LEADING THE WAY
To Reward or To Punish? … That Is the Question
Suppose you are the general manager of a supermar-
ket
and
you’ve
just
finished
a
department-
by-department year-end review of your managers’
performance. Every department—meats, dairy, sea-
food, deli, bakery, and so forth—has performed up to
or beyond expectations. All except one: Produce fell
12 percent short of your forecast. You decide to
reward all your managers with healthy bonuses
except for your produce manager. In other words,
you plan to use
punishment
to motivate your pro-
duce manager and
positive reinforcement
to moti-
vate all of your other managers. You congratulate
yourself for having reached a fair and logical
decision.
According to Daniel Kahneman, a psychologist
who won the Nobel Prize in economics for his work
on behavioral and decision-making models, your
decision is probably not fair (at least not altogether),
and it’s certainly not logical—at least not when the
reality of the situation is taken into consideration.
Here’s how Kahneman sees your two-pronged
decision-making model:
•
Manager’s department performs well
You
reward manager
Department continues to
perform well
•
Manager’s department
performs poorly
You punish
manager
Department
performs better
The key to Kahneman’s per-
spective is called
regression to
the mean
—the principle that,
from one performance measure-
ment to the next, the change in
performance will be toward the
overall average level of perfor-
mance. Say, for example, that
you’re a par golfer and that par
for your course is 72. If you shoot
68 in one round, your next round
will probably be
in the direction
of
72—not
necessarily
72
exactly, which is your average,
or 76, which would bring you
exactly back to a two-round
average
of
72.
Technically,
regression to the mean is a
law
,
and not a
rule
: You could shoot a
second round of 70 or even 67, but
most of the time
,
your second-round score won’t be as good as your
first-round score.
Why does regression to the mean occur?
Because a complex combination of factors usually
determines any outcome. And because this combi-
nation is complex, it’s not likely that the same
combination will repeat itself the next time you
measure the outcome. Which brings us back to
your produce manager:
It’s not likely that his
managerial performance was the sole (or even nec-
essarily the primary) factor in his department’s
poor performance.
Other factors might include
variations in competition, economic and market
conditions, and decisions made by managers
above him—all of which are largely random and
which will undoubtedly be different from one
performance measurement to the next.
Now that you understand a little about the reality
of regression to the mean, compare your decision-
making model to a model that reflects reality:
•
Manager’s department performs well
Department probably does not perform as well
•
Manager’s department performs poorly
Department probably performs better
Fuse/
G
etty
Images
Organizations use a variety of rewards to motivate their employees.
This employee, for example, is receiving an award recognizing his high
performance. His facial expression clearly communicates his sense of
accomplishment and happiness at being recognized for his contributions.
(
continued
)
Do'stlaringiz bilan baham: