HBR
FROM THE HARVARD BUSI NESS REVI EW
OnPoint
A R T I C L E
How to motivate
Why Incentive Plans
employees?
Cannot Work
Don t pay them for
by Alfie Kohn
good performance.
New sections to
guide you through
the article:
" The Idea in Brief
" The Idea at Work
" Exploring Further. . .
P R ODUC T NUMB E R 2 7 9 9
DO NOT COPY
T H E I D E A I N B R I E F Why Incentive Plans Cannot Work
If you want to build a committed, collabora- Whether it s piecework pay, stock options,
tive, and creative workforce, you have to pay commissions, or Employee of the Month privi-
employees for excellence, right? Not necessarily. leges, pay-for-performance gains you one thing:
Though most U.S. corporations use incentive temporary compliance. It may change people s
programs, trying to reward quality may be a behavior in the short run, but it doesn t alter
fool s errand. the attitudes driving behavior. It can t create
an enduring commitment to your company s
Why? Studies show that people who expect to
values or lasting, meaningful change.
receive a reward for completing a task typically
underperform compared to those who expect no So how can you build an exceptional work-
reward particularly if the task requires sophis- force? Understand the real costs of pay-for-
ticated thinking. At the executive level, studies performance. Then consider more potent
reveal minimal or even negative correlations strategies including long-term goal setting
between pay and performance, as measured and training.
by corporate profitability and other criteria.
T H E I D E A A T W O R K
Why Rewards Don t Work
Pay-for-performance carries a high price for
equately prepared? Unable to collaborate?
your organization in six respects:
Burned out? Too many managers use rewards
as substitutes for what workers really need:
1. Pay doesn t motivate. People need money,
useful feedback, social support, and room for
of course. But when asked what they care
self-determination. Dangling bonuses may be
about most, pay typically ranks only fifth or
easy but it impedes managers ability to ful-
sixth. Though cutting pay would damage
fill their real responsibilities.
morale, increasing it won t necessarily
improve performance.
5. Rewards kill creativity. Incentives encourage
people to focus on precisely what they ll get
2. Rewards punish.  Do this and you ll get that
for completing a task not what might be
rewards aren t too different from  Do this and
gained by taking risks, exploring new possi-
here s what ll happen to you punishments.
bilities, and playing hunches. Rewards pull
People make little distinction between not
people s attention away from excellence.
receiving an expected reward and being
Employees may manipulate task schedules
punished. Pay-for-performance usually makes
or behave unethically to  make the numbers.
people feel manipulated rather than moti-
To finish the task as expediently as possible,
vated to explore, learn, and progress.
they ll opt for simplicity and predictability,
3. Rewards rupture relationships. When you
not challenge.
force people to compete for rewards, team-
6. Rewards undermine interest. If your goal is
work evaporates. Viewing teammates as
excellence, no artificial incentive can match
obstacles to their own success, employees
the power of intrinsic motivation: people
pressure the system for individual gain. And
working because they love what they do.
instead of asking for help from managers
Rewards undermine intrinsic motivation by
essential for enhancing performance they
making people feel controlled and devaluing
conceal problems and present themselves
their work especially when tied to interest-
as infinitely competent.
ing or complicated work. When people view
4. Rewards ignore the causes behind problems.
their work as externally directed and unwor-
To solve workplace problems, managers must
thy, they won t approach it with a commit-
understand their causes: Are employees inad- ment to excellence.
HBR OnPoint © 2002 by Harvard Business School Publishing Corporation. All rights reserved.
DO NOT COPY
I N Q U E S T I O N
When reward systems fail,
don t blame the program 
look at the premise behind it.
Why Incentive Plans
Cannot Work
thinking  those who promote team-
work, participative management,
continuous improvement, and the
By Alfie Kohn
like  urge the use of rewards to
institute and maintain these very
reforms. What we use bribes to ac-
It is difficult to overstate the ex- failure of any given incentive pro- complish may have changed, but the
tent to which most managers and gram is due less to a glitch in that reliance on bribes, on behaviorist
the people who advise them believe program than to the inadequacy of doctrine, has not.
in the redemptive power of rewards. the psychological assumptions that Moreover, the few articles that ap-
Certainly, the vast majority of U.S. ground all such plans. pear to criticize incentive plans are
corporations use some sort of pro-
Temporary
gram intended to motivate employ-
Compliance
ees by tying compensation to one in-
Incentives do not alter
dex of performance or another. But
Behaviorist theory,
more striking is the rarely examined
derived from work the attitudes that underlie
belief that people will do a better job
with laboratory ani-
our behaviors.
if they have been promised some
mals, is indirectly re-
sort of incentive. This assumption
sponsible for such pro-
and the practices associated with it
grams as piece-work pay for factory invariably limited to details of im-
are pervasive, but a growing collec- workers, stock options for top exec- plementation. Only fine-tune the
tion of evidence supports an oppos- utives, special privileges accorded to calculations and delivery of the in-
Employees of the Month, centive  or perhaps hire the author
and commissions for as a consultant  and the problem
salespeople. Indeed, the
Most managers too often
livelihood of innumer- Alfie Kohn is the author of four
believe in the redemptive
able consultants has long books, including No Contest: The
been based on devising Case Against Competition and the
power of rewards.
fresh formulas for com- newly published Punished by Re-
puting bonuses to wave wards: The Trouble with Gold Stars,
ing view. According to numerous in front of employees. Money, va- Incentive Plans, A s, Praise, and Oth-
studies in laboratories, workplaces, cations, banquets, plaques  the list er Bribes, from which this article
classrooms, and other settings, re- of variations on a single, simple be- is adapted. Kohn lectures widely
wards typically undermine the very haviorist model of motivation is lim- at universities, conferences, and
processes they are intended to en- itless. And today even many peo- corporations on education and man-
hance. The findings suggest that the ple who are regarded as forward agement.
Copyright © 1993 by the President and Fellows of Harvard College. All rights reserved.
PHOTOS BY TONY RINALDO
DO NOT COPY
I N Q U E S T I O N
cally, the absence of such a relation-
ship is interpreted as evidence of
links between compensation and
something other than how well peo-
ple do their jobs. But most of these
data could support a different con-
clusion, one that reverses the causal
arrow. Perhaps what these studies
As for productivity, at least two reveal is that higher pay does not
dozen studies over the last three produce better performance. In other
decades have conclusively shown words, the very idea of trying to re-
that people who expect to receive a ward quality may be a fool s errand.
reward for completing a task or for Consider the findings of Jude T.
doing that task successfully simply Rich and John A. Larson, formerly
do not perform as well as those of McKinsey & Company. In 1982,
will be solved, we are told. As Her- who expect no reward
bert H. Meyer, professor emeritus in at all. These studies
Rewards do not create a
the psychology department at the examined rewards for
College of Social and Behavioral Sci- children and adults,
lasting commitment. They
ences at the University of South males and females,
Florida, has written,  Anyone read- and included tasks
merely, and temporarily,
ing the literature on this subject ranging from memo-
change what we do.
published 20 years ago would find rizing facts to creative
that the articles look almost identi- problem-solving to de-
cal to those published today. That signing collages. In general, the using interviews and proxy state-
assessment, which could have been more cognitive sophistication and ments, they examined compensa-
written this morning, was actually open-ended thinking that was re- tion programs at 90 major U.S. com-
offered in 1975. In nearly forty years, quired, the worse people performed panies to determine whether return
the thinking hasn t changed. when working for a reward. Interest- to shareholders was better for cor-
Do rewards work? The answer de- ingly enough, the researchers them- porations that had incentive plans
pends on what we mean by  work. selves were often taken by surprise. for top executives than it was for
Research suggests that, by and large, They assumed that rewards would those companies that had no such
rewards succeed at securing one produce better work but discovered plans. They were unable to find any
thing only: temporary compliance. otherwise. difference.
When it comes to producing lasting The question for managers is Four years later, Jenkins tracked
change in attitudes and behavior, whether incentive plans can work down 28 previously published stud-
however, rewards, like punishment, when extrinsic motivators more ies that measured the impact of fi-
are strikingly ineffective. Once the generally do not. Unfortunately, as nancial incentives on performance.
rewards run out, people revert to author G. Douglas Jenkins, Jr., has (Some were conducted in the labora-
their old behaviors. Studies show noted, most organizational studies tory and some in the field.) His anal-
that offering incentives for losing to date  like the articles published ysis,  Financial Incentives, pub-
weight, quitting smoking, using seat have tended  to focus on the effects lished in 1986, revealed that 16, or
belts, or (in the case of children) act- of variations in incentive condi- 57%, of the studies found a positive
ing generously is not only less effec- tions, and not on whether perfor- effect on performance. However, all
tive than other strategies but often mance-based pay per se raises perfor- of the performance measures were
proves worse than doing nothing at mance levels. quantitative in nature: a good job
all. Incentives, a version of what psy- A number of studies, however, consisted of producing more of
chologists call extrinsic motivators, have examined whether or not pay, something or doing it faster. Only
do not alter the attitudes that under- especially at the executive level, is five of the studies looked at the qual-
lie our behaviors. They do not cre- related to corporate profitability and ity of performance. And none of
ate an enduring commitment to any other measures of organizational those five showed any benefits from
value or action. Rather, incentives performance. Often they have found incentives.
merely  and temporarily  change slight or even negative correlations Another analysis took advantage
what we do. between pay and performance. Typi- of an unusual situation that affected
3
HARVARD BUSINESS REVIEW September-October 1993
DO NOT COPY
I N Q U E S T I O N
a group of welders at a Midwestern
manufacturing company. At the re-
On Incentives
quest of the union, an incentive sys-
tem that had been in effect for some
years was abruptly eliminated. Now,
 The Pay-for-Performance  Why Merit Pay Doesn t Work:
Implications from Organization if a financial incentive supplies mo-
Dilemma
Theory
by Herbert H. Meyer
tivation, its absence should drive
by Jone L. Pearce
Organizational Dynamics
down production. And that is exact-
in New Perspectives on
Winter 1975.
ly what happened, at first. Fortu-
Compensation
nately, Harold F. Rothe, former per-
edited by David B. Balkin and
 Financial Incentives
Luis R. Gomez-Mejia
by G. Douglas Jenkins, Jr. sonnel manager and corporate staff
Englewood Cliffs, NJ:
in Generalizing from Laboratory to
assistant at the Beloit Corporation,
Prentice-Hall, 1987.
Field Settings
tracked production over a period of
edited by Edwin A. Locke
months, providing the sort of long-
 The New Performance Measures
Lexington, MA: Lexington
by Monroe J. Haegele term data rarely collected in this
Books, 1986.
in The Compensation Handbook
field. After the initial slump, Rothe
Third Edition
 Why Some Long-Term
found that in the absence of incen-
edited by Milton L. Rock and
Incentives Fail
tives the welders production quick-
Lance A. Berger
by Jude T. Rich and John A. Larson
ly began to rise and eventually
New York: McGraw-Hill, 1991.
in Incentives, Cooperation, and
Risk Sharing reached a level as high or higher than
 Intrinsic and Extrinsic
edited by Haig R. Nalbantian
it had been before.
Motivational Orientations: Reward-
Totowa, NJ: Rowman &
One of the largest reviews of how
Induced Changes in Preference for
Littlefield, 1987.
intervention programs affect worker
Complexity
by Thane S. Pittman, Jolee Emery, productivity, a meta-analysis of
 Output Rates Among Welders:
and Ann K. Boggiano
Productivity and Consistency
some 330 comparisons from 98 stud-
Journal of Personality and Social
Following Removal of a Financial
ies, was conducted in the mid-1980s
Psychology
Incentive System
by Richard A. Guzzo, associate pro-
March 1982.
by Harold F. Rothe
fessor of psychology at the Universi-
Journal of Applied Psychology
 Enemies of Exploration: Self-
December 1970. ty of Maryland, College Park, and
Initiated Versus Other-Initiated
his colleagues at New York Univer-
Learning
 The Effects of Psychologically
sity. The raw numbers seemed to
by John Condry
Based Intervention Programs on
suggest a positive relationship be-
Journal of Personality and Social
Worker Productivity:
Psychology tween financial incentives and pro-
A Meta-Analysis
July 1977.
by Richard A. Guzzo, Richard D.
ductivity, but because of the huge
Jette, and Raymond A. Katzell
variations from one study to anoth-
 Toward a Theory of Task
Personnel Psychology
er, statistical tests indicated that
Motivation and Incentives
Summer 1985.
there was no significant effect over-
by Edwin A. Locke
Organizational Behavior and
 One More Time: How Do You all. What s more, financial incen-
Human Performance
Motivate Employees?
tives were virtually unrelated to the
Volume 3, 1968.
by Frederick Herzberg
number of workers who were absent
Harvard Business Review
or who quit their jobs over a period
Intrinsic Motivation and Self-
January-February 1968.
?
Determination in Human Behavior of time. By contrast, training and
by Edward L. Deci and
 An Elaboration on Deming s
goal-setting programs had a far
Richard M. Ryan
Teachings on Performance
greater impact on productivity than
New York: Plenum Press, 1985.
Appraisal
did pay-for-performance plans.
by Peter R. Scholtes
 Inferred Values and the
in Performance Appraisal:
Why Rewards Fail
Reverse-Incentive Effect in
Perspectives on a Quality
Induced Compliance
Management Approach
Why do most executives continue
by Jonathan L. Freedman, John A.
edited by Gary N. McLean, et al.
to rely on incentive programs? Per-
Cunningham, and Kirsten Krismer
Alexandria, VA: University of
haps it s because few people take the
Journal of Personality and Social
Minnesota Training and
Psychology time to examine the connection be-
Development Research Center and
March 1992.
American Society for Training and
tween incentive programs and prob-
Development, 1990.
lems with workplace productivity
The Battle for Human Nature:
and morale. Rewards buy temporary
Science, Morality and Modern Life
People, Performance, and Pay
compliance, so it looks like the prob-
by Barry Schwartz
by Carla O Dell
New York: W.W. Norton and
Houston: American Productivity lems are solved. It s harder to spot
Company, 1986.
Center, 1987.
the harm they cause over the long
term. Moreover, it does not occur to
most of us to suspect rewards, given
that our own teachers, parents, and
4
HARVARD BUSINESS REVIEW September-October 1993
DO NOT COPY
I N Q U E S T I O N
managers probably used them.  Do But it doesn t necessarily follow that environment conducive to explo-
this and you ll get that is part of the doubling that person s pay would re- ration, learning, and progress.
fabric of American life. Finally, by sult in better work. 3. Rewards rupture relationships.
clinging to the belief that motiva- 2. Rewards punish. Many man- Relationships among employees are
tional problems are due to the par- agers understand that coercion and often casualties of the scramble for
ticular incentive system in effect at fear destroy motivation and create rewards. As leaders of the Total
the moment, rather than to the psy- defiance, defensiveness, and rage. Quality Management movement
chological theory behind all incen- They realize that punitive manage- have emphasized, incentive pro-
tives, we can remain optimistic that ment is a contradiction in terms. As grams, and the performance ap-
a relatively minor adjustment will Herzberg wrote in HBR
repair the damage. some 25 years ago ( One
Punishment and rewards
Over the long haul, however, the More Time: How Do
potential cost to any organization of You Motivate Employ-
are actually two sides of
trying to fine-tune reward-driven ees? January-Febru-
compensation systems may be con- ary 1968), a  KITA 
the same coin. Both have
siderable. The fundamental flaws of which, he coyly ex-
a punitive effect because
behaviorism itself doom the pros- plains, stands for  kick
pects of affecting long-term behavior in the pants  may pro-
they are manipulative.
change or performance improve- duce movement but
ment through the use of rewards. never motivation.
Consider the following six-point What most executives fail to rec- praisal systems that accompany
framework that examines the true ognize is that Herzberg s observa- them, reduce the possibilities for co-
costs of an incentive program. tion is equally true of rewards. Pun- operation. Peter R. Scholtes, senior
management consultant at Joiner
1.  Pay is not a motivator. W. Ed- ishment and rewards are two sides of
ward Deming s declaration may the same coin. Rewards have a puni- Associates Inc., put it starkly,  Ev-
seem surprising, even absurd. Of tive effect because they, like out- eryone is pressuring the system for
individual gain. No one is improving
course, money buys the things peo- right punishment, are manipulative.
the system for collective gain. The
ple want and need. Moreover, the  Do this and you ll get that is not
system will inevitably crash. With-
less people are paid, the more con- really very different from  Do this
out teamwork, in other words, there
cerned they are likely to be about or here s what will happen to you.
can be no quality.
financial matters. Indeed, several In the case of incentives, the reward
The surest way to destroy cooper-
studies over the last few decades itself may be highly desired; but
ation and, therefore, organizational
have found that when people are by making that bonus contingent
asked to guess what matters to their on certain behaviors, managers ma- excellence, is to force people to com-
pete for rewards or recognition or to
coworkers  or, in the case of man- nipulate their subordinates, and that
rank them against each other. For
agers, to their subordinates  they experience of being controlled is
each person who wins, there are
assume money heads the list. But likely to assume a punitive quality
many others who carry with them
put the question directly   What do over time.
the feeling of having lost. And the
you care about?  and pay typically Further, not receiving a reward
more these awards are publicized
ranks only fifth or sixth. one had expected to receive is also
Even if people were principally indistinguishable from being pun- through the use of memos, newslet-
ters, and awards banquets, the more
concerned with their salaries, this ished. Whether the incentive is
detrimental their impact can be.
does not prove that money is moti- witheld or withdrawn deliberately,
Furthermore, when employees com-
vating. There is no firm basis for the or simply not received by someone
pete for a limited number of incen-
assumption that paying people more who had hoped to get it, the effect is
tives, they will most likely begin to
will encourage them to do better identical. And the more desirable
see each other as obstacles to their
work or even, in the long run, more the reward, the more demoralizing it
own success. But the same result can
work. As Frederick Herzberg, Dis- is to miss out.
occur with any use of rewards; intro-
tinguished Professor of Management The new school, which exhorts us
ducing competition just makes a bad
at the University of Utah s Graduate to catch people doing something
thing worse.
School of Management, has argued, right and reward them for it, is not
Relationships between supervi-
just because too little money can ir- very different from the old school,
ritate and demotivate does not mean which advised us to catch people do- sors and subordinates can also col-
lapse under the weight of incentives.
that more and more money will ing something wrong and threaten
Of course, the supervisor who pun-
bring about increased satisfaction, to punish them if they ever do it
ishes is about as welcome to em-
much less increased motivation. It again. What is essentially taking
ployees as a glimpse of a police car in
is plausible to assume that if some- place in both approaches is that a lot
one s take-home pay was cut in half, of people are getting caught. Man- their rearview mirrors. But even the
supervisor who rewards can produce
his or her morale would suffer agers are creating a workplace in
some damaging reactions. For in-
enough to undermine performance. which people feel controlled, not an
5
HARVARD BUSINESS REVIEW September-October 1993
DO NOT COPY
I N Q U E S T I O N
ment, University of California at
Recommended Reading
Irvine, wrote in  Why Merit Pay
Doesn t Work: Implications from
 A Model of Creativity and The Market Experience
Organization Theory, pay for per-
Innovation in Organizations by Robert E. Lane
formance actually  impedes the
by Teresa M. Amabile Cambridge, England:
in Research in Organizational Cambridge University Press, 1991.
ability of managers to manage.
Behavior, Volume 10
5. Rewards discourage risk-taking.
edited by Barry M. Staw and The Hidden Costs of Reward:
 People will do precisely what they
L.L. Cummings New Perspectives on the Psychology
are asked to do if the reward is signif-
Greenwich, CT: JAI Press, Inc., 1988. of Human Motivation
edited by Mark R. Lepper and icant, enthused Monroe J. Haegele,
Out of the Crisis David Greene
a proponent of pay-for-performance
by W. Edwards Deming Hillsdale, NJ: L. Erlbaum
programs, in  The New Perfor-
Cambridge, MA: MIT Center for Associates, 1978.
mance Measures. And here is the
Advanced Engineering Study, 1986.
root of the problem. Whenever peo-
The Great Jackass Fallacy
 Merit Pay, Performance Targeting, by Harry Levinson
ple are encouraged to think about
and Productivity Cambridge, MA: Harvard University
what they will get for engaging in a
by Arie Halachmi and Marc Holzer Press, 1973.
task, they become less inclined to
Review of Public Personnel
take risks or explore possibilities, to
Administration The Human Side of Enterprise
Spring 1987. by Douglas McGregor play hunches or to consider inciden-
New York: McGraw-Hill, 1960.
tal stimuli. In a word, the number
No Contest: The Case Against
one casualty of rewards is creativity.
Competition, Revised Edition Wealth Addiction
Excellence pulls in one direction;
by Alfie Kohn by Philip Slater
rewards pull in another. Tell people
Boston: Houghton Mifflin, 1992. New York: Dutton, 1980.
that their income will depend on
Punished by Rewards: The Trouble Money and Motivation: An
their productivity or performance
with Gold Stars, Incentive Plans, Analysis of Incentives in Industry
rating, and they will focus on the
A s, Praise, and Other Bribes by William Foote Whyte and
numbers. Sometimes they will ma-
by Alfie Kohn Melville Dalton, et al.
Boston: Houghton Mifflin, 1993. New York: Harper, 1955. nipulate the schedule for complet-
ing tasks or even engage in patently
unethical and illegal behavior. As
stance, employees may be tempted ing problems and bring about mean-
Thane S. Pittman, professor and
to conceal any problems they might ingful change.
chair of the psychology department
be having and present themselves as Moreover, managers often use in-
at Gettysburg College, and his col-
infinitely competent to the manager centive systems as a substitute for
leagues point out, when we are moti-
in control of the money. Rather than giving workers what they need to do
vated by incentives,  features such
ask for help  a prerequisite for opti- a good job. Treating workers well 
as predictability and simplicity are
mal performance they might opt in- providing useful feedback, social
desirable, since the primary focus as-
stead for flattery, attempting to con- support, and the room for self-deter-
sociated with this orientation is to
vince the manager that they have mination  is the essence of good
get through the task expediently in
everything under control. Very few management. On the other hand,
order to reach the desired goal. The
things threaten an organization as dangling a bonus in front of employ-
late Cornell University professor,
much as a hoard of incentive-driven ees and waiting for the results re-
John Condry, was more succinct: re-
individuals trying to curry favor quires much less effort. Indeed,
wards, he said, are the  enemies of
with the incentive dispenser. some evidence suggests that produc-
exploration.
4. Rewards ignore reasons. In or- tive managerial strategies are less
Consider the findings of organiza-
der to solve problems in the work- likely to be used in organizations
tional psychologist Edwin A. Locke.
place, managers must understand that lean on pay-for-performance
When Locke paid subjects on a
what caused them. Are employees plans. In his study of welders perfor-
piece-rate basis for their work, he
inadequately prepared for the de- mance, Rothe noted that supervisors
noticed that they tended to choose
mands of their jobs? Is long-term tended to  demonstrate relatively
easier tasks as the payment for suc-
growth being sacrificed to maximize less leadership when incentives
cess increased. A number of other
short-term return? Are workers un- were in place. Likewise, author
studies have also found that people
able to collaborate effectively? Is the Carla O Dell reports in People, Per-
working for a reward generally try to
organization so rigidly hierarchical formance, and Pay that a survey of
minimize challenge. It isn t that hu-
that employees are intimidated 1,600 organizations by the American
man beings are naturally lazy or that
about making recommendations Productivity Center discovered lit-
it is unwise to give employees a
and feel powerless and burned out? tle in the way of active employee in-
voice in determining the standards
Each of these situations calls for a volvement in organizations that
to be used. Rather, people tend to
different response. But relying on in- used small-group incentive plans. As
lower their sights when they are en-
centives to boost productivity does Jone L. Pearce, associate professor at
couraged to think about what they
nothing to address possible underly- the Graduate School of Manage-
6
HARVARD BUSINESS REVIEW September-October 1993
DO NOT COPY
I N Q U E S T I O N
are going to get for their efforts.  Do ly destructive when tied to interest- comes in two flavors, the logic goes,
this and you ll get that, in other ing or complicated tasks. and both together must be better
words, focuses attention on the Deci and Ryan argue that receiv- than either alone. But studies show
 that instead of the  this. Empha- ing a reward for a particular behavior that the real world works differently.
sizing large bonuses is the last strat- sends a certain message about Some managers insist that the
egy we should use if we care about what we have done and controls, or only problem with incentive pro-
innovation. Do rewards motivate attempts to control, our future be- grams is that they don t reward the
people? Absolutely. They motivate havior. The more we
people to get rewards. experience being con-
The number one casualty
6. Rewards undermine interest. If trolled, the more we
our goal is excellence, no artificial will tend to lose inter-
of rewards is creativity. As
incentive can ever match the power est in what we are do-
of intrinsic motivation. People who ing. If we go to work
the late John Condry put it,
do exceptional work may be glad to thinking about the
rewards are the  enemies
be paid and even more glad to be well possibility of getting
paid, but they do not work to collect a bonus, we come to
of exploration.
a paycheck. They work because they feel that our work
love what they do. is not self-directed.
right things. But these managers fail
Few will be shocked by the news Rather, it is the reward that drives
to understand the psychological fac-
that extrinsic motivators are a poor our behavior.
substitute for genuine interest in Other theorists favor a more sim- tors involved and, consequently, the
one s job. What is far more surprising ple explanation for the negative ef- risks of sticking with the status quo.
is that rewards, like punishment, fect rewards have on intrinsic mo- Contrary to conventional wis-
dom, the use of rewards is not a re-
may actually undermine the intrin- tivation: anything presented as
sic motivation that results in opti- a prerequisite for something else  sponse to the extrinsic orientation
exhibited by many workers. Rather,
mal performance. The more a man- that is, as a means toward another
ager stresses what an employee can end  comes to be seen as less desir- incentives help create this focus on
financial considerations. When an
earn for good work, the less interest- able. The recipient of the reward
organization uses a Skinnerian man-
ed that employee will be in the work assumes,  If they have to bribe me
agement or compensation system,
itself. to do it, it must be something I
The first studies to establish the wouldn t want to do. In fact, a se- people are likely to become less in-
terested in their work, requiring
effect of rewards on intrinsic moti- ries of studies, published in 1992 by
extrinsic incentives before expend-
vation were conducted in the early psychology professor Jonathan L.
ing effort. Then supervisors shake
1970s by Edward Deci, professor and Freedman and his colleagues at the
their heads and say,  You see? If you
chairman of the psychology depart- University of Toronto, confirmed
that the larger the incen- don t offer them a reward, they won t
do anything. It is a classic self-ful-
tive we are offered, the
Do rewards motivate
filling prophecy. Swarthmore Col-
more negatively we will
lege psychology professor Barry
view the activity for
people? Absolutely.
which the bonus was re- Schwartz has conceded that behav-
ior theory may seem to provide us
ceived. (The activities
They motivate people to
with a useful way of describing what
themselves don t seem to
get rewards.
goes on in U.S. workplaces. Howev-
matter; in this study,
they ranged from partici- er,  It does this not because work is
ment at the University of Rochester. a natural exemplification of behavior
pating in a medical experiment to
By now, scores of experiments across theory principles but because behav-
eating unfamiliar food.) Whatever
the country have replicated the find- the reason for the effect, however,
ior theory principles...had a signifi-
ing. As Deci and his colleague
any incentive or pay-for-perfor- cant hand in transforming work into
Richard Ryan, senior vice president an exemplification of behavior theo-
mance system tends to make people
of investment and training manager ry principles.
less enthusiastic about their work
at Robert W. Baird and Co., Inc., Managers who insist that the job
and therefore less likely to approach
wrote in their 1985 book, Intrinsic won t get done right without re-
it with a commitment to excellence.
Motivation and Self-Determination wards have failed to offer a convinc-
Dangerous Assumptions
in Human Behavior,  the research ing argument for behavioral ma-
has consistently shown that any
Outside of psychology depart- nipulation. Promising a reward to
contingent payment system tends to
ments, few people distinguish be- someone who appears unmotivated
undermine intrinsic motivation.
tween intrinsic and extrinsic moti- is a bit like offering salt water to
The basic effect is the same for a va- vation. Those who do assume that
someone who is thirsty. Bribes in the
riety of rewards and tasks, although workplace simply can t work.
the two concepts can simply be added
extrinsic motivators are particular- together for best effect. Motivation Product no. 2799
7
HARVARD BUSINESS REVIEW September-October 1993
DO NOT COPY
E X P L O R I N G F U R T H E R . . . Why Incentive Plans Cannot Work
ARTICLES
 Introducing T-Shaped Managers: equity, regardless of tenure; and rewards them
Knowledge Management s Next Generation with profit shares based solely on length of
by Morten T. Hansen and Bolko von Oetinger tenure. Nonpartners receive base salaries
(Harvard Business Review, March 2001, and annual bonuses based on their support
Product no. 6463) of their colleagues and their enhancement of
One reason so many managers embrace EZI s reputation, for example, through pub-
pay-for-performance is that it s much easier lishing articles. Results? Annual profit expan-
to  dangle carrots before employees than to sion for 40 consecutive years, 60% of business
fulfill managers real responsibilities: develop- from repeat customers, and yearly employee
ing direct reports and creating horizontal turnover of 2% in an industry averaging 30%.
value by sharing ideas and expertise across
the company.  The Best-Laid Incentive Plans by Steve Kerr
(Harvard Business Review, January 2003,
This article takes a closer look at the benefits
Product no. R0301A)
of horizontal value including better deci-
In this fictional case study, the CFO of
sion making through peer advice, increased
Rainbarrel Products experiences firsthand the
revenues through shared expertise and cross-
perils of incentive programs. Rather than
pollination of ideas, and well-coordinated
enhancing productivity, his new performance-
implementation of bold strategic moves.
management system seems to be destroying
How to spur the creation of horizontal value? the company: Rainbarrel fails to bring a
Formalize cross-unit interactions and clearly breakthrough product to market quickly
communicate the corporate value of sharing enough, a survey reveals a demoralized work-
knowledge outside individual units. force, and customers lament the quality of
Rainbarrel s service.
 A Simpler Way to Pay by Egon Zehnder
Four experts offer explanations, including the
(Harvard Business Review, April 2001,
 you get what you pay for rule. For example,
Product no. 6765)
if you compensate salespeople only for sales
Though Zehnder agrees that pay-for-individual-
dollars, they ll achieve more sales but at
performance programs can backfire, he does
lower prices or with expensive extra services.
recommend compensating workers for com-
Define your criteria for long-term success
panywide results. At his company, executive
before launching an incentive program. Is it
search firm Egon Zehnder International (EZI),
sales? Profits? Retained business? Then select
success hinges on employees networks of
appropriate performance metrics. Use clear,
contacts, sharp intuition, and commitment
two-way communication so employees
to the firm all qualities derived from lengthy
understand and accept your company s strat-
tenure. For this reason, EZI links compensa-
egy and can align their performance behind it.
tion directly to seniority.
EZI pays partners a base salary; gives them
an equal number of shares in the company s
Visit us on the Web at:
U.S. and Canada: 800-988-0886
617-783-7500 " Fax: [phone]
DO NOT COPY