Organizational Design 607Once again, the corporate reward system faces a trade-off. Rather than tie
compensation solely to aggregate performance measures, many companies
turn to subjective measures to evaluate individual employee performance. For
example, supervisors evaluate the performance of employees on an annual or
semiannual basis. The supervisor may give numerical ratings (say, on a 10-point
scale) for a number of aspects of job performance. Alternatively, the employer
may give the employee a number of annual goals and then evaluate the
employee on whether and how well he or she accomplishes these goals.
There is an additional evaluation problem. Sometimes it is difficult to know
“how much performance is good performance.” This is sometimes known as
the “benchmark” problem. How does the evaluation system set realistic goals
on which to calibrate performance? Companies have a number of sources of
information in establishing the correct benchmark. Industrial engineers and
efficiency experts could perform studies that examine efficient ways of com-
pleting tasks. Their findings could provide the appropriate benchmark.
Alternatively, firms may use actual past performance as a benchmark. This gives
realistic data as to performance possibilities, but it also can lead to strategic
behavior on the part of workers who are torn between the benefit of exceed-
ing the current benchmark and the cost of establishing higher benchmarks for
the future. Frequently, firms base benchmarks on the historic average per-
formance of a large group of comparable workers.EVALUATING GROUP PERFORMANCE Frequently, group performance is eas-
ier to measure than individual performance for many of the reasons noted ear-
lier. Rewarding group performance may encourage cooperation among
employees who can all share in the fruits of their collective achievement. How-
ever, rewarding group performance does introduce new uncertainties into the
compensation of the employee. That is, an employee’s compensation and pro-
motion are now tied to the efforts of others. More important, rewarding group
performance may discourage optimal effort (indeed, encourage shirking),
especially when the firm cannot easily observe individual effort.
For example, suppose that a team consists of five workers whose annual
bonuses will depend on the measured success of the group. Suppose that the
group cannot observe individual effort but that if all team members exert 15
percent extra effort, group performance will rise by a like amount and each
member will reap an additional $25,000 in bonus. Because each member reck-
ons the disutility of extra effort at $10,000, agreeing to become a high-
performing team benefits all (the net advantage is $15,000 each). But there
is a catch. (Have you spotted the prisoner’s dilemma yet?) Each member’s
personal incentive is to “free ride” on the efforts of the others. By exerting
extra effort, the member raises the average performance of the group by only
one-fifth of 15 percent, or 3 percent. (Remember, there are fiveteam mem-
bers.) In turn, the marginal individual benefit to exerting extra effort is only
one-fifth of $25,000, or $5,000. The benefit of the extra effort is not worth thec14AsymmetricInformationandOrganizationalDesign.qxd 9/26/11 11:03 AM Page 607