Corporate social responsibility in emerging and developing economies in Central and Eastern Europe – a measurement model from the stakeholder theory perspective

Abstract

Based on the idea that the development of complex relationships with stakeholders generates significant benefits for organisations, this article aims to create an original model for assessing corporate social responsibility from the stakeholder theory perspective by building a Stakeholder Satisfaction National Index. Our model identified eight stakeholder groups, defined their expectations and suggested 26 indicators to assess the extent to which these expectations are met by organisations in emerging and developing economies in Central and Eastern Europe. A linear regression model for panel data with fixed/random effects was employed in order to get statistically significant results while investigating the correlation between categories of stakeholders. Results revealed that there is a certain correlation between the National Index of Stakeholder Satisfaction and the Index of Human Development and also allowed a two-dimensional categorisation of emerging and developing economies in Central and Eastern Europe, highlighting their potential to align the global requirements of corporate social responsibility and sustainable development.

Keywords:

• stakeholders
• stakeholder theory
• corporate social responsibility (CSR)
• Stakeholder Satisfaction National Index (SSNI)
• emerging and developing economies
• Central and Eastern Europe

JEL classification:

• M14
• C43
• O52
• O53

1. Introduction

Traditional views on organisations put stakeholder expectations and interests, materialised by managers who must generate added value for them, at the core of an organisation’s relationship with its environment (Baleanu, Chelcea, & Stancu, 2011). Considerations that organisations carry out legal activities, develop complex relationships with a number of interested partners and introduce products or services to consumers, led to the development of an important theory: stakeholder theory.

As the creator of this theory, Edward Freeman (1984) replaced the principle according to which managers are accountable to shareholders for an innovative concept, and stated that managers develop relationships based on trust with a variety of stakeholders that are directly affected by the actions of the organisations they manage, while shareholders cease to be regarded as the sole beneficiaries of corporate activities. Putting this theory into practice generates two types of effects (Stieb, 2009, p. 405): benefits are redistributed among several categories of beneficiaries, while organisational decision makers are reconsidered and all stakeholders are invested with decision-making power.

For Edward Freeman’s followers, the survival and success of an organisation depend on its ability to generate satisfaction for all its stakeholders, not just shareholders (Clarkson, 1988), and it being possible to evaluate corporate social responsibility by identifying stakeholders and by analysing corporate responsibility toward stakeholder groups (Hopkins, 2004).

But there is also research that shows that corporate social responsibility is unjustified, as it has a neutral impact on financial performance (McWilliams & Siegel, 2000) or even a negative impact (Aupperle, Carroll, & Hatfield, 1985; Wright & Ferris, 1997). Even in this context, McWilliams and Siegel (2000) showed that different stakeholders encourage companies to invest in social responsibility initiatives.

Our belief is that stakeholders have an important role in corporate social responsibility activities and act as an instrument used to operationalise the theory of extended corporate responsibility and as support for the organisation’s CSR strategy. We believe that determining the extent to which the organisation meets the expectations of its stakeholders can help in assessing the performance of CSR actions. Using this belief as a starting point, this paper aims to develop a methodology for measuring the CSR performance of emerging and developing economies in Central and Eastern Europe, from the perspective of the stakeholder theory.

This research focuses on presenting the various stages in the development of the stakeholder theory in the Literature review section and establishes the intimate connection between the stakeholder theory and CSR actions. The paper then continues by describing the methodology and research objectives. It suggests a conceptual model to assess corporate social responsibility from the stakeholder theory perspective and describes the methodology used for determining the Stakeholder Satisfaction National Index (SSNI). Finally, emerging and developing economies in Central and Eastern Europe are ranked based on SSNI, assumptions are validated and conclusions are drawn.

2. Literature review

2.1. Stages in the development of stakeholder theory

In scholarly literature, there are many attempts to conceptualise and identify stakeholders. These attempts fall into two types of perceptions on the stakeholder theory: the narrow vision and the broad vision.

According to the narrow perception, stakeholders have been defined based on their ability to support or limit the results and activities of organisations and were seen as groups that are vital to the survival and success of organisations (Freeman & Reed, 1983) or as ‘people or groups that are important for the survival and success of organisations’ (Dessler, 2004, p. 42). Thus, Freeman (1984) believed that capital owners, suppliers, employees, customers, local communities and managers were all stakeholders.

According to the broad perception, organisations became aware of their impact on stakeholders, which were thought to be groups, or individuals, who benefit from or are harmed by corporate actions, whose rights are violated or, on the contrary, respected (Freeman, 1998), or individuals, groups and organisations directly or indirectly affected by corporate actions, objectives and policies (Miron, Petcu, & Sobolevschi, 2011). Clarkson (1995, p. 106) suggests a distinction between primary stakeholders (shareholders and investors, employees, customers and suppliers, governments, and communities) and secondary ones (entities affected by the actions of the company, but not involved in direct transactions with it).

In the last few decades, developments of stakeholder theory were made in the context of acknowledging stakeholder–organisation interrelations. In this context, Donaldson and Preston (1995) define stakeholders as individuals or groups of people with legitimate interests in procedural and/or substantive issues of corporate activity whether or not the corporation has any corresponding functional interest in them. Donaldson and Preston (1995) introduced the idea that stakeholders have inherent values, as each group is important by itself and not because of its ability to directly or indirectly promote and support the interests of the company or of another group.

Furthermore, Mitchell, Agle, and Wood (1997) have offered identification and salience to the stakeholder theory. The theory suggests that power, legitimacy and urgency are the three attributes that affect the degree to which company managers give priority to competing stakeholder claims (Matilainen, 2012). Power refers to the stakeholder’s ability to influence the company. Legitimacy means that stakeholders’ claims are accepted or expected in a given society, and urgency refers to the degree to which stakeholder claims call for immediate attention and to the importance of the claim or of the relationship with the stakeholder. The more stakeholders possess these three attributes (power, legitimacy and urgency), the higher is the importance company management grants them. A stakeholder may have a legitimate claim on the company, but it will not be deemed important by company management if the stakeholder does not have the power to enforce its claim or the claim is not perceived as urgent (Mitchell et al., 1997).

Post, Preston, and Sachs (2002) restored balance to the relationship of an organisation with its stakeholders, seen as individuals and groups who contribute voluntarily or involuntarily to building capacity and corporate activities that generate wealth, and who represent potential beneficiaries or, on the contrary, risk bearers. Organisations are expected to responsibly manage an extended web of stakeholder interests across increasingly permeable organisational boundaries and acknowledge a duty of care towards traditional interest groups, as well as silent stakeholders – such as local communities and the environment (Simmons, 2004). Acceptance that a wider range of stakeholders has legitimate expectations has resulted in proposals to align profit-centred and social responsibility models of corporate governance (Waring, 2008), and to balance shareholder value creation with stakeholder value protection (Law, 2011).

The most comprehensive vision on the stakeholder theory belongs to Starik, according to whom ‘interested parties are any natural entity that affects or is affected by the activities of the organisation’ (Starik, 1995, p. 216) – an approach criticised by some researchers or that is even considered absurd (Orts & Strudler, 2009, p. 607), as it doesn’t allow precise identification of legitimate stakeholders.

In the present research we used Freeman’s (1984, p. 25) vision, viewing stakeholders as ‘any group or individual who can affect or is affected by the achievement of corporate objectives’. Our study focuses on eight stakeholder groups: investors, employees, customers, suppliers, creditors, government, natural environment and local community. We chose these eight categories of stakeholders for the following reasons: (a) a large body of research has identified them as the most influential stakeholders in organisational activity (Agle, Mitchell, & Sonnenfeld, 1999; Clarkson, 1995; Donaldson & Preston, 1995; Jenkins, 2006; Maignan & Ferrell, 2000; Maignan & Ralston, 2002; Weaver, Trevino, & Cochran, 1999; Williamson, Lynch-Wood, & Ramsay, 2006); (b) although the importance of these stakeholders has been established and their relationship with organisations was demonstrated in research conducted in developed countries or even in emerging or developing ones (Fryxell & Lo, 2003, McCarthy & Puffer, 2008; Turker, 2009, Tang & Tang, 2012; Matilainen, 2012), for Central and Eastern Europe these issues have not been sufficiently addressed (Miron et al., 2011) .

2.2. CSR from the perspective of stakeholder theory

Based on stakeholder theory, we believe that corporate social responsibility is the financial or non-financial, direct or indirect degree of empathy shown by an organisation in relation to stakeholders during the organisation’s activity. From this perspective, we believe that identifying stakeholders and identifying their commitment to corporate social responsibility are fundamental.

The first relevant connection between corporate social responsibility and stakeholder theory belongs to Archie Carroll (1991, p. 43) who personalised corporate social responsibility by specifying which groups or individuals organisations are accountable to and should show responsiveness to. Carroll (2004) introduced the concept of ‘stakeholder’ in its quadripartite model of corporate social responsibility, considering that organisations do what is asked of them by global capitalism, what is necessary, what is expected and what is desired by stakeholders.

Social responsibility can be seen as a kind of management that takes into account the economic and social effects of its decisions (Boone & Kurtz, 1992, p. 38) and as a specific way of making decisions according to specific standards in order to have favourable effects on shareholders (Puiu, 2007, p. 88).

For Michael Hopkins (2004), corporate social responsibility involves treating stakeholders in a responsible or ethical manner, while maintaining corporate profitability, and requires a specific attitude of the company towards society, in which success is achieved by observing the law, by behaving in an ethical fashion, by focusing on the environment and by taking into account the needs and interests of all partners (Oprea, 2005, p. 47), reflecting the extent to which organisations should use their resources to increase the welfare of one or more segments of society, other than investors (Dessler, 2004, p. 41).

The European Commission (Commission of the European Communities, 2001) states that companies should, on a voluntary basis, use corporate social responsibility to integrate social and environmental priorities and relationships with stakeholders into their business operations, because being socially responsible means not only complying with legal requirements, but also going beyond them.

ISO 26000 (2010), a recent global initiative, advocates the accountability of organisations for the impact of their decisions and activities on society and the environment, manifested through a transparent and ethical behaviour that contributes to sustainable development, including health and welfare of society, takes stakeholder expectations into account, obeys laws, is compliant with international behaviour norms, and is integrated throughout the organisation and into its relationships.

Nowadays, CSR is no longer about individual projects or programmes, but rather about how the totality of business activities impacts on stakeholders such as customers, suppliers, employees, communities, government and the environment (Rake & Grayson, 2009). Stakeholder theory facilitates a heightened awareness of CSR, business ethics and business practices that enable more informed decisions on stakeholder salience (Fassin, 2010) and more robust CSR evaluations (Fassin and Buelens, 2011).

In order to build a CSR reputation and stakeholders’ trust, organisations must demonstrate genuine concern and evidence of long-term enhancement of CSR and also inform stakeholders about the social influence of their daily operations on the environment (Peterlin, Dimovski, Uhan, & Penger, 2011, p. 139).

Thus, stakeholder theory becomes a necessary process in operationalising corporate social responsibility (Matten, Crane, & Chapple, 2003, p. 111), through which managers in an organisation think about and discuss relations with the stakeholders and their roles in relation to the common good (Basu & Palazzo, 2008). Businesses will align their values and behaviour with the needs and expectations of their social partners by implementing this process on a large scale (Borza, 2011).

Measuring stakeholders’ satisfaction should focus on determining the social impact of corporate social responsibility actions and on the creation of added value (creating shared value), which can be seen as a long-term investment in future corporate competitiveness (Porter & Kramer, 2006).

Creating shared value should supersede corporate social responsibility in guiding the investments of companies in their communities. CSR programmes focus mostly on reputation and have only a limited connection to the business, making them hard to justify and maintain in the long run. (Porter & Kramer, 2011)

Scholarly literature highlights a number of studies that suggest methodologies for assessing corporate social responsibility. Thus, Turker (2009) identified five categories of methods for measuring CSR: reputation indices or databases; single and multiple issue indicators; content analysis of corporate reports; scales measuring CSR at individual level and scales measuring CSR at organisational level. The fifth and most relevant category tries to overcome this limitation and to measure CSR at organisational level. One of the most widely known methodologies applied at organisational level was proposed by Maignan and Ferrell (2000) and is based on the concept of corporate citizenship, defined as the economical, legal, ethical and discretionary responsibilities imposed on the company by its stakeholders. The weak point of this model derives from the reduced number of identified stakeholders. Turker (2009) however eliminated this shortcoming, taking into account the perspective of a larger number of stakeholders, such as employees, customers, government, competitors, the natural environment, future generations and NGOs, but did not consider the shareholders’ point of view, thus excluding the economic dimension of CSR.

We believe that stakeholder theory justifies corporate social responsibility actions, while the efficiency of CSR actions can be assessed based on the degree of satisfaction felt by stakeholders.

Thus, we recommend a model to analyse the expectations of major stakeholder groups and assess the extent to which organisations in emerging and developing economies in Central and Eastern Europe meet them. This model eliminates the drawback of Maignan and Ferrell’s model by considering a relatively large number of stakeholders and builds on Turker’s model by taking into account the interests of shareholders. Our model is the basis for creating a methodology to calculate a national index of corporate social responsibility based on stakeholder theory and a ranking of these countries depending on the index values.

3. Methodology and research objectives

The purpose of this study is to assess the performance of emerging and developing economies in Central and Eastern Europe using the determinants of corporate responsibility towards stakeholders and to define a national index of corporate social responsibility based on stakeholder theory. Thus, we formulated the following objectives.

1. Identify the groups of stakeholders, their expectations and justify these choices.

2. Develop a methodology to calculate the national index of corporate social responsibility based on stakeholder theory for emerging and developing economies in Central and Eastern Europe.

3. Rank emerging and developing economies in Central and Eastern Europe depending on the values of the national index of corporate social responsibility based on stakeholder theory.

4. Identify the degree of association between the index of human development (HDI) and the national index of corporate social responsibility based on stakeholder theory through the Pearson correlation coefficient.

Study hypotheses are:

• Hypothesis 1. There is a positive correlation between categories of stakeholders.

• Hypothesis 2. There is a relationship between the index of human development (HDI) and the national index of corporate social responsibility based on stakeholder theory.

3.1. A conceptual model of assessment of corporate social responsibility based on stakeholder theory

Although, in theory, all stakeholders are equally important to any organisation, corporations tend to rank them based on their ability to understand and meet their expectations and also on the power, legitimacy and urgency of these expectations (Jamali, 2008). With regards to this view, we believe that every organisation should consider the following stakeholder categories (Figure 1).

Figure 1. Categories of stakeholders.

Source: Authors’ own creation.

Investors own the business capital and the right to receive the full benefits of capitalisation. Employees seek functional, economic, psychological and ethical benefits from their employing organisations (Simmons, 2004). Functional benefits are obtained if employment provides challenging, stimulating and fulfilling work, economic benefits derive from competitive compensation, psychological benefits accrue from employee involvement in a valued work role, and ethical benefits are anticipated from the equitable treatment that employees hope to experience. Meeting customers’ expectations stimulates their loyalty and attracts new customers. Suppliers, interpreted in a stakeholder sense, are vital to the success of the firm, as raw materials will determine the quality and price of the final product. In turn, the organisation is a customer of the supplier and is therefore vital to its success and survival (Freeman, 1998).

Creditors invest in the organisation’s operations and consequently take on an economic risk, which is why they hold direct, legitimate, economic interests regarding the results generated by corporate operations (Orts & Strudler, 2009, p. 607). The state sets the rules and requirements designed to encourage thriving economic activity that generates benefits for all partners involved, and organisations are expected to comply with legal requirements and with those imposed by the sustainable development and competitiveness of the national economy in a global business environment.

The degradation of the natural environment as a result of the aggressive manner in which organisations have used it during wild capitalism favoured the emergence of an environmental movement and the implementation of environmental legislation. Social investments made by organisations to support local communities generate social outcomes, such as improved overall social welfare and higher corporate economic results.

3.2. Defining the SSNI

Based on the vision outlined above, we identified the expectations of each stakeholder category and assessed how these expectations were met by using a national index of corporate social responsibility based on stakeholder theory (Stakeholder Satisfaction National Index – SSNI), which explains the focus of companies from analysed states on corporate social responsibility activities, based on stakeholder theory (Table 1).

Table 1. Indicators for evaluating various expectations of stakeholders.

Stakeholder category	Stakeholder expectations	Indicators for assessing how stakeholder expectations were met
Investors (Sub-index 1)	Maximise benefits Business development •Investors’ rights	Effectiveness of corporate governance Confidence in the management’s professionalism Investor protection Protection of the interests of minority investors
Employees (Sub-index 2)	Job security Workplace health and safety Fair salary / benefits Professional development opportunities Authority	Hiring and firing practices Linking wage to labour productivity Cooperation in the employee–employer relationship The scale of investment in staff development
Consumers (Sub-index 3)	Variety, quality and safety of products Complete and accurate information Just prices	Customer focus Consumer sophistication Extent of marketing Intensity of local competition
Suppliers (Sub-index 4)	Partnership with suppliers Selection and analysis of supplier	Index of responsible management of logistics channels
Creditors (Sub-index 5)	Recover funds Collect interests and commissions	Legal protection of rights of providers and beneficiaries of financial services
State (Sub-index 6)	Tax collection Fairness in relations with the state Social inclusion Highly competitive business environment Business contribution to the development of the national economy	Development of clusters Participation of women in the labour market Employment rate Use of new technologies in companies The presence of illegal payments and bribes in business Nature of competitive advantage
Natural environment (Sub-index 7)	Control of air, water and soil pollution Preservation of energy and natural resources	Index of environmental performance
Local community(Sub-index 8)	Urban development Support for education, arts and culture Ethical behaviour Human development	Overall quality of infrastructure Quality of education system Average life expectancy Ethical behaviour of organisations Gross domestic product

Source: Authors’ own creation.

The methodology used to determine the SSNI is based on content analysis of the 2009–2012 Global Competitiveness Report published by the World Economic Forum (World Economic Forum, 2009; World Economic Forum, 2010; World Economic Forum, 2011; World Economic Forum, 2012), the 2009–2012 Environmental Performance Index (Emerson et al., 2012) and statistical reports of the World Bank (World Bank, 2013). We aimed to analyse the performance of emerging and developing economies in Central and Eastern Europe because they are marked by interesting economic developments, while corporate social responsibility has gained global meaning and is integrated into the competitive strategies of developed countries.

In calculating the SSNI, we initially established a list of 14 emerging and developing markets in Central and Eastern Europe (International Monetary Fund, 2013): Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Hungary, Latvia, Lithuania, Macedonia, Romania, Poland, Serbia and Turkey. Montenegro and Kosovo were excluded from the list due to lack of data for some of the analysed indicators. A database was put together by documenting the indicators that are part of each sub-index assigned to a certain group of stakeholders. The values of most of these indicators were taken from the Global Competitiveness Reports published by the World Economic Forum and are expressed as scores on a scale from 1 to 7. We used the index of responsible management of logistic channels to assess how suppliers’ expectations are met (Gănescu, Asandei, Gangone, & Chirilă, 2013), and the environmental performance index to highlight the contribution of the analysed states to preserving the integrity of the natural environment (Emerson et al., 2012).

Because certain sub-indexes contain variables expressed using other measurement units, we unified and transformed data using the min-max method to preserve order and relative distance between the scores of different countries. Normalisation was achieved by applying the following formula:(1)

where X_i is the value of indicator to be normalised, val_max denotes best values, and val_min worst value.

Thirdly, we determined sub-indexes for each category of stakeholders. To determine the value of the SSNI composite index, we used an arithmetic average of the eight sub-indexes using the following formula:(2)

where: SSNI is the value of the Stakeholder Satisfaction National Index; I₁, I₂, ..., I₈ are sub-indexes for each category of stakeholders.

Based on these results, we ranked emerging and developing economies in Central and Eastern Europe depending on the capacity of their organisations to meet the expectations of the eight categories of stakeholders, the state with the highest SSNI value being the one with a higher degree of performance in terms of corporate social responsibility based on stakeholder theory (Table 2).

Table 2. Top emerging and developing economies in Central and Eastern Europe according to corporate social responsibility performance based on stakeholder theory.

State/Acronym	Year	Investors Subindex	Employees Subindex	Consumers Subindex	Suppliers Subindex	Creditors Subindex	Government Subindex	Environment Subindex	Local Community Subindex	SSNI
Latvia	2012	0.74	0.79	0.65	0.90	1	0.73	1	0.54	0.79
	2011	0.64	0.70	0.63	0.88	1	0.76	1	0.49	0.77
	2010	0.61	0.70	0.57	0.68	1	0.64	1	0.60	0.73
	2009	0.70	0.71	0.53	0.67	1	0.70	1	0.54	0.73
Albania	2012	0.91	0.95	0.70	0.73	0.83	0.41	0.88	0.83	0.78
	2011	0.96	1	0.80	0.99	1	0.53	0.88	0.91	0.88
	2010	0.81	0.94	0.60	0.68	1	0.47	0.88	0.89	0.79
	2009	0.72	0.87	0.29	0.44	1	0.47	0.90	0.48	0.65
Poland	2012	0.72	0.57	0.80	0.91	0.83	0.69	0.82	0.74	0.76
	2011	0.73	0.48	0.80	1	1	0.68	0.82	0.64	0.77
	2010	0.82	0.57	0.86	1	1	0.72	0.82	0.77	0.82
	2009	0.8	0.56	0.92	1	0.80	0.70	0.82	0.75	0.79
Lithuania	2012	0.83	0.64	0.78	0.98	0.17	0.77	0.87	0.69	0.71
	2011	0.73	0.55	0.74	0.97	0.20	0.76	0.87	0.57	0.68
	2010	0.76	0.68	0.71	0.94	0.20	0.82	0.87	0.64	0.7
	2009	0.73	0.79	0.70	0.92	0.20	0.80	0.87	0.58	0.7
Turkey	2012	0.71	0.65	0.96	1	0	0.59	0.32	0.59	0.60
	2011	0.59	0.42	0.88	0.97	0	0.53	0.32	0.45	0.52
	2010	0.55	0.49	0.85	0.84	0	0.53	0.32	0.46	0.51
	2009	0.46	0.52	0.71	0.75	0	0.55	0.31	0.56	0.48
Hungary	2012	0.50	0.49	0.57	0.63	0.50	0.65	0.65	0.46	0.56
	2011	0.62	0.53	0.66	0.67	0.60	0.64	0.65	0.46	0.6
	2010	0.58	0.64	0.62	0.62	0.60	0.59	0.65	0.61	0.61
	2009	0.66	0.58	0.50	0.61	0.60	0.61	0.65	0.55	0.59
Bulgaria	2012	0.44	0.55	0.49	0.50	0.67	0.54	0.62	0.19	0.50
	2011	0.44	0.46	0.50	0.51	0.80	0.45	0.62	0.23	0.5
	2010	0.40	0.43	0.44	0.38	0.80	0.40	0.62	0.37	0.48
	2009	0.44	0.40	0.42	0.43	0.80	0.50	0.62	0.37	0.5
Croatia	2012	0.27	0.10	0.38	0.53	0.33	0.62	0.83	0.60	0.46
	2011	0.19	0.15	0.39	0.57	0.40	0.62	0.83	0.62	0.47
	2010	0.21	0.14	0.35	0.48	0.40	0.52	0.84	0.71	0.46
	2009	0.37	0.28	0.43	0.60	0.40	0.56	0.84	0.77	0.53
Romania	2012	0.40	0.21	0.41	0.33	0.83	0.49	0.41	0.05	0.39
	2011	0.57	0.38	0.49	0.50	0.80	0.47	0.41	0.18	0.47
	2010	0.70	0.53	0.58	0.56	0.80	0.53	0.41	0.37	0.56
	2009	0.76	0.61	0.58	0.64	0.80	0.55	0.41	0.38	0.59
Macedonia	2012	0.47	0.40	0.29	0.26	0.50	0.35	0.37	0.38	0.38
	2011	0.44	0.52	0.40	0.39	0.60	0.45	0.37	0.44	0.45
	2010	0.48	0.60	0.41	0.45	0.60	0.41	0.38	0.70	0.5
	2009	0.40	0.52	0.26	0.43	0.60	0.31	0.37	0.58	0.43
Bosnia and Herzegovina	2012	0.52	0.47	0.32	0.25	0.17	0.26	0.10	0.15	0.28
	2011	0.31	0.26	0.37	0.18	0.20	0.40	0.10	0.28	0.26
	2010	0.16	0.36	0.17	0	0.20	0.28	0.10	0.17	0.18
	2009	0.08	0.48	0	0	0.20	0.22	0.13	0.20	0.16
Serbia	2012	0.10	0.09	0	0	0.67	0.20	0.35	0.09	0.19
	2011	0.10	0.14	0	0	0.80	0.14	0.35	0.16	0.21
	2010	0.13	0.23	0.01	0.07	0.80	0.21	0.35	0.33	0.27
	2009	0.30	0.37	0.13	0.26	0.60	0.28	0.35	0.41	0.34

Source: Authors’ own creation.

The SSNI produced some interesting distributions of scores of analysed countries. National differences in the performance of CSR activities can be seen as differences in determining the indicators included in the index. Latvia and Albania reached a balance between stakeholder expectations and the extent to which these expectations are met. The lower end of the ranking includes countries such as Macedonia, Bosnia and Herzegovina or Serbia, which have a low level of stakeholder satisfaction. These results are due to poor awareness and manifestation of stakeholder expectations, lack of foreign capital, reduced activity of NGOs in empowering businesses and low involvement of communities in social responsibility actions.

3.3. Testing the study hypothesis

To test our study hypothesis we used the panel data method because the sample has a cross-sectional dimension represented by countries (i = 1,…,12) and a longitudinal dimension represented by a time series (t = 1,…,4) (Gujarati, 2004, p. 636). The sample comprises balanced panel data since there are no time periods missing from the units included in the population of interest. We used data analysis techniques such as the panel data model and Eviews software to perform a statistical analysis of the data.

To test Hypothesis 1, we used the random effects (REM) and the fixed effects (FEM) estimation methods, which allow us to deal with the problem of unobserved heterogeneity. While the FEM capture country-specific effects with α_i, which do not change over time, the REM incorporates heterogeneity among the countries by including a specific unobservable effect u_it in the error term. All models were tested using the Hausman test to decide between REM and FEM and standard errors that are robust to heteroscedasticity and autocorrelation (Kersan-Skabic, 2013). The use of these methods was justified as it ensures statistically superior results compared with the Pearson correlation method.

Data processing began with the introduction of eight sub-indexes of SSNI values, which were previously calculated with Eviews software, as a data panel consisting of information related to 12 states and four years, which has led to 48 observations.

Next, we generated eight linear regression models with fixed effects and each of the eight sub-indexes was used, in turn, as a dependent variable, while the others were used as independent variables. We then estimated eight linear regression models with random effects. Failure to obtain data for the eight sub-indexes for more than four years required us to select the ‘Random’ option for the cross-section and the ‘None’ option for the Period in the Panel Options tab. The Hausman test was used to select between the linear regression model with fixed effects and the linear regression model with random effects. The null hypothesis is: ‘The random effects regression model is appropriate’ and the alternative hypothesis is: ‘The fixed effects regression model is appropriate’. A statistically significant probability (p < 0.05) justifies the use of the linear regression model with fixed effects, while a probability of p>0.05 requires the use of a linear regression model with random effects. We analysed the coefficients associated with explanatory variables. The t-test value and the corresponding probability show whether the coefficient associated with an explanatory variable is statistically significant and different from 0 and also if the explanatory variable associated with the coefficient can explain the endogenous variable.

The results of the statistical analysis are summarised in Table 3, highlighting the following.

• The Hausman test failed to reject the null hypothesis that the difference in coefficients is not systematic, thus favouring random effects in six of eight regression models.

• The eight regression equations are detailed below. The t-test value and the corresponding probability enabled us to identify exogenous variables, which explain the endogenous variable.

Table 3. Results: linear regression model for panel data.

Endogenous variable	Hausman test probability	Regression model	Exogenous variables with high impact on endogenous variable	t-Statistic	Sub-index Probability according to the regression model	Coefficient	R-squared	F-statistic	Durbin-Watson stat
Investors Sub-index	0.7958	REM	Sub-index Employees	4.4458	0.0001	0.442243	0.805811	23.71209	1.762888
			Sub-index Consumers	2.7163	0.0097	0.493023
Employees Sub-index	0.0247	FEM	Sub-index Investors	2.3003	0.0288	0.402845	0.947784	29.24336	2.257176
			Sub-index Local Community	3.4233	0.0019	0.404642
Consumers Sub-index	0.1192	REM	Sub-index Investors	3.0432	0.0041	0.341831	0.869579	38.09990	2.219581
			Sub-index Employees	−2.285211	0.0.277	−0.205141
			Sub-index Suppliers	5.72165	0.0000	0.591980
			Sub-index Environment	−2.106402	0.0415	−0.218426
Suppliers Sub-index	0.3925	REM	Sub-index Consumers	5.688189	0.0000	0.724154	0.904979	54.42303	2.120014
Creditors Sub-index	0.1555	REM	–	–	–	–	0.151284	1.018575	1.656506
Government Sub-index	0.7122	REM	Sub-index Environment	3.768049	0.0005	0.376245	0.641348	10.21839	1.893670
Environment Sub-index	0.0000	FEM	–	–	–	–	0.999792	7730.952	2.699004
Local Community Sub-index	0.3106	REM	Sub-index Employees	2.697853	0.0102	0.468174	0.494130	5.581672	1.907783

Source: Authors' own creation.

(3)

The Employees Sub-index and the Consumers Sub-index explain the Investors Sub-index. Thus, an increase by one unit of the Employees Sub-index and of the Consumers Sub-index leads to an increase of 0.442243, and 0.493023 respectively of the Investors Sub-index.(4)

The Investors Sub-index and the Local Community Sub-index explain the Employees Sub-index. Thus, an increase by one unit of the Investors Sub-index and of the Local Community Sub-index generates an increase of 0.402845 and 0.404642 respectively of the Employees Sub-index.(5)

The Investors Sub-index, the Employees Sub-index, the Suppliers Sub-index and the Environment Sub-index explain the Consumers Sub-index. Thus, an increase by one unit of the Investors Sub-index and of the Suppliers Sub-index will generate an increase of 0.341831 and 0.591980 respectively of the Consumers Sub-index; the increase by one unit of the Employees Sub-index and of the Environment Sub-index will decrease the Consumers Sub-index by 0.205141 and 0.218426 respectively.(6)

The Consumers Sub-index explains the Suppliers Sub-index. Thus, an increase by one unit of the Consumers Sub-index will help increase the Suppliers Sub-index by 0.724154.(7)

None of the independent variables significantly explain the evolution of the dependent variable.(8)

The Environment Sub-index explains the Government Sub-index(9)

None of the independent variables significantly explain the evolution of the dependent variable.(10)

The Employees Sub-index explains the Local Community Sub-index. Thus, an increase by one unit of the Employees Sub-index increases the Local Community Sub-index by 0.468174.

where: SI=Investors Sub-index, SE = Employees Sub-index, SCo = Consumers Sub-index, SS = Suppliers Sub-index, SCr = Creditors Sub-index, SG = Government Sub-index, SEn = Environment Sub-index, SLC = Local Community Sub-index, i = country, t = year, u = standard error.

• R squared (R²) shows how the variation of the dependent variable is explained by the influence of the seven independent variables. The recorded values of R² for regression equations (3(3) ), (4(4) ), (5(5) ), (6(6) ) and (9(9) ) show that a significant proportion of the variance of the dependent variable is explained by the influence of the other seven independent variables on it. In regression models (8) and (10), the R² shows that only 64.13% and 49.41% of the variance in the dependent variable is explained by the influence of the other seven independent variables. The R² value for regression model (7) shows that only 15.12% of the variance in the dependent variable is explained by the influence of the other seven independent variables on it (see Table 3).

• Fisher statistics calculated for regression models (3), (4), (5), (6), (8), (9) and (10) shown in Table 3 are greater than 2.25 (the critical value). Thus, we reject the idea that models are not valid and conclude that the seven regression models are valid. The Fisher statistic value calculated for regression model (7) is 1.018575 < 2.25 (critical value), confirming the null hypothesis that the regression model is not valid.

• The Durbin-Watson Statistic, which is approximately equal to 2, in every one of the eight regression models, indicates no autocorrelation in the residuals.

Therefore, Hypothesis 1 according to which ‘There is a positive correlation between categories of stakeholders’ is partially validated. Meeting the expectations of a category of stakeholders to a greater extent supports the idea of better meeting the expectations of other categories of stakeholders, enhancing the performance of CSR initiatives. Organisations interested in obtaining superior CSR performance should pay equal attention to meeting the expectations of all stakeholders.

To test the second hypothesis, we performed a two-dimensional grouping of countries with emerging and developing economies in Central and Eastern Europe, using the SSNI and the Human Development Index (HDI), taken from the Human Development Report (Malik, 2013; United Nations Development Programme, 2011). We created a panel with the values of the two indexes for four years (2009–2012).

Using Eviews, we studied the relationship between the two indexes using the statistical correlation method. The correlation coefficient with a value of 0.427393 proves the direct, positive, moderate connection between SSNI and HDI (see Table 4).

Table 4. SSNI and HDI correlation matrix.

	SSNI	HDI
SSNI	1	0.427393
HDI	0.427393	1

Source: Authors' own creation.

The study of the relationship between SSNI and HDI implies causality, namely the existence of a cause before the effect. Unfortunately, statistical testing of causality cannot be carried out because the data series is too short, as some of the data necessary to determine SSNI before 2009 is missing. For example, the ‘Irregular payments and bribes’ used in the evaluation of the Local Community Sub-index is not evaluated in the Global Competitiveness Reports issued by the World Economic Forum before 2009, and Montenegro and Kosovo are not in the list of countries assessed in the Global Competitiveness Reports issued by the World Economic Forum. We believe that more time is required to pass to statistically answer the question regarding the direction of the relationship between SSNI and HDI.

Therefore, Hypothesis 2, according to which there is a relationship between the human development index (HDI) and the national index of corporate social responsibility based on the stakeholder theory, is validated. Applying descriptive statistics (Adams, Khan, Raeside, & White, 2007) we plotted the data and identified four categories of performance in terms of meeting stakeholder expectations.

States were grouped into four performance categories (see Figure 2):

• ‘The competitive ones’ (Performers) are countries in which organisations best meet stakeholder expectations and thereby contribute to human development. Poland, Hungary, Latvia, Lithuania and Bulgaria have the greatest potential to align to global requirements of corporate social responsibility and to achieve a high level of sustainable development.

• ‘The responsible ones’ have above-average values of the SSNI, but also have a below-average level of human development. Albania has a high level of stakeholder satisfaction that supports only to a little extent the process of human development. We believe that this is due to the qualitative nature of the data on which the SSNI is based, as these data emphasise individual perceptions.

• ‘The cautious ones’ are a group of member states that exceed the average level of human development of emerging and developing economies in Central and Eastern Europe, although the degree of stakeholder satisfaction is below average. Romania and Croatia fall into this category as well. As more organisations will integrate the principles of corporate social responsibility in their work, they will be able to better meet stakeholder expectations and, in the long run, will contribute to human development in their communities.

• ‘The late ones’, Serbia, Macedonia and Bosnia and Herzegovina, have low levels of both the SSNI and human development. We believe that one of the factors that explain this distribution is the low level of expression of stakeholder expectations and reduced corporate focus on corporate social responsibility in these countries.

Figure 2. Two-dimensional ranking of emerging and developing economies in Central and Eastern Europe.

Source: Authors’ own creation.

We believe that the methodology used to determine the SSNI and create the two-dimensional ranking of states can also be applied to developed countries, not just to emerging economies.

4. Conclusions

Corporate social responsibility can be one of the most effective tools to operationalise stakeholder theory. Stakeholders provide organisations with critical resources for business objectives, invest value in corporate activities, and have enough power to affect corporate performance. As a consequence of interrelations with organisations, stakeholders would either benefit or lose.

The utility of stakeholder theory in analysing corporate social responsibility is indisputable. The main success of this theory is that it clearly defines potential beneficiaries of corporate social responsibility actions and convincingly identifies which categories are under the social umbrella. Being socially responsible is synonymous with proactively considering how your business affects various stakeholders, regardless of whether they are primary or secondary. From this perspective, corporate social responsibility is important in ensuring long-term commercial advantage, facilitating the process of building the trust of customers and other stakeholders.

This study responds to the challenge of measuring how stakeholder expectations are met. Because of our results, this paper contributes to the development of an index that allows comparisons between the performances of states in terms of corporate social responsibility. The SSNI shows significant differences between countries, for reasons related to the proportion of responsible companies in those host countries.

The results of the regression models developed and tested during our study have shown that greater satisfaction of one stakeholder category supports higher satisfaction of other categories of stakeholders, enhancing the performance of CSR initiatives. Therefore, achieving superior CSR performance is possible while meeting the expectations of all categories of stakeholders.

Statistical analysis shows a relationship between corporate social responsibility performance and human development – a relationship whose direction can be determined only by extending the data series.

Limitations of the study derive from the qualitative nature of most data, due to lack of data before 2009, and from difficulties in identifying relevant information related to corporate social responsibility. As organisations and states improve annual reports on their economic, political and social development, the methodology for calculating this index can be improved by using a broader range of variables. The index will acquire more complex meaning and relevance in substantiating sustainable development strategies.

Acknowledgments

The authors are grateful to the anonymous referees for the very helpful and constructive suggestions on the first draft of this paper.