“Quid pro Quo? Corporate Sustainability as a Rational
Marketing Strategy”
Master Thesis Suzanne N. Groen
10003288
MSc Business Economics: Finance
Supervisor: Dr. T. Jochem
Statement of Originality
This document is written by Student S.N. Groen, who declares to take full responsibility for the contents of this document
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This paper analyses the relationship between sustainability performance and financial performance from a firm’s perspective. The main research question concerns whether
sustainable performance and corporate spending on sustainability (CS) can be understood as a rational marketing strategy. Based on data collected from the accounting databases and the Dow Jones Sustainability Index (DJSI) for the period 2005-2014, this thesis first analyses the impact of index inclusions and exclusions on the daily returns of a U.S. company. The abnormal daily returns associated with an event are calculated and their significance are tested. This thesis cannot find statistically significant abnormal returns for companies being added or deleted to the DJSI. However, more striking results are found when the sample is decomposed into industries. There exists some weak evidence for a negative significant impact on stock returns across industries. This can be explained by industry characteristics that play a role in CS (McWilliams and Siegel, 2001). In addition, this paper examines the CS spending behaviour of a company’s industry concerning the distance towards end-consumers using panel estimation techniques. For a sample of U.S. firms in the period of 2005-2012, the results show a highly significant impact of the distance to end-consumer on CS spending. An industry that is closer to the end consumer has a positive association with CS spending, while for industries further away, this is negative. In that way, the results indicate that the amount of CS spending is larger for industries that are closer to end-consumer. A reason for this is the expected benefits derived from consumers valuing corporate sustainability.
Acknowledgements
As part of accomplishing my academic career I would like to express my gratitude to my supervisor Prof dr. T. Jochem at the University of Amsterdam, for his support, feedback
throughout the process, valuable input and patience.
Furthermore, a very special thanks goes to my parents, who offered me their house in a remote region in France. This gave me the new energy and inspiration all I needed to finish
my Master thesis.
Thank you! I appreciated it a lot.
‘’Your ability to learn faster than your competition is your only sustainable competitive advantage’’ – Arie de Geus
Table of Contents
1. Introduction ... 6
2. Literature Review ... 9
2.1 Corporate Sustainability Definition ... 9
2.2 Why engage in CS ... 10
2.3 The link between CS and financial performance ... 12
2.4 The role of consumer’s behaviour in sustainable performance ... 18
2.5 Hypotheses development... 19
3. Research Design ... 22
3.1 Sample and Data: Event study ... 22
3.2 Research Method: Event study ... 25
3.3 Descriptive statistics: Event study... 29
3.4 Sample and Data: Panel regression ... 31
3.5 Research Method: Panel regression ... 32
3.6 Descriptive statistics: Panel regression ... 34
4. Results ... 36
4.1 Event study ... 36
4.2 Panel regression... 44
5. Conclusion and Limitations ... 51
References ... 54
1. Introduction
The recognition of the concept of corporate sustainability started around the 70’s. A debate concerning CS, touched upon issues relevant to the modern economy and their consequences for individuals, society and organizations. This was because the society started to demand for firms that invest in sustainable development and satisfy their stakeholders. Since the last 15 years, the demands of the society are rising; they demand for more and more sustainable investments and products. Not only the financial performance of a firm is interesting, an increasing number of investors start also valuing the way firms meet their sustainable responsibilities (Barnett and Salomon, 2006). As a result, the pressure for accountability on various aspects (e.g., legal, social, financial) is increasing. And thus the pressure of the government to develop stricter regulations is increasing as well. All these developments are partly related to the corporate scandals and failures that have been predominated in the media (e.g., Enron and WorldCom). These accounting irregularities and corporate scandals have given a loss of trust in the corporate management. For these reasons, the investment in
sustainability is a sort of way to guarantee the stability and continuity of the financial system, an answer to societal uncertainties that an organization has to cope with (Lo and Sheu, 2007; Salzmann et al., 2005; Van Dijken, 2007). As response, an increasingly popular investment strategy of firms is aiming to satisfy the social responsibilities combined with maximizing the financial returns (Cheung, 2010; Curran and Moran 2007; Van Beurden and Gossling, 2008). This thesis centers on the link between sustainability and financial performance, which is empirically justified by the inconclusive results of existing literature. This issue is particularly relevant from a strategic point of view. Even so, much of the research in this field of CS, attempt to provide an answer on the question of doing socially good has a financial payoff. And thus focuses only on the altruistic (sustainability participation) CS activities instead of also the strategic (stakeholder management) activities. So far, the results of the previous studies on this question are still mixed (Lourenco et al., 2011; Van Beurden and Gossling, 2008). This leaves the firms without a clear direction regarding the desirability of investment in CS. A primary reason for the mixed results is the recognition of the relevance of multiple stakeholders. These various stakeholders have different responses and
expectations, which will influence how social performance is defined and what the
performance of the firm will be. This divergence in expectations and response has shifted the focus from a merely financial orientation to a much broader one; how do firms do good. More importantly, this shift to a much broader understanding is of greater importance to
firms (McWilliams and Siegel, 2001). Similarly, there is a growing interest among firms over the strategic activities of CS and specifically consumer’s behaviour in this relation. Recent researches suggest that there is a positive relation between firms CS activities and several consumer related outcomes, such as the attitude towards the company and its products (Bhattacharya and Sen, 2003). There exist complex interrelationships between consumer behaviour and firm performance, and this makes it hard to implement effective strategies towards CS spending. Although already a stream of research has contributed to a great insight, there is still a limited understanding about some key factors of consumer behaviour that affect the CS spending attitude of firms. In summary, the literature still suffers from some of the implications of CS and lack of understanding from various perspectives. This research responds to this by providing and expanding the framework for analysing CS within the context of the firm and industry. In order words, this thesis tries to shed further light on the strategic implications of CS. More specific, understand when and how a consumer reacts to the CS spending’s of a firm by focusing on the distance of a company’s industry to the end-consumer. And thus the rational marketing strategy of the firms to spend on CS. In order to understand these strategic CS activities, first a basic knowledge and updated view about the impact of the altruistic CS on firm performance needs to be understood. Because, according to Ruf et al. (2001), it is not known how reliable this
evidence is from previous studies on the question of doing good pays off. This is because the outcomes from the investment in CS change with new developments and expectations. Besides, analysis of strategic implications of firms CS investments is hampered by cross-industry differences. Corporate leaders know that the business norms and standards, regulation and consumers demand for and expectations of CS can vary substantially across firms and industries. For firms that operate in certain industries, this complicates the process of determining if it should invest in CS and how much. And thus when it is clear if the impact of CS activities on the performance vary across industries (i.e., imply industry
characteristics), this thesis is better able to analyse the moderator variable; distance of a company’s industry to end-consumer. Motivated by this, one focus of this thesis is to empirically re-examine and analyse the impact of CS investment on the financial
performance of firms and across various industries. Afterwards, this research studies the effect of the distance between a company’s industry and end-consumer on the behaviour of consumers and therefore the investment practice of company’s industries. The question that
arises is: To what extent is the level by which a company’s industry is removed from
end-consumers important for the CS spending behaviour
Based on this framework, the methodology of this analysis is carried out in two distinct parts. First, an event study has been conducted in order to examine the stock price changes of US companies and industries surrounding an inclusion or exclusion from the DJSI over the period 2005-2014. Second, this paper proposes a panel regression in which a company’s industry distance to end-consumer is the intermediary variable for the effect on CS spending. The results implicate that for both, firm specific and industry-specific; there exist no direct impact of CS activities on the performance. More specifically, there is no significant
abnormal return generated by an index change for companies and across industries. Possible explanations for these non-significant results are the used research method (estimation window, event window, announcement day) and sample. However, there is evidence for a positive relation between CS spending and the distance between a company’s industries to end-consumer. Particularly, this thesis shows that, the distance between a company’s
industries to end-consumer, under certain conditions, increase or decrease the CS spending of firms.
This study extends and contributes to the debate on the role of CS in corporate strategy in several ways. The first analyse provides additional and updated evidence on the question if firms do well by doing good. This thesis considers the various concerns obtained over the years from prior studies in order to provide a clear answer to this question (Cheung, 2010; Consolandi et al., 2009; Karlsson and Chakarova, 2008). Second, the impact of CS
investment per industry for one country is given. Finally, this research contributes with a new moderator variable for the strategy of CS activities. Although there is a large body of
literature on the CS strategy of companies, it lacks empirical analyses and understanding. Besides, it illustrates the complexity of the mechanisms that play a role in CS investments. So this research examines a new mediator variable that has been added to the regression analyses of various previous related papers (e.g. Bhattacharya and Sen 2003; Klein and Dawar 2004; Valor, 2008).
The remainder of the study is organised as follow. Section two, provides the concept of CS and a selected literature review. The third section discusses the research design, including the data and methodology. The fourth section presents the empirical results and analyse those findings. And finally, the fifth section concludes and offers limitations.
2. Literature Review
This section provides a definition for the different sustainable concepts used in previous studies. Subsequently, an explanation is given why firms invest in sustainable performance. This is followed by a review on the existing evidence regarding the relation between
sustainable performance and either accounting or market-based financial performance.
Moreover, the role of consumer’s behaviour is discussed. Finally the hypotheses development is described.
2.1 Corporate Sustainability Definition
In 1987, the World Commission on Economic Development has defined sustainable
developments (SD) as ‘’meeting the needs of the present without compromising the ability of future generations to meet their own needs’’. However, a more recent definition of
sustainable development rely on the intersection of three principles environment, economic growth and social equity and is part of the social responsibility and sustainability of a business. This means a company contributes to sustainable development if it manage their operations in such a way that it increases the competitive advantage while it still secure social responsibility (Bansal, 2005; Lourenco et al., 2011). Moreover, the strategic engagement in SD has become an important dimension of corporate sustainable voluntary practice and constitutes a way of creating value for the company (Hart and Milstein, 2003; Lopez et al. 2007). While some argue that SD is the integration of economic, environmental and social principles, incorporated by the firm this interpretation is quite different and is called Corporate Sustainability (CS). Nevertheless, the European Commission argues that even though the interpretation is different, CS and SD are intrinsically linked and thus can be seen as a business contribution to SD (European Commission, 2002).
In recent years a broad list of the terminology of Corporate Sustainability (CS) and Corporate Social Responsibility (CSR) has arisen by various authors. Back in the days, the neoclassical economics theory from Friedman (1970) defines Corporate Sustainability primarily by the economic principle. The goal of firms exists of economic performance and not environmental or social equity. The focus is on market share and shareholder value, in order words; Friedman focuses only on a very precise aspect of corporate and social
responsibility (Bansal, 2005; Van Beurden and Gossling, 2008). These days, the concept of CS contains the social, environmental and economics aspect and thus environmental
this we avoid the classical economic theory and go beyond just maximizing shareholder’s interest (Lopez et al., 2007). The European commission (2007) defines CSR as the voluntary integration of environmental and social principles in a firms operation. Curran and Moran (2007) add something to this definition and say that firms are self-responsible for the integration of social and environmental principles and thereby their impact. Moreover, Lo and Sheu (2007) see CS as a business approach that enhances long-term shareholder value by managing the risk from the three principles and embracing various opportunities. While Adams and Zutshi (2006) define CS as firms that use their resources in such a way that it benefits and improve the welfare of the society and environment.
To sum up, CS can be seen as a valuable resource that firms can use to improve their firm performance but it stilllacks a standard definitions and interpretation. However, in general, the various concepts are all about voluntary business activities and a firms impact on relationships with and responsibilities to society (Van Marrewijk, 2003). Therefore, this paper uses the concepts interchangeably as it is considered that they all address the same basic issues. And thus a firm is called sustainable if it contributes to sustainable
developments through presenting environmental, economic and social benefits, likewise the triple bottom line (Hart and Milstein, 2003).
2.2 Why engage in CS
A company considers engaging in CS and reporting their activities and impacts, if it expects that reporting and investment in environment and social strategies brings specific benefits and improved performance. Or in other words creates value, in terms of sufficient profits and satisfy the demands of stakeholders (Hart and Milstein, 2003). Besides the various benefits that are created by investment in CS, in general there are two key drivers why firms act in a social responsibly manner and report it. Recent years, firms have enormously been increased in size and operate in more countries than ever before. This development has led to a
situation of increased power and impacts that firms have on society and environment in various countries. Therefore, one of the drivers is that companies acknowledge that they have a broader responsibility for those impacts. The second driver is that firms recognize that is it in their interest to report. Several parties such as investors, customers, politicians and
government exert pressure on companies because they want more public information. Therefore, companies start to report their activities to improve their reputation with these parties (Adams and Zutshi, 2006).
While some papers argue that social and environmental welfare expenditures are costly1 and only in the self-interest of the manager, there also exist evidence that it should be considered as a strategic investment and thus creates opportunities of competitive advantages and increased profits (McWilliams et al., 2006, 2001). Last decade, various authors have
outstandingly analysed the engagement in social and environmental responsible activities and argue that it creates benefits for the company as a whole (Lourenco et al., 2011; Porter and Kramer, 2006). In addition, those various benefits ensure improved performance over firms that do not adopt CS (Adams and Zutshi, 2006; Lopez et al., 2007). The benefits created from CS investment could be separated into internal and external benefits (Orlitzky et al., 2003). There are several internal benefits due to the investment in social and environmental practices. First, the government encourage the investment in CS activities and therefore supports the companies frequently (Wagner, 2010). A second benefit is that a firm attracts and retain more talented people, which lead to more development of capabilities, motivation and productivity. That in turn results in new resources and surrogate the innovation of the company. In addition, this results in a cost reduction of training for new employees (Lourenco et al., 2011). Thirdly, companies that develop new intangible resources could differentiate its products from competitors and thereby raise their demand, have a competitive advantage and enhance its performance (Brickley et al., 2002). As fourth, the reporting of firm’s social and environmental practices ensures better internal control, decision-making and more alignment with the business goals. Because of this there are improvements in the
efficiency of the operations and processes, which results in more predictability and reduced risk. Finally, due to the investment in CS, firms also adapt the norms and expectations of stakeholders. This is followed by improved communication with the stakeholders and thereby building a good relationship with them. All together this brings increased financial returns for the company. Nevertheless, the external benefits are all associated with corporate reputation. By reporting information of social and environmental activities in public, several parties gets a better understanding of the CS practices and thus reduces the criticism from in- and outside the company. This results in an improved reputation of the firm (Adams and Zutshi, 2006). Companies should care about their reputation because it provides a company with
competitive advantage. Reputation is a valuable intangible assets that affects the market price of its shares due to the positive relations with investors, banks, suppliers and customers and
therefore more loyalty and commitment to the firm (Branco and Rodrigues, 2006; Lourenco et al., 2011; Orlitzky et al., 2003). Building a good reputation leads to more negotiation power of the company to suppliers and governments. For example, a company can negotiate with a supplier about the contract and prices of goods and services and therefore reduces its cost of capital (Fombrun et al., 2000). In conclusion, investment in CS activities has long-term benefits because a firm becomes more attractive for investors due to reputation, relations and employee activities. From a financial performance perspective, the reason for
engagement in sustainability lies in improving efficiency in their operations and more alignment with firm’s business goals. As well as reduce costs and become a more profitable company in the future (Artiach et al., 2010). Altogether, companies engage in CS is because they generate competitive advantage by controlling and manipulating their valuable and rare resources effectively and that in turn is linked with positive firm performance and value creation in the long-term.
On the other hand, the investment in CS could as well go wrong. One example that can occur is that the costs are greater than the benefits. To put it in another way, the social responsible actions reduce not only the profits but have a negative result on the social welfare too. Another example is when firms only engage in CS as a kind of image advertisement. In this case the reputation of the firm is strengthen, however if it does not follow up with the activities it put forward, the damage will dominate the benefits (Van Dijken, 2007).
As shown, there remains disagreement about the specific contribution, meaning of and motivation for CS. There does not exist a standard recipe, the process of sustainability vary per company. Therefore managers should choose themselves which contribution is the best match with the aims and strategies of the firm. And thus are firms self-responsible for their created benefits and social and environmental impacts. That’s why they need to manage and monitor those impacts and benefits properly (Curran and Moran, 2007; Van Marrewijk, 2003).
2.3 The link between CS and financial performance
This section gives an overview of the theoretical arguments, accounting-based studies and market-based analysis regarding the link between CS and financial performance. Researchers have hypothesized that this link has a negative, neutral or positive association. Since it is
important that the included literature is up to date2, this paper excludes most of the research published in the 70’s, 80’s and 90’s and concentrates only on the most influential papers. Two specific reasons for this exclusion are the Brundtland Report in 1987 and the already existing quality reviews3 (e.g., Cochran and wood, 1984; McGuire et al., 1988).
Various studies analyses the link between CS and financial performance using theoretical models. One view argues that there exists a negative relation because investment in CS is costly and thereby reduces profits, shareholders wealth and economics advantages compared to other firms (Alexander and Buchholz, 1978; Becchetti et al., 2007; Preston and O’Bannon, 1997; Vance, 1975; Waddock and Graves, 1997). Hence, a firm that invest in CS has
additional costs from the adaption of environmentally and social friendly practices such as charitable donations, improved employee conditions and the establishment of environmental protection procedures (Artiach et al., 2010; Aupperle et al., 1985; McGuire et al., 1988). Besides, the investment might limit a company’s strategic alternatives, for example
investment opportunities that will be missed by acting responsible. A second view claims that CS is not directly related to financial performance. Ullmann (1985) and Waddock and Graves (1997) explain this neutral relationship through the complexity of the relation between firm performance and society. There are many influences that step in; the situation is so complex that there does not exist any simple direct relation. Since it is difficult to control for these interventions that step in, Ullmann (1985) argues that there is not an adequate amount of theoretical research, in order to support for a direct relation. Finally, a third view says that there exists a positive relation. A first argument for this is that the potential benefits to the firm exceed the actual costs of investment in CS. The costs of environmental and social practices might be way less than the gains that result due to more and efficient productivity. For example through employee motivation, morale, goodwill, better access to capital and above all improved relationships with investors, bankers and government (Barnett, 2005; McGuire et al., 1988; Waddock and Graves, 1997). A consistent view described by Preston
2 According to Roman et al. (1999), previous papers about CS in relation to firm performance could be used as
theoretical background, yet not as empirical justice
3 The year around 1987 could be seen as a turning point in the awareness towards CS because of the World
Commission on Environment and Development that published the Brundtland Report. This has brought the attention towards the risks and problems in the world and changed the way in which organizations do business. This change together with the reactions, consequences and responses from various parties is with most certainty not covered in research papers before the 90’s (Van Beurden and Gossling, 2008).
and O’Bannon (1997) and supported by the stakeholder theory is, that investment in CS and managing the stakeholders generates a positive impact on the financial performance of the firm (Artiach et al., 2010; Simpson and Kohers, 2002). This is because the significant cost for the investment in CS is offset by various other costs of the firm that are reduced. Lastly, successful firms that invest in CS practices have superior resources and this results in higher financial performances. In order words, only firms that have superior and sufficient resources have the availability to invest in CS. Therefore, CS is positively related with financial
performance because the investing firms in CS have better resources that generates superior financial performance (Alexander and Buchholz, 1978; Preston and O’Bannon, 1997; Waddock and Graves, 1997).
Studies using market-based measures show mixed results over the years. Early studies from Moskowitz (1972), Vance (1975) and Alexander and Buchholz (1978) all use the social responsible firms rated by Moskowitz (of around the same period) for their research. While Moskowitz (1972) finds higher average returns for high ranked sustainable firms, Vance (1975) however, find that some firms have lower market performance compared to the NYSE or S&P500. Furthermore, Alexander and Buchholz (1978) adjust for risk and finds no
significant impact. More recent studies rely on sustainable indexes such as DJSI, FTSE4Good and Domini 400 Social Index. The majority of those papers do not find any significance impact of sustainable performance on stock returns and risk, aside from studies by Dilling (2008) and Consolandi et al. (2009). The former finds positive abnormal returns and these returns change across industries. Consolandi et al. (2009) looks into the European firms and finds positive abnormal returns around the announcement day that ends at the effective day. Papers examining the link between sustainable performance and accounting-based financial performance generally find positive results. For example, the studies by Lo and Sheu (2007), Waddock and Graves (1997) and Ruf et al. (2001) find a positive relation, controlling for size, risk and industry but do not consider any possible effect of other variables. Later studies by Dowell et al. (2000) and Wagner (2010) offer support for a positive relation considering R&D and advertising expense. In contrast, Aupperle et al. (1985), Lopez et al. (2007) and McWilliams and Siegel (2000) find either no significant or a negative impact. Table 1 below gives a comprehensive overview of the most influential papers that have been published over the years.
To be concluded, theoretically it is not obvious whether social behaviour is financially and economically beneficial. However, among the empirical studies, the outcomes are
contradictory. While the majority of the previous research on accounting based measures tends to support a positive reaction, the market-based papers show mostly no significant impact. Notable, across the various studies there is a considerable variation in the research design.
Author(s) and Year Sample Reseach Method Corporate Sustainability Financial performance Moderator variables Findings
Positive relation
Artiach et al. (2010) Cross-sectional Data US firms 2002-2006 OLS Dow Jones Sustainability Index TA, ROE, ROA Leverage, FCF, PB Positive relation Barnett and Salomon (2006) Panel data SRI funds 1972-2000 OLS Social Investment Forum Risk Adjusted Performance Globality of fund Positive relation Cochran and Wood (1984) US firms 1970-1974 & 1975-1979 OLS Moskowitz reputation index ROE, ROA, EV Age, Turnover Positive relation Dowell et al. (2000) US MNE firms 1994-1997 OLS Investor Responsibility Research Center Tobin's Q R&D, Advertising Positive relation
Garcia-Castro et al. (2010) Panel data firms in KLD 1991-2005 OLS,FE, IV Kinder Lydenberg Domini ROE,ROA,MVA,Tobin's Q Size, Risk, Industry Positive for OLS, non-significant others Griffin and Mahon (1997) Chemical companies KLD, survey 1992 Ranked, Stock returns (CAPM) KLD and Fortune Survey ROE,ROA, TA Size, Growth Positive relation
Herremans et al. (1993) US firms 1982-1987 Ranked, Stock returns (CAPM) Survey, firms from Fortune 500 Stock market returns Industry Positive relation Lo and Sheu (2007) Panel data US S&P500 firms 1999-2002 OLS, pooled, fixed effects Dow Jones Sustainability Index Tobin's Q Positive relation Lourenco et al. (2011) Panel data US&Canada firms 2007-2010 OLS Dow Jones Sustainability index MV BV, NI Positive relation Luo and Bhattacharya (2006) Panel data US firms 2001-2003 OLS FAMA Tobin's Q, Stock return Customer Satisfaction Positive relation McGuire et al. (1988) US firms 1983-1985 OLS Survey, firms from Fortune 500 ROE,ROA, Growth & Return Positive relation Pava and Krausz (1995) US firms 1985-1987 & 1989-1991 OLS Council on Economic and priority ranking P/E ratio, ROE, ROA, EPS, DIV Investment intensity,size Positive relation
Preston and O'bannon (1997) US Firms 1982-1992 OLS Survey, firms from Fortune 500 ROE,ROA, ROI Positive relation
Ruf et al. (2001) US firms 1991-1995 OLS Kinder Lydenberg Domini ROE, ROS, Growth in sales Size, Industry, Sales Positive relation Simpson and Kohers (2002) US Banks 1993-1994 OLS Community Reinvestment Act Ratings ROA, loan losses Loan ratio, capital ratio Positive relation Surroca et al. (2010) Panel data industrial firms 28 countries 2002-2004 Three Regression model Baron and Kenny’s (1986) Sustainalytics Platform Database Tobin's Q Intangible resources Positive relation Waddock and Graves (1997) Cross-sectional data US S&P500 firms 1989 -1991 OLS Kinder Lydenberg Domini ROE, ROA, ROS Size, Risk, Industry Positive relation
Wagner (2010) Panel data US S&P500 firms 1992-2003 OLS Kinder Lydenberg Domini Tobin's Q, R&D, Advertising Positive relation, adv. Intensity moderates it
No relation
Abbott and Monsen (1979) Panel data US firms 1971-1975 Comparison Beresford's Social Involvement Disclusure Av. Annual % return to investors No significant impact Aupperle et al. (1985) 818 sent surveys 1985 OLS Survey, listed in Forbes 1981 (risk-adjusted) ROA No significant impact McWilliams and Siegel (2000) Cross-sectional data KLD firms 1991 -1996 OLS Kinder Lydenberg Domini ROE, ROA R&D No significant impact
Negative relation
Lopez et al. (2007) Panel data European firms 1998-2004 OLS Dow Jones Sustainability Index PBT REV Negative relation
This table presents an overview of the most important and influential papers of Corporate Sustainability in relation with either accounting or market-based financial performance, that started around the 1970's. Panel A Table 1: Overview of studies in the relation between CS and either accounting or market-based financial performance
Panel A: Overview of studies in the relation with CS and accounting-based financial performance
contains the overview of studies for the accounting-based financial performance. Panel B shows the overview of studies for the market-based financial performance and the meta-analysis. Both panels give for each author a summary of the specific sample, research method and variables that are used in the their paper. Besides, the overall findings from the papers about the link between CS and financial performance are showed.
Author(s) and Year Sample Reseach Method Corporate Sustainability Financial performance Moderator variables Findings Positive relation
Consolandi et al. (2009) European firms from DJSSI 2001-2006 Event Study Dow Jones Sustainability Stoxx Index CAAR Positive impact
Dilling (2008) All companies from DJSI World 2002-2005 Event study per industry Dow Jones Sustainability Index CAR Positive impact and varies per industry Moskowitz (1972) Moskowitz sample Comparison Moskowitz reputation index Stock market returns Positive relation
Van Dijken (2007) US firms DJSI, S&P500, NASDAQ Comparison Dow Jones Sustainability Index Stock market returns CS returns outperform market No relation
Alexander and Buchholz (1978) US Firms 1970-1974 Ranked, Stock returns (CAPM) Survey, firms from Fortune 500 Stock market returns No significant impact
Bechetti et al. (2007) US Firms from Domini 400 index 1999-2004 Event study Domini 400 Social Index CAR No significant impact for exclusions Cheung (2011) US firms from DJSI 2002-2008 Event study Dow Jones Sustainability Index CAR, risk and liquidity No significant impact
Curran and Moran (2007) UK firms from FTSE4Good 2001-2002 Event study FTSE4Good index Stock market returns No significant price movements Karlsson and Chakarova (2008) Nine countries from DJSI World 2002-2007 Event study Dow Jones Sustainability Index CAAR No significant impact Negative relation
Vance (1975) Moskowitz sample firms 1972-1975 Comparison Moskowitz reputation index Change in share price Negative relation Brammer et al. (2006) UK quoted firms 2002-2003 OLS Ethical Investment Research Services Stock market returns Industry Negative relation Meta-analysis
Orlitzky et al. (2003) Meta-analysis Comparison Average positive relation
Roman et al. (1999) Meta-analysis Comparison Positive relation
Margolis and Walsh (2007) Meta-analysis Comparison Positive relation
Salzmann et al. (2005) Meta-analysis Comparison No significant impact
Van Beurden van Gossling (2008)Meta-analysis Comparison Positive relation
Wu (2006) Meta-analysis Comparison Average positive effect
2.4 The role of consumer’s behaviour in sustainable performance
This section gives a short review about the role of consumer’s behaviour in relation with sustainability investments. This is followed by the importance of the industries characteristics and consumers awareness. Finally, the role of advertising is described.
Several marketing studies find that sustainable activities have significant influence on various consumer-related outcomes and therefore the business benefits of a firm (e.g., Bhattacharya et al., 2006; Du et al. 2010; Fombrun et al. 2000; Bhattacharya and Sen 2001; Murray and Vogel 1997). More specific, CS practices positive effect consumer behaviour for a range of cognitive and affective (e.g., awareness, attributions, customer identification, attitudes) as well as behavioural factors (e.g., loyalty, relationship) (Bhattacharya and Sen, 2001; Klein and Dawar, 2004). For example if a firm behaves in a good social responsible way and still produce quality, a firm can foster product differentiation, consumer identification and loyalty and turn them into consumers who engage in advocacy behaviour (e.g., purchase and
willingness to pay a price premium, positive word-of-mouth and resilience to negative company news) (Du et al., 2010). As can be seen in figure 1 below, consumers react in multiple-ways and thus it is quite hard for firms to understand all the range of factors that leads to this positive relation. However, some key prerequisite for this range of factors to result in positive returns and benefits are the awareness creation and most importantly the firms and industry characteristics. McWilliams and Siegel (2001) and Ullmann (1985) argue that the provision of CS attributes depends on certain characteristics of the industry. For example, to which extend a company can differentiate their products and the certain industry life cycle. These various industry characteristics affect the sustainable performance of the firms, consumer’s behaviour and in return, the financial performance of the company.
Besides, the positive attitude of the consumers and thus the provision of CS activities are also highly related to consumer’s awareness of sustainability practices from firms. Firms with high consumer awareness can enhance firm value by increasing their CS spending
(Bhattacharya and Sen, 2001; Du et al., 2010; Murray and Vogel, 1997). Much of the
academic research about the insights into the psychological mechanisms and outcomes of CS-drive consumer behaviour even presumed CS awareness in their analysis. And thus the
question is; how do consumers become aware. This is where the role of advertising in relation with CS becomes interesting. Advertising enlarges a company’s information environment, because it reduces the information gap between the company and the consumers. Thereby
increasing the awareness of consumers about a company’s involvement in CS. And thus is it more likely that consumer’s reward and react positively to firms CS investment. According to McWilliams and Siegel (2000,2001), CS advertising increases the awareness and thus the demand for corporate sustainable behaviour and the returns to engage in such behaviour. Therefore, several studies analyses this relation between CS and advertising expenses (McWilliams and Siegel, 2000). And indeed, the majority find that firms with a high
advertising intensity have positive impact on firm value. While a low intensity of advertising, has a negative impact. And thus advertising is highly related to the positive benefits the firm gets due to the awareness.
To summarize, CS activities have a positive effect on divers factors of the consumer’s behaviour. Nevertheless, to benefit from CS investment via these factors, various key
determinants are a prerequisite. For example, the advertising to promote consumer awareness of CS activities, the attribution and attachment of the consumer (e.g., belief, causal reasoning, identification), industry characteristics and product differentiation.
2.5 Hypotheses development
Since 1972, the increased interest in CS practices has led to a situation where a continuous stream of researchers study the link between sustainable performance and either accounting or market-based financial performance. They try to find evidence for the question of doing good pays. Some previous studies suggest that investment in CS can be value adding, while
Figure I: Internal and External Outcomes
Input Outcomes Internal Outcomes external
This figure gives an overview of all the different internal and external outcomes for the company and consumer by the investment in CS activities.
CS Activity Type: Community Support Diversity Environment Product Investment: Money/Goods Time Personnel Expertise Company Awareness Attitude Attribution Attachment Consumer Well-being Company Purchase Price Loyalty Resilience word of mouth Consumer Behaviour Modification
others do not see a direct link or even propose a negative relation (Artiach et al., 2010; Orlitzky et al., 2003; Simpson and Kohers, 2002). Two concerns explain this ambiguous result created over the years. The first concern is the value-laden literature. Definitions, assumptions, objectives, interpretation of the concepts and the results are affected by the values of the various authors and new expectations and developments. The second concern is related to the various research designs (i.e., measures, control variable, mediator variable, specific test, time period) used in the studies (Griffin and Mahon, 1997; Lopez et al., 2007). These concerns influence the results because accounting and market-based performance measures focus on different aspects of performance. According to Griffin and Mahon (1997), the upfront used measures for CS and financial performance already predetermine the
relationship outcome. Therefore, it is not surprisingly that the results of the accounting and market-based performance vary. While the accounting-based measures tend to support a positive relation, the market-based outcomes are mostly contradictory. Striking is, that the market-based analyse makes use of the event study. This methodology is supposed to be the most successful in the area of corporate finance and it has several advantages over other methodologies. For example, it presents a quick and direct insight whether investors care about CS, the reactions on the stock market can be analysed within several dimensions (e.g., liquidity, risk and return) and the short and long term behaviour of CS investments is
available if the event window length will be adjusted several times (Cheung, 2010). Even so, the results are conflicting. Because of these inconsistent outcomes for market-based (event) studies this thesis gives an updated evidence for the relation between CS investment and market-based financial performance, considering an appropriate research design (i.e., event window, sustainable index, estimation window) based on previous analyses. Therefore the first hypothesis that can be proposed:
H1 = There is no significant abnormal return generated by an entry or exit of a US company from the DJSI between 2005-2014
However, Artiach et al. (2010) argues that not all company’s invest in CS activities and those that do invest do so at various levels. A company’s investment in sustainability is partly determined by its environment. As argued by various authors (e.g., McWilliams and Siegel, 2001), the provision of CS depends on characteristics of the industry. Therefore, the impact of firms CS activities might differ across industries. But, this impact of CS across various industries has not been analysed much, especially not for one country specific. And thus it is
interesting to show these impacts for various industries, separately for one country. It follows that:
H2 = There is a significant abnormal return generated by changes in the DJSI within various industries in the period 2005-2014.
This paper posits that any announcement of an index change in the DJSI has no significant impact. This is because it is difficult to determine the actual impact of CS activities due to different research designs and various other events of around the same time period that might influence the results. Besides, the important underlying processes are omitted by this market-based method. However, consistent with Ullmanm (1985), this thesis proposes that the impact do vary across the different industries. In particular, industries with high levels of product differentiation (e.g., consumer goods, financial services) should show more impact by any index change. This new evidence is important for the main purpose of this study. It provides a stable basis for the recent developments in the sustainable area of market-based financial performance as a whole and across industries. Moreover, it completes the theoretical and empirical arguments to take steps along this direct link of CS and financial performance. In other words, the need for other research methods to consider the underlying processes and missing variables.
Moreover, because one of the explanations for these conflicting market-based findings could be the largely omitted underlying processes, this relationship between CS and financial performance ask for a deeper understanding in order to give a more certain prediction and explanation for the outcome of CS activities. It is important to acknowledge that the benefits provided by CS investments also deals with the underlying processes by consumers.
Especially because several studies already find evidence for the importance of consumer reactions on firm’s profitability and spending behaviour (Bhattacharya and Sen, 2003). These processes are complex and not straightforward, numerous psychological factors play hereby a role. That means, the market-based measures of performance may not be sufficient for this research. The market-based measure implies that the financial stakeholder valuation of firm performance is a proper performance measure. However, firms face multiple constituencies and thus other stakeholders, such as consumers are also affected by a company’s CS practices (Ullmann, 1985). Given this growing concerns of current methodology and potential diversity in consumer’s behaviour to CS practices it is interesting to also use the accounting-based
determinants determine the level of CS spending, the various consumers reactions and in the end, financial performance. In other words, these key prerequisite could be the ultimate test for the success or failure of any strategic initiative. As one of the key determinants is the industry characteristic, it is worthwhile to consider this in more detail. More specific, analyse the distance between a company’s industry to end-consumer in relation with the CS
spending’s. This leads to the two main hypotheses:
Hypothesis 3: The distance between a company’s and end-consumer has no significant effect on the CS spending behaviour
Hypothesis 4: The CS spending is larger for a company’s industry that is closer to end-consumer
This study proposes that the distance between a company’s industry and end-consumer has a positive impact for industries that are close to end-consumer and negative for industries further away. Besides, the CS spending is larger for a company’s industry that is closer to the end-consumer because it will obtain more benefits due to positive consumer behaviour. This follows the discussion from Artiach et al. (2010) and McWilliams and Siegel (2001), that industry characteristics may explain the variations in the levels of investment in corporate sustainability.
3. Research Design
This section provides an overview of the sample and data. Subsequently, the specific research method is given. In the end, the descriptive statistics is illustrated. Because this thesis consists of two different analyses whereas the research design differs, this section is divided into two parts.
3.1 Sample and Data: Event study
The empirical analysis of the index changes on the stock market of a company and across various industries uses an event study for a set of US firms. The set of firms in the DJSI is used to define the sample of firms to be analysed. The sample for this study is composed of U.S. companies that have been added to or deleted from the DJSI in the period of 2005 to 2014. A further description about the Dow Jones Sustainability index is at the end of this section.
In order to examine the impact of an index inclusion or exclusion from the DJSI on the stock market returns, various market data is required. The data required should match the event that has been studied. First, data from the sustainability index (DJSI) is retrieved. Since not all the prerequisite information for this study was publicly available, a request to the DJSI was sent. Therefore, the information about the additions and deletions of the various US companies and their industry classification have been sent by a PDF-file. The remaining information about the press release dates and effective dates is collected from the website of the DJSI. The press release dates, implementation dates and industry classification can be find in the descriptive statistic section (3.3). Afterwards, in order to calculate the normal and abnormal returns, market data from the CRSP database is necessary. Particularly, the relevant stock prices of the sustainable companies in the DJSI and the returns of U.S. companies in the S&P500 index are gathered from CRSP. The S&P500 index is representative for the market
developments of all the U.S. firms. For the right stock prices, the daily stock files from CRSP with the variables price is used. Further, to get the same sustainable companies as in the DJSI, the companies are manually added by their unique gvkey codes. So all daily data of the prices from the annual updates of CRSP is collected over the period of 2005-2014. As well as the returns from the S&P500 index. This is done by downloading the daily value-weighted return of the entire S&P500 index file over the same period. Moreover, to give a reliable comparison of reactions to inclusions or exclusions from the DJSI per industry, each stock is allocated to one of the ten standard industries. However, the DJSI gives the super sectors instead of the industries and thus every company is placed into the correct industry based on their GICS code4. One striking and important point that needs to be clarified, the DJSI has changed the super sectors of various companies, started from 2013. This means that some of the companies in 2013 and 2014 were allocated to a different industry according to the DJSI. This thesis did not follow the DJSI industry change and allocated the various companies to the same industries as the years before. Finally, these two files of the daily prices of the sustainable companies plus the S&P500 returns are merged together.
In order to get the abnormal returns, the returns of the sustainable firms needs to be
calculated first. The returns of the sustainable firms are easily calculated from the historical prices by the following formula:
𝑅𝑖,𝑡 =𝑃𝑖,𝑡 − 𝑃𝑖,𝑡−1 𝑃𝑖,𝑡−1
Where 𝑅𝑖,𝑡 is the individual stock return for a company and 𝑃𝑖,𝑡 is the individual stock price for a company. Afterwards, each individual return of the sustainable firm is matched with the S&P500 index return. Data is omitted from the sample when the observations of a single index change are not complete for that particular year. And thus, this data cannot be used for the analysis. From the period of 2005 until 2014, data of four stocks is not available.
Therefore, the final sample consists of 302 events from which 168 inclusions and 134 exclusions.
Dow Jones Sustainability Index (DJSI)
The importance of sustainability activities has led to the origin of several sustainability indexes. Among these are for example, FTSE4good index, Domini Social index, ARESE Sustainable Performance Indices, Ethibel Sustainability Index and the Dow Jones
Sustainability Group Index. These indexes are designed to measure the performance of companies worldwide that meet the global corporate sustainability criteria and helps with the investment in those companies (Lo and Sheu, 2007; Lopez et al., 2007). This thesis makes use of the DJSI as proxy for CS, because numerous studies accept and believe that the DJSI is a significant proxy for CS (Artiach et al., 2010; Cheung, 2010; Consolandi et al., 2009; Lo and Sheu, 2007; Lopez et al., 2007). Two of the motives for the use of DJSI that those authors mention, are that firms are evaluated on different criteria’s for each dimension independent of their characteristics and that the companies are rated over time. The Dow Jones Sustainability Index (DJSI) is the first global sustainability index that keeps track of the financial performance of worldwide leading companies that invest in CS practices. The index was launched in cooperation with the Sustainable Asset Management (SAM) in the year 1999. Every year, ten percent of the leading sustainable companies across different industries are selected from the Dow Jones Global Index (DJGI). In order to identify the ten percent leading firms, a rating5 is given to each company that will be altered annually. The DJSI family consist of 20 different indexes and five of them are based on the geographical
perspective: Global markets, Asia, North America, Europe and the US. The determination of sustainable companies is based on the corporate sustainability assessment applied by SAM. The assessment includes three major dimensions of CS: environmental, social and economic
5 This rating is based on a questionnaire, the so called best-in-class approach. In addition, more information
dimension. For companies to be on the list in the DJSI, it must satisfy the criteria of the sustainability assessment (Lo and Sheu, 2007; Lopez et al., 2007). Those criteria’s have been determined with a general procedure and applies to all industries. Table 2 below shows the various criteria, divided into the three dimensions. The companies included in the DJSI are monitored throughout the year and if necessary, they will be deleted from the list.
3.2 Research Method: Event study
Event Study
The first event study from MacKinlay (1997) dates back to the 1930’s. Nevertheless, the methodology how it has been used as today, is introduced by Fama et al. (1969). While there have been a number of new approaches developed the last couple of year, the overall
formulation is remained the same. The main goal is to measure to what extent a firms stock prices respond to the announcement of new information (Karlsson and Chakarova, 2008). Many research fields such as accounting, finance, economics and law, all use an event study to measure the impact of a change in the environment on the value of the firm. Up till now, event studies are the most successful in the area of corporate finance. Therefore, research in the area of corporates finance is dominated by the event study methodology. And thus, various previous studies that examine the impact of CS on the stock price of a company uses the event study as well. The general event study methodology focuses on the abnormal returns around the date of the announcements. The abnormal returns are the returns of an
Dimension Critera Weighting (%)
Economic Codes of conduct, Compliance, Corruption&Bribery 5.5
Corporate Governance 6.0
Risk and Crisis Management 6.0
Industry Specific Criteria Depends
Environment Environmental Reporting 3.0
Industry Specific Criteria Depends
Environmental Performance 7.0
Social Corporate Citizenship/ Philanthropy 3.5
Labor Practice Indicators 5.0
Human Capital Development 5.5
Social Reporting 3.0
Talent Attraction and Retention 5.5
Industry Specific Criteria Depends
Table 2: Dow Jones Sustainability Index Criteria
This table shows the various criteria based on three dimensions, that are required for a firm to be in the DJSI. Each criteria has its own weighting, which shows for what percent it counts. The criteria assessed are based on publicly available information.
individual firm minus the index returns. For example for US firms, the index returns are those from the S&P500 index. Those abnormal returns indicate the stock market reacts after the announcement of new information. The adjusted rate of return is the expected return minus the actual returns. When there is a significant difference, it is assumed to be an abnormal return. In order to get the cumulative abnormal returns for each company, the standardised abnormal returns must be cumulated over the event window. These cumulative abnormal returns are used to examine the average effect of the event on the total number of firms in the sample. When the abnormal returns are significant, this means that the event has an impact on the value of the company (Curran and Moran, 2007). The event study methodology has several advantages over other methodologies in the field of corporate finance. However, the event-study relies on various assumptions such as market efficiency, no other disruptive events in the event window and the event is unpredictable. Thus, for an event study to have valid outcomes, the authors must exactly identify the abnormal returns related with the specific event and date of the event (Cheung, 2010; Curran and Moran, 2007).
Generally, the event study methodology consists of three phases. First, determine the event of interest and make certain what the time period of the event should be. In other words, the total period wherein the share prices are analysed: the event window. According to Karlsson and Chakarova (2008), the event window is crucial. This is because there is evidence that the stock market response to any positive or negative information about environmental issues up to one week after the news has been published. It is common that the event window is larger than the real period of interest. This means that the event window has often been expanded with a few days before the news has come out and one or two days after. This way of event window determination is to capture any price effects that happens prior to the announcement, latecomers and when the stock exchange has closed. However, to avoid noise from other events and thus isolate the impact of that specific event, the event window should be close around the event of interest (Curran and Moran, 2007; Karlsson and Chakarova, 2008). In this thesis, the event of interest is the announcement of an inclusion or exclusion of a company from the DJSI. A study by Karlsson and Chakarova (2008), has carried out an interview with a head of public relations from the DJSI. From this interview, this study assumes that the press release date of the annual review of DJSI is the most appropriate date as the announcement date of the event. In the DJSI, the press releases always takes place in the first week of September, every year. Besides, the implementation day of these changes take place on the third Friday of September and become effective on the next trading day.
Notwithstanding, the headquarter of DJSI is in Zurich, Switzerland. It is important to keep in mind that the US stock market is closed when the announcement occurs in Zurich. Therefore, the dataset is adjusted to this time difference. This study calculates and analyse the
cumulative abnormal return (0,2) as main event window. In order to test if the result stays the same when the event window changes and thus check for robustness, other event windows are analysed: CAAR (0,3) and CAAR (-1,2)
The second phase is: specify a model for the normal returns of a stock. But first of all, in order to determine the normal stock returns, an estimation window should be defined. The estimation window is the period wherein the estimation is performed and consists of data of the period before the event window. Regardless of the numerous event studies in the area of index inclusions or exclusions from previous research, there still does not exist an appropriate length for the estimation window. While, various previous research use as estimation window a period of trading days, others use weeks or months. For example, Cheung (2010) uses a period of 235 trading days. Curran and Moran (2007) uses a period of 310 trading days. While Consolandi et al. (2009) even use a time span of 52 weeks. Karlsson and Chakarova (2008) and Becchetti et al. (2007) use a period of five and eight calendar months respectively. This study uses an estimation period length somewhere in the middle of the previous papers. The length of the estimation window is 125 trading days, lasting from [-130] to [-6].
Whereby time 0 is the time of the event. With this specification, there is a gap between the estimation window and event window. According to Klassen and McLaughin (1996), these days have been excluded from the analysis to prevent and limit impurity, such as insider trading. After the determination of this relevant information, a model for the normal stock returns is determined. This is because the purpose of this event study is to calculate the abnormal return and therefore the normal stock returns have to be calculated. Since abnormal returns are the actual returns minus the normal returns. There are several approaches to calculate the abnormal returns; the most common ones are the Market Model and Characteristic-based model. The various models differ in terms of benchmark returns,
however in case of short event windows is does not really matters, which model you use. The choice of the model for the normal returns has a little effect. This paper uses the market model because it is the most frequent used model by previous papers in this field as well as the model has a primary advantage above the others. The market model is easy to apply and it takes into account the differences in the beta of the stocks. The normal stock returns are
𝐸(𝑅𝑖,𝑡) = 𝛼𝑖,𝑡 + 𝛽𝑖𝑅𝑚𝑡+ ɛ𝑖,𝑡
Where 𝐸(𝑅𝑖,𝑡) is the rate of the normal return of an individual firm on day 𝑡. Or in other words, the expected return. Further, 𝛼𝑖 and 𝛽𝑖 are the intercept and slope ordinary least squares estimators of the regression coefficient. 𝑅𝑚𝑡 is the return of the market index on day 𝑖. Finally, ɛ𝑖,𝑡 is disturbance term.
Once the normal returns have been calculated, the third phase starts with the calculation of the abnormal returns. The abnormal return is the difference between the return of a certain company on a certain day and the normal return of that stock in the absence of the event. The calculation for the abnormal returns is as follow:
𝐴𝑅𝑖,𝑡 = 𝑅𝑖,𝑡− 𝛼ᵢ − 𝛽𝑖𝑅𝑚𝑡
Where 𝐴𝑅𝑖,𝑡 is the abnormal return of an individual firm on day 𝑡. 𝑅𝑖,𝑡 is the rate of return of an individual firm on day 𝑡. 𝑅𝑚𝑡 is the return of the S&P500 index on day 𝑖. Finally,
𝛼𝑖 and 𝛽𝑖 are the parameter estimates for an individual firm as computed from the window [-130, -6] before the event at date 𝑡. Afterwards, the abnormal returns have been aggregated across stocks over the whole event window. This is done because it is often more interesting to examine the performance close to the event window. This aggregation of returns is called the cumulative abnormal returns and is calculated by the formula:
𝐶𝐴𝑅𝑖(𝑡,𝑡+𝑘) = ∑ 𝐴𝑅𝑖,𝑡 𝑡+𝑘
𝑡=𝑑𝑎𝑦−1
Where 𝐶𝐴𝑅 is the cumulative abnormal return and 𝑘 is the number of days in the event window. Because it does not really make sense to analyse the returns for an individual
company separately, the average of the abnormal and cumulative return have been calculated. The reason why the returns have not been analysed separately, is because the stock price changes could be caused by some other event instead of the index change of sustainable companies inclusions or exclusions. Due to the average of the abnormal returns, all
information that is not related to this event will cancel out on average. The average abnormal return is formulated as:
𝐴𝐴𝑅𝑡 = 1
𝑁∑ 𝐴𝑅𝑖,𝑡 𝑁
Where 𝐴𝐴𝑅𝑡 is the average abnormal return on day 𝑡. As with the abnormal returns, the CAR’s can be averaged as well:
𝐶𝐴𝐴𝑅𝑡= 1
𝑁∑ 𝐶AR
𝑁
𝑖=1
Where 𝐶𝐴𝐴𝑅 is the cumulative average abnormal return. Finally, the t-test is used to test the CAAR for significance. Specifically, the two-sided sample t-tests from STATA is carried out. The purpose is to statistically test whether the average abnormal returns are significantly different from zero. In general a significance level of 5% is used in the t-test. Therefore, this study tests the hypotheses using the t-test with a significance level of 5%.
3.3 Descriptive statistics: Event study
Table 3 displays the press release (announcement) and implementation (effective) dates from 2005-2014. The DJSI always announces the results of the assessment in the first week of September. The implementation date (effective day) of the index inclusions or exclusions is in the fourth week of September on a Monday.
The sample in table 4 displays the number of companies, events and frequency of index additions and deletions per year. The total sample from the DJSI includes 83 firms and 302 events in the period 2005-2014. The total number of firms is considerable lower than the total number of events, because many of the same firms were added and deleted in different years. In other words, not only new companies come and go, but also old companies come in and go
Year Press Release Dates Effective Dates
2005 7-sep 19-sep 2006 6-sep 18-sep 2007 6-sep 24-sep 2008 4-sep 22-sep 2009 3-sep 21-sep 2010 9-sep 20-sep 2011 8-sep 19-sep 2012 13-sep 24-sep 2013 12-sep 23-sep 2014 11-sep 22-sep
This table shows the annoucement and effective dates of an index change for each year seperate from the period 2005-2014.
index changes have occurred in 2013, in total 41 companies were added or deleted. The majority of the years show more additions than the deletions, except in year 2010 and 2013. Remarkable is that for those two years (where the most events happen), the number of
deletions exceed that of the additions. One other interesting point for the year 2010 is that the number of deletions is extremely high. What has become apparent is that almost every
company that came in new in the DJSI in 2009 did not survive the next year. In other words, the majority of the companies that were added in 2009 to the DJSI were the next year already deleted.
All the various companies are assigned to one of the ten standard industries instead of the super sectors from the DJSI. Table 5 shows the number of additions, deletions and total events in the DJSI per industry. The majority of the events occur in the Financial Industry (14.9%). This is followed by companies from the Consumer Goods and Services industry. The telecommunication industry has by far the lowest events.
Year Number of companies Number of additions Number of deletions Total number of events
2005 92 13 12 25 2006 93 15 14 29 2007 99 15 9 24 2008 101 15 13 28 2009 116 24 9 33 2010 109 16 23 39 2011 118 18 9 27 2012 117 15 15 30 2013 116 20 21 41 2014 124 17 9 26 Total 1085 168 134 302
Table 4: Number of events per year
This table shows the amount of companies, additions and deletions and the total number of events per year.
Industry Number of additions Number of deletions Total number of events
Basic Materials 9 6 15 Consumer Goods 25 19 44 Consumer Services 23 21 44 Financials 26 19 45 Health Care 17 16 33 Industrials 23 17 40
Oil and Gas 10 7 17
Technology 15 12 27
Telecommunication 4 3 7
Utilities 16 14 30
Total 168 134 302
Table 5: Number of event per industry
This table shows the amount of additions and deletions and the total number of events per industry
The appendix presents more detailed information about the additions and deletion of the various industries per year.
3.4 Sample and Data: Panel regression
The empirical analysis of the hypothesis derived in the previous section uses panel data for a set of US firms. The set of firms in the DJSI is used to define the sample of firms to be analysed. This study choses only firms that are persistently included in the DJSI because they have made a more financial and strategic investment in CS than firms that are included just for a while. Moreover, the data is only available for a period of eight years as of 2005, therefore the time period in this analysis is shortened. And thus, the total sample consists of 44 US firms included in the DJSI during the entire period of 2005-2012. However, in the year 2010 the number of deletions from the DJSI was extremely high. Since this research only includes firms that are in the sample for the entire period, these deletions might have
influenced the sample. Therefore, to control for this, a second sample is composed of 55 US firms that are included in the DJSI during the entire time period of 2005-2009.
In developing this research, a series of accounting data is required. This data is retrieved from the Compustat database for the two time periods of 2005-2012 and 2005-2009. As first step, to obtain the accounting data for the right sample of firms (i.e., belonging to the DJSI during the entire time period), these selected sustainable firms are manually added by their unique gvkey code in the North-America annual update of Compustat. Afterwards data about the sustainability spending of those firms is collected. More specific, the advertising expenses and sales variable are gathered in order to generate the advertising intensity of firms. In addition, the data for the distance between a company’s industry and end-consumer is obtained by the identifying information in Compustat. Particularly, the GIC industries, sectors and sub-industries codes are included in the research. These Global Industry Classification Standard (GICS) codes, classify the various firms into a specific industry. Based on the descriptions of the various sub-industries from the GICS website, all the firms in the sample are logically divided into four ranked groups of distance between a company’s industry and end-consumer. In that manner, four dummy variables are created as proxy for whom industry is closer to end-consumer. Next to the dependent and independent variables data, of a number of additional variables are included based on a review of extant literature (Artiach et al., 2010; Lo and Sheu, 2007; Lopez et al., 2007; Wagner, 2010). These include firm size, profitability, leverage and R&D. As firm assets, liabilities, net income and
For both samples, data is omitted when there are no observations for certain periods. From the period of 2005-2012, 28 observations are not available and result in a total dataset of 324 observations for 44 companies. For the period of 2005-2009, 38 observations are not
available and result in a dataset of 292 observations for 55 companies.
3.5 Research Method: Panel regression
A widely used statistical method for the relation between sustainable performance and the accounting-based financial performance is the panel regression. The advantage of this model is that the estimation technique largely captures the effects of unobserved heterogeneity (Wagner, 2010). Therefore, this analysis also uses a panel regression in order to examine if the distance of a company’s industry to end-consumer (hereafter, distance) has an impact and to what extent the impact is on the CS spending. In other words, a t-test is used to statistically test the significance of both hypotheses. The proposed model includes CS spending as
dependent variable, distance as independent variable and size, profitability, leverage and R&D as control variable. The specific regression model tested is:
𝐶𝑆 𝑠𝑝𝑒𝑛𝑑𝑖𝑛𝑔𝑖,𝑡
= 𝛼0+ 𝛽1𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒1𝑖,𝑡+ 𝛽2𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒2𝑖,𝑡+ 𝛽3𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒3𝑖,𝑡 + 𝛽4𝑃𝑟𝑜𝑓𝑖𝑡𝑎𝑏𝑙𝑖𝑡𝑦𝑖,𝑡 + 𝛽5𝑆𝑖𝑧𝑒𝑖,𝑡+ 𝛽6𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡+ 𝛽7𝑅&𝐷𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦𝑖,𝑡+ ɛ𝑖,𝑡
Where 𝛼𝑖,𝑡 is a constant representing the effects of the omitted variables to the firm and constant over time. The various beta’s are the panel regression coefficients and ɛ𝑖,𝑡 is the error term, which is assumed to be an independent and identical distributed variables with a mean of zero and a constant variance.
Measures
Corporate Sustainability spending
There exist several reasons why it is arguable that the advertising intensity can be used as proxy for CS spending. Already a decade ago Drumwright (1996) observed that indeed there exist sustainable advertising and it was on the rise. More importantly, along with the
increasing media coverage of CS issues, firms themselves also take direct visible steps to promote their CS activities to various stakeholders. What this means is that sustainable advertising nowadays is a well-known and important factor in the sustainability spending of firms. Besides, both CS activities and advertising can be seen as a form of reputation building and thus are related concepts. The mature and strong evidence for a highly positive relation
