As we saw above, Smith (1776) already discussed the problems associated with piece-rate contracts in the industry. Babbage (1835) went a step further by under- standing the need for precise measurement of performances to set up efficient piece-rate or profit-sharing contracts:
It would, indeed, be of great mutual advantage to the industrious work- man, and to the master manufacturer in every trade, if the machines employed in it could register the quantity of work which they perform, in the same manner as a steam-engine does the number of strokes it makes. The introduction of such contrivances gives a greater stimulus to honest industry than can readily be imagined, and removes one of the sources of disagreement between parties.
—Babbage (1835, p. 297)
Also, Babbage proposed various principles to remunerate labor:
The general principles on which the proposed system is founded, are
- That a considerable part of the wages received by each person should depend on the profits made by the establishment; and,
- That every person connected with it should derive more advantage from applying any improvement he might discover than he could by any other course.
(Babbage 1989, Vol. 8, p. 177)
However, Barnard (1938) is the one who can probably be credited with the first attempt to define a general theory of incentives in management, in chapter 11 (the economy of incentives) and chapter 12 (the theory of authority) of his cele- brated book The Functions of the Executive, which he wrote after a long career in management, most notably as president of the New Jersey Bell Telephone Com- pany:
[A]n essential element of organizations is the willingness of persons to contribute their individual efforts to the cooperative system. Inade-quate incentives mean dissolution, or changes of organization purpose, or failure to cooperate. Hence, in all sorts of organizations the affording of adequate incentives becomes the most definitely emphasized task in their existence. It is probably in this aspect of executive work that failure is most pronounced.
—Barnard (1938, p. 139)
Actually, Barnard had a broad view of incentives, involving both what we would now call monetary and nonmonetary incentives:
An organization can secure the efforts necessary to its existence, then, either by the objective inducements it provides or by changing states of mind. We shall call the process of offering objective incentives “the method of incentives”; and the processes of changing subjective attitudes “the method of persuasion.”
—Barnard (1938, p. 142)
The incentives may be specific or general:
The specific inducements that may be offered are of several classes, for example: a) material inducements; b) personal non material oppor- tunities; c) desirable physical conditions; d) ideal benefactions. Gen- eral incentives afforded are, for example: e) associational attractiveness; f) adaptation of conditions to habitual methods and attitudes; g) oppor- tunity of enlarged participation; h) the condition of communion.
—Barnard (1938, p. 142)
Barnard also stressed the ineffectiveness of material incentives, which at the time were almost exclusively considered by economic theory:
[E]ven in purely commercial organizations, material incentives are so weak as to be almost negligible except when reinforced by other incen- tives.
—Barnard (1938, p. 144)
Persuasion . . . includes: a) the creation of coercive conditions (as forced exclusion of indesirables); b) the rationalization of opportunities (if the conviction that material things are worth while . . . succeeds in capturing waste effort and wasted time . . . it is clearly advantageous); c) the incul- cation of motives.
—Barnard (1938, p. 149)
Barnard pointed out the necessary delicate balance of the various types of incentives for success. Furthermore, such a good balance is highly dependent on an unstable environment (through competition in particular) and on the internal evolution of the organization itself (growth, change of personnel). Finally, in his chapter on authority, Barnard recognized that incentive contracts do not rule all the activities within an organization. The distribution of authority along communi- cation channels is also necessary to achieve coordination and promote cooperation:
Authority arises from the technological and social limitations of cooper- ative systems on the one hand, and of individuals on the other.
—Barnard (1938, p. 184)
In modern language, Barnard is saying that the incompleteness of contracts and the bounded rationality of members in the organization require that some leaders be given authority to make decisions in circumstances not addressed specif- ically by the contracts. His main point is to stress the need to satisfy ex post partic- ipation constraints of members who accept noncontractual orders only if they are compatible with their own long-run interests:
A person can and will accept a communication as authoritative only when . . . at the time of his decision, he believes it to be compatible with his personal interest as a whole.
—Barnard (1938, p. 165)
Barnard’s work emphasized the need to induce appropriate effort levels from members of the organization—the moral hazard problem—and to create authority relationships within the organization to deal with the necessary incompleteness of incentive contracts. Not until a few decades later did Arrow (1963a) intro- duce into the literature on the control of management, the idea of moral hazard, borrowed from the world of insurance. This work would be further extended by Wilson (1968) and Ross (1973), who redefined it explicitly as an agency problem. The chapter on authority written by Barnard directly inspired Simon’s (1951) for- mal theory of the employment relationship. Finally, Williamson (1975) followed Barnard and Simon to develop his transaction costs theory for the case of sym- metric but nonverifiable information between two parties.6 Grossman and Hart (1986) modeled this paradigm, which led to the large body of recent literature on incomplete contracts.
Source: Laffont Jean-Jacques, Martimort David (2002), The Theory of Incentives: The Principal-Agent Model, Princeton University Press.