American Sociological Review 2004 Zajac 433 57

American http://asr.sagepub.
com/ Sociological Review
The Social Construction of Market Value: Institutionalization and Learning Perspectives on Stock Market Reactions
Edward J. Zajac and James D. Westphal American Sociological Review 2004 69: 433 DOI: 10.1177/000312240406900306 The online version of this article can be found at: http://asr.sagepub.com/content/69/3/433
Published by:
http://www.sagepublications.com
On behalf of:
American Sociological Association
Additional services and information for American Sociological Review can be found at: Email Alerts: http://asr.sagepub.com/cgi/alerts Subscriptions: http://asr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://asr.sagepub.com/content/69/3/433.refs.html
>> Version of Record - Jun 1, 2004 What is This?
Downloaded from asr.sagepub.com by Daniel Silva on November 1, 2013
#1600-ASR 69:3 filename:69306-zajac
Edward J. Zajac
Northwestern University
James D. Westphal
University of Texas at Austin
This study advances a social constructionist view of financial market behavior. The paper suggests that the markets reaction to particular corporate practices, such as stock repurchase plans, are not, as financial economists contend, simply a function of the inherent efficiency of such practices. Rather, stock market reactions are also influenced by the prevailing institutional logic and the degree of institutionalization of the practice. The theory first predicts that the emergence of the agency perspective on corporate governance in the mid-1980s represented a powerful new institutional logic that would lead the market to reverse its prior aggregate reaction to stock repurchase plans in the United States. The paper then considers the potential for institutional decoupling of repurchase plans and develops competing hypotheses about how the market value of these policies might have changed as more firms formally adopted, but did not implement, the plans over time. In contrast to a financial economic perspective on market valuation, which suggests that markets should discount the value of a policy as evidence of non-implementation accumulates, this study posits that institutionalization processes might increase the market value of a policy as more firms adopt it, despite growing evidence of decoupling. Implications for institutional theory and theoretical perspectives on capital markets are discussed.
conomic and organizational sociologists have devoted increased scholarly attention to processes of institutionalization, or mechanisms by which organizational structures, policies, and practices acquire social legitimacy and ultimately become taken-for-granted as
Both authors contributed equally. Direct all correspondence to Edward J. Zajac, Department of Management and Organizations, Northwestern University, Evanston, IL 60208-2001 (e-zajac@ northwestern.kellogg.edu). The helpful comments and suggestions of Randy Beatty, Ranjay Gulati, Andrew Henderson, Paul Hirsch, William Ocasio, Don Palmer, Huggy Rao, and seminar participants at Duke University, University of Chicago, University of Southern California, University of Minnesota, University of Maryland, Emory University, and the Pennsylvania State University are greatly appreciated.
normatively appropriate in a population. From a neoinstitutional perspective, organizational structures and practices acquire legitimacy with an organizations stakeholders to the extent that they are consistent with prevailing institutional logics, or historically-variant sets of assumptions, beliefs, values, and rules by which individuals . . . interpret organizational reality and what constitutes appropriate behavior (Thornton and Ocasio, 1999:804; see also Friedland and Alford 1991; Scott 2001). Empirical studies in this growing literature have examined how historical change in prevailing institutional logics can lead to change in organizational policies and practices or to different interpretations of a given policy. Thornton and colleagues, for instance, showed that change in institutional logics in the publishing industry from an editorial logic (which conceived publishing as a profession) to a market logic (which conceived publishing as a business) led to
AMERICAN SOCIOLOGICAL REVIEW, 2004, VOL. 69 (June:433457)

434AMERICAN SOCIOLOGICAL REVIEW
change in executive succession practices, organizational structure, and acquisition targets in the industry (Thornton 2001, 2002; Thornton and Ocasio 1999). Moreover, Fligstein and colleagues demonstrated how a shift in prevailing conceptions of corporate control from a manufacturing logic to a finance logic contributed to the spread of multidivisional forms of organization (Fligstein 1990; Fligstein and Brantley 1992). While organizational policies and practices can acquire legitimacy and spread in a population to the extent that they are consonant with prevailing institutional logics, such policies also acquire further legitimacy by virtue of their prevalence in a population. Fligstein (1990, 1991) found not only that firms were more likely to adopt the multidivisional form following the advent of a finance conception of control but also that firms were most likely to adopt this innovation when many other firms in their organizational field had already done so. This finding can be interpreted as suggesting that as more firms adopted the multidivisional form it became progressively institutionalized, or takenfor-granted, as an appropriate means of organizing production (cf., Chaves 1996; Davis and Greve 1997; Tolbert and Zucker 1983). Zajac and Westphal (1995) extended this literature by showing that as prevailing beliefs about corporate governance changed to suggest a more positive view of certain corporate policies (e.g., CEO incentive plans), and as such policies became more prevalent among large firms, a focal firm was more likely to formally adopt the plans but less likely to implement them. They suggested that as executive incentive plans became institutionalized, building symbolic value as normatively appropriate elements of corporate governance policy, firms were more likely to formally adopt the plans (to enhance their external legitimacy) while decoupling the plans from actual governance practices (to preserve informal routines and political interests of executives) (Edelman et al. 1991; Meyer and Rowan 1977). This literature has thus demonstrated that institutional processes can influence how organizational actors perceive corporate policies, affecting firm-level adoption of policies and structures and field-level change in prevailing organizational practices. Yet, little research has examined whether and how these processes
influence the perceptions of actors in financial markets and ultimately determine the market value of corporate policies. Theories of capital markets are dominated by a financial economics perspective in which the stock markets reaction to a policy adoption is conceived as a reliable, historically invariant indicator of the efficiency benefits gained from adopting the policy (David 1997; Fama 1970; Timmerman 1993). The market is seen as responding to the value of corporate policies according to consistent criteria of technical efficiency. This response would only change over time to the extent that market actors acquire new information about the degree to which the policy satisfies those criteria and if they update their assessment of the policys efficiency benefits accordingly. This perspective does not recognize the potential influence on market actors of historical change in prevailing beliefs about the sources or criteria of organizational efficiency. Moreover, from this perspective, if there is evidence that firms often adopt, but do not implement a corporate policy (i.e., efficiency benefits are not realized), then the stock market should incorporate this evidence and discount the value of the policy (i.e., the policy would receive a less positive market reaction). We develop a sociological perspective on the process and outcomes associated with the stock market valuation of corporate policies and practices, thus providing an alternative to the dominant, financial economics perspective. Our perspective highlights how investment behavior is governed by social dynamics that can compromise market efficiency. This theoretical approach is consistent with recent sociological forays into the study of financial markets. For example, studies of trader markets have shown that investors base their investment decisions on the actual or anticipated decisions of other market players rather than on independent forecasts of the assets economic performance (Abolafia 1996; Knorr Cetina and Bruegger 2002; MacKenzie 2003). The importance of imitation or social referencing in investment decisions can, in fact, be traced to statements by Keynes (1936) and Merton (1948), as others have noted (see Zuckerman 1999). We build on this literature by considering theoretically and empirically how different historical and institutional contexts condition these social dynamics, and

SOCIAL CONSTRUCTION OF MARKET VALUE435
thus how institutionalization processes can shape investment behavior. Specifically, we begin by introducing the notion that historical shifts in institutional logics have the potential to dramatically change the prevailing interpretation of an organizational policy, influencing the reaction of investors to the policy, and ultimately affecting the market value of the policy to the firm. We focus on the domain of corporate governance and propose that a change in institutional logics toward an agency logic of governance, in which managers are assumed to prefer strategies that often conflict with shareholder interests, will affect positively the markets assessment of policies (such as stock repurchase plans), interpretable as consistent with this emerging institutional logic. We then consider the potential for the institutional decoupling of repurchase plans and develop hypotheses about how the market value of these policies may have changed as more firms formally adopted and decoupled these plans over time. In contrast to a financial economic perspective on market reactions, which suggests that the market value of repurchase plans should diminish as evidence of nonimplementation accumulates, we propose a sociological perspective on institutionalization processes that addresses how and why the market value of these policies may actually increase as more firms adopt them over time, despite accumulated evidence of decoupling. Thus, we seek to contribute to the existing literature on the sociology of markets and organizations by analyzing theoretically and empirically when, how, and why the market value of corporate policies are historically contingent (given different dominant institutional logics), and how processes of institutionalization that occur in industrial markets also occur in capital markets (cf., Fligstein 1991; Tolbert and Zucker 1983; Zajac and Westphal 1995). We show how our perspective leads to the notion that stock market reactions to corporate policies are socially constructed and do not necessarily reflect expected efficiency benefits. We contrast this line of reasoning with dominant economic assumptions about capital markets, which are viewed by economists as the paragon of allocative efficiency. In particular, we seek to extend neoinstitutional theory in two ways: (1) We link the social dynamics of financial markets with the processes and outcomes of institutionaliza-
tion, and (2) we show how the phenomenon of institutional decoupling (Meyer and Rowan 1977) is related to the process of institutionalization (Zucker 1983). We propose below that investors are likely to reference prior market reactions to similar events in estimating the reactions of other investors to the adoption of the focal policy, and we further propose that this social estimation process causes the value of corporate policies to become increasingly takenfor-granted, even as the rate of decoupling increases over time. THEORY AND HYPOTHESES
OF
SHIFT FROM CORPORATE TO AGENCY LOGICS GOVERNANCE
Prior research has documented a significant change in prevailing beliefs and assumptions about corporate governance in the mid-1980s. In particular, several authors have suggested that dominant beliefs about corporate governance shifted toward an agency model of corporate control (Davis and Thompson 1994; Useem 1993, 1996; Zajac and Westphal 1995:283). We characterize the emergence of this new conception of governance as a shift in institutional logics. It entailed change in fundamental assumptions about the very purpose of corporate governance, thus shaping normative beliefs about a wide range of business practices. Moreover, these new assumptions and beliefs about governance were linked to higher-order societal logics of economic activity (Friedland and Alford 1991; Scott 2001). Table 1 summarizes key elements of the agency logic and compares it with the model it replaced. The agency logic proceeds from different assumptions about managers. Useem (1993) showed that in the mid-1980s discourse about corporate governance was increasingly premised on the belief that managers, if left to their own devices, tend to pursue strategies that advance their own interests at the expense of shareholders. There was a growing perception that the managerial revolution had afforded executives the privilege of ignoring shareholders if they so chose (Useem 1993:6; see also Davis and Thompson 1994; Donaldson 1990; Zajac and Westphal 1995). Managers self-interested strategies were thought to generate agency costs that reduce shareholder value (Fama and Jensen 1983:304). As the agency conceptual-

Table 1. Agency Logic Versus Corporate Logic of Governance X Assumptions about: Top managers X Agency Logic X Relatively fungible agents of shareholders Pursue strategies that advance personal interests at expense of shareholders. A nexus of contracts; legal fiction. Investor capitalism: shareholders can diversify better and more easily than firms. Corporate Logic X Professionals with unique strategic knowledge that is required for efficient allocation of corporate resources. Stewards of their organizations. An institution with unique core competencies. Managerial capitalism: top managers have primary responsibility for allocating resources to different businesses in the corporation. Norms of professional autonomy Managerialist theory (Chandler, 1962) X Use salary and other rewards to attract and retain scarce management talent. Retain and reinvest.
The firm Concept of resource allocation
Links to high-order cultural frames Links to theories of organization Implications for governance practices: Compensation
Logic of capitalist markets Agency theory (Jensen and Meckling, 1976) X Use incentives to align management and shareholder interests. Redistribute to shareholders.
Allocation of cash flow
ization of managers as self-interested grew, this view displaced the pre-agency emphasis on valuing managers for their professional expertise. Whereas top executives had previously been viewed as professionals who possessed the unique strategic knowledge required for efficient allocation of corporate resources, in the 1980s they were increasingly characterized as relatively fungible agents of shareholders. Useem (1993:21) quoted investor Carl Icahn as referring to executives as drones. Useems qualitative research, together with Zajac and Westphals (1995) archival analyses of proxy statements and Davis and colleagues (Davis, Diekmann, and Tinsley 1994; Davis and Thompson 1994) analyses of Securities and Exchange Commission (SEC) documents, court decisions and publications of various investor groups, all provided evidence that this shift in assumptions about top executives was prevalent among the full spectrum of corporate stakeholders. The agency logic also proceeds from different assumptions about the firm. In the agency conception, there is nothing uniquely valuable about the corporation itself. It is merely a nexus of contracts between individuals (Jensen and Meckling 1976:311); the very notion of a cor-
poration is dismissed as a legal fiction (Davis et al. 1994:548). As Davis and colleagues have suggested, this conception of the firm eclipsed the longstanding notion that corporations were institutions with unique core competencies. Agency assumptions led to a different model of economic resource allocation, which several authors have characterized as investor capitalism, in contrast to managerial capitalism (Davis and Thompson 1994; Useem 1996:1; Westphal and Zajac 1998). Specifically, if managers are merely fungible agents with no particular strategic expertise, and if firms are merely nexus of contracts without unique core competencies, then resources can be allocated by investors in capital markets rather than by executives in corporations. If executives are also presumed to generate agency costs, then capital allocation would be better left to investors. Agency assumptions also led to different interpretations of specific governance practices. In a study of verbal explanations for executive incentive plans in proxy statements, Zajac and Westphal (1995) found that incentive plans in the 1970s were typically justified as a mechanism to attract and retain scarce executive talent (consistent with the traditional corporate

conception of managers as professionals who have unique strategic expertise). In the mid-tolate 1980s, these plans were more frequently justified as a mechanism to align manager and shareholder interests (consistent with the agency logic). Governance researchers have generally attributed the rise of the agency logic to investor dissatisfaction with the performance of large U.S. companies beginning in the mid- 1970s. Useem (1993) provided qualitative evidence that when corporate performance problems continued into the early 1980s, despite large-scale deregulation and a generally favorable political climate for business, the search for causes veered in on management itself (p. 21). In particular, performance problems of large companies were widely attributed to excessive levels of corporate diversification, which in turn were increasingly attributed to empire-building by top executives (Useem 1993:21). According to Useem (1993) and Davis and colleagues (Davis, Diekmann, and Tinsley 1994; Davis and Thompson 1994), this attribution of firm performance problems to self-interested decision making by managers was reinforced by (and to some extent inspired by) the propagation of agency theory by f inancial economists. Ultimately, however, the popularity of agency theory as a perspective on corporate governance derives from its consonance with fundamental, normative beliefs about the organization of economic activity. Friedland and Alford (1991:250) and Scott (2001) have suggested that institutional logics draw legitimacy from higher-order logics or cultural frames to which they are connected. While the corporate logic of governance draws legitimacy from norms of professional autonomy (Zajac and Westphal 1995), the agency logic draws legitimacy from its connection to the logic of capitalist markets (Friedland and Alford 1991), given its emphasis on allocative efficiency through the invisible hand of the stock market rather than through the visible hand of cor porate managers (Donaldson 1990). To the extent that such a market logic of resource allocation became more prevalent in some sectors of the economy in the 1980s (e.g., Scott 2000; Thornton 2001; Thornton and Ocasio 1999), the legitimacy of the agency logic may have been further enhanced.
INSTITUTIONAL LOGICS OF GOVERNANCE AND MARKET REACTIONS TO REPURCHASE PLANS

We now consider how the described shift in institutional logics of governance may have changed the perceived value of a specific corporate governance policy, namely stock repurchase plans, resulting in changing stock market reactions to this policy over time. A stock repurchase or buyback plan is a formal (i.e., written) policy approved by the board of directors to buy a portion of the firms outstanding shares back from investors, typically in the open market, and retire them (Franz, Rao, and Tripathy 1995). The plan typically specifies the number of shares authorized by the board for repurchase and describes how the buybacks will be funded (typically with cash). Repurchase plans grew more prevalent in the mid-1980s and remained quite common throughout the 1990s among mid- and large-sized companies (Grullon and Ikenberry 2000). From the mid-1980s to the mid-1990s, U.S. firms announced plans to buy back more than $1 trillion of their stock. Financial economists have developed a literature on stock repurchase plans that consists primarily of event studies, which in this context assess the value of a policy according to the stock markets reaction to its adoption. The findings of this literature are remarkably consistent. Published studies typically demonstrate positive, and often very large and persistent stock market reactions to the adoption of repurchase plans (cf., Lee, Mikkelson, and Partch 1992; Medury, Bowyer, and Srinivasan 1992; Raad and Wu 1995; Ratner, Szewczyk, and Tsetsekos 1996). In explaining these findings, financial economists have typically invoked an agency logic. For instance, Jensen (1989) explained the markets enthusiastic response to repurchase plans by suggesting that when managers have free cash flow at their disposal they are tempted to waste it on empire-building projects or other perks that benefit themselves at the expense of shareholders (p. 64); thus, shareholders benefit from policies such as repurchase plans that return free cash flow to investors and allow them to invest it themselves (Bagwell and Shoven 1988; Lazonick and OSullivan 2000:17; Medury et al. 1992). Financial gurus and members of the business press have offered similar interpretations. For example, the renowned investor Warren Buffett gave a widely quoted explanation for positive

reactions to stock buybacks that invoked the agency logic:

By making repurchases . . . management clearly demonstrates that it is given to actions that enhance the wealth of shareholders, rather than to actions that expand managements domain but do nothing (or even harm) shareholders. Seeing this, shareholders and potential shareholders increase their estimates of future returns from the business, which produces higher market values. (San Francisco Chronicle, July 31, 1995, p. B1 1995)
From a financial economics perspective, event studies indicate the consequences of policy adoptions for economic efficiency (David 1997; Fama 1970). We suggest, however, that stock market reactions are based on investor perceptions about the sources of efficiency, rather than on any unchanging standard of efficiency. More specifically, we contend that investor perceptions, like the perceptions of other corporate stakeholders, are structured by historically variant institutional logics. As Friedland and Alford (1991) suggest, any behavior/activity can carry with it alternative meanings. . . . There is no one-to-one relationship between an institution and the meaning carried by the practices associated with it (p. 250). In this case, while the agency logic of governance suggests a positive interpretation of stock repurchase plans, the earlier corporate logic suggests a different interpretation. From the latter perspective, as discussed above, top managers are professionals with unique strategic knowledge that enables them to allocate corporate resources efficiently. Accordingly, they should tend to allocate free cash flow only to profitable projects. If such projects are not available, they would return the cash to shareholders to invest in other companies with more attractive prospects. Thus, from the corporate logic perspective, a stock repurchase plan could be perceived negatively as evidence that a firm has only limited prospects for investing its money. We suggest, therefore, that stock market reactions to repurchase plan adoptions have changed over time, driven by changes in the prevailing institutional logic of governance. Specifically, we predict that financial investors perceptions of stock repurchase plans starting in the mid1980s (i.e., the period when the agency logic of governance became a dominant institutional logic) will be generally positive, resulting in
positive market reactions to the adoption of repurchase plans. We also expect, however, that prior to this time period (i.e., when the corporate logic of governance was dominant), financial investors perceptions of stock buybacks will be generally negative, resulting in negative market reactions to the adoption of these plans. Some financial economists have suggested that investors react positively to repurchase plan adoption in part because they view adoption as a signal that managers believe the firm is undervalued (Ikenberry, Lakonishok, and Vermaelen, 2000). However, this perspective does not address the potential for market reactions to change from negative to positive over time, which is the focus of our theory. This suggests the following, initial hypothesis for the periods of our study: Hypothesis 1: Stock market reactions to repurchase plan adoptions shift from negative to positive in the mid-1980s.
THE (NON)IMPLEMENTATION OF STOCK REPURCHASE PLANS

Our discussion thus far has not distinguished between the formal adoption of stock repurchase plans and the actual implementation of those plans. From a neoinstitutional perspective, when firms adopt policies in an act of conformity to prevailing institutional logics, they may also decouple the formally adopted policy from informal routines in the organization. Decoupling reflects the overriding [imperative] of all systems .|.|. to maintain the integrity and continuity of the system itself .|.|. including the stability of informal [routines and power relationships] within the organization (Scott 1998:117), while still conforming to institutional environments (Selznick 1957). In a previous study (Westphal and Zajac 2001), we posited that institutional decoupling can occur in the context of stock repurchase plans. We demonstrated that firms may formally announce a plan to repurchase stock and then subsequently buy back only a small fraction of the shares targeted for repurchase in the formal plan; or they may make no repurchases at all (see also Stephens and Weisbach, 1998). We suggested that such nonimplementation constitutes an institutional decoupling, whereby managers symbolically demonstrate commitment to agency values by formally adopting repurchase

plans, and then the managers neglect to redistribute free cash to shareholders to preserve informal routines and discretionary control of corporate resource allocation within the firm. In our multivariate analyses, significant predictors of whether firms were likely to adopt but not implement repurchase plans included CEO power, network ties, and a firms history of prior decoupling, but not post-announcement stock returns or market-to-book value. Our present study includes data on repurchase plans from 1980 to 1994 among a sample of 463 Forbes 500 companies. Figure 1 shows repurchase plan adoption and implementation during this time period. The graph shows a sharp increase in adoptions in the mid1980s. It also suggests that the prevalence of nonimplemented repurchase plans increased during this period. Thus, firms were more likely to formally adopt repurchase plans during a period in which the agency logic of governance prevailed, but they were also less likely to actually implement the formally adopted plans. From a neoinstitutional perspective, it appears that repurchase plan adoptions became more symbolic and less substantive over time (Edelman et al. 1991; Pfeffer 1981). While firms may gain legitimacy benefits in the stock market from adopting policies that conform to prevailing institutional logics, we must also consider whether such policies might lose their symbolic value as evidence of nonimplementation emerges. Would any positive market reactions to repurchase plans resulting from the rise
of an agency logic of governance in the mid1980s diminish in later time periods, or perhaps disappear altogether, as evidence of nonimplementation accumulates? We develop competing hypotheses regarding the effects of prior instances of adopting, but not implementing, stock repurchase plans on the market reaction to these plans.
MARKET LEARNING VERSUS INSTITUTIONALIZATION PERSPECTIVES

From a financial economics perspective, stock market reactions to repurchase plan adoptions reflect the markets assessment of the efficiency benefits of adopting and implementing the plans (Bagwell and Shoven 1988; Raad and Wu 1995; Ratner et al. 1996). That assessment, in turn, is influenced by prior evidence regarding the economic benefits of adoption. The larger event study literature typically assumes semistrong capital market efficiency, wherein all publicly available information about a given policy at the time of adoption is reflected in the stock markets response (Brealy and Myers 1991; Fama 1970). Although this literature has not formally examined change in market reactions over time, it is assumed that the market has responded less positively over time to corporate strategies, such as unrelated acquisitions, because of an absence of economic benefits from the policy. In effect, the market reaction to a particular policy adoption is assumed to derive from a sophisticated and rapid statistical
100 90 80 70 60 50 40 30 20 10 0 1980 1982 1984 1986 1988 1990 1992 1994

Year
Figure 1. The Adoption and Decoupling of Stock Repurchase Programs: 19801994
Number of Plans
Not Implemented Implemented

analysis, based on all publicly available data on prior adoptions and subsequent changes in economic performance. As more firms adopt the policy and their economic efficiency increases or decreases, these cases are added to the analysis, and the expected benefits from future adoptions are modified based on the revised results. A corollary of this perspective is that investors who are less informed about the efficiency benefits of particular policies (i.e., investors who do not receive information about the consequences of prior adoptions or who receive the information relatively late) are less likely to remain in the market over the long term, because they earn a lower return. Thus, market knowledge about the efficiency benefits that firms have enjoyed from adopting a particular policy should tend to rise to the level of the most knowledgeable investor (Brealy and Myers 1991; Fama 1970). In this way, markets should learn about the value of a particular policy through prior experience with it. There is a growing theoretical literature in information economics on stock market learning. However, theorists have typically assumed that market learning does occur. The focus in this literature is on varying key parameters of the learning process, such as the speed with which investors respond to new information (e.g., David 1997; King et al. 1993; Timmermann 1993). Consequently, there is little empirical evidence for (or against) the underlying assumption that the market learns (Bradley, Desai, and Kim 1983). Some research has examined whether market returns from adopting a policy that is assumed to benefit shareholders, such as a related acquisition, appears to dissipate if the policy is not subsequently consummated (Bradley et al. 1983). However, as several studies in the event study literature have shown, the measurement of market reactions over extended time periods is fraught with difficulties (McWilliams and Siegel 1997). Moreover, prior research has not directly examined whether markets appear to learn about particular policies, such that the markets reaction to a policy is informed by prior experience with the same policy at other firms. Thus, we posit an initial hypothesis based on the market-learning perspective. As discussed above, in the mid-1980s, firms increasingly adopted stock repurchase plans without actually implementing the plans. When these repur-
chase plans were not implemented, they did not, of course, enhance economic efficiency as was expected by agency theorists. Consequently, from a market-learning perspective, as more firms neglected to implement their repurchase plans over time, the expected efficiency benefits from adoption should diminish, and the markets reaction to the adoption of these plans should diminish accordingly. Note that this would be true from a market-learning perspective even if other motives for non-implementation were sometimes also involved, since investors would, on average, come to expect fewer agency benefits from announced repurchase plans, given growing accumulated evidence of non-implementation. Thus we have the following hypothesis: Hypothesis 2 (H2): The number of firms that have adopted, but not implemented, stock repurchase plans is negatively associated with the stock market reaction to repurchase plan adoption at the focal firm. A neoinstitutional perspective suggests a competing prediction regarding the change in market reactions to repurchase plans as more firms adopt (but do not implement) these plans over time. From this perspective, repurchase plans may accumulate symbolic value over time, despite repeated instances of decoupling, through a process of institutionalization. Institutionalization refers to the social construction process by which organizational policies become instilled with value and ultimately taken-for-granted among external constituents as normatively appropriate (Selznick 1996; Tolbert and Zucker 1983:25). Institutionalization is more likely in the context of a supraorganizational belief system that offers a positive interpretation of the policy (DiMaggio 1997; Scott 1994). In this case, the advent of an agency logic of governance suggested a more positive interpretation of repurchase plans. Moreover, as more firms adopt repurchase plans over time, the plans should become more strongly associated with the agency values that they affirm. From a neoinstitutional perspective, policies build symbolic value through reciprocated interpretations in which a policy is presented to constituents as furthering their values and that interpretation is validated by the constituents (Berger and Luckmann 1967; Meyer and Rowan 1977; Scott 1987:496, 1994). As this pattern is

repeated, the connection between the policy and the belief system or institutional logic tightens and may ultimately become taken-for-granted. Thus, stock repurchase plans may gradually accumulate symbolic value and thus become institutionalized as more firms adopt them over time. We consider how purely ceremonial acts of conformity to agency values, in the form of decoupled stock repurchase plans, could accumulate symbolic value in this way. Through examining change in market reactions to decoupled repurchase plans as more firms adopt them, we are able to directly assess change in their symbolic value. As noted above, although market reactions are viewed in the financial economics literature as providing objective data that reflect expected efficiency gains from policy implementation, from an institutional perspective, we view such reactions as subjective data that reflect the symbolic value of adoption, neatly quantified and aggregated, and we expect that these numbers will increase as the policy becomes institutionalized. How could repurchase plans accumulate symbolic value, as indicated by more positive investor reactions, despite growing evidence that they are often not implemented? We suggest that, since market reactions to announced policy adoptions occur quickly and are based on imperfect communication among participants, investors are likely to predict the response of other investors to a policy adoption by referring to prior market responses to similar events. This sociohistorical estimation process serves to perpetuate market reactions. As more firms adopt a policy and receive a favorable market response, an individual investors uncertainty about the likely response to the current adoption is reduced, which should tend to result in a more positive reaction (Westphal and Zajac 1998). Thus, the market value of repurchase plans can build over time through a self-perpetuating, social construction process despite the potential for decoupling. As more firms adopt the plans and receive a positive reaction, the value of the plans becomes increasingly takenfor-granted among individual investors in that their individual beliefs about the plans become less relevant to the evaluation process and are less likely to be referenced in any meaningful way. In this way, the social referencing process that underlies stock market reactions to policy adoptions can contribute to the institutional-
ization of corporate policies, and this occurs even as evidence accumulates that the policies do not enhance economic efficiency as originally promised (e.g., due to nonimplementation). Moreover, from an institutional perspective, as a policy becomes institutionalized such that it becomes closely identified with an institutional logic, the collective reaction of external constituents will increasingly be governed by a logic of confidence and good faith, whereby constituents (e.g., investors) accept it on faith that organizational actors are behaving in accordance with the legitimate goals of the policy (Meyer and Rowan 1977:357). This shared assumption that legitimate formal policies are faithfully implemented not only simplifies the decision-making process for investors, but also has the ultimate function of maintaining confidence in organizations and in the extent to which they are efficiently governed (Meyer and Rowan 1977:358). Thus, in opposition to Hypothesis 2, the market-learning hypothesis (a neo-institutional perspective on stock market dynamics) leads to the following hypothesis: Hypothesis 2a: The number of firms that have adopted, but not implemented, stock repurchase plans is positively associated with the stock market reaction to repurchase plan adoption at the focal firm. MARKET
LEARNING VERSUS INSTITUTIONAL-
Prior event studies have treated different corporate governance policies and strategies as independent events. Research on repurchase plans, for instance, has examined the stock market reaction to such plans independently of other, related policy adoptions. However, the markets reaction to a particular policy may be influenced by the prior experience of investors with other policies that had similar objectives. From a market-learning perspective, stock repurchase plans can be viewed as one element of a larger set of policies that purport to control agency costs in the firm. To the extent that market actors recognize the common objectives that underlie different policies, the potential for market learning is broadened. Thus, market reactions to repurchase plans could be influenced by the failure to implement other governance
IZATION ACROSS RELATED POLICIES.

policies that were ostensibly adopted to control agency costs. Prior to widespread adoption and decoupling of repurchase plans, many large firms had adopted executive incentive plans known as LTIPs (long-term incentive plans). These were alleged to reduce agency costs by aligning CEO interests with the interests of shareholders (Kumar and Sopariwala 1992).1 Many firms explicitly invoked an agency logic in announcing LTIP adoption to shareholders in proxy statements, describing the plans as incentive alignment or control mechanisms that discourage executives from making decisions that advance their own interests at the expense of shareholders (Wade, Porac, and Pollack 1997; Zajac and Westphal 1995:283). Thus, the rationale for LTIP adoption, like the rationale for stock repurchase plans, was rooted in the agency logic of governance. Whereas LTIPs partially resolve the agency problem by giving CEOs a financial incentive to pursue shareholder objectives (i.e., rather than wasting corporate cash on empire-building strategies [Rajagopalan 1996]), repurchase plans prevent CEOs from wasting cash flow by taking it out of their hands entirely and returning it to investors (Jensen and Warner 1988). Moreover, while implementation of repurchase plans decreased in the mid-to-late 1980s, in the early 1980s, many firms formally adopted an LTIP without making grants under the plan (Westphal and Zajac 1998). This represents an earlier instance of decoupling a corporate governance policy purportedly designed to control agency costs. Thus, the market-learning perspective suggests that as more firms adopt LTIPs with an agency rationale and then fail to implement the plans (i.e., agency costs are not actually reduced), the market may not only discount the expected value of LTIPs but it may also lower its estimation of other policies that purport to reduce agency costs, including stock repurchase plans. In effect, the market learns to recognize the potential for nonimplementation of governance policies in general, as investors learn from their experience with LTIPs to be skeptical about policies that
claim to control agency costs. This suggests the following hypothesis: Hypothesis 3: The number of firms that have adopted LTIPs with an agency explanation but have not implemented the plans is negatively associated with the stock market reaction to repurchase plan adoption at the focal firm. However, a neoinstitutional perspective suggests an additional, competing prediction regarding the effect of prior LTIP adoptions on the stock markets reaction to repurchase plan adoptions. Just as repurchase plans can be seen as acquiring greater symbolic value as more firms adopt them over time, the symbolic value that accumulates to a policy from repeated instances of adoption and endorsement by an external audience may spread to other policies that can be interpreted as having a similar objective. As discussed earlier, institutionalization occurs through a social construction process in which a policy is presented to constituents as furthering a set of objectives, and that interpretation is validated by constituents. As this pattern is repeated over time, the benefits of the policy become increasingly taken-for-granted. An institutional perspective also suggests that as more firms adopt a particular policy and invoke the same or similar logic for adoption, not only the policy but also the stated objective itself can build institutional value and become taken-forgranted as a legitimate rationale for action (i.e., the assumptions that underlie the alleged benefits from adoption are less likely to be questioned) (Scott 2001; Zajac and Westphal 1995). Thus, subsequent policies that appear to conform to the same logic enjoy greater social acceptance, and firms realize greater legitimacy benefits from adopting them (Meyer, Boli, and Thomas 1987; Scott 1994, 2001; Suchman 1995). By virtue of this process, the prior diffusion of executive incentive plans for top managers may have enhanced the legitimacy of stock repurchase plans. As more firms adopted LTIPs with an agency rationale, the agency logic itself acquired greater legitimacy as a socially accepted rationale for corporate action. Thus, as LTIPs became widely diffused in the mid-1980s, firms realized greater legitimacy benefits from adopting other policies that appeared to control agency costs, including stock repurchase plans.
LTIPs have been formally defined as incentive programs that grant executives the right to receive common stock or cash on a particular date in the future to the extent that specific performance objectives are met (Crystal 1991).

In this way, repurchase plans may have borrowed symbolic value from the prior diffusion of LTIPs. Moreover, this process could operate despite prior decoupling of LTIPs. The social estimation process that underlies the institutionalization of stock market reactions, in which investors reference prior reactions to similar policies in determining their own response (i.e., regardless of their personal assessment of the likelihood of implementation), should apply to LTIPs as well as to repurchase plans. Thus, market reactions to LTIPs should grow more positive over time as more firms adopt LTIPs with an agency rationale, despite prior instances of nonimplementation, and this increasingly positive reaction should then generalize to stock repurchase plans. This suggests the following alternative to Hypothesis 3: Hypothesis 3a: The number of firms that have adopted LTIPs with an agency explanation but have not implemented the plans is positively associated with the stock market reaction to repurchase plan adoption at the focal firm. METHOD
We collected complete data for the period 1980 to 1994. Later data were also collected to measure repurchase plan implementation, and earlier data were collected on LTIP adoption (1972 to 1980). Data on the adoption and implementation of repurchase plans were obtained primarily from an extensive database compiled by the Securities Data Company. We carefully checked the accuracy of these data using the Wall Street Journal Index, Reuters, and The Investment Dealers Digest, and we obtained additional data on repurchase plan implementation from COMPUSTAT (see Westphal and Zajac 2001). We also obtained information on possible confounding events that were announced during the event window from the Wall Street Journal Index (see discussion of control variables below). Data on stock market reactions and other financial information were provided by COMPUSTAT and the Center for Research in Security Prices (CRSP). We obtained data on board structure and ownership from Compact Disclosure, Standard and Poors Register of Corporations, Directors, and Executives, and corporate proxy statements. Data on CEO incentive plan adoption and implementation were also collected directly from proxies.
DATA
The initial sample for this study included all open-market stock repurchase plans adopted from 1980 through 1994, among firms listed in the 1980 Forbes 500 or Fortune 500 indexes. We chose this time period because nonimplemented repurchase plans became increasingly prevalent in the mid-1980s and previously were very rare, as shown in Figure 1. We excluded cases for which complete archival data were unavailable, leaving a final sample of 860 repurchase plans adopted by 463 firms. These firms were not significantly different in size (measured by sales and assets) or performance (measured by return on equity and return on assets in the previous year) from firms in the larger population that adopted repurchase plans, as indicated by Kolmogorov-Smirnov two-sample tests.2
INDEPENDENT VARIABLES
PRIOR REPURCHASE PLAN ADOPTIONS. We created a dichotomous variable to indicate the adoption and nonimplementation of repurchase plans, coded as 1 if firms adopted a plan and did not subsequently repurchase any shares within three years. We also created a separate variable to indicate implemented repurchase plans, coded as 1 if firms adopted a repurchase plan and actually repurchased shares within three years. In separate models we measured plan implementation over different time periods (e.g., two years or five years following adoption), and the results presented below were substantively unchanged. When firms adopted a new plan during the implementation window, the original plan was coded as nonimplemented, because firms rarely had multiple plans operating simultaneously. For each case, we then developed count variables to indicate the number of repurchase plans
We followed prior studies in the financial economics literature in excluding repurchase plans that were adopted during the two-week period after the 1987 stock market crash (e.g., Raad and Wu 1995). These adoptions were made under extraordinary cir-
cumstances in which many firms were grossly undervalued by the market.

that were implemented or not implemented prior to the focal adoption (prior non-implemented repurchase plans; prior implemented repurchase plans). We also developed a continuous measure of prior repurchase plan implementation. For each prior adoption, we calculated the percentage of shares available for repurchase that were actually repurchased under the plan, and we then calculated the average level of implementation across all prior adoptions (average implementation of prior adoptions). PRIOR LTIP ADOPTIONS WITH AGENCY EXPLALTIPs are routinely announced to shareholders in proxy statements. The announcements often include an introduction that summarizes the ostensible rationale for adopting the plan. We conducted a basic content analysis of these announcements (Holsti 1968; Zajac and Westphal 1995). Following Holsti (1968) we did not provide coders with a manual dictating how all possible phrases should be classified; such a procedure can overstate reliability. Instead, we gave coders a capsule description of the agency perspective, including key concepts that characterize the theory (Zajac and Westphal 1995). Three people coded the announcements: two were doctoral candidates in business, and the third was an undergraduate student who had taken no business coursework. Thus, the coders had different backgrounds and different levels of exposure to organizational theory, which should provide a conservative test of interrater reliability (Holsti 1968). Coders were asked to indicate whether an agency explanation was present in relevant sections of the proxy. Interrater reliability was high (95 percent agreement rate), suggesting that coders faced minimal ambiguity in classifying the explanations. We analyzed proxies throughout the diffusion period of LTIPs to identify whether long-term incentive grants were made under each formally adopted plan (see Westphal and Zajac 1998). A count variable was created to indicate the number of LTIPs that were adopted with an agency explanation but without actually granting any incentives under the plans, prior to the focal repurchase plan adoption (prior non-implemented LTIPs). A second variable represents the number of LTIPs that had been adopted and implemented (i.e., with grants made under the plans) (prior implemented LTIPs).
NATIONS .
CONTROL VARIABLES. We controlled for financial and governance characteristics that might influence the markets reaction to repurchase plan adoption. The financial variables include cash flow per share (i.e., income before extraordinary items, divided by total common shares), long-term debt to equity ratio, return on assets, and log of sales. The governance variables included two indicators of apparent board independence from management (the ratio of outside to inside directors, or outsider ratio, and separation of the CEO and board chair positions, or CEO/chair separation), and the level of ownership by institutional investors (institutional ownership). Moreover, we controlled for the prior adoption of repurchase plans or long-term incentive plans at the focal firm (prior adoption of non-implemented/implemented repurchase plans; and prior adoption of non-implemented/implemented LTIPs). We controlled for major policy announcements and other publicized incidents that occurred during the event period (other events), including the possible confounding events listed by McWilliams and Siegel (1997:640). Finally, we included dummy variables for year and industry (using primary two-digit SIC codes of adopting firms) (coefficients for these dummy variables are not reported and are available from the authors).
DEPENDENT VARIABLES
We measured stock market reactions to repurchase plan adoption with excess stock returns, or the cumulative difference between a firms observed return and its expected return during a specified time period surrounding adoption (Brown and Warner 1985; Patell 1976; Gaver, Gaver, and Battistel 1992). It is assumed that in the absence of stock price effects resulting from repurchase plan adoption, stock returns are described by the following market model: Rjt = j + jRmt + jt (1)
where Rjt is the return for firm j on day t, Rmt is the market return on day t; j is the beta, or systematic risk, of firm j (i.e., the market-adjusted variance in stock returns for firm j), j is the rate of return for firm j when Rmt equals zero; and jt is a serially independent disturbance term (E(jt) = 0). The market model parameters for each firm (j and j) are estimated over the period from day 259 to day 21 relative to

each adoption date t (Gaver et al. 1992). The excess daily return (ejt) for each firm is then estimated as follows: ejt = Rjt aj bjRmt, (2) where aj and bj are least squares estimates of j and j. Intuitively, this measure calculates the stock returns for a particular firm on a particular day that exceed the returns that would have been expected based on the recent returns of firms with comparable betas (i.e., comparable variance in their stock returns). To correct for heteroskedasticity, we estimated standardized excess returns using the Jaffe-Mandelker portfolio method (Binder 1998). An important variable in event study design is the event period, or the period over which excess returns are cumulated. Relatively long event periods allow for the possibility of gradual diffusion of information about an event following adoption. However, research has generally shown that stock prices adjust very quickly to the announcement of significant corporate events such as stock repurchase plans (e.g., within 15 minutes [Dann, Mayers, and Raab 1977; Ryngaert and Netter 1990]). Moreover, longer event periods increase the likelihood of contamination from extraneous organizational events during the time period (we control for such events in this study) (McWilliams and Siegel 1997). Thus, we estimated excess returns over a 2-day period (t1 to t0) and an 11-day period (t5 to t+5). These event periods are commonly used in the event study literature (e.g., Franz et al. 1995; Raad and Wu 1995). In the interest of thoroughness, we also ran analyses of excess returns using a 31-day event period (t5 to t+25), and these results are presented separately below.
ANALYSIS
Event studies in the financial economics literature often analyze the effects of independent variables on excess returns using subgroup analyses. Several authors have advocated the use of multiple regression analysis instead, in order to control for possible third variables (e.g., McWilliams and Siegel 1997). Given that our data set includes excess returns from different time periods, autocorrelation could occur in the data (cf., Binder 1998). We use the CochraneOrcutt transformation to correct for first-order
autocorrelation. In separate analyses we use the Prais-Winsten method, and the results are very similar (Johnston and DiNardo 1997). Aside from the Cochrane-Orcutt transformation, we address the potential for spuriousness resulting from time trends in the data in two ways: (1) in separate models we control for higher-order autocorrelation and find that the results are substantively unchanged, and (2) we control for time effects with robust standard errors clustered by year or by longer time periods (e.g., 1980 to 1984 vs. 1985 to 1994) given evidence that the agency logic emerged in the mid-1980s (Davis and Thompson 1994; Useem 1993; Zajac and Westphal 1995), and again we find that the results are unchanged. As discussed above, prior repurchase plan adoptions were coded as nonimplemented if no repurchases had been made within three years of adoption. Given this implementation window, the first three years of the sample are excluded from the regression analysis. Thus, the sample for this analysis includes all repurchase plans adopted from 1983 through 1994 (N = 778). To ensure that our results are not sensitive to this research design, we conducted two sets of additional analyses. First, we ran analyses with different implementation windows, including windows of (1) one year and (2) six months, and the results were consistent with those we report below. A relatively short implementation window seems justified, given that firms typically have no obvious economic reason for delaying implementation, and investors would not expect such a delay. And from an empirical standpoint, firms that implement their plans typically do so shortly after adoption (e.g., the portion of plans that are not implemented increases by less than 10 percent when implementation is measured over a one-year window).3 Finally, to ensure that sample selection biases would not affect our results, we estimate sep-
3 We also conducted separate analyses that includ-
ed the first three years of repurchase plan adoptions, with additional data on prior adoptions in the late 1970s provided by COMPUSTAT. Although complete data were unavailable for 19 percent of the companies in our sample during this earlier period, the overall findings were substantively unchanged from the results of our primary analyses

arate Heckman selection models, in which one equation estimates the likelihood of repurchase plan adoption, and parameters from that model are included in a second equation that estimates excess returns from adoption (Heckman and Borjas 1980). The results are similar, and the hypothesized effects are unchanged, suggesting that selection biases are not affecting our results. RESULTS Hypothesis 1 is tested in an event study analysis of repurchase plan adoptions. The results are displayed in Figure 2. We assess the significance of excess returns using the following test statistic, which is commonly used in the event study literature (Brown and Warner 1985): At /S(At), where At is the average cumulated excess return over the relevant observation period, and S(At) is the time-series standard deviation of excess returns over a 238-day estimation period. Given that average excess returns are independent, identically distributed, and normal, this statistic has a student-t distribution under the null hypothesis. As shown in Figure 2, market reactions were significantly positive in the late
1980s, consistent with results from the prior event study literature, and this response continued through the 1990s. However, the results also clearly show that market reactions became more positive over time and were significantly negative in the early years of the time period. These negative responses are significantly different from the positive responses in the later period, consistent with Hypothesis 1 (the difference in t-statistics between the beginning and the end of the study period is significant at the p .01 level). This shift in market reactions from negative to positive occurred even as the cumulative evidence of nonimplementation and the yearly rate of nonimplementation were growing over the same period, as shown in Figure 1. Descriptive statistics and correlation coefficients for variables included in the regression models are displayed in Table 2. Results of the Cochrane-Orcutt regression analyses are provided in Table 3. For this analysis, excess returns are calculated using three different event periods, and results are presented separately for each. Results of models that estimate excess returns over a 2-day period (t1 to t0) are shown in the first column, while models based on an 11-day period (t5 to t+5) and a 31-day event period (t5 to t+25) are shown in the second and third columns, respectively. These models control for
2 1.5 1 0.5 0 -0.5 -1 -1.5

(-2.34*) (-2.13*) (-2.51**) (-1.53) (.97) (2.61**) (1.72) (1.88) (2.08*) (2.25*) (2.77**) (2.83**) (2.78**) (2.71**)
(2.75**)
Figure 2. Excess Returns from Stock Repurchase Plan Adoptions: 19801994 Note: Average excess returns are presented on the vertical axis; t statistics appear in parentheses. For years with less than 50 observations, statistics are derived from a bootstrap distribution of excess returns (Dodd and Warner 1983; McWilliams and Siegel 1997). * p .05; ** p .01; *** p .001 (one-tailed tests).
Average Excess Returns
-2
1980
1982
1984
1986
Year
1988
1990
1992
1994

Table 2. Descriptive Statistics and Pearson Correlation Coefficients Independent Variable 0(1) 0(2) 0(3) 0(4) 0(5) 0(6) 0(7) 0(8) 0(9) (10) (11) (12) (13) Mean SD .(1) . .53 .31 .20 .03 .00 .07 .19 .07 .17 .03 .02 .21 .13 .16 .08 .36 .31 .28 .(2) .(3) .(4) .(5) .(6) .(7) .(8)
Prior nonimplemented repurchase plans 59.20 63.22 Prior implemented repurchase plans 210.7 135.7 Prior nonimplemented LTIPs 98.47 45.13 Prior implemented LTIPs 66.42 41.13 Cash flow per share 2.62 2.80 Long-term debt/equity .73 1.21 Outsider ratio .26 .16 CEO/chair separation .23 .42 Institutional ownership .31 .21 Return on assets .06 .06 Log of sales 7.92 1.25 Other events .14 .35 Prior adoptions: (a) Nonimplemented repurchase program .12 .32 (b) Implemented repurchase program .30 .46 (c) Nonimplemented LTIP .14 .35 (d) Implemented LTIP .09 .29 (14) Excess returns: 2 Days .71 .63 (15) Excess returns: 11 Days .84 .71 (16) Excess returns: 30 Days .91 .75 Independent Variable (10) (11) (12) (13)
. . . . . . . . . . . . . . .24 . . . . . . .28 .41 . . . . . .02 .07 .00 . . . . .03 .16 .06 .11 . . . .02 .15 .00 .06 .11 . . .00 .01 .04 .04 .03 .06 . .02 .03 .03 .00 .09 .04 .10 .02 .03 .08 .09 .21 .02 .05 .07 .12 .04 .34 .01 .16 .05 .01 .05 .01 .03 .08 .03 .21 .15 .26 .05 .17 .29 .28 .35 .14 .10 .30 .17 .18 .13 .14 .12 .13 .16 .20 .25 .23 .17 . . . . . .04 .05 .04 .03 .07 .03 .03 .01 .01 .23 .19 .20 . . . . . . .40 .03 .01 .02 .08 .05 .02 .01 .08 .12 .13 . . . . . . . .01 .05 .02 .03 .03 .05 .07 .02 .02 .00 .04 .07 .04 .09 .08 .09 .05 . . . . . . . . . . . .48 .31 . . . . . . .36
.(9) .(10) .(11) .(12) .(13a) .(13b) .(13c) .(13d).(14) .(15)
Return on assets .03 . . . . Log of sales .12 .01 . . . Other events .01 .07 .05 . . Prior adoptions: (a) Nonimplemented repurchase program .05 .02 .02 .03 . (b) Implemented repurchase program .03 .05 .01 .06 .24 (c) Nonimplemented LTIP .14 .03 .01 .00 .09 (d) Implemented LTIP .07 .01 .04 .01 .02 (14) Excess returns: 2 Days .10 .02 .02 .04 .02 (15) Excess returns: 11 Days .06 .00 .02 .05 .00 (16) Excess returns: 30 Days .10 .01 .01 .05 .01 Note: N = 778
year effects and industry differences, as well as the other control variables listed in the table. The results consistently support Hypothesis 2a: The number of firms that have previously adopted but not implemented stock repurchase plans is positively associated with the stock market reaction to adoption of the focal plan, for all three event periods. Conversely, Hypothesis 2, the market-learning hypothesis, which argued that markets would begin to discount repurchase announcements given accumulating evidence of nonimplementation, is not supported. Note that the number of prior implemented repurchase plans is also associated with more positive market reactions. In effect, the number of previously adopted plans predicts market reactions regardless of whether those plans were implemented.
The results in Table 3 also show support for Hypothesis 3a. In particular, after controlling for year effects, the number of firms that have previously adopted LTIPs with an agency explanation but not implemented the plans is positively associated with the stock market reaction to repurchase plan adoption at the focal firm. This result, which holds for all three event periods, is consistent with our contention that the use and the spread of the agency logic of governance as the interpretive lens for one corporate governance policy (i.e., top executive incentive plans) can positively color investor perceptions of another corporate policy (stock repurchase plans), even in the face of growing evidence that the incentive plans were often not implemented. Accordingly, the market-learning hypothesis (Hypothesis 3) is not supported.

Table 3. Cochrane-Orcutt Regression Analyses of Excess Stock Returns for Firms Adopting Stock Repurchase Plans: 19801994 Independent Variable 0(1) 0(2) 0(3) 0(4) 0(5) 0(6) 0(7) 0(8) 0(9) (10) (11) (12) (13) Prior nonimplemented repurchase Plans Prior implemented repurchase Plans Prior nonimplemented LTIPs Prior implemented LTIPs Cash flow per share Long-term debt/equity Outsider ratio CEO/chair separation Institutional ownership Return on assets Log of sales Other events Prior adoptions-focal firm: (a) Nonimplemented repurchase plan (b) Implemented repurchase plan (c) Nonimplemented LTIP (d) Implemented LTIP Constant Absorbed year effects (F) Model F R2 Durbin-Watson statistic (D) 2-Day Period 11-Day Period 30-Day Period .0016*** (.0005) .0008*** (.0002) .0019*** (.0007) .0013*** (.0007) .022***0 (.008) .0004*** (.0002) .204***0 (.179) .147***0 (.088) .290***0 (.126) .0008*** (.0052) .008***0 (.025) .215***0 (.117) .016***0 (.146) .095***0 (.064) .063***0 (.077) .088***0 (.146) .341***0 (.224) 3.68*** 7.39*** 0.32*** 1.59***
.0014*** (.0004) .0013*** (.0004) .0007*** (.0002) .0006*** (.0002) .0016*** (.0005) .0015*** (.0006) .0013*** (.0006) .0011*** (.0007) .020***0 (.007) .016***0 (.007) .0004*** (.0002) .0004*** (.0002) .256***0 (.144) .270***0 (.162) .094***0 (.071) .176***0 (.079) .261***0 (.101) .194***0 (.114) .0006*** (.0042) .0039*** (.0047) .013***0 (.020) .015*** (.022) .154**0* (.094) .210***0 (.105) .012***0 (.067) .058***0 (.051) .074***0 (.062) .002***0 (.001) .169***0 (.180) 3.83*** 8.02*** 0.36*** 1.43*** .001***0 (.075) .074***0 (.057) .016***0 (.131) .114***0 (.069) .360***0 (.201) 3.90*** 7.23*** 0.30*** 1.55***
Note: Coefficients shown with standard errors in parentheses. N = 778. * p .05; ** p .01; *** p .001 (one-tailed tests).
We also examined the robustness of these findings in several ways. In supplementary regression models, we used an alternative approach to test Hypotheses 2 and 2a by examining the interaction between prior adoptions and the average implementation of previously adopted plans. These results were consistent with the results in Table 3, showing that the effect of prior adoptions is not contingent on whether the plans were implemented (i.e., the market learning argument is not supported). Results of these analyses are provided in Table 1 of the Appendix. We also used a moving threeyear window to measure the number of prior nonimplemented (and implemented) repurchase plans, and we found results that essentially mirrored those reported in Table 3. Last, we reanalyzed our data using a sample that included all repurchase plans announced during the exceptional two-week period following the market crash of 1987 (these had been excluded to ensure consistency with prior research, as noted above); we found that the Table 3 results remained robust.
It might be suggested that stock market reactions to repurchase plans become more positive over time despite a lower likelihood of implementation because (1) the efficiency benefits from plans that were implemented increased over time, or (2) the implementation of repurchase plans generally had a positive effect on efficiency and investors learned about those efficiency benefits over time. To examine these possibilities, we conducted supplementary analyses of the performance effects of implemented repurchase plans. We used data on stock repurchases and performance for a random sample of 400 firms from our sample frame (i.e., firms listed in the 1980 Forbes 500 or Fortune 500 indexes) from 1980 through 1994. We estimated two measures of firm performance return on assets and return on equityusing cross-sectional time-series regression models (Greene 1997). In the primary models we used the random-effects (GLS) estimator, but results were robust to the fixed-effects estimator. As shown in in Table 2 of the Appendix, the adoption and implementation of repurchase plans do not have a significant effect on either meas-

ure of firm profitability. The interaction of repurchase plan implementation and time of adoption is also not significant: The effect of repurchase plans on profitability does not increase or decrease over time. This result was robust to the specification of alternative lag structures: The table shows effects on profitability lagged by one to three years, and similar results were obtained with performance lagged by four or five years. These models included controls for industry effects as well as the other control variables listed in the table. Thus, our findings regarding stock market reactions to repurchase plans are not an artifact of the efficiency benefits of implementing the plans or of the change in such benefits over time.4 DISCUSSION Overall, our results provide strong support for our sociological perspective on the market valuation of corporate policies. The first set of results showed a dramatic change in stock market reactions to stock repurchase plan announcements over time, shifting markedly from a negative view in the late 1970s and early 1980s to a positive view from the mid-1980s onward. This finding supports our theoretical contention that, in the mid-1980s, as prevailing beliefs about corporate governance shifted toward an agency conception of control (Davis and Thompson 1994; Useem 1993, 1996; Zajac and Westphal 1995), repurchase plans were increasingly viewed positively as a means of preventing managers from wasting free cash flow on empire-building projects. This resulted in more positive market reactions to adopted plans. In contrast, in earlier periods, when a corporate logic prevailed, these repurchase plans were
4 Additional analyses showed that repurchase plan announcements (including implemented and nonimplemented plans) were also not associated with subsequent performance, and this finding was similarly robust to different lag structures and both measures of performance. Thus, our finding that the market reacts more positively to repurchase plans even as the rate of implementation decreases cannot be attributed to any tendency for repurchase plan announcements to signal that management is dedicated to increasing the firms value to shareholders in some way (i.e., other than repurchase plan implementation).
seen as indications that managers lacked attractive investment prospects, and this resulted in negative market reactions. Thus, our findings suggest that changing institutional logics of governance led to a shift over time in the financial markets assessment of stock repurchase plans. Of course, one might contend that the increasingly positive reaction to buyback announcements simply reflects the education of the market as to the technical merits of buybacks for shareholder value. Therefore, it is particularly revealing that we find a growing positive reaction to repurchase plans despite a decreasing rate of implementation of adopted repurchase plans over the same time period. In fact, our multivariate analyses show that the number of previously adoptedbut not implemented repurchase plans is positively associated with the market reaction to adoption at the focal firm. Prevailing assumptions in the financial economics literature are that market learning occurs and that all publicly available information about a corporate policy (including accumulated evidence about the efficiency benefits associated with adopting the policy) are reflected in the stock markets response to adoption (David 1997; Fama 1970; King et al. 1993; Timmermann 1993). While theorists have explored the implications of varying key parameters of the market-learning process (e.g., the speed with which investors respond to new information), the occurrence of learning itself is generally taken-for-granted. Our findings, however, show that as more firms adopted but did not implement repurchase plans over time (thus reducing the efficiency benefits expected from adoption), the market value of repurchase plans actually increased. Additional analyses also ruled out the possibility that our results might simply reflect increasing efficiency benefits from implementation over time or increasing investor learning about efficiency benefits over time. While our findings do not support a marketlearning perspective, they are consistent with a sociological perspective on stock market reactions to corporate policy adoptions. From this perspective, repurchase plans acquired greater symbolic value over time, despite increased decoupling, through a process of institutionalization. Institutionalization is more likely in the presence of a prevailing belief system or insti-

tutional logic that suggests a sanguine interpretation of the policy (DiMaggio 1997; Scott 1994). In this context, the advent of an agency logic of governance provided a foundation for institutionalization by offering a positive interpretation of repurchase plans. These findings are consistent with our contention that as more firms adopted repurchase plans over time, the plans became more strongly associated with the agency values they affirm. In effect, repurchase plans built symbolic value through reciprocated interpretations, in which plans were presented to constituents as furthering the agency logic, and that interpretation was validated by constituents (Scott 1987:496). As this pattern was repeated, gradually the connection between repurchase plans and agency values was strengthened and may have ultimately become taken-for-granted as normatively appropriate. This resulted in more positive investor reactions to plan adoptions over time, despite a decreasing rate of plan implementation (Scott 1994; Selznick 1996; Zucker 1983). We also argued theoretically, and found empirically, that the earlier diffusion of executive incentive plans (LTIPS), which had advanced the agency logic of governance, contributed to this reinterpretation of stock repurchase plans over time. This result is again consistent with our institutional perspective, which suggests that the prior adoption of LTIPs with an agency explanation serves to reinforce, as well as reflect, the symbolic value of other policies that appear to reduce agency costs, including repurchase plans (Meyer et al. 1987; Scott 1994). That is, it appears that the growing predominance of an agency logic among members of the financial community in the 1980s, which is reflected in (and reinforced by) the use of agency explanations for executive incentive plans, conferred increased legitimacy on stock repurchase plans. It appears, in effect, that repurchase plans borrowed symbolic value from the earlier diffusion of LTIPs. More generally, our findings support our institutional perspective on how the stock market values corporate policies and challenges predominant perspectives rooted in financial economics. Whereas the dominant, financial economics perspective on capital markets conceives the markets reaction to a policy adoption as a reliable, historically invariant indicator of the efficiency benefits from adoption, our the-
ory and findings demonstrate the influence on market actors of historical change in prevailing beliefs or institutional logics about the sources of organizational efficiency. Moreover, assumptions of market learning in financial economics suggest that markets should discount the value of a policy given accumulated evidence that efficiency benefits from the policy often are not realized (i.e., due to nonimplementation), so that the policy would effectively lose its legitimacy in the stock market. However, our theory and consistent findings suggest how the market value of a corporate policy can instead increase as more firms adopt the policy over time, despite a decreasing rate of implementation. Thus, our results suggest how institutional processes that have been shown to affect policy adoptions in industrial markets can also influence policy assessments in capital markets, a context that is held up by financial economists as the closest approximation of allocative efficiency. Our theory and findings ultimately suggest how stock market reactions to corporate policiesand thus the market value of those policiesare socially constructed. Our theoretical account of the social construction of market value complements recent sociological research on the micro-social dynamics of financial markets. This nascent literature has provided evidence that in some market contexts, trading and investment behavior is governed by a social referencing process in which market actors make investment decisions by imitating or anticipating the decisions of other market participants (i.e., rather than independently forecasting assets economic performance), such that the financial valuation of economic assets may not reflect their true economic value (Abolafia 1996; Westphal and Zajac 1998; Zuckerman 1999). While this literature has yielded important insights, it begs the question of how investors estimate the beliefs and reactions of other market participants; that is, what are the inputs to the social estimation process that drives stock market valuation? Our study addresses this question by considering the macro-historical and institutional context in which the micro-social processes of stock market valuation occur. Specifically, we suggest that investor perceptions of the value of corporate policies are influenced by prevailing institutional logics and prior market reac-

tions to the adoption of similar policies, and our findings support our theoretical claims. Thus, our study advances the understanding of the social construction of financial markets by considering the institutional determinants of market value. Our theoretical perspective and findings suggest how the social referencing process that underlies stock market reactions to policy adoptions can contribute to the institutionalization of corporate policies. We began by suggesting that, given imperfect communication among market participants, investors are likely to reference prior market reactions to similar events in estimating the reactions of other participants to a focal policy adoption. As more firms adopt a policy and receive a favorable market response, each investors uncertainty about the likely response of other investors to the current adoption decreases, which should tend to produce a more positive reaction. As a result, the market value of corporate policies can increase over time through a self-perpetuating, social construction process. As more firms adopt and receive positive reactions, the perceived value of the policy becomes increasingly takenfor-granted by individual investors, and thus the policy becomes institutionalized. Moreover, this social referencing process can contribute to institutionalization despite persistent or increasing rates of decoupling of formally adopted policies from actual practices. Thus, our theory and findings also advance neoinstitutional theory by suggesting how policies can become institutionalized, despite growing evidence of nonimplementation, by virtue of the sociohistorical estimation process that drives market reactions. In this way, our perspective integrates Meyer and Rowans (1977) decoupling thesis with Zuckers (1983) thesis of institutionalization. The theory and findings of our study also suggest new directions for research on corporate governance. The governance literature has been dominated by financial economic perspectives, especially agency perspectives, and behavioral studies have focused primarily on political and micro-social factors that impede the implementation of agency prescriptions. Our study suggests the need for further research that considers how governance practices, as well as the determinants and financial consequences of those practices, are conditioned by prevailing institutional logics and processes of institu-
tionalization. Future research could also examine the key actors and actions involved in the establishment of an institutitonal logic. For example, Zajac and Fiss (2003) perform a comparative analysis of the adoption of a shareholder value orientation among U.S. and German firms and describe how entire academic and applied/practitioner journals (whose founders were committed to the agency perspective) helped push this change in the United States. They suggest that the Journal of Applied Corporate Finance, founded in the mid-1980s, utilized the rapid rise of agency-based research in corporate finance as an opportunity to translate agency theory into more accessible terms for senior executives and policy makers. Much of this research had appeared previously in the Journal of Financial Economics , a highly regarded academic journal closely associated with agency theory (and with agency theorists). Our study has implications for research on stock buybacks. Given that market reactions to repurchase plans are historically contingent, event studies should model the effects of time and prior adoptions in estimating market reactions to adoption. Moreover, while some researchers have suggested that investors react positively to repurchase plans in part because they interpret the plans as signals that managers believe their firm is undervalued, our finding that market reactions change from negative to positive over time disconfirms the undervaluation interpretation, which cannot explain negative reactions early in the time period, nor a gradual increase over time. Our study also addresses a centrally important question in the financial economics literature, albeit one that has curiously not attracted empirical attention: To what extent do financial markets learn? Perhaps strongly held belief structures in financial economics have led researchers to simply assume the answer is yes, but we suggest the answer is not obvious. One interpretation of our findings is that, at the broadest level, financial markets are teachablethey did reverse their earlier responses to stock buybacks, viewing them much more positively during the period in which a new dominant ideology (i.e., agency theory) took root. However, our central findings suggest that financial markets only slowly incorporated available information regarding the decoupling of stock buybacks and discounted the informa-

tional content of their announcements. Putting these two results together suggests that the effort expended by financial economics researchers, business school professors, and the business media to emphasize the rationality of stock buybacks may have been too successful, paradoxically limiting the markets subsequent opportunity to learn about institutional decoupling. Future research could also examine the confluence of events or tipping points that lead one institutional logic to begin to be supplanted by another. In the context of corporate governance, for example, scandals involving Enron, Worldcom, and other major U.S. corporations have led to numerous public and political discussions regarding the role of governance policies such as the use of stock options for senior executives. It is interesting that these discussions reveal a reinterpretation of stock options. Specifically, while academic research and public discourse first viewed stock options through the lens of agency logic as a valued tool for aligning the incentives of top managers and owners, options are now viewed as making managers greedy and leading them to develop schemes to inflate stock prices. The investment community has begun to advocate executive compensation vehicles, such as restricted stock, that were previously criticized from an agency perspective as failing to adequately align executive compensation with stock returns but that are alleged to more effectively satisfy other corporate goals, such as executive retention (Crystal 1991). The apparent shift in the debate
about executive incentive compensation (and possibly future challenges to stock repurchase plans) suggest that the predominance of agency logic in the domain of corporate governance may soon give way to a more pluralistic perspective on corporate governance in which multiple institutional logics are represented. Finally, while irrational stock market behavior has increasingly been a focus of study in the financial economics literature (under the umbrella term of behavioral finance), we stress that our sociological approach is quite different from the typical psychological emphasis on human decision-making limitations stemming from individual decision-making biases. Specifically, we seek to advance a logic of sociological finance that emphasizes how macrolevel ideologies and institutionalization processes affect financial markets and other constituent groups in society. However, opportunities remain for a joint sociological and psychological understanding of financial market behavior. For example, the fact that psychologists have found that individuals prefer to seek confirming, rather than disconfirming, evidence when evaluating their cognitive schemas (Nisbett and Ross, 1980) can add micro-level detail supporting the more sociological analyses of market behavior we offer here. In conclusion, we hope our focus on the social construction of financial markets contributes to a more complete and accurate understanding of the collective perceptions and behaviors of market participants.

APPENDIX
Table 1. Supplementary Regression Analyses of Excess Stock Returns for Firms Adopting Stock Repurchase Plans: 19801994 Independent Variable 0(1) Prior repurchase plans 0(2) Average implementation of prior adoptions 0(3) Prior repurchase plans X average implementation 0(4) Prior nonimplemented LTIPs 0(5) Prior implemented LTIPs 0(6) Cash flow per share 0(7) Long-term debt/equity 0(8) Outsider ratio 0(9) CEO/chair separation (10) Institutional ownership (11) Return on assets (12) Log of sales (13) Other events (14) Prior adoptions-focal firm: (00)Nonimplemented repurchase plan (00)Implemented repurchase plan (00)Nonimplemented LTIP (00)Implemented LTIP Constant Absorbed year effects (F) Model F R2 Durbin-Watson statistic (D) Main Effects Model .0004*** (.0001) .183***0 (.315) .0016*** (.0005) .0013*** (.0006) .020***0 (.007) .0004*** (.0002) .490***0 (.256) .088 ***0(.069) .264**** (.102) .003**** (.004) .006**** (.020) .126**** (.095) .013**** (.069) .061**** (.050) .072**** (.063) .002**** (.001) .110**** (.183) 3.75*** 7.51*** 0.33*** 1.46*** Interaction Model .0006*** (.00015) .211***0 (.321) .0002*** (.0011) .0015*** (.0005) .0013*** (.0006) .020***0 (.007) .0004*** (.0002) .488***0 (.257) .072***0 (.070) .268***0 (.102) .004***0 (.004) .006***0 (.020) .130***0 (.095) .015***0 (.069) .058***0 (.051) .075***0 (.064) .002***0 (.001) .116***0 (.182) 3.67*** 7.41*** 0.32*** 1.52***
Note: Coefficients shown with standard errors in parentheses. N = 778. LTIP = long-term incentive plans * p .05; ** p .01; *** p .001 (one-tailed tests).
Table 2. Cross-sectional Time-series Regression Analyses of Firm Performance X Independent Variable Return on Assets Adoption and implementation of repurchase plan Time of adoption Adoption and implementation of plan X time of adoption Prior return on assets Prior return on equity Log of sales Diversification Constant 2 Cases (N) 1-Year Lag .0X .017 (.035) .009 (.005) . .0Y 2-Year Lag .0X .0Y 3-Year Lag .0X .0Y
.023 .062 .057 .097 .090 (.036) (.055) (.056) (.095) (.096) .010 .008 .014 .013 .023 (.005) (.008) (.009) (.014) (.015) .007 . .016 . .024 (.008) (.012) (.021) .043*** .043*** .065*** .065*** .095*** .096*** (.002) (.002) (.0035) (.0035) (.006) (.006) . . . . . . .047 .046 .024 .021 .008 .003 (.057) (.060) (.086) (.086) (.137) (.137) .136 .135 .284* .283* .206 .206 (.070) (.070) (.120) (.120) (.120) (.120) .995 .999 2.215** 2.205* 1.756 1.743 (.535) (.534) (.751) (.750) (1.140) (1.139) 144.97*** 150.93*** 128.72*** 132.28*** 125.43*** 135.96*** 4674 4674 4430 4430 4180 4180 (Continued on next page)

Table 2. (Continued) X Independent Variable Return on Equity Adoption and implementation of repurchase plan Time of adoption Adoption and implementation of plan X time of adoption Prior return on assets Prior return on equity Log of sales Diversification Constant 2 Cases (N) 1-Year Lag .0X .004 (.006) .0002 (.0008) . .0Y 2-Year Lag .0X .0Y 3-Year Lag .0X .0Y
.004 .017 .018 .029 .034 (.006) (.015) (.015) (.027) (.027) .0002 .0001 .001 .00006 .002 (.0009) (.002) (.002) (.0035) (.004) .0001 . .002 . .006 (.001) (.003) (.006) . . . . . . .049*** .047*** .119*** .116*** .178*** .181*** (.002) (.002) (.005) (.005) (.010) (.010) .008 .008 .006 .005 .019 .018 (.009) (.009) (.020) (.020) (.031) (.031) .026* .026* .025 .026 .032 .033 (.011) (.011) (.022) (.022) (.026) (.026) .176* .176* .177 .174 .349 .339 (.084) (.084) (.171) (.170) (.251) (.251) 189.94*** 192.77*** 167.53*** 177.99*** 146.38*** 154.90*** 4674 4674 4430 4430 4180 4180
Note: Coefficients shown with standard errors in parentheses. * p .05; ** p .01; *** p .001 (one-tailed tests).
Edward J. Zajac is the James F. Ber Distinguished Professor of Management and Organizations at the Kellogg School of Management, Northwestern University. His research emphasizes the integration of economic and behavioral perspectives on corporate governance, organizational adaptation, and strategic alliances. Recent publications include Status Evolution and Competition: Theory and Evidence, with Marvin Washington (Academy of Management Journal, forthcoming), and Corporate Elites and Corporate Strategy: How Demographic Preferences and Structural Position Shape the Scope of the Firm, with Michael Jensen (Strategic Management Journal, forthcoming). James D.Westphal is the Ed and Molly Smith Chair in Business Administration at the McCombs School of Business, University of Texas at Austin. His current research interests include corporate governance, institutional processes, and corporate elites. Recent publications include Keeping directors in line: Social distancing as a control mechanism in the corporate elite, with Poonam Khanna (Administrative Science Quarterly 48:36198), and Getting by with the advice of their friends: CEOs advice networks and firms strategic responses to poor performance, with Michael McDonald (Administrative Science Quarterly 48:132).
REFERENCES
Abolaf ia, Mitchel Y. 1996. Making Markets:
Opportunism and Restraint on Wall Street. Cambridge, MA: Harvard University Press. Bagwell, Laurie S. and John B. Shoven. 1988. Share Repurchases and Acquisitions: An Analysis of Which Firms Participate. Pp. 191220 in Corporate Takeovers: Causes and Consequences, edited by A. J. Auerbach. Chicago, IL: University of Chicago Press. Berger, Peter L. and Thomas Luckmann. 1967. The Social Construction of Reality: A Treatise in the Sociology of Knowledge. New York: Doubleday Anchor. Binder, John J. 1998. The Event Study Methodology Since 1969. Review of Quantitative Finance and Accounting 11:11137. Bradley, Michael, Anand Desai, and F. Han Kim. 1983. The Rationale behind Interfirm Tender Offers: Information or Synergy. Journal of Financial Economics 11:183206. Brealy, Richard A. and Stewart C. Myers. 1991. Principles of Corporate Finance. 4th ed. New York: McGraw-Hill. Brown, Stephen J. and Jerold B. Warner. 1985. Using Daily Stock Returns: The Case of Event Studies. Journal of Financial Economics 14:331. Chandler, Alfred D. 1962. Strategy and Structure: Chapters in the History of the American Industrial Enterprise. Cambridge, MA: MIT Press. Chaves, Mark. 1996. Ordaining Women: The Diffusion of an Organizational Innovation. American Journal of Sociology 101:84073. Crystal, Graef S. 1991. In Search of Excess: The

Overcompensation of American Executives. New York: Norton. Dann, Larry Y., David Mayers, and Robert J. Raab Jr. 1977. Trading Rules, Large Blocks and the Speed of Price Adjustment. Journal of Financial Economics 4:322. David, Alexander. 1997. Fluctuating Confidence in Stock Markets: Implications for Returns and Volatility. Journal of Financial and Quantitative Analysis 32:42762. Davis, Gerald F., Kristina A. Diekmann, and Catherine H. Tinsley. 1994. The Decline and Fall of the Conglomerate Firm in the 1980s: The Deinstitutionalization of an Organizational Form. American Sociological Review 59:54770. Davis, Gerald F. and Henrich R. Greve. 1997. Corporate Elite Networks and Governance Changes in the 1980s. American Journal of Sociology 103:137. Davis, Gerald F. and Tracy A. Thompson. 1994. A Social Movement Perspective on Corporate Control. Administrative Science Quarterly 39:14173. DiMaggio, Paul. 1997. Culture and Cognition. Annual Review of Sociology 23:26387. Dodd, Peter and Jerold B. Warner. 1983. On Cor porate Governance: A Study of Proxy Contests. Journal of Financial Economics 11:40138. Donaldson, Lex. 1990. The Ethereal Hand: Organizational Economics and Management Theory. Academy of Management Review 15:36981. Edelman, Lauren B., Stephen Petterson, Elizabeth Chambliss, and Howard S. Erlanger. 1991. Legal Ambiguity and the Politics of Compliance: Affirmative Action Officers Dilemma. Law and Policy 13:7397. Fama, Eugene F. 1970. Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance 25:383417. Fama, Eugene F. and Michael C. Jensen. 1983. Separation of Ownership and Control. Journal of Law and Economics 26(June):30125. Fligstein, Neil. 1990. The Transformation of Corporate Control. Cambridge, MA: Harvard University Press. . 1991. The Structural Transformation of American Industry: An Institutional Account of the Causes of Diversification in the Largest Firms, 19191979. Pp. 31136 in The New Institutionalism in Organizational Analysis, edited by W.W. Powel and P.J. DiMaggio. Chicago, IL: University of Chicago Press. Fligstein, Neil. and Peter Brantley. 1992. Bank Control, Owner Control, or Organizational Dynamics: Who Controls the Large Modern Corporation? American Journal of Sociology 98:280307.
Franz, Diana R., Ramesh P. Rao, and Niranjan Tripathy. 1995. Informed Trading Risk and Bidask Spread Changes around Open Market Stock Repurchases in the NASDAQ Market. Journal of Financial Research 18:31127. Friedland, Roger and Robert R. Alford. 1991. Bringing Society Back In: Symbols, Practices, and Institutional Contradictions. Pp. 23266 in The New Institutionalism in Organizational Analysis, edited by W.W. Powel and P.J. DiMaggio. Chicago, IL: University of Chicago Press. Gaver, Jennifer J., Kenneth M. Gaver, and George P. Battistel. 1992. The Stock Market Reaction to Performance Plan Adoptions. Accounting Review 67:17282. Greene, William H. 1997. Econometric Analysis. 3d ed. Upper Saddle River, NJ: Prentice-Hall. Grullon, Gustavo and David L. Ikenberry. 2000. What Do We Know about Stock Repurchases? Journal of Applied Corporate Finance 13:3151. Grullon, Gustovo and Roni Michaely. 2003. The Information Content of Share Repurchase Programs. Journal of Finance 59:65180. Heckman, James and George J. Borjas. 1980. Does Unemployment Cause Future Unemployment: Definitions, Questions, and Answers from a Continuous Time Model of Heterogeneity and State Dependence. Econometrica 47:24783. Holsti, Ole R. 1968. Content Analysis. Pp. 596692 in The Handbook of Social Psychology. 2d ed. Edited by Gardner Lindzey and Elliot Aronson. Reading, MA: Addison-Wesley. Ikenberry, David, Josef Lakonishok, and Theo Vermaelen. 2000. Stock repurchases in Canada: Performance and strategic trading. Journal of Finance 55:237398. Jensen, Michael C. 1989. The Eclipse of the Public Corporation. Harvard Business Review 67:6174. Jensen, Michael C. and William H. Meckling. 1976. Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure. Journal of Financial Economics 3:30550. Jensen, Michael C. and Jerome Warner. 1988. The Distribution of Power among Corporate Managers, Shareholders, and Directors. Journal of Financial Economics 20:324. Johnston, John and John DiNardo. 1997. Econometric Methods. 4th ed. New York: McGraw-Hill. Keynes, John M. 1936. The General Theory of Employment, Interest and Money. New York: Harcourt Brace. King, Ronald R., Vernon L. Smith, Arlington W. Williams, and Mark Van Boening. 1993. The Robustness of Bubbles and Crashes in Experimental Stock Markets. Pp. 183200 in Nonlinear Dynamics and Evolutionary Economics, edited by R. H. Day and P. Chen. Oxford, England: Oxford University Press. Knorr Cetina, Karin and Urs Bruegger. 2002. Global

Microstructures: The Virtual Societies of Financial Markets. American Journal of Sociology 107:90550. Kumar, Raman and Parvez R. Sopariwala. 1992. The Effect of Adoption of Long-term Performance Plans on Stock Prices and Accounting Numbers. Journal of Finance 67:56173. Lazonick, William and Mary OSullivan. 2000. Maximizing Shareholder Value: A New Ideology for Corporate Governance. Economy and Society 29:1335. Lee, D. Scott, Wayne H. Mikkelson, and H. Megan Partch. 1992. Managers Trading Around Stock Repurchases. Journal of Finance 47:194761. Mackenzie, Donald. 2003. Long-term Capital and the Sociology of Arbitrage. Economy and Society 32:34980. McWilliams, Abagail and Donald Siegel. 1997. Event Studies in Management Research: Theoretical and Empirical Issues. Academy of Management Journal 40:62657. Medury, Prasad V., Linda E. Bowyer, and Venkat Srinivasan. 1992. Stock Repurchases: A Multivariate Analysis of Repurchasing Firms. Quarterly Journal of Business and Economics 31:2144. Merton, Robert K. 1948. The Self-Fulf illing Prophecy. Pp. 24867 in Social Research and the Practicing Professions, edited by A. Rosenblatt and T.F. Gieryn. Cambridge, MA: Abt Books. Meyer, John W. and Brian Rowan. 1977. Institutional Organizations: Formal Structure as Myth and Ceremony. American Journal of Sociology 83:34063. Meyer, John W., John Boli, and George M. Thomas. 1987. Ontology and Rationalization in the Western Cultural Account. Pp. 1237 in Institutional Structure: Constituting State, Society, and the Individual, edited by G. M. Thomas, J. W. Meyer, F. O. Ramirez, and J. Boli. Newbury Park, CA: Sage. Nisbett, Richard E., and Lee Ross. 1980. Human Inference: Strategies and Shortcomings of Social Judgment. Englewood Cliffs, NJ: Prentice-Hall. Nohel, Tom and Vefa Tarhan. 1998. Share repurchases and firm performance: new evidence on the agency costs of free cash flow. Journal of Financial Economics 49:187222. Patell, James M. 1976. Corporate Forecasts of Earnings per Share and Stock Price Behavior: Empirical Tests. Journal of Accounting Research 14:24676. Pfeffer, Jeffrey. 1981. Management as Symbolic Action: The Creation and Maintenance of Organizational Paradigms. Pp. 152 in Research in Organizational Behavior, vol. 3, edited by L.L. Cummings and B.M. Staw. Greenwich, CT: JAI Press. Raad, Elias, and H.K. Wu. 1995. Insider Trading
Effects on Stock Returns around Open-Market Stock Repurchase Announcements: An Empirical Study. Journal of Financial Research 18:4557. Rajagopalan, Nandini. 1996. Strategic Orientations, Incentive Plan Adoption, and Firm Performance: Evidence From Eletrical Utility Firms. Strategic Management Journal 18:76186. Ratner, Mitchell, Samuel H. Szewczyk, and George P. Tsetsekos. 1996. The Information Role of Tender Offer Stock Repurchases: Evidence From Institutional Ownership. Journal of Business Finance and Accounting 23:86980. Ryngaert, Michael and Jeffrey Netter. 1990. Shareholder Wealth Effects of the 1986 Ohio Antitakeover Law Revisited: Its Real Effects. Journal of Law, Economics, and Organization 6:25362. San Francisco Chronicle. 1995. Theyre Not as Sexy as IPOs, but Stock Buybacks are Hot. July 31: B1. Scott, W. Richard. 1987. The Adolescence of Institutional Theory. Administrative Science Quarterly 32:493511. . 1994. Institutions and Organizations: Toward a Theoretical Synthesis. Pp. 12136 in Institutional Environments and Organizations: Structural Complexity and Individualism, edited by W.R. Scott and J.W. Meyer. Thousand Oaks, CA: Sage. . 1998. Organizations: Rational, Natural, and Open Systems. 4th ed. Englewood Cliffs, NJ: Prentice-Hall. . 2000. Institutional Change and Healthcare Organizations: From Professional Dominance to Managed Care. Chicago, IL: University of Chicago Press. . 2001. Institutions and Organizations. 2d ed. Thousand Oaks, CA: Sage. Selznick, Philip. 1957. Leadership in Administration. New York: Harper and Row. . 1996. Institutionalism Old and New. Administrative Science Quarterly 41:27077. Stephens, Clifford P. and Michael S. Weisbach. 1998. Actual Share Reacquisitions in Open-Market Repurchase Programs. The Journal of Finance 53:31333. Suchman, Mark C. 1995. Managing Legitimacy: Strategic and Institutional Approaches. Academy of Management Review 20:571610. Thornton, Patricia H. 2001. Personal versus Market Logics of Control: A Historically Contingent Theory of the Risk of Acquisition. Organization Science 12:294311. . 2002. The Rise of the Corporation in a Craft Industry: Conflict and Conformity in Institutional Logics. Academy of Management Journal 45:81101. Thornton, Patricia H. and William Ocasio. 1999. Institutional Logics and the Historical

Contingency of Power in Organizations: Executive Succession in the Higher Education Publishing Industry, 19581990. American Journal of Sociology 105:80143. Timmermann, Allan G. 1993. How Learning in Financial Markets Generates Excess Volatility and Predictability in Stock Prices. Quarterly Journal of Economics 104:77797. Tolbert, Pamela S. and Lynne G. Zucker. 1983. Institutional Sources of Change in the Formal Structure of Organizations: The Diffusion of Civil Service Reform, 18801935. Administrative Science Quarterly 28:2239. Useem, Michael. 1993. Executive Defense: Shareholder Power and Corporate Reorganization. Cambridge, MA: Harvard University Press. . 1996. Investor Capitalism: How Money Managers Are Changing the Face of Corporate America. New York: Basic Books. Wade, James B., Joseph F. Porac, and Timothy G. Pollock. 1997. Worth, Words, and the Justification of Executive Pay. Journal of Organizational Behavior 18:64164. Westphal, James D. and Edward J. Zajac. 1998. The
Symbolic Management of Stockholders: Corporate Governance Reforms and Shareholder Reactions. Administrative Science Quarterly 43:12753. . 2001. Explaining Institutional Decoupling: The Case of Stock Repurchase Programs. Administrative Science Quarterly 46:20228. Zajac, Edward J. and James D. Westphal. 1995. Accounting for the Explanations of CEO Compensation: Substance and Symbolism. Administrative Science Quarterly 40:283308. Zajac, Edward J. and Peer C. Fiss. 2003. Practices as Ideology: The Definition and Diffusion of Shareholder Value Maximization in Germany. Working paper. Department of Management and Organizations, Kellogg School of Management, Northwestern University, Evanston, IL. Zucker, Lynne G. 1983. Organizations as Institutions. Pp. 142 in Research in the Sociology of Organizations, vol. 2, edited by S. B. Bacharach. Greenwich, CT: JAI Press. Zuckerman, Ezra W. 1999. The Categorical Imperative: Securities Analysts and the Illegitimacy Discount. American Journal of Sociology 104:13981438.
#1844-ASR 69:5 filename:69507-erratum
ERRATUM
Edward J. Zajac Northwestern University James D. Westphal University of Texas at Austin
In: AMERICAN SOCIOLOGICAL REVIEW 2004, Vol. 69 (June:43357) The following text does not appear in the accepted draft of the Edward Zajac and James Westphal paper published in the June 2004 issue of American Sociological Review. While normally changes of this sort are routine at the copyediting stage, in this case the changes might inadvertently have had some bearing on the interpretation of the comment published in the same issue by Ezra Zuckerman. In the case of the simultaneous publication of an article and a comment, no changes may be permitted at the copyediting stage. We regret any misinterpretation of the exchange between Zuckerman and Zajac and Westphal that may have resulted. 1. P. 438, right column:
Some financial economists have suggested that investors react positively to repurchase plan adoption in part because they view adoption as a signal that managers believe the firm is undervalued (Ikenberry, Lakonishok, and Vermaelen, 2000). However, this perspective does not address the potential for market reactions to change from negative to positive over time, which is the focus of our theory.
2. P. 438, right column:

(see also Stephens and Weisbach 1998)
3. P. 439, left column:

In our multivariate analyses, significant predictors of whether firms were likely to adopt but not implement repurchase plans included CEO power, network ties, and a firms history of prior decoupling, but not post-announcement stock returns or market-to-book value.

Note that this would be true from a market-learning perspective even if other motives for nonimplementation were sometimes also involved, since investors would, on average, come to expect fewer agency benefits from announced repurchase plans, given growing accumulated evidence of non-implementation.

Our study has implications for research on stock buybacks. Given that market reactions to repurchase plans are historically contingent, event studies should model the effects of time and prior adoptions in estimating market reactions to adoption. Moreover, while some researchers have suggested that investors react positively to repurchase plans in part because they interpret the plans as signals that managers believe their firm is undervalued, our finding that market reactions change from negative to positive over time disconfirms the undervaluation interpretation, which cannot explain negative reactions early in the time period, nor a gradual increase over time. (Continued on next page)
AMERICAN SOCIOLOGICAL REVIEW, 2004, VOL. 69 (October:748749)
#1844-ASR 69:5 filename:69507-erratum

ARTICLE TITLE749
6. References:
Ikenberry, David, Josef Lakonishok, and Theo Vermaelen. 2000. Stock repurchases in Canada: Performance and strategic trading. Journal of Finance 55:237398. Stephens, Clifford P. and Michael S. Weisbach. 1998. Actual Share Reacquisitions in Open-Market Repurchase Programs. The Journal of Finance 53:31333. There are also two additional references that are not cited in the published text: Grullon, Gustovo and Roni Michaely. 2003. The Information Content of Share Repurchase Programs. Journal of Finance 59:65180. Nohel, Tom and Vefa Tarhan. 1998. Share repurchases and firm performance: new evidence on the agency costs of free cash flow. Journal of Financial Economics 49:187222.

American Sociological Review 2004 Zajac 433 57

Загружено:

Сведения о документе

Исходное описание:

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

American Sociological Review 2004 Zajac 433 57

Загружено:

Авторское право:

Доступные форматы

American http://asr.sagepub.

com/ Sociological Review

American Sociological Association

>> Version of Record - Jun 1, 2004 What is This?

Downloaded from asr.sagepub.com by Daniel Silva on November 1, 2013

#1600-ASR 69:3 filename:69306-zajac

AMERICAN SOCIOLOGICAL REVIEW, 2004, VOL. 69 (June:433457)

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

SHIFT FROM CORPORATE TO AGENCY LOGICS GOVERNANCE

#1600-ASR 69:3 filename:69306-zajac

The firm Concept of resource allocation

Allocation of cash flow

#1471-ASR 69:2 filename:69306-zajac

INSTITUTIONAL LOGICS OF GOVERNANCE AND MARKET REACTIONS TO REPURCHASE PLANS

#1600-ASR 69:3 filename:69306-zajac

reactions to stock buybacks that invoked the agency logic:

THE (NON)IMPLEMENTATION OF STOCK REPURCHASE PLANS

#1471-ASR 69:2 filename:69306-zajac

MARKET LEARNING VERSUS INSTITUTIONALIZATION PERSPECTIVES

100 90 80 70 60 50 40 30 20 10 0 1980 1982 1984 1986 1988 1990 1992 1994

Figure 1. The Adoption and Decoupling of Stock Repurchase Programs: 19801994

Not Implemented Implemented

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

IZATION ACROSS RELATED POLICIES.

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

cumstances in which many firms were grossly undervalued by the market.

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

3 We also conducted separate analyses that includ-

#1600-ASR 69:3 filename:69306-zajac

2 1.5 1 0.5 0 -0.5 -1 -1.5

Average Excess Returns

#1471-ASR 69:2 filename:69306-zajac

.(9) .(10) .(11) .(12) .(13a) .(13b) .(13c) .(13d).(14) .(15)

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

#1600-ASR 69:3 filename:69306-zajac

#1471-ASR 69:2 filename:69306-zajac

#1844-ASR 69:5 filename:69507-erratum

2. P. 438, right column:

3. P. 439, left column:

4. P. 440, right column:

5. P. 451, right column:

AMERICAN SOCIOLOGICAL REVIEW, 2004, VOL. 69 (October:748749)

#1844-ASR 69:5 filename:69507-erratum

Вам также может понравиться