Robert Myers, Revision Date: 12/25/01
(This note was widely circulated via email in the World Bank in October 1994. It stimulated a lot of comments --author's note.)
1. There is a memo floating around the Bank, obviously not from our PR people, that acknowledges that there are problems with the quality of IBRD and IDA lending. The fault, according to the memo, lies with the Bank's incentive system, not with its structure. The current incentive system stresses lending volumes or the quantity rather than quality of lending. The same quantity vs. quality problem is ascribed to our non-lending work, including Economic and Sector Work (ESW) and Operations Evaluation Department (OED) output. The conclusion is that it is the incentive system, not the structure or staff which ought to be changed. If I remember correctly the memo does not delve into how to change the Bank's incentive system. However, an article in the Sept. 1993 American Economic Review by Canice Prendergast, entitled. "A Theory of `Yes Men'", contains some interesting insights.
2. The abstract of the article, which elaborates on something called the "Principal-Agent" (P/A) problem, reads as follows:
"This paper illustrates an incentive for workers to conform to the opinion of their supervisors when firms use subjective performance evaluations. This desire to conform arises endogenously from the firm's need to induce the worker to exert effort. I show that firms may optimally eschew the use of incentive contracts to retain workers' incentives for honesty. I illustrate that the incentive to conform implies inefficiencies, even when workers are risk-neutral, and is likely to lead to more centralized decision-making than in the absence of the desire to conform."3. The article is tightly reasoned and somewhat mathematical, particularly in its requirement that the reader understand that honesty is defined in terms of attempting to know the true value of a random variable, which is unknowable, ex anti. This means that the "correctness" of the information provided by the agent can only be determined subjectively. Given this, the article presents a P/A problem in which the agent must choose to allocate his/her work effort between learning and reporting honestly vs. learning and conforming to a superior's prejudices. You guessed it! Conformity wins every time. This is not necessarily bad (the superior may already know what the worker will learn after a lot of work) but it is nearly always inefficient (because there is duplication -- inefficiency is defined in terms of spending resources without raising the probability of a "correct" decision). Furthermore, if the manager doesn't know very much, it can be disastrous.
4. A potpourri of problems suggested by the article which apply to the Bank are as follows:
(ii) "This paper, therefore, provides another reason why employees may [should] be allowed considerable discretion over their behavior instead of [being given] explicit performance measures." (p. 759) [] added;
(iii) "...the existence of `yes men' is a natural implication of subjective performance evaluations." (p. 763);
(iv) "Hence, in organizations where workers know what their managers want to hear, incentive contracts should not be used..." [they encourage dishonesty, compared to "set" or "civil service" salary conditions] []=added (p.764);
(v) "...`toadyism' is likely to be concentrated amongst the less able, who trust their own opinion less than do able workers." (p.767);
(vi) "...this paper illustrates inefficiencies associated with 'yes men'. First information becomes noisier, implying, for example, less efficient project selection..."(p.769); and,
(vii) "It was also shown that, conditional on incentives being offered, 'yes men' are likely to be concentrated among less able workers, among workers with less able managers, [and] in organizations with much interaction [teamwork] between management and workers...". [] added, (p.770).