables, the parameters of the model are estimated following a Bayesian approach, which, as Lee [127] shows, enables us to process this kind of information by random imputations that reflect the uncertainty associated with their values. Furthermore, the Bayesian approach enables exact inferences about model parameters, increasing the reliability of the analysis.

Finally, the constructs obtained by SEM are used to present a graphic representation, using the Quazi and O’Brien [41] two-dimensional model, which has been used many times in the literature [46,82]. Quazi and O’Brien [41] established a two-dimensional CSR representation based on two Cartesian axes. On the one hand, the horizontal axis shows the organization’s conception of social responsibility with two extremes: A narrow conception and a wide conception of social responsibility. On the other hand, the vertical axis shows expectations or the expected results of this social commitment and the two ends of the axis are: the benefits from and the cost of social responsibility. In this model, the firm is finally located in one of the four quadrants formed by the two axes.

2.2.2. Sources of Information and Sample of Firms

In collaboration with the Regional Government of Aragon, and within the framework of its Competitiveness Plan, the Business Confederation of Aragon [128] promoted the creation of a stable group to study CSR in businesses which began to work in 2006. The group comprised voluntary experts on the subject and representatives of leading business organizations, NGOs and public and private institutions. The information about CSR in the firms used in this study was obtained by this group through mail, telephone and face-to-face surveys with the chief executive officers (CEOs) of a representative random sample of Spanish companies located in Aragon from a database of 11,251 companies. From the information provided by the Business Confederation of Aragon, we obtained data about 416 firms in the region’s most representative sectors and in all the firm categories established by the European Commission [129,130] micro-sized enterpises, small-sized enterpises, medium-sized enterpises and large-sized enterpises Aragon is regularly selected in socioeconomic studies as representative of Spain because it has a similar sectorial distribution and representative socioeconomic indicators [131,132]. We were subsequently authorized to make use of the information obtained that was significant for our research. The use of partial information from a questionnaire designed with objectives much broader than those established in this research contributes to avoiding the bias of the common method, more endemic in questionnaires in which the research objective is evident.

Of the 416 analyzed firms, 185 belonged to the industrial sector, 48 to the construction sector, 83 to hostelry and commerce and 100 to other services. Furthermore, 122 were micro-sized enterprises, 177 were small-enterprises, 73 were medium-sized enterprises and 45 were large-sized enterprises

Table 3. Sample distribution by sector and size.

	Number	%
Sample distribution by sector
Industry	185	44.5%
Construction	48	11.5%
Hostelry and Commerce	83	20.0%
Other Services	100	24.0%
TOTAL	416	100.0%
Sample distribution by size of organization following EU criteria
Micro-sized enterprises	122	29.3%
Small-sized enterprises	177	42.5%
Medium-sized enterprises	72	17.3%
Large-sized enterprise	45	10.8%
TOTAL	416	100.0%

shows their distribution by sector and size.

2.2.3. Description of the Variables

Below is a description of the variables, distinguishing between those that enable CSR measurement and those that enable the measurement of the drivers/barriers determining CSR initiatives. The complexity of the measurement models is useful to avoid the bias of the common method that is usually attributed to the study of relations based on primary data.

(1)

Measuring CSR: The evaluation of each firm’s CSR was carried out following the methodology described in Agudo et al. [133] and based on the model proposed by Carroll [40]. The former authors use a set of 53 items, on a 0 to 10 continuous scale, that cover aspects of the different types of responsibility considered by Carroll [40] and later researchers [66,134,135] related to the different stakeholder groups to which measurements of the organization’s responsibility apply. Agudo et al. [133] propose a second-order factor model where the 53 items are grouped in 13 first-order factors and, finally, in a single second order factor denoted here as “Corporate Social Responsibility (FCSR)”. The construct finally obtained as the second-order factor thus summarizes the information contained in the variables related to the organization’s activities and commitments related to CSR. The FCSR determines the organization’s CSR, the variable now estimated for the rest of the study.

(2)

Measuring factors determining CSR initiatives: Following the previous literature, we chose a set of variables that enabled objective measurement of the extent to which managers are aware of the favorable or unfavorable conditions for the development of CSR in their organization, and their conception of CSR. We thus consider eight subjective variables related to management’s conception of CSR and 10 objective conditions for the development of CSR in the organization. Measurement of the 18 variables was based on a seven-point Likert scale, according to the degree of agreement with the issue represented by the variable in the firm, where 1 represents “agree completely” and 7 “disagree completely”. All computations were made using SPSS 15.0.

We identify two different blocks of subjective management CSR drivers/barriers, denoted “subjective drivers of CSR” (S_D) and “subjective barriers to CSR” (S_B) and another two blocks of objective drivers, denoted “objective drivers of CSR” (O_D) and “objective barriers to CSR” (O_B).

Table 4. Variables used to identify managers’ conception of CSR and the conditions that firm managers believe affect CSR development.

Factor Score	Survey Variable Contemplated in the Study	Mean	Typical Deviation	% Missing	Alpha Cronbach
Subjective drivers of CSR (SD)	YSD1 CSR implies integrating ethics into a firm’s management.	5.82	1.73	5.3%	0.581
	YSD2 CSR integrates transparency and affects investor confidence.	5.37	1.75	5.3%
	YSD3 CSR contemplates aspects related to social, environmental and economic performance.	5.68	1.68	5.1%
Subjective barriers to CSR (SB)	YSB1 CSR is especially applicable in large corporations.	3.25	2.04	5.1%	0.738
	YSB2 CSR only focuses on undertaking social and solidarity-based projects.	2.75	1.69	5.3%
	YSB3 CSR is an image strategy rather than a true strategic conviction.	3.46	1.90	4.8%
	YSB4 CSR is only a passing fashion that arose after the latest scandals.	2.56	1.71	5.5%
	YSB5 CSR is a utopian value that cannot be integrated into or managed by firms.	3.24	1.78	5.3%
Objective drivers of CSR (OD)	YOD1 Sectorial trends.	3.58	2.10	6.5%	0.773
	YOD2 Improvement of the firm’s reputation.	5.11	1.95	5´8%
	YOD3 Social demand for more transparency, honesty, ethics.	5.10	1.93	4.3%
	YOD4 Publication of sustainability reports by leading corporations.	4.27	2.13	7.2%
	YOD5 Associated legal and institutional framework.	4.21	2.14	7.4%
Objective Barriers to CSR (OB)	YOB1 The little time available for the subject makes CSR difficult in the firm.	5.29	1.95	2.9%	0.812
	YOB2 CSR cannot be implanted in my firm due to a lack of structure and human resources.	4.77	2.14	3.6%
	YOB3 CSR is not applied in my firm because of the lack of official grants.	4.86	2.14	3.6%
	YOB4 The firm does not develop CSR because of difficulties in interpreting what it involves together with a shortage of CSR training opportunities.	4.87	2.11	3.6%
	YOB5 The lack of economic resources makes CSR difficult in the firm.	4.90	2.09	3.6%

shows the descriptive statistics of the 18 original variables and the score factors for the items corresponding to each construct.

In general, we find significant differences in the scores reflecting the existence of two groups. The first have a wide conception, scoring above 5 (integration of ethics 5.82, relationship with social, environmental and economic balance 5.68 and confidence 5.37). The second group has a narrow conception, scoring under 4 (image strategy 3.46, utopia 3.24, this is limited to large corporations 3.25, it is only social action 2.75 and a passing fashion 2.56). With regard to objective CSR drivers (see Table 4), the managers surveyed believe that the aspects that most favor CSR are related to reputation enhancement (5.11) and social demand (5.10). Aspects related to sector trends (3.58), on the other hand, are seen as being less relevant in the decision regarding active social response. In intermediate positions are organizations that are leaders in these practices (4.27) and the legal and institutional framework (4.21). Likewise, with regard to objective barriers to CSR (see Table 4), there is a medium to high level of consensus regarding their relevance in hindering firms’ social response, with mean scores ranging from 5.29 to 4.77. The objective barriers that prevent CSR development are, in this order: lack of time (5.29); shortage of economic resources (4.90); difficult interpretation (4.87); lack of official grants (4.86); and lack of necessary human resources (4.77). Finally, the reliability of our analysis is considered adequate given that all the Cronbach alpha values are larger than 0.5 (see Table 4) and most of them are larger than 0.7, justifying the selection of the variables considered in the study.

2.2.4. Structural Model

We now consider a structural equation model to analyze the impact of the organization’s factors determining CSR initiatives (S_F, S_U, O_F and O_U) on its CSR (F_CSR). We describe the model and the results obtained using the AMOS 7.0 program. The model used to test the hypotheses established in Section 2 is described by the sequence diagram in Figure 2, where a CFA separate measurement model is specified for constructs S_D, S_B, O_D, O_B and F_CSR and a structural regression model for the relationship between them.

The model was estimated using a Bayesian approach and a diffuse prior distribution. In the Bayesian approach, inferences about the model’s parameters are based on their posterior distribution which is calculated using the Bayes Theorem [127,136]. In particular, point estimations of the parameters of the model (factor loadings and regression coefficients of the structural model) are calculated using their posterior median, and 95% Bayesian credibility intervals through their posterior 2.5 and 97.5 quantiles [137]. Additionally inferences about the factor score vector of each organization (F_Sd,ℓ, F_Sb,ℓ, F_Od,ℓ, F_Ob,ℓ)’; ℓ = 1, …, 416 [138] can be made, where F_Sd,ℓ, F_Sb,ℓ, F_Od,ℓ and F_Ob,ℓ represent the scores for wide concept, narrow concept, objective drivers and objective barriers, respectively. As this posterior distribution is not analytically tractable, it has to be calculated by approximate methods, specifically, Gibbs sampling (see, for instance, Lee [127]). Gibbs sampling is a Monte Carlo method that draws a sample from the above posterior distribution. From this sample, it is possible to calculate posterior quantiles using the respective quantiles of the sample. All these inferences are exact and do not depend on asymptotic results, increasing the reliability of the estimation process.

3. Empirical Results

The results of the empirical analysis are described in two different subsections: first, the goodness of fit of the model is analyzed, then the model estimation results are presented and, finally, the results of the joint analysis of the subjective and objective drivers of CSR are shown.

The goodness of fit of the model is analyzed in

Table 5. Goodness of fit of the model.

Criterion	Independent	Model Proposed	Complete
BIC	2.43 × 1012	2.38 × 1012	2.44 × 1012
DIC	4.52 × 1012	2.40 × 1012	4.03 × 1012
LPRED	−1.73 × 1012	−1.18 × 1012	−1.56 × 1012
R2		0.4499
COV95		95.04%
COV99		98.97%

. First, a comparison of the proposed model is carried out with two alternatives: the independence model, where it is assumed that all the considered variables are independent (and, therefore, their covariance matrix is diagonal), and the complete model, where it is assumed that all the variables are interrelated (and, therefore, their covariance matrix is unrestricted). These are usually considered as reference models in SEM literature. Both models were estimated using a diffuse prior with AMOS 7.0.

We use three comparison criteria: BIC (Bayesian Information Criterion) which is asymptotically equivalent to the Bayes factor and is consistent; DIC (Deviance Information Criterion) proposed by Spiegelhalter et al. [139] which evaluates the predictive ability of the model with samples similar to that considered in the paper and LPRED (logarithm of the posterior predictive density evaluated in the observed data). These three criteria take into account the goodness of fit of the model to data and its complexity, and result in lower accuracy in the estimation of its parameters, that the lower their value (in case of BIC and DIC) or the higher (in case of LPRED) the better the goodness of fit to data. It can be noted that, according to these three criteria, our proposed model has a better goodness of fit. The empirical coverage of the 95% and 99% posterior predictive intervals of the observations is also shown. It can be see that these coverages do not significantly differ from the nominal credibility which reflects that the observations are compatible with the predictions made by our model. Finally, to measure the explanatory power of the structural model, the R² is shown. Its value is equal to 0.45, which shows a moderate explanatory power because the model explains 45% of the total data variation.

3.1. Estimation of the Model

Table 6. Estimation of the parameters.

a. Posterior Factors Loading Estimations of the CFA Measurement Model
Factor	Factor Loading	Median	Limits to 95% of the Distribution
			Lower	Upper
Subjective drivers of CSR (SD)	SD1	1
	SD2	0.724	0.550	0.900
	SD3	0.694	0.525	0.863
Subjective barriers to CSR (SB)	SB1	1
	SB2	0.991	0.813	1.171
	SB3	1.046	0.847	1.247
	SB4	0.837	0.664	1.011
	SB5	0.861	0.675	1.044
Objective driver of CSR (OD)	OD1	1
	OD2	1.163	1.006	1.325
	OD3	0.876	0.692	1.059
	OD4	1.392	1.210	1.586
	OD5	1.218	1.024	1.418
Objective barriers to CSR (OB)	OB1	1
	OB2	1.251	1.069	1.440
	OB3	1.408	1.224	1.598
	OB4	1.193	1.013	1.377
	OB5	1.063	0.880	1.252
b. Posterior Estimation of the Parameters of the Structural Model
	Median	Limits to 95% of the Distribution
		Lower	Upper
γSD	0.203	0.126	0.302
γSB	−0.153	−0.233	−0.079
γOD	0.069	0.012	0.125
γOB	−0.091	−0.148	−0.037

a,b shows the estimations of the factor loadings λ of the CFA measurement model (Table 6a) and the total effects λ of the structural model (Table 6b). More specifically, they show the point estimation and the 95% Bayesian credibility intervals. As can be seen in Table 6a, all the factor loadings of the measurement model are 95% significantly positive, as their confidence interval limits are positive, showing the validity convergence of the CFA measurement model.

Table 6b and Figure 3 show the estimations of the total effects of factors F_Sd,ℓ, F_Sb,ℓ, F_Od,ℓ and F_Ob,ℓ on CSR.

As expected, the total effects are positive in the case of the subjective drivers (0.203) and of the objective drivers (0.069), and negative for the subjective barriers (−0.153) and for the objective barriers (−0.091). They are all significant at the 95% credibility level. Regarding the subjective barriers, the total effect γ_SB (−0.153) is negative and significant, showing an inverse relationship between the subjective barriers and CSR. Therefore, firms with a narrower concept of CSR tend to have lower CSR, thus confirming Hypothesis 1A. The total effect γ_SD (0.203), however, is positive and significant, showing a direct relationship between the subjective drivers of CSR and CSR. Firms with a wider concept of CSR tend to have better CSR, thus confirming Hypothesis 1B.

Regarding the objective drivers, the total effect γ_OD (0.069) is positive and significant, showing a direct relationship between the objective drivers of CSR and CSR. Therefore, firms with a greater perception of objective drivers of CSR tend to present better CSR, thus confirming Hypothesis 2A. The total effect γ_OB (−0.091) is negative and significant, showing an inverse relation between the objective barriers of CSR and CSR. Therefore, firms with a greater perception of the objective barriers of CSR tend to present a worse CSR, thus confirming Hypothesis 2B.

3.2. Joint Analysis of Subjective and Objective Drivers of CSR

Having estimated the model and confirmed its validity, reliability and goodness of fit, this section studies individual firms’ subjective drivers of CSR, their objective drivers and their CSR itself in order to find groups of firms presenting homogeneous patterns. To that end, we use the factor scores of each firm, estimated by their posterior median value. In this analysis, we adapt the results obtained from the model proposed by Quazi and O’Brien [41], and construct the standard subjective drivers score of a firm calculated as $\frac{F_{Sd}-F_{Sb}}{{\sigma }_{F_{Sd}-F_{Sb}}}$ where ${\sigma }_{F_{Sd}-F_{Sb}}$ denotes the standard deviation of $F_{Sd}-F_{Sb}$, and its standard objective drivers score, calculated as $\frac{F_{Od}-F_{Ob}}{{\sigma }_{F_{Od}-F_{Ob}}}$ where ${\sigma }_{F_{Od}-F_{Ob}}$ denotes the standard deviation of $F_{Od}-F_{Ob}$. We can thus represent each firm’s position according to the two-dimensional scatter of the Quazi and O’Brian model [41] where, on the abscissa is the “CSR subjective drivers minus CSR subjective barriers” axis and on the ordinate is the “CSR objective drivers minus CSR objective barriers” axis. The results are shown in Figure 4.

We can see that the firms with a greater CSR (greater Factor CSR) are in the first quadrant (“modern view” according to Quazi and O’Brian model), shown in a darker color. These are firms in which the favorable subjective drivers and favorable objective drivers predominate. The opposite occurs with the firms in the third quadrant (“classic view” according to Quazi and O’Brian model). The second (“socioeconomic view” according to Quazi and O’Brian model) and fourth (“philanthropic view” according to Quazi and O’rian model) quadrants contain firms in intermediate positions.

Table 7. Characterization of groups by size of organization following EU criteria and by sector.

		Approach to CSR								Total
		Modern		Socioeconomic		Classical		Philanthropic
		Nº	%	Nº	%	Nº	%	Nº	%	Nº	%
Size	Micro-sized enterprises	27	22.1%	16	13.1%	66	54.1%	13	10.7%	122	100%
	Small-sized enterprise	62	35.0%	27	15.3%	61	34.5%	27	15.3%	177	100%
	Medium-sized enterprise	44	61.1%	10	13.9%	10	13.9%	8	11.1%	72	100%
	Large-sized enterprise	24	53.3%	10	22.2%	7	15.6%	4	8.9%	45	100%
	Total	157	37.7%	63	15.1%	144	34.6%	52	12.5%	416	100%
Sector	Industry	74	40.0%	26	14.1%	68	36.8%	17	9.2%	185	100%
	Construction	15	31.3%	6	12.5%	20	41.7%	7	14.6%	48	100%
	Hostelry and commerce	24	28.9%	11	13.3%	36	43.4%	12	14.5%	83	100%
	Other Services	44	44.0%	20	20.0%	20	20.0%	16	16.0%	100	100%
	Total	157	37.7%	63	15.1%	144	34.6%	52	12.5%	416	100%
By Size; p-value χ2 = 0
By Sector; p-value χ2 = 0,040

shows the results obtained when we relate firm size and sector to the approach to CSR, determined from the results shown in Figure 4.

The independence hypothesis that assumes that there are no differences between the groups compared is clearly rejected, as the p-value of the chi-square test is 0, showing that the organizations with a modern approach to CSR clearly tend to be larger (61.1% medium and 53.3% large enterprises). This situation changes drastically in organizations that have a classic approach to CSR, which are fundamentally micro enterprises (54.1%). The modern approach (35%) and the classic approach (34.5%) have the same presence in small companies. At the same time, the socioeconomic and philanthropic approaches (15.3%) have a greater presence in small companies than in larger companies.

Performing the same analysis by sector, we find that there are differences regarding the approach to CSR. The service sector (not including hostelry and commerce) tends to have a modern approach to CSR (44%), while hostelry and commerce (43.4%) and construction (41.7%) present a classic approach. However, in industry, the modern approach (40%) and the classic approach (36.8%) have the same presence and it also has the lowest presence (9.2%) of all the sample in the philanthropic approach. The results indicate that there are important differences between firms of different sizes and sectors in the modern and classical approaches, while the differences between firms of different sizes and sectors are smaller in the socioeconomic and philanthropic approaches.

4. Discussion

The behavior of businesses in CSR is strongly influenced by subjective and objective drivers and barriers according to their potential effect for CSR. Firstly, from a subjective perspective, managers’ subjective interpretation of what CSR means affects the degree of CSR development achieved in firms. Secondly, from an outcome perspective, managers’ perceptions of the benefits (objective drivers) and costs (objective barriers) of CSR affect the degree of CSR development achieved in firms.

The results of this study suggest that CSR subjective drivers can promote the development of CSR in Spanish companies. We find that there is a positive and significant relationship between the subjective drivers of CSR and CSR, confirming the hypothesis that the wider a firm’s conception of CSR, the better its CSR. On the other hand, there is a negative and significant relationship between the subjective barriers of CSR and CSR, confirming the hypothesis that the more limited a firm’s conception of CSR, the worse its CSR.

With regard to objective drivers, aspects related to instrumental stakeholder and integrative theories appear to be the most important, together with the firm’s resources and capabilities. In this case, there is a positive and significant relationship between objective drivers and CSR, confirming the hypothesis that the greater a manager’s expectations of the economic benefits derived from CSR, the better the resulting CSR. The relationship between objective barriers and CSR, however, is negative and significant, confirming the hypothesis that the greater a manager’s expectations of the economic or organizational costs derived from CSR, the worse the resulting CSR.

Our findings provide evidence that firms with higher CSR ratings present a statistically significant higher level of subjective CSR drivers (integrating ethics into organizational culture, sustainable development) and a larger perception of objective CSR drivers (reputation management, stakeholder pressure, institutional framework), compared to firms with lower CSR ratings. Likewise, the firms with lower CSR ratings present a statistically significant higher level of subjective CSR barriers (a public relations exercise, philanthropy, and a fashion) and a larger perception of objective CSR barriers (the availability of the resources and the difficulties involved in interpreting CSR), compared to firms with higher CSR ratings.

These results do not significantly differ from those related to objective CSR drivers in companies in other environments according to the reviewed literature. For example, the empirical results corroborate the findings of Lozano [70], who interviewed some managers and experts and found that ethics, reputation management or legal and institutional framework are some of the most important drivers. In addition, these results are consistent with previous research showing that motivational factors tend to vary according to the type of industry and firm size [140].

5. Conclusions

5.1. General Conclusions

In this article, by examining the differences between sets of CSR drivers and barriers, we proposed a methodology that could be useful for research and business applications to formalize CSR strategies. Through this process, we were able to classify them according to the impact of the drivers and barriers perceived by managers on an organization´s development of CSR, from subjective and objective perspectives.

The results obtained in this study, conducted on a sample of 416 Spanish firms, provide objective information about the different CSR drivers considered by firms. However, our study also sheds light on how these factors lead a company not simply to be socially responsible but also to undertake a CSR approach embedded in corporate strategy. Two elements act as mediators: the subjective CSR drivers/barriers and objective CSR drivers/barriers. Both factor blocks are interconnected and appear to play a fundamental role in shaping CSR strategy.

In Spanish firms, the integration of ethics and sustainable development principles is the aspect of the subjective CSR drivers most perceived by managers, while the objective CSR drivers most commonly mentioned by managers are, in this order, stakeholder pressure, institutional framework and reputation management. In contrast, public relations, charitable behavior and philanthropy are the aspects of the subjective CSR barriers most perceived by managers, while the objective CSR barriers developed by the companies are, in this order, availability of financial, time-related and human resources.

We also found that companies of different sizes and in different industries present determined behaviors. Regarding the effect of firm size, we find that managers of smaller firms tend to have a more limited conception of CSR and their firms tend to present more barriers in their development, which explains their lower CSR. In contrast, larger firms have more drivers for the development of CSR and their managers have a wider conception of CSR, which explains their higher CSR. This is completely understandable from the resource perspective.

We also found differences between sectors. Firms in the construction and hostelry and commerce sectors tend to present a classic approach to CSR, with greater value given to objective barriers and with a narrow conception of the idea. However, the service sector (not including hostelry and commerce) tends to present a more modern approach to CSR.

5.2. Managerial Implications

The output from our model can be helpful in devising operational strategies to improve a firm’s CSR and to achieve competitive advantages with stakeholders as it allows organizations to identify the drivers and measure their degree of compliance with social responsibility objectives. The model can analyze the different ways in which CSR is perceived by managers and the drivers of and barriers to CSR because they are all related to CSR and have statistically significant relationships.

Our analysis of CSR drivers also has significant practical implications. By identifying the objective and subjective drivers of CSR practices, we suggest a framework for practitioners to develop a CSR strategy to align their CSR practices with their multiple environments. This information is also relevant for government, business organizations and stakeholders in general, who are able to learn details about the different CSR drivers considered by firms, and the extent to which they are considered, as they advance in their CSR policies.

5.3. Limitations and Directions for Future Research

The results of this study are subject to some limitations, which need to be taken into account when generalizing the validity of the model. First, the main limitation concerns the geographical context; the empirical test of the model is based on data from samples drawn from only one country, Spain. This is a potential limitation of this research, as it may not be representative of a European setting. Therefore, the results of the study can only be generalized in this country. However, future research can test the applicability of the findings to different contexts and countries.

Second, there are limitations with the analytical techniques used in this study and its rigid structure. Future research may wish to extend the scope of this study to explore the structural relationships between other factors and to include the analysis of companies selected based on geographic criteria.