Corporate Social Responsibility

Hypothesis 16: There is a negative and significant relationship between media accident and the level of CCSR

Hypothesis 17: There is a positive and significant relationship between media accident and the level of LCSR

Hypothesis 18: There is a positive and significant relationship between media accident and the level of ECSR

1.3 Development of hypotheses, step 2

Hypotheses built in the first step of the model aim to show factors that affect the presence of various strategies, which constitute socially responsible actions of the firm, however they do not show the motivation for the engagement in CSR activities. Motivational aspects for corporations regarding responsible investments differ from that of individuals in that sense that these social contributions should still provide a certain tangible benefit, apart from satisfaction of individual preferences. As it was emphasized earlier, involvement in the sphere of CSR does not show improvements in financial performance of the firm. However, that doesn't necessarily mean that CSR activities are proved to be inappropriate means to attain a “win, win” scenario. Possible explanation here lies in the fact that usually companies suffer from a short-term monetary bias, in other words, incentives to boost profits in the short-run exceed that for the long-run, which in turn disadvantageously affects long-term performance. Such short-term bias implies not only possible loss in the long-term performance, but also deteriorations in corporate relationships with stakeholder groups (R. Benabou, J. Tirole, 2009). For example, a company may spare money on the safety and quality of a product, quality of working conditions or on the environmental issues in order to boost short-term profits, but these actions in the long-run may lead to a loss of consumers, difficulties in attracting competent labor and to increased regulations. Thus, it is reasonable to conclude that CSR strategies should be tested in the context of long-run opportunity to raise profit. In other words, CSR should show to influence the level of corporate sustainability. Taking this logic into account, the second step of the model is to be constructed.

Consumer CSR

Socially responsible investments aimed at improving relationships with customers vary in possible strategies and provide different benefits. Consumer CSR in the form of, for example, certification of plants and suppliers may be aimed at signaling the quality of a product, which ensures long-term commitment of customers. This type of CSR may also generate a new group of “eco-consumers”, implementing recycled packaging or promoting waste reduction to the public. Consumer CSR is especially significant for retail companies, who have a direct “dialogue” with their customers. Thus:

Hypothesis 1: There is a positive and significant relationship between level of Consumer CSR and sustainability of a company

Labor CSR

The main function of CSR activities implemented in order to address labor concerns is to attract qualified labor. The ability to retain competent workers directly affects company's effectiveness, thus:

Hypothesis 2: There is a positive and significant relationship between level of Labor CSR and sustainability of a company

Environment CSR

CSR in environment address several issues. First of all, these investments serve as a way to improve relationship with regulatory authorities and may be considered as an insurance against disadvantageous actions and punishments implemented by public and market in case of a negative event. Furthermore, investments in related programs may decrease costs of production and lead to a more efficient usage of resources, thus:

Hypothesis 3: There is a positive and significant relationship between level of Environment CSR and sustainability of a company

2. Design of the study and methodology

2.1 Model construction

This part of the paper is devoted to the presentation of the model that is constructed in order to reveal factors that influence three large sub-groups, which in sum constitute level of CSR activities implemented by the company and show that these sub-groups have a direct impact on the sustainability of a company.

As it was stated earlier, the underlying model is constructed as a two-step model, where the first part is made in order to show the relationships between different firm-level characteristics, macroeconomic conditions and levels of CSR sub-groups.

Regarding firm-level characteristics, it was hypothesized that company's size, age, profitability and industry type have a positive relationship with the levels of Consumer CSR, Labor CSR and Environment CSR. As a macroeconomic condition, country status is used to account for the differences among countries, which influences the level CSR. It is also hypothesized that exposure of companies to media influences levels of CSR sub-groups since higher levels of media exposure increases the visibility of a company by its stakeholder groups. Thus, three following Pool models are constructed:

CCSR_ti= б+в₁ASSETS_ti +в₂AGE_ti+в₃EPS_ti+ в₄INDUSTRY_ti+в₅COUNTRYST_ti+в₆MED_A_ti+е_t

LCSR_ti= б+в₁ASSETS_ti+ в₂AGE_ti+в₃EPS_ti+в₄INDUSTRY_ti +в₅COUNTRYST_ti+в₆MED_A_ti+е_t

ECSR_ti= б+в₁ASSETS_ti+ в₂AGE_ti+в₃EPS_ti+ в₄INDUSTRY_ti+в₅COUNTRYST_ti+в₆MED_A_ti+е_t,

where:

CCSR - level of consumer CSR

LCSR - level of labor CSR

ECSR - level of environment CSR

ASSETS - measure of the size of a company

AGE - refers to the age of a company

EPS - measure of a profitability of a company

INDUSTRY - provides a distinction between types of industries

COUNTRYST - provides a distinction between country statuses

MED_A - a measure of a number of articles devoted to negative events caused by company's operations

The second step of the model is aimed at showing relationship between three sub-groups of CSR and sustainability of a company, hypothesizing that underlying sub-groups influence the sustainability positively, thus, the following Pool model is constructed:

SALES_ti= б+в₁CCSR_ti+в₂LCSR_ti+в₃ECSR+е_t,

where:

SALES - a measure of sustainability of a company

CCSR - level of consumer CSR of a company

LCSR - level of labor CSR of a company

ECSR - level of environment CSR of a company

2.2 Data and sample construction

Since the research in this paper is organized, generally, as a two-step model, data construction is comprehensive and consists of quantitative and qualitative characteristics. The main difficulty in obtaining relevant data was the inability to access a unique database, KLD Research, which addresses social and environmental performance of firms, emphasizing different dimensions and possible CSR sub-categories. Such situation influenced the way in which sample and dependent variable data was measured.

2.2.1 Sample construction

In order to address main targets of this paper, it was needed to obtain data for the maximum period possible. As it was already mentioned, due to the absence of access to a specific database, data on CSR could be obtained only from a specific type of reporting, these are CSR reports or Sustainability reports, which are available on the companies' websites. In order to get the information on the amount of companies that are involved in the CSR strategies for a relevant period of time, GRI database was used, which is a service provided by the Global Reporting Initiative organization. This organization promotes CSR and Sustainability reporting to be as vital as financial reporting and offers reporting standards for different industries and regions, however, it should be noted here that such type of reporting is not obligatory and appears as a voluntary initiative that increases transparency. GRI database consists of general information on the firm and several CSR reports. Despite the existence and growing interest to guidelines provided by this organization, very small amount of companies are engaged in this process from its beginning. Taking this fact into account, it was possible to obtain a sample of 20 companies with annual CSR reports being available for a six-year period (years 2007-2012). Another important sample characteristic is that companies are generally divided into two large categories, with 10 companies in each, where categories refer to the “consumer-oriented” and “production-oriented”. Consumer-oriented category includes such companies as Nestle, Procter and Gamble, J.P. Morgan Chase&Co, Apple etc., whereas production-oriented one consists of Arcellor Mittal, BP, E.ON and other. Both categories include broad range of industries such as food and beverage, computer, financial services, mining, oil and gas and other. The main differences between these two categories is that production-oriented category refers to the heavy equipment industries and that companies within categories showed to have different understanding of CSR risks they face. Such difference is evident from materiality matrixes, which is a tool to access particular fields of CSR that should be addressed. The matrix is plotted in two dimensions, with different issues considered under stakeholders' needs and business performance. Evaluating these matrixes, it could be noted that companies which are allocated to production-oriented category perceive environmental and labor safety as “high, high” level of importance, whereas consumer-oriented companies, on average, state product safety to be of “high, high” level. Additionally, in order to avoid a single country analysis, which is done in many previous papers (R. Gamerschlag, K. Moller, F. Verbeeten, 2010; N. S. Bayoud, M. Kavanagh, G. Slaughter, 2012), companies from U.S.A, Europe and Russia were chosen. Given the novelty and voluntary nature of such reporting, 15% of the sample refers to Russian companies, 40% refer to U.S. and the remaining part is allocated to EU companies, this resulting allocation is consistent with a general view regarding the fact that European countries have leading CSR practices, demonstrating a good quality of reports and better accessibility as compared with U.S. and emerging countries (KPMG, 2011).

2.2.2 Dependent Variables, step 1

In order to measure CSR performance of the firm, coding procedure was used. The concept of CSR was divided into three large categories of influence with many different sub-categories. These categories are: Consumer CSR, Labor CSR and Environment CSR. The former refers to corporate responsible practices that have particular influence on product: product safety, packaging, suppliers monitoring, certification and other, and on a target consumer' view of a company, which is related to a responsible investment in local community. Labor CSR refers to the implementation of responsible actions towards employee, these could be: non-discriminating recruitment processes, improved working conditions, safety issues, education, training and other. The latter category consists of strategy aimed at decreasing environmental impact and improving operational practices in terms of energy usage, waste reduction, recycling and other (Table 1, Appendix). Total CSR activity includes 93 sub-items, with 34 points allocated to Consumer CSR, 27 - to Labor CSR and 32 - to Environment CSR.

Coding procedure involved a detailed analysis of each of 120 Sustainability reports and assignment of 1 point in case if a certain activity was recognized. Such procedure was used for several reasons. First reason is that companies do not provide costs allocated to CSR activities neither in CSR reports, nor in profit and loss accounts, which leads to difficulties in assessing total cost of CSR activities implemented during a year.

The second one refers to the core of the concept itself, that is: since firm's actions can be regarded as socially responsible only in case when they possess “beyond the law” principle, it is impossible to apportion the obligatory part of a certain improvement from a voluntary one, especially if it is incorporated in operating process. These two reasons influenced the way in which the table of total sub-items was created.

Generally, it is conducted in such a way that allows accounting for both the progress effect and the diversity of possible CSR activities. For example, if a company has a waste reduction program, it gets 1 point for the presence of a program and 1 point if it showed to reduce total waste, but if the next year it still implements a program, but didn't show to actually reduce waste for the year, 0 point is allocated.

However, this method alone doesn't address the size effect. The essence is that, for example, two companies may initiate an educational program for its labor, in this case both of the companies get 1 point, but companies may have spent different amounts of money on that, which influences the scale of implemented activity. Thus, in order to account for a scale of CSR activities, final scores for CCSR, LCSR and ECSR were calculated as a proportion of annual expenses of a company, considering the results derived from the coding method (Table 1).

This particular method to calculate levels of CSR sub-groups results in a case, when CSR appears as a static cost parameter, however, coding procedure used to determine the actual involvement of a company in CSR strategies makes it possible to account also for the diversity of underlying activities and for the progress a company makes.

2.2.3 Dependent Variable, step 2

As it was showed in the meta-analysis written by Margolis, Elfenbein and Walsh (2007), there is an extremely small positive correlation between CSR and financial performance of the firm. Yet, more and more companies around the world, irrespective of preferences of managers, involve in CSR activities. As it was emphasized earlier, CSR activities affect different “hidden” sides of a business, strengthening relationships with relevant stakeholder groups, attracting labor and relaxing attention paid by regulatory authorities and public in general. All that boost long-term opportunities to raise profits, leading to sustainability. Level of sales of a company is considered to be a good measure of sustainability, thus, annual data from Compustat Global and Compustat North America was used.

2.2.4 Control Variables, 1^st regression

A different set of firm-level measures was used in order to account for characteristics that influence CSR level and diversity of activities. These measures, collected from Compustat Global and Compustat North America for 20 companies within a given 6-year period, include firm's size, age and profitability, where firm's size is reflected as asset value and profitability is introduced by Earnings per Share ratio. Dummy variable for the type of industry was also included with 0 allocated to consumer-oriented firm and 1 to production-oriented company. Another firm-level characteristic, which is Media Accident variable, was calculated using Factiva database, a source that provides an access to leading periodical publications all around the world. Media accident variable was estimated as a number of articles, published during each year in a given time horizon, that publicize accidents, emergency or injury cases with reference to a given company in the sample. News line provided by a company itself was excluded from the sample in order to account only for the real attention paid to a firm by the public. As it could be expected, values are higher for firms of heavy equipment industries than for those of consumer-oriented ones. Apart from firm-level variables, Country Status factor was also included in order to account for the differences between levels of CSR across developed and emerging countries. Country Status variable is a dummy variable, where 1 indicates that a country is OECD member and 0 otherwise. Membership of a country in the Organization for Economic Co-operation and Development addresses primarily policy implication regarding sustainable development and responsible investments. The aim of organization is to recommend different policies on various topics that influence social well-being, including: finance, competition, bribery, corruption and other. Organization also pays attention to global environmental issues, launching a Green Economy Initiative, which is aimed at providing guidance on country-level reforms in order to achieve sustainable green growth. Belonging of a firm to OECD country may explain higher level of CSR diversity, such as the implementation of LEED certified buildings, regeneration of land and forests etc.

2.2.5 Control Variables, step 2

As the second step of the model is aimed at identifying correlation between Sales and components of total CSR level, control variables are calculated according to the approach used to estimate CSR expenditures. Levels of diversity and completeness of CSR for companies for each year were calculated, using scoring technique. Thus, levels of Consumer CSR, Labor CSR and Environment CSR are:

CCSR_ti= У X_Cti?N_C

LCSR_ti = У X_Lti?N_L

ECSR_ti = У X_Eti?N_E,

where: N_C, N_L, N_E - indicate maximum sub-items possible under Consumer CSR, Labor CSR and Environment CSR sub-categories a firm may address

X_C, X_L, X_E - indicate the amount of sub-item activities actually performed.

3. Results and discussions

3.1 First step of the model

The results of the first step of the model are presented in the following three tables. Since the implemented methodology required the usage of a panel data, three Pool models were constructed. Sorting between possible specifications of a panel data model, such as the fixed-effect model and random-effect model, and using OLS, WLS and GLS, the best models were chosen based on the R² and Durbin-Watson parameters. Taking into account the fact that E-views does not provide the opportunity to test panel data model for the presence of heteroscedasticity, White cross-section weighted method was used in order to prevent estimation bias.

3.1.1 Consumer CSR

The resulting regression for the Consumer CSR is as follows with the results provided in the Table 3:

CCSR = 1.22803751763*AGE + 22.8073389384*EPS - 0.18620226096*MEDIA_A + 766.526924176*COUNTRYST + 0.00131269112442*ASSETS - 443.095361103*INDUSTRY

Variable	Coefficient	Std. Error	t-Statistic	Prob.
AGE_	1.228038	0.314893	3.899854	0.0002
EPS_	22.80734	6.231033	3.660282	0.0004
MEDIA_A_	-0.186202	0.076364	-2.438346	0.0163
COUNTRYST_	766.5269	46.18782	16.59587	0.0000
ASSETS_	0.001313	0.000143	9.178363	0.0000
INDUSTRY_	-443.0954	83.55462	-5.303062	0.0000
	Weighted Statistics
R-squared	0.560509	Mean dependent var	3652.099
Adjusted R-squared	0.541233	S.D. dependent var	5320.887
S.E. of regression	2181.303	Sum squared resid	5.42E+08
Durbin-Watson stat	0.587658
	Unweighted Statistics
R-squared	0.070407	Mean dependent var	1316.159
Sum squared resid	6.90E+08	Durbin-Watson stat	1.419681

As it can be seen from the table, R², which shows the proportion of variations in response variable, explained by the model, is higher for the weighted statistics in comparison to the unweighted statistics, however it still is not very high, which is most probably caused by the relatively small time horizon. Durbin-Watson statistics, examined to identify possible autocorrelation in the residuals, is higher for the unweighted statistics. If to consider the control variables, coefficients for all factors are shown to be significant at all reasonable significance levels, except for the MEDIA_A factor, which is significant only at 5% level. It can be concluded that country status and industry type affect heavily level of Consumer CSR activities, which is generally consistent with the logic emphasized during the development of hypotheses. Dummy variable for the Industry Type is constructed in a way, when 0 is assigned in case if a company is estimated to be consumer-oriented and 1 is assigned if a company is considered to be production-oriented. The following result that coefficient for this variable is very high and negative implies that companies, belonging to heavy equipment industries tend to invest drastically less in Consumer CSR than companies that belong to consumer-oriented industries. The intuition under such result is that production-oriented companies have different set of stakeholder groups in comparison to consumer-oriented companies, since, usually, consumer goods or final goods do not constitute a big proportion of their turnover, thus, such companies do not address much of the Consumer CSR issues. Another dummy variable coefficient - Country status - also shows to be an extremely high, but positive. As 1 was assigned in the case, when company operates in OECD-member country and 0 otherwise, the resulting effect is consistent with hypothesis, which assumes that membership of a country in different organizations, which promote different social and ethical strategies, directly and positively affects CSR levels. Age, EPS and Assets showed to have an intended positive effect, whereas Media Accident variable showed a negative effect on the level of CCSR. Such situation may be explained by the fact that the underlying data, used to measure this variable, was constructed in a way, where only accidents, injuries and fatality cases mentioned in the press for the companies in a sample were used, which is more related to Labor CSR and Environment CSR, since injuries account for a negative events with labor and accidents may account for the environmental distortions.

3.1.2 Labor CSR

The resulting regression for Labor CSR is as follows, with results provided in the Table 4:

LCSR = 0.681641387676*AGE + 24.8092243746*EPS + 0.204192985799*MEDIA_A + 891.690614186*COUNTRYST + 0.000210708673908*ASSETS - 504.149065001*INDUSTRY

Variable	Coefficient	Std. Error	t-Statistic	Prob.
AGE_?	0.681641	0.268367	2.539958	0.0124
EPS_?	24.80922	6.096301	4.069554	0.0001
MEDIA_A_?	0.204193	0.121844	1.675857	0.0965
COUNTRYST_?	891.6906	131.6469	6.773351	0.0000
ASSETS_?	0.000211	3.06E-05	6.888921	0.0000
INDUSTRY_?	-504.1491	57.63637	-8.747066	0.0000
R-squared	0.579273	Mean dependent var	800.3077
Adjusted R-squared	0.543276	S.D. dependent var	1074.115
S.E. of regression	994.1941	Akaike info criterion	16.69045
Sum squared resid	1.13E+08	Schwarz criterion	16.82982
Log likelihood	-995.4269	Hannan-Quinn criter.	16.74705
Durbin-Watson stat	0.995148

R² and Durbin-Watson statistics for the following panel data model are not very high, which again might be connected to sub-optimal time horizon. However, most of the coefficients are significant at all reasonable significance levels, except for the coefficient for AGE, which is significant at 5% level and for the coefficient for MEDIA_A, which is significant at 10% level. It is again possible to observe high and negative coefficient for the industry type and high and positive coefficient for the Country Status.

In case of the country status, the same logic can be implemented as in the case with Consumer CSR. As for the industry type variable, it can be concluded, that production-oriented industries are extremely affected by the governmental constraints and expectations from the public, regarding their environmental activities, leading to decreased incentives to voluntary invest in other sub-groups of CSR. Media accident variable showed to have an intended effect, emphasized by the hypothesis, which is also consistent with an intuition behind its negative value in the Consumer CSR Pool model.

EPS and Assets variables showed consistent results, however Assets variable coefficient is smaller than in case with Consumer CSR, which might be the case, since older companies have tighter relationships with customer stakeholder group, so that CCSR should be addressed in order to maintain these relationships, but in case with Labor CSR, smaller coefficient value may account for the fact that ways in which labor issues are addressed change a little with time.

3.1.3 Environment CSR

The resulting regression for Labor CSR is as follows, with results provided in the Table 5:

ECSR = 1.39428167296*AGE + 3.95280786462*EPS + 0.161782630355*MEDIA_A - 49.8891229593*COUNTRYST + 0.00301234363213*ASSETS + 139.844599738*INDUSTRY

Table 5.

Variable	Coefficient	Std. Error	t-Statistic	Prob.
AGE_?	1.394282	0.637797	2.186089	0.0309
EPS_?	3.952808	1.579623	2.502374	0.0138
MEDIA_A_?	0.161783	0.080162	2.018186	0.0459
COUNTRYST_?	-49.88912	22.47797	-2.219468	0.0284
ASSETS_?	0.003012	0.000128	23.62454	0.0000
INDUSTRY_?	139.8446	26.22611	5.332266	0.0000
	Weighted Statistics
R-squared	0.726601	Mean dependent var	3945.784
Adjusted R-squared	0.714610	S.D. dependent var	4631.824
S.E. of regression	2276.242	Sum squared resid	5.91E+08
Durbin-Watson stat	0.402473
	Unweighted Statistics
R-squared	0.169016	Mean dependent var	1733.806
Sum squared resid	2.10E+09	Durbin-Watson stat	0.626399

R² for this panel data model is higher than for previous ones, however Durbin-Watson statistics is not very high. All coefficients of the underlying variables showed to be significant at 5% level. Here, we observe changes in the sign of the coefficient for the Industry type dummy variable, for the Environmnet CSR model it is high and positive, which perfectly goes in line with intuitions provided previously for its negative sign for Consumer CSR and Labor. So, it might be the situation when managers of production-oriented industries tend to invest heavily in socially responsible activities aimed at improving environmental conditions. Country Status dummy variable also changed its sign from positive to negative. Which may partially be explained by the fact that OECD-membership countries tend to have tougher environmental regulations which leads to a decline in incentives for voluntary environmental enhancement programs. Media accident variable coefficient is also positive, which means that an increased amount of articles devoted to accidents, injuries, fatalities and other negative events leads to an increased investment in Environmental CSR, which suits the proposed hypothesis and logic behind it.

3.2 Second step of the model

Considering the second step of the model development, a fixed-effect panel data model was constructed and the resulting regression is:

SALES=С_i+ 123530.931205 + 5.25538005069*ECSR + 30.7113491663*LCSR +7.45303324343*CCSR

Table 6.

Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	123530.9	6969.842	17.72363	0.0000
ECSR_	5.255380	1.546857	3.397457	0.0010
LCSR_	30.71135	3.673044	8.361280	0.0000
CCSR_	7.453033	2.402389	3.102343	0.0025
	Effects Specification
Cross-section fixed (dummy variables)
R-squared	0.965080	Mean dependent var	147411.9
Adjusted R-squared	0.957160	S.D. dependent var	169053.4
S.E. of regression	34990.42	Akaike info criterion	23.93409
Sum squared resid	1.19E+11	Schwarz criterion	24.46836
Log likelihood	-1413.045	Hannan-Quinn criter.	24.15106
F-statistic	121.8532	Durbin-Watson stat	1.807204
Prob(F-statistic)	0.000000

As it is evident from the Table 6, R² is nearly reaches 1 and Durbin-Watson statistics also show superior results. All variable coefficients are significant at 1% level, means that Consumer CSR, Labor CSR and Environment CSR investments positively affect Sales, which is a measure of sustainability of a company. It is also interesting to analyze the size allocation of coefficients. Investments in Labor CSR showed to have a larger positive effect on the company's sustainability, then go investments in Consumer CSR and then - investments in Environment CSR. Such situation is generally consistent with the logic that investments in labor help to attract and retain competent labor, enhancing its safety, health, education and working conditions, which leads to a superior effectiveness of company's production, while investment in consumer CSR achieve its objectives to maintain relationships with the underlying stakeholder group. Investments in Environment CSR are shown to affect the long-term ability to generate profits, however, its' relatively small coefficient, in comparison to CCSR and LCSR, may be explained by the fact that usually, voluntary investments in ECSR, given the presence of excessive regulations, are considered to serve as an insurance against public pressures in case of a negative events, which was demonstrated by the significant and positive Media accident coefficient in the regression tested to identify factors that influence level of ECSR.

Policy implications

Following the results and findings of this paper, it is possible to introduce an evidence for managers that a theoretical “win, win” scenario is possible. That is, taking into account that shareholders possess monetary preferences, while stakeholders, represented by different groups that can affect the performance of a company, possess social preferences, implementation of various CSR activities may enhance the long-run opportunity to raise profits via different channels, where each channel is determine by the specific characteristics of a particular stakeholder group. Thus, CSR can be viewed not as the way to satisfy particular shareholder's social preferences in the form of charitable contributions, which may lead to a sub-optimal value of the firm, but, conversely, CSR can be implemented as a strategy aimed at attaining sustainability. These conclusions are relevant, since it was shown by the model, that such large sub-groups of CSR as Consumer CSR, Labor CSR and Environment CSR are all proved to influence significantly and positively the sustainability of a company. The only case to be considered here is the appropriate evaluation of relevant stakeholder groups and understanding of the CSR channels, through which a firm may address this group, so that these means could be improved, enhancing the quality of these interactions, which in turn will lead to a long-term financial superiority.

Another possible implication that might be introduced should be addressed to governmental authorities. Taking into account the presence of tough environmental regulations and a trend towards its further implementation and the results of the current model, it might be argued that government need to relax such regulations due to several reasons. First of all, it was showed that country status, introduced as a country membership in OECD organization, influences significantly and negatively the level of Environment CSR, which may imply that excessive weight put on environmental regulations may depress voluntary motives, which leads to a decrease in the Environmental CSR level. Moreover, as Environmental CSR is shown to constitute a significant part of companies' sustainability, excessive regulations may, thus, influence negatively the long-term opportunity to generate profits, distorting general economic conditions.

Conclusion

Theoretical literature written in the field of CSR is aimed at the identifying various channels and strategies through which implementation of CSR activities may enhance market positions of the firm and improve its financial performance. These theories agree upon the importance of identifying and addressing appropriate stakeholder groups. Building relationships with stakeholders, who generally possess social preference, a classical firm may reach a “win, win” condition, which implies maximization of profits, while meeting demand for various social interactions. However, such theoretical framework, when testing in order to reveal relationships between CSR strategies and financial performance of the firm, generally do not provide the evidence for such relationship, meaning that CSR does not show to increase profits of the firm.

Taking all that into account, this paper was developed in order to reveal what motivation may stand behind CSR activities. Basing on the stakeholder theory and a “win, win” scenario, a two-step model was developed, where the first step was devoted to identifying factors that influence levels of CSR sub-groups, each of which represent a channel to address main stakeholders groups, and the second step was devoted to showing that these channels indeed influence sustainability of a firm.

Following this research method, 3 panel data models were constructed with results, showing that companies' size, age, profitability positively affect all three sub-groups of CSR, while other factors such as industry type, country status and media accident revealed different relationships with three CSR dimensions, for example, companies that operate in a heavy equipment type of industry, tend not to invest in the Consumer CSR and Labor CSR, while Environment CSR is positively affected by the industry type. Country membership in OECD organization negatively affects the level of Environment CSR, while Media exposure of a company in case of a negative even, such as accident that affects labor or environmental conditions, have a positive effect on the level of Environmental CSR. The results of the second step fixed-effect model revealed that actually all three sub-groups that constitute total CSR level have a positive effect on the firms' sustainability, showing the motivations to involve in CSR strategies.

Speaking about policy implications, apart from proposing a way to managers to enhance sustainability, an important policy implication lies in a fact that it highly recommended for governmental authorities to relax the increasing implementation of different environmental regulations and let the companies to implement voluntary socially responsible activities towards environment in the ways that benefit stakeholders and society at large, enjoying a “win, win” scenario.

corporate social responsibility manager

References

Academic sources:

K. Mattew Gilley, Dan L. Worrell, Wallace N. Davidson, Abuzar Et-Jelly, 2000, “Corporate Environmental Initiatives and Anticipated Firm Performance: The Differential Effects of Process-Driven Versus Product-Driven Greening Initiatives”, Journal of Management, 26, pp. 1199-1216.

Don Grant, Andrew W. Jones, Mary Nell Trautner, 2004, “Do facilities with Distant Headquarters Pollute More? How Civic Engagement Conditions the Environmental Performance of Absentee Managed Plants”, Social Forces, 83, pp. 189-214.

Wilma Rose, Q. Anton, George Deltas, and Madhu Khanna, 2003, “Incentives for environmental self-regulation and implications for environmental performance”, Journal of Environmental Economics and Management, 48, pp. 632-654.

Robert Innes and Abdoul G. Sam, “Voluntary Pollution Reductions and the Enforcement of Environmental Law: an Empirical Study of the 33/50 Program”????

Charles Eesley and Michael J. Lenox, 2006, “Firm responses to secondary stakeholder action”, Strategic Management Journal, 27, 765-781.

Jacob Brower and Vijay Mahajan, 2011, “Driven to Be Good: A Stakeholder Theory Perspective on the Drivers of Corporate Social Performance”, Journal of Business Ethics

Forest L. Reinhardt, Robert N. Stavins, and Richard H.K. Vietor, 2008, “Corporate social Responsibility Through an Economic Lens”, Review of Environmental Economics and Policy, 2, pp. 219-239.

Stefan Ambec, Philippe Barla, 2002, “A theoretical foundation of the Porter hypothesis”, Economic Letters, 335-360.

Richard D. Morgenstern, William A. Pizer, and Jhih-Shyang Shih, 2002, “Jobs Versus the Environment: An Industry-Level Perspective”, Journal of Environmental Economics and Msnagement, pp. 412-436.

Richard D. Morgenstern, William A. Pizer, and Jhih-Shyang Shih, 2001, “The Cost of Environmental Protection”, The Review of Economics and Statistics, pp. 732-738.

David Pearce and Charles Palmer, 2001, “Public and Private Spending for Environmental Protection: A Cross-Country Policy Analysis, Fiscal Studies, pp. 403-456.

Karen Palmer, Wallace E. Oates, and Paul R. Portney, 1995, “Tightening Environmental Standards: The Benefit-Cost or the No-Cost Paradigm?”, Journal of Economic Perspectives, pp. 119-132.

Carl Pasurka, 2008, “Perspectives on Pollution Abatement and Competitivness: Theory, Data, and Analyses”, Review of Environmental Economics and Policy, 2, pp. 194-218.

Joshua D. Margolis, Hillary Anger Elfenbein, James P. Walsh, 2007, “Does is pay to be good? A Meta-Analysis and redirections of research on the the relationship between corporate social and financial performance”

Paul C. Godfrey, Craig B. Merril, and Jared M. Hansen, 2009, “ The relationship between corporate social responsibility and stakeholder value: an empirical test of the risk management hypothesis”, Strategic Management Journal, 30, pp. 425-445.

Markus Kitzmueller and Jay Shimshack, 2012, “Economic Perspectives on Corporate Social Responsibility”, Journal of Economic Literature, pp. 51-84.

Michael E. Porter and Mark R. Kramer, 2002, “The Competitive Advantage of Corporate Philantropy”, Harward Business Review, pp. 56-68.

Размещено на Allbest.ru