Thesis MSc Business Administration: Strategy
"
When One Plus One Does Not Equal Three"
The Elusive Effects of Combining Corporate Social
Responsibility and Corporate Political Activity.
University of Amsterdam
Faculty of Economics and BusinessStudent: P.S Zonderop (11657146)
Supervisor: Dr. P. Vishwanathan
Statement of originality
This document is written by student Pepita Zonderop 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.
Table of contents
Introduction 5
Literature Review 8
Corporate Reputation and Institutional Theory 9
Reputation as a competitive advantage 11
Nonmarket behaviors 12
Corporate Social Responsibility and reputation 13
Corporate Political Activity and reputation 15
Reputational synergies between CSR – CPA 18
Conceptual model 22
Methodology 23
Sample and Data Collection 23
Variables and Measures 24
Dependent Variable 24 Independent Variables 25 Control Variables 27 Statistical analysis 28 Results 28 Descriptive analytics 28 Exploratory analyses 32 Discussion 40 Conclusion 48 References 50 Appendix 65
Abstract
Companies are increasingly judged on their social and political behaviors, which results in reputational gains or losses. Prevailing literature argues how these two elements of nonmarket strategies influence firm performance, especially focusing on potential synergies that occur when both strategies are integrated. These studies are less certain about how corporate social responsibility (CSR) and corporate political activity (CPA) exactly influence corporate reputation. Where CSR is usually expected to positively signal stakeholders with regards to its image, CPA has a more negative association among the public. Besides, integrating in both strategies should increase reputation due to a decrease in decoupling risk. Using data from a sample of the largest firms in the U.S., this study specifies how CSR and CPA, separately and combined, help build a positive reputation. It finds that a high performance CSR, as well as significant investments in CPA, can increase firm reputation. The frequently stated synergies of combining CSR and CPA were not found within this study. These findings may possibly indicate that, the long presumed complementariness of these two nonmarket strategies does not exist for building a positive reputation.
1. Introduction
Nowadays businesses face an extensive pattern of nonmarket risk and opportunities coming from an increasingly complicated and multi-polar world (Frynas, Child, & Tarba, 2017) (e.g. PriceWaterhouseCoopers, 2016; World Economic Forum, 2016). Governments, activists, and global media have become proficient at holding companies accountable for the social outcomes of their activities. Consequently, corporate social responsibility (CSR) has emerged as a crucial concern for business leaders worldwide (M. Porter & Kramer, 2006). At the same time, businesses have started to engage more in activities which are traditionally seen as governmental activities (Margolis & Walsh, 2003; Matten & Crane, 2005; Scherer & Palazzo, 2011) Having said that, many firms already operate ‘politically’ in a more traditional sense, through interactions with governmental decision-makers; referred to as corporate political activity (CPA; den Hond, Rehbein, de Bakker, & Lankveld, 2014).
These social and political activities are strategically combined at times to make serious progress on a specific issue, yet usually they may very well be decoupled. This results in firms symbolically marketing their CSR commitments, while simultaneously lobbying for corporate interests. If revealed, these incongruent practices could lead to great reputational damage, as is seen in frequently exposed scandals. In 2015 the Policy Studies Institute (PSI) discovered that numerous multi-national firms, usually eager to promote their sustainability efforts, were also part of influential trade associations lobbying against EU climate policy (Fagan-Watson, 2015). When this issue came to light, several corporates such as BP, Procter & Gamble and Glencore were publically challenged to justify their behaviors and terminate their association memberships (Paddison, 2015). This showcases how dealing with CSR as just a corporate goal, threatens the company’s reputation and its legitimacy from institutional contexts (Anastasiadis, 2014; den Hond et al., 2014).
Recent news surrounding Mr. Cohen, President Trump’s personal lawyer, illustrates how moving in the nonmarket environment can be treacherous and requires specific skills (Frynas, Mellahi, & Pigman, 2006; Oliver & Holzinger, 2008). Last year, the Swiss pharmaceutical company Novartaris paid Mr. Cohen, $1.2 billion to help the company navigate the U.S. administration on coming reforms in the US healthcare policy (Wilkie, 2018). Novartis omitted to report it as lobbying expenditures, even though the arrangement was very similar to a lobbying agreement (Staton, 2018), resulting in a dark stain on the firm its image.
Prevailing literature on nonmarket strategies and their competitive performance outcomes is highly isolated and disintegrated into a separate social and political domain (Frynas & Stephens, 2015). Evidence of CSR influencing corporate performance through reputation building or financial measures is contradictory (Griffin & Mahon, 1997; Porter & Kramer, 2006; Schnietz & Epstein, 2005). Only a limited amount empirical studies discuss the impact of CSR on reputation (Brammer & Pavelin, 2006; Turban & Greening, 1997), and even less literature has investigated the relation between reputation and CPA. The need for integration of these two concepts has long been called upon (Baron, 2001; McWilliams, Fleet, & Cory, 2002), although just recently scholars have started to explore this integration (den Hond et al., 2014; Frynas & Stephens, 2015; Mellahi, Frynas, Sun, & Siegel, 2016). It is argued that through deliberately combining CSR and CPA strategies, both can become more effective and consequently enhance reputation. This can happen through signals of reliability and trustworthiness (den Hond et al., 2014), or by reducing the decoupling risk (Marquis & Qian, 2013)). Empirical evidence is yet to support this idea. Academic insights on the reputation effect of these increasingly common behaviors could help firms shape their competitive environment, while maintaining a good relationship with stakeholders.
both activities. In order to examine these relationships this research looks at a sample from the 500 largest U.S. firms from 2014 to 2016 with regards to their social and political performance. Through the lens of signaling theory and institutional theory this paper will seek an answer to the following research question:
What is the influence of combining corporate social responsibility activities and corporate political activities on the reputation of a firm?
The findings of this paper suggest the combination of CPA with CSR does not generate any additional effects on corporate reputation. In our sample, the stand-alone effect of a high social performance on reputation was found to be positive. This was especially true for the Human Rights dimension of CSR. Additionally, the relationship between the firm’s political efforts and reputation seems to work in a directly opposite direction from what was expected. The analysis indicated that CPA to positively influence corporate reputation, rather than negatively.
This paper will makes three contributions to the current field. To start with, it addresses the gap on how CSR contributes to firm performance. It does so by extending the current knowledge the valuable role of CSR in reputational frameworks with regards to signaling trustworthy characteristics towards stakeholders. Moreover, it enhances understanding of the specific CSR dimensions that contribute positively to reputation. This should help managers in their decisions of which CSR activities to focus on for a beneficial corporate image. Furthermore, this paper provides a fresh perspective on the reputational effects of political activities in modern day. Despite increasing public concerns that CPA is manipulative or even egocentric (Ashforth & Gibbs, 1990; Liedong, Ghobadian, Rajwani, & O’Regan, 2015), the results indicate that the prevailing negative associations of CPA may not harm firm reputation. Finally, the study adds to the field of nonmarket strategies by presenting an, empirically grounded, opposing claim on the conceptualized potential synergies of integrating CSR and CPA. While several theoretical
Hond et al., 2014; Liedong et al., 2015; Mellahi et al., 2016), this paper indicates the puzzle of integrating strategies may be more difficult in practice. These contributions are relevant to fill the lack of insight, and moreover to help business managers get a better understanding of how to organize their nonmarket strategies in ways to best benefit the firm and society. Shining a light on reputational effects of social and political pursuits will hopefully put current managerial behaviors in a new light and encourage them to take an integrated approach on both strategies for an increased corporate image.
This paper begins by examining the prevailing literature on the formation of corporate reputation, focusing on signaling theory and the institutional theory perspective. The literature chapter will then move on to discuss nonmarket strategies, of which corporate social responsibility and corporate political activity in special. To answer the research question it then dives deeper into the potential synergies that arise from the combination of CSR and CPA. The third section describes the methodology used by this paper to research our question. It gives an overview of the operationalization of the variables and explains the analytical strategy. The results of this methodology are presented in the fourth section of this paper. A discussion of these results follows after and any problems regarding the method are addressed, along with some suggestions for future research. The conclusions of this research are drawn in the final section.
2. Literature review
This section discusses the main insights of the existing literature on corporate reputation and nonmarket strategies, where the concepts of corporate social responsibility and corporate political activity are further outlined and the hypotheses are presented. First, it introduces the dominant perspectives on corporate reputation by applying the principles of signaling theory
followed by the first two hypotheses. Afterwards, the literature gap on synergies between CSR and CPA is addressed and the last hypothesis is introduced.
2.1 Corporate reputation
The generally accepted definition of the term reputation is “a perceptual representation of a firm’s past actions and future prospects that describe the firm’s overall appeal to all its key stakeholders when compared to other leading rivals” (Fombrun, 1996: 72). This definition emphasizes three fundamental attributes to reputation. Firstly, reputation is based on perceptions. Secondly, it is the accumulated perception of all stakeholders, and, thirdly, this evaluation is comparative (Walker, 2010; Wartick, 2002). More recent literature proposes two additive attributes; corporate reputation can be positive or negative (Brown, Dacin, Pratt, & Whetten, 2006; Mahon, 2002; Rhee & Haunschild, 2006) and reputation is stable and enduring (Gray & Balmer, 1998; Mahon, 2002; Rhee & Haunschild, 2006; Roberts & Dowling, 2002). According to the signaling theory, stakeholder impressions are formed through strategic choices of organizations that serve as informative signals on which stakeholders evaluate that organization (Basdeo, Smith, Grimm, Rindova, & Derfus, 2006; Fombrun & Shanley, 1990; Turban & Greening, 1997; Walker, 2010). They enable stakeholders to form an opinion on the firm’s ability to create value (Clark & Montgomery, 1998; Rindova & Fombrun, 1999) and indicate the quality of a firm’s products and services (Shapiro, 1983). Stakeholders can receive direct and indirect signals concerning an organization’s behaviors, either directly from the company itself or through information channels like the media or stock market (Brammer & Pavelin, 2006). The delay in competitors’ responses to a firm’s actions and the similarity of the actions with those of the competitors also functions as a signal to evaluate reputation (Basdeo et al., 2006). The more information stakeholders get, the easier it becomes to make reputational evaluations (Dowling & Moran, 2012).
Firms can develop a high reputation by demonstrating consistent behaviors that produce valued and recognized outcomes from a stakeholder perspective (Pfarrer, Pollock, & Rindova, 2010). Constituencies are particularly interested in the suitability of an organization as an exchange partner. Firms’ actions that signal trustworthiness and credibility are especially relevant for improving reputation, as these provide a basis for predicting future firm behavior (Fombrun, 1996; Fombrun & Van Riel, 2004; Love & Kraatz, 2009). Actions perceived to hint towards opportunism, unreliability, or lack of integrity will damage reputations (Love & Kraatz, 2009). It can therefore be recommended for firms to focus on signals that suggest trustworthiness when trying to build a positive reputation.
2.1.2 Institutional Theory
The institutional theory focuses more on the context of building reputations, which happens before the phase of actively sending signals (Walker, 2010). This theory claims organizations are positioned within broader institutional contexts and focuses on the corresponding cultural processes in these environments (DiMaggio & Powell, 1983; Meyer & Rowan, 1977). It highlights how organizations should endorse certain structures and take actions in response to field-level pressures, obtaining legitimacy and support in return for their compliance (Westphal, Gulati, & Shortell, 1997). Stakeholders, from an institutional context, will assess corporate actions as symbolic indicators of the cultural fit of a firm and will alter their reputational evaluation correspondingly (Love & Kraatz, 2009). An organization is believed to be legitimate to the degree that its means and ends are in line with social norms, values, and expectations (Ashforth & Gibbs, 1990). Without making appropriate endorsements and taking appropriate actions, effort toward building a reputation may become be futile (Walker, 2010).
institutional environment to build reputations (Deephouse & Carter, 2005; Love & Kraatz, 2009; Rao, 1994; Staw & Epstein, 2000).
2.1.3. Reputation as a competitive advantage
A vast amount of literature has shown the importance of reputation in relation to competitive advantage (Basdeo et al., 2006; Fombrun, 1996; Fombrun & Shanley, 1990; Roberts & Dowling, 2002; Turban & Greening, 1997; Walker, 2010). A good reputation can lead to various strategic advantages, such as the ability to charge premium prices (Deephouse, 2000; Fombrun, 1996; Fombrun & Shanley, 1990), attract and retain customers (Fombrun, 1996) and talented employees (Bhattacharya, Sen, & Korschun, 2008; Turban & Greening, 1997), provide better access to external financiers (Engelen & Essen, 2012), and increase profitability (Roberts & Dowling, 2002).
Moreover, reputation is considered to offer a sustained advantage, as scholars consider a good reputation as difficult or even impossible to replicate in the short term (Dierickx & Cool, 1989; Roberts & Dowling, 2002; Rumelt, 1987). Additionally, a strategy-based reputation offers a better position for providing sustainable competitive advantage (Dowling & Moran, 2012). Strategy-based reputations are an integral part to the organization’s DNA (i.e. ‘built-in’) and therefore tend to be significantly different from those of the competitors. This type of reputation stands in contrast to those based on tactical initiatives, like ad hoc philanthropy and non-strategic CSR programs (Dowling & Moran, 2012). These can be categorized as being ‘bolted on’ to the core operations of the organization, as they mostly do not reflect the fundamental objectives of the organization. These ‘bolted on’ activities sends conflicting signals about the firm’s purpose, thus negatively effecting reputation in the long run. Moreover, it is unlikely they provide sustainable advantage, since they are easily copied by competitors.
The findings of Dowling & Moran (2012)point out the importance of building a sustained reputation by engaging in strategy-based activities, also with regards to philanthropic behaviors. Not surprisingly, reputation is the single most valued organizational asset (Gibson, 2006) and it is becoming increasingly important in competitive markets (Vallaster & Abimbola, 2007).
2.2 Nonmarket strategies
Besides traditional firm stakeholders, governments, activists and media are becoming more skilled at holding firms accountable for the social outcomes of their activities (Doh, Lawton, & Rajwani, 2012). Consequently, nonmarket strategy is becoming an inevitable priority for business leaders around the world (Baron, 1995). Even more so, since nonmarket forces can influence economic as well reputational performance (Lock & Seele, 2016). Therefore, previous research argues and demonstrates that an effective nonmarket strategy is of critical importance to a company’s survival, performance, and sustainable competitive advantage (Baron, 2001; Frynas et al., 2006; McWilliams et al., 2002; McWilliams & Siegel, 2001; Mellahi et al., 2016; Oliver & Holzinger, 2008).
Nonmarket strategies consist of the firm’s collaborative composition of actions aimed at improving its performance by directing the institutional and/or social context of economic competition (Baron, 1995). In contrast to the market environment, this context can be characterized mainly by all the social, cultural, legal, and political forces that either restrain or ease firm activities. Where firms compete in the market environment for resources, revenues and profits, they regard more extensive dimensions of performance in the nonmarket environment, such as ethical behavior, policy realization and social responsibility (Doh et al., 2012). To influence the nonmarket environment, many firms engage in CPA and/or CSR (Doh et al., 2012; Liedong et al., 2015; Mellahi et al., 2016). These have thus far been studied as two
to corporate goals, threaten a firm’s image and reputation, and above all prevent organizational synergies (Anastasiadis, 2014; den Hond et al., 2014; Liedong et al., 2015). Both concepts and their effect on reputation, solitary and combined, are explained in the following sections.
2.3 Corporate Social Responsibility
Managers continually encounter conflicting demands from multiple stakeholder groups to devote resources to CSR. These different goals and objectives are complicating a clear definition of CSR. The traditional, broadly accepted definition sees CSR as “voluntary actions that appear to further some social good, beyond the interests of the firm and that which is required by law” (Carroll, 1979; McWilliams & Siegel, 2001). Though CSR is often seen as an aggregate concept, it is essentially multi-dimensional and symbolizes a large variety of corporate behaviors with respect to its resources, processes, and outputs (S. J. Brammer & Pavelin, 2006; Carroll, 1979; Waddock & Graves, 1997; Wood, 1991). There is a large audience for corporate well-doing, as the two foremost sources of CSR demand come from consumers and from other stakeholders, like investors, employees, and local communities (McWilliams & Siegel, 2001). The performance on social responsibility can be explained as the degree to which a firm’s principles, processes, policies, programs, and observable outcomes relate to its societal relationships (Orlitzky, Schmidt, & Rynes, 2003; Wood, 1991). For many years, researchers have mostly focused on highlighting this relationship between CSR and corporate financial performance (CFP; Orlitzky et al., 2003; Schnietz & Epstein, 2005; Waddock & Graves, 1997). Still, the finding of these studies have been largely inconsistent (Melo & Garrido-Morgado, 2012).
The concept of CSR was introduced in this paper as being a nonmarket strategy, although some caution concerning this typology is appropriate. CSR may be imbedded in the nonmarket context of the organization for the most part, as firms’ social activities are traditionally aimed
at improving the relationship with the social, legal, and cultural institutions. Yet, certain aspects of CSR have come to be a fundamental part of the firm as they can be beneficially valued in the marketplace (Baron, 1995), for instance, as a marketing feature or as a driver for innovation. From an economic perspective, CSR can be regarded simply as a profit-maximizing strategy (McWilliams & Siegel, 2001) to alter the competitive position of the firm (Baron, 2001). However, reflecting on CSR from an institutional perspective, this paper argues that CSR is driven by the need to respond to external demands, such as stakeholder expectations (Mitchell, Agle, & Wood, 1997), that require them to engage in socially acceptable and legitimate activities (Campbell, 2007; Chiu & Sharfman, 2011). This paper will therefore consider CSR as a nonmarket strategy, since the original purpose of CSR comes from a philanthropic, non-economic perspective and this still provides the core values upon which stakeholders evaluate a company’s CSR performance.
2.3.1 Reputational effect of CSR
It is generally thought that CSR also contributes positively to corporate reputation, which is seen as a driver of superior financial performance (S. Brammer & Millington, 2005; Melo & Garrido-Morgado, 2012; Orlitzky et al., 2003; M. E. Porter & Kramer, 2002; Roberts & Dowling, 2002; Turban & Greening, 1997). To begin, social responsiveness can be seen as impacting the welfare of key stakeholders, hence sending an information signal that the firm is striving towards a reciprocal relationship with potentially influential stakeholder groups (Fombrun & Shanley, 1990) and that it stands apart from its competition. This signal, consequently, improves the firm’s reputation (S. J. Brammer & Pavelin, 2006; Turban & Greening, 1997), for example, from consumers’ positive evaluations of the organization and its outputs (Aguinis & Glavas, 2012; Brown & Dacin, 1997; Sen, Bhattacharya, & Korschun,
Furthermore, CSR can also serve as an insurance mechanism. A strong reputation for CSR will allow firms to better maintain their value after negative events (Godfrey, Merrill, & Hansen, 2009; Schnietz & Epstein, 2005). It is argued that, when negative events happen, stakeholders will keep in mind their positive perception of the firm when considering the negative effects of the act (Godfrey et al., 2009); while believing the damaging events are probably caused due to bad luck rather than bad management (Minor & Morgan, 2011). Likewise, a firm can use social activities to restore its reputation after carrying out illegal, damaging acts (Williams & Barrett, 2000). In consumer-oriented companies this often translates to cause-related marketing practices, while in stigmatized industries, like chemicals and energy, firms pursue CSR as a form of insurance. These firms hope that a good reputation on social issues will temper public criticism in the event of a crisis.
As summarized above, a good performance on CSR can work to build and protect corporate reputation (Liedong et al., 2015; Minor & Morgan, 2011; Park et al., 2014), hence leading to the first hypotheses of this paper:
Hypothesis 1: A high Corporate Social Responsibility has a positive effect on Corporate Reputation
2.4 Corporate Political Activity
The other strand of nonmarket strategy is Corporate Political Activity (CPA), which refers to corporate efforts to manage political institutions and influence political actors in beneficial ways for the firm (Hillman et al., 2004;Lux et al., 2011; Baysinger, 1984). This can be either for defensive reasons, , in the case that an imminent government decision could threaten a firm’s competitiveness (Getz, 1997),or, it can be for pro-active reasons, in the case that a firm is pursuing an opportunity to influence regulatory or legislative processes towards an outcome that better reflects the internal goals of the firm (Baysinger, 1984). Different behaviors can be
used to convey corporate preferences to policymakers: information provision, financial contributions, or constituency building (Hillman & Hitt, 1999). This primarily translates to tactics like political action committee (PAC) contributions and lobbying (Doh et al., 2012). A company’s political actions shape their competitive space and facilitate them to exploit economic activities (Capron & Chatain, 2008; McWilliams et al., 2002). In general, the goal of engaging politically as a firm can be either an aspiration to pursue the organization’s private interest or to control and shape public policy for defensive or conservative reasons (Baines & Viney, 2010). Possible beneficial results for the organization may involve reduced transaction cost, reduced environmental uncertainty and enhanced long-term sustainability.
Governmental policy and business interests are ever closely connected and have been regarded at times as controversial (MacGillivray, Raynard, & Zadek, 2005). Although some organizations use lobbying as a driver for good (Alzola, 2013; Scherer, Palazzo, & Seidl, 2013), too often corporate promises to sustainability are disregarded in place of lobbying for commercial interest (MacGillivray et al., 2005). For organizations, the leading perspective on political activity is that it should maintain or add value to its business (Baron, 2001; Bonardi, Holburn, & Bergh, 2006; Frynas et al., 2006). Naturally, the relationship between CPA on firm performance is a vastly growing research topic(Adhikari, Derashid, & Zhang, 2006; Claessens, Feijen, & Laeven, 2008; De Figueiredo & Silverman, 2006; Liedong et al., 2015). However, this relationship is rather elusive. Among the recent empirical papers more than half did not find a positive effect of CPA on firm performance (Mellahi et al., 2016). For instance, Hadani & Schuler (2013) find that a firm’s political investments are actually negatively related to both their market- and accounting performance.
2.4.1 Reputational effect of CPA
As empirical research on CPA and its subsequent performance is relatively new, literature on the relation of political activity on corporate reputation is even more limited. CPA is generally considered to have a ‘legitimate and important role to play in the public policy process’, even if it is ‘not always a welcome reality in western politics’ (Coen, 2007. p.334).
On a darker side, the political environment is one where corruption and immoral practices can easily prevail. This is the outcome of a profit-seeking mindset often found in nonmarket strategy, which leads from legal forms of lobbying to illegal forms of bribery and exploitation (Doh et al., 2012). Accordingly, the concerns are growing about firms crossing the line of legitimate political participation and the opportunistic pursuit of self-interest. The general thought being that the power balance between business and government is lost (Drutman, 2015), as well as the balance between democratic power in between businesses (Roe, 2010). Political actions of firms have been well-known to suppress competition (Robertson, Gilley, & Crittenden, 2008) by means of triumphant agitations for trade protecting and anti-dumping protocols (Evans & Sherlund, 2011; Lee & Baik, 2010; Liedong et al., 2015). In emerging and developing countries with weak and inexperienced institutional environments, CPA regularly involves corruption (Doh et al., 2012), widespread use of informal connections for organizational benefits (Lawton, McGuire, & Rajwani, 2013), and cronyism (Gul, 2006). Consequently, engaging in any of these socio-political actions will create mistrust and damage reputation (Liedong et al., 2015).
In recent years, firms are increasing their lobbying expenditures as part of their CPA strategies, even though these practices have a negative sentiment to them. Lobbying is often seen as the involvement of backroom deals among powerful, corrupt companies pursuing ‘special interests’ (Liedong et al., 2015; Mills, 1956 and countless media reports). From this view, many believe that CPA is ever self-serving, or even manipulative; benefitting only the firms that
pursue it (Ashforth & Gibbs, 1990; Liedong et al., 2015). As a matter of fact, certain political maneuvers, such as PAC contributions, are considered by the public as enticements to influence political processes and provide unjustifiable political power to businesses (den Hond et al., 2014; Milyo, Primo, & Groseclose, 2000). On the other hand, firm’s political efforts can be disguised from the stakeholders’ view, which prevents any reputational effect. Though, when the activities are revealed the potential negative effect on firm reputation can be more severe (den Hond et al., 2014).
Precisely this usual lack of transparency concerning a firm’s political efforts is another factor of CPA undermining corporate reputation, whether organized individually or collectively (den Hond et al., 2014). According to MacGillivray et al., (2005), it is this lack of transparency of firms involved in the political arena that notably contributes to a negative perception of CPA. Building on the arguments described above, this study hypothesizes the following relationship:
Hypothesis 2: A high Corporate Political Activity has a negative effect on Corporate Reputation
2.5 Combining CSR and CPA strategies
In the past decade, research on CSR and CPA has emerged predominantly in isolation, although there is an overlap between the social and political aspects of firm strategies (Mellahi et al., 2016). Engaging in the two nonmarket strategies without integration can be highly problematic. From an institutional theory perspective firms are looking to acquire legitimacy from their institutional environments. Yet, social values and expectations are frequently changing and contradictory, which makes it difficult to operationalize. Moreover, there are usually ambiguities and inconsistencies in their communication through laws, media and interest group campaigns (Ashforth & Gibbs, 1990). For large and complex organizations this entails having
to answer numerous scattered stakeholders with conflicting expectations and perceptions (Zald, 1978).
Such inconclusive legitimacy pressures can lead to decoupling processes whereby the organization’s responses to external demands vary in the degree to which they are symbolic or substantive (Ashforth & Gibbs, 1990; Marquis & Qian, 2013; Meyer & Rowan, 1977; Okhmatovskiy & David, 2012; Oliver, 1991). This symbolic conformity with stakeholder demands without creating substantive changes is called decoupling (Marquis & Qian, 2013; Meyer & Rowan, 1977). It is a useful strategy for firms, since it enables them to sustain legitimating, formal structures, while their activities change in response to practical concerns (Meyer & Rowan, 1977). Decoupling can be linked to social and political firm behavior in the following ways.
As mentioned above, managers continually encounter contradictory demands with regards to CSR behaviors. Instead of actually changing its CSR behaviors, the organization might solely portray them symbolically in order to appear in harmony with social values and expectations (Ashforth & Gibbs, 1990; Westphal & Zajac, 1995). A suitable example for this can be found in the insurance strategy for CSR,. In this case, the firm is only trying to change stakeholder perceptions in a favorable way, rather than truly making an effort to change its way or to stop future social scandals from happening.
It should be noted that not all CSR behaviors can be assumed to be symbolic. Evidently, firms can also engage in CSR in a substantive way by thoroughly changing business practices. Then again, for larger firms with complex internal processes this is a challenging approach, uncertain of financial returns (Wickert, Scherer, & Spence, 2016). Although, big corporations generally receive significant visibility and stakeholder scrutiny with regards to CSR (Chiu & Sharfman, 2011), assessing substantive social actions is awfully difficult for stakeholders (Siegel & Vitaliano, 2007). These two rationales reduce the motivation for organizations to invest in CSR
substantively. Therefore, this paper assumes that large firms are more inclined to invest in a symbolic impression of their social performance to satisfy demands from stakeholders (Perez-Batres, Doh, Miller, & Pisani, 2012).
On the other hand, this paper views CPA as substantive behavior towards a firm’s institutional environment, as it entails significant changes in business practices and even requires some risk-taking behavior in order to achieve real improvements in the firm’s performance (Ashforth & Gibbs, 1990). Firms mostly engage in political activity in order to maintain or add value to the organization (Baron, 2001; Bonardi, Holburn, & Bergh, 2006; Frynas et al., 2006). Deciding to devote resources to lobbying or financially contribute to PAC’s involves a real, material change in the organizations goals and processes (Ashforth & Gibbs, 1990). By engaging in these activities the firm is attempting to bring prevailing opinions and institutionalized laws and practices into conformity with its own goals (Ashforth & Gibbs, 1990). While doing this, the firm is more concerned with achieving the appointed goals than with trying to change the perceptions of stakeholders in a favorable way.
2.5.1 Reputational effect of combining CSR and CPA strategies
An increasing number of studies have started to theoretically explore the possible overlap of
these two corporate activities and assessed the impact of alignment between and within CSR and CPA activities with regards to performance (den Hond et al., 2014; Frynas et al., 2017; Liedong et al., 2015; Mellahi et al., 2016).
It is argued that complementary resources from CSR can support CPA and vice versa (den Hond et al., 2014). A good CSR reputation can boost the firm’s political pursuits by facilitating access
to political entities (den Hond et al., 2014; Gao & Hafsi, 2017; H. Wang & Qian, 2011),
knowledge (den Hond et al., 2014; Yaziji, 2014), and reducing the potentially unwarranted reputational effects of CPA (Sun, Mellahi, & Wright, 2012). Alternatively, strong political ties can support CSR activities by helping firms select salient social issues and problems due to ‘social and political intelligence’ from politicians and administrators (Post, Dickie, Mahon, & Murray, 1983), enhancing the feasibility of CSR strategies by acquiring political and regulatory support through CPA (Post et al., 1983), and increasing the credibility of CSR commitments through association with authoritative external parties (den Hond et al., 2014; Jedrzej George Frynas et al., 2017; Peterson & Pfitzer, 2009). Selecting more social salient issues and increasing the effectiveness of working on them should increase corporate reputation due to signaling an enhanced cultural fit with the institutional context.
Moreover, reputation damages can occur when following a decoupling strategy. Organizations that try to pursue all demands from the institutional environment, without integrating them into one consistent strategy, run the risk of ‘protesting too much’ and consequently being marked as manipulative and illegitimate (Ashforth & Gibbs, 1990). Organizations also face a ‘decoupling risk’ (Marquis & Qian, 2013) when exposed as responding only with symbolic changes, without adjusting substantive practices. This may possibly harm the firm’s reputation. If discovered, decoupled actions result in stakeholders retracting their support and loyalty, which will reduce a firm’s legitimacy and ultimately even its financial performance (Schons & Steinmeier, 2015; Walker and Wan, 2012). For that reason, it would not be advisable to focus solely on symbolic behaviors when trying to build a sustained competitive reputation.
Consequently, through engaging in symbolic behavior (CSR) as well as substantive behavior (CPA) organizations decrease the risk of decoupling exposure. In other words, combining social and political practices should increase corporate reputation as the degree of incongruent and opportunistic signals significantly decreases. Joining CSR and CPA resources to work on the
same social topics is expected to boost the sense of reliability and trustworthiness towards the stakeholder environment.
Building on the theory of den Hond et al., (2014) and the perspectives from institutional theory and signaling theory, this paper investigates the complementariness of CSR and CPA, particularly with regards to improving credibility, reliability, and trustworthiness, which subsequently improve reputation (Liedong et al., 2015). Hence, leading to the following hypothesis:
Hypothesis 3: When Corporate Social Responsibility and Corporate Political Activity are strategically combined, it will
increase Corporate Reputation
2. 6 Conceptual model
Figure 1 presents the overall conceptual model of the individual reputational effect of CSR and CPA respectively, and the combination of both, as specified above. All three models contain three control variables: firm size, corporate financial performance and industry.
Figure 1: Conceptual models with variables CSR, CPA and corporate reputation H1
H2
CSR Corporate Reputation
CPA
• Intensity of lobbying • Total PAC donations
3. Methodology
This chapter outlines the research approach and design used for this paper. It starts with discussing the sampling and data collection strategy. Then, the chapter continues with operationalization of the dependent, independent and control variables. The last section includes the models used for data analysis.
3.1 Sample and Data Collection
The sample of this research consists of the largest publicly listed firms from the United States in the years 2010 to 2016. This sample was chosen due to the fact that these types of firms are obligated to have all company information publicly available, which allows us insight into their financials, strategies and other performances. It also ensures that these firms are comparable to each other in terms of size and resources. Furthermore, US companies have to make political spending data publicly available according to national regulations. Moreover, the MSCI ESG KLD STATS database, that is used to measure CSR, mainly reports on US publicly listed firms. Since we need data on both CPA and CSR for all companies, the chosen sample offers the largest data availability. This motivation for the sample used is in line with various studies on CPA and CSR (Aggarwal, Meschke, & Wang, 2012; Hadani & Schuler, 2013; Walker, 2010). Data was obtained using secondary datasets from third-parties. The dataset consists of about 635firms that were all listed in the Fortune 500 ranking for the chosen years and evaluated by the Kinder, Lydenberg, Domini (KLD) firm. Corporate reputation scores were extracted from the Fortune’s Most Admired Companies (MAC) database, which is used by the majority of studies on firm reputation (Walker, 2010). All information about CSR performance was gathered using the MSCI ESG KLD STATS database, following prior research performed by CSR scholars (S. J. Brammer & Pavelin, 2006; Griffin & Mahon, 1997; Turban & Greening, 1997; Waddock & Graves, 1997). Data regarding CPA is obtained through the website of the Center for Responsive Politics, called Open Secrets. This website stores data on firm campaign
contributions and lobbying behavior. Using this data source, this paper follows many previous studies on CPA (Hillman et al., 2004). To finish, financial information for all firms was gathered using the COMPUSTAT database, which is a frequently used database in CSR and CPA papers (Fombrun & Shanley, 1990; Godfrey et al., 2009; Schuler, Rehbein, & Cramer, 2002).
All data retrieved was from the years 2010 to 2016 inclusive. For the final sample some firms had to be excluded from population as too much data was missing, for instance due to mergers, acquisitions, bankruptcy and various other reasons. Data from 2014 to 2016 proved to be most comprehensive for all variables and thus this data was used for the final dataset. After the data collection and removal of incomplete entries, the final sample consisted of 543 firms.
3.2 Variables and measures 3.2.1. Dependent variables
Corporate Reputation. To operationalize reputation, scores from the Fortune’s Most Admired Companies (FMAC) ranking are used. By doing so, this study follows many others (Basdeo et al., 2006; Fombrun & Shanley, 1990; Fryxell & Wang, 1994; Walker, 2010). The FMAC represents an aggregate perception on one firm for a specific year, measured by the perceptions of executives, directors, and financial analysts from other companies in the same industry (Walker, 2010). These perceptions are based on nine criteria that together form an overall score of the firm’s reputation, ranking from one to ten. For the analysis, the reputation scores from 2015 to 2017 were taken per firm, since data for the year 2014 was missing in the database. The year 2017 is also included, as reputation literature states that it building a reputation takes time and stakeholder perceptions are built on the firm’s past actions (Fombrun, 1996; Walker, 2010). To operationalize this data, an average of the reputation scores of those years is taken, as this
provides the most inclusive measure with regards to certain firms that were missing values for one or more of these years.
3.2.2. Independent variables
Corporate social responsibility. To measure corporate social responsibility activities, ratings from the MSCI ESG KLD STATS are used. This is the largest multidimensional corporate social performance database available to the public and has become the standard for quantitative measurement of CSR (Mattingly & Berman, 2006). The KLD is an independent rating service that exclusively works on assessment of CSR across a range of dimensions related to stakeholder concerns (Waddock & Graves, 1997) and is based on objective measures (S. J. Brammer & Pavelin, 2006). The investigated dimensions are the five issues of CSR, being Environment, Community, Human Rights, Employees and Products. These dimensions consist of multiple indicators of ‘strengths’ and ‘concerns’ that are assessed using binary codes. All indicators scored with a 1 are seen as present, and all indicators scored with a 0 are seen as not present. These ones and zeros add up to a total strengths and weaknesses score per dimension, where weaknesses are subtracted from strengths (Chin, Hambrick, & Treviño, 2013). The resulting CSR score per dimension indicates the quality of the firm’s corporate social performance in this dimension and in that specific year.
Unfortunately, most dimensions suffered from missing data in various indicators. Most of these were strength indicators, and, remarkably, almost none were weakness indicators. Deleting all indicators with missing data, therefore, turned out to be a non-option, as it would produce a distorted image of the firms’ CSR performances, leaning towards the negative end of the spectrum. The first operationalization method of this paper uses all data available and calculates a total CSR score per dimension by subtracting the total weaknesses score from the total
strengths score (Chin et al., 2013). By summing the CSR score of all dimensions in a given year, a total CSP score per year is formed.
The second operationalization that is used in this analysis is based on ratios. This allows for an easier comparison among the companies in the dataset with regards to judgment on strong and weak CSR. Here, CSR strengths and weaknesses performance is calculated by dividing the obtained score by the maximum achievable score. The weaknesses ratio is then subtracted from the strengths ratio to create a total score on the dimension. Again, the total CSR per year is calculated by summing overall CSR dimension scores in that year. Finally, the average of CSR performance scores is taken over the years 2014 to 2016 for all firms.
CPA activities. To operationalize a firms political activity, data is used on the total political action committees (PAC) contributions and total lobbying expenses of all firms in the sample. Following Hadani & Schuler (2013), these are summed to compute a corporate political investment measure. PACs are committees that gather data from corporations and individuals and allocate this to various political parties, candidates, and other politically oriented sources. Donations to PACs are a means for firms to promote their political agendas in an indirect way. The data on firm PAC contributions is obtained from the OpenSecrets website. The total height of the bids is measured to get a sense of the company’s CPA engagement. Analyzing PAC data is a common measure in CPA research (Aggarwal et al., 2012; Cho, Patten, & Roberts, 2006;
Hadani & Schuler, 2013). Additionally, CPA is measured through the total lobbying expenses.
Firms spend money on hiring professional campaigners to lobby with political players on issues that the firm wants to have some bearing on. Firm-specific lobbying expenses are also collected from OpenSecrets, containing the overall expense made on lobbying per year. The CPA data represents PAC contributions for two election cycles (2014 and 2016) and lobbying data for
each year from 2014 until 2016. The average of the total corporate political investment was determined following the methodology of Hadani & Schuler (2013).
3.2.3. Control variables
Several factors are controlled for as they potentially influence the corporate reputation score. Firm size. As firms grow as a result of some effectiveness in their corporate history, larger firms can be expected to have a better reputation than their smaller rivals (S. J. Brammer & Pavelin, 2006). Moreover, larger firms are known to have higher CSR scores (Waddock & Graves, 1997), have additional resources available to use for CPA (Hillman et al., 2004), and have a higher visibility, which can influence reputation (Fombrun & Shanley, 1990). Therefore, firm size is controlled for by using a logarithm of total employees (Wang, Dou, & Jia, 2016). CFP. Strong corporate financial performance usually indicates an effective corporate strategy, good management, and good resource allocations. It is therefore expected to help a firm establish or maintain a good reputation (Roberts & Dowling, 2002; Sabate & Puente, 2003). Financial performance is operationalized by using the COMPUSTAT data on return on assets (RoA), which is a ratio of net income to total assets. This is a commonly used measure for firm performance (Barnett & Salomon, 2012; Waddock & Graves, 1997).
Industry. Reputation may also vary systematically across sectors. Meaning that some core business activities may incline a firm to a better reputation than do other activities (Wang et al., 2016). Previous research also controls for industry, since a specific industry can have certain reputational benefits or tensions for CSR and CPA with regards to reputation (Cho et al., 2006; Hillman et al., 2004; T. Wang & Bansal, 2012). Hence, the variable industry orders all firms in the sample in respective industry groups, based on the four digit SIC codes. These are operationalized by building dummy variables for all 10 different industry groups. Eight of these industries are included in the analysis, as only these eight were represented in the final sample.
3.3. Statistical Analysis
The above-mentioned hypotheses are analyzed using statistical analysis software “SPSS Statistics 24” from IBM. The first two hypotheses are tested using single linear regression analysis. These specify if a high CSR has a positive effect on corporate reputation, and if CPA has a negative effect on corporate reputation. In both analyses, firm size, industry and corporate financial performance will be controlled for. The third hypothesis will be examined using the PROCESS add-in for SPSS, developed by Hayes (2013). This software is helpful for testing moderation, because it automatically mean-centers variable scores and corrects for heteroscedasticity. Model 1 of PROCESS is used to test whether a high CPA has a moderating effect on the effect of CSR on corporate reputation, while controlling for firm size, industry, and financial performance. These three analyses are performed with both operationalization methods of CSR. The results are presented in the next chapter.
4. Results
This chapter shows the results of the analyses performed in this research. It starts with portraying the descriptive statistics and correlation for all variables of the study to get an overview of the data. Afterwards, the findings of the multiple regression analyses are reported in order to test the hypotheses of this paper. Each section is divided by means of operationalization of CSR.
4.1 Descriptive statistics and correlation analysis
Table 1 and 2 portray the descriptive statistics and correlations for all variables used to test the hypotheses of this paper, while respectively using the first and the second operationalization of CSR.
The average reputation score in this first sample is around 6.50, with Walt Disney performing best at an 8.64 maximum value and Chesapeake Energy performing worst with the minimum score of 4.40. With regards to the average total CSR score of 1.44, Microsoft scored highest with a total of 8.33 and Monsanto scored lowest with a striking minimum of -4.00. The mean corporate political investment of $1.9 billion almost seems like nothing when compared to the maximum investment of $20.7 million by Boeing. However, there do appear to be many firms that spent zero dollars on political activities, such as lobbying and PAC’s.
The results in Table 1 show that the strongest correlation is between CPA and corporate reputation with a Pearson correlation coefficient of r = .342; p < .01. In other words, firms with a high CPA also seem to have a high reputation. Another strong correlation can be seen between firm size and reputation with a coefficient of r = .163; p < .05. Looking at these results, CSR seems to be less correlated to reputation than CPA, but still with a significant (p < .05) correlation coefficient of r = .130. Table 1 shows that all the significant correlations have a small to medium effect, with none reporting a large effect over r = .50 (Cohen 1988, 1992). The average reputation score in the second sample is the same as in the first sample. Due to different operationalization, the second sample reports an average total CSR score of 0.44. Alphabet scored highest with a total of 2.73 and Monsanto scored lowest with a striking minimum of -1.41. Descriptive statistics for CPA were almost completely similar to those in the first sample. The results in Table 2 also show that the strongest correlation is between CPA and corporate reputation, with a Pearson correlation coefficient of r = .338; p < .01. In contrast to the first sample, this sample reports the second strongest correlation between CSR and reputation, with a coefficient of r = .250; p < .01. This indicates that, in the second sample, CPA is still more strongly correlated to reputation, but CSR proves to be of comparable strength. In Table 2 all significant correlations have a small to medium effect (Cohen 1988, 1992).
T abl e 1: M ean s, S tan dar d D ev iat ion s, C or re lat ion s Me an SD 1 2 3 4 5 6 7 8 9 10 11 12 13 1. 44 1. 91 - 192 2.55 318 1. 87 .121 ** - 6. 50 0.67 .130 * .34 2** - 1. 42 .5 23 .1 78 ** .377 ** .163 * - .0 47 .0 55 .142 ** .096 * .152 * .256 ** - .032 .177 .082 -0.1 5 .0 92 -1.73** -.324** - .020 .141 -.0 96* -.069 -.08 7 -.068 -.00 9 -.0 26 - .357 .480 .212 ** .0 48 .144 * .01 2 .1 79 ** -.136** -. 107* - .144 .351 -0.82 .135 ** -.12 3 -.092 * -.07 1 -.075 -. 059 -. 305 ** - tr ade .067 .250 -.145 ** -.142** -.08 6 -.129** -.039 -.049 -. 03 8 -. 199 ** -. 110* - .110 .312 -.117 ** -.10 3* -.10 1 .314 ** .200 ** -.064 -. 050 -. 261** -. 144** -. 094* - .158 .365 .03 0 -.00 4 .03 2 -.133** -.151 ** -.079 -. 062 -. 322 ** -. 177 ** -. 116** -. 152** - .10 3 .30 5 .0 09 -.0 18 .02 0 .1 30 ** -.0 09 -.062 -.04 9 -.252 ** -.139 ** -. 091 * -. 119* * -. 14 ** - si gni fi ca nt a t t he 0. 01 le ve l ( 2-ta ile d) si gni fi ca nt a t t he 0. 05 le ve l ( 2-ta ile d) p re se nt ed in th ou sa nd s ( x1 00 0) e fi rs t ope ra tiona liz at ion m et ho d for CS R
T abl e 2: M ean s, S tan dar d D ev iat ion s, C or re lat ion s Me an SD 1 2 3 4 5 6 7 8 9 10 11 12 13 .44 3 .624 - 1907. 23 3166. 908 .124 ** - ion 6. 50 .6 68 .250* * .338 ** - iz e b 1. 42 .5 26 .104 * .374 ** .160 * - .0 47 .0 53 .044 .092 * .150 * .254 ** - .032 .177 .288** -0.14 .0 92 -1.73** -.324** - uc tion .020 .141 -.059 -.069 -.086 -.068 -.008 -.0 26 - ac tur ing .357 .480 .1 79 ** .051 .147* .01 4 .181** -.136** -. 107* - ta tion .144 .351 -.121** .138** -.122 -.091* -.070 -.075 -. 059 -. 306** - es al e tr ade .067 .250 -.148** -.142** -.085 -.129** -.039 -.049 -. 039 -. 200** -. 110* - l tr ade .110 .312 -.136** -.101* -.100 .316** .201** -.064 -. 050 -. 261** -. 144** -. 094* - e .158 .365 .000 -.002 .034 -.133** -.150** -.079 -. 062 -. 323** -. 178** -. 116** -. 152** - ce s .101 .302 .030 -.032 .010 .123** -.015 -.062 -.048 -.250** -.138** -. 090* -. 118** -. 146** - tion is si gni fi ca nt a t t he 0. 01 le ve l ( 2-ta ile d) tion is si gni fi ca nt a t t he 0. 05 le ve l ( 2-ta ile d) ue s a re p re se nt ed in th ou sa nd s ( x1 00 0) tin g th e se co nd ope ra tiona liz at io n m et ho d for CS R
4.2 Exploratory analyses 4.2.1. Model 1
The first analysis is a hierarchical regression analysis that is used to examine the relation between corporate social responsibility and corporate reputation. This analysis is performed to check the ability of CSR to predict a firm’s reputation, after controlling for firm size, financial performance and industry. As hypothesized in the literature review, it is expected that CSR will positively influence corporate reputation.
The results of the test on this hypothesis, using the first measure of CSR, can be found in Table 3 and Table 4. The model containing only control variables turned out to be significant (F(10, 233) = 4.269, p < .001) and explains 15.5% of variance in corporate reputation. After entry of the total CSR score the model accounted for 15.6% of total variance, however it showed no significance (F(11, 232) = 3.899, p = 0.569). In this model there were two significant predictors, with financial performance (β = .279, p < .001) recording a higher Beta value than firm size (β = .239, p < .001). This means that when financial performance increases by one point, the corporate reputation increases by .279 points, and when firm size increases by one point the corporate reputation increases by .239 points. The industry group Retail Trade also proved to be significant, indicating that companies in this group score .373 lower on reputation compared to companies that are not in this industry group.
The first analysis did not show a significant relationship between CSR and reputation, although this was highly expected from the literature (Orlitzky et al., 2003). Therefore, an analysis was also performed on a deconstruction of the total CSR score. In another hierarchical regression the relationship between all five CSR dimensions (Environment, Employees, Community, Human Rights and Product) and corporate reputation is further investigated. These results can be found in Table 4. In the first step, the control variables were entered. This model showed
Table 3: Linear Regression Total CSR1 – Corporate Reputation R R2 R2 Change B SE β T Model 1 .394 .155*** Firm Size .330 .096 .246** 2.64 CFP 3.58 .912 .283*** 2.70 Mining .948 . 480 .181* 1.97 Construction -.446 .405 -.112 -1.10 Manufacturing -.196 .324 -.141 -.606 Transportation -.366 .338 -.192 -1.08 Wholesale trade -.346 .365 -.121 -.948 Retail trade -.806 .339 -.373* -2.37 Finance -.106 .337 -.057 -.314 Services -.219 .337 -.108 -.649 Model 2 .395 .156 .001 Firm Size .321 .098 .239** 2.29 CFP 3.53 .918 .279*** 3.84 Mining .879 .496 .167 1.77 Construction -.477 .409 -.119 -1.17 Manufacturing -.253 .340 -.182 -.746 Transportation -.406 .346 -.214 -1.17 Wholesale trade -.384 .372 -.134 -1.03 Retail trade -.843 .346 -.390* -2.44 Finance -.158 .350 -.085 -.453 Services -.263 .346 -.129 -.759
reputation. Next, the five CSR dimensions were entered. This last model accounted for 18,7% of variance in corporate reputation and approaches marginal levels of significance (F(15,228) = 3.497, p = .113). Human Rights (β = .202, p < .01), financial performance (β = .275, p < .001) and firm size (β = .295, p < .01) were the significant predictors in this model. Although the second regression analysis showed a significant relationship between one CSR dimension and corporate reputation, the first hypothesis must be rejected as there is no statistical correlation between total CSR score and corporate reputation.
Table 4: Linear Regression Deconstructed CSR1 – Corporate Reputation R R2 R2 Change B SE β T Model 1 .394 .155*** Firm Size .330 .096 .246** 2.64 CFP 3.58 .912 .283*** 2.70 Mining .948 . 480 .181* 1.97 Construction -.446 .405 -.112 -1.10 Manufacturing -.196 .324 -.141 -.606 Transportation -.366 .338 -.192 -1.08 Wholesale trade -.346 .365 -.121 -.948 Retail trade -.806 .339 -.373* -2.37 Finance -.106 .337 -.057 -.314 Services -.219 .337 -.108 -.649 Model 2 .432 .187 .032 Firm Size .396 .109 .295** 3.64 CFP 3.48 .910 .275*** 3.82 Mining .167 .575 .032 .290 Construction -.538 .413 -.135 -1.30 Manufacturing .330 .345 -.273 -.956 Transportation -.406 .349 -.272 -1.48 Wholesale trade -.456 .375 -.159 -1.22 Retail trade -.973 .353 -.450* -2.76 Finance -.279 .354 -.150 -.786 Services -.405 .354 -.199 -1.14 Environment1 -.023 .036 -.049 -.635 Employees1 .063 .045 .093 1.05 Community1 .118 .166 .047 .708 Human Rights1 .392 .151 .202** 2.59 Product1 -.032 .068 -.032 -.475
The same test is performed using the second measure of CSR, results of which can be found in Table 5. The model containing only control variables turned out to be significant (F(10, 232) = 4.182, p < .001) and explains 15.3% of variance in corporate reputation. After entry of the total
Table 5: Linear Regression Total CSR2 – Corporate Reputation
R R2 R2 Change B SE β T Model 1 .391 .153** Firm Size .327 ,097 .243** 3.38 CFP 3,54 .916 .281*** 3.87 Mining .940 .481 .179 1.95 Construction -.449 .406 -.113 -1.10 Manufacturing -.197 .325 -.142 -.608 Transportation -.368 .339 -.194 -1.08 Wholesale trade -.349 .366 -.122 -.955 Retail trade -.805 .340 -.373* -2.36 Finance -.108 .337 -.058 -.322 Services -.231 .338 -.112 -.682 Model 2 .422 .178** .025** Firm Size .294 .096 .219** 3.05 CFP 3.28 .909 .260*** 3.61 Mining .461 .508 .088 .907 Construction -.633 .407 -.159 -1.56 Manufacturing -.470 .336 -.338 -1.40 Transportation -.558 .342 -.294 -1.63 Wholesale trade -.512 .366 -.179 -1.40 Retail trade -.978 .342 -.453* -2.86 Finance -.363 .346 -.196 -1.05 Services -.457 .345 -.222 -1.33 CSR2 .183 .069 .181* 2.66 Statistical significance: *p <.05; **p <.01; ***p<.001
231) = 4.546, p < .01). In this model three predictors were significant. Financial performance (β = .260, p < .001) and firm size (β = .219, p < .01) reported a higher Beta value than CSR (β = .181, p < .05). This indicates that, according to the second analysis, corporate reputation increases by .260 points when financial performance increases by one point, and when firm size increases by one point the reputation increases by .219. More importantly, this analysis showed that when CSR score improves by one point, the company’s reputation increases by .181 point. A test on the deconstructed CSR score using the second operationalization method is also performed for completeness. These results can be found in Appendix A. In the second model the five different CSR dimensions were entered. This last model accounted for 22.4% of variance in corporate reputation and was statistically significant (F(15,227) = 4.369, p < .01). This model reported the same significant predictors, though with different coefficients. These were Human Rights (β = .288, p < .001), financial performance (β = .263, p < .001) and firm size (β = .291, p < .001). When looking at the model with the second measurement of CSR, the positive relation between CSR and reputation proves the first hypothesis can be accepted.
4.2.2. Model 2
The following analysis is a hierarchical regression analysis that is used to examine the relationship between the CPA and corporate reputation (H2). This analysis is performed to check the ability of CPA to predict corporate reputation, after controlling for firm size, financial performance and industry. The results of this analysis can be found in Table 6. In the first step of the hierarchical multiple regression, three predictors were entered: firm size, industry, and financial performance. This model was statistically significant F(10, 232) = 4.284, p < .001 and explained 15.6% of variance in corporate reputation. After entry of corporate political activity in step 2 the total variance explained by the total model was 23%, F(11, 231) = 6.287; p < .001. The introduction of CPA explained additional 7.5% variance in corporate reputation, after
controlling for firm size, industry, and financial performance (R2 change = .075; F(1, 231) =
22.372; p < .001).
In the final model, three variables were statistically significant, with CPA recording a higher Beta value (β = .316; p < .001) than financial performance (β = .279; p < .001). In other words, if a firm’s political activities increase by one point, the corporate reputation increases by 0.316 and if a firm’s financial performance increases for one point the reputation increases by 0.279. The industry group ‘Mining’ also proved to be significant, indicating that companies in this group invest .200 more in CPA compared to companies that are not in this industry group. This paper also investigated at the effect of the two separate components of CPA, lobbying and PAC contributions. After entering these two variables, in place of total CPA, the total variance explained by the model was 23.1% (F(12, 230) = 11.269; p < .001). By adding lobbying and PAC to the model an additional 7.5% of variance in corporate reputation was explained. Out of these two variables only lobbying was statistically significant (β = .317; p < .001), while PAC contributions showed a negative, but insignificant relationship (β = -.002; p = .980).
The exact same analysis was done for the other sample with the second CSR measure. As these results (R2 change = .074, F(11, 230) = 6.165; p < .001; Appendix B) barely differ from the
results that are mentioned above, no further explanation is given here.
The results on model 2 show that corporate political activity influences corporate reputation, although the coefficient reports a positive influence rather than a negative one. Therefore, hypothesis 2 should be rejected.
Table 6: Linear Regression CPA – Corporate Reputation R R2 R2 Change B SE β T Model 1 .395 .156*** Firm Size .334 .096 .249** 3.47 CFP 3.54 .912 .280*** 3.88 Mining .946 . 480 .180* 1.97 Construction -.442 .405 -.111 -1.09 Manufacturing -.194 .324 -.140 -.600 Transportation -.363 .338 -.191 -1.08 Wholesale trade -.344 .365 -.120 -.941 Retail trade -.805 .339 -.373* -2.38 Finance -.085 .337 -.045 -.252 Services -.217 .337 -.107 -.645 Model 2 .480 .230*** .075*** Firm Size .142 .101 .106 1.42 CFP 3.53 .873 .279*** 4.04 Mining 1.05 .460 .200* 2.27 Construction -.186 . 391 -.047 -.474 Manufacturing .013 .313 .009 .040 Transportation -.258 .324 -.136 -.794 Wholesale trade -.054 .355 -.019 -.153 Retail trade -.408 .335 -.189 -1.22 Finance .129 .326 .069 .396 Services .061 .328 .030 .187 CPA 5.31 .000 .316*** 4.73 Statistical significance: *p <.05; **p <.01; ***p<.001
4.2.3. Model 3
Lastly, a moderation analysis was performed to determine if the effect of corporate social responsibility on a firm’s reputation is influenced by the firm’s level of political activities. To test the moderating effect of political activities, the PROCESS macro from Hayes (2013) is used in SPSS.
Table 7 shows an overview of results of the PROCESS Model 1 analysis, where M reflects the moderator, X reflects the independent variable and the last three variables were the variables controlled for. No significant interaction was found between CPA and first measure of CSR on corporate reputation (C3 = .000, p = .651). The moderation effect was also tested separately for
Table 7: Moderation Regression Model
Coefficient Std.
Error T P LLCI ULCI
Intercept 6.17 .322 19.2 .000 5.546 6.814 CPA (M) .000 .000 4.80 .000 .0000 .0001 CSR1 (X) .014 .022 .622 .535 -.0303 -.0583 CPA*CSR1 .000 .000 .146 .884 .0000 .0000 Firm Size .133 .116 1.15 .251 -.0950 -.3616 CFP 3.46 1.01 3.42 .001 1.466 5.452 Mining .955 .601 1.59 .113 -.2291 2.139 Construction -.234 .287 -.814 .416 -.800 .3319 Manufacturing -.063 .240 -.264 .729 -.5360 .4093 Transportation -.313 .244 -1.28 .201 -.7933 .1674 Wholesale trade -.109 .302 -.360 .719 -.7034 .4861 Retail trade -.460 .287 -1.60 .111 -1.026 .1061 Finance .057 .254 .225 .823 -.4430 .5570 Services -.004 .260 -.0.14 .989 -.5158 .5086 R2 = .0005; P = .6888
