As the process of globalization increases, an interesting, yet unanswered question, is: What is the impact of international expansion of business activities on the social orientation of firms? The recent public reaction to the meeting of the World Trade Organization in Seattle, Washington, in 1999, would indicate that some elements of the public are concerned that further expansion will have a deleterious impact on worker’s rights, civil liberties, and the environment, in many countries around the world. While it may be possible to dismiss the public demonstrations against the World Trade Organization as the work of misguided individuals, it does focus attention on the role that multinational corporations play in society. It also presents a challenge to researchers to examine these relationships more closely.

Historically viewed in a very negative light, MNCs have been found guilty of violations of human rights, exploitation of the environment, and corruption of governments. However, there are signs that this has changed in recent times, and that the reputation of MNCs has improved dramatically. More importantly, there are indications that MNCs are including social issues management as an integral part of their corporate strategy (Dechant and Altman, 1994). An encouraging result of this is that the trend seems to be a shift from simple compliance with governmental regulations and avoidance of liabilities, toward a more proactive stance representing adaptation to country-specific needs and innovation in management practices as a means of establishing a competitive advantage. We can see that where countries once passed laws limiting their influence, MNCs are now being courted by national and local governments eager for the economic opportunities that they bring.

Further, ethical issues unique to MNCs arose from the diversity of national operating locations set against the uniformity of the multinational organization (Tavis, 1996). This invited MNCs to act in an egocentric manner, in part, because of a lack of any other uniform set of standards to apply across multiple cultures and operating situations. Over the last three decades changes in corporate attitude can be attributed to the establishment of numerous laws and regulations within the U.S. (Ausubel and Sladovich, 1989), and to the creation of a variety of international accords.

Modern international environmental law dates from 1972 with the establishment of the United Nations Environment Program (Weiss, 1994). Since that time, a number of agreements have been established which are setting the pace for cooperation among nations in the development of international environmental law. In fact, there are over nine hundred international legal instruments intended to protect various elements of the environment (Choucri, 1993; Weiss, 1994). It can be argued that MNCs now have more of a level playing field in that all corporations, regardless of country of origin, are expected to play by the same rules and regulations.

Despite these changes in the competitive landscape, we are not aware of any empirical research to examine the relationship between the degree of multinational involvement and corporate social performance. An examination of the ABI data base revealed that while there have been over two dozen papers published in the last two decades addressing multinationality and corporate social performance, all were prescriptive in nature. The purpose of this study was to extend our knowledge in this important area.

This study examined the corporate social orientation of U.S. multinational corporations using a longitudinal design to determine if there was a relationship between the degree of social orientation and the exposure of firms to the multinational competitive environment.

One of the reasons suggested for the lack of research was that there is no theoretical framework specific to the international dimensions of corporate social performance (Wood and Pasquero, 1992). Gnyawali (1996) suggests that corporate social performance (CSP) is shaped by a combination of firm specific, country specific, and global factors. Further, each firm’s unique set of characteristics is seen to shape the responses of a firm to specific challenges. While this is no doubt true at the micro level, we believe that in the aggregate firm behavior should exhibit similarities that can be observed and generalized to a larger population. This is reflected in the definition of corporate social performance suggested by Wood: "a business organization’s configuration of principles of social responsibility, processes of social responsiveness, and policies, programs, and observable outcomes as they relate to the firm’s societal relationships" (Wood, 1991:693).

Freeman (1984) argued that as the social and political issues of importance to organizations became more complex, firms would of necessity consider and evaluate their relationships with more of their stakeholders. It would seem logical to then conclude that the changes in both U.S. and international laws governing environmental and social issues would have a greater impact on those firms that have a greater international presence. The act of attempting to comply with various laws would require firms to develop an understanding of how their actions affected peoples and organizations in various cultural settings.

Also, the traditional view that the primary purpose of the corporation is to maximize shareholder wealth (Friedman, 1962) is being replaced by the view that the purpose is to satisfy needs within society (Freeman, 1984). From this perspective, shareholders comprise only one of many groups who’s needs must be addressed. Other constituencies, which include customers, suppliers, governments, and society, have a right to voice concerns when their needs are not being adequately addressed. As these groups gain power through established laws corporations are obliged to respond appropriately.

We can view corporations as social communities that serve as mechanisms for the creation and transformation of knowledge into economically rewarded products and services (Kogut and Zander, 1993). As corporations interact with different constituencies they institutionalize those patterns of actions which prove to be most rewarding from both an economic and socially responsible perspective. This process contributes to establishing the legitimacy of the organization within its socio-economic environment, and encourages the constant flow of resources necessary for its continued survival. This implies that we should be able to see a direct relationship between the degree of adaptation to the needs of constituents and growth in the economic rewards received by the firm.

As firms expand beyond their original national borders, the first stage is normally to export products from their home base. While they can be classified as being international firms, there is relatively little need for interaction with the cultures, customs or laws of the countries to which they export. Adler (1991) classified such firms as being multi-domestic. When firms move production facilities beyond their borders they are becoming what can be classified as multinational corporations (MNCs). At this stage they must take into consideration national differences in customer needs, which are closely related to the culture and customs of that country (Herbert, 1984). By establishing business units in other countries firms are also exposed to a greater extent to the laws, rules and regulations governing trade within different countries.

Extending this to the social issues arena, we can say that this expanded awareness of other cultures, values and laws should contribute to a greater understanding of the role the firm plays in the larger international setting. This increased awareness and experience should contribute to improving overall corporate social performance. We then hypothesize that:

While this hypothesis provides some general guidance for empirical testing, it is important to note that we have not been able to pinpoint the precise nature of the relationship between the degree of involvement in international involvement and corporate social performance. As identified in the literature review and theory building process, two forces are at work which have an impact on the social responsiveness of MNCs. First, there is the "push" from the external environment; particularly various governmental agencies and international groups that develop and implement regulations regarding MNC activities, which could force firms to be more responsive socially. Secondly, the "pull" from MNCs as they recognize the importance of social responsibility in global settings. As we noted earlier, more firms are beginning to integrate social responsibility as an integral part of their strategic orientation.

These two forces may have differing effects on firms’ social performance as they expand international operations. Specifically, the forces of the "push" from the external environment would lead us to suggest that at a given period, the greater the multinational activities, the greater the social performance after controlling for other factors such as industry and size. This would be a general reaction to, and compliance with, country-specific laws and regulations, as well as to the greater international community which demands increased socially responsibility performance.

On the other hand, when firms begin to actively integrate social responsibility into their respective strategic orientation, it takes time for them to monitor and scan the environment in order to understand various dimensions of the social issues being faced, develop meaningful responses to those social issues, and implement their responses. Therefore, if the forces of the "pull" from firms are at play, we would see the multinational activities observed in an earlier stage being linked with corporate social performance observed in a later period.

This is an important consideration as firms explore social issues as part of their strategic management process, and discover issues important to different stakeholder groups, develop responses, and implement those responses over time. A related consideration is whether social performance will increase as firms increase their multinational activities. An understanding of the results of such changes will provide more insights to the relationship between multinational efforts and corporate social performance. Thus, we formulate three research questions related to the main hypothesis:

The purpose of this study was to examine empirically the nature of the relationship between involvement in international competitive activities and Corporate Social Performance (CSP). Specifically, we wanted to understand the relationship between expanded international presence and its impact on social responsibility.

Data for the corporate social performance measure were taken from the Kinder, Lydenberg and Domini (KLD) social index. Our performance measure was taken from the Stern Stewart Performance 1000. Each of these is discussed in more detail below. Other data were collected from Compustat. The years chosen for this study were 1991 and 1996, the latest year for which complete data were available. Our data sample consisted of 350 firms operating across 56 industries.

Variables

Corporate social performance. Researchers have historically made difficult choices in an effort to measure this construct. Some prior studies have used the Fortune reputation data set (e.g., McGuire, Sundgren and Schneeweis, 1988), forced choice survey instruments (e.g., Aupperle, Carroll and Hatfield, 1985), and content analysis of documents and case study data (e.g., Abbott and Monsen,1979; Clarkson, 1991). Sturdivant and Ginter (1977) used an index from the Council of Concerned Businessmen, while Moskowitz (1975) used rankings from Business and Society.

However, each of these approaches has significant limitations. The Fortune data set has been criticized as being biased toward the financial performance of the firms being rated rather than being objective indicators of purely social performance (Fombrun and Shanley, 1990). Survey based approaches can be influenced by a desire to appear socially acceptable, thus creating self-report bias. Further, Gatewood and Carroll (1991) pointed out that the validity of the indexes and ranking instruments depends on the knowledge of the evaluators, an issue that is difficult to judge.

In an effort to overcome some of these limitations, this study employed the Kinder, Lydenberg and Domini (KLD) database to operationalize the corporate social performance construct. Our measure is intended as a surrogate for corporate social performance, and we feel that it is a fair representation of the overall social orientation of the firm. There has been growing acceptance of the KLD data base among social issues researchers (e.g., Graves, 1988; Graves and Waddock, 1992; Sharfman, 1993; Thomas and Simerly, 1995; Waddock and Graves, 1997). As Sharfman (1993) notes, it captures the multidimensional nature of the CSP construct. Because the data is gathered by individuals outside the firms being evaluated, it is less vulnerable to self-report bias. While this measure has its limitations, it is the best researched and most comprehensive measure available (Wood and Jones, 1995).

KLD corporation is an investment advisory firm which compiles data used to evaluate some 700 firms on eight separate dimensions: community relations, employee relations, environmental, product quality and liability, women and minority issues, nuclear power, military investments and investments in South Africa. They use a five point rating scale, ranging from -2 (major weakness) to +2 (major strength) for the first five of the eight dimensions. Since the last three dimensions are not scaled, this study only used the first five dimensions.

One limitation to the KLD data set is that it assigns equal weight to each of the dimensions. This means that it is not possible to capture the incremental differences across the dimensions. Ruf, Muralidhar and Paul (1993) surveyed practicing managers and asked them to evaluate the relative importance of the dimensions using an analytical hierarchy modeling and measuring process. The results were used as inputs to an algorithm to create vector weights which reflect the relative weights given to each dimension of the KLD data set. The following weights were developed, and used in this study: .22 for product quality and liability, .18 for employee relations, .15 for women and minority issues, .14 for environment, and .12 for community relations. In other words, we took the CSP score created by KLD for each firm and multiplied it by the weights and summed the scores to derive an index of social performance. This process was repeated for each firm over a two year period. This resulted in a single value of CSP for each firm in the sample ranging from -2 to +1.67. This process was performed for each year of the study.

MNC activities: The degree to which firms had achieved multinational status was measured by dividing firm total sales by the proportion of sales which occurred outside the United States. The larger the number, the greater the percentage of sales from multinational activities. The Compustat data set provided information on the sales of firms in their various geographic segments together with total sales for 1991 and 1996

Controls: Two aspects are important for control purposes. On the one hand we need to control firm level factors. On the other hand we also need to control industrial level factors.

At the firm level, firm size is an important factor. Larger firms may have more resources, hence greater ability to pursue socially responsible behavior. However, small firms may enjoy greater agility and flexibility. On the other hand, studies have found that larger firms tend to engage in socially irresponsible activities (Baucus & Near, 1991; Hill, Kelley, Agle, Hitt & Hoskisson, 1992). While the potential impact is unclear, it is clear that we should control size when explaining firm social performance. In this study, we used the log transformation of total employees as the indicator for firm size. Due to factor cost differences related to the multiple industry sample in this study we also considered other measures of firm size (e.g., log transformation of total sales revenue, and log transformation of total assets). The substantive findings remained essentially the same, therefore we only report the results using log transformation of total employees as the indicator for firm size.

Industry level characteristics are believed to influence a firm’s social orientation and subsequently social performance as well. For example different industries may experience different risks. Prior studies have made a clear connection between risk and CSP. For example, Fombrun and Shanley (1990) found a strong relationship between risk and the judgment that external stakeholders held of firms. Also, McGuire, et. al. (1988) found that market based risk measures were negatively associated with social responsibility. In addition, other industry environment characteristics such as munificence and dynamism are also believed to influence a firm’s social performance. As the central focus of this study was to investigate the nature of the relationship between MNC activities and CSP, we chose to use industry codes as a fixed control factor to capture the potential impact of industry level risk, munificence and dynamism. Our industry code was a dummy variable provided by Stern Stewart representing the industry which provided the majority of a firm’s revenue. A similar measure was used by Waddock and Graves (1997) as a control for industry effects.

Economic performance: We felt that it was necessary to control for the effects of economic performance because studies have demonstrated a positive relationship between social performance and economic performance (e.g. Waddock & Graves, 1997) but there is still a question as to the causality. Since this is an unanswered question beyond the scope of this paper, we felt it better to include economic performance as a control. The question of the relationship between corporate social performance and economic performance is still one of the more difficult questions within the social issues field (Wood and Jones, 1995; Steiner and Steiner, 1994). For purposes of this paper, we wish to avoid problems resulting from past research and take a perspective which reflects the holistic nature of our research. According to Wood and Jones (1995) market measures of performance have been most consistent in finding positive relationships between CSP and performance. For this reason we have chosen such a measure. The measure of economic performance used in this study was Market Value Added. The data for this measure were provided by the Stern Steward Corporation. This measure has been used in prior empirical research (e.g., Baliga, Moyer, and Rao, 1996).

Market value added is the difference between a company's market value and its invested capital. This measure was averaged across a five year time period. This measure reflects the market's assessment of the net present value of all the firm's past and projected capital investment projects. Therefore, this measure should capture the reputational element of corporate activities, as well as the response of stakeholder groups to the firm's varied activities. This would not seem possible with more traditional measures of performance, such as return on investment (ROI) or return on equity (ROE).

Analytical Methods.

For the statistical analyses we employed a general linear modeling approach. Industry effect was controlled by introducing industry code as a fixed factor in Analysis of Covariance Models (ANOCOVA). MNC Activities, MVA, and Firm Size were introduced as covariates in the respective analyses. As reported earlier, we developed a two-wave database (with necessary variables in both 1996 and 1991). This database design enabled us to probe the nature of the relationship between MNC activities and CSP from three different angels. First we asked, cross-sectionally do greater MNC activities relate to greater CSP? Second, we asked, does greater prior MNC activities lead to greater CSP? Finally we asked, would the change in MNC activities over time lead to greater CSP? The first relationship was assessed by two linear models (Models 1 and 2) with independent variables from the same period respectively. The second relationship was assessed by a linear model (Model 3) with independent variables from the period prior to the measure of the dependent variable. The third relationship was assessed by introducing the changes in MNC activities as the central independent variable (Model 4).

Table 1 reports both descriptive statistics and correlation coefficients.

a N=350.

RESULTS

In linear modeling, it is important to have a reasonable data set size in order to avoid over-fitting data, and to assess the possible presence of multicollinearity (Hair, Anderson, Tatham & Black, 1995). As reported earlier, the final data set contains 350 firms, with three independent variables, thus the data set was reasonably large for linear model analysis (Hair et al., 1995). We also conducted an assessment of multicollinearity using the conventional procedures of coefficient variance decomposition analysis with condition indices (SPSS Win, 7.5), and found that the modeling variables employed in this study did not present a threat of multicollinearity.

The dependent variables (CSP1996 and CSP1991) were standardized using a z-transformation to reduce the possible confounding impact of size differences.

Three paired t-tests present some interesting preliminary look at the study sample. First of all, the mean of CSP 1996 is statistically significantly greater than that of CSP 1991 (t = 1.855, p < 0.05 single tailed), indicating that for firms in our data set there was a significant increase in corporate social performance. Secondly, the mean of MVA 1996 is statistically significantly greater than that of MVA 1991 (t = 6.618; p < 0.000), indicating that these firms increased their economic performance across time. Finally the mean of MNC activities 1996 is statistically greater than that of MNC activities 1991 (t = 5.710, p < 0.000), indicating an increase in the degree of international involvement.

Table 2 presents two linear models (ANOCOVA) that probed how firms’ MNC activities influenced CSP on a cross-sectional basis. It is clear that the fixed factor, industry, accounts for a portion of the variance explained. Furthermore, for each time period, the economic performance as measured by MVA exhibits a statistically significant positive impact on a firm’s CSP. This supports other studies which have found a positive relationship between economic performance and CSP (e.g., Waddock and Graves, 1997). MNC activities on the other hand do not produce any statistically significant impact on CSP. These results lead us to conclude that cross-sectionally firms greater in MNC activities do not necessarily have greater CSP. Finally, firms’ size does not appear to have any statistically significant impact on firm CSP.

Having obtained some evidence regarding the first type of MNC activities and CSP relationship, we now turn to the next two questions.

Table 3 presents two further models. Model 3 assesses how prior MNC activities will influence CSP. Model 4 focuses on how the changes on MNC activities over time will influence firm performance. Note that both models identified the impact of industry as significant. In addition, the earlier economic performance seems to exhibit a statistically significantly positive impact on CSP. The firm size appears to have a statistically significant negative impact on CSP. Furthermore, Model 3 reveals that the prior level of MNC activities produces a statistically significant positive impact on firm CSP level. Model 4, designed to answer if the change in MNC activities over time will influence CSP, suggests that it does not.

While the paper began with a single hypothesis, it moved to answer three questions related to two explanations of increased social awareness among multinational firms. In this process we identified two important forces at play. The first is the "push" from the external environment. Clearly when governments, international bodies and various special interest groups are closely monitoring the activities of multinational firms, they have little alternative but to respond. These laws, regulations and social pressures are particularly important for they provide a leveling field for firms of all nations competing in the global economy.

On the other hand, firms are of their own volition beginning to heed calls for greater social responsibility. As the literature suggests more firms are beginning to integrate their social responsibilities worldwide into their strategic orientations. However, it must be noted here that our study suggests that firms are exploring strategic issues, developing responses and implementing them as they expand their international presence.

Our study suggests that firms are indeed developing strategies to cope with social issues. We believe that it takes time to scan and monitor the environment, to develop meaningful responses, and finally to implement them successfully. This process is reflected in how the prior multinational activities affect corporate social performance later. The results of this study suggest that as more firms are exposed to international involvement, and become a part of the international competitive environment, they will also adopt more socially responsible practices.

Obviously this study focused on U.S. based firms only. As a first step this is important as we were able to remove other confounding factors, such as differences in national culture (Hofstede, 1980, 1991) and patterns of multinational evolution (Bartlett and Ghoshal, 1989), from entering into the study. Future studies, however, should expand this line of inquiry to include firms from other countries.

These findings suggest other opportunities for further research. While there is an increase in the international awareness of the importance of responsible behavior, there is also pressure for improved corporate economic performance. We see within the U.S. an increase in the level of investor activism to improve bottom line performance. How will this affect the social issues movement? There is the possibility that institutional investor activism operates counter to social issues activism either nationally or internationally.

Another question of interest would be how CEO characteristics impact the international social performance of firms. Do firms that employ foreign nationals have a better record on social issues than those which take a more ethnocentric approach to hiring and promoting senior management? It would be hoped that this study points the way to expanded research in these areas.