Power and Paradigms: The Dutch Response to Pressures for Shareholder Value
Abstract
Manuscript Type
Empirical
Research Question/Issue
In recent years, research aimed at identifying and relating the antecedents and consequences of diffusing organizational practices/ideas has turned its attention to debating the international adoption and implementation of the Anglo-American model of corporate governance, i.e., a shareholder value orientation (SVO). While financial economists characterize the adoption of an SVO as necessary and performance-enhancing, behavioral scientists have disputed such claims, invoking institutional contingencies in the appropriateness of an SVO. Our study seeks to provide some resolution to the debate by developing an overarching sociopolitical perspective that links the antecedents and consequences of the adoption of the contested practice of SVO.
Research Findings/Insights
We test our framework using extensive longitudinal data from 1992–2006 from the largest listed corporations in the Netherlands. We find a negative relationship between SVO adoption and subsequent firm performance, although this effect is attenuated when accompanied by greater SVO alignment among major owners and a firm's visible commitment to an SVO.
Theoretical/Academic Implications
This study extends prior research on the diffusion of contested organizational practices that has taken a sociopolitical perspective by offering an original contingency perspective that addresses how and why the misaligned preferences of corporate owners will affect (i) a company's inclination to espouse an SVO, and (ii) the performance consequences of such misalignment.
Practitioner/Policy Implications
This study suggests that, when board members are considering the adoption of new ideas/practices (e.g., SVO), they should consider the contextual fitness of the idea/practice with the firm's owners and their interests.
“There are of course competing claims about the meaning of a paradigm, indeed, about the meaning and significance of any body of beliefs. But none is correct, and which among them secures local ‘hegemony’ is a matter of social, political, economic and other sorts of power” (Rosenberg, 2005: 177)
Introduction
Despite the long history of the public corporation, its fundamental purpose remains a topic of debate among scholars and practitioners around the world (Freeman, Wicks, & Parmar, 2004; Sundaram & Inkpen, 2004). While coordinated market economies (e.g., Germany and Japan) conceptualized corporations as communities of interests whose stakeholders should be served, liberal market economies (e.g., the United States) typically viewed corporations as instruments for creating value for shareholders (Meyer & Höllerer, 2010; Stadler, Matzler, Hinterhuber, & Renzl, 2006). Recent scholarship suggests growing evidence that the latter model is becoming more popular around the globe and executives in contexts with a historical emphasis on stakeholder value, ranging from Germany (Fiss & Zajac, 2004; Tüschke & Sanders, 2003) and Austria (Meyer & Höllerer, 2010) to Japan (Yoshikawa & McGuire, 2008; Yoshikawa, Tsui-Auch, & McGuire, 2007) and Sweden (Lazonick & O'Sullivan, 2000), are struggling with pressures for the adoption of a shareholder value orientation (SVO).
The specific antecedents and consequences of a firm's adoption of an SVO in a national context with a pre-existing competing logic has become a focused subject of recent academic inquiry (Fiss & Zajac, 2004, 2006; Sanders & Tüschke, 2007; Yoshikawa et al., 2007). Some scholars have suggested that the rise of the shareholder value model is driven by capital- and product-market pressures (Lazonick & O'Sullivan, 2000; Yoshikawa & Rasheed, 2009). Others have pointed to the impact of sociopolitical and institutional processes. For instance, Fiss and Zajac (2004, 2006) illustrate how the varying preferences of major domestic owners influenced the adoption of an SVO among German companies, and Sanders and Tüschke (2007) describe how a firm's exposure to institutional contexts where shareholder value practices are more legitimate increases SVO diffusion across national boundaries. Even greater variation exists in examinations of performance consequences of SVO adoption: while financial economists have emphasized the performance-enhancing character of an SVO, behavioral scientists have suggested that the contested nature of an SVO could invoke political conflict and misalignment among its main stakeholders, negating any presumed performance benefit.
While this growing literature has expanded our broad understanding of the diffusion of an SVO, a number of fundamental questions regarding this diffusion process and its consequences remain unanswered. For example, while large-scale diffusion studies have recently added a focus on contestation in diffusing organizational practices/ideas (Fiss, Kennedy, & Davis, 2012; Fiss & Zajac, 2004; Sanders & Tüschke, 2007), they do not analyze directly the balance of power between forces that are for or against such a change. Indeed, “diffusion can be a highly political process … driven by the political dynamics of opposition, argumentation, contestation and compromise” (Schneiberg & Soule, 2005: 156). Moreover, there is little research that addresses the performance consequences of diffusing practices under conditions of contestation; when changes are needed but also particularly contested, does this also portend a subsequent firm performance improvement or decline? In this study we use a sociopolitical perspective to examine these processes in the Dutch context. In doing so, we hope to make the following two contributions to the existing literature.
First, we extend prior research on the diffusion of organizational practices that has taken a sociopolitical perspective (Fiss & Zajac, 2004; Westphal & Zajac, 1998, 2001; Zajac & Westphal, 1995) by directly examining how the balance of power between major owners affects a firm's inclination to espouse an SVO in a national context dominated by an alternate logic. In contrast to prior diffusion research that has generally examined the influence of specific actors (Fiss & Zajac, 2004, 2006; Lazonick & O'Sullivan, 2000), we investigate how contestation regarding adoption emerges from the combined impact of owners for and owners against SVO change. We contextualize this analysis by exploring two potential sources of SVO contestation in the traditionally stakeholder-oriented Dutch context: technical misalignment between the preferences of major owners on the basis of their relationship with the firm (i.e., independent institutional investors versus dependent investors), and cultural misalignment between the preferences of major owners (i.e., Anglo-American versus non-Anglo-American investors).
In addition, while prior research has focused primarily on the antecedents of SVO adoption, we also analyze the performance consequences of a company's adoption of an SVO. While there is no dearth of opinion on the consequences of SVO adoption – e.g., strong and universalistic negative statements such as, “End shareholder value tyranny: Put the corporation first” (Raynor, 2008), and “So if others are stupid enough to do it [adopting an SVO], that will only help North American businesses” (Mintzberg, 2000: 38) – there is relatively little evidence either supporting or refuting such strongly held views. Moreover, the situation suggests the need for a more nuanced framework that seeks to specify the mechanisms that would generate (or attenuate) such an expected negative outcome. In this study, we address explicitly these shortcomings by providing a theoretical and empirical analysis of the conditions that can exert a downward performance pressure following the adoption of an SVO.
The remainder of this paper is structured as follows. In the next section, we advance our sociopolitical perspective and develop our hypotheses regarding SVO adoption and its performance implications. The third section describes our data and methods, i.e., the Dutch research context, our use of extensive panel data on top-100 listed firms in the Netherlands from 1992 to 2006, and our empirical analyses. In the fourth section we discuss the results emerging from our study and in the final section we elaborate on our findings, their implications, and avenues for future research.
Power Struggles and Contested Change
A growing body of research on organizational change has emphasized the importance of sociopolitical processes driving the diffusion of organizational practices/ideas (Ansari, Fiss, & Zajac, 2010; DiMaggio & Powell, 1983; Kennedy & Fiss, 2009; Westphal & Zajac, 1998, 2001, 2013; Zajac & Westphal, 1995, 2004), i.e., companies do not only need to deal with demands from their technical environments, but also with demands from their cultural environments. Accordingly, the interpretations of organizational and societal actors about “the way things are” and “the way things are to be done” (Scott, 1987: 496) have a significant imprint on a company's response to novel organizational practices and ideas. In fact, firms may adopt these practices and ideas for motives of societal legitimacy rather than efficiency (Davis, 1991; DiMaggio & Powell, 1983; Westphal & Zajac, 1994) or even with the knowledge that they are inefficient (Yoshikawa & Rasheed, 2009). Moreover, scholars have shown that symbolism, i.e., espousing the adoption of new normative beliefs without implementing them, may already generate market reactions, because prevailing financial market logics also tend to be socially constructed (Westphal & Zajac, 1998; Zajac & Westphal, 2004).
Given that the contexts in which practices diffuse are far from neutral (Ansari et al., 2010; Schneper & Guillen, 2004) and often highly political (Davis & Thompson, 1994; Fiol & O'Connor, 2002), prior studies have indicated that these sociopolitical processes tend to become particularly pronounced when new organizational practices/ideas are contested, i.e., pioneered by supporting actors, yet strongly opposed by others. Resistance is most likely to come from actors that have operated according to the “old” logic and for whom new organizational practices/ideas resemble a disruption which threatens their vested interests and identities (Hannan, Baron, Hsu, & Koçak, 2006; Pardo del Val & Martinez Fuentes, 2003). This is consistent with growing broader recognition that when institutionalized practices are highly resistant to change, it is due to their embeddedness in social phenomena such as belief systems, which themselves reflect the distribution of power (Jürgens, Naumann, & Rupp, 2000; Yoshikawa & Rasheed, 2009). In those circumstances the entrance of a powerful actor with a different background is often necessary to overthrow prevailing practices and initiate change (Fiol & O'Connor, 2002; Kraatz & Moore, 2002).
Given our focus on how corporate practices are embedded in social phenomena such as belief systems and reflect the distribution of power, it should not be surprising that our sociopolitical perspective sees corporate control as the outcome of the struggle between varying preferences of the actors involved. While several actors have an interest to be involved in this process (Schneper & Guillen, 2004; Yoshikawa & Rasheed, 2009), controlling shareholders will have the strongest incentives to participate in the light of the limited liquidity of their investments (Dharwadkar, Goranova, Brandes, & Khan, 2008; Schnatterly, Shaw, & Jennings, 2008; van Essen, van Oosterhout, & Heugens, 2013). Accordingly, several studies have indicated that the dispositions of major owners are reflected in a firm's value orientation. For example, Fiss and Zajac (2004) illustrate that varying preferences of major domestic owners influenced the espousal of an SVO among German corporations. Furthermore, Brouthers, Gelderman, and Arens (2007) describe that government ownership has a significant imprint on a company's strategic orientation and managers' leadership style. In the remainder of this section, we build on this argument and introduce the notion that the espousal of an SVO can be understood as driven by the combined impact of the technical and cultural preferences of owners for and owners against change. Accordingly, we suggest that when misalignment exists among elites, both the likelihood of SVO adoption and its performance benefit will diminish.
Technical and Cultural Preferences of Major Owners and SVO Espousal
Following prior work on the diffusion of organizational practices/ideas (Ansari et al., 2010; Kennedy & Fiss, 2009; Zeitz, Mittal, & McAulay, 1999), two different accounts can be given for the identification of major owners with the shareholder value model: a technical and a cultural one. Regarding the first account, scholars have referred to an SVO in rather technical terms, i.e., as an “instrument of corporate control” (Pfeiffer, 2000: 68) and “product and promise that purposive management action will be rewarded” (Froud, Haslam, Johal, & Williams, 2000: 80). Originating from the field of economics, an SVO emphasizes profit maximization for the shareholders, and is generally associated with legitimacy in global financial markets (Ghoshal, 2005; Lazonick & O'Sullivan, 2000). Given these roots, we posit that independent institutional investors, i.e., the ones that have no business relations with the company and a clear value maximization perspective, will be the advocates for SVO. Indeed, private equity, hedge funds, and pension funds are among the frontrunners striving for a stronger emphasis on shareholder value in financial markets (Clifford, 2008; Klein & Zur, 2009), and act autonomously of the management's agenda (Almazan, Hartzell, & Starks, 2005; Chen, Harford, & Li, 2007; De Jong, Mertens and Roosenboom, 2006).
Whereas independent institutional investors should be the most inclined to stimulate SVO change, two other groups of major owners might be particularly resistant to their initiatives; corporate (non-financial) and family owners. First, corporate owners are characterized by reciprocal business relationships (Allen & Phillips, 2000). Furthermore, they may also anticipate contagion effects in their own context, i.e., they might expect to become the next “target” if the new logic increasingly gets a foothold. Hence, corporate owners are likely to oppose pressures for shareholder value. Second, family owners might resist change due to the fact they are risk averse and care more about the longevity of the company, rather than short-term profits, which is believed to be an important characteristic of an SVO. Indeed, long-term survival and prosperity of the family seems at odds with the more short-term oriented SVO. Accordingly, scholars have illustrated that family owners often have a more traditional, long-term orientation and often act as “protective stewards” of a firm (Fiss & Zajac, 2004: 509; Miller, Le Breton-Miller, & Lester, 2013; Young, Peng, Ahlstrom, Bruton, & Jiang, 2008).
A second stream of research has referred to an SVO in cultural terms, arguing that this orientation is deeply embedded in socioeconomic culture, i.e., in the ways in which actors give meaning to the public company and its main purposes (e.g., Ansari et al., 2010; Zeitz et al., 1999). In particular, literature on economic nationalism suggests that differences between nations are quite profound and persistent (e.g., Levi-Faur, 1997) and that domestic actors are likely to respond to foreign contested practices by “strong assertions of national differences and identity” (Yoshikawa & Rasheed, 2009: 394). Therefore, in the context of the spread of an SVO, we expect that blockholders that are exposed to the stakeholder model in their own domestic context will strongly oppose the new logic and will act as “protective stewards” of the traditional stakeholder logic.
Conversely, we expect that major owners from abroad that have the shareholder value logic as taken for granted in their domestic context, might, in particular, contribute to the diffusion of an SVO. They might force managers to reflect on the appropriateness and international sustainability of the old logic and/or share their successful experiences with the alternative logic (Sanders & Tüschke, 2007; Yoshikawa et al., 2007). As the shareholder model has its historical origins in the United States (Lazonick & O'Sullivan, 2000), we expect that Anglo-American-oriented blockholders will be the prime propagators of the shift towards an SVO in the context of the Dutch market. Hence, the cultural distribution across blockholders will play an important role regarding the company's readiness to adopt the shareholder logic, at the expense of the prevalent stakeholder logic in the Netherlands.
Performance Implications of SVO Espousal
While the introduction of the Anglo-American shareholder model in stakeholder-oriented contexts has been criticized extensively (Froud et al., 2000; Mintzberg, 2000; Raynor, 2008), the empirical evidence regarding the performance implications is sparse and rather mixed. On the one hand, scholars have highlighted that the espousal of an SVO is associated with disturbed labor and industrial relations, underinvestment, and a short-term (vs. long-term) optimization of a firm's performance (e.g., Ghoshal, 2005; Lane, 2003). On the other hand, some suggest that espousing an SVO will have a positive impact on a firm's financial performance, because “associations around value creation are powerful, because they suggest focused, effective management, delivering improved performance in the interests of shareholders” (Froud et al., 2000: 85). By referring to an SVO, listed companies may signal their commitment to the interests of shareholders and, as such, enhance their legitimacy in international capital markets. Moreover, espousing companies may be more focused on the financial bottom line as they might be more sensitive to the preferences of shareholders, being the actors that have the strongest incentives to maximize total firm value as residual claimants (Sundaram & Inkpen, 2004).
Methods
We conduct our research within the Netherlands, a small and open economy heavily reliant on the global economy. The roots of the contemporary Dutch open economy can be found in the glory days of the Golden Age (1600s). In this era, when the Netherlands was one of the largest trading nations, the Dutch founded the “Dutch United East India Company”, one of the first joint stock companies in the world. With a small group of large, internationally diversified firms and a large part of gross domestic product earned abroad, the Dutch trade origins and its international orientation are still prominent. The Netherlands is a welfare state with a long tradition of balancing the interests of societal groups. Accordingly, Dutch company law explicitly defines publicly listed corporations as legal entities that must take into account the rights of all stakeholders affected by the firm. The institutionalized stakeholder approach is in most cases supported by a two-tier board model consisting of a management board and a supervisory board. The supervisory board consists of non-executive directors to assure its independence and has the duty by law to supervise and advise the management board while acting in the best interests of the company and its stakeholders.3
While prior research in this area has mainly focused on larger economies, such as Japan (Yoshikawa & McGuire, 2008; Yoshikawa et al., 2007) and Germany (Fiss & Zajac, 2004, 2006; Sanders & Tüschke, 2007), we view the Dutch setting as well suited to the study of technical and cultural power struggles and contested SVO change. First, given our emphasis on analyzing possible cultural power struggles between Anglo-American and non-Anglo-American investors, we needed a national context whereby large listed firms are both sensitive to international capital market pressures, yet also historically wedded to the stakeholder-oriented model. The stakeholder-oriented context in the Netherlands fits these criteria very well, as the Dutch economy has been more reliant on foreign ownership than other typically stakeholder-oriented countries such as Japan and Germany (Huizinga and Nicodème, 2006; Pedersen and Thomsen, 1997; Yoshikawa et al., 2007). In fact, the Netherlands is among the countries with the highest foreign ownership in Europe (i.e., around 70 percent), while Germany is among the countries with the lowest foreign ownership (i.e., around 20 percent) (Federation of European Securities Exchanges, 2008). The Dutch setting is thereby more likely to show the required cultural diversity in ownership to adequately test whether cultural power struggles (next to technical pressures) affect the adoption of contested SVO change. Second, in the Dutch context power struggles among owners are not played out via formal exercising of voting rights at general meetings (De Jong et al., 2006). Hence, looking at outcomes like SVO adoption is one way to address the way power plays when new corporate practices are being adopted.
The empirical research focuses on the top-100 listed corporations at the Euronext Stock Exchange in Amsterdam during the period 1992–2006. The top-100 listed firms were selected on the basis of their membership of the AEX, AMX and AScX indices. These three main indices contain the top-75 firms (each index consists of 25 companies) and membership is based on the annual value of shares traded. The top-100 was completed by including 25 of the remaining firms based on their absolute market capitalization at year's end. Given new entrants, mergers and acquisitions, bankruptcies, and delistings, 201 different companies met one of these criteria during our observation window. Of these, 17 corporations were excluded as their annual reports were unavailable. Overall, this resulted in an unbalanced panel of 184 listed firms for which a total of 1,216 complete firm-year observations were available during the period 1992–2006. Together, these companies represent seven industries (construction, manufacturing, transport and communication, wholesale, retail, financial services, and other services) and constituted the backbone of the Dutch economy during this period. While several companies already espoused an SVO in 1992, the availability of annual reports and ownership data unfortunately restricted us from going further back in time and capturing the dynamics during the initial phase.
Dependent Variables
In this study, we use two dependent variables: the espousal of an SVO and firm performance. Following prior research (e.g., Fiss & Zajac, 2004, 2006), data on the espousal of an SVO were collected through a content analysis of companies' annual reports, still one of the major corporate tools in use to communicate with shareholders in the Dutch context. As such, the language used inside these reports regarding shareholder value is an important reflection of managerial predispositions towards this “normative governance paradigm” (Fiss & Zajac, 2004: 512; Pye, 2002) and of the way in which a firm publicly presents itself to capital markets. Two independent raters, both native speakers, coded the annual reports by searching for the terms “shareholder value” and its Dutch equivalent “aandeelhouderswaarde.” There appeared to be a high overlap in the ratings of the coders: the percentage agreement (96.2 percent) and Cohen's Kappa (.924) were both above the acceptance threshold (Dewey, 1983). The following two statements are illustrations of SVO identified sections of text: “Our key objective is to increase shareholder value” (Wessanen, 1999: 1) and “Going forward, we will continue to identify opportunities in emerging markets that offer growth potential consistent with AEGON's requirements of long-term profitability and the creation of shareholder value” (AEGON, 2006: 47).
To investigate the performance implications of the espousal of an SVO, we use two widely used performance indicators: return on assets (ROA) and total shareholder return (TSR). ROA is one of the most often used accounting-based measures of performance that provides insight into the short-term, historical profitability of a firm (Anderson, Fornell, & Mazvancheryl, 2004; Thomsen & Pedersen, 2000). TSR is one of the most often used market-based measures of performance that also incorporates the long-term profitability expectations (Anderson et al., 2004; Fiss & Zajac, 2006). The financial data were collected via the databases Thomson One Banker Worldscope and Reach.
Independent Variables and Moderating Variables
Data on the capital stakes by major owners were collected via “Handboek Nederlandse Beursfondsen,” an annual publication that provides an overview of blockholdings. According to the Dutch “Law on the Disclosure of Shareholdings,” these firms are legally obliged to disclose ownership stakes above 5 percent to the AFM, the regulator of Dutch financial markets. To calculate the cultural balance of power, we first summed the blockholdings by (i) major owners from contexts in which an SVO is generally accepted, e.g., the USA, UK, Australia, Canada, Hong Kong, and Singapore (Groot, 1998; Lazonick & O'Sullivan, 2000) and (ii) major owners from all remaining other countries, including the Netherlands, i.e., shareholders that have the strongest association with the Dutch stakeholder model. To calculate the technical balance of power, we summed the blockholdings by (i) independent institutional investors (private equity firms, hedge funds, venture capitalists, and pension funds), owners that are used to applying the economic logic of the SVO model, and (ii) corporate blockholders and families, i.e., owners that can identify themselves, for a variety of reasons, with a firm's business. As the concept of power is core to our arguments and the rights of owners increase when certain ownership thresholds are reached, we use the same ordinal measure as Fiss and Zajac (2004, 2006): 0 if that category owned less than 5 percent, 1 if that category owned more than 5 percent but less than 25 percent, 2 if that category owned more than 25 percent but less than 50 percent, 3 if that category owned more than 50 percent but less than 75 percent, and 4 if that category owned more than 75 percent. Finally, we subtracted the scores of cultural and technical owners against change from the ones of cultural and technical owners for change, to obtain our scores for both balances of power.
The implementation of an SVO was measured by a formative scale of four practices that have often been associated with such a logic: share buy-back programs, stock option compensation schemes, value-based management tools, and internationally accepted accounting standards (Fiss & Zajac, 2004; Froud et al., 2000; Lazonick & O'Sullivan, 2000). All practices either align the interests of managers and shareholders or provide benefits to owners such as more transparency or a stronger focus on economic value. All data were collected through a content analysis of the annual reports. Two other independent raters, both native speakers, coded the available annual reports by searching for terms related to these four practices. Again there appeared to be enough overlap in the ratings of the two coders: the percentage agreement varied between 85.6 percent and 96.3 percent, and Cohen's Kappa ranged from .712 to .925 for the varying practices.
Control Variables
By including various additional variables, we controlled for possible confounding effects. As scholars have suggested that the diffusion of an SVO is driven by capital- and product-market pressures (Lazonick & O'Sullivan, 2000), we included a firm's debt-to-equity ratio (Debt/ext. market capitalization), due to data availability using the amount of equity raised externally as a proxy for total shareholders' equity (Fiss & Zajac, 2004). Furthermore, we included the ratio of foreign sales to total sales (Foreign sales/total sales) to control for international product-market pressures, i.e., by operating abroad firms may become more sensitive to alternative corporate governance models (Sanders & Tüschke, 2007). As the espousal of an SVO might also depend on the dynamics of the market for corporate control, we included the percentage of dispersed shares (Dispersed shares) to control for the disciplining role this market might fulfill (Fiss & Zajac, 2004). Furthermore, we included ownership by major banks (Ownership by major banks) and managers/directors (TMT/board ownership), as we remained agnostic about whether these groups would be for or against SVO change. Indeed, banks are known to have a conflict of interest by also being financial service and debt providers to the firms they own (Romano, 2003).4 Indeed, this conflict of interest can prevent them from being for SVO given the fact that they are aligned to management (Brickley, Lease, & Smith, 1988; Pound, 1988). However, in the Dutch setting, De Jong et al. (2006) do not find evidence that banks are pressure-sensitive and consistently aligned to management. Regarding the management-director ownership there is also a conflict of interest; while it can drive share value maximization due to convergence of interest, at the same time it allows the management to indulge in other behavior than value maximization, due to lack of entrenchment (Morck, Shleifer, & Vishny, 1990). Finally, as the Dutch context has been known for its anti-investor protection devices (e.g., Chirinko, Garretsen, Van Ees, & Sterken, 2004; De Jong, Douglas, Mertens, & Wasley, 2005; Kabir, Cantrijn, & Jeunink, 1997), we included a dummy (Board protection devices) to capture whether a firm used one of the two most prevalent investor protection devices in the Netherlands, i.e., preference shares (shares that can be issues against 25 percent of the nominal value to a friendly holder in case management feels threatened) and tradable depository receipts (shares with cash flow rights, but no voting power). In both cases board members are (somewhat) shielded from market pressures, potentially limiting their inclination to adopt an SVO.
We also added several firm-level controls. First, as larger companies may be less dependent on the financial markets and, as a result, might perceive less short-term pressures (Groot, 1998), we included the natural logarithm of the total sales during a year (Log sales). Second, poorly performing firms might be more inclined to espouse an SVO to restore their public reputation. Therefore, we added a lagged measure of TSR and ROA to our models (Return on Assets and Total return). Third, prior studies have shown that older corporations may have higher cumulative experience enhancing change; however, they might also encounter problems in keeping abreast with new developments (Sorensen & Stuart, 2000). We included firm age (Firm age), which was operationalized as the number of years passed after founding to control for these effects. Fourth, while Dutch listed companies usually operate a two-tier board, several of them have adopted the one-tier board, i.e., a structure that is more common around the globe. As these firms may be more responsive to international developments, we included a dummy variable capturing the (non)presence of a two-tier board (Leadership structure). Fifth, we control for firms that have an Anglo-American cross-listing. That is, Anglo cross-listed firms expose themselves to monitoring by the relevant stakeholders, like security analysts, rating agencies, and other informational intermediaries. These market-based mechanisms impose reputational bonding, due to which cross-listed companies might voluntarily act in accordance with the norms in the Anglo-American financial marketplace (Siegel, 2005). Hence, cross-listing can be an important determinant of SVO espousal. We collected the cross-listing data using the “Gids bij de Officiele Prijscourant van de Amsterdamse Effectenbeurs.” A dummy variable (Anglo listing) captures whether or not a listed company has a cross-listing in an Anglo-oriented country. Sixth, to account for industry- and time-specific effects, we included dummy variables for the company's two-digit BIK code (i.e., the Dutch equivalent of the SIC codes) and for the years of our observation window, in all our models. Given the large number of variables, the industry- and time-specific effects are not reported in the tables. Seventh, given the unbalanced nature of our sample and the likelihood that firms entering or dropping out of the sample during the period 1992–2006 could bias the results, we included a dummy variable capturing whether or not a company remained in the sample for the whole period (Firm present 1992–2006).
Analysis
To examine the antecedents of company's espousal of an SVO, we employed random-effects logistic regression models, as our dependent variable is a dummy capturing a firm's (non)espousal (Maddala, 2005). A Hausman test indicated that the random-effects model is the appropriate choice to analyze our data (χ2 = 18.21; d.f. = 19; p > .05). Furthermore, to examine the performance implications of the espousal of an SVO, we ran several random-effects pooled time series regression models for ROA. A Hausman test showed that the random-effects model is the appropriate way to examine our data (χ2 = 33.52; d.f. = 26; p > .05). To investigate the impact on a firm's TSR, we conducted several fixed-effects pooled time series regression models. A Hausman test indicated that this model was most appropriate (χ2 = 688.65; d.f. = 27; p < .05). Finally, all the independent and control variables were one year lagged in our analyses in order to address potential issues of reverse-causality.
Results
Using explicit statements referring to an SVO in annual reports of the top-100 listed corporations in the Netherlands, Figure 1 illustrates the extent to which these companies have espoused this orientation during the period 1992–2006. As depicted, a gradual and significant increase is visible, revealing the advance of this alternative orientation in a traditionally stakeholder-oriented context: while 13 percent of the top-100 listed firms referred to the shareholder value model in 1992, 74 percent of them did in 2006.
Espousal of an SVO among Dutch Firms (1992–2006)
Table 1 provides an overview of the pooled descriptive statistics of our panel of 184 listed firms in the Netherlands during the period 1992–2006. Table 2 reports the random effects logistic models predicting the espousal of an SVO among listed firms in the Netherlands. The results partially support our notion that the espousal of an SVO can be understood as driven by the balance of power between technical and cultural owners for and against change. In accordance with H1, our empirical results indicate that when the balance of power shifts toward independent institutional investors, the likelihood that a firm will espouse an SVO increases. In contrast to H2, our findings also highlight that a predominance of major Anglo-American owners does not influence a firm's inclination to adopt an SVO. Hence, the cultural influence seems not to be related to SVO espousal. This finding is also supported by the fact that Anglo-American listing is also not related to SVO espousal. Taken together, these findings suggest that (i) understanding power struggles among major owners helps to explain contested SVO change in the Netherlands, and (ii) it is more a technical phenomenon than a cultural one.
| Mean | St. Dev | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | (11) | (12) | (13) | (14) | (15) | (16) | (17) | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) SVO espousal | .50 | .50 | |||||||||||||||||
| (2) SVO implementation | 1.60 | 1.15 | .44 | ||||||||||||||||
| (3) Technical balance of power | .23 | 1.43 | .14 | .14 | |||||||||||||||
| (4) Cultural balance of power | −1.47 | 1.30 | .22 | .31 | .27 | ||||||||||||||
| (5) National network centrality | 1.65 | .88 | .18 | .25 | .09 | .12 | |||||||||||||
| (6) TMT/board ownership | .31 | .75 | −.08 | −.12 | −.25 | −.33 | −.19 | ||||||||||||
| (7) Ownership by major banks | .66 | .61 | .06 × 10−1 | −.02 | .29 | −.18 | .15 | −.10 | |||||||||||
| (8) Board protection devices | .49 | .50 | .10 | .03 | .25 | .06 | .06 | −.09 | .21 | ||||||||||
| (9) Debt/ext. market capitalization | 417.42 | 5805.73 | −.03 | .07 × 10−1 | .04 | −.04 | .03 | −.01 | −.02 | −.04 | |||||||||
| (10) Foreign sales/total sales | 50.15 | 30.36 | .16 | .22 | .06 | .17 | .20 | −.10 | .06 × 10−1 | .04 | −.04 × 10−1 | ||||||||
| (11) Anglo listing | .29 | .53 | .12 | .36 | −.04 | .20 | .29 | −.12 | −.04 | −.09 | −.01 × 10−1 | .26 | |||||||
| (12) Dispersed shares | 61.79 | 25.97 | .15 | .24 | .23 | .69 | .24 | −.35 | −.20 | −.02 | −.10 | .20 | .24 | ||||||
| (13) Log sales | 6.93 | 1.83 | .26 | .45 | −.03 | .16 | .61 | −.22 | .10 | .04 | .01 | .20 | .44 | .27 | |||||
| (14) Leadership structure | .96 | .21 | −.08 | −.20 | .07 | −.04 | −.02 × 10−1 | −.05 | .05 | .16 | .01 | −.04 | −.21 | −.13 | −.26 | ||||
| (15) Return on assets | 6.89 | 15.49 | −.04 | −.04 | −.05 | −.06 | .04 | .04 | −.01 | −.01 | −.04 | −.04 | −.06 | −.02 | .08 | −.07 × 10−1 | |||
| (16) Total return | 18.10 | 44.32 | .02 | .07 × 10−2 | .02 | .01 | −.05 | −.01 × 10−2 | .04 × 10−1 | −.04 | −.04 | −.05 | −.08 × 10−2 | −.01 | −.05 | .01 | .09 | ||
| (17) Firm age | 67.48 | 63.56 | .08 | .07 | .03 | .06 | .12 | −.08 | .09 × 10−1 | −.07 × 10−1 | −.01 | −.04 × 10−1 | .02 | .09 | .09 | −.12 | −.02 | −.04 | |
| (18) Firm present 1992–2006 | .33 | .47 | .11 | .23 | .04 × 10−2 | .14 | .34 | .04 | −.04 | .03 | −.05 | .10 | .21 | .17 | .39 | −.14 | .03 | .01 | .16 |
- All correlations above |.05| are significant at p < .05.
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| Main effects | ||||
| Technical balance of power | .23* (.11) | .22* (.11) | ||
| Cultural balance of power | .18 (.12) | .15 (.13) | ||
| Controls | ||||
| Debt/ext. market capitalization | −.02 × 10−4 (.08 × 10−3) | .06 × 10−4 (.08 × 10−3) | −.01 × 10−3 (.08 × 10−3) | −.02 × 10−4 (.08 × 10−3) |
| Foreign sales/total sales | .03 × 10−1 (.05 × 10−1) | .03 × 10−1 (.05 × 10−1) | .02 × 10−1 (.05 × 10−1) | .03 × 10−1 (.05 × 10−1) |
| Anglo listing | .35 (.41) | .36 (.40) | .32 (.40) | .34 (.40) |
| Dispersed shares | .01* (.06 × 10−1) | .01 (.07 × 10−1) | .09 × 10−1 (.07 × 10−1) | .07 × 10−1 (.07 × 10−1) |
| TMT/board ownership | .06 × 10−1 (.22) | .05 (.22) | .03 (.22) | .07 (.22) |
| Ownership by major banks | −.15 (.24) | −.29 (.25) | −.13 (.24) | −.27 (.25) |
| National network centrality | .10 (.19) | .08 (.19) | .10 (.19) | .08 (.19) |
| Board protection devices | .27 (.29) | .23 (.29) | .30 (.29) | .25 (.29) |
| Leadership structure | −.63 (1.13) | −.69 (1.13) | −.72 (1.14) | −.76 (1.13) |
| Log of sales | .44*** (.14) | .47*** (.14) | .46*** (.14) | .48*** (.14) |
| Return on assets | −.01 (.01) | −.02 (.01) | −.01 (.01) | −.01 (.01) |
| Total return | .03 × 10−1 (.02 × 10−1) | .03 × 10−1 (.02 × 10−1) | .03 × 10−1 (.02 × 10−1) | .03 × 10−1 (.02 × 10−1) |
| Firm age | .04 × 10−1 (.03 × 10−1) | .04 × 10−1 (.03 × 10−1) | .04 × 10−1 (.03 × 10−1) | .04 × 10−1 (.03 × 10−1) |
| Firm present 1992–2006 | −.10 (.51) | −.16 (.50) | −.17 (.51) | −.21 (.51) |
| Constant | −.05 (2.59) | .14 (2.57) | .55 (2.62) | .63 (2.61) |
| Wald Chi-square | 165.3*** | 165.3*** | 165.5*** | 165.3*** |
| d.f. | 33 | 34 | 34 | 35 |
- All models also control for industry and year effects (dummy variables). Standard errors are in parentheses. LR-tests revealed that the difference between the Wald Chi-squares of the control model and models 2 and 4 are (weakly) statistically significant (p = .03; p = .05). The maximum VIF was not higher than 2.31 in any model, which is well below the rule-of-thumb cut-off of 10. Significance test are one-tailed for directional hypothesis and two-tailed for control variables.
- *p < .05
- **p < .01
- ***p < .001
Table 3 provides our results for the random-effects regressions predicting the accounting performance implications of the espousal of an SVO. Recall that while we did not make a main effect hypothesis for the consequences of SVO adoption, it is of course noteworthy that Table 3 shows that the main effect is in fact negative, suggesting that espousing an SVO is not beneficial to a corporation in a stakeholder-oriented context. Furthermore, we find mixed support for the notion that the technical and cultural balance of power (i.e., engaging in SVO without the support of a firm's major owners) will exert a downward pressure on company performance. As predicted in H3A, the relationship between SVO espousal and subsequent firm performance is made worse the more the balance of power shifts away from major independent institutional investors. In contrast, we do not find any evidence that the predominance of non-Anglo-American investors similarly affects this relationship (H3B). Finally, as predicted in H3C, the negative relationship between a firm's SVO espousal and ROA is amplified by a lack of implementation of practices associated with the new SVO logic. In conclusion, this suggests that a company's espousal of an SVO will create downward pressure on a firm's accounting performance in a stakeholder oriented context. This relationship is intensified by two sources of misalignment: the predominance of technical owners against change and a lack of implementation of SVO practices. Finally, Table 4 reports our findings for the fixed-effects regressions predicting the market performance implications of the espousal of an SVO. Interestingly, none of the relationships between a firm's espousal and TSR is supported, suggesting that the stock market has been rather skeptical and unresponsive to the espousal and implementation of an SVO in the Dutch context during the period 1992–2006.
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |
|---|---|---|---|---|---|
| Main effect: | |||||
| SVO espousal | −.89* (.50) | −.84** (.51) | −1.00* (.51) | −.92* (.51) | |
| Moderators (substance): | |||||
| SVO implementation | −.66 (.38) | −.77* (.38) | |||
| SVO espousal × SVO implementation | .98* (.47) | 1.10* (.49) | |||
| Moderators (fit): | |||||
| Cultural balance of power | .20 (.34) | .30 (.35) | |||
| Cultural balance of power × SVO espousal | −.51 (.50) | −.75 (.51) | |||
| Technical balance of power | .27 (.30) | .33 (.30) | |||
| Technical balance of power × SVO espousal | 1.16* (.54) | 1.12* (.54) | |||
| Controls: | |||||
| Debt/ext. market capitalization | .02 × 10−5 (.03 × 10−2) | .03 × 10−3 (.02 × 10−2) | .05 × 10−3 (.02 × 10−2) | −.04 × 10−3 (.02 × 10−2) | −.04 × 10−3 (.02 × 10−2) |
| Foreign sales/total sales | −.06** (.02) | −.09 × 10−1 (.01) | −.01 (.01) | −.08 × 10−1 (.01) | −.09 × 10−1 (.01) |
| Dispersed shares | .03 (.02) | −.03* (.02) | −.03* (.02) | −.04* (.02) | −.04* (.02) |
| National network centrality | −.94 (.61) | −.11 (.43) | −.20 (.44) | −.03 (.43) | −.13 (.44) |
| Log of sales | −.18 (.59) | −.19 (.46) | −.10 (.46) | −.16 (.46) | −.09 (.46) |
| Leadership structure | −1.26 (3.30) | −1.71 (2.40) | −1.42 (2.40) | −2.12 (2.40) | −1.91 (2.40) |
| Total return | .03*** (.07 × 10−1) | .03*** (.05 × 10−1) | .03*** (.05 × 10−1) | .03*** (.05 × 10−1) | .03*** (.05 × 10−1) |
| Firm age | .05 × 10−1 (.01) | .06 × 10−1 (.01) | .07 × 10−1 (.01) | .06 × 10−1 (.01) | .07 × 10−1 (.01) |
| Firm present 1992–2006 | 2.08 (3.54) | 2.76 (3.28) | 2.83 (3.26) | 2.51 (3.25) | 2.61 (3.25) |
| Constant | 15.31 (9.12) | 17.01* (7.10) | 16.96* (7.10) | 18.52** (7.13) | 19.00** (7.14) |
| Wald Chi-square | 138.8*** | 203.7*** | 206.8*** | 214.4*** | 219.1*** |
| d.f. | 28 | 29 | 31 | 33 | 35 |
- All models also control for industry and year effects (dummy variables). Standard errors are in parentheses. The maximum VIF was not higher than 3.81 in any model, which is well below the rule-of-thumb cut-off of 10. Significance test are one-tailed for directional hypothesis and two-tailed for control variables.
- *p < .05
- **p < .01
- ***p < .001
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |
|---|---|---|---|---|---|
| Main effect: | |||||
| SVO espousal | −1.36 (3.17) | −1.78 (3.29) | −1.98 (3.29) | −2.30 (3.32) | |
| Moderators (substance): | |||||
| SVO implementation | 1.01 (2.45) | 1.01 (2.49) | |||
| SVO espousal × SVO implementation | 2.05 (3.05) | 1.77 (3.15) | |||
| Moderators (fit): | |||||
| Cultural balance of power | 2.17 (2.24) | 2.19 (2.27) | |||
| Cultural balance of power × SVO espousal | 2.01 (3.26) | 1.76 (3.36) | |||
| Technical balance of power | 1.22 (1.98) | 1.40 (1.99) | |||
| Technical balance of power × SVO espousal | .74 (3.49) | .44 (3.51) | |||
| Controls: | |||||
| Debt/ext. market capitalization | .02 × 10−1 (.01 × 10−1) | .02 × 10−1* (.01 × 10−1) | .02 × 10−1 (.01 × 10−1) | .02 × 10−1 (.01 × 10−1) | .02 × 10−1 (.01 × 10−1) |
| Foreign sales/total sales | −.15 (.10) | −.16 (.10) | −.17 (.10) | −.16 (.10) | −.16 (.10) |
| Dispersed shares | −.33** (.10) | −.31** (.10) | −.31** (.11) | −.38** (.12) | −.39** (.12) |
| National network centrality | −3.05 (2.78) | −3.54 (2.82) | −3.51 (2.84) | −3.61 (2.83) | −3.60 (2.86) |
| Log of sales | −10.78** (3.49) | −11.19** (3.54) | −11.32** (3.56) | −10.67** (3.56) | −10.82** (3.58) |
| Leadership structure | −10.46 (17.27) | −11.97 (17.38) | −11.81 (17.42) | −11.86 (17.42) | −11.87 (17.47) |
| Return on assets | −.15 (.18) | −.15 (.19) | −.17 (.21) | −.17 (.19) | −.20 (.21) |
| Firm age | 1.98 (1.32) | 2.19 (1.33) | 2.10 (1.35) | 2.10 (1.34) | 2.01 (1.35) |
| Constant | 2.07 (104.40) | −8.08 (105.47) | −3.76 (106.97) | 2.20 (105.59) | 7.37 (107.13) |
| Wald Chi-square | 17.45*** | 16.53*** | 20.16*** | 14.15*** | 13.08*** |
| d.f. | 21 | 22 | 24 | 26 | 28 |
- All models also control for year effects (dummy variables). Standard errors are in parentheses. The maximum VIF was not higher than 3.80 in any model, which is well below the rule-of-thumb cut-off of 10. Significance test are one-tailed for directional hypothesis and two-tailed for control variables.
- *p < .05
- **p < .01
- ***p < .001
Robustness and Endogeneity
We performed several post-hoc analyses to assess the robustness of our results. First, a central notion in our study has been that it is the technical balance of power between major owners that drives SVO espousal, not the presence or absence of independent institutional investors per se. Therefore, we ran a post-hoc analysis to examine whether the technical owners for SVO change and technical owners against SVO change would have an independent impact on SVO espousal. Our results indicate that this is not the case, i.e., technical owners for change (i.e., independent institutional investors) have a positive, yet insignificant effect (p = .173) on SVO espousal, whereas technical owners against SVO change (i.e., corporate and family owners) have a negative, yet insignificant effect (p = .405) on SVO espousal. Similar results are obtained if we rerun the analysis for Anglo-American-oriented (p = .819) and Rhineland-oriented investors (p = .135) separately. Overall this affirms the important role of power constellations in SVO diffusion.
Second, scholars have noted that the ownership data of listed firms in the Netherlands is less reliable before 1997 (see De Jong, Kabir, Marra, & Roell, 1998). We therefore reran our analyses for the period 1997–2006 only. Again the results indicate that our SVO models remain identical (i.e., the effect of the technical balance of power remains significant at p < .05). Third, as a further robustness check, we also assessed the market performance impact of SVO espousal using Tobin's Q instead of TSR. A fixed-effect regression (excluding financial service providers) again yielded insignificant results for all SVO effects on market performance.
Next to these robustness checks we ran several additional analyses to address two specific endogeneity concerns. First, the relationship between the technical balance of power and SVO adoption might be the result of reverse causality. SVO adoption might be a signal to the market that a listed firm is changing its direction and that it wants to attract institutional shareholders that support such an approach. To check for this possibility, we ran an additional analysis in which we regressed the SVO variable (lagged) on the technical balance of power. A fixed-effects regression indicated that SVO espousal does not have an impact on the technical balance of power (p > .10). As such, the likelihood of SVO espousal driving ownership appears to be rather small.
Second, the relationship between SVO adoption and ROA might contain a similar issue. Poor performance could prompt firms to espouse an SVO to reassure investors that the company is committed to economic value creation. Thus, the question is whether SVO adoption hampers performance or poor performance leads to long-lasting SVO adoption? However, the argument for poor performance causing SVO adoption is questionable for several reasons. First, the results of our SVO models (see Table 2) indicate that prior performance does not have any significant impact on the likelihood of SVO adoption. Second, as SVO espousal is very consistent after first-time adoption (more than 90 percent of the firms still espouse an SVO three years later), firms seem to stick to it in good and bad times. For example, the 32 firms that were present in our sample during the full observation period, on average espoused an SVO for 8.54 years after their first adoption. This time often included both the positive period before the dotcom bubble as well as the tougher years afterwards. Moreover, in this context it is particularly notable that in the years after the dotcom bubble, SVO espousal decreased for the first time since 1992. Third, to address the potential issue of an omitted variable regarding the SVO–ROA relationship, we followed Andres (2008) and employed linear instrumental regressions. As SVO espousal is very time-consistent and we only expect a short-term performance impact of SVO espousal, our one-year lagged endogenous SVO variable was instrumented by its two-year and three-year lagged values. The results of this analysis show that the relationship between SVO adoption and ROA remains identical (see Table 5). This finding tentatively indicates that the possible endogeneity of SVO adoption does not appear to drive the observed ROA performance consequences.
| Model 1 | Model 2 | |
|---|---|---|
| Main effect: | ||
| SVO espousal | −.2.75* (1.55) | −.3.13* (1.68) |
| Moderators (substance): | ||
| SVO implementation | −.76* (.40) | |
| SVO espousal × SVO implementation | 1.32** (.55) | |
| Moderators (fit): | ||
| Cultural balance of power | −.98* (.49) | |
| Cultural balance of power × SVO espousal | .16 (.65) | |
| Technical balance of power | .22 (.34) | |
| Technical balance of power × SVO espousal | .94 (.71) | |
| Controls: | ||
| Debt/ext. market capitalization | −.02 × 10−3 (.05 × 10−2) | −.03 × 10−3 (.05 × 10−2) |
| Foreign sales/total sales | −.01 (.01) | −.02 (.01) |
| Dispersed shares | .02 (.02) | .05* (.02) |
| National network centrality | .70 (.50) | .81 (.50) |
| Log of sales | .12 (.35) | .13 (.34) |
| Leadership structure | −1.91 (2.15) | −1.46 (2.11) |
| Total return | .04*** (.05 × 10−1) | .04*** (.05 × 10−1) |
| Firm age | −.02 × 10−1 (.05 × 10−1) | −.01 × 10−1 (.05 × 10−1) |
| Firm present 1992–2006 | .59 (1.23) | .64 (1.19) |
| Constant | 11.44* (5.57) | 8.66 (5.69) |
| Wald Chi-square | 177.0*** | 189.1*** |
| d.f. | 27 | 33 |
- All models also control for industry and year effects (dummy variables). Standard errors are in parentheses. Our one-year lagged endogenous SVO variable was instrumented by its two-year and three-year lagged values in this analysis. Significance test are one-tailed for directional hypothesis and two-tailed for control variables.
- *p < .05
- **p < .01
- ***p < .001
Discussion and Conclusion
We began by suggesting that research on the diffusion of a contested organizational practice could benefit from a contingent, sociopolitical framework that gave prominence to identifying technical and cultural power struggles among major owners and their subsequent impact on practice adoption and firm performance. We developed and tested such a framework in the context of the diffusion of an SVO among listed companies in the Netherlands. Our findings indicate that the balance of ownership power indeed helps to explain the diffusion of contested beliefs in the Netherlands, and further, that this contestation was only notable among technical owners for/against change. Interestingly, we found a negative main effect relationship between SVO espousal and firm performance, although this relationship is attenuated (as predicted) under specific conditions of greater SVO alignment among major technical owners, as well as a firm's greater visible commitment to an SVO.
Theoretical and Practical Implications
Our study's approach and findings suggests at least three specific implications for theory and practice. First, our results highlight how the predominance of a specific group of major owners might have a significant impact on the value orientation of public corporations. While this may at first glance seem intuitive, we see this as important for the study of the diffusion of contested ideas in two ways. First, instead of focusing on more traditional principal-agent problems generally studied under agency theory, it highlights the significance of principal-principal problems (Kim & Lee, 2008; Young et al., 2008), i.e., the struggle and costs associated with obtaining alignment among major owners of the public corporation. Importantly, we do acknowledge that when a practice becomes widely diffused and taken for granted (not contested) then political struggles are likely to be settled. Hence, it is when a practice is contested and has not reached a taken-for-granted status that the relative power balance of owners would appear to play its largest role in the adoption process. Thus, our research highlights the importance of conceptualizing, operationalizing, and modeling power constellations among major actors (in this case major owners) when trying to understand the adoption of contested organizational practice.
Moreover, while the board of directors is often seen as the most powerful body in the public corporation (at least in many stakeholder-oriented countries), our results highlight how power struggles between major owners influence a company's value orientation and related decision making. Interestingly, the increasing impact of blockholders has coincided with corporate governance reforms in Continental Europe with a stronger focus on transparency and investor participation, raising the question whether these institutional changes might have increased the likelihood of boards of directors being captured by the agendas of institutional investors (e.g., Klein & Zur, 2009; Nicholson & Cook, 2009). For scholars researching boards of directors, this hints to the importance of studying corporate decision making from a more encompassing perspective, as the balance of power among major owners might seriously affect and constrain a board's (and by implication, top management's) decision-making processes.
Second, our sociopolitical contingency framework allows us to connect the antecedents and consequences of contested organizational change, i.e., our results indicate that the power constellations for and against contested change not only affect the likelihood of adoption, but also the performance consequences of such a change. When the adoption of a contested practice finds powerful actors unprepared, this seems to significantly inhibit the desired performance improvements of such a change. In our context, this suggests that there is no single optimal governance model for all large, publicly traded Dutch firms; rather, our contingency framework and findings illustrate how and why the appropriateness of a particular corporate governance model will depend on its fit with the specific technical and political situations in which listed firms reside (Ansari et al., 2010; Rosenberg, 2005). This implies that directors should be very careful in transferring and applying new, contested corporate governance beliefs in other settings without critically analyzing their contextual fitness and making sure there is support from the most powerful actors. This also resonates well with the observation of scholars that the role of contextual factors is rather ill understood in corporate governance research and calls for future studies that more structurally compare and contrast the processes through which contested beliefs diffuse within and across various technical, political, and cultural contexts (e.g., Aguilera, Filatotchev, Gospel, & Jackson, 2008; Hambrick, von Werder, & Zajac, 2008; Yoshikawa & Rasheed, 2009).
Third, our research findings question the sustainability of the SVO logic in a stakeholder-oriented context such as that of the Netherlands. While we were agnostic as to the overall main-effect relationship between SVO adoption and firm performance, it is noteworthy that the main-effect results support criticism that the adoption of this contested practice/idea hampers firm performance (Clarke, 2013, 2014; Mintzberg, 2000; Raynor, 2008). While this may surprise some observers more enamored with the shareholder primacy perspective, we would note that an SVO does reflect a significant institutional shift for firms in a stakeholder-oriented society, such as those in our study. Our findings are not definitive, of course, but they do suggest that most listed firms in the Netherlands have been unable to derive the presumed benefits of adopting an SVO. This can be viewed as consistent with prior studies that have noted that beliefs and practices can still diffuse when they are technically inefficient (Yoshikawa & Rasheed, 2009); our study provides an explanation for this phenomenon by highlighting the role of local power constellations pushing for change. However, there may also be additional contextual factors that might discriminate between those few listed corporations that actually benefit from contested change versus the many that do not. Future research could assess this notion by examining to what extent the timing of adoption (Tüschke & Sanders, 2003; Zajac & Westphal, 2004) and the tailoring of practices to the national context (Ansari et al., 2010; Fiss et al., 2012; Yoshikawa & Rasheed, 2009) affect a firm's subsequent financial performance.
Limitations and Avenues for Future Research
The study has several limitations that also provide avenues for future research. For example, we examined whether and why companies did or did not espouse an SVO, consistent with other research that views adoption as essentially dichotomous. However, very recent research has suggested that processes associated with the adoption of (contested) corporate governance beliefs may also be more subtle (Ansari et al., 2010; Yoshikawa & Rasheed, 2009). Studies have suggested that firms may actually adopt “hybrid models” in which new beliefs are adapted and tailored to the specifics of a company's context (Fiss et al., 2012; Fiss & Zajac, 2006; Guillen, 2001; Yoshikawa et al., 2007). As such, there may be additional nuances in the rhetoric that listed firms use while espousing an SVO in a stakeholder context that future research could investigate.
In methodological terms, it would also be interesting to apply more fine-grained tools to analyze a firm's discourse, and examine boardroom discussions when a corporation's response to contested change is socially constructed. Relatedly, future research could examine how SVO-espousing firms talk about other stakeholders in the rest of their annual report, which might indicate to what extent managers perceive the preferences of shareholders and other stakeholders as additive rather than substitutive (Freeman et al., 2004; Sundaram & Inkpen, 2004). Such studies could examine the extent to which companies balance their rhetoric by referring to other stakeholders (Fiss & Zajac, 2006) or activities linked to stakeholder-oriented practices, such as non-financial performance targets tied to executive remuneration, and corporate social responsibility reports (Harrison, Bosse, & Phillips, 2010).
In addition, while our decision to investigate the diffusion of an SVO in the Netherlands contributes to our understanding of communalities and differences between institutional contexts, it also constrains the generalizability of our findings, at least with respect to the role different shareholders play in the adoption of SVO. The Dutch context has several particularities, i.e., a strong international orientation and a history of blockholdings, which may have affected the observed adoption dynamics. Moreover, we have only focused on the top-100 largest firms in the Netherlands and the transferability of our findings to smaller listed firms might therefore be questionable. Future replication studies in different (national) contexts are necessary to gauge how related political struggles among major owners vary within (broader) institutional settings.
Conclusion
We see our analysis of the contested SVO diffusion in the Netherlands as having benefited considerably from applying our contingent sociopolitical framework, which emphasizes the sources and consequences of such contestation. In particular, we found value in advancing the notion that contestation can be identified, measured, and related specifically to the technical misalignment of major owners, and that this misalignment, if not considered, can result in an SVO adoption that results in poorer firm performance. Taken together, this suggests that assessing the contextual fitness of contested beliefs is an important activity for directors when considering their adoption (Ansari et al., 2010; Volberda, van der Weerdt, Verwaal, Stienstra, & Verdu, 2012).
Our study's theoretical and empirical contributions regarding the contingent value of an SVO also suggest the value of examining whether the observed development in the Netherlands is part of a two-way convergence of governance models. In other words, we would welcome research that examines whether Anglo-American-oriented firms have also started to incorporate elements of the stakeholder model in response to the internationalization of their shareholder base and the global financial crisis. The diffusion of corporate governance models is an ongoing process whose continued study will show the extent to which national differences will persist or global governance models will converge. We believe that our sociopolitical perspective on diffusion, with its emphasis on misalignment and the joint impact of technical and political factors, can provide needed insights into understanding where and when contested corporate governance change is more likely to take root, and whether the performance effect of such changes will be beneficial or detrimental, in terms of both the involved firms and the societies in which they are situated.
Acknowledgements
We highly appreciate the helpful comments of Kevin Boeh, Lionel Page, Amedeo Pugliese, Michiel Tempelaar and Dennis Veltrop, and the guidance provided by the Editor, Associate Editor and anonymous reviewers. Furthermore, we thank Spencer Stuart Amsterdam for the data from the Netherlands Board Indexes, the Erasmus Research Institute of Management for their financial support and Abe De Jong for his generosity in sharing his Dutch listing data. Moreover, we appreciate the excellent research assistance by Martin Dongelmans, Johannes Drees, Arlette Guijt, Mariano Heyden II, Rick Hollen, and Lotte van Houwelingen, without whom the extensive data collection process would not have been possible. Earlier versions of this paper have been presented at the Academy of Management, European Academy of Management and Strategic Management Society.
Notes
Biographies
Pieter-Jan Bezemer is a Post-Doctoral Research Fellow in the School of Accountancy at Queensland University of Technology. His research interests encompass several governance issues of corporations, with a particular focus on board dynamics and the diffusion of contested governance beliefs.
Edward J. Zajac is the James F. Beré Professor of Management and Organizations at the Kellogg School of Management, Northwestern University. His research focuses on the integration of economic and behavioral science perspectives on corporate governance, organizational adaptation, and strategic alliances.
Ivana Naumovska is a Post-Doctoral Fellow at the Kellogg School of Management, Northwestern University. Her research revolves around the organizational and strategic questions that emerge when considering how firms and financial markets interact in a world influenced by institutional and social psychological factors.
Frans A. J. Van Den Bosch is Professor of Management Interfaces between Organizations and Environment, Rotterdam School of Management, Erasmus University. He has published several books and over 165 articles in scientific journals and book chapters in the areas of strategy, international business, general management, economics, and industry studies. His scientific papers have appeared in, among others, Academy of Management Journal, Business and Society, Journal of Management Studies, Management Science, and Organization Science. He is a board member of several scientific journals, including Long Range Planning and Organization Studies.
Henk W. Volberda is Professor of Strategic Management and Business Policy and Director of Knowledge Transfer at the Rotterdam School of Management, Erasmus University. He is also Scientific Director of the top institute INSCOPE: Research for Innovation. He has been a visiting scholar at the Wharton School at the University of Pennsylvania and Cass Business School, London. His research on strategic flexibility and renewal received many awards, including the Igor Ansoff Strategic Management Award and the SAP Strategy Award 2005. His work on strategic renewal, coevolution of firms and industries, new organizational forms, and innovation has been published in many high-impact journals such as Academy of Management Journal, Management Science, Organization Science, and Strategic Management Journal. He is serving as a member of the Editorial Review Board of Long Range Planning, Organization Studies, and Organization Science.
