Categorical Delegitimation: Theory and Empirical
Exploration
How Innocent Firms Face Legitimacy Loss in Response to an External Organizational Crisis
by
Simon Westmaas
Newcastle University Business School Faculty of Economics and Business Supervisor: Dr. Peter Edward Supervisor: Dr. Kees van Veen Module: Dissertation – Dual Award (NBS8199) Course: Master Thesis IB&M
Student Number: B2014709 Student Number: S1657682
Acknowledgments
Looking back at the writing process of this master thesis, I can not escape the realization that without a number of people helping me out and motivating me when necessary, the end product might have never seen the light of day. It is here that I would like to say thank you to these people for supporting me when needed; whether it was because it is their job to do so, or simply because they cared enough to be there for me without getting anything in return other than my eternal gratitude. First of all, I must thank my supervisors, dr. Kees van Veen from the Rijksuniversiteit Groningen, and dr. Peter Edward from Newcastle University. Although I might have not requested their feedback often, when I did need it, a response was almost often swift and informative. In particular I would like to thank them for the fruitful hour-long discussions we had early on in the research process, which have guided the further development of this master thesis.
Secondly, I would like to fellow students, in particular the double degree students from Groningen, with whom I spent an extremely fun time in Newcastle. A special thanks in this respect goes out to Martijn Mullink, who was willing to act as a second coder.
A special mention and thank you is also in order for Dr. Tracy Scurry, degree program director for the dual master awards at Newcastle University. My eternal gratitude for helping me sort things out when my own life experienced a crisis, and for the fun and insightful discussions we had on a number of occasions.
For help with the choices that had to be made in multivariate data analysis, I must thank Mr. Melvyn Perry, who spent some of his spare time quickly and precisely answering the questions I had
regarding modeling options.
Last, but (as the cliché goes) definitely not least, I thank my close friends and family here. Although they have often questioned my work planning (and not without reason), they were always ready to speak words of encouragement and support. A special thanks here should go out to my mother, for putting up with my highly irregular sleeping patterns, and to Hannah Perry, who I could count on for a stern reality check whenever necessary.
Simon Westmaas
Abstract
This master thesis is split up in two parts. Firstly, an in-depth review of the concept of organizational legitimacy is presented. In particular, use of the concept in organizational research, and the
formation of legitimacy judgments are explored. Finally, a theoretical argument for the existence of categorical delegitimation is presented in this section.
The second part of this thesis presents an exploratory empirical approach to the research of categorical delegitimation. Two separate, high-profile, recent organizational crises are examined. Using both univariate and multivariate methods, the links between salient issue dependence in, and response of, innocent firms facing potential spillover effects of an organizational crisis and the constructed categorical delegitimation index are explored.
Analyzing a dataset of 408 firm-date observations from two industries (Oil and Energy producing firms), univariate results indicate that the existence of a categorical delegitimation effect is not the case for the oil industry, but is present for the subset of energy producers. Also, firm responses of any type generally seem to lead to a stronger delegitimating effect for the responding firm.
Table of Contents
Title Page...1
Acknowledgements………..………..2
Abstract………3
List of Figures………..………...………..…………..5
List of Tables………..………..……5
Introduction……….………..………..6
Conceptual Model and Research Question……….…………..………..7
1.Literature Review……….………....…..8
1.1: Social Judgment of Organizations………..8
1.2: The Concept of Legitimacy in Different Branches of Organization Theory………..10
1.3: Legitimacy Typologies………14
1.4: The Formation and Importance of Legitimacy Judgments………..16
1.5: Organizational Crises……….21
1.6: Consequences of Legitimacy Loss……….23
1.7: Categorical Delegitimation……….24
2.Hypotheses……….……….26
2.1: Salient Issue Prevalence and Categorical Delegitimation……….26
2.2: Organizational Crisis Responses and Categorical Delegitimation……….……….27
3.Research and Methodology………29
3.1: Study Context; Two High-Profile Organizational Crises………..……….29
3.2: The Dataset – Selection of Firms………31
4.Results and Analysis………..……….36
4.1: Univariate Analysis and Results……….36
4.2: Multivariate Analysis and Results……….38
5.Discussion and Conclusions……….………..41
References……..………...………43
List of Figures
Figure 1: Conceptual Model………...7
Figure 2: Model of Legitimacy Formation (Based on Tost, 2010)……….….….…18
List of Tables
Table 1: Frequency Count of CD Threshold Exceeding by Firms………...…………..36
Table 2:Descriptives and Pearson Correlations: Full Dataset………36
Table 3:Descriptives and Pearson Correlations: Energy Companies…..……….37
Table 4:Descriptives and Pearson Correlations: Oil Companies…………..………...37
Table 5:Regression Coefficients Multivariate OLS Regression: Full Dataset………..38
Table 6:Regression Coefficients Multivariate OLS Regression: Energy Firms………39
Table 7:Regression Coefficients Multivariate OLS Regression: Oil Firms……...……..40
Table 8:Summary of Hypothesis testing outcomes and related evidence……….41
Introduction
The concept of organizational legitimacy is one the of the most studied subjects in organizational research, yet very few scholars agree on its’ definition and use within organizational theory (Suchman, 1995). In general however, it seems that most scholars agree on one thing: without legitimacy, an organization is unable to survive in its’ contemporary form. As such, legitimacy is often referred to as a ‘license to operate’.
Both the antecedents and results of legitimacy gain and loss have been extensively studied over the past decades. Firms that lose legitimacy for example are likely to experience more trouble obtaining essential inputs, and as a result will face deteriorating of performance and share price (Hamilton, 2006). Firms that gain legitimacy in contrast are more trusted by their audiences and will see an increase in share price and performance (Suchman, 1995).
As Massey (2001) explains, legitimacy loss is especially likely to occur when the organization’s behavior can be directly linked to negative social outcomes. These effects are already well described in existing literature: not surprisingly, companies involved in industrial disasters are shown to face legitimacy loss (Hamilton, 2006), as do companies that engage in corporate misbehavior (Jonsson e.a., 2009).
While the impact of organizational crises and misbehavior on those organizations that are at the centre of the crises has thus been documented on many occasions, the impacts that such crises can have on firms operating within the same industry has received a lot less attention (Yu e.a., 2008; Jonsson e.a., 2009 are notable exceptions). A number of existing organizational and marketing theories however can be readily used to theorize the existence of such effects. Population Ecology for example argues that firms gain legitimacy by adopting similar structures and behavior, leading to the argument that misbehavior would also be seen as a category feature rather than as a firm feature. This type of legitimacy by innocent firms is labeled categorical delegitimation (Jonsson e.a., 2009), Spillover (Yu e.a., 2008), or ‘loss of field legitimacy (Desai, 2011). This paper is focused on further exploring the categorical delegitimation effects, in particular examining two that theorists in the field have hypothesized to influence the magnitude of the spillover effect: salient issue
dependence and firm responses.
The following research question can be formulated based on this short overview:
The conceptual model of this paper is presented below (Figure 1). In the sections following the model, each term in the model will be further clarified and their inclusion in the model explained.
Figure 1: The Conceptual Model
This paper will now proceed with an extensive literature review involving several topics related to categorical delegitimation, of which the concept of organizational legitimacy is the first. It is then followed by the conceptual model and the hypotheses to be tested, a description of the
1. Literature Review
The concept of organizational legitimacy has been intensively researched by academics from various social and organizational sciences backgrounds. Attempts to define the concept can for example be found in publications by institutional theorists (e.g. Suchman, 1995), resource dependence theorists (e.g. Pfeffer, 1981), stakeholder theorists (e.g. Mitchell e.a., 1997), organizational ecologists (e.g. Hannan & Carroll, 1992) and social psychologists (e.g. Johnson e.a., 2006). Definitions of legitimacy vary widely between these different disciplines, and the lenses used to study the phenomenon are equally divergent. However, there is general consensus that legitimacy plays a critical role in determining the development and endurance of social systems in general, and organizations in particular (Tost, 2011). As such, having legitimacy is seen as being granted a license to operate by the environment of the organization. Due to this consensus, legitimacy is considered ‘perhaps the most central concept in institutional research’ (Colyvas & Powell, 2006). Before organizational legitimacy can be discussed in more detail however, I will explain how the concept of legitimacy differs from related social judgment concepts, status and reputation. Subsequently, legitimacy will be further defined and we will look at how the concept used by theorists from different fields of social science research.
1.1: Social Judgment of Organizations: Legitimacy, Status and Reputation
Multiple types of social judgments of organizations exist, of which legitimacy is just one (Deephouse & Suchman, 2008; Washington & Zajac, 2006). This fact is often overlooked by organizational theorists, who have the tendency to use terms as organizational legitimacy, reputation and status interchangeably (Bitektine, 2011). In this section the difference between these terms will shortly be discussed. Later on in the literature review, models of social judgment formation will be discussed, with special attention paid to the formation of legitimacy judgments.
Although more definitions will be discussed in more detail in the next sections, most definitions of organizational legitimacy focus on social acceptance. This acceptance is a result of the organization meeting social norms and requirements (Deephouse & Carter, 2005). In effect, legitimacy is thus attributed to an organization because it possesses many of the features other organizations also have (as these organizations meet the same requirements). Although legitimacy can be seen as a continuous variable, one of the more important characteristics of legitimacy is that a threshold level is needed for an organization to survive (Dowling & Pfeffer, 1975; Meyer & Rowan, 1977; Ruef & Scott, 1998; Bruderl e.a., 1992). A legitimacy judgment can only go two ways; either legitimacy is granted, or it is not; by what margin it is granted (or not granted) is of no real value and will not provide the organization with any additional benefits (Bitektine, 2011).
Most definitions of reputation on the other hand are focused on relative comparisons of organizations. Consider for example the definition given by Washington and Zajac (2005;p283):
‘Reputation is an economic concept that captures differences in perceived or actual quality or merit that generate earned, performance-based rewards.’
Rather than focusing on organizations having the same features, a reputation judgment is based on the differences between organizations (see for example Rindova e.a., 2005 or Lange e.a., 2011). Reputation is said to be based on the audiences’ perceptions and past experiences with the organization which serve as a proxy used to anticipate the future behavior of the organization (Whetten & Mackey, 2002; Bitektine, 2011). In contrast to legitimacy, reputation judgments can be placed on a continuous instead of a dichotomous scale; having a better reputation will indeed provide economic benefits to the organization. For example, both Phillips and Zuckerman (2001), and Deephouse and Carter (2005) find that firms with higher reputation can deviate from ‘normal’ strategic behavior without losing legitimacy or status. As it seems, firms with a higher reputation can ask higher prices for their products; they can thus extract more profit and sustain this over a long period of time (Roberts & Dowling, 2002).
Status is similar to legitimacy to the extent that it refers to social acceptance by an audience. A difference with legitimacy is that status judgments lead to multiple status groups, making the outcome of a legitimacy judgment ordinal rather than dichotomous (Bitektine, 2011). Washington and Zajac (2005; p284) define status as follows:
Based on this definition, status resembles legitimacy closer than it does reputation. Being classified as high status grants an organization benefits that Weber (1978) calls ‘privileges’. These privileges are not based on perceived quality of an organization’s products or services, but rather on stereotyping of organizations. These stereotypes might persist even when the organization’s output no longer meets the high standards it did before, a situation that leads to the organization possessing low reputation but high status.
1.2: The concept of legitimacy in different branches of organization theory 1.2.1: Legitimacy in institutional theory
Institutional theory explains how organizational actions and responses are often repetitive, and products of past actions and practices. These practices are seen as forced on the organization by means of social pressure and the cultural definitions of those institutions that surround it. When these pressures have a significant impact on organizational behavior, they are considered ‘institutionalized’ (Robbins & Barnwell, 2002). Because only a limited set of socially acceptable behaviors becomes institutionalized, firms in the same organizational field are likely to resemble each other in both form and behavior. This process is called institutional isomorphism, and manifests itself by increased firm similarity (Deephouse & Carter, 2005). For some attributes of organizations this similarity can be a simple yes or no question, such as whether or not a firm has a department dealing with Corporate Social Responsibility (CSR) issues. Other issues, such as corporate strategy, are more complicated and can thus show degrees of conformity rather than a simple yes or no answer.
One of the key propositions of institutional theory is that isomorphism leads to legitimacy (DiMaggio & Powell, 1983). By mimicking strategies, structures and practices used by competitors and/or imposed by authorities, the organization appears rational and thus ‘makes sense’ to others. This leads to the following definition of legitimacy commonly cited among institutional theorists:
“Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper or appropriate within some socially constructed system of norms, values, beliefs and definitions” (Suchman, 1995).
based on what the audience can see. Legitimacy is socially constructed to the extent that it is not bestowed upon an organization by a single observer, but rather by a social system (or subsystem). In a further clarification of the concept, Tost (2011) labels this individual level legitimacy propriety, whereas legitimacy at the collective level is labeled validity. In the remainder of this paper the term legitimacy will be used to discuss the concept on the collective level. Whenever the individual level is discussed, the term propriety will be used.
1.2.2: Legitimacy in Resource Dependence Theory
Resource dependence theory (RDT) asserts that organizations depend on their environment to supply them with the necessary resources that the organization can transform into output (Pfeffer and Salancik, 1978). This environment contains other organizations, but also individuals and other social actors such as governments and media institutions. As resources are owned by an
organization’s environment, the organization must find a way to obtain these resources. This empowers the resource suppliers to the extent that they can choose whether or not to allow the organization to use them; they are able to ask something in exchange. In many cases, the unit of exchange is monetary, for example a shoe store buying shoes for its inventory from a factory, or the same shoe store hiring staff that require a wage. However, exchanges can also be non-monetary. For example, a factory might be required to meet certain safety standards before being granted the right to operate.
Based on Ashforth & Gibbs (1990), the following definition of Legitimacy in RDT can be constructed:
Legitimacy is a socially constructed, contestable resource that justifies the organization’s role in the social system and helps attract resources from constituents.
In RDT, legitimacy is a thus a resource like many others; like in institutional theory it is not owned by an organization by bestowed upon it by its contingent social systems (Suchman, 1995). However, whereas institutional theory leaves little room for managerial control over the process of obtaining legitimacy, RDT researchers consider organizations to able to actively extract this resource from their environments. It is argued that rather than by influencing the real processes and outcomes of organizational activity, organizations prefer to use symbols and rituals in order to meet expectations of their constituents (Ashforth & Gibbs, 1990). This in turn can lead to clashes between the
1.2.3: Legitimacy in Stakeholder Theory
Stakeholder theory, as popularized by Freeman (1984), attempts to explain which constituents (stakeholders) of an organization deserve or require management attention, and which do not (Mitchell e.a., 1997). Two core concepts in deciding the stakeholder relationships with the firm are power and legitimacy. Power in this context can be considered as the extent to which the
stakeholder can make an organization change its’ behavior. As such, power might be quantified as the potential harm the stakeholder can do to an organization if the latter is not responsive to the stakeholder’s demands.
Based on Mitchell e.a. (1997), the following definition of legitimacy in stakeholder theory can be defined:
‘Legitimacy is a normative concept describing a social judgment about the fairness of stakeholders’ exercising of power within stakeholder the stakeholder relationship.’
Whereas power is purely a concept that describes the relative strength of stakeholders, legitimacy in stakeholder theory describes a normative element of the relationship. In short, legitimacy in this case is a judgment of social actors of the morality, responsibility and desirability of the use of power by stakeholders vis-à-vis the organization. Illegitimate use of power will inevitably lead to loss of power (Davis, 1973).
Legitimacy of a constituent’s claim will not necessarily lead to management attention (that is, being considered a salient stakeholder). If the constituent does not have the necessary power over the organization, an organization’s management can safely disregard any claims made. Together however, power and legitimacy create authority (Weber, 1947), which will place the constituent firmly on the organization’s radar and will lead to the constituent being considered a stakeholder.
1.2.4: Legitimacy in Population Ecology Theory
In contrast to theories discussed earlier, population ecology (PE) theories are not aimed at explaining the existence, form and structure of single organizations. Rather, PE is concerned with aggregates of organizations, which are labeled populations (Hannan & Freeman, 1977). The theory of population ecology relies heavily on Darwinian survival-of-the-fittest doctrines as can be found in biological theories of evolution (Robbins & Barnwell, 2002). The core argument of PE is that organizations survive on the fit with their environment; those organizations that have the best fit and/or are the most adaptable will be the ones to survive.
The role of legitimacy in PE is that of a potential external source of structural inertia, or barrier to change (Hannan & Freeman, 1977). When the environment of an organization changes in a way that threatens the organization’s survival, radical structural change of the organization might be required in order to better fit the new environment. However, if the structure of the organization is an important determinant of the firms legitimacy (as also predicted by institutional theory), changing this structure can have devastating consequences for a firms survival. This results in a classic catch-22 where firms have to chose between likely extinction because they do no longer fit the environment, versus choosing to change structure but lose legitimacy in the short run. Luckily, although a difficult and intensive process, regaining legitimacy is not a hopeless task per se (Suchman, 1995).
1.2.5: Legitimacy in Social Psychology
Social psychologists study how people’s thoughts, feelings, and behaviors are influenced by the presence of others, regardless of this presence being real or imagined (Allport, 1985). As the name suggests, it combines the fields of sociology and psychology. Rather than being theorizing on how firms are influenced by their social environment, social psychology entails research at the level of the individual.
Based on the work of Tost (2011), the following definition of legitimacy in social psychology can be constructed:
Legitimacy is a function of instrumental, relational and moral considerations that influences the formation of social judgments made by individuals.
Social psychologists have used the construct of legitimacy to explain the stability of and behavioral reactions to a broad range of social entities (Tost, 2011). Like in stakeholder theory, the distinction between power and legitimacy is very much present in social psychology. More to the point, social psychologists argue that seeking to gain influence through the use of power alone is a costly and inefficient strategy. Legitimacy can be used as a complementary or even replacement mechanism for power to gain influence, and it is less costly to maintain than power once obtained (Suchman, 1995). Also, as influence based on legitimacy is voluntarily accepted by the person or organization
1.3: Legitimacy Typologies
Not surprisingly, because of the centrality of the concept of legitimacy in many different
organizational theories, different typologies of legitimacy have been defined. There are several bases for the distinction of organizational legitimacy types, as summarized by Bitektine (2011). These will be discussed in the following sections.
1.3.1: Typologies based on Legitimating Audiences
Legitimating audiences are the social actors that grant or withhold legitimacy to organizations, ultimately influencing the latter’s survival. In line with stakeholder theory, not all audiences are to be considered equally important (Oliver, 1991; Mitchell e.a., 1997), leaving managerial discretion to the choice of which audiences should be appeased and which ignored. A legitimacy typology distinction based on legitimating audiences has lead researchers to identify and use media legitimacy (Pollock & Rindova, 2003; Lamin & Zaheer, 2012), legitimacy with regulators (Deephouse & Carter, 2005; Johnson e.a., 2006), legitimacy with investors (Certo, 2003; Lounsbury & Glynn, 2001), legitimacy with advocacy groups (Zald, 2005; Rao e.a., 2000) and internal legitimacy (Kostova & Zaheer, 1999; MacLean & Benham, 2010). Which evaluating audience should be catered too often depends on context. Consider for example the case of a criminal syndicate: regulative legitimacy is not of
importance as by their very nature these organizations are illegal. Internal legitimacy however is very important, as it fosters a sense of belonging and faith in the syndicate. This in turn will keep
syndicate members from threatening the survival of the organization by for example opening up to law enforcement. In order for a criminal syndicate to survive, it is thus legitimacy with its
organizational insiders that syndicate leaders should maintain first and foremost. 1.3.2: Typologies based on evaluated feature(s)
Organizational activity brings outcomes that might provide benefits, losses or hazards of different magnitudes for potentially multiple distinct audiences. By analyzing how social actors value the outcomes of organizational behavior, Suchman (1995) isolates three feature based types of organizational legitimacy: Pragmatic, Moral and Cognitive
Firstly, there is pragmatic legitimacy, through which social actors grant legitimacy based on private cost/benefit calculations. Ashforth & Gibbs (1990) describe the same concept as audiences granting organizations legitimacy because of the concentrated benefits they bring. Within pragmatic
of the legitimacy created by what stakeholder theorists call ‘symbolic management’. Although these actions do not necessarily mean that this will change the outcome of organizational behavior, legitimacy can be granted based on the organization being seen as responsive to critics’ demands (Meyer & Rowan, 1991). The third and final form of pragmatic legitimacy is labeled Dispositional
Legitimacy. Social actors increasingly personify organizations to the extent they are seen as having
their own ‘character’ or ‘personality’. This leads to audiences to legitimate those organizations that share their own values (Scott, 1992). In times of adversity, organizations possessing high levels of dispositional legitimacy might be insulated from potential legitimacy loss in the case of small-scale and isolated failures (Suchman, 1995).
The second feature-based legitimacy type identified by Suchman (1995) is Moral Legitimacy. By attributing moral legitimacy to a firm, audiences go beyond a calculation of personal benefit towards a more normative judgment of firms doing ‘the right thing to do’. This type of legitimacy brings the most diffuse benefits to society (Ashforth & Gibbs, 1990). Moral concerns by audiences are more difficult to manipulate and are thus considered to require substantial rather than symbolic
management to overcome. This makes moral legitimacy more difficult and costly to obtain, but also more durable and thus desirable as an organizational resource. Within the realm of moral legitimacy judgments, there are four notable subtypes identified. Consequential legitimacy is granted if the legitimating audience is satisfied with the accomplishments of an organization. This legitimacy type is of specific importance for firms in markets where most decisions are characterized by low
involvement, such as firms operating in the fast-moving consumer goods segment. Legitimacy in these markets is based on quality and value (or, in micro-economic terms; utility). Organizations who provide their audiences with good quality products and services at reasonable prices will thus easily obtain consequential legitimacy. In addition to providing audiences with socially valued
consequences of organizational behavior, moral legitimacy can also be gained on the basis of adopting socially accepted techniques and procedures. This type of legitimacy is called Procedural
Legitimacy, and becomes significant in the event that outcomes are hard to measure. Corporate
their personal legitimacy to the organization they lead purely on the basis of their presence. It is in effect the only type of moral legitimacy that can be placed in the realm of symbolic rather than substantive management, as the appointment of a superstar CEO does not necessarily mean that any organizational features, structures or behaviors change significantly.
Where the previous two feature-based types of legitimacy are based on active evaluation and support for an organization, cognitive legitimacy does not involve this process of scrutiny by legitimating audiences. Tost (2011) argues that cognitive legitimacy is created on the base of either moral or pragmatic legitimacy, and such is not based on organizational features. Institutional theorists and resource dependence theorists however have a different explanation. Even though evaluations of the organization can be negative, legitimacy might still be granted if the organization is considered to be inevitable and/or necessary. This makes cognitive the most valuable feature-based type of legitimacy for an organization to possess: it means the organization’s existence is not questioned and legitimacy is granted without scrutiny (Suchman, 1995).
1.4: The Formation and Importance of Legitimacy Judgments
1.4.1: A model of Legitimacy Judgment; Judgment FormationNow that definitions and bases of legitimacy have been discussed, the next step is to examine the process of formation of this particular type of judgment. Social judgments, such as the assessment of an organization’s legitimacy, status, or reputation, are useful to audiences because they can guide economic exchanges (‘should I buy a product from this manufacturer?’), or simply because having judged an organization can help audience members to engage in social exchanges (Bitektine, 2011). Take for example a member of a gentleman’s club where the smoking of cigars is commonplace. The social judgment will not just guide the decision of what cigars to buy (optimizing utility of the economic exchange by making an informed decision, Mankiw, 2011). As a matter of fact, it will also facilitate a lengthy discussion about different cigar brands with other club members (Thus, enabling social exchange). Being able to participate in social exchange fosters relationships, and can earn esteem and recognition by other participants. These socio-emotional resources in turn can be called upon somewhere in the future, for example when the other participants can recommend another person for a prestigious position in their organization (Rupp & Cropanzano, 2002).
Figure 2: A model of legitimacy judgment formation Adapted from Tost (2010)
In the active mode of judgment formation, the evaluator chooses to engage in an (intensive) process of personally gathering information and processing this information. The ‘generalized legitimacy judgment’ is then formed on the basis of how well the processed information fits the potential legitimacy dimensions: the Pragmatic (or, instrumental), Moral and Cognitive beliefs of the evaluator. These three dimensions are not mutually exclusive; a positive judgment on one dimension does not mean that the other two dimensions will also be judged positive. Depending on context, one or more dimensions will be evaluated, the weight of each dimension in the final legitimacy judgment is not necessarily equal. However, the legitimacy granted to the organization under evaluation on each dimension that is considered necessary to form the general legitimacy judgment must pass the yes/no threshold in order for the final judgment to be positive. Consider for example three recently graduated university students, trying to find a job. After writing countless application letters, and having had dozens of interviews, they each receive three job offers. In both cases, Company A and C offer a lower wage than company B, but company A is seen as a ‘caring’ company that ‘gives back to the environment’ by promoting cultural and sports-related activities in their city. Company B
top management. Company C is not known for its social engagement, nor is it involved in any
scandals, but each of the three graduates has done a summer internship here. As it turns out, each of the graduates have use a different evaluative dimension in making their judgment: The first graduate decides that he wants to earn as much as possible, thus joining company B; his final evaluation is based on pragmatic legitimacy more than on the other dimensions. The second graduate, having thoroughly enjoyed his summer internship in company C, decides that this is the best place for him to work. Having previously worked at Company C, this company has earned cognitive legitimacy (in this case, based on familiarity) in the eyes of the second graduate. The third graduate decides that if he really has to start working for an organization, this organization better be one that does more than taking inputs from the environment solely focused on extracting the optimal profit. He decides to join company A because this company has moral legitimacy to the extent that it offers the most diffuse benefits out of his three options.
The passive mode in contrast has the evaluator simply granting the organization legitimacy because others do so, or because there is no reason to question the legitimacy. Judgment from other social actors can be used a proxy in many situations. The advantage of doing this is that it requires much less time and resource, a concept that is termed ‘cognitive economy’ (Singh, 1994). These other social actors include for example family members, co-workers, and friends, but also organizations such as the sports club one might be a member of, the media organizations or lobby groups behind publications one reads, and one’s employing organization. These influences are what Tost (2011) calls ‘validity cues’ and thus can replace the three content based dimensions that play a role in the active mode of legitimacy judgment formation.
The evaluator’s choice of which mode to apply to the judgment of an organization’s legitimacy is based on two questions: First off, are validity cues available?, and secondly, to what extent does the
organization evaluated conform to the evaluator’s social expectations (thus, does the organization fit
within current institutional boundaries)? If sufficient validity cues are sent by social actors that the evaluator has already legitimated, the passive mode is much more likely to dominate. The active mode become more likely when there are sufficient validity cues, but these are conflicting; not all social actors agree on whether or not the organization under evaluation is legitimate. This can give the evaluator to believe it is desirable to gather and process his or her own information: the evaluative mode takes over (Kahneman & Frederick, 2002)
1.4.2: A Model of Legitimacy Judgment: The Use Stage
The second stage in Tost’s model of legitimacy judgment is labeled the use stage. In this stage of the process, the evaluator acts upon the generalized legitimacy judgment formed in the previous stage: he or she either supports or seeks to change the organization in order for it to align with the evaluator’s pragmatic, moral and/or cognitive considerations. This judgment in turn can act as a validity cue to other evaluators, in a fashion similar to how a reputation cascade works (See Kuran & Sunstein, 1999). New information that becomes available to the evaluator during the use stage is processed in one of two ways. In most cases, the original legitimacy judgment becomes the basis of a process of assimilation (Tost, 2011; Kunda, 1990): any new information that might influence the legitimacy of the organization is processed in a way that it fits with the evaluators judgment as created in the judgment formation stage. In essence, the judgment acts as a primer, making it much more likely that new information is processed in a way consistent with the earlier judgment (Förster e.a., 2008). This process of judgment replication is also called motivated reasoning (Rousseau & Tijiorwala, 1999), and it is argued that the main reason evaluator’s engage in this process is because it reduces the cognitive energy needed to evaluate an organization continuously (Tost & Lind, 2010). In simple terms, people assimilate information because it is easier, and will leave them more time to make other social judgments (Tost, 2011). Assimilation will also lead to any original negative
elements in the legitimacy formation stage to fade, so any negative pieces of information processed before will no longer affect the overall judgment. In the long run, assimilation will also make the basis of the original judgment fade, leading to a taken-for-granted type of legitimacy. This means that eventually, as long as all new information fits with the evaluators perceptions of the organization, taken-for-granted (thus, cognitive) legitimacy will emerge (See also Greenwood e.a., 2002)
Consider however the case where the new information is in contrast with the information on which the general legitimacy judgment was based. Because it is so different, the evaluator will not be able to process it in a fashion that supports his original judgment. If this is the case, the evaluator moves on to the third stage of the process, the judgment reassessment stage. This stage is in many ways similar to the first stage, with the exception that in the reassessment stage the evaluator is much less likely to use legitimacy cues in the process of forming a new judgment. Thus, the judgment
reassessment stage is much like the active mode of judgment formation (Tost, 2011). The outcome of the reassessment stage is a generalized legitimacy judgment; this however does not have to be one opposite to the original judgment. New information is not necessarily negative, and negative
1.4.3: A Model of Legitimacy Judgment: Reassessment Triggers
Three types of reassessment triggers can be identified (Tost, 2011): Contradictions in institutional logics, reflexivity and jolts (or, shocks). Contradictions in institutional logics can occur when social actors become involved in different institutional systems. Consider for example an employee of a large multinational company. Having been part of this multinational for quite a while now, he has become accustomed with the ways in which his organization does business in developing countries, where it extracts most of its basic material resources from. In return for this depletion of the developing countries’ resource base, the multinational firm has set up an extensive and costly Corporate Social Responsibility (CSR) project, which aims at helping local communities by providing funds for construction projects and schooling. It is also basis for the employee to grant legitimacy to his firm, as he believes that the firm is ‘caring’ and ‘giving back’. However, at some point the
employee decides to stay in for the evening, switches on the TV and starts changing the channels, ending up seeing a documentary showing the devastating impact his employer’s resource extracting operations have on the life of the local community. Being drawn into this new institutional reality, his new-found insight clashes with the institutional reality he has grown so accustomed to; a
contradiction is formed. Of course, this contradiction does not necessarily mean the employee will change his earlier legitimacy judgment: the wage might be too good, he might believe that the company isn’t to blame or he simply does not care about the lives of these people living far away. Another trigger is reflexivity, or the evaluator’s ability to distance oneself from a judgment to reflect on the arguments and backgrounds of how one came to make that judgment (Mutch, 2007). In essence, these people converse with themselves, trying to flesh out the antecedents of their judgment and bringing up counterarguments. It is suggested that by far not all people possess this trait (Tost, 2011), although it could also be argued that people choose not to be reflexive because of the cognitive energy such a continuous self-deliberation might cost.
that results in little discomfort of any variety. Another scenario would be the same car owner being called by his local dealership that his car his being recalled (along with 1000s of other cars of the same type) because of a large oversight or error in the manufacturer’s production facility. In this case, there seems to be a structural problem. This might indicate to the car owner that the legitimacy he originally granted to the manufacturer might have to be revoked. He thus enters the legitimacy reassessment stage. Once a new judgment is formed (based on the available new information, for example about how the defect came to be), the car owner will move on to the use stage again. It is thus this judgment reassessment stage where a loss of organizational legitimacy can occur. In the next section of this paper we will look at one of the ways in which the legitimacy judgment of organizations can become reconsidered by legitimating audiences.
1.5: Organizational Crisis
As mentioned in the previous section, one of the triggers that can cause legitimating audiences to reassess the legitimacy of an organization is the occurrence of jolts (or, shocks) within the
environment of said organization. Another reason is a contradiction in institutional logics. Massey (2001) argues that organizations may experience a crisis of legitimacy when negative societal outcomes can be directly linked with the behavior of an organization. In turn then, organizational crises can be considered such jolts. It is also postulated by some organizational theorists that
organizational crises lead to a breakdown of meaning (Pearson & Clair, 1998); they are hard to make sense of because they fit so badly with the institutional logic that is present in society at the time of the crisis. In effect, they thus provide with the contradictions in institutional logic that Tost (2011) describes. Because of its’ dual link to judgment reassessment triggers, I argue that organizational crises provide an excellent setting in which to study organizations facing a (potential) loss of
legitimacy. In the following section, I will further describe the concept of organizational crisis and the impact crises can have on firms.
Organizational activities often have unintended side-effects that can harm their stakeholders or even the natural environment. In most cases, these side-effects (referred to by economists as
Organizational crises can come in many forms, including but not limited to for example product malfunctions (e.g., rupturing breast implants) , environmental damage (BP’s Deepwater Horizon oil-leak), CEO illness (Apple’s Steve Jobs), copyright wars (Samsung vs. Apple) and terrorist attacks (London subway attacks). Although these crises are indeed very different in nature, they do share a number of common elements, as brought together by Pearson and Clair (1998; p61) in their often used definition of organizational crisis:
‘An organizational crisis is a low probability, high-impact event that threatens the viability of the
organization and is characterized by ambiguity of cause, effect and means of resolution, as well as by a belief that decisions must be made swiftly’
Organizational crises are thus rare events that can have great effect on the organization, to the extent that some crises can have dire consequences for organizational survival (Massey, 2001). As Yu e.a. (2008) argue, crises are often characterized by indicators that can only in retrospect be exposed as causes. The full effect of a crisis event might be unknown for long periods of time, which means that years after the event an organization can still suffer from consequences. The high-profile case of the Softenon medicine is a good example of this. Prescribed to pregnant women in the 1950s and 60s against morning sickness, the substance was later found to cause a rare congenital disorder
Phocomelia. Although the producer of Softenon already contributed to a victims fund in the late 1960s, they have recently again done so, proving that the backlash of a large organizational crisis can still hit the organization decades later (NOS.nl). As crisis are rare events, organizations are often not prepared to deal with them, leading to uncertainty about the best way to respond. Yet, as scholars agree, a quick response to a crisis is necessary in order for the organization to be able to contain the crisis as much as possible (e.g. Horsley & Barker, 2002).
Many organizational theorists in the past three decades have focused their attention on the antecedents to organizational crises. One particularly influential strand of literature focuses on explaining how crises are an unintentional result of the structure of organizations, especially relating to the ever increasing complexity of organizational technologies. Perrow (1984) for example argues that because of the way society in general and organizations in particular are structured, the
technologies safeguarding increasingly complex technological systems continues to widen as industrial development progresses.
1.6: Consequences of Organizational Legitimacy Loss
A number of organizational consequences following legitimacy loss are summarized by Hamilton (2006). First of all, organizations in crisis are likely to react in defensive terms, which in many cases does not reassure stakeholders that legitimacy loss is not warranted (see also Ashforth & Gibbs, 1990). As organizations with low perceived legitimacy respond defensively, audiences become more and more suspicious about the organization; the need for an organization to legitimize itself is thus negatively linked to audiences’ acceptance of such legitimacy attempts, a mechanism that has been dubbed the ‘self-promoter’s paradox’. Repairing organizational legitimacy is thus a more complex task than the initial acquiring of it, because of the vicious circle that exists when organizations are forced to respond to a loss of legitimacy (Suchman, 1995).
A second set of important outcomes for an organization facing legitimacy loss are the actions of those stakeholders that control the organization’s critical resource flows. An organization’s
legitimacy plays a vital role in acquiring resources such as financial and human capital, as it is argued that audiences are much more inclined to provide such resources to organizations that they trust and value (Suchman, 1995). As legitimacy is a socially constructed resource, important stakeholders to the stricken organization such as suppliers and employees and political connections can become stigmatized by association with the stricken firm. Stakeholders that fear the risk of such contagion are likely to attempt to sever ties with the organization, potentially leaving the organization without the resources it needs to continue business as usual (Hamilton, 2006). This retraction cascade bring large costs to organizations, not just by the short-term loss of business as usual, but also by the time and money necessary to rebuild relationships with new suppliers of the lost resources (Suchman, 1995). Another way in which stakeholders can retract support is through product boycotts, where (potential) buyers of the organization’s outputs decide to no longer do so because they do no longer deem the organization desirable.
Finally, a loss of legitimacy can also lead to lower stock prices when investor’s deem the legitimacy loss as potentially harmful to future profitability (see for example Baucus & Baucus, 1997).
When the downward spiral of legitimacy loss can’t be broken, the organization can face death
order to shake off the persistent negative social judgments. Complete seizing of operation might also occur, if the loss of legitimacy has so badly scarred the organization that no buyer or merging
organization can be found, and critical resource flows are cut off to such an extent that the organization simply can no longer function.
1.7: Categorical Delegitimation
Loss of legitimacy has for a long time been a topic of research in academia. Much is known regarding how individual organizations respond when facing potential legitimacy loss as result of for example accidents, scandals or other ‘disruptions’ that arise directly through their own operations (Desai, 2011). In some cases of these disruptions however, not only the focal firm is affected, but also other firms that share one or more characteristics of the focal firm. An often cited example of this is the Enron scandal, which not only lead to the demise of auditor Arthur Andersen, but also resulted in tighter regulations and more cautious customers for other accounting firms (Jonsson e.a., 2009). Organizational crises can thus become contagious to the extent that they will bring about more uncertainty (for example financial and legal) for firms other than the focal firm. Recent studies into this phenomenon have dubbed this contagion effect ‘categorical delegitimation’ (Jonsson e.a., 2009), ‘Spillover’ (Yu e.a., 2008), loss of ‘Field Legitimacy’ (Desai, 2011).
The concept of spillover has been more extensively studied in a marketing context, but on the level of brands rather than organizations. Lei e.a. (2008) for example argue that once similarities between brands are established within the consumers’ minds, exposure to information of one brand from other brands is inevitable. When new information about a brand becomes available, this will trigger the individual to remember similarities and linkages to other brands. The new information will then to some extent also be applied to these similar brand, where the size of the spillover is determined by both the quantity and quality of similarities, and the audiences motivation to process the new information.
In a comprehensive study of the contagion effect on the organizational level in the Swedish insurance industry, Jonsson e.a. (2009) argue that there are two mechanisms that facilitate the spread of legitimacy loss beyond the focal firm, categorization and generalization. Categorization is used by individuals as a means of sense making. By considering a firm to belong to a certain category, any information about the firm that the individual does not possess is being replaced by the
2. Hypotheses
The following section is devoted to the development and justification of the hypotheses to be tested in the data analysis section of this paper. Based on the conceptual model presented in the
introduction, there are two different set of hypotheses. The first deals with how the prevalence (in an innocent firm) of the attribute made salient by the crisis affects an innocent organization’s legitimacy. The second set of hypotheses is based on the relationship between the responses of innocent firms and their effect on legitimacy.
2.1: Salient Issue Prevalence and Categorical Delegitimation
Categorical delegitimation can spread in three generic ways according to Jonsson e.a. (2009): Firstly within an organizational form, secondly across organizational forms with similar characteristics as the ‘deviant’ organization(s), and across organizations with characteristics made salient by the deviant act. As such, it stands to reason that when an organizational crisis occurs in one firm in an industry, other firms in the same industry are likely to be facing negative effects of this crisis. The following hypothesis can then be formulated:
Hypothesis 1
The occurrence of an organizational crisis in a corporation will lead to spillover effects (categorical delegitimation) for firms within the same industry.
When a crisis occurs, audiences will likely focus their attention on those issues made salient by the crisis, in an effort to save cognitive energy. It stands to reason then that innocent firms that are more similar in operations as the firms initially stricken by the crisis will also be more strongly affected by categorical delegitimation effects. The following hypothesis can now be formulated:
Hypothesis 2
The more similar an innocent firm is to the crisis-stricken firm in terms of operations, the stronger the effect of categorical delegitimation will be.
from the early 17th century, how investors value stocks remains a debate among scholars. In general, two different beliefs can be distinguished from the literature on this subject.
Firstly, there are those scholars that argue that investors value stocks based on fundamental criteria. These are calculable criteria such as Earnings per Share (EPS) ratios, Price/Earnings (P/E) ratios and Return on assets (ROA) statistics. Fundamental stock valuation is thus based on facts and predictions about the company’s performance, and assumes rational behavior from investors in terms of their stock valuation methods (Fama e.a. 1965)
Secondly, there are scholars that argue investors base their valuation of stocks based on market criteria. This type of stock valuation is a more subtle art, as it implies trying to model behavior of social, irrational actors as well as simple mathematical ratios (Shiller, 2003).
Whichever method is right does not however influence the central premise of this paper: that firm responses to an organizational crisis (see next section) and the existence of potential spillover effects can influence the stock price of a corporation. In the case of the fundamental valuation method, if a firm indeed stands to damage from legitimacy loss (for example through being cut off from critical resource flows), this will undoubtedly affect the future earnings of the corporation, and thus in effect, its’ stock price. If the market based valuation is considered to be true, the same holds, as fundamental considerations are not completely dismissed. In fact, market based valuation would indeed leave more room for news to have an effect on stock prices as irrational actors are believed to be led also by their sentiments towards a corporation.
2.2: Organizational Crisis Responses and Categorical Delegitimation
When organizational crises occur, the stricken organization must respond in order to satisfy the information needs of the different stakeholders involved. Sharing information about causes, consequences and crisis management efforts can indeed help prevent a total loss of legitimacy (Pearson and Clair, 1998). A number of different channels are available to the organization through which to send the response message, for example press releases or commercials. In addition to that, legitimacy judgments can also be influenced indirectly, through external institutional intermediaries (Yu e.a., 2008). These intermediaries do not necessarily own vital tangible resources for thethat the external institutional intermediaries can influence the organization’s stakeholders in two ways: Firstly, they serve as opinion leaders because they are perceived to have better access to information than other stakeholders, and more knowledge interpreting this information. Secondly, they focus stakeholders’ attention on a small number of salient details, rendering firms legitimate or illegitimate on the basis of their performance on these salient details (Yu e.a., 2008).
Because of the potential influence external institutional intermediaries can have on their legitimacy, organizations are likely to try and influence the reporting external institutional intermediaries where possible. This can be done for example through industrial self-regulation and corporate social responsibility (to appease activist groups), and also through the release of press statements, use of spokespersons and public appearances of key management figures on radio and television (Desai, 2011). Because the firm responses are new information, they are likely to influence stock prices. As such, the following hypothesis can be formulated:
Hypothesis 3
Crisis responses by ‘innocent’ firms will influence the magnitude of the effect categorical
delegitimation following the occurrence of an organizational crisis in a firm within the same industry
Building on the extensive literature on strategic communication, Lamin and Zaheer (2012) identify five different response strategies: no action, denial, defiance, decoupling and accommodation. In the case of no action, the organization decides not to respond to any questions about the firms
legitimacy, possibly because the organization’s decision makers are aware of the existence of the self-promoter’s paradox. The other responses can be seen in a continuum from denial to
accommodation. Denial is a simple case of dismissing the problem as being non-existent; no action is said to be undertaken by the organization simply because the organizational decision makers do not agree that a challenge of legitimacy exist. Accommodation responses are those responses in which firms take substantive action, altering behavior as a result of the legitimacy challenge raised by an organizational crisis in one (or more) other firm(s).
It must be noted that each response strategy might have different effects on different stakeholders. Lamin and Zaheer (2012) show this by examining the effect of different strategies on investors and the general public. Because consistent information provision is important in legitimacy defense (Ulmer, 2001), only a single strategy or combination of strategies should be followed for each
strategy to follow, and that this choice will be influenced by the decision makers’ perceptions about the importance of different stakeholders.
An important concept that can guide us in arguing the effect of these different types of responses is the concept of stakeholder ‘thought worlds’ (Lamin & Zaheer, 2012). Based on Freeman’s (1984) observation of the existence of heterogeneous sets of stakeholders for each firm, Lamin and Zaheer (2012) show that because of this heterogeneity, stakeholders filter, process, and interpret
information in vastly different ways. Their research has two findings that are relevant to this study. Firstly, it is argued that there is a positive relationship between defensive responses and stock price movements following a crisis. Opposite goes for the accommodative responses, which are claimed to have a negative relationship to stock price movements following a crisis. Although this research is done in the context of a focal firm in crisis, rather than in the context of firms facing categorical delegitimation, I argue that the same relationships exist. This follows from the fact that defensive responses provide investors with assurance that business as usual will continue (thus lowering uncertainty), whereas in contrast accommodative responses lead to change that often is accompanied by extra costs or lower revenues. The following hypotheses are thus formulated:
Hypothesis 4A
Defensive responses by ‘innocent’ firms will have a positive impact on these firms’ stock returns following the occurrence of an organizational crisis in a firm within the same industry.
Hypothesis 4B
3. Data and Methodology
The first phase of the ‘original’ research element in this paper was to identify a number of
organizational crises that would allow for the hypotheses to be tested. A few factors were taken into account for a crisis to qualify. First and foremost, in order to ensure that enough data would be available, the crises to be under investigation had to be high-profile cases involving relatively large organizations with stock market listings. Secondly, the industries needed to either be global by nature, or at least firms active in these industries should be largely similar in operations, to facilitate cross-country comparison of firms. Finally, the salient issues for the crises studied to answer the first three hypotheses should be measurable either in absolute value or replaceable by proxy. In the next section the three selected crises are discussed.
3.1: Study context: Two High-Profile Organizational Crises
3.1.1: British Petroleum and the Deepwater Horizon CatastropheThe Deepwater Horizon mobile offshore drilling unit was leased by British Petroleum (BP) early in 2008. While drilling in the Gulf of Mexico on a location about 66 kilometers from the US coast on the 20th of April, 2010, disaster struck when an explosion occurred on the platform. Eleven people that were present at the time were killed in the explosion and its aftermath. Despite extensive efforts to put out the flames now engulfing the platform, the fire kept raging for two days. On the 22nd of April, Deepwater Horizon sunk to the ocean floor. This however was only the onset of a larger
organizational crisis to hit BP. Despite initial belief no oil had spilled from the well into the ocean, on the 24th of April a leak was announced to be found. While initial estimates of the size of the leak were relatively small, over the course of a couple of months BP had to concede that the damage was much larger than it had originally pictured. A number of failed attempts to close the leaking well and BP’s initial persistent denial of the impact the spill would have, created a high-profile organizational crises that dominated news reports around the world for months. The salient issue in this organizational crisis is oil production, and in particular offshore drilling for oil. As offshore drilling accounted for approximately 30% of the total daily production worldwide (Reuters, 2010), it is likely that this organizational crisis has had a market-wide impact, making it an excellent incident to study in the context of organizational delegitimation.
3.1.2: Fukushima Daiichi nuclear disaster
Construction on the Fukushima Daiichi (1) nuclear power plant began in 1967, and the plant was started operating at full capacity in 1971. Owned and operated solely by the Tokyo Electric Power Company (TEPCO), it was built on the Japanese coastline near the town of Okuma. To prevent potential tsunami damage, the plant was initially scheduled to be built on a height of 35 meters above sea level, although in the final design meant it was constructed only 10 meters above sea level, as this was deemed enough protection against any tsunami’s under normal circumstances. On March 11, 2011 however, an earthquake of a 9.0 magnitude occurred some miles off Japan’s northeast coast. This earthquake lead to a tsunami with waves up to 14 meters. Although the reactors were automatically shut down, the tsunami disabled the emergency generators that
supplied energy necessary to cool down the radioactive fuel used in the main generators. As a result, partial nuclear meltdowns occurred in several of Fukushima 1’s reactors, leading to the release of highly dangerous radiation and the subsequent evacuation of people in a 20 kilometer radius around the plant. The salient issue in this organizational crisis is the generation of nuclear energy, a practice that invariably constitutes large risks. Some countries however are largely reliant on this type of energy; in France for example in 2011 77.7% of total energy production involved nuclear power (Nuclear Energy Institute, 2012). It is not unlikely that an incident like this has led to a reevaluation of the nuclear activities of many power companies around the world, making this organizational crisis a good candidate to study categorical delegitimation.
(Story Sources were obtained through the LexisNexis Database and are available from the author upon request)
3.2: The Dataset – selection of firms
38 in the energy sector. A total of 408 firm-day observations together encompassed the final dataset. A full list of the companies can be found in appendix A
To identify firm responses, the LexisNexis database was used. Every company in database was used as a search term, limiting the responses to a timeframe of the day of the crisis event (for oil
companies 22/04/2010; for energy companies 11/3/2011) until 5 trading days after the event. For example, the keyword BP was entered and searched for in the period spanning 22/04/2010 up to and including 29/04/2010. Any hits were subsequently scanned for whether or not they entailed a
response by a company official related to the crisis event. If such a response was found, the full article was read and saved. (All responses used in this study can be obtained from the author on request.) A total number of 44 articles was found for the period, referring to 47 company responses. Six of these articles were related to oil companies (seven responses in total), and 38 articles referred to energy companies (a total of 41 responses). Responses were only coded once; hence, if the same response was quoted in multiple sources on different days, only the earliest response would be taken into account. Each response was subsequently coded by the author for being a defensive or an accommodative response. After coding of the responses, a second coder independently coded the articles again, agreeing on all 44 accounts with the original coding, leading to a high level of intercoder reliability.
3.3: Variables
3.3.1: Dependent variable; Categorical Delegitimation Index
The measurement of categorical delegitimation in stock prices brings a difficult question of how exactly this is best achieved. Previous studies in spillover effects and categorical delegitimation did not focus on stock prices (e.g. Jonsson e.a., 2010). Lamin and Zaheer (2012) in their study of
legitimacy loss do include stock price measurements, by operationalizing their dependent variable of ‘investor perceptions of the firm’ as the cumulative abnormal daily stock return (CAR) for the stricken firm. This CAR is calculated by observing the firms’ stock returns starting from 200 days before the event until 50 days after the event.
The CD index value is thus 1 when the daily stock return movement for a company is a decline equal to the mean movement plus two standard deviations. This threshold is chosen because based on the normal distribution, 97.5% of the observations should be below this value. This makes an occurrence of a CD Index value of 1 or higher a likely indicator of delegitimation.
3.3.1: Predictor Variable: Issue Salience
Measured as the percentage of business that the ‘innocent’ firm has from the sort of operation that has been made salient by the organizational crisis. For oil companies, this constitutes the percentage of operating income attributable to exploration (upstream) activities in the year before the relevant crisis (2009). For energy companies, this measure constitutes the percentage of the firm’s power generation capacity in MW attributable to nuclear power generation, in the year before the relevant crisis (2010). In line with Jonsson e.a. (2009), this variable is expected to have a positive relationship to the dependent variables.
3.3.2: Predictor Variable: Response
This variable answers the question of whether or not a company has responded to the crisis through an official press release or through a response by a company employee in the newspapers. It is modeled as a dummy variable, where the value 0 indicates no response, and 1 indicates there has been a response on that particular day. This variable is used in order to establish if responding, regardless of the tone of the response, has an influence on categorical delegitimation
3.3.3: Predictor Variable: Defensive Response
This variable is used to model one of two possible response types that are identified in this study. The defensive response type is characterized by firm officials claiming that the crisis event will not interfere with business as usual, and/or will not change any future commitments the firm has expressed in the past. Consider the following example, a quote from a press release from Entergy Corporation, on the 15th of March 2011:
“Entergy's nuclear plants were designed and built to withstand the effects of natural disasters, including earthquakes and catastrophic flooding.”
The defensive response variable is modeled as a dummy variable, with a value of 0 if there was no defensive response, and a value of 1 if there was a defensive response on the particular day of the observation.
3.3.4: Predictor Variable: Accommodative Response
This variable models the other end of the response spectrum. In contrast to the defensive response type, accommodative responses do imply that the firm has taken or will change its operations following the crisis event. An example of such a response can be seen in the following quote taken from a press release of German energy giant E.ON, released on March 16, 2011:
'Even though the nuclear power station Isar-1is not directly affected by the moratorium at the current time, we have decided to suspend the operation of E.ON's oldest reactor…’
The firm thus intends not just to react symbolic but by changing firm behavior following the events that had no direct influence on its’ operations. As hypothesis 3c states, it is argued that this variable will have a positive relationship to the dependent variable.
This variable is modeled as a dummy variable; a value of 0 indicates no accommodative response, whereas a value of 1 indicates that the company has reacted in an accommodative way on said observation date.
3.3.5: Control Variable: Organizational Age
It is often argued in research on organizational legitimacy that the age of an organization is positively related to the audiences’ judgment of an organization (Deephouse & Carter, 2005). Verhoef et al. (2002) for example argue that the length of a relationship has a positive effect on partner
evaluations, indicating that it is likely that older firms are more likely to get positive reviews, as they have had longer exchange relationships with their audiences. As investor legitimacy is based on how investors judge about the exchange between them and the organization (the financial returns), it can be argued that organizational age can provide a form of insulation against categorical delegitimation. Organizational age is calculated in either of the following two ways: the year of founding (as
indicated in the Orbis database), or the year of privatization (as indicated on the organization’s website). I have chosen to model age this way because in both industries, there is a long history of state ownership. This would have barred investors from building an exchange relationship with the organization, and as such investor legitimacy can have only developed once private investors were able to buy a stake in the organization.
3.3.6: Control Variable: Organizational Size
As is the case for age, organizational size is also argued to be positively related to the social
