D 'ofT$P'' M.I.T. LIBRARIES -DEWE^ Digitized by the Internet Archive in 2011 with funding from Boston Library Consortium Member Libraries http://www.archive.org/details/formalrealauthor00aghi2 HB31 .M415 f\0> / working paper department of economics FORMAL AND REAL AUTHORITY IN ORGANIZATIONS Philippe Aghion Jean Tirole 95-8 June, 1994 massachusetts institute of technology 50 memorial drive Cambridge, mass. 02139 FORMAL AND REAL AUTHORITY IN ORGANIZATIONS Philippe Aghion Jean Tirole June, 1994 95-8 MASSAC? !US£TTS INSTI i"U'f£ OF TECHNOLOGY MAR 2 9 1995 USRARiES Formal and Real Authority in Organizations* Philippe Aghiont and June "The authors 1994 and Martin Hellwig for helpful discussions, to and especially Denis Gromb and David Martimort for helpand to the Centre National d'Etudes des Telecommunications are grateful to Oliver Hart Patrick Bolton, Leonardo ful 1, Jean Tirole* comments on a first Felli, draft, for financial support. 'Nuffield College, Oxford tlDEI, Toulouse, France; and EBRD, London CERAS, Paris; and MIT, Cambridge, MA USA Abstract The paper develops a theory cide) and real authority (the effective control over decisions), integrated structure can is accommodate various degrees determined by the structure of information, which formal authority. a of the separation between formal authority (the right to de- An loss of control for and illustrates how a formally of "real" integration. Real authority in turn depends on the allocation of increase in an agent's real authority promotes initiative but results in the principal. After showing that firm boundaries and information structures are intertwined, the paper examines a number of factors that increase the subordinates' real authority in an integrated structure: overload, lenient rules, urgency of decision, reputation, performance measurement, and multiplicity of superiors. Lastly, the amount of communication in an organization allocation of formal authority. is shown to depend on the Introduction 1 Over forty years ago, Herbert Simon defined authority part or the whole of an organization. Moore ( 1 1 As pointed out by Grossman-Hart (1986) and Hart- may be conferred by 990), authority as the right to select actions affecting the ownership of an asset, which gives the owner may more the right to take decisions concerning the use of this asset. Authority result to a from an member members of the organization. 2 This formal authority however need not confer real authority, that over decisions, upon its holder. For example, it is commonplace have limited control over their board of directors, which nation of the top executives, who in itself an is effective control to observe that shareholders may be subject to the domi- turn often rubber-stamp the divisions' projects, and so Similarly, the president of a country really controls only a small forth. on specified matters explicit or implicit contract allocating the right to decide or group of generally number of the deci- sions taken by the executive branch. This paper develops a theory of the separation formal authority and real authority, accommodate various degrees Our approach Weber notes that follows Max officials, 3 and illustrates how a formally between integrated structure can of "real" integration. Weber's (1968) description of "rational" or "legal" authority. employees, and workers attached to the administrative staff of a bureaucracy do not themselves own the non-human means of production and administration, yet they may exert substantial control over the bureaucratic machinery (p217-225). example, he observes that For : "Bureaucratic administration means fundamentally domination through knowlis the feature of it which makes it specifically rational. This consists on the one hand in technical knowledge which, by itself, is sufficient to ensure it a position of extraordinary power. But in addition to this, bureaucratic organizations, or the holders of power who make use of them, have the tendency to increase their power still further by the knowledge growing out of experience in the service. For they acquire through the conduct of office a special knowledge of facts and have available a store of documentary material peculiar to themselves. While not peculiar to bureaucratic organizations, the concept of "official secrets" is certainly typical of them. It stands in relation to technical knowledge in somewhat the same position as commercial secrets do to technological training. It is edge. This 1 [a "We will say that the boss exercises authority over the worker "behavior", i.e., any element of the the worker accepts authority will accept authority only acceptance") of in all if when Xq, tne his if the worker permits the boss to select x worker performs on the job]. That is, determined by the boss's decision. In general, the worker set of specific actions that the behavior is x chosen by the boss, the possible values. This modern administrative theory" [Simon ( is is restricted to some subset the definition of authority that is (the worker's "area of most generally employed 1951 ,p294)]. This corresponds to Weber's (1968) notion of "rational" or "legal" authority. Weber distinguishes among three types of authority, the other two being "traditional authority" and "charismatic authority". 2 3 and control is reflected in the principal-agent literature, starting Holmstrom with Mirrlees (1975), (1979) and Shavell (1979). This literature studies the structuring of compensation and rewards, assuming full control by the agent, and does not investigate the allocation of The distinction between ownership authority, formal or real. 1 a product of the striving for power." As The tion. so real authority is asymmetric informa- superior can always reverse her subordinate's decision, but will refrain from doing the subordinate if We Weber, the key to our analysis of formal vs in is much better informed and formalize this idea in a straightforward way. The tiative) to suggest a project to the superior. is The subordinate exerts effort (shows latter as well chooses how much ini- to learn Once informed, the subordinate recommends a project that about the potential project. sometimes their objectives are not too antinomic. if not optimal for the superior, because from the point of view of the agent this project creates a higher private benefit, yields better career opportunities, or requires less effort to superior may be implemented than the optimal project. is Formal authority prevails when the informed, since the superior then chooses her preferred project (which may or not coincide with the subordinate's proposal). In contrast, a poorly informed superior optimally rubber-stamps the subordinate's proposal by fear of picking a worse alternative. The subordinate then has A superior who real, although no formal authority. for instance is overloaded and therefore has subordinates' decisions loses control and involuntarily endorses Conversely, too much what The information also hurts the superior. creates initiative from the subordinate. little time to monitor her many suboptimal projects. very prospect of influencing The subordinate loses motivation decisions is when the superior's incentives (for instance, idleness, obsession or high competency) lead to "stand constantly There is behind the subordinate's shoulders", that to monitor very thoroughly. thus a general trade-off between initiative and loss of control. degree of separation of ownership and control task, is on who performs and monitors it, is endogenous; him Furthermore, the depends on the nature of the it and on the organizational structure. 4 This new approach to ownership and control based on the distinction between formal and real authority delivers a First, it number of interesting conclusions. provides a logical relationship between the information structure and the alloca- tion of decision rights within an organization: 5 The reallocation of formal authority to the agent prevents the principal from overruling the agent in those situations (a la Grossman- Hart (1986)) where both parties have acquired the relevant information about the potential projects' payoffs. The agent's initiative. On 4 A more transfer of formal authority to the agent thus credibly increases the the other hand, the principal has straightforward application of this idea likely to is more incentives to become informed the prediction that the delegation of day-to-day activities be observed than the delegation of key investments. This of day-to-day activities, but rather to the fact that they are very is is not due to a relative irrelevance numerous and therefore more costly to monitor. 3 Note also that an agent who contract with the principal, decision making. may is "outside the organization" in the sense of not having signed an initial may bring information that influences also exert "real authority", in that he Firm boundaries and information she has the right to reverse decisions. if thus intertwined, and the size of a firm structures are ultimately determined by the above trade-off be- is tween the subordinate's initiative and the principal's cost of losing control over the choice of projects. Second, whilst conferring decision rights on the agent increases a number of factors that encourage initiative factor of initiative and loss of control when his initiative, there are also the principal retains formal authority. One a wide span of control, which raises the principal's is We marginal cost of monitoring each agent. show that there a sense in which optimal is organizations always function in a situation of overload. Another raise-the-monitoring-cost way of committing not to from imposing stifle initiative is to refrain rules that constrain the agent within a set of easily monitorable activities. Alternatively, the gain from intervention can be reduced by spreading also be made more its among benefits several principals/owners; by splitting property rights difficult among of a matrix organization or of multiministry oversight) 6 intervention can several superiors (as in the case and by requiring that intervention be unanimously agreed upon. Other factors that encourage does not give the superior initiative include the much time for urgency of decision making, which a thorough investigation; repeated interaction, which allows the superior to develop a reputation for not intervening fairly in matters that are inconsequential to her and for intervening only in important matters; and improved performance measurement. Third, our approach enables us to provide a modest, but of "'collective bounded plications of imperfect rationality"' and incentives. A first, step toward the integration recent literature has studied the im- communication on the functioning of an organization. 7 This research has taken the imperfection of communication as given and in general has assumed that bers of the organization pursue the same objective. endogenizing the limits to communication. team theoretic framework of the The communication because the agent is is would basic idea literature to allow of information ship. In particular, less The We is like to go one step beyond by to depart members to mem- from the traditional have dissonant objectives. then strategic and depends on the authority relation- communication may take place if the principal has formal authority concerned that the principal might abuse her authority once she informed. This will typically be the case if sufficiently dissonant. In the opposite case is well the principal's and subordinate's objectives are where these objectives are we show that communication may instead be encouraged by the sufficiently congruent agent's subordination to the principal. 6 a This theme number 7 is developed in a corporate finance framework by Burkart, Gromb and Panunzi (1994), with of interesting implications. See, e.g., Marschak-Radner (1992). Sah-Stiglitz (1986), Bolton- Dewatripont (1992) and Radner (1992). Although our analysis The moral hazard and property blocks developed by other authors. A the two polar cases. our knowledge, new, to the best of is, Riordan (1990), and Schmidt (1991) develop may hurt a principal. 9 rights literatures supplied Cremer 8 MIT in Papers by Cremer (1992), in different contexts the idea that too much Riordan (1990) argues that information allows principals to expropriate the agents' specific investments. based on information. makes use of building seminar given by Diego Rodriguez and Dimitri Vayanos at 1991 contained several seeds of the basic model described here. information it (in He provides a definition of vertical integration the context of a corporation) and Schmidt (in a paper on privatization) show that poor information allows principals to avoid (ex ante costly) renegotiation of long term contracts with agents. In Cremer's paper, the principal publicly chooses the accuracy of a technology used to monitor the agent's type. technology reduces the agent's incentive to work to signal high discuss property rights and, in the tradition of establish) a link The paper it: first It Arrow (1975), accurate Riordan and Schmidt posit (but do not formally between property rights and information structure. is organized as follows: Section 2 presents the model. relates the information structure formal authority; thority ability. A more Section 3 analyzes and equilibrium payoffs to the allocation then demonstrates a tight link between our incomplete contract/au- it methodology and a complete contract approach. Section 4 looks the agent's initiative of when the at factors favoring principal has formal authority (overload, lenient rules, urgency of decision, reputation, better performance measurement, multiple principals), and derives implications for business management. Section 5 endogenizes communication, and section 6 concludes. 2 A The model hierarchy composed of a principal (she) and an agent (he) can implement one or zero project. project. The principal hires the agent to collect information about Examples of hierarchies we have in mind and implement the are board of directors or trustees/mana- gement, CEO/division manager, thesis advisor/student, foreman/worker, or supranational authority/country. • Projects: The agent screens among n > 3 potential behalf of the principal. To each project k 6 {1, gain or profit gross of any Bk for the principal monetary • ,n} and a private benefit transfer between the two parties]. perks on the job, acquisition of human and a priori identical projects on is associated a verifiable monetary &<, for the agent. [These payoffs are The agent's private benefit includes capital, the possibility of signalling ability, or (minus) "Their 1993 discussion paper focuses on different themes than the ones considered here. 3 The literature on the ratchet effect also emphasizes a cost for a principal from being well informed. the disutility of implementing the project. private benefit are both equal to zero. with known payoffs B = = b 10 If no project is implemented, the profit "No project" can formally be treated and the as project 0, 0. Only two of the n potential projects are actually "serious" or "relevant". projects yield "sufficiently negative" payoffs to both parties. All other Anticipating somewhat, this imply that an uninformed agent prefers to confess ignorance and to recommend inaction will recommend a rather than to specific project, and that an uninformed principal similarly would not choose by herself to undertake a project. One Similarly, other B > of the two relevant projects yields profit one of the two relevant projects yields private benefit The ex ante 0. to the principal probability that the same project is b > and the other to the agent, preferred by both is a6 0. and the [0, 1], the parameter of congruence. 11 The • Preferences: and w The agent's utility is the principal wage paid risk neutral to the agent. then u(w) is is + 6t, and has The agent where u(-) is is utility wage 4.5 is project k is liability, so chosen, w > 0. increasing and concave. of zero. Alternatively, the agent the principal's benefit if is infinitely risk averse to income. monetary incentives and receives a constant wage equal therefore does not respond to his reservation —w protected by limited For expositional simplicity we will assume that the agent He Bk noncontractible; the agent's may wage to not be infinitely risk averse, but is then again a constant. [Section shows that the model can be straightforwardly extended to allow the agent to respond to monetary incentives. Profit sharing then lowers the principal's and raises the agent's utility from picking a profitable project.] • Information: The agent acquires information in a binary form. At private cost ^(e), he perfectly learns the payoffs of (1 — 10 and all candidate projects with probability the agent learns nothing and e), For example, a R&D common units put too applications ("D"). complaint in large much emphasis on The emphasis on still e. With probability views the projects as identical. conglomerates is basic research ("R") basic research stems that R&D and too operations get out of control little from the units' private on developing commercial benefit. "This congruence parameter will be treated as exogenous in the following analysis. However, one could think of various methods whereby the principal might affect congruence with her subordinates: for example, investments in the recruiting and training of new employees, design of career profiles, enforcement of [contractual] rules restricting the subordinates' set of possible actions. An interesting determinant of by the activity of this agent. vis-a-vis their units. IBM A is less a is the possibility that other agents working for the principal are affected is supplied by the difference in behavior of IBM and Fujitsu case in point successful in keeping an arms-length relationship with its units than Fujitsu According to The Economist (Management Focus. April 10, 1993), one reason is that Fujitsu and ICL are not serious rivals anywhere while IBM's units are often competitors. Under the plausible hypothesis that the existence of externalities on other units reduces the congruence parameter, this view is consistent with our theory, which predicts that the subordinate has more discretion, with its British subsidiary ICL. the higher the congruence. A systematic exploration of the various ways of endogenizing the congruence parameter scope of this paper. lies beyond the Similarly, the principal chooses how much time or effort to devote to learning payoffs. At private cost gp{E), she becomes perfectly informed about the payoffs with probability and learns nothing with probability The (1 — E). be contemporaneous with the agent's or principal's acquisition of information can else start after the agent makes his report. We will refer to these simultaneous and sequential models, respectively. Which variant on the context. Sequential investigations usually are who can already build on an existing report. On is two possibilities as the more relevant depends time consuming for the principal, less the other hand, the principal to wait until the report accrues to start her investigation, as otherwise she to accept the agent's proposal by lack of time. same cases yield essentially the results, E we 12 may not want may be forced Because the simultaneous and sequential will focus on the simultaneous case and content ourselves with illustrating the sequential case in section 4.3. We leave the endogenization of the timing (simultaneous vs sequential) for future research. The gi {0) • = Q,g' (Q) i = verified = 0,g'i (\) Communication: or soft. and gp(-) are increasing and disutilities of effort fiu(-) oo,i most of the paper we can assume that information In by the other party if communicated by the party who The as a pure suggestion for a project choice. and is it. Soft information specific results of section 4.3 rely on the we will occasionally label "inte- the main focus of this paper), the principal has the formal authority and She indeed does so if . she The is principal may always informed and In this case, the principal has fully collected soft. called the '''superior'' and can either hard communication must be interpreted Authority: In the case of P-formal authority (which gration'' is its is be costlessly and instantaneously project's payoffs can cannot be verified by the other party and therefore • convex and satisfy = A,P. Hard information about a information being strictly if overrule the agent (the "subordinate"). the agent's recommendation both the formal and is not "congruent". real authority over the choice of project dispense with the agent's information and recommendation. Otherwise, she (optimally) rubber-stamps the agent's proposal. We will then say that the agent has real authority. Our payoff structure subordinate; The standard for, implies that there is no need to include an the superior always takes a decision that yields nonnegative utility to both. institution of letting subordinates quit decisions emerges naturally in the variant of our 12 For instance, directors of a stamp the annual report or involved. "exit option" for the company or external to accept the thesis if if they are unhappy with their superiors' model members in which the principal's preferred of a thesis jury are usually forced to rubber- they have waited until receiving the documents to become may impose project a substantial loss of utility for the agent. Under A-formal authority (which we dent agenf That is, agent is the agent now has formal Note that authority. this covers the situation in which the this particular 13 will assume that whoever has authority zero, but benefits the other party to doing two relevant projects yield • the "indepen- informed, picks his preferred project and cannot be overruled by the principal. an employee who contractually receives an irrevocable right to take decision. We if , will occasionally label "delegation"), Contracts. In prefers choosing a project that gives nothing (equivalently, we could assume that the strictly positive payoffs to most of the analysis, him both parties). we adopt the incomplete contracting approach (Grossman-Hart (1986)) by positing that projects cannot be described and contracted upon ex ante. to The initial one of the two The timing is contract only specifies an allocation of formal authority (control rights) parties. as follows: (i) The principal proposes a contract that allocates formal authority to her or the agent over the future choice of projects; 14 gather information about the n projects' payoffs; (iii) the party (ii) who the parties privately does not have formal authority communicates a subset of the relevant projects' payoffs he has learned to the controlling party; given that there are only two relevant projects, this subset zero, one or two project payoffs descriptions; (iv) may comprise the controlling party picks a project (or none) on the basis of his information and the information communicated by the other party. To put the incomplete contracting approach somewhat in perspective, section 3.4 explores the polar assumption that projects can be described and contracted though their payoffs are ex ante unknown to both parties. We upon ex will see that ante, even under specific circumstances, the optimal complete contract corresponds exactly to a (possibly random) authority allocation scheme, and we will point at the implications of contract incompleteness when these circumstances do not • Payoffs obtain. under the two allocations of authority. Under P -formal authority (integration), 13 The allocation of authority must therefore be thought of as issue specific. For instance, the chairperson of a department can pick the salary or the teaching load of a professor but has no right to choose the professor's research agenda. Another case in point is that of Du Pont in 1921, in which "headquarters did not have the [formal] authority to step in and interfere with divisional decisions unless something was clearly going wrong" (Chandler, in Continental Bank (1993), p56, bracket added). Our model has a single project/issue, but we find its extension to multiple decision rights and clusters of authority potentially fascinating. Note also that the delegation of formal authority is sometimes partial in that the principal may be able to regain authority at a large cost: An official may be impeached, or a board of directors overruled by shareholders through a takeover or a proxy l4 fight. we assume that there is ex ante a competitive supply of potential agents, so that the allocation of authority between the two parties is the one that maximizes the principal's ex ante expected utility That is, the utilities are = EB + uP {1 -E)eaB-gp(E) (1) and = Eab + {l-E)eb-gA (e). uA That is, with probability E, the principal probability (1 and suggests — E), the principal is is informed and picks her preferred project. With uninformed. With probability The principal then his preferred project. (2) either learns the payoffs attached to this project (hard information) or is still e, the agent from his is informed recommendation uncertain about whether the agent proposes her preferred project (soft information). Either way, the principal optimally rubber-stamps the agent's proposal. 15 Under A-formal authority (delegation) when informed, the agent simply chooses ferred project. When the agent is uninformed, and the principal suggests her preferred project, which is is his pre- informed, the principal then implemented by the agent. So, using the su- perscript "d" for "delegation", preferences are = eaB + up {\ -e)EB-gp(E) (3) and d uA =eb+{l-e)Eab-gA (e). (4) Note that the agent's lack of responsiveness to monetary incentives precludes any renegotiation of the exercise of authority. Remark: That the principal is principal's formal authority uninformed follows from the becomes entirely ineffective when specific projects payoff structure, in particular the from the principal's weakly preferring a (relevant) noncongruent project to no project at all. To > 4), say see this, consider the following example: there are three relevant projects (thus k = 1 and 2 1,2 and 3. n Project 3 yields a strictly negative profit to the principal, whilst projects yield (as above) a nonnegative profit. whenever uninformed can still Then, if information use her formal authority in order to project 3 vs projects {1,2} from the agent. The is elicit hard, the principal information about principal will then rubber-stamp the agent's decision only conditionally on information ruling out the negative-profit project having been disclosed to her. 13 Note that the superior is always better off, the higher the agent's effort. This need not hold in a multi- may then crowd out his effort on other one task may not be positively valued by the principal. task extension of our model; for, the agent's effort on this task Consequently, the agent's initiative in tasks. Analysis 3 The 3.1 The reaction curves under P- formal authority first-order conditions when the principal has formal authority are: - ea)B = g'P (E) (1 (5) and - (I The = g'A (e). principal supervises more, the higher her stake and the agent's and E)b effort. (6) and the lower the congruence parameter The agent demonstrates more initiative, the higher his private benefit the lower the principal's interference. We assume that the two systems (E.e) 16 [Such an assumption acquires information only stability condition The is if is of equations {(5), (6)} has a unique, stable intersection not needed in the sequential case: the agent makes a proposal, automatically satisfied. sloping, the principal independent of e and the ] In contrast, would never want is 17 fact that the agent's reaction curve (6) of this (or any) delegation model. E Because the principal is if downward sloping is a crucial feature the agent's reaction curve were upward to reduce her degree of interference E for strategic reasons. This latter case might correspond either to a situation of strategic complementarity if the principal's reaction curve were also upward sloping or to a supervision situation principal's reaction curve were Finally, downward if the sloping. note the first-order condition (5) implies that the principal's reaction to the agent's effort is also downward sloping. an incentive to substitute for the agent that the agent is indispensable. That The interpretation in case is, is that the principal can and has the agent does not work. Suppose in contrast the principal cannot invent a project if the agent does not come up with a proper suggestion; on the other hand, the principal can carefully read any curve is file then gathered by the agent and modify the existing project. The principal's reaction likely to slope up. [The comparative statics results obtained in this paper carry over to this case for shifts in the principal's reaction function, but not for shifts in the agent's reaction function.! 16 That is, abB < y P (E)g^{e). '"For example, with soft information, the principal's payoff in the sequential case up = e[EB + (1 - E)aB - g P (E)]. is: The 3.2 basic tradeoff between loss of control Suppose that some "exogenous" reason (e.g., initiative because of overload), the marginal cost of P ) increases. The effect on the principal's expected payoff is a ambiguous. On the one hand, ceteris paribus, the principal's probability of becoming effort of priori for and the principal (i.e., g' informed about the projects' payoffs (E) decreases (see equation loses real authority (i.e., control) over the choice of project, having to endorse suboptimal projects. intervention E On with a higher resulting risk of the other hand, the reduction in the principal's encourages initiative from the subordinate (see equation raises the principal's the principal thus (5)); (6)), which in turn expected (monetary) benefit. This basic tradeoff between loss of control and initiative determines the optimal allocation of real authority when the initial how the question then arises as to amount (optimal) This question is principal can actually commit The herself to delegate an of real authority to her subordinate while retaining formal authority. taken up foster the agent's initiative The optimal 3.3 contract allocates formal authority to the principal. One (extreme) way in detail in section 4. is to relinquish her for the principal to formal authority, as we shall now see. allocation of formal authority Whether formal authority should optimally be allocated to the principal or to the agent hinges on the tradeoff between loss of control and initiative. More formally, when the agent has formal authority, the first-order conditions are: (l-e)B = g'P (E) (7) and (l-aE)b = Assuming again that that {(7), (8)} yields (8) 18 one can show a unique, stable equilibrium (E^e*), : E > Ed Our model thus explains why and is agent, she has less incentive to is The e d . (e.g., in Williamson Because the principal cannot overrule the become informed and thus giving the formal authority to a credible way for the principal to commit not to intervene. The cost of leaving initiative to the agent, decisions < often suggested in the literature (1975,1985)), to an arms-length relationship. the agent e the absence of integration (where "integration" is understood as P-formal authority) corresponds, as ls g' (e). A more on the other hand, is a loss of control. The agent often. stability condition is the same as in the case of P-formal authority. 10 takes suboptimal Example: Let us compare the two governance structures tion, in One which gA (e) = e 2 =£ /2 and gP (E) 2 in the quadratic cost specifica- /2. 19 has B{\-ab) E_ 1 6(1 - abB ' 1 - B) - abB ' and d = B(l-b) I b(l-aB) d \-abB -abB' ' Furthermore, UP ~ P -formal authority is ^ = o'-ai"^ optimal for low " I (1 + 11 levels of Q) " a(1 " a6B)1 congruence (a small), and A-formal authority dominates when preferences almost coincide (a close to a" 6 3.4 such that P-formal authority (0, 1) optimal if 1). Indeed, there and only if a < a*. is a cutoff value 20 Complete vs incomplete contracts The purpose One is ' of this subsection common of the is to step back and reflect on our methodological approach. 21 motivations for using incomplete contracting models is that the standard complete contract paradigms (moral hazard, adverse selection or Nash implementation) to account for notions such as ownership and authority. We fail have adopted an incomplete contracting approach to analyze authority and its conceptualize our intuitions and obtain others. Yet, in view of the current lack of proper delegation and found it very useful to foundations for the incomplete contracting methodology, an investigation of what could be achieved with complete contracts will bring us a deeper perspective. Interestingly, while the notion of authority finds its natural habitat in contracting framework, it does not rely on such an interpretation. As we have seen, choosing who should decide amounts A to answering two subsidiary questions. First, who is informed? party with private information can manipulate decision making by revealing a coarser information structure than his structure. More concretely, he will typically disclose his preferred project and abstain from revealing information that of projects he likes less. answer to 19 as an incomplete this Second, who gets his second question hinges on the way when both classic tradeoff may lead to the adoption parties are informed? between ex post The efficiency of This case does not satisfy the assumption that g'i(l) = oc, but is still amenable to our analysis as long that, in the relevant range of parameters, the parties do not become informed with probability we assume 1. 20 To prove this, one uses the fact that the probabilities e and E must be less than one. One can also show when congruence is low, but may prefer not to have formal that the agent prefers having formal authority authority in specific circumstances (high congruence, private benefit small relative to the principal's payoff.) -'This section is not required reading for the rest of the paper, and may be skipped by readers who are primarily interested in the incomplete contracting approach. 11 making and ex ante decision as well in a world of These two questions surface incentives to acquire information. We now complete contracting with multiple moral hazard. state the circumstances under which the outcome can alternatively be interpreted as resulting from an optimal complete contract. a • In we maintain our assumption step, first only two parties (or that B is that the principal and the agent are the noncontractible); as we rules out the use of the sharing of the principal's profit the agent's initiative. The timing We assume limited as follows: is liability for The two (i) with outsiders both parties. parties sign a contract; information about the n projects' payoffs (for simplicity, a party also knows whether the other party has learned them they possibly renegotiate the new contract is either (if ("I the initial contract; (v) renegotiation has taken place) is assumption essentially shall see, this as well); initial in order to (ii) they privately gather who has (iii) learned the payoffs they send messages; contract implemented. encourage 22 A (if still in message (iv) force) or the m,-, i = A, P, have not learned the payoffs"), or a pair of payoffs for the principal and the agent for each project k specifies probabilities 23 = 1, • • • ,n ("I have learned the payoffs, and they pk(mp,m A ) are..."). A contract of implementing project k, such that n Y^,Pk(™p,m A ) < 1. k=i There are three states of the world therefore a project £(1 — e)); 2: is to be implemented: congruent project In state j is preferred project = 1, is £ {1,2,3}, implemented principal's preferred project r which at 1: is if let • , T one party learns the payoffs and only the principal learns the payoffs (probability (1 — E)e); 3: both learn the payoffs Xj denote the equilibrium probability that the there indeed is congruence; in case of noncongruence, the implemented with equilibrium probability implemented with equilibrium probability assume that the renegotiation process • least only the agent learns the payoffs (probability (probability Ee). We in is Zj, with y_, yj and the agent's + Zj < 1. a finite bargaining game. In each subperiod of this process, one of the parties suggests a project or doing nothing (or more generally, a probability distribution over these decisions); the other party accepts or turns down the offer. In the latter case, bargaining proceeds to the next subperiod. specified by the initial contract for the messages sent at stage agreement is (iii) is The decision implemented if no reached by subperiod T. Furthermore an informed party can disclose payoffs about particular projects at the start of each subperiod. For concreteness we "Alternatively, the two parties might be infinitely risk averse under zero income. An will assume arbitrarily small probability of error would then endogenize the limited liability assumption. 23 no leeway in using monetary transfers to adjust thrown away (since renegotiation is feasible), the agent, does not respond to monetary incentives. t'nlike in conventional mechanism design, there is incentives because there are only two parties, profits cannot be and one of the parties, 12 g that information is The argument can be extended with minor hard. modifications to soft information. The possibility of renegotiation 24 implies that ij = for all 1 compatibility plus the possibility of renegotiation imply that y t is, in case of Furthermore, incentive i. = 1 and z2 = renegotiation, is then implemented with probability irrelevant projects yield very negative payoffs, 1. To see this, note first because of that, because when both announce they have not the payoffs, the initial contract yields status quo expected payoffs equal to zero probability one no project is bargaining game and suppose and, assuming that amount positive = 0, = i A, P) at stage (if (Hi). Now consider the stage when last offer necessarily indifferent a party chooses we could assume what proposes preferred by the is that each relevant project yields at least a small Backward induction then shows to each party), the other party accepts. that by disclosing only his preferred project, the informed party necessarily gets his way. Last, the possibility of renegotiation implies that y3 when the information investments e is with that the informed party has disclosed only his preferred Then, whoever makes the project at the last subperiod T. other party (alternatively learned implemented) or negative (otherwise.) Second, suppose that the informed party feigns ignorance (m< this project That asymmetric information and noncongruence, the informed party can guaran- tee himself his preferred payoff by disclosing only his preferred project, which, (iv) 1. symmetric (state 3) + z3 = 1. Who gets his own own way determined by the relative desirability of the is and E. 25 To sum up, the optimal contract is a (random) authority scheme, in which authority is ex post conferred upon the principal with probability y3 and upon the agent with proba(note that the allocation of authority was shown in section 3.1 to be irrelevant in bility z3 states of asymmetric information, in which the informed party always gets his own way). Because random authority mechanisms are allowed by the incomplete contracting approach 24 The possibility of renegotiation rules out some contracting features that might emerge otherwise, such throwing away profit or not implementing profitable projects. The absence of renegotiation would also prevent parties from concealing information; for, the optimal contract would specify that no project is implemented unless two relevant projects are proposed and would costlessly ensure truthful revelation. This policy eliminates asymmetries of information. [To reintroduce real asymmetries of information in the absence of renegotiation, one may assume that each party observes with some probability the payoffs of only a subset as of projects 25 That and therefore can claim that he is, is aware only of one relevant project.] {y 3 ,z 3 ,e,E) solves: max{fl[£(l - e) + e(l - E)a + eE{a + (1 - a)y3 )} - g P {E)} s.t. E 6 argmaxlfl^l -e) + e(l - E)a+ eE{a + e€ argmax + -3 = I. {b[E{[ - e)a + e(l - E) J/3 13 + eE(a + (1 (I - a)y3 )] - gp(E)}, - a)z 3 )]- A (e)}, (although they are usually not considered in the literature), we conclude that the incomplete contracting approach cannot involve a loss of generality under these specific circumstances. What about • third parties? B is contractible (call it a profit), such parties who shareholders, headquarters) do exist in practice, (e.g., function by sharing her profit. A Assuming that contract now The presence specifies probabilities monetary rewards {Bp(mp,mA),B when profit some fraction in We leave We B. is it assume that of third parties — Bp(mp,m,A)} to the principal third parties by the renegotiation. to the reader to check that the optimal complete contract in the presence of (random) authority relationship absence of third parties) cum a linear profit sharing scheme receives a fraction in [0, 1] of the profit B. Profit sharing serves to encourage initiative profit sharing and the with the agent, the principal gets in case of renegotiation third parties takes the following simple form: It specifies a (as in the expands the contract space. {pfc(mp,m /i)} of implementing projects as well as of the increase in profit brought about [0, 1] alter the principal's objective if the principal has too is in therefore the much which the principal new instrument, and incentives to monitor. 26 Because schemes involving third parties are perfectly consistent with the incomplete contracting approach, we conclude that the incomplete contracting approach cannot involve a loss of generality in the model of section • We 2. by no means want to argue that optimal complete contracts always boil down to simple authority relationships (possibly augmented by profit sharing schemes). In particular, if the agent responded to monetary incentives as in section 4.5, he could be rewarded simply for bringing information to the fore rather which is impossible in than by implementing the project he prefers, the standard incomplete contracting approach followed here. ability to contract ex ante on projects then considerably alters the picture because it a disconnection between rewards for information acquisition and project choice. But it allows we find comforting that authority can under some circumstances be given the complete contracting interpretation of who gets his way in case of discordant announcements. when the Factors favoring initiative formal authority 4 Span of 4.1 It The is and initiative often argued that the planning and allocation process of the large conglomerates that were formed 26 control, overload principal has The in the "60s became bureaucratized, and third parties' ex post expected profit the shares. The reason why the is that the headquarters were responsible received ex ante by the principal through an auction of third parties act as a "sink" rather than a "source" (to use Holmstrom when that from the principal's viewpoint, the principal's effort is never suboptimal the principal receives the full profit B, but may be excessive in that it reduces the agent's initiative. (1982)'s terminology) is 14 for too many units, whose strategy they could not understand or The purpose a refocus on "core businesses". of this subsection is influence. This called for to introduce the superior's span of control and overload into the analysis. Suppose that a superior has authority over identical subordinates. Each subordinate screens in a set of tasks as described in section 2 i and learns the corresponding payoff structure with probability gp(EiEi) where of efforts is activity. The £,- is is: up = E,[£,-fl Each agent's reaction curve is still + (1 - Ei)tiaB principal's disutility is z's a fixed cost / per subordinate. 27 is (9) given by i assume that the equilibrium The -/}- gp(EiEi). (l-E )b = g'A (e We e,-. the principal's probability of learning the payoffs of agent subordinates' tasks are independent. There So, the principal's payoff m symmetric 28 and (1 (10) i ). stable. 29 - ae)B = g'P (mE), (11) and (1 Let {E(m),e(m)} denote the notation, - E)b = g' (e). A (12) solution of the system of equations {(11), (12)}. Abusing let up(m) = mR(E(m),e(m)) — gp(mE(m)), where R{E(m), e(m)) is = E{m)B + [1 - E{m)]e(m)aB - f the revenue per subordinate. Using the envelope theorem and treating the optimal span of control ^ The The [R(E(m), e(m)) There agents, An is - E(m)g P (mE(m))} + 0. extra agent brings revenue R, but requires attention number of subordinates of a positive reservation wage of the subordinates.) also exist who m^^- = asymmetric equilibria in A in the all In the quadratic case (see the which raises obtained when / > her attention to a subset of To eliminate asymmetric equilibria, one can sufficiently nonsubstitutable in the principal's disutility of effort function. 29 E absence of a fixed cost finite size is which the principal devotes therefore lack initiative, and none to the others. assume that the probabilities {£,} are (13) the marginal profit associated with a unit increase in superior would choose to have an infinite (or, equivalently, 28 obtained from expression in brackets in (13) the span of control. 27 = is m as a real number, example in subsection 3.3), this amounts to 15 abB < m. the cost of supervision by Eg'P "initiative effect", the "overload cost". , and measures the increase The second term, in ff j in the agents' effort associated 5- > 0, is the with a reduction in oversight. We will say that a firm is in a situation of overload employee, keeping employee behavior constant, always optimal for the firm initiative. to be in a situation is if the marginal profit of an extra negative. Equation (13) shows that it is of overload so as to credibly commit to reward 30 Remark 1: The analysis in this subsection has an interesting dynamic application: Sup- pose that the implementation of projects takes place continuously over time, and that at each point in time the principal can freely adjust the span of control by hiring or firing subordinates. Assume furthermore time passes by (there is that the principal acquires experience about her subordinates as learning-by-doing in monitoring). Then, the above trade-off between overload costs and initiative has the following dynamic equivalent: Letting the firm grow fast (that is, hiring new subordinates at a high speed) involves high overload costs a loss of control for the principal; on the other hand, a slow growth policy stifle is and therefore more likely to the subordinates' initiative as the principal acquires experience on monitoring them. Remark 2: We have assumed that all agents are subordinates. of "partial integration" determines the level of overload, with principal's authority corresponding to Remark 3: As noted in subsection 3.4, the principal to share B, provided CEOs managers or more B is More generally, the extent more agents subject to the overload. another credible way to promoting initiative verifiable, is for with a third party. [Recall that division are not residual claimants in practice.] While it has the drawback of reducing the principal's incentives on other tasks as well, profit sharing may reduce the desirability of overload. Remark 4: Overload is one way of raising the marginal cost of monitoring. Another way consists in refraining from investing in the monitoring structure (on this, see also the next subsection.) A case in point is the relationship between Fujitsu and its British subsidiary ICL. Fujitsu wanted to commit to preserve ICL's independence and initiative after acquiring it in 1990 and seems to have been successful in this respect. fact that only two Japanese managers are resident in 30 31 Of particular interest is London. [Fujitsu also wants to the float Because the marginal profit is negative, the principal would be better off committing herself, say, to playing golf rather than reaching overload. The problem with this is that playing golf is not a credible commitment (recall that the g P () function summarizes the principal's disutility of supervision and therefore already includes the cost of forgone opportunities). Overload is a credible commitment not to stifle initiative. 31 See The Economist (Management Focus. April 10, 1993). 16 shares of ICL 1995 while retaining a majority holding, in order to reduce in further guarantee ICL's independence. More 4.2 The purpose of this subsection we literature. Here, we concretely, rules. to provide a link is The concept will define a rule as N, but in will A "), r between the distribution of decision rights of "rule" has been given several meanings in the assume that the can be screened by the agent set of projects that N and N. The agent will screen projects either in not in both. There are therefore three contractual possibilities for a given two rules ("the agent must screen allocation of authority: in remark 3 above.] a contractual constraint on the agent's action space. 32 can ex ante be divided into two subsets N or this, see stake and lenient rules and the existence of More On its and the under which the agent lack of rule, Naturally, one of the rules, choose by himself, is namely the one that is in JV" "the agent must screen , free to pick his screening strategy. forces the agent to pick the subset he would equivalent to the lack of rule. So, there are really two choices: constrain or not constrain the agent by a rule. Concretely, we only one respect: will assume that the two candidate research strategies The principal has less expertise on research strategy therefore has a higher cost of monitoring projects in more result authority, that An reason. is is that a rule is more likely to be independent agent is here, the agent even strictly gains may likely to come up with a and loss of control is expertise in N and congruence is N, We then ask, is a rule imposed when the principal has formal not threatened by the principal's being well informed, Actually, in the simultaneous model studied by choosing research strategy N, because the principal N in order to protect himself against interference Whether the superior should of initiative strategy N. is useful idea. In contrast, in the absence of rule, a subordinate well choose research strategy his superior. in within an organization rather than across organizations, for the following because the principal cannot overrule him. more than be imposed when the principal has formal authority? likely to Our main N N and N differ in N than on N and from allow her subordinate to do so hinges on which the more important concern. If the superior has very little low, the superior wants to force the agent to adopt research N. So, let us assume that the g' higher than that of investigating recommendations in P {E), is agent searches in N or N is principal's marginal cost of investigating contractible. In each research line ( N, recommendations g'p(E). in Whether the N or N), there are as earlier As we will predict, such contractual restrictions are mainly enforced within organizations, even though they are sometimes observed across organizations, for example in research contracts between independent laboratories and product manufacturers, which sometimes restrict the research line to be pursued by the 32 laboratory. 17 two relevant projects and a bunch of negative value projects. The agent can investigate only one research Consider stead of line. first the case of P-formal authority. Having the subordinate screen within N shifts the superior's reaction curve downwards without affecting the subordinate's reaction curve. The equilibrium efforts therefore satisfy: E(N) < E{N) and e(N) As expected, the subordinate demonstrates more control when N. Furthermore, the of intervention, at least On agent, thus to is if e(N). One way the other hand, a lack of congruence makes initiative for the principal to N in order less relevant nate most often recommends projects that are useless to the superior. prefers to impose a rule when principal has less commit N to limit the threat 33 low. is and the the agent choose whether to screen within let given the freedom, chooses congruence if > initiative the superior has a higher monitoring cost. to a lower degree of intervention or N in- the congruence parameter is The small (up(N) An independent Consider now the case of A-formal authority. because the subordi- agent superior therefore > is up(N)). 34 not threatened by the principal's being well informed, because the principal cannot overrule him. Actually, simultaneous model studied here, the agent even strictly gains by choosing research in the strategy N, because the principal more is 33 In the absence of a congruence is high. rule, the It may subordinate may not want to inflict N = gp(E) h'P (E,K) and decreasing with 34 is As in the previous footnote, ( h P {Q, K) = for all K = 6 [0, is low. To But the see this, l],hp(E, K), with hp(E, 0) let = dE a close to zero since j^ > and lim<,_o e = 9a~ [(1 — d'p [B))b] > 0. us index the principal's disutility-of-effort function by a parameter K, g P (E),h P (E,l) = gp(E), and both h P {E,K) and h'P (E,K) decreasing with unambiguously negative h P {EJ<), withAp(£,0) K a K. Then, by the envelope theorem: du A which Rules high monitoring costs on the superior the congruence parameter if us index the principal's disutility-of-effort function by a parameter 1) 35 turn out that he benefits from the superior's being well informed. subordinate definitely prefers research strategy <jp(E),hp(E, useful idea. upon an independent agent. 36 therefore are therefore never imposed if come up with a likely to for let K and K) P {E, li' decreasing in K imply that h P (E, K) is also decreasing in K). The envelope theorem implies that: dup — - = ,, (1 - E)aBn ,-,> i//\ Thus, for a .More formally, since we have 36 to dh P (E,K) oK - > for a , close to zero. close to zero: u P (N) 35 Q— — de dl< u^N) = = u P (K E d (N) > E d (N) d = and since max, u%(E (N),e) > max e > 1) u P (K any u A (E (N),e) d for d Rules could arise under .4-forrnal authority for instance monitor them, but in the payoffs. IS = e = if 0) = u P (N). the function E >-> u dA (E,e) is increasing in E, d u A (N)._ N and JV differed not in the principal's ability Urgency and delegation 4.3 It is sometimes observed that the need to adapt quickly to customer requirements has forced firms to decentralize decision-making. some insights The purpose of this subsection is We superior can investigate only once the project proposal has been made. A Abusing terminology, we A > t and date T, him yields to start production even more time The time in t life and T; the other relevant project in this it in S€ choosing a stopping time [0, T] at which her investigations have not been successful by then. some date r, paper, F(t), is makes a it trivial Waiting namely, the probability that the principal learns is in monitoring, so the case of hard information principal learns stand for the product Similarly, the subordinate's preferred project profit. for monitoring; however decreasing returns first denote the per unit of time between the starting date for problem consists if the payoffs herself before For the T Let benefit. principal's decision longer gives her T will let private benefit b per unit of time, between no private The Suppose that the which time a superior substitute arrives on the market. at noncongruent project generates no gives B congruent project yields profit production it the time elapsed between the proposal (date 0) and the date at which the is product becomes obsolete. cycle. It offers urgency of the decision by the length will formalize the cycle, but several alternative interpretations are possible. life horizon, that two-fold: on how delegation might be affected by the urgency of the decision and illustrates the sequential case. of a product 37 hazard rate /(t)/[1 and uninteresting. it — F(r)] whether information difference from the agent and implements There are increasing, with density /(r). If the project immediately (S = 0); is is is decreasing. hard or soft. congruent, the on the other hand, case of noncongruence, the superior never implements the subordinate's preferred project, because there let us is always some hope that she assume that information flow profit B soft. is aB between S and T and t if uP rS e -rt = B 7' is e therefore: T she learns payoffs at if The g-rS -rT }f(t)dt 37 See, e.g., the discussions of This would not be so if < t S, 38 So and has principal's utility, conditional + aB[l - _ S and on the -rT F(S)}[ ], r the principal's rate of time preference. This objective function 38 T). 39 r Jo where _ [- is = she has not learned payoffs by date thus rubber-stamps the agent's project at date S. agent's having proposed a project, her preferred one (S For a given stopping rule 5, the principal obtains between the date of learning expected flow profit will discover Wyman-Gordon and WalMart in is quasi-concave, Continental Bank (1993). the subordinate's preferred project yielded a strictly positive profit to the superior. 39 In this formulation, the principal's cost of investigating duction of the product. The is simply forgone profit due to delayed intromore standard disutility of the sequential model can also be formulated with a principal's effort: See footnote 17. 19 and the optimal stopping time either zero is principal rubber-stamps without even checking), or f(S) The left-hand side of (14) (divided by B); its is ' ^~ > if i * a (^or % i-f(o) ^ ar g e enough, the is given by the first-order condition: - e - r ( T - 5) the marginal cost of delaying the introduction of the product right-hand side is equal to the conditional density of discovering the S and T (divided T (1 > dS/dT > payoffs times the value of overruling the agent's choice between by B). The optimal stopping time 0) decreases with a (dS/da < strictly positive increases if 0), as we would Our main expect. That horizon, the principal conducts a cursory investigation. to rubber-stamp, the Last, more urgent the we have been silent decision. result is that for a short the principal is, and is more likely 40 about the agent's behavior that his search time (which could be with in that we have random) was exogenously given. implicitly assumed In general this time could depend on the urgency of the decision due to altered incentives of the agent. Another interesting question (in a world of random time whether the agent would ever want is the date of adoption (recall that overruled (because dS/dT > of acquisition of information to delay a proposal. dS/dT < 0). Clearly, 1), Delaying the proposal delays but also reduces the probability of being an agent with a congruent project would not want to delay the proposal, but an agent with a noncongruent project might. late proposal could signal a principal (since dS/da < of positive correlation by the agent) In this case, a noncongruent project and be given substantial attention by the 0). We conjecture that this would not reverse our main insight between urgency and rubber-stamping, 41 but a formal treatment lies outside the scope of this paper. 4.4 As is Reputation and random delegation usual, an alternative to contracting or authority allocation is reputation. In practice, superiors try to develop reputations for "not intervening too often", or in a context with a larger project diversity, for "intervening only when not develop a formal model of reputation building, but the familiar 40 lines. We justified". it is In our terminology, Bolton will straightforward to do so along can sketch the broader idea of "intervening only when justified". Bolton and Farrell (1990) find that urgent decisions are more notion of centralization and For conciseness, we likely to the set of issues studied there are quite different and be centralized. However the from the ones analyzed here. Farrell's "centralization vs decentralization" refers more to the allocation of formal authority than to the degree of the subordinate's real authority. Bolton and Farrell analyze a multi-agent investment problem and contrast the duplication-and-delay inefficiency of decentralization with the incompetence of a bumbling central decision maker. 4 'No type'' of agent would choose to delay the proposal investigation. 20 if this also increased the principal's length of Suppose that the superior faces a sequence of agents. For each agent, the payoff structure is as described above except that with some probability a noncongruent project imposes a nonnegligible loss on the principal instead of yielding profit zero. For incentives purposes, may then be optimal for the superior to commit to overrule the agent only project yields a negative profit, in that overruling in the other case much the principal but reduces initiative too optimal in section A 3.3). to be worth is if if the noncongruent ex post optimal for (A-formal authority would be it patient superior facing enough subordinates a reputation for overruling agents only it may then develop the noncongruent project yields a negative profit. So, the superior uses her authority to overrule the subordinate "in important matters", but voluntarily relinquishes this authority (which is that are less important to her. This behavior would not be credible in from rubber-stamping) different which the superior would systematically overrule if in matters in a one-shot situation, informed. Performance measurement and subordinate's responsiveness to monetary incentives 4.5 The economics literature has emphasized the effect of the allocation of control on incentives. This subsection shows that incentives feed back on control. To this purpose, we generalize The profit > u" our theory to allow the agent to respond to monetary incentives. u(w) the agent's utility for project k is of generality, the agent receives w> The + 6jt (where u(0) when the = 0, u' principal's profit principal's net profit in her preferred project is now average gain from being informed and having real authority and b = decision; u(w) + ab u(w) > b: when congruent, the agent "aligned incentives". his private benefit is for The When u(w) < also receives b, is is congruent decision-making by raising the wage first-order conditions (5) and (6) and 0). Without loss B, and otherwise. B = B — w. 6 = b + au(w) wage w. The case u(w) > agent's monetary incentives are powerful and always recommends the < verifiable The for agent's u(w) < 6, the agent always picks his preferred b can be labelled enough that he forgoes principal's preferred decision. never optimal for the principal to set a wage just below u The 0, is -1 (6), Note that it because she can obtain slightly. under P-formal authority become: (l-ae)B = g' (E), P (15) and (1 The main it more (16) conclusion of this section can be drawn from these two equations. wage increases makes - E)b = g'A (e). real authority likely that the for two reasons: First, agent will be able to 21 A higher by raising the agent's incentives, recommend a project. Second, it it reduces the principal's incentive to monitor and therefore the probability that the principal overrules the agent. Letting profit {E{w) e(w)} denote the 1 with respect to w is: = ^f The term on the first second term not aligned) Remark (1 - E)a(B - w)^- - hand right is 1: a wage bill. if Chandler Continental Bank, financial controls + (1 - E)ea\. is is effective in closely involved in industries in The optimal wage (when 1993., in incentives are mature industries while the corporate which new product development office The is critical. management approach, "applying decentralized financial controls to mature businesses and a more centralized form of "strategic" control to capital-intensive, high-tech businesses". serves that at (generally deemed successful) GE from the corporate office, He ob- during the '80s, the managers of the "core" businesses - the long-established, mature businesses - received gets The p54-58) argues that decentralization application of this idea within a firm yields the so-called dual top which consists (17) small, but can be positive in general. equal to zero (in [E side of (17) corresponds to the increase in initiative. reflects the increase in the combined with must be solution of {(15), (16)}, the derivative of the principal's and were run instead through and budget-based bonuses). The corporate strict little planning or attention monetary incentives (bud- office in contrast was very involved in the high-tech businesses (aerospace, aircraft engines, medical equipment), for which monetary incentives are harder to design (due to the uncertainty Our argument that better performance measurement raises an agent's real authority offers a rationale for the dual top Remark 2: and the novelty of the products.) The management approach. equilibrium described above may not be immune to the possibility of renegotiation. For, suppose that the principal and the agent have ex ante agreed on a wage w and that the agent has learned the payoffs while the principal has not. Suppose that 42 In case of b > u(w). Suppose further that the agent's information is hard information. noncongruence, the principal must raise the wage to w* = u~ (b) in order to get a profit. She l will be willing to do so commit to wage w" ex if w~ < B. While renegotiation occurs, the principal may not want ante, because she can get away with a lower wage when she informed or when the projects are congruent. The analysis to that 42 It is developed in the is absence of renegotiation. equally straightforward to study the case of soft information. 22 herself to is otherwise qualitatively similar Multiple principals 4.6 Having multiple principals providing a generally believed to impact on an agent's behavior. is treatment of this topic full already yields a number out of the scope of this paper, a short discussion and authority. The benefit B, provided principals. Consider for instance the case of to receive B/n want to maximize all whole set of principals as a profit, it is monetary, can be n equal partners in case of success of the project. The principals. There are two dimensions to the deconcen- of useful observations. tration of ownership: returns a) Splitting returns. lies While is We among split several (or "co-owners"), each entitled assume the same cost function informed the allocation of authority if any of them among them Because they is. is for all Each irrelevant. and the agent's reaction curves are respectively given by principal's (1 - Er-/(l- ea)- = g'p (E) (5') -E) n b = g'A (e). (6') n and (1 Spreading monetary benefits among several principals has two it effects on initiative. First, generates free riding and therefore reduces monitoring. This effect dominates principal's cost function In this case, On control. an increase not too convex, as is in the number is the case for instance for a quadratic cost. 43 results in a loss of the other hand, with a very convex cost function, the multiplication of monitors 1: tiplying the and of principals raises initiative substantially improves the monitoring structure, which Remark when the We may reduce initiative. have assumed that the principals have similar monitoring number of monitors may increase monitoring if 44 abilities. Mul- the principals' talents are complementary. 45 43 To show and this, rewrite (5') (6') in terms of the probability £ that the principals be informed: Let E(£, n) be defined by (1 The -E) n = \-£. first-order conditions are then: (1 -£)(1 - ea)B = n(l - E(£,n))g'P (E(£,n)) (5") and (l-£)b = g'A (e). In the {£ e) space, , hand side of (5")- an increase We leave it in n (6") shifts only the principals' reaction curve, to the reader to check that for a quadratic gp through a change in the right- function, the right-hand side of (5") increases with n. for E < Eq, and = oo demonstrated with the following functions: gp(E) = n is given by 1 — £ = (1 — E ) Consider a project that yields a strictly positive profit only if both its marketing and manufacturing sides are satisfactory. The use of two experts in the two fields increases the efficiency of the monitoring 44 This point E>E for 45 . is Then most easily the probability that the principals are informed, £, process. 23 . Remark The 2: insights of this subsection should also apply to multi-layers hierarchies where, for instance, downstream agents are monitored by both a principal and a supervisor The monetary (middle-man). benefits are then spread systematic analysis of complex organizations however Splitting authority b) Splitting authority: consequence thority is if lies among among beyond the scope of this paper. several principals obviously has among manufacturing divisions in no But au- the principals' objectives are aligned as in the previous example. often split A several (upper) layers. principals with imperfectly aligned objectives (marketing and a matrix organization, multiple ministries, chambers in congress, Who partners in a joint venture, creditors in a bankruptcy process). hinges on the matrix of congruence parameters among principals has real authority then and agent, as well as on the governance mechanism (for example, each principal can have veto power, or there can be majority voting with or without the participation of the agent.) considerations, the agent's initiative A conflict of interest It may way. principals may recommendations. may be On or reduced by the split of authority. increase the probability of veto by one of them. also raise each principal's incentive to principals' agent among may be enhanced Depending on these become informed and not on the other to rely the other hand, for more collegial decision processes, the able to ''play" his multiple principals against each other and thereby get his 46 Authority and communication 5 This section extends the basic framework by introducing the possibility that the agent com- municates some prior information he question then is may privately hold about the projects. whether the allocation of formal authority affects the A natural communication of (relevant) information by the agent. We assume that at the beginning the agent can the principal's marginal cost of investigation from all E > 0. communicate information that reduces g' P to [For example, the agent might privately P such that g'p(E) g' know that the two to the principal section 2 except that the agent to the principal. Then first is noncontractible. The timing P {E) g' for relevant projects belong to a subset N^ of vV and decide whether to reveal Ni to the principal.] communicating information > is The action of as described in chooses whether to communicate his private information the two parties choose noncooperatively to invest in learning the projects' payoffs. how much effort (E and e) Depending on the allocation of formal authority the equilibrium efforts are given by the first-order conditions (5) through (8) for the relevant marginal-disutility-of-effort function for the principal (g'P {-) or g'P (-))lS See Davis-Lawrence (1977) for a description of such behaviors. 24 Communication curve up (regardless of the allocation of shifts the principal's reaction The formal authority) and has no effect on the agent's reaction curve. E effort to communicate information principal's marginal disutility function h'P (E,0) = question of whether the agent wants down to the principal thus boils Without from the principal's being better informed. with The thus increases in a stable equilibrium. g'p(E), h'P {E,l) = to whether the agent gains loss of generality let us by a communication parameter P (E) g' noted that the principal's equilibrium principal's monitoring for all E and E(K) and efforts h'P (E,K) E d K€ index the P (E, K) [0, 1], h' decreasing in K. (K) increase with K. Using the envelope theorem, the impact of communication on the agent's utility (see equations (4)) is We just and (2) given by: du A , 7K= {Q .,dE - e)b (18) dK and § An independent M (>-*<<. = agent always benefits from the principal's being better informed. In contrast, communicate information a subordinate wants to if his expected gain from the superior becoming informed, ab, exceeds the expected benefit from having a > e. When congruence his information. low, is a < e, 47 so the agent In case of low congruence there is real authority, eb, or better off not communicating more communication by an independent is agent. In this framework, there formal authority. This is may no actually always at least as longer be the case if much communication under A- congruence is high and the agent incurs a direct (fixed) cost of communicating the information, as can be seen from equation (19). The agent no he himself that e d is is longer derives a benefit from the principal's being ex ante well informed well informed (e indeed close to 1, d close to 1). If the agent's private benefit In contrast, with quadratic payoffs) that, provided a is it may be the case (this can be checked high enough, **>*£~0. dl< centives to ' formal authority affects the agent's prior information to the principal. of formal authority on communication depends ticular the degree of 0ne has lim a _ e = The impact upon the parameters in- of the allocation of the model, in par- congruence between the principal's and the agent's objectives. More communication may take place under P-formal authority 47 (20) K dl< this discussion, the allocation of communicate high enough so the independent agent does not bother incurring the cost of communicating ex ante information. To summarize is if g'^ l [{l - l g'p (B))]b > 0. 25 if these objectives are sufficiently congruent, communication less take place will if they are too dissonant. We conclude this section with two remarks: Remark J: The discussion above has focused on the effects of the allocation of formal authority on communication. One can reasonably conjecture that the organizational factors that affect the distribution of real authority under P-formal authority (see section 4) will thereby also affect the amount of communication within the same organization. amount the equilibrium of in section 4 Remark 2: ( communication varies "comonotonically" with the amount of by the principal through various commitment devices of the kind initiative left to the agent mentioned Whether span of control, multiple principals, The above analysis shows that etc.), is left for future research. communication can never be detrimental to an independent agent. Such a strong conclusion, however, is unlikely to be robust to various extensions of our basic model, in particular to the introduction of the agent's responsiveness to monetary incentives. For example, one could imagine that an independent agent might prefer not to help the principal find out that the two parties' preferences are congruent in order to credibly blackmail the principal ex post and thereby obtain a higher monetary compensation. Summary and 6 Let us first extensions summarize our main of information, the are endogenous amount of points. In an organization, the structure and the distribution communication among members, and the existence of and depend on the allocation of decision raises one's incentives to become informed; consequently improves the principal's information about the agent's agent's initiative. It may rights. also (although it Having the right to decide vertical integration activity. Relatedly, rules all endogenously it reduces the need not) jeopardize communication by making the agent concerned about being overruled. Similarly, rules are more likely to be imposed within an organization than across organizations because integration makes the agent fearful of being overruled We and therefore less also identified factors that trustworthy. may increase a subordinate's real authority: large span of control, lenient rules, urgency, reputation for surement and multiple principals. which we leave There doubtless are other factors, the investigation of for future research. This paper aims only at being a its moderate interventionism, performance mea- delegation. There are many first step toward a more general theory desirable extensions. For instance, of authority and we have been concerned with a single decision right. Organizations in general allocate multiple decision rights their members. A fascinating question is whether there 26 among exist forces that lead to the clustering of certain types of decision rights in a single hand. Another crucial extension complex webs of authority relationships to in organizations. be answered here, suppose two subordinates potential projects and Whose make different The obvious answer (possibly endogenous) preferences are illustrated That is in learning who is is (real) authority over delegated to the agent whose principal's. This point b^ and his wage in case of success the payoffs with probability E) is his preferences well is as described in the two-tier model. The is w^. not performed by enjoys no private benefit and receives wage principal in case of success. In this model, the supervisor clearly should because about their superior (the principal). most congruent with the is collect information which agent has that authority private benefit attached to his preferred project the principal, but by a supervisor is, to study the kind of questions illustrate the an organization in by the following hierarchy: The agent Monitoring (which consists To recommendations to advice will the principal follow? the other agent? and to allow for multi-layered hierarchies, is ws from the be given authority, with respect to project choice (maximize the probability of success) perfectly coincide with the principal's, even though his preferences over his monitoring effort from the differ principal's. issue contingent. of the The Note that, more principal decides one whose objective is less generally, among whether the supervisor gets who the parties his way is collect information, in favor biased by private benefits or can be most easily aligned with the principal's through formal incentives. 48 Another question hierarchies, project. and The directly related to authority its relates to a negative externality exerted delegation is that of liability in on a third party by the choice of existing legal literature on "vicarious" liability rules essentially argues that in situations of two-sided moral hazard upon both a and principal's where the externalities exerted by a hierarchy depend and an agent's efforts, responsibility for damages to a third party should be shared between the two contracting parties. Even in situations where the agent alone can directly affect the third party, making him entirely liable principal can monitor and control the agent's level of care (that agent only has a limited ability to pay for damages. Thus, if is, is suboptimal if the has authority) and the the principal has a direct power of intervention, she should be held responsible as well (see Shavell (1987, chapter 7) for a statement of the argument). While providing some support for this basic principle (which has 4S For example, a principal may arbitrate in favor of the agent when he feels the supervisor is biased (due a prejudice against the agent, collusion with a rival agent, supervisor's own private benefit, and so forth). Along the same lines, one can wonder whether the principal necessarily gains from choosing a "protege" as a supervisor, that is someone with high congruence with her. It might be the case that the principal be better off with a supervisor sharing the same culture (congruent) with the agent even if this entails some to The "protege' might stifle initiative. focus on private benefits does not preclude the existence of other factors influencing the allocation cover-ups. Our competency of the two parties. Who gets his way may also depend on the power of the various parties, or on the desire of the organization to keep key indispensability/bargaining personnel in the long run (Rotemberg (1993,1994)). of authority such as the relative 27 been applied, e.g., to the new environmental liability rules ("CERCLA") 1980), our analysis suggests at least two further considerations for the degree of liability imposed upon a liable if First, companies or creditors should an extensive account of the courts' views thus provide support for liability rules that can be accommodate various organizational our analysis suggests that in (i) in they are more directly involved in supervising the agent's activities (see Strasser-Rodosevich (1993) for We enforcement: US principal in an integrated firm should vary with her ability to exert real authority: In particular, directors, parent be held more its enacted in the integrated structures sufficiently flexible in order to characteristics that affect real authority, liability rules may have made in this respect). (ii) Second, which allocate much responsibility to principals undesirable consequences "ex ante" an endogenous choice of the authority structure. Such liability rules when we may allow for indeed induce excessive divestiture in situations where coordination considerations would naturally favor the emergence of integrated structures. 49 There is thus a trade-off between responsabilizing the owners or managers in integrated firms and at the same time avoiding inefficiencies in the allocation of formal authority. The tion 49 analysis of these and of other exciting questions related to authority and its delega- must await future research. Such phenomena appear liability rules (see to have occurred in the US Ringleb-Wiggins (1990)). 28 following the introduction of the new environmental MIT LIBRARIES 3 9080 00922 6710 References Arrow, K. [1975], "Vertical Integration and Communication", Bell Journal of Economics, 6:173-183. Bolton, P. and litical J. Farrell [1990] "Decentralization, Duplication, and Delay", Journal of Po- Economy, 98:803-826. M. Dewatripont [1992] "The Firm as a Communication Network", mimeo, LSE, and Universite Libre de Bruxelles, forthcoming, Quarterly Journal of Economics. Bolton, P. and Burkart, M. Gromb, D. and F. Panunzi [1994] "Large Shareholders, Monitoring and Fidumimeo, LSE, and MIT. , ciary Duty", Continental Bank [1993] "Executive Roundtable on Corporate Strategy in the '90s", Journal of Applied Corporate Finance, 6:42-60. Cremer, Davis, S. J. [1992], and P. "Arm's-length Relationships", mimeo, IDEI, Toulouse. Lawrence [1977], Matrix, Reading: Addison- Wesley. Grossman, S. and O. Hart [1986], "The Costs and Benefits of Ownership: A Theory of Lateral and Vertical Integration", Journal of Political Economy, 94: 691-719. Moore [1990], "Property Rights and the Theory of the Firm", Journal Economy, 98: 1119-1158. Hart, O. and J. Political Holmstrom, B. [1979], of "Moral Hazard and Observability", Bell Journal of Economics, 10:74- 91. Holmstrom, B. Marschak, J. [1982], "'Moral and R. Radner Hazard [1972] in Teams", Bell Journal of Economics, 13:324-340. Economic Theory of Teams, New Haven: Yale Univer- sity Press. Mirrlees, J. [1975], "The Theory of Moral mimeo, Nuffield College, Oxford. Hazard and Unobservable Behaviour, Part I", Radner, R. [1992] "Hierarchy: The Economics of Managing", Journal of Economic Literature, 1382-1415. Riordan, M. [1990], "Vertical Integration and the Strategic Management of the Enterprise", in The Firm as a Nexus of Treaties, Aoki, Gustafsson and Williamson, eds, Sage Publications. Ringleb, A., and S. Wiggins [1990] "Liability and Large-Scale, nal of Political Economy, 98:574-595. Long-Term Hazards", Jour- Rodriguez, D. and D. Vayanos [1993], "Decentralization and the Management of Competition", mimeo, MIT. Maximizing Organizations", mimeo, MIT. Rotemberg, J. [1993], "Power Rotemberg, J. [1994]. "Formal Authority vs Power uiimeo, in Profit in Profit Maximizing Organizations", MIT. Sah, R., and J. Stiglitz [1986] "The Architecture of Economic Systems: Hierarchies and Polyarchies", American Economic Review, 76:716-727. 29 Schmidt, K [1991], "The Costs and Benefits of Privatization", mimeo, Universitat Bonn. Shavell, S. [1979], "Risk Sharing and Incentives in the Principal Bell Journal of Economics, 10:55-73. Shavell, S. [1987], Press. and Agent Relationship", Economic Analysis of Accident Law, Cambridge: Harvard University Simon, H. [1951] "A Formal Theory of the Employment Relationship", Econometrica, 19:293305. Strasser, K., ity", and D. Rodosevich [1993] "Seeing the Forest for the Trees in CERCLA Liabil- Yale Journal on Regulation, 11:493-560. Williamson, O. [1975], Market and Hierarchies: Analysis and Antitrust Implications, York: Free Press. Williamson, 0. [1985], The Economic Institutions of Capitalism, Weber, M. [1968], Economy and Society, New " 906 30 New New York: Free Press. York: Bedminster Press. Date Due SIGN at this card the M.I.T. and present it with book CIRCULATION DESK. LIBRARIES - DEWEY MASSACHUSETTS INSTITUTE OF TECHNOLOGY LIBRARIES