1 Introduction

Corporate social responsibility (CSR) has been defined in numerous ways (Wood 1991, Garriga/Mele 2004). Nevertheless, nearly all these definitions share the view that CSR is founded on business “actions that appear to further some social good” (McWilliams/Siegel 2001, p. 117). There has been significant research as to whether CSR contributes to the firm’s interests in addition to contributing to the social good. In other words, assuming that a firm wants to be socially responsible, the different ways of achieving this objective will have varying consequences for the firm’s financial performance. The aim of this research is to demonstrate under what conditions CSR contributes to firm value creation among multinational enterprises in Mexico (Burke/Logsdon 1996).

One of the principal issues this research has faced is demonstrating a positive relationship of CSR to financial performance (Waddock/Graves 1997, Griffin/Mahon 1997, McWilliams/Siegel 2000, Margolis/Walsh 2001). Unfortunately, to date the results are mixed, in some cases showing a positive relationship between the two; in others, a negative relationship; and in still others, no relationship. Recent meta-analysis of several decades of CSR-firm performance research suggests that there might be a positive relationship after all (Orlitzky/Schmidt/Rynes 2003), though the weakness of the correlation indicates that we may be debating this issue for some time to come.

Rather than continue this debate, we take a different tack by arguing that we are more likely to find a positive relationship between CSR and financial performance when executives design CSR programs in ways that will lead to the creation of competitive advantages for the firm (Liedtka 2000, Burke/Logsdon 1996). Working within a framework of strategic management, this paper looks at how different, strategic features of CSR programs may create value for the firm. Taking into account these strategic features allows us to consider how firms can manage CSR opportunities and link social action to social and financial performance.

Borrowing from Burke and Logsdon (1996), we look into the potential relationship of five strategic dimensions of CSR programs (centrality, visibility, specificity or appropriability, proactivity, and voluntarism) with value creation. We then present the results of a survey carried out among multinational enterprises (MNEs) in Mexico that tests these five dimensions. We discuss the importance of these results for the strategic management of CSR and make suggestions for future research.

2 Theory

2.1 Value Creation

Economic value is created when consumers are willing to pay more for products and/or services provided by companies than the cost of their inputs (Barney 2001). Value creation in the CSR realm has been defined as “identifiable, measurable economic benefits that the firm expects to receive” (Burke/Logsdon 1996, p. 497). Value creation occurs by combining firm resources in new ways so as to increase the potential productivity of those resources (Moran/Ghoshal 1999, Schumpeter 1934). Thus, value creation is fundamentally, although not exclusively, about innovation (Schrage 2007, Jacobides/Knudesen/Augier 2006, Edwards/Battisti/Neely 2004).

2.2 CSR as a Driver of Value Creation

Several authors have claimed that CSR innovation is possible under a specific set of circumstances and will create economic value for firms (Burke/Logsdon 1996, Kanter 1999). Clearly, not all CSR programs create economic value (Margolis/Walsh 2001). Social initiatives may increase costs and, although they may create value for different stakeholder groups, stockholders may see the value of their shares decline. In order to identify which CSR programs may create economic value, Burke and Logsdon (1996) developed a model of five strategic dimensions of CSR that affect the ability of these programs to create value: centrality, appropriability, proactivity, visibility, and voluntarism. These five strategic CSR dimensions help to explain how resources and capabilities may create value for the firm. The Burke and Logsdon (1996) framework is particularly relevant given its influence on the field (de Bakker/Groenewegen/den Hond 2006).

Burke and Logsdon (1996, p. 496) define centrality as “a measure of the closeness of fit between a CSR policy or programme and the firm’s mission and objectives.” CSR programs usually deal with the social or environmental problems of importance to stakeholders – those groups that affect or are affected by the firm (Freeman 1984). Sometimes these programs have little to do with a firm’s mission. For example, Banamex, one of the largest banks in Mexico and a member of Citigroup, has a CSR project to support the restoration of the chapel of Saints Peter and Paul in Teposcolula, Oaxaca. Though laudable, corporate philanthropy that contributes to chapel restoration is clearly not related to Banamex’s core business. However, Cemex, a Mexican-based MNE and one of the largest cement manufacturers in the world, has instituted a program called “Patrimonio Hoy” [Patrimony Today], which organizes needy families into savings clubs in order to purchase Cemex products and build decent, affordable housing. This project is central to Cemex’s business as a cement manufacturer.

Firms that participate in CSR programs that are highly central to their business missions are more likely to create business value because the firm develops resources and capabilities in the solution of social problems that can then be applied to its business activities. The more closely related the social projects are to the core business mission, the more easily transferable are these resources and capabilities (Kanter 1999). For example, in the case of Cemex, the products used and distribution systems employed in “Patrimonio Hoy” were already in place. As Cemex learned to deal with the poor in Mexico, it was able to apply these business skills to its business expansion in developing countries, especially among the low-income segments within those markets (Hart/Sharma 2004).

A second avenue of value creation is through cost reductions available to the firm by focusing CSR projects on activities within the expertise of the firm. Projects within the business domain of the firm are subject to greater oversight and monitoring from the firm, thus reducing costs compared to initiatives in areas little understood by the firm.

Therefore, highly central programs are likely to create greater value over time than projects that are marginal to the business mission.

Hypothesis 1: The greater the centrality of a firm’s CSR programs, the greater the value creation of such programs.

Visibility refers to the extent to which social initiatives may be observed by the firm’s stakeholders. Visible CSR projects seem to have a positive effect on firm reputation (Fombrun/Shanley 1990). Some research has demonstrated that a good corporate reputation has a significant potential for value creation and is, moreover, difficult for competitors to imitate (Roberts/Dowling 2002). For products that are essentially commodities, reputation can serve as a way to differentiate a corporation and its products in the minds of consumers (Thompson/Thompson 2006). Hence, we should expect that the visibility of CSR programs would enable the firm to differentiate its products from its competitors and, in this way, create value by increasing market share or obtaining a price premium from consumers. For example, Danone Mexico, a subsidiary of the French Groupe Danone, used the “Let’s Build Their Dreams” campaign to make donations to charitable organizations with a mission to help needy children based on the quantity of their products purchased by consumers (Lozano/Moxon/Maas 2003). As a result of the campaign, the public image of the company changed from being cold and unfriendly to being warm and involved in the community. This change seems to have had a positive impact on Danone’s market share, while maintaining their price at a time when the competition offered deep price discounts of up to 40% (Lozano/Moxon/Maas 2003). As a result of the campaign, Danone was apparently able to create customer loyalty and avoided dropping their prices. Thus, it appears that the high visibility of its CSR programs was instrumental to differentiating Danone’s products.

Hypothesis 2: The greater the visibility of a firm’s CSR programs, the greater the value creation of such programs.

Proactivity is the extent to which corporate social initiatives anticipate social trends. Empirical evidence indicates that there tends to be a significant relationship between proactive environmental and social policy and proactive business strategy (Aragón-Correa 1998). In addition, socially proactive firms tend to engage stakeholders effectively and adapt to emerging societal expectations (Meznar/Nigh 1995). These capabilities have been found to be positively associated with value creation by the firm (Sharma/Vredenburg 1998). Thus, firms that employ proactive social practices seem to detect changing social trends and needs more quickly than those firms that do not employ such practices (Meznar/Nigh 1995). The anticipation of changes in social trends and values improves business intelligence, which is often a key element for innovation (Schmidheiny 2006). Continuing with the Cemex example, their innovative housing project combined elements of micro-financing as pioneered by the Grameen Bank in Bangladesh with the construction needs of the poor (Prahalad 2004). By anticipating the social needs of the poor, Cemex has preempted the competition in reaching this underserved segment of the market (Hart/Sharma, 2004).

Hypothesis 3: The greater the proactivity of the firm’s CSR programs, the greater the contribution of those programs to value creation.

Specificity or appropriability, using the term commonly employed in the strategic management literature, may be defined as the ability of the firm to link financial benefits to the achievement of social objectives (Burke/Logsdon 1996). One kind of appropriability that is especially relevant to value creation is product differentiation. Commodities are notoriously difficult to turn into branded products with appropriable rents. CSR can sometimes be treated as a form of product differentiation because it fulfills the needs of some consumers for socially responsible products (McWilliams/Siegel 2001, Reinhardt 1998). Products can be differentiated by tying CSR attributes to a product (product innovation) or using CSR-consistent processes in its production (process innovation) (McWilliams/Siegel 2001). In either case, the firm develops a new market for such CSR products or a willingness of the consumer to pay a price premium for products with CSR attributes. For example, cause-related marketing and “fair trade” initiatives tie CSR attributes to products. Cemex seems to have been able to take cement, a commodity, and increase its market position through CSR initiatives with appropriable value creation (Hart/Sharma 2004).

Hypothesis 4: The greater the appropriability of a firm’s CSR programs, the greater the contribution of those programs to value creation for the firm.

Finally, voluntarism, as defined by Burke and Logsdon (1996), requires that firm social activities be undertaken freely, rather than in response to legal constraints, fiscal incentives, or industry practice. In both the CSR and strategic management literatures, voluntarism is associated with choice. For CSR, choice is defined in terms of exceeding legal and fiscal constraints (McWilliams/Siegel 2001); in strategic management, choice is linked to innovation (Hrebiniak/Joyce 1985). Despite this difference, in both approaches, voluntarism may result in value creation. In strategic management, this process is clear. We speak of first-mover advantage, product differentiation, and so on. However, in the area of CSR, we are unaccustomed to treating it as a source of value creation.

In order to treat CSR as a source of value creation, we must place CSR in its proper context. First, voluntary CSR behaviors are less common than coerced CSR behavior. In most cases, firms simply decide to comply with legal and fiscal requirements rather than go beyond them (Stone 1975). Accordingly, a firm that chooses to engage in voluntary CSR behavior may take advantage of the opportunity to build firm-specific resources and capabilities, necessary to value creation (Barney 1986, Dierickx/Cool 1989). Second, customers value voluntary non-market action more than non-voluntary action (Gulbrandsen 2006). Finally, voluntarism reflects managerial commitment to CSR, which is vital to successful implementation and to ensuring that the organization benefits from more loyal and thus more productive employees (Khoo/Tan 2002, Thomas/Simerly 1995, Ostlund 1977). Thus, we hypothesize:

Hypothesis 5: The more voluntary the CSR programs of a firm, the greater the contribution of those programs to value creation.

3 Methods

3.1 Research Setting

This study examines strategic corporate social responsibility in multinational enterprises (MNEs) in Mexico. Mexico provides an interesting site for this kind of study because of the relative paucity of research regarding corporate social responsibility and performance among MNEs in Latin America (Acutt/Medina-Ross/O’Riordan 2004). Mexico, with 104 million inhabitants, is considered an upper-middle-income country with a gross national income per capita of US$ 7,870 (World Bank 2007). It suffers from the usual problems of industrializing economies. For example, only 78% of the population has access to safe drinking water, while 66% has access to proper sanitation (World Bank 1996). Mexico represents both a very large emerging market as well as a country that is characterized by some of the greatest social and economic inequities in the world (United Nations Development Program 2006).

Manufacturing MNEs operating in Mexico are often involved in the maquiladora tax regimen, which permits the tax-free importation of parts from the United States for assembly in Mexico and their re-exportation to the United States. However, some manufacturing goes beyond simple assembly to the making of parts, as in the automotive industry, where Daimler-Chrysler, Volkswagen, and Ford factories produce parts as well as assemble them. Research, development, and design are not common activities for MNEs operating in Mexico.

Indigenous expressions of CSR have a long history in Mexico (Logsdon/Thomas/van Buren 2006). However, the global CSR discussion has influenced Mexican firms and some are beginning to adopt global practices like corporate social reporting and ISO 14000 environmental certification (Paul et al. 2006). Given the enormity of the social and environmental problems in the country, these issues have taken center stage in the efforts of large Mexican firms to be socially responsible (Logsdon/Thomas/van Buren 2006). Multinational enterprises in Mexico face especially significant public pressure to become good corporate citizens (Acutt/Medina-Ross/O’Riordan 2004, Paul et al. 2006). This paper focuses on the extent to which MNEs are responding to these demands strategically and whether these efforts are helping to create value for the firm.

3.2 Data Collection

The survey instrument was developed to measure the basic constructs of value creation, centrality, appropriability, proactivity, voluntarism, and visibility. The face validity of the instrument was determined by a detailed examination of the instrument by a group of ten academics and business people who reviewed the instrument for items that may have been unclear. The group was shown the different variables and the items used to measure them. Following the method used by Abrahamson (1983), we asked them whether they thought the items looked like they would work as indicators of the variables. As a result of this process, we made a few adjustments in the wording of some items in order to improve their clarity. We then conducted a small pilot study by sending the survey to thirteen firms and asked the chief executive officers to respond. This preliminary study was carried out with firms where the CEO or other high-level executive was an alumni of one of the sponsoring institutions of this study. We analyzed these preliminary responses by using factor analysis to examine the extent to which the items loaded on the variable they were intended to represent. We also examined the correlations among the items for each variable. Based on the positive outcome of these preliminary analyses, which seemed to support the validity of these measures, we then proceeded to distribute the survey as will be described shortly. Although we have not included the results of these very preliminary analyses, these results are available from the authors upon request. We have included a copy of the items used to measure the variables in Appendix A.

As a general introduction to the questionnaire, we explained that the survey studies the way in which firms in Mexico conceive their role in community development. The dependent variable, value creation, was measured by asking the extent to which the firm derives benefits from CSR due to increased customer loyalty, future customers, new products, and new markets.

In order to assess the strategic features of CSR, we developed measures of the five dimensions proposed by Burke and Logsdon (1996). Given the constraints of a survey instrument, we were unable to ask questions about specific CSR projects. Centrality was measured by asking the extent to which social objectives like collaborating in community projects, protecting the environment, and helping solve social problems coincided with the firm’s mission.Footnote 1 Appropriability was a single-item measure that probed the extent to which the firm links its social objectives to the achievement of economic profits (Gardner et al. 1998). Visibility was determined by asking the extent to which corporate social programs are important for improving the image of the firm and increasing the firm’s presence in the news media. Voluntarism was evaluated by asking whether firms participated in social programs because of tax incentives, legal constraints, or industry practice. These items were reverse scored. Finally, proactivity was measured by asking whether the firm analyzes its social environment to respond to expectations, is among the first to adapt corporate practices to expectations, complies with the latest legislative changes, and develops corporate standards that exceed regulatory requirements. Each of these variables was measured using five-point Likert scales.

In order to assess the reliability of the variables developed for this project, Cronbach’s alpha was calculated for each variable. Value creation (α = 0.86), centrality (α = 0.75), proactivity (α = 0.86), and voluntarism (α = 0.83) all were found to have satisfactory levels of reliability. Visibility (α = 0.65) was less than the desired level of 0.70, but sufficient for use in this exploratory kind of research (Nunnally/Bernstein 1994). Since appropriability was a single-item construct, reliability was not calculated.

Convergent validity of the measures was assessed looking at the pairwise correlations between the items for each construct. All are significant at the p < 0.05 level, while 96% are significant at the p < 0.01 level. Therefore, there is evidence of convergent validity for the different measures (Gardner et al. 1998).

Discriminant validity is problematic in a survey instrument based on subjective measures because of problems related to common method variance (Podsakoff/Organ 1986). In order to reduce these problems, we employed a number of methods. Among others, we avoided implying that one response was preferable to another, made all responses of equal effort, paid attention to item wording, used items that were less subject to bias, and provided clear instructions (Nunnally/Bernstein 1994, p. 391).

Although subjective measures of business performance and value creation are not ideal, there is considerable evidence that such measures enjoy high levels of validity (Dess/Robinson 1984, Geringer/Hebert 1989, Venkatraman/Ramanujan 1987). In some contexts, such as in Mexico, gathering actual financial data on firm performance or value creation is almost impossible. Thus, in Mexico, reliance on subjective measures of firm performance and value creation is an appropriate approach (Robins/Tallman/Fladmoe-Lindquist 2002).

We also included firm size, industry, and U.S. origin, as control variables. We measured firm size as the number of employees. Industry was measured as a dummy variable according to the Mexican system of industrial classification. Country of origin was a dummy variable with “1” representing U.S. origin and “0” representing some other country of origin.

There is no comprehensive list of MNEs operating in Mexico. Firms were selected from the corporate membership directory of the American Chamber of Commerce (Amcham). Although other countries have chambers of commerce or similar organizations in Mexico, those organizations are much smaller in size. The U.S. is the largest single source of foreign direct investment in Mexico, accounting for 78% of the total in 2001 (Pacheco-Lopez 2004). In addition, the Amcham directory has already been successfully used for studies of MNEs in Mexico (Robins/Tallman/Fladmoe-Lindquist 2002).

Questionnaires were sent to the general managers of the subsidiaries of the 478 multinational firms listed in the directory, most of which were U.S. firms, although there was a mixture of firms from other countries, including Canada, Germany, France, and the United Kingdom, among others. The sample consisted of 111 firms that have answered the survey, either after the initial mailing or as a result of the follow-up. This level of response represented a response rate of 23.2%. This rate is typical of surveys done with firms in Mexico (Robins/Tallman/Fladmoe-Lindquist 2002). A comparison of the early responders with late responders showed no significant difference in firm size, participation in CSR projects, the use of social strategy, or in competitive environmental factors. In fact, there were no significant differences in the responses to any of the survey questions. Some analysts suggest that late responders are similar to non-responders (Armstrong/Overton 1977). The fact that no significant differences in responses were found between early and late responders suggests that non-response bias is not a problem.

3.3 Data Analysis

Table 1 shows the correlation matrix for the variables. The correlation matrix suggests that a moderate level of multicollinearity may exist among the measures of centrality, appropriability, and proactivity. Multicollinearity occurs when any independent variable is highly correlated with any of the other independent variables. The effect of multicollinearity is to depress the significance of the affected variables. However, moderate levels of correlation should not be damaging to the assumptions of ordinary least squares (OLS) regression (Hanushek/Jackson 1977, p. 90).

Table 1 Correlations
Full size table

The validity of the constructs was analyzed using factor analysis. The hypotheses were analyzed using regression analysis. The dependent variable was value creation. The independent variables included centrality, appropriability, visibility, proactivity, voluntarism, and the control variables. Since the dependent variable was continuous and the data were cross-sectional, a regression model appeared appropriate.

A potential problem that may occur with such data is heteroskedasticity, which is a relationship between the error terms over a range of independent variables (Hair et al. 1992). In order to test for the possibility of heteroskedasticity, we conducted White’s test. The chi-square statistic for the test of first and second moment specification was 52.53 (p = 0.64), but the associated p-value was insignificant. Thus, we cannot reject the assumption of homoskedasticity, which is necessary for OLS regression analysis.

4 Results

An OLS regression was run on all of the independent and control variables. The results appear in Table 2. It appears that the model as a whole was quite significant and explained 68% of the variance in the dependent variable. The regression coefficient for centrality was both positive and significantly different from zero (t = 1.97, p = 0.055). The regression coefficient for appropriability was quite small (0.01) and thus was not significantly different from zero (t = 0.08, p = 0.939). The coefficient for proactivity was both negative and relatively small (−0.08). The standard t-test was insignificant (t = −0.84, p = 0.408). Visibility was positive as hypothesized and the coefficient was significantly different from zero (t = 3.01, p = 0.004). Finally, voluntarism was quite large, but negative (−0.69). Although the coefficient was significant (t = −7.48, p = 0.000), the sign was the opposite of what was expected in the work of Burke and Logsdon. Thus the hypotheses for two of the theoretical variables (centrality and visibility) were supported by the data, while the impact of voluntarism was negative. The two hypotheses regarding the impact of appropriability and proactivity on value creation were not supported. None of the control variables was significant.

Table 2 Results of O.L.S. Regression Analysis
Full size table

As mentioned earlier, one problem that may violate the assumptions of OLS regression is multicollinearity. The variance inflation factor (VIF) is one measure of the effect the other independent variables have on the variance of a regression coefficient. Large VIF values indicate high collinearity. All values of the VIF were below the suggested cutoff of 10, except for the dummy control variable for the chemical, petroleum, plastic, glass, and cement industries (Hair et al. 1992). Thus, the problem of multicollinearity does not appear to be significant for the theoretical variables of interest. The consequence for the dummy variable is that it may actually be somewhat more significant than reported in Table 2.

5 Discussion

As predicted, voluntarism is an essential element for the creation of value; however, not in the direction hypothesized. According to the original work by Burke and Logsdon (1996), one would expect that greater voluntarism would lead to greater creation of value from strategic CSR programs. On the contrary, among MNEs in Mexico, value creation is perceived to be associated with constraints such as legal requirements, industry practice, and fiscal incentives. The implications of this finding for public policy in Mexico are enormous. These results suggest that MNEs in Mexico are more likely to create value from CSR programs when such programs arise as the result of industry, tax, or regulatory constraints. Voluntary CSR activity may be more altruistic in nature, but appears less likely to create value for the firm in Mexico. This result is consistent with the argument advanced by Porter and van der Linde (1995) that environmental regulation can stimulate innovation and competitiveness among firms.

Not all competitive environmental contexts provide such opportunities. An examination of the Mexican legal environment provides insight into why this is so and the anomalous result found in our study. For example, Mexican law mandates employee education programs, which may be considered a form of CSR. In addition, firms can avoid paying social security taxes if they provide their employees with appropriate medical services. Given the social protections afforded by Mexican law and in light of its lax enforcement, simply obeying the law sets a company apart as socially responsible. Consequently, compliance with the law in Mexico may represent voluntary behavior and thus the original Burke and Logsdon framework may hold in Mexico. Further research would be required to confirm this contention.

Centrality, as operationalized in this study, is also a relevant dimension of CSR programs that significantly affects value creation. In other words, the greater the extent to which certain objectives of social programs coincide with the firm’s business mission, the more likely these social programs will generate value. This finding is consistent with the argument that firms that develop resources and capabilities through CSR may be able to leverage benefits for their core business. Although beyond the scope of this study, we might speculate that centrality enables the redeployment of resources and capabilities developed for CSR to product and process innovation – a source of significant and lasting competitive advantage (Kanter 1999). For this reason, as Cemex has demonstrated, aligning CSR objectives and operations with the firm’s business products, services, and operations may be at the core of value creation. This is an important area for future research.

Visibility is also significantly related to value creation. This result is consistent with the expectation that much of CSR is developed in order to improve the firm’s image and maintain the firm’s presence in the news media, possibly allowing it to differentiate its products in the market.

Somewhat surprisingly, the hypotheses related to appropriability and proactivity were not confirmed. The lack of significance of proactivity is consistent with the finding that non-voluntarism is related to value creation among MNEs in Mexico. Given the generally low level of CSR in Mexico, it might make sense for MNEs not to be proactive. Quite possibly CSR initiatives are developed at home country headquarters and some local subsidiaries may not carefully monitor the host-country environment in order to respond proactively to local needs (Husted/Allen 2006). Stakeholders at home countries may be driving CSR activity more than stakeholders in host countries, especially when we are speaking of newly industrializing economies like Mexico. Certainly, more research would be necessary to confirm these suspicions.

As argued earlier, appropriability would ordinarily seem to be an essential element in order to create value from CSR projects. Two possible explanations may account for this unexpected result. First, given what appears to be a lack of concern by MNEs to understand CSR needs proactively, there is undoubtedly a lack of attention to the design of CSR projects in ways that will generate profits. Many markets in Mexico are not competitive, and thus firms do not need to worry about deliberately linking the achievement of social objectives to profits. Second, appropriability was measured with a single item and further refinement of its measurement may be necessary.

The overall picture painted by these results seems to indicate that MNEs in Mexico do create value from CSR projects as a result of the fact that CSR is related to the firm’s business mission and is related to programs that enhance the image of the firm. However, it is not clear that CSR value creation is a result of strategic planning on the part of MNEs. CSR is spurred by industry and regulatory constraints, and in this respect MNEs in Mexico may differ somewhat from their parent companies. Burke and Logsdon (1996) wrote within a U.S. context, where domestic MNEs may more proactively engage in “beyond-compliance” behavior and design CSR in ways that will extract benefits.

In our work with MNEs in Mexico, we have seen that firms rarely evaluate social initiatives in this way, though larger global MNEs do understand that social initiatives play an increasingly important role in corporate strategy. Telefonica, the Spanish telecommunications leader in Mexico, Brasil and other Latin American markets, manages CSR as part of its ongoing global corporate reputation project. The project began in 2001 with an assessment of corporate risk across stakeholder groups in all the markets where Telefonica operates. This analysis determined that Telefonica’s principal business risks were associated with reputation loss with key stakeholders, in particular customers. The firm’s response was to engage in a company-wide reputation project which resulted in developing social initiatives to respond to each risk. In few cases were the individual projects central to the firm’s strategy, nor did they provide direct opportunities for value appropriation. Rather Telefonica understood that its very survival depended on maintaining a reputation for creating social welfare and translating that perception into improved relations with key stakeholders, starting with customers, regulators and investors.

Accordingly, Telefonica organized its social projects to extract maximum reputation value from these programs. The firm brought all social projects under the same umbrella, including philanthropy, sponsorships, community involvement, employee programs, as well as mandated social action projects (free service to rural areas, for example).

Telefonica’s astute reputation-building strategy focused firstly on the importance of visibility. However, in one specific area, customer service, centrality was tied to social welfare. Mandated service levels were linked to CSR, as Telefonica competed for licenses with other telecoms on both price and service levels for all customers – e.g., wait time for new lines for regular customers and no-cost or low-cost service to rural areas. Telefonica decided to include these commitments to customers and their communities within the corporate responsibility chapter of its agreements with local governments, a practice that it has continued to date.

In our discussions with Alberto Andreu, Director of the Department of Corporate Reputation, Brand and CSR, he has described Telefonica’s customer strategy as follows. The customer is considered firstly as a stakeholder. The company meets this stakeholder’s need via innovation, superior service levels, and by extending free or low-price service to disadvantaged groups. Responsibility builds reputation and brand and directly leads to superior market performance. In this respect, corporate responsibility with a primary stakeholder is directly linked to centrality and appropriability. And though, in principle, much of what Telefonica does is linked to mandated requirements, and hence not voluntary, the company has sought to re-position these mandated activities as value-added activities.

A much more direct example of creating competitive advantage through social initiatives is HEB, a large grocery chain based in Texas, which entered the Mexican market over ten years ago. One of their social projects involves the donation of food products about to expire to local food banks (Austin et al. 2004). Beyond simple donations, HEB managers worked closely with food bank officials in Monterrey, Mexico to streamline the process and reduce waste as much possible. The concept of a food bank is not a novel one in Mexico, and although participation is not required by law, it is a relatively common industry practice. Despite the potential of this project to create value, because of the close relation of the food bank’s objectives to HEB’s core business mission as a supermarket and the appropriation of some benefits through reduced costs due to close collaboration with food bank personnel, this potential was limited by the reluctance of HEB to publicize its activities and raise the project’s visibility in the wider community. Given the significance of visibility to value creation as confirmed in this study, the low visibility of the HEB-Monterrey Food Bank project calls into question its strategic value for the firm.

Despite the lack of evidence that firms undertake and evaluate their CSR initiatives in accordance with the Burke and Logsdon (1996) model for creating competitive advantage via social projects, the cases of Telefonica and HEB illustrate how pursuing one or more of the elements may limit or enhance the potential for value creation. As Burke and Logsdon argue, the incorporation of multiple elements of competitive advantage increases the likelihood that a particular CSR initiative will succeed and create value for the firm. The results of this study and our work with individual firms indicate that to date the focus has been on visibility and centrality as the key drivers of value creation for social initiatives in Mexico. Clearly, there is an opportunity for firms to move ahead to seek competitive advantage and value creation in a richer and more robust fashion.

6 Conclusions

This paper operationalizes and examines the causes of CSR value creation via social action projects among MNEs. We can say that value creation appears to be related to centrality and visibility, as predicted by Burke and Logsdon (1996). This confirmatory result contrasts with the finding that in the Mexican context, MNEs actually benefit from involuntary constraints such as governmental regulation, tax incentives, and industry practice. MNEs in Mexico do not appear to be proactive in strategically monitoring their environment or designing CSR projects in ways that capture economic benefits for the firm.

Together these findings raise important issues for CSR management. One possible interpretation of these results is that CSR as practiced by MNEs in Mexico may very well be less sophisticated than that practiced by MNEs in their home countries and may be more similar to that practiced by Mexican companies in Mexico. This lack of sophistication may be due to Mexican customers who are not as demanding as home-country customers. This possible explanation needs to be treated as a hypothesis subject to further study.

The discussion surrounding voluntarism suggests that in Mexico, at least, CSR may simply mean complying with legal regulation – contrary to the original definition of CSR laid out in the introduction of this paper, which defines CSR as “going beyond the interests of the firm and that which is required by law.” Of course, this interpretation needs to be tested rigorously.

In this context, Cemex emerges as a crucial counter-example. As a Mexican-based MNE, Cemex has been especially innovative in CSR, with a clear agenda that appears to encompass clearly all the strategic elements set out by Burke and Logsdon. Moreover, outside of Mexico, the Cemex program has achieved enormous visibility (Prahalad 2004, Hart/Sharma 2004) and the firm is quite proud of its achievements and believes that its strong CSR record is a key element in its successful international expansion (Cemex 2005). One area of future research will be to examine the influence of such successful cases of strategic CSR by local MNEs on domestic and foreign MNE behavior in the future, in both Mexico and, by extension, other developing countries, such as Brazil and India, where local MNE’s have begun to develop strategic CSR programs.

Clearly, further research is necessary to understand more fully the puzzling behavior of multinational enterprises in Mexico. The expectations set up by the Burke and Logsdon (1996) framework are only partially supported by the MNEs in Mexico. It would be useful to study whether MNEs behave similarly in other countries that are economically and socially like Mexico. One of the limitations of this study is that while we measure the extent to which the objectives of a firm’s CSR programs coincide with its mission or are visible to the public, we do not have more specific data about how CSR actually supports the firm’s strategy in Mexico, nor the extent to which CSR is integrated into the firm’s strategy in its home country. We have not controlled for specific value chain activities, so we do not know the impact that different activities might have on value creation. Some firms are only involved in the assembly of parts manufactured in other countries. Others are only engaged in marketing and distribution. Still others undertake manufacturing. Future studies should take these different activities into account. In addition, it would be important to include more objective measures of value creation beyond the subjective, self-reports of the respondents. In the United States, for example, financial data is much more readily available and research there could take advantage of these more objective measures.

Much more detailed case-based examination is required to understand more deeply the interaction of strategic variables at the firm level. Although our questionnaire asks executives to evaluate their CSR programs along the Burke and Logsdon dimensions, further research should analyze CSR at the level of specific initiatives in order to better understand the processes through which these dimensions create economic value.

Other areas for further research include examining whether MNEs from developed home countries differentially engage in CSR depending upon whether the host country is a developed or less developed country. By the same token, it would be equally useful to examine how MNEs from developing home countries, such as Cemex, understand their CSR role in developed countries with fewer, or at least less dramatic, social needs. Another issue to be explored is the impact of local corruption on CSR behavior and the extent to which possible rewards for non-compliance affect CSR programs. In the case of Mexico, it appears that compliance is a differentiating factor; however, it is possible that extreme levels of corruption might simply make compliance an unattractive source of value creation.

Finally, given the pressures brought to bear on MNEs to be the vanguard of CSR and to demonstrate that they are capable of using their extraordinary economic power to help cure the world’s ills, research into how MNEs can strategically manage CSR to create value is vital to ensuring that MNEs continue to take on CSR as part of their mission. The CSR boom of the last decade is under continual pressure (Economist 2005); unless it is demonstrated how CSR “pays off”, it is unlikely that most MNEs will maintain their commitment.