Saam 2007.pdf

The Journal of Socio-Economics 36 (2007) 825–840

Asymmetry in information versus asymmetry in power:Implicit assumptions of agency theory?

Nicole J. Saam ∗Universitat Erfurt, Staatswissenschaftliche Fakultat, Sozialwissenschaften, Nordhauser Str. 63,

99089 Erfurt, Germany

Abstract

The hypothesis of this article is that agency theory makes implicit assumptions on the power relationbetween principal and agent. It implicitly assumes an asymmetry in power in favour of the principal. Thebases 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. Apower/interaction model of interpersonal influence: French and Raven 30 years later. Journal of Social Behav-ior and Personality 7, 217–244] of social psychology is used to prove this argument. Surprisingly, the socialpower perspective does not only reveal this implicit assumption. It also paves the way for a new mechanismthat resolves agency problems that had not yet been discovered by agency theory: identification systems.© 2007 Elsevier Inc. All rights reserved.

JEL classification: B52; D23; D70; D80

Keywords: Agency theory; Social power; Identification systems

1. Introduction

Together with transaction cost economics and the property rights approach, agency theoryforms the new institutional economics (Furubotn and Richter, 1991). In the centre of analysis arethose institutions in which economic exchange takes place, like markets, organizations, and legalnorms. The aim of institutional analysis is to explain the structure of institutions, their influenceon individual action, their efficiency, as well as the change of institutions. Although it is mainlyspread in economics and looked at as an economic theory, its explanatory power advances to theclassical domains of social psychology and sociology: the new institutional economics definesitself 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].

1053-5357/$ – see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.socec.2007.01.018

mailto:[email protected]

dx.doi.org/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 agencyrelations have been of continuous interest: the relation between employee and employer, betweenshareholder and manager, and between creditor and stockholder. In recent years, agency hasbeen applied in analysing research fields as diverse as the ownership structure of the firm (Shortet al., 2002; Chen, 2001; Wright et al., 2001a,b; Jensen, 2000; O’Sullivan, 2000), incentiveproblems in accounting, foreign aid, or between shareholder and policyholder (Martens et al.,2002; Krishnaswami and Pottier, 2001; Lambert, 2001; Hubbard, 2000; Laffont, 2000), execu-tive compensation (Hermalin and Wallace, 2001), organizational forms and the transformationof corporate forms (Bento and White, 2001; Zey and Swenson, 1999; Zey and Camp, 1996),corporate governance (Beyer and Hassel, 2002), changing interorganizational patterns in sup-ply 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 pol-icy implementation (Kiser, 1999), nonmajoritarian institutions (Majone, 2000), the relationshipbetween constituents and legislators (Lord, 2000; Strom, 2000; Longley, 1999, 1998), controlof the executive (Franchino, 2000), civil–military relations (Feaver, 1998), and boundary orga-nizations between politics and science (Guston, 1999). Critical contributions to agency theoryas 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 assumptionsand 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 andpublic choice and put forward the hypothesis that their integration could generate a general the-ory of bureaucratic organizations. In sum, agency theory has become an inspiring and integratingtheory perspective to many at first sight very different fields of research.

My hypothesis is that agency theory makes implicit assumptions on the power relation betweenprincipal and agent. Especially, I will prove that agency theory implicitly assumes an asymmetryin 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 socialpower perspective does not only reveal this implicit assumption. It also paves the way for a newmechanism 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 classicalwriters: Jensen and Meckling define an agency relationship “as a contract, under which one ormore persons (the principal(s)) engage another person (the agent), to perform some service ontheir 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 betweenprincipal and agent, and “delegated choice” (Rees, 1985, p. 3). Its formal structure is applicable toan 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 actionaffects the welfare of both the agent and another person, the principal. The principal, . . ., has theadditional function of prescribing payoff rules.”

There are three main differences respectively asymmetries in the relation between principaland agent. Agency theory explains the realization of the division of labour between both actors bythese 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) tofulfil a task which the principal does not have (Pratt and Zeckhauser, 1985, p. 3); or both have thecompetences, but the potential agent can fulfil the task at lower cost. Informational asymmetriesarise 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 agentdepending on her effort. Additionally, the principal needs information on environmental states orprocesses which have influenced the agent’s work performance. Therefore, in the standard scenarioof agency theory, asymmetrical information in favour of the agent is assumed. (2) Different riskpreferences. Principal and agent have different attitudes towards risks which are related to differentcompensation schemes (outcome-based versus behaviour-based). In the standard scenario, agentsare assumed to be risk averse. In comparison to the principal the agent’s income is low. Reductionsin 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 tomaximize their individual utility. The agent wants to maximize his income, whereas the principalwants to maximize her returns. Both income and returns are based on the agent’s effort andexogenous random elements. Agency theory assumes that greater effort by the agent results ingreater outcome. However, greater effort is also related to a greater disutility on the side of theagent. As a consequence, the agent wants to maximize his income while minimizing his disutilityof 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 characterizethe optimal forms of agency contracts the agency problems have to be solved. Agency theory solvesthem 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 areproblems 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 thepractice and theory of insurance and has reconceptualised them in more general form as hiddenaction and hidden information. Meanwhile hidden information has differentiated into hiddencharacteristics, hidden intentions, and hidden knowledge (Spremann, 1990). However, in theagency literature, there is not yet a uniform use of the concepts. The normative agency literatureoften sticks to the concepts of moral hazard and adverse selection, whereas positive agency theoryprefers the reconceptualised concepts.

(1) Hidden characteristics. Before contracting the principal has incomplete information on theagent’s qualification, like her skills or abilities. The agent may claim to have certain skillswhen he is hired. A problem arises because the principal cannot completely verify these skillsor 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 holdup (Goldberg, 1976). The principal cannot simply resign from the contract relation becauseshe 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 knowledgeon exogenous facts like environmental states and processes which are relevant to evaluatehis work’s outcome. This knowledge is part of the agent’s professional or process knowledgewhich he may use opportunistically. This is Arrow’s (1985) hidden information definition.


(4) Hidden action. After contracting the agent can choose among different actions. The agentmay shirk. He may choose to work less but to pretend to work hard which cannot easily beevaluated 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 moreimportant are incentives. Incentive compensation systems coalign the preferences of the agentwith those of the principal because the rewards for both depend on the same actions. Efficientincentive compensation systems give a share in the returns to the agent. Both are then interestedin high returns. The implementation of incentive compensation systems generates costs tothe principal and risks to the agent. Therefore, a trade off between agency costs and increasein 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. Theagent is more likely to behave in the interests of the principal if the principal has informa-tion to verify the agent’s behaviour (Eisenhardt, 1989, p. 60). Again, the implementation ofmonitoring systems such as budgeting systems, reporting procedures, boards of directors,and additional layers of management generates costs to the principal. Perfect monitoringsystems are unrealistic. Pure information systems are often not efficient. They have to becomplemented by control systems or incentive compensation systems. Monitoring systemsare typical solutions to hidden action.

(3) Vertical integration (Klein et al., 1978; Alchian and Woodward, 1988) establishes a relationof authority between the originally equal partners of principal and agent. Now, the agent hasto comply with the instructions of the principal. Compliance is fostered by sanctions. Verticalintegration is prevalent in labour contracts. Again, the implementation of vertical integrationgenerates costs to the principal. Vertical integration does not function as an incentive. Agencytheorists often neglect hierarchy and norms as solutions to agency problems. Vertical inte-gration is often only interpreted in a monetary way. Then, it becomes difficult to differentiatevertical integration from incentive systems. For example, promises concerning pensions andbonuses are granted only after the agent has been loyal to the principal for many years. Verticalintegration is the typical solution to hidden intentions.

(4) Game theoretical solutions. Agency problems can be represented as games like in gametheory, and vice versa (Rasmusen, 1990). Agency relations are considered as “hybrid betweena cooperative and non-cooperative game” (Myerson, 1982, p. 68). In situations in whichvertical integration is not effective game theoretical solutions may help. For example, anagency problem may be modelled as a prisoner’s dilemma. Then, one can alter the situationon 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 canbe increased. Sanctions can be imposed on one-sided defective actions. Game theoreticalsolutions are typical solutions to hidden intention.

(5) Self-selection (Arrow, 1986). The principal can offer different contracts to the agent, e.g. anoutcome-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


willingness to work hard. Again, the implementation of self-selection generates costs to theprincipal. 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 char-acteristics in order to convince the principal that he is willing to work hard. In contrast tothe other solution mechanisms, the agent pays the cost of signalling. For example, the choiceof 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 tocertain actions, e.g. he produces information on his work performance, and he agrees uponsanctions if he should not produce them. The income of the agent depends at least partiallyon his performance. Again, bonding generates costs to the principal because to be efficientbonding has to be controlled.

(8) Screening reduces hidden characteristics as well as hidden information. The principalimproves her selection procedures like tests, or assessment centres (Spence, 1974) whichgenerates 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 whichthe principal intends to change the behaviour of the agent if the latter should be willing to behavein an opportunistic way that harms the principal. The potential for changing an other person’sbehaviour is captured by the social psychological definition of power. Raven defines influenceas “a change in the belief, attitude or behaviour of a person – the target of influence, whichresults from the action, or presence, of another person or group of persons – the influencingagent” (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 ofthe influencing agent is irrelevant. The objective effects on the side of the target of influence areconsidered. 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 itremains unclear how to verify that an agent has realized his or her will. The price of objectivityis the renunciation of the influencing agent’s will. As a consequence, Raven’s actors exert powereven if they do not have the will to do it. For example, someone will have exerted power if he hastold his new fellow at the bus stop that all busses are late and this person decides not to wait atthe bus stop, but to walk by foot. The exertion of power thus becomes a phenomenon of everydaylife and its negative connotation is removed.

French and Raven (1959) have developed a bases of power concept that is widely acceptedin 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 andRaven, 1969; Raven and Kruglanski, 1970; Raven and Rubin, 1976). This concept has been appliedto 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, andinteraction among politicians.

Since 1965, French and Raven have differentiated six bases of power, “resources which aninfluencing 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 con-ceptualisation presumes a dyad (French and Raven, 1959, p. 150). The interacting agents are


called influencing agent O (other) and target of influence P (person). They are assumed to berational (Raven, 1992, p. 228). Reward power is based on P’s perception that O has the abilityto mediate rewards for him. Coercive power is based on P’s perception that O has the ability tomediate punishments for him. Legitimate power is based on the perception of P that O has a legit-imate 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 (Frenchand Raven, 1959, p. 155f). Informational power is based on an information which O forwards toP. 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 includesall actions which the influencing agent can evoke on the side of the target of influence throughher usage of power:

1. Reward power and coercive power are limited to situations in which the influencing agent canreward or punish.

2. Legitimate power has a broader range because it may be exerted in all situations in which thevalue system is valuable. Furthermore, within the scope of legitimate power the influencingagent is allowed to reward or sanction.

3. Expert power and informational power are limited to specific situations in which expertise orinformation 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 orinformational power. Then, an expert may exert referent power.

4. Referent power depends on attractivity of the influencing agent. It may be far reaching becauseit 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 ininfluencing 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 whodefined that “when O influences P by ecological control, he takes some action which modifies P’ssocial or physical environment, on the assumption that the new environment will subsequentlybring about the desired change in P” (Cartwright, 1965, p. 19f). This power base is most importantfor executives. “Since much of an executive’s responsibility is to design behaviour settings, he isbound 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 exertspower if she changes the belief, attitude or behaviour of the agent—e.g. by implementing amonitoring system; the agent exerts power if he changes the belief, attitude or behaviour of theprincipal—e.g. by giving some information to the principal.

This raises the question whether all agency problems are power phenomena (Saam, 2002). Areagency problems problems of power, or are they informational problems as the traditional viewstates? Assume a principal who does not relate her decisions to information of the agent. Then,


hidden information is not a problem of power, but only one of information. The agent possessesthe base of power information, but he cannot exert informational power. The same holds for thesolution 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 ofpower to do it. The next two propositions capture these situations.

Proposition 1a. Each agency problem becomes a problem of power to the principal if shechanges 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 theagent changes his beliefs, attitudes, or behaviour as a result of the implementation of a solutionmechanism 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 andagent? Especially, does agency theory implicitly assume an asymmetry in power in favour of theprincipal? In the following, I analyse one agency problem after the other and I show which basesof power are available to principal and agent.

3.2.1. The hidden characteristics-situationThere 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 getsinto contact with many individuals, potential agents, on the other side. The principal has tomake 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 ofcharacteristics, she usually will be uncertain about the particular type of agent. How to makean 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 nocontract or a contract which the agent dislikes. Then, the agent would like to change the prin-cipal’s behaviour. However, at this point of time, the only resource of the agent is informationon himself. Signalling is the solution mechanism of the agent that is based on informationalpower.

In the hidden characteristics-situation, the principal mediates reward or punishment to theagent, 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 forproblems of hidden characteristics. There is a nexus between self-selection and screening on oneside 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 ofthe agent. She designs behaviour settings in which the agent may not only give information whichwas hidden before, but also may change the agent’s behaviour. The following proposition capturesthat situation.

Proposition 2a. In the hidden characteristics-situation the principal possesses quantitativelymore bases of power than the agent.


3.2.2. The hidden intentions-situationThere 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 realizehidden intentions. He may act intentionally unfair with the principal because there is always somefreedom in interpreting the contract which cannot specify any detail of action and interaction. Byacting fair or unfair the agent can mediate reward or punishment to the principal. Raven himselfrelated hidden intention to referent power—although he did not know agency theory: “The termmanipulation is generally used to refer to social influence wherein the intentions of the agent arehidden from the influencee” (Raven, 1965, p. 377). He gives the example of a political campaignworker who “deliberately makes himself extremely obnoxious at a bar, then leaves, advisingothers to ‘Vote for Blotz’, the opposing candidate”. Raven states that “when it is discovered bythe influencee, it is likely to establish the manipulator as a person with very strong negativeexpert and negative referent power” (Raven, 1965, p. 377). The example does not represent anagency relation. Imagine a project manager who intentionally sets deadlines which his engineerscannot hold. After the engineers have discovered this, e.g. in a talk with the secretary, he willhave established negative referent power. Both examples suggest that the principal may solve thehidden intentions problem by establishing positive referent power. If the agent identifies positivelywith 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 notyet considered.

Agency theory suggests vertical integration and game theoretical solutions as the principal’ssolution mechanisms for problems of hidden intentions. There is a nexus between vertical inte-gration 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 authoritybetween the originally equal partners of principal and agent. Legitimate power is based on theperception of P that O has a legitimate right to prescribe behaviour for him. This is the agent’sperception in case of vertical integration. Within the scope of legitimate power the principal canreward or sanction the agent. Thus, vertical integration provides the principal with the bases ofpower legitimacy, reward, and coercion. There is also a nexus between game theoretical solutionson one side and reward power and coercive power on the other side. Agency problems can berepresented as games. One can try to alter the situation on the base of the suggestions which gametheory offers to generate ongoing cooperation. Thibaut and Kelley (1959, p. 101) have interpretedthe ability of A to influence the returns of B as power of A over B. In the concepts of French andRaven, this is reward power and coercive power. Since vertical integration and game theoreticalsolutions 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 morebases of power than the agent, and she possesses those power bases with the broader range.

3.2.3. The hidden information-situationThere is a nexus between hidden information and informational power. “In the hidden infor-

mation problems, the agent has made some observation that the principal has not made. Theagent uses (and should use) this observation in making decisions; however, the principal cannotcheck whether the agent has used his or her information in the way that best serves the principal’sinterest.” (Arrow, 1985, p. 39). The agent exerts (informational) power if the principal changesher beliefs, attitudes, or behaviours as a result of the agent’s information. There is also a nexus


between hidden knowledge and expert power. Hidden knowledge is defined as the agent’s pro-fessional or process knowledge which he may use opportunistically. Analogous to the previouscase, the agent exerts expert power if the principal changes her beliefs, attitudes, or behavioursas a result of the agent’s knowledge. Agency theory suggests signalling as the agent’s solutionmechanism for problems of hidden information. The only resource of the agent is information onhimself. Signalling is based on informational power.

Agency theory suggests incentive compensation systems, self-selection, and screening as theprincipal’s solution mechanisms for problems of hidden information. There is a nexus betweenincentive compensation systems on one side and reward power and coercive power on the otherside. Incentive systems are mechanisms to solve hidden information and hidden action. Raven andRubin 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 thereturns to the agent. Then, principal and agent are interested in high returns. Low returns reducethe income of the agent. By implementing incentive systems the principal mediates rewards andpunishment to the agent. Since incentive compensation systems, self-selection, and screeningare behaviour settings, that the principal designs, she exerts environmental manipulation. Thefollowing proposition captures that situation.

Proposition 2c. In the hidden information-situation the principal possesses quantitatively morebases of power than the agent.

3.2.4. The hidden action-situationThere 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 exogenousrisk (state of nature). Although the principal knows the probability distribution of that risk, shemight be unable to come to know which state nature was actually realizing. Consequently, she isunable to separate low effort from bad luck. If results turned out to be poor, the principal cannotconclude that the agent’s effort must have been low. So it is the environmental uncertainty thatexplains 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 agentuses to sanction the principal. Coercive power is based on P’s perception that O has the abilityto mediate punishments for him. The agent may use the resources which the principal forwardedto him to fulfil the delegated task in a way which punishes the principal. The agent may also usethese resources to mediate rewards for the principal.

Agency theory suggests incentive compensation systems, bonding, and monitoring systemsas the principal’s solution mechanisms for problems of hidden action. There is a nexus betweenmonitoring 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 collectsinformation on the agent’s activities. The agent is more likely to behave in the interest of theprincipal if the principal obtains information on the agent’s behaviour. However, pure informationsystems 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 toexert power on the base of informational power. If monitoring systems are complemented bycontrol systems or incentive compensation systems the principal will additionally use her rewardand coercive power. There is a nexus between bonding on one side and legitimate power, reward


power and coercive power on the other side. Bonding provides the principal with the bases ofpower legitimacy, reward, and coercion. Bonding reduces hidden action. The agent binds himselfto certain actions. For example, he produces information on his work performance, and he agreesupon sanctions if he should not produce them. As stated above, the introduction of incentivecompensation systems allows the principal to exert reward and coercive power. Since incentivecompensation systems, bonding, and monitoring systems are behaviour settings that the principaldesigns she exerts environmental manipulation. The following proposition captures that situation.

Proposition 2d. In the hidden action-situation the principal possesses quantitatively more basesof power than the agent, and she possesses those power bases with the broader range.

3.2.5. SummaryUnder 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 topossess certain bases of power. However, relying on French and Raven who define social influenceas “a change in the belief, attitude or behaviour of a person—the target of influence, which resultsfrom the action, or presence, of another person or group of persons—the influencing agent”, andsocial power as “potential for such influence” (Raven, 1992, p. 218), we summarize that theprincipal has more potential for such influence than the agent. Not only does the principal possessquantitatively more power bases than the agent, she also often possesses those power bases withthe broader range. The next proposition summarizes Propositions 2a–2d.

Proposition 2. In all situations in which an agency problem exists the principal possessesquantitatively 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.

In other words, agency theory makes implicit assumptions on the power relation betweenprincipal and agent. Agency theory assumes implicitly an asymmetry in power in favour of theprincipal.

4. Novel insights: new solution mechanism

During the discussion of the hidden intentions-situation an essential element of a solutionmechanism which agency theory had not yet considered was found: a principal could find a wayto establish positive referent power. The example of Raven (1965, p. 377) shows that the principalmay prevent hidden intentions by exerting referent power. If the agent identifies positively withthe principal he will behave like the referent. That is, the agent will behave non-opportunisticallylike the principal. Then, not only the problem of hidden intentions, but all agency problems areresolved. Referent power has the broadest range among all bases of power. The more attractivethe principal, the more far reaching will her referent power be. A new solution mechanism shouldbe elaborated which is based on the principal’s referent power.

Therefore, the establishment of identification systems should be a promising solution mecha-nism to agency problems. Identification means the take over, the internalisation of real or imaginedcharacteristics of an object. For example, a daughter may take over the attitudes and behaviour ofher mother (she may “identify with her”). The term has spread from psychoanalysis to the socialsciences and has a similarity to the concept of imitation in learning theory. Imitation signifies abehaviour of a person A that may be shown suddenly after it has been observed from anotherperson B without having ever been shown before by person A. The range of imitation depends on


the characteristics of the model which is imitated. Such characteristics may be the social powerof the model, the relation of the imitating person to the model or the observation that the model isrewarded or sanctioned for his or her behaviour. The central mechanism of identification is thata 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 agentmay identify with an organization’s mission or goals, with its leaders or even with the organizationitself 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 typesrefer to the causes of identification: natural attraction (“when the new member gets attracted to theideology of the organization”), the result of a selection process (when the recruiting considers thatthe new member’s values match with the ideology of the organization), specific organizationalattempts to evoke it (when organizations use social mechanisms like indoctrination or socializa-tion to evoke identification), and the calculated cultivation of it by the individual (when a persondislikes 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 cal-culated and therefore fragile one). Calculated identification is a type of identification which isespecially suited for rational egotists. There is a continuum from weakest to strongest among thesefour 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, andcalculated identification. Mintzberg does not consider the case of negative identification whichhas been introduced by Collins and Raven (1969, p. 175). In this case, the influencing agent haspower over the target of influence because she is a negative model to him. The target of influencedoes not want to behave like the model. Natural negative identification is the strongest type ofnegative identification, followed by selected negative, evoked negative and calculated negativeidentification. Using referent power is therefore ambiguous. Learning on the model may evokepositive or negative identification.

I define identification systems as systems of means which the principal implements on the baseof her positive referent power. They make the principal attractive vis-a-vis the agent. They areefficient without support from other solution mechanisms. The effect of identification systemsis 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 havebeen given have not yet been categorized. For example, Furubotn and Richter (1991) includeconcepts like the organizational culture or weak solidarity into the new institutional economics.They give a contractual interpretation of organizational culture. Hermalin (1998) has presentedan economic theory of leadership which is based on leading by example. Witt (1998) discussesleadership in the framework of the new institutional economics and evolutionary economicsas an activity which generates imagination and inspiration among the employees. The idea ofidentification systems is discussed as an alternative to incentive systems in the debate on intrinsicand 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 targetdecides whether to identify with the influencing agent or not. Identification systems like orga-nizational cultures cannot be designed mechanically. Schein (1985, pp. 122, 135) has pointedout that two organizations may have the same structure but different cultures. Culture cannotbe derived from structure. These statements leave many degrees of freedom to the principal. Itremains open which system of means makes the principal attractive to the agent. A principal may


already become attractive if she does not behave in an opportunistic way. However, this is alreadyassumed by agency theory. There is no general answer to the question which values evoke theidentification of the agent. Identification systems should be strong to be perceived. Only if theyare 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 referentpower. Identification systems are systems of means which the principal implements on the baseof her positive referent power. Identification systems are a new solution mechanism to all agencyproblems.

From this discussion, the following hypotheses derive which can be empirically verified infurther studies:

Hypothesis 1. If the principal implements identification systems the agent will with higherprobability act in the interest of the principal.

Hypothesis 2. Goal conflicts between principal and agent correlate negatively with strong andpositively with weak identification systems.

Hypothesis 3. The duration of agency relations correlates positively with strong and negativelywith weak identification systems.

5. Conclusion

The concern of agency theory is that the welfare of the principal may not be maximized becauseof the different goals and risk preferences of principal and agent. Therefore, agency theory hasclassified typical problems in agency relations and suggestions for their solution from the pointof view of the principal. A social psychological perspective on agency relations and solutionmechanisms reveals that each agency problem (each solution of an agency problem) becomes aproblem of social power to the principal (agent) if she (he) changes her (his) beliefs, attitudes, orbehaviour 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 agencyproblem exists the principal possesses quantitatively more bases of power than the agent. In thehidden intentions and hidden action-situations, she additionally possesses those power bases withthe broader range. Agency theory assumes implicitly an asymmetry in power in favour of theprincipal.

The social power perspective paves the way to a new solution mechanism that agency theoryhad not yet classified. In all situations in which an agency problem exists the principal possessesreferent power. Identification systems are systems of means which the principal implements onthe 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 agencyproblems.

5.1. Considerations about empirical testing

The thesis of this article still lacks empirical grounding. Until now, I have presented some casesand real life examples in order to illustrate the argument, as well as some new Hypotheses 1–3that have been derived and may be empirically verified in further studies. I conclude with some


suggestions that indicate how this can be done. My suggestion is to choose the duration of orga-nizational 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 member-ships should ceteris paribus last longer in organizations that have created a strong organizationalculture.

5.2. Application to politics

The proposed three propositions approach does not only concern an objective definition ofpower that is useful for organization theory. Identification systems defined as systems of meanswhich the principal implements on the base of her positive referent power can be used as definitionof ideology. Then, the proposed three propositions concern ideology and its use on the micro-and macro-level of the society. It could turn out that the three propositions are more meaningfulwhen applied to politics than when applied to organizations. However, as agency theory is widelyapplied in economics and economists tend to give power a wide berth the three propositions maynevertheless 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 possessionof bases of power. In his power-interaction model, Raven (1992) describes the process of trans-forming the bases of power into the exertion of power. Based on her motivation to influence, theinfluencing agent assesses which bases of power are available to her, especially in relation to thetarget of influence. If the principal makes a wrong assessment the agent may turn out to be morepowerful.

The proposition on the asymmetry in power in favour of the principal was derived for the dyad ofprincipal and agent. The dyad is the smallest type of social interaction. However, in organizationsand in many other social interaction systems in which agency theory is applied principals andagents interact in multiple principal-agent structures. Multiple principal-agent settings will notterminate the efficacy of identification systems. However, they may lead to another distributionof social power between principals and agents than in the dyad. An analysis of the asymmetry inpower 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 assumptionswhich agency theory completely ignores. For example, Perrow (1981, 1986) has criticized thatagency theory ignores power. This paper may be a first step in the development of an extendedagency theory that includes an explicit concept of power. Pillars of this extended agency the-ory could be: agency relations are characterized by informational asymmetry, asymmetry inpower, outcome uncertainty, and risk. In the normal case, the solution of the agency problemis a power phenomenon namely if either principal or agent changes his or her belief, attitudeor behaviour as a consequence of the action, or presence, of the other. The agency problemsof the principal are based on the bases of power of the agent. The solution mechanisms of theprincipal are based on the bases of power of the principal. Power comes across as an interveningvariable.


References

Alchian, A.A., Woodward, S., 1988. The firm is dead–long live the firm: a review of Oliver E. Williamson’s The economicinstitutions 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 ofBusiness. 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, Tubingen.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.


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, Tubingen.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.


Spremann, K., 1987. Agent and principal. In: Spremann, K., Bamberg, G. (Eds.), Agency Theory, Information, andIncentives. Springer, Berlin, pp. 3–37, Corrected reprint, 1989.

Spremann, K., 1990. Asymmetrische Information. Zeitschrift fur 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: extensions and 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.

Saam 2007.pdf

Documents

Transcript of Saam 2007.pdf