The Journal of Socio-Economics 36 (2007) 825–840
Asymmetry in information versus asymmetry in power: Implicit assumptions of agency theory? Nicole J. Saam
∗
Universit¨ at Erfurt, Staatswissenschaftliche Fakult¨ at, Sozialwissenschaften, Nordh¨ auser Str. 63, 99089 Erfurt, Germany
Abstract
The hypothesis of this article is that agency theory makes implicit assumptions on the power relation between principal and agent. It implicitly assumes an asymmetry in power in favour of the principal. The bases of power approach [French Jr., J.R.P., Raven, B.H., 1959. The bases of social power. In: Cartwright, D. (Ed.), Studies in Social Power. University of Michigan, Ann Arbor, pp. 150–167; Raven, B.H., 1992. A power/ power/inte interac ractio tion n model model of interp interpers ersona onall influenc influence: e: French French and Rave Raven n 30 years years later later.. Journa Journall of Social Social Behav Behav-ior and Personality 7, 217–244] of social psychology is used to prove this argument. Surprisingly, the social power perspective does not only reveal this implicit assumption. It also paves the way for a new mechanism that resolves agency problems that had not yet been discovered by agency theory: identification systems. © 2007 Elsevier Inc. All rights reserved. JEL classification: c lassification: B52; D23; D70; D80 Keywords: Agency theory; Social power; Identification systems
1. Introduct Introduction ion
Together with transaction cost economics and the property rights approach, agency theory forms the new institutional economics (Furubotn ( Furubotn and Richter, 1991). 1991 ). In the centre of analysis are those institutions in which economic exchange takes place, like markets, organizations, and legal norms. The aim of institutional analysis is to explain the structure of institutions, their influence on individual action, their efficiency, as well as the change of institutions. Although it is mainly spread in economics and looked at as an economic theory, its explanatory power advances to the classical domains of social psychology and sociology: the new institutional economics defines itself as a theory of interaction based on rational choice behaviour. ∗
Tel.: +49 361 737 4901; fax: +49 361 737 4909. E-mail address:
[email protected] [email protected]..
1053-5357/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.socec.2007.01.018 doi:10.1016/j.socec.2007.01.018
826
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
White (1985) has elaborated the historical reach of principal-agent relations. Three agency relations have been of continuous interest: the relation between employee and employer, between shareholder and manager, and between creditor and stockholder. In recent years, agency has been applied in analysing research fields as diverse as the ownership structure of the firm (Short et al., 2002; Chen, 2001; Wright et al., 2001a,b; Jensen, 2000; O’Sullivan, 2000 ), incentive problems in accounting, foreign aid, or between shareholder and policyholder (Martens et al., 2002; Krishnaswami and Pottier, 2001; Lambert, 2001; Hubbard, 2000; Laffont, 2000 ), executive compensation (Hermalin and Wallace, 2001), organizational forms and the transformation of corporate forms (Bento and White, 2001; Zey and Swenson, 1999; Zey and Camp, 1996), corporate governance (Beyer and Hassel, 2002), changing interorganizational patterns in supply chains (Senter and Flynn, 1999), profit and nonprofit boards (Miller, 2002; Sapienza et al., 2000; Ott, 2000), the analysis of union leaders’ behaviour (Bemmels and Lau, 2001), state policy implementation (Kiser, 1999), nonmajoritarian institutions (Majone, 2000), the relationship between constituents and legislators (Lord, 2000; Strom, 2000; Longley, 1999, 1998), control of the executive (Franchino, 2000), civil–military relations (Feaver, 1998), and boundary organizations between politics and science (Guston, 1999). Critical contributions to agency theory as applied to political and administrative science have come from Miller and Whitford (2002), Eisner et al. (1996), and Worsham et al. (1997). A reexamination of agency theory assumptions and extensions on the base of behavioural theories have been presented by Wright et al. (2001a,b). Breton (1995) has elaborated the common features and the complementarity of agency theory and public choice and put forward the hypothesis that their integration could generate a general theory of bureaucratic organizations. In sum, agency theory has become an inspiring and integrating theory perspective to many at first sight very different fields of research. My hypothesis is that agency theory makes implicit assumptions on the power relation between principal and agent. Especially, I will prove that agency theory implicitly assumes an asymmetry in power in favour of the principal (Section 3). The bases of power approach (French and Raven, 1959; Raven, 1992) of social psychology is used to prove this argument. Surprisingly, the social power perspective does not only reveal this implicit assumption. It also paves the way for a new mechanism that resolves agency problems that had not yet been discovered by agency theory (Section 4). In the following, Section 2 will give a short overview on agency theory. 2. Agency problems and solution mechanisms—the traditional view
There is a wider and a more narrow definition of the agency relation among its classical writers: Jensen and Meckling define an agency relationship “as a contract, under which one or more persons (the principal(s)) engage another person (the agent), to perform some service on their behalf which involves delegating some decision making to the agent” (Jensen and Meckling, 1976, p. 308). This definition makes delegation explicit. There is a division of labour between principal and agent, and “delegated choice” (Rees, 1985, p. 3). Its formal structure is applicable to an even wider class of problems, “where no formal delegation relationship is explicitly involved” (Rees, 1985, p. 3). This definition is given by Arrow (1985, p. 37): there are “. . . two individuals. One (the agent) must choose an action from a number of alternative possibilities. The action affects the welfare of both the agent and another person, the principal. The principal, . . ., has the additional function of prescribing payoff rules.” There are three main differences respectively asymmetries in the relation between principal and agent. Agency theory explains the realization of the division of labour between both actors by these differences. (1) Informational asymmetries. It is rational for both principal and agent to enter
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
827
an agency relation. Either, a potential agent has competences (or more general information) to fulfil a task which the principal does not have (Pratt and Zeckhauser, 1985, p. 3); or both have the competences, but the potential agent can fulfil the task at lower cost. Informational asymmetries arise because the principal cannot monitor the competences (“hidden characteristics”), intentions (“hidden intention”), knowledge (“hidden knowledge”), and actions (“hidden action”) of the agent, or he can monitor them only at high costs. The principal needs this information to pay the agent depending on her effort. Additionally, the principal needs information on environmental states or processes which have influenced the agent’s work performance. Therefore, in the standard scenario of agency theory, asymmetrical information in favour of the agent is assumed. (2) Different risk preferences. Principal and agent have different attitudes towards risks which are related to different compensation schemes (outcome-based versus behaviour-based). In the standard scenario, agents are assumed to be risk averse. In comparison to the principal the agent’s income is low. Reductions in income endanger the existence of the agent. The principal can diversify, while the agent cannot. Therefore, the principal is assumed to be risk neutral (Eisenhardt, 1989, p. 60f). Meanwhile, there have also been specified models with risk neutral agents ( Milgrom and Roberts, 1992, p. 236ff). (3) Goal conflicts arise out of different preferences of principal and agent. Both want to maximize their individual utility. The agent wants to maximize his income, whereas the principal wants to maximize her returns. Both income and returns are based on the agent’s effort and exogenous random elements. Agency theory assumes that greater effort by the agent results in greater outcome. However, greater effort is also related to a greater disutility on the side of the agent. As a consequence, the agent wants to maximize his income while minimizing his disutility of effort whereas the principal wants to maximize her returns while the agent maximizes his effort. Problems which arise from this goal conflict are called agency problems. In order to characterize the optimal forms of agency contracts the agency problems have to be solved. Agency theory solves them from the perspective of the principal and suggests mechanisms which are outlined below. The number of agency problems has increased from originally two to four. Agency problems are problems of the principal, and they arise to the principal because the agent has private information. Arrow (1985, p. 38) has borrowed the problems of moral hazard and adverse selection from the practice and theory of insurance and has reconceptualised them in more general form as hidden action and hidden information. Meanwhile hidden information has differentiated into hidden characteristics, hidden intentions, and hidden knowledge ( Spremann, 1990). However, in the agency literature, there is not yet a uniform use of the concepts. The normative agency literature often sticks to the concepts of moral hazard and adverse selection, whereas positive agency theory prefers the reconceptualised concepts. (1) Hidden characteristics. Before contracting the principal has incomplete information on the agent’s qualification, like her skills or abilities. The agent may claim to have certain skills when he is hired. A problem arises because the principal cannot completely verify these skills or abilities either at the time of hiring or while the agent is working (Eisenhardt, 1989, p. 61). (2) Hidden intentions. After contracting the agent has the opportunity to realize hidden intentions. He may act intentionally unfair with the principal. This agency problem has been called hold up (Goldberg, 1976). The principal cannot simply resign from the contract relation because she will lose the investment that she had when she contracted with the agent (sunk costs). (3) Hidden knowledge or hidden information. After contracting the agent has private knowledge on exogenous facts like environmental states and processes which are relevant to evaluate his work’s outcome. This knowledge is part of the agent’s professional or process knowledge which he may use opportunistically. This is Arrow’s (1985) hidden information definition.
828
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
(4) Hidden action. After contracting the agent can choose among different actions. The agent may shirk. He may choose to work less but to pretend to work hard which cannot easily be evaluated by the principal. This is the former moral hazard definition. Agency theory has elaborated general mechanisms for the solution of each agency problem: (1) Incentive compensation systems (Jensen and Meckling, 1976) or reward systems (Eisenhardt, 1989). The more difficult and expensive the monitoring of the agent’s activities is the more important are incentives. Incentive compensation systems coalign the preferences of the agent with those of the principal because the rewards for both depend on the same actions. Efficient incentive compensation systems give a share in the returns to the agent. Both arethen interested in high returns. The implementation of incentive compensation systems generates costs to the principal and risks to the agent. Therefore, a trade off between agency costs and increase in productivity by incentive compensation systems has to be considered by the principal. Incentive compensation systems are typical solutions to hidden information and hidden action. (2) Monitoring systems (Jensen and Meckling, 1976; or information systems, Eisenhardt, 1989). Monitoring systems serve to control the agent (Jensen and Meckling, 1976, p. 323; Eisenhardt, 1989, p. 64). The principal systematically collects information on the agent’s activities. The agent is more likely to behave in the interests of the principal if the principal has information to verify the agent’s behaviour (Eisenhardt, 1989, p. 60). Again, the implementation of monitoring systems such as budgeting systems, reporting procedures, boards of directors, and additional layers of management generates costs to the principal. Perfect monitoring systems are unrealistic. Pure information systems are often not efficient. They have to be complemented by control systems or incentive compensation systems. Monitoring systems are typical solutions to hidden action. (3) Vertical integration (Klein et al., 1978; Alchian and Woodward, 1988) establishes a relation of authority between the originally equal partners of principal and agent. Now, the agent has to comply with the instructions of the principal. Compliance is fostered by sanctions. Vertical integration is prevalent in labour contracts. Again, the implementation of vertical integration generates costs to the principal. Vertical integration does not function as an incentive. Agency theorists often neglect hierarchy and norms as solutions to agency problems. Vertical integration is often only interpreted in a monetary way. Then, it becomes difficult to differentiate vertical integration from incentive systems. For example, promises concerning pensions and bonuses are granted only after the agent has been loyal to the principal for many years. Vertical integration is the typical solution to hidden intentions. (4) Game theoretical solutions. Agency problems can be represented as games like in game theory, and vice versa (Rasmusen, 1990). Agency relations are considered as “hybrid between a cooperative and non-cooperative game” (Myerson, 1982, p. 68). In situations in which vertical integration is not effective game theoretical solutions may help. For example, an agency problem may be modelled as a prisoner’s dilemma. Then, one can alter the situation on the base of the suggestions which game theory offers to generate ongoing cooperation. Interaction frequency and length can be increased. The importance of future interactions can be increased. Sanctions can be imposed on one-sided defective actions. Game theoretical solutions are typical solutions to hidden intention. (5) Self-selection (Arrow, 1986). The principal can offer different contracts to the agent, e.g. an outcome-based and a behaviour-based one, which vary the risk and the returns of the agent. By choosing one of the contracts the agent makes transparent her risk preference and her
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
829
willingness to work hard. Again, the implementation of self-selection generates costs to the principal. Self-selection is a typical solution to hidden characteristics and hidden information. (6) Signalling . The agent may herself signal her risk preference and willingness to work hard (Spence, 1973, 1974; Riley, 1979; Grinblatt and Hwang, 1989). He offers his hidden characteristics in order to convince the principal that he is willing to work hard. In contrast to the other solution mechanisms, the agent pays the cost of signalling. For example, the choice of a high level education and the presentation of certificates are interpreted as signalling. Signalling is a typical solution to hidden characteristics and hidden information. (7) Bonding (Jensen and Meckling, 1976) reduces hidden action. The agent binds himself to certain actions, e.g. he produces information on his work performance, and he agrees upon sanctions if he should not produce them. The income of the agent depends at least partially on his performance. Again, bonding generates costs to the principal because to be efficient bonding has to be controlled. (8) Screening reduces hidden characteristics as well as hidden information. The principal improves her selection procedures like tests, or assessment centres (Spence, 1974) which generates costs to her. 3. A social psychological perspective on agency problems and solution mechanisms
From a social psychological perspective, each agency problem describes a situation in which the principal intends to change the behaviour of the agent if the latter should be willing to behave in an opportunistic way that harms the principal. The potential for changing an other person’s behaviour is captured by the social psychological definition of power. Raven defines influence as “a change in the belief, attitude or behaviour of a person – the target of influence, which results from the action, or presence, of another person or group of persons – the influencing agent” (Raven, 1992, p. 218). He defines social power as “ potential for such influence” (Raven, 1992, p. 218). This is, in contrast to Weber (1947), an objective definition of power. The will of the influencing agent is irrelevant. The objective effects on the side of the target of influence are considered. However, it is quite difficult to measure the change of beliefs and attitudes objectively. Nevertheless, this facilitates the identification of the exertion of power. In Weber’s conception it remains unclear how to verify that an agent has realized his or her will. The price of objectivity is the renunciation of the influencing agent’s will. As a consequence, Raven’s actors exert power even if they do not have the will to do it. For example, someone will have exerted power if he has told his new fellow at the bus stop that all busses are late and this person decides not to wait at the bus stop, but to walk by foot. The exertion of power thus becomes a phenomenon of everyday life and its negative connotation is removed. French and Raven (1959) have developed a bases of power concept that is widely accepted in social psychology (Podsakoff and Schriesheim, 1985). It has been applied, verified, criticized, and further developed in several steps for nearly 40 years (Raven, 1965, 1974, 1992; Collins and Raven, 1969; Raven and Kruglanski, 1970; Raven and Rubin, 1976). This concept has been applied to diverse social interactions, e.g. parent–children interaction, interaction of wife and husband, of teacher and pupil, physician and patient, of buyer and seller, of franchisee and franchiser, and interaction among politicians. Since 1965, French and Raven have differentiated six bases of power, “resources which an influencing agent can utilize in changing the beliefs, attitudes, or behaviours of a target: Reward, Coercion, Legitimacy, Expertise, Reference, and Information” (Raven, 1992, p. 218). Their conceptualisation presumes a dyad (French and Raven, 1959, p. 150). The interacting agents are
830
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
called influencing agent O ( other ) and target of influence P ( person). They are assumed to be rational (Raven, 1992, p. 228). Reward power is based on P’s perception that O has the ability to mediate rewards for him. Coercive power is based on P’s perception that O has the ability to mediate punishments for him. Legitimate power is based on the perception of P that O has a legitimate right to prescribe behaviour for him. Referent power is based on P’s identification with O. Expert power is based on the perception that O has some special knowledge or expertness (French and Raven, 1959, p. 155f). Informational power is based on an information which O forwards to P. Here, the contents of the information is most important, not the expertise of O (Raven, 1965, p. 372). Raven differentiates the bases of power with respect to their range. The range of power includes all actions which the influencing agent can evoke on the side of the target of influence through her usage of power: 1. Reward power and coercive power are limited to situations in which the influencing agent can reward or punish. 2. Legitimate power has a broader range because it may be exerted in all situations in which the value system is valuable. Furthermore, within the scope of legitimate power the influencing agent is allowed to reward or sanction. 3. Expert power and informational power are limited to specific situations in which expertise or information are relevant. However, expert power and informational power may be generalized. The target of influence may learn about the influencing agent who exerts expert power or informational power. Then, an expert may exert referent power. 4. Referent power depends on attractivity of the influencing agent. It may be far reaching because it may be exerted in all situations which the target of influence relates to the influencing agent. French and Raven have stated that “for any type of power the size of the range may vary greatly, but in general referent power will have the broadest range” (French and Raven, 1959, p. 165). Raven underlines that in a situation several different bases of power may be operating (1965, p. 379f). He acknowledges that these six bases of power are “not the only means which people use in influencing one another” (Raven, 1992, p. 222). He discusses environmental manipulation (Raven, 1992, p. 229) and integrates it into his power-interaction model. Raven refers to Cartwright who defined that “when O influences P by ecological control, he takes some action which modifies P’s social or physical environment, on the assumption that the new environment will subsequently bring about the desired change in P” (Cartwright, 1965, p. 19f). This power base is most important for executives. “Since much of an executive’s responsibility is to design behaviour settings, he is bound to exert influence by ecological control, whether he is fully aware of it or not” ( Cartwright, 1965, p. 21). 3.1. Agency problems and solution mechanisms as power phenomena
If a principal and an agent interact, both may exert power on each other: the principal exerts power if she changes the belief, attitude or behaviour of the agent—e.g. by implementing a monitoring system; the agent exerts power if he changes the belief, attitude or behaviour of the principal—e.g. by giving some information to the principal. This raises the question whether all agency problems are power phenomena (Saam, 2002). Are agency problems problems of power, or are they informational problems as the traditional view states? Assume a principal who does not relate her decisions to information of the agent. Then,
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
831
hidden information is not a problem of power, but only one of information. The agent possesses the base of power information, but he cannot exert informational power. The same holds for the solution mechanisms of the principal. The agent needs not necessarily change his behaviour, e.g. if he did not have an opportunistic attitude he may simply show the same behaviour as before. That is, then the principal does not exert power on the agent although she possesses the bases of power to do it. The next two propositions capture these situations. Proposition 1a. Each agency problem becomes a problem of power to the principal if she changes her beliefs , attitudes, or behaviour as a result of the action , or presence of the agent . Proposition 1b. Each solution of an agency problem becomes a power phenomenon if the agent changes his beliefs , attitudes, or behaviour as a result of the implementation of a solution mechanism by the principal . 3.2. Asymmetry of power in favour of the principal
Does agency theory make implicit assumptions on the power relation between principal and agent? Especially, does agency theory implicitly assume an asymmetry in power in favour of the principal? In the following, I analyse one agency problem after the other and I show which bases of power are available to principal and agent. 3.2.1. The hidden characteristics-situation There is a nexus between hidden characteristics and informational power. Spremann describes the hidden characteristics problem as follows: “A principal on the one side of the market gets into contact with many individuals, potential agents, on the other side. The principal has to make an offer in the moment of getting into contact with one of these agents. The agents, however, differ in their characteristics. Although the principal might know the distribution of characteristics, she usually will be uncertain about the particular type of agent. How to make an offer that is appropriate without knowing the individual characteristics?” (Spremann, 1987, p. 11). Before contracting, the agent will be motivated to exert power if the principal offers no contract or a contract which the agent dislikes. Then, the agent would like to change the principal’s behaviour. However, at this point of time, the only resource of the agent is information on himself. Signalling is the solution mechanism of the agent that is based on informational power. In the hidden characteristics-situation, the principal mediates reward or punishment to the agent, because she decides whether one of the many potential agents will be offered a contract. Agency theory suggests self-selection and screening as the principal’s solution mechanisms for problems of hidden characteristics. There is a nexus between self-selection and screening on one side and environmental manipulation on the other side. By offering different contracts to the agent, e.g. an outcome-based and a behaviour-based one, which vary the risk and the returns of the agent, or by running tests like assessment centres, the principal manipulates the social environment of the agent. She designs behaviour settings in which the agent may not only give information which was hidden before, but also may change the agent’s behaviour. The following proposition captures that situation. Proposition 2a. In the hidden characteristics-situation the principal possesses quantitatively more bases of power than the agent .
832
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
3.2.2. The hidden intentions-situation There is a nexus between hidden intentions on one side and reward power, coercive power and referent power on the other side. After contracting the agent has the opportunity to realize hidden intentions. He may act intentionally unfair with the principal because there is always some freedom in interpreting the contract which cannot specify any detail of action and interaction. By acting fair or unfair the agent can mediate reward or punishment to the principal. Raven himself related hidden intention to referent power—although he did not know agency theory: “The term manipulation is generally used to refer to social influence wherein the intentions of the agent are hidden from the influencee” (Raven, 1965, p. 377). He gives the example of a political campaign worker who “deliberately makes himself extremely obnoxious at a bar, then leaves, advising others to ‘Vote for Blotz’, the opposing candidate”. Raven states that “when it is discovered by the influencee, it is likely to establish the manipulator as a person with very strong negative expert and negative referent power” (Raven, 1965, p. 377). The example does not represent an agency relation. Imagine a project manager who intentionally sets deadlines which his engineers cannot hold. After the engineers have discovered this, e.g. in a talk with the secretary, he will have established negative referent power. Both examples suggest that the principal may solve the hidden intentions problem by establishing positive referent power. If the agent identifies positively with the principal he will behave like the referent. Then, the principal may exert referent power. This may be interpreted as a first hint to a new solution mechanism which agency theory has not yet considered. Agency theory suggests vertical integration and game theoretical solutions as the principal’s solution mechanisms for problems of hidden intentions. There is a nexus between vertical integration on one side and legitimate power, reward power and coercive power on the other side. Vertical integration is the typical solution to hidden intentions. It establishes a relation of authority between the originally equal partners of principal and agent. Legitimate power is based on the perception of P that O has a legitimate right to prescribe behaviour for him. This is the agent’s perception in case of vertical integration. Within the scope of legitimate power the principal can reward or sanction the agent. Thus, vertical integration provides the principal with the bases of power legitimacy, reward, and coercion. There is also a nexus between game theoretical solutions on one side and reward power and coercive power on the other side. Agency problems can be represented as games. One can try to alter the situation on the base of the suggestions which game theory offers to generate ongoing cooperation. Thibaut and Kelley (1959, p. 101) have interpreted the ability of A to influence the returns of B as power of A over B. In the concepts of French and Raven, this is reward power and coercive power. Since vertical integration and game theoretical solutions are behaviour settings that the principal designs she exerts environmental manipulation. The following proposition captures that situation. Proposition 2b. In the hidden intentions-situation the principal possesses quantitatively more bases of power than the agent , and she possesses those power bases with the broader range . 3.2.3. The hidden information-situation There is a nexus between hidden information and informational power. “In the hidden information problems, the agent has made some observation that the principal has not made. The agent uses (and should use) this observation in making decisions; however, the principal cannot check whether the agent has used his or her information in the way that best serves the principal’s interest.” (Arrow, 1985, p. 39). The agent exerts (informational) power if the principal changes her beliefs, attitudes, or behaviours as a result of the agent’s information. There is also a nexus
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
833
between hidden knowledge and expert power. Hidden knowledge is defined as the agent’s professional or process knowledge which he may use opportunistically. Analogous to the previous case, the agent exerts expert power if the principal changes her beliefs, attitudes, or behaviours as a result of the agent’s knowledge. Agency theory suggests signalling as the agent’s solution mechanism for problems of hidden information. The only resource of the agent is information on himself. Signalling is based on informational power. Agency theory suggests incentive compensation systems, self-selection, and screening as the principal’s solution mechanisms for problems of hidden information. There is a nexus between incentive compensation systems on one side and reward power and coercive power on the other side. Incentive systems are mechanisms to solve hidden information and hidden action. Raven and Rubin themselves describe the relation between incentive systems and reward and coercive power: “examples of reward include: . . . a supervisor offering higher pay for increased productivity” (Raven and Rubin, 1976, p. 408). Efficient incentive compensation systems give a share in the returns to the agent. Then, principal and agent are interested in high returns. Low returns reduce the income of the agent. By implementing incentive systems the principal mediates rewards and punishment to the agent. Since incentive compensation systems, self-selection, and screening are behaviour settings, that the principal designs, she exerts environmental manipulation. The following proposition captures that situation. Proposition 2c. In the hidden information-situation the principal possesses quantitatively more bases of power than the agent . 3.2.4. The hidden action-situation There is a nexus between hidden action on one side and reward power and coercive power on the other side. Spremann describes the hidden action problem of the principal as follows: “the principal’s welfare depends on the agent’s effort, but additionally also on some exogenous risk (state of nature). Although the principal knows the probability distribution of that risk, she might be unable to come to know which state nature was actually realizing. Consequently, she is unable to separate low effort from bad luck. If results turned out to be poor, the principal cannot conclude that the agent’s effort must have been low. So it is the environmental uncertainty that explains why the principal is unable to deduce the agent’s effort from the resulting team output” (Spremann, 1987, p. 10). However, hidden action may also be an alternative which the agent uses to sanction the principal. Coercive power is based on P’s perception that O has the ability to mediate punishments for him. The agent may use the resources which the principal forwarded to him to fulfil the delegated task in a way which punishes the principal. The agent may also use these resources to mediate rewards for the principal. Agency theory suggests incentive compensation systems, bonding, and monitoring systems as the principal’s solution mechanisms for problems of hidden action. There is a nexus between monitoring systems on one side and informational, reward and coercive power on the other side. Monitoring systems are typical solutions to hidden action. The principal systematically collects information on the agent’s activities. The agent is more likely to behave in the interest of the principal if the principal obtains information on the agent’s behaviour. However, pure information systems are often not efficient: the principal tries to get information which the agent already has. Therefore, although the principal may have some success in this she will often not be able to exert power on the base of informational power. If monitoring systems are complemented by control systems or incentive compensation systems the principal will additionally use her reward and coercive power. There is a nexus between bonding on one side and legitimate power, reward
834
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
power and coercive power on the other side. Bonding provides the principal with the bases of power legitimacy, reward, and coercion. Bonding reduces hidden action. The agent binds himself to certain actions. For example, he produces information on his work performance, and he agrees upon sanctions if he should not produce them. As stated above, the introduction of incentive compensation systems allows the principal to exert reward and coercive power. Since incentive compensation systems, bonding, and monitoring systems are behaviour settings that the principal designs she exerts environmental manipulation. The following proposition captures that situation. Proposition 2d. In the hidden action-situation the principal possesses quantitatively more bases of power than the agent , and she possesses those power bases with the broader range . 3.2.5. Summary Under certain circumstances, a principal may not use her bases of power. She may also be incompetent. Also, the agent may build up a reputation for being stubborn or may pretend to possess certain bases of power. However, relying on French and Raven who define social influence as “a change in the belief, attitude or behaviour of a person—the target of influence, which results from the action, or presence, of another person or group of persons—the influencing agent”, and social power as “ potential for such influence” (Raven, 1992, p. 218), we summarize that the principal has more potential for such influence than the agent. Not only does the principal possess quantitatively more power bases than the agent, she also often possesses those power bases with the broader range. The next proposition summarizes Propositions 2a–2d. Proposition 2. In all situations in which an agency problem exists the principal possesses quantitatively more bases of power than the agent. In the hidden intentions and hidden actionsituations , she additionally possesses those power bases with the broader range .
In other words, agency theory makes implicit assumptions on the power relation between principal and agent. Agency theory assumes implicitly an asymmetry in power in favour of the principal. 4. Novel insights: new solution mechanism
During the discussion of the hidden intentions-situation an essential element of a solution mechanism which agency theory had not yet considered was found: a principal could find a way to establish positive referent power. The example of Raven (1965, p. 377) shows that the principal may prevent hidden intentions by exerting referent power. If the agent identifies positively with the principal he will behave like the referent. That is, the agent will behave non-opportunistically like the principal. Then, not only the problem of hidden intentions, but all agency problems are resolved. Referent power has the broadest range among all bases of power. The more attractive the principal, the more far reaching will her referent power be. A new solution mechanism should be elaborated which is based on the principal’s referent power. Therefore, the establishment of identification systems should be a promising solution mechanism to agency problems. Identification means the take over, the internalisation of real or imagined characteristics of an object. For example, a daughter may take over the attitudes and behaviour of her mother (she may “identify with her”). The term has spread from psychoanalysis to the social sciences and has a similarity to the concept of imitation in learning theory. Imitation signifies a behaviour of a person A that may be shown suddenly after it has been observed from another person B without having ever been shown before by person A. The range of imitation depends on
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
835
the characteristics of the model which is imitated. Such characteristics may be the social power of the model, the relation of the imitating person to the model or the observation that the model is rewarded or sanctioned for his or her behaviour. The central mechanism of identification is that a dyad becomes emotionalised. The emotionalisation may be explained by social learning theory (Bandura, 1977) or by psychoanalytic theories. As Simon (1957, p. 205) states, a rational agent may identify with an organization’s mission or goals, with its leaders or even with the organization itself as an entity distinct from its purpose. Referring to Simon, Mintzberg (1983, pp. 155–161) has introduced four types of identification: natural, selected, evoked, and calculated. These types refer to the causes of identification: natural attraction (“when the new member gets attracted to the ideology of the organization”), the result of a selection process (when the recruiting considers that the new member’s values match with the ideology of the organization), specific organizational attempts to evoke it (when organizations use social mechanisms like indoctrination or socialization to evoke identification), and the calculated cultivation of it by the individual (when a person dislikes these forms of identification and remains firmly committed to his own goals; however, this person may find that his self-interest can best be served by an identification—albeit a calculated and therefore fragile one). Calculated identification is a type of identification which is especially suited for rational egotists. There is a continuum from weakest to strongest among these four types of identification that integrate individual and organizational goals (Mintzberg, 1983, p. 160f). Natural identification is the strongest of these means, followed by selected, evocated, and calculated identification. Mintzberg does not consider the case of negative identification which has been introduced by Collins and Raven (1969, p. 175). In this case, the influencing agent has power over the target of influence because she is a negative model to him. The target of influence does not want to behave like the model. Natural negative identification is the strongest type of negative identification, followed by selected negative, evoked negative and calculated negative identification. Using referent power is therefore ambiguous. Learning on the model may evoke positive or negative identification. I define identification systems as systems of means which the principal implements on the base of her positive referent power. They make the principal attractive vis- a` -vis the agent. They are efficient without support from other solution mechanisms. The effect of identification systems is that the agent learns from the model of the principal not to behave in an opportunistic way. Identification systems are not absolutely new for agency theory. However, the examples that have been given have not yet been categorized. For example, Furubotn and Richter (1991) include concepts like the organizational culture or weak solidarity into the new institutional economics. They give a contractual interpretation of organizational culture. Hermalin (1998) has presented an economic theory of leadership which is based on leading by example. Witt (1998) discusses leadership in the framework of the new institutional economics and evolutionary economics as an activity which generates imagination and inspiration among the employees. The idea of identification systems is discussed as an alternative to incentive systems in the debate on intrinsic and extrinsic motivation (Deci, 1997). A strong organizational culture (see e.g. Deal and Kennedy, 1982; Schein, 1985) is an example of an identification system, as well as a strong corporate identity. There are some specific problems in the design of identification systems. French and Raven (1959) have pointed out that referent power has its origin in the target of influence. The target decides whether to identify with the influencing agent or not. Identification systems like organizational cultures cannot be designed mechanically. Schein (1985, pp. 122, 135) has pointed out that two organizations may have the same structure but different cultures. Culture cannot be derived from structure. These statements leave many degrees of freedom to the principal. It remains open which system of means makes the principal attractive to the agent. A principal may
836
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
already become attractive if she does not behave in an opportunistic way. However, this is already assumed by agency theory. There is no general answer to the question which values evoke the identification of the agent. Identification systems should be strong to be perceived. Only if they are perceived they can evoke identification. I submit the following proposition to address this. Proposition 3. In all situations in which an agency problem exists the principal possesses referent power. Identification systems are systems of means which the principal implements on the base of her positive referent power. Identification systems are a new solution mechanism to all agency problems.
From this discussion, the following hypotheses derive which can be empirically verified in further studies: Hypothesis 1. If the principal implements identification systems the agent will with higher probability act in the interest of the principal. Hypothesis 2. Goal conflicts between principal and agent correlate negatively with strong and positively with weak identification systems. Hypothesis 3. The duration of agency relations correlates positively with strong and negatively with weak identification systems. 5. Conclusion
The concern of agency theory is that the welfare of the principal may not be maximized because of the different goals and risk preferences of principal and agent. Therefore, agency theory has classified typical problems in agency relations and suggestions for their solution from the point of view of the principal. A social psychological perspective on agency relations and solution mechanisms reveals that each agency problem (each solution of an agency problem) becomes a problem of social power to the principal (agent) if she (he) changes her (his) beliefs, attitudes, or behaviour as a result of the action, or presence of the agent (principal). Agency theory makes implicit assumptions on the power relation between principal and agent. An analysis of one agency problem after the other revealed that in all situations in which an agency problem exists the principal possesses quantitatively more bases of power than the agent. In the hidden intentions and hidden action-situations, she additionally possesses those power bases with the broader range. Agency theory assumes implicitly an asymmetry in power in favour of the principal. The social power perspective paves the way to a new solution mechanism that agency theory had not yet classified. In all situations in which an agency problem exists the principal possesses referent power. Identification systems are systems of means which the principal implements on the base of her positive referent power. Identification systems are a new solution mechanism. They are not restricted to certain agency problems; instead they are suited for solving all agency problems. 5.1. Considerations about empirical testing
The thesis of this article still lacks empirical grounding. Until now, I have presented some cases and real life examples in order to illustrate the argument, as well as some new Hypotheses 1–3 that have been derived and may be empirically verified in further studies. I conclude with some
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
837
suggestions that indicate how this can be done. My suggestion is to choose the duration of organizational membership in firms as an indicator for the duration of agency relations and the quality (strong or weak) of each firm’s organizational culture as the indicator for identification systems. In a data set of varying quality of the firms’ organizational cultures, organizational memberships should ceteris paribus last longer in organizations that have created a strong organizational culture. 5.2. Application to politics
The proposed three propositions approach does not only concern an objective definition of power that is useful for organization theory. Identification systems defined as systems of means which the principal implements on the base of her positive referent power can be used as definition of ideology. Then, the proposed three propositions concern ideology and its use on the microand macro-level of the society. It could turn out that the three propositions are more meaningful when applied to politics than when applied to organizations. However, as agency theory is widely applied in economics and economists tend to give power a wide berth the three propositions may nevertheless be more innovative in agency theory as applied by economists and to organizations. 5.3. What are the limitations of this perspective?
One has to keep in mind that the exertion of power has to be differentiated from the possession of bases of power. In his power-interaction model, Raven (1992) describes the process of transforming the bases of power into the exertion of power. Based on her motivation to influence, the influencing agent assesses which bases of power are available to her, especially in relation to the target of influence. If the principal makes a wrong assessment the agent may turn out to be more powerful. The proposition on the asymmetry in power in favour of the principal was derived for the dyad of principal and agent. The dyad is the smallest type of social interaction. However, in organizations and in many other social interaction systems in which agency theory is applied principals and agents interact in multiple principal-agent structures. Multiple principal-agent settings will not terminate the efficacy of identification systems. However, they may lead to another distribution of social power between principals and agents than in the dyad. An analysis of the asymmetry in power in more complex social interactions is a theme for further studies. 5.4. What further research questions are raised by this perspective?
Critics of agency theory have argued that there are a number of behavioural assumptions which agency theory completely ignores. For example, Perrow (1981, 1986) has criticized that agency theory ignores power. This paper may be a first step in the development of an extended agency theory that includes an explicit concept of power. Pillars of this extended agency theory could be: agency relations are characterized by informational asymmetry, asymmetry in power, outcome uncertainty, and risk. In the normal case, the solution of the agency problem is a power phenomenon namely if either principal or agent changes his or her belief, attitude or behaviour as a consequence of the action, or presence, of the other. The agency problems of the principal are based on the bases of power of the agent. The solution mechanisms of the principal are based on the bases of power of the principal. Power comes across as an intervening variable.
838
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
References Alchian, A.A., Woodward, S., 1988. The firm is dead–long live the firm: a review of Oliver E. Williamson’s The economic institutions of capitalism. Journal of Economic Literature 26, 65–79. Arrow, K., 1985. The economics of agency. In: Pratt, J., Zeckhauser, R. (Eds.), Principals and Agents: The Structure of Business. Harvard Business Scholl Press, Boston, pp. 37–51. Arrow, K., 1986. Agency and the market. In: Arrow, K.J., Intriligator, M.D. (Eds.), Handbook of Mathematical Economics, Bd. III. North-Holland, Amsterdam, pp. 1183–1195. Bandura, A., 1977. Social Learning Theory. Prentice-Hall, Englewood Cliffs, NJ. Bemmels, B., Lau, D.C., 2001. Local union leaders’ satisfaction with grievance procedures. Journal of Labor Research 22, 653–667. Bento, A.M., White, L.F., 2001. Organizational form, performance and information costs in small businesses. Journal of Applied Business Research 17, 41–61. Beyer, J., Hassel, A., 2002. The effects of convergence: internationalization and the changing distribution of net value added in large German firms. Economy and Society 31, 309–332. Breton, A., 1995. Organizational hierarchies and bureaucracies: an integrative essay. European Journal of Political Economy 11, 411–440. Cartwright, D., 1965. Influence, leadership, control. In: March, J. (Ed.), Handbook of Organization. Rand McNally, Chicago, pp. 1–47. Chen, J., 2001. Ownership structure as corporate governance mechanism: evidence from Chinese listed companies. Economics of Planning 34, 53–72. Collins, B.E., Raven, B.H., 1969. Group structure: attraction, coalitions, communication, and power. In: Lindzey, G., Aronson, E. (Eds.), The Handbook of Social Psychology, vol. 4. Addison-Wesley, Reading, MA, pp. 102– 204. Deal, T.E., Kennedy, A.A., 1982. Corporate cultures. In: The Rites and Rituales of Corporate Life. Addison-Wesley, Reading, MA. Deci, E.L., 1997. Why We Do What We Do: Understanding Self-motivation. Penguin Books, New York. Eisenhardt, K., 1989. Agency theory. An assessment and review. Academy of Management Review 14, 57–74. Eisner, M.A., Worsham, J., Ringquist, E., 1996. Crossing the organizational void: the limits of agency theory in the analysis of regulatory control. Governance 9, 407–428. Feaver, P.D., 1998. Crisis a shirking: an agency theory explanation of the souring of American civil–military relations. Armed Forces and Society 24, 407–434. Franchino, F., 2000. Control of the [EU] Commission’s executive functions: uncertainty, conflict and decision rules. European Union Politics 1, 63–92. French Jr., J.R.P., Raven, B.H., 1959. The bases of social power. In: Cartwright, D. (Ed.), Studies in Social Power. University of Michigan, Ann Arbor, pp. 150–167. Furubotn, E.G., Richter, R. (Eds.), 1991. The New Institutional Economics. Mohr, T u¨ bingen. Goldberg, V.P., 1976. Regulation and administered contracts. Bell Journal of Economics and Management Science 7, 439–441. Grinblatt, M., Hwang, C.Y., 1989. Signalling and the price of new issues. Journal of Finance 44, 393–420. Guston, D.H., 1999. Stabilizing the boundary between US politics and science: the role of the office of technology transfer as a boundary organization. Social Studies of Science 29, 87–111. Hermalin, B.E., 1998. Toward an economic theory of leadership. Leading by example. American Economic Review 88, 1188–1206. Hermalin, B.E., Wallace, N.E., 2001. Firm performance and executive compensation in the savings and loan industry. Journal of Financial Economics 61, 139–170. Hubbard, T.N., 2000. The demand for monitoring technologies: the case of trucking. Quarterly Journal of Economics 115, 533–560. Jensen, M., 2000. A Theory of the Firm: Governance, Residual Claims, and Organizational Forms. Harvard University Press, Cambridge. Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs, and ownership structure. Journal of Financial Economics 3, 305–360. Kiser, E., 1999. Comparing varieties of agency theory in economics, political science, and sociology: an illustration from state policy implementation. Sociological Theory 17, 146–170. Klein, B., Crawford, R.G., Alchian, A.A., 1978. Vertical integration, appropriable rents, and the competitive contracting process. Journal of Law and Economics 22, 297–326.
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
839
Krishnaswami, S., Pottier, S., 2001. Agency theory and participating policy usage evidence from stock life insurers. Journal of Risk and Insurance 68, 659–683. Laffont, J.-J., 2000. Incentives and Political Economy. Oxford University Press, Oxford. Lambert, R.A., 2001. Contracting theory and accounting. Journal of Accounting and Economics 32, 3–87. Longley, N., 1998. Legislative systems with absolute party discipline: implications for the agency theory approach to the constituent-legislator link. Public Choice 97, 121–140. Longley, N., 1999. Voting on abortion in the [Canadian] house of commons: a test for legislator shirking. Canadian Public Policy 25, 503–521. Lord, M.D., 2000. Constituency-based lobbying as corporate political strategy [in the US]: testing an agency theory perspective. Business and Politics 2, 289–308. Majone, G., 2000. Nonmajoritarian institutions and the limits of democratic governance: a political transaction-cost approach. Journal of Institutional and Theoretical Economics 157, 57–78. Martens, B., et al., 2002. The Institutional Economics of Foreign Aid. Cambridge University Press, Cambridge. Milgrom, P., Roberts, J., 1992. Economics, Organization and Management. Prentice-Hall, Englewood Cliffs, NJ. Miller, J.L., 2002. The board as a monitor of organizational activity: the applicability of agency theory to nonprofit boards. Nonprofit Management and Leadership 12, 429–450. Miller, G.J., Whitford, A.B., 2002. Trust and incentives in principal-agent negotiations: the “insurance/incentive trade-off”. Journal of Theoretical Politics 14, 231–267. Mintzberg, H., 1983. Power in and Around Organizations. Prentice-Hall, Englewood Cliffs, NJ. Myerson, R., 1982. Optimal coordination mechanisms in generalized principal-agent problems. Journal of Mathematical Economics 10, 67–81. O’Sullivan, N., 2000. The determinants of non-executive representation on the boards of large UK companies. Journal of Management and Governance 4, 283–297. Ott, U.F., 2000. International joint ventures: a common agency problem. Global Business and Economics Review 2, 67–84. Perrow, C., 1981. Markets, hierarchies, and hegemony: a critique of Chandler and Williamson. In: Van de Ven, A., Joyce, J. (Eds.), Perspectives in Organisation Design and Behaviour. Wiley, New York, pp. 371–386. Perrow, C., 1986. Complex Organizations. A Critical Essay, 3rd ed. Random House, New York. Podsakoff, Ph.M., Schriesheim, Ch.A., 1985. Field studies of French and Raven’s bases of power: critique, reanalysis, and suggestions for future research. Psychological Bulletin 97, 387–411. Pratt, J.W., Zeckhauser, R. (Eds.), 1985. Principals and Agents. The Structure of Business. Harvard Business School Press, Boston, MA. Rasmusen, E., 1990. Games and Information. An Introduction to Game Theory. Blackwell, Oxford. Raven, B.H., 1965. Social influence and power. In: Stein, Ivan, D., Martin, Fishbein (Eds.), Current Studies in Social Psychology. Holt, Rinehart and Winston Inc., New York, pp. 371–382. Raven, B.H., 1974. The comparative analysis of power and power preference. In: Tedeschi, James, T. (Eds.), Perspectives on Social Power. Aldine Publishing Co., Chicago, pp. 172–198. Raven, B.H., 1992. A power/interaction model of interpersonal influence: French and Raven 30 years later. Journal of Social Behavior and Personality 7, 217–244. Raven, B.H., Kruglanski, A.W., 1970. Conflict and power. In: Swingle, P. (Ed.), The Structure of Conflict. Academic Press, New York, pp. 69–109. Raven, B.H., Rubin, J.Z., 1976. Social Psychology: People in Groups. Wiley, New York. Rees, R., 1985. The theory of principal and agent, Part I. Bulletin of Economic Research 37, 3–26 and 75–95. Riley, J., 1979. Informational equilibrium. Econometrica 47, 331–359. Saam, N.J., 2002. Prinzipale, Agenten und Macht. Mohr Siebeck, T u¨ bingen. Sapienza, H.J., et al., 2000. Effects of agency risks and procedural justice on board processes in venture capital-backed firms. Entrepreneurship and Regional Development 12, 331–351. Schein, E.H., 1985. Organizational Culture and Leadership. A Dynamic View. Jossey-Bass, San Francisco. Senter Jr., R., Flynn, M.S., 1999. Changing interorganizational patterns in the North American automotive supply chain. Applied Behavioral Science Review 7, 59–80. Short, H., Keasey, K., Duxbury, D., 2002. Capital structure, management ownership and large external shareholders: a UK analysis. International Journal of the Economics of Business 9, 375–399. Simon, H.A., 1957. Administrative Behavior, 2nd ed. Macmillan, New York. Spence, M., 1973. Job market signaling. Quarterly Journal of Economics 87, 355–374. Spence, A.M., 1974. Market Signaling: Informational Transfer in Hiring and Related Screening Processes. Harvard University Press, Cambridge, MA.
840
N.J. Saam / The Journal of Socio-Economics 36 (2007) 825–840
Spremann, K., 1987. Agent and principal. In: Spremann, K., Bamberg, G. (Eds.), Agency Theory, Information, and Incentives. Springer, Berlin, pp. 3–37, Corrected reprint, 1989. Spremann, K., 1990. Asymmetrische Information. Zeitschrift f ¨ur Betriebswirtschaft 60, 561–586. Strom, K., 2000. Delegation and accountability in parliamentary democracies. European Journal of Political Research 37, 261–289. Thibaut, J.W., Kelley, H.H., 1959. The Social Psychology of Groups. Wiley, New York. Weber, M., 1947. The Theory of Social and Economic Organization. Oxford University Press, Oxford. White, H., 1985. Agency as control. In: Pratt, J., Zeckhauser, R. (Eds.), Principals and Agents: The Structure of Business. Harvard Business Scholl Press, Boston, pp. 187–214. Witt, U., 1998. Imagination and leadership. The neglected dimension of an evolutionary theory of the firm. Journal of Economic Behavior & Organization 35, 161–177. Worsham, J., Eisner, M.A., Ringquist, E.J., 1997. Assessing the assumptions. A critical analysis of agency theory. Administration and Society 28, 419–440. Wright, M., et al., 2001a. Finance and management buyouts: agency versus entrepreneurship perspectives. Venture Capital 3, 239–261. Wright, P., Mukherji, A., Kroll, M.J., 2001b.A reexamination of agency theory assumptions: extensionsand extrapolations. Journal of Socio-Economics 30, 413–429. Zey, M., Camp, B., 1996. The transformation from multidimensional form to corporate groups of subsidies in the 1980s: capital crisis theory. The Sociological Quarterly 37, 327–351. Zey, M., Swenson, T., 1999. The transformation of the dominant corporate form from multidivisional to multisubsidiary: the role of the 1986 Tax Reform Act. The Sociological Quarterly 40, 241–267.
